Crespi and Badcock are charlatans with no experience in the relevant fields, who cobbled together some things that sounded intuitively plausible at first blush and defended them despite overwhelming evidence to the contrary from just about everything adjacent to the relevant neurotypes. The sole reason their pseudoscience lived so long is that in 2011 the Wikipedia article for the hypothesis was written as a glowing endorsement by a supporter (violating a whole pile of site policies and guidelines with it) and stayed more or less the same for a decade, being read by about as many people as you'd expect a pop-psych summary on the fifth most visited website to be read.
In the vein of "interesting things about autism that cause it to not be a failure", besides the obvious engineer and intelligence cases, is one about inability to distinguish fantasy from reality! Speaking from experience here as mildly autistic, i often have very very strong emotional reactions to fiction in a way that is similar to if those events happened in real life, and this is apparently quite common for aspies. This offers the interesting benefit that if a fictional story tries to relay an important message or lesson or moral, i'm unable to simply disregard is as "just a story", because it feels as emotionally important to me as if it happened irl. Thus, this aspie specific ability allows us to not just learn things, but truly experience things in ways that otherwise would only be possible through actual lived experiences. Of course, this comes with the tradeoff you may learn the wrong lessons from fiction, but it's an interesting power nonetheless.
This is speculation and probably not original, but it might be interesting to consider how e.g. autism fares now compared to in the past. Your mildly autistic Google engineer is probably doing quite well for themselves now, but might not have had the same opportunities before engineering was a thing. So certain genes that evolution was in the process of weeding out might not be selected again, or selected against as heavily, as societal incentives change, which also has implications as to which genes are considered failures by your previous definition.
Maybe it is neither a failure nor a tradeoff, maybe it is an evolutionary beneficial trait of evolution itself (let's call this meta-evolution).
Let's imagine that you're a drug researcher trying to determine the best dosage of a particular drug. You'd probably run an AB test with a range of different dosages and pick the one that makes the best tradeoff in terms of desired effect vs undesirable side-effects. But what happens if you expect environmental changes to alter the sensitivity to this drug over time and you want to keep your drug performing optimally. One thing that you could do is give a narrow spectrum of different dosages to different patients and continuously monitor differences in outcomes to tune the mean dosage that you use as your reference. You trade off optimal performance now against adaptability to future changes.
It's entirely possible that evolvability is itself an important evolutionary trait and what you're seeing here is the result of Mother Nature's (extremely unethical) AB testing framework. You see a spectrum of different levels of psychiatric disorder because without such a spectrum, evolution becomes unresponsive to changing environmental conditions and we are highly evolved in favor of being adaptive.
I think you did a disservice here by with the explanation of tradeoffs involved in autism. For instance, I know some autistic people (particularly autistic AFAB people) who are definitely autistic, and definitely very able to function in society, but also are very creative and good at gestalt thinking. There is a tradeoff involved, but simplifying it to "people person" versus "average person" versus "stereotypical aspie engineer" reduces that tradeoff to the point of meaninglessness, in addition to making it seem more like a binary (trinary?).
In addition, what role is the "autism spectrum" in this explanation? Does it match with the triangle up and right from the "stereotypical aspie engineer" dot?
This is a really good read, with lots of "oh, that's me!" moments. Just making this comment because I wanted to offer some more concrete appreciation than a like.
This is almost certainly just...motivated reasoning. I want this to be true, so I'm going to ask.
My mother has schizophrenia -- the very bad, refusing-medicine, wanting-to-kill-people, writing-screeds-to-the CIA kind -- and so did my great-aunt on the other side of the family, less severely. My husband's dad had a psychotic episode about a decade ago that he seemed to white-knuckle himself out of, somehow. I'm terrified that my daughter or my future children will be schizophrenic, since the genetic dice seem loaded against them.
BUT! My husband and I aren't like this: "creative, attuned to interesting patterns, and charismatic, but also a bit odd and superstitious." If anything, I'm "responsible and perfectionist, but has trouble letting things go," and he is chill and was a math major.
Soooo does that mean we maybe don't have those pesky psychosis-causing genes?
P.S. I first found SSC when I was trying to understand my mom's illness. Scott was incredibly kind and generous with his expertise when I, a total stranger, sent him a frantic email asking "will my children be nuts??"
> I think most psychiatric disorders exist on a spectrum from mostly-tradeoff to mostly-failure (what we might call "high-functioning" and "low-functioning" versions of the same phenotype).
For completeness's sake, there's hypothesis that anorexia (and maybe other eating disorders) fit this category to. Gusinger 2003 (https://pubmed.ncbi.nlm.nih.gov/14599241/) argues that there's a tradeoff between "being able to tolerate starvation in order to flee famine" and the pathological "self-induced starvation and hyperactivity" behaviors that constitute anorexia. (This came up in a comment thread or two on some SSC articles)
I think Guisinger's formulation is a little "just-so," but I think there's something to this. The same behaviors can be induced in laboratory animals (see this article, which was recently cited in the Subreddit https://pubmed.ncbi.nlm.nih.gov/22231828/).
>In fact, just being conceived in March raises your autism risk a bit - it means you were in an especially vulnerable developmental stage duriing flu season!
So, is it plausible that flu shots (for parents-to-be) actually reduce autism?
>Overall genes that increase risk for ADHD decrease risk for OCD, and vice versa, suggesting that at least one advantage of ADHD is lowering OCD risk.
With this assumption how would one account for people who have both ADHD-like tendencies (forgetfulness, distractibility, impulsivity) and OCD tendencies (obsessive checking and the like)? Would these contradict the "opposite" model, or could one explain the obsessive behaviors as a coping method for ADHD symptoms ("I have to check to make sure I turned the stove off because I know how easily I forget!")?
It doesn't make sense to say a trait X is Y% genetic. And if you're talking about heritability, the measure is population dependent and has little to do with genetic determinism.
One of the things I struggle with is the notion that schizotypy correlates with charisma, especially because schizotypal personality disorder is also characterized with strange communication patterns, inappropriate responses to social situations, inappropriate dress and presentation. Is there a study correlating schizotypy with charisma?
Sometimes I find your writing wonderfully nuanced and thought-provoking, other times the oversimplifications and unexamined assumptions perplex me.
In the first place I don't know whether to admire or be appalled at the attempt to tackle this complicated subject in these tiny snippets, but this is our culture and I guess at least someone is talking about it.
But treating mental illness as a homogeneous entity as you sometimes do is problematic. There is zero evidence that they all share anything but some degree of inconvenience to the sufferer. And as you rightly pointed out in your last post, far different etiologies can lead to the exact same symptoms (and similar etiologies to different symptoms). Sometimes the "health" in mental health is what appears to be central. Other times it appears to be an understandable (if unchosen) reaction to problems of living. Other times based on continuing once adaptive behaviors that are no longer so. And so on.
And it presumes that those who do not fit current DSM diagnoses are healthy, that there are no illnesses disguised by their social acceptabilty. I don't know how anyone with any experience of human beings can deny that something akin to the pathology of normalcy affects a significant minority, and is far worse and more sickly a thing than many DSM diagnoses. The foundation of your analysis ad hoc embraces psychiatric conventions of thought and "mental illness" when the discipline doesn't even define what is meant by mental or mind, much less any other foundational term.
And the conclusion that this heterogeneous concept of "mental illness" across the board is "mostly" a sign of failure genes is on its face questionable. Over half the population has a mental illness at some point. So... 40% or so of the population is a genetic maladaptation?
That a significant number of syndromes are a result of mostly maladaptive genes and flat out bad environment factors is incontestable. But a) these are overcounted due to methodological biases towards genetic explanations and b) that they look so similar to "trade-off" based syndromes is a sign of the deep problems of the current system.
And your discussion doesn't mention the most obvious source of trade-offs: sensitivity. The same exact "bad genes" implicated in some psychiatric symptoms are also responsible for thriving. In other words, the presence of some gene variants mean that, depending on the environment, they will lead to either sickness or its opposite [1].
And this corresponds precisely with everything we know about basic biology. You can have all the genes variants for something and not get it. We are district in the animal kingdom in the degree to which we are "programmed" to be highly biologically attuned to culture, such that it becomes sedimented in the body. That nature is nurture and vice versa, to a significant extent.
But your overwhelming focus seems to be on a simplistic treatment of genes and obviously unhealthy very early experiences at the behest of all else. I don't even know where to begin responding to a phrase like "it looks like evolution has been trying hard to get rid of them". I'm hoping that's just lazy phrasing.
The problem with the overemphasis on the "failures" is that a) individuals amenable to turning their "syndrome" into an advantageous trade-off are under-emphasized and b) the degree to which any psychiatric symptoms can be damaging, regardless of whether it meets a DSM threshold, is brushed under the table.
For an example of the latter, the vast majority of individuals with any ADHD symptoms to any extent are impaired, and impaired to the extent to which they display those symptoms [2]. The common sense conclusion from this is that ADHD is not an illness, its symptoms are signs that they are already sick. And full-fledged ADHD is simply an extreme form of that illness (that often emerges as consequence of disadvantage [3])
Successful individuals are simply those who go from being ADHD - having no ability to distance themselves from their worst excesses - to having ADHD. It's a developmental achievement in those without the worst genes/environ combo, and involves calming/channeling its co-occuring cyclothymic temperament [4] into an asset.
But does every syndrome operate like this? Of course not. But when you lump all these random syndromes together, it obscures more than it reveals.
In my experience, a lot of lawyers have ADHD (especially in the field of criminal law). This makes sense for similar reasons to why ER doctors would have ADHD - if you have trouble doing things before the last minute, doing long-term followup, and have a whole bunch of coping strategies around organizing chaos already, those fields are a much better fit than other areas of the appropriate profession would be.
The problem, of course, is that the difficulties of the disorder are still very much there. They just get in the way less.
On tradeoff is suggested by how all minorities usually fare in a society, with most of them repressed, but with one or a small group rising to the top by the very virtue of their unique perspective. The outlier of the outlier might be king for having something like two eyes.
It beautifully ties together all the different symptoms of autism into the predictive processing model, postulating that autistic phenotypes arise from weak inputs from top-down predictions, at different time scales and layers.* I really recommend this read.
It also fits the observation that many of these risk genes are in synaptic proteins, and their phenotypes often include abnormal dendritic spine physiology.
I think it fits very well into what you say. To the extent that autism is a tradeoff, it's a tradeoff of stronger bottom-up input (and weaker top-down priors) which makes your thinking less biased, lets you see things others ignore, and lets you forgo the intuitive wrong answers and actually do the math. The price is that you can't adjust yourself to sensory input, have trouble predicting where the flying ball will go, and find it hard to not take someone's words literally.
To the extent that autism is a failure, it's a failure of what Clark calls "precision-based weighing" - i.e. deciding accurately how much weight to give to bottom-up input vs top-down priors. And I suspect that schizophrenia shares much of that in common, and may well be a similar failure and the opposite tradeoff.
*At one level, it makes a person unable to adjust away a constant sensory input, so they stay highly sensitive to it and can't ignore. It also makes them really want a predictable, self-caused sensory input, i.e. stimming. At even shorter timescales, it leads to the very idiosyncratic motor problems. At longer timescales, it makes a person unable to use priors on human behavior when interpreting other people.
When u write a long comment after reading the first part of the post thinking it's going off in another direction only to find out that's the point Scott was making :-( :-)
But I would add that there is another kind of trade-off to be considered. The general performance/tolerance for errors trade-off that happens when you overclock a computer. It might not be that autisticy genes are ever helpful but the overclocked brains that tend to make the brightest engineers just have turned down the tolerance for that kind of screwup way down.
You made me really uncomfortable with your justice system analogies. They seem so inapt that it's hard for me to accept the surrounding discussion.
> In a country biased towards finding people innocent, it only takes a tiny failure to let a murderer go free. In a country biased towards finding people guilty, it would take a huge failure, or many different failures in parallel.
For this to be correct, it would require a bizarre measurement of failure size.
The problem is this: the simplest way to let a murderer go free is to convict an innocent person instead. Problem solved! Nobody gets murdered twice; as soon as you convict one murderer, you're done. But the "we find everyone guilty" system will do this *all the time*. They do it on purpose! That's the whole point of a "we find everyone guilty" policy.
In order for this mistake to require a huge failure, suspecting the wrong person would have to be considered a huge failure. But that can't be right. For just one example, if a married woman dies, her husband is automatically suspect. I don't think that's a mistake. I don't think many people at all think that's a mistake. But it would certainly be a mistake if her husband was automatically guilty!
> Go back to the two ways a justice system can go wrong. First, it sets too many guilty people free. Second, it convicts too many innocent people. If you ask which tradeoffs cause each problem, you'll find they're opposites.
Well, no. There are more than two ways the justice system can go wrong. This perspective only makes sense if you consider "the justice system" to consist of nothing but court trials, into which suspects are deposited by some ineffable force. And in the larger perspective, we see that these two problems are not opposites at all! Every time you convict an innocent person, you automatically let a guilty person go free! Increasing the one metric necessarily increases the other! (Though not vice versa.)
There's a mathematical formalism called "detection theory" which separates bias (what is called "tradeoff" here) and sensitivity (the inverse of "failure" here), starting only from success or error rates to two opposite problem types. Its results are usually represented on a ROC curve, i.e. a graph with false alarm rate on the x axis and hit rate on the y axis; this can easily be adapted to put "rate of using precise cognition when the situation doesn't call for it" on the x asis and "rate of using precise cognition when the situation doesn't call for it" on the y axis. Let me refer to this image I've drawn in the usual style of the blog: https://imgur.com/a/7ucjzUA
There's a few differences of interpretation between this formulation and Scott's: Detection theory says you can have super-high bias even with high sensitivity, unlike Scott's model which tightens the possibilities for "tradeoff" as we approach no failure. It's just that bias matters less and less as you reach towards infinite sensitivity (no failure). At infinite sensitivity, bias is indeterminate, not zero. (At infinite bias on either side, sensitivity is also indeterminate.)
Interestingly, though Scott applies the model primarily to mental health here, he briefly touches on its applicability to social issues such as crime. One thing I've noticed is that an enormous amount of energy, political will, debate, advocacy, resources and thought are spent on where we set the bias term, and discussions that touch on the sensitivity term (anti-failure) tend to be rare and difficult to find. Presumably this is because moving the bias term is relatively easy and improving the sensitivity of anything can be very difficult, but this still tends to annoy me at times.
And I think it's worth adding that the neurodiversity movement often conflates several things. Is x a desireable trait or not, should we search for a cure for x and should society avoid regarding x as a disability or negative so it's not viewed with pity and we don't treat those with x less seriously. I mean it's really hard for any trait to be totally neutral.
I mean consider being very short. Every person I've met under 5'1" wishes they were taller, being that short is almost purely a negative in today's society (even if it just keeps you from reaching high objects) yet we don't view being short as a disease or medical deficiency and it might not be worth looking for a cure for being only 5'. In other cases, like ADHD it's probably good to find a treatment (even if on net mild ADHD is beneficial it's even better to be able to turn off) but we don't view those who have it with much pity or treat their opinions as less important. But that's only possible because they (we) seem mostly like everyone else.
Rhetorically and psychologically it's really hard to say yes, all things considered, severe autism makes life worse and a cure would be good but I do think want you to tag sufferers with all the negative attitudes we associate with mental disability.
I mean hell, even if your a high functioning autistic person surely it would be great to have at least a temporary treatment letting you switch on the way normals see the world. I mean even you don't think it's better surely it would be useful to experience it did say 6 months.
I find it interesting ironic that it's pretty stereotypically high functioning autistic to ignore all the social implications and overtones of saying we should try to cure autism and/or avoid carrying fetuses with high risk of autism to term because autistic individuals have lower expected utility in our existing society.
Not saying that is true just find it ironic that the kind of claim which often provokes the most negative response by the autism nuerodiversity folks is something that I've found to be frequently voiced by individuals with a degree of Aspergers.
I find it fascinating to apply the tradeoff/failure framework to hiring, and was just having a debate with my household about this - the similarities are striking (innocent : guilty :: bad hire : good hire, where you want to catch the good hires and let the bad hires go, but it's unclear if a hiring process that is failing to hire as many engineers as it wants to get is failing or just very far on the spectrum).
We know, or at least strongly suspect, that a minority of people experience weird immune reactions to viral infection, some of which include psychiatric effects. This is getting much more attention now, with covid, but it has long been associated with viral infection in general, because it is more about the defective immune response than a feature of a specific virus.
I read a lot of 19c medical history, so I can compare today with a time in which everyone got a lot of infections, which they had to fight off without help. Obviously, we'd expect to see a lot more of this, and while I can't prove it, there seems to have been an accepted connection between viral illness and psychiatric symptoms in adults and teens, and many probable cases.
But there are surprisingly few mentions of anything resembling low-functioning childhood autism, and even fewer relating such a condition to a recent illness or illness during pregnancy. Both early childhood illness and illness during pregnancy were quite common at the time. This puzzles me. If it is at all related to immune function, which the March birth thing suggests, we should have seen more of it even well in to the 20th century--the 1918 flu caused a lot of weird psychiatric reactions, and other diseases were prevalent. Early 20th century records do have many more cases, under the name childhood schizophrenia, but few seem to have made any connection to viral infection. Maybe it's just not properly documented, but it strikes me as a real puzzle. Everything attributed to covid I'm used to reading about in historical accounts, and seem to be well-established immune responses. I think we've massively underrated the role of viruses in triggering certain medical conditions, and the wide variety of possible symptoms. But I don't see many descriptions matching childhood autism. Just something I wonder about, as there are other reasons to think there is a link between autism and some sort of immune dysfunction.
It seems that this is an evolutionary version of a bias-variance tradeoff, which makes sense, since in any optimization system you'll find a tradeoff like this. And as with the examples here, the system can in theory minimize either, but typically there will be some of both.
Two comments... first, anyone here read Marco del Giudice's "Evolutionary Psychopathology: A Unified Approach"? He tackles the problem of the ontology of psychiatric conditions head on, with an attempt at a global hypothesis involving life strategies (fast and slow) and many other things. I skimmed through it, but it's way over my head! Any opinions?
Second, about depression, I've read a few articles lately converging on the idea that depression is a response to a distress. Quote from Johann Hari: "This pain you are feeling is not a pathology. It’s not crazy. It is a signal that your natural psychological needs are not being met. It is a form of grief – for yourself, and for the culture you live in going so wrong." Any opinions?
>which I’m tempted to cynically attribute to their being less likely to remember to use contraception
Would it be considered out of line (or just more holistic medical practice than anyone can currently be arsed to provide) to talk to your ADHD patients about long-acting reversible contraception? That feels like a no brainier.
You talk about autism and Schizophrenia probably being under negative selection by evolution. I'd be interested in seeing a list of other things under negative selection. Maybe once we have genetic engineering, we could consider skipping the evolution and eliminating those traits straight away (in some cases, obviously evolution doesn't have our values).
I was taught that ADHD and OCD were frequently co-morbid conditions, rather than inversely related, such as when we see those two and Tics associated (Venn diagram ish) especially in Tourette’s. Also higher than optimal stimulants can convert inattention in ADHD intro laser focus on tasks vs unmediated ADHD translates into laser focused on repetitive play.
The “failure” hypothesis for genetics is bound to be mathematically sensical vs the “failure” hypothesis for epigenetic factors, eg in utero virus exposure with schizophrenia.
Finally I was taught that the generic advantage of ADHD ( see Barkley) was related to something like the “twitch” reflex in animals, conferring advantage to humanoids coexisting with predator animals, where scanning and distraction, impulsivity were advantages or at least had a valued place in the village.
I've heard that historically many that religious figures have personalities that sound a lot like schizophrenia. It's possible the tradeoffs were more worthwhile in the past.
Also it's possible that since many of these iq and creativity tests were designed with neurotypicals in mind that you would need different tests to get a good measure of the tradeoffs.
I'm not terribly well versed on this, but doesn't the idea that a gene is selecting for its propagation, not your propagation come into play? That is, if a clan produces 90% normies that have babies, 5% autists that make great tools but no babies, and 5% OCDers that make great use those tools but no babies, that the clan could outcompete a clan with 100% normies who make babies, ok tools, and ok use of the tools? So each of the first can's normie parents produce 5% autist and 5% OCD babies, ensuring a continual supply of them without the autist / OCD parents producing any?
Something I am much more familiar with is that I think the role of autist in security is understated in this article. For 1, OH MY GOD SOMEONE IS BREAKING THE RULLLLEES! and for 2, don't high functioning autists, definitionally, develop formalized rules for behavioral and social interactions instead of just intuiting them? Well, if you want to develop a security procedure, you need to generate formalized rules. Actually, there's another - things like threat analysis require you to really, really, really understand all the bits and pieces of a system, and then identify all the ways you could subvert them. Bruce Scnierer calls this the Security Mindset. And it requires reducing people to things and behaviors to rules. (I spent 10 years in this field, fwiw.)
The tradeoff is presented here as being due to intraorganismal constraints (e.g., I can't maximise size and speed at the same time). Some recent theory has suggested that things like autism and psychosis arise from intragenomic conflict, leading to genomic imprinting. The tradeoff here is between genes from different sources (e.g. maternal and paternal), which have different evolutionary interests (e.g. the paternal genes have lower expected relatedness to siblings than maternal genes). I haven't thought how about how it might change your interpretation (if at all), but it does change what patterns we expect to see, both in behaviour and in traces of selection in the genome. It also might mean that some of the things which look like failures at the genetic level, aren't just deleterious mutations in the traditional sense, but imprinted genes that increase their own fitness at the expense of the largest co-replicon (the organism).
Here's another analogy for both trade-offs and failures working together to create a psychiatric condition.
When amateurs discuss tank design (as in war machines not water tanks,) the first thing that they reference is usually the gun-size/armor-thickness/mobility trade-off; Generally, the large the gun, the thicker the armor and the faster the tank, the more useful it is, (In the real world, there are other factors that are also important but for this analogy this is good enough.) Bigger guns and more armor make a tank more formidable, but all things being equal, they also make it heavier and therefore slower. A tank that has a bigger gun can engage an enemy at a longer range, which is a big advantage, but a tank which is faster and more maneuverable has a better chance of choosing the ground in which an engagement occurs, also a big advantage. So there's an obvious trade-off between the size of the vehicle and it's mobility.
The thing is, other factors also affect the mobility of a vehicle, such as the quality of the suspension system and the power the main engine. So if you improve the engineering of a vehicle by using better suspension or a better engine, you can get around the weight trade-off and back the tank faster without sacrificing armament or armor. This is why the M4 Sherman (WWII era tank) weighed about 30 tons, but could only do about 30 mph, while the modern M1 Abrams often weighs about 60 tons but can go 45 mph, so twice as heavy but half-again as fast. The modern vehicle has a lot of engineering improvements over the old vehicle.
So I think here we have a model for two different kind of underlying trait. If an "autistic" tank was a tank that was too slow to be useful on a battlefield, then you could achieve that either by accidentally making the gun or armor too heavy, ie, taking a trade-off too far, or by making a mistake in the design of the suspension system or engine, ie, a failure without any upside.
It could be that many if not most of the genes that cause psychiatric conditions are usually design failures but a few of them are trade-offs that occasionally go too far, or combine with one of the design failures in a tragic way to cause an identifiable psychiatric condition. Perhaps the presence of a trade-off creates a situation where the presence failures are more likely or more of a problem. So maybe someone is close to the edge for having autism because of the presence of a lot of genes for smartness (assuming that's actual trade-off) but it's only the presence of a failure that puts them over the edge, but without the trade-off the failure wouldn't have been enough to put them over the edge? Or maybe sometimes it's just one cause or the other?
Wouldn't it be more accurate to say senior corporate management? IIRC psychopathy is far more common in senior corporate leadership than it is in among the general public.
Bipolar disorder can be formulated with some relation to the trade-off/failure paradigm but I think it's more a failure that happens to have some trade-offs, complicated by the difficult trade-offs of the system it is a failure in.
To dramatically oversimplify, before engaging in any action some calculation of whether it is "worth it," i.e. are the cost and risk of an action outweighed by the probable benefit. This is very complex process and mistakes either way can be harmful, although false positives tend to be more harmful than false negatives (e.g. "I can take on that lion" is more likely to get you killed than "I can't catch that turtle", at least in the short run).
In depression, the bias is towards "not worth it" usually due to an overestimation of risks and cost with an underestimation of benefit. This tends towards a state of inaction and arguably many of the symptoms of depression such as anhedonia and decreased energy and concentration are meant to keep the individual from engaging in actions. In mania there is the opposite bias, with dramatic overestimations of ability and underestimations of danger. Everything seems like a good idea and is acted upon. This is very dangerous state and untreated manic episodes have a very high mortality rate (~20% IIRC).
I would make the argument that the mood states are not the causes but the RESULTS of the biases, i.e. the system that assesses risk/reward comes to a general conclusion about which way to bias the conclusions and sets the mood accordingly (similar to setting the DEFCON level). In bipolar disorder, there is a tendency of this system to introduce biases inappropriately. There is sometimes benefit to this, particularly when one is mildly on the manic part of the spectrum since they may have more energy and creativity without too much impairment in judgement.
In this formulation, I think bipolar disorder is best characterized as a failure in the decision assessment system, related to the underlying difficulty in optimizing it. There hjust appen to be some trade-offs because the decision assessment system is inherently about trade-offs.
"Instead of insisting that psychiatric conditions cannot possibly be bad for anybody and it's always just social prejudice all the way down, they would be better off admitting that this can be true for some people and false for others."
... they do admit this tho? you are asking them to do a thing they already do.
I've considered similar multifactor models and one of the advantages is in explaining the somewhat paradoxical trend of an increasingly even sex ratio as autism severity increases, since it might be the case that only some of the factors show sex differences and that those factors might be the ones less related to general brain goodness.
Using for instance a 2-factor model where factor 1 is sexually dimorphic (higher in males) and unrelated to mental deficiency whereas factor 2 is not sexually dimorphic and related to mental deficiency, one would expect to find that generally speaking those who’s autism was primarily caused by factor 1 would be highly disproportionally male, whereas those who’s autism was primarily caused by factor 2, and who would be by dint of their mental deficiency deemed a more severe case, would be more even in terms of sex ratio.
I think the counterbalancing advantage of depression is conflict avoidance and lack of aggression. This used to be adaptive when at the bottom of a social hierarchy. I call it the slavery response.
On the observation around genetic factors being evolutionarily selected against, is the following possible:
There are a subset of these genetic factors which are not mistake factors as you have described them per se, but which still influence in which direction the bias goes in your graph. You could view them as having no particular advantage either way until they reach extreme levels - a more creative person may fill one role in society, a more precise person another but both might find reproductive success that way. Such a factor would have either no effect on an individual's fitness or a negative one, and thus would be expected to be slowly selected out of the gene pool, but then quite slowly because most instances of it do not harm reproductive success. I guess ultimately I am questioning whether the axes on your graph can realistically be thought of as orthogonal?
(Then on another level, a group or species might be expected to be more successful with individuals of both these types than a species without either, so that confuses my question further : ( Sorry! )
I think addiction may be thought of as an attractor state as well - a failure of self regulating. On that note - so can compulsive behaviors in OCD. The difference might be that compulsions generally alleviate negative emotions, while addictive substances or behaviors may also enhance positive emotions (though eventually one of the hallmarks of substance use disorders is that the 'alleviate negative emotions' part becomes dominant)
"In the modern day, people with genes for psychiatric disorders tend to have fewer, rather than more children than people without those genes - except in the case of ADHD, which I'm tempted to cynically attribute to them being less likely to remember to use contraception."
I doubt you meant this seriously, but if it were the case that ADHD-prone people were less likely to remember to use contraception, and that resulted in having more children, *that would be an evolutionary advantage* and therefore would support ADHD as a tradeoff, or even just a beneficial trait that hasn't yet fixed.
Is type 2 diabetes primarily due to genetic tradeoffs or failures? While there are certainly genetic predispositions to type 2 diabetes, I expect most would agree that evolutionary mismatch is a more productive framing. In the environment of evolutionary adaptation, our access to refined sugars was far more limited than it is today. There may not have been significant tradeoffs nor should a predisposition to type 2 diabetes in our contemporary environment be regarded as a genetic "failure" in the context of our environment of evolutionary adaptation.
What is most remarkable about this post is that the evolutionary mismatch framing is an entirely invisible category. While it is fair to speculate regarding the extent to which any particular psychiatric condition is due to genetic tradeoff, genetic failure, or evolutionary mismatch (i.e. the genetic sources of a predisposition did not, or were less likely to, result in pathology in the environment of evolutionary adaptation), that third category is not even considered here.
Here I provide a number of independent strands of evidence suggestive of the idea that evolutionary mismatch is a causal factor in adolescent dysfunction, including some psychiatric conditions,
In it I quote Harvard's Daniel Lieberman, a leading expert in evolutionary mismatch diseases, who claims,
“There is good reason to believe that modern environments contribute to a sizable percentage of mental illnesses, such as anxiety and depressive disorders.”
In addition to anxiety and depression, he also adds ADHD, eating disorders, chronic insomnia, and OCD as mental illnesses whose modern prevalence is likely due in part to an evolutionary mismatch.
Aaron Blaisdell, Editor-in-Chief of the Journal of Evolution and Health, UCLA Department of Psychology and Neuroscience, left a comment,
"What a great essay! I just came across it now, and it really resonates with my own thinking on this topic."
I generally avoid citing authorities, but the fact that such a significant category (i.e. evolutionary mismatch) was entirely neglected is striking.
The "autism as overly-precise thinking" characterization is fascinating, but feels narrow. I wonder if a developmental perspective could broaden it. Consider what a tendency for overly-precise cognition does to a toddler. All of their interactions with the world are colored by this tendency, each interaction shapes their minds and shapes future interactions, feeding back into itself.
And then you have the fact that schizophrenia doesn't generally manifest until early/mid twenties. What's up with that?
At the risk of oversimplifying, I prefer to view ADHD as a relative underdevelopment of the prefrontal cortex (as opposed to, say, OCD's overactive cingulate gyrus).
It's highly comorbid with OCD, bipolar disorder, and autism among other things, and I believe that the genes that lead to these comorbidities end up in ADHD families because they give them the high anxiety or stimulation necessary to cope with it.
"If, as Badcock and Crespi claim, schizophrenia is the reverse of autism, it might involve being too imprecise - too willing to declare the identity of unlike data - too quick to pattern-match. If autistics are too quick to mistake signal for noise, schizophrenics are quick to mistake noise for signal. In a well-functioning brain, this makes them creative and socially adept; in a poorly-functioning brain that is constantly getting things wrong, they connect everything to everything else and it's a mess.
I don't know if this is quite right. But it would reconcile two seemingly contradictory sets of findings."
Personality research backs this up: The personality trait most associated with aesthetic receptivity and artistic creativity, Openness to Experience, seems to be partially based on a system called latent inhibition. This is an unconscious process by which your brain essentially decides what associations are real and what are fanciful nonsense. High inhibition, and the associations have to be pretty concrete to count. The more it's loosened up the more upside you gain from being attracted to tenuous associations that may have a point, but your chance of making Type 1 errors also increases. High openness is associated with various measures of apophenia, and apophenia is a risk factor for full blown break-from-reality psychosis.
I think the loosened association is why things like poetry and song lyrics sometimes feel like they have something there, even if it doesn't quite make sense, and other times just inane bullshit. They tickle our apophenic, pattern-confirming tendencies. The best stuff are things that you wouldn't write off the bat yourself, but are still "in range", so to speak.
Wouldn't it make sense that neurodiversity can be adaptive at the population level, even if particular traits are often negative for individuals in particular situations?
Maybe this is a little off-topic, but I want to explain why I'm pro-neurodiversity. The argument about whether autism is a good thing or not is actually completely moot, because there is no autism cure, or even a way of turning low-functioning autistics into high-functioning ones, and there won't be for the foreseeable future. The primary effects of the search for a cure have been wasted life savings and kids traumatized by abusive "therapies," with some unpredictable effects like the resurgence of disease caused by the anti-vaxx movement. I support neurodiversity even for severely disabled autistics because it's the only practical, humane thing.
Also, every theory I've seen for the cause of autism seems oversimplified, including this one. My theory is that autism will result from any major, low-level difference in how the brain processes data, which is why autistics are so diverse.
I think the model being used by some of the people in neurodiversity (and, at the very least, some of the people in my university when I did a subject on this) goes something like this.
Autism is basically a switch. Most people have it turned off. Some people have it turned on (genetically or epigenetically; I seem to recall something about certain methylation patterns having 100% correlation with autism).
Due to the fact that most people have the switch turned off, having high mutational load will correlate with having the switch turned on. High mutational load is bad, because it means a lot of switches will get flipped, and most of those switches are in one position at fixation for a good reason; typically this means low intelligence. However, the correlation is not 100%; people can inherit the switch being turned on from a parent, or can have it flipped by freak accident at low mutational load (or can have high mutational load and *not* have this particular switch flipped).
The neurodiversity people, or at least the neurodiversity people I agree with, think that having the specific autism switch flipped is not bad and shouldn't be eugenically removed from the population. In this model, cutting down on mutational load (somehow) would stop new low-functioning autistics from being born, but have little effect on high-functioning autistics (at least, in the short-term; in the longer-term if there is significant selection against HFA the lack of replenishment via mutation would cause a decline in numbers). That is to say, it would target the failure and leave the tradeoff.
I may have been lied to. Maybe taxometrically the distinction they fed us between low-functioning and high-functioning autism is not real; I don't know. I'm just trying to sketch in the argument as I understand it.
Thank you for this writeup. It perfectly summarizes how I feel about characterizing autism as 'neurodiversity' and I'll link to this rather than trying to re-explain it myself in the future.
Is it possible that 'the genes which put you at risk for autism' are actually the old versions, and what we're seeing is 'genes which lower risk of autism' haven't hit 100% penetration in the general population yet?
Also in terms of tradeoffs, isn't there strong evidence of tradeoffs in terms of blindness vs schizophrenia and depression vs epilepsy? It's possible that our overdeveloped vision system damages our executive functioning, due to the accident of them happening next to each other in the brain and competing for territory over the same neurons, leading to a tradeoff and possibly explaining the bizarre observation that nobody who was born stone blind has ever been schizophrenic. This makes me wonder about a hypothetical where someone invents a compound which can be injected into the eyeball which completely clouds vision until the appropriate enzyme is injected at which point it clears up completely. If there's a newborn whose parents are both severely schizophrenic thus putting them at very high risk would it be ethical to use this to deprive them of vision until the age of 2 at which point hopefully their risk of schizophrenia has gone way down?
On depression vs epilepsy it's possible that our huge oversized brains are at risk for epilepsy for the simple reason that a malfunction in any part of the brain can cause a seizure of the whole brain, so the bigger your brain the more likely they are, so we have things to slow down our brain functioning which cause problems when there's too much of them or the epilepsy they're supposed to counter isn't there. The utility of shock therapy seems to hint in this direction.
Long-time reader, first-time commenter here. Big fan!
Speaking as someone who maybe possesses - ok definitely maybe - one or more of the mentioned traits, the old saying, “Bi-polar is the worst, it’s awesome.” is mostly true. I also have ADHD and five kids, and so the part about contraception is also accurate. The thing i like about your framing is the attempt to contextualize intrinsic and extrinsic attributes. Schizophrenia, generally not pleasant, but there are some potential extrinsic benefits to the heard. Still, extrapolating extrinsic values is difficult without a longview and reasonable longitudinal data. Yet we make trade-offs. You allude to these trade-offs in the triangle thingy, but I wonder if you could provide more thoughts on these extrinsic pros and cons? A triangle would probably be insufficient. Which would displease many sales types. But they’re probably not your core readers.
You speak of evolution as though it has intention, is purposeful. Perhaps this is what you believe. Perhaps it is just a convenient anthropomorphism. But I think that evolutionary trends are emergent functions, inferred by our observation of the laws of nature and the balance of probabilities playing out. This is important because it is easy for us to project the values that shape the sense we make of evolution onto the evolutionary process itself. I believe the value of evolutionary outcomes is determined by humans, not evolution. Evolution is value neutral.
The example of the bipolar coder is very good. Contemporary society demands a lot of redundance out of people and being capable of that is “functioning.” Whereas I think some conditions contribute very specific things that society might only need once in a while, but it makes a really big difference to get it, leading to positive notoriety for that individual and lots of descendants. Historically, some cultures have been willing to completely subsidize the existence of people who bring them something they appreciate - monks, wandering minstrels, hermits, an author who produces two or three incredible books in a short period but lives 80 years. Being in their presence allows observation about aspects of life one wouldn’t know otherwise.
Right now we like intelligence and “functioning.” Past societies may have valued the ability to stay awake for a few days hunting, but only needed that once a month, the fortunate individual being free to do whatever in between. Evolutionary pressures are changing.
I am not sure that the extreme states would have ever been valued, but the trade offs they generate might have been more valued in differently structured societies.
I know a dental hygienist and I think this person may be a sociopath, but they can stay focused for hours on a procedure with a wincing patient and blood and drool everywhere. Narrow trade off but very valuable.
Do you think that when psychotherapy is used to treat autism-spectrum disorders or schizoid disorders, the goal is usually try to move people horizontally on this chart (trying to adjust their prior between gestalt and precise cognition), or downward (trying to improve meta-strategies so as to just be more intelligent overall), or both? Or does it depend on the type of therapy?
Maybe someone has made this point already, but I am getting a slight sense that there's a confusion between fitness and 'good' in the standard English meaning here. Take autistic body language, i.e. stimming by rocking gently, as I do. It's plausible that genes that make you more likely to stim are fitness-damaging, because it looks weird to people, and things that look weird are sexually off-putting. But there's nothing inherently bad or bodily or mental function-impairing or painful about stimming. It's just not the norm. More relevantly there's no reason to think that 'makes you more likely to have stereotypical autistic strengths that would make you a google engineer' and 'fitness-damaging' are mutually exclusive categories. A gene could even in theory be fitness-damaging *because* it did the former. I.e. suppose for example that a gene made you more likely to have a particular 'cold-blooded rationalist' type of personality. It's really not that implausie that it would both make you better at certain kinds of intellectual labour, and overall decrease fitness in an evolutionary sense because most people find that personality-type Not Hot.
I don't doubt that many genes associated with autism really are bad in an everyday sense for the same reason they are fitness negative though, so don't mistake this for an attack on the whole post or a dogmatic assertion of the sort of neurodiversity view on which nothing about autism is bad except how neurotypicals treat us.
Thanks for writing this Scott - I've often felt that the high-functioning end of the autistic spectrum was a strengths and weaknesses model (i.e. trade-offs) but also that the low functioning end is a weakness only condition (i.e. a failure mode). I think that many in the neuro-diversity movement are high functioning individuals, who have a point, but perhaps don't realise that there are non-verbal institutionalised people who are not reading and posting on the internet about neuro-diverse culture. This makes sense to me both from knowing a lot of high functioning geeks in my personal life, and also from my work as a Clinical Psychologist in neuropsychology
Enjoyed this post quite a bit - it’s my first intro to your work. I believe these interactions of human behavior (genetically moderated or not) with the environment are even more recursive and polymorphic than you’ve suggested. It’s probably complex enough that it would require some form of advanced complexity science to even begin interpreting it. There are just too many nodes. The same problem plagues many modern fields of study.
I do admire your observations and logical thinking! And I thank you for that! It is the initial step when questioning, well, everything.
However, I could only go so far reading your article. While I do appreciate it, I had so many scruples trying to get through each paragraph, until I had to stop.
I feel obliged to mention a few of my own thoughts. (Go figure, behavior is so predictable.) I found your thoughts very self-centered, as were the ancient Greeks, and everyone else, say, earlier than the 1950s? That belief being that humans are the center of the universe; the world was created for “us” rather than the other way around. In another sense, it’s sort of a “creationist” way of thinking.
Do you rememberer, for example, the creationist who proclaimed on television that the banana is curved in such a way for the human mouth to consume? Anyways, that’s the image I couldn’t escape while reading your rant… I’ll call it a rant, like Henry Miller’s Tropic of Cancer, because it isn’t scientific, but it is the impetus for questioning, which I am thankful for.
I’m being very lazy right now, and I do apologize! But, when you discuss psychiatric conditions as if the conditions themselves are conscious, thoughtful beings -the ways in which humans are thoughtful- is troubling. Evolution isn’t conscious like you and I are. (I apologize for any non-humans reading this article, but this is strictly a human experience!)
I’m not going to bore everyone with the philosophy, (unless you ask specifically, then I’d love to discuss!), or biology (please inquire!), or break down everything I know about every single psychiatric condition, but I feel I should at least mention one or two conditions, for reference.. perhaps two that were mentioned in the article? Ugh, how predictable of me.
Psychiatric conditions are sometimes the brain’s attempt to protect us. Think of PTSD, for example. If one suffers from ongoing childhood abuse, their brain’s priority is survival, well, because survival was a constant threat. In this case, whether or not the abused’s brain cognitively understands that they are safe, they cannot convince their body and heart otherwise. Think of it as, say, a celebrity death match where the celebrities are the limbic system and the frontal lobe. It doesn’t matter what your cognition thinks about the situation you are in, your heart will start pounding, you will sweat, and you will “go offline.”
In other words, it’s impossible for the frontal lobe to convince the limbic system otherwise. It’s debilitating, thus a disorder. The “hilarious” part, (trying to stay on par with your sense of humor), is that an ignorant person witnessing a person with PTSD experiencing an acute, distressing event will misinterpret their experience as, simply, “geez, that person is dumb.”
Other psychiatric conditions occur during brain development. I was overcome with joy that you recognized that environmental factors also play a part during development; although, you did call these factors “failures.” (I’m not familiar with these hypotheses you mention. Anyways.)
Yes, neuronal developmental also depends on environmental factors, accordingly. For example, toxins such as alcohol penetrate the blood brain barrier of a developing fetus (an environmental factor) and can cause very significant damage, in addition to genetic factors. The neural tube (brain and spinal cord) develop in the first month of pregnancy. Often, people don’t even know they’re pregnant at that time.
With this is mind, I’d encourage you to research the difference between critical and sensitive periods. It’s actually quite a complicated topic for a novice. This is where specialists come in.
Different researches focus and measure different, umm, “things.” These measurements are known as a dependent variable. Intelligence, for instance, in academia, must be defined by the researcher. Dualities such as “dumb” and “smart” are not acceptable.
As you can see, I’ve already written to much and I doubt anyone will read this.
So, last sentence here: The most difficult part about reading your rant was, “Why the ***k am I wasting my time on this?” And then I remembered. You are an intelligent person with logical thoughts, and there is merit in that!
The trade-off between gestalt vs precision thinking reminds me of the distractable vs hyperfocused states commonly experienced by ADHD people. In the case of ADHD it's generally thought of as an attention regulation problem which has to be clunkily coped with by overcorrecting too far to either one side or the other. Maybe there's something similar going on in autism vs schizophrenia, where the mechanism that regulates which type of thinking to use is defective, so the brain settles into overusing one or the other?
I think disabled activists would offer a few additions to the statements made above. A major argument in the disability rights movement is that folks with disabilities suffer largely because society does not accommodate them through attitudes of prejudice, or through lack of consideration for the range of human needs (building stairs instead of ramps everyone can use), and treats disability as an exception, rather than the eventual rule (for all of us, as we age, have accidents, etc.). So in this sense, in many cases, there is not "high functioning" vs. "low functioning" autism, but there are attitudes and barriers that are greater for those with more differences than there are for those with fewer and for those living under more or less restrictive social norms. Many people with autism who would be considered "low functioning" at some point in their lives make a connection that allows more access to the world around them (Temple Grandin is an example of this, mostly nonverbal as a child, had violent episodes, now a leader in the field of animal behavior). In my experiences working with people with significant disabilities for the last 20 years, I would say very few of them need to be institutionalized so they "don't bite their own fingers off." In fact, most of the problems I see people with disabilities face are socially created.
This article does a great job with showing how some “failure states” are real, objective things. All of your articles are excellent. This one is no exception and has given me tools for navigating my own thoughts on my own disorder. Thank you again.
But something was bugging me, and I couldn’t solve it until I got to the “Hypothetical Perfect Person.”
The objective failure states – problems in the womb, seizures, birth environment issues, etc. are obviously factors. And you’ve given proof that (some of) these traits are being selected against.
But the conclusions drawn here actually prove more than I think you intended – and more than I think you believe. Falling in various places on the trade-off scale is some species of failure – in the example you give, if you tend to be more precise or gestalt, you become more failure-prone. You can only reach certain “low failure” levels if you are close enough to the hypothetical idea between precision and gestalt cognition because otherwise you will be less likely to choose the correct cognitive algorithm.
There isn’t a hypothetical perfect person, and there isn’t one “correct” brain algorithm for any given problem. A single problem can have multiple “good” outcomes. In fact, a single problem can have outcomes that are “good” on entirely different axes, providing value that may not even be perceived as value to someone with different priorities.
I mention this because in my experience the working world is trying to standardize humans the same way we standardize McDonalds’ hamburgers. It’s true that in Seattle a precision-thinker on the autism spectrum may be a successful programmer. In Arkansas he’s likely to be a “seriously don’t hire that guy, he’s weird.” And there’s more Arkansas than Seattle in the world.
I’ve never bitten on the neurodivergence movement or the opposition to it, and this article finally gives me the words to explain why – some neurodivergence is just plain bad, and some is not. But the conclusion then treats the other kind, the trade-offs, as though they are deviations from a “correct” norm. And having people at all points along those trade-off spectrums is an unmitigated *good.* It will allow you to find value that people on other points on the spectrum could not see.
This all leaves me wondering to what extent my diagnosis and the diagnosis of others I’ve known are based on failures that are just plain bad, and to what extent it’s based on just deviating from the norm on various tradeoff scales. Because if we’re pathologizing “being different from the standard worker,” that has a lot of very worrying implications.
This is well-written, obviously, but not such a profound concept. The thesis is essentially: An evolutionary flawed genotype can produce distinct, even behaviorally opposed, evolutionarily unfit phenotypes. The best example is cancer. Oncogenes or tumor suppressor genes can either have gain of function (GOF) mutations or loss of function (LOF) mutations. Let's say p53 (tumor suppressor gene) has an LOF mutation. The organism would develop tumors and expire. Interestingly, there is an asymmetry in a germ cell GOF tumor suppressor mutation in that the organism would not develop at all as the cell replication cycle would be inhibited. But if crisper-cas9 was used to insert a mutated GOF p53 in an adult organism, that organism would be less fit because it could not make enough stem cells (blood cells, GI epithelium, hair, etc.) to function optimally. That is, it would phenotypically resemble a cancer patient on chemotherapy. The same is true for psychiatric illness. Mutating the same genes can result in distinct, even opposing, phenotypes. This piece conjured in me the work of Carl Jung. Many psychiatrists have armchair diagnosed Jung with "good psychosis." I tend to agree with Scott that balance provides, in general, the highest fitness over the long term. However, at the very top - which is what people tend to focus on - are the outliers who are usually unbalanced in some way but are able to wield their mania, or ADHD, or autism to perform at the highest of levels. We must also leave room for the counter hypothesis though. That as psychiatrists we only see those with decompensated OCD or schizophrenia and that there are many individuals who thrive with these "mental illnesses" that simply do not make their way to the ward. Thus resulting in a bias. I do not think this is the case, but this alternative hypothesis, that many Jungs and Nashes are out there under the radar crushing it, is still possible.
Worth noting that what Scott describes as a "failure" for the individual (i.e. a "bad" de novo mutation or prenatal infection) is still a manifestation of a tradeoff in a broader systems sense. Thanks to population-level selective pressures, DNA proofreading is not so accurate that the species can never benefit from random mutations, but it's accurate enough that every child isn't born with disabling disease. Prenatal immunity is aggressive enough to protect the fetus from most infections, but not so aggressive that it attacks fetal tissue. Even the "failures" of his justice system example are just reflections of how a hypothetical society has decided to balance investments between their justice systems and other worthy societal goals. In many cases, the "tradeoff" that is being optimized for may not be obvious to us from our perspective of limited biological understanding. And I don't make this point just to be nitpicky -- tinkering with balances we don't understand (and may not even recognize as tradeoffs in the first place) can sometimes backfire.
How much do social mores play into definitions of psychiatric disorders? If ADHD for most of human history was a desirable trait (hyper-focused "hunters" would be a huge advantage), then is it fair to call it a disorder because societal expectations have changed fairly rapidly on the evolutionary time scale? Is our modern context ("civil society") penalizing traits that 10,000 years ago were necessary and vital?
It would be really cool if you would write an article about creativity in the future though too, cause your insights and depth of research into these topics are really appreciated
The conceptual tradeoff model wouldn't apply to mental illness, as it can't be successfully used in somatic medicine. What's the trade-off of emphysema, diabetes, or coronary thrombosis? Disease is a harmful dysfunction and debilitating distress, not a zero-sum game. It's very natural and human to look for the silver lining; however, nature is neither generous nor fair. Evolution cares little about individual suffering; it favors the survival of the whole species and is willing to sacrifice a certain percentage of individuals (to diseases, suffering, and death) to achieve reproductive success.
Also, I would not support the idea that schizophrenia and autism are on the same continuum. I see them orthogonal or, perhaps, operating on independent dimensions. By analogy, both the brakes and engine could fail and affect the car's movement, but they are not on the opposite sides and "the reverse" of each other.
Does Scott mean "allele" basically every place he says "gene" in this post?
For example, "Evolution is imperfect, so some bad genes manage to slip through." We're really talking about bad alleles, aren't we? People have (mostly?) the same genes; but there may be good or bad alleles. Right?
To understand the origin of mental illness (etiology) one might look at the evolution, adaptability and procreation of the species. To understand the mechanics of psychiatric diseases (pathology) one should look at the working and malfunctions of the brain. Treatment of mental diseases consists of the restoration of function and relief of debilitating distress and often doesn't have much to do with either etiology or pathology. It's a mistake, in my opinion, to conflate all these.
I'm a high-functioning autistic mathematician, and although you could say that it's because I'm precise, I think most people who succeed in math do so because of their ability to see patterns and make generalizations from specific examples. We're definitely precise in how we describe these patterns and how we prove them, but I'm just hesitant about your triangle which puts pattern-matching on the opposite side of autism. I would love to hear from other mathematicians about this!
I think there's another possibility that you're overlooking with regards to your personal experience:
Namely, that it isn't that these things are tradeoffs, but rather, that these downsides are more severe in some places than others.
In other words, normal function is better. But people who have some sort of problem which is NOT a problem (or not as much of a problem) in one field will disproportinately end up in that field, because that field feels comfortable to them (it is finally a place where their problem ISN'T a problem, or isn't as much of one).
This will give you exactly the impression you got - that these things are somehow advantageous sometimes - but instead of actually being advantageous, they're just less disadvantageous sometimes.
Humans have an innate sense of fairness. People seem to think that life is a balanced RPG system, and are dangerously biased towards such ideas.
People who have these issues are going to be biased towards thinking that their issues are helping them in some way, rather than viewing them as a hinderance to their ability to function. And outside observers will likely see it the same way.
This makes the "it's just straight up bad" explanation much more likely.
Even the existence of beneficial genes that have associations with negative outcomes sometimes can be explained in this way.
For example, a stronger immune system is a good thing. But if you get some deletorious gene that causes you to have an auto-immune disorder, having a good immune system makes it much worse. But the actual original gene has no downside - the downside is a separate mutation that gets selected against.
And there's also the possibility that the beneficial gene allows the subject to exist at all with deletorious mutations. If you have some gene that boosts your intelligence massively, and then some other gene that screws up your brain, in most people, it would render you totally non-functional and you die - but in the person who has the huge intelligence boost, you end up at least marginally functional, and so get diagnosed with autism.
It could also be that the effect is not on the individual but on their *parent*. For example, smarter parents may be more likely to be able to keep defective defective children alive. Autistic people are much more likely to engage in pica, among other things, and autistic children have a significantly higher disease burden. So genes for high intelligence may appear more often in autistic children because we only test living children for autism. A more intelligent parent may also be more likely to recognize that there is something wrong with their child and seek out a diagnosis, which would also result in children with more smart genes ending up disproportionately likely to be diagnosed with autism.
I'm not really sure I understand the logic of differentiating between trade-offs and failures for mental disorders. I would see the tradeoffs discussed here as *part* of the failures; e.g. increased intelligence in autism is a failure as it allocates too much energy and processing power to a certain type of rigid intelligence, decreasing the amount available for other things that promote quality of life. The same for the other examples listed - they may be more successful at a particular job because of the trade-off, but if you consider life satisfaction or well-being on the whole, it would be a net negative, wouldn't it?
I was immediately reminded of this factor being reflective of the 'failure' axis, but the authors don't really mention this as a potential explanation. Instead, they consider that p might reflect a diffuse unpleasant affective state, poor impulse control over emotions, deficits in intellectual function-, or generally disordered thought processes, but these could ALL be true and STILL originate from accumulated early life environmental stressors that mean your brain just doesn't work like it's supposed to. Maybe you could start a conversation with them about this.
I'm trying to wrap my head around the question of how well the "failure-tradeoff triangle" maps onto the pretty-well-explored ideas of precision-recall and sensitivity-specificity from domains like machine learning. https://en.wikipedia.org/wiki/Precision_and_recall
I don't know anyone outside of AA who uses the 'alcoholism as a disease' model because there are so many other reasonable explanations for why people get addicted (if we view nicotine/heroin/cocaine/X drug addiction as a result of nicotine/heroin/cocaine/X drug being addictive, why do we view alcoholism as special?)
Genes vs Environment: the first portion of the article focuses on genes as primary factors contributing to some mental illnesses. However, neither the scientific literature at this time, nor clinical experience, (in my perspective) lend much support to the idea of genes being responsible directly for these illnesses, certainly in the vast, vast majority of cases. "Genes load the gun, environment pulls the trigger," as some say. Why no thought on environmental factors? Make no mistake, I would be thrilled if mental illnesses could be addressed with genetic approaches, but the likelihood that this is either fundamentally possible, nor feasible, is very slim. Something about the philosophy here is troubling. Why is environment broadly, and environmental factors specifically, not deserving of greater attention?
I guess one potential advantage of depression is that you might be less susceptible to certain cognitive biases, e.g. just world fallacy.
"If, as Badcock and Crespi claim..."
Crespi and Badcock are charlatans with no experience in the relevant fields, who cobbled together some things that sounded intuitively plausible at first blush and defended them despite overwhelming evidence to the contrary from just about everything adjacent to the relevant neurotypes. The sole reason their pseudoscience lived so long is that in 2011 the Wikipedia article for the hypothesis was written as a glowing endorsement by a supporter (violating a whole pile of site policies and guidelines with it) and stayed more or less the same for a decade, being read by about as many people as you'd expect a pop-psych summary on the fifth most visited website to be read.
In the vein of "interesting things about autism that cause it to not be a failure", besides the obvious engineer and intelligence cases, is one about inability to distinguish fantasy from reality! Speaking from experience here as mildly autistic, i often have very very strong emotional reactions to fiction in a way that is similar to if those events happened in real life, and this is apparently quite common for aspies. This offers the interesting benefit that if a fictional story tries to relay an important message or lesson or moral, i'm unable to simply disregard is as "just a story", because it feels as emotionally important to me as if it happened irl. Thus, this aspie specific ability allows us to not just learn things, but truly experience things in ways that otherwise would only be possible through actual lived experiences. Of course, this comes with the tradeoff you may learn the wrong lessons from fiction, but it's an interesting power nonetheless.
This is speculation and probably not original, but it might be interesting to consider how e.g. autism fares now compared to in the past. Your mildly autistic Google engineer is probably doing quite well for themselves now, but might not have had the same opportunities before engineering was a thing. So certain genes that evolution was in the process of weeding out might not be selected again, or selected against as heavily, as societal incentives change, which also has implications as to which genes are considered failures by your previous definition.
"In fact, just being born in March raises your autism risk a bit"
The linked article seems to refer to being *conceived* in March as higher risk.
-Someone born in March
Maybe it is neither a failure nor a tradeoff, maybe it is an evolutionary beneficial trait of evolution itself (let's call this meta-evolution).
Let's imagine that you're a drug researcher trying to determine the best dosage of a particular drug. You'd probably run an AB test with a range of different dosages and pick the one that makes the best tradeoff in terms of desired effect vs undesirable side-effects. But what happens if you expect environmental changes to alter the sensitivity to this drug over time and you want to keep your drug performing optimally. One thing that you could do is give a narrow spectrum of different dosages to different patients and continuously monitor differences in outcomes to tune the mean dosage that you use as your reference. You trade off optimal performance now against adaptability to future changes.
It's entirely possible that evolvability is itself an important evolutionary trait and what you're seeing here is the result of Mother Nature's (extremely unethical) AB testing framework. You see a spectrum of different levels of psychiatric disorder because without such a spectrum, evolution becomes unresponsive to changing environmental conditions and we are highly evolved in favor of being adaptive.
Autism is a spectrum disorder, right? It seems to me that ought to be an important part of considering whether or not its's maladaptive.
Autism is a spectrum disorder, right? It seems to me that ought to be an important part of considering whether or not its's maladaptive.
"If autistics are too quick to mistake signal for noise, schizophrenics are quick to mistake noise for signal."
Shouldn't this be the other way around? (autistics see noise and think it's signal, and viceversa).
This entire writeup reminds me of Clans of the Alphane Moon by Philip K. Dick. (Who, as far as I can tell, really hated people in mania.)
I think you did a disservice here by with the explanation of tradeoffs involved in autism. For instance, I know some autistic people (particularly autistic AFAB people) who are definitely autistic, and definitely very able to function in society, but also are very creative and good at gestalt thinking. There is a tradeoff involved, but simplifying it to "people person" versus "average person" versus "stereotypical aspie engineer" reduces that tradeoff to the point of meaninglessness, in addition to making it seem more like a binary (trinary?).
In addition, what role is the "autism spectrum" in this explanation? Does it match with the triangle up and right from the "stereotypical aspie engineer" dot?
This is a really good read, with lots of "oh, that's me!" moments. Just making this comment because I wanted to offer some more concrete appreciation than a like.
This is almost certainly just...motivated reasoning. I want this to be true, so I'm going to ask.
My mother has schizophrenia -- the very bad, refusing-medicine, wanting-to-kill-people, writing-screeds-to-the CIA kind -- and so did my great-aunt on the other side of the family, less severely. My husband's dad had a psychotic episode about a decade ago that he seemed to white-knuckle himself out of, somehow. I'm terrified that my daughter or my future children will be schizophrenic, since the genetic dice seem loaded against them.
BUT! My husband and I aren't like this: "creative, attuned to interesting patterns, and charismatic, but also a bit odd and superstitious." If anything, I'm "responsible and perfectionist, but has trouble letting things go," and he is chill and was a math major.
Soooo does that mean we maybe don't have those pesky psychosis-causing genes?
P.S. I first found SSC when I was trying to understand my mom's illness. Scott was incredibly kind and generous with his expertise when I, a total stranger, sent him a frantic email asking "will my children be nuts??"
> I think most psychiatric disorders exist on a spectrum from mostly-tradeoff to mostly-failure (what we might call "high-functioning" and "low-functioning" versions of the same phenotype).
For completeness's sake, there's hypothesis that anorexia (and maybe other eating disorders) fit this category to. Gusinger 2003 (https://pubmed.ncbi.nlm.nih.gov/14599241/) argues that there's a tradeoff between "being able to tolerate starvation in order to flee famine" and the pathological "self-induced starvation and hyperactivity" behaviors that constitute anorexia. (This came up in a comment thread or two on some SSC articles)
I think Guisinger's formulation is a little "just-so," but I think there's something to this. The same behaviors can be induced in laboratory animals (see this article, which was recently cited in the Subreddit https://pubmed.ncbi.nlm.nih.gov/22231828/).
> I used to work in the business district of San Francisco, meaning I got to see a lot of very high-functioning people with mental disorders.
Why is it a disorder then?
>In fact, just being conceived in March raises your autism risk a bit - it means you were in an especially vulnerable developmental stage duriing flu season!
So, is it plausible that flu shots (for parents-to-be) actually reduce autism?
>Overall genes that increase risk for ADHD decrease risk for OCD, and vice versa, suggesting that at least one advantage of ADHD is lowering OCD risk.
With this assumption how would one account for people who have both ADHD-like tendencies (forgetfulness, distractibility, impulsivity) and OCD tendencies (obsessive checking and the like)? Would these contradict the "opposite" model, or could one explain the obsessive behaviors as a coping method for ADHD symptoms ("I have to check to make sure I turned the stove off because I know how easily I forget!")?
It doesn't make sense to say a trait X is Y% genetic. And if you're talking about heritability, the measure is population dependent and has little to do with genetic determinism.
uhhhh section numbering go I,II,IV, whaaaat?
One of the things I struggle with is the notion that schizotypy correlates with charisma, especially because schizotypal personality disorder is also characterized with strange communication patterns, inappropriate responses to social situations, inappropriate dress and presentation. Is there a study correlating schizotypy with charisma?
Sometimes I find your writing wonderfully nuanced and thought-provoking, other times the oversimplifications and unexamined assumptions perplex me.
In the first place I don't know whether to admire or be appalled at the attempt to tackle this complicated subject in these tiny snippets, but this is our culture and I guess at least someone is talking about it.
But treating mental illness as a homogeneous entity as you sometimes do is problematic. There is zero evidence that they all share anything but some degree of inconvenience to the sufferer. And as you rightly pointed out in your last post, far different etiologies can lead to the exact same symptoms (and similar etiologies to different symptoms). Sometimes the "health" in mental health is what appears to be central. Other times it appears to be an understandable (if unchosen) reaction to problems of living. Other times based on continuing once adaptive behaviors that are no longer so. And so on.
And it presumes that those who do not fit current DSM diagnoses are healthy, that there are no illnesses disguised by their social acceptabilty. I don't know how anyone with any experience of human beings can deny that something akin to the pathology of normalcy affects a significant minority, and is far worse and more sickly a thing than many DSM diagnoses. The foundation of your analysis ad hoc embraces psychiatric conventions of thought and "mental illness" when the discipline doesn't even define what is meant by mental or mind, much less any other foundational term.
And the conclusion that this heterogeneous concept of "mental illness" across the board is "mostly" a sign of failure genes is on its face questionable. Over half the population has a mental illness at some point. So... 40% or so of the population is a genetic maladaptation?
That a significant number of syndromes are a result of mostly maladaptive genes and flat out bad environment factors is incontestable. But a) these are overcounted due to methodological biases towards genetic explanations and b) that they look so similar to "trade-off" based syndromes is a sign of the deep problems of the current system.
And your discussion doesn't mention the most obvious source of trade-offs: sensitivity. The same exact "bad genes" implicated in some psychiatric symptoms are also responsible for thriving. In other words, the presence of some gene variants mean that, depending on the environment, they will lead to either sickness or its opposite [1].
And this corresponds precisely with everything we know about basic biology. You can have all the genes variants for something and not get it. We are district in the animal kingdom in the degree to which we are "programmed" to be highly biologically attuned to culture, such that it becomes sedimented in the body. That nature is nurture and vice versa, to a significant extent.
But your overwhelming focus seems to be on a simplistic treatment of genes and obviously unhealthy very early experiences at the behest of all else. I don't even know where to begin responding to a phrase like "it looks like evolution has been trying hard to get rid of them". I'm hoping that's just lazy phrasing.
The problem with the overemphasis on the "failures" is that a) individuals amenable to turning their "syndrome" into an advantageous trade-off are under-emphasized and b) the degree to which any psychiatric symptoms can be damaging, regardless of whether it meets a DSM threshold, is brushed under the table.
For an example of the latter, the vast majority of individuals with any ADHD symptoms to any extent are impaired, and impaired to the extent to which they display those symptoms [2]. The common sense conclusion from this is that ADHD is not an illness, its symptoms are signs that they are already sick. And full-fledged ADHD is simply an extreme form of that illness (that often emerges as consequence of disadvantage [3])
Successful individuals are simply those who go from being ADHD - having no ability to distance themselves from their worst excesses - to having ADHD. It's a developmental achievement in those without the worst genes/environ combo, and involves calming/channeling its co-occuring cyclothymic temperament [4] into an asset.
But does every syndrome operate like this? Of course not. But when you lump all these random syndromes together, it obscures more than it reveals.
1. http://dx.doi.org/10.1038/mp.2016.114 & https://doi.org/10.3390/genes11111248 2. https://doi.org/10.1016/j.psychres.2019.02.003 & https://doi.org/10.1111/j.1469-7610.2011.02467.x 3. https://doi.org/10.1016/j.ssmph.2020.100548 4. https://doi.org/10.1016/j.jad.2012.04.034
In my experience, a lot of lawyers have ADHD (especially in the field of criminal law). This makes sense for similar reasons to why ER doctors would have ADHD - if you have trouble doing things before the last minute, doing long-term followup, and have a whole bunch of coping strategies around organizing chaos already, those fields are a much better fit than other areas of the appropriate profession would be.
The problem, of course, is that the difficulties of the disorder are still very much there. They just get in the way less.
On tradeoff is suggested by how all minorities usually fare in a society, with most of them repressed, but with one or a small group rising to the top by the very virtue of their unique perspective. The outlier of the outlier might be king for having something like two eyes.
Re: Autism, I've been very persuaded by this review: https://www.pnas.org/content/111/42/15220
"Autism as a disorder of prediction"
It beautifully ties together all the different symptoms of autism into the predictive processing model, postulating that autistic phenotypes arise from weak inputs from top-down predictions, at different time scales and layers.* I really recommend this read.
It also fits the observation that many of these risk genes are in synaptic proteins, and their phenotypes often include abnormal dendritic spine physiology.
I think it fits very well into what you say. To the extent that autism is a tradeoff, it's a tradeoff of stronger bottom-up input (and weaker top-down priors) which makes your thinking less biased, lets you see things others ignore, and lets you forgo the intuitive wrong answers and actually do the math. The price is that you can't adjust yourself to sensory input, have trouble predicting where the flying ball will go, and find it hard to not take someone's words literally.
To the extent that autism is a failure, it's a failure of what Clark calls "precision-based weighing" - i.e. deciding accurately how much weight to give to bottom-up input vs top-down priors. And I suspect that schizophrenia shares much of that in common, and may well be a similar failure and the opposite tradeoff.
*At one level, it makes a person unable to adjust away a constant sensory input, so they stay highly sensitive to it and can't ignore. It also makes them really want a predictable, self-caused sensory input, i.e. stimming. At even shorter timescales, it leads to the very idiosyncratic motor problems. At longer timescales, it makes a person unable to use priors on human behavior when interpreting other people.
When u write a long comment after reading the first part of the post thinking it's going off in another direction only to find out that's the point Scott was making :-( :-)
But I would add that there is another kind of trade-off to be considered. The general performance/tolerance for errors trade-off that happens when you overclock a computer. It might not be that autisticy genes are ever helpful but the overclocked brains that tend to make the brightest engineers just have turned down the tolerance for that kind of screwup way down.
You made me really uncomfortable with your justice system analogies. They seem so inapt that it's hard for me to accept the surrounding discussion.
> In a country biased towards finding people innocent, it only takes a tiny failure to let a murderer go free. In a country biased towards finding people guilty, it would take a huge failure, or many different failures in parallel.
For this to be correct, it would require a bizarre measurement of failure size.
The problem is this: the simplest way to let a murderer go free is to convict an innocent person instead. Problem solved! Nobody gets murdered twice; as soon as you convict one murderer, you're done. But the "we find everyone guilty" system will do this *all the time*. They do it on purpose! That's the whole point of a "we find everyone guilty" policy.
In order for this mistake to require a huge failure, suspecting the wrong person would have to be considered a huge failure. But that can't be right. For just one example, if a married woman dies, her husband is automatically suspect. I don't think that's a mistake. I don't think many people at all think that's a mistake. But it would certainly be a mistake if her husband was automatically guilty!
> Go back to the two ways a justice system can go wrong. First, it sets too many guilty people free. Second, it convicts too many innocent people. If you ask which tradeoffs cause each problem, you'll find they're opposites.
Well, no. There are more than two ways the justice system can go wrong. This perspective only makes sense if you consider "the justice system" to consist of nothing but court trials, into which suspects are deposited by some ineffable force. And in the larger perspective, we see that these two problems are not opposites at all! Every time you convict an innocent person, you automatically let a guilty person go free! Increasing the one metric necessarily increases the other! (Though not vice versa.)
There's a mathematical formalism called "detection theory" which separates bias (what is called "tradeoff" here) and sensitivity (the inverse of "failure" here), starting only from success or error rates to two opposite problem types. Its results are usually represented on a ROC curve, i.e. a graph with false alarm rate on the x axis and hit rate on the y axis; this can easily be adapted to put "rate of using precise cognition when the situation doesn't call for it" on the x asis and "rate of using precise cognition when the situation doesn't call for it" on the y axis. Let me refer to this image I've drawn in the usual style of the blog: https://imgur.com/a/7ucjzUA
There's a few differences of interpretation between this formulation and Scott's: Detection theory says you can have super-high bias even with high sensitivity, unlike Scott's model which tightens the possibilities for "tradeoff" as we approach no failure. It's just that bias matters less and less as you reach towards infinite sensitivity (no failure). At infinite sensitivity, bias is indeterminate, not zero. (At infinite bias on either side, sensitivity is also indeterminate.)
Interestingly, though Scott applies the model primarily to mental health here, he briefly touches on its applicability to social issues such as crime. One thing I've noticed is that an enormous amount of energy, political will, debate, advocacy, resources and thought are spent on where we set the bias term, and discussions that touch on the sensitivity term (anti-failure) tend to be rare and difficult to find. Presumably this is because moving the bias term is relatively easy and improving the sensitivity of anything can be very difficult, but this still tends to annoy me at times.
And I think it's worth adding that the neurodiversity movement often conflates several things. Is x a desireable trait or not, should we search for a cure for x and should society avoid regarding x as a disability or negative so it's not viewed with pity and we don't treat those with x less seriously. I mean it's really hard for any trait to be totally neutral.
I mean consider being very short. Every person I've met under 5'1" wishes they were taller, being that short is almost purely a negative in today's society (even if it just keeps you from reaching high objects) yet we don't view being short as a disease or medical deficiency and it might not be worth looking for a cure for being only 5'. In other cases, like ADHD it's probably good to find a treatment (even if on net mild ADHD is beneficial it's even better to be able to turn off) but we don't view those who have it with much pity or treat their opinions as less important. But that's only possible because they (we) seem mostly like everyone else.
Rhetorically and psychologically it's really hard to say yes, all things considered, severe autism makes life worse and a cure would be good but I do think want you to tag sufferers with all the negative attitudes we associate with mental disability.
I mean hell, even if your a high functioning autistic person surely it would be great to have at least a temporary treatment letting you switch on the way normals see the world. I mean even you don't think it's better surely it would be useful to experience it did say 6 months.
I find it interesting ironic that it's pretty stereotypically high functioning autistic to ignore all the social implications and overtones of saying we should try to cure autism and/or avoid carrying fetuses with high risk of autism to term because autistic individuals have lower expected utility in our existing society.
Not saying that is true just find it ironic that the kind of claim which often provokes the most negative response by the autism nuerodiversity folks is something that I've found to be frequently voiced by individuals with a degree of Aspergers.
I find it fascinating to apply the tradeoff/failure framework to hiring, and was just having a debate with my household about this - the similarities are striking (innocent : guilty :: bad hire : good hire, where you want to catch the good hires and let the bad hires go, but it's unclear if a hiring process that is failing to hire as many engineers as it wants to get is failing or just very far on the spectrum).
We know, or at least strongly suspect, that a minority of people experience weird immune reactions to viral infection, some of which include psychiatric effects. This is getting much more attention now, with covid, but it has long been associated with viral infection in general, because it is more about the defective immune response than a feature of a specific virus.
I read a lot of 19c medical history, so I can compare today with a time in which everyone got a lot of infections, which they had to fight off without help. Obviously, we'd expect to see a lot more of this, and while I can't prove it, there seems to have been an accepted connection between viral illness and psychiatric symptoms in adults and teens, and many probable cases.
But there are surprisingly few mentions of anything resembling low-functioning childhood autism, and even fewer relating such a condition to a recent illness or illness during pregnancy. Both early childhood illness and illness during pregnancy were quite common at the time. This puzzles me. If it is at all related to immune function, which the March birth thing suggests, we should have seen more of it even well in to the 20th century--the 1918 flu caused a lot of weird psychiatric reactions, and other diseases were prevalent. Early 20th century records do have many more cases, under the name childhood schizophrenia, but few seem to have made any connection to viral infection. Maybe it's just not properly documented, but it strikes me as a real puzzle. Everything attributed to covid I'm used to reading about in historical accounts, and seem to be well-established immune responses. I think we've massively underrated the role of viruses in triggering certain medical conditions, and the wide variety of possible symptoms. But I don't see many descriptions matching childhood autism. Just something I wonder about, as there are other reasons to think there is a link between autism and some sort of immune dysfunction.
It seems that this is an evolutionary version of a bias-variance tradeoff, which makes sense, since in any optimization system you'll find a tradeoff like this. And as with the examples here, the system can in theory minimize either, but typically there will be some of both.
Two comments... first, anyone here read Marco del Giudice's "Evolutionary Psychopathology: A Unified Approach"? He tackles the problem of the ontology of psychiatric conditions head on, with an attempt at a global hypothesis involving life strategies (fast and slow) and many other things. I skimmed through it, but it's way over my head! Any opinions?
Second, about depression, I've read a few articles lately converging on the idea that depression is a response to a distress. Quote from Johann Hari: "This pain you are feeling is not a pathology. It’s not crazy. It is a signal that your natural psychological needs are not being met. It is a form of grief – for yourself, and for the culture you live in going so wrong." Any opinions?
"They seem to be general genes for having mental disorders, with a wide variety of negative effects"
I've been told that there is a correlation between these "general genes for having mental disorders" and homosexuality. Is that true?
>which I’m tempted to cynically attribute to their being less likely to remember to use contraception
Would it be considered out of line (or just more holistic medical practice than anyone can currently be arsed to provide) to talk to your ADHD patients about long-acting reversible contraception? That feels like a no brainier.
You might be interested in my latest paper with Abdel, which estimates natural selection for a bunch of polygenic scores: https://ideas.repec.org/p/uea/ueaeco/2021-02.html
Japan has 99% conviction rate: https://www.journals.uchicago.edu/doi/abs/10.1086/468111?journalCode=jls
You talk about autism and Schizophrenia probably being under negative selection by evolution. I'd be interested in seeing a list of other things under negative selection. Maybe once we have genetic engineering, we could consider skipping the evolution and eliminating those traits straight away (in some cases, obviously evolution doesn't have our values).
Much in the same way as we might do the genome "spell checking" thing (https://slatestarcodex.com/2016/05/04/myers-race-car-versus-the-general-fitness-factor/)
I was taught that ADHD and OCD were frequently co-morbid conditions, rather than inversely related, such as when we see those two and Tics associated (Venn diagram ish) especially in Tourette’s. Also higher than optimal stimulants can convert inattention in ADHD intro laser focus on tasks vs unmediated ADHD translates into laser focused on repetitive play.
The “failure” hypothesis for genetics is bound to be mathematically sensical vs the “failure” hypothesis for epigenetic factors, eg in utero virus exposure with schizophrenia.
Finally I was taught that the generic advantage of ADHD ( see Barkley) was related to something like the “twitch” reflex in animals, conferring advantage to humanoids coexisting with predator animals, where scanning and distraction, impulsivity were advantages or at least had a valued place in the village.
"Almost all de novo mutations are bad"
Is this even true? Don't most of them just have no effect? Although it would hold true if we limit ourselves to ones that DO have an effect.
I've heard that historically many that religious figures have personalities that sound a lot like schizophrenia. It's possible the tradeoffs were more worthwhile in the past.
Also it's possible that since many of these iq and creativity tests were designed with neurotypicals in mind that you would need different tests to get a good measure of the tradeoffs.
"a certain especially bad type of de novo mutation was found in 20% of autistic patients and only 10% of their siblings"
It's not clear from this sentence whether it's 10% of all autistic patients' siblings, or only of those that had the mutation.
Is this "trade-off" between autism and schizophrenia very different from the trade-off between up-down processing and down-up processing, as explained in your article on Perceptual Control Theory? https://slatestarcodex.com/2017/09/06/predictive-processing-and-perceptual-control/
A good read, presenting an appealing big picture, but I am having trouble believing the main point can be new. (Literature search, anybody?)
I'm not terribly well versed on this, but doesn't the idea that a gene is selecting for its propagation, not your propagation come into play? That is, if a clan produces 90% normies that have babies, 5% autists that make great tools but no babies, and 5% OCDers that make great use those tools but no babies, that the clan could outcompete a clan with 100% normies who make babies, ok tools, and ok use of the tools? So each of the first can's normie parents produce 5% autist and 5% OCD babies, ensuring a continual supply of them without the autist / OCD parents producing any?
Something I am much more familiar with is that I think the role of autist in security is understated in this article. For 1, OH MY GOD SOMEONE IS BREAKING THE RULLLLEES! and for 2, don't high functioning autists, definitionally, develop formalized rules for behavioral and social interactions instead of just intuiting them? Well, if you want to develop a security procedure, you need to generate formalized rules. Actually, there's another - things like threat analysis require you to really, really, really understand all the bits and pieces of a system, and then identify all the ways you could subvert them. Bruce Scnierer calls this the Security Mindset. And it requires reducing people to things and behaviors to rules. (I spent 10 years in this field, fwiw.)
(TLDR: Some of this disorders might arise from something called genomic imprinting, which can arise when different parts of the genome have different evolutionary "interests" https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1558-5646.2010.01015.x)
The tradeoff is presented here as being due to intraorganismal constraints (e.g., I can't maximise size and speed at the same time). Some recent theory has suggested that things like autism and psychosis arise from intragenomic conflict, leading to genomic imprinting. The tradeoff here is between genes from different sources (e.g. maternal and paternal), which have different evolutionary interests (e.g. the paternal genes have lower expected relatedness to siblings than maternal genes). I haven't thought how about how it might change your interpretation (if at all), but it does change what patterns we expect to see, both in behaviour and in traces of selection in the genome. It also might mean that some of the things which look like failures at the genetic level, aren't just deleterious mutations in the traditional sense, but imprinted genes that increase their own fitness at the expense of the largest co-replicon (the organism).
Here's another analogy for both trade-offs and failures working together to create a psychiatric condition.
When amateurs discuss tank design (as in war machines not water tanks,) the first thing that they reference is usually the gun-size/armor-thickness/mobility trade-off; Generally, the large the gun, the thicker the armor and the faster the tank, the more useful it is, (In the real world, there are other factors that are also important but for this analogy this is good enough.) Bigger guns and more armor make a tank more formidable, but all things being equal, they also make it heavier and therefore slower. A tank that has a bigger gun can engage an enemy at a longer range, which is a big advantage, but a tank which is faster and more maneuverable has a better chance of choosing the ground in which an engagement occurs, also a big advantage. So there's an obvious trade-off between the size of the vehicle and it's mobility.
The thing is, other factors also affect the mobility of a vehicle, such as the quality of the suspension system and the power the main engine. So if you improve the engineering of a vehicle by using better suspension or a better engine, you can get around the weight trade-off and back the tank faster without sacrificing armament or armor. This is why the M4 Sherman (WWII era tank) weighed about 30 tons, but could only do about 30 mph, while the modern M1 Abrams often weighs about 60 tons but can go 45 mph, so twice as heavy but half-again as fast. The modern vehicle has a lot of engineering improvements over the old vehicle.
So I think here we have a model for two different kind of underlying trait. If an "autistic" tank was a tank that was too slow to be useful on a battlefield, then you could achieve that either by accidentally making the gun or armor too heavy, ie, taking a trade-off too far, or by making a mistake in the design of the suspension system or engine, ie, a failure without any upside.
It could be that many if not most of the genes that cause psychiatric conditions are usually design failures but a few of them are trade-offs that occasionally go too far, or combine with one of the design failures in a tragic way to cause an identifiable psychiatric condition. Perhaps the presence of a trade-off creates a situation where the presence failures are more likely or more of a problem. So maybe someone is close to the edge for having autism because of the presence of a lot of genes for smartness (assuming that's actual trade-off) but it's only the presence of a failure that puts them over the edge, but without the trade-off the failure wouldn't have been enough to put them over the edge? Or maybe sometimes it's just one cause or the other?
"Psychopaths, local government. "
Wouldn't it be more accurate to say senior corporate management? IIRC psychopathy is far more common in senior corporate leadership than it is in among the general public.
Bipolar disorder can be formulated with some relation to the trade-off/failure paradigm but I think it's more a failure that happens to have some trade-offs, complicated by the difficult trade-offs of the system it is a failure in.
To dramatically oversimplify, before engaging in any action some calculation of whether it is "worth it," i.e. are the cost and risk of an action outweighed by the probable benefit. This is very complex process and mistakes either way can be harmful, although false positives tend to be more harmful than false negatives (e.g. "I can take on that lion" is more likely to get you killed than "I can't catch that turtle", at least in the short run).
In depression, the bias is towards "not worth it" usually due to an overestimation of risks and cost with an underestimation of benefit. This tends towards a state of inaction and arguably many of the symptoms of depression such as anhedonia and decreased energy and concentration are meant to keep the individual from engaging in actions. In mania there is the opposite bias, with dramatic overestimations of ability and underestimations of danger. Everything seems like a good idea and is acted upon. This is very dangerous state and untreated manic episodes have a very high mortality rate (~20% IIRC).
I would make the argument that the mood states are not the causes but the RESULTS of the biases, i.e. the system that assesses risk/reward comes to a general conclusion about which way to bias the conclusions and sets the mood accordingly (similar to setting the DEFCON level). In bipolar disorder, there is a tendency of this system to introduce biases inappropriately. There is sometimes benefit to this, particularly when one is mildly on the manic part of the spectrum since they may have more energy and creativity without too much impairment in judgement.
In this formulation, I think bipolar disorder is best characterized as a failure in the decision assessment system, related to the underlying difficulty in optimizing it. There hjust appen to be some trade-offs because the decision assessment system is inherently about trade-offs.
"Instead of insisting that psychiatric conditions cannot possibly be bad for anybody and it's always just social prejudice all the way down, they would be better off admitting that this can be true for some people and false for others."
... they do admit this tho? you are asking them to do a thing they already do.
I've considered similar multifactor models and one of the advantages is in explaining the somewhat paradoxical trend of an increasingly even sex ratio as autism severity increases, since it might be the case that only some of the factors show sex differences and that those factors might be the ones less related to general brain goodness.
Using for instance a 2-factor model where factor 1 is sexually dimorphic (higher in males) and unrelated to mental deficiency whereas factor 2 is not sexually dimorphic and related to mental deficiency, one would expect to find that generally speaking those who’s autism was primarily caused by factor 1 would be highly disproportionally male, whereas those who’s autism was primarily caused by factor 2, and who would be by dint of their mental deficiency deemed a more severe case, would be more even in terms of sex ratio.
Obligatory sf reference for the combined psychiatric clinic / job placement agency: clans of the alfane moon ;-)
I think the counterbalancing advantage of depression is conflict avoidance and lack of aggression. This used to be adaptive when at the bottom of a social hierarchy. I call it the slavery response.
On the observation around genetic factors being evolutionarily selected against, is the following possible:
There are a subset of these genetic factors which are not mistake factors as you have described them per se, but which still influence in which direction the bias goes in your graph. You could view them as having no particular advantage either way until they reach extreme levels - a more creative person may fill one role in society, a more precise person another but both might find reproductive success that way. Such a factor would have either no effect on an individual's fitness or a negative one, and thus would be expected to be slowly selected out of the gene pool, but then quite slowly because most instances of it do not harm reproductive success. I guess ultimately I am questioning whether the axes on your graph can realistically be thought of as orthogonal?
(Then on another level, a group or species might be expected to be more successful with individuals of both these types than a species without either, so that confuses my question further : ( Sorry! )
I think addiction may be thought of as an attractor state as well - a failure of self regulating. On that note - so can compulsive behaviors in OCD. The difference might be that compulsions generally alleviate negative emotions, while addictive substances or behaviors may also enhance positive emotions (though eventually one of the hallmarks of substance use disorders is that the 'alleviate negative emotions' part becomes dominant)
"In the modern day, people with genes for psychiatric disorders tend to have fewer, rather than more children than people without those genes - except in the case of ADHD, which I'm tempted to cynically attribute to them being less likely to remember to use contraception."
I doubt you meant this seriously, but if it were the case that ADHD-prone people were less likely to remember to use contraception, and that resulted in having more children, *that would be an evolutionary advantage* and therefore would support ADHD as a tradeoff, or even just a beneficial trait that hasn't yet fixed.
Is type 2 diabetes primarily due to genetic tradeoffs or failures? While there are certainly genetic predispositions to type 2 diabetes, I expect most would agree that evolutionary mismatch is a more productive framing. In the environment of evolutionary adaptation, our access to refined sugars was far more limited than it is today. There may not have been significant tradeoffs nor should a predisposition to type 2 diabetes in our contemporary environment be regarded as a genetic "failure" in the context of our environment of evolutionary adaptation.
What is most remarkable about this post is that the evolutionary mismatch framing is an entirely invisible category. While it is fair to speculate regarding the extent to which any particular psychiatric condition is due to genetic tradeoff, genetic failure, or evolutionary mismatch (i.e. the genetic sources of a predisposition did not, or were less likely to, result in pathology in the environment of evolutionary adaptation), that third category is not even considered here.
Here I provide a number of independent strands of evidence suggestive of the idea that evolutionary mismatch is a causal factor in adolescent dysfunction, including some psychiatric conditions,
https://flowidealism.medium.com/evolutionary-mismatch-as-a-causal-factor-in-adolescent-dysfunction-and-mental-illness-d235cc85584
In it I quote Harvard's Daniel Lieberman, a leading expert in evolutionary mismatch diseases, who claims,
“There is good reason to believe that modern environments contribute to a sizable percentage of mental illnesses, such as anxiety and depressive disorders.”
In addition to anxiety and depression, he also adds ADHD, eating disorders, chronic insomnia, and OCD as mental illnesses whose modern prevalence is likely due in part to an evolutionary mismatch.
Aaron Blaisdell, Editor-in-Chief of the Journal of Evolution and Health, UCLA Department of Psychology and Neuroscience, left a comment,
"What a great essay! I just came across it now, and it really resonates with my own thinking on this topic."
I generally avoid citing authorities, but the fact that such a significant category (i.e. evolutionary mismatch) was entirely neglected is striking.
The "autism as overly-precise thinking" characterization is fascinating, but feels narrow. I wonder if a developmental perspective could broaden it. Consider what a tendency for overly-precise cognition does to a toddler. All of their interactions with the world are colored by this tendency, each interaction shapes their minds and shapes future interactions, feeding back into itself.
And then you have the fact that schizophrenia doesn't generally manifest until early/mid twenties. What's up with that?
At the risk of oversimplifying, I prefer to view ADHD as a relative underdevelopment of the prefrontal cortex (as opposed to, say, OCD's overactive cingulate gyrus).
It's highly comorbid with OCD, bipolar disorder, and autism among other things, and I believe that the genes that lead to these comorbidities end up in ADHD families because they give them the high anxiety or stimulation necessary to cope with it.
"If, as Badcock and Crespi claim, schizophrenia is the reverse of autism, it might involve being too imprecise - too willing to declare the identity of unlike data - too quick to pattern-match. If autistics are too quick to mistake signal for noise, schizophrenics are quick to mistake noise for signal. In a well-functioning brain, this makes them creative and socially adept; in a poorly-functioning brain that is constantly getting things wrong, they connect everything to everything else and it's a mess.
I don't know if this is quite right. But it would reconcile two seemingly contradictory sets of findings."
Personality research backs this up: The personality trait most associated with aesthetic receptivity and artistic creativity, Openness to Experience, seems to be partially based on a system called latent inhibition. This is an unconscious process by which your brain essentially decides what associations are real and what are fanciful nonsense. High inhibition, and the associations have to be pretty concrete to count. The more it's loosened up the more upside you gain from being attracted to tenuous associations that may have a point, but your chance of making Type 1 errors also increases. High openness is associated with various measures of apophenia, and apophenia is a risk factor for full blown break-from-reality psychosis.
I think the loosened association is why things like poetry and song lyrics sometimes feel like they have something there, even if it doesn't quite make sense, and other times just inane bullshit. They tickle our apophenic, pattern-confirming tendencies. The best stuff are things that you wouldn't write off the bat yourself, but are still "in range", so to speak.
Wouldn't it make sense that neurodiversity can be adaptive at the population level, even if particular traits are often negative for individuals in particular situations?
A
Maybe this is a little off-topic, but I want to explain why I'm pro-neurodiversity. The argument about whether autism is a good thing or not is actually completely moot, because there is no autism cure, or even a way of turning low-functioning autistics into high-functioning ones, and there won't be for the foreseeable future. The primary effects of the search for a cure have been wasted life savings and kids traumatized by abusive "therapies," with some unpredictable effects like the resurgence of disease caused by the anti-vaxx movement. I support neurodiversity even for severely disabled autistics because it's the only practical, humane thing.
Also, every theory I've seen for the cause of autism seems oversimplified, including this one. My theory is that autism will result from any major, low-level difference in how the brain processes data, which is why autistics are so diverse.
I feel like this post might have been a bit improved by quoting liberally from Evolutionary Psychopathology, particularly some of the first sections.
I think the model being used by some of the people in neurodiversity (and, at the very least, some of the people in my university when I did a subject on this) goes something like this.
Autism is basically a switch. Most people have it turned off. Some people have it turned on (genetically or epigenetically; I seem to recall something about certain methylation patterns having 100% correlation with autism).
Due to the fact that most people have the switch turned off, having high mutational load will correlate with having the switch turned on. High mutational load is bad, because it means a lot of switches will get flipped, and most of those switches are in one position at fixation for a good reason; typically this means low intelligence. However, the correlation is not 100%; people can inherit the switch being turned on from a parent, or can have it flipped by freak accident at low mutational load (or can have high mutational load and *not* have this particular switch flipped).
The neurodiversity people, or at least the neurodiversity people I agree with, think that having the specific autism switch flipped is not bad and shouldn't be eugenically removed from the population. In this model, cutting down on mutational load (somehow) would stop new low-functioning autistics from being born, but have little effect on high-functioning autistics (at least, in the short-term; in the longer-term if there is significant selection against HFA the lack of replenishment via mutation would cause a decline in numbers). That is to say, it would target the failure and leave the tradeoff.
I may have been lied to. Maybe taxometrically the distinction they fed us between low-functioning and high-functioning autism is not real; I don't know. I'm just trying to sketch in the argument as I understand it.
Evolution is nature's way of tinkering. It's not even a blunt instrument, much less a scalpel (or a carefully knapped flint).
Thank you for this writeup. It perfectly summarizes how I feel about characterizing autism as 'neurodiversity' and I'll link to this rather than trying to re-explain it myself in the future.
Is it possible that 'the genes which put you at risk for autism' are actually the old versions, and what we're seeing is 'genes which lower risk of autism' haven't hit 100% penetration in the general population yet?
Also in terms of tradeoffs, isn't there strong evidence of tradeoffs in terms of blindness vs schizophrenia and depression vs epilepsy? It's possible that our overdeveloped vision system damages our executive functioning, due to the accident of them happening next to each other in the brain and competing for territory over the same neurons, leading to a tradeoff and possibly explaining the bizarre observation that nobody who was born stone blind has ever been schizophrenic. This makes me wonder about a hypothetical where someone invents a compound which can be injected into the eyeball which completely clouds vision until the appropriate enzyme is injected at which point it clears up completely. If there's a newborn whose parents are both severely schizophrenic thus putting them at very high risk would it be ethical to use this to deprive them of vision until the age of 2 at which point hopefully their risk of schizophrenia has gone way down?
On depression vs epilepsy it's possible that our huge oversized brains are at risk for epilepsy for the simple reason that a malfunction in any part of the brain can cause a seizure of the whole brain, so the bigger your brain the more likely they are, so we have things to slow down our brain functioning which cause problems when there's too much of them or the epilepsy they're supposed to counter isn't there. The utility of shock therapy seems to hint in this direction.
Long-time reader, first-time commenter here. Big fan!
Speaking as someone who maybe possesses - ok definitely maybe - one or more of the mentioned traits, the old saying, “Bi-polar is the worst, it’s awesome.” is mostly true. I also have ADHD and five kids, and so the part about contraception is also accurate. The thing i like about your framing is the attempt to contextualize intrinsic and extrinsic attributes. Schizophrenia, generally not pleasant, but there are some potential extrinsic benefits to the heard. Still, extrapolating extrinsic values is difficult without a longview and reasonable longitudinal data. Yet we make trade-offs. You allude to these trade-offs in the triangle thingy, but I wonder if you could provide more thoughts on these extrinsic pros and cons? A triangle would probably be insufficient. Which would displease many sales types. But they’re probably not your core readers.
You speak of evolution as though it has intention, is purposeful. Perhaps this is what you believe. Perhaps it is just a convenient anthropomorphism. But I think that evolutionary trends are emergent functions, inferred by our observation of the laws of nature and the balance of probabilities playing out. This is important because it is easy for us to project the values that shape the sense we make of evolution onto the evolutionary process itself. I believe the value of evolutionary outcomes is determined by humans, not evolution. Evolution is value neutral.
These have been fascinating, cheers
Maybe DNA has a finite information capacity. This is the Occam's razor.
The example of the bipolar coder is very good. Contemporary society demands a lot of redundance out of people and being capable of that is “functioning.” Whereas I think some conditions contribute very specific things that society might only need once in a while, but it makes a really big difference to get it, leading to positive notoriety for that individual and lots of descendants. Historically, some cultures have been willing to completely subsidize the existence of people who bring them something they appreciate - monks, wandering minstrels, hermits, an author who produces two or three incredible books in a short period but lives 80 years. Being in their presence allows observation about aspects of life one wouldn’t know otherwise.
Right now we like intelligence and “functioning.” Past societies may have valued the ability to stay awake for a few days hunting, but only needed that once a month, the fortunate individual being free to do whatever in between. Evolutionary pressures are changing.
I am not sure that the extreme states would have ever been valued, but the trade offs they generate might have been more valued in differently structured societies.
I know a dental hygienist and I think this person may be a sociopath, but they can stay focused for hours on a procedure with a wincing patient and blood and drool everywhere. Narrow trade off but very valuable.
Do you think that when psychotherapy is used to treat autism-spectrum disorders or schizoid disorders, the goal is usually try to move people horizontally on this chart (trying to adjust their prior between gestalt and precise cognition), or downward (trying to improve meta-strategies so as to just be more intelligent overall), or both? Or does it depend on the type of therapy?
Maybe someone has made this point already, but I am getting a slight sense that there's a confusion between fitness and 'good' in the standard English meaning here. Take autistic body language, i.e. stimming by rocking gently, as I do. It's plausible that genes that make you more likely to stim are fitness-damaging, because it looks weird to people, and things that look weird are sexually off-putting. But there's nothing inherently bad or bodily or mental function-impairing or painful about stimming. It's just not the norm. More relevantly there's no reason to think that 'makes you more likely to have stereotypical autistic strengths that would make you a google engineer' and 'fitness-damaging' are mutually exclusive categories. A gene could even in theory be fitness-damaging *because* it did the former. I.e. suppose for example that a gene made you more likely to have a particular 'cold-blooded rationalist' type of personality. It's really not that implausie that it would both make you better at certain kinds of intellectual labour, and overall decrease fitness in an evolutionary sense because most people find that personality-type Not Hot.
I don't doubt that many genes associated with autism really are bad in an everyday sense for the same reason they are fitness negative though, so don't mistake this for an attack on the whole post or a dogmatic assertion of the sort of neurodiversity view on which nothing about autism is bad except how neurotypicals treat us.
Thanks for writing this Scott - I've often felt that the high-functioning end of the autistic spectrum was a strengths and weaknesses model (i.e. trade-offs) but also that the low functioning end is a weakness only condition (i.e. a failure mode). I think that many in the neuro-diversity movement are high functioning individuals, who have a point, but perhaps don't realise that there are non-verbal institutionalised people who are not reading and posting on the internet about neuro-diverse culture. This makes sense to me both from knowing a lot of high functioning geeks in my personal life, and also from my work as a Clinical Psychologist in neuropsychology
If autism was solely evolution's way of giving us Dirac, I'd say it's a great deal.
Enjoyed this post quite a bit - it’s my first intro to your work. I believe these interactions of human behavior (genetically moderated or not) with the environment are even more recursive and polymorphic than you’ve suggested. It’s probably complex enough that it would require some form of advanced complexity science to even begin interpreting it. There are just too many nodes. The same problem plagues many modern fields of study.
I do admire your observations and logical thinking! And I thank you for that! It is the initial step when questioning, well, everything.
However, I could only go so far reading your article. While I do appreciate it, I had so many scruples trying to get through each paragraph, until I had to stop.
I feel obliged to mention a few of my own thoughts. (Go figure, behavior is so predictable.) I found your thoughts very self-centered, as were the ancient Greeks, and everyone else, say, earlier than the 1950s? That belief being that humans are the center of the universe; the world was created for “us” rather than the other way around. In another sense, it’s sort of a “creationist” way of thinking.
Do you rememberer, for example, the creationist who proclaimed on television that the banana is curved in such a way for the human mouth to consume? Anyways, that’s the image I couldn’t escape while reading your rant… I’ll call it a rant, like Henry Miller’s Tropic of Cancer, because it isn’t scientific, but it is the impetus for questioning, which I am thankful for.
I’m being very lazy right now, and I do apologize! But, when you discuss psychiatric conditions as if the conditions themselves are conscious, thoughtful beings -the ways in which humans are thoughtful- is troubling. Evolution isn’t conscious like you and I are. (I apologize for any non-humans reading this article, but this is strictly a human experience!)
I’m not going to bore everyone with the philosophy, (unless you ask specifically, then I’d love to discuss!), or biology (please inquire!), or break down everything I know about every single psychiatric condition, but I feel I should at least mention one or two conditions, for reference.. perhaps two that were mentioned in the article? Ugh, how predictable of me.
Psychiatric conditions are sometimes the brain’s attempt to protect us. Think of PTSD, for example. If one suffers from ongoing childhood abuse, their brain’s priority is survival, well, because survival was a constant threat. In this case, whether or not the abused’s brain cognitively understands that they are safe, they cannot convince their body and heart otherwise. Think of it as, say, a celebrity death match where the celebrities are the limbic system and the frontal lobe. It doesn’t matter what your cognition thinks about the situation you are in, your heart will start pounding, you will sweat, and you will “go offline.”
In other words, it’s impossible for the frontal lobe to convince the limbic system otherwise. It’s debilitating, thus a disorder. The “hilarious” part, (trying to stay on par with your sense of humor), is that an ignorant person witnessing a person with PTSD experiencing an acute, distressing event will misinterpret their experience as, simply, “geez, that person is dumb.”
Other psychiatric conditions occur during brain development. I was overcome with joy that you recognized that environmental factors also play a part during development; although, you did call these factors “failures.” (I’m not familiar with these hypotheses you mention. Anyways.)
Yes, neuronal developmental also depends on environmental factors, accordingly. For example, toxins such as alcohol penetrate the blood brain barrier of a developing fetus (an environmental factor) and can cause very significant damage, in addition to genetic factors. The neural tube (brain and spinal cord) develop in the first month of pregnancy. Often, people don’t even know they’re pregnant at that time.
With this is mind, I’d encourage you to research the difference between critical and sensitive periods. It’s actually quite a complicated topic for a novice. This is where specialists come in.
Different researches focus and measure different, umm, “things.” These measurements are known as a dependent variable. Intelligence, for instance, in academia, must be defined by the researcher. Dualities such as “dumb” and “smart” are not acceptable.
As you can see, I’ve already written to much and I doubt anyone will read this.
So, last sentence here: The most difficult part about reading your rant was, “Why the ***k am I wasting my time on this?” And then I remembered. You are an intelligent person with logical thoughts, and there is merit in that!
Please, send me your inquiries.
The trade-off between gestalt vs precision thinking reminds me of the distractable vs hyperfocused states commonly experienced by ADHD people. In the case of ADHD it's generally thought of as an attention regulation problem which has to be clunkily coped with by overcorrecting too far to either one side or the other. Maybe there's something similar going on in autism vs schizophrenia, where the mechanism that regulates which type of thinking to use is defective, so the brain settles into overusing one or the other?
I think disabled activists would offer a few additions to the statements made above. A major argument in the disability rights movement is that folks with disabilities suffer largely because society does not accommodate them through attitudes of prejudice, or through lack of consideration for the range of human needs (building stairs instead of ramps everyone can use), and treats disability as an exception, rather than the eventual rule (for all of us, as we age, have accidents, etc.). So in this sense, in many cases, there is not "high functioning" vs. "low functioning" autism, but there are attitudes and barriers that are greater for those with more differences than there are for those with fewer and for those living under more or less restrictive social norms. Many people with autism who would be considered "low functioning" at some point in their lives make a connection that allows more access to the world around them (Temple Grandin is an example of this, mostly nonverbal as a child, had violent episodes, now a leader in the field of animal behavior). In my experiences working with people with significant disabilities for the last 20 years, I would say very few of them need to be institutionalized so they "don't bite their own fingers off." In fact, most of the problems I see people with disabilities face are socially created.
This article does a great job with showing how some “failure states” are real, objective things. All of your articles are excellent. This one is no exception and has given me tools for navigating my own thoughts on my own disorder. Thank you again.
But something was bugging me, and I couldn’t solve it until I got to the “Hypothetical Perfect Person.”
The objective failure states – problems in the womb, seizures, birth environment issues, etc. are obviously factors. And you’ve given proof that (some of) these traits are being selected against.
But the conclusions drawn here actually prove more than I think you intended – and more than I think you believe. Falling in various places on the trade-off scale is some species of failure – in the example you give, if you tend to be more precise or gestalt, you become more failure-prone. You can only reach certain “low failure” levels if you are close enough to the hypothetical idea between precision and gestalt cognition because otherwise you will be less likely to choose the correct cognitive algorithm.
There isn’t a hypothetical perfect person, and there isn’t one “correct” brain algorithm for any given problem. A single problem can have multiple “good” outcomes. In fact, a single problem can have outcomes that are “good” on entirely different axes, providing value that may not even be perceived as value to someone with different priorities.
I mention this because in my experience the working world is trying to standardize humans the same way we standardize McDonalds’ hamburgers. It’s true that in Seattle a precision-thinker on the autism spectrum may be a successful programmer. In Arkansas he’s likely to be a “seriously don’t hire that guy, he’s weird.” And there’s more Arkansas than Seattle in the world.
I’ve never bitten on the neurodivergence movement or the opposition to it, and this article finally gives me the words to explain why – some neurodivergence is just plain bad, and some is not. But the conclusion then treats the other kind, the trade-offs, as though they are deviations from a “correct” norm. And having people at all points along those trade-off spectrums is an unmitigated *good.* It will allow you to find value that people on other points on the spectrum could not see.
This all leaves me wondering to what extent my diagnosis and the diagnosis of others I’ve known are based on failures that are just plain bad, and to what extent it’s based on just deviating from the norm on various tradeoff scales. Because if we’re pathologizing “being different from the standard worker,” that has a lot of very worrying implications.
This is well-written, obviously, but not such a profound concept. The thesis is essentially: An evolutionary flawed genotype can produce distinct, even behaviorally opposed, evolutionarily unfit phenotypes. The best example is cancer. Oncogenes or tumor suppressor genes can either have gain of function (GOF) mutations or loss of function (LOF) mutations. Let's say p53 (tumor suppressor gene) has an LOF mutation. The organism would develop tumors and expire. Interestingly, there is an asymmetry in a germ cell GOF tumor suppressor mutation in that the organism would not develop at all as the cell replication cycle would be inhibited. But if crisper-cas9 was used to insert a mutated GOF p53 in an adult organism, that organism would be less fit because it could not make enough stem cells (blood cells, GI epithelium, hair, etc.) to function optimally. That is, it would phenotypically resemble a cancer patient on chemotherapy. The same is true for psychiatric illness. Mutating the same genes can result in distinct, even opposing, phenotypes. This piece conjured in me the work of Carl Jung. Many psychiatrists have armchair diagnosed Jung with "good psychosis." I tend to agree with Scott that balance provides, in general, the highest fitness over the long term. However, at the very top - which is what people tend to focus on - are the outliers who are usually unbalanced in some way but are able to wield their mania, or ADHD, or autism to perform at the highest of levels. We must also leave room for the counter hypothesis though. That as psychiatrists we only see those with decompensated OCD or schizophrenia and that there are many individuals who thrive with these "mental illnesses" that simply do not make their way to the ward. Thus resulting in a bias. I do not think this is the case, but this alternative hypothesis, that many Jungs and Nashes are out there under the radar crushing it, is still possible.
Worth noting that what Scott describes as a "failure" for the individual (i.e. a "bad" de novo mutation or prenatal infection) is still a manifestation of a tradeoff in a broader systems sense. Thanks to population-level selective pressures, DNA proofreading is not so accurate that the species can never benefit from random mutations, but it's accurate enough that every child isn't born with disabling disease. Prenatal immunity is aggressive enough to protect the fetus from most infections, but not so aggressive that it attacks fetal tissue. Even the "failures" of his justice system example are just reflections of how a hypothetical society has decided to balance investments between their justice systems and other worthy societal goals. In many cases, the "tradeoff" that is being optimized for may not be obvious to us from our perspective of limited biological understanding. And I don't make this point just to be nitpicky -- tinkering with balances we don't understand (and may not even recognize as tradeoffs in the first place) can sometimes backfire.
How much do social mores play into definitions of psychiatric disorders? If ADHD for most of human history was a desirable trait (hyper-focused "hunters" would be a huge advantage), then is it fair to call it a disorder because societal expectations have changed fairly rapidly on the evolutionary time scale? Is our modern context ("civil society") penalizing traits that 10,000 years ago were necessary and vital?
On the topic of creativity and mental illness/personality these articles might be of interest-
https://www.theatlantic.com/science/archive/2016/02/myth-of-the-neurotic-creative/471447/
https://blogs.scientificamerican.com/beautiful-minds/the-real-link-between-creativity-and-mental-illness/
It would be really cool if you would write an article about creativity in the future though too, cause your insights and depth of research into these topics are really appreciated
The conceptual tradeoff model wouldn't apply to mental illness, as it can't be successfully used in somatic medicine. What's the trade-off of emphysema, diabetes, or coronary thrombosis? Disease is a harmful dysfunction and debilitating distress, not a zero-sum game. It's very natural and human to look for the silver lining; however, nature is neither generous nor fair. Evolution cares little about individual suffering; it favors the survival of the whole species and is willing to sacrifice a certain percentage of individuals (to diseases, suffering, and death) to achieve reproductive success.
Also, I would not support the idea that schizophrenia and autism are on the same continuum. I see them orthogonal or, perhaps, operating on independent dimensions. By analogy, both the brakes and engine could fail and affect the car's movement, but they are not on the opposite sides and "the reverse" of each other.
Terminology question:
Does Scott mean "allele" basically every place he says "gene" in this post?
For example, "Evolution is imperfect, so some bad genes manage to slip through." We're really talking about bad alleles, aren't we? People have (mostly?) the same genes; but there may be good or bad alleles. Right?
What does it look like if you have mostly-tradeoff versions of both autism AND schizophrenia?
To understand the origin of mental illness (etiology) one might look at the evolution, adaptability and procreation of the species. To understand the mechanics of psychiatric diseases (pathology) one should look at the working and malfunctions of the brain. Treatment of mental diseases consists of the restoration of function and relief of debilitating distress and often doesn't have much to do with either etiology or pathology. It's a mistake, in my opinion, to conflate all these.
I'm a high-functioning autistic mathematician, and although you could say that it's because I'm precise, I think most people who succeed in math do so because of their ability to see patterns and make generalizations from specific examples. We're definitely precise in how we describe these patterns and how we prove them, but I'm just hesitant about your triangle which puts pattern-matching on the opposite side of autism. I would love to hear from other mathematicians about this!
I think there's another possibility that you're overlooking with regards to your personal experience:
Namely, that it isn't that these things are tradeoffs, but rather, that these downsides are more severe in some places than others.
In other words, normal function is better. But people who have some sort of problem which is NOT a problem (or not as much of a problem) in one field will disproportinately end up in that field, because that field feels comfortable to them (it is finally a place where their problem ISN'T a problem, or isn't as much of one).
This will give you exactly the impression you got - that these things are somehow advantageous sometimes - but instead of actually being advantageous, they're just less disadvantageous sometimes.
Humans have an innate sense of fairness. People seem to think that life is a balanced RPG system, and are dangerously biased towards such ideas.
People who have these issues are going to be biased towards thinking that their issues are helping them in some way, rather than viewing them as a hinderance to their ability to function. And outside observers will likely see it the same way.
This makes the "it's just straight up bad" explanation much more likely.
Even the existence of beneficial genes that have associations with negative outcomes sometimes can be explained in this way.
For example, a stronger immune system is a good thing. But if you get some deletorious gene that causes you to have an auto-immune disorder, having a good immune system makes it much worse. But the actual original gene has no downside - the downside is a separate mutation that gets selected against.
And there's also the possibility that the beneficial gene allows the subject to exist at all with deletorious mutations. If you have some gene that boosts your intelligence massively, and then some other gene that screws up your brain, in most people, it would render you totally non-functional and you die - but in the person who has the huge intelligence boost, you end up at least marginally functional, and so get diagnosed with autism.
It could also be that the effect is not on the individual but on their *parent*. For example, smarter parents may be more likely to be able to keep defective defective children alive. Autistic people are much more likely to engage in pica, among other things, and autistic children have a significantly higher disease burden. So genes for high intelligence may appear more often in autistic children because we only test living children for autism. A more intelligent parent may also be more likely to recognize that there is something wrong with their child and seek out a diagnosis, which would also result in children with more smart genes ending up disproportionately likely to be diagnosed with autism.
I'm not really sure I understand the logic of differentiating between trade-offs and failures for mental disorders. I would see the tradeoffs discussed here as *part* of the failures; e.g. increased intelligence in autism is a failure as it allocates too much energy and processing power to a certain type of rigid intelligence, decreasing the amount available for other things that promote quality of life. The same for the other examples listed - they may be more successful at a particular job because of the trade-off, but if you consider life satisfaction or well-being on the whole, it would be a net negative, wouldn't it?
I came across this paper about a general factor of psychopathology dubbed p, akin to the general factor of intelligence g: https://ajp.psychiatryonline.org/doi/pdf/10.1176/appi.ajp.2018.17121383
I was immediately reminded of this factor being reflective of the 'failure' axis, but the authors don't really mention this as a potential explanation. Instead, they consider that p might reflect a diffuse unpleasant affective state, poor impulse control over emotions, deficits in intellectual function-, or generally disordered thought processes, but these could ALL be true and STILL originate from accumulated early life environmental stressors that mean your brain just doesn't work like it's supposed to. Maybe you could start a conversation with them about this.
I'm trying to wrap my head around the question of how well the "failure-tradeoff triangle" maps onto the pretty-well-explored ideas of precision-recall and sensitivity-specificity from domains like machine learning. https://en.wikipedia.org/wiki/Precision_and_recall
I don't know anyone outside of AA who uses the 'alcoholism as a disease' model because there are so many other reasonable explanations for why people get addicted (if we view nicotine/heroin/cocaine/X drug addiction as a result of nicotine/heroin/cocaine/X drug being addictive, why do we view alcoholism as special?)
Typo: "duriing" in "developmental stage duriing flu season"
Genes vs Environment: the first portion of the article focuses on genes as primary factors contributing to some mental illnesses. However, neither the scientific literature at this time, nor clinical experience, (in my perspective) lend much support to the idea of genes being responsible directly for these illnesses, certainly in the vast, vast majority of cases. "Genes load the gun, environment pulls the trigger," as some say. Why no thought on environmental factors? Make no mistake, I would be thrilled if mental illnesses could be addressed with genetic approaches, but the likelihood that this is either fundamentally possible, nor feasible, is very slim. Something about the philosophy here is troubling. Why is environment broadly, and environmental factors specifically, not deserving of greater attention?