Bryan Caplan  

Genetic Determinism vs. Parental Irrelevantism

Is Iran a Threat?... Dynamic Free Markets...
I think that parents affect their kids in lots of ways.  Parents have big effects on religious affiliation and political party, small effects on many other traits, and a clear effect on the quality of the parent-child relationship.  And those are just long-run effects within vaguely normal, First World families.  In the short-run, and outside the vaguely normal First World range, parents matter far more.  Still, if someone were to caricature my position as "parental irrelevantism," I would take no offense.  It's an inaccurate summary of my position, but not wildly inaccurate.  I really do think that parents' influence on kids' long-run outcomes is greatly over-rated.  Calling me a "parental irrelevantist" is arguably a useful simplification of my actual, subtler position.

Unfortunately, when people caricature me, they usually try to pin a totally inaccurate label on me: "genetic determinist."  I am not a genetic determinist.  Not even close.  The simplest twin study conclusively disproves genetic determinism.  Identical twins are genetically identical, but they're not identical for any complex trait.  Identical twins don't have the same lifespan, the same IQ, the same happiness, the same friends, or the same income - even when they're raised together.  Since genetic determinism predicts that identical twins will be identical for all traits, genetic determinism is demonstrably false.

Not only am I not a genetic determinist; I'm not a determinist of any kind.  The empirical evidence in favor of my own free will is overwhelming: Introspection reveals free will to me during my every waking moment.  There are many reasons why identical twins aren't phenotypically identical.  But free will has to be the most overlooked.  As I write in Selfish Reasons to Have More Kids:
I doubt that scientists will ever account for my sons' differences, because I think their primary source is free will. Despite genes, despite family, despite everything, human beings always have choices--and when we can make different choices, we often do. Some choices are moment-to-moment: To keep working or give up, lie or tell the truth, abandon or defend your views on immigration policy. Other choices are cumulative: You can't change your weight, education, or income by snapping your fingers, but in the long run they depend on diet, study, and effort--all of which you're free to choose.
I really don't mind caricatures.  I'm a fan of simplification, and what's good for the goose is good for the gander.  But if you're going to caricature me, please give your caricature a kernel of truth.

Happy Father's Day!

Comments and Sharing

COMMENTS (19 to date)
PrometheeFeu writes:

I think free-will is most often ignored because from the point of view of studying human behavior, it is entirely uninteresting. As students of human behaviors, what is interesting to us is what we can study and predict. It is how we can take the unknown, the unpredictable or the incomprehensible and turn it into the known, the predictable and the understood. The bounds of free-will are what we can study scientifically. Free will itself is 1-R^2: A topic best left to philosophy.

A. writes:

So I assume you are not a physical determinist either, then you think there is some kind of soul? Or are you trying to advocate some type of compatibilism?

Pandaemoni writes:

I think the obvious alternative answer to why you sons are different is "sensitivity to initial conditions". If determinism is an accurate model of the human mind, then I think the mind is still obviously a (mathematically) complex system, and would almost certainly display at least a high degree of sensitivity to initial conditions.

No matter how similarly two children are raised, there will always be minor differences that would lead them down different paths, even if free will is an illusion.

Chris Koresko writes:

Bryan, I have a question that's probably a little off-topic, but I'm really interested in this and hope you will be patient.

I haven't read the parenting literature your book is based on, but from your descriptions of it, it sounds like it implicitly assumes that parenting styles are rigidly determined by the parents. But it seems to me on the basis of my own experience as a father that parenting style is an emergent phenomenon: it is a relationship between parents and children, with each side adapting to the other, and the rapid development of the child preventing it from ever reaching equilibrium.

So if you look at what happens to a pair of twins raised in different families, and accept that for genetic reasons there will be correlations between the personalities and abilities of the twins, then the parenting styles they experience will be correlated as well. That would tend to reduce the strength of the signal the study is looking for, i.e., it will make parenting look less important than it is.


James writes:


LessWrong has a series of posts from Eliezer Yudkowsky that "dissolves" the concept of free will. I think he does the job pretty well. It's not a full-length sequence, just 10 posts. I'm genuinely interested in your response to them, if you have the time and interest.

(I tried to follow links until I got to a page fully explaining your reasoning in favor of free will, but this page is the only one I saw, and it doesn't work.)

J Storrs Hall writes:

A robot's brain contains a model of the world, used for predicting the outcomes of actions. It uses the predictions to plan its own actions. The world model must contain a model of the robot. But this self-model can't be deterministic because it would need the outputs of the decision-making process it is a part of. Thus the self-model has a different user interface than the rest of the world model, and that's what we experience as free will (the rest of the world model being deterministic or random).

H/t Drew McDermott

Tracy W writes:

Chris Koresko - the literature that Brian is drawing on does look at the interaction between parenting styles and child behaviour. What scientists have attempted to do is to discover ways in which parenting behaviour varies independently of the child's genetic starting point - eg looking at birth order, or single parents or single-sex couples versus traditional couples. This doesn't find much effect. (Divorce does seem to have an effect, but as divorce often involves the kids moving home, the research hasn't been done yet into whether it's the divorce itself, or the change in the kids' peer groups as a result).

I don't quite understand what you mean though by this correlation making parenting looking less important than it is. If parents normally adjust their parenting style to the child, then that's what they do. How does that make their parenting styles any more effective, or important, once scientists get around to measuring it?

Chris Koresko writes:

Tracy W I don't quite understand what you mean though by this correlation making parenting looking less important than it is.

Thanks for responding to my question. Sorry it wasn't more clear.

My point was that if the way a child is raised depends on the character of the child, then moving the child from home to home will have less impact on his upbringing than you'd expect from looking at the ex ante character of the parents. So if you make a plot of some variable which describes an aspect of the outcome against some independent variable related to the parents, you will tend to underestimate the role the parents played in producing the outcome.

The question is whether this makes sense, and if so, whether the studies Bryan relies on take this effect into account.

jb writes:

I can't find any rational reason to believe in free will. I can believe in unpredictable will (i.e. no one within this universe can correctly predict my actions), but not free will.

I wish I could believe, (Heh) it would be much better than feeling like I'm a self-deluded chain-reaction of events. But the same logic that pushes me away from religion pushes me away from magical thinking of any kind.

But we're not talking about free will, so much as parenting. There is one, extremely well documented way that parenting has significant effect on a child's future: Abuse. A sufficiently abusive (even just emotionally abusive) parent will obliterate any chance that a child will have of living a normal life.

Can we postulate an argument then - does this pendulum only swing one way? Is there any equivalent (but as of yet unknown) mechanism by which parents can be sufficiently nurturing to improve their child's chances of having a normal life (or even better than normal)?

I don't know that it exists, but it's interesting to think about.

Tracy W writes:

I'm sorry Chris, but I still don't understand the distinction you are making. I follow your point that parenting styles depend in part on the nature of the child, but I don't see how that results in underestimating the role that parents played in producing the outcome. Not accounting for this interaction overestimates the role - Judith Harris points out that a study that finds that "Parents who know where their teenagers are, and what they are doing, have teenagers who are less likely to be doing drugs and having sex than parents who don't", might well mean only that "teenagers who are doing drugs and having sex are less likely to tell their parents where they are, and what they are doing." Or "children whose mothers smacked them are more likely to be badly behaved" could be "children who are badly behaved are more likely to be smacked".

So as far as I can tell, the effect you are talking about would lead to overestimating of parenting importance.

Tracy W writes:

Is there any equivalent (but as of yet unknown) mechanism by which parents can be sufficiently nurturing to improve their child's chances of having a normal life (or even better than normal)?

Well, human babies are utterly dependent, so all of us who have survived to adulthood have had at least one person who supplied sufficient parenting, starting with food, water, etc, to improve their child's chances of having a normal life. (There's the odd case of children apparently raised by wolves, so perhaps "person" shouldn't be limited to humans).

Then there are the parents who sacrifice greatly for the chance to get their children to a better life, such as the Jews who sent their children away out of the reach of Nazi Germany, or the parents who leave their friends and families to immigrate to places where they maybe don't even speak the language. Looking at the life expectancies of say Russian Jews and their descendants in the USA in the 20th century, versus Russians in the Soviet Union, these immigrants mostly did improve their children's chances of having a normal, or better-than-normal life. Although the mechanisms by which this happened are reasonably well-known, of course.

eccdogg writes:

Fort those interested here is Dr Caplan on Free Will.

Evan writes:
Unfortunately, when people caricature me, they usually try to pin a totally inaccurate label on me: "genetic determinist." I am not a genetic determinist. Not even close.
I think the reason a lot of people do this is because when we think of "environment" in our culture when talking about child development it is usually regarded as a synonym for "parenting." People often don't consider that environment also includes your friends, your culture, etc. So when you say "Parents are less relevant than people think" it gets translated in people's minds to "Environment is less important that people think," even though you didn't say that.
Clay writes:

Bryan claims that "genetic determinist" is "a totally inaccurate label", but says things like this:

"A vast literature finds that heredity is not merely part of the reason for family resemblance, but virtually the whole story."

Of course, he doesn't believe that genetics are completely deterministic, but he does believe that they have a very strong causual effect on long run individual outcomes. "genetic determinist" isn't the pefect label and it would be nice to offer a superior alternative, but it's mostly accurate.

Mr. Econtarian writes:

"Free will" is kind of a cop-out. Non-genetic effects in twin studies imply some kind of experiential difference. Which is of course true, no two twins exist in the same location in space, they don't have the same name, etc., thus they have to have experiential differences.

It is possible for very small differences to be amplified by the non-linear processing of the brain. My anecdotal experience is that twins (identical or not) will sometimes amplify small behavioral differences between them to achieve a higher feeling of individuality. One will insist on a green dress and the other will want purple, for example. This difference could have started out as a very small thing (maybe one saw something she liked that was green), but over time it evolves into a deep internalized clothing color preference.

What I think it fairer to say is that research shows experiential differences in genetically identical twins. While parenting strategy may be a part of that experiential difference 1) parenting strategy could likely play a small role compared to the rest of the universe of experiential differences and 2) it is unclear whether parenting strategy tends to achieve its desired outcomes.

Anyone who has trained a dog knows that there is always a high risk of training undesired behaviors when you try to train for desired behaviors. No doubt a lot of parenting strategy of children often does the same thing.

Floccina writes:

Yeah kids don't listen to teachers and to parents only so much. I think of that some times when I hear that in school in North Korea and Cuba tell the kids that Americans are devils.

Noah Yetter writes:

The empirical evidence in favor of my own free will is overwhelming: Introspection reveals free will to me during my every waking moment.

This is not evidence. In fact what the evidence shows is that we act independently of our consciousness, then subsequently our consciousness experiences a rationalization of that action.

At any rate simple physics is all that's needed to show that free will is highly unlikely at best. Cause and effect leaves no room for choice.

Philo writes:

"Introspection reveals free will to me during my every waking moment." Does introspection also reveal to you that the "free will" you are finding in yourself is incompatible with determinism, or do you have to do a bit of philosophizing to discover that? I expect the latter, in which case there may be doubt about how well you did that philosophizing. (The doubter need not impugn your introspective faculty.)

Tim writes:

I love the movie Gattaca.

Tagline: There is no gene for the human spirit.

Comments for this entry have been closed
Return to top