Guest Blog by Yasha Hartberg. Rules as Genotype: Let’s not declare the idea dead too quickly



Join 36.9K other subscribers

I was another participant at the Rules as Genotype workshop held at Indiana University recently.  Unlike Peter Turchin, however, I came away with a very different perspective on the usefulness of the metaphor.  I am undoubtedly somewhat biased since some of my own research explores the idea of a sacred text as a kind of cultural genotype, research which Peter Turchin referred to in his April 24 blog entry.  Nevertheless, the stimulating conversations that we had at the workshop reinforced in my mind the usefulness of the idea, especially since the metaphor seems to provoke interesting and productive questions whether we are talking about religious social groups such as Christian congregations operating within the local scale of Binghamton, NY or whether we are talking about constitutional arrangements at the level of nations.  I would therefore like to take some space to respond to Peter Turchin’s reflections on the workshop.  In this first post, I will address some of his specific concerns, most particularly with his objection that cultural evolution is just “too different” from genetic evolution to make “rules as genotype” a useful construct.  In my second post I will explore what I feel is likely common ground between our two positions and give a few reasons why I think the metaphor is not only useful, but necessary for a complete understanding of the evolution of rules.

I suppose I would first question the contention that because human phenotypes are the result of interactions between genetic, epigenetic, cultural and symbolic influences (to borrow Jablonka and Lamb’s framework) that it doesn’t make sense to talk “separately of ‘phenotype resulting from genetic influences’ or ‘phenotype resulting from cultural influences.’”  As Robert Boyd argued during the workshop, not all possible sources of inheritance responsible for human behavior have equal explanatory power.  While I disagree with him that epigenetic and symbolic systems have a negligible influence, I nevertheless feel that his point here is well made.  For example, while humans have evolved genetic systems that make them susceptible to acquiring traits such as foot binding, the genetic influences on foot binding as a particular practice are likely minimal.  Cultural systems give you far more explanatory power for that phenomenon.  Foot binding, then, would provide an example of a phenotype stemming primarily from cultural influences.  On the other hand, if we wanted to understand the evolution of the ice cream sundae, we would need to consider the genetic influences that allow lactose tolerance as well as the cultural influences that led to the development of ice cream.  Discussions, then, of what phenotypes result from is largely a matter of choosing an appropriate level of analysis for the specific questions being addressed.

But what about the contention that cultural systems are just too different from genetic systems?  I both agree and disagree with this argument.  Yes, cultural systems are different.  But we need to be careful to consider specifically what aspects make them different and whether those really negate the usefulness of the metaphor.  For instance, Peter Turchin argues that one of the properties that make cultural systems too different is that cultural information can come from many sources and be stored in many ways.  Yet, prokaryotic organisms are notoriously promiscuous in terms of their horizontal exchange of genetic information.  Does this mean that there is no useful distinction between genotype and phenotype in bacteria?  Of course not.

I suppose it could be argued that even though bacteria exchange genetic material with wild abandon that the information is still all encoded in DNA.  This strikes me as problematic for a couple of reasons.  First, many viruses store their genetic material as RNA rather than DNA and this information can become embedded within the genomes of prokaryotes and eukaryotes.  Prions, while not imparting information in the way we think of genes nevertheless can be passed between individuals and affect phenotype.  And, of course, epigenetics involves all manner of DNA modification as well as changes in associated protein factors that affect transcription and translation.  The picture of a uniform genetic material, then, seems to me not to be quite as clear as it is often portrayed.  Second, most of the alternative kinds of cultural storage media cited such as writing, YouTube clips, etc. are very recent developments.  Rules, on the other hand, predate all of these.  Regardless of where one stands on the debate about whether these rules would have been stored in people’s brains or in distributed social networks (I couldn’t quite follow that debate either), the fact remains that those rules were not stored in multiple forms of media for most of human evolutionary development.  Even if multiple media for cultural storage were a valid strike against rules as genotype in the modern context, it couldn’t have been in the (pre)historical context.  This argument may not serve to validate the concept of rules as genotype, but it does caution us to be careful not to extrapolate too far from the modern condition.

Where I share some of Peter Turchin’s anxiety, I think, is in the inherent danger of arguing from metaphor, a concern expressed by several scholars at the workshop.  While I may disagree that the differences between cultural and genetic evolution are so different as to make the metaphor of rules as genotype meaningless, I can hardly deny that they are different in many important ways.  The notion of a cultural genotype is one laden with all the theoretical baggage of genetic evolution, only some of which may apply to cultural systems.  It is an evocative idea, but one that risks not only misapplying biological theories to social phenomenon but also causing an unintentional blindness to novel processes at work in cultural evolution that have no counterparts in biological systems.  Therefore, while I’m happy enough with the conversations sparked by the cultural genotype metaphor I would like to see the discussion ultimately move to the use of more substrate neutral language.

There is nothing particularly new about this call.  Richard Dawkins is often credited with coining the term “universal Darwinism,” but he was by no means the first to extend general Darwinian principles to domains outside genetic processes.  Moreover, a host of scholars have worked on the problem of generalizing Darwinism since the term was coined and I think it would be productive to consider rules from within this broader framework.  For instance, in Darwin’s Conjecture:  The Search for General Principles of Social and Economic Evolution, Geoffrey Hodgson and Thørbjorn Knudsen develop the idea of generative replicators.  According to their definition, genotypes are generative replicators by virtue of the fact that they 1) are causally implicated in their replication, 2) produce copies that are similar in their generative mechanisms to the parent, 3) transfer information and 4) are conditionally activated through their interaction with the environment.  From Hodgson and Knudsen’s perspective, however, there is nothing unique about genotypes in fulfilling these four requirements.  While they do not address rules in the specific sense employed by Elinor Ostrom’s IAD framework, they do explore the idea of whether judicial laws can be considered as generative replicators and conclude that they can.

So is there still life in the rules as genotype metaphor?  I certainly think so and I think it is a conceptually tidy way of considering institutional rules which are, after all, devised in order to affect human behavior.  In other words, rules are conceived and adopted specifically as a way of creating the cultural phenotypes that Peter Turchin argues are the only important things to consider in cultural evolution.  Anyone studying institutional development, however, is keenly aware that even the most carefully crafted rules often have unexpected outcomes, either failing to make any changes in human behavior at all or making profound, yet unintended changes, often with perverse consequences.  This is because institutional rules interact within a complex biological and social environment, not unlike how genes interact with the environment to create phenotypes.  Yes, the mechanisms by which phenotypes are generated are very different between genes and rules.  Nevertheless, within a generalized Darwinian framework they are conceptually similar.  Creating better institutional rules, then, becomes a process not unlike site-directed mutagenesis in which changes to specific genes are made and the resultant changes (if any) on phenotype are measured.  Instead of using a DNA template, policy makers use statutes.  By carefully monitoring the behavioral changes that accompany new policy implementation, researchers can gain a deeper understanding of the mechanisms of cultural evolution.  This is particularly true if they are able to take the social and biological context into account.  I find it difficult to see how the same level of precision could be reached if the cultural “genotype” were disregarded as unimportant.  I’m happy to call it something else, such as a generative replicator, but I’m not quite willing to give up on the general concept.

Notify of
Most Voted
Newest Oldest
Inline Feedbacks
View all comments

Memotypes (cultural genotypes) are surely memetics 101. No insult intended – but I think if your application of evolution to culture doesn’t even get this far, then you have barely really got off the ground.


Regarding the call for “substrate neutral” language – that was the point of Dawkins’ “replicator” terminology. Alas this has caused at least as much confusion as enlightenment – with eager critics missing the point, ignoring the “Dawkins” definition of a “replicator” and proclaiming that evolution is not necessarily based on high fidelity transmission processes .

A precursor was the “mneme” terminology from Semon – though this never really took off.

I think the number one alternative is to expand the terminology of genetics to cover all forms of inheritance. So, then “genes”, “genetics” and “genotype” would *become* substrate-neutral. This project goes back to the definitions of “gene” from Williams:

“In evolutionary theory, a gene could be defined as any hereditary information for which there is a favorable or unfavorable selection bias equal to several or many times the rate of endogenous change” – Williams 1966, page 25.

We *already* have a science of heredity: genetics. We don’t need another one. There’s more about all this in my “Informational Genetics” article.

Yasha Hartberg

In principle, I have no strong objection to sticking with biological terminology. Given my background in biochemistry it’s certainly the most comfortable vocabulary for me to use. In practice, though, I do worry that something important might be missed if we use genetics as our starting point. Moreover, as an educator, I’m keenly aware of how dangerous analogy can be in transferring ideas between domains. Analogies are useful tools, but they are by no means perfect.

Moreover, I think interesting new questions arise if we develop more substrate neutral language. For instance, I presume that a necessary component of any inheritance system is some sort of information storage system. In biological systems, this is most often thought of as DNA (though I think we agree this is a simplification), a relatively concentrated, compact, almost digital system. Cultural systems, by contrast, seem to store information far more diffusely in distributed networks. An obvious question from this line of thinking is, what are the constraints of each sort of system on information storage and how do those constraints affect a system’s response to natural selection? I’m not sure such a question arises as naturally when thinking from a genetics perspective.


Propose away. However, genetics seems like a formidable competitor to me. My guess is that it will win, hands down.

Peter Turchin

Yasha, you say

“Yet, prokaryotic organisms are notoriously promiscuous in terms of their horizontal exchange of genetic information.”

This misses the point that I was making. Yes, there is a lot of genetic exchange in bacteria, but the carrier is invariably RNA (or DNA). Whereas cultural information – for example, tunes – can be recorded in human brains, written in musical notation, on a phongraph, on a vynil record, on a magnetic tape, on a CD, and as digital audio. This is a big difference,

Yasha Hartberg

I did try to address that point, though. In particular, I pointed out that things like musical notation, phonographs, vinyl records, magnetic tapes, CD’s, digital audio, etc. are new phenomena, far newer than institutional rules. At some point, and likely for the bulk of our evolutionary history, it was all just stored in human brains. As such, I have a hard time seeing how the variety of media now can be a strong argument against what was happening then.


Peter, non-cultural inheritance is not always nucleic acid-based. Ant nests, rabbit warrens and the soils produced by earthworms are all inherited – yet few would declare these to be “cultural”. Some forms of stress are inherited. Prion diseases are passed from one generation to the next without nucleic acid. Also, gene sequences are stored in databases by genetic engineers these days.

The alleged “big difference” between genes and memes that you refer to is a myth.

Peter Turchin


My point is that the same cultural element, a tune, can be stored in 7 completely different media, but still remain the same tune. Wheras genetic information can only be stored in the form of nucleic acids.

But I agree with a more general point that information can be transmitted in a variety of ways, which means that there is a great variety of Darwinian-type processes (in addition to genetic and cultural). George Price famously noted that his equation can describe pretty much any kind of a Darwinian process.


Consider the UCSC Genome Bioinformatics Site. Genetic information is still information – and information is portable. These days genetic information can be stored in any medium – and then go on to be inserted back into organisms again using retroviral engineering – or other techniques. Genes are catching up with culture in this particular respect.

  1. Home
  2. /
  3. Cliodynamica
  4. /
  5. Guest Posts
  6. /
  7. Guest Blog by Yasha...

© Peter Turchin 2023 All rights reserved

Privacy Policy