The Stupidity of the “Smart Gene”

Posted by Jessica Cussins September 17, 2014
Biopolitical Times

Science writer David Dobbs has definitively described the voracious appetite of the “selfish gene” meme, pointing out that the notion of individual genes exercising power on the outcome of events has been so good at mass producing itself, that “the selfish gene has become a selfish meme.”

At CGS, we have another name for this phenomenon: the gene of the week.

All “genes of the week” have something in common: they never actually live up to their billing. For starters, it is never true that a single gene just does something. Genes work together, and genomes work with their environments. But this inconvenient reality has done amazingly little to stem genetic determinism, or the funding of research that relies upon its framework.

How could this be true? Is it really the case, as Dobbs posits, that “the gene-centric model survives because simplicity is a hugely advantageous trait for an idea to possess?”

The adventures of “smart genes” provide an illuminating case study. 

There are some people who really want to find the genetic basis of intelligence. The sole effort of BGI’s Cognitive Genomics Lab, for example, is to investigate the genetics of human cognition. The fascinating documentary film DNA Dreams sheds light on the hopes of those involved with this project. They harbor no doubt that there is a genetic basis for intelligence; the only question is when they’ll uncover it and how limitless the possibilities will be once they do. Embryo selection and modification to ensure smarter babies are among the future scenarios they envision.

But the future doesn’t seem to be cooperating. Project leader Bowen Zhao confidently promised data in three months time in February, 2013; it has now been nineteen months and we are still waiting.

In the meantime, two of the Cognitive Genomics Lab’s illustrious “collaborators and advisors,” Robert Plomin and Steven Pinker, are among the 59 co-authors of a brand new PNAS study, which Nature has called “one of the largest, most rigorous genetic studies of human cognition.”

The paper, published on September 5, is called “Common genetic variants associated with cognitive performance identified using the proxy-phenotype method.” It claims that three gene variants are associated with both educational attainment and higher IQ scores. The study is a follow-up of the 2013 Science study, “GWAS of 126,559 Individuals Identifies Genetic Variants Associated with Educational Attainment.”

What is remarkable about these studies is how little what they “identify” actually matters. Over 100,000 genomes were studied in each, but the results are so limited that they could easily fail to be replicated and become one of the many of a decade’s worth of publications on gene associations that the editors of Behavior Genetics disparage as “wrong or misleading and have not contributed to real advances in knowledge.”

In the Science study, three gene variants were found that each correlated with roughly one month’s difference in a person’s total amount of schooling.  In the PNAS study, three gene variants were found that each correlated with about 0.3 points on an IQ test. Results this inconsequential could easily be no more than false positives.

Study co-leader Daniel Benjamin was upfront about this fact. He told Nature, “We haven’t found nothing.”

Whatever mechanisms are working on the partial heritability of intelligence, it doesn’t seem to be a defined group of “smart genes.” The authors of the PNAS study have suggested that looking into more than one million people’s genomes could help explain the genetic basis of cognitive ability a little bit better.

But the problem isn’t in the numbers, it’s in the framework. Genes are not selfish; they don’t act on their own for virtuous or nefarious ends. Heritability is much more complex than genes.

So, can we finally put the notion of “smart genes” behind us?

Not quite. Here’s how The Sydney Morning Herald covered the exact same study last week: “Scientists discover 'smart' genes.”

These are powerful memes. And when information is so beautifully simple, who cares if it says basically nothing and could well be wrong?

Previously on Biopolitical Times: