The author of blueprint, Robert Plomin is an American psychologist, geneticist and neuroscientist and perhaps the most important voice, over many years, in the field of behavioral genetics. It is difficult today to imaging how scientifically taboo it was to study the genetics of human behavior after the racist horrors, bogus research and eugenics projects carried out by the Germans in the Nazi period. The field of behavioral genetics got off to a politically rocky beginning in the 1960s, but has gradually gained respectability, although some of its applications, particularly in the area of race, have been controversial (I would argue, misguided). The great achievement of the field is to show without any doubt that understanding human behavior must include the factor of genetic predispositions. Robert Plomin is to be admired for his contributions and his courage. What he writes deserves attention.

Two general points about the book. First, it is not written for specialists. For example, there is a chapter that clearly, and at a very basic level, summarizes the basics of DNA. And a major point of the book is a societal question of how to reconcile the idea of genetic influences upon behavior with beliefs in meritocracy, free will and others.  Second, ideas of innate or genetic differences among people, including human groupings, have a real potential to do harm if not discussed clearly and applied accurately. This is particularly true today as we witness a dangerous rise in racist and nationalist politics. These two points put a difficult burden upon Robert Plomin to write very clearly so that the reader does not misunderstand his conclusions and put them to wrong-headed use. I do not believe that Plomin has met this burden to state clearly some of the results and implications of behavioral research. I begin this review, therefore, with a discussion of what behavioral genetics is and what its key concept “heritability” means (and does not mean).

Behavioral genetics is famous for its studies of identical twins, a phenomenon that carries an immensely wide popular interest (who hasn’t stared shamelessly at the appearance of identical twins?). Plomin’s and others’ interest in identical and fraternal twins stems from the facts that that the former carry identical DNA and latter share 50% of DNA. This provides a basis for behavioral geneticists to measure the contribution of genetics to understanding human variabilities. (I ignore here the recent discussions of epigenetic change and (perhaps) inheritance.)

Take a little time with the three notes below and then the definition of behavioral genetics. Confusions about it have led to very common misunderstandings. I’ve italicized key words to focus upon.

First Note: Roughly, variability is a measurement of how widely different a set of traits is in a group. Think of the trait of the body height in a group of 500 New York City 25-year-olds males. The heights will vary widely because of the city’s ethnic variety. Contrast this with results from a more ethnically homogeneous city, for example Stockholm. (Variance is the specific measure of variability in behavioral genetics research, but that’s not important here.)

Second Note: The two causes of differences in levels of a trait (e.g., height) among people in a group are difference in genetics (nature, DNA) and differences in the environment (nurture, external factors) in the group. So, “Genetics + Environment produce Traits” (e.g., artistic ability, depression, weight, autism)

Third Note: Here is an example of the kind of question that behavioral genetics seeks to answer. Randomly select 500 first graders in a city and test each for the ability to draw, rated on a 1 – 100 scale. There will be a differences (variability) in the scores in the group. Why are the scores not all the same? Some children were born with an ability to draw well (DNA), and parents of some children draw with their children from an early age (Environment). So, there is genetic variability and environmental variability in the group of 500 that produces the variability in the ratings. Here’s the question: Which of the two factors, genetic or environmental variability, is greater and by how much?

Definition: Behavior genetics is the study of the percentage of the variability of a trait (drawing ability, height) in a group (fist graders, 30-year-old men) that is caused by the variability of genes (DNA) in that group.

The resulting percentage level (0 to 100%) is called the Heritability (or H²) of that trait in that group. The contribution of the group’s environmental variability to the variability of a trait in the group is calculated simply as 100% – H² (In other words, what’s left over.) There is no direct measurement procedure for the % of environmental variability’s effects on the variability for a trait. It is necessary to grasp that what is to be explained is not a trait, but a trait’s variability in a group. And what explains the trait’s variability is not genes, but genetic variability plus environmental variability. This is important because it is not easy to get from these variability results to talking about you or me as individuals.

Because we know the degree of sharing of DNA in identical twins, fraternal twins and unrelated people, these twins can be used to calculate H² for different traits in different groups of identical and fraternal twins. These results can then be generalized to groups with no twins.

Example: If you put two identically cloned tomato plants in different places in your garden, the variability of “tomato weight”(the genes’ expression) in the two places would be due 100% to environment, H² = 0%, since there was no variability of genes. Note. This does not say that DNA made no contribution to the tomato weight. Its zero contribution was to the variability of the weight. If you put two plants cloned to be different into two identical environments, the variability would be entirely due to DNA, H² = 100%. If you did the same for two differently cloned plants in different environments, the tomato weight variability would be due, in some proportions, to DNA and environment, 0 < H² < 1. That number, say it’s 60%, would depend upon the genetics of the plants and the quality of the soils. Change the soils (the environment) and the 60% H² number might well change.

Although the variability of a trait can be 100% genetic in the sense of H², there is no trait in a human person (or any individual living thing) that is 100% genetic. Every human trait requires an environment that renders it possible.

There are important implications of this definition of what behavioral genetics studies.

Heritablility (H²) It is a property of groups, not of individuals.  Neither you nor have an H² score for any of our traits. There is no way of determining as of yet what the % of my terrible stick figure drawings rating is the result of my DNA or my environment. The H² rating for a trait will often be very different in different groups. In a homogeneous society where people’s nutritional intake is very similar, differences in height will be more likely due to DNA differences than in a society with great variability in nutritional intake. Note again, this says nothing about to degree to which DNA contributes to the height of any individual in these society.

A second point. It is a very common error to think that if a trait is “genetic”, it will be more difficult to change than a trait that is “environmentally caused.”  So, it is erroneously thought, if we want to change a trait in a group that is heritable, say at 60%, we have only 40% of something (the trait’s environment) to work with. Here’s the British economist (Lord) Richard Layard (2003) discussing the difficulty of boosting social happiness, the kind of happiness or life satisfaction that gets measured and compared among countries. He writes about why this is difficult, “This is of course partly because of the strong role which our genes play in determining our temperament. Identical twins reared apart have happiness levels with a correlation as high as 0.5. But there is still plenty of room for nurture as well …” In other words, what’s accounted for by DNA cannot be changed whereas what’s accounted for by environmental factors can be changed. Even the great Stephen Pinker makes this error. See for both Layard and Pinker, my essay “Nature, Nurture, and Individual Change.” Behavior and Philosophy 34 (2006): 1- 17. But Layard makes another error.  He takes the heritability of 0.5, which refers to variabilities in groups and applies it to the power that DNA has individuals in the group.

Of course, some traits are more difficult than others to alter. But knowing only that one trait is strongly genetic and the other is strongly environmental tells us nothing about how difficult it is to alter the trait. (No surprise that I dislike the expression “wired in” to refer to largely genetic-based traits). Knowing that a trait in a group is highly heritable or that (what is different) a trait in an individual is inherited, tells us nothing about how difficult the trait is to alter. Our DNA often bestows us (in the proper environments) with wonderful effects for our lives, allowing us, and not other animals, to write poems, do math and grow food. And our DNA sometimes burdens us with tragic effects, Downs disease and Phenylketonuria (PKU) – both leading to retardation. In the above example both Downs and PKU are largely genetic. Downs has no known cure but PKU is easily cured by low a protein diet in a person’s youth. Having ten fingers is highly genetic but in some occupational environments (logging and meat packing) it is easily and often changed (by its environment). Similarly, a high H² for a trait in some group says nothing about how resistant to change the trait may be.   

I wrote above that (1) this book is not directed to a readership of the author’s peers but to one that is less prepared to understand his claims, and (2) he has not made his terms or the meaning of his numbers clear enough. What are his readers to think of these heritability numbers? 

Stomach ulcers  ………. 70%

          Autism          ……….    70%

          Reading Disability  ….. 60%

          School Achievement .. 60%

          General Intelligence … 50%

          Personality ………………. 40%

Do they mean: That there’s a 70% chance for a newborn to be autistic? No. That there’s a strong chance? No. That there’s a 70% chance that a child with autism in her family will be autistic? No. None of these follows from the numbers. Suppose the 70% became reduced to 50%, wouldn’t it be correct to think that the likelihood of one’s child’s being autistic is less than thought previously. No. Why not? Individuals do not have heritabilities for traits. Individuals may have genes that make a trait more or less likely to be expressed, but H² does not measure that.

In light of this, let’s look at some claims that Plomin makes, that I find either unclear or questionable:

Claims Page 5: Inherited DNA differences are the main reason we are who we are. … Estimates of genetic influence are called heritability … [which] … describes how much of the differences between individuals can be explained by inherited DNA differences. [This book] Blueprint is about what makes us different psychologically.

Page 26: “Heritability is central to this book because it indicates the extent to which DNA makes us who we are.”

Response: In this, it is too easy to read that inherited DNA are the main reason that we (you and I) are “who we are”. He states that the measure of influence that our DNA has over us is “heritability.” But this is not what heritability is. Heritability is a measure of the contribution of genetic variability to the variability of a trait in a group. This cannot be transferred to the power that my DNA has to influence me.  There is no evidence presented that DNA is the main reason that Plomin and you and I are who we are. But finally, buried within the above Plomin says of heritability, is that it “describes how much of the difference between individuals can be explained by inherited DNA differences. This is incomplete and vague, but closer to what heritability really is than talk of “making us who we are.”

Claim, Page 71: “We now know that genetics makes siblings 50 percent similar, which means that it also makes them 50% different.”

Response: The fact that 50% of the genes of fraternal siblings are identical does not make the siblings themselves 50% similar. In fact, it’s hard to know what it would mean for any two people to be 50% similar or different. Every person has an indefinitely large number of characteristics. List them all and half will be the same as a brother’s? Of course not. Plomin is transferring what are group properties only to individuals.

Claim, page 72: “We resemble our parents and our siblings because we are similar to them genetically, not because we grow up with them in the same environment and experience the same opportunities or traumas.”

“Growing up with siblings does not make you similar to them beyond the similarity due to genetics.”

Response: If you separate identical twins at birth, as they grow, they will resemble their biological parents, in many ways, more than they resemble their adoptive parents. This shows some of the limitations of “parenting to child” effects. Still, in important ways in which a person is who she is, this is likely exaggerated by the fact that children are generally adopted out to families that share the broader cultural mores. If a child from a poor Appalachian family in West Virginia was adopted out at birth to an observant Muslim family in Morocco or an Orthodox Jewish family in Jerusalem would that child be more like his biological parents? In height, weight, posture, artistic ability, yes. But in the important ways a person is who she is, very unlikely. Brothers could share all sorts of traits, aggressiveness, cleverness, ambition, introversion, height, weight, but perhaps one is a famous surgeon and the other a famous embezzler of friends and family. The latter difference requires a vastly different answer to the question, “Who are you?” than “I’m 5’10”, IQ of 123, extroverted in the extreme and cautious.” I grant that we come into life with a bag of DNA. But who a person is develops gradually as the person choses and acts in response to what they face in the world.Claim, Page 85: “It’s also important for parents to know that, beyond genetics, most of what happens to children involves random experiences over which parents have no control.”

Response: I hope that the author has no inclination to write a parenting book. I accept the idea that parenting styles; authoritative, permissive, authoritarian, etc. are unlikely to have lasting, transferable effects outside the home as long as the parenting styles are within the realm of the non-pathological. (as Judith Harris has so well argued in The Nurture Assumption. (See my review, originally in this site. https://www.helenfarabee.org/poc/view_doc.php?type=book&id=37&cn=82.). But I reject the idea that parents have no control over a child’s non-shared events, the events that seems to come out of nowhere, that can damage a child physically and otherwise over many years. Protective vigilance is an obligation of parenthood and some carry it out far more effectively than others. Protective vigilance can have a large and varied effect upon the life of the child. It would be a complicated but very interesting study to isolate such life-changing events (rape, football concussion-based dementia, handgun accidents, petty crime, poorly educated or troubled teachers, etc.) and investigate the degree to which they correlate with non-vigilant parenting.

Claim, Page 89 – 90: “Life experiences matter, and can affect us profoundly, but they don’t make a difference in terms of who we are … good and bad things just happen … [but] … their impact is not long-lasting. “

 

Response: It’s difficult here to know what Plomin is thinking. A teenager in a wheelchair who had dived into shallow water and broken his neck as his father sat nearby reading The Wall Street Journal. A soldier comes home with PTSD or with no legs and one arm. These experiences have great power to change who we are? They perhaps have the same gene-based inclinations resembling what were present before the traumas, but their ability to express them may have been lost. Plomin notes that the results he reports, of lack of parenting effects on or the fleetingness of “random environmental events” do not include “outliers.” But the environment of pathological, traumatic parenting (or peer aggression) are too common to be characterized as outliers and will have life changing effects. Similarly, with tens of millions of combat veterans and even more millions of civilian war casualties. These are people whose lives, bodies, outlook and personalities have been altered by their environment. There is, in fact, a temptation to think that the outliers are the samples that behavioral geneticists’ study.

With the latest edition of blueprint, the author responds to some critics. Defending himself against the charge of genetic determinism (we are only what our genes make us.) He points to statements he’s written, “I talk about genetic influences as nudges and whispers … I mean them.” Page 191. But it’s hard to reconcile this with, “Growing up with siblings does not make you similar to them beyond the similarity due to genetics.” This seems more than a nudge. And, explaining why some children are “nice” “Put crudely, nice parents have nice children because they are all nice genetically.” Seems more that a whisper.    

On a positive side, here are three really interesting things that Plomin notes in this book.

1.    He coins a term, “the nature of nurture.” During the days when suggestions of genetic influence were all but banned from psychological research, what counted as “environment” was consistently oversimplified. (my example) Prior to the acceptability of introducing genetically-based tendencies or shared genes (between child and parent, for example) environmental factors were erroneously characterized. For example, to the extent that children were considered blank slates to be written upon only by the environment it would not have made sense to consider “child to parent effects”. For example, children who are read to by parents are more likely to acquire reading as a habit. Instead of assuming that the nightly reading conditioned the child to be a reader, researchers could have considered that perhaps the child was genetically conditioned to enjoy stories and therefore unknowingly conditioned the parent to read to her. This is an important caution for researchers and complicates the idea of “environmental factors”.

2.    Concerning again environmental factors, Plomin distinguishes shared versus unshared factors. If trying to explain why even fraternal twins raised together differ in so many ways, e.g., aggressiveness, competitiveness, scholarliness, it’s necessary to distinguish the environments they share, e.g., parents, neighborhoods, religion, socio-economic class from environmental factors that are unshared, e.g., a traumatic accident experienced by one, a young love affair, an illness, an introduction to something she loves, (music, basketball). The shared experiences tend to be stable and long term, the unshared ones tend to be random, unpredictable. Plomin is convinced that the random, unshared factors have a greater effect upon the particulars of development than the stable factors of parenting style, class, etc. I’m not sure one way or the other, but it’s important to look into it.

3.    Plomin has a very clear, but brief, discussion of ways that science research can go wrong, even within the boundaries set by journal editors. He discusses the replication crisis in today’s (and yesterday’s) research, “… that the results of many studies, including classic studies that are the backbone of today’s textbooks, [are] creating gaping cracks in today’s science. This is not unique to the social sciences and is quite common in drug efficacy research. An example is “chasing the classic probability (P) value”. The conventional criterion for statistical significance is that the result would re-appear in 95% of the trials, if repeated many times (the remaining 5% due to randomness). If a study does not meet this criterion, the researcher can restart and restart until it does. (In drug testing, the failed tests cannot be demanded to be seen even by Congress – so Congress has decreed.) Or researchers can change statistical tests, etc.

I recommend this book for those interested in the question of trait distributions among groups. But for reasons stated above, I believe that caution is necessary.

This is an important book written by an excellent scholar. I recommend it for reading.

 

© 2019 John Mullen

 

John Mullen is a philosopher and writer living in Gloucester, Massachusett