Steven Pinker's "probabilistic" genes

It makes absolutely no sense that I should take on Steven Pinker on the subject of genetics. A distinguished Harvard professor of psychology, best-selling writer and popular lecturer, Pinker is one of the best-known and presumably most-respected scientists in the world. 

I have a B.A. in English from some college in Rhode Island.

Pinker has been a director of Cognitive Neuroscience at MIT and a few years ago was named by Time magazine as one of the 100 most influential people in the world. He has twice been a finalist for the Pulitzer Prize. 

I was recently invited on an internet radio show to talk about my obsession with Jerry Garcia.

On stage in a debate with Pinker, I would last about 90 seconds. And even that would be a privilege, since my chance of ever getting to debate him directly are about as good as becoming an Olympic archer. (I don't arch).

And yet, I am simply compelled by what I know to write that there is something fundamentally misleading -- or at least missing -- about Pinker's characterization of how genes affect us. 

To be sure, he gets most of it right. But in his 8000-word article published in yesterday's New York Times Magazine, he never once acknowledges gene-environment interaction or epigenetics, and gives no real attention to, or explanation of, genetic expression. He refers once to the "probabilistic" nature of genes, but does not begin to explain how a gene can lead to probability as opposed to a certainty.

Maybe I'm reading his piece unfairly -- please tell me if you think so -- but to my eye he puts by far the strongest emphasis on the unrelenting power of genes in the formation of human traits. Yes, the piece is filled with small caveats. But this paragraph leaves, I think, an overwhelming impression:

"The most prominent finding of behavioral genetics has been summarized by the psychologist Eric Turkheimer: “The nature-nurture debate is over. . . . All human behavioral traits are heritable.” By this he meant that a substantial fraction of the variation among individuals within a culture can be linked to variation in their genes. Whether you measure intelligence or personality, religiosity or political orientation, television watchingor cigarette smoking, the outcome is the same. Identical twins (who share all their genes) are more similar than fraternal twins (who share half their genes that vary among people). Biological siblings (who share half those genes too) are more similar than adopted siblings (who share no more genes than do strangers). And identical twins separated at birth and raised in different adoptive homes (who share their genes but not their environments) are uncannily similar."

Turkheimer did say that. The quote is correct. But Turkheimer has produced a study showing close to zero heritability with respect to some of these same traits. Heritability studies are frought with complexity and ambiguity. Pinker must know this and is choosing for some reason not to acknowledge it. The last line in the paragraph above is particularly problematic. I'm trying to choose my words carefully, but I would go so far as to say that Pinker's sentence is more false than true. Yes, some identical twins separated at birth have some uncanny similarities -- but many of them also turn out to have not actually been separated at birth. Others are not similar at all. 

I have none of Pinker's credentials, but I've read enough to say confidently that the word "heritable" should never be used so casually in a New York Times article, for the very simple reason that it does not mean what non-scientists think it means. Pinker's quick follow-up definition doesn't help. He allows the quote and the entire paragraph to leave the impression that scientists have proven beyond a shadow of a doubt that people inherit portions of their intelligence, personality, and religiosity. 

If you read his caveats elsewhere very carefully, he backtracks by saying that a lot of different elements go into it. But in the public explanation of science, it's all a matter of emphasis and Pinker's emphasis is clear. He's still spending most of his energy trying to fight against the blank-slaters who once-upon-a-time naively argued against any genetic influence whatever. We're past that argument now -- or should be. We know that genes influence everything but directly determine very little. We know that most traits, particularly complex character traits and abilities, are the consequence of development. From the moment of conception, genes interact with the environment. People are not born with fixed IQs. Talents are not determined by a genetic lottery. Each one of us is conceived with immense potential.

The nature-nurture debate has not been settled in favor of nurture, as Pinker implies. It's been rendered obsolete. 

  

Comments

[Jacob Mack]

Oh and I love your blog! I just stumbled upon it:)

[Jacob Mack]

The journal Nature and Scientific Aamerican late last year both highlighted the importance of understanding biological evolution and also de-emphasized the findings of evolutionary psychology as often not based on sound science. There are elements of truth or should I say evidence in both evo psychology and socio-biology, but there are severe limitations.

[Jacob Mack]

I do not as of yet have a PhD but my Biology and genetics background is considerable and my psychology/brain and behavior background is impeccable. As is my statistics background and Pinker cites research and conducted research that did not control for important confounfing variables. Also in true genetics textbooks and many in behavioral genetics textbooks hot of the press a far more conservative interpretation in made than Pinker's even when one can test for an average range of 40-50% and far less than 80% of behavior being accounted to genes. To the best of my knowledge Pinker does not well understand alternative splicing and epigentic changes resulting in RNA pre-translational modification and DNA nucleotide mutations, though he must have heard of them. Also, environmental influences greatly influence points of view and perspectives which in turn affects neural connections and long term potentiations of neron communication. Simply put how long and well and where neurons "learn" to talk a longer time together:) Statistical averaging of even 1000 twins in 2-4 studies is still not a large enough sample from a population when some of the outcomes are not consistent and I do not just mean outliers (high and low ends of individual results that can change the average or conatrdict a generalized finding) but to many cases that do not fit Pinker's model.

It is clear that genes play important roles in childhood and personality development, however, multiple robust studies come to contraictory conclusions like: genes are more important in personality development, genes are more important in development of intelligence or how well one does on an IQ test. The problem is most of the myopic interpretation of such studies are not robust and well evidenced and oftetimes the methods were biased from the get go. It does seem from my own analysis and most studies that a great poportion of persoanlity is based upon pre-natal, post natal and genetic factors, however, Pinker's assertion that parents have only a 10% effect on their chilren's personality or different but related coping skills, (well, he never actually mentions coping skills, but coping skills, over time do tend to become ingrained as a personality component)is unfounded unless he is cherry picking data and only data he likess. I would love to debate Pinker. I have read all of his books, seen all of his lectures available online and I have studied the man very carefully. He is no doubt, intelligent and educated... but I see his agenda. I am not saying it is wrong to be an atheist either, or agnostic, or even a moderate religious person:) I am saying Pinker himself with a high IQ and legendary status is using it to initimidate the competition and to further his own gain.

[tumor cells]

HI
there is so many genious in our time even when we dont see them the perfect example is of a great medic that same me to a great problem head tumour that i had it very deep in my head he same me for me he is a genius
thanks DR john beringer

[William]

Pinker has done this before. In one of his early books, "The Language Instinct" he minimizes evidence from a study of eye movements. He describes a study in which people did not notice the"dual meaning" provided by the syntax of a sentence, but instead read it in only one way because of their background knowledge of the subject matter (p. 215). My sense on reading this part of the book was that he didn't care for the idea that general knowledge might control understanding more than syntax so he tried to minimize the findings from this study. Perhaps his attitude would have been the same if the study had found the opposite, but I have my doubts.

[David Shenk]

Thanks for your remarks, everyone. DD, I see that you disagree with my characterization of Pinker's piece. Fair enough, but a few points:

First, just a tiny correction for anyone following this back and forth. The paragraph you are referring to is three paragraphs before the one I quoted it. Here is the graph you're referring to, I believe. I'm reprinting it so I can refer to it.

****
Nor should the scare word “determinism” get in the way of understanding our genetic roots. For some conditions, like Huntington’s disease, genetic determinism is simply correct: everyone with the defective gene who lives long enough will develop the condition. But for most other traits, any influence of the genes will be probabilistic. Having a version of a gene may change the odds, making you more or less likely to have a trait, all things being equal, but as we shall see, the actual outcome depends on a tangle of other circumstances as well.
****

I think it's incorrect to say that Pinker "decries" determinism in this paragraph or in any other graph in the piece. That's not to say he embraces it either -- and I never claimed in my initial post that he did. What he's specifically criticizing in the first line of the graph above is actually the scare-tactic use of the word determinism. I read a lot of material on genetics, including most of the stuff attacking genetic determinism, and I don't think any of these critics are using "determinism" as a scare word. I think it's simply the most accurate word they can come up with to characterize the old view of genetics. If someone suggests that X gene causes Y condition, what better word for that is there than "determinism"? So on that particular point of Pinker's, which is essentially a quick political attack on certain geneticists, I think he's off the mark.

Pinker then goes on to embrace the word determinism for some conditions -- he gives Huntington's disease as an example. I certainly don't disagree with his statement that "everyone with the defective gene who lives long enough will develop the condition." And in the context of discussing Huntington's disease, it's perfectly appropriate to use the shorthand of saying that the disease is "caused" by a defective gene. But I do find it odd that, in an 8000-word piece about genes and how they work, Pinker would go out of his way to embrace the concept of genetic determinism and say that it is sometimes "simply correct." In my view, it's not only misleading; it more importantly misses a huge opportunity to explain that genes really don't determine things on their own. They interact with other genes, and with the environment; they get turned on and off; they are one vital player in an inherently dynamic system. All of that is true even if, on occasion, a specific gene alteration often -- or even *always* -- leads to a specific disorder. (It needs to be pointed out here that Huntington's disease is not exactly the same disease each time; its symptoms and age of onset vary.)

My original complaint about Pinker's piece was not that he prints outright falsehoods, but that he leaves the wrong impression and fails to explain the most important ideas. Embrace of the word "heritable" is one example, and neglecting to give any explanation to the mechanism of "probabilistic" genes is another. He acknowledges it but doesn't explain why it is.

[dd]

He also says:

"But for most other traits, any influence of the genes will be probabilistic. Having a version of a gene may change the odds, making you more or less likely to have a trait, all things being equal, but as we shall see, the actual outcome depends on a tangle of other circumstances as well. "

This was in the paragraph preceding yours, in which he decries the deterministic view of genes.

[Cana]

I like this post. I totally agree that "Talents are not determined by a genetic lottery. Each one of us is conceived with immense potential."

However, we know that genes do have impact on one's life. Parents and educators need to find out a child's unique personality and find a way tailoring to his or her unique individuality in order to nurture his or her character traits and abilities towards success, and to accomplish this is not easy. I believe researchers, educators and parents will learn better and do better.

I think that the nature and nurture have very close and complex relationship, need us to learn more and explore more.

[Kevin Maloney]

David,
I was leafing through a back-copy of The Economist(June 21, 2008) and in the Science and Tech section they have a piece on how lifestyle changes appear to have an influence on the expression of genes - cancer genes to be specific.

The article was based upon Dr. Dean Ornish's recent study into how healthy living (low fat vegetarian diet, plenty of exercise, no smoking, etc) can 'switch-off' tumor-promoting genes (as many as 500 of them) and 'switch-on' cancer-fighting genes. The article gets into how they used gene-chip technology to look for a shift in RNA messengers. So, even though a patient may possess the genes (DNA) of a prostatic cancer patient, by altering external factors, you can exert control over the expression of the genes (ie: increasing or decreasing the production of corresponding RNA molecules) and ultimately you can exert control over the physiological (and/or pathological) changes (ie: the expression of tumor growth vs. tumor atrophy).

Though not directly related to the development of 'genius', with respect to the nature vs. nurture debate, it is, I think, very relevant to your research. I'd like your take.