You don’t need “genes” for genetics

By Razib Khan | October 12, 2012 2:04 am

After yesterday’s post I feel it is important again to reiterate that there is an unfortunate tyranny of the gene-as-physical-entity when it comes to our understanding of human heredity. To clarify what I mean, I think it is useful to borrow a framework from Andrew Brown. On the one hand you have a conventional modern mainstream understanding of the gene as a molecular biological entity, fundamentally derived from DNA and its role as envisaged by Francis Crick and James Watson, but with roots deeper back into the physiological genetic tradition which Sewall Wright was embedded within. In contrast to this concrete and biophysical conception of the gene there are those who conceive of the gene as an abstract unity of analysis. Richard Dawkins is the primary proponent of this viewpoint on the public intellectual scene, though men such as William D. Hamilton self-consciously understood the difference between their own genetics, and that which arose out of the insights of Crick and Watson.

The key point which you have to remember is that the gene was conceived of before its substrate, DNA, was understood. Genes, and therefore genetics, predates molecular biology. For example, when you look at the pedigree above you see an autosomal recessive trait (dark). This is inferred by the pattern of inheritance. It is not adduced from a feature of molecular biology. If it is an autosomal recessive trait with a locus of high penetrance, then no doubt there is a clear molecular biological grounding for how the consequent phenotype manifests. But an understanding of molecular biology is not necessary to infer the genetic character of the trait. More colloquially: you don’t need to know the exact gene of major effect to conclude that a trait is genetic.

And the tyranny of the gene is even more pernicious when it comes to a concept which predates genetics itself, but is clearly ultimately inextricably tied to a genetic understanding of the inheritance of traits: heritability. Heritability is simply the proportion of variation in the trait that is due to variation in the genes. One infers it in a classical sense by tracing correlations of the trait of interest across relatives of differing degrees of relatedness. Fundamentally it is a “gene blind” summary of statistical patterns and associations.

Why does all this matter? Because today people want a specific gene, elucidated by contemporary molecular biological methods, before they accept that a trait is “genetic” or “heritable.” This is the fundamental tragedy of behavior genomics. It is clear that many behavioral and cognitive traits have a heritable component, but the public is always hungry for the “X gene.”


MORE ABOUT: Human Genetics
  • AG

    My simple deduction: Only 4 DNA codes A, C, G, T -> many codons -> even more genes ->much much more exression through different conbination and interaction of genes -> ? -> numerous phenotypic presentations including species.

  • jose

    Wouldn’t genotype be a better word to talk about heritable traits? If a gene is a concrete, physical thing, it’s only logical people want to know what genes are being talked about. If you talk about the genotype for this or that trait, however, you’re abstracting the specific genetic makeup for it and focusing on its inheritance.

  • Razib Khan

    #2, true, but trivial. do you think discussing a genotype of 5,000 loci with 15,000 allelic variants with minor allele frequency thresholds of 0.05 or more is intelligible to most people? there’s a reason that a domain of genetics uses quantitative statistical methods rather than verbal logic and concepts.

  • Dm

    People “want a specific gene before they accepts that a trait is heritable”?

    I think they actually want a specific gene before they accept that the geneticists have anything non-trivial to say about a trait. They actually knew all along that like anestors, like progeny; and a nearly-continuous, sort of low-penetrance polygenic model of inheritance is very intuitive. People are very much at ease with terms such as “1/8th of one’s great grandfather’s blood” etc.

    One of the biggest surprises of genetics comes when it turns out the inheritance is actually highly discrete, all or none; or when it follows along intuitively unexpected lines, like X-linkage or even something as simple as heritability of male traits through mothers or female traits through fathers. Now that doesn’t resemble simple “serial dilution of blood”, and the listeners are like, aha, now the geneticists have something to tell us what we didn’t already know!

  • Chad

    Its hard enough explaining genetics to people….now throw in epigenetics 😉

  • Razib Khan

    #4, this is not the right model for very many educated people who are not biologists. i can give you plenty of concrete examples, as i have to explain this shit to people constantly who think if there ain’t a SNP it ain’t heritable. in some ways it is easier to explain to less educated people because you only have to explain the discrete part.

  • Razib Khan

    e.g., the post was triggered by a long, though productive, exchange with a computer science person why falsification of candidate gene studies DO NOT imply that a trait is not heritable.

  • Chad

    I blame the media, but I am Conservative so that is nothing new….but seriously, when the news reports always talk of “the gene for…” it sends a wrong message.

  • nooffensebut

    “It is clear that many behavioral and cognitive traits have a heritable component, but the public is always hungry for the ‘X gene.’”

    And, yet, when scientists do discover an allele of profound impact on an important behavior, though it be rare in all but black people, four years can go by before it receives meaningful attention again (after some direct persuasion from yours truly). Therefore, I would say that the GWAS jihadists have had a rather successful campaign to silence candidate gene research. What little research this allele has received was through low-cost data mining of an existing database, but I would expect that to change now that MAOA is also a cancer gene.

  • gcochran

    This happens more when people don’t like the answer.

  • Dm

    #7 It isn’t uncommon for people to think that, once you estabished heritability of a trait (typically by a study of twins) but couldn’t find a mechanism of heritability, then you are just reaffirming old wives tales. Like everybody knew that the twins are alike, and that children resemble parents, and at a first glance, it doesn’t look like your study has achieved anything new.

    What happens afterwards (after your corresponded has already mentally equated a heritability study with the old wives tales) sort of depends on what one thinks about this specific “old wives’ tale”. If the old belief contradicts with this person’s body of beliefs, then woe to genetics, it must be as stupid and as wrong as the “old wives”. I venture to guess that your software engineer corresponded was concerned about some mental ability or societal success trait … areas where we’ve been acculturated to believe in equal abilities / 100% power of nurture, and to reject the traditional, discriminatory crackpot lore.

  • Razib Khan

    #11 & #10, you’ve inferred correctly a lot of the concerns (they were citing chabris’ paper as refutation of heritability). that being said, they seemed surprised that height has the same issues, since height obviously is heritable. so there was a broader confusion about what heritability is.

  • Dm

    It wouldn’t be the first time when genetic “overpromises and underdelivers” and the people’s response is to call genetics a fraud. I guess most people consider themselves to be Great Experts in the fields of breeding and heredity, and freely rely on their own intuition and beliefs (perhaps we evolved that way, with some sort of innate heritability compass which, alas, isn’t tuned quite right).

    Some well-intentioned – but lacking good scientific data and/or extremely willing to substitute superstitions of class and race for proof – souls practiced eugenics, and tainted the idea (probably) forever.

    Others promised a rapid agrigenetic revolution in the former Soviet Union, delivered next to nothing, and ended up in jails, with the whole enterprise of genetics condemned and substituted by neo-Lamarckism for over 20 years.

    And more recently, underpowered candidate gene studies, shooting for SNPs associated with multiple IQ poin effects, yielded “statistically signifcant” flukes thanks to liberally applied multitesting, and now Chabris tries to salvage the tainted reputation of IQ genetics, but my guess is that the field will remain tainted for some time to come.

  • Merm

    Post #999 about heritability. Can you just show me a diagram instead?

  • rich lawler

    @2…not to state the obvious, but we don’t inherit genotypes, we inherit alleles. Unless I’m misunderstanding what you are trying to say.

    And a somewhat hastily written thought:
    It seems recently that there have been astounding technological advances in science, particularly genetics, as well as a better scientific awareness in lay people. These two things seem to have combined into a sociological demand for empirical proof–that we measure and provide evidence for the latent variables we’ve long assumed to exist; such variables include “additive genetic variance” in genetics, “general factor (g) of intelligence” in psychology, “quality of life” in economics ” and “frailty” in demography. In some cases there is actually a technological way to measure and get at the latent variable we’ve assumed to exist, for example when we find a particular genetic marker of additive effect causally associating with a phenotype, but in other cases there is not a straight forward mechanistic relationship between the latent variable and its presumed causal influence on the things we can actually measure. For example, while we assume that “g” controls aspects of intelligence, describing how g does this is not something that we can empirically point to in the brain.

  • Tim Tyler

    You don’t need DNA “genes” for genetics. You do, however need some kind of heritable information. Williams (1966) defined sections of these to be “genes”. In that context, genetics would be seen as the science of genes and gene changes.

  • gillt

    “But an understanding of molecular biology is not necessary to infer the genetic character of the trait.”

    People find that trivial and besides the point because they want control over it, especially if it’s disease-related and to get that you need the tyranny of behavioral genetics.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar