The opposite is the case, Caplan hardly wants to talk about those at all. He's a libertarian economist much fonder of Szasz' ideas than the typical libertarian economist, and so modeling mental illnesses as preferences follows naturally from him. The reclassification of homosexuality is just a piece of evidence that Szasz/Caplan can cite to claim psychology/psychiatry is political rather than scientific.
But it's obviously some amount of both. I guess it's understandable that they're miffed about their opponents (also understandable) reluctance to admit the political component, but going off the deep end in response doesn't seem like the greatest idea.
No, I think if there's a specific aspect of this that excites him, it's alcoholism, degenerate gambling, drug abuse, etc. Certainly, alcoholism is his go-to referent. Over the past several decades, these pathologies have been re-imagined, so they are now seen as mental diseases rather than personal and moral failings, and Caplan doesn't like it.
Mental illness is physical illness is illness. It’s all the same stuff at rock bottom. Maybe Caplan is in some senses just a dualist. Or he believes in a “faculty of will” that just decides things.
And it’s hard to believe he’s ever really known a severely mentally ill person in his life given how he misunderstands it.
I think he is a dualist. He believes in free-will, and rejects cryonics because he doesn't think an emulation behaving exactly like him will really contain his consciousness.
I believe Greg Cochran would agree that mental illness is physical illness, and his primary criteria for determining if something qualifies is its effect on Darwinian fitness.
If it only applies in our unusual environment with contraception, then he might make an exception (his criteria for assuming a pathogen involves both the loss of fitness and having been around a long time). He has mentioned that normal mating preferences we've evolved would result in men finding a robot attractive if it actually resembled Sean Young in Bladerunner, even though that wouldn't be conducive to fitness (he wrote that prior to the release of Bladerunner 2049).
Is there a reason why it's obvious which of ADHD and not-ADHD we should consider the diseased state?
Not-ADHD seems more adaptive for the exact modern first-world environment with schools and office jobs and taxes, etc. But ADHD seems perfectly adaptive to plenty of other actual human environments that still exist, have existed, and will exist again.
I don't see how this is true? Non-ADHD seems to mean you can shift your attention when you want to, vs ADHD which seems to mean your attention shifts without your control. What environments is that more adaptive for?
I think that's not the most productive way of looking at it. Non-ADHD means it takes longer for your attention to shift without your control, but it will shift eventually. Moreover, your attention shifting without your control can be beneficial, because it prevents you getting stuck in a rut, or eaten by a hostile predator while you're gazing at interesting pebbles. I've seen it likened to a safety mechanism for your attention.
So the question is do you want a relatively relaxed safety mechanism, or a relatively trigger-happy mechanism. Compared to the ancestral environment, today we spend more time poring over boring spreadsheets and less time running from predators [citation needed] so ADHD has become less adaptive.
This might be controversial, but it seems like ADHD is also correlated with some positive traits like creativity and responding well (and quickly) to surprising and unfamiliar situations. These are definitely adaptive traits.
And, given that the negative effects of ADHD are almost invisible in a large number of environments (notable excluding the environment of modern highly-organized first world society), the benefit doesn't need to be very large for ADHD to be more adaptive than not-ADHD in these other contexts.
Since becoming a mom to two kids, I’ve found my mild ADHD feels so much like an adaptation for motherhood I’ve actually put off trying Ritalin so I can leverage it until they start school. As a parent you get interrupted constantly. For my husband who has excellent focus, this is *exhausting*. For me, it kind of slots into the “normal”(typical?) functioning of my brain. There was space in my head for incessant gear switching and noise, and while it is tiring and annoying at times, it doesn’t leave me wiped out the like it does to my husband. I’m convinced that it the tendency to ADHD must have served some purpose, at least when its expression is fairly mild.
That's very interesting to hear! I hadn't considered the other side of the coin (that being interrupted with 'normal' focus levels leads to exhaustion).
This is my experience as well. Whether I have ADD or not, I was diagnosed with it as a kid and put on medication... As a mom-of-four I don't need any medication to get through my day. Starting things? Yes, I have hundreds of things to do each day. Finishing them? Housework is never done. There will always be more dirty clothes or hungry kids or books to read... Many of my duties are simultaneous, like cooking breakfast and feeding the pets and watching the baby. Focusing on getting one task "all done" would mean not doing any of the other tasks, which would be bad!
I think of it as "working in circles": doing a little bit of each task until I arrive back at the beginning.
I suspect this is true for many mental illnesses, and for that matter, some physical illnesses. PTSD is likely more adaptive in an environment where you experience trauma with no hope of escape, such as slavery. Anxiety is likely adaptive if conditions are constantly life or death. Obesity is likely adaptive in societies with frequent famines.
> Is there a reason why it's obvious which of ADHD and not-ADHD we should consider the diseased state?
You wouldn't want to consider either of them a diseased state. That's just natural variation. (Of course, that's exactly what Caplan wants to say about everything.)
The fact that a particular strategy outcompetes every other one in a particular environment does not mean that following other strategies is a symptom of a disease. We maintain reservoirs of genetic variation because (1) innovating new adaptations is extremely difficult, and (2) the environment frequently changes. Thus, we hold on to our old adaptations, in the form of people who struggle to reproduce, against the day when they turn out to be superior to us in the face of some new feature of the world.
As I point out in a root-level comment sidethread, the better way to think about "disease" or "not disease" is whether or not you are functioning as intended.
For another example, would that mean that Alzheimer's is often not an illness, because at the onset of symptoms the person involved has already reproduced as much as they were ever going to?
(Based on my limited understanding of when Alzheimer's usually manifests and human fertility rates as a function of age)
I think one's ability to boost one's fitness declines more gradually than that. Women have menopause, but it's believed the reason that evolved was to push them toward boosting the fitness of their existing offspring rather than trying to have more, and Alzheimer's would reduce their ability to do that.
^ this. Everything bad that happens to a member of such a social species as ours decreases fitness. Fortunately for us we've had enough extra fitness that we've been able to absorb the cost of our best hunter or spearmaker slowly declining from old age, or even our best warrior losing his right arm and requiring subsidy as an honored hero. But cost there always is.
In a sense, it's obvious that our "healthy" state, as judged by ancestral environment standards, is sub-optimal in terms of fitness in the post-industrial environment. The not-having-an-urge to donate to sperm banks is a popular "disability" to mention around these parts.
The ADHD people having more children thing. Is this sex differentiated?
I could see it being forgetfulness/inattentiveness/in-ability to keep to a routine/schedule/habit. If so I would say forgetting to take the pill sometimes would differentiate from males "forgetting to put on/bring a condom" if we are presuming the partner is on average neurotypical and it takes two to tango.
Lack of condom etiquette in my opinion would be much much harder to go by unnoticed in the moment. I would expect in a theory like this for there to be more adhd women than ADHD men having a higher than average number of kids?
Scott has proposed a theory and you are investigating his theory, but is there any evidence that the underlying assertion (ADHD people have allegedly have more children) is actually true? Do you have citation(s) for the assertion?
As an assertion about humanity globally, I can't think that that can possibly be true. (Shrug.)
I have no citation! Honestly I just kinda went with the assertion based pretty much on respect for Scott, and the assumption he had seen something credible and I got caught up on modeling the causes and was too distracted/polite/cowardly to ask for a source.
I would love if someone could link on any source for the base claim!
I mostly just jumped on this as, over the last two weeks, I've been investigating ADHD stuff since it turns out if you draw a box around all my "personality quirks", issues, idosyncracies etc I have been coping with undiagnosed ADHD for 30 years now. Just got out of a psychotherpist who specializes in it and yuuuup.
The last 10 years would have been a looooot different with some Ritalin or what-have-you.
Turns out fundamentally not being able to connect your knowledge and values with your actions is not something one is supposed to be hugely struggling with every moment of every day to this degree????
Hard to speak for Cochran, he's a very clever man. But I wonder if he *would* think not-having-ADHD is a disease in modern conditions. After all, give it a few generations and all-other-things-being-equal beyond the realms of sanity, and the whole population will have ADHD, and the few weirdos who refrain from breeding when they get the chance because of something that happens in their heads will look mysteriously asexual.
Is asexuality a disease?
I'm far from asexual but as far as I know my preferences have resulted in me having no children, and I'm not signed up for cryonics either. I'd be happy to call that a disease in the Darwinian sense. I don't expect there to be many copies of me in the future.
I'd be willing to say it's a disorder, and I'm speaking as one. Given that 99% of the humans around me would crawl over broken glass to get their ashes hauled, being "nope, not one glimmer of interest in anyone or anything" does look unnatural or at the least 'something went wrong in the wiring'.
I'm very happy the way I am, but from the way the rest of yiz go on about "what makes us human is the ability to love" (where love means romantic/erotic love), then I'm a robot or an alien not a human. Fine by me!
I wouldn't say it's a choice, as such; I didn't choose to be this way. Caplan might say I could always choose to have sex, so it's not whatever. Well yeah, and I could also choose to stick my hand in a blender, but I'm not gonna do that either, Bryan.
"Too impulsive" is one explanation, "staying on top of birth control appointments and timing isn't trivial and requires executive function" is another part of an explanation.
This is a very careless sort of metric, but perhaps there's a difference between a trait which harms Darwinian fitness and the lack of a trait which aids it.
I’d argue strict physicalism rules out true emulations of our consciousness. Our consciousness is literally in the physical stuff of our brains; the substrate does matter. There’s no uploading. An emulation of a hurricane doesn’t get you wet, as they say. And an inorganic robotic arm that is functionally indistinguishable from a human arm doesn’t make it a replica of your arm.
I dunno, what if consciousness is literally just electrons in a neural net? Then all we'd have to do is connect your brain to the new substrate before killing you.
They seem generally important to computation, in both digital computers and possibly in brains. It does raise the question of whether non-electrical computers could have consciousness I guess. But it seems like as good a starting place as any to me.
I guess the basis of the intuition besides that is just that electricity is the only other thing besides the whole network which is involved in the whole network.
It's true that electric charges are important in the brain, but they are carried by ions and not by electrons.
Excitation is generally caused by an influx of Sodium (or calcium) ions, inhibition by an influx of Chloride ions, getting back to the resting state from a depolarized state is caused by an outflow of potassium Ion. Release of neurotransmitter after an action potential is triggered by an influx of calcium ions.
EDIT: What are the artificial neurons made of? If they are made of the exact same stuff as the rest of me then there's no difference. If they're made of inorganic material then that's an (interesting) empirical question we may be able to test someday as to whether your thought experiment is even possible.
Wait, so you're saying that consciousness is tied to the form the matter takes? You're mostly made up of carbon, hydrogen, oxygen, and a few other things. If I made an artificial neuron out of those things that didn't look like a natural one, but performed the exact same function, would you count that as preserving consciousness? What if I took a neuron from another person instead? What if I used a particle collider to synthesize atoms from hydrogen and put together an atomically identical neuron, and replaced yours with those?
"Wait, so you're saying that consciousness is tied to the form the matter takes?"
This is far from rare - it's what John Searle uses to argue against conscious AI, for instance, that matter and not merely abstract function is what's relevant. I don't *think* this is the case, but it isn't silly.
" If I made an artificial neuron out of those things that didn't look like a natural one, but performed the exact same function, "
That's a big if. Compare with proteins: They only work because of the shape which is dependent on the their composition and the environment in which they exist.
Given the number of neurons in a typical human brain and the typical human lifespan, "when I die of old age" is probably the correct answer here.
Though to be fair it's less than two orders of magnitude off if you're replacing one per second.
The worry is that there's one neuron in particular in there that, when you disconnect it, everything goes black for *you*, and after that it's just something else wired up to the meatsuit you were piloting. And, for obvious reasons, "you" are not going to have any way to protest that they've made a mistake, because "you" are lost in an increasingly large sea of thirty-or-so excised neurons per second.
There's uploading in the sense that it creates two instances of your consciousness, with all the gnarly moral questions about what, if anything, to do with multiple instances of "yourself", at what point they sufficiently diverge and become unique etc. etc.
You can't just say "the substrate does matter" without giving any reason to believe that. "An emulation of a hurricane doesn’t get you wet" is an analogy, and analogies aren't evidence, just hints. An inorganic robotic arm is not a replica of a human arm, because the definition of "replica" includes being made of the same material.
Making claims about other people's internal mental states is fraught at the best of times. The strength of behaviorism is that it only talks about observable phenomena. The weakness (of the strong version) is that its predictions are shit.
In the language I prefer, behaviors are the observables, but there ARE hidden-variables.
When it’s an illness, yes. The problem is that it’s not always clear when something is an illness. Is it an illness for the second toe to be longer than the first? It doesn’t make someone’s life much worse, so we say no (though it is annoying how many shoes aren’t shaped well for this condition). Similarly for being gay.
We say no mostly because the liberal individualist paradigm has crushed the conservative collectivist one as far as elite tastemaker opinions are concerned. LGBTetc. expression was (more or less) suppressed for thousands of years on conservative grounds. It's not clear whether this was ever optimal even on conservatives' own terms, but the toe analogy would've never been considered relevant until the last 60 years or so.
Hm, probably misunderstanding. Elite tastemakers hold high-status opinions. Previously, "homosexuality is a society-corrupting sin" was an elite opinion, whereas now it's one that only an irrelevant fringe holds. Nobody has ever thought that having an unusually-sized toe has any society-wide implications.
Yeah, it’s inconvenient to have Greek toe. I do have trouble finding shoes that fit, but at least I have good company. The lady portrayed in the Statue of Liberty is in the same boat.
Disagree. I read several papers by an obscure mid-twentieth century cardiologist that explain very clearly how what mainstream medicine derisively calls "heart attacks" are really just the body's physical expression of a preference for reduced blood flow. You're welcome to read me sledgehammer every modern example of physical illness into this conceptual framework at great length at my new Substack, SitOnIt.
Right. Just because I can't tell you exactly how many grains of sand constitute "a heap" does not mean that "heap" is meaningless or a false category. And no amount of pointing to edge cases on the definition of "heap" will make the Sleeping Bear Dunes cease to be a major geographic feature.
I don't think Scott is arguing that the distinction between constraints and preferences is meaningless; he gives examples of things that are clearly on each side. It's more the claim that Caplan's attempt to draw a clear boundary doesn't work because there is no clear boundary, so if we're considering edge cases we need to take into account that they are genuine edge cases where the categories don't work well.
To make this a bit more precise, if Caplan is arguing that a condition has some property of a preference (you could stop it if you had a gun to your head) therefore it can't also have properties of constraints (like deserving help to cure it), pointing to the existence of edge cases is a reasonable counter-argument. What matters is not whether the category exists at all, but whether it's fuzzy or mutually exclusive.
So to steelman his model, we need to be trying to determine for each individual patient and illness to what extent they have a constraint vs preference and then tailor our treatment (aid/gun to head) based on that? I am wondering how that will do useful work.
If diseases vary in the extent to which they are preferences vs constraints then that DOES indeed affect how we should respond. The late Mark Kleiman in a Bloggingheads episode noted that the "disease model of addiction" is falsified by Mao's success in clamping down on opium addiction, and that this fit with the most successful treatment for opiate-addicted doctors being monitoring with consistent immediate consequences.
I'm beginning to see the problem with behaviorism, with which I have limited familiarity. The problem is in a clinical setting it serves the function, or is the thing that currently best serves the function, of preventing neurotypical midwits from making bad and abusive decisions based on bad theory of mind. If you let the average behavioral health professional make any inferences at all into the behavior of an autistic person, or a mentally handicapped person, or whatever, that will lead to dark and dangerous roads pretty consistently. If you want examples of this you can just look at the troubled teen industry, which relies on folk psychology and religious theories of the human mind rather than behaviorism.
Complete digression but I have to say it as it reflects an important takeaway I got from practicing ABA for three years.
The distinction between behavior as disease and behavior as preference that I keep coming back to is whether it preserves optionality or not. This is probably strictly as bad as Caplan's because I have never discussed it with anyone (as a general rule I never have an opportunity to discuss anything with anyone: I am immediately dismissed as aggressive and insane whenever I articulate anything).
So anything that causes death trivially fails to preserve optionality. Anything that cripples a person or reduces their intelligence or health trivially fails to preserve optionality. Getting into a machine that injects an optimized dose of heroin into you continuously while keeping you alive indefinitely trivially fails to preserve optionality, etc. Environmentally I think of this as a problem that can sometimes be solved by reducing the strength of incentives or attractors. Some behavioral incentives are basically singularities to humans in some or even most mental states. I can't think of a good reason any environment should be engineered in a way that creates such singularities.
Given how many people have accounts of being abused by ABA practicioners and how openly abusive Ivar Lovaas was, I think you're either wrong about behaviorism preventing abuse, or gesturing at the existence of some dramatically more abusive environment than all the behaviorism-based abuse I've heard of. I wish I could be more confident that you were just wrong.
Abuse is the norm in the setting. It was before ABA and if ABA is abolished it will be afterward as well. But yes, the troubled teen industry is worse than the average ABA setting. Diamond Ranch Academy alone has something like a half dozen easily preventable deaths now.
It's possible my experiences were moderated by unique factors and the overall situation is worse. I don't believe that the people who killed these victims would have not killed them if it weren't for ABA, but you're steering me away from the belief that ABA might have any advantages.
There is a certain type of person for whom any theory, organizational schema, or ideology is just a pretext under which to conduct and camouflage abuse. There are a lot of these people in certain professions. I don't think ABA solves this problem. I don't think the average person even notices or believes this problem exists for various reasons. My only observation is that for some, theory of mind doesn't exist, they are only capable of falsely inferring antisocial inner states or motives, and getting them to rotely follow mechanical procedures would be a step up for everyone involved given the limited workforce available (the 90-105 IQ pseudo professional range)
Still feels less bad to me and like it can probably be placed on the grounds of physics and just needs to avoid physical nihilism. I also suspect most definitions of optionality converge rather than diverge and it's just a book keeping problem.
Ironically, Caplan often criticizes his fellow economists for being too inclined toward behaviorism and not taking seriously intuition & reports about mental states. I'm more behaviorist than Caplan and think he relies too much on the latter sort of evidence.
I do think Caplan's gun to the head test is somewhat stronger than you give it credit: even with a gun to the head you will have reduced capacity with a physical illness relative to your capacity without that, and we don't need to look at a bright line like traveling a specific distance when we have other observables like how fast & steadily you're moving, how heavily you're breathing etc.
But as Scott points out: even with a gun to the head you will have reduced capacity with many mental illnesses relative to your capacity without that. So this argument is refuted.
(And even with a gun to your head, you will have reduced capacity of you don't exercise regularly, or if you eat a big Christmas dinner. Is not exercising regularly or eating a big dinner diseases?)
1) Several mental illnesses pass the "gun to the head" test, as Scott clearly illustrated in the post, both in terms of having cognitive effects that can't be overcome by any incentive and by empirically preferring the gunshot in extreme cases.
2) In the case of ADHD and I suspect other disorders involving executive dysfunction, the whole nature of the disorder is that the sufferer's ability to respond to incentives is malfunctioning. They can respond to ICNU incentives, but not other kinds. Note that this is a statement about their incentive structure, not their preference structure. Often someone with ADHD has the same preferences as anybody else, but can only take action towards those preferences by reframing that action to satisfy ICNU. (Healthy version: an instant-gratification point system for errands can slot them into I. Unhealthy version: neglecting your work until your boss threatens to fire you can slot it into U.) That is, an ADHD person is operating under the constraint that they are essentially incapable of responding to non-ICNU incentives. So gun-to-the-head, plainly an ICNU incentive, doesn't reveal anything about their preferences, it just removes that constraint.
Interest, Challenge, Novelty, Urgency. It's a framework for understanding ADHD that comes from ADHD specialist William Dodson. I used it because it's expressed very concisely; I've seen other specialists use different language to convey similar ideas about incentive structures and intention versus action. I'm not sure how widespread it is among non-specialists, actually; I'm not a psychiatrist and I've sometimes been mildly surprised at the way Scott writes about ADHD.
In other words they just have poor discipline. That is not an illness. Just because you can change a behavior with drugs doesn't make it an illness. If I had spring shoes I could dunk a basketball, that doesn't mean I have "no dunk" illness.
Kid who I believe has a mental illness related to attention: Kid where you tell them "just stand here for one minute and watch your sister play soccer" and you turn your head. When you turn around ten seconds later the kid is off running in the street. This type of lack of focus and ability to follow instructions is consistent throughout their life, and severe consequences change nothing.
Kid who I DON'T believe has a mental illness related to attention: Kid where they are perfectly functional in most environments except work, homework or anything involving self discipline, but severe consequences see immediate improvement in discipline.
This is just an undisciplined person who doesn't actually have incentives to change their behavior, and also seems like it is currently a majority of the people "diagnosed" with this "illness". And the accommodations that are made for them just increase the problem, since it is provides incentives to be "ill" and insulates them from consequences.
You're far from alone in that opinion. Thankfully, it doesn't really matter what you think unless you're put in charge of some unfortunate ADHD kid or employee. People with the ability to research, diagnose and/or treat ADHD seem to disagree with you on balance.
I mean everyone's entitled to their opinion and certainly some mental disorders are incorrectly or over diagnosed. Same with people who get orthopedic surgery, etc etc. But I would want someone trained to diagnose ADHD to be drawing the lines and making determinations even if they get it wrong sometimes. Your second example, depending on other factors, is potentially a central example of ADHD and not even an edge case. Functional in some environments but impaired at school or work and non-responsive to medium-level consequences is a lot of ADHD. Distinguishing that from malingering, trauma, depression, or anxiety isn't easy to do just from watching.
What would you think about my kid's friend, who had impulse control problems where he could not stop himself from getting enraged at his classmates, even when threatened with severe consequences? (The consequences eventually encompassed being expelled from a school that he really loved and had good friends at.) He'd try and try to keep hold of his temper, and he'd even succeed for a little while (so there's your improvement -- but only temporarily, like in Scott's post), and then he'd explode in a way that was at least as bad, maybe worse, as if there hadn't been consequences in the first place.
This kid then was diagnosed with "classic" ADHD and put on medication, and now he is perfectly functional and doesn't do that kind of thing anymore.
(Interestingly, he was mostly functional in regard to schoolwork even when he was having his worst anger issues -- he's an extremely smart kid, so even though he also did have the classic lack of focus, he was smart enough that he could get by without focusing too hard. So "perfectly functional in most environments" might fit him, actually.)
> In other words they just have poor discipline. That is not an illness.
Something feels wrong about this type of argument. Not sure I can put my finger on it, but it sounds like:
"I have a flu. It makes me sneeze all the time."
"In other words, you just sneeze all the time. That is not an illness."
Sounds like: if X can be reduced to components, it is not a true X.
The problem is that "poor discipline" is a symptom that can have many possible causes. Tired people probably have poorer discipline on average that well-rested ones, but that shouldn't make us conclude that being tired is not a thing (because it is "just poor discipline").
From certain perspective, the diagnosis of ADHD is like saying: these people have poor discipline *because of* their shape of brain or their internal chemistry. As opposed to people who have poor discipline merely because of their habits or the incentives they have at the moment.
Basically, by saying that ADHD is "just poor discipline", you seem to imply that a relatively simple set of incentives applied consistently for a few weeks could easily make the ADHD go away. Which I believe is wrong.
What do you make of adults who want to change their behavior and focus, but find themselves drifting off at work and they're worried that it's going to impact their job performance, so they go to the doctor on their own to ask for help? Are they faking or is it real?
That’s not a mental illness that is a performance enhancing drug. And like I said, from a public policy perspective it is a bad idea since it creates an arms race.
People should believe it if the person has demonstrated that larger capacity earlier. And certainly determining someone's maximum capacity is going to be more difficult than checking if they can meet some minimum, but so what?
Same, I wonder if it served as an inspiration, or if "ravenously hungry to the point that you will eat until you explode" is a more general trope that people would come to even without knowledge of that or similar illnesses. Another historical example, although I don't think it's known which disorder in particular caused it, is Tarrare: https://en.wikipedia.org/wiki/Tarrare
Some dog breeds sometimes eat themselves to death if given unlimited food. Many pet owners have heard about this, so that could be another source of inspiration.
The phenomenon is called "eating to gorging". It's not so much that the animal will kill itself. It's whether or not it will hurt itself by eating excessive food if that much is available. People who work with animals (e.g. the large majority of all premodern people) need to know which ones will eat to gorging and which won't. It is, as illustrated here, most typically thought of as a trait of the species, though there is abundant evidence that some individual humans will eat to gorging in the presence of sufficient food and others won't.
I said something to this effect on Caplan's post itself, but his ideas on mental illness are the perfect example of the incoherence that results from applying market-based reasoning to areas outside its proper domain. Markets and economistic modes of analysis are useful tools, but when you insist on applying them to literally everything you end up with nonsense like insisting mental illness doesn't exist because something something preferences. Remember people: just because you have a model that can explain a lot of things in a really cool way, doesn't mean you have to apply it to literally everything in the universe!
This is precisely what Caplan is responding to in his post on Scott getting married (the mental illness angle is just that Scott had made a similar argument about economics in an earlier post).
Looking at that post, Caplan seems to be saying "look, you admitted to applying economics to an area people don't usually apply it to! That means I'm right and you should apply it to literally everything!" which seems like a gotcha that still isn't strong enough to imply the very extreme claim he wants it to. You can say people should apply economistic reasoning to more areas than they currently do, while still recognizing there are some areas it simply doesn't apply. All that one would have to say in response to this post is that marriage is an example of the former, and mental illness the latter.
Now we're getting somewhere (closer to behaviorism!), although in one section above Scott acknowledged that Caplan could make the same predictions as him (and just sound silly).
Did he? It seems to me that Scott's main thrust is that Caplan's model doesn't make good predictions (but can "explain" anything after the fact), because the preference/constraint dichotomy is incoherent.
The goal is to find the "correct" model, one that DOES work well literally everywhere, to which all others are approximations in the suitable limit. Sure, this is difficult to achieve in practice, but abandoning the pursuit in favor in collecting stamps seems like … well, you can call it "science" if you like.
Is this really necessary though? Why can't we have multiple models, each of which is the most accurate in its own domain? Sure, this might not be as parsimonious or "pretty," and pursuing a "Theory of Everything" has value in the deep insights it might reveal if we really find one, but until then I don't see why we can't be happy with models that respect their own areas of applicability. It reminds me of how in geodesy, you have multiple different geoids, or approximate models of the shape of the Earth, depending on whether you want to model the whole Earth (in which case it will be pretty good but not great everywhere), or a specific area (in which case it will work great there but terribly in other places). You don't need to insist on using literally only one geoid for everything, and in fact would be severely handicapping yourself to do so.
That's a bad example: we know there IS a real shape we're approximating, and choosing an approximation we like. What it reminds ME of is the theory of gravity: you can treat the motion of cannonballs and the motion of the planets as entirely different, but it's still worth knowing they're ultimately the same phenomenon.
But surely there are domains where market dynamics simply don't apply. How would you explain the motions of subatomic particles in economic terms? Surely talk of preferences and markets don't apply there (saying a particle has a "preference" to follow a field is the sort of cute rhetorical anthropomorphism I don't think any physicist would seriously endorse), and yet there are areas this form of analysis clearly does apply, so this seems like proof it is a model that is both predictive and domain-restricted, in which case knowing what those domain restrictions are would be useful. Denying this would seem like the same style of thought (insisting your guiding theory must apply to absolutely everything) which led the Soviet Union to denounce the existence of resonant Benzene-ring bonds, because it somehow contradicted the ideas of Dialectical Materialism, their pet economic theory (I know you could say this is just proof DiaMat is wrong, but the point is more how this style of denying domain restrictions on theories can lead to silly outcomes).
I didn't know that bit about the Soviet Union and resonant benzene bonds. Amusing.
The relationship between human preferences and the fundamental physics of the particle interactions is the other way around: the latter DOES ultimately cause the former, but yes, it's obviously not a useful way to describe the emergent behavior.
Conflict is only between theories on the same … hierarchical level, if you will permit some hand-waving. Only if one doesn't subsume the other entirely do you need to worry about reconciling them, as in the case of economic preferences vs. mental illness (honestly, it's pretty easy to reconcile them here; Scott's theory is simply better).
A better example might be: all chemistry is fundamentally quantum mechanics, but using the latter is prohibitively difficult, so you need the heuristics of chemistry. Or if you prefer, all fluid dynamics is fundamentally just intermolecular forces, so there's no conflict between the Navier–Stokes equations and electromagnetism.
And then why would you insist on using economics to explain psychology, when it's psychology that's more fundamental than economics? That really is like trying to explain particle physics in terms of economics!
> How would you explain the motions of subatomic particles in economic terms?
By saying they move towards their lowest-energy configuration and then stay there. There is a reason the motto of Marginal Revolution appears to be "solve for the equilibrium".
> Denying this would seem like the same style of thought [...] which led the Soviet Union to denounce the existence of resonant Benzene-ring bonds
Man, the number of people who would like to denounce the existence of resonant bonds is not low. Same reason quantum mechanics is so frequently denounced. Same reason hidden variable theories are so popular.
The Soviet Union denounced all kinds of empirical facts. Why pick one of the ones that everyone really does hate? Your example in isolation could easily be explained by the natural desire of all people to denounce resonant chemical structures by whatever means they have available to them.
Is that the goal? I would say the goal should be primarily which model is the most accurate, not taking your fancy hammer and seeing how many objects you can treat as nails. Treating elevating a single model above all others as the ultimate goal seems to lead the sort of bullheaded dickishness that you see on the part of Bryan Caplan and String Theorists.
Even look at Mathematics, where there is a very famous theorem that proves that you cannot derive an answer to every meaningful possible question you have from a set of first principles (among other things). Even though ZFC set theory is generally widely used, there are significant questions that cannot be said to have an answer one way or another.
> The goal is to find the "correct" model, one that DOES work well literally everywhere, to which all others are approximations in the suitable limit. Sure, this is difficult to achieve in practice
No, it's functionally impossible, not merely difficult. The evidence is quite clear that this strategy - seeking the "correct" model of human behavior - is in exactly the wrong direction. The people who are best at predicting the future (Tetlock's Superforecasters) are good at predicting the future because they mix and synthesize and use different models for different situations. The people who have tried to develop big singular models of human models all suck at predicting the future basically exactly as much as one another.
"Markets and economistic modes of analysis are useful tools, but when you insist on applying them to literally everything you end up with nonsense"
Would nontransitive preferences be sufficient to rule out a economic analysis? If someone prefers A to B, B to C, and C to A, then they cannot be correctly modeled as a rational actor with a consistent utility function (at least in the domain of choices that include A, B, and C). Would that be sufficient to say: "Don't hand this case to an economist." ?
I think he can be a source of useful perspectives on some issues, but he's a quintessential hedgehog. He has one intellectual framework he tries to fit everything into, and if something doesn't fit, he hammers on it until it wedges into place and pointedly ignores all the bits which stick out. The framework doesn't work for everything he wants it to, but that doesn't mean that it's not a useful thing for a fox to be mindful of.
Horses have no appetite control and can't vomit. If you put enough horse feed in front of a horse, it will continue to eat past the point where it starts injuring its stomach and will eventually die from it.
That sounds implausible (not the vomiting part; that's true), and more like a folk-story cautioning against overfeeding horses (for health reasons similar to human overeating) than an actual feature of equine biology.
This is flatly untrue. Horses eat grass. You can put a horse in a paddock with more grass than it can eat, and it will not eat itself to death. Indeed, it's hard to imagine how they could have survived as a species if it were not so.
Equally untrue of hard feed (concentrates, oats, etc.) If you are trying to catch a hard-to-catch horse by enticing it with a bucket of feed, success depends on whether it is feeling hungry or not. The claim they cannot vomit is true but not relevant.
True of many, perhaps most, horses and concentrates (corn, grain, pellets) but the cause is more a mismatch between high starch/carbs and the horse's evolutionary gut, which is designed for constant intake of low quality food. Horse overeating is generally pathological only for concentrates, and is of a different severity and process than human overeating.
They don't specify what the consequence is, but apparently it can make them very sick. It's certainly folk wisdom among horse people that it's possible for a horse to eat itself to death on grains.
I stand corrected! Thank you for enlightening me. It is very strange though. How did horses ever evolve? How could they make it through periods when humans weren't controlling their food supply?
I believe the issue is that compound feed swells in the intestines. At a mountain dairy I worked at one summer, just before my arriving a goat had gotten into the compound feed and eaten enough for the swelling feed to rupture its innards.
As a longtime equestrian hobbyist, my understanding is that horses evolved to eat low-calorie density wild steppe grass. Under these circumstances, they had to eat almost continuously (when not sleeping, escaping predators, socializing/mating, etc.) to satisfy their caloric needs. So, their continuous(ish) eating behavior is rational in their natural environment. Prader-Willi syndrome in humans is presumably not.
The problem occurs when they're confonted with high-calorie-density human-supplied foods (oats, corn, even grass that's been selected, watered, fertilized, etc. to be much more rich than wild steppe grass). Saying they have *no* appetite control is probably an overstatement, but they're not evolutionary equipped to regulate their consumption of such unnaturally (for them) rich food. Some will indeed eat until their bowels are obstructed, which can result in death.
The condition is colloquially known as "colic" (I know nothing about human babies, so I'm not sure if it's the same as the "colic" that human babies get.)
Perhaps, but if being reasonably polite but moderately rude is considered a mind-bogglingly high level of courtesy, then I would hate to encounter discourtesy.
This entire post came off as extremely spicy. I had never heard of Caplan before, and I felt a little bad for him by the end of reading this. Imagine being that obsessed with Scott (he seems like an avid Slate Star/Astral Codex fan), and that roasted by Scott.
Caplan's gun-to-the-head distinction between a "preference" and a "constraint" seems so wildly, obviously silly that it's hard for me to believe he actually means it. Reading his original paper, it seems like he's just trying to plug a hole in libertarian philosophy that can't actually be plugged.
Wow this makes sense. Of course Caplan is especially interested in preferences because he’s an economist. But it’s deeper than that, his entire ideology is that the perfect world is one in which everyone is free to follow their preferences. Mental illness destroys this perfect world so he’s forced to either moderate his ideology or pretend mental illness doesn’t exist.
That' a really interesting point, actually. Since moving to Canada, which is much more individualistic/libertarian than where I come from, I notice people with mental illness are much less well taken care off. You're fine if you get yourself to a therapist or doctor and sit down for treatments, but most people with severe mental illnesses will not be able to do these things, and they are left out to dry because it's 'their preference' not to get treatment. A friend of mine in a severe manic episode was refusing to take her medication, terrorising her roommates, and putting herself in dangerous situations - but the authorities refused to do anything because it was 'her choice' (not to take her meds).
What's a bit ironic, perhaps, is that his kind of behaviourism can be harnessed to abet, if not justify, a very different kind of political philosophy. If we insist that people can function just fine absent a set of objectively-verifiable 'constraints', and internal states are just bourgeois indulgence/slave morality, then we can design quite the clockwork society.
Caplan himself has noted that his belief in “a dualistic philosophy of mind, free will, moral objectivism, and an optimistic view of human potential” is linked to his political philosophy. I responded to that here:
Thanks for that. I don't think Caplan's essay on Hobbes was particularly good, to be honest, which might account for some of your bemusement at the time. It reads like he begins with his presumed ingredients of totalitarianism and attempts to stuff chunks of Hobbes into the framework, rather hastily.
The argument collapses, for me, amid the confusion between natural rights and natural law, the distinction between ius and lex that Hobbes explicitly warns about. Caplan writes: "[Hobbes] qualifies this by pointing out that even without a sovereign, people are still bound by natural law. But if we remember that natural law gives everyone the right to do whatever they have the might to do, we can see that this qualification doesn't make any difference." Natural law in Hobbes is, in a sense, the opposite of the natural rights to do whatever you want. Natural law (once a mechanism of enforcement arises via the sovereign) exists to protect us from each other's brutal exercise of those natural rights.
I note in passing that Caplan makes no mention of Hobbes's list of immutable natural laws (which would scuttle the relativism argument), nor of those natural rights that, according to Hobbes, are inalienable and cannot be transferred by contract - self-defence, freedom from imprisonment, freedom from personal harm, etc. Nor does he mention Hobbes's provisions for rights of the subject vis a vis the sovereign. Nor does he make much emphasis on the principle that natural rights are laid down freely, rather than seized. There is also a bit of goal-post shifting between 'totalitarianism' and 'authoritarianism'.
I think your instincts in your reply post were correct - you can readily be a Hobbesian minarchist. But the more interesting question: in the intervening sixteen years, did you become a fully paid-up Stirnerite?
In the time since I wrote that I took seriously Paul Graham's advice to "keep your identity small" and moved away from an explicit ideology like libertarianism into a more consequentialist decentralist bent. This was written the next year (as you can guess from the url) https://entitledtoanopinion.wordpress.com/2008/02/09/rhymes-with-shmashmortion/
Once you are getting into these either or spaces within nosology you have missed the boat in medicine and psychiatry. The huntingtons hemiballism, the tourettic tic and the OCD compulsion are all on a spectrum, with the inevitability of performing each action modulated by shades of increasing agency, but all vastly to the seemingly inevitable and ego-dystonic side of one's perceived agency. Interestingly neuroimaging and demographic data (OCD/tourettes ~30% comorbid with eachother) point to common potential imbalances between the direct and inhibitory pathway in the basal ganglia.
Thus, it's just shades and penumbras between what psychiatrists and neurologists are doing, psychiatrists are like epi-neurologists. On this note, in thinking of preferences and constraints- late stage addiction involves changes in the dorsal striatum and almost becomes like a movement disorder- and the withdrawal state becomes a physiologic constraint that drives the inevitability of use through negative reinforcement- (not to mention the avoidance of hypertensive states, dysphoria, seizures etc). A preference clearly can become a constraint, like eating turns morbid obesity. These are all intricately related. Anyone who is not just armchair commenting and has experience with dual dx populations though, will know the experience well of hearing tales of trauma and hyperarousal etc, driving the original use- was it a preference? Was it not the cause of an original imbalance that constrained the inevitability of pursuing one action over another? It's diffuse and penumbral and deserves individual assessment and formulation, patient by patient, and that's the beauty of psychiatry.
To be charitable to the guy, after all the shade I've thrown at him, Caplan may well be proceeding on the lines that if the 'talking cures' worked, then that must mean there is some element of choice: in response to certain incentives, you chose to be hysterical or phobic or in love with your mother, and then after talking it over with a shrink, you then chose to be mentally healthy.
So there wasn't an "illness" there constraining you to be hysterically blind, it was a choice.
Which okay, yeah. But full-on paranoid schizophrenia is a horse of a very different colour, and it's definitely nothing to do with "choosing" to throw off all your clothes and run naked into traffic.
From your references I assume you are a psychiatrist or psychologist, and i doubt caplan would know what most of your language means in a nuanced manner. I see your point and do think there a levels of CLEAR BRAIN DYSFUNCTION in certain psychiatric illness mania, catatonia, schizophrenia, acute withdrawal, delirium, things secondary to medical conditions, etc which you cannot pin down as clearly with factitious disorders, anxiety, depression. But we work within a physiology which is neuroplastic, and under epigenetic modulation- with real time interaction between behavior and genes. And "talking cures" can impinge on this system as much as anything, as much as pharmacology.
So for example you take someone with a sustained childhood trauma hx or prolonged exposure to combat and PTSD, they have epigenetically, with no choice involved, downregulated feedback inhibition of their HPA axis d/t consistency of threat in their environment (fkbp15 hypo-methylated along with other implicated physiological changes). They likely have salivary cortisol measures anywhere from 3-5x the general population. Thus the tendency toward aggression, toward externalizing psychopathology, toward hitting the ground AS SOON AS they hear a car backfire (no sense of choice, reflexive loops between thalamus-amygdala- PAG means almost no frontal lobe evaluation of their response). Say they engage in 'the talking cure' and all these memories are unlocked, new, comforting information added to each, now within the context of a therapeutic frame, and this alone methylates fkbp15 and restores HPA homeostasis.
This doesn't mean that there was more choice in the original system that led to hitting the ground with loud sounds, but that choice and constraint are modulated with opposite turns of a screw in a dynamic system, and experiences- war and therapy, can turn this screw in opposite directions. War turns up constraint and therapy relieves it. Kind of like a rational evidence based treatment for a disease!
hey you talk the talk! Well some of the neuroendocrine physiology i outlined may not be familiar, but it's all very well established neurphysiological outcomes to trauma and part f the etiologies of ptsd
Behavioral coaching works on many physical illnesses, too.
“Give away your cat, move to an uncarpeted dwelling, and wash your bedding every three to four days in steaming hot water,” is advice that can control some people’s sniffles.
Mental and physical illness are not distinguished by whether mundane (if sometimes costly) individual choices result in better disease control.
I am happy to play along with Caplan by calling physical disease “weird preference”, too. I’ve noticed that Caplan seems reluctant to acknowledge how much physical disease can be modeled as a weird preference which simply must be managed in order to avoid burdening others.
But mental illness being preference while physical is not? No. As uncharitable as it is to call some mental-health crises “preferences”, much physical disease is hardly exempt from being modeled as a “weird preference”, too.
> Caplan may well be proceeding on the lines that if the 'talking cures' worked, then that must mean there is some element of choice: in response to certain incentives
Be wary of conflating habits with choice. Clearly we have have well-ingrained thinking habits, and a big component of therapy is establishing a feedback mechanism to change thinking habits. Ingrained thinking habits are not a form of deliberate choice in the usual sense.
If the difference is supposed to be that a "preference" can be changed with incentives and a "constraint" can't, then I don't think the dichotomy makes sense even in purely economic terms. Any budget "constraint" can be overcome with enough creativity and disregard for the law - it's just that the incentive would have to be much much stronger to match.
Or, in Caplanian terms: I can't afford a house, but if I had a gun to my head, I could probably make it happen.
Right, but Scott's point about it in the article is that Caplan's framing makes it unfalsifiable. Sure you can _claim_ that every person who you put a gun to their head and told them to travel faster than light didn't do it because it's physically impossible, but I can just counter with they just had a stronger preference for not traveling faster than light.
This was the point Scott was making about putting a gun to a deeply psychotic person's head and asking them to reason normally. Caplan is (apparently according to Scott, I haven't followed this debate very closely, certainly not Caplan's side of it) unwilling to admit that this could possibly be the result of a constraint. It proves itself.
The gun-to-the-head test doesn't actually prove anything about constraints vs. preferences, even though it is true that _some_ things are actually physically impossible and others are not.
Caplan hasn't actually made it unfalsifiable, because he hasn't claimed anyone has a stronger preference than being shot in the head... but this does leave him vulnerable to accepting suicidal behavior as a constraint rather than a preference (I don't know that he's addressed that).
No, but in all seriousness: I think budget constraints work as an economic concept because people are so averse to breaking the law that once you go past your budget there's a steep spike in the incentives required to alter behavior. But it's still not *physically impossible* to go beyond that type of constraint, and pretending it is means you have to redefine "constraint" to a point where it makes no practical sense in either psychology or economics.
Even in a complete state of anarchy, budget constraints still apply. Budget constraints work as an economic concept because of the nature of reality. You can't create a perpetual motion machine, no matter how strongly motivated you are.
My point is that there's a wide gulf between "unreasonable/unacceptable" and "literally impossible", and budget constraints are usually defined by the former limit because (a) it's a way easier line to draw and (b) only psychopaths would dare venture into that gulf anyway. Caplan is bent on using the "impossible" definition in his mental illness argument, even though turning that around would break his model of economics by implying that he "can afford" a giant gold statue of himself but merely "prefers" not to sell everything he has and go on a bank robbing spree to make it happen.
Economists are interested in people's willingness to pay, and that applies even to the completely mindless. Distributed Republic (which unfortunately is no longer online) had a great post on this illustrated with a Magic: The Gathering Card of a mindless entity, citing Gary Becker on the minimal requirements for supply/demand curves to have their normal slopes. You can't pay more than you have, and you can't supply more than you can physically produce.
I don't think that breaks his model. He could afford a smaller-scale gold statue without the bank robbing, but in both cases the cost is above what he is willing to pay for it.
Hmm, I do not disagree with your point that preferences vs constraints is a very limited framework for evaluating whether something is an illness. However, your parsimony argument is on the weak side: the universe is complex and the apparently simplest models are often wrong. Whether to consider something an illness that needs to be cured/accommodated/empathized with depends on a complex interplay of internal and external factors.
For example, homosexuality can be considered an illness to be cured in one specific society if every member's procreation is essential for the survival of the society (and assuming artificial insemination is not an option).
Having lethal cancer at the age of 30 can be considered normal in a society where people over 30 are a burden for whatever reason, and not getting a cancer is a disease. Taking steps to avoid getting cancer by 30 could be considered a mental illness or a crime.
Star Trek: TNG has a whole episode on Geordi not being considered ill.
The subculture of hearing-impaired often looks askance at those who choose to have an implant. Without implants available poor hearing was an illness, with them available it's a choice.
Many things that we normally think of as symptoms of physical illness, are not in fact directly caused by the disease, but are an evolved behavioral suite that activates in response to the disease!
I thought the debate about whether depression is a disease as a debate mainly about whether depression is a distinct entity, or whether the real problems depressed people have couldn’t be classified better by other disease pathologies.
Or something else. Maybe some set of people who have depression diagnoses are actually lonely, maybe that is their real problem. ”Lonely” is a real problem but it isn’t a disease. You might say, well, they are depressed too. But is that really the most useful way of thinking about it?
But Caplan seems to be arguing that people can’t actually be lonely without choosing it, without its being their revealed preference?
I think the preferences are more supposed to be over doing something rather than being something (although I suppose the preference for doing something would sensibly be related to the effect of the action).
I think you've won this against Brian Caplan-Szasz. (I have some expertise in criminal insanity law, and I'd occasionally see citations to Szasz well past the point where it was defensible.)
the interesting thing about szasz is that he was a pretty normal psychiatrist who practiced much like the other psychiatrists from his day. I learned from many of the people who learned from him at suny upstate medical center where he was chair, they said he was a great teacher of psychiatry who taught them to think abt what/why but basically practiced normal psychiatry.
I haven't read Szasz since grad school but my memory is thinking that he had a lot of interesting and rich clinical material as well as some refreshing perspectives. It seems like his views years later and out of context are being weaponized for new arguments.
"Instead, I would rather describe things that make an action difficult and unpleasant as in some sense real constraints."
You'd think an economist would understand that friction (in the metaphorical sense) making activities more costly (in the economic sense of all costs) doesn't have to be a hard constraint to alter people's behavior. Illness is, well, taxing. Many illnesses, like taxes, just make stuff more costly without imposing hard boundaries.
I've developed, against my will, a strong preference for avoiding cats and smokers. It's nothing personal, but I like breathing better. Medicine has advanced enough that I could live with a cat or smoker and not die, but it would be an unproductive life, since struggling to breathe turns out to be really distracting.
If my only asthma triggers were cats and smokers, I could lead a normal, un-asthmatic life in a society that shunned smoking and cat-keeping – politics! In this society, though, where cat-keeping is normal and smoking is stigmatized but still common, I'll sometimes drug myself extra to disguise my airway's weird preferences as much as possible while I visit loved ones who live with cats or smokers. I'll usually still have symptoms, but I consider it polite to hide them if I can – though "polite" can be overdone:
I once "polited" my way all the way to the ER as a child. I was wheezing at school but didn't want to make a scene, so I hid it for as long as I could. As my parents pointed out, hiding it that much ended up making even *more* of a scene. The etiquette of minimizing the burden we impose on others due to our weird bodily preferences (like asthma or diabetes) can get rather complicated.
I would also think an economist has gotten away from the sharp distinction between “want” and “need”, and therefore shouldn’t draw a sharp distinction between “preference” and “constraint”!
As my husband (a UofC-trained economist) observed several years ago, when he and I first discussed Caplan’s theory of mental illnesses, budget constraints are usually set by preferences, anyhow.
We can all imagine that there must be some limiting constraints “out there” that can’t be further changed by rejiggering our preferences, but our knowledge of these “absolute constraints” is uncertain, and in practice, budgeting is very much influenced by preference:
How much risk or debt are we comfortable taking on? How far would we go to augment our income in a pinch? (Pawning stuff, sex work, selling organs? — preferences!) Are we comfortable with bankruptcy? With moving (or going homeless — preference!) if we can’t make rent or mortgage?…
People budget in order to reach goals. We all know there are limits to rejiggering our goals to stretch our budget, but few of us really know where those limits are — and that’s mostly a good thing: the kind of hardship that requires radically reconfiguring your goals just to eke through is, well, hard, sapping resources from other productive activities.
> Many illnesses, like taxes, just make stuff more costly without imposing hard boundaries.
Exactly this. Let me try it as a dialog...
A: "So, what is your problem?"
B: "I was cursed by an economist witch. Every time I eat an ice cream, I have to pay $1000 extra tax."
A: "And why exactly is that a problem?"
B: "Because I like ice cream!"
A: "Are you saying that you can't buy ice cream?"
B: "No. I am saying that I can -- I just have to pay $1000 extra for each piece. That's a lot of money for me... but I like ice cream very much... so I typically just buy it once in a few months."
A: "Well, no one can buy unlimited amounts of ice cream anyway, so I don't see what is your problem. We live in a world of limited resources, and we need to choose how to spend our money, according to our preferences."
B: "I don't want unlimited amounts of ice cream. Once a day, even once a week, would be enough. All my friends are having ice cream at least once a week. I can't because of this stupid curse. I wonder if there is a ritual that could remove it... and whether my insurance would pay for such ritual..."
A: "Hm. Let me ask you. This curse... is it a constraint, or a preference?"
B: "I don't understand."
A: "If I put a gun to your head and said 'if you don't buy an ice cream, I will shoot you', would you buy one?"
B: "Of course I would!"
A: "If I told you to buy ten, or one hundred, would you?"
B: "Well, yes... although I would need to increase my mortgage for the one hundred ice creams. That is too much!"
A: "Then I don't see what is your problem. You say you want to buy an ice cream. You can buy an ice cream. You can even buy one hundred ice creams, although you prefer to keep the money. You are free to follow your preferences, within the constraints of your budget, just like everyone else."
B: "But also, quite importantly, within the constrains of the curse. Which I want to get rid of."
A: "I thought we have just agreed that the curse is *not* a constraint."
What's up with this guy Caplan? After reading Scott's essay, above, I thought he was a mentally ill person who felt dehumanized by being diagnosed as a something-or-other -- and that's a common reaction, and one I have no trouble sympathizing with. So, I thought, this mentally ill guy, maybe somebody bipolar and often manic, has fixated on Scott as a representative of the profession who diagnosed him as a something, rather than seeing *him.* He wants Scott to say that diagnoses are bullshit so that he can have a win against the profession. And I wondered why Scott was continuing to debate with somebody like that.
But in fact Caplan's a professor of economics & has written several books that sound entertaining and politically incorrect: The Case against Education, Selfish Reasons to Have More Kids. Glanced over his blog for about 3 mins and read a few random bits that were smart, witty and well-written.
But he has this weird bee in his bonnet about Scott. In his Twitter post asking whether Scott's take on witches stealing penises shows Scott's a closert Szaszian he ends with a survey asking "am I crazy or not?" and most common response was "yes, you are crazy." Then I looked at his post about Scott's announcement of his marriage, and didn't see any spinning of that regarding mental illness -- in fact mental illness wasn't mentioned. So whattup?
It's easy to see that Caplan's wrong. I'm more interested in why he's got this weird take on Scott. Anyone know more? Scott, do you?
This is the relevant section in his reaction to Scott's marriage post:
No snark intended, but Scott’s write-up is a big wedding present from him to me. Why? Well, some years ago, Scott almost entirely denied the broad applicability of basic economics:
[...]
So what? Years ago, Scott told us that the “preference/budget distinction is a bad way of dealing with anything more complicated than which brand of shampoo to buy.” Yet he used this standard economic framework to deal with something vastly more complicated: his quest for a life partner.
The preference/budget distinction pervades Scott’s diction. [...] Instead of telling people that there’s only one possible outcome because everything is “like a constraint,” Scott acknowledges that all of us have a vast array of choices – and must strategize to make the most of them. For shampoo and marriage alike.
As if that weren’t enough, Scott explicitly declares that, “Marriage is a contract, no different in theory than an airline’s contract with an airplane manufacturer.” Once again, he’s embracing the standard economic framework of preferences and constraints for life-defining decisions. Shampoo indeed.
But what does this have to do with whether there is such a thing as mental illness? Do you mean because Scott seems to being buying into the preference/budget distinction? So would the idea be that the preference/budget distinction, as applied to mental illness, forces one to say that there is such thing as no mental illness, only unusual choices? If so, that's far from obvious, and I would have expected Caplan to spell that out, and then crow for a coupla paragraphs about how he's been right all along about Scott actually being a Szazian. But he doesn't even spell it out, much less crow. Seems to me kind of a big stretch for you to equate Scott's apparent acceptance of an economic framework with Scott's losing the argument with Caplan (at least in Caplan's opinion) about whether there's such a thing as mental illness. And the fact that Caplan never mentions mental illness here argues against this stretch being valid.
Scott argued against Caplan's logic by saying that the preference/budget distinction of economic logic only applies to a narrow domain. If Scott no longer believes that, then he can no longer rely on that argument when it comes to mental illness and must make a different argument. His argument above is more sophisticated than "just choices like shampoo".
A big stretch for me? There's hardly any of my own writing in my comment above, just quotes from Caplan. Argue against him if you want to. I'll add another quote which linked to that post: "I say his marital search algorithm shows I was right all along."
When I said it was a big stretch for you, I meant a big conceptual stretch to point to the paragraph about preference/budget as an instance of Caplan's spinning Scott's marriage announcement as an acknowledgement of defeat in the mental illness debate. As I said, the little I've seen of Caplan's style in posts about this issue suggest that if he spies something he takes as evidence that Scott is losing the mental illness debate he says very clearly and with great excitement, but here he doesn't mention mental illness, Szasz or any other key words. Maybe Caplan believes that he and Scott are also having a mostly separate many years long debate about economics?
Scott and Caplan are on very friendly terms (mostly) - in no small part because they DID also have "a mostly separate many years long debate about economics". Scott was the only one to blurb Caplan's recent book ("feminism") and had Bryan on his SSC blogroll.. Caplan calls him "Alexander the Great". - The economic debate began (I think) with a rather naive post of Scott he reposted 2017 on SSC https://slatestarcodex.com/2017/02/22/repost-the-non-libertarian-faq/
So he wants to extend this simple economic model to explain most of human psychology and behavior?
Why is this even about mental illness? Doesn’t it apply to a wider variety of issues? Does Caplan think that if I do a math problem and am confused and get the answer wrong, that reveals my revealed preference?
Because in my opinion the broader point is that behavior is cognitively complex, and a variety of mental states make them harder. They can be illness but it could be anything else, like just ordinary tiredness or low mood.
Throwing a party is a complex behavior. It may not even occur to someone to throw a party. Holding a gun to someone’s head and saying, throw a party, is already applying a type of therapy.
I think the main argument under discussion is that some of them do. Scott mentions Caplan defending "all" which seems like an obviously false position. But surely some of the people who have ADHD or depression or Narcissism or whatever would really change with just a small changes in the incentives around them, in a way others with more deep seated/biological issues would not.
Maybe they would change with a small change in the incentives around them, but they don’t control the incentives around them, so their condition still exists in the circumstances where they are. It isn’t their ”preference”.
If I am struggling for years to solve a puzzle, maybe the slightest change in circumstances could help me solve it all at once, or a change in incentives, but that does not mean my inability to solve it initially was my ”preference”.
If I understand it correctly, Caplan’s argument is almost tautological: it just requires defining ”preference” to mean something completely absurd.
Hmmm. Perhaps the current state of marriage can indeed be put down to people deciding to choose and change partners the same way they'd choose a new brand of shampoo?
What's up with Caplan is he is a very smart guy who is emotionally drawn to adopting extreme contrarian positions. I'm not psychoanalyzing him here, he's said it about himself: "Confession: I have been enamored of extreme policies for as long as I can remember. When I was around ten years old, for example, I decided that all smokers should be summarily executed." https://www.econlib.org/archives/2014/11/on_the_complexi.html
Now he's a libertarian of course, but while "execute smokers" and "legalise heroin" are policy goals that would seem to be utterly diametrically opposed to each other, they each at various points won Caplan as an adherent due to their extremity and distance from mainstream discourse.
"Mental illnesses are just socially unacceptable preferences" is just another weird, extreme opinion he holds in large part because its weirdness and extremeness makes it appealing for him.
Thank you. That makes sense. Do you know whether he has long-running weird one-sided debates going on with lots of people, similar to the one he has with Scott? Seems like there's something up with him and Scott, because someone like Caplan, very smart, energetically putting lots of his extreme ideas out there in books and blogs and Tweets, is certainly not going to have any trouble finding people to argue with. So why the fixation on Scott? And it's pretty wacko for him not to get that Scott doesn't even see what's going on as a long-running debate, because from his point of view Caplan's wrong about various points having to do with mental illness, Scott has explained very clearly why he thinks Caplan's wrong, Scott has not heard anything back from Caplan that constitutes a worthwhile rebuttal, so Scott's done with the matter. Why does Caplan think Scott's sneakily signalling that he's really a Szaszian? Either Caplan's kind of crazy, at least as regards Scott, or else he really knows Scott's not signalling anything at all, bit is claiming that Scott is as a way of teasing him. What do you think about all that?
I mean, sort of? I think the Alexander/Caplan mental health debate is the longest running one in terms of numbers of blog posts, but he definitely gets into a lot of arguments with a lot of people and quite often he keeps these arguments going for a long time in one form or another. He's particularly keen on making bets (and has a very good track record).
It's extremely clear that Caplan likes and respects Scott a lot. I think he engages with him a lot (not just on this issue!) in large part because he very much enjoys reading Scott's writing. And even though Scott clearly wrote this particular post in a bit of a frustrated mood, I think the feeling is mutual and Scott likes and respects Caplan as well.
I also think it's clear that Caplan and Alexander do both see this as a long-running exchange, and that this is not the one-sided relationship you're seeing it as. This is probably the 4th post Scott has written about why Caplan is wrong about mental illness.
I agree with Ash. Caplan loves to engage in debate and is one of the biggest and earliest proponents of making public bets. I also think he is genuinely a good natured person but unafraid of engaging in debates that many people would fine uncomfortable or enraging. He is a quirky guy!
I think the one-sided-ness of this mental illness debate is the other way around. If anything, Caplan is UN-fixated.
Scott has now written three lengthy essays on why Caplan is wrong about mental illness, and Caplan has given one brief reply, and then will occasionally casually refer offhand to Scott regarding mental illness as if there is some ongoing "debate." The reason there is no debate is that Caplan doesn't respond properly!
Thanks to all who answered my questions. I am now much better oriented, and can read Scott's piece with the right mental framework about where this debate came from.
People often have a problem with the parts of reality that go against their beliefs. For example, it is difficult for many religious people to accept evolution, because it violates a few assumptions they have (e.g. that all animal species were created at some moment of history, then Adam named them).
As far as I know, Caplan is a libertarian. He makes good points where libertarianism makes good points, and he has a problem accepting the parts of reality that go against the libertarian assumptions.
From libertarian perspective (I am simplifying here, but I believe I am pointing in approximately correct direction), individuals are the most perfect things that exist, and everyone following their own reason and their own will is the best way the society could be organized. The idea of mentally illness just goes against this all. -- It means that for some people, therapy may be better than leaving them alone, even if they disagree. It also means that if you leave people alone, many of them predictably will *not* do the things that they want most, and some of them would even welcome some kind of external influence.
How convenient it would be, for libertarians, if it turned out that the very idea of mental illness is actually just a big hoax promoted by the evil governments! Yeah, it doesn't seem to match some things we can observe, but maybe we can find a clever argument to make those inconvenient facts go away...
One is that he's basically stopped making interesting bets (https://docs.google.com/document/d/1qShKedFJptpxfTHl9MBtHARAiurX-WK6ChrMgQRQz-0/edit). In particular, Caplan made a bunch of extremely strong claims about the signalling theory of education, and then the pandemic came along with a bunch of novel disruptions to schools: tons of natural experiments and data to analyze. And as far as I can tell Caplan didn't even try to bet on any of it, nor has he updated at all on the data that's come out since. Admittedly his argument rests on the idea that in the long-term adults retain very little book-knowledge from their school days, but even with that consideration I'd expect some update based on new empirical findings or specific falsifiable predictions, and they're just not there.
As with his "debate" with Scott, it's like the guy has just abandoned incorporating real-world evidence and now just thinks about how something could conceivably work without checking in with reality to make sure that his thought experiments actually check out.
Far from an expert, so feel free to correct errors.
I heard recently that for many mental illnesses, they are more correctly thought of "syndromes" or collections of symptoms. Having a known, identifiable, physical cause means that it _not_ that mental illness, even if all the actual symptoms are identical. If they find a brain tumor, that means you are not depressed, you just have a tumor that is causing exactly all of the standard depression symptoms. You only have depression if a specific physical cause can _not_ be identified.
As far as this is correct, it may be causing Caplan's issue (and to be clear, I think I disagree with him on this topic). We have explicitly defined many (if not all) mental illnesses as "thing that we do not fully understand". As soon as we fully understand it, it is not longer a mental illness and moves into the realm of standard physical illnesses.
The obvious answer to me is to just admit that we don't fully understand them, not to try and wave them away as nothing but preferences. If you moved to various points in time, nearly every single currently understood physical disease would fit into a similar space.
There was another post where I recalled him being relatively sympathetic to that idea for mental illnesses, but he backed away from it later and I remember asking him why but can't remember the response.
In the practical sense of how medical specialties divvy up the illnesses they treat, there's truth in this. When general paresis of the insane was finally known to be caused by syphilis and a bit later we developed antibiotics, it stopped being a "psychiatric" issue in practice because an injection of penicillin does the trick. If we discovered tomorrow that schizophrenia was caused by some weird infection (it isn't but just for argument's sake), then it would suddenly become the domain of infectious disease doctors or something like that.
Kind of but the reasons physical illnesses cause mental symptoms is not usually clear either, and it would be odd to say that the hallucinations caused by tumors were more real than ones caused by a mental illness.
Also mental illnesses are not ”defined” as what we don’t understand.
Even before viruses were discovered, it could be presumed that most viral diseases were physical illnesses for those people who had a mechanistic model of the universe (ie who didn’t believe in demons and the like).
Some types of OCD can be treated by burning a small brain area neurosurgically. Does that make it a physical illness? Or do we need to know what is going on in that area before we could say that?
We mostly try and group diseases by (medical) cause, not by symptom. This is ultimately because the treatment (if any) depends on the cause. The symptom is a headache, but the disease might be a cracked skull, or bacterial meningitis, or viral meningitis, and the treatment will vary accordingly.
However, before a disease is properly understood, researchers typically try and cluster symptoms together, to get a grip with its broad outline. For example, doctors had a broad notion of what was then called consumption, long before anyone identified Koch's bacillus. This is a necessary step - how do you identify what TB sufferers have in common without a rudimentary notion of TB? - but always tentative until you have the cause. Your proposed categories might not line up with reality.
Depression is a classic mental illness that is not yet well understood. We don't know what causes it, or even if it's really one thing, but we know it isn't caused by tumours. Someone who has a lot of the same symptoms as sufferers of the disease, but whose symptoms are caused by a brain tumour, clearly has a different disease, in the same way that a cracked skull and bacterial meningitis are two different things, even though they may both show up as a severe headache.
If we found a specific physical cause for depression (e.g. something in your brain chemistry) that would not stop it from being perceived as a mental illness.
>If they find a brain tumor, that means you are not depressed, you just have a tumor that is causing exactly all of the standard depression symptoms.
Depression is a bad example here because "depressed" is both a description of a symptom-cluster and a shorthand for a psychiatric diagnosis. If we find a a brain tumor that's causing your depression symptoms, you're still depressed, but you don't have Major Depressive Disorder.
Your central point is mostly correct: psychiatric disorders are defined as not having identifiable organic causes. But this actually breaks down to a certain extent with depression and anxiety: someone who has depressive symptoms due to neurodegenerative disease, for example, like an autoimmune disorder or dementia, often gets sent back to psychiatry for ongoing treatment under a mental health code even after their organic disease has been diagnosed and is being treated.
So psychiatric diagnoses are really more about whether the "mental health" treatment paradigm is helpful in a particular case than any sort of claim about the underlying reality.
I posted this comment elsewhere but seems relevant (sorry if Scott has already used this example in one of his other rebuttals)—
Bryan needs to explain why someone with schizophrenia who is in the midst of a psychotic episode is fundamentally different than someone in the midst of an epileptic seizure shaking on the floor, in the sense than in both cases the person's brain is in a highly abnormal state which causes various behaviors we wouldn't normally expect. Does the epileptic have a preference for flailing around on the ground? Obviously not.
One central problem (as I see it) is Bryan's belief in free will, which totally clouds his entire analysis. Many of the conditions we are talking about systematically warp and distort peoples' decision-making and judgment. Viewing the mentally ill as just making different choices based on different preferences is the kind of things I would expect to hear from someone who has never actually met a severely mentally ill person in their lives, which I suspect may actually be the case for Bryan (but that's just my suspicion from the way he talks about this stuff).
I think schizophrenics are more responsive to a gun-to-the-head (although not completely). I sometimes watch the "Living Well With Schizophrenia" youtube channel and the creator of that has discussed things like pretending to be healthy to avoid being hospitalized while experiencing psychosis, which is beyond the ability of someone in a seizure.
Of course, if you're having a generalized seizure you aren't even conscious. It's the "abnormal brain state" point I was stressing. Epilepsy is an intentionally extreme comparison. I could use tardive dyskinesia as another example (involuntary movements which are suppressible). Or various kinds of intoxication. Or hell, just plain old anxiety. The examples are endless because agency exists on a spectrum, as another commenter already pointed out.
This seems to just reinforce Scott's point that it's most useful to think of as a sliding scale with the cutoff being decided pragmatically on what actually produces the best outcome, rather than there being a clear separation between preference and constraint.
Probably. Although I think there is sometimes a boundary, if I understand the preference-constraint thing correctly. For example, someone who is psychotic and has a paranoid delusion may be unshakable in that particular set of beliefs. That is effectively a constraint on their thinking. In an overly narrow sense you could say they "prefer" acting in accordance with the delusion but I think that obviously does violence to what everyone understands preferences to actually mean.
A psychotic person can alter their behavior in response to incentives, but they can't cease to experience their delusions or hallucinations in the same way. I've experienced this. No gun to my head would have altered my perceptions (and delusions are quasi-perceptual phenomena imo) any more than a gun to a sane persons's head could get them to see a green apple where there was a red one. But Caplan seems at best unconcerned with such subjective phenomena as the color of an apple.
I'm familiar with these results. I think some far simpler explanations for that small minority of subjects who seemed to be unaware that their answers were incorrect can be offered. They may have misunderstood the task, they may have been too embarrassed to admit to a copying strategy after the fact, or they may have been supplying the answer they thought the experimenter wanted to hear. These are behaviors you might expect to emerge from the group that displayed the highest levels of conformity in the actual trials. I find these possiibilities more likely because of, well, everything else we know about the non-defeasibility of perception. If it were indeed the case that peer pressure could shift the perception of even a minority of people, I think we'd have some every day evidence of that phenomena. (You might have heard a similar argument about social priming results: if social priming were as powerful as some researchers would like to claim, we'd have noticed it!) If there's been any work since supporting Asch's "distortion of perception" interpretation I'd be very curious to read it though.
Could be. The copying strategy case is pretty close to what some of the other subjects admitted. I wouldn't be surprised if some of Asch's subjects truly did have their perceptions changed, since there are other situations where high level information can modify what nominally look like purely bottom-up sensory processing (Necker cube, duck/rabbit ambiguous figure, figure-ground ambiguous interpretations, the dalmatian dog figure etc.).
I don't think that instances of bi-stable or ambiguous perception are comparable to perceiving a shorter line as longer. You'd have me properly intrigued if eg people could be convinced that the Necker Cube was a really a Necker Sphere, that the Duck-Rabbit was just as good a Rat-Mongoose, or that a cow popped out of those blobs as readily as a Dalmatian. But they can't and they aren't and they don't. The fact that it takes a particular set of stimuli to induce these kinds of perceptual ambiguities suggest to me that while top-down information isn't wholly irrelevant, it's heavily constrained by the bottom-up.
Let me wrap this back to hallucinations. Most of the auditory hallucinations I've experienced have been kind of inherently ambiguous and vague (lucky in this regard), and it has very much felt like the verbal interpretations I supply are somewhat superimposed over them. But still, if I 'hear' something like swwshhiwooohppeerf and I grind that out as "Hi what's your preference?" (rarely do I have this robust a hallucination) while I might have alternately heard "Shh, I'm woofing, derp", I would not have heard "Ack! The war machine is coming!" because the internal stimulus didn't really have the capacity to be moulded into that shape.
There are people with untreated psychosis who just get lucky and can function okay even when actively psychotic. He may have talked to some of them, and never to the ones who end up in hospitals.
Caplan seems to suggest we should doubt psychotic people's accounts of their hallucinations and delusions:
"I don’t doubt that people sometimes have genuine hallucinations. Yet as I explain in the original piece, it is reasonable to doubt people when they have shown themselves to be broadly unreliable. Thus, I assume that Scott disbelieves self-reports of alien abduction. Why? Because the kind of people who report alien abduction seem unreliable."
I agree that it seems unlikely Caplan has much experience with any severely mentally ill people. Nor does he seem curious enough to investigate this issue by asking psychiatrists how the psychotic syndrome was conceptualized and how it has been validated. Or how they differentiate psychotic patients from malingerers (something that's usually not too difficult because most malingerers have very aberrant ideas about what constitutes psychosis and put on a poor performance). It's easy (almost definitionally) to account for the behavior of a malingerer (a preference not to be jailed, say), but I don't know how Caplan would describe the preferences and constraints of the typical psychotic patient who suffers through profligate negative outcomes thanks to their symptoms.
I think the fact that Caplan seems eager to deny the reality of interior phenomena like the positive symptoms of psychosis in some cases suggests he does find it a little problematic for his model.
Also: the typical reason we disbelieve alien abduction stories is the content of the claim, not the unreliability of the claimant. We might infer their unreliability after they make such a claim, but this is because we already place low credence on alien abductions, and usually not because of other cues in their behavior that render them untrustworthy. It's an odd example.
Moreover, people who experience psychosis ofter get better, become quite reliable, and report on the experiences they had at the time.
I think maybe an even better analogy than physical illness is injury.
There is an obvious and blatant malfunction when you break your ankle, or lose a finger, or get shot.
And yet, despite the pain, in extreme circumstances we have many examples of people who walk on a broken limb or continue to function for many minutes despite being shot multiple times.
The ability to damage yourself or accept pain to further an important goal in no way contradicts the existence of the injury. The change in behavior while injured is not the same thing as just having a preference for not walking. The treatment for injury is generally obvious, simple, and results in behavioral changes, like walking again, but the treatment is not about changing preferences, it's about changing the actual physical state of the body.
Maybe I don't understand your point, but a key difference (I think?) is that with mental illness people routinely do things they really really wouldn't want to do when in a state of sound mind. So it isn't just about compensating for an injury, but an injury that can lead you to make very bad decisions.
People with an injury often do things they really really wouldn’t want to do when in sound physical health, like remaining flat on the ground while their attacker reloads the gun.
So do e.g. depressed people, which is Scott's point. Unless you claim that they simply have a lesser dis-preference against getting killed, but then the symmetry with physical injury still holds.
"One of the diagnostic criteria for depression is cognitive and memory problems; people with these problems cannot choose to remember things better, even with a gun to their head."
The repeated focus on genuine psychosis when a person's behavior does not respond to any incentives, does not seem like a good example to focus on in the context of Caplan's argument. If indeed it acts purely as a constraint and lacks any of the characteristics of a preference, then fine - it's an illness - and not subject to Caplan's point.
Caplan repeatedly states or implies that some so called mental illnesses are actual illnesses (although unfortunately he rarely elaborates on which these are). So if (some cases of) psychosis qualify, so be it.
But level of psychosis is on a sliding scale, even for the same person and even at different points in the same day, and whether or not they'd respond to the "gun to the head" test at a given point in the day based on level of psychosis is probably a question of varying probability, not one where there are points that would always be yes or always be no. So it seems like it's still a good argument against Caplan, in that he would be forced to admit that a psychotic person is suffering a constraint past some arbitrary cutoff point of how bad their psychosis is at the moment, with their milder moments of psychosis merely being a preference, which seems incoherent.
For Szasz (and therefore presumably Caplan) schizophrenia is the paradigm case of “not really an illness”. I don’t think he accepts that psychosis is a disease.
I suspect Caplan is admitting things like epilepsy, dementia, traumatic brain injury, Parkinson’s, etc.
You're right - I was imprecise in saying "are" instead of "may be," but the point is the same. Since Caplan acknowledges that *some* so called mental illnesses may be real illnesses, while asserting that "a large fraction of what is called mental illness is nothing other than unusual preferences," then focusing on the most extreme edge cases, where Caplan could agree is misleading.
How are psychosis and depression extreme edge cases? They're some of the most common mental illnesses, and symptoms of a lot more. What would you consider a non-edge case?
Caplan's claim was that "a large fraction of what is called mental illness is nothing other than unusual preferences."
Within the set of all cases called mental illnesses, only a very small minority would be under the umbrella category "psychosis," and a smaller minority yet would match the example of "A psychotic person runs out in the street naked claiming to be God."
Psychosis is not a diagnosis; it is a symptom. It's a fairly common symptom (e.g. for BPD from your list), and schizophrenia and other psychosis-heavy mental illnesses carry a very large burden of the cost of mental illness.
It still seems strange to consider it an 'edge case'; a model of mental illness that does not explain psychosis would be a pretty bad one.
Fine, that doesn't affect my point at all. Let's get a ballpark idea of frequency. About a quarter of Americans suffer from a diagnosable mental disorder. What percentage of that 26% run out in the street naked claiming to be God?
It strains credulity to think this is more than e.g. 5% of that group (a whopping 4.4 million Americans) exhibits such behaviors.
This is clearly rare, even within the set of mentally ill people, although the umbrella term psychosis is surely much higher.
Even the set of *all* serious mental illnesses combined including (but not limited to schizophrenia and major depression), only affect 4% of all Americans (https://www.cdc.gov/mentalhealth/learn/index.htm), which is 80+% less than the percentage with any mental illness.
Meaning the vast majority of mental illness cases are minor - far from the extreme examples I highlighted.
>Again, Caplan could say that this is just a preference for attacking cops and then being killed. But in that case he should stop touting the “gun to the head test” as meaningful.
This really seems to miss Caplan's point. The "gun to the head" is a common metaphor for duress, given that typically people don't want be shot in the head. The question is whether incentives can affect someone's behavior.
In the rare case where someone wants to be shot in the head, then being shot in the head is obviously not a disincentive. Such a case tells us nothing about whether incentives could sway the person's behavior like preferences.
But as Scott's counterexamples point out, if you draw the line at "whether incentives can affect someone's behavior" then you can disqualify migraines, rashes, colds, cancer, etc. as being diseases too.
Okay, fine. Caplan's position seems very weak. I had the same objection to Caplan before this post was written. But the fact that an idea is bad doesn't mean that a particular argument against it is good.
My point about misunderstanding the "gun to the head" test is still correct, even if Caplan is wrong because of migraines.
His response makes his position seem more reasonable than presented here, but still seems lacking. I have discovered a truly marvelous proof of this, which this comment is too narrow to contain.
>But as Scott's counterexamples point out, if you draw the line at "whether incentives can affect someone's behavior" then you can disqualify migraines, rashes, colds, cancer, etc. as being diseases too.
No no no no no.
"I am mentally ill." What is wrong with you? "I can't focus enough to hold down a job." What happens if you cannot hold down a job? "Nothing my parents pay my rent for me and make sure I do fine so I lay around all day and play videogames."
This person doesn't have ADHD. They don't need pills, they need life changes.
"I am physically ill." What is wrong with you? "my temperature is 2* above normal and there is mucus streaming out of my body constantly." Ok you really are sick. No amount of futzing with their incentives makes them *not sick*.
Sure EVERYONE replies to incentives. My barely verbal "autistic" sociopath half brother responds to incentives.
The point about incentives is that for some people who claim to be mentally ill, the primary problem in their life is their incentive structure. No amount of incentives are going to make my half brother functional. On the other hand when my sister was a junkie she didn't have "addictive personality disorder", she was just a junkie. As evidenced by her immediately going totally clean once she got pregnant.
If either medication *or* a change to incentive structures can alleviate a person's symptoms, is there necessarily an automatic reason to prefer the latter? Should changing a person's incentive structure be assumed to be less costly (both economically and in terms of preserving family relationships and stability) than just getting some pills? Do you think that anybody in person 1's situation would end up just laying around and playing video games, or do you think that many people could potentially have that safety net but don't need it?
I mean I certainly don't want my taxes/insurance premiums going towards "fixing" person 1 with expensive medication and counseling.
If they want to buy medication/counseling as an easy way out of their lack of motivation that should be on them.
I additionally think there is a moral hazard in medicalizing so many things and giving people excuses where it actively discourages them from having high expectations of themselves. "Oh X is hard, I guess I won't try". Nope dumb dumb valuable things generally are pretty hard, that is why they are seen as valuable.
That you find hour 200 this year of studying organic chemistry boring doesn't mean you are mentally deficient, it means you are a normal person. And if you don't want to stick with it it doesn't mean there is something wrong with you and you need pills. You just weren't cut out to be a chemist. That is fine.
It gets a bit more complicated when the question of who's bearing the costs expands beyond a single family to the whole country (or at least all of the insurance company's clients), yeah, but I'm not sure you can confidently say that the 'just change the incentive structure' solution will keep the costs within that single family either. A person I knew as a child did end up 'having their incentives changed' by getting kicked out of their house, and they've remained homeless - with all the costs to society and the state that implies - to this day, so far as I know.
As far as the moral hazard - yeah, I don't disagree, but there are hazards in pushing the line too far in the other direction, too. "Anyone should be able to do this; plenty of people far more dysfunctional than you in other ways manage it; why can't you?" That way lies bullying, ostracism, depression, suicide. There are going to be tradeoffs no matter where you draw the line.
(And what if the value to society created by having one additional organic chemist, enabled by having those pills, ends up matching or even exceeding the costs of those pills?)
"'I am mentally ill.' What is wrong with you? 'I can't focus enough to hold down a job.' What happens if you cannot hold down a job? 'Nothing my parents pay my rent for me and make sure I do fine so I lay around all day and play videogames.'"
Funny thing is, I was in that situation myself for a few months - and Adderall was the exact thing that fixed it. The whole thing with executive dysfunction is it dulls your ability to respond to incentives in the first place: you simply don't have the mental RAM to actually model/evaluate the consequences of option A over option B. For a long time I couldn't get anything done unless I convinced myself my life was on the line. Now I'm in a much better spot.
(And before you ask, I'm American so I have to pay for the meds myself.)
Again, this seems to deliberately misrepresent Caplan to make him seem sillier. The post keeps talking about people choosing their preferences. When does Caplan talk about choosing preferences? What does it even mean to choose a preference?
I don't think that's what Scott means. Throughout the post he caricatures Caplan as asserting that preferences arise from free will.
E.g. "Maybe depressed people completely voluntarily choose to lie in bed for a few years while falsely claiming to be miserable and then shoot themselves in the head, and all of this is a perfectly free choice that they are happy with."
It sure seems that Scott's point is to imply that Caplan asserts that people choose their preferences and are therefore happy about them.
Caplan explicitly disavowed this portrayal, having written "People often wish they had different preferences, but this hardly shows that what appear to be preferences are actually diseases."
That's exactly the point: they are voluntarily choosing to stay in bed; they are not voluntarily choosing to have the preference to stay in bed all day everyday.
>this hardly shows that what appear to be preferences are actually diseases
Is Caplan claiming that real diseases must impose constraints, and if something happens to only change your preferences in a way that you don't endorse, then tough luck, but it's not a disease? Doesn't seem like a particularly useful definition.
As coming from a family of alcoholics, this example getting brought up again and again by Caplan is especially irritating, given that once you have maintained a certain blood-alcohol level for a very long time, it can be lethal to suddenly stop drinking. Without meds, best case scenario is delirium tremens, or you will have a 45 minute long seizure before you finally drop dead.
>His current post says that either you have to believe that mental illness doesn’t exist and is just voluntary preferences which are stigmatized by society, or you have to believe that homosexuality is objectively a mental illness.
>Not only are each of these incoherent ideas, they’re not even the same incoherent idea! You could easily accept one of the incoherent ideas and reject the other!
Of course they're different ideas! That's the whole point. If they were the same idea, then it wouldn't be either / or.
Caplan explicitly rejects the idea that homosexuality is an illness.
In between heterosexuality and coke v Pepsi: cilantro tasting like soap. Seems a lot like a preference, but we've found the gene and we know the subjective experience is different.
I don't usually hear it classified as a generic disease, probably because it usually doesn't cause much distress, but in a hypothetical culture that put cilantro in everything, or used it for sacred purposes...
This is starting to sound like Social Model of Disability. Come to think of it, a lot of this post sounded like SMoD. Could you write a full post on that some time? I think it could benefit from your analysis.
I am not sure if you have covered this argument somewhere, but even in terms of observable, economically meaningful behavior, Caplan's approach contradicts reality - I mean, how does he explain depressed patients spending considerable time, money and effort on therapy and various medications (that come with nasty side effects) TO GET RID OF THEIR DEPRESSION if being depressed is just something they chose to do?
Thanks. Yeah, that covers it. I am not sure how Caplan's worldview accomodates the possibility that someone hires someone to help them make better choices. If it's all just a matter of preferences that govern how you act within the constraints, and they are not to be questioned, how and why would you prefer, and put in time and effort, to have other preferences than the ones you have? Just have them and be done with it?
Hm. I'm not sure whether there's an observable difference between an "illness" and "an unusual, strong preference that you'd like to change but just can't get rid of", but the former seems conceptually clearer.
Indeed, in the linked post, Caplan writes explicitly in this context "People often wish they had different preferences, but this hardly shows that what appear to be preferences are actually diseases."
He also links to a longer post about this point that I haven't looked at.
Obviously Caplan is arguing from a fallacious, dualist view of human behavior, but it also strikes me that he's performing a bit of rhetorical legerdemain by treating mental illnesses as merely matters of conflicting preferences, whereas I imagine psychotics and schizophrenics would encounter difficulties stemming from their unusual perceptions of reality even if they were living as hermits, or if they inhabited societies exclusively comprising fellow psychotics or schizophrenics
this seems to be an issue Caplan has often. I agree with him on a lot of things (immigration entirely) and find his work on education to be really interesting; but a lot of the time he seems to want to have a debate in a specific way and not in the way most of society sees a problem. His latest book "Don't Be a Feminist" should really be titled "Don't Be a Feminist (as I define feminist which is a very specific type of feminism that is easiest for me to be against but doesn't necessarily describe a large group of feminist thinkers and definitely not in a way they describe themselves)". It's really to his detriment that he does this!
The debate with Scott is a great example of how this is detrimental. There seems to be a large divide between Scott and Caplan on what preferences means and what constraints means. Instead of clarifying how he defines them or using different language that may better communicate his point, Caplan just doubles down on definitions he has written before which doesn't clear anything up!
A masterly rebuttal, but your footnote 4 is wrong. You can prefer something because it is less bad than the alternative; what you think is, the only thing worse than staying in this bed, would be getting out of it. Presumably Caplan has to say that you have chosen to believe that.
Caplan's view just seems sort of crazy. Like, you just read his description of schizophrenia as just being a preference to lie around all day claiming one sees demons, and it's so obviously false. Good job with the reply--very convincing.
Caplan : "Here's what I think. Anyone see anything wrong with it?"
Anyone: "Well, since you ask, here is the GIANT MOUNTAIN OF EVIDENCE you ignore, and the dozen major authors on this subject you clearly haven't read or engaged with."
Caplan: "...no one? Guess I am right about everything."
What confuses me about this whole debate is that Caplan long ago admitted that some things can change both preferences and budgets: "Still, does this mean that the flu isn’t “really” an illness either? No. Rather it means that physical illness often constrains behavioral and changes preferences. When sick, the maximum amount of weight I can bench press falls. (Yes, I’ve actually tried this). Yet in addition, I don’t feel like lifting weights at all when I’m sick. Anyone who has worked while ill should be able to appreciate these dual effects. If you literally get sick, your ability and desire to work both go down. When you metaphorically get “sick of your job,” in contrast, only your desire goes down." https://www.econlib.org/scott-alexander-on-mental-illness-a-belated-reply/
Surely mental illness also affects both ability and preferences, though? E.g. if someone has serious anxiety, that will both affect their preference for avoiding stressful situations, and also affect what they are physically able to do (e.g. speak confidently in front of a crowd without having a panic attack). Gun to the head will not calm them down.
So I kind of have no real problem with Caplan's framework, but also just think he isn't actually applying it?
Caplan asserts that he really can't lift the same amount of weight with the flu. Even if a gun to the head would cause him to lift more weight with the flu, the same gun would cause him to lift even MORE without the flu.
If he can lift 10 tons (look, I have no idea how much weight is involved in lifting or benchpressing and I am profoundly uninterested in finding out) when he's well, and then if he has the flu he feels all ouchy and can only lift 8 tons, but if I put a gun to his head and tell him I'll blow his brains out then he can make himself lift 10 tons even if he's all ouchy - he's choosing to be ouchy. Man up and stop choosing to be ouchy, Bryan! It's all in your mind!
I get that you're intending to ask about willpower in #2, and that in that sense my answer violates the "spirit" of the question, but I also think that the fact that it actually isn't about willpower for a lot of people is relevant and highlights where the spirit has gone astray.
Gun to my head, I'm definitely blacking out due to hypoxia before breathing. This is certainly true for at least any semi-serious freediver, but I suspect far more than that. At some point you're just not "will power" limited and it becomes a question of whether increased incentive can lower heart rate and energy expenditure further -- which is a bit like "As long as you manage to not be afraid, I won't shoot you". Is the person who feels fear there showing a preference for not dying or a preference for dying?
Or just that they're constrained by their ability to organize their preferences?
You might be pleased to learn that when I got as far as the 'restored' painting of Jesus, I laughed out loud (and I'm at work, good job it's early in the morning when there's only one other person in the building as yet).
Mostly this lines up with my views of Bryan Caplan (and to an extent Tyler Cowen) and their views on matters, which are - well, I can't put it the way I feel because that would violate the "Kind/True/Necessary and Principle of Charity" on here.
Let's just say that Caplan annoys me with his total self-assurance that being an expert in one domain means he's an expert on everything under the sun. Ditto Cowen and education.
Would Caplan claim that being diabetic is a choice? You can certainly make that argument about Type II diabetes, and maybe (if you feel like being a particular son of a bitch) about gestational diabetes - after all, nobody *forced* you to get pregnant, but Type I is being born with a deficient pancreas. Unless we're going to invoke karma and the Wheel of the Law and reincarnation, it's tough to argue a Type I diabetic 'chose' that.
" You can imagine the depressed person choosing to throw parties and work hard instead."
This man has his head up his arse, and I'm sorry if this phrasing forces Scott to discipline me, but fucking hell. I'm not formally diagnosed (mostly because I can't convince my doctor) but I'm sorta-kinda depressed at times (at really bad times I have spent something like three days lying in bed sleeping) and yeah, sure, Bryan: I can totes leap up, decide to throw a party (for all zero of my non-existent friends) and have a great fun time getting those extroverted tendencies of mine online and having the ol' happy-reward neurotransmitters flowing! It's only a choice to be crying my fool eyes out over my dumb life!
"One can imagine Bryan Caplan not being a feckin' eejit" but I admit, it's only a thought experiment and not reality.
"In particular, he claims I am FORCED to either accept that all mental illnesses are just “preferences” and so not illnesses at all, or as posited in a response by Emil Kierkegaard, that homosexuality is a mental illness and therefore bad."
I'm willing to bite the bullet on that one. "But - but - that's mean to the gays!" Yeppers, Bryan, I'm not a secular liberal. Whatcha gonna do now?
Or, that's a false dichotomy. Consider someone who is short and fat. You can argue all day about fat - is it an illness (an eating disorder) or a preference (for pizza and doughnuts), but short is just the way things are. Homosexuality is a tall/short, not fat/thin, analog.
Fat/thin attitudes may be changing. Clearly, in cases like the Prader-Willi example, the usual nostrum of "just stop eating so much! have some willpower!" doesn't work.
But for us ordinary chubbos, there may be something more going on as well that isn't down to "just stop shoving junk into your piehole":
"Obesity may damage the brain’s ability to recognize the sensation of fullness and be satisfied after eating fats and sugars, a new study found.
Further, those brain changes may last even after people considered medically obese lose a significant amount of weight — possibly explaining why many people often regain the pounds they lose.
“There was no sign of reversibility — the brains of people with obesity continued to lack the chemical responses that tell the body, ‘OK, you ate enough,’” said Dr. Caroline Apovian, a professor of medicine at Harvard Medical School and codirector of the Center for Weight Management and Wellness at Brigham and Women’s Hospital in Boston."
Note that we usually don’t say that being short *or* being fat is a disease! When someone is short enough or fat enough, we sometimes do, but not just when they’re shorter or fatter than the size that would work best in their social context.
Not really relevant to your larger point, but when you wrote "ditto Cowen and education", were you perhaps also thinking of Bryan Caplan, who wrote The Case Against Education?
I have kind of find hard time calling Down syndrome an illness. It is called a syndrome for a reason.
And also it doesn't really matter if we call it an illness or not. It is relevant only for people who care about policies and so on. They can use their own definitions if they wish, it shouldn't impact scientific thought.
This whole discussion sounds more like Plutonians debating if Pluto is a planet or not (a reference from Rick and Morty). If by scientific calculations their activities are damaging their environment what does it matter how they call it? It only matters for politicians who don't care about science but votes even if it leads to the ruin of the whole planet.
The same thing about Down syndrome. It is bad, we know it. We don't need to call it an illness to understand this fact. If we could invent a drug that fixes the genetics of a new-born baby, we would gladly use it. Maybe one day we will have such genetic cure. Maybe for those scientists who work on this problem it is important to call it an illness because those who give grants for these studies, follow bureaucratic rules – if it is an illness, you get money, if it is called something less – nada. But once you get money, you don't need to call it an illness anymore. Just study the thing whatever it is called.
"We would gladly use it" - that is you and me and most. Not all. We do have the option now to check for Down very early in pregnancy. And there are a few people saying: "foul! Insurances should not pay and people should not test!" Not because they anti-abortion (we are in Europe), but they do not want Downies to be aborted, because: less Down is bad.
Pregnancy, foetus, abortion are a very sensitive issues. I will not delve deeper into them. But once a child is born the legal position is much less controversial. In this case you have a child with a serious deficiency. If we could cure it, it would be mandatory.
Parents who refuse to accept the treatment would be accused of neglect and deprived custody of the child. It would be regarded in the same way as parents who starve their children or deprive them of a normal physical and mental development or do not provide necessary medical help if they happen to have a treatable disease.
Some people get jumpy when they see phrases like "fixes the genetics." Perhaps that's because they have a mental illness that can, in principle, be fixed too.
Perhaps. I have personal reasons to be wary of eliminating conditions like depression and Down's syndrome, so I may be overgeneralizing from my experience. I have severe problems with depression and concomitant addiction. I'm currently trying to recover from a crisis. All the same, I'm not sure I would take a pill that eliminated my depression because I don't know what else would go with it. Perhaps the tendency toward harmful rumination is a maladaptive extension of our capacity for introspection, and so the elimination of depression would also undercut a major source of human creativity. Robert Sapolsky's argument about the evolutionary value of a little, but not too much schizotypal disorder in a population leaps to mind.
Going further down this road, my maternal grandparents' youngest child, my Uncle Joe, had Down's syndrome. My grandparents did a remarkable thing by raising him in their own home, because he was born at the tail end of the era in which institutionalization was the standard. Joe was the kindest, gentlest, most Christ-like person I've ever known. His devotion to his family was unstinting. He needed a lot of help, but he lived well. It's hard for me to accept the idea that it would be better had Joe never been born.
I might be jumpy because of my familiarity with the history of eugenics. I might also be jumpy because I have observed humanity's tendency to apply Fordist principles of standardization in realms where they are not appropriate, such as biology & ecology, where diversity provides resilience. Heck, I might just be jumpy because I remember my Tolkien: "Many that live deserve death. Some that die deserve life. Can you give it to them, Frodo? Do not be too eager to deal out death in judgment. Even the very wise cannot see all ends."
I use "death" here cautiously, but with purpose. In the current framing of American politics, I am staunchly pro-choice, since the question of when a fetus gains the rights of personhood is a terribly complex question, best answered by the women who most directly bear the risks and responsibilities of pregnancy. The standard of fetal viability used in my home state of Illinois seems reasonable to me. All the same, I have qualms about abortion, since if it does not end a human life, it certainly ends a process that will lead to a human life absent problems that lead to miscarriage. I despise Bill Clinton, but his formulation "safe, legal, and rare" seems entirely correct to me.
I don't understand what Scott's stated opposition to congenital conditions like depression and Down's syndrome means. Even if we fully understood the genetic predicates of these conditions, how can this belief be made operational without resorting to eugenicist techniques?
What if we invented a pill that makes everyone's IQ to increase to 100 if it's currently below that threshold or stay the same if they're already above 100 IQ? Would we then classify everyone below 100 IQ as "diseased"?
It makes sense to take a pill to improve IQ, if it is a safe medicine. I mean, there are always some effects and some people will suffer from them. But as with vaccines we draw a line when very rare side effects are considered an acceptable risk so that everybody has to get them.
An adult can still choose not to take them. It is a free choice. But with children it would be considered a bad parenting to not give such a pill to your children if they need it. Probably at certain threshold could be considered a child abuse and parental rights to be taken away.
With the current ADHD medication I would say that side effects can still be quite severe, like stunting of growth and tolerance, even addiction to amphetamines, so I don't think they will be ever mandated to those who are veritably diagnosed with ADHD. We also don't have much evidence that they improve life outcomes, life expectancy or whatever. I remember that some studies even claim that they don't even improve learning just makes life easier for kids and their carers. Which is a good thing but insufficient for mandate.
A quick web search on Bryan Caplan reveals that he is an economist. As such he must have had rational actor decision-related concepts, such as "choice" and "preference", drummed into him for years and years and sees the world through that lens, and has a desperate urge to shoehorn everything into the same rubrik.
So it sounds like he is trying to extend this outlook from economics into areas where it is often not appropriate (just as it often isn't in economics itself - hence all economists' notoriously duff predictions!)
Although I find the intro to your post amusing (Caplin is a relatively well-known individual adjacent to the rationalist-verse) I think I have to agree with your conclusion: the compulsion to reject stated preferences in favor of revealed preferences is a strong one for the modern economist, as it should be, and I think Caplin is taking this framework and running with it where he should not tread.
"In fact, if you put a gun to the alcoholic’s head and threatened to shoot him if he ever drank again, probably he would stop drinking."
He's never been around hardcore alkies, has he? I know that he's trying to pull the Chestertonian gambit as in "Manalive" collection of short stories, where the protagonist pulls a gun on a professor who is droning on about non-existence being better than life, and shocking him into wanting to live, but that's not a good method in reality (not unless you're sure that the person is only repeating fashionable absurdities and doesn't really hold to them deep down).
Put a gun to a depressed person's head, say you'll shoot unless they start throwing wild parties, and it's likely (if you pick the right - which is to say the wrong -time) that they'll say "Pull the trigger, I'd rather be dead". I've had those moments and I've had to claw my way out of them, and fuck you Caplan, it's NOT a 'choice' to do so. It's a struggle and a fight and I wish there were no necessity for it. I'm not "choosing" to feel like this, and you can't wave about "but you admit you fight your way out of it" as a choice. Mostly it's because I was raised that suicide is a mortal sin. Knock that pillar out from under me, and I'll take the other path one of the dark days.
"If we don’t like them, we should ask the people who have them to choose differently, instead of treating them as diseased."
Okay, I'm crying laughing here. So back in my social housing days, we should have said to our schizophrenic client "I'm sorry, but instead of sending the workmen out to change the locks on your doors yet again, can you instead *choose* not to believe that your neighbours are breaking into your house to smear shit on the kitchen walls and that we are sending cameras up through the toilet in order to spy on you on behalf of the government? Thanks!"
Yes, Bryan, *that* would *surely* work!
Same person on her meds: was stable, was able to hold down a job, did not phone or call into the office in a highly excited state
Same person off her meds: see above for sample of behaviour/thoughts.
As someone who is in this post, I feel like I need to speak up for Caplan here, because while I agree with you that while I don't feel like I can choose my mental states (I mean: technically I can; otherwise there's no way CBT could work), what to do about them very much does feel like a choice to me.
I have been very hard on Caplan, but he (and Cowen) do have this breezy way of "now, if we just apply economics, then every problem can be solved" that annoys me (it's not confined to them, of course; every tuppenny-ha'penny expert in his own field thinks that expertise in one area transfers over to all areas, particularly if those other areas are considered 'well you don't need to be smart, not like what I am in this area, to get on well in this').
So yeah, there is an element of choice there. CBT is a good example; in the end, it's up to *you* to put the principles into practice and work on your problems.
But somebody in the grip of full-blown paranoid episode is not really capable of choosing not to be a screaming maniac. You can say they're 'choosing' to break windows or strip naked and run into the middle of the street ranting or cover their walls in tinfoil because the Greys have replaced their family members with replicants who are beaming mind-control rays into their brain, but that's not really a choice as most people would understand it, and you certainly can't talk them down from that - after all, the person trying to persuade them to take off the tinfoil and accept their family members as really their family members, well, that's what an agent of the Greys would say, wouldn't they?
You have to wait until they're medicated out of the episode or otherwise come back to their senses. There's no "choosing" not to be out of their gourd at times like that, anymore than there is "choosing" to be thirsty when you're in the desert at noon under the sun sweating with no water in your bottle.
I think part of the problem here is that both sides in the discussion are more than happy to ride past one another tilting valiantly at the strawman they set up.
The steelman version of Caplan's argument is that while not all mental illnesses are amenable to choosing, a great many are. Alcoholics Anonymous aren't pumping their members up with heroic doses of drugs, to my knowledge. If they are good for anything (and I gather they are at least moderately successful at their core purpose), it's through persuading their members to choose not to drink. Plus, I understand that AA doesn't treat alcoholism as a disease to begin with.
We aren't forced to adopt an all-or-nothing approach to the question. We can, in fact, examine each individual case on its own merits.
>One might as well claim that a paralyzed person could walk if they wanted, but chooses not to.
The kicker being, some of them can. My uncle pulled it off after three years of paralysis (before I was born). He's mentioned a few paralyzed people have had more feeling in their legs than when he started out, but they didn't recover because they didn't have the ridiculous drive necessary to get back up (the kind that lets you focus on standing up for a full year without making any visible progress to standing up, and then fall on your face a thousand times).
Kaplan's a dualist. He's talking about the preferences of a homunculus, not about the preferences of a body. The human brain has been dissected many times and no one has ever found a homunculus. The homunculus is an illusion (a kind of narrative component) that focusses the brain's modelling capacity on the successful survival of the organism. Our modelling is always imperfect, limited and faulty - mental illness makes it even worse.
Modelling human organisms as disembodied preferences is useful in economics. It's still a model.
Non-dualists can endorse a "revealed preferences" model too. Caplan's problem, as far as I can tell, is that he treats those preferences as irreducible, which, one, is obviously wrong, and two, cripples your model for no good reason. A bad reason might be emotional attachment to the incoherent "free will" concept, in which case, well, like the saying goes, you can't argue somebody out of something they weren't argued into.
Agree. It's the irreducible bit that is weird, like if it's not in the economic model it doesn't exist. The difference between stated or presumed preferences and and revealed preferences can be very illuminating at times but they are skin deep, by definition. In this context, we might say that psychology is where you go when preferences fail.
Caplan is making two fundamental errors, which I think Scott realises but doesn't explicitly name. Caplan shows that it is POSSIBLE to account for the behaviours that mentally ill people exhibit by invoking preferences, but mistakes this for showing that it is PLAUSIBLE. Scott’s counterexamples are basically ways of saying that the way ‘preferences’ are being invoked by Caplan are a) stretching the concept in ad hoc ways one wouldn’t do unless one was trying to win a very specific argument, b) basically unfalsifiable and c) simply not the best explanation of what we observe. The reason they exhibit these flaws is because Caplan treats preference as ‘thing that produces action’ and not noticing. No matter what change in behaviour is observed, it will be trivially true that preferences can explain it on this way of thinking. The only way you could even notice the flaw with this characterisation is by realising that it does a bad job of accounting for how we describe our other observations, and that requires noticing that ‘it is possible to explain their behaviour in this fashion’ is no where near showing that it’s a satisfying explanation. If someone’s behaviour can be explained by preferences no matter what they do, then ‘preference’ isn’t doing any useful explanatory work – it generates no predictions and constrains no anticipations. It’s the same flaw that affected early psychological work on traits. If you say someone is courageous in setting A but not in B, you can salvage any given trait in your explanation by just saying they have the trait of ‘Courageousness-in-A-but-not-in-B’, which works for literally any observation.
The second error is that Caplan’s test for ‘is a preference’ is ‘is voluntary’, but his test for ‘is voluntary’ is simply ‘is changeable by will given a certain stimuli’ [namely a gun to the head]. But he needs an argument for why we should think of ‘voluntary’ and ‘changeable by will given some stimuli’ are the same given they clearly are not the same concept at all, evidenced by a) folk understanding of those concepts, b) volitional will as a capacity clearly comes in degrees, and c) volitional capacity in some circumstances cleraly does not translate to volitional capacity in all cases (gun to head is likely to spike adrenaline which clearly increases some capacities you would otherwise lack).
> ‘voluntary’ and ‘changeable by will given some stimuli’ are the same given they clearly are not the same concept at all, evidenced by a) folk understanding of those concepts
'Voluntary' invokes will but it also typically is understood as something like 'uncoerced'. We typically say that if I put a gun to your head and say 'give me your money or I'll kill you' you do not hand over your money voluntarily, even if the means by which you hand over the money rather than get shot is by using your will. It would be silly to say that since there was will involved, it was voluntary and therefore just a preference / choice like any other. One can quibble about whether this is how we collectively *ought* to think of 'voluntariness' (maybe you think most people are wrong about what the concept entails) but they're going to need other criteria / arguments *independent* of 'here's how I define voluntary'
What is the bottom line? How does this help us find out what criteria justify treating adults as incapable of choosing for themselves, or how how best to deal with them?
It almost makes sense to say, we should treat diseases and have compassion for those infected, and should not treat preferences and perhaps not have compassion in some instances. But then aren’t we ruling out diets or diet pills as treatments for being overweight? Why shouldn’t we treat preferences, if the person with the preference consents to altering that preference?
Szasz was against electroshock therapy, lobotomy, and probably would have resisted many pharmaceutical therapies for depression and other mental complaints. But again it isn’t the preference/disease distinction that is driving things. For adults that seem able to take responsibility for themselves (or whatever we should call the criteria), treating them should depend on do they want to be treated, is the treatment likely to be safe and effective, etc. But there are some people that do not wish to be treated, but also seem unable to take care of themselves. Szasz, Caplan (presumably), and maybe Foucault think that society transforms a social problem into a medical problem, using medical style treatments to control some who are difficult to control otherwise. In the extreme cases, there is something to be said in favor of this, just as society puts violent criminals in jail. The important question concerns when this is justified and when it is not. If murder was a symptom of an untreatable disease, isolation of murderers would still be justified.
So why are we arguing over disease versus preference?
>society transforms a social problem into a medical problem
Only if your definition of medical problems excludes social problems. Which is of course the underlying issue with all the brouhaha about homosexuality. Society changed its opinion leaving medicine hung out to dry.
Admire and love both of you (met Bryan&family last Sunday). The controversy pains me - though, as it leads to more posts by both of you, it must be a good thing. Scott "wins" by 42 to 3, I'd say. Nitpicking, here the three: 1. Scott's former "preference is just good for shampoo" was silly, (could count double for not admitting). Scott does much better now, saying "The middle three (straight, gay, even depression!) aren’t obvious, which is why we’re having this debate."
2. Scott about Caplan: "his position: that ALL mental illness is just voluntary preferences" - misrepresentation: even in the last post Caplan writes: "most". (Counting also against Bryan, as he never seems to point out which are "obviously" not preferences and which may not be). Oh, and not sure, Caplan agrees with "just".
3. oops, forgot ... maybe a few very tiny things adding up? Oh: Maybe the Szaszi/Caplan view has its uses?! Like: those addicted/"crazy" enough to defecate on the streets of San Francisco might be more able to control that urge if it resulted in sanctions - maybe birching (Singaporean style - any numbers on how often shit happens there?). But assuming "oh, they can not help it"/"we should help them more" may mostly result in: more crap?
End note: "free will/voluntary/preferences" are ambiguous - see Albert Einstein: "Schopenhauer's saying 'A man may do what he wills, but he cannot will what he wills' has filled me vividly since my youth and has always been a comfort to me seeing and suffering the hardships of life and it remains an inexhaustible source of tolerance."
Caplan says maybe some mental disorders are not entirely a matter of preferences, but he's clear that he's open to the possibility that they all are. That doesn't make much more sense than the view that Scott ascribes to him, and it's open to the same criticisms that Scott makes (just rephrase them as criticisms of the claim that "it's possible that all mental illnesses are just voluntary preferences," which is how Caplan clarifies his view).
His post started: "I have a long-running debate with Scott Alexander about mental illness. I take the Szaszian view that most so-called “mental illnesses” are not illnesses at all, but socially disapproved preferences". Now, I assume someone needs to count the entries in DSM-5-TR. Which is fetishism? Which agoraphobia? - Even if only 30% turn out to come with a relevant measure of "preference"-character - it might turn out that 86% of individual cases do: OCDs/phobias/sex-stuff/depression/ADHD are huge. And even some manic/shizo et al. may. And most addicts knew well enough Marlboro, J. Walker, Js, H and those painkillers to be addictive before they took their first dose. Plus we all are crazy (except Tyler). And man, do we still have some choice. "Most" may not be as mad as it may sound.
Hm fair enough, that's different from what he says later in the essay. But still, I think Scott's examples show that depression, ADHD etc can't be completely characterized in terms of preferences although one can argue that some of their aspects amount to preferences. (As I say elsewhere in these comments, though, it seems much more likely that what's actually happening in these cases is that people's *perceptions* about what will satisfy their preferences are distorted by the disorders, and they actually have quite normal preferences.) So he is still saying something extremely implausible here.
Mostly (!), I agree. (As I mostly agree with Scott in the debate:) - Nitpicking: a) Not sure, Caplan would subscribe to "can be COMPLETELY characterized". ;) He would argue: "preferences" play an important role/ are a not-useless concept with most issues. b) "quite normal preferences": drugs are universal in human cultures. Yanomami-men do love their highs ( David Good ). And I really do like my beer in summer. And winter. And not just beer. (all bottles on my blog are real). otoh: I do spent more time on CK3 than I might want to. Or do I? (Which is NOT my approach to the first wave of the opioid-crisis. Still, if I ever get the cancer that killed my dad, I WANT heroin, not those dumb opioids he got 24/6 (he found those hallucinations annoying - on 7 he died).
I'm trying not to fall into the trap of believing I know Caplan's actual point better than he himself has expressed, but your number three touches on it. I'm really perplexed by all of this - if I'm right I can't explain why Scott and his commentariat are so confused. Perhaps dueling mental inflexibilities regarding terminologies?
Caplan is brilliant at explaining non-purely-economic things in economic terms. He appears to be doing the same here. Perhaps his own terminological inflexibility prevents him from saying, "Look, I don't actually care what these things are. I'm saying we should pretend they're preferences for the sake of making market-based societal improvements.". I don't know what his improvements would be or if they would improve anything, but at least calling it all preference would surface and make legible the problems we have as a society in addressing mental illness.
“Why not say that […] colds are just a preference for lying in bed and blowing your nose a lot?”
Because colds also prevent you from stopping your nose from running, various tissues from swelling, lungs from congesting, etc., even if a gun is held to your head.
Depression causes low serotonin levels and whatever even if a gun is held to your head. Schizophrenia causes hallucinations even if a gun is held to your head. Anorexia causes warped body perception even if etc etc.
AN causes distorted perceptions like how OCD, anxiety, and depression do. People’s perceptions are not entirely subject to will. Distorted perceptions impact capacity.
While AN, OCD, anxiety, and depression are not psychotic, they produce their own distortion fields in a way that could be seen as on the same spectrum in terms of reality testing.
In fact depression can come with psychotic features when it’s really severe and OCD is specified by degree of insight. Panic disorder comes with its own perceptual distortions. I think perceptual distortion may be an important factor that clarifies how the preference frame as it relates to mental disorders is flawed.
If you perceive your thighs to be fat despite objective evidence to the contrary, that continued perceptual distortion is going to create continued barriers to behavioral change and so limit capacity which a different person with less perceptual distortion would have in that situation.
I still don't get it: what is a specific example of something someone with anorexia CANNOT do even with a gun to their head that they could do in that same situation if they didn't have anorexia?
I'm making no claim about capacities for anorexia specifically. The parent comment seemed to be implying that colds are real diseases because their effects persist even if someone is threatened at gunpoint. I was pointing out that many mental illnesses also have effects that persist at gunpoint, and couldn't be arsed to think of a wittier third example than that.
I'll add OCD, ADHD, anxiety, and phobias to your list. Personality disorders pervasively affect people's perceptions in a way that affects their behavior and capacities. Substance abuse also.
If the whole underlying motivation behind Caplan-side arguments is that people have more responsibility for their actions then they feel like people generally acknowledge, then it seems to me it would be helpful to have that conversation instead and as it relates to specific situations. What I often hear running behind these debates is "who is entitled to my sympathy" and I'd rather we talked about that as a moral/philosophical question than deflecting it into this more abstract tussle over definitions.
I guess Caplan would call depression a constraint, and schizophrenia and anorexia preferences. I interpret him as distinguishing physical symptoms from mental ones. This may not hold up, but at least conceptually, we can imagine 3 categories: one where a biological hypothesis seems hard to refute, one where symptoms are purely subjective, and a third where the mechanism is not well understood yet.
I still don’t know how this all cashes out in terms of treatment. I don’t think Caplan is suggesting that people should ignore conditions where persons have preferences that they don’t want, or even that doctors should resist prescribing pharmaceuticals to treat a preference.. What should change depending on which side of the debate is correct, if either?
That is, existing laws pretty much ignore this analysis unless we think that only the legislature and the medical establishment count as competent adults who should determine the parameters of treatment for everyone else.
What would Caplan say in reply? He is unlikely to respond directly for a long time, if ever, so I will try to imagine his response. I found in composing it that Caplan has already spoken to almost all of Scott's points, albeit perhaps not to Scott's (or my!) satisfaction. As a result, I fear Scott and Caplan are talking almost entirely past each other.
Disclaimer: this is my summary of Caplan's views, not mine. I am not a Caplanian!
Scott thinks I have crudely mischaracterised his views, like an ugly restoration of a beautiful medieval painting. Unfortunately, his picture of my views is no more accurate. I am not asking you to adopt a crude behaviourism, but rather the most parsimonious account of mental illness. It is extremely surprising that Scott would attribute behaviourism to me of all people! But while our models should explain, they should not over-explain. We have strong reasons to be sceptical of people's self-report, in circumstances in which they are strongly incentivised to answer in a particular fashion.
On Internal States
Scott's reply bafflingly attributes a strong behaviourism to me. I cheerfully agree that pain is real, that people have internal mental states, and so on. Scott returns again and again to this theme, but I'll just pick out one example as particularly illuminating:
"Maybe depressed people completely voluntarily choose to lie in bed for a few years while falsely claiming to be miserable and then shoot themselves in the head, and all of this is a perfectly free choice that they are happy with."
No, the depressed person is (presumably) genuinely miserable, and not happy with his choices, because his preference (lie in bed) don't line up with his meta-preference - have a fulfilling life. But that doesn't mean his choices weren't voluntary, or indeed, that there was anything unusual about his situation. This is the normal human condition! No doubt many people wish they were more conscientious, or ambitious, or loving. No doubt many people struggle with time-consistency. There is nothing mysterious or diseased about any of this.
On The Analogy To Physical Diseases
Despite what Scott says, I agree that migraines are a real disease. They are (in part) a constraint, that the sufferer can't look at a light without feeling pain. With a gun to the head, the migraineur would still feel the pain. But sitting in a darkened room is no part of that disease. That is a voluntary response which - as Scott is forced to concede - would change under different incentives. Once again, my gun-to-the-head test is vindicated - headaches are a disease, but sitting in the dark is a preference.
More generally, both physical and mental illness can both constrain behaviour and changes preferences. Scott tries to reject this model for physical diseases, writing:
"The essence of colds is feeling unwell and ugh and wanting to stay in bed and having unpleasant congestion in your nasal passages. None of these particularly change your preferences."
But what is "wanting to stay in bed" other than a change in preferences? Similarly, what is "craving for alcohol" other than a very strong preference for it? I do not deny that mental illnesses may go beyond a change in preferences, and to the extent they do, then they are diseases. My argument is that no preference is a disease, not that all mental illnesses are merely preferences.
When the Gun-To-The-Head test fails
I proposed a gun-to-the-head test not because it represents metaphysical truth, but because it's a quick thought experiment to clarify matters. I'll go further than Scott - the gun-to-the-head test can never fully clarify! Just because someone doesn't change their behaviour in response to a gun to the head doesn't mean they couldn't. But my test is far from unfalsifiable. It correctly identifies a subset of behaviours that respond to incentives, and as such are preferences, even if it does not identify others.
Are Preferences And Constraints the Same Thing?
No, and I'm surprised Scott makes the argument. Every concept has some grey area in the extremes, but for this argument to work for Scott, he needs to demonstrate that these grey areas represent the preponderance of mental illness - which he can't, because they don't. If there are some narrow grey areas, we should tread carefully in them, but it does nothing to impugn the vast rolling fields of black and white.
There is no difficulty whatsoever in categorising his 6 examples. An internal state of desire (i.e. a preference) is not the same kind of thing as an internal state of fatigue. Yes, a state of fatigue may lead to an internal state of desire for low activity (i.e. a preference), but that is perfectly normal - all our preferences are grounded somehow. Examples 4, 5 and 6 are purely preferences, and so cannot be diseases. 1, 2, and 3 involve both constraints and preferences, so the constraint aspects of them (fatigue, pain) are diseases, but the preference aspects of them are not. A skin rash is not merely an out-of-nowhere desire to scratch the skin, but if it were, that would not be a disease.
On Justification
We have already established that the constraints and preferences model works extremely well across a wide domain of applicability. But, it seems, if using it suggests that some people may have strong, socially disapproved preferences, then it becomes vital that we discard it for an ad hoc framework with minimal explanatory power that justifies almost anything. We must pretend that these people are somehow compelled beyond their will to take actions, because that makes it easier for us to:
- Force them to make other choices.
- Pretend that their social preferences are normal and that as such they are "good people."
- Indulge our own neuroticism and avoid blame.
I take no position on whether this pretence is socially useful. I am merely pointing out its dishonesty.
> Every concept has some grey area in the extremes, but for this argument to work for Scott, he needs to demonstrate that these grey areas represent the preponderance of mental illness - which he can't, because they don't. If there are some narrow grey areas, we should tread carefully in them, but it does nothing to impugn the vast rolling fields of black and white.
That's a very good way to express my reaction to reading a lot of this. For someone who speaks of rationalism as much as Scott does, I found it quite disappointing to see just how heavily he relied on the all or nothing fallacy here, insinuating that because an idea does not perfectly model reality that it's worthless.
To draw an analogy from the harder sciences, in (macroscopic-scale) physics, Einstein's theory of relativity perfectly models reality as far as we've been able to test. Newton's laws of motion do not. But when engineers need to calculate something, almost every time they'll reach for Newton's equations and not Einstein's, because Newton's work well enough under most circumstances and they're significantly simpler. At extreme edge cases they do break down, and that's where you reach for the more complicated relativity equations, but outside of that very specific territory, Newtonian physics work just fine.
You have come closer than any other commenter (or Scott) to telling me what is at stake in the controversy. You seem to be saying, if these conditions are diseases, we have a better justification for forcing treatments upon the patients (or forcibly preventing them from using a treatment).
I don’t really see it, unless we make additional paternalistic assumptions. We can imagine forcing a treatment upon patients for their own good, or for the good of other people who must deal with them. Whether their condition counts as a disease or a preference seems irrelevant in either case. Someone with a preference may wish to change it using treatment, and someone with a disease may wish not to treat it. We can force a treatment on someone because we think it will improve their lives, or because we think it will improve the social environment. So the disease/preference distinction doesn’t help, unless one or the other has exclusive effects on such outcomes. A narrow consequentialist approach would base the conclusion on the expected outcome of the treatment. A more deontological or broader approach would worry about how to draw the line between competent adults that must consent, and others who need not, since maybe we would not like the outcomes resulting from a world where patients have no rights.
Is there something at stake here that I have missed, or is this all about terminology, with no implications for what ought to happen differently?
These do not follow from Caplan's position in a way that is obvious to me, or conflict with Scott's.
“maybe it's not a problem if the alcoholic drinks himself to death”
Or maybe, even if it is not a disease, it is a problem for the alcoholics and their families. The families certainly, and often the alcoholics, would like to change the behavior, whatever its cause.
“- if alcoholics cause trouble for wider society, maybe instead of treatment we should change the incentives (e.g. harsher punishments)”
We currently do not have punishments specifically for being an alcoholic. Most jurisdictions make public drunkenness illegal, but one can be an alcoholic for decades without ever being arrested for public drunkenness. Many other factors are involved, meaning that punishment would only make sense at all if it worked, which it might not. This is not inconsistent with Caplan's idea, though I have not seen him discuss it explicitly. Maybe thinking of alcoholism as a disease or a preference has implications for how it might effectively be treated, but I am unaware of those implications so far. I would prefer to test cures or treatments, rather than deduce them logically.
“maybe we should talk more about willpower and less about psychiatry”
This could be true or false, whatever we concluded about the preference/disease debate. Nothing prevents psychiatrists from talking about will power, or other non-pharmaceutical treatments, and some do. I think exercise is being advised often.
Or maybe pharmaceuticals sometimes work great to help some people who have no disease with their non-disease problems. Even if we decide they have no disease, why would we conclude they should not use treatments they find helpful?
- “maybe we should abolish insanity defences to crimes”
I have been scanning the article Caplan published that began the controversy, and he at least mentions that Szasz opposed the insanity defense. If so, I think that was perhaps a mistake. My perception is that those who get off on the insanity defense are not let free the next day, but depending on the crime and circumstances, forcibly sent to a mental hospital. I can imagine that defendants exist for whom this would be more appropriate than ordinary prison, whether we think of them as diseased or something else. It might cause legal or ethical complications, if doctors are legally or ethically prevented from treating persons that do not have a diagnosed disease. As noted above, I don’t think it would be unethical if it actually helps the person, or at least if prison is the only alternative, and yet prison is inappropriate some reason. In extreme cases, social control is what we want, and we should admit it. There are dangerous persons, who should be isolated from people who are not prepared to deal with them. I don’t think that the physical cause of the dangerous behavior is particularly relevant by itself.
> Or we could stop thinking like behaviorists, a philosophy which nobody has taken seriously since the 1970s.
I'm not a psychiatrist, so there might be some nuances to this that I'm missing, but the first thing I thought of when I read that sentence is that that would definitely be news to my wife. She's a social worker who works with adults with disabilities who are "in services," and sometimes she talks with me about work. An unnamed psychiatrist who she calls "the behaviorist" is an integral part of setting up and reviewing each client's "plan," and she frequently uses the term "behaviors" as a euphemism/jargon for "doing things they shouldn't be doing."
It certainly seems as if the agency she works for takes behaviorists quite seriously in the 2020s.
The latter is the theory that internal mental states don't exist, only behavior does. I assume the former is a specialist in changing people's behavior.
That seems a little bit absurd, then. Where does the behaviorist think behavior comes from, if not mental states? Does "behavior" somehow abrogate the law of cause and effect? Or is your description simply an oversimplification? (Not trying to attack you or anything; just trying to understand what this theory is.)
The weaker, more sane interpretation of behaviorism is as a plea for parsimony: That a lot of facts about behavior can be explained using models that make no reference to mental states (canonical example: reinforcement learning), and that it is more parsimonious/productive to study psychology through such a lens, treating internal states as a black box, than to propose mentalist explanations for everything.
That said, the strong interpretation of behaviorism (that mental states literally do not exist, or at least are completely un-observable and therefore outside the domain of science) was influential for a while in psychology. I agree that this position is fairly absurd. I think it's best viewed as an over-correction against the early psychoanalytic school, which had a tendency to explain everything in terms of unobservable mental states.
The problem I have with the behaviorist approach is that it leans so heavily into objectivism to the point of effectively denying subjective experience. Or at the least it throws up its hands in the face of overwhelming subjective experience, preferring instead to squint at a tiny sliver of objective data because that's what it's comfortable with.
Scientists will always prefer to lean on objectivism because empiricism is a cornerstone of how we study the world for the past couple hundred years. But when it comes to actually treating a single patient (who isn't the definition of a statistical mean), most often it's the subjective elements that we really care about, and the objective measures ("your biomarker levels are down") don't matter so long as the subjective experience is unchanged. No patient walks into the doctor's office and says, "Doc, my triglycerides are way out of whack, can you help me out?" They might complain about a headache and we check and their blood pressure is high. Then we treat them for hypertension, check the PB is down, but also *confirm with the patient* before claiming that the job is done. If the headache is still there, it doesn't matter what the biomarker says.
In this sense, behaviorism is uniquely unqualified for application to clinical practice.
It is not the theory that internal states don't exist, but rather the approach to psychology where you ignore internal states and focus on observable behaviours. This made a lot of sense a century ago when there was no way to observe what was going on in someone's head, and so you just had to rely on what they told you, which might be complete nonsense - much better to focus on their observed behaviours. Since improvements in medical imaging in the last few decades, behaviourism has mostly been dropped.
Like many things, there is a motte-and-bailey of behaviorism. The "motte" is what you described: the attitude that if we can't measure something directly (using the technology that existed a century ago), let's first check whether we can make a good model without using it. The "bailey" is a positive statement that things outside your model are not real, and anyone talking about them is unscientific and wrong.
Behaviorism can explain many things that people have in common, but cannot explain things where individuals differ, because things like genes or beliefs (or mental illnesses) are outside its domain of study. Therefore, fans of behaviorism are often likely to assume that people are the same... it's just that you are not applying the right incentives consistently. (Which may be true in some cases, but in other cases there may be a genuine difference that is outside the behaviorist model.)
This falls into my area of (not expertise, but "hey I heard about this!") due to where I work.
There is a lot of talk about "behaviours" when dealing with children with special and additional needs, and it's not Behaviourism a la B. F. Skinner. It's more "Little Johnny is going through the biting phase and we need to cope with this". Because it's normal for kids to go through a biting phase, but if little Johnny is biting all the time, biting everyone, and really sinking his teeth into the other kids so they get hurt, then it's a Behaviour of Concern (this is why daycares etc. have Policies On Everything):
Behaviour is anything that a person does that we can see or hear. It is how we act and how we respond in different situations. Behaviour can include laughing, eating, sitting, standing, walking, talking, singing and dancing. Behaviour is learned over time and through experience. This is common to everyone. Behaviour is also communication. This includes talking to others and using body language such as signs and gestures to express how we think and feel. We sometimes use less acceptable behaviours to communicate. These can include using bad language, nagging, shouting and complaining. These behaviours can occur when we are striving to communicate, especially when we feel we are not being understood or listened to. They are often an expression of feeling upset, frustrated and overwhelmed and often children with additional needs may not have developed a set of calming strategies to respond to these feelings.
Are all behaviours a problem?
No, not all behaviours are a problem. Sometimes behaviours may seem unusual e.g., flapping hands, saying the same word or phrase lots of times, throwing items in the air and watching them fall. Your child may display these behaviours because it feels good or helps him or her to self-regulate. Even though these behaviours may appear bothersome to you and/or family members, you need to ask yourself if it is beneficial to worry about them at this time? If however, your child’s behaviours are preventing him or her from taking part in ordinary family life, these would be considered to be behaviours of concern. These behaviours can become a very powerful tool for your child, causing him or her and your family a lot of stress.
What do behaviours of concern look like?
Some children and young people with additional needs display behaviours of concern. Behaviours of concern can include any behaviour that puts either the person themselves at risk and/or those around them. For your child, a behaviour of concern might be a bite, pinch, kick or head bang to either themselves or others. Other examples might be messing up a room in the house, namecalling, using bad language, screaming or refusing to do what you ask. Although these behaviours can be upsetting, worrying and challenging, it is important to remember that your child is using them to communicate with you about how they feel or what they need. It is important to ‘tune in’ to your child’s behaviours and think about what he or she is trying to tell you."
Caplan openly acknowledges that preferences can be changed with drugs. Almost all the objections (in the post and comments) have already been raised and responded to by Caplan.
Everyone appears to be missing Caplan's actual point, but I don't know how to surface this in the comments. I'll go back through and see where I should reply.
Well, Caplan otherwise seems mostly sane, but on this issue he looks like an alien barely trying to impersonate a human. Maybe he has an actual point, which can be rephrased in human language, but I'm lost at what it might be.
I responded higher up to the only other comment that appeared willing to entertain the possibility that he had an actual point to make. My comment was highly speculative and certainly not worth posting twice, but you can find it with ctrl-f.
The one about cities-infesting crazies who just might respond to incentives? I don't think that either Scott or most of the commenters would disagree about this, I certainly wouldn't. What's lacking are not the bright ideas, but the will to implement them.
That's merely one tiny example. The point is that thinking of mental illness as preference might from an economist's perspective open up options that aren't possible from an illness perspective. I won't speculate about what those options might be due to the fact that I've never been once able to predict what people like Caplan or Robin Hanson will come up with. They are just further out in that area of idea-space.
So suppose that I want Scott to write an essay about (say) giant sloths. At this point it feels like a viable strategy would be for me to write an essay saying how one of his posts says that he must have a bunch of incorrect beliefs about sloths (but I present these as true facts). This requires 0 research on my part (I just have to pick random quotes and make up facts), but seems like it has pretty good odds of getting him to do research on giant sloths for me.
This would only work if you were sufficiently famous. But I think it would work on almost anyone! If Elon Musk kept accusing me of having "dangerous and damaging" beliefs about giant sloths, you bet I would write to try and clarify!
>Consider Prader-Willi syndrome, caused by damage to a region of chromosome 15.
Prader-Willi is quite interesting biologically, since it's an epigenetic disorder. In this region of chromosome 15, the father's copy is unmethylated and the mother's copy is methylated (and therefore silenced). If the father's copy is deleted, or the person accidentally gets both chromosome 15 copies from their mother, the genes aren't expressed.
The opposite can also happen, which causes Angelman syndrome.
"Certain neurological conditions involve intense and frequent pain. One example is trigeminal neuralgia, also known as tic douloureux. The term neuralgia stands for pain with a neural origin, and the term trigeminal refers to the trigeminal nerve, the nerve which supplies face tissues and which ferries signals from the face to the brain. Trigeminal neuralgia affects the face, generally on one side and in one sector, for instance the cheek. Suddenly an innocent act such as touching the skin or an even more innocent breeze caressing the same skin may trigger a sudden excruciating pain. People afflicted complain of the sensation of knives' stabbing their flesh, of pins sticking in their skin and bone. Their whole lives may become focused on the pain; they can do or think of nothing else while the jabbing lasts, and the jabbing may come on frequently. Their bodies close in a tight, defensive coil.
For patients in whom the neuralgia is resistant to all available medication, the condition is classified as intractable or refractory. In such cases, neurosurgery can come to the rescue and offer the possibility of relief with a surgical intervention. One modality of treatment attempted in the past was prefrontal leucotomy (described in chapter 4). The results of this intervention illustrate better than any other fact the distinction between pain itself, that is, the perception of a certain class of sensory signals, and suffering, that is, the feeling that comes from perceiving the emotional reaction to that perception.
Consider the following episode, which I witnessed personally, when I was training with Almeida Lima, the neurosurgeon who had helped Egas Moniz develop cerebral angiography and prefrontal leucotomy and in fact had performed the first such operation. Lima, who was not only a skillful surgeon but a compassionate man, had been using a modified leucotomy for the management of intractable pain and was convinced the procedure was justifiable in desperate cases. He wanted me to see an example of the problem from the very beginning.
I vividly recall the particular patient, sitting in bed waiting for the operation. He was crouched in profound suffering, almost immobile, afraid of triggering further pain. Two days after the operation, when Lima and I visited on rounds, he was a different person. He looked relaxed, like anyone else, and was happily absorbed in a game of cards with a companion in his hospital room. Lima asked him about the pain. The man looked up and said cheerfully: "Oh, the pains are the same, but I feel fine now, thank you." Clearly, what the operation seemed to have done, then, was abolish the emotional reaction that is part of what we call pain. It had ended the man's suffering. His facial expression, his voice, and his deportment were those one associates with pleasant states, not pain. But the operation seemed to have done little to the image oflocal alteration in the body region supplied by the trigeminal nerve, and that is why the patient stated that the pains were the same. While the brain could no longer engender suffering, it was still making "images of pain," that is, processing normally the somatosensory mapping of a pain land scape. In addition to what it may tell us about the mechanisms of pain, this example reveals the separation between the image of an entity (the state of biological tissue which equals a pain image) and the image of a body state which qualifies the entity image by dint of juxtaposition in time."
-- Antonio R. Damasio, "Descartes' Error: Emotion, Reason, and the Human Brain"
> “Should we set every tree in the United States on fire, then dump the entire Strategic Uranium Reserve in the Mississippi River?”
It's got to be telling us something that an Englishman who tries as much as possible to ignore US politics and indeed politics in general, immediately thinks: "I bet that's a Republican policy." Bad system-one. Bad.
Yeah, you guys totally need to get another civil war going, you're all slavering for it... Have fun and try not to nuke anywhere else while you're enjoying yourselves.
In Argentina we have a word for people who think like how Caplan is thinking in this situation: we call them "boludos". This word does not necessarily mean a lack of intelligence. Very intelligent people can be boludos. Boludity stems from applying the wrong model of the world to a given situation, often because that model was successful in other situations. This can very easily happen to specialists, people who are very effective in some domain and are led by that success to believe their expertise transfers to other domains. Nobel Prize winners often hang on to very odd beliefs because of this. Chemistry describes everything, until it does not. Models of the word are optimized to obtain certain kinds of information from the world, and often suck at everything else. So you need to be able to switch between different models and frameworks to operate broadly in the world, and that we call wisdom. If you can't figure out which model to apply to each situation, or if you always apply the same model no matter what, like if you had a hammer and the world was made of nails, then you are a boludo.
Two others come to mind: In German, there’s “Fachidiot”. Fach translates to “subject matter”. Someone who might be very apt in a specific field but has a poor grasp of others.
It also reminds me of the French “déformation professionelle”, an expression that you’d use for a teacher who corrects their friends constantly, or a police officer who is looking for rule breakers at all times.
All of this probably falls into the “if you have a hammer, every problem looks like a nail” category
"you’ll sleep less, not more" is linking to the same paper four times. You surely intended to either link four different things or make the whole phrase a single hyperlink
This is an artifact of italicizing a word inside of a hyperlink, part of how browsers parse HTML, not something Scott did specifically. As far as I know there's no way to avoid this, other than avoiding heterogeneous formatting inside of <a href> tags.
- Psychiatric conditions can be modeled using consumer theory.
- Psychiatric conditions can be modeled as changing both budgets and preferences.
- Decisions about classification of conditions, syndroms, diseases, normalcy, etc. involve politics.
- Some psychiatrists engage sometimes engage in what can be modeled as rent seeking behavior.
- The field of psychiatry can be modeled as engaging in some rent seeking behavior (especially historically).
- Some psychiatrists use relatively bad and incoherent models of mental illness.
- Mainstream views in psychiatry (especially historically) have included very flawed models.
The bailey:
- Consumer theory is the best way to model psychiatric conditions.
- All/the vast majority of psychiatric conditions only change preferences.
- Decisions about classification of conditions, syndromes, diseases, normalcy, etc. are entirely political.
- (All) psychiatrists are rent seekers.
- Psychiatry is a homogeneous field (across time and practitioners) and is best understood as primarily engaged in rent seeking behavior.
- (All) psychiatrists use relatively bad and incoherent models of mental illness.
- Models and understanding from the field of psychiatry have nothing to add to our understanding of mental illness and have no benefit over consumer theory.
I'd also appreciate if Caplan would give concrete examples of applying consumer theory to mental illness. This would include giving the concrete illnesses and showing data from which a demand curve can be extracted (including numbers), testing to see how well that demand curve fits various situations, etc.
For example, when quickly skimming I found this study: https://pubmed.ncbi.nlm.nih.gov/26216390/ which estimates the price elasticity of methamphetamine. However, a fully fleshed out consumer theory model of addiction would need a lot more than the study above seems to provide. Among other things, it would seem to need to include:
- The price elasticity of *individuals* in addition to aggregate populations.
- Change in individuals demand curve over time. For example, what does their demand curve look like prior to using substance? After using it 10 times? After using it 100 times? After using it 1000 times?
- Substitute goods.
- Discussion of wellbeing. [To the best of my understanding, one of the central claims of consumer theories in economics is something like the efficient market hypothesis, which claims that consumers maximize their wellbeing by optimizing against their personal demand curves. If this was the case, you would expect addicts to have higher wellbeing with cheap access to drugs rather than expensive access. Is this actually the case? Further, does coercive rehab change preferences? Do addicts have higher or lower wellbeing after coercive rehab and while they don't have access to drugs?]
It is very easy to propose an idea for a model, it is much harder to do the work of turning that idea for a model into an actionable model, and harder still to test the performance of that actionable model in the world. However, the primary value of models comes not from the mere idea of them, but from their fleshed-out and actionable implementation. The field of psychiatry has done a lot of research trying to make empirical models and study outcomes of various interventions. Economics seems to have done much less of this with regards to mental illness. Perhaps the best response to Caplan is "Hmm, seems interesting. I [as a practitioner] will consider this more when you have hard data for it.".
There are two options I can see: One, utility correlates with some sort of internal state of a consumer (e.g. wellbeing or pleasure); two, utility does not correlate with any single internal state or experience.
It sounds like Caplan takes the view that utility falls under option two, in that utility doesn't correlate with any internal state, but instead is a measure constructed to explain the preference-like behavior of consumers. However, this is very strange: it is claiming that humans can be modeled as being utility maximizers, where utility is a single coherent quantitative scale, but also that this utility does not correlate with any single internal state or part of experience (e.g. pleasure). In other words, this utility *is a fiction*. This should call the accuracy of the model into question quite significantly.
I will propose, as an alternative that I believe lies closer to the truth, the model that humans do not maximize utility as a single coherent quantity, but instead perform actions that are reinforced by reward. Reward need not be coherent (i.e., people can prefer A over B, B over C, and C over A). In terms of calculus, one can think of standard conceptions of utility being modeled by a conservative vector field, whereas in reality I believe human behavior is better captured by reward, which is can be modeled by a *non*-conservative vector field.
Perhaps Caplan would be fine with modeling preferences as a non-conservative vector field. However, to the best of my understanding, this would be a pretty significant departure from standard economics, and would require substantially different analysis to understand things like market equilibria, efficiency, and other macroscopic properties that economists typically care about.
I probably should have said "Discussion of utility" instead of "Discussion of wellbeing". I will claim additionally, as above, that I'm not sure that utility makes sense if there does not exist an experiential correlate to utility, which could be phenomena such as wellbeing or pleasure.
"Reward need not be coherent (i.e., people can prefer A over B, B over C, and C over A)."
Yes, I expect that having incoherent preferences would be damning for attempts to model them as rational utility maximizers.
"this utility does not correlate with any single internal state or part of experience (e.g. pleasure). In other words, this utility *is a fiction*."
This, in and of itself, could be less damning for modeling people as utility maximizers. For example, suppose that there were two brain centers with neural firing rates that met all of the consistency criteria for valid utility functions, and the person acted as if they were maximizing the 50:50 average of these two functions (which would then, itself, meet the consistency criteria) - but which does not match any single internal state. Everything would still work.
It seems to me that if there was a part of the brain that averaged these two utility functions, then there should be something that correlates to the averaging.
As a toy example, if we assume people have two parts of their brain, one of which is "sensual pleasure" and the other is "emotional pleasure", and that these to parts are averaged to get "average pleasure", it would make sense that a person feels overall pleasure, but could also subcategorize different feelings as being either sensual pleasure or emotional pleasure.
Now, it could be the case that the utility part of the brain is sub-conscious and not experienced. However, at a meta-level, this might make us believe that utility maximization matters less. Why would I rationally maximize utility if it doesn't make me feel good? This is especially pertinent if we are considering whether society should support people in utility maximization.
I do think path-dependence is a pretty big strike against utility maximization models, though (or, at least the very simple models). Personal actions and the world seem to change the utility function quite frequently, which makes things much more complicated. I.e., if human action modifies our utility function, then most of the complex and interesting behavior in humans might not come from the utility maximization itself, but from the dynamics that govern the evolution of the utility function. Such a situation would seem to be better modeled by differential equations/non-conservative vector fields and not gradient descent on a single-valued function.
"Personal actions and the world seem to change the utility function quite frequently, which makes things much more complicated."
Ouch! This makes it much harder to have a crisply testable theory too. If the utility function "existed" but has a lot of dynamics, trying to extract the differential equations from observations of choices gets very difficult...
"During the course of any given day, individuals diagnosed with substance abuse, ADHD and ASPD act contrary to their impulses because giving in to them would be too expensive. Studies of demand elasticity normally find that consumption of hard drugs is
quite sensitive to price (van Ours 1995); in fact, the psychiatric literature on ‘contingency management’ shows that a high percentage of heavy users of alcohol and drugs will go cold turkey for a moderate price (Higgins and Petry 1999)."
Unfortunately, I don't think that that presents a satisfactory theory of addiction. Heavy users are not necessarily the same as addicts, and those studies are contradicted by the study above (https://pubmed.ncbi.nlm.nih.gov/26216390/) that finds relatively low price elasticity for methamphetamine and "some evidence of a positive cross-price effect for cocaine".
A robust theory of addiction (which would be useful, but is also one of many things that are sometimes modeled as mental illnesses) would need to do the work of comparing and explaining all of these studies and offering robust predictions that perform in the real world.
I think it would do much to clarify the discussion if Caplan provides/has provided a list saying "these are the things that are best modeled by consumer theory, and here are links to the best particular models for each of those typologies, including concrete and testable predictions, and these are the things that are not modeled well by consumer theory".
I read that as various organized groups jostle and argue over definitions, policies, and benefits conferred as it relates to problems people have. Those groups include professional associations, legislative bodies, and citizen groups. The definitions, policies, and benefits change across time and space as a result of this jostling. Scientific research is one input into the jostling but the outcomes are more determined by political processes.
Illness only refers to things we consider *bad*. (For example, mental retardation is considered an illness, high IQ is not. Both are a non-standard value of IQ, but having lower IQ is perceived as bad, therefore an illness, and having higher IQ is perceived as good - or maybe as irrelevant by people like Taleb - therefore not an illness.) This has various implications, for example many people believe that healthcare should help people get rid of the bad things.
But what is "bad", and what is "not bad"? The answer depends on people's personal values and cultural background.
For example, pain would be considered bad by most people, but maybe Mother Theresa would object that it brings you closer to Jesus, therefore chronic pain should be considered a blessing, not an illness. Talking to invisible spirits is considered an illness, because we do not believe in the invisible spirits; in a different culture, a person with the same condition might be a respected shaman. If someone is unable to focus enough to keep a job, that's bad, because we need jobs today. But maybe, in a future where everything is done by robots, no one would even notice this condition. Shortly, for any illness, we can probably imagine a society, where given condition would not be considered a problem, and therefore would not be considered an "illness".
(Though this ignores the part that people with conditions called "not bad" might still suffer from their consequences; even if everyone, hypothetically including themselves, agreed that it is not a problem. The people "close to Jesus" would still feel the pain. The people talking to invisible spirits would still... talk to something that is not there. People who are unable to keep a job now, might be unable to complete some of their personal projects.)
Gonna be honest, I don’t think you can actually arrive at a super robust “mental illness is very real” position without also embracing teleology. (This is no bitter pill for me, cause I DO embrace it, but I think it’s a bitter pill for a lot of people, and a lot of counter-intuitive, non-parsimonious theories abound as a result.)
Could you explain what you mean by "teleology" in this context? Do you mean it in an Aristotelian sense that people and things have specific ends that they exist towards, with the relevance being that mental illness is something that limits the ability to achieve these ends?
Correct! And that it is very difficult to talk about mental illness without both an existent idea of the end of humans and the end of human cognition (as directed towards the former end.)
Does this not apply equally to all forms of illness, injury, or disability? A human is meant to be able to use their bodies for various purposes, not just their minds, and where these purposes are impeded, we focus our attention there and call it illness or disease or disorder or syndrome? Some of it is based on suffering (the reduction of which I suppose comes with its own teleological implications) and some of it based on impaired functioning.
I have to believe that if Caplan were to speak to someone with actual, real-deal, no-foolin' psychosis for more than five minutes he would have a much harder time maintaining his position. Thomas Szasz did not have a lot of experience with psychosis during his training, and I think that is reflected in the positions he arrived at as well. I suspect genuine mania would also challenge his conviction, or very severe OCD.
Please explain to me the set of preferences that leads a Conservative Jewish woman to stop sleeping, repetitive give the Nazi salute to passersby on the street, and rant at the top of her lungs about how she would have been the best prison cafeteria cook at Auschwitz while sobbing and taking a hammer to a mailbox.
Please explain to the me the set of preferences that leads a man who loved working as a car mechanic more than just about anything in the world to become so constrained and confined by his rituals that he lost his job because he it took him 7-10 hours to feel prepared to leave the house, started picking at his skin so severely that he opened up patent holes in his cheeks, and begged to be referred for deep brain stimulator implantation because it was absolutely the last hope he had left for living any life he regarded as worth living.
I don't think it is possible to articulate a coherent idea of "preference" that encompasses these things that bears any relationship to what use the word "preference" to mean in ordinary language. At some point you can use the same syllables to describe a concept but it is far enough away that I think it is perfectly legitimate to say then that some "preferences" are, in fact, disordered.
Some truths are difficult to grasp in a non-experiential fashion.
If satisfying the preferences in question is recognized to be worthless and does not actually contain or provoke any rewarding stimuli of any kind (the OCD) or does not pass the gun to the head test (manic people get shot to death by the police all the time), and both of them are radically different from an honest and true accounting of preferences (that have been true for the rest of the agent's lives and are internally consistent with each other), precisely what information does calling it a preference convey?
I reject the characterization of one's accounting of one's own preferences as "honest and true." Revealed preferences are the true ones. And satisfying those is the only thing of value, by definition, so it can't be "worthless."
The information is the assertion that there is no meaningful distinction between the revealed preferences of the so-called "mentally ill," such as those with "depression" who lie in bed all day, and those considered sane, such as people who to prefer Pepsi to coke, and a characterization of one set of preferences as a disease is an expression of one's subjective judgment.
I agree that if you stipulate your conclusions you are justified in drawing them.
If one cannot regulate one's behavior in response to 'these armed men will shoot me if I do not', and prospectively and retrospectively the agent is confident in their desire to regulate the behavior, in what sense is this a preference?
There is a reason even the radical behaviorists (like the ACT people) feel the need to talk about internal, 'private' events in contradistinction to strict stimulus-response pairings. They feel they are forced to do so by existing empirical evidence and to provide a useful account of individual human behavior. You may want to reflect/investigate what that is so.
You also seem to assume that human minds are unitary and do not have competing and often conflicting drives subserved by relatively independent/encapsulated systems. Behavior that is in some since influenced by choice but is not a preference emerges naturally out of such a system. On what grounds do you maintain that there is a coherent self that even has a single specifiable set of preferences?
I think it helps to be conceptually clear what you mean by the term "preference." Scott is using it to mean a consciously attended mental state that some state of affairs should obtain that one can consciously reject. There clearly are numerous mental conditions, including examples listed above, where that is not reasonably called a preference.
I suspect you mean something more like goal-directed behavior, where the goals are one's preferences and some goals the human organism has just seem odd. This isn't really how the term is used in common parlance or philosophy, though.
When we talk about mental illnesses or physical ones as a "disease" what we mean is that the state one is in is reasonably thought of as undesirable. That is to say, if someone fully understood what their desires are and how to fulfill them, they wouldn't want to be in that state. When it comes to something like "cancer" you can get easy, near universal agreement. When it comes to some things described as mental illness, that's not as obviously true, though when it comes to severe psychosis, it is.
The most popular modern conception of what a mental disorder is relies on a person functioning poorly in their social environment. It should be immediately obvious to you that we can say the problem is more the social environment than it is the person, and if that's the case, then we don't consider that a mental disorder. For instance, if a society banned women from reading books, but some women kept doing so anyway despite risking going to jail, we might call these women as having "book use disorder" but we shouldn't, because a society that bans women from reading books is morally wrong. So it is with, say, homosexuality.
Per the "functioning poorly in their social environment" definition, the obvious analogy to your hypothetical "book use disorder" and homosexuality is pedophilia. Is it that, by miraculous coincidence, your beliefs just happen to be objectively morally true, or are your statements about what mental disorders are really just your own version of what is sinful and what isn't, gussied up in scientific terms?
Given the context, the comment section of an essay about the term used in an economics sense, it ought to be clear that the term "preference" is here used differently than in common parlance. (Among philosophers, your mileage might vary; I do not grant that none of them are intelligent enough to use the term in a more meaningful way, but you might be right.)
The argument was that when you define mental disorders in terms of one's global functioning, as clinicians do, you necessarily import judgments about whether you consider their social environment to be morally appropriate. A compulsive desire to read books despite a ban on reading books or have romantic relationships with members of the same sex despite societal sanction on them is apt to cause dysfunction, but we wouldn't consider that a disorder because we recognize that the problem is with environment.
You can reply to those judgments with, "Oh, so you think you're morally right?" with the answer being, "Yes, yes I do." That's not a miraculous coincidence as I don't think moral judgment is a random walk. But, again, my point is you cannot avoid the moral judgment, if only implicitly.
This does not mean that mental disorders are just a description of what is "sinful." It's not saying that cyclothymia is immoral. Rather, they are a subset of states of being we've described as undesirable. We just have to be careful about thinking *why* they are undesirable which requires us to think about the context in which they exist.
Looking at Caplans argument -- as Scott presents it -- I feel that Caplan started with writing the bottom line "Therefore, mental illness does not exist." before filling the stuff above. Credit where credit is due, I don't think I could come up with a better argument for that proposition.
Just like about every other concept, the concept of illness has some blurry outlines. Preferring Coke to Pepsi is not generally considered an illness, though if Coke had terrible side effects or external costs not present in Pepsi some might give it a fancy Latin name and call it a disease.
Of course, somewhere, someone will call a Deficit in Appreciation for our Supreme Leader a mental illness and try to place people in reeducation camps or something. I think the obvious place to make a stand against this sort of thing is to be against any treatment without (at least prior) consent, not the definition of illness.
Also, why is Scott the one feuding with Caplan over this? Scott's side is not the fringe position here, his opinion is probably shared by more than 80% of the medical professionals. Is it simply that Scott was the most high-profile person to refute Caplan?
Caplan's position is so funny - he's just very obviously taking the abstracted models about human behavior we use to make economic models work and explicitly using them as his understanding of psychology.
Yes, he's taking models that explicitly assume the "rational, self-interested actor" and attempting to apply them to conditions where the whole point is that the person in question ISN'T rational.
I have no horse in this race and no expertise in this field but surely talking cures can’t work unless the brain/mind has a choice. My dad had a major depressive episode a few years back which was cured, only and finally, by pills. Now he is back to normal - and normal is happy and loquacious so that was quite a change and quite a reversal. He had no choice, perhaps, without the pills.
My uncle was treated for alcoholism and as far as I know he just did AA and some retreat or other. Now he doesn’t drink. ( Ok some people say that he still has alcoholism - but I never really believed that - it’s like saying that somebody who eats well still has an eating disorder. Sure both the alcoholic and the ex-anorexic can fall off the bandwagon but while on it they don’t have the disorder).
So in some cases it looks like there is a choice to overcome some mental illness where choice is “this can be fixed without chemical intervention”.
I guess the way I think about this (as a psychotherapist) is that we have varying degrees of influence over the circumstances of our inner and outer lives. It's rare that we have total control and it's rare that we have no control. Most experience happens in the space between where it's about maximizing our influence within constraining conditions (again, conditions being both inner and outer).
One's default mode perceptions and default mode coping repertoire in the face of stressors are inner conditions that act as constraints. It's possible to change those inner conditions and thus loosen constraints (or increase capacities) through medication as well as through learning new ways to perceive and respond to those conditions. Some of that learning happens through words as much learning does, but I wouldn't reduce the intervention to "talk" since the actual change happens in the moment-to-moment way the person moves inside their daily life.
Alfred Marshall, inventor of modern mainstream price theory, used as a motto of his textbook Principles of Economics Latin phrase "Natura non facit saltum", i.e. Nature does not do jumps, attributed to Charles Darwin. Meaning that there are no hard lines between (some) phenomena, just one concept gradually morphing into another through edge cases. Perhaps Caplan needs to be reminded of that, since it obviously applies to a distinction between constraints and preferences.
I just want to say: this post particularly changed my mind. I read Caplan's original piece on how mental illnesses are actually just stigmatized preferences, found it compelling, then went to read your rebuttal but got sidetracked and kind of forgot about the whole thing. As a natural contrarian Bryan's position seemed reasonable to me. Particularly compelling is the parallel to arguments from disgust: people argue condition x is disgusting and therefore immoral, when such a mental model would produce virulent homophobia 70 years ago. Then I read this, remembered the whole thing, and it's pretty clearly wrong. I'm sure you don't particularly enjoy writing posts about this stuff as a psychiatrist, so just know you changed at least one person's mind.
Intellectually dishonest in the sense that he clearly has pretty severe biases that he isn't even trying to resist. Or else a person as smart as he is would acknowledge, or probably even anticipate, the sort of objections Scott raises and abandon this stupid preference view.
Guess what? He did anticipate many of the objections Scott raises, and subsequently responded to them, as well.
Maybe deciding that someone's view is stupid and that they are dishonest should follow reading their actual work, rather than the paraphrases of their work by someone disagreeing with it.
I'm pretty sure I've read every Caplan essay on this topic and I don't agree with your characterization, but it's possible I've missed something! In some cases he's responded in the technical sense that he wrote something he claimed was relevant, but the responses I've seen from him completely missed the point.
How we define "disease" doesn't matter one iota politically. What matters is whether the individual is hurting others. If they are not, you should respect their freedom and not lock them up for thought crime, however irrational or abnormal you think their thoughts are.
If you disagree, fuck you and everything you stand for.
I agree with not locking people up, but it's still useful to call things diseases. Or to rephrase, to say that we should work towards trying to eliminate certain conditions instead of just saying "well the only harm is to yourself, so good luck to you bye". Where the politics comes in is deciding which things are worth eliminating and which are just normal differences. For example, I would say severe OCD is pretty clearly in the former, while disliking alcohol is in the latter. Homosexuality switched from the former to the latter in the popular mindset, while things like schizophrenia and deafness (more of a discussion about "disability" than "disease" for that one) are heavily debated today.
For a more interesting example, consider a person who begs you to restrain them and prevent them from hurting themselves, when outside of the episode -- but during the episode, they yell at you to let them be, and to give them back their guns.
I would say the spectrum you describe is strictly a constraint spectrum and preferences are *informed* by that.
Another way to look at the constraint spectrum: if doing X will result in Y bad thing, the degree to which you are constrained from doing X is proportional to the degree to which Y is bad.
I think Bryan has unusually high willpower/self control/whatever you want to call it, so when he sees people behaving a certain way, he concludes they must prefer this.
One example that should also apply to Bryan himself, is tiredness though. He should imagine he's really tired and can't think straight anymore, such that if he was left alone he would go to sleep. But now we put a gun to his head and force him to stay awake. Fearing for his life, he manages. This looks like a preference. But then we pull out a math test and say if he gets everything right he'll get $100. If he wasn't so tired he would be able to do it, but because he's so tired he's careless and makes many mistakes. That sounds like a constraint!
Furthermore, if we offer him a magical pill which would make his tiredness go away without any side effects, just like if he got a good 8 hours of sleep (i.e., a much better pill than caffeine), then he would probably take it in order to earn that $100! How does that fit into his framework?
So Caplan's model is nuts but your response has given me a framework to think about my own concerns with mental illness diagnosis. I generally think destigmatizing efforts are good but also think there's a cost we're ignoring and was having trouble articulating it.
That cost is that while the line between a constraint and a preference is extremely blurred in reality, we tend to dualistically categorize things in mental models. Imagine a universe in which we have a concept of a lower stigma, culturally accepted Can Only Run 5K Disease. In such a universe children whose legs will never carry them 6k won't suffer horrific pain in gym class - they'll be excused. They'll be mocked less for the inability to run. The diagnosis would make life better for thousands.
People with a strong preference for not running 6K can now model it as a constraint though. Maybe it makes them feel tired and gross and they dislike feeling tired and gross more than the average person. They're correct this mental state can be modeled as either preference or constraint, but modeling it as constraint changes how they think of "problems to be solved by running 6K." Where before it was "It would be extremely difficult for me to run 6K" now it's "I cannot run 6K." I can't prove it but I suspect that if you put a gun to the head of a person absolutely convinced they had this condition, who had made this condition part of their identity and made Facebook posts about how it impacted their life, they would be able to run exactly 5K before being shot.
Can you keep going and say what the cost is we're ignoring that you're concerned about? Is it loss of "grittiness" in kids? Is it more malingering? Is it moral offense at what people do for attention? Is it cost to the healthcare system? Some other thing?
None of those, except maybe something adjacent to "grittiness." Just simply the loss of choice - mentally converting something you don't want to do into something you can't do turns "achievements with high cost" into "achievements eternally out of reach." I think the ground lost here is probably less than the ground lost by assuming everyone can do anything. I have just debilitating ADHD and tried to be a lawyer - I was paying far far far too high a price to accomplish a goal that didn't make sense for me and in a culture that treated that as a moral failure (like U.S. culture only 10 years ago) I would never have recognized that. Nonetheless, some ground is lost.
Trying to understand your personal comment. Your assessment about trying to be a lawyer while having debilitating ADHD is that the trying so hard didn't make sense but you had a hard time seeing that because our culture considers any shortfall of trying hard to be a moral failing? I think I'm not understanding that relative to what ground is being lost.
The ground being lost if I understand you is that people with mild mental health type impairments will say "can't" when they should be saying "could" and that medical support for "giving up" in a way will deny society and individuals the benefits of their trying harder?
Why is trying so hard valorized so much I wonder? I get that we like hero stories and people fighting against all odds to overcome impossible hurdles. But isn't it good that we also find ways to make functioning well in good health easier for more people to accomplish? We could take a stand that we should only be treating people with moderate to severe mental illness but everyone with mild impairment should buck up or be shamed for laziness. But does that not come with its own costs, that approach?
My experience working clinically is that the people who have mild impairment are one life stressor away from moderate to severe impairment. This is also supported by research on the value of treating psychosis in its early prodromal stages before it blossoms into the full-blown creature. We know this about pain management as well. The resources it takes to treat something when it moves to moderate and severe are a lot higher.
Yes I don't think I was very clear with that. My point was that this is a spectrum - on one extreme we assume everything is an issue of willpower, on the other we deny willpower altogether. In the first case people (including my example) endure a lot of distress for no good reason. In the latter, people are led to believe they literally cannot do things that are simply difficult.
The point of sharing the personal anecdote was to make it clear I think it's *good* we're moving away from a "pull yourself up by your bootstraps" mentality - that the fact that I think we're ignoring a cost doesn't mean the cost isn't worth paying.
But ignoring the cost altogether as some subcultures are doing will eventually cause us to choose a place on the spectrum that is too far away from the personal responsibility side of things. And I've had a lot of trouble articulating why this bothers me. But the "preference/constraint" dichotomy helps me to articulate this - constraints, generally, are bad. Like the rabbit/duck illusion, many issues tackled by mental illness diagnoses can be viewed either way. Viewing a constraint as a strong preference or costly expenditure is sometimes more useful than viewing a strong preference as an absolute constraint. I'd just hate to live in a world that didn't recognize that.
A personal note that might cloud that point yet again - I legitimately don't know whether my inability to practice law is better modeled as a constraint or a strong preference. Certainly if my boss came to me and said I'd be fired if I didn't shape up, I could be more successful for a few weeks or months. I don't know whether I would have been capable of sustaining that level of effort long-term. But crucially for me what actually caused me to move on was the belief that even if I was successful it would have been at much too high a cost. That let me model it as a choice instead of a failure.
Thank you for saying all this, totally clear to me now. Or at least I think so. :)
I can add on to what you say from a couple of other angles. I agree with all you say.
One angle is one of my favorite Buddhist quotes which is "The great way is not difficult for those who have no preferences." (3rd Zen patriarch). I think about that up against what I saw for a time as the dominant mode of parenting among lots of people I knew which involved offering up a small child a huge number of meaningless choices. "Would you like apple, grape, orange, or pear juice? Would you like it in the blue, yellow, or purple sippy cup? Would you like to have it while watching this, that, or the other video?" In certain corners of society, it seems to me we're behaviorally programming people to have absurdly specific preferences in a way that can only contribute to delusion and suffering. It's good to learn to travel light and know you'll be okay.
The other thing is I think people having a sense of agency over their own lives is really important for mental health. So like your modeling not being a lawyer anymore (if I understand that right) as a choice instead of a failure seems really healthy to me. The whole notion of failure seems unhealthy to me unless we're talking about in the sense of "I tried that experiment and it failed, so now I'm trying another one." I think it's really important for people of whatever mental condition to have sense of capacity to learn to shape their lived experience. To not get stuck in victimhood stories and to identify choice wherever we can.
Despite being essentially a perfect rebuttal, this is obviously not going to change Caplan’s mind. Perhaps a day of shadowing in a psychiatric ward would do this trick? After the requisite HIPAA training of course.
I can’t image meeting a poorly controlled Schizophrenic patient and then still thinking mental illness doesn’t exist.
One thing seems to have clearly emerged from all this, uh, debate: we should really be using the “point a gun at the patient’s head” treatment for a whole swathe of somatic and mental illnesses, since it appears relatively effective in many circumstances.
Feels a lot like the Harris/Dennett "debate" about Free Will. Harris says "particles, therefore no free will". Dennett says "I object to the definition you use, because of the following reasons (...)" and then Harris replies with "but particles, therefore no free will!"
I don't think either Harris or Caplan intends to be foil for a more careful thinker.
Scott, you write as if Bryan is doing something nefarious, in his manner of pursuing this issue.
But I think the truth is, he's just not paying that much attention to you. As in, I bet he's not even really reading (at least, with much attention) those long essays you write. He simply isn't *crediting* what you write all that much. He then goes on to reassert his position in a manner that a third-party might (perhaps rudely?) describe as a bit autistic.
Now, you might think that such behavior *is* nefarious. And maybe it is, at some meta-level concerning modes of discourse. But my point is that he's not deliberately trolling, or anything like that. He's just lacking self-awareness in a certain characteristic manner.
Many people, very much including rationalists, behave in exactly this way on other topics. It's just cognitively really difficult to closely attend to arguments that one viscerally does not credit. I've encountered that very mode here, e.g., on the topic of whether witches were historically real (in terms of their own self-understanding and practice).
I think the view that "mental illness is a preference" makes a sort of sense in a dualistic framework. If the incorruptible soul/perfect mind controls the body, then any disorder beyound brute physical constraints (i.e. failure to run due to arthritis) must be a choice, by definition.
If, however, the mind is a function of the body just like everything else a person does, then the mind can be damaged just like every other function. Arthritis makes it impossible for people to run, and depression makes it impossible for them to choose to run, but we have medications to at least mitigate both of these conditions; and in cases where we don't, we're working on it.
So, if that's what it comes down to, then the two positions (mental illness as a preference vs. as an illness) are irreconcilable, because dualism requires a kind of faith by definition (not necessarily religious faith, but still), and faith cannot be adequately communicated (though we're working on it). It just depends on which side of the coin-flip you happened to land, ontologically speaking.
>that all mental illness is just voluntary preferences
is a terrible position, I cannot imagine he actually believes that. That said I do tend to think he is correct that a big portion (possibly a majority) of what is treated as "mental illness" is really just people with poor preferences/behavioral ruts (particularly ADHD/depression).
Though perhaps that is not the central case as I think that is mostly the "frosting" of mental illness treatment, and the cake is the truly crazy people.
He doesn't. Scott is deliberately misrepresenting him.
While normally I'm loathe to attribute to malice that which is adequately explained by incompetence, Caplan writes explicitly that "a large fraction of what is called mental illness is nothing other than unusual preferences," and already explicitly corrected Scott for accusing him of asserting that all mental illness is a matter of preference.
Why is Scott so careful/thoughtful sometimes and so....well not...others? I don't understand. I guess when you have made as many blog posts as he has you start treating them with the casualness I treat my comments.
When somebody attacks things that Scott considers valuable/important he sadly loses his usual dispassionate rigor. It still makes for entertaining snarky reads, so I'm not complaining. Nobody's perfect, and he's far better than most.
I don't think Scott is being excessively casual. The fact is that while Caplan does acknowledge that there are some "genuine" mental disorders (in his thinking, one that impose genuine constraints), he sees these as very unusual and niche situations. Saying "Caplan thinks all mental illness is voluntary preferences" is kind of like saying "The earth is round". It's not exactly true, but it's true enough for all practical purposes.
No, Caplan doesn't think that literally all mental illness is unusual preferences. But he does think that's true of virtually any example of mental illness that you care to think of.
Explicitly, Caplan is saying that while he is leaving open the possibility that some mental illnesses are real, he cannot identify one. This is not a limited, modest position only applying to to dubious ADHD claims. This is an expansive position applying to every mental illness he can think of - the only modesty or restraint here is that he acknowledges that there might be some case or situation that is escaping his awareness.
I am not making excuses for Scott. I have no problem disagreeing with Scott and do so quite often. But you are misreading Caplan and imagining him to be saying something saner and more reasonable than he actually is saying.
Caplan doesn't argue that *all* mental illness is voluntary preferences, but he sure does argue that a whole lot of it is:
"Most glaringly, a large fraction of what is called mental illness is nothing other than unusual
preferences – fully compatible with basic consumer theory."
And he explicitly expands his analysis beyond things like ADHD to include things like schizophrenia as one example of a not-mental-illness, and even hallucination (he acknowledges that if you really do see things that aren't there that's a genuine constraint, but goes on to argue that most claims of hallucination are false).
Thinking further, the one thing I think is missing from your argument is a discussion of involuntary cognitive distortions that come along with even the "milder" mental illnesses like anxiety and depression. It's not just that part of being socially anxious is that you feel affectively worse about being in crowds of people. The disorder also causes you to make systematically bad predictions about how you'll react to social situations, how given activities will feel, etc. You become irrationally pessimistic in your beliefs.
This sort of cognitive distortion is clearly involuntary--we might choose our actions, but we don't choose our thoughts/beliefs. It can be counteracted with higher-order cognitive tools. But this is another respect in which mental illnesses can't be pigeonholed as "just about preferences"
My comments below agree with yours. I think the perception distortion we see in many mental disorders is key to loss of capacity. Interventions solely at the level of behavior are limited in their efficacy. Drugs and psychotherapy are both aimed at changing perception which then makes it much much easier to change behavior.
Isn't the same thing true of other preferences? E.g. people who strongly prefer one political team are often irrationally pessimistic in their beliefs regarding the outcome if the other team wins the election.
I'd like to provide a small possible link between willpower and the preference/constraint distinction.
Preference might tentatively be thought as that which one may change under their own willpower and practice. For example, there are foods I used to dislike, then I learned to like by forcing myself to eat them until I got used to them, and afterwards became so cognizant of, they acquired preferences about which variety is like more than others.
And a constraint may be tentatively thought as that which one cannot change under one's own willpower, though it may be changed, in some cases, via external interventions, such as medications, surgeries etc., if available.
And then the overlap may similarly be linked to cases that require both willpower and external interventions.
This would turn the dichotomy into a false one that's in fact a continuum of that which can and cannot be effected via how much willpower.
Sidenote: the word that's usually translated as "desire" in Buddhist sources, "tanha", is better translated as "craving". It doesn't include small desires, such as Coke over Pepsi, or the desire for Enlightenment, but only those things that cause compulsions, and whose absence would lead the person to experience "dukkha". "Dukkha" in turn is usually translated as "suffering", but that's also inaccurate. It's a broader category that includes from dissatisfactions all the way to literal suffering.
Therefore, the lesson isn't that having desires of any kind lead to suffering, but that being overwhelmed by cravings leads to a continuous state of dissatisfaction that can grow into full suffering, which in turn is a major roadblock for the pursuit of that clarity of mind required for achieving Enlightenment. Hence overcoming cravings, addictions, and the like, is a necessary step on the path.
As someone who likes ecological psychology, I am wary of putting too much emphasis on internal states.
Credences, for example, can only be defined insofar as they are coherent; if someone makes different bets about the same thing in different situations, for example if the bet is presented differently.
But it's not like preferences are a perfectly defined concept either! They too can only be defined insofar as they are coherent.
Why is Scott engaging with this? As far as I can tell, Kaplan has identified a theory that the majority of people believe to be true, determined that Scott is somehow exceptional in being a proponent of this theory, and then repeatedly challenged Scott with inane arguments against the theory without actually considering Scott's defense.
I mean, I can do that too. Scott thinks we shouldn't eat babies, but either babies are more delicious than adults, or adults are more delicious than babies. So Scott must concede that if we shouldn't eat babies, we should eat adults. This isn't something Scott is willing to concede, so baby-eating is morally permissible. I'm going to take Scott's failure to argue about this as evidence that he has no counter-argument, and I'm also going to take any counter-argument Scott might make as evidence that he has no counter-argument.
Babies are delicious, but I couldn't eat a whole one. Not in one go. Maybe I should portion out the baby meat so it can be used to make several meals. What are some of the recommendations I've heard of about that?
"I have been assured by a very knowing American of my acquaintance in London, that a young healthy child well nursed, is, at a year old, a most delicious nourishing and wholesome food, whether stewed, roasted, baked, or boiled; and I make no doubt that it will equally serve in a fricasee, or a ragoust.
... A child will make two dishes at an entertainment for friends, and when the family dines alone, the fore or hind quarter will make a reasonable dish, and seasoned with a little pepper or salt, will be very good boiled on the fourth day, especially in winter."
Scott is the one who started this exchange. It began with a paper that Caplan wrote in 2006 laying out his position. Scott read it and wrote a post about why it was wrong nine years later. It's been back and forth between them since then.
Is he not representing a common view of mental illness from a libertarian view? (I might be really wrong about that -- the politics side of this isn't my area)
Kinda, but this sounds is more the economist's view than a libertarian's, insofar as there is much of a difference, as interpreted by a psychiatrist and translated for a rationalist audience. How much is preserved through this game of telephone is dubious.
Edited to add: Strike that - reverse it! I had only read the beginning of Caplan's article and assumed it was representative of the rest. Now that I've read more, I am 100% team Scott.
Caplan starts out saying that you can model someone with mental illness as a rational actor with unusual preferences - that is correct. Then he moves into endorsing Szasz and going after psychiatry, and IMHO ends up someplace crazy.
You can model someone who is an alcoholic or has ADHD as just having unusual preferences, but when you say they shouldn't receive psychiatric treatment, as Caplan seems to, you're ignoring that a lot of people with alcoholism or ADHD want to change, in the sense that they voluntarily undergo treatment in an effort to change, they sign up for Beeminder, etc.
Now whether psychiatry is effective at introducing change for a particular condition is an empirical question, but Caplan isn't attacking psychiatry by arguing it's ineffective, he's just arguing that it's inappropriate to change people's "preferences," even voluntarily. That seems crazy.
-------- Old Post ----
I went back and skimmed Caplan's original article, and I'm now more sympathetic to Caplan than I was before, unless the exchanges since then have materially changed the discussion.
The question Caplan originally was trying to address was whether classical economic modeling can apply to the decision making progress of a mentally ill person - not whether an illness should be covered by health insurance or whether we should call it an illness in everyday speech, etc.
Under this framework, a person has priorities and resources, and they spend their resources based on their priorities and their beliefs about the world.
Some examples:
1) If my leg is trapped under a rock and I have the choice of sawing my leg off or staying where I am and hoping help finds me, I won't like that choice at all, but I will make that choice based on my preferences and my beliefs. If you know my preferences and beliefs, you can model my behavior.
2) If I mistakenly believe that my family is spying on me in an effort to prevent me from attending business school, you can model my choices based on those preferences, even if the belief itself is based on a mental disorder.
3) If cilantro tastes like soap to me and I don't like the taste of soap, you can model my choices based on that - I'll pay less for a burrito with cilantro is it, but there might be a sale price where I'll eat the cilantro burrito. This model works whether my dislike of cilantro is genetic or based on some other cause.
The ultimate question is whether the economic preference and budgeting models apply to the mentally ill, not something else. On that front, I wonder if Caplan and Alexander have a misunderstanding, not a disagreement.
"The ultimate question is whether the economic preference and budgeting models apply to the mentally ill, not something else"
Thing is though, don't we base (economic) choices on a rational actor? And when your decision-making apparatus/reason is impaired, then the choices downstream of that are not so predictable.
If I think my family are spying on me, what am I likely to do? Predicting my course of action by modelling my choices is going to differ whether "well Deiseach is not crazy so she's not going to burn the house down with them inside it" or not.
As I understand it, a "rational actor" just means you make choices rationally given your preferences. It grants that someone's preferences can be very odd.
IMHO, there's nothing wrong with the model, but some people might not find it very useful.
I've definitely had experiences where I was so angry or so sad that I made choices that I later found irrational from my normal viewpoint. You can model that as I had an unusual preference in that moment that valued telling the other driver what I thought of them more than it valued not getting my car headlights smashed out, or you can model it as me being so angry that I was "irrational." Either model works, so I can't get upset if someone prefers one or the other.
As someone who's been on and off stimulant medications for ADHD at various times I've thought a lot about where boundary between "wanting something but not being able to bring myself to do it" and "not actually wanting something" lies. On stimulant medications I'm much more inclined to classify "things I can't do" into the second bucket rather than the first.
I think this is because on stimulants there's a much shorter path between vaguely thinking about doing a thing, and your brain/body actually doing it. This results in there being a much smaller pool of things that you vaguely want to do but just can't seem to make happen, and therefore you're more likely to classify these things as "not actually wanting them".
Parkinson's is probably the most direct way of exploring this connection. The neurotransmitter that turns thought into movement is literally missing. Does this result in a subjective experience of feeling paralyzed, or of feeling like you "want" to move less?
Caplan is applying the "revealed preferences" principle in economics too broadly. Revealed preferences is a term that means that we should look at the behavior of individuals to determine demand, not what they say in surveys or what is assumed by policy makers. For instance, people may say they don't drink alcohol, but a superficial empirical investigation will reveal that alcohol consumption exceeds that which is predicted by surveys. Drug addicts might not want to continue using drugs, but they do and policy makers should assume that they will continue, even if it means addicts will break laws prohibiting drug use, resulting in a cascade of other negative social outcomes (violence, petty crime, etc.).
Revealed preferences is a useful principle in predicting economic behavior in aggregate. Economists are not really concerned with the internal state of individuals, because it doesn't really predict policy outcomes.
The term is not useful in evaluating the internal state of individuals, or what actually drives them to do what they do. That's the realm of psychology.
Caplan is too smart to interpret the term too literally, in that I doubt he really believes that a schizophrenic is just choosing to hear voices because that is what he prefers. Instead, I think Caplan has another model of behavior in mind that he is not articulating well. Without putting words into his mouth, I would suspect he has not integrated mental illness concepts particularly well into his overall thought process, given his training in economics. Also, embracing nonvoluntary mental illness may undermine his commitment to libertarianism, certain versions of which emphasize the voluntary desired nature of individual behavior.
My thought is yes and no -- that the same mental processes inside two different people can be differently "sticky" and sticky may speak to how much voluntary influence a person can exert over that process. It's common with respect to OCD and ruminative thinking to say we may not control the arising of thoughts but we have some influence over how much we keep directing our attention to particular ones. How sticky some thoughts are though is the result of the usual mix of genes and environment, and to some extent the learned skills of the thinker (and how well they can learn the skills is also a result of a mix of genes and environment).
For someone with OCD, some of their mental processes may indeed be more nonvoluntary than someone else's mental processes around the same content. They may have to exert more conscious effort to not get stuck on the sticky thoughts. Maybe someone with ADHD has the opposite problem with respect to some thought arenas, where their thoughts aren't sticky enough and slide too easily out of awareness.
This connects for me back to that canalization paper Scott shared. It speaks to a quality of fixatedness when it comes to certain thought arenas. People in manic episodes can get very fixated on thoughts, as well as people with paranoid delusions, phobias, etc. People with severe depression can get very fixated on how worthless they think they are. How subject to change all that is has more to do with severity of the disorder, degree of psychological flexibility the person has overall, or the level of stress the person is under than the type of mental process they're in.
From a practical standpoint, isn’t it more objective to cure a physical illness than an internal state?
The policy implications of this discussion sorta rest on the tangible nature of influencing physical versus mental states. I think Caplan’s metapoint is that you can disincentivize or incentivize internal states (preferences) and the preferences which cause illness.
It's legitimate to debate at what point cancer is a 'disease' and also at what point that disease should be treated. Many people die in their old age from some other condition, yet they had a benign tumor somewhere in their body. It really matters whether a cancerous growth is either (a) malignant, or (b) likely to become malignant before you die.
Does this ever matter in the real world? Yes. Back in the day, they used to use PSA alone to determine whether you likely had (a) or (b). Too high a PSA, and they'd resect your prostate. Sure, you'd struggle with incontinence for the rest of your life, but then you could celebrate being a 'cancer survivor'! Except later studies suggested there was an unacceptably high false-positive rate with this test alone, and that hundreds of thousands of men got their prostates out for no good reason.
Yes, cancer is bad and should be treated. But we should also be paying attention whenever we set out on a medical intervention. ESPECIALLY preventative treatments. Those approaches have the highest probability of medical reversal later, when we realize it was all harm for little to no benefit.
I think any conceptual categorizations beyond a purely descriptive one on the basis of atom configurations and observed behaviors places a burden of proof on the part of the part of the categorizer to explain why their classification scheme does useful work. If you steelmanned Bryan's position, what work is the constraints/preference classification doing? It sounds like it's just trying to do political prescriptive work in terms of telling us how to treat the mentally ill. I would like to see an argument based on data from him that treating the mentally ill as if they have a preference rather than a constraint did things like improve recovery rates or speed, or reduce their burden on society, or have some other effect that he might be able to pitch as positive. If the classification doesn't do work then throw it out, since all models are only as good as the work they do.
Th grain of truth in what Caplan is saying is that people with mental illnesses often feel as though the doctors categorizing them as a whatever -- borderline, bulemic, depressive, obsessive, etc -- are sort of denying that they are valid people. Their pain, their memories, their opinions, their sense of humor all sort of do not count, in the minds of the people diagnosing them. And there is a fair amount of truth in that. I used to work on an inpatient unit, and when an upcoming admission was announced, the person was generally identified by their diagnosis: "We're getting another eating disorder later today, she's 19, bulemic, has a very significant abuse history." Doctors hearing about criticisms patients had made of them often rolled their eyes, as though to say, "c'mon, consider the source." Awful, right? And yet, viewed from the outside, people with these disorders do often seem very controlled by their disorders, to the point where you feel as though all the rest of then is hidden and controlled by it.
Are there any therapies that encourage patients to view their mental illnesses as a set of simultaneous constraints and preferences? It's a low bar but I could see that benefitting some people as a way to model their internal states. Would be interesting to try, and surely better than internal family conversations or one of those other weird therapy systems.
There probably are, but I don't know their names. But I think I, and probably a lot of other people who do cognitive-behavioral treatment, think about things in a way that's pretty compatible with that model. For instance it's very helpful to depressed people to recognize that the reason they aren't doing anything is that all their thoughts and mental movies depict every action they might take as pointless, boring and extremely unpleasant to do. And while I don't use the word constraint, I do say that their mental state makes it impossible to believe that taking a walk or whatever might leave them feeling a bit better. So then then they might say something like "even if that's the explanation, there's nothing I can do about it. Everybody keeps telling me to think more positively and I can't." So I ask them whether taking a walk (or whatever) is impossible, or merely looks very unpleasant and pointless, and work on getting them to do it as an experiment. Most people acknowledge that it's not impossible to do. So our model is that your thoughts about action are truly constrained, but your aversion to action is a preference. Anxiety disorder work has an analogous model.
Things like internal family conversations are silly as hell if the therapist is convinced that they are the truth, the whole truth, and the explanation for every problem people have. But for some people an internalized harsh parent really is the problem, and understanding that is very helpful.
This fascinating Invisibilia episode profiles an interesting case where a chronic “augmented pain” syndrome is worsened by paying attention to the pain, and treated by ignoring and therefore desensitization.
Sometimes your preferences change because your circumstances change. It's difficult to say that just because you have a certain preference in the moment that this is an enduring preference, or even to predict what your preference would be when those circumstances change.
Just one more reason that the whole idea of 'preferences' is not practical to drive policy.
I just want to say that I once held my breath until I blacked out. It's a mildly unusual thing that is tangentially related to this article. I was motivated by a competitive spirit and desire to win.
On terminally ill people "holding on" (note 3) this has been studied at a population level.
In countries where Christmas is important, there is an increase in deaths from natural causes in the week immediately following it, with a corresponding dip the preceding week.
Two UCSD sociologists, David Phillips and Elliot King, did a similar study with Passover which is published in the Lancet:
They compared Jews and non-Jews in the same place, which provides a handy control group for factors like the healthcare system not functioning as well the week after a major holiday.
A likely explanation is simply that people "hold on" to reach a major annual family-gathering holiday.
Is there an implicit assumption (or explicit somewhere I have forgotten) that doctors may/should never prescribe treatments, especially pharmaceuticals, for a condition we categorize as a preference rather than a disease? Is this a legal constraint, or medical ethics, or what? Does Caplan take a stance on this?
This leads to the issue of the ethics of self-medication, where presumably Caplan, as a libertarian, sees no objection to any adult treating themselves however they wish, or hiring whoever’s they wish to advise them. I am not sure what Scott's position is. From this perspective, the legal restrictions on prescription are the main sticking point, since whether a condition is a disease or a preference, in a world without legal restrictions on prescription drugs, competent adults in either category could try to improve their lives using whatever treatment they believed would help.
The other issue is under what circumstances bystanders are justified in forcing a treatment upon an unwilling patient (or similarly, to forcefully prevent them from applying a treatment). The current dominant opinion among medical ethicists is that treatment should be voluntary for competent adults. So the interesting question becomes, who counts as a competent adult, and how may we treat infants and invalids? (Of course, existing laws pretty much ignore this analysis.)
It seems to me that the categorization of conditions as diseases or preferences has little relevance for these issues. What is the issue where this controversy does have relevance?
Insurance companies use the language of "medical necessity" to determine whether a medication or procedure will be covered. Doctors' licenses to some degree are at risk if they are seen to be prescribing or procedure-ing outside of medical necessity. Some parts of medicine get to be outside of this, like elective cosmetic surgery or orthodontics.
I don't believe any part of mental health treatment within the medical model gets to operate in this elective surgery way. Unlike a nose job, you don't get to pay a psychiatrist to prescribe ritalin just because you'd like to have it. Though of course some of that does go on anyway.
In the land of talk therapy, you can go talk to a psychotherapist (in the US anyway) and pay out of pocket and do that without a diagnostic paper trail. Some psychedelic-assisted psychotherapy also operates in this less regulated space, though at some risk to the providers' licenses. The current growth of at-home ketamine treatment is also pushing the bounds in a way that can look more like the unregulated wild west.
I meant to say, I like your question and think it's worth asking anytime one's about to get into a big debate: "What is the issue where this controversy does have relevance?"
So someone might not be able to get a prescription, or might not be able to get insurance to pay for it, depending on the result of this debate. I guess if we take those institutional constraints as given, that gives this terminological debate some relevance.
But I would rather criticize the institutional constraints that prevent good outcomes (while treating patients paternalistically).
Though the people who have power in this situation -- healthcare systems, insurance companies, pharmaceutical companies, and the politicians they support -- are unlikely I think to be swayed by libertarian arguments to radically shrink the definition of mental illness and who qualifies for treatment.
Economic precarity also plays a role by feeding parental anxiety which puts pressure on schools and doctors to intervene hard and fast on kids who aren't measuring up. There's so much more measuring that goes on these days than when I was a kid (70s and 80s).
Which is just to say I don't think there's much at stake in this debate and that it's more circling around definitions -- what's a preference and what's a constraint and does holding a gun to someone's head actually reveal anything useful about those terms?
If Caplan came out and said "I think we ought to stop covering any kind of mental health treatment under insurance plans except for people with severe schizophrenia or bipolar disorder, and everyone else ought to be shamed for their poor choices and told to buck the hell up," then Scott might have had something more to work with. As it is, we're parsing definitions and keeping our actual opinions mostly to ourselves.
I'd love to talk more about what we consider good outcomes and what gets in the way of those.
Would Caplan say that? Only if he buys into the presumption that only doctors should prescribe treatments, and then only if they have diagnosed a specific (non-preference) disease (and insurance should only cover disease).
Oh I don't know what Caplan would say, which is part of the frustration with this whole non-debate debate.
I found the piece of Caplan's that Scott was most directly responding to to be somewhat incoherent. Caplan doesn't show much understanding of mental illness, like at any level. He doesn't support with any kind of specifics his general statement that a lot of currently defined mental illnesses are voluntary preferences.
His argument seems a bit all over the map and offering no support other than clever rhetorical moves by quoting Scott out of context. He doesn't explain what supports his choice to use a gun-to-the-head standard to define voluntary preference. There's more missing from his argument than is there I would say. Which leaves him open it seems to me to speculating what he's actually up to, what's motivating whatever this ongoing conversation is. When someone makes an argument but doesn't support it, but then keeps hammering at it across multiple rounds, it suggests there are other motives at play that are being satisfied by keeping it going without taking it any deeper.
I haven’t been following the debate from the beginning. I just today scanned the original journal article by Caplan, which QuintusQuark kindly provided a link to: https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf. I was skimming to try to find out what Caplan thinks is at stake, which I didn’t find. A bad summary might be, he felt he had an opportunity to show how economic analysis applied in an area where no previous economist had thought to apply it. If he thinks there are practical implications of his insight, I missed them.
He probably would while supporting people’s right to pay for their own treatment. Insurance covering treatments sounds like a special privilege for extreme preferences he thinks would discourage people from bucking up. (Look at page 361 here: https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf)
I was just looking at that page, but I’m not sure what you mean. Maybe you mean, if I have insurance against x, I will not pay the full social cost of x? Sort of a moral hazard thing? That's still a long way from “we ought to stop covering any kind of mental health treatment under insurance plans except for people with severe schizophrenia or bipolar disorder, and everyone else ought to be shamed for their poor choices and told to buck the hell up.” What part of his argument implies that no one should sell or buy insurance against bad preferences, or that shaming would be an effective alternative?
From Categories Made for Man: "Hume’s ethics restrict “bad” to an instrumental criticism – you can condemn something as a bad way to achieve a certain goal, but not as morally bad independent of what the goal is. In the same way, borders can be bad at fulfilling your goals in drawing them, but not bad in an absolute sense or factually incorrect. Namibia’s border is bad from the perspective of Germans who want access to the Indian Ocean. But it’s excellent from the perspective of Englishmen who want to watch Germans plummet into the Lower Zambezi and get eaten by hippos."
I suspect Caplan wants to discuss (and has discussed https://betonit.substack.com/p/lgbt-explosion ) empirical questions of " (1) Are increasing numbers of people falling into mental health/sexual/etc. categories through marketing and self-selection, and (2) is this driven by discussing the categories themselves, (3) thus accruing outcomes XYZ (which are somewhat no-benefit with some cost, or sub-optimal)." This is a very common question, something that doesn't necessarily interact with the inherent political nature of categories so much as whether those politics are good or bad in the goal-fulfillment sense.
Bryan continually names Scott, because he likes him! He wants his favorite psychiatrist to say "I estimate that x% of diagnoses are due to self-selection/marketing/discussion of the diagnosis, and we would be better off limiting discussions of XYZ to prevent social/memetic contagion towards sub-optimal outcomes." Scott could pick any number for x, that'd be an entertaining post, but wildly more political (is this good/bad) than Scott normally does. I think Bryan is an excited and friendly-combative debater who refuses to ask that question, in adherence to the inscrutable norms of online-debate etiquette, and doesn't want to draw a crowd that wouldn't otherwise sit for a discussion of category definition. Caplan, and most of the GMU gang, ran into the same problem while discussing feminism a while back - just a bunch of posts discussing "What does feminism mean" while all dancing around having a stance on whether feminism-as-I-define-it-in-this-post (rather than naming policies XYZ) is good or bad.
Well, there is a qualitative difference between a person who feels excruciating pain when they walk, and a person who has had all the nerves in their legs severed. I'm... just not sure if that qualitative difference is particularly useful for reasoning about anything (except maybe figuring out what medical treatment to apply to get that person walking).
I haven't gone to read what Caplan has written in the past (maybe I should), but in my experience the line of reasoning presented as his in this post is essentially never put forward by someone who has a nuanced contribution to make to philosophy, or who wants academics in medicine or elsewhere to have a more expressive taxonomy available for the conditions of homo sapiens. In my experience it is always, and only, *that they're intent on proposing a way that communities of people can reduce their support and respect for some group of people they've identified as deserving cruelty*.
You’re right about Caplan. His 2006 article calls psychiatrists rent-seekers and seems to recommend that mentally ill people shouldn’t get any kind of accommodations because they’ve chosen to be mentally ill.
You could say that the only people who are truly mentally ill are people who are completely not viable. Meaning, whatever mental adjustments they are making will not allow them to survive in the world. that of course, is very cultural and environmental. Generally speaking in western culture today there is a lot of outreach towards people who are challenged in this way, and there is much more possibility for their viability. As a result, more people become viable. One could imagine a very different culture that has much less interest or means in reaching out. You might even care to call that political. A village can tolerate an idiot, but a village cannot tolerate hundreds of them just to make a crude example.
You could say mental illness is doing anything that doesn’t work.
I’m a psychologist and a therapist, and agree with everyone who’s saying there really isn’t a clear distinction between constraint and preference (or you could call it a distinction between can’t and won’t). So here's a list of :
*7 Denizens of the Constraint-Preference Murk Zone*
(1) Anxiety disorders: expectation of terrible suffering if you have a close encounter with the feared thing.
Ex: I had a patient who feared broken glass as much as I fear jumping off a 100 foot cliff. I think if somebody put a gun to my head and told me to jump, I probably would not, because my picture of dying via the fall is more vivid and terrible than my picture of a bullet to the head. (Yet when I finally convinced my patient to handle some broken glass, her worst moments of fear were in the 3 seconds before her hand touched the first piece of glass. Once she started it wasn’t so bad, and by piece 10 she was hardly scared at all.)
(2) Procrastination: A habit of asking oneself before doing something whether one is up to doing it, whether there’s chance the in the future that the task will look more attractive, and whether it’s possible that doing it now might be a mistake. (If most of us asked ourselves these questions before we did something we’d never do anything). I see this habit a lot in people who are Aspergerish, and the habit is often coupled with a felt certainty that it is important to ask these questions, otherwise life will be too unpredictable and get out of control. Not asking these questions feels to them like I feel about jumping off a cliff.
(3) Addiction: People have addiction-related “delusions” about the consequences of giving up the substance you’re addicted to. I had these about nicotine. Lighting up had gotten so deeply associated for me with reading, writing, thinking and talking to people that I more than half believed that I would not be able to do any of these things nearly as well without nicotine. What made it hard to stop smoking wasn’t inability to face cravings, it was dread of becoming unable to do the activities that were most important to me. (As Stephen KIng said after recovery from addiction to alcohol and cocaine, I found to my surprise that my mind worked fine without the drug.)
(4) Depression: It’s not exactly inability to act, it’s inability to have any positive expectations about acting, and of course that produces failure to act. If the mind had suits, the way a deck of cards does, and there was a fearful suit, a sad suit, an excited suit and a hopeful suit, depression would be playing with a deck where the last 2 suits are missing.
(5) Depression: Inability to hear encouragement from others as anything except invalidation and criticism. Consequently, something that’s often pretty available to unhappy people is unhelpful and toxic to the depressed person.
(6) Avoidant and dependent take on life: People with any of the above who are also seriously dependent or avoidant have an extra constraint/preference: They believe that if they overcome the condition that has them stuck they will lose their status as a beloved and sheltered person, and be cast out into the lonesome cold.
It seems to me that it’s too simple to think of there being a spectrum running from constraint to preference, with lots of things in between. You can see in the above examples how people can be constrained by the false expectations ofwhat would happen if they did or didn’t do something — or by the sheer inability to conceive of certain things.
Yes, yes. See my post about braces applied to limbs at childhood. In my experience, depression is a very much internalized narrative about outcome that trumps the immediate circumstances of one’s life. It is an internalized memory that controls the present. And it is not just mental, it is physical. It is the feeling of being defeated.
"Depression: Inability to hear encouragement from others as anything except invalidation and criticism."
Oh, yes. If someone criticises me, I believe it because sure, it must be true, they're seeing the real me, they can see the faults and lacks and general awfulness. If someone praises me, I can't believe it, because they're only saying that to be nice, or they have no idea what I'm really like and if they knew they wouldn't say that nice thing.
So praise can be *worse* than criticism, because criticism is only saying what I already say to myself, I know it, I'm comfortable with it. But praise? What do I say there? "You're an idiot who has no idea, what you just said is stupid and dumb and false"? You can't say that to people who are just trying to be nice to you, so you shut up and smile and have the thorns of praise prick you to the bone.
Oh this is all quite lovely, thank you for writing it!
This has turned out to be true for so many people I've worked with and for me as well: "her worst moments of fear were in the 3 seconds before her hand touched the first piece of glass." It's really hard to learn this, though I think if a person gets a somewhat profound chance to learn it, it maybe transfers pretty well to other situations. It has for me anyway. I think the learning that sticks may have to happen on the really scary stuff though.
Also, I love this that you say about procrastination: "I see this habit a lot in people who are Aspergerish, and the habit is often coupled with a felt certainty that it is important to ask these questions, otherwise life will be too unpredictable and get out of control." That's so well said.
I've been circling around a similar idea having to do with anxiety/OCD/rumination producing a kind of habit of mental "due diligence" which I think is the same thing you're talking about here maybe. I see it in a lot of people who aren't Aspergerish but are more your average, bright neurotic. And I agree that it's an effort to bring into control things that are in fact not in one's control via a kind of relentless mental scanning effort, "what am I missing? what am I missing?" that has a stove knob checking quality about it. It leads to a kind of default place of self-doubt that undermines a person's ability to trust their own experience (which impedes learning). The relentless self-scrutiny has an OCD feel to me and I hadn't thought of that in connection with procrastination which I've mainly thought of as a form of emotional avoidance that's part of performance anxiety. But it seems like the emotional avoidance is being sustained by a lot of active rumination of the kind you describe as opposed to just by binging Netflix (which is what the emotional avoidance frame brings to mind).
Moving an experience from constraint to something a person can exert some influence over (I don't like the preference language at all) is a learning process and no one can predict ahead of time how much progress any one person can make along that road because it depends on so many variables that are not in their control. People can bring willingness to try things, but the outcomes are not in anyone's control, something which seems to elude Caplan's tidy gun-to-the-head framework.
It’s kind of interesting to break the word disease down into its components.
Dis-ease.
A lack of ease.. that is a big umbrella.
Mental illness is a big umbrella.
My problem with this conversation is the generalities implicit in the vocabulary.
Let’s take two things; liver cancer, and diabetes. They are both diseases.
Diabetes has been narrowed down to a very particular chemical deficiency that can be remedied.
Liver cancer on the other hand, as far as I know, is way more wacky than that.
One form of Mental illness is tricky, because as anyone who’s ever dealt with it knows, one’s own participation in its cure (or management? )is essential. See the classic joke about how many psychiatrists it takes to change a lightbulb.
But
The brain is also a physical organ and can malfunction. Godspeed to the people who are trying to crack that nut… I have to believe that some of the pharmaceutical interventions developed over the last 40 years have actual relevance in that world. I am not at all in a position to speak to it. I am more interested in what I consider the other half of mental illness..
What ever happened to the word neurotic? It really seems to of fallen off the bus. I never see it in any of these discussions anymore. Perhaps it is a concept that is extremely out of favor. I happen to think it’s an important distinction.
A disease if you will, of emotional dysregulation? Is this a failure of will or a condition to be treated as a disease, in the classic sense?
What if someone was born and had a brace put on one of their legs, a brace that dictated the growth and shape of their limb over a period of time? What if eventually that brace was removed? Would the affected limb then spontaneously assume it’s rightful shape?
Or is it more likely that a lot of physiotherapy and conscious exercise would be necessary to even begin to restore it to what it should’ve been?
There are lots of people who I believe are mentally ill in the most fundamental sense of that expression; meaning, they have very little volition to bring to the task of understanding who they are. And there are others, who I would refer to as neurotic, that given the opportunity can direct their attention towards their own management or recovery. It’s very difficult to find a border between those two things.
So finally, is this a philosophical discussion or a discussion about social policy?
I would be very interested in some granular information about what it is to pay attention in some technical sense. The ability to direct your self towards paying attention to something. I think it’s a very significant thing, this ability to redirect one’s attention.
Caplan skimmed the DSM and now he thinks he’s a psych expert. I took a look at his 2006 article that started all this. He linked to it this month, so he still endorses it.
His section on ADHD is based on a shallow reading of the DSM diagnostic criteria with almost no context. He uses this to suggest economically modeling ADHD as “a high disutility of work combined with a strong taste for variety.” The first part doesn’t make sense. The DSM-listed symptom regarding avoiding tasks requiring sustained mental effort can apply to work, but it can also inhibit people from returning to books they were enjoying or even watching TV episodes with a lot of hard-to-process stimuli.
He also claims the forgetfulness symptom refers to people with ADHD “conveniently forgetting to do things they would rather avoid.” But ADHD forgetfulness also includes things like, I don’t know, forgetting to eat dinner and drink water because of reading an article by someone too lazy to do thorough research on ADHD. Would this be modeled as a high disutility of eating and drinking?
The DSM criteria also don’t include commonly reported symptoms like rejection sensitivity or hyperfixations. These would indicate that ADHD isn’t “medicalized laziness” but a distinct type of brain with characteristics that tend to come together for a non-obvious reason.
I learned the hard way that preferences *are* constraints.
I was in a community that put very strong social pressure on me to disavow and avoid satisfying many of my preferences. I tried very hard to conform, and to tell myself "they're just preferences, they aren't real tangible things, I can just not listen to them."
This seems to me (with the usual caveat that it's very difficult to have much certainty about the causes of one's own internal states, because n=1 and it's all being seen from the inside) to have caused me to enter a very bad mental state that has lasted 6+ years after I both left that community and recognized the problem and started making serious efforts to improve my mental health.
"they're just preferences, they're not real" transitioned smoothly into "I'm not real, I don't exist" which caused a lot of dissociation and loss of executive function.
I am gradually getting over this, but it's ongoing, and I'm not fully convinced I'll ever feel as real and as sure of myself as I did before.
(ETA: I do recognize that my desire to be accepted by the community was also "just a preference", but I'm not sure why I accepted that one as more real than the ones it was trading off against.)
Need to write a large response here later today, as i have a lot of thoughts on this. Ive also had email conversations with bryan caplan, that i need to look through
Caplan had a point about constraints, but was wrong about constraints and preferences being opposite.
The "if you can do it with a gun to your head, it's not a constraint" example is horrible. Lots of people have something like arthritis where some actions are impossible and some actions are entirely possible but incredibly painful. That situation is a restriction on your life by any sensible colloquial, moral or legal standard, and correspondingly is a disease.
I think the distinction he was looking for was, if we think a restriction is an IMPORTANT restriction, we're likely to think of the cause as a disease rather than not. That's not universal, the restriction will be important in some situations and not in others, depending on our judgement of the person, the judgement of people around them, and the amount society is accessible to people with that limitation.
I haven't read what Caplan wrote, so I'm probably missing the good bits, but I suspected it was a combination of:
1. He says there's not a neat objective criteria for what is a disease and what isn't. (Scott agrees, I agree) And he thought that Scott thought that WAS an objective criteria (he was wrong, that's why he thought Scott changed his mind)
2. He thinks that he has a better definition with constraint vs preference (Scott disagrees repeatedly, I mostly disagree as described at the start of this comment). I suspect that he saw Scott disagreeing with his particular definition as if Scott disagreed with point #1.
3. He thinks that modern categorisation of mental illnesses include too many things that shouldn't be included, because of these disagreements. I think that that HAS BEEN true and is true of SOME things, but that it's much less true now. I suspect if he talks about specifics, he'd talk about lots of things which psychiatrists think are illnesses which he doesn't, and I would probably agree with which are which, but that he thinks that the false-diagnoses are much more prevalent than I do.
On the last point, I have ADHD. I know first, second and third hand that there are people with a real, treatable disease which manifests in a variety of ways which are very real but hard to measure from the outside (e.g. inability to concentrate, inability to start doing important tasks. And yes, most people can do them for a few minutes if they have a gun to their head, but can't KEEP doing them every day without an immediate threat to their existence, and can't be threatened into doing them without long term damage).
I also know that there are people who suffer unable to fit into a regimented school routine, but flourish perfectly well in jobs which suit them (which jobs vary, but programming, EMT, crafter, financial trader, etc depending on their combination of needs).
I also know that there are SOME people who have adhd-like problems in life and seek out or are forced into treatment or into attitudes about it which are counterproductive in helping them. I think that's something to be aware of, especially if people are forced into treatment which is bad for them, but I don't know HOW widespread that is, and that doesn't constitute ADHD being fake for everyone else.
While there are conditions associated with intellectual disability that do predispose people to being, crudely, "happy and kind" this is often just a stereotype of people with intellectual disability that is simply not true. While it may seem harmless because it's a positive stereotype, it does tend to promote a kind of dehumanization where people aren't seen as part of the full range of human experience. Down syndrome is an example of this. In fact, people with Down Syndrome have a higher prevalence of depression.
There was an SSC article several years ago (IIRC re autism?) where Scott made the point that the general public sees the "happy" sufferers walking around on the street, holding down jobs, etc. but does not see the "unhappy" ones strapped to their beds in asylums -- even when the latter are far more numerous.
"The far more numerous" part isn't accurate and I think you are misremembering that specific point. That said, I think the article you are referring to was a good corrective to people who don't see the more damaging side of developmental disorders because it is out of their interactions and thus invisible to them. A *lot* of people over the past decade have gotten some distorted ideas about autism spectrum disorders because they're reading the commentary of self-advocates who are talking about their experiences that aren't fully representative.
To the point you're making, I'd go further. Lots of people you see on the street, holding down jobs, etc. may seem "happy" to you when reality is more complex behind closed doors. People can keep it together in superficial public interactions. This is true of people with developmental disabilities as well. It's just a common stereotype that people that read as "simple" to others are happy-go-lucky, and just like everyone else, sometimes this is true, and sometimes it isn't. It's more likely to be untrue than of your average person in the case of Down Syndrome because the condition is more apt to predispose mood disorder as are the still relatively pervasive noxious social conditions people with intellectual disabilities often face. I think part of why people are attracted to this stereotype is because it feels karmic. People will such disabilities are perceived as having gotten a raw deal, and this is imagined to be "made up" for with a pleasant disposition.
I've attended funerals for people with ID/DD where preachers who know very little about the person being eulogized have drunk deeply from this stereotype and give a talk describing people who were not all that happy or nice this way. It's profoundly awkward to listen to hear someone reduced to a stereotype that isn't even ballpark true for someone who nonetheless had traits worth memorializing.
“While the person may choose whether or not to scratch themselves, they cannot choose whether or not to feel the suffering”
It’s interesting to me that you even think it’s a choice to scratch. As an itchy person I can maybe delay it 15 seconds tops, and the whole time I won’t be thinking of anything but “it itches so bad but I’m not gonna scratch it.” If some other thought grabs my attention, it doesn’t mean it distracts me and the itch goes away on its own, it means I impulsively/unconsciously scratch because I’m no longer focusing on not scratching.
Once when I was 13 I was on a Boy Scout canoeing trip and my legs had gotten literally dozens of bug bites. I’d scratched them to the point that I had polka dots of blood all over my pants, and one of my friends asked me why I didn’t “just stop scratching”. I think I just looked at him confusedly like he had two heads, wondering how he even thought that was humanly possible, and immediately recommenced scratching. But I’ve since learned that apparently I’m the weird one? I don’t know, I can’t imagine how people are apparently able to just ignore it.
>I don’t know, I can’t imagine how people are apparently able to just ignore it.
Typically you don't try to ignore it or push it away, that will only make it worse. Instead you're supposed to calmly experience it with equanimity for as long as it wishes to control your attention. That's not easy to do at all, but that's the theory behind a lot of mindfulness training. You train yourself to allow negative thoughts, stimuli, etc. to have their place in your attention sphere while maintaining a separate fuller awareness. Eventually you get better and better at holding an awareness while allowing the negative thought/stimuli to come into attention and fade away on its own. The better you get the less strongly you'll feel them when they do happen, and typically they happen less frequently because you train your mind that you aren't going to respond so there's no point in bringing them into attention.
If you've never attempted such a practice it probably does seem like an impossibility to feel an itch or some other stimuli and not react to it.
Though there probably are some people that naturally have less sensitivity to various stimuli and are able to just ignore itches, etc. without much effort. Or for whatever reason naturally have a more stoic and/or mindful mental state without ever specifically training that skill.
This is a great example because it also clarifies that there's a ton of variation between individuals for experiences we call by the same name -- itching, worry, regret, sadness, etc.
And one of the things you're pointing to about that variation is that some people experience thoughts (and perhaps sensations and feelings) as a whole lot stickier than other people. Some people are more able to dissociate (for good and bad) from their felt experiences. Some people have trouble hanging onto thoughts while other people have trouble letting them go. This isn't a measure of preferences or character strength -- a whole lot of this is down to genetics, with some nurture/environment thrown in.
If someone wants to change their "preferences" -- or let's say rather features of their default mode -- we simply don't know ahead of time how possible that's going to be for them to do. So the folks who don't have your specific problem or your specific way of experiencing things, it's easy for them to say, "well, obviously you're just not trying hard enough, you must be kinda lazy" or "if you'd just do some more meditation, this wouldn't be a problem for you" or "if you just accepted that you're hanging onto unhealthy preferences, then you wouldn't be this way."
All kinds of change is possible. In my experience, both more and less than we think. But we have no earthly idea for another person (and nor do they for themselves until they try) how much change is possible for them. All of which should invite humility when people speak about other people's experiences, but doesn't seem to always.
Caplan is a provocateur who believes he is always the smartest man in the room. Too often he is out of his lane and does not know what he is talking about.
Cowan seems to have two versions of Mental Illness is Preference. One is that it just unusual or disapproved preference; the other a matter of conflicted preferences.
"Question for Scott: Would your reaction be any different if someone said, “I’ve gotten tired of my self-destructive preferences, please help me”
"People often feel inner conflict and remorse. Consider the Biblical account of Judas’ betrayal of Jesus. A theologically conflicted Jew betrays his rabbi, feels guilty about it, then hangs himself. If this doesn’t show that Judas was sick, why does Scott’s vignette show the alcoholic was sick?"
But the conflicted preferences version doesn't support Szaszian claims anywhere nearly as well. Someone with conflicting preferences could voluntarily seek treatment to resolve the conflict, without any of their preferences being disapproved of.
I'd argue that in basically all cases, it's more illuminating and promoting of empathy if we think of the issue as a constraint or a fact about the relevant person. For example, gay people have the same preference for emotional and physical intimacy as straights, but while straight people can best/only satisfy these preferences through a relationship with someone of the opposite sex, gays can best/only satisfy the exact same preferences through a relationship with someone of the same sex. This avoids ascribing to gays a poor taste in sexual partners and instead facilitates empathy with their situation. More generally, preferences often carry moral or aesthetic overtones, which facts don't have.
I have read both bryan caplan and scott alexander for a long time, and i have high respect for both. But mental health is the part where i think bryan caplan is almost entirely wrong.
I have written a bit to caplan, and have gotten a few responses back and forth. I cant find the original responses, so i have to remember them, with the unreliability that comes with it.
(Personal note: i have adhd, autism, and is gay and have had depression and suicedal thoughts struggles for most of my life. Im a 27 year old swedish guy.)
I think this mental health, “preference vs constraints” debacle can be broken down into 3 parts worth digging into
1: “preference vs constrainsts” is a flawed model of humans
2: bryan caplans ideological beliefs and worries that economic and personal freedom is reduced by common disability and mental health narratives, and my thoughts on that
3: bryan caplans personal psychology and philosophy
I will then go over the thought experiments and illustrate how these points helps explain key disagreements
Lets start with point 1:
1: “preference vs constraints” is a flawed model of humans
I think this model is highly useful as a tool of analysis. But all models are useful because they SIMPLIFY reality, and remove other variables.
While I believe its much more relevant then most people here, i do think that there actually 2 MORE ASPECTS of it that is neglected. Those are…
“Harmony”:
And…
“Outside signals”
Lets go into harmony first.
Preference and constraint is seeing a person as one single entity that is consistent, and dont conflict.
But as split brain patients show, we are actually a ton of smaller modules that work together, and often conflict.
Anyone that has tried to quit smoking or to diet have some 3
Experience with feeling that they have different preferences and constraints at different times.
Most normal people dont have too much disharmony in them. But some people have very big disharmony. An alcoholic can have parts of him that desperately wants alcohol, while most parts dont.
Other people have very much harmony in them: i think that naturally conscientious people are like this: the differents parts of themself dont have much conflict.
We have limited energy and willpower, and if there is too much stuff happening mentally or physically we become overwhelmed. So if you have disharmony in you, then things usually require more willpower and work to make the parts of you work together, which makes you less conscientious.
I personally have low contentiousness, while i believe bryan caplan to be unusually extremely high on conscientiousness and harmony: so he has never to a significant degree experienced that clash of motives, and thus see everyone else as being the same
“Outside signals” is more straightforward; its basically that you always have to interact with other agents, and being able to consistently show what preferences or constraints you have heavily influences the interactions you have. If you have very low outside signals (like having an invisible disability) then your ability to negotiate or get social support is lowered.
One note: outside threats(gun to the head) or desperation can temporarily bring you disharmony’s parts into an alliance, the same way an outside threat makes polarized peoples cooperate. But this requires tons of effort, and as soon as the threat is removed the disharmony is back
Point 2:
2: bryan caplans ideological beliefs and worries that economic and personal freedom is reduced by common disability and mental health narratives, and my thoughts on that
Ive noticed that in climate change discussions, climate change deniers are usually worried that if they say that climate change IS a problem, then their freedooms will be taken away simce its usually seen as collective problem, outside of personal responsibility and agency.
I think many libertarians get stuck into this mentality on the topic of disabilities or mental health, cause mental health is in a way built on the idea that you are not fully rational or consistent. That weakens the individuals strength and rationality, and has to be fair in the past been used to completely imprison and abuse people since mental ilness has been a way to say “their opinions dont matter”
As i earlier said, ive written to bryan caplan a number of times. And one of them i wrote along the lines of “i feel super depressed and i wonder if im a burden since im on welfare, am i that?”
And bryan caplan wrote something along the lines of “im busy and cant respond to much right now, but im sorry your feeling horrible. I suggest reading the book doing great. Good luck”
This seems at first to be a contradiction to his deppression preference article. And i think it is!! But i think that when BC is interacting with one individual that expresses suffering and advice, that it doesnt really challenge the individual liberty intuition he has, it is strengthened as he sees someone taking personal responsibility and putting in effort.
But when it comes to the abstract, his intuition is that adults are adults and should be left alone first and foremost, and only jn extreme circumstances have liberty taken from them.
To be fair, the mental heath and disability rights movements are generally left leaning, pro redistribution, and somewhat anti econ freedom. I think this is a mistaken view myself; im libertarian, socially liberal, and think capitalism and econ freedom is awesome for disabled people. Anyhow, the last point;
3: bryan caplans personal psychology and philosophy
Bryan caplan is a huge fan of michael huemer, who is a philosopher that advocates for conservative epistemics; the idea that before we use logic, we must appraise what appears in front of us. Its skeptical towards principle based or utility based formulas, and thinks we often fool ourselfs. Its heavily leaned towards trusting intuition and appearances first, and only trusting abstract reasoning if it has repeatedly and clearly demontrably outperformed appearances
From what i can tell, bryan caplan also has a very strong intuition about just desserts; people should get what they deserve, and etc. he also is heavily skeptical about abstract moral or philosophy claims made by people like peter singer
He seems to go by “revealed preferences shows what people actually care about” in morality. Which he uses to dismiss the idea that animals morally matter, since we treat them like garbage, and even vegans prioritize small comforts for themself over animals.
He wants very very clear and obvious proof that someone is disabled and constrained; if its vague then his intuition about people faking or should be left alone is much stronger then the intuition to help someone in pain.
To summarize:
1; {preference vs constraints} has two more prts: {harmony and outside signals} (lets shorten this as PCHO)
2: {BC worrys that mental health ideas takes personal agency away from people}
3:{BC goss with intuition first, and distrusts vague abstract and hard to judge things that are hypothetical}
So i just went over all of this. But how does this interact with the thought experiments and debates BC and SA have? Well, lets see:
A: the alcoholic.
SA thinks he is a victim but BC thinks hes blameworthy
On the PCHO spectrum…
BC dont see the alcoholic as being disharmonious, and sees him as consistently and visibly prefering alcohol to the family.
SA sees him as having conflicting parts and because of his therapist background, can see a lot more of the signals the alcoholic gives that he actually hates this situation.
Additionally, BC is going off “is this person worth associating with or a pain in the butt to help?” Intuition, for which the answer is “yeah get away from them”
While SA dont have these as much
B: tweaking the variable
Lets say now that we have a completely normal non alcoholic person. If they decide to start drinking copiously, or deliberately takes a pill that turns them into an alcoholic, then i think both SA and BC would see him as fully responsible and preferring alcoholism
But if he person is at first normal, but then an goverment agency feeds him a drug that turns him into an alcoholic, then i think even BC would agree with SA that this alcoholic is a victim and deserves help or sympathy
In these cases, the first one of personal deliberation has HIGH HARMONY and HIGH OUTSIDE SIGNALS. Its very easy to see that this person wanted this consistently, and its very easy to judge this as an outside viever.
The latter one has an LOW HARMONY and HIGH OUTSIDE SIGNALS: aside from just normal morals about coercion, its easy to see that this person did not at all consistently want this alcoholism, and its obvious viewers.
Ok ive written a lot and is tired now. Hopefully this is interesting and helpful
What most bothers me here is the conflation of different kinds of things under the heading of "disease" or "illness". There are often good reasons to treat these things together, but there are also important reasons to treat them separately.
Prader-Willi syndrome [side note: it is not caused by "damage to a region of chromosome 15". It is caused exclusively by damage to that region in the paternal copy of chromosome 15; the same damage on the copy inherited from the mother gets you the radically different Angelman syndrome], Down syndrome, and cancer are all examples of the body breaking down. There is an intended state of affairs and an intended functionality, and those things have failed to be. But the failure is spontaneous.
The stomach flu this essay mentions in passing is a phenomenon of a radically different kind. There the body is functioning according to plan, but it's been invaded by hostile forces which impair its functionality anyway.
Both types of problem commonly cause the victim to complain that something is wrong. It's also common that they don't. The archetype for problem-without-complaint is of course the brain slug that takes over your mind, causing you to do things like protecting the brain slug and resisting efforts to have it removed and incubating its children.
So we can reject the idea that if someone isn't complaining about a problem they have, then no such problem actually exists.
Aging is similar to the two categories above in that it involves a breakdown in bodily functionality and sufferers complain that things have gone wrong.
But it's unlike them in that it's not a deviation from the plan. Aging is normal, intended functionality. It's just bad functionality that we wish we didn't have. I strongly suspect that this is why people reject the idea of calling aging a disease. I can be certain that that's why *I* reject the idea of calling aging a disease.
This is a teleological argument: it assumes that a state of affairs is intended, and what deviates from the intended plan is an anomaly that we are justified in correcting.
My question would be, who does the intending, and why should we care about their opinion on the matter?
No one does the intending. That doesn't mean the intent isn't there.
This is the old chestnut about the evolution of an eye. Any investigation of the anatomy of an eye will tell you conclusively that it's designed to provide vision. Do we believe that that's because the eye developed over time due to the usefulness of vision, or because someone with a plan ordained that eyed animals should have vision?
People make both choices. There is not a lot to recommend the second one, except for the obvious purposefulness of the eye. The answer is not that in fact eyes do not come with a purpose. It's that purposes can exist in the absence of a planner.
> and why should we care about their opinion on the matter?
Because this is, by definition, the answer to the question of whether something has gone wrong. There are many implications for how likely you are to be able to change the outcomes you see that depend on this.
> This is a teleological argument: it assumes that a state of affairs is intended, and what deviates from the intended plan is an anomaly that we are justified in correcting.
This is just a gross error on your part. My argument states overtly that a state of affairs is intended.
There is no assumption that what deviates from the intended plan is something that we are justified in correcting. My argument is that using the term "disease" to follow this natural phenomenon is a way of carving nature at the joints. (And in fact, that "disease" needs to be further subdivided.) This is a difference that reflects into other aspects of reality in ways that matter. Aging should not be considered a "disease" because it is unlike diseases in almost every way, and the lessons that are useful as to one are generally not useful as to the other.
I think that by using words like intended you are engaging in at least unconscious anthropomorphization, assuming that the process that created us has goals that we should be loyal to over our own goals.
I'm curious, other than this 'intention' argument, in what ways are aging different from other diseases?
Philosophically, you may prefer to call this a functionalist argument (avoiding the word 'intent') or (a-la-Dennett) taking the intentional stance. Either way, it is an indispensable way of understanding the world that is perfectly compatible with evolution.
The problem is that there's a smuggling of what is into what ought to be. Diseases are things that we think ought to be cured. Why should we defer to the function of evolution when deciding whether we ought to seek to cure death?
I think Bryan Caplan may be a philosophical zombie. He appears human and so we assume that he, like us, has interior experiences. But he does not. It's the only way to understand the depth of delusion to which his argument takes him.
also there is absolutely no reason why homosexuality would be a disease and heterosexuality would not. caplan is wrong on so many different levels that it's difficult to even consider how to reply.
Sociologically, I think something is probably a disease when it is an obstacle for satisfying the values of either the person with the condition or of society at large.
For homosexuality, we decided as a society that we should reorganize such that gays have a place within society, and a lot of people's values shifted, and so being gay is no longer considered a disease.
For Down Syndrome, we're still sorting out what we value, and even the people with the condition probably vary from being frustrated by their limitations to being completely content with the way they are.
For let's say a panic disorder basically no one values the condition of randomly suffering and freaking out, and panic attacks get in the way of quite a lot of things that people and society do value, so we consider it a disease non-controversially. If we believed as a society that people in a state of panic were getting closer to God or something like that, we probably would have a different opinion.
ADHD is just going to be an easy target for a long time. In the last hundred years the society has changed in terms of how much attention needs to be paid to activities, particular ones you get paid for. Sometimes the needs are intense, and people who had some difficulty adapting to the slower pace in other eras (even within memory) are now thriving. But this doesn't stretch indefinitely. A fourth-grade teacher from the 1950s might have to adjust her metrics a bit, and everyone might start looking at various professions with different eyes, but the ability to focus on a task and see it through ain't that different. This mild ambiguity in our era allows clever people to create fun anecdotes in their attempt to prove that it is all hogwash. But they are just fools with large vocabularies.
I was a psychiatric social worker in an institutional setting for forty years, and you have heard most of the arguments by the end of year one.
You may have mentioned in your writing, but I will mention now. Szasz never treated a schizophrenic patient and may never have met one. At this point I have to wonder what people are hiding.
This is interesting, but can you please address how this might apply to something like obesity? Is that the result of someone having constant suffering unless they eat (like Prader-Willi) and constant suffering if they try to exercise? Or is it the result of someone clearly and rationally choosing to overeat and not exercise fully aware of the tradeoffs? Is it a little of each? Is it different for different people? Should we just politely assume that it's the Prader-Willi-esque explanation for everybody? Is that even more polite than assigning them agency? Should we just give everyone semaglutide and stop talking about it?
One additional point is pain as a warning for future damage. I once kept working despite increasing pain — until I literally couldn't any more (finally passing the constraint test). I had reached the point where I would drop coffee cups etc because of damage done by continuing. The preceding pain was screaming at me STOP DOING THIS RIGHT NOW, but I kept going, stupidly, in part because I though of the pain as just an experience and not as a warning that I was in the process of destroying myself.
And people do this all the time. Some tear muscles at the gym because they ignore the pain, others burn out, and so on. They CAN keep going so they DO keep going — until they can't. You can even die by ignoring pain. It's clearly not just an experience or a preference.
The opposite is the case, Caplan hardly wants to talk about those at all. He's a libertarian economist much fonder of Szasz' ideas than the typical libertarian economist, and so modeling mental illnesses as preferences follows naturally from him. The reclassification of homosexuality is just a piece of evidence that Szasz/Caplan can cite to claim psychology/psychiatry is political rather than scientific.
But it's obviously some amount of both. I guess it's understandable that they're miffed about their opponents (also understandable) reluctance to admit the political component, but going off the deep end in response doesn't seem like the greatest idea.
If the field was "obviously" scientific then pseudoscientists like Freud & Jung wouldn't have been such big deals :)
I'm very sympathetic to this line of criticism, except coming from economists it's prime pot calling kettle black territory.
You might be surprised by how much better economists are than other social scientists:
https://twitter.com/cremieuxrecueil/status/1650932077267767297
I'm not sure I understand what's going on there. Can you help me figure it out?
No, I think if there's a specific aspect of this that excites him, it's alcoholism, degenerate gambling, drug abuse, etc. Certainly, alcoholism is his go-to referent. Over the past several decades, these pathologies have been re-imagined, so they are now seen as mental diseases rather than personal and moral failings, and Caplan doesn't like it.
That was my sense as well. I'd love to hear from him about that so we don't have to guess about it.
If true, then that's a tale as old as Aesop. "Bah! The grapes were probably sour, I didn't want them anyways!"
He’s been on it for many years, academic articles with this argument going back decades.
No, he was making Szaszian arguments long before TikTok existed. I have no idea where you got your idea.
Right, the first piece Caplan published on this was in 2006 https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf
Mental illness is physical illness is illness. It’s all the same stuff at rock bottom. Maybe Caplan is in some senses just a dualist. Or he believes in a “faculty of will” that just decides things.
And it’s hard to believe he’s ever really known a severely mentally ill person in his life given how he misunderstands it.
Great article thanks for writing.
I think he is a dualist. He believes in free-will, and rejects cryonics because he doesn't think an emulation behaving exactly like him will really contain his consciousness.
I believe Greg Cochran would agree that mental illness is physical illness, and his primary criteria for determining if something qualifies is its effect on Darwinian fitness.
I'm sure Cochran's real criteria are more complex, but as written doesn't that suggest that not-having-ADHD is an illness?
(ADHD people tend to have more children - my theory is that they're too impulsive to use contraceptives)
If it only applies in our unusual environment with contraception, then he might make an exception (his criteria for assuming a pathogen involves both the loss of fitness and having been around a long time). He has mentioned that normal mating preferences we've evolved would result in men finding a robot attractive if it actually resembled Sean Young in Bladerunner, even though that wouldn't be conducive to fitness (he wrote that prior to the release of Bladerunner 2049).
Is there a reason why it's obvious which of ADHD and not-ADHD we should consider the diseased state?
Not-ADHD seems more adaptive for the exact modern first-world environment with schools and office jobs and taxes, etc. But ADHD seems perfectly adaptive to plenty of other actual human environments that still exist, have existed, and will exist again.
I don't see how this is true? Non-ADHD seems to mean you can shift your attention when you want to, vs ADHD which seems to mean your attention shifts without your control. What environments is that more adaptive for?
I think that's not the most productive way of looking at it. Non-ADHD means it takes longer for your attention to shift without your control, but it will shift eventually. Moreover, your attention shifting without your control can be beneficial, because it prevents you getting stuck in a rut, or eaten by a hostile predator while you're gazing at interesting pebbles. I've seen it likened to a safety mechanism for your attention.
So the question is do you want a relatively relaxed safety mechanism, or a relatively trigger-happy mechanism. Compared to the ancestral environment, today we spend more time poring over boring spreadsheets and less time running from predators [citation needed] so ADHD has become less adaptive.
fair, I can see how it might be more like a spectrum!
This might be controversial, but it seems like ADHD is also correlated with some positive traits like creativity and responding well (and quickly) to surprising and unfamiliar situations. These are definitely adaptive traits.
And, given that the negative effects of ADHD are almost invisible in a large number of environments (notable excluding the environment of modern highly-organized first world society), the benefit doesn't need to be very large for ADHD to be more adaptive than not-ADHD in these other contexts.
Since becoming a mom to two kids, I’ve found my mild ADHD feels so much like an adaptation for motherhood I’ve actually put off trying Ritalin so I can leverage it until they start school. As a parent you get interrupted constantly. For my husband who has excellent focus, this is *exhausting*. For me, it kind of slots into the “normal”(typical?) functioning of my brain. There was space in my head for incessant gear switching and noise, and while it is tiring and annoying at times, it doesn’t leave me wiped out the like it does to my husband. I’m convinced that it the tendency to ADHD must have served some purpose, at least when its expression is fairly mild.
That's very interesting to hear! I hadn't considered the other side of the coin (that being interrupted with 'normal' focus levels leads to exhaustion).
This is my experience as well. Whether I have ADD or not, I was diagnosed with it as a kid and put on medication... As a mom-of-four I don't need any medication to get through my day. Starting things? Yes, I have hundreds of things to do each day. Finishing them? Housework is never done. There will always be more dirty clothes or hungry kids or books to read... Many of my duties are simultaneous, like cooking breakfast and feeding the pets and watching the baby. Focusing on getting one task "all done" would mean not doing any of the other tasks, which would be bad!
I think of it as "working in circles": doing a little bit of each task until I arrive back at the beginning.
I suspect this is true for many mental illnesses, and for that matter, some physical illnesses. PTSD is likely more adaptive in an environment where you experience trauma with no hope of escape, such as slavery. Anxiety is likely adaptive if conditions are constantly life or death. Obesity is likely adaptive in societies with frequent famines.
> Is there a reason why it's obvious which of ADHD and not-ADHD we should consider the diseased state?
You wouldn't want to consider either of them a diseased state. That's just natural variation. (Of course, that's exactly what Caplan wants to say about everything.)
The fact that a particular strategy outcompetes every other one in a particular environment does not mean that following other strategies is a symptom of a disease. We maintain reservoirs of genetic variation because (1) innovating new adaptations is extremely difficult, and (2) the environment frequently changes. Thus, we hold on to our old adaptations, in the form of people who struggle to reproduce, against the day when they turn out to be superior to us in the face of some new feature of the world.
As I point out in a root-level comment sidethread, the better way to think about "disease" or "not disease" is whether or not you are functioning as intended.
For another example, would that mean that Alzheimer's is often not an illness, because at the onset of symptoms the person involved has already reproduced as much as they were ever going to?
(Based on my limited understanding of when Alzheimer's usually manifests and human fertility rates as a function of age)
I think one's ability to boost one's fitness declines more gradually than that. Women have menopause, but it's believed the reason that evolved was to push them toward boosting the fitness of their existing offspring rather than trying to have more, and Alzheimer's would reduce their ability to do that.
^ this. Everything bad that happens to a member of such a social species as ours decreases fitness. Fortunately for us we've had enough extra fitness that we've been able to absorb the cost of our best hunter or spearmaker slowly declining from old age, or even our best warrior losing his right arm and requiring subsidy as an honored hero. But cost there always is.
In a sense, it's obvious that our "healthy" state, as judged by ancestral environment standards, is sub-optimal in terms of fitness in the post-industrial environment. The not-having-an-urge to donate to sperm banks is a popular "disability" to mention around these parts.
The ADHD people having more children thing. Is this sex differentiated?
I could see it being forgetfulness/inattentiveness/in-ability to keep to a routine/schedule/habit. If so I would say forgetting to take the pill sometimes would differentiate from males "forgetting to put on/bring a condom" if we are presuming the partner is on average neurotypical and it takes two to tango.
Lack of condom etiquette in my opinion would be much much harder to go by unnoticed in the moment. I would expect in a theory like this for there to be more adhd women than ADHD men having a higher than average number of kids?
Scott has proposed a theory and you are investigating his theory, but is there any evidence that the underlying assertion (ADHD people have allegedly have more children) is actually true? Do you have citation(s) for the assertion?
As an assertion about humanity globally, I can't think that that can possibly be true. (Shrug.)
I have no citation! Honestly I just kinda went with the assertion based pretty much on respect for Scott, and the assumption he had seen something credible and I got caught up on modeling the causes and was too distracted/polite/cowardly to ask for a source.
I would love if someone could link on any source for the base claim!
I mostly just jumped on this as, over the last two weeks, I've been investigating ADHD stuff since it turns out if you draw a box around all my "personality quirks", issues, idosyncracies etc I have been coping with undiagnosed ADHD for 30 years now. Just got out of a psychotherpist who specializes in it and yuuuup.
The last 10 years would have been a looooot different with some Ritalin or what-have-you.
Turns out fundamentally not being able to connect your knowledge and values with your actions is not something one is supposed to be hugely struggling with every moment of every day to this degree????
Hard to speak for Cochran, he's a very clever man. But I wonder if he *would* think not-having-ADHD is a disease in modern conditions. After all, give it a few generations and all-other-things-being-equal beyond the realms of sanity, and the whole population will have ADHD, and the few weirdos who refrain from breeding when they get the chance because of something that happens in their heads will look mysteriously asexual.
Is asexuality a disease?
I'm far from asexual but as far as I know my preferences have resulted in me having no children, and I'm not signed up for cryonics either. I'd be happy to call that a disease in the Darwinian sense. I don't expect there to be many copies of me in the future.
"Is asexuality a disease?"
I'd be willing to say it's a disorder, and I'm speaking as one. Given that 99% of the humans around me would crawl over broken glass to get their ashes hauled, being "nope, not one glimmer of interest in anyone or anything" does look unnatural or at the least 'something went wrong in the wiring'.
I'm very happy the way I am, but from the way the rest of yiz go on about "what makes us human is the ability to love" (where love means romantic/erotic love), then I'm a robot or an alien not a human. Fine by me!
I wouldn't say it's a choice, as such; I didn't choose to be this way. Caplan might say I could always choose to have sex, so it's not whatever. Well yeah, and I could also choose to stick my hand in a blender, but I'm not gonna do that either, Bryan.
"get their ashes hauled"
I'm not familiar with that saying, but in context is that a euphemism for having sex?
I learned it all from seeing this performance on TV back in the 80s 😁
https://www.youtube.com/watch?v=ZLoPKQZRjOQ
"Too impulsive" is one explanation, "staying on top of birth control appointments and timing isn't trivial and requires executive function" is another part of an explanation.
This is a very careless sort of metric, but perhaps there's a difference between a trait which harms Darwinian fitness and the lack of a trait which aids it.
"ADHD people tend to have more children" What?
Could you please provide citation(s) to support this assertion?
"Both men and women with ADHD have
been shown to become parents earlier than adults without ADHD and to
have more children compared to adults
without the disorder (Barkley et al., 2008)"
I don't have access to the initial study, but the study that cites it is here: https://guilfordjournals.com/doi/pdf/10.1521/adhd.2016.24.7.1#:~:text=Both%20men%20and%20women%20with,context%20of%20a%20supportive%20romantic
I’d argue strict physicalism rules out true emulations of our consciousness. Our consciousness is literally in the physical stuff of our brains; the substrate does matter. There’s no uploading. An emulation of a hurricane doesn’t get you wet, as they say. And an inorganic robotic arm that is functionally indistinguishable from a human arm doesn’t make it a replica of your arm.
I dunno, what if consciousness is literally just electrons in a neural net? Then all we'd have to do is connect your brain to the new substrate before killing you.
Why electrons?
They seem generally important to computation, in both digital computers and possibly in brains. It does raise the question of whether non-electrical computers could have consciousness I guess. But it seems like as good a starting place as any to me.
I guess the basis of the intuition besides that is just that electricity is the only other thing besides the whole network which is involved in the whole network.
Even an abacus contains electrons
It's true that electric charges are important in the brain, but they are carried by ions and not by electrons.
Excitation is generally caused by an influx of Sodium (or calcium) ions, inhibition by an influx of Chloride ions, getting back to the resting state from a depolarized state is caused by an outflow of potassium Ion. Release of neurotransmitter after an action potential is triggered by an influx of calcium ions.
What if I start replacing your neurons one at a time with artificial versions? At what point do you stop being you?
There is no essential “me”.
EDIT: What are the artificial neurons made of? If they are made of the exact same stuff as the rest of me then there's no difference. If they're made of inorganic material then that's an (interesting) empirical question we may be able to test someday as to whether your thought experiment is even possible.
No one mentioned an essential you, they mentioned a physical you
Wait, so you're saying that consciousness is tied to the form the matter takes? You're mostly made up of carbon, hydrogen, oxygen, and a few other things. If I made an artificial neuron out of those things that didn't look like a natural one, but performed the exact same function, would you count that as preserving consciousness? What if I took a neuron from another person instead? What if I used a particle collider to synthesize atoms from hydrogen and put together an atomically identical neuron, and replaced yours with those?
"Wait, so you're saying that consciousness is tied to the form the matter takes?"
This is far from rare - it's what John Searle uses to argue against conscious AI, for instance, that matter and not merely abstract function is what's relevant. I don't *think* this is the case, but it isn't silly.
" If I made an artificial neuron out of those things that didn't look like a natural one, but performed the exact same function, "
That's a big if. Compare with proteins: They only work because of the shape which is dependent on the their composition and the environment in which they exist.
Given the number of neurons in a typical human brain and the typical human lifespan, "when I die of old age" is probably the correct answer here.
Though to be fair it's less than two orders of magnitude off if you're replacing one per second.
The worry is that there's one neuron in particular in there that, when you disconnect it, everything goes black for *you*, and after that it's just something else wired up to the meatsuit you were piloting. And, for obvious reasons, "you" are not going to have any way to protest that they've made a mistake, because "you" are lost in an increasingly large sea of thirty-or-so excised neurons per second.
What if you start replacing my neurons one at a time with tiny pieces of plastic? At what point do I die?
There's uploading in the sense that it creates two instances of your consciousness, with all the gnarly moral questions about what, if anything, to do with multiple instances of "yourself", at what point they sufficiently diverge and become unique etc. etc.
Have you ever heard of the Moravec Transfer? An interesting thought experiment that you might enjoy philosophising about!
This kind of "physicalism" is ruled out by physics, since fundamental particles have no persistent identities over time.
You can't just say "the substrate does matter" without giving any reason to believe that. "An emulation of a hurricane doesn’t get you wet" is an analogy, and analogies aren't evidence, just hints. An inorganic robotic arm is not a replica of a human arm, because the definition of "replica" includes being made of the same material.
Making claims about other people's internal mental states is fraught at the best of times. The strength of behaviorism is that it only talks about observable phenomena. The weakness (of the strong version) is that its predictions are shit.
In the language I prefer, behaviors are the observables, but there ARE hidden-variables.
When it’s an illness, yes. The problem is that it’s not always clear when something is an illness. Is it an illness for the second toe to be longer than the first? It doesn’t make someone’s life much worse, so we say no (though it is annoying how many shoes aren’t shaped well for this condition). Similarly for being gay.
We say no mostly because the liberal individualist paradigm has crushed the conservative collectivist one as far as elite tastemaker opinions are concerned. LGBTetc. expression was (more or less) suppressed for thousands of years on conservative grounds. It's not clear whether this was ever optimal even on conservatives' own terms, but the toe analogy would've never been considered relevant until the last 60 years or so.
What work is “elite tastemaker” doing here? Do elites have different views about toe length or am I misunderstanding something?
Hm, probably misunderstanding. Elite tastemakers hold high-status opinions. Previously, "homosexuality is a society-corrupting sin" was an elite opinion, whereas now it's one that only an irrelevant fringe holds. Nobody has ever thought that having an unusually-sized toe has any society-wide implications.
I am pretty sure there is a history of body modifications or corrections that involve changing indivduals feet to match elite tastes.
Yeah, it’s inconvenient to have Greek toe. I do have trouble finding shoes that fit, but at least I have good company. The lady portrayed in the Statue of Liberty is in the same boat.
TIL my toe is weird
Or left-handed (in the West, and fairly recently) https://www.smithsonianmag.com/smart-news/two-thirds-of-the-world-still-hates-lefties-64727388/
Disagree. I read several papers by an obscure mid-twentieth century cardiologist that explain very clearly how what mainstream medicine derisively calls "heart attacks" are really just the body's physical expression of a preference for reduced blood flow. You're welcome to read me sledgehammer every modern example of physical illness into this conceptual framework at great length at my new Substack, SitOnIt.
Right. Just because I can't tell you exactly how many grains of sand constitute "a heap" does not mean that "heap" is meaningless or a false category. And no amount of pointing to edge cases on the definition of "heap" will make the Sleeping Bear Dunes cease to be a major geographic feature.
I don't think Scott is arguing that the distinction between constraints and preferences is meaningless; he gives examples of things that are clearly on each side. It's more the claim that Caplan's attempt to draw a clear boundary doesn't work because there is no clear boundary, so if we're considering edge cases we need to take into account that they are genuine edge cases where the categories don't work well.
To make this a bit more precise, if Caplan is arguing that a condition has some property of a preference (you could stop it if you had a gun to your head) therefore it can't also have properties of constraints (like deserving help to cure it), pointing to the existence of edge cases is a reasonable counter-argument. What matters is not whether the category exists at all, but whether it's fuzzy or mutually exclusive.
Caplan explicitly said that the flu changes both your constraints & preferences.
So to steelman his model, we need to be trying to determine for each individual patient and illness to what extent they have a constraint vs preference and then tailor our treatment (aid/gun to head) based on that? I am wondering how that will do useful work.
If diseases vary in the extent to which they are preferences vs constraints then that DOES indeed affect how we should respond. The late Mark Kleiman in a Bloggingheads episode noted that the "disease model of addiction" is falsified by Mao's success in clamping down on opium addiction, and that this fit with the most successful treatment for opiate-addicted doctors being monitoring with consistent immediate consequences.
That seems interesting! Has Bryan ever dug into or expounded on that data and what studies/interventions/personal techniques it might point toward?
Ah, I see how you could get the impression I was critiquing Scott, but my "Right" was, in fact. a genuine statement of actual agreement with Scott.
My intended point was rather that Capalan's "no preference is a disease" amounts to "no number of grains of sand is a heap".
I'm beginning to see the problem with behaviorism, with which I have limited familiarity. The problem is in a clinical setting it serves the function, or is the thing that currently best serves the function, of preventing neurotypical midwits from making bad and abusive decisions based on bad theory of mind. If you let the average behavioral health professional make any inferences at all into the behavior of an autistic person, or a mentally handicapped person, or whatever, that will lead to dark and dangerous roads pretty consistently. If you want examples of this you can just look at the troubled teen industry, which relies on folk psychology and religious theories of the human mind rather than behaviorism.
Complete digression but I have to say it as it reflects an important takeaway I got from practicing ABA for three years.
The distinction between behavior as disease and behavior as preference that I keep coming back to is whether it preserves optionality or not. This is probably strictly as bad as Caplan's because I have never discussed it with anyone (as a general rule I never have an opportunity to discuss anything with anyone: I am immediately dismissed as aggressive and insane whenever I articulate anything).
So anything that causes death trivially fails to preserve optionality. Anything that cripples a person or reduces their intelligence or health trivially fails to preserve optionality. Getting into a machine that injects an optimized dose of heroin into you continuously while keeping you alive indefinitely trivially fails to preserve optionality, etc. Environmentally I think of this as a problem that can sometimes be solved by reducing the strength of incentives or attractors. Some behavioral incentives are basically singularities to humans in some or even most mental states. I can't think of a good reason any environment should be engineered in a way that creates such singularities.
Please critique, develop, or steal this.
Given how many people have accounts of being abused by ABA practicioners and how openly abusive Ivar Lovaas was, I think you're either wrong about behaviorism preventing abuse, or gesturing at the existence of some dramatically more abusive environment than all the behaviorism-based abuse I've heard of. I wish I could be more confident that you were just wrong.
Abuse is the norm in the setting. It was before ABA and if ABA is abolished it will be afterward as well. But yes, the troubled teen industry is worse than the average ABA setting. Diamond Ranch Academy alone has something like a half dozen easily preventable deaths now.
Wikipedia says that the JRC practices ABA, and they sure also have a death rate: https://en.wikipedia.org/wiki/Judge_Rotenberg_Educational_Center#Deaths
It's possible my experiences were moderated by unique factors and the overall situation is worse. I don't believe that the people who killed these victims would have not killed them if it weren't for ABA, but you're steering me away from the belief that ABA might have any advantages.
There is a certain type of person for whom any theory, organizational schema, or ideology is just a pretext under which to conduct and camouflage abuse. There are a lot of these people in certain professions. I don't think ABA solves this problem. I don't think the average person even notices or believes this problem exists for various reasons. My only observation is that for some, theory of mind doesn't exist, they are only capable of falsely inferring antisocial inner states or motives, and getting them to rotely follow mechanical procedures would be a step up for everyone involved given the limited workforce available (the 90-105 IQ pseudo professional range)
The problem is that “optionality” isn’t that clear either, as all of Scott’s examples show.
Still feels less bad to me and like it can probably be placed on the grounds of physics and just needs to avoid physical nihilism. I also suspect most definitions of optionality converge rather than diverge and it's just a book keeping problem.
Ironically, Caplan often criticizes his fellow economists for being too inclined toward behaviorism and not taking seriously intuition & reports about mental states. I'm more behaviorist than Caplan and think he relies too much on the latter sort of evidence.
I do think Caplan's gun to the head test is somewhat stronger than you give it credit: even with a gun to the head you will have reduced capacity with a physical illness relative to your capacity without that, and we don't need to look at a bright line like traveling a specific distance when we have other observables like how fast & steadily you're moving, how heavily you're breathing etc.
But as Scott points out: even with a gun to the head you will have reduced capacity with many mental illnesses relative to your capacity without that. So this argument is refuted.
(And even with a gun to your head, you will have reduced capacity of you don't exercise regularly, or if you eat a big Christmas dinner. Is not exercising regularly or eating a big dinner diseases?)
The problem is
1) Several mental illnesses pass the "gun to the head" test, as Scott clearly illustrated in the post, both in terms of having cognitive effects that can't be overcome by any incentive and by empirically preferring the gunshot in extreme cases.
2) In the case of ADHD and I suspect other disorders involving executive dysfunction, the whole nature of the disorder is that the sufferer's ability to respond to incentives is malfunctioning. They can respond to ICNU incentives, but not other kinds. Note that this is a statement about their incentive structure, not their preference structure. Often someone with ADHD has the same preferences as anybody else, but can only take action towards those preferences by reframing that action to satisfy ICNU. (Healthy version: an instant-gratification point system for errands can slot them into I. Unhealthy version: neglecting your work until your boss threatens to fire you can slot it into U.) That is, an ADHD person is operating under the constraint that they are essentially incapable of responding to non-ICNU incentives. So gun-to-the-head, plainly an ICNU incentive, doesn't reveal anything about their preferences, it just removes that constraint.
What does ICNU stand for?
Interest, Challenge, Novelty, Urgency. It's a framework for understanding ADHD that comes from ADHD specialist William Dodson. I used it because it's expressed very concisely; I've seen other specialists use different language to convey similar ideas about incentive structures and intention versus action. I'm not sure how widespread it is among non-specialists, actually; I'm not a psychiatrist and I've sometimes been mildly surprised at the way Scott writes about ADHD.
In other words they just have poor discipline. That is not an illness. Just because you can change a behavior with drugs doesn't make it an illness. If I had spring shoes I could dunk a basketball, that doesn't mean I have "no dunk" illness.
Kid who I believe has a mental illness related to attention: Kid where you tell them "just stand here for one minute and watch your sister play soccer" and you turn your head. When you turn around ten seconds later the kid is off running in the street. This type of lack of focus and ability to follow instructions is consistent throughout their life, and severe consequences change nothing.
Kid who I DON'T believe has a mental illness related to attention: Kid where they are perfectly functional in most environments except work, homework or anything involving self discipline, but severe consequences see immediate improvement in discipline.
This is just an undisciplined person who doesn't actually have incentives to change their behavior, and also seems like it is currently a majority of the people "diagnosed" with this "illness". And the accommodations that are made for them just increase the problem, since it is provides incentives to be "ill" and insulates them from consequences.
You're far from alone in that opinion. Thankfully, it doesn't really matter what you think unless you're put in charge of some unfortunate ADHD kid or employee. People with the ability to research, diagnose and/or treat ADHD seem to disagree with you on balance.
>People with the ability to research, diagnose and/or treat ADHD seem to disagree with you on balance.
Well sure if you can make normal human variability into a big business requiring counseling and constant medication why not? Pills pills pills!
I mean everyone's entitled to their opinion and certainly some mental disorders are incorrectly or over diagnosed. Same with people who get orthopedic surgery, etc etc. But I would want someone trained to diagnose ADHD to be drawing the lines and making determinations even if they get it wrong sometimes. Your second example, depending on other factors, is potentially a central example of ADHD and not even an edge case. Functional in some environments but impaired at school or work and non-responsive to medium-level consequences is a lot of ADHD. Distinguishing that from malingering, trauma, depression, or anxiety isn't easy to do just from watching.
>Functional in some environments but impaired at school or work and non-responsive to medium-level consequences is a lot of ADHD.
That isn't an illness it is a personality defect. You want to "treat" it with pills whatever, but don't act like it it is some "illness".
I am not someone with "non-dunking illness". I am 5'10".
What would you think about my kid's friend, who had impulse control problems where he could not stop himself from getting enraged at his classmates, even when threatened with severe consequences? (The consequences eventually encompassed being expelled from a school that he really loved and had good friends at.) He'd try and try to keep hold of his temper, and he'd even succeed for a little while (so there's your improvement -- but only temporarily, like in Scott's post), and then he'd explode in a way that was at least as bad, maybe worse, as if there hadn't been consequences in the first place.
This kid then was diagnosed with "classic" ADHD and put on medication, and now he is perfectly functional and doesn't do that kind of thing anymore.
(Interestingly, he was mostly functional in regard to schoolwork even when he was having his worst anger issues -- he's an extremely smart kid, so even though he also did have the classic lack of focus, he was smart enough that he could get by without focusing too hard. So "perfectly functional in most environments" might fit him, actually.)
> In other words they just have poor discipline. That is not an illness.
Something feels wrong about this type of argument. Not sure I can put my finger on it, but it sounds like:
"I have a flu. It makes me sneeze all the time."
"In other words, you just sneeze all the time. That is not an illness."
Sounds like: if X can be reduced to components, it is not a true X.
The problem is that "poor discipline" is a symptom that can have many possible causes. Tired people probably have poorer discipline on average that well-rested ones, but that shouldn't make us conclude that being tired is not a thing (because it is "just poor discipline").
From certain perspective, the diagnosis of ADHD is like saying: these people have poor discipline *because of* their shape of brain or their internal chemistry. As opposed to people who have poor discipline merely because of their habits or the incentives they have at the moment.
Basically, by saying that ADHD is "just poor discipline", you seem to imply that a relatively simple set of incentives applied consistently for a few weeks could easily make the ADHD go away. Which I believe is wrong.
> In other words they just have poor discipline. That is not an illness.
ADHD is directly related to a physiological issue with dopamine. That's why stimulants calm people with ADHD down, while it amps "normal" people up.
How many people who get pills are actually physically diagnosed with dopamine disorder?
What do you make of adults who want to change their behavior and focus, but find themselves drifting off at work and they're worried that it's going to impact their job performance, so they go to the doctor on their own to ask for help? Are they faking or is it real?
That’s not a mental illness that is a performance enhancing drug. And like I said, from a public policy perspective it is a bad idea since it creates an arms race.
But why should people believe you have a healthy capacity that is greater than your current one just because you say you’re impaired?
Verifying the greater, healthy capacity can be tough, even if one is really present.
People should believe it if the person has demonstrated that larger capacity earlier. And certainly determining someone's maximum capacity is going to be more difficult than checking if they can meet some minimum, but so what?
The Prader–Willi syndrome reminds me of the taxxons from the Animorphs series.
Same, I wonder if it served as an inspiration, or if "ravenously hungry to the point that you will eat until you explode" is a more general trope that people would come to even without knowledge of that or similar illnesses. Another historical example, although I don't think it's known which disorder in particular caused it, is Tarrare: https://en.wikipedia.org/wiki/Tarrare
Some dog breeds sometimes eat themselves to death if given unlimited food. Many pet owners have heard about this, so that could be another source of inspiration.
I have heard this about horses as well.
The phenomenon is called "eating to gorging". It's not so much that the animal will kill itself. It's whether or not it will hurt itself by eating excessive food if that much is available. People who work with animals (e.g. the large majority of all premodern people) need to know which ones will eat to gorging and which won't. It is, as illustrated here, most typically thought of as a trait of the species, though there is abundant evidence that some individual humans will eat to gorging in the presence of sufficient food and others won't.
That story essentially sounds like a real-life case of the first victim in the movie "Seven".
I said something to this effect on Caplan's post itself, but his ideas on mental illness are the perfect example of the incoherence that results from applying market-based reasoning to areas outside its proper domain. Markets and economistic modes of analysis are useful tools, but when you insist on applying them to literally everything you end up with nonsense like insisting mental illness doesn't exist because something something preferences. Remember people: just because you have a model that can explain a lot of things in a really cool way, doesn't mean you have to apply it to literally everything in the universe!
This is precisely what Caplan is responding to in his post on Scott getting married (the mental illness angle is just that Scott had made a similar argument about economics in an earlier post).
Looking at that post, Caplan seems to be saying "look, you admitted to applying economics to an area people don't usually apply it to! That means I'm right and you should apply it to literally everything!" which seems like a gotcha that still isn't strong enough to imply the very extreme claim he wants it to. You can say people should apply economistic reasoning to more areas than they currently do, while still recognizing there are some areas it simply doesn't apply. All that one would have to say in response to this post is that marriage is an example of the former, and mental illness the latter.
That just raises the question of why we should believe it applies in some areas like marriage but not others.
I suppose because upon observation, in some cases it makes predictions that mesh with empirical reality, while in other cases it doesn't.
Now we're getting somewhere (closer to behaviorism!), although in one section above Scott acknowledged that Caplan could make the same predictions as him (and just sound silly).
Did he? It seems to me that Scott's main thrust is that Caplan's model doesn't make good predictions (but can "explain" anything after the fact), because the preference/constraint dichotomy is incoherent.
The goal is to find the "correct" model, one that DOES work well literally everywhere, to which all others are approximations in the suitable limit. Sure, this is difficult to achieve in practice, but abandoning the pursuit in favor in collecting stamps seems like … well, you can call it "science" if you like.
I'm in favor of using the appropriate approximations, but KNOWING the most correct theory.
Is this really necessary though? Why can't we have multiple models, each of which is the most accurate in its own domain? Sure, this might not be as parsimonious or "pretty," and pursuing a "Theory of Everything" has value in the deep insights it might reveal if we really find one, but until then I don't see why we can't be happy with models that respect their own areas of applicability. It reminds me of how in geodesy, you have multiple different geoids, or approximate models of the shape of the Earth, depending on whether you want to model the whole Earth (in which case it will be pretty good but not great everywhere), or a specific area (in which case it will work great there but terribly in other places). You don't need to insist on using literally only one geoid for everything, and in fact would be severely handicapping yourself to do so.
That's a bad example: we know there IS a real shape we're approximating, and choosing an approximation we like. What it reminds ME of is the theory of gravity: you can treat the motion of cannonballs and the motion of the planets as entirely different, but it's still worth knowing they're ultimately the same phenomenon.
But surely there are domains where market dynamics simply don't apply. How would you explain the motions of subatomic particles in economic terms? Surely talk of preferences and markets don't apply there (saying a particle has a "preference" to follow a field is the sort of cute rhetorical anthropomorphism I don't think any physicist would seriously endorse), and yet there are areas this form of analysis clearly does apply, so this seems like proof it is a model that is both predictive and domain-restricted, in which case knowing what those domain restrictions are would be useful. Denying this would seem like the same style of thought (insisting your guiding theory must apply to absolutely everything) which led the Soviet Union to denounce the existence of resonant Benzene-ring bonds, because it somehow contradicted the ideas of Dialectical Materialism, their pet economic theory (I know you could say this is just proof DiaMat is wrong, but the point is more how this style of denying domain restrictions on theories can lead to silly outcomes).
I didn't know that bit about the Soviet Union and resonant benzene bonds. Amusing.
The relationship between human preferences and the fundamental physics of the particle interactions is the other way around: the latter DOES ultimately cause the former, but yes, it's obviously not a useful way to describe the emergent behavior.
Conflict is only between theories on the same … hierarchical level, if you will permit some hand-waving. Only if one doesn't subsume the other entirely do you need to worry about reconciling them, as in the case of economic preferences vs. mental illness (honestly, it's pretty easy to reconcile them here; Scott's theory is simply better).
A better example might be: all chemistry is fundamentally quantum mechanics, but using the latter is prohibitively difficult, so you need the heuristics of chemistry. Or if you prefer, all fluid dynamics is fundamentally just intermolecular forces, so there's no conflict between the Navier–Stokes equations and electromagnetism.
And then why would you insist on using economics to explain psychology, when it's psychology that's more fundamental than economics? That really is like trying to explain particle physics in terms of economics!
> How would you explain the motions of subatomic particles in economic terms?
By saying they move towards their lowest-energy configuration and then stay there. There is a reason the motto of Marginal Revolution appears to be "solve for the equilibrium".
> Denying this would seem like the same style of thought [...] which led the Soviet Union to denounce the existence of resonant Benzene-ring bonds
Man, the number of people who would like to denounce the existence of resonant bonds is not low. Same reason quantum mechanics is so frequently denounced. Same reason hidden variable theories are so popular.
The Soviet Union denounced all kinds of empirical facts. Why pick one of the ones that everyone really does hate? Your example in isolation could easily be explained by the natural desire of all people to denounce resonant chemical structures by whatever means they have available to them.
Is that the goal? I would say the goal should be primarily which model is the most accurate, not taking your fancy hammer and seeing how many objects you can treat as nails. Treating elevating a single model above all others as the ultimate goal seems to lead the sort of bullheaded dickishness that you see on the part of Bryan Caplan and String Theorists.
Even look at Mathematics, where there is a very famous theorem that proves that you cannot derive an answer to every meaningful possible question you have from a set of first principles (among other things). Even though ZFC set theory is generally widely used, there are significant questions that cannot be said to have an answer one way or another.
"It is impossible to tell" is a perfectly reasonable answer. As is "the question is ill-posed."
> The goal is to find the "correct" model, one that DOES work well literally everywhere, to which all others are approximations in the suitable limit. Sure, this is difficult to achieve in practice
No, it's functionally impossible, not merely difficult. The evidence is quite clear that this strategy - seeking the "correct" model of human behavior - is in exactly the wrong direction. The people who are best at predicting the future (Tetlock's Superforecasters) are good at predicting the future because they mix and synthesize and use different models for different situations. The people who have tried to develop big singular models of human models all suck at predicting the future basically exactly as much as one another.
"Markets and economistic modes of analysis are useful tools, but when you insist on applying them to literally everything you end up with nonsense"
Would nontransitive preferences be sufficient to rule out a economic analysis? If someone prefers A to B, B to C, and C to A, then they cannot be correctly modeled as a rational actor with a consistent utility function (at least in the domain of choices that include A, B, and C). Would that be sufficient to say: "Don't hand this case to an economist." ?
Well, I've never gotten much value from Bryan Caplan in any context. YMMV.
I think he can be a source of useful perspectives on some issues, but he's a quintessential hedgehog. He has one intellectual framework he tries to fit everything into, and if something doesn't fit, he hammers on it until it wedges into place and pointedly ignores all the bits which stick out. The framework doesn't work for everything he wants it to, but that doesn't mean that it's not a useful thing for a fox to be mindful of.
Horses have no appetite control and can't vomit. If you put enough horse feed in front of a horse, it will continue to eat past the point where it starts injuring its stomach and will eventually die from it.
Do horses have Prader-Willi syndrome,?
That sounds implausible (not the vomiting part; that's true), and more like a folk-story cautioning against overfeeding horses (for health reasons similar to human overeating) than an actual feature of equine biology.
This is flatly untrue. Horses eat grass. You can put a horse in a paddock with more grass than it can eat, and it will not eat itself to death. Indeed, it's hard to imagine how they could have survived as a species if it were not so.
Equally untrue of hard feed (concentrates, oats, etc.) If you are trying to catch a hard-to-catch horse by enticing it with a bucket of feed, success depends on whether it is feeling hungry or not. The claim they cannot vomit is true but not relevant.
True of many, perhaps most, horses and concentrates (corn, grain, pellets) but the cause is more a mismatch between high starch/carbs and the horse's evolutionary gut, which is designed for constant intake of low quality food. Horse overeating is generally pathological only for concentrates, and is of a different severity and process than human overeating.
I don't believe this, can you cite?
https://horsesidevetguide.com/drv/Observation/23/grain-overload-horse-got-into-feed-room/
They don't specify what the consequence is, but apparently it can make them very sick. It's certainly folk wisdom among horse people that it's possible for a horse to eat itself to death on grains.
I stand corrected! Thank you for enlightening me. It is very strange though. How did horses ever evolve? How could they make it through periods when humans weren't controlling their food supply?
They evolved to eat grass. The problem is when they get into a large supply of grain, which is grown and stored by humans.
I believe the issue is that compound feed swells in the intestines. At a mountain dairy I worked at one summer, just before my arriving a goat had gotten into the compound feed and eaten enough for the swelling feed to rupture its innards.
As a longtime equestrian hobbyist, my understanding is that horses evolved to eat low-calorie density wild steppe grass. Under these circumstances, they had to eat almost continuously (when not sleeping, escaping predators, socializing/mating, etc.) to satisfy their caloric needs. So, their continuous(ish) eating behavior is rational in their natural environment. Prader-Willi syndrome in humans is presumably not.
The problem occurs when they're confonted with high-calorie-density human-supplied foods (oats, corn, even grass that's been selected, watered, fertilized, etc. to be much more rich than wild steppe grass). Saying they have *no* appetite control is probably an overstatement, but they're not evolutionary equipped to regulate their consumption of such unnaturally (for them) rich food. Some will indeed eat until their bowels are obstructed, which can result in death.
The condition is colloquially known as "colic" (I know nothing about human babies, so I'm not sure if it's the same as the "colic" that human babies get.)
It's hard for me to be polite about Caplan's last post on mental illness. I imagine it was even harder for Scott.
Kudos for such a thorough, mostly-not-sarcastic dismantling of such an obviously silly, half-baked position.
The herculean effort of being that thoroughly professional and courteous... it boggles my mind. I wish I was half the man Scott is.
Oh come on, he was not thoroughly courteous. He was reasonably polite, but there are plenty of rude and sarcastic bits here.
I think the word "that" in Philos comment indicates that they did not think that Scott was maximally professional and courteous.
Perhaps, but if being reasonably polite but moderately rude is considered a mind-bogglingly high level of courtesy, then I would hate to encounter discourtesy.
This entire post came off as extremely spicy. I had never heard of Caplan before, and I felt a little bad for him by the end of reading this. Imagine being that obsessed with Scott (he seems like an avid Slate Star/Astral Codex fan), and that roasted by Scott.
Caplan's gun-to-the-head distinction between a "preference" and a "constraint" seems so wildly, obviously silly that it's hard for me to believe he actually means it. Reading his original paper, it seems like he's just trying to plug a hole in libertarian philosophy that can't actually be plugged.
Wow this makes sense. Of course Caplan is especially interested in preferences because he’s an economist. But it’s deeper than that, his entire ideology is that the perfect world is one in which everyone is free to follow their preferences. Mental illness destroys this perfect world so he’s forced to either moderate his ideology or pretend mental illness doesn’t exist.
That' a really interesting point, actually. Since moving to Canada, which is much more individualistic/libertarian than where I come from, I notice people with mental illness are much less well taken care off. You're fine if you get yourself to a therapist or doctor and sit down for treatments, but most people with severe mental illnesses will not be able to do these things, and they are left out to dry because it's 'their preference' not to get treatment. A friend of mine in a severe manic episode was refusing to take her medication, terrorising her roommates, and putting herself in dangerous situations - but the authorities refused to do anything because it was 'her choice' (not to take her meds).
It's guns all the way down!
That's very much how it came across to me, too.
What's a bit ironic, perhaps, is that his kind of behaviourism can be harnessed to abet, if not justify, a very different kind of political philosophy. If we insist that people can function just fine absent a set of objectively-verifiable 'constraints', and internal states are just bourgeois indulgence/slave morality, then we can design quite the clockwork society.
Caplan himself has noted that his belief in “a dualistic philosophy of mind, free will, moral objectivism, and an optimistic view of human potential” is linked to his political philosophy. I responded to that here:
https://entitledtoanopinion.wordpress.com/2007/10/11/hey-just-why-am-i-not-a-hobbesian/
Thanks for that. I don't think Caplan's essay on Hobbes was particularly good, to be honest, which might account for some of your bemusement at the time. It reads like he begins with his presumed ingredients of totalitarianism and attempts to stuff chunks of Hobbes into the framework, rather hastily.
The argument collapses, for me, amid the confusion between natural rights and natural law, the distinction between ius and lex that Hobbes explicitly warns about. Caplan writes: "[Hobbes] qualifies this by pointing out that even without a sovereign, people are still bound by natural law. But if we remember that natural law gives everyone the right to do whatever they have the might to do, we can see that this qualification doesn't make any difference." Natural law in Hobbes is, in a sense, the opposite of the natural rights to do whatever you want. Natural law (once a mechanism of enforcement arises via the sovereign) exists to protect us from each other's brutal exercise of those natural rights.
I note in passing that Caplan makes no mention of Hobbes's list of immutable natural laws (which would scuttle the relativism argument), nor of those natural rights that, according to Hobbes, are inalienable and cannot be transferred by contract - self-defence, freedom from imprisonment, freedom from personal harm, etc. Nor does he mention Hobbes's provisions for rights of the subject vis a vis the sovereign. Nor does he make much emphasis on the principle that natural rights are laid down freely, rather than seized. There is also a bit of goal-post shifting between 'totalitarianism' and 'authoritarianism'.
I think your instincts in your reply post were correct - you can readily be a Hobbesian minarchist. But the more interesting question: in the intervening sixteen years, did you become a fully paid-up Stirnerite?
Checking the old post, I see that my link to Caplan's essay rotted and the correct url is now at https://econfaculty.gmu.edu/bcaplan/hobbes.txt
In the time since I wrote that I took seriously Paul Graham's advice to "keep your identity small" and moved away from an explicit ideology like libertarianism into a more consequentialist decentralist bent. This was written the next year (as you can guess from the url) https://entitledtoanopinion.wordpress.com/2008/02/09/rhymes-with-shmashmortion/
Once you are getting into these either or spaces within nosology you have missed the boat in medicine and psychiatry. The huntingtons hemiballism, the tourettic tic and the OCD compulsion are all on a spectrum, with the inevitability of performing each action modulated by shades of increasing agency, but all vastly to the seemingly inevitable and ego-dystonic side of one's perceived agency. Interestingly neuroimaging and demographic data (OCD/tourettes ~30% comorbid with eachother) point to common potential imbalances between the direct and inhibitory pathway in the basal ganglia.
Thus, it's just shades and penumbras between what psychiatrists and neurologists are doing, psychiatrists are like epi-neurologists. On this note, in thinking of preferences and constraints- late stage addiction involves changes in the dorsal striatum and almost becomes like a movement disorder- and the withdrawal state becomes a physiologic constraint that drives the inevitability of use through negative reinforcement- (not to mention the avoidance of hypertensive states, dysphoria, seizures etc). A preference clearly can become a constraint, like eating turns morbid obesity. These are all intricately related. Anyone who is not just armchair commenting and has experience with dual dx populations though, will know the experience well of hearing tales of trauma and hyperarousal etc, driving the original use- was it a preference? Was it not the cause of an original imbalance that constrained the inevitability of pursuing one action over another? It's diffuse and penumbral and deserves individual assessment and formulation, patient by patient, and that's the beauty of psychiatry.
Agree very much on the spectrum of agency idea.
thanks, I have previously read and enjoyed your writing on psychiatry!
So well said all of this.
To be charitable to the guy, after all the shade I've thrown at him, Caplan may well be proceeding on the lines that if the 'talking cures' worked, then that must mean there is some element of choice: in response to certain incentives, you chose to be hysterical or phobic or in love with your mother, and then after talking it over with a shrink, you then chose to be mentally healthy.
So there wasn't an "illness" there constraining you to be hysterically blind, it was a choice.
Which okay, yeah. But full-on paranoid schizophrenia is a horse of a very different colour, and it's definitely nothing to do with "choosing" to throw off all your clothes and run naked into traffic.
From your references I assume you are a psychiatrist or psychologist, and i doubt caplan would know what most of your language means in a nuanced manner. I see your point and do think there a levels of CLEAR BRAIN DYSFUNCTION in certain psychiatric illness mania, catatonia, schizophrenia, acute withdrawal, delirium, things secondary to medical conditions, etc which you cannot pin down as clearly with factitious disorders, anxiety, depression. But we work within a physiology which is neuroplastic, and under epigenetic modulation- with real time interaction between behavior and genes. And "talking cures" can impinge on this system as much as anything, as much as pharmacology.
So for example you take someone with a sustained childhood trauma hx or prolonged exposure to combat and PTSD, they have epigenetically, with no choice involved, downregulated feedback inhibition of their HPA axis d/t consistency of threat in their environment (fkbp15 hypo-methylated along with other implicated physiological changes). They likely have salivary cortisol measures anywhere from 3-5x the general population. Thus the tendency toward aggression, toward externalizing psychopathology, toward hitting the ground AS SOON AS they hear a car backfire (no sense of choice, reflexive loops between thalamus-amygdala- PAG means almost no frontal lobe evaluation of their response). Say they engage in 'the talking cure' and all these memories are unlocked, new, comforting information added to each, now within the context of a therapeutic frame, and this alone methylates fkbp15 and restores HPA homeostasis.
This doesn't mean that there was more choice in the original system that led to hitting the ground with loud sounds, but that choice and constraint are modulated with opposite turns of a screw in a dynamic system, and experiences- war and therapy, can turn this screw in opposite directions. War turns up constraint and therapy relieves it. Kind of like a rational evidence based treatment for a disease!
"From your references I assume you are a psychiatrist or psychologist"
Dear co-commenter on ACX, that is probably the most flattering thing ever said to me 😀
In reality, I am just an ignorant, over-opinonated autodidact who can't keep her mouth shut!
hey you talk the talk! Well some of the neuroendocrine physiology i outlined may not be familiar, but it's all very well established neurphysiological outcomes to trauma and part f the etiologies of ptsd
This is the kind of picture I would really like to hear Caplan respond to. Even the concept of "dynamic system" seems to be missing from his model.
Behavioral coaching works on many physical illnesses, too.
“Give away your cat, move to an uncarpeted dwelling, and wash your bedding every three to four days in steaming hot water,” is advice that can control some people’s sniffles.
Mental and physical illness are not distinguished by whether mundane (if sometimes costly) individual choices result in better disease control.
I am happy to play along with Caplan by calling physical disease “weird preference”, too. I’ve noticed that Caplan seems reluctant to acknowledge how much physical disease can be modeled as a weird preference which simply must be managed in order to avoid burdening others.
But mental illness being preference while physical is not? No. As uncharitable as it is to call some mental-health crises “preferences”, much physical disease is hardly exempt from being modeled as a “weird preference”, too.
> Caplan may well be proceeding on the lines that if the 'talking cures' worked, then that must mean there is some element of choice: in response to certain incentives
Be wary of conflating habits with choice. Clearly we have have well-ingrained thinking habits, and a big component of therapy is establishing a feedback mechanism to change thinking habits. Ingrained thinking habits are not a form of deliberate choice in the usual sense.
If the difference is supposed to be that a "preference" can be changed with incentives and a "constraint" can't, then I don't think the dichotomy makes sense even in purely economic terms. Any budget "constraint" can be overcome with enough creativity and disregard for the law - it's just that the incentive would have to be much much stronger to match.
Or, in Caplanian terms: I can't afford a house, but if I had a gun to my head, I could probably make it happen.
No, it's not the case that every constraint can be overcome. Some things really are impossible.
Right, but Scott's point about it in the article is that Caplan's framing makes it unfalsifiable. Sure you can _claim_ that every person who you put a gun to their head and told them to travel faster than light didn't do it because it's physically impossible, but I can just counter with they just had a stronger preference for not traveling faster than light.
This was the point Scott was making about putting a gun to a deeply psychotic person's head and asking them to reason normally. Caplan is (apparently according to Scott, I haven't followed this debate very closely, certainly not Caplan's side of it) unwilling to admit that this could possibly be the result of a constraint. It proves itself.
The gun-to-the-head test doesn't actually prove anything about constraints vs. preferences, even though it is true that _some_ things are actually physically impossible and others are not.
Caplan hasn't actually made it unfalsifiable, because he hasn't claimed anyone has a stronger preference than being shot in the head... but this does leave him vulnerable to accepting suicidal behavior as a constraint rather than a preference (I don't know that he's addressed that).
maybe for you but I'm built different :^)
No, but in all seriousness: I think budget constraints work as an economic concept because people are so averse to breaking the law that once you go past your budget there's a steep spike in the incentives required to alter behavior. But it's still not *physically impossible* to go beyond that type of constraint, and pretending it is means you have to redefine "constraint" to a point where it makes no practical sense in either psychology or economics.
Even in a complete state of anarchy, budget constraints still apply. Budget constraints work as an economic concept because of the nature of reality. You can't create a perpetual motion machine, no matter how strongly motivated you are.
My point is that there's a wide gulf between "unreasonable/unacceptable" and "literally impossible", and budget constraints are usually defined by the former limit because (a) it's a way easier line to draw and (b) only psychopaths would dare venture into that gulf anyway. Caplan is bent on using the "impossible" definition in his mental illness argument, even though turning that around would break his model of economics by implying that he "can afford" a giant gold statue of himself but merely "prefers" not to sell everything he has and go on a bank robbing spree to make it happen.
Economists are interested in people's willingness to pay, and that applies even to the completely mindless. Distributed Republic (which unfortunately is no longer online) had a great post on this illustrated with a Magic: The Gathering Card of a mindless entity, citing Gary Becker on the minimal requirements for supply/demand curves to have their normal slopes. You can't pay more than you have, and you can't supply more than you can physically produce.
I don't think that breaks his model. He could afford a smaller-scale gold statue without the bank robbing, but in both cases the cost is above what he is willing to pay for it.
Hmm, I do not disagree with your point that preferences vs constraints is a very limited framework for evaluating whether something is an illness. However, your parsimony argument is on the weak side: the universe is complex and the apparently simplest models are often wrong. Whether to consider something an illness that needs to be cured/accommodated/empathized with depends on a complex interplay of internal and external factors.
For example, homosexuality can be considered an illness to be cured in one specific society if every member's procreation is essential for the survival of the society (and assuming artificial insemination is not an option).
Having lethal cancer at the age of 30 can be considered normal in a society where people over 30 are a burden for whatever reason, and not getting a cancer is a disease. Taking steps to avoid getting cancer by 30 could be considered a mental illness or a crime.
Star Trek: TNG has a whole episode on Geordi not being considered ill.
The subculture of hearing-impaired often looks askance at those who choose to have an implant. Without implants available poor hearing was an illness, with them available it's a choice.
Useful term worth linking here (one of your earlier posts mentions it but this one doesn't): Sickness behavior https://en.wikipedia.org/wiki/Sickness_behavior
Many things that we normally think of as symptoms of physical illness, are not in fact directly caused by the disease, but are an evolved behavioral suite that activates in response to the disease!
Great essay.
I thought the debate about whether depression is a disease as a debate mainly about whether depression is a distinct entity, or whether the real problems depressed people have couldn’t be classified better by other disease pathologies.
Or something else. Maybe some set of people who have depression diagnoses are actually lonely, maybe that is their real problem. ”Lonely” is a real problem but it isn’t a disease. You might say, well, they are depressed too. But is that really the most useful way of thinking about it?
But Caplan seems to be arguing that people can’t actually be lonely without choosing it, without its being their revealed preference?
I think the preferences are more supposed to be over doing something rather than being something (although I suppose the preference for doing something would sensibly be related to the effect of the action).
I think you've won this against Brian Caplan-Szasz. (I have some expertise in criminal insanity law, and I'd occasionally see citations to Szasz well past the point where it was defensible.)
the interesting thing about szasz is that he was a pretty normal psychiatrist who practiced much like the other psychiatrists from his day. I learned from many of the people who learned from him at suny upstate medical center where he was chair, they said he was a great teacher of psychiatry who taught them to think abt what/why but basically practiced normal psychiatry.
I haven't read Szasz since grad school but my memory is thinking that he had a lot of interesting and rich clinical material as well as some refreshing perspectives. It seems like his views years later and out of context are being weaponized for new arguments.
"Instead, I would rather describe things that make an action difficult and unpleasant as in some sense real constraints."
You'd think an economist would understand that friction (in the metaphorical sense) making activities more costly (in the economic sense of all costs) doesn't have to be a hard constraint to alter people's behavior. Illness is, well, taxing. Many illnesses, like taxes, just make stuff more costly without imposing hard boundaries.
I've developed, against my will, a strong preference for avoiding cats and smokers. It's nothing personal, but I like breathing better. Medicine has advanced enough that I could live with a cat or smoker and not die, but it would be an unproductive life, since struggling to breathe turns out to be really distracting.
If my only asthma triggers were cats and smokers, I could lead a normal, un-asthmatic life in a society that shunned smoking and cat-keeping – politics! In this society, though, where cat-keeping is normal and smoking is stigmatized but still common, I'll sometimes drug myself extra to disguise my airway's weird preferences as much as possible while I visit loved ones who live with cats or smokers. I'll usually still have symptoms, but I consider it polite to hide them if I can – though "polite" can be overdone:
I once "polited" my way all the way to the ER as a child. I was wheezing at school but didn't want to make a scene, so I hid it for as long as I could. As my parents pointed out, hiding it that much ended up making even *more* of a scene. The etiquette of minimizing the burden we impose on others due to our weird bodily preferences (like asthma or diabetes) can get rather complicated.
I would also think an economist has gotten away from the sharp distinction between “want” and “need”, and therefore shouldn’t draw a sharp distinction between “preference” and “constraint”!
As my husband (a UofC-trained economist) observed several years ago, when he and I first discussed Caplan’s theory of mental illnesses, budget constraints are usually set by preferences, anyhow.
We can all imagine that there must be some limiting constraints “out there” that can’t be further changed by rejiggering our preferences, but our knowledge of these “absolute constraints” is uncertain, and in practice, budgeting is very much influenced by preference:
How much risk or debt are we comfortable taking on? How far would we go to augment our income in a pinch? (Pawning stuff, sex work, selling organs? — preferences!) Are we comfortable with bankruptcy? With moving (or going homeless — preference!) if we can’t make rent or mortgage?…
People budget in order to reach goals. We all know there are limits to rejiggering our goals to stretch our budget, but few of us really know where those limits are — and that’s mostly a good thing: the kind of hardship that requires radically reconfiguring your goals just to eke through is, well, hard, sapping resources from other productive activities.
> Many illnesses, like taxes, just make stuff more costly without imposing hard boundaries.
Exactly this. Let me try it as a dialog...
A: "So, what is your problem?"
B: "I was cursed by an economist witch. Every time I eat an ice cream, I have to pay $1000 extra tax."
A: "And why exactly is that a problem?"
B: "Because I like ice cream!"
A: "Are you saying that you can't buy ice cream?"
B: "No. I am saying that I can -- I just have to pay $1000 extra for each piece. That's a lot of money for me... but I like ice cream very much... so I typically just buy it once in a few months."
A: "Well, no one can buy unlimited amounts of ice cream anyway, so I don't see what is your problem. We live in a world of limited resources, and we need to choose how to spend our money, according to our preferences."
B: "I don't want unlimited amounts of ice cream. Once a day, even once a week, would be enough. All my friends are having ice cream at least once a week. I can't because of this stupid curse. I wonder if there is a ritual that could remove it... and whether my insurance would pay for such ritual..."
A: "Hm. Let me ask you. This curse... is it a constraint, or a preference?"
B: "I don't understand."
A: "If I put a gun to your head and said 'if you don't buy an ice cream, I will shoot you', would you buy one?"
B: "Of course I would!"
A: "If I told you to buy ten, or one hundred, would you?"
B: "Well, yes... although I would need to increase my mortgage for the one hundred ice creams. That is too much!"
A: "Then I don't see what is your problem. You say you want to buy an ice cream. You can buy an ice cream. You can even buy one hundred ice creams, although you prefer to keep the money. You are free to follow your preferences, within the constraints of your budget, just like everyone else."
B: "But also, quite importantly, within the constrains of the curse. Which I want to get rid of."
A: "I thought we have just agreed that the curse is *not* a constraint."
What's up with this guy Caplan? After reading Scott's essay, above, I thought he was a mentally ill person who felt dehumanized by being diagnosed as a something-or-other -- and that's a common reaction, and one I have no trouble sympathizing with. So, I thought, this mentally ill guy, maybe somebody bipolar and often manic, has fixated on Scott as a representative of the profession who diagnosed him as a something, rather than seeing *him.* He wants Scott to say that diagnoses are bullshit so that he can have a win against the profession. And I wondered why Scott was continuing to debate with somebody like that.
But in fact Caplan's a professor of economics & has written several books that sound entertaining and politically incorrect: The Case against Education, Selfish Reasons to Have More Kids. Glanced over his blog for about 3 mins and read a few random bits that were smart, witty and well-written.
But he has this weird bee in his bonnet about Scott. In his Twitter post asking whether Scott's take on witches stealing penises shows Scott's a closert Szaszian he ends with a survey asking "am I crazy or not?" and most common response was "yes, you are crazy." Then I looked at his post about Scott's announcement of his marriage, and didn't see any spinning of that regarding mental illness -- in fact mental illness wasn't mentioned. So whattup?
It's easy to see that Caplan's wrong. I'm more interested in why he's got this weird take on Scott. Anyone know more? Scott, do you?
This is the relevant section in his reaction to Scott's marriage post:
No snark intended, but Scott’s write-up is a big wedding present from him to me. Why? Well, some years ago, Scott almost entirely denied the broad applicability of basic economics:
[...]
So what? Years ago, Scott told us that the “preference/budget distinction is a bad way of dealing with anything more complicated than which brand of shampoo to buy.” Yet he used this standard economic framework to deal with something vastly more complicated: his quest for a life partner.
The preference/budget distinction pervades Scott’s diction. [...] Instead of telling people that there’s only one possible outcome because everything is “like a constraint,” Scott acknowledges that all of us have a vast array of choices – and must strategize to make the most of them. For shampoo and marriage alike.
As if that weren’t enough, Scott explicitly declares that, “Marriage is a contract, no different in theory than an airline’s contract with an airplane manufacturer.” Once again, he’s embracing the standard economic framework of preferences and constraints for life-defining decisions. Shampoo indeed.
But what does this have to do with whether there is such a thing as mental illness? Do you mean because Scott seems to being buying into the preference/budget distinction? So would the idea be that the preference/budget distinction, as applied to mental illness, forces one to say that there is such thing as no mental illness, only unusual choices? If so, that's far from obvious, and I would have expected Caplan to spell that out, and then crow for a coupla paragraphs about how he's been right all along about Scott actually being a Szazian. But he doesn't even spell it out, much less crow. Seems to me kind of a big stretch for you to equate Scott's apparent acceptance of an economic framework with Scott's losing the argument with Caplan (at least in Caplan's opinion) about whether there's such a thing as mental illness. And the fact that Caplan never mentions mental illness here argues against this stretch being valid.
Scott argued against Caplan's logic by saying that the preference/budget distinction of economic logic only applies to a narrow domain. If Scott no longer believes that, then he can no longer rely on that argument when it comes to mental illness and must make a different argument. His argument above is more sophisticated than "just choices like shampoo".
A big stretch for me? There's hardly any of my own writing in my comment above, just quotes from Caplan. Argue against him if you want to. I'll add another quote which linked to that post: "I say his marital search algorithm shows I was right all along."
When I said it was a big stretch for you, I meant a big conceptual stretch to point to the paragraph about preference/budget as an instance of Caplan's spinning Scott's marriage announcement as an acknowledgement of defeat in the mental illness debate. As I said, the little I've seen of Caplan's style in posts about this issue suggest that if he spies something he takes as evidence that Scott is losing the mental illness debate he says very clearly and with great excitement, but here he doesn't mention mental illness, Szasz or any other key words. Maybe Caplan believes that he and Scott are also having a mostly separate many years long debate about economics?
No, Caplan really is referring to the mental illness debate.
I think that probably this Caplan post is the context you are missing: https://www.econlib.org/scott-alexander-on-mental-illness-a-belated-reply/.
Scott and Caplan are on very friendly terms (mostly) - in no small part because they DID also have "a mostly separate many years long debate about economics". Scott was the only one to blurb Caplan's recent book ("feminism") and had Bryan on his SSC blogroll.. Caplan calls him "Alexander the Great". - The economic debate began (I think) with a rather naive post of Scott he reposted 2017 on SSC https://slatestarcodex.com/2017/02/22/repost-the-non-libertarian-faq/
Caplan reacted on econlib, his post-site then.
You're reading an awful lot into a hyperbolic throwaway line about shampoo.
So he wants to extend this simple economic model to explain most of human psychology and behavior?
Why is this even about mental illness? Doesn’t it apply to a wider variety of issues? Does Caplan think that if I do a math problem and am confused and get the answer wrong, that reveals my revealed preference?
Because in my opinion the broader point is that behavior is cognitively complex, and a variety of mental states make them harder. They can be illness but it could be anything else, like just ordinary tiredness or low mood.
Throwing a party is a complex behavior. It may not even occur to someone to throw a party. Holding a gun to someone’s head and saying, throw a party, is already applying a type of therapy.
Even with a gun to the head, people who are bad at math don't become good.
Even with a gun to the head, people with personality disorders don’t become more functional in relationships.
That might depend on the relational competence of the person ordering them around at gunpoint :)
I think the main argument under discussion is that some of them do. Scott mentions Caplan defending "all" which seems like an obviously false position. But surely some of the people who have ADHD or depression or Narcissism or whatever would really change with just a small changes in the incentives around them, in a way others with more deep seated/biological issues would not.
Maybe they would change with a small change in the incentives around them, but they don’t control the incentives around them, so their condition still exists in the circumstances where they are. It isn’t their ”preference”.
If I am struggling for years to solve a puzzle, maybe the slightest change in circumstances could help me solve it all at once, or a change in incentives, but that does not mean my inability to solve it initially was my ”preference”.
If I understand it correctly, Caplan’s argument is almost tautological: it just requires defining ”preference” to mean something completely absurd.
Hmmm. Perhaps the current state of marriage can indeed be put down to people deciding to choose and change partners the same way they'd choose a new brand of shampoo?
I'm contemplating someone having irreconcilable differences with their shampoo. :-)
What's up with Caplan is he is a very smart guy who is emotionally drawn to adopting extreme contrarian positions. I'm not psychoanalyzing him here, he's said it about himself: "Confession: I have been enamored of extreme policies for as long as I can remember. When I was around ten years old, for example, I decided that all smokers should be summarily executed." https://www.econlib.org/archives/2014/11/on_the_complexi.html
Now he's a libertarian of course, but while "execute smokers" and "legalise heroin" are policy goals that would seem to be utterly diametrically opposed to each other, they each at various points won Caplan as an adherent due to their extremity and distance from mainstream discourse.
"Mental illnesses are just socially unacceptable preferences" is just another weird, extreme opinion he holds in large part because its weirdness and extremeness makes it appealing for him.
Thank you. That makes sense. Do you know whether he has long-running weird one-sided debates going on with lots of people, similar to the one he has with Scott? Seems like there's something up with him and Scott, because someone like Caplan, very smart, energetically putting lots of his extreme ideas out there in books and blogs and Tweets, is certainly not going to have any trouble finding people to argue with. So why the fixation on Scott? And it's pretty wacko for him not to get that Scott doesn't even see what's going on as a long-running debate, because from his point of view Caplan's wrong about various points having to do with mental illness, Scott has explained very clearly why he thinks Caplan's wrong, Scott has not heard anything back from Caplan that constitutes a worthwhile rebuttal, so Scott's done with the matter. Why does Caplan think Scott's sneakily signalling that he's really a Szaszian? Either Caplan's kind of crazy, at least as regards Scott, or else he really knows Scott's not signalling anything at all, bit is claiming that Scott is as a way of teasing him. What do you think about all that?
I mean, sort of? I think the Alexander/Caplan mental health debate is the longest running one in terms of numbers of blog posts, but he definitely gets into a lot of arguments with a lot of people and quite often he keeps these arguments going for a long time in one form or another. He's particularly keen on making bets (and has a very good track record).
It's extremely clear that Caplan likes and respects Scott a lot. I think he engages with him a lot (not just on this issue!) in large part because he very much enjoys reading Scott's writing. And even though Scott clearly wrote this particular post in a bit of a frustrated mood, I think the feeling is mutual and Scott likes and respects Caplan as well.
I also think it's clear that Caplan and Alexander do both see this as a long-running exchange, and that this is not the one-sided relationship you're seeing it as. This is probably the 4th post Scott has written about why Caplan is wrong about mental illness.
I agree with Ash. Caplan loves to engage in debate and is one of the biggest and earliest proponents of making public bets. I also think he is genuinely a good natured person but unafraid of engaging in debates that many people would fine uncomfortable or enraging. He is a quirky guy!
I think the one-sided-ness of this mental illness debate is the other way around. If anything, Caplan is UN-fixated.
Scott has now written three lengthy essays on why Caplan is wrong about mental illness, and Caplan has given one brief reply, and then will occasionally casually refer offhand to Scott regarding mental illness as if there is some ongoing "debate." The reason there is no debate is that Caplan doesn't respond properly!
Thanks to all who answered my questions. I am now much better oriented, and can read Scott's piece with the right mental framework about where this debate came from.
People often have a problem with the parts of reality that go against their beliefs. For example, it is difficult for many religious people to accept evolution, because it violates a few assumptions they have (e.g. that all animal species were created at some moment of history, then Adam named them).
As far as I know, Caplan is a libertarian. He makes good points where libertarianism makes good points, and he has a problem accepting the parts of reality that go against the libertarian assumptions.
From libertarian perspective (I am simplifying here, but I believe I am pointing in approximately correct direction), individuals are the most perfect things that exist, and everyone following their own reason and their own will is the best way the society could be organized. The idea of mentally illness just goes against this all. -- It means that for some people, therapy may be better than leaving them alone, even if they disagree. It also means that if you leave people alone, many of them predictably will *not* do the things that they want most, and some of them would even welcome some kind of external influence.
How convenient it would be, for libertarians, if it turned out that the very idea of mental illness is actually just a big hoax promoted by the evil governments! Yeah, it doesn't seem to match some things we can observe, but maybe we can find a clever argument to make those inconvenient facts go away...
Somebody needs to tell Caplan about Procrustes. (Also Narcissus.)
This was a cathartic read as someone whose opinion of Caplan has been going downhill for a few years now.
Care to explain why your opinion of him has been going downhill?
One is that he's basically stopped making interesting bets (https://docs.google.com/document/d/1qShKedFJptpxfTHl9MBtHARAiurX-WK6ChrMgQRQz-0/edit). In particular, Caplan made a bunch of extremely strong claims about the signalling theory of education, and then the pandemic came along with a bunch of novel disruptions to schools: tons of natural experiments and data to analyze. And as far as I can tell Caplan didn't even try to bet on any of it, nor has he updated at all on the data that's come out since. Admittedly his argument rests on the idea that in the long-term adults retain very little book-knowledge from their school days, but even with that consideration I'd expect some update based on new empirical findings or specific falsifiable predictions, and they're just not there.
As with his "debate" with Scott, it's like the guy has just abandoned incorporating real-world evidence and now just thinks about how something could conceivably work without checking in with reality to make sure that his thought experiments actually check out.
Thank you for your response.
FWIW, I always thought Arnold Kling was the most interesting of the two.
He did have this:
https://betonit.substack.com/p/school-choice-sorry-i-underrated
Far from an expert, so feel free to correct errors.
I heard recently that for many mental illnesses, they are more correctly thought of "syndromes" or collections of symptoms. Having a known, identifiable, physical cause means that it _not_ that mental illness, even if all the actual symptoms are identical. If they find a brain tumor, that means you are not depressed, you just have a tumor that is causing exactly all of the standard depression symptoms. You only have depression if a specific physical cause can _not_ be identified.
As far as this is correct, it may be causing Caplan's issue (and to be clear, I think I disagree with him on this topic). We have explicitly defined many (if not all) mental illnesses as "thing that we do not fully understand". As soon as we fully understand it, it is not longer a mental illness and moves into the realm of standard physical illnesses.
The obvious answer to me is to just admit that we don't fully understand them, not to try and wave them away as nothing but preferences. If you moved to various points in time, nearly every single currently understood physical disease would fit into a similar space.
Scott has written about that:
https://slatestarcodex.com/2019/12/04/symptom-condition-cause/
There was another post where I recalled him being relatively sympathetic to that idea for mental illnesses, but he backed away from it later and I remember asking him why but can't remember the response.
In the practical sense of how medical specialties divvy up the illnesses they treat, there's truth in this. When general paresis of the insane was finally known to be caused by syphilis and a bit later we developed antibiotics, it stopped being a "psychiatric" issue in practice because an injection of penicillin does the trick. If we discovered tomorrow that schizophrenia was caused by some weird infection (it isn't but just for argument's sake), then it would suddenly become the domain of infectious disease doctors or something like that.
Kind of but the reasons physical illnesses cause mental symptoms is not usually clear either, and it would be odd to say that the hallucinations caused by tumors were more real than ones caused by a mental illness.
Also mental illnesses are not ”defined” as what we don’t understand.
Even before viruses were discovered, it could be presumed that most viral diseases were physical illnesses for those people who had a mechanistic model of the universe (ie who didn’t believe in demons and the like).
Some types of OCD can be treated by burning a small brain area neurosurgically. Does that make it a physical illness? Or do we need to know what is going on in that area before we could say that?
We mostly try and group diseases by (medical) cause, not by symptom. This is ultimately because the treatment (if any) depends on the cause. The symptom is a headache, but the disease might be a cracked skull, or bacterial meningitis, or viral meningitis, and the treatment will vary accordingly.
However, before a disease is properly understood, researchers typically try and cluster symptoms together, to get a grip with its broad outline. For example, doctors had a broad notion of what was then called consumption, long before anyone identified Koch's bacillus. This is a necessary step - how do you identify what TB sufferers have in common without a rudimentary notion of TB? - but always tentative until you have the cause. Your proposed categories might not line up with reality.
Depression is a classic mental illness that is not yet well understood. We don't know what causes it, or even if it's really one thing, but we know it isn't caused by tumours. Someone who has a lot of the same symptoms as sufferers of the disease, but whose symptoms are caused by a brain tumour, clearly has a different disease, in the same way that a cracked skull and bacterial meningitis are two different things, even though they may both show up as a severe headache.
If we found a specific physical cause for depression (e.g. something in your brain chemistry) that would not stop it from being perceived as a mental illness.
>If they find a brain tumor, that means you are not depressed, you just have a tumor that is causing exactly all of the standard depression symptoms.
Depression is a bad example here because "depressed" is both a description of a symptom-cluster and a shorthand for a psychiatric diagnosis. If we find a a brain tumor that's causing your depression symptoms, you're still depressed, but you don't have Major Depressive Disorder.
Schizophrenia is a better example because "schizophrenic" doesn't have that dual meaning, so if we find, say, an autoimmune disorder that's been causing your symptoms (https://www.washingtonpost.com/wellness/2023/06/01/schizophrenia-autoimmune-lupus-psychiatry/) it's accurate to say that you were never really schizophrenic.
Your central point is mostly correct: psychiatric disorders are defined as not having identifiable organic causes. But this actually breaks down to a certain extent with depression and anxiety: someone who has depressive symptoms due to neurodegenerative disease, for example, like an autoimmune disorder or dementia, often gets sent back to psychiatry for ongoing treatment under a mental health code even after their organic disease has been diagnosed and is being treated.
So psychiatric diagnoses are really more about whether the "mental health" treatment paradigm is helpful in a particular case than any sort of claim about the underlying reality.
>who thought were were
Should be "who thought *we* were.
I posted this comment elsewhere but seems relevant (sorry if Scott has already used this example in one of his other rebuttals)—
Bryan needs to explain why someone with schizophrenia who is in the midst of a psychotic episode is fundamentally different than someone in the midst of an epileptic seizure shaking on the floor, in the sense than in both cases the person's brain is in a highly abnormal state which causes various behaviors we wouldn't normally expect. Does the epileptic have a preference for flailing around on the ground? Obviously not.
One central problem (as I see it) is Bryan's belief in free will, which totally clouds his entire analysis. Many of the conditions we are talking about systematically warp and distort peoples' decision-making and judgment. Viewing the mentally ill as just making different choices based on different preferences is the kind of things I would expect to hear from someone who has never actually met a severely mentally ill person in their lives, which I suspect may actually be the case for Bryan (but that's just my suspicion from the way he talks about this stuff).
I think schizophrenics are more responsive to a gun-to-the-head (although not completely). I sometimes watch the "Living Well With Schizophrenia" youtube channel and the creator of that has discussed things like pretending to be healthy to avoid being hospitalized while experiencing psychosis, which is beyond the ability of someone in a seizure.
Of course, if you're having a generalized seizure you aren't even conscious. It's the "abnormal brain state" point I was stressing. Epilepsy is an intentionally extreme comparison. I could use tardive dyskinesia as another example (involuntary movements which are suppressible). Or various kinds of intoxication. Or hell, just plain old anxiety. The examples are endless because agency exists on a spectrum, as another commenter already pointed out.
This seems to just reinforce Scott's point that it's most useful to think of as a sliding scale with the cutoff being decided pragmatically on what actually produces the best outcome, rather than there being a clear separation between preference and constraint.
Probably. Although I think there is sometimes a boundary, if I understand the preference-constraint thing correctly. For example, someone who is psychotic and has a paranoid delusion may be unshakable in that particular set of beliefs. That is effectively a constraint on their thinking. In an overly narrow sense you could say they "prefer" acting in accordance with the delusion but I think that obviously does violence to what everyone understands preferences to actually mean.
A psychotic person can alter their behavior in response to incentives, but they can't cease to experience their delusions or hallucinations in the same way. I've experienced this. No gun to my head would have altered my perceptions (and delusions are quasi-perceptual phenomena imo) any more than a gun to a sane persons's head could get them to see a green apple where there was a red one. But Caplan seems at best unconcerned with such subjective phenomena as the color of an apple.
"any more than a gun to a sane persons's head could get them to see a green apple where there was a red one"
Peer pressure can sometimes do this... https://en.wikipedia.org/wiki/Asch_conformity_experiments#Attitudes_of_responders_conforming_on_one_or_more_trials
I'm familiar with these results. I think some far simpler explanations for that small minority of subjects who seemed to be unaware that their answers were incorrect can be offered. They may have misunderstood the task, they may have been too embarrassed to admit to a copying strategy after the fact, or they may have been supplying the answer they thought the experimenter wanted to hear. These are behaviors you might expect to emerge from the group that displayed the highest levels of conformity in the actual trials. I find these possiibilities more likely because of, well, everything else we know about the non-defeasibility of perception. If it were indeed the case that peer pressure could shift the perception of even a minority of people, I think we'd have some every day evidence of that phenomena. (You might have heard a similar argument about social priming results: if social priming were as powerful as some researchers would like to claim, we'd have noticed it!) If there's been any work since supporting Asch's "distortion of perception" interpretation I'd be very curious to read it though.
Could be. The copying strategy case is pretty close to what some of the other subjects admitted. I wouldn't be surprised if some of Asch's subjects truly did have their perceptions changed, since there are other situations where high level information can modify what nominally look like purely bottom-up sensory processing (Necker cube, duck/rabbit ambiguous figure, figure-ground ambiguous interpretations, the dalmatian dog figure etc.).
I don't think that instances of bi-stable or ambiguous perception are comparable to perceiving a shorter line as longer. You'd have me properly intrigued if eg people could be convinced that the Necker Cube was a really a Necker Sphere, that the Duck-Rabbit was just as good a Rat-Mongoose, or that a cow popped out of those blobs as readily as a Dalmatian. But they can't and they aren't and they don't. The fact that it takes a particular set of stimuli to induce these kinds of perceptual ambiguities suggest to me that while top-down information isn't wholly irrelevant, it's heavily constrained by the bottom-up.
Let me wrap this back to hallucinations. Most of the auditory hallucinations I've experienced have been kind of inherently ambiguous and vague (lucky in this regard), and it has very much felt like the verbal interpretations I supply are somewhat superimposed over them. But still, if I 'hear' something like swwshhiwooohppeerf and I grind that out as "Hi what's your preference?" (rarely do I have this robust a hallucination) while I might have alternately heard "Shh, I'm woofing, derp", I would not have heard "Ack! The war machine is coming!" because the internal stimulus didn't really have the capacity to be moulded into that shape.
There are people with untreated psychosis who just get lucky and can function okay even when actively psychotic. He may have talked to some of them, and never to the ones who end up in hospitals.
Caplan seems to suggest we should doubt psychotic people's accounts of their hallucinations and delusions:
"I don’t doubt that people sometimes have genuine hallucinations. Yet as I explain in the original piece, it is reasonable to doubt people when they have shown themselves to be broadly unreliable. Thus, I assume that Scott disbelieves self-reports of alien abduction. Why? Because the kind of people who report alien abduction seem unreliable."
I agree that it seems unlikely Caplan has much experience with any severely mentally ill people. Nor does he seem curious enough to investigate this issue by asking psychiatrists how the psychotic syndrome was conceptualized and how it has been validated. Or how they differentiate psychotic patients from malingerers (something that's usually not too difficult because most malingerers have very aberrant ideas about what constitutes psychosis and put on a poor performance). It's easy (almost definitionally) to account for the behavior of a malingerer (a preference not to be jailed, say), but I don't know how Caplan would describe the preferences and constraints of the typical psychotic patient who suffers through profligate negative outcomes thanks to their symptoms.
I think the fact that Caplan seems eager to deny the reality of interior phenomena like the positive symptoms of psychosis in some cases suggests he does find it a little problematic for his model.
Also: the typical reason we disbelieve alien abduction stories is the content of the claim, not the unreliability of the claimant. We might infer their unreliability after they make such a claim, but this is because we already place low credence on alien abductions, and usually not because of other cues in their behavior that render them untrustworthy. It's an odd example.
Moreover, people who experience psychosis ofter get better, become quite reliable, and report on the experiences they had at the time.
>>who thought were were unfairly stigmatizing <<
--> we were
I think maybe an even better analogy than physical illness is injury.
There is an obvious and blatant malfunction when you break your ankle, or lose a finger, or get shot.
And yet, despite the pain, in extreme circumstances we have many examples of people who walk on a broken limb or continue to function for many minutes despite being shot multiple times.
The ability to damage yourself or accept pain to further an important goal in no way contradicts the existence of the injury. The change in behavior while injured is not the same thing as just having a preference for not walking. The treatment for injury is generally obvious, simple, and results in behavioral changes, like walking again, but the treatment is not about changing preferences, it's about changing the actual physical state of the body.
Maybe I don't understand your point, but a key difference (I think?) is that with mental illness people routinely do things they really really wouldn't want to do when in a state of sound mind. So it isn't just about compensating for an injury, but an injury that can lead you to make very bad decisions.
People with an injury often do things they really really wouldn’t want to do when in sound physical health, like remaining flat on the ground while their attacker reloads the gun.
You still have reduced capacity with the injury, even with a gun to the head.
So do e.g. depressed people, which is Scott's point. Unless you claim that they simply have a lesser dis-preference against getting killed, but then the symmetry with physical injury still holds.
I don't know what capacity the depressed have. What was the evidence cited on that?
"One of the diagnostic criteria for depression is cognitive and memory problems; people with these problems cannot choose to remember things better, even with a gun to their head."
The repeated focus on genuine psychosis when a person's behavior does not respond to any incentives, does not seem like a good example to focus on in the context of Caplan's argument. If indeed it acts purely as a constraint and lacks any of the characteristics of a preference, then fine - it's an illness - and not subject to Caplan's point.
Caplan repeatedly states or implies that some so called mental illnesses are actual illnesses (although unfortunately he rarely elaborates on which these are). So if (some cases of) psychosis qualify, so be it.
But level of psychosis is on a sliding scale, even for the same person and even at different points in the same day, and whether or not they'd respond to the "gun to the head" test at a given point in the day based on level of psychosis is probably a question of varying probability, not one where there are points that would always be yes or always be no. So it seems like it's still a good argument against Caplan, in that he would be forced to admit that a psychotic person is suffering a constraint past some arbitrary cutoff point of how bad their psychosis is at the moment, with their milder moments of psychosis merely being a preference, which seems incoherent.
For Szasz (and therefore presumably Caplan) schizophrenia is the paradigm case of “not really an illness”. I don’t think he accepts that psychosis is a disease.
I suspect Caplan is admitting things like epilepsy, dementia, traumatic brain injury, Parkinson’s, etc.
Caplan is more saying that some so-called mental illnesses MAY be actual illnesses. He had specified that any definitely are.
You're right - I was imprecise in saying "are" instead of "may be," but the point is the same. Since Caplan acknowledges that *some* so called mental illnesses may be real illnesses, while asserting that "a large fraction of what is called mental illness is nothing other than unusual preferences," then focusing on the most extreme edge cases, where Caplan could agree is misleading.
How are psychosis and depression extreme edge cases? They're some of the most common mental illnesses, and symptoms of a lot more. What would you consider a non-edge case?
"Depression" as an umbrella category, is not rare. "Psychosis" as an umbrella category is much less common (It is not one of the top 10 most common mental health diagnoses: https://www.definitivehc.com/blog/most-common-mental-health-diagnoses).
Caplan's claim was that "a large fraction of what is called mental illness is nothing other than unusual preferences."
Within the set of all cases called mental illnesses, only a very small minority would be under the umbrella category "psychosis," and a smaller minority yet would match the example of "A psychotic person runs out in the street naked claiming to be God."
Psychosis is not a diagnosis; it is a symptom. It's a fairly common symptom (e.g. for BPD from your list), and schizophrenia and other psychosis-heavy mental illnesses carry a very large burden of the cost of mental illness.
It still seems strange to consider it an 'edge case'; a model of mental illness that does not explain psychosis would be a pretty bad one.
Fine, that doesn't affect my point at all. Let's get a ballpark idea of frequency. About a quarter of Americans suffer from a diagnosable mental disorder. What percentage of that 26% run out in the street naked claiming to be God?
It strains credulity to think this is more than e.g. 5% of that group (a whopping 4.4 million Americans) exhibits such behaviors.
This is clearly rare, even within the set of mentally ill people, although the umbrella term psychosis is surely much higher.
Even the set of *all* serious mental illnesses combined including (but not limited to schizophrenia and major depression), only affect 4% of all Americans (https://www.cdc.gov/mentalhealth/learn/index.htm), which is 80+% less than the percentage with any mental illness.
Meaning the vast majority of mental illness cases are minor - far from the extreme examples I highlighted.
>Again, Caplan could say that this is just a preference for attacking cops and then being killed. But in that case he should stop touting the “gun to the head test” as meaningful.
This really seems to miss Caplan's point. The "gun to the head" is a common metaphor for duress, given that typically people don't want be shot in the head. The question is whether incentives can affect someone's behavior.
In the rare case where someone wants to be shot in the head, then being shot in the head is obviously not a disincentive. Such a case tells us nothing about whether incentives could sway the person's behavior like preferences.
But as Scott's counterexamples point out, if you draw the line at "whether incentives can affect someone's behavior" then you can disqualify migraines, rashes, colds, cancer, etc. as being diseases too.
Okay, fine. Caplan's position seems very weak. I had the same objection to Caplan before this post was written. But the fact that an idea is bad doesn't mean that a particular argument against it is good.
My point about misunderstanding the "gun to the head" test is still correct, even if Caplan is wrong because of migraines.
Incidentally, I now see that Scott already raised this argument and Caplan already responded to it, here: https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply.
His response makes his position seem more reasonable than presented here, but still seems lacking. I have discovered a truly marvelous proof of this, which this comment is too narrow to contain.
>But as Scott's counterexamples point out, if you draw the line at "whether incentives can affect someone's behavior" then you can disqualify migraines, rashes, colds, cancer, etc. as being diseases too.
No no no no no.
"I am mentally ill." What is wrong with you? "I can't focus enough to hold down a job." What happens if you cannot hold down a job? "Nothing my parents pay my rent for me and make sure I do fine so I lay around all day and play videogames."
This person doesn't have ADHD. They don't need pills, they need life changes.
"I am physically ill." What is wrong with you? "my temperature is 2* above normal and there is mucus streaming out of my body constantly." Ok you really are sick. No amount of futzing with their incentives makes them *not sick*.
Sure EVERYONE replies to incentives. My barely verbal "autistic" sociopath half brother responds to incentives.
The point about incentives is that for some people who claim to be mentally ill, the primary problem in their life is their incentive structure. No amount of incentives are going to make my half brother functional. On the other hand when my sister was a junkie she didn't have "addictive personality disorder", she was just a junkie. As evidenced by her immediately going totally clean once she got pregnant.
If either medication *or* a change to incentive structures can alleviate a person's symptoms, is there necessarily an automatic reason to prefer the latter? Should changing a person's incentive structure be assumed to be less costly (both economically and in terms of preserving family relationships and stability) than just getting some pills? Do you think that anybody in person 1's situation would end up just laying around and playing video games, or do you think that many people could potentially have that safety net but don't need it?
I mean I certainly don't want my taxes/insurance premiums going towards "fixing" person 1 with expensive medication and counseling.
If they want to buy medication/counseling as an easy way out of their lack of motivation that should be on them.
I additionally think there is a moral hazard in medicalizing so many things and giving people excuses where it actively discourages them from having high expectations of themselves. "Oh X is hard, I guess I won't try". Nope dumb dumb valuable things generally are pretty hard, that is why they are seen as valuable.
That you find hour 200 this year of studying organic chemistry boring doesn't mean you are mentally deficient, it means you are a normal person. And if you don't want to stick with it it doesn't mean there is something wrong with you and you need pills. You just weren't cut out to be a chemist. That is fine.
It gets a bit more complicated when the question of who's bearing the costs expands beyond a single family to the whole country (or at least all of the insurance company's clients), yeah, but I'm not sure you can confidently say that the 'just change the incentive structure' solution will keep the costs within that single family either. A person I knew as a child did end up 'having their incentives changed' by getting kicked out of their house, and they've remained homeless - with all the costs to society and the state that implies - to this day, so far as I know.
As far as the moral hazard - yeah, I don't disagree, but there are hazards in pushing the line too far in the other direction, too. "Anyone should be able to do this; plenty of people far more dysfunctional than you in other ways manage it; why can't you?" That way lies bullying, ostracism, depression, suicide. There are going to be tradeoffs no matter where you draw the line.
(And what if the value to society created by having one additional organic chemist, enabled by having those pills, ends up matching or even exceeding the costs of those pills?)
"If they want to buy medication/counseling as an easy way out of their lack of motivation that should be on them."
Hmm...
What do you think about employers who supply their employees with coffee?
"'I am mentally ill.' What is wrong with you? 'I can't focus enough to hold down a job.' What happens if you cannot hold down a job? 'Nothing my parents pay my rent for me and make sure I do fine so I lay around all day and play videogames.'"
Funny thing is, I was in that situation myself for a few months - and Adderall was the exact thing that fixed it. The whole thing with executive dysfunction is it dulls your ability to respond to incentives in the first place: you simply don't have the mental RAM to actually model/evaluate the consequences of option A over option B. For a long time I couldn't get anything done unless I convinced myself my life was on the line. Now I'm in a much better spot.
(And before you ask, I'm American so I have to pay for the meds myself.)
>is just voluntary preferences
Again, this seems to deliberately misrepresent Caplan to make him seem sillier. The post keeps talking about people choosing their preferences. When does Caplan talk about choosing preferences? What does it even mean to choose a preference?
I think "voluntary preferences" is shorthand for "preference you can choose whether or not to satisfy" not "preference you chose to have".
I don't think that's what Scott means. Throughout the post he caricatures Caplan as asserting that preferences arise from free will.
E.g. "Maybe depressed people completely voluntarily choose to lie in bed for a few years while falsely claiming to be miserable and then shoot themselves in the head, and all of this is a perfectly free choice that they are happy with."
It sure seems that Scott's point is to imply that Caplan asserts that people choose their preferences and are therefore happy about them.
Caplan explicitly disavowed this portrayal, having written "People often wish they had different preferences, but this hardly shows that what appear to be preferences are actually diseases."
That's exactly the point: they are voluntarily choosing to stay in bed; they are not voluntarily choosing to have the preference to stay in bed all day everyday.
>this hardly shows that what appear to be preferences are actually diseases
Is Caplan claiming that real diseases must impose constraints, and if something happens to only change your preferences in a way that you don't endorse, then tough luck, but it's not a disease? Doesn't seem like a particularly useful definition.
As coming from a family of alcoholics, this example getting brought up again and again by Caplan is especially irritating, given that once you have maintained a certain blood-alcohol level for a very long time, it can be lethal to suddenly stop drinking. Without meds, best case scenario is delirium tremens, or you will have a 45 minute long seizure before you finally drop dead.
Try to solve that with a gun.
yeah, the guy is doing spherical cow thought experiments without ever having seen a real cow in real life
>His current post says that either you have to believe that mental illness doesn’t exist and is just voluntary preferences which are stigmatized by society, or you have to believe that homosexuality is objectively a mental illness.
>Not only are each of these incoherent ideas, they’re not even the same incoherent idea! You could easily accept one of the incoherent ideas and reject the other!
Of course they're different ideas! That's the whole point. If they were the same idea, then it wouldn't be either / or.
Caplan explicitly rejects the idea that homosexuality is an illness.
In between heterosexuality and coke v Pepsi: cilantro tasting like soap. Seems a lot like a preference, but we've found the gene and we know the subjective experience is different.
I don't usually hear it classified as a generic disease, probably because it usually doesn't cause much distress, but in a hypothetical culture that put cilantro in everything, or used it for sacred purposes...
This is starting to sound like Social Model of Disability. Come to think of it, a lot of this post sounded like SMoD. Could you write a full post on that some time? I think it could benefit from your analysis.
Yes I agree!
I am not sure if you have covered this argument somewhere, but even in terms of observable, economically meaningful behavior, Caplan's approach contradicts reality - I mean, how does he explain depressed patients spending considerable time, money and effort on therapy and various medications (that come with nasty side effects) TO GET RID OF THEIR DEPRESSION if being depressed is just something they chose to do?
Scott previously raised that objection, and Caplan already responded to it here: https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply.
Thanks. Yeah, that covers it. I am not sure how Caplan's worldview accomodates the possibility that someone hires someone to help them make better choices. If it's all just a matter of preferences that govern how you act within the constraints, and they are not to be questioned, how and why would you prefer, and put in time and effort, to have other preferences than the ones you have? Just have them and be done with it?
> I am not sure how Caplan's worldview accomodates the possibility that someone hires someone to help them make better choices.
I don't think Caplan's worldview would have a problem with people hiring consultants.
> Just have them and be done with it?
I don't think Caplan ever said people can change their preferences that easily.
Hm. I'm not sure whether there's an observable difference between an "illness" and "an unusual, strong preference that you'd like to change but just can't get rid of", but the former seems conceptually clearer.
Indeed, in the linked post, Caplan writes explicitly in this context "People often wish they had different preferences, but this hardly shows that what appear to be preferences are actually diseases."
He also links to a longer post about this point that I haven't looked at.
Obviously Caplan is arguing from a fallacious, dualist view of human behavior, but it also strikes me that he's performing a bit of rhetorical legerdemain by treating mental illnesses as merely matters of conflicting preferences, whereas I imagine psychotics and schizophrenics would encounter difficulties stemming from their unusual perceptions of reality even if they were living as hermits, or if they inhabited societies exclusively comprising fellow psychotics or schizophrenics
->rhetorical legerdemain
this seems to be an issue Caplan has often. I agree with him on a lot of things (immigration entirely) and find his work on education to be really interesting; but a lot of the time he seems to want to have a debate in a specific way and not in the way most of society sees a problem. His latest book "Don't Be a Feminist" should really be titled "Don't Be a Feminist (as I define feminist which is a very specific type of feminism that is easiest for me to be against but doesn't necessarily describe a large group of feminist thinkers and definitely not in a way they describe themselves)". It's really to his detriment that he does this!
The debate with Scott is a great example of how this is detrimental. There seems to be a large divide between Scott and Caplan on what preferences means and what constraints means. Instead of clarifying how he defines them or using different language that may better communicate his point, Caplan just doubles down on definitions he has written before which doesn't clear anything up!
Typo, missing "not" - You can imagine the depressed person choosing to throw parties and work hard instead.
A masterly rebuttal, but your footnote 4 is wrong. You can prefer something because it is less bad than the alternative; what you think is, the only thing worse than staying in this bed, would be getting out of it. Presumably Caplan has to say that you have chosen to believe that.
Caplan's view just seems sort of crazy. Like, you just read his description of schizophrenia as just being a preference to lie around all day claiming one sees demons, and it's so obviously false. Good job with the reply--very convincing.
Bryan Caplan has this wonderful way of debating:
Caplan : "Here's what I think. Anyone see anything wrong with it?"
Anyone: "Well, since you ask, here is the GIANT MOUNTAIN OF EVIDENCE you ignore, and the dozen major authors on this subject you clearly haven't read or engaged with."
Caplan: "...no one? Guess I am right about everything."
Case in point:
https://skepticink.com/prussian/2020/03/24/seal-the-borders/
What confuses me about this whole debate is that Caplan long ago admitted that some things can change both preferences and budgets: "Still, does this mean that the flu isn’t “really” an illness either? No. Rather it means that physical illness often constrains behavioral and changes preferences. When sick, the maximum amount of weight I can bench press falls. (Yes, I’ve actually tried this). Yet in addition, I don’t feel like lifting weights at all when I’m sick. Anyone who has worked while ill should be able to appreciate these dual effects. If you literally get sick, your ability and desire to work both go down. When you metaphorically get “sick of your job,” in contrast, only your desire goes down." https://www.econlib.org/scott-alexander-on-mental-illness-a-belated-reply/
Surely mental illness also affects both ability and preferences, though? E.g. if someone has serious anxiety, that will both affect their preference for avoiding stressful situations, and also affect what they are physically able to do (e.g. speak confidently in front of a crowd without having a panic attack). Gun to the head will not calm them down.
So I kind of have no real problem with Caplan's framework, but also just think he isn't actually applying it?
Well, let's apply the gun-to-the-head test here.
If Bryan has the flu and feels he can't lift the same amount of weights as when he's well, if we put a gun to his head *could* he lift those weights?
If he can, then 'having the flu' is just a choice, not a constraint!
Caplan asserts that he really can't lift the same amount of weight with the flu. Even if a gun to the head would cause him to lift more weight with the flu, the same gun would cause him to lift even MORE without the flu.
If he can lift 10 tons (look, I have no idea how much weight is involved in lifting or benchpressing and I am profoundly uninterested in finding out) when he's well, and then if he has the flu he feels all ouchy and can only lift 8 tons, but if I put a gun to his head and tell him I'll blow his brains out then he can make himself lift 10 tons even if he's all ouchy - he's choosing to be ouchy. Man up and stop choosing to be ouchy, Bryan! It's all in your mind!
No, I really don't believe he can lift the same amount when sick as when healthy, even with a gun to his head, nor does he.
I get that you're intending to ask about willpower in #2, and that in that sense my answer violates the "spirit" of the question, but I also think that the fact that it actually isn't about willpower for a lot of people is relevant and highlights where the spirit has gone astray.
Gun to my head, I'm definitely blacking out due to hypoxia before breathing. This is certainly true for at least any semi-serious freediver, but I suspect far more than that. At some point you're just not "will power" limited and it becomes a question of whether increased incentive can lower heart rate and energy expenditure further -- which is a bit like "As long as you manage to not be afraid, I won't shoot you". Is the person who feels fear there showing a preference for not dying or a preference for dying?
Or just that they're constrained by their ability to organize their preferences?
You might be pleased to learn that when I got as far as the 'restored' painting of Jesus, I laughed out loud (and I'm at work, good job it's early in the morning when there's only one other person in the building as yet).
Mostly this lines up with my views of Bryan Caplan (and to an extent Tyler Cowen) and their views on matters, which are - well, I can't put it the way I feel because that would violate the "Kind/True/Necessary and Principle of Charity" on here.
Let's just say that Caplan annoys me with his total self-assurance that being an expert in one domain means he's an expert on everything under the sun. Ditto Cowen and education.
Would Caplan claim that being diabetic is a choice? You can certainly make that argument about Type II diabetes, and maybe (if you feel like being a particular son of a bitch) about gestational diabetes - after all, nobody *forced* you to get pregnant, but Type I is being born with a deficient pancreas. Unless we're going to invoke karma and the Wheel of the Law and reincarnation, it's tough to argue a Type I diabetic 'chose' that.
" You can imagine the depressed person choosing to throw parties and work hard instead."
This man has his head up his arse, and I'm sorry if this phrasing forces Scott to discipline me, but fucking hell. I'm not formally diagnosed (mostly because I can't convince my doctor) but I'm sorta-kinda depressed at times (at really bad times I have spent something like three days lying in bed sleeping) and yeah, sure, Bryan: I can totes leap up, decide to throw a party (for all zero of my non-existent friends) and have a great fun time getting those extroverted tendencies of mine online and having the ol' happy-reward neurotransmitters flowing! It's only a choice to be crying my fool eyes out over my dumb life!
"One can imagine Bryan Caplan not being a feckin' eejit" but I admit, it's only a thought experiment and not reality.
"In particular, he claims I am FORCED to either accept that all mental illnesses are just “preferences” and so not illnesses at all, or as posited in a response by Emil Kierkegaard, that homosexuality is a mental illness and therefore bad."
I'm willing to bite the bullet on that one. "But - but - that's mean to the gays!" Yeppers, Bryan, I'm not a secular liberal. Whatcha gonna do now?
Or, that's a false dichotomy. Consider someone who is short and fat. You can argue all day about fat - is it an illness (an eating disorder) or a preference (for pizza and doughnuts), but short is just the way things are. Homosexuality is a tall/short, not fat/thin, analog.
Fat/thin attitudes may be changing. Clearly, in cases like the Prader-Willi example, the usual nostrum of "just stop eating so much! have some willpower!" doesn't work.
But for us ordinary chubbos, there may be something more going on as well that isn't down to "just stop shoving junk into your piehole":
https://edition.cnn.com/2023/06/12/health/obesity-changes-brain-wellness/index.html
"Obesity may damage the brain’s ability to recognize the sensation of fullness and be satisfied after eating fats and sugars, a new study found.
Further, those brain changes may last even after people considered medically obese lose a significant amount of weight — possibly explaining why many people often regain the pounds they lose.
“There was no sign of reversibility — the brains of people with obesity continued to lack the chemical responses that tell the body, ‘OK, you ate enough,’” said Dr. Caroline Apovian, a professor of medicine at Harvard Medical School and codirector of the Center for Weight Management and Wellness at Brigham and Women’s Hospital in Boston."
Note that we usually don’t say that being short *or* being fat is a disease! When someone is short enough or fat enough, we sometimes do, but not just when they’re shorter or fatter than the size that would work best in their social context.
Yes, it was purely an example of somewhere that the illness/preference schema can *plausibly* (not necessarily correctly) be applied.
Not really relevant to your larger point, but when you wrote "ditto Cowen and education", were you perhaps also thinking of Bryan Caplan, who wrote The Case Against Education?
I have kind of find hard time calling Down syndrome an illness. It is called a syndrome for a reason.
And also it doesn't really matter if we call it an illness or not. It is relevant only for people who care about policies and so on. They can use their own definitions if they wish, it shouldn't impact scientific thought.
This whole discussion sounds more like Plutonians debating if Pluto is a planet or not (a reference from Rick and Morty). If by scientific calculations their activities are damaging their environment what does it matter how they call it? It only matters for politicians who don't care about science but votes even if it leads to the ruin of the whole planet.
The same thing about Down syndrome. It is bad, we know it. We don't need to call it an illness to understand this fact. If we could invent a drug that fixes the genetics of a new-born baby, we would gladly use it. Maybe one day we will have such genetic cure. Maybe for those scientists who work on this problem it is important to call it an illness because those who give grants for these studies, follow bureaucratic rules – if it is an illness, you get money, if it is called something less – nada. But once you get money, you don't need to call it an illness anymore. Just study the thing whatever it is called.
"We would gladly use it" - that is you and me and most. Not all. We do have the option now to check for Down very early in pregnancy. And there are a few people saying: "foul! Insurances should not pay and people should not test!" Not because they anti-abortion (we are in Europe), but they do not want Downies to be aborted, because: less Down is bad.
Pregnancy, foetus, abortion are a very sensitive issues. I will not delve deeper into them. But once a child is born the legal position is much less controversial. In this case you have a child with a serious deficiency. If we could cure it, it would be mandatory.
Parents who refuse to accept the treatment would be accused of neglect and deprived custody of the child. It would be regarded in the same way as parents who starve their children or deprive them of a normal physical and mental development or do not provide necessary medical help if they happen to have a treatable disease.
Some people get jumpy when they see phrases like "fixes the genetics." Perhaps that's because they have a mental illness that can, in principle, be fixed too.
Perhaps. I have personal reasons to be wary of eliminating conditions like depression and Down's syndrome, so I may be overgeneralizing from my experience. I have severe problems with depression and concomitant addiction. I'm currently trying to recover from a crisis. All the same, I'm not sure I would take a pill that eliminated my depression because I don't know what else would go with it. Perhaps the tendency toward harmful rumination is a maladaptive extension of our capacity for introspection, and so the elimination of depression would also undercut a major source of human creativity. Robert Sapolsky's argument about the evolutionary value of a little, but not too much schizotypal disorder in a population leaps to mind.
https://www.youtube.com/watch?v=4WwAQqWUkpI
Going further down this road, my maternal grandparents' youngest child, my Uncle Joe, had Down's syndrome. My grandparents did a remarkable thing by raising him in their own home, because he was born at the tail end of the era in which institutionalization was the standard. Joe was the kindest, gentlest, most Christ-like person I've ever known. His devotion to his family was unstinting. He needed a lot of help, but he lived well. It's hard for me to accept the idea that it would be better had Joe never been born.
I might be jumpy because of my familiarity with the history of eugenics. I might also be jumpy because I have observed humanity's tendency to apply Fordist principles of standardization in realms where they are not appropriate, such as biology & ecology, where diversity provides resilience. Heck, I might just be jumpy because I remember my Tolkien: "Many that live deserve death. Some that die deserve life. Can you give it to them, Frodo? Do not be too eager to deal out death in judgment. Even the very wise cannot see all ends."
I use "death" here cautiously, but with purpose. In the current framing of American politics, I am staunchly pro-choice, since the question of when a fetus gains the rights of personhood is a terribly complex question, best answered by the women who most directly bear the risks and responsibilities of pregnancy. The standard of fetal viability used in my home state of Illinois seems reasonable to me. All the same, I have qualms about abortion, since if it does not end a human life, it certainly ends a process that will lead to a human life absent problems that lead to miscarriage. I despise Bill Clinton, but his formulation "safe, legal, and rare" seems entirely correct to me.
I don't understand what Scott's stated opposition to congenital conditions like depression and Down's syndrome means. Even if we fully understood the genetic predicates of these conditions, how can this belief be made operational without resorting to eugenicist techniques?
Scott addresses your concern about eugenics in his essay "Galton, Ehrlich, Buck."
What if we invented a pill that makes everyone's IQ to increase to 100 if it's currently below that threshold or stay the same if they're already above 100 IQ? Would we then classify everyone below 100 IQ as "diseased"?
It makes sense to take a pill to improve IQ, if it is a safe medicine. I mean, there are always some effects and some people will suffer from them. But as with vaccines we draw a line when very rare side effects are considered an acceptable risk so that everybody has to get them.
An adult can still choose not to take them. It is a free choice. But with children it would be considered a bad parenting to not give such a pill to your children if they need it. Probably at certain threshold could be considered a child abuse and parental rights to be taken away.
With the current ADHD medication I would say that side effects can still be quite severe, like stunting of growth and tolerance, even addiction to amphetamines, so I don't think they will be ever mandated to those who are veritably diagnosed with ADHD. We also don't have much evidence that they improve life outcomes, life expectancy or whatever. I remember that some studies even claim that they don't even improve learning just makes life easier for kids and their carers. Which is a good thing but insufficient for mandate.
A quick web search on Bryan Caplan reveals that he is an economist. As such he must have had rational actor decision-related concepts, such as "choice" and "preference", drummed into him for years and years and sees the world through that lens, and has a desperate urge to shoehorn everything into the same rubrik.
So it sounds like he is trying to extend this outlook from economics into areas where it is often not appropriate (just as it often isn't in economics itself - hence all economists' notoriously duff predictions!)
Although I find the intro to your post amusing (Caplin is a relatively well-known individual adjacent to the rationalist-verse) I think I have to agree with your conclusion: the compulsion to reject stated preferences in favor of revealed preferences is a strong one for the modern economist, as it should be, and I think Caplin is taking this framework and running with it where he should not tread.
That doesn’t work well as a refutation, since it does not specify where one should stop running, or why.
"In fact, if you put a gun to the alcoholic’s head and threatened to shoot him if he ever drank again, probably he would stop drinking."
He's never been around hardcore alkies, has he? I know that he's trying to pull the Chestertonian gambit as in "Manalive" collection of short stories, where the protagonist pulls a gun on a professor who is droning on about non-existence being better than life, and shocking him into wanting to live, but that's not a good method in reality (not unless you're sure that the person is only repeating fashionable absurdities and doesn't really hold to them deep down).
Put a gun to a depressed person's head, say you'll shoot unless they start throwing wild parties, and it's likely (if you pick the right - which is to say the wrong -time) that they'll say "Pull the trigger, I'd rather be dead". I've had those moments and I've had to claw my way out of them, and fuck you Caplan, it's NOT a 'choice' to do so. It's a struggle and a fight and I wish there were no necessity for it. I'm not "choosing" to feel like this, and you can't wave about "but you admit you fight your way out of it" as a choice. Mostly it's because I was raised that suicide is a mortal sin. Knock that pillar out from under me, and I'll take the other path one of the dark days.
"If we don’t like them, we should ask the people who have them to choose differently, instead of treating them as diseased."
Okay, I'm crying laughing here. So back in my social housing days, we should have said to our schizophrenic client "I'm sorry, but instead of sending the workmen out to change the locks on your doors yet again, can you instead *choose* not to believe that your neighbours are breaking into your house to smear shit on the kitchen walls and that we are sending cameras up through the toilet in order to spy on you on behalf of the government? Thanks!"
Yes, Bryan, *that* would *surely* work!
Same person on her meds: was stable, was able to hold down a job, did not phone or call into the office in a highly excited state
Same person off her meds: see above for sample of behaviour/thoughts.
Where's the "choosing" there?
As someone who is in this post, I feel like I need to speak up for Caplan here, because while I agree with you that while I don't feel like I can choose my mental states (I mean: technically I can; otherwise there's no way CBT could work), what to do about them very much does feel like a choice to me.
Now, it so happens that unlike you I haven't received the grace of faith, but I find that "Maul halten und weiter dienen" (https://translate.google.com/?sl=auto&tl=en&text=Maul%20halten%20und%20weiter%20dienen&op=translate) works just as well.
On reflection, I would probably do better to avoid people, and - failing that - keep my big mouth shut, but since I just failed at both, here we are.
Anyone who'd like to No True Scotsman what I just wrote is welcome to do so, if it makes them feel better.
I have been very hard on Caplan, but he (and Cowen) do have this breezy way of "now, if we just apply economics, then every problem can be solved" that annoys me (it's not confined to them, of course; every tuppenny-ha'penny expert in his own field thinks that expertise in one area transfers over to all areas, particularly if those other areas are considered 'well you don't need to be smart, not like what I am in this area, to get on well in this').
So yeah, there is an element of choice there. CBT is a good example; in the end, it's up to *you* to put the principles into practice and work on your problems.
But somebody in the grip of full-blown paranoid episode is not really capable of choosing not to be a screaming maniac. You can say they're 'choosing' to break windows or strip naked and run into the middle of the street ranting or cover their walls in tinfoil because the Greys have replaced their family members with replicants who are beaming mind-control rays into their brain, but that's not really a choice as most people would understand it, and you certainly can't talk them down from that - after all, the person trying to persuade them to take off the tinfoil and accept their family members as really their family members, well, that's what an agent of the Greys would say, wouldn't they?
You have to wait until they're medicated out of the episode or otherwise come back to their senses. There's no "choosing" not to be out of their gourd at times like that, anymore than there is "choosing" to be thirsty when you're in the desert at noon under the sun sweating with no water in your bottle.
I think part of the problem here is that both sides in the discussion are more than happy to ride past one another tilting valiantly at the strawman they set up.
The steelman version of Caplan's argument is that while not all mental illnesses are amenable to choosing, a great many are. Alcoholics Anonymous aren't pumping their members up with heroic doses of drugs, to my knowledge. If they are good for anything (and I gather they are at least moderately successful at their core purpose), it's through persuading their members to choose not to drink. Plus, I understand that AA doesn't treat alcoholism as a disease to begin with.
We aren't forced to adopt an all-or-nothing approach to the question. We can, in fact, examine each individual case on its own merits.
>One might as well claim that a paralyzed person could walk if they wanted, but chooses not to.
The kicker being, some of them can. My uncle pulled it off after three years of paralysis (before I was born). He's mentioned a few paralyzed people have had more feeling in their legs than when he started out, but they didn't recover because they didn't have the ridiculous drive necessary to get back up (the kind that lets you focus on standing up for a full year without making any visible progress to standing up, and then fall on your face a thousand times).
Kaplan's a dualist. He's talking about the preferences of a homunculus, not about the preferences of a body. The human brain has been dissected many times and no one has ever found a homunculus. The homunculus is an illusion (a kind of narrative component) that focusses the brain's modelling capacity on the successful survival of the organism. Our modelling is always imperfect, limited and faulty - mental illness makes it even worse.
Modelling human organisms as disembodied preferences is useful in economics. It's still a model.
Non-dualists can endorse a "revealed preferences" model too. Caplan's problem, as far as I can tell, is that he treats those preferences as irreducible, which, one, is obviously wrong, and two, cripples your model for no good reason. A bad reason might be emotional attachment to the incoherent "free will" concept, in which case, well, like the saying goes, you can't argue somebody out of something they weren't argued into.
Agree. It's the irreducible bit that is weird, like if it's not in the economic model it doesn't exist. The difference between stated or presumed preferences and and revealed preferences can be very illuminating at times but they are skin deep, by definition. In this context, we might say that psychology is where you go when preferences fail.
Caplan is making two fundamental errors, which I think Scott realises but doesn't explicitly name. Caplan shows that it is POSSIBLE to account for the behaviours that mentally ill people exhibit by invoking preferences, but mistakes this for showing that it is PLAUSIBLE. Scott’s counterexamples are basically ways of saying that the way ‘preferences’ are being invoked by Caplan are a) stretching the concept in ad hoc ways one wouldn’t do unless one was trying to win a very specific argument, b) basically unfalsifiable and c) simply not the best explanation of what we observe. The reason they exhibit these flaws is because Caplan treats preference as ‘thing that produces action’ and not noticing. No matter what change in behaviour is observed, it will be trivially true that preferences can explain it on this way of thinking. The only way you could even notice the flaw with this characterisation is by realising that it does a bad job of accounting for how we describe our other observations, and that requires noticing that ‘it is possible to explain their behaviour in this fashion’ is no where near showing that it’s a satisfying explanation. If someone’s behaviour can be explained by preferences no matter what they do, then ‘preference’ isn’t doing any useful explanatory work – it generates no predictions and constrains no anticipations. It’s the same flaw that affected early psychological work on traits. If you say someone is courageous in setting A but not in B, you can salvage any given trait in your explanation by just saying they have the trait of ‘Courageousness-in-A-but-not-in-B’, which works for literally any observation.
The second error is that Caplan’s test for ‘is a preference’ is ‘is voluntary’, but his test for ‘is voluntary’ is simply ‘is changeable by will given a certain stimuli’ [namely a gun to the head]. But he needs an argument for why we should think of ‘voluntary’ and ‘changeable by will given some stimuli’ are the same given they clearly are not the same concept at all, evidenced by a) folk understanding of those concepts, b) volitional will as a capacity clearly comes in degrees, and c) volitional capacity in some circumstances cleraly does not translate to volitional capacity in all cases (gun to head is likely to spike adrenaline which clearly increases some capacities you would otherwise lack).
> ‘voluntary’ and ‘changeable by will given some stimuli’ are the same given they clearly are not the same concept at all, evidenced by a) folk understanding of those concepts
Could you elaborate on that?
'Voluntary' invokes will but it also typically is understood as something like 'uncoerced'. We typically say that if I put a gun to your head and say 'give me your money or I'll kill you' you do not hand over your money voluntarily, even if the means by which you hand over the money rather than get shot is by using your will. It would be silly to say that since there was will involved, it was voluntary and therefore just a preference / choice like any other. One can quibble about whether this is how we collectively *ought* to think of 'voluntariness' (maybe you think most people are wrong about what the concept entails) but they're going to need other criteria / arguments *independent* of 'here's how I define voluntary'
@Bryan Caplan
Same thing I told Kyrie Irving
Stick to basketball buddy
What is the bottom line? How does this help us find out what criteria justify treating adults as incapable of choosing for themselves, or how how best to deal with them?
It almost makes sense to say, we should treat diseases and have compassion for those infected, and should not treat preferences and perhaps not have compassion in some instances. But then aren’t we ruling out diets or diet pills as treatments for being overweight? Why shouldn’t we treat preferences, if the person with the preference consents to altering that preference?
Szasz was against electroshock therapy, lobotomy, and probably would have resisted many pharmaceutical therapies for depression and other mental complaints. But again it isn’t the preference/disease distinction that is driving things. For adults that seem able to take responsibility for themselves (or whatever we should call the criteria), treating them should depend on do they want to be treated, is the treatment likely to be safe and effective, etc. But there are some people that do not wish to be treated, but also seem unable to take care of themselves. Szasz, Caplan (presumably), and maybe Foucault think that society transforms a social problem into a medical problem, using medical style treatments to control some who are difficult to control otherwise. In the extreme cases, there is something to be said in favor of this, just as society puts violent criminals in jail. The important question concerns when this is justified and when it is not. If murder was a symptom of an untreatable disease, isolation of murderers would still be justified.
So why are we arguing over disease versus preference?
>society transforms a social problem into a medical problem
Only if your definition of medical problems excludes social problems. Which is of course the underlying issue with all the brouhaha about homosexuality. Society changed its opinion leaving medicine hung out to dry.
Admire and love both of you (met Bryan&family last Sunday). The controversy pains me - though, as it leads to more posts by both of you, it must be a good thing. Scott "wins" by 42 to 3, I'd say. Nitpicking, here the three: 1. Scott's former "preference is just good for shampoo" was silly, (could count double for not admitting). Scott does much better now, saying "The middle three (straight, gay, even depression!) aren’t obvious, which is why we’re having this debate."
2. Scott about Caplan: "his position: that ALL mental illness is just voluntary preferences" - misrepresentation: even in the last post Caplan writes: "most". (Counting also against Bryan, as he never seems to point out which are "obviously" not preferences and which may not be). Oh, and not sure, Caplan agrees with "just".
3. oops, forgot ... maybe a few very tiny things adding up? Oh: Maybe the Szaszi/Caplan view has its uses?! Like: those addicted/"crazy" enough to defecate on the streets of San Francisco might be more able to control that urge if it resulted in sanctions - maybe birching (Singaporean style - any numbers on how often shit happens there?). But assuming "oh, they can not help it"/"we should help them more" may mostly result in: more crap?
End note: "free will/voluntary/preferences" are ambiguous - see Albert Einstein: "Schopenhauer's saying 'A man may do what he wills, but he cannot will what he wills' has filled me vividly since my youth and has always been a comfort to me seeing and suffering the hardships of life and it remains an inexhaustible source of tolerance."
>his position: that ALL mental illness is just voluntary preferences"
It's unfortunate that Scott misrepresents Caplan like this, as Scott *already made this mistake.* and Caplan *already corrected it* here: https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply.
Caplan says maybe some mental disorders are not entirely a matter of preferences, but he's clear that he's open to the possibility that they all are. That doesn't make much more sense than the view that Scott ascribes to him, and it's open to the same criticisms that Scott makes (just rephrase them as criticisms of the claim that "it's possible that all mental illnesses are just voluntary preferences," which is how Caplan clarifies his view).
His post started: "I have a long-running debate with Scott Alexander about mental illness. I take the Szaszian view that most so-called “mental illnesses” are not illnesses at all, but socially disapproved preferences". Now, I assume someone needs to count the entries in DSM-5-TR. Which is fetishism? Which agoraphobia? - Even if only 30% turn out to come with a relevant measure of "preference"-character - it might turn out that 86% of individual cases do: OCDs/phobias/sex-stuff/depression/ADHD are huge. And even some manic/shizo et al. may. And most addicts knew well enough Marlboro, J. Walker, Js, H and those painkillers to be addictive before they took their first dose. Plus we all are crazy (except Tyler). And man, do we still have some choice. "Most" may not be as mad as it may sound.
Hm fair enough, that's different from what he says later in the essay. But still, I think Scott's examples show that depression, ADHD etc can't be completely characterized in terms of preferences although one can argue that some of their aspects amount to preferences. (As I say elsewhere in these comments, though, it seems much more likely that what's actually happening in these cases is that people's *perceptions* about what will satisfy their preferences are distorted by the disorders, and they actually have quite normal preferences.) So he is still saying something extremely implausible here.
Mostly (!), I agree. (As I mostly agree with Scott in the debate:) - Nitpicking: a) Not sure, Caplan would subscribe to "can be COMPLETELY characterized". ;) He would argue: "preferences" play an important role/ are a not-useless concept with most issues. b) "quite normal preferences": drugs are universal in human cultures. Yanomami-men do love their highs ( David Good ). And I really do like my beer in summer. And winter. And not just beer. (all bottles on my blog are real). otoh: I do spent more time on CK3 than I might want to. Or do I? (Which is NOT my approach to the first wave of the opioid-crisis. Still, if I ever get the cancer that killed my dad, I WANT heroin, not those dumb opioids he got 24/6 (he found those hallucinations annoying - on 7 he died).
I'm trying not to fall into the trap of believing I know Caplan's actual point better than he himself has expressed, but your number three touches on it. I'm really perplexed by all of this - if I'm right I can't explain why Scott and his commentariat are so confused. Perhaps dueling mental inflexibilities regarding terminologies?
Caplan is brilliant at explaining non-purely-economic things in economic terms. He appears to be doing the same here. Perhaps his own terminological inflexibility prevents him from saying, "Look, I don't actually care what these things are. I'm saying we should pretend they're preferences for the sake of making market-based societal improvements.". I don't know what his improvements would be or if they would improve anything, but at least calling it all preference would surface and make legible the problems we have as a society in addressing mental illness.
Bryan still needs to tell us how long his magical gun to the head test lasts.
“Why not say that […] colds are just a preference for lying in bed and blowing your nose a lot?”
Because colds also prevent you from stopping your nose from running, various tissues from swelling, lungs from congesting, etc., even if a gun is held to your head.
Depression causes low serotonin levels and whatever even if a gun is held to your head. Schizophrenia causes hallucinations even if a gun is held to your head. Anorexia causes warped body perception even if etc etc.
I don't see how anorexia is changing capacities rather than preferences.
AN causes distorted perceptions like how OCD, anxiety, and depression do. People’s perceptions are not entirely subject to will. Distorted perceptions impact capacity.
While AN, OCD, anxiety, and depression are not psychotic, they produce their own distortion fields in a way that could be seen as on the same spectrum in terms of reality testing.
In fact depression can come with psychotic features when it’s really severe and OCD is specified by degree of insight. Panic disorder comes with its own perceptual distortions. I think perceptual distortion may be an important factor that clarifies how the preference frame as it relates to mental disorders is flawed.
Again, how is capacity impacted?
If you perceive your thighs to be fat despite objective evidence to the contrary, that continued perceptual distortion is going to create continued barriers to behavioral change and so limit capacity which a different person with less perceptual distortion would have in that situation.
I still don't get it: what is a specific example of something someone with anorexia CANNOT do even with a gun to their head that they could do in that same situation if they didn't have anorexia?
I'm making no claim about capacities for anorexia specifically. The parent comment seemed to be implying that colds are real diseases because their effects persist even if someone is threatened at gunpoint. I was pointing out that many mental illnesses also have effects that persist at gunpoint, and couldn't be arsed to think of a wittier third example than that.
I'll add OCD, ADHD, anxiety, and phobias to your list. Personality disorders pervasively affect people's perceptions in a way that affects their behavior and capacities. Substance abuse also.
If the whole underlying motivation behind Caplan-side arguments is that people have more responsibility for their actions then they feel like people generally acknowledge, then it seems to me it would be helpful to have that conversation instead and as it relates to specific situations. What I often hear running behind these debates is "who is entitled to my sympathy" and I'd rather we talked about that as a moral/philosophical question than deflecting it into this more abstract tussle over definitions.
I guess Caplan would call depression a constraint, and schizophrenia and anorexia preferences. I interpret him as distinguishing physical symptoms from mental ones. This may not hold up, but at least conceptually, we can imagine 3 categories: one where a biological hypothesis seems hard to refute, one where symptoms are purely subjective, and a third where the mechanism is not well understood yet.
I still don’t know how this all cashes out in terms of treatment. I don’t think Caplan is suggesting that people should ignore conditions where persons have preferences that they don’t want, or even that doctors should resist prescribing pharmaceuticals to treat a preference.. What should change depending on which side of the debate is correct, if either?
That is, existing laws pretty much ignore this analysis unless we think that only the legislature and the medical establishment count as competent adults who should determine the parameters of treatment for everyone else.
What would Caplan say in reply? He is unlikely to respond directly for a long time, if ever, so I will try to imagine his response. I found in composing it that Caplan has already spoken to almost all of Scott's points, albeit perhaps not to Scott's (or my!) satisfaction. As a result, I fear Scott and Caplan are talking almost entirely past each other.
Disclaimer: this is my summary of Caplan's views, not mine. I am not a Caplanian!
Scott thinks I have crudely mischaracterised his views, like an ugly restoration of a beautiful medieval painting. Unfortunately, his picture of my views is no more accurate. I am not asking you to adopt a crude behaviourism, but rather the most parsimonious account of mental illness. It is extremely surprising that Scott would attribute behaviourism to me of all people! But while our models should explain, they should not over-explain. We have strong reasons to be sceptical of people's self-report, in circumstances in which they are strongly incentivised to answer in a particular fashion.
On Internal States
Scott's reply bafflingly attributes a strong behaviourism to me. I cheerfully agree that pain is real, that people have internal mental states, and so on. Scott returns again and again to this theme, but I'll just pick out one example as particularly illuminating:
"Maybe depressed people completely voluntarily choose to lie in bed for a few years while falsely claiming to be miserable and then shoot themselves in the head, and all of this is a perfectly free choice that they are happy with."
No, the depressed person is (presumably) genuinely miserable, and not happy with his choices, because his preference (lie in bed) don't line up with his meta-preference - have a fulfilling life. But that doesn't mean his choices weren't voluntary, or indeed, that there was anything unusual about his situation. This is the normal human condition! No doubt many people wish they were more conscientious, or ambitious, or loving. No doubt many people struggle with time-consistency. There is nothing mysterious or diseased about any of this.
On The Analogy To Physical Diseases
Despite what Scott says, I agree that migraines are a real disease. They are (in part) a constraint, that the sufferer can't look at a light without feeling pain. With a gun to the head, the migraineur would still feel the pain. But sitting in a darkened room is no part of that disease. That is a voluntary response which - as Scott is forced to concede - would change under different incentives. Once again, my gun-to-the-head test is vindicated - headaches are a disease, but sitting in the dark is a preference.
More generally, both physical and mental illness can both constrain behaviour and changes preferences. Scott tries to reject this model for physical diseases, writing:
"The essence of colds is feeling unwell and ugh and wanting to stay in bed and having unpleasant congestion in your nasal passages. None of these particularly change your preferences."
But what is "wanting to stay in bed" other than a change in preferences? Similarly, what is "craving for alcohol" other than a very strong preference for it? I do not deny that mental illnesses may go beyond a change in preferences, and to the extent they do, then they are diseases. My argument is that no preference is a disease, not that all mental illnesses are merely preferences.
When the Gun-To-The-Head test fails
I proposed a gun-to-the-head test not because it represents metaphysical truth, but because it's a quick thought experiment to clarify matters. I'll go further than Scott - the gun-to-the-head test can never fully clarify! Just because someone doesn't change their behaviour in response to a gun to the head doesn't mean they couldn't. But my test is far from unfalsifiable. It correctly identifies a subset of behaviours that respond to incentives, and as such are preferences, even if it does not identify others.
Are Preferences And Constraints the Same Thing?
No, and I'm surprised Scott makes the argument. Every concept has some grey area in the extremes, but for this argument to work for Scott, he needs to demonstrate that these grey areas represent the preponderance of mental illness - which he can't, because they don't. If there are some narrow grey areas, we should tread carefully in them, but it does nothing to impugn the vast rolling fields of black and white.
There is no difficulty whatsoever in categorising his 6 examples. An internal state of desire (i.e. a preference) is not the same kind of thing as an internal state of fatigue. Yes, a state of fatigue may lead to an internal state of desire for low activity (i.e. a preference), but that is perfectly normal - all our preferences are grounded somehow. Examples 4, 5 and 6 are purely preferences, and so cannot be diseases. 1, 2, and 3 involve both constraints and preferences, so the constraint aspects of them (fatigue, pain) are diseases, but the preference aspects of them are not. A skin rash is not merely an out-of-nowhere desire to scratch the skin, but if it were, that would not be a disease.
On Justification
We have already established that the constraints and preferences model works extremely well across a wide domain of applicability. But, it seems, if using it suggests that some people may have strong, socially disapproved preferences, then it becomes vital that we discard it for an ad hoc framework with minimal explanatory power that justifies almost anything. We must pretend that these people are somehow compelled beyond their will to take actions, because that makes it easier for us to:
- Force them to make other choices.
- Pretend that their social preferences are normal and that as such they are "good people."
- Indulge our own neuroticism and avoid blame.
I take no position on whether this pretence is socially useful. I am merely pointing out its dishonesty.
Some sources I used:
https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf
https://www.econlib.org/the-depression-preference/
https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply
https://betonit.substack.com/p/the-szaszian-fork-another-reply-to
> Every concept has some grey area in the extremes, but for this argument to work for Scott, he needs to demonstrate that these grey areas represent the preponderance of mental illness - which he can't, because they don't. If there are some narrow grey areas, we should tread carefully in them, but it does nothing to impugn the vast rolling fields of black and white.
That's a very good way to express my reaction to reading a lot of this. For someone who speaks of rationalism as much as Scott does, I found it quite disappointing to see just how heavily he relied on the all or nothing fallacy here, insinuating that because an idea does not perfectly model reality that it's worthless.
To draw an analogy from the harder sciences, in (macroscopic-scale) physics, Einstein's theory of relativity perfectly models reality as far as we've been able to test. Newton's laws of motion do not. But when engineers need to calculate something, almost every time they'll reach for Newton's equations and not Einstein's, because Newton's work well enough under most circumstances and they're significantly simpler. At extreme edge cases they do break down, and that's where you reach for the more complicated relativity equations, but outside of that very specific territory, Newtonian physics work just fine.
I appreciate your steelman of Caplan. Cheers!
You have come closer than any other commenter (or Scott) to telling me what is at stake in the controversy. You seem to be saying, if these conditions are diseases, we have a better justification for forcing treatments upon the patients (or forcibly preventing them from using a treatment).
I don’t really see it, unless we make additional paternalistic assumptions. We can imagine forcing a treatment upon patients for their own good, or for the good of other people who must deal with them. Whether their condition counts as a disease or a preference seems irrelevant in either case. Someone with a preference may wish to change it using treatment, and someone with a disease may wish not to treat it. We can force a treatment on someone because we think it will improve their lives, or because we think it will improve the social environment. So the disease/preference distinction doesn’t help, unless one or the other has exclusive effects on such outcomes. A narrow consequentialist approach would base the conclusion on the expected outcome of the treatment. A more deontological or broader approach would worry about how to draw the line between competent adults that must consent, and others who need not, since maybe we would not like the outcomes resulting from a world where patients have no rights.
Is there something at stake here that I have missed, or is this all about terminology, with no implications for what ought to happen differently?
Forcible treatment is just part of it. If these conditions are simply preferences that respond to incentives, then:
- maybe it's not a problem if the alcoholic drinks himself to death
- if alcoholics cause trouble for wider society, maybe instead of treatment we should change the incentives (e.g. harsher punishments)
- maybe we should talk more about willpower and less about psychiatry
- maybe we should abolish insanity defences to crimes
- etc
These do not follow from Caplan's position in a way that is obvious to me, or conflict with Scott's.
“maybe it's not a problem if the alcoholic drinks himself to death”
Or maybe, even if it is not a disease, it is a problem for the alcoholics and their families. The families certainly, and often the alcoholics, would like to change the behavior, whatever its cause.
“- if alcoholics cause trouble for wider society, maybe instead of treatment we should change the incentives (e.g. harsher punishments)”
We currently do not have punishments specifically for being an alcoholic. Most jurisdictions make public drunkenness illegal, but one can be an alcoholic for decades without ever being arrested for public drunkenness. Many other factors are involved, meaning that punishment would only make sense at all if it worked, which it might not. This is not inconsistent with Caplan's idea, though I have not seen him discuss it explicitly. Maybe thinking of alcoholism as a disease or a preference has implications for how it might effectively be treated, but I am unaware of those implications so far. I would prefer to test cures or treatments, rather than deduce them logically.
“maybe we should talk more about willpower and less about psychiatry”
This could be true or false, whatever we concluded about the preference/disease debate. Nothing prevents psychiatrists from talking about will power, or other non-pharmaceutical treatments, and some do. I think exercise is being advised often.
Or maybe pharmaceuticals sometimes work great to help some people who have no disease with their non-disease problems. Even if we decide they have no disease, why would we conclude they should not use treatments they find helpful?
- “maybe we should abolish insanity defences to crimes”
I have been scanning the article Caplan published that began the controversy, and he at least mentions that Szasz opposed the insanity defense. If so, I think that was perhaps a mistake. My perception is that those who get off on the insanity defense are not let free the next day, but depending on the crime and circumstances, forcibly sent to a mental hospital. I can imagine that defendants exist for whom this would be more appropriate than ordinary prison, whether we think of them as diseased or something else. It might cause legal or ethical complications, if doctors are legally or ethically prevented from treating persons that do not have a diagnosed disease. As noted above, I don’t think it would be unethical if it actually helps the person, or at least if prison is the only alternative, and yet prison is inappropriate some reason. In extreme cases, social control is what we want, and we should admit it. There are dangerous persons, who should be isolated from people who are not prepared to deal with them. I don’t think that the physical cause of the dangerous behavior is particularly relevant by itself.
"Diseases are constraints, not preferences" is a completely incoherent definition.
How so?
> Or we could stop thinking like behaviorists, a philosophy which nobody has taken seriously since the 1970s.
I'm not a psychiatrist, so there might be some nuances to this that I'm missing, but the first thing I thought of when I read that sentence is that that would definitely be news to my wife. She's a social worker who works with adults with disabilities who are "in services," and sometimes she talks with me about work. An unnamed psychiatrist who she calls "the behaviorist" is an integral part of setting up and reviewing each client's "plan," and she frequently uses the term "behaviors" as a euphemism/jargon for "doing things they shouldn't be doing."
It certainly seems as if the agency she works for takes behaviorists quite seriously in the 2020s.
I think "behaviorist" as a job is different from "behaviorist" as a scientific model.
All right. Do you know what the distinction is?
The latter is the theory that internal mental states don't exist, only behavior does. I assume the former is a specialist in changing people's behavior.
That seems a little bit absurd, then. Where does the behaviorist think behavior comes from, if not mental states? Does "behavior" somehow abrogate the law of cause and effect? Or is your description simply an oversimplification? (Not trying to attack you or anything; just trying to understand what this theory is.)
https://en.wikipedia.org/wiki/Behaviorism
The weaker, more sane interpretation of behaviorism is as a plea for parsimony: That a lot of facts about behavior can be explained using models that make no reference to mental states (canonical example: reinforcement learning), and that it is more parsimonious/productive to study psychology through such a lens, treating internal states as a black box, than to propose mentalist explanations for everything.
That said, the strong interpretation of behaviorism (that mental states literally do not exist, or at least are completely un-observable and therefore outside the domain of science) was influential for a while in psychology. I agree that this position is fairly absurd. I think it's best viewed as an over-correction against the early psychoanalytic school, which had a tendency to explain everything in terms of unobservable mental states.
The problem I have with the behaviorist approach is that it leans so heavily into objectivism to the point of effectively denying subjective experience. Or at the least it throws up its hands in the face of overwhelming subjective experience, preferring instead to squint at a tiny sliver of objective data because that's what it's comfortable with.
Scientists will always prefer to lean on objectivism because empiricism is a cornerstone of how we study the world for the past couple hundred years. But when it comes to actually treating a single patient (who isn't the definition of a statistical mean), most often it's the subjective elements that we really care about, and the objective measures ("your biomarker levels are down") don't matter so long as the subjective experience is unchanged. No patient walks into the doctor's office and says, "Doc, my triglycerides are way out of whack, can you help me out?" They might complain about a headache and we check and their blood pressure is high. Then we treat them for hypertension, check the PB is down, but also *confirm with the patient* before claiming that the job is done. If the headache is still there, it doesn't matter what the biomarker says.
In this sense, behaviorism is uniquely unqualified for application to clinical practice.
It is not the theory that internal states don't exist, but rather the approach to psychology where you ignore internal states and focus on observable behaviours. This made a lot of sense a century ago when there was no way to observe what was going on in someone's head, and so you just had to rely on what they told you, which might be complete nonsense - much better to focus on their observed behaviours. Since improvements in medical imaging in the last few decades, behaviourism has mostly been dropped.
OK, that makes a lot more sense. Thanks.
Like many things, there is a motte-and-bailey of behaviorism. The "motte" is what you described: the attitude that if we can't measure something directly (using the technology that existed a century ago), let's first check whether we can make a good model without using it. The "bailey" is a positive statement that things outside your model are not real, and anyone talking about them is unscientific and wrong.
Behaviorism can explain many things that people have in common, but cannot explain things where individuals differ, because things like genes or beliefs (or mental illnesses) are outside its domain of study. Therefore, fans of behaviorism are often likely to assume that people are the same... it's just that you are not applying the right incentives consistently. (Which may be true in some cases, but in other cases there may be a genuine difference that is outside the behaviorist model.)
This falls into my area of (not expertise, but "hey I heard about this!") due to where I work.
There is a lot of talk about "behaviours" when dealing with children with special and additional needs, and it's not Behaviourism a la B. F. Skinner. It's more "Little Johnny is going through the biting phase and we need to cope with this". Because it's normal for kids to go through a biting phase, but if little Johnny is biting all the time, biting everyone, and really sinking his teeth into the other kids so they get hurt, then it's a Behaviour of Concern (this is why daycares etc. have Policies On Everything):
https://ncse.ie/wp-content/uploads/2020/04/NCSE-Resource-1-Behaviour-Support-Pack.pdf
"What is Behaviour?
Behaviour is anything that a person does that we can see or hear. It is how we act and how we respond in different situations. Behaviour can include laughing, eating, sitting, standing, walking, talking, singing and dancing. Behaviour is learned over time and through experience. This is common to everyone. Behaviour is also communication. This includes talking to others and using body language such as signs and gestures to express how we think and feel. We sometimes use less acceptable behaviours to communicate. These can include using bad language, nagging, shouting and complaining. These behaviours can occur when we are striving to communicate, especially when we feel we are not being understood or listened to. They are often an expression of feeling upset, frustrated and overwhelmed and often children with additional needs may not have developed a set of calming strategies to respond to these feelings.
Are all behaviours a problem?
No, not all behaviours are a problem. Sometimes behaviours may seem unusual e.g., flapping hands, saying the same word or phrase lots of times, throwing items in the air and watching them fall. Your child may display these behaviours because it feels good or helps him or her to self-regulate. Even though these behaviours may appear bothersome to you and/or family members, you need to ask yourself if it is beneficial to worry about them at this time? If however, your child’s behaviours are preventing him or her from taking part in ordinary family life, these would be considered to be behaviours of concern. These behaviours can become a very powerful tool for your child, causing him or her and your family a lot of stress.
What do behaviours of concern look like?
Some children and young people with additional needs display behaviours of concern. Behaviours of concern can include any behaviour that puts either the person themselves at risk and/or those around them. For your child, a behaviour of concern might be a bite, pinch, kick or head bang to either themselves or others. Other examples might be messing up a room in the house, namecalling, using bad language, screaming or refusing to do what you ask. Although these behaviours can be upsetting, worrying and challenging, it is important to remember that your child is using them to communicate with you about how they feel or what they need. It is important to ‘tune in’ to your child’s behaviours and think about what he or she is trying to tell you."
Someone should give Caplan some psychiatric drugs with known side effects to see how easily his "preferences" can be changed.
Caplan openly acknowledges that preferences can be changed with drugs. Almost all the objections (in the post and comments) have already been raised and responded to by Caplan.
Everyone appears to be missing Caplan's actual point, but I don't know how to surface this in the comments. I'll go back through and see where I should reply.
Well, Caplan otherwise seems mostly sane, but on this issue he looks like an alien barely trying to impersonate a human. Maybe he has an actual point, which can be rephrased in human language, but I'm lost at what it might be.
I responded higher up to the only other comment that appeared willing to entertain the possibility that he had an actual point to make. My comment was highly speculative and certainly not worth posting twice, but you can find it with ctrl-f.
The one about cities-infesting crazies who just might respond to incentives? I don't think that either Scott or most of the commenters would disagree about this, I certainly wouldn't. What's lacking are not the bright ideas, but the will to implement them.
That's merely one tiny example. The point is that thinking of mental illness as preference might from an economist's perspective open up options that aren't possible from an illness perspective. I won't speculate about what those options might be due to the fact that I've never been once able to predict what people like Caplan or Robin Hanson will come up with. They are just further out in that area of idea-space.
> There are lots of stories of some cancer patient who “holds on” until their beloved spouse can be by their bedside.
Those are the stories we hear because those are the ones that get told.
https://www.theonion.com/loved-ones-recall-local-mans-cowardly-battle-with-cance-1819565052
I had a friend who was a mortician. We were making plans and she said she wanted to hold off "until after the busy season" to plan some things.
Wait, what? Busy season? For people dying?
"Yeah, lots of people hold on until after the holidays."
So suppose that I want Scott to write an essay about (say) giant sloths. At this point it feels like a viable strategy would be for me to write an essay saying how one of his posts says that he must have a bunch of incorrect beliefs about sloths (but I present these as true facts). This requires 0 research on my part (I just have to pick random quotes and make up facts), but seems like it has pretty good odds of getting him to do research on giant sloths for me.
This would only work if you were sufficiently famous. But I think it would work on almost anyone! If Elon Musk kept accusing me of having "dangerous and damaging" beliefs about giant sloths, you bet I would write to try and clarify!
This topic is a bit more relevant to Scott's career interests than are sloths.
>Consider Prader-Willi syndrome, caused by damage to a region of chromosome 15.
Prader-Willi is quite interesting biologically, since it's an epigenetic disorder. In this region of chromosome 15, the father's copy is unmethylated and the mother's copy is methylated (and therefore silenced). If the father's copy is deleted, or the person accidentally gets both chromosome 15 copies from their mother, the genes aren't expressed.
The opposite can also happen, which causes Angelman syndrome.
"Certain neurological conditions involve intense and frequent pain. One example is trigeminal neuralgia, also known as tic douloureux. The term neuralgia stands for pain with a neural origin, and the term trigeminal refers to the trigeminal nerve, the nerve which supplies face tissues and which ferries signals from the face to the brain. Trigeminal neuralgia affects the face, generally on one side and in one sector, for instance the cheek. Suddenly an innocent act such as touching the skin or an even more innocent breeze caressing the same skin may trigger a sudden excruciating pain. People afflicted complain of the sensation of knives' stabbing their flesh, of pins sticking in their skin and bone. Their whole lives may become focused on the pain; they can do or think of nothing else while the jabbing lasts, and the jabbing may come on frequently. Their bodies close in a tight, defensive coil.
For patients in whom the neuralgia is resistant to all available medication, the condition is classified as intractable or refractory. In such cases, neurosurgery can come to the rescue and offer the possibility of relief with a surgical intervention. One modality of treatment attempted in the past was prefrontal leucotomy (described in chapter 4). The results of this intervention illustrate better than any other fact the distinction between pain itself, that is, the perception of a certain class of sensory signals, and suffering, that is, the feeling that comes from perceiving the emotional reaction to that perception.
Consider the following episode, which I witnessed personally, when I was training with Almeida Lima, the neurosurgeon who had helped Egas Moniz develop cerebral angiography and prefrontal leucotomy and in fact had performed the first such operation. Lima, who was not only a skillful surgeon but a compassionate man, had been using a modified leucotomy for the management of intractable pain and was convinced the procedure was justifiable in desperate cases. He wanted me to see an example of the problem from the very beginning.
I vividly recall the particular patient, sitting in bed waiting for the operation. He was crouched in profound suffering, almost immobile, afraid of triggering further pain. Two days after the operation, when Lima and I visited on rounds, he was a different person. He looked relaxed, like anyone else, and was happily absorbed in a game of cards with a companion in his hospital room. Lima asked him about the pain. The man looked up and said cheerfully: "Oh, the pains are the same, but I feel fine now, thank you." Clearly, what the operation seemed to have done, then, was abolish the emotional reaction that is part of what we call pain. It had ended the man's suffering. His facial expression, his voice, and his deportment were those one associates with pleasant states, not pain. But the operation seemed to have done little to the image oflocal alteration in the body region supplied by the trigeminal nerve, and that is why the patient stated that the pains were the same. While the brain could no longer engender suffering, it was still making "images of pain," that is, processing normally the somatosensory mapping of a pain land scape. In addition to what it may tell us about the mechanisms of pain, this example reveals the separation between the image of an entity (the state of biological tissue which equals a pain image) and the image of a body state which qualifies the entity image by dint of juxtaposition in time."
-- Antonio R. Damasio, "Descartes' Error: Emotion, Reason, and the Human Brain"
> “Should we set every tree in the United States on fire, then dump the entire Strategic Uranium Reserve in the Mississippi River?”
It's got to be telling us something that an Englishman who tries as much as possible to ignore US politics and indeed politics in general, immediately thinks: "I bet that's a Republican policy." Bad system-one. Bad.
https://twitter.com/BDSixsmith/status/1630170015050657792
https://twitter.com/JUNlPER/status/1653206203856461827
Yech twitter. What do the links say?
I'd endorse that policy for the lulz if it makes the right people sufficiently mad, so you might be onto something.
Yeah, you guys totally need to get another civil war going, you're all slavering for it... Have fun and try not to nuke anywhere else while you're enjoying yourselves.
In Argentina we have a word for people who think like how Caplan is thinking in this situation: we call them "boludos". This word does not necessarily mean a lack of intelligence. Very intelligent people can be boludos. Boludity stems from applying the wrong model of the world to a given situation, often because that model was successful in other situations. This can very easily happen to specialists, people who are very effective in some domain and are led by that success to believe their expertise transfers to other domains. Nobel Prize winners often hang on to very odd beliefs because of this. Chemistry describes everything, until it does not. Models of the word are optimized to obtain certain kinds of information from the world, and often suck at everything else. So you need to be able to switch between different models and frameworks to operate broadly in the world, and that we call wisdom. If you can't figure out which model to apply to each situation, or if you always apply the same model no matter what, like if you had a hammer and the world was made of nails, then you are a boludo.
That’s an interesting term.
Two others come to mind: In German, there’s “Fachidiot”. Fach translates to “subject matter”. Someone who might be very apt in a specific field but has a poor grasp of others.
It also reminds me of the French “déformation professionelle”, an expression that you’d use for a teacher who corrects their friends constantly, or a police officer who is looking for rule breakers at all times.
All of this probably falls into the “if you have a hammer, every problem looks like a nail” category
"you’ll sleep less, not more" is linking to the same paper four times. You surely intended to either link four different things or make the whole phrase a single hyperlink
This is an artifact of italicizing a word inside of a hyperlink, part of how browsers parse HTML, not something Scott did specifically. As far as I know there's no way to avoid this, other than avoiding heterogeneous formatting inside of <a href> tags.
The motte:
- Psychiatric conditions can be modeled using consumer theory.
- Psychiatric conditions can be modeled as changing both budgets and preferences.
- Decisions about classification of conditions, syndroms, diseases, normalcy, etc. involve politics.
- Some psychiatrists engage sometimes engage in what can be modeled as rent seeking behavior.
- The field of psychiatry can be modeled as engaging in some rent seeking behavior (especially historically).
- Some psychiatrists use relatively bad and incoherent models of mental illness.
- Mainstream views in psychiatry (especially historically) have included very flawed models.
The bailey:
- Consumer theory is the best way to model psychiatric conditions.
- All/the vast majority of psychiatric conditions only change preferences.
- Decisions about classification of conditions, syndromes, diseases, normalcy, etc. are entirely political.
- (All) psychiatrists are rent seekers.
- Psychiatry is a homogeneous field (across time and practitioners) and is best understood as primarily engaged in rent seeking behavior.
- (All) psychiatrists use relatively bad and incoherent models of mental illness.
- Models and understanding from the field of psychiatry have nothing to add to our understanding of mental illness and have no benefit over consumer theory.
I'd also appreciate if Caplan would give concrete examples of applying consumer theory to mental illness. This would include giving the concrete illnesses and showing data from which a demand curve can be extracted (including numbers), testing to see how well that demand curve fits various situations, etc.
For example, when quickly skimming I found this study: https://pubmed.ncbi.nlm.nih.gov/26216390/ which estimates the price elasticity of methamphetamine. However, a fully fleshed out consumer theory model of addiction would need a lot more than the study above seems to provide. Among other things, it would seem to need to include:
- The price elasticity of *individuals* in addition to aggregate populations.
- Change in individuals demand curve over time. For example, what does their demand curve look like prior to using substance? After using it 10 times? After using it 100 times? After using it 1000 times?
- Substitute goods.
- Discussion of wellbeing. [To the best of my understanding, one of the central claims of consumer theories in economics is something like the efficient market hypothesis, which claims that consumers maximize their wellbeing by optimizing against their personal demand curves. If this was the case, you would expect addicts to have higher wellbeing with cheap access to drugs rather than expensive access. Is this actually the case? Further, does coercive rehab change preferences? Do addicts have higher or lower wellbeing after coercive rehab and while they don't have access to drugs?]
It is very easy to propose an idea for a model, it is much harder to do the work of turning that idea for a model into an actionable model, and harder still to test the performance of that actionable model in the world. However, the primary value of models comes not from the mere idea of them, but from their fleshed-out and actionable implementation. The field of psychiatry has done a lot of research trying to make empirical models and study outcomes of various interventions. Economics seems to have done much less of this with regards to mental illness. Perhaps the best response to Caplan is "Hmm, seems interesting. I [as a practitioner] will consider this more when you have hard data for it.".
Caplan rarely talks about "wellbeing", he defers to the subjective preference of consumers.
There are two options I can see: One, utility correlates with some sort of internal state of a consumer (e.g. wellbeing or pleasure); two, utility does not correlate with any single internal state or experience.
It sounds like Caplan takes the view that utility falls under option two, in that utility doesn't correlate with any internal state, but instead is a measure constructed to explain the preference-like behavior of consumers. However, this is very strange: it is claiming that humans can be modeled as being utility maximizers, where utility is a single coherent quantitative scale, but also that this utility does not correlate with any single internal state or part of experience (e.g. pleasure). In other words, this utility *is a fiction*. This should call the accuracy of the model into question quite significantly.
I will propose, as an alternative that I believe lies closer to the truth, the model that humans do not maximize utility as a single coherent quantity, but instead perform actions that are reinforced by reward. Reward need not be coherent (i.e., people can prefer A over B, B over C, and C over A). In terms of calculus, one can think of standard conceptions of utility being modeled by a conservative vector field, whereas in reality I believe human behavior is better captured by reward, which is can be modeled by a *non*-conservative vector field.
Perhaps Caplan would be fine with modeling preferences as a non-conservative vector field. However, to the best of my understanding, this would be a pretty significant departure from standard economics, and would require substantially different analysis to understand things like market equilibria, efficiency, and other macroscopic properties that economists typically care about.
"Wellbeing" and "pleasure" are not normally considered the same concept.
I probably should have said "Discussion of utility" instead of "Discussion of wellbeing". I will claim additionally, as above, that I'm not sure that utility makes sense if there does not exist an experiential correlate to utility, which could be phenomena such as wellbeing or pleasure.
"Reward need not be coherent (i.e., people can prefer A over B, B over C, and C over A)."
Yes, I expect that having incoherent preferences would be damning for attempts to model them as rational utility maximizers.
"this utility does not correlate with any single internal state or part of experience (e.g. pleasure). In other words, this utility *is a fiction*."
This, in and of itself, could be less damning for modeling people as utility maximizers. For example, suppose that there were two brain centers with neural firing rates that met all of the consistency criteria for valid utility functions, and the person acted as if they were maximizing the 50:50 average of these two functions (which would then, itself, meet the consistency criteria) - but which does not match any single internal state. Everything would still work.
It seems to me that if there was a part of the brain that averaged these two utility functions, then there should be something that correlates to the averaging.
As a toy example, if we assume people have two parts of their brain, one of which is "sensual pleasure" and the other is "emotional pleasure", and that these to parts are averaged to get "average pleasure", it would make sense that a person feels overall pleasure, but could also subcategorize different feelings as being either sensual pleasure or emotional pleasure.
Now, it could be the case that the utility part of the brain is sub-conscious and not experienced. However, at a meta-level, this might make us believe that utility maximization matters less. Why would I rationally maximize utility if it doesn't make me feel good? This is especially pertinent if we are considering whether society should support people in utility maximization.
I do think path-dependence is a pretty big strike against utility maximization models, though (or, at least the very simple models). Personal actions and the world seem to change the utility function quite frequently, which makes things much more complicated. I.e., if human action modifies our utility function, then most of the complex and interesting behavior in humans might not come from the utility maximization itself, but from the dynamics that govern the evolution of the utility function. Such a situation would seem to be better modeled by differential equations/non-conservative vector fields and not gradient descent on a single-valued function.
That's fair.
"Personal actions and the world seem to change the utility function quite frequently, which makes things much more complicated."
Ouch! This makes it much harder to have a crisply testable theory too. If the utility function "existed" but has a lot of dynamics, trying to extract the differential equations from observations of choices gets very difficult...
You'll be pleased to learn that Caplan wrote here: https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf:
"During the course of any given day, individuals diagnosed with substance abuse, ADHD and ASPD act contrary to their impulses because giving in to them would be too expensive. Studies of demand elasticity normally find that consumption of hard drugs is
quite sensitive to price (van Ours 1995); in fact, the psychiatric literature on ‘contingency management’ shows that a high percentage of heavy users of alcohol and drugs will go cold turkey for a moderate price (Higgins and Petry 1999)."
Unfortunately, I don't think that that presents a satisfactory theory of addiction. Heavy users are not necessarily the same as addicts, and those studies are contradicted by the study above (https://pubmed.ncbi.nlm.nih.gov/26216390/) that finds relatively low price elasticity for methamphetamine and "some evidence of a positive cross-price effect for cocaine".
A robust theory of addiction (which would be useful, but is also one of many things that are sometimes modeled as mental illnesses) would need to do the work of comparing and explaining all of these studies and offering robust predictions that perform in the real world.
I think it would do much to clarify the discussion if Caplan provides/has provided a list saying "these are the things that are best modeled by consumer theory, and here are links to the best particular models for each of those typologies, including concrete and testable predictions, and these are the things that are not modeled well by consumer theory".
I don't think Caplan has yet granted any behavior as not being best modeled by consumer theory, only that genuine mental illnesses MAY exist.
Yes, I know, economists are retardeds.
SA writes: "But it’s a political question whether or not to classify any condition, including physical conditions, as illnesses."
Could I get some clarity about how the phrase "political question" is being used here?
I read that as various organized groups jostle and argue over definitions, policies, and benefits conferred as it relates to problems people have. Those groups include professional associations, legislative bodies, and citizen groups. The definitions, policies, and benefits change across time and space as a result of this jostling. Scientific research is one input into the jostling but the outcomes are more determined by political processes.
Centrally, because public money is used in treatment. Presumably, disease would not be political in a all-private libertopia.
Illness only refers to things we consider *bad*. (For example, mental retardation is considered an illness, high IQ is not. Both are a non-standard value of IQ, but having lower IQ is perceived as bad, therefore an illness, and having higher IQ is perceived as good - or maybe as irrelevant by people like Taleb - therefore not an illness.) This has various implications, for example many people believe that healthcare should help people get rid of the bad things.
But what is "bad", and what is "not bad"? The answer depends on people's personal values and cultural background.
For example, pain would be considered bad by most people, but maybe Mother Theresa would object that it brings you closer to Jesus, therefore chronic pain should be considered a blessing, not an illness. Talking to invisible spirits is considered an illness, because we do not believe in the invisible spirits; in a different culture, a person with the same condition might be a respected shaman. If someone is unable to focus enough to keep a job, that's bad, because we need jobs today. But maybe, in a future where everything is done by robots, no one would even notice this condition. Shortly, for any illness, we can probably imagine a society, where given condition would not be considered a problem, and therefore would not be considered an "illness".
(Though this ignores the part that people with conditions called "not bad" might still suffer from their consequences; even if everyone, hypothetically including themselves, agreed that it is not a problem. The people "close to Jesus" would still feel the pain. The people talking to invisible spirits would still... talk to something that is not there. People who are unable to keep a job now, might be unable to complete some of their personal projects.)
Gonna be honest, I don’t think you can actually arrive at a super robust “mental illness is very real” position without also embracing teleology. (This is no bitter pill for me, cause I DO embrace it, but I think it’s a bitter pill for a lot of people, and a lot of counter-intuitive, non-parsimonious theories abound as a result.)
Would you say more?
Just because teleology doesn't explain _everything_, doesn't mean that it's always useless. Just like economics.
Could you explain what you mean by "teleology" in this context? Do you mean it in an Aristotelian sense that people and things have specific ends that they exist towards, with the relevance being that mental illness is something that limits the ability to achieve these ends?
Correct! And that it is very difficult to talk about mental illness without both an existent idea of the end of humans and the end of human cognition (as directed towards the former end.)
Does this not apply equally to all forms of illness, injury, or disability? A human is meant to be able to use their bodies for various purposes, not just their minds, and where these purposes are impeded, we focus our attention there and call it illness or disease or disorder or syndrome? Some of it is based on suffering (the reduction of which I suppose comes with its own teleological implications) and some of it based on impaired functioning.
I have to believe that if Caplan were to speak to someone with actual, real-deal, no-foolin' psychosis for more than five minutes he would have a much harder time maintaining his position. Thomas Szasz did not have a lot of experience with psychosis during his training, and I think that is reflected in the positions he arrived at as well. I suspect genuine mania would also challenge his conviction, or very severe OCD.
Please explain to me the set of preferences that leads a Conservative Jewish woman to stop sleeping, repetitive give the Nazi salute to passersby on the street, and rant at the top of her lungs about how she would have been the best prison cafeteria cook at Auschwitz while sobbing and taking a hammer to a mailbox.
Please explain to the me the set of preferences that leads a man who loved working as a car mechanic more than just about anything in the world to become so constrained and confined by his rituals that he lost his job because he it took him 7-10 hours to feel prepared to leave the house, started picking at his skin so severely that he opened up patent holes in his cheeks, and begged to be referred for deep brain stimulator implantation because it was absolutely the last hope he had left for living any life he regarded as worth living.
I don't think it is possible to articulate a coherent idea of "preference" that encompasses these things that bears any relationship to what use the word "preference" to mean in ordinary language. At some point you can use the same syllables to describe a concept but it is far enough away that I think it is perfectly legitimate to say then that some "preferences" are, in fact, disordered.
Some truths are difficult to grasp in a non-experiential fashion.
What is there to explain? Some people have strange preferences. How would you explain any of yours to someone who thinks them peculiar?
If satisfying the preferences in question is recognized to be worthless and does not actually contain or provoke any rewarding stimuli of any kind (the OCD) or does not pass the gun to the head test (manic people get shot to death by the police all the time), and both of them are radically different from an honest and true accounting of preferences (that have been true for the rest of the agent's lives and are internally consistent with each other), precisely what information does calling it a preference convey?
If.
I reject the characterization of one's accounting of one's own preferences as "honest and true." Revealed preferences are the true ones. And satisfying those is the only thing of value, by definition, so it can't be "worthless."
The information is the assertion that there is no meaningful distinction between the revealed preferences of the so-called "mentally ill," such as those with "depression" who lie in bed all day, and those considered sane, such as people who to prefer Pepsi to coke, and a characterization of one set of preferences as a disease is an expression of one's subjective judgment.
I agree that if you stipulate your conclusions you are justified in drawing them.
If one cannot regulate one's behavior in response to 'these armed men will shoot me if I do not', and prospectively and retrospectively the agent is confident in their desire to regulate the behavior, in what sense is this a preference?
There is a reason even the radical behaviorists (like the ACT people) feel the need to talk about internal, 'private' events in contradistinction to strict stimulus-response pairings. They feel they are forced to do so by existing empirical evidence and to provide a useful account of individual human behavior. You may want to reflect/investigate what that is so.
You also seem to assume that human minds are unitary and do not have competing and often conflicting drives subserved by relatively independent/encapsulated systems. Behavior that is in some since influenced by choice but is not a preference emerges naturally out of such a system. On what grounds do you maintain that there is a coherent self that even has a single specifiable set of preferences?
I think it helps to be conceptually clear what you mean by the term "preference." Scott is using it to mean a consciously attended mental state that some state of affairs should obtain that one can consciously reject. There clearly are numerous mental conditions, including examples listed above, where that is not reasonably called a preference.
I suspect you mean something more like goal-directed behavior, where the goals are one's preferences and some goals the human organism has just seem odd. This isn't really how the term is used in common parlance or philosophy, though.
When we talk about mental illnesses or physical ones as a "disease" what we mean is that the state one is in is reasonably thought of as undesirable. That is to say, if someone fully understood what their desires are and how to fulfill them, they wouldn't want to be in that state. When it comes to something like "cancer" you can get easy, near universal agreement. When it comes to some things described as mental illness, that's not as obviously true, though when it comes to severe psychosis, it is.
The most popular modern conception of what a mental disorder is relies on a person functioning poorly in their social environment. It should be immediately obvious to you that we can say the problem is more the social environment than it is the person, and if that's the case, then we don't consider that a mental disorder. For instance, if a society banned women from reading books, but some women kept doing so anyway despite risking going to jail, we might call these women as having "book use disorder" but we shouldn't, because a society that bans women from reading books is morally wrong. So it is with, say, homosexuality.
Per the "functioning poorly in their social environment" definition, the obvious analogy to your hypothetical "book use disorder" and homosexuality is pedophilia. Is it that, by miraculous coincidence, your beliefs just happen to be objectively morally true, or are your statements about what mental disorders are really just your own version of what is sinful and what isn't, gussied up in scientific terms?
Given the context, the comment section of an essay about the term used in an economics sense, it ought to be clear that the term "preference" is here used differently than in common parlance. (Among philosophers, your mileage might vary; I do not grant that none of them are intelligent enough to use the term in a more meaningful way, but you might be right.)
The argument was that when you define mental disorders in terms of one's global functioning, as clinicians do, you necessarily import judgments about whether you consider their social environment to be morally appropriate. A compulsive desire to read books despite a ban on reading books or have romantic relationships with members of the same sex despite societal sanction on them is apt to cause dysfunction, but we wouldn't consider that a disorder because we recognize that the problem is with environment.
You can reply to those judgments with, "Oh, so you think you're morally right?" with the answer being, "Yes, yes I do." That's not a miraculous coincidence as I don't think moral judgment is a random walk. But, again, my point is you cannot avoid the moral judgment, if only implicitly.
This does not mean that mental disorders are just a description of what is "sinful." It's not saying that cyclothymia is immoral. Rather, they are a subset of states of being we've described as undesirable. We just have to be careful about thinking *why* they are undesirable which requires us to think about the context in which they exist.
Looking at Caplans argument -- as Scott presents it -- I feel that Caplan started with writing the bottom line "Therefore, mental illness does not exist." before filling the stuff above. Credit where credit is due, I don't think I could come up with a better argument for that proposition.
Just like about every other concept, the concept of illness has some blurry outlines. Preferring Coke to Pepsi is not generally considered an illness, though if Coke had terrible side effects or external costs not present in Pepsi some might give it a fancy Latin name and call it a disease.
Of course, somewhere, someone will call a Deficit in Appreciation for our Supreme Leader a mental illness and try to place people in reeducation camps or something. I think the obvious place to make a stand against this sort of thing is to be against any treatment without (at least prior) consent, not the definition of illness.
Also, why is Scott the one feuding with Caplan over this? Scott's side is not the fringe position here, his opinion is probably shared by more than 80% of the medical professionals. Is it simply that Scott was the most high-profile person to refute Caplan?
Caplan says that he believes most "mental illnesses" are just extreme preferences, leaving the possibility that some really are illnesses.
Caplan's position is so funny - he's just very obviously taking the abstracted models about human behavior we use to make economic models work and explicitly using them as his understanding of psychology.
I don't think that's quite right. What he's done is cast Szasz's understanding of psychology in economic language.
Yes, he's taking models that explicitly assume the "rational, self-interested actor" and attempting to apply them to conditions where the whole point is that the person in question ISN'T rational.
My advice is to not engage with this person.
I have no horse in this race and no expertise in this field but surely talking cures can’t work unless the brain/mind has a choice. My dad had a major depressive episode a few years back which was cured, only and finally, by pills. Now he is back to normal - and normal is happy and loquacious so that was quite a change and quite a reversal. He had no choice, perhaps, without the pills.
My uncle was treated for alcoholism and as far as I know he just did AA and some retreat or other. Now he doesn’t drink. ( Ok some people say that he still has alcoholism - but I never really believed that - it’s like saying that somebody who eats well still has an eating disorder. Sure both the alcoholic and the ex-anorexic can fall off the bandwagon but while on it they don’t have the disorder).
So in some cases it looks like there is a choice to overcome some mental illness where choice is “this can be fixed without chemical intervention”.
I guess the way I think about this (as a psychotherapist) is that we have varying degrees of influence over the circumstances of our inner and outer lives. It's rare that we have total control and it's rare that we have no control. Most experience happens in the space between where it's about maximizing our influence within constraining conditions (again, conditions being both inner and outer).
One's default mode perceptions and default mode coping repertoire in the face of stressors are inner conditions that act as constraints. It's possible to change those inner conditions and thus loosen constraints (or increase capacities) through medication as well as through learning new ways to perceive and respond to those conditions. Some of that learning happens through words as much learning does, but I wouldn't reduce the intervention to "talk" since the actual change happens in the moment-to-moment way the person moves inside their daily life.
Alfred Marshall, inventor of modern mainstream price theory, used as a motto of his textbook Principles of Economics Latin phrase "Natura non facit saltum", i.e. Nature does not do jumps, attributed to Charles Darwin. Meaning that there are no hard lines between (some) phenomena, just one concept gradually morphing into another through edge cases. Perhaps Caplan needs to be reminded of that, since it obviously applies to a distinction between constraints and preferences.
I just want to say: this post particularly changed my mind. I read Caplan's original piece on how mental illnesses are actually just stigmatized preferences, found it compelling, then went to read your rebuttal but got sidetracked and kind of forgot about the whole thing. As a natural contrarian Bryan's position seemed reasonable to me. Particularly compelling is the parallel to arguments from disgust: people argue condition x is disgusting and therefore immoral, when such a mental model would produce virulent homophobia 70 years ago. Then I read this, remembered the whole thing, and it's pretty clearly wrong. I'm sure you don't particularly enjoy writing posts about this stuff as a psychiatrist, so just know you changed at least one person's mind.
Well done. Caplan is comically bad and intellectually dishonest in his thinking when it comes to this subject.
How is he "dishonest"?
Intellectually dishonest in the sense that he clearly has pretty severe biases that he isn't even trying to resist. Or else a person as smart as he is would acknowledge, or probably even anticipate, the sort of objections Scott raises and abandon this stupid preference view.
Guess what? He did anticipate many of the objections Scott raises, and subsequently responded to them, as well.
Maybe deciding that someone's view is stupid and that they are dishonest should follow reading their actual work, rather than the paraphrases of their work by someone disagreeing with it.
I'm pretty sure I've read every Caplan essay on this topic and I don't agree with your characterization, but it's possible I've missed something! In some cases he's responded in the technical sense that he wrote something he claimed was relevant, but the responses I've seen from him completely missed the point.
How we define "disease" doesn't matter one iota politically. What matters is whether the individual is hurting others. If they are not, you should respect their freedom and not lock them up for thought crime, however irrational or abnormal you think their thoughts are.
If you disagree, fuck you and everything you stand for.
I agree with not locking people up, but it's still useful to call things diseases. Or to rephrase, to say that we should work towards trying to eliminate certain conditions instead of just saying "well the only harm is to yourself, so good luck to you bye". Where the politics comes in is deciding which things are worth eliminating and which are just normal differences. For example, I would say severe OCD is pretty clearly in the former, while disliking alcohol is in the latter. Homosexuality switched from the former to the latter in the popular mindset, while things like schizophrenia and deafness (more of a discussion about "disability" than "disease" for that one) are heavily debated today.
Whoa. Escalated quickly. We also sometimes “lock up” people who are a danger to themselves.
For a more interesting example, consider a person who begs you to restrain them and prevent them from hurting themselves, when outside of the episode -- but during the episode, they yell at you to let them be, and to give them back their guns.
I would say the spectrum you describe is strictly a constraint spectrum and preferences are *informed* by that.
Another way to look at the constraint spectrum: if doing X will result in Y bad thing, the degree to which you are constrained from doing X is proportional to the degree to which Y is bad.
I think Bryan has unusually high willpower/self control/whatever you want to call it, so when he sees people behaving a certain way, he concludes they must prefer this.
One example that should also apply to Bryan himself, is tiredness though. He should imagine he's really tired and can't think straight anymore, such that if he was left alone he would go to sleep. But now we put a gun to his head and force him to stay awake. Fearing for his life, he manages. This looks like a preference. But then we pull out a math test and say if he gets everything right he'll get $100. If he wasn't so tired he would be able to do it, but because he's so tired he's careless and makes many mistakes. That sounds like a constraint!
Furthermore, if we offer him a magical pill which would make his tiredness go away without any side effects, just like if he got a good 8 hours of sleep (i.e., a much better pill than caffeine), then he would probably take it in order to earn that $100! How does that fit into his framework?
The math test shows it to be a constraint under his framework. And even with a gun to the head, he would eventually fall asleep.
So Caplan's model is nuts but your response has given me a framework to think about my own concerns with mental illness diagnosis. I generally think destigmatizing efforts are good but also think there's a cost we're ignoring and was having trouble articulating it.
That cost is that while the line between a constraint and a preference is extremely blurred in reality, we tend to dualistically categorize things in mental models. Imagine a universe in which we have a concept of a lower stigma, culturally accepted Can Only Run 5K Disease. In such a universe children whose legs will never carry them 6k won't suffer horrific pain in gym class - they'll be excused. They'll be mocked less for the inability to run. The diagnosis would make life better for thousands.
People with a strong preference for not running 6K can now model it as a constraint though. Maybe it makes them feel tired and gross and they dislike feeling tired and gross more than the average person. They're correct this mental state can be modeled as either preference or constraint, but modeling it as constraint changes how they think of "problems to be solved by running 6K." Where before it was "It would be extremely difficult for me to run 6K" now it's "I cannot run 6K." I can't prove it but I suspect that if you put a gun to the head of a person absolutely convinced they had this condition, who had made this condition part of their identity and made Facebook posts about how it impacted their life, they would be able to run exactly 5K before being shot.
Can you keep going and say what the cost is we're ignoring that you're concerned about? Is it loss of "grittiness" in kids? Is it more malingering? Is it moral offense at what people do for attention? Is it cost to the healthcare system? Some other thing?
None of those, except maybe something adjacent to "grittiness." Just simply the loss of choice - mentally converting something you don't want to do into something you can't do turns "achievements with high cost" into "achievements eternally out of reach." I think the ground lost here is probably less than the ground lost by assuming everyone can do anything. I have just debilitating ADHD and tried to be a lawyer - I was paying far far far too high a price to accomplish a goal that didn't make sense for me and in a culture that treated that as a moral failure (like U.S. culture only 10 years ago) I would never have recognized that. Nonetheless, some ground is lost.
Trying to understand your personal comment. Your assessment about trying to be a lawyer while having debilitating ADHD is that the trying so hard didn't make sense but you had a hard time seeing that because our culture considers any shortfall of trying hard to be a moral failing? I think I'm not understanding that relative to what ground is being lost.
The ground being lost if I understand you is that people with mild mental health type impairments will say "can't" when they should be saying "could" and that medical support for "giving up" in a way will deny society and individuals the benefits of their trying harder?
Why is trying so hard valorized so much I wonder? I get that we like hero stories and people fighting against all odds to overcome impossible hurdles. But isn't it good that we also find ways to make functioning well in good health easier for more people to accomplish? We could take a stand that we should only be treating people with moderate to severe mental illness but everyone with mild impairment should buck up or be shamed for laziness. But does that not come with its own costs, that approach?
My experience working clinically is that the people who have mild impairment are one life stressor away from moderate to severe impairment. This is also supported by research on the value of treating psychosis in its early prodromal stages before it blossoms into the full-blown creature. We know this about pain management as well. The resources it takes to treat something when it moves to moderate and severe are a lot higher.
Yes I don't think I was very clear with that. My point was that this is a spectrum - on one extreme we assume everything is an issue of willpower, on the other we deny willpower altogether. In the first case people (including my example) endure a lot of distress for no good reason. In the latter, people are led to believe they literally cannot do things that are simply difficult.
The point of sharing the personal anecdote was to make it clear I think it's *good* we're moving away from a "pull yourself up by your bootstraps" mentality - that the fact that I think we're ignoring a cost doesn't mean the cost isn't worth paying.
But ignoring the cost altogether as some subcultures are doing will eventually cause us to choose a place on the spectrum that is too far away from the personal responsibility side of things. And I've had a lot of trouble articulating why this bothers me. But the "preference/constraint" dichotomy helps me to articulate this - constraints, generally, are bad. Like the rabbit/duck illusion, many issues tackled by mental illness diagnoses can be viewed either way. Viewing a constraint as a strong preference or costly expenditure is sometimes more useful than viewing a strong preference as an absolute constraint. I'd just hate to live in a world that didn't recognize that.
A personal note that might cloud that point yet again - I legitimately don't know whether my inability to practice law is better modeled as a constraint or a strong preference. Certainly if my boss came to me and said I'd be fired if I didn't shape up, I could be more successful for a few weeks or months. I don't know whether I would have been capable of sustaining that level of effort long-term. But crucially for me what actually caused me to move on was the belief that even if I was successful it would have been at much too high a cost. That let me model it as a choice instead of a failure.
Thank you for saying all this, totally clear to me now. Or at least I think so. :)
I can add on to what you say from a couple of other angles. I agree with all you say.
One angle is one of my favorite Buddhist quotes which is "The great way is not difficult for those who have no preferences." (3rd Zen patriarch). I think about that up against what I saw for a time as the dominant mode of parenting among lots of people I knew which involved offering up a small child a huge number of meaningless choices. "Would you like apple, grape, orange, or pear juice? Would you like it in the blue, yellow, or purple sippy cup? Would you like to have it while watching this, that, or the other video?" In certain corners of society, it seems to me we're behaviorally programming people to have absurdly specific preferences in a way that can only contribute to delusion and suffering. It's good to learn to travel light and know you'll be okay.
The other thing is I think people having a sense of agency over their own lives is really important for mental health. So like your modeling not being a lawyer anymore (if I understand that right) as a choice instead of a failure seems really healthy to me. The whole notion of failure seems unhealthy to me unless we're talking about in the sense of "I tried that experiment and it failed, so now I'm trying another one." I think it's really important for people of whatever mental condition to have sense of capacity to learn to shape their lived experience. To not get stuck in victimhood stories and to identify choice wherever we can.
Despite being essentially a perfect rebuttal, this is obviously not going to change Caplan’s mind. Perhaps a day of shadowing in a psychiatric ward would do this trick? After the requisite HIPAA training of course.
I can’t image meeting a poorly controlled Schizophrenic patient and then still thinking mental illness doesn’t exist.
His position is that most don't exist, so he can make allowances for some. He might need more than a day.
One thing seems to have clearly emerged from all this, uh, debate: we should really be using the “point a gun at the patient’s head” treatment for a whole swathe of somatic and mental illnesses, since it appears relatively effective in many circumstances.
I came away with exactly opposite thought: one should NEVER use "gun to head" as test for anything.
We should never desensitize ourselves to violence including self harm.
I find it very disturbing and harmful.
Feels a lot like the Harris/Dennett "debate" about Free Will. Harris says "particles, therefore no free will". Dennett says "I object to the definition you use, because of the following reasons (...)" and then Harris replies with "but particles, therefore no free will!"
I don't think either Harris or Caplan intends to be foil for a more careful thinker.
Scott, you write as if Bryan is doing something nefarious, in his manner of pursuing this issue.
But I think the truth is, he's just not paying that much attention to you. As in, I bet he's not even really reading (at least, with much attention) those long essays you write. He simply isn't *crediting* what you write all that much. He then goes on to reassert his position in a manner that a third-party might (perhaps rudely?) describe as a bit autistic.
Now, you might think that such behavior *is* nefarious. And maybe it is, at some meta-level concerning modes of discourse. But my point is that he's not deliberately trolling, or anything like that. He's just lacking self-awareness in a certain characteristic manner.
Many people, very much including rationalists, behave in exactly this way on other topics. It's just cognitively really difficult to closely attend to arguments that one viscerally does not credit. I've encountered that very mode here, e.g., on the topic of whether witches were historically real (in terms of their own self-understanding and practice).
I'm uncertain you and Caplan actually disagree about anything except what the word "preference" means.
I think the view that "mental illness is a preference" makes a sort of sense in a dualistic framework. If the incorruptible soul/perfect mind controls the body, then any disorder beyound brute physical constraints (i.e. failure to run due to arthritis) must be a choice, by definition.
If, however, the mind is a function of the body just like everything else a person does, then the mind can be damaged just like every other function. Arthritis makes it impossible for people to run, and depression makes it impossible for them to choose to run, but we have medications to at least mitigate both of these conditions; and in cases where we don't, we're working on it.
So, if that's what it comes down to, then the two positions (mental illness as a preference vs. as an illness) are irreconcilable, because dualism requires a kind of faith by definition (not necessarily religious faith, but still), and faith cannot be adequately communicated (though we're working on it). It just depends on which side of the coin-flip you happened to land, ontologically speaking.
>that all mental illness is just voluntary preferences
is a terrible position, I cannot imagine he actually believes that. That said I do tend to think he is correct that a big portion (possibly a majority) of what is treated as "mental illness" is really just people with poor preferences/behavioral ruts (particularly ADHD/depression).
Though perhaps that is not the central case as I think that is mostly the "frosting" of mental illness treatment, and the cake is the truly crazy people.
> I cannot imagine he actually believes that.
He doesn't. Scott is deliberately misrepresenting him.
While normally I'm loathe to attribute to malice that which is adequately explained by incompetence, Caplan writes explicitly that "a large fraction of what is called mental illness is nothing other than unusual preferences," and already explicitly corrected Scott for accusing him of asserting that all mental illness is a matter of preference.
See here: https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply.
[There is another subtler misrepresentation of Caplan in those 8 words that you quoted, that I describe here: https://astralcodexten.substack.com/p/sure-whatever-lets-try-another-contra/comment/17888326.]
Why is Scott so careful/thoughtful sometimes and so....well not...others? I don't understand. I guess when you have made as many blog posts as he has you start treating them with the casualness I treat my comments.
When somebody attacks things that Scott considers valuable/important he sadly loses his usual dispassionate rigor. It still makes for entertaining snarky reads, so I'm not complaining. Nobody's perfect, and he's far better than most.
I don't think Scott is being excessively casual. The fact is that while Caplan does acknowledge that there are some "genuine" mental disorders (in his thinking, one that impose genuine constraints), he sees these as very unusual and niche situations. Saying "Caplan thinks all mental illness is voluntary preferences" is kind of like saying "The earth is round". It's not exactly true, but it's true enough for all practical purposes.
No, Caplan doesn't think that literally all mental illness is unusual preferences. But he does think that's true of virtually any example of mental illness that you care to think of.
That really doesn't seem to be what Caplan is saying at all. You are going out of your way to make excuses for Scott here.
From Caplan's rebuttal:
"Thus, while I’m open to the possibility that every single alleged mental illness is non-existent, I’m not convinced that this is so."
https://betonit.substack.com/p/scott-alexander-on-mental-illness-a-belated-reply
Explicitly, Caplan is saying that while he is leaving open the possibility that some mental illnesses are real, he cannot identify one. This is not a limited, modest position only applying to to dubious ADHD claims. This is an expansive position applying to every mental illness he can think of - the only modesty or restraint here is that he acknowledges that there might be some case or situation that is escaping his awareness.
I am not making excuses for Scott. I have no problem disagreeing with Scott and do so quite often. But you are misreading Caplan and imagining him to be saying something saner and more reasonable than he actually is saying.
Caplan doesn't argue that *all* mental illness is voluntary preferences, but he sure does argue that a whole lot of it is:
"Most glaringly, a large fraction of what is called mental illness is nothing other than unusual
preferences – fully compatible with basic consumer theory."
And he explicitly expands his analysis beyond things like ADHD to include things like schizophrenia as one example of a not-mental-illness, and even hallucination (he acknowledges that if you really do see things that aren't there that's a genuine constraint, but goes on to argue that most claims of hallucination are false).
https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf
Thinking further, the one thing I think is missing from your argument is a discussion of involuntary cognitive distortions that come along with even the "milder" mental illnesses like anxiety and depression. It's not just that part of being socially anxious is that you feel affectively worse about being in crowds of people. The disorder also causes you to make systematically bad predictions about how you'll react to social situations, how given activities will feel, etc. You become irrationally pessimistic in your beliefs.
This sort of cognitive distortion is clearly involuntary--we might choose our actions, but we don't choose our thoughts/beliefs. It can be counteracted with higher-order cognitive tools. But this is another respect in which mental illnesses can't be pigeonholed as "just about preferences"
See: https://www.youtube.com/watch?v=lGZCZUUtLGQ&ab_channel=PennMindCORE - "All Mental Disorders Involve Systematic Misrepresentation of Value: Mental Illness from a Neuroeconomic Point of View"
My comments below agree with yours. I think the perception distortion we see in many mental disorders is key to loss of capacity. Interventions solely at the level of behavior are limited in their efficacy. Drugs and psychotherapy are both aimed at changing perception which then makes it much much easier to change behavior.
Nice, I didn't see your points. Good stuff.
Isn't the same thing true of other preferences? E.g. people who strongly prefer one political team are often irrationally pessimistic in their beliefs regarding the outcome if the other team wins the election.
I'd like to provide a small possible link between willpower and the preference/constraint distinction.
Preference might tentatively be thought as that which one may change under their own willpower and practice. For example, there are foods I used to dislike, then I learned to like by forcing myself to eat them until I got used to them, and afterwards became so cognizant of, they acquired preferences about which variety is like more than others.
And a constraint may be tentatively thought as that which one cannot change under one's own willpower, though it may be changed, in some cases, via external interventions, such as medications, surgeries etc., if available.
And then the overlap may similarly be linked to cases that require both willpower and external interventions.
This would turn the dichotomy into a false one that's in fact a continuum of that which can and cannot be effected via how much willpower.
Sidenote: the word that's usually translated as "desire" in Buddhist sources, "tanha", is better translated as "craving". It doesn't include small desires, such as Coke over Pepsi, or the desire for Enlightenment, but only those things that cause compulsions, and whose absence would lead the person to experience "dukkha". "Dukkha" in turn is usually translated as "suffering", but that's also inaccurate. It's a broader category that includes from dissatisfactions all the way to literal suffering.
Therefore, the lesson isn't that having desires of any kind lead to suffering, but that being overwhelmed by cravings leads to a continuous state of dissatisfaction that can grow into full suffering, which in turn is a major roadblock for the pursuit of that clarity of mind required for achieving Enlightenment. Hence overcoming cravings, addictions, and the like, is a necessary step on the path.
As someone who likes ecological psychology, I am wary of putting too much emphasis on internal states.
Credences, for example, can only be defined insofar as they are coherent; if someone makes different bets about the same thing in different situations, for example if the bet is presented differently.
But it's not like preferences are a perfectly defined concept either! They too can only be defined insofar as they are coherent.
Why is Scott engaging with this? As far as I can tell, Kaplan has identified a theory that the majority of people believe to be true, determined that Scott is somehow exceptional in being a proponent of this theory, and then repeatedly challenged Scott with inane arguments against the theory without actually considering Scott's defense.
I mean, I can do that too. Scott thinks we shouldn't eat babies, but either babies are more delicious than adults, or adults are more delicious than babies. So Scott must concede that if we shouldn't eat babies, we should eat adults. This isn't something Scott is willing to concede, so baby-eating is morally permissible. I'm going to take Scott's failure to argue about this as evidence that he has no counter-argument, and I'm also going to take any counter-argument Scott might make as evidence that he has no counter-argument.
Babies are delicious, but I couldn't eat a whole one. Not in one go. Maybe I should portion out the baby meat so it can be used to make several meals. What are some of the recommendations I've heard of about that?
"I have been assured by a very knowing American of my acquaintance in London, that a young healthy child well nursed, is, at a year old, a most delicious nourishing and wholesome food, whether stewed, roasted, baked, or boiled; and I make no doubt that it will equally serve in a fricasee, or a ragoust.
... A child will make two dishes at an entertainment for friends, and when the family dines alone, the fore or hind quarter will make a reasonable dish, and seasoned with a little pepper or salt, will be very good boiled on the fourth day, especially in winter."
Scott is the one who started this exchange. It began with a paper that Caplan wrote in 2006 laying out his position. Scott read it and wrote a post about why it was wrong nine years later. It's been back and forth between them since then.
Well then, I stand corrected. I admit that I do find these essays repetitive but it was unfair for me to blame Caplan.
What's the value of even paying attention to Bryan Caplan at this point? It seems mostly like a waste of everybody's time.
Is he not representing a common view of mental illness from a libertarian view? (I might be really wrong about that -- the politics side of this isn't my area)
Kinda, but this sounds is more the economist's view than a libertarian's, insofar as there is much of a difference, as interpreted by a psychiatrist and translated for a rationalist audience. How much is preserved through this game of telephone is dubious.
Caplan is a very smart, intellectually honest guy and his opinions are worth paying attention to IMO. Even when they are wildly incorrect.
Edited to add: Strike that - reverse it! I had only read the beginning of Caplan's article and assumed it was representative of the rest. Now that I've read more, I am 100% team Scott.
Caplan starts out saying that you can model someone with mental illness as a rational actor with unusual preferences - that is correct. Then he moves into endorsing Szasz and going after psychiatry, and IMHO ends up someplace crazy.
You can model someone who is an alcoholic or has ADHD as just having unusual preferences, but when you say they shouldn't receive psychiatric treatment, as Caplan seems to, you're ignoring that a lot of people with alcoholism or ADHD want to change, in the sense that they voluntarily undergo treatment in an effort to change, they sign up for Beeminder, etc.
Now whether psychiatry is effective at introducing change for a particular condition is an empirical question, but Caplan isn't attacking psychiatry by arguing it's ineffective, he's just arguing that it's inappropriate to change people's "preferences," even voluntarily. That seems crazy.
-------- Old Post ----
I went back and skimmed Caplan's original article, and I'm now more sympathetic to Caplan than I was before, unless the exchanges since then have materially changed the discussion.
https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf
The question Caplan originally was trying to address was whether classical economic modeling can apply to the decision making progress of a mentally ill person - not whether an illness should be covered by health insurance or whether we should call it an illness in everyday speech, etc.
Under this framework, a person has priorities and resources, and they spend their resources based on their priorities and their beliefs about the world.
Some examples:
1) If my leg is trapped under a rock and I have the choice of sawing my leg off or staying where I am and hoping help finds me, I won't like that choice at all, but I will make that choice based on my preferences and my beliefs. If you know my preferences and beliefs, you can model my behavior.
2) If I mistakenly believe that my family is spying on me in an effort to prevent me from attending business school, you can model my choices based on those preferences, even if the belief itself is based on a mental disorder.
3) If cilantro tastes like soap to me and I don't like the taste of soap, you can model my choices based on that - I'll pay less for a burrito with cilantro is it, but there might be a sale price where I'll eat the cilantro burrito. This model works whether my dislike of cilantro is genetic or based on some other cause.
The ultimate question is whether the economic preference and budgeting models apply to the mentally ill, not something else. On that front, I wonder if Caplan and Alexander have a misunderstanding, not a disagreement.
"The ultimate question is whether the economic preference and budgeting models apply to the mentally ill, not something else"
Thing is though, don't we base (economic) choices on a rational actor? And when your decision-making apparatus/reason is impaired, then the choices downstream of that are not so predictable.
If I think my family are spying on me, what am I likely to do? Predicting my course of action by modelling my choices is going to differ whether "well Deiseach is not crazy so she's not going to burn the house down with them inside it" or not.
As I understand it, a "rational actor" just means you make choices rationally given your preferences. It grants that someone's preferences can be very odd.
IMHO, there's nothing wrong with the model, but some people might not find it very useful.
I've definitely had experiences where I was so angry or so sad that I made choices that I later found irrational from my normal viewpoint. You can model that as I had an unusual preference in that moment that valued telling the other driver what I thought of them more than it valued not getting my car headlights smashed out, or you can model it as me being so angry that I was "irrational." Either model works, so I can't get upset if someone prefers one or the other.
Sorry, I thought I had skimmed Caplan sufficiently, but I was not even close, and I completely misunderstood him. I've edited my original post.
As someone who's been on and off stimulant medications for ADHD at various times I've thought a lot about where boundary between "wanting something but not being able to bring myself to do it" and "not actually wanting something" lies. On stimulant medications I'm much more inclined to classify "things I can't do" into the second bucket rather than the first.
I think this is because on stimulants there's a much shorter path between vaguely thinking about doing a thing, and your brain/body actually doing it. This results in there being a much smaller pool of things that you vaguely want to do but just can't seem to make happen, and therefore you're more likely to classify these things as "not actually wanting them".
Parkinson's is probably the most direct way of exploring this connection. The neurotransmitter that turns thought into movement is literally missing. Does this result in a subjective experience of feeling paralyzed, or of feeling like you "want" to move less?
Caplan is applying the "revealed preferences" principle in economics too broadly. Revealed preferences is a term that means that we should look at the behavior of individuals to determine demand, not what they say in surveys or what is assumed by policy makers. For instance, people may say they don't drink alcohol, but a superficial empirical investigation will reveal that alcohol consumption exceeds that which is predicted by surveys. Drug addicts might not want to continue using drugs, but they do and policy makers should assume that they will continue, even if it means addicts will break laws prohibiting drug use, resulting in a cascade of other negative social outcomes (violence, petty crime, etc.).
Revealed preferences is a useful principle in predicting economic behavior in aggregate. Economists are not really concerned with the internal state of individuals, because it doesn't really predict policy outcomes.
The term is not useful in evaluating the internal state of individuals, or what actually drives them to do what they do. That's the realm of psychology.
Caplan is too smart to interpret the term too literally, in that I doubt he really believes that a schizophrenic is just choosing to hear voices because that is what he prefers. Instead, I think Caplan has another model of behavior in mind that he is not articulating well. Without putting words into his mouth, I would suspect he has not integrated mental illness concepts particularly well into his overall thought process, given his training in economics. Also, embracing nonvoluntary mental illness may undermine his commitment to libertarianism, certain versions of which emphasize the voluntary desired nature of individual behavior.
Is mental illness any less "nonvoluntary" than any other mental process?
My thought is yes and no -- that the same mental processes inside two different people can be differently "sticky" and sticky may speak to how much voluntary influence a person can exert over that process. It's common with respect to OCD and ruminative thinking to say we may not control the arising of thoughts but we have some influence over how much we keep directing our attention to particular ones. How sticky some thoughts are though is the result of the usual mix of genes and environment, and to some extent the learned skills of the thinker (and how well they can learn the skills is also a result of a mix of genes and environment).
For someone with OCD, some of their mental processes may indeed be more nonvoluntary than someone else's mental processes around the same content. They may have to exert more conscious effort to not get stuck on the sticky thoughts. Maybe someone with ADHD has the opposite problem with respect to some thought arenas, where their thoughts aren't sticky enough and slide too easily out of awareness.
This connects for me back to that canalization paper Scott shared. It speaks to a quality of fixatedness when it comes to certain thought arenas. People in manic episodes can get very fixated on thoughts, as well as people with paranoid delusions, phobias, etc. People with severe depression can get very fixated on how worthless they think they are. How subject to change all that is has more to do with severity of the disorder, degree of psychological flexibility the person has overall, or the level of stress the person is under than the type of mental process they're in.
This seems to be a conflict between mind–body dualism and mental-illness–mind dualism. If you reject both, this entire debate seems to be ill-posed.
From a practical standpoint, isn’t it more objective to cure a physical illness than an internal state?
The policy implications of this discussion sorta rest on the tangible nature of influencing physical versus mental states. I think Caplan’s metapoint is that you can disincentivize or incentivize internal states (preferences) and the preferences which cause illness.
A minor point about cancer:
It's legitimate to debate at what point cancer is a 'disease' and also at what point that disease should be treated. Many people die in their old age from some other condition, yet they had a benign tumor somewhere in their body. It really matters whether a cancerous growth is either (a) malignant, or (b) likely to become malignant before you die.
Does this ever matter in the real world? Yes. Back in the day, they used to use PSA alone to determine whether you likely had (a) or (b). Too high a PSA, and they'd resect your prostate. Sure, you'd struggle with incontinence for the rest of your life, but then you could celebrate being a 'cancer survivor'! Except later studies suggested there was an unacceptably high false-positive rate with this test alone, and that hundreds of thousands of men got their prostates out for no good reason.
Yes, cancer is bad and should be treated. But we should also be paying attention whenever we set out on a medical intervention. ESPECIALLY preventative treatments. Those approaches have the highest probability of medical reversal later, when we realize it was all harm for little to no benefit.
I think any conceptual categorizations beyond a purely descriptive one on the basis of atom configurations and observed behaviors places a burden of proof on the part of the part of the categorizer to explain why their classification scheme does useful work. If you steelmanned Bryan's position, what work is the constraints/preference classification doing? It sounds like it's just trying to do political prescriptive work in terms of telling us how to treat the mentally ill. I would like to see an argument based on data from him that treating the mentally ill as if they have a preference rather than a constraint did things like improve recovery rates or speed, or reduce their burden on society, or have some other effect that he might be able to pitch as positive. If the classification doesn't do work then throw it out, since all models are only as good as the work they do.
I support this!
Th grain of truth in what Caplan is saying is that people with mental illnesses often feel as though the doctors categorizing them as a whatever -- borderline, bulemic, depressive, obsessive, etc -- are sort of denying that they are valid people. Their pain, their memories, their opinions, their sense of humor all sort of do not count, in the minds of the people diagnosing them. And there is a fair amount of truth in that. I used to work on an inpatient unit, and when an upcoming admission was announced, the person was generally identified by their diagnosis: "We're getting another eating disorder later today, she's 19, bulemic, has a very significant abuse history." Doctors hearing about criticisms patients had made of them often rolled their eyes, as though to say, "c'mon, consider the source." Awful, right? And yet, viewed from the outside, people with these disorders do often seem very controlled by their disorders, to the point where you feel as though all the rest of then is hidden and controlled by it.
Are there any therapies that encourage patients to view their mental illnesses as a set of simultaneous constraints and preferences? It's a low bar but I could see that benefitting some people as a way to model their internal states. Would be interesting to try, and surely better than internal family conversations or one of those other weird therapy systems.
There probably are, but I don't know their names. But I think I, and probably a lot of other people who do cognitive-behavioral treatment, think about things in a way that's pretty compatible with that model. For instance it's very helpful to depressed people to recognize that the reason they aren't doing anything is that all their thoughts and mental movies depict every action they might take as pointless, boring and extremely unpleasant to do. And while I don't use the word constraint, I do say that their mental state makes it impossible to believe that taking a walk or whatever might leave them feeling a bit better. So then then they might say something like "even if that's the explanation, there's nothing I can do about it. Everybody keeps telling me to think more positively and I can't." So I ask them whether taking a walk (or whatever) is impossible, or merely looks very unpleasant and pointless, and work on getting them to do it as an experiment. Most people acknowledge that it's not impossible to do. So our model is that your thoughts about action are truly constrained, but your aversion to action is a preference. Anxiety disorder work has an analogous model.
Things like internal family conversations are silly as hell if the therapist is convinced that they are the truth, the whole truth, and the explanation for every problem people have. But for some people an internalized harsh parent really is the problem, and understanding that is very helpful.
This fascinating Invisibilia episode profiles an interesting case where a chronic “augmented pain” syndrome is worsened by paying attention to the pain, and treated by ignoring and therefore desensitization.
https://www.npr.org/programs/invisibilia/701219878/the-fifth-vital-sign
IMO this complicates even a simple continuum of preference-constraint.
There's a great discussion about changing preferences from EconTalk a few years back: https://www.econtalk.org/l-a-paul-on-vampires-life-choices-and-transformation/ (Yes, it includes a serious discussion on vampires.)
Sometimes your preferences change because your circumstances change. It's difficult to say that just because you have a certain preference in the moment that this is an enduring preference, or even to predict what your preference would be when those circumstances change.
Just one more reason that the whole idea of 'preferences' is not practical to drive policy.
I just want to say that I once held my breath until I blacked out. It's a mildly unusual thing that is tangentially related to this article. I was motivated by a competitive spirit and desire to win.
Would Scott *prefer* that Caplan respond to these four points?
Or is he suffering from a terrible constraint where he suffers unless Caplan actually responds to the content of his argument?
It seems to me this might be the real question.
On terminally ill people "holding on" (note 3) this has been studied at a population level.
In countries where Christmas is important, there is an increase in deaths from natural causes in the week immediately following it, with a corresponding dip the preceding week.
Two UCSD sociologists, David Phillips and Elliot King, did a similar study with Passover which is published in the Lancet:
https://www.sciencedirect.com/science/article/abs/pii/S0140673688901985
They compared Jews and non-Jews in the same place, which provides a handy control group for factors like the healthcare system not functioning as well the week after a major holiday.
A likely explanation is simply that people "hold on" to reach a major annual family-gathering holiday.
Is there an implicit assumption (or explicit somewhere I have forgotten) that doctors may/should never prescribe treatments, especially pharmaceuticals, for a condition we categorize as a preference rather than a disease? Is this a legal constraint, or medical ethics, or what? Does Caplan take a stance on this?
This leads to the issue of the ethics of self-medication, where presumably Caplan, as a libertarian, sees no objection to any adult treating themselves however they wish, or hiring whoever’s they wish to advise them. I am not sure what Scott's position is. From this perspective, the legal restrictions on prescription are the main sticking point, since whether a condition is a disease or a preference, in a world without legal restrictions on prescription drugs, competent adults in either category could try to improve their lives using whatever treatment they believed would help.
The other issue is under what circumstances bystanders are justified in forcing a treatment upon an unwilling patient (or similarly, to forcefully prevent them from applying a treatment). The current dominant opinion among medical ethicists is that treatment should be voluntary for competent adults. So the interesting question becomes, who counts as a competent adult, and how may we treat infants and invalids? (Of course, existing laws pretty much ignore this analysis.)
It seems to me that the categorization of conditions as diseases or preferences has little relevance for these issues. What is the issue where this controversy does have relevance?
Insurance companies use the language of "medical necessity" to determine whether a medication or procedure will be covered. Doctors' licenses to some degree are at risk if they are seen to be prescribing or procedure-ing outside of medical necessity. Some parts of medicine get to be outside of this, like elective cosmetic surgery or orthodontics.
I don't believe any part of mental health treatment within the medical model gets to operate in this elective surgery way. Unlike a nose job, you don't get to pay a psychiatrist to prescribe ritalin just because you'd like to have it. Though of course some of that does go on anyway.
In the land of talk therapy, you can go talk to a psychotherapist (in the US anyway) and pay out of pocket and do that without a diagnostic paper trail. Some psychedelic-assisted psychotherapy also operates in this less regulated space, though at some risk to the providers' licenses. The current growth of at-home ketamine treatment is also pushing the bounds in a way that can look more like the unregulated wild west.
I meant to say, I like your question and think it's worth asking anytime one's about to get into a big debate: "What is the issue where this controversy does have relevance?"
+++++
That makes sense.
So someone might not be able to get a prescription, or might not be able to get insurance to pay for it, depending on the result of this debate. I guess if we take those institutional constraints as given, that gives this terminological debate some relevance.
But I would rather criticize the institutional constraints that prevent good outcomes (while treating patients paternalistically).
Though the people who have power in this situation -- healthcare systems, insurance companies, pharmaceutical companies, and the politicians they support -- are unlikely I think to be swayed by libertarian arguments to radically shrink the definition of mental illness and who qualifies for treatment.
Economic precarity also plays a role by feeding parental anxiety which puts pressure on schools and doctors to intervene hard and fast on kids who aren't measuring up. There's so much more measuring that goes on these days than when I was a kid (70s and 80s).
Which is just to say I don't think there's much at stake in this debate and that it's more circling around definitions -- what's a preference and what's a constraint and does holding a gun to someone's head actually reveal anything useful about those terms?
If Caplan came out and said "I think we ought to stop covering any kind of mental health treatment under insurance plans except for people with severe schizophrenia or bipolar disorder, and everyone else ought to be shamed for their poor choices and told to buck the hell up," then Scott might have had something more to work with. As it is, we're parsing definitions and keeping our actual opinions mostly to ourselves.
I'd love to talk more about what we consider good outcomes and what gets in the way of those.
+++++
Would Caplan say that? Only if he buys into the presumption that only doctors should prescribe treatments, and then only if they have diagnosed a specific (non-preference) disease (and insurance should only cover disease).
Oh I don't know what Caplan would say, which is part of the frustration with this whole non-debate debate.
I found the piece of Caplan's that Scott was most directly responding to to be somewhat incoherent. Caplan doesn't show much understanding of mental illness, like at any level. He doesn't support with any kind of specifics his general statement that a lot of currently defined mental illnesses are voluntary preferences.
His argument seems a bit all over the map and offering no support other than clever rhetorical moves by quoting Scott out of context. He doesn't explain what supports his choice to use a gun-to-the-head standard to define voluntary preference. There's more missing from his argument than is there I would say. Which leaves him open it seems to me to speculating what he's actually up to, what's motivating whatever this ongoing conversation is. When someone makes an argument but doesn't support it, but then keeps hammering at it across multiple rounds, it suggests there are other motives at play that are being satisfied by keeping it going without taking it any deeper.
I haven’t been following the debate from the beginning. I just today scanned the original journal article by Caplan, which QuintusQuark kindly provided a link to: https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf. I was skimming to try to find out what Caplan thinks is at stake, which I didn’t find. A bad summary might be, he felt he had an opportunity to show how economic analysis applied in an area where no previous economist had thought to apply it. If he thinks there are practical implications of his insight, I missed them.
He probably would while supporting people’s right to pay for their own treatment. Insurance covering treatments sounds like a special privilege for extreme preferences he thinks would discourage people from bucking up. (Look at page 361 here: https://econfaculty.gmu.edu/bcaplan/pdfs/szasz.pdf)
I was just looking at that page, but I’m not sure what you mean. Maybe you mean, if I have insurance against x, I will not pay the full social cost of x? Sort of a moral hazard thing? That's still a long way from “we ought to stop covering any kind of mental health treatment under insurance plans except for people with severe schizophrenia or bipolar disorder, and everyone else ought to be shamed for their poor choices and told to buck the hell up.” What part of his argument implies that no one should sell or buy insurance against bad preferences, or that shaming would be an effective alternative?
From Categories Made for Man: "Hume’s ethics restrict “bad” to an instrumental criticism – you can condemn something as a bad way to achieve a certain goal, but not as morally bad independent of what the goal is. In the same way, borders can be bad at fulfilling your goals in drawing them, but not bad in an absolute sense or factually incorrect. Namibia’s border is bad from the perspective of Germans who want access to the Indian Ocean. But it’s excellent from the perspective of Englishmen who want to watch Germans plummet into the Lower Zambezi and get eaten by hippos."
I suspect Caplan wants to discuss (and has discussed https://betonit.substack.com/p/lgbt-explosion ) empirical questions of " (1) Are increasing numbers of people falling into mental health/sexual/etc. categories through marketing and self-selection, and (2) is this driven by discussing the categories themselves, (3) thus accruing outcomes XYZ (which are somewhat no-benefit with some cost, or sub-optimal)." This is a very common question, something that doesn't necessarily interact with the inherent political nature of categories so much as whether those politics are good or bad in the goal-fulfillment sense.
Bryan continually names Scott, because he likes him! He wants his favorite psychiatrist to say "I estimate that x% of diagnoses are due to self-selection/marketing/discussion of the diagnosis, and we would be better off limiting discussions of XYZ to prevent social/memetic contagion towards sub-optimal outcomes." Scott could pick any number for x, that'd be an entertaining post, but wildly more political (is this good/bad) than Scott normally does. I think Bryan is an excited and friendly-combative debater who refuses to ask that question, in adherence to the inscrutable norms of online-debate etiquette, and doesn't want to draw a crowd that wouldn't otherwise sit for a discussion of category definition. Caplan, and most of the GMU gang, ran into the same problem while discussing feminism a while back - just a bunch of posts discussing "What does feminism mean" while all dancing around having a stance on whether feminism-as-I-define-it-in-this-post (rather than naming policies XYZ) is good or bad.
Well, there is a qualitative difference between a person who feels excruciating pain when they walk, and a person who has had all the nerves in their legs severed. I'm... just not sure if that qualitative difference is particularly useful for reasoning about anything (except maybe figuring out what medical treatment to apply to get that person walking).
I haven't gone to read what Caplan has written in the past (maybe I should), but in my experience the line of reasoning presented as his in this post is essentially never put forward by someone who has a nuanced contribution to make to philosophy, or who wants academics in medicine or elsewhere to have a more expressive taxonomy available for the conditions of homo sapiens. In my experience it is always, and only, *that they're intent on proposing a way that communities of people can reduce their support and respect for some group of people they've identified as deserving cruelty*.
You’re right about Caplan. His 2006 article calls psychiatrists rent-seekers and seems to recommend that mentally ill people shouldn’t get any kind of accommodations because they’ve chosen to be mentally ill.
The obvious compromise solution is for both Scott and Caplan to accept that all preferences are in fact mental illnesses.
You could say that the only people who are truly mentally ill are people who are completely not viable. Meaning, whatever mental adjustments they are making will not allow them to survive in the world. that of course, is very cultural and environmental. Generally speaking in western culture today there is a lot of outreach towards people who are challenged in this way, and there is much more possibility for their viability. As a result, more people become viable. One could imagine a very different culture that has much less interest or means in reaching out. You might even care to call that political. A village can tolerate an idiot, but a village cannot tolerate hundreds of them just to make a crude example.
You could say mental illness is doing anything that doesn’t work.
I’m a psychologist and a therapist, and agree with everyone who’s saying there really isn’t a clear distinction between constraint and preference (or you could call it a distinction between can’t and won’t). So here's a list of :
*7 Denizens of the Constraint-Preference Murk Zone*
(1) Anxiety disorders: expectation of terrible suffering if you have a close encounter with the feared thing.
Ex: I had a patient who feared broken glass as much as I fear jumping off a 100 foot cliff. I think if somebody put a gun to my head and told me to jump, I probably would not, because my picture of dying via the fall is more vivid and terrible than my picture of a bullet to the head. (Yet when I finally convinced my patient to handle some broken glass, her worst moments of fear were in the 3 seconds before her hand touched the first piece of glass. Once she started it wasn’t so bad, and by piece 10 she was hardly scared at all.)
(2) Procrastination: A habit of asking oneself before doing something whether one is up to doing it, whether there’s chance the in the future that the task will look more attractive, and whether it’s possible that doing it now might be a mistake. (If most of us asked ourselves these questions before we did something we’d never do anything). I see this habit a lot in people who are Aspergerish, and the habit is often coupled with a felt certainty that it is important to ask these questions, otherwise life will be too unpredictable and get out of control. Not asking these questions feels to them like I feel about jumping off a cliff.
(3) Addiction: People have addiction-related “delusions” about the consequences of giving up the substance you’re addicted to. I had these about nicotine. Lighting up had gotten so deeply associated for me with reading, writing, thinking and talking to people that I more than half believed that I would not be able to do any of these things nearly as well without nicotine. What made it hard to stop smoking wasn’t inability to face cravings, it was dread of becoming unable to do the activities that were most important to me. (As Stephen KIng said after recovery from addiction to alcohol and cocaine, I found to my surprise that my mind worked fine without the drug.)
(4) Depression: It’s not exactly inability to act, it’s inability to have any positive expectations about acting, and of course that produces failure to act. If the mind had suits, the way a deck of cards does, and there was a fearful suit, a sad suit, an excited suit and a hopeful suit, depression would be playing with a deck where the last 2 suits are missing.
(5) Depression: Inability to hear encouragement from others as anything except invalidation and criticism. Consequently, something that’s often pretty available to unhappy people is unhelpful and toxic to the depressed person.
(6) Avoidant and dependent take on life: People with any of the above who are also seriously dependent or avoidant have an extra constraint/preference: They believe that if they overcome the condition that has them stuck they will lose their status as a beloved and sheltered person, and be cast out into the lonesome cold.
It seems to me that it’s too simple to think of there being a spectrum running from constraint to preference, with lots of things in between. You can see in the above examples how people can be constrained by the false expectations ofwhat would happen if they did or didn’t do something — or by the sheer inability to conceive of certain things.
Yes, yes. See my post about braces applied to limbs at childhood. In my experience, depression is a very much internalized narrative about outcome that trumps the immediate circumstances of one’s life. It is an internalized memory that controls the present. And it is not just mental, it is physical. It is the feeling of being defeated.
"Depression: Inability to hear encouragement from others as anything except invalidation and criticism."
Oh, yes. If someone criticises me, I believe it because sure, it must be true, they're seeing the real me, they can see the faults and lacks and general awfulness. If someone praises me, I can't believe it, because they're only saying that to be nice, or they have no idea what I'm really like and if they knew they wouldn't say that nice thing.
So praise can be *worse* than criticism, because criticism is only saying what I already say to myself, I know it, I'm comfortable with it. But praise? What do I say there? "You're an idiot who has no idea, what you just said is stupid and dumb and false"? You can't say that to people who are just trying to be nice to you, so you shut up and smile and have the thorns of praise prick you to the bone.
Oh this is all quite lovely, thank you for writing it!
This has turned out to be true for so many people I've worked with and for me as well: "her worst moments of fear were in the 3 seconds before her hand touched the first piece of glass." It's really hard to learn this, though I think if a person gets a somewhat profound chance to learn it, it maybe transfers pretty well to other situations. It has for me anyway. I think the learning that sticks may have to happen on the really scary stuff though.
Also, I love this that you say about procrastination: "I see this habit a lot in people who are Aspergerish, and the habit is often coupled with a felt certainty that it is important to ask these questions, otherwise life will be too unpredictable and get out of control." That's so well said.
I've been circling around a similar idea having to do with anxiety/OCD/rumination producing a kind of habit of mental "due diligence" which I think is the same thing you're talking about here maybe. I see it in a lot of people who aren't Aspergerish but are more your average, bright neurotic. And I agree that it's an effort to bring into control things that are in fact not in one's control via a kind of relentless mental scanning effort, "what am I missing? what am I missing?" that has a stove knob checking quality about it. It leads to a kind of default place of self-doubt that undermines a person's ability to trust their own experience (which impedes learning). The relentless self-scrutiny has an OCD feel to me and I hadn't thought of that in connection with procrastination which I've mainly thought of as a form of emotional avoidance that's part of performance anxiety. But it seems like the emotional avoidance is being sustained by a lot of active rumination of the kind you describe as opposed to just by binging Netflix (which is what the emotional avoidance frame brings to mind).
Moving an experience from constraint to something a person can exert some influence over (I don't like the preference language at all) is a learning process and no one can predict ahead of time how much progress any one person can make along that road because it depends on so many variables that are not in their control. People can bring willingness to try things, but the outcomes are not in anyone's control, something which seems to elude Caplan's tidy gun-to-the-head framework.
It’s kind of interesting to break the word disease down into its components.
Dis-ease.
A lack of ease.. that is a big umbrella.
Mental illness is a big umbrella.
My problem with this conversation is the generalities implicit in the vocabulary.
Let’s take two things; liver cancer, and diabetes. They are both diseases.
Diabetes has been narrowed down to a very particular chemical deficiency that can be remedied.
Liver cancer on the other hand, as far as I know, is way more wacky than that.
One form of Mental illness is tricky, because as anyone who’s ever dealt with it knows, one’s own participation in its cure (or management? )is essential. See the classic joke about how many psychiatrists it takes to change a lightbulb.
But
The brain is also a physical organ and can malfunction. Godspeed to the people who are trying to crack that nut… I have to believe that some of the pharmaceutical interventions developed over the last 40 years have actual relevance in that world. I am not at all in a position to speak to it. I am more interested in what I consider the other half of mental illness..
What ever happened to the word neurotic? It really seems to of fallen off the bus. I never see it in any of these discussions anymore. Perhaps it is a concept that is extremely out of favor. I happen to think it’s an important distinction.
A disease if you will, of emotional dysregulation? Is this a failure of will or a condition to be treated as a disease, in the classic sense?
What if someone was born and had a brace put on one of their legs, a brace that dictated the growth and shape of their limb over a period of time? What if eventually that brace was removed? Would the affected limb then spontaneously assume it’s rightful shape?
Or is it more likely that a lot of physiotherapy and conscious exercise would be necessary to even begin to restore it to what it should’ve been?
There are lots of people who I believe are mentally ill in the most fundamental sense of that expression; meaning, they have very little volition to bring to the task of understanding who they are. And there are others, who I would refer to as neurotic, that given the opportunity can direct their attention towards their own management or recovery. It’s very difficult to find a border between those two things.
So finally, is this a philosophical discussion or a discussion about social policy?
I would be very interested in some granular information about what it is to pay attention in some technical sense. The ability to direct your self towards paying attention to something. I think it’s a very significant thing, this ability to redirect one’s attention.
Why do you care a 'rat's ass' what Caplan thinks? Seems like it would be better for you to ignore him.
(But I have no idea, so feel free to ignore this.)
https://www.productgems.io/blog/tag/mark-twain/
Is the boy crazy?
Caplan skimmed the DSM and now he thinks he’s a psych expert. I took a look at his 2006 article that started all this. He linked to it this month, so he still endorses it.
His section on ADHD is based on a shallow reading of the DSM diagnostic criteria with almost no context. He uses this to suggest economically modeling ADHD as “a high disutility of work combined with a strong taste for variety.” The first part doesn’t make sense. The DSM-listed symptom regarding avoiding tasks requiring sustained mental effort can apply to work, but it can also inhibit people from returning to books they were enjoying or even watching TV episodes with a lot of hard-to-process stimuli.
He also claims the forgetfulness symptom refers to people with ADHD “conveniently forgetting to do things they would rather avoid.” But ADHD forgetfulness also includes things like, I don’t know, forgetting to eat dinner and drink water because of reading an article by someone too lazy to do thorough research on ADHD. Would this be modeled as a high disutility of eating and drinking?
The DSM criteria also don’t include commonly reported symptoms like rejection sensitivity or hyperfixations. These would indicate that ADHD isn’t “medicalized laziness” but a distinct type of brain with characteristics that tend to come together for a non-obvious reason.
I learned the hard way that preferences *are* constraints.
I was in a community that put very strong social pressure on me to disavow and avoid satisfying many of my preferences. I tried very hard to conform, and to tell myself "they're just preferences, they aren't real tangible things, I can just not listen to them."
This seems to me (with the usual caveat that it's very difficult to have much certainty about the causes of one's own internal states, because n=1 and it's all being seen from the inside) to have caused me to enter a very bad mental state that has lasted 6+ years after I both left that community and recognized the problem and started making serious efforts to improve my mental health.
"they're just preferences, they're not real" transitioned smoothly into "I'm not real, I don't exist" which caused a lot of dissociation and loss of executive function.
I am gradually getting over this, but it's ongoing, and I'm not fully convinced I'll ever feel as real and as sure of myself as I did before.
(ETA: I do recognize that my desire to be accepted by the community was also "just a preference", but I'm not sure why I accepted that one as more real than the ones it was trading off against.)
Need to write a large response here later today, as i have a lot of thoughts on this. Ive also had email conversations with bryan caplan, that i need to look through
Hmmmmmm
> Call it unibating, or monobating, or another word ending in -bating which is less polite but as far as I can tell equally appropriate.
I would use an homophone: baiting.
Caplan had a point about constraints, but was wrong about constraints and preferences being opposite.
The "if you can do it with a gun to your head, it's not a constraint" example is horrible. Lots of people have something like arthritis where some actions are impossible and some actions are entirely possible but incredibly painful. That situation is a restriction on your life by any sensible colloquial, moral or legal standard, and correspondingly is a disease.
I think the distinction he was looking for was, if we think a restriction is an IMPORTANT restriction, we're likely to think of the cause as a disease rather than not. That's not universal, the restriction will be important in some situations and not in others, depending on our judgement of the person, the judgement of people around them, and the amount society is accessible to people with that limitation.
I haven't read what Caplan wrote, so I'm probably missing the good bits, but I suspected it was a combination of:
1. He says there's not a neat objective criteria for what is a disease and what isn't. (Scott agrees, I agree) And he thought that Scott thought that WAS an objective criteria (he was wrong, that's why he thought Scott changed his mind)
2. He thinks that he has a better definition with constraint vs preference (Scott disagrees repeatedly, I mostly disagree as described at the start of this comment). I suspect that he saw Scott disagreeing with his particular definition as if Scott disagreed with point #1.
3. He thinks that modern categorisation of mental illnesses include too many things that shouldn't be included, because of these disagreements. I think that that HAS BEEN true and is true of SOME things, but that it's much less true now. I suspect if he talks about specifics, he'd talk about lots of things which psychiatrists think are illnesses which he doesn't, and I would probably agree with which are which, but that he thinks that the false-diagnoses are much more prevalent than I do.
On the last point, I have ADHD. I know first, second and third hand that there are people with a real, treatable disease which manifests in a variety of ways which are very real but hard to measure from the outside (e.g. inability to concentrate, inability to start doing important tasks. And yes, most people can do them for a few minutes if they have a gun to their head, but can't KEEP doing them every day without an immediate threat to their existence, and can't be threatened into doing them without long term damage).
I also know that there are people who suffer unable to fit into a regimented school routine, but flourish perfectly well in jobs which suit them (which jobs vary, but programming, EMT, crafter, financial trader, etc depending on their combination of needs).
I also know that there are SOME people who have adhd-like problems in life and seek out or are forced into treatment or into attitudes about it which are counterproductive in helping them. I think that's something to be aware of, especially if people are forced into treatment which is bad for them, but I don't know HOW widespread that is, and that doesn't constitute ADHD being fake for everyone else.
While there are conditions associated with intellectual disability that do predispose people to being, crudely, "happy and kind" this is often just a stereotype of people with intellectual disability that is simply not true. While it may seem harmless because it's a positive stereotype, it does tend to promote a kind of dehumanization where people aren't seen as part of the full range of human experience. Down syndrome is an example of this. In fact, people with Down Syndrome have a higher prevalence of depression.
There was an SSC article several years ago (IIRC re autism?) where Scott made the point that the general public sees the "happy" sufferers walking around on the street, holding down jobs, etc. but does not see the "unhappy" ones strapped to their beds in asylums -- even when the latter are far more numerous.
"The far more numerous" part isn't accurate and I think you are misremembering that specific point. That said, I think the article you are referring to was a good corrective to people who don't see the more damaging side of developmental disorders because it is out of their interactions and thus invisible to them. A *lot* of people over the past decade have gotten some distorted ideas about autism spectrum disorders because they're reading the commentary of self-advocates who are talking about their experiences that aren't fully representative.
To the point you're making, I'd go further. Lots of people you see on the street, holding down jobs, etc. may seem "happy" to you when reality is more complex behind closed doors. People can keep it together in superficial public interactions. This is true of people with developmental disabilities as well. It's just a common stereotype that people that read as "simple" to others are happy-go-lucky, and just like everyone else, sometimes this is true, and sometimes it isn't. It's more likely to be untrue than of your average person in the case of Down Syndrome because the condition is more apt to predispose mood disorder as are the still relatively pervasive noxious social conditions people with intellectual disabilities often face. I think part of why people are attracted to this stereotype is because it feels karmic. People will such disabilities are perceived as having gotten a raw deal, and this is imagined to be "made up" for with a pleasant disposition.
I've attended funerals for people with ID/DD where preachers who know very little about the person being eulogized have drunk deeply from this stereotype and give a talk describing people who were not all that happy or nice this way. It's profoundly awkward to listen to hear someone reduced to a stereotype that isn't even ballpark true for someone who nonetheless had traits worth memorializing.
“While the person may choose whether or not to scratch themselves, they cannot choose whether or not to feel the suffering”
It’s interesting to me that you even think it’s a choice to scratch. As an itchy person I can maybe delay it 15 seconds tops, and the whole time I won’t be thinking of anything but “it itches so bad but I’m not gonna scratch it.” If some other thought grabs my attention, it doesn’t mean it distracts me and the itch goes away on its own, it means I impulsively/unconsciously scratch because I’m no longer focusing on not scratching.
Once when I was 13 I was on a Boy Scout canoeing trip and my legs had gotten literally dozens of bug bites. I’d scratched them to the point that I had polka dots of blood all over my pants, and one of my friends asked me why I didn’t “just stop scratching”. I think I just looked at him confusedly like he had two heads, wondering how he even thought that was humanly possible, and immediately recommenced scratching. But I’ve since learned that apparently I’m the weird one? I don’t know, I can’t imagine how people are apparently able to just ignore it.
>I don’t know, I can’t imagine how people are apparently able to just ignore it.
Typically you don't try to ignore it or push it away, that will only make it worse. Instead you're supposed to calmly experience it with equanimity for as long as it wishes to control your attention. That's not easy to do at all, but that's the theory behind a lot of mindfulness training. You train yourself to allow negative thoughts, stimuli, etc. to have their place in your attention sphere while maintaining a separate fuller awareness. Eventually you get better and better at holding an awareness while allowing the negative thought/stimuli to come into attention and fade away on its own. The better you get the less strongly you'll feel them when they do happen, and typically they happen less frequently because you train your mind that you aren't going to respond so there's no point in bringing them into attention.
If you've never attempted such a practice it probably does seem like an impossibility to feel an itch or some other stimuli and not react to it.
Though there probably are some people that naturally have less sensitivity to various stimuli and are able to just ignore itches, etc. without much effort. Or for whatever reason naturally have a more stoic and/or mindful mental state without ever specifically training that skill.
This is a great example because it also clarifies that there's a ton of variation between individuals for experiences we call by the same name -- itching, worry, regret, sadness, etc.
And one of the things you're pointing to about that variation is that some people experience thoughts (and perhaps sensations and feelings) as a whole lot stickier than other people. Some people are more able to dissociate (for good and bad) from their felt experiences. Some people have trouble hanging onto thoughts while other people have trouble letting them go. This isn't a measure of preferences or character strength -- a whole lot of this is down to genetics, with some nurture/environment thrown in.
If someone wants to change their "preferences" -- or let's say rather features of their default mode -- we simply don't know ahead of time how possible that's going to be for them to do. So the folks who don't have your specific problem or your specific way of experiencing things, it's easy for them to say, "well, obviously you're just not trying hard enough, you must be kinda lazy" or "if you'd just do some more meditation, this wouldn't be a problem for you" or "if you just accepted that you're hanging onto unhealthy preferences, then you wouldn't be this way."
All kinds of change is possible. In my experience, both more and less than we think. But we have no earthly idea for another person (and nor do they for themselves until they try) how much change is possible for them. All of which should invite humility when people speak about other people's experiences, but doesn't seem to always.
Caplan is a provocateur who believes he is always the smartest man in the room. Too often he is out of his lane and does not know what he is talking about.
Cowan seems to have two versions of Mental Illness is Preference. One is that it just unusual or disapproved preference; the other a matter of conflicted preferences.
"Question for Scott: Would your reaction be any different if someone said, “I’ve gotten tired of my self-destructive preferences, please help me”
"People often feel inner conflict and remorse. Consider the Biblical account of Judas’ betrayal of Jesus. A theologically conflicted Jew betrays his rabbi, feels guilty about it, then hangs himself. If this doesn’t show that Judas was sick, why does Scott’s vignette show the alcoholic was sick?"
But the conflicted preferences version doesn't support Szaszian claims anywhere nearly as well. Someone with conflicting preferences could voluntarily seek treatment to resolve the conflict, without any of their preferences being disapproved of.
"or another word ending in -bating which is less polite but as far as I can tell equally appropriate"
On yerself, Scot
I'd argue that in basically all cases, it's more illuminating and promoting of empathy if we think of the issue as a constraint or a fact about the relevant person. For example, gay people have the same preference for emotional and physical intimacy as straights, but while straight people can best/only satisfy these preferences through a relationship with someone of the opposite sex, gays can best/only satisfy the exact same preferences through a relationship with someone of the same sex. This avoids ascribing to gays a poor taste in sexual partners and instead facilitates empathy with their situation. More generally, preferences often carry moral or aesthetic overtones, which facts don't have.
I have read both bryan caplan and scott alexander for a long time, and i have high respect for both. But mental health is the part where i think bryan caplan is almost entirely wrong.
I have written a bit to caplan, and have gotten a few responses back and forth. I cant find the original responses, so i have to remember them, with the unreliability that comes with it.
(Personal note: i have adhd, autism, and is gay and have had depression and suicedal thoughts struggles for most of my life. Im a 27 year old swedish guy.)
I think this mental health, “preference vs constraints” debacle can be broken down into 3 parts worth digging into
1: “preference vs constrainsts” is a flawed model of humans
2: bryan caplans ideological beliefs and worries that economic and personal freedom is reduced by common disability and mental health narratives, and my thoughts on that
3: bryan caplans personal psychology and philosophy
I will then go over the thought experiments and illustrate how these points helps explain key disagreements
Lets start with point 1:
1: “preference vs constraints” is a flawed model of humans
I think this model is highly useful as a tool of analysis. But all models are useful because they SIMPLIFY reality, and remove other variables.
While I believe its much more relevant then most people here, i do think that there actually 2 MORE ASPECTS of it that is neglected. Those are…
“Harmony”:
And…
“Outside signals”
Lets go into harmony first.
Preference and constraint is seeing a person as one single entity that is consistent, and dont conflict.
But as split brain patients show, we are actually a ton of smaller modules that work together, and often conflict.
Anyone that has tried to quit smoking or to diet have some 3
Experience with feeling that they have different preferences and constraints at different times.
Most normal people dont have too much disharmony in them. But some people have very big disharmony. An alcoholic can have parts of him that desperately wants alcohol, while most parts dont.
Other people have very much harmony in them: i think that naturally conscientious people are like this: the differents parts of themself dont have much conflict.
We have limited energy and willpower, and if there is too much stuff happening mentally or physically we become overwhelmed. So if you have disharmony in you, then things usually require more willpower and work to make the parts of you work together, which makes you less conscientious.
I personally have low contentiousness, while i believe bryan caplan to be unusually extremely high on conscientiousness and harmony: so he has never to a significant degree experienced that clash of motives, and thus see everyone else as being the same
“Outside signals” is more straightforward; its basically that you always have to interact with other agents, and being able to consistently show what preferences or constraints you have heavily influences the interactions you have. If you have very low outside signals (like having an invisible disability) then your ability to negotiate or get social support is lowered.
One note: outside threats(gun to the head) or desperation can temporarily bring you disharmony’s parts into an alliance, the same way an outside threat makes polarized peoples cooperate. But this requires tons of effort, and as soon as the threat is removed the disharmony is back
Point 2:
2: bryan caplans ideological beliefs and worries that economic and personal freedom is reduced by common disability and mental health narratives, and my thoughts on that
Ive noticed that in climate change discussions, climate change deniers are usually worried that if they say that climate change IS a problem, then their freedooms will be taken away simce its usually seen as collective problem, outside of personal responsibility and agency.
I think many libertarians get stuck into this mentality on the topic of disabilities or mental health, cause mental health is in a way built on the idea that you are not fully rational or consistent. That weakens the individuals strength and rationality, and has to be fair in the past been used to completely imprison and abuse people since mental ilness has been a way to say “their opinions dont matter”
As i earlier said, ive written to bryan caplan a number of times. And one of them i wrote along the lines of “i feel super depressed and i wonder if im a burden since im on welfare, am i that?”
And bryan caplan wrote something along the lines of “im busy and cant respond to much right now, but im sorry your feeling horrible. I suggest reading the book doing great. Good luck”
This seems at first to be a contradiction to his deppression preference article. And i think it is!! But i think that when BC is interacting with one individual that expresses suffering and advice, that it doesnt really challenge the individual liberty intuition he has, it is strengthened as he sees someone taking personal responsibility and putting in effort.
But when it comes to the abstract, his intuition is that adults are adults and should be left alone first and foremost, and only jn extreme circumstances have liberty taken from them.
To be fair, the mental heath and disability rights movements are generally left leaning, pro redistribution, and somewhat anti econ freedom. I think this is a mistaken view myself; im libertarian, socially liberal, and think capitalism and econ freedom is awesome for disabled people. Anyhow, the last point;
3: bryan caplans personal psychology and philosophy
Bryan caplan is a huge fan of michael huemer, who is a philosopher that advocates for conservative epistemics; the idea that before we use logic, we must appraise what appears in front of us. Its skeptical towards principle based or utility based formulas, and thinks we often fool ourselfs. Its heavily leaned towards trusting intuition and appearances first, and only trusting abstract reasoning if it has repeatedly and clearly demontrably outperformed appearances
From what i can tell, bryan caplan also has a very strong intuition about just desserts; people should get what they deserve, and etc. he also is heavily skeptical about abstract moral or philosophy claims made by people like peter singer
He seems to go by “revealed preferences shows what people actually care about” in morality. Which he uses to dismiss the idea that animals morally matter, since we treat them like garbage, and even vegans prioritize small comforts for themself over animals.
He wants very very clear and obvious proof that someone is disabled and constrained; if its vague then his intuition about people faking or should be left alone is much stronger then the intuition to help someone in pain.
To summarize:
1; {preference vs constraints} has two more prts: {harmony and outside signals} (lets shorten this as PCHO)
2: {BC worrys that mental health ideas takes personal agency away from people}
3:{BC goss with intuition first, and distrusts vague abstract and hard to judge things that are hypothetical}
So i just went over all of this. But how does this interact with the thought experiments and debates BC and SA have? Well, lets see:
A: the alcoholic.
SA thinks he is a victim but BC thinks hes blameworthy
On the PCHO spectrum…
BC dont see the alcoholic as being disharmonious, and sees him as consistently and visibly prefering alcohol to the family.
SA sees him as having conflicting parts and because of his therapist background, can see a lot more of the signals the alcoholic gives that he actually hates this situation.
Additionally, BC is going off “is this person worth associating with or a pain in the butt to help?” Intuition, for which the answer is “yeah get away from them”
While SA dont have these as much
B: tweaking the variable
Lets say now that we have a completely normal non alcoholic person. If they decide to start drinking copiously, or deliberately takes a pill that turns them into an alcoholic, then i think both SA and BC would see him as fully responsible and preferring alcoholism
But if he person is at first normal, but then an goverment agency feeds him a drug that turns him into an alcoholic, then i think even BC would agree with SA that this alcoholic is a victim and deserves help or sympathy
In these cases, the first one of personal deliberation has HIGH HARMONY and HIGH OUTSIDE SIGNALS. Its very easy to see that this person wanted this consistently, and its very easy to judge this as an outside viever.
The latter one has an LOW HARMONY and HIGH OUTSIDE SIGNALS: aside from just normal morals about coercion, its easy to see that this person did not at all consistently want this alcoholism, and its obvious viewers.
Ok ive written a lot and is tired now. Hopefully this is interesting and helpful
> Bryan Caplan is a huge fan of Michael Huemer,
Both are fans of Ayn Rand.
funnily enough im 90% certain raynd had autism
I wish I had both your intelligence and humor!
mental illness - when someone adopts one or more preferences which produce a significant negative impact on their lives vs. their alternatives.
We're done here.
What most bothers me here is the conflation of different kinds of things under the heading of "disease" or "illness". There are often good reasons to treat these things together, but there are also important reasons to treat them separately.
Prader-Willi syndrome [side note: it is not caused by "damage to a region of chromosome 15". It is caused exclusively by damage to that region in the paternal copy of chromosome 15; the same damage on the copy inherited from the mother gets you the radically different Angelman syndrome], Down syndrome, and cancer are all examples of the body breaking down. There is an intended state of affairs and an intended functionality, and those things have failed to be. But the failure is spontaneous.
The stomach flu this essay mentions in passing is a phenomenon of a radically different kind. There the body is functioning according to plan, but it's been invaded by hostile forces which impair its functionality anyway.
Both types of problem commonly cause the victim to complain that something is wrong. It's also common that they don't. The archetype for problem-without-complaint is of course the brain slug that takes over your mind, causing you to do things like protecting the brain slug and resisting efforts to have it removed and incubating its children.
So we can reject the idea that if someone isn't complaining about a problem they have, then no such problem actually exists.
Aging is similar to the two categories above in that it involves a breakdown in bodily functionality and sufferers complain that things have gone wrong.
But it's unlike them in that it's not a deviation from the plan. Aging is normal, intended functionality. It's just bad functionality that we wish we didn't have. I strongly suspect that this is why people reject the idea of calling aging a disease. I can be certain that that's why *I* reject the idea of calling aging a disease.
This is a teleological argument: it assumes that a state of affairs is intended, and what deviates from the intended plan is an anomaly that we are justified in correcting.
My question would be, who does the intending, and why should we care about their opinion on the matter?
> who does the intending
No one does the intending. That doesn't mean the intent isn't there.
This is the old chestnut about the evolution of an eye. Any investigation of the anatomy of an eye will tell you conclusively that it's designed to provide vision. Do we believe that that's because the eye developed over time due to the usefulness of vision, or because someone with a plan ordained that eyed animals should have vision?
People make both choices. There is not a lot to recommend the second one, except for the obvious purposefulness of the eye. The answer is not that in fact eyes do not come with a purpose. It's that purposes can exist in the absence of a planner.
> and why should we care about their opinion on the matter?
Because this is, by definition, the answer to the question of whether something has gone wrong. There are many implications for how likely you are to be able to change the outcomes you see that depend on this.
> This is a teleological argument: it assumes that a state of affairs is intended, and what deviates from the intended plan is an anomaly that we are justified in correcting.
This is just a gross error on your part. My argument states overtly that a state of affairs is intended.
There is no assumption that what deviates from the intended plan is something that we are justified in correcting. My argument is that using the term "disease" to follow this natural phenomenon is a way of carving nature at the joints. (And in fact, that "disease" needs to be further subdivided.) This is a difference that reflects into other aspects of reality in ways that matter. Aging should not be considered a "disease" because it is unlike diseases in almost every way, and the lessons that are useful as to one are generally not useful as to the other.
I think that by using words like intended you are engaging in at least unconscious anthropomorphization, assuming that the process that created us has goals that we should be loyal to over our own goals.
I'm curious, other than this 'intention' argument, in what ways are aging different from other diseases?
> I'm curious, other than this 'intention' argument, in what ways are aging different from other diseases?
Perhaps you might benefit from reading the comment of mine that you responded to.
Philosophically, you may prefer to call this a functionalist argument (avoiding the word 'intent') or (a-la-Dennett) taking the intentional stance. Either way, it is an indispensable way of understanding the world that is perfectly compatible with evolution.
The problem is that there's a smuggling of what is into what ought to be. Diseases are things that we think ought to be cured. Why should we defer to the function of evolution when deciding whether we ought to seek to cure death?
I think Bryan Caplan may be a philosophical zombie. He appears human and so we assume that he, like us, has interior experiences. But he does not. It's the only way to understand the depth of delusion to which his argument takes him.
also there is absolutely no reason why homosexuality would be a disease and heterosexuality would not. caplan is wrong on so many different levels that it's difficult to even consider how to reply.
> also there is absolutely no reason why homosexuality would be a disease and heterosexuality would not.
This is, um, not an argument that can withstand half a second of scrutiny.
It's not an argument at all, just a statement. I do have an argument but I have yet to put it together. Thanks for your interest!
Sociologically, I think something is probably a disease when it is an obstacle for satisfying the values of either the person with the condition or of society at large.
For homosexuality, we decided as a society that we should reorganize such that gays have a place within society, and a lot of people's values shifted, and so being gay is no longer considered a disease.
For Down Syndrome, we're still sorting out what we value, and even the people with the condition probably vary from being frustrated by their limitations to being completely content with the way they are.
For let's say a panic disorder basically no one values the condition of randomly suffering and freaking out, and panic attacks get in the way of quite a lot of things that people and society do value, so we consider it a disease non-controversially. If we believed as a society that people in a state of panic were getting closer to God or something like that, we probably would have a different opinion.
I am wondering if anyone knows whether David Smail is part of the anti-psychiatry "movement".
ADHD is just going to be an easy target for a long time. In the last hundred years the society has changed in terms of how much attention needs to be paid to activities, particular ones you get paid for. Sometimes the needs are intense, and people who had some difficulty adapting to the slower pace in other eras (even within memory) are now thriving. But this doesn't stretch indefinitely. A fourth-grade teacher from the 1950s might have to adjust her metrics a bit, and everyone might start looking at various professions with different eyes, but the ability to focus on a task and see it through ain't that different. This mild ambiguity in our era allows clever people to create fun anecdotes in their attempt to prove that it is all hogwash. But they are just fools with large vocabularies.
I was a psychiatric social worker in an institutional setting for forty years, and you have heard most of the arguments by the end of year one.
You may have mentioned in your writing, but I will mention now. Szasz never treated a schizophrenic patient and may never have met one. At this point I have to wonder what people are hiding.
This is interesting, but can you please address how this might apply to something like obesity? Is that the result of someone having constant suffering unless they eat (like Prader-Willi) and constant suffering if they try to exercise? Or is it the result of someone clearly and rationally choosing to overeat and not exercise fully aware of the tradeoffs? Is it a little of each? Is it different for different people? Should we just politely assume that it's the Prader-Willi-esque explanation for everybody? Is that even more polite than assigning them agency? Should we just give everyone semaglutide and stop talking about it?
One additional point is pain as a warning for future damage. I once kept working despite increasing pain — until I literally couldn't any more (finally passing the constraint test). I had reached the point where I would drop coffee cups etc because of damage done by continuing. The preceding pain was screaming at me STOP DOING THIS RIGHT NOW, but I kept going, stupidly, in part because I though of the pain as just an experience and not as a warning that I was in the process of destroying myself.
And people do this all the time. Some tear muscles at the gym because they ignore the pain, others burn out, and so on. They CAN keep going so they DO keep going — until they can't. You can even die by ignoring pain. It's clearly not just an experience or a preference.