Against Replicator Terminology

Against Replicator Terminology

Against Replicator Terminology

Hi, I'm Tim Tyler - and today I will be discussing the idea that the formulations of evolution that build on the idea of a "replicator" have some serious problems - and should probably be abandoned.

Here I will present my case against "replicator" terminology - in the context of the foundations of evolutionary biology.

I have got two main objections:

  • The first is that the term "replicator" largely duplicates the function of the term "gene" - which is its rival for being the foundation stone of evolutionary biology. The term "gene" came first, has more support from associated terminology - and so has got a better claim on the central role;

  • The second is that the term "replicator" has too strong implications of high-fidelity copying - and those implications are inappropriate.

My view is that - between them, these objections render the concept of a "replicator" still-born in the context of the foundations of evolutionary biology.

I think that biologists should continue to use the term "gene" instead - to refer to small chunks of heritable information.


Richard Dawkins appears to have been primarily responsible for introducing the term "replicator" into evolutionary biology.

He did this in his 1976 book, "The Selfish Gene".

In his 1982 book, "The Extended Phenotype" he defined the term, as follows:

I define a replicator as anything in the universe of which copies are made.

My first objection

The term "replicator" is a late-comer to the scene of evolutionary biology. Before its arrival everyone was quite happily talking about genes.

Dawkins correctly recognised that a key concept in biology was the idea of the coping of heritable information - and that this concept applied to human culture just as well as to DNA-protein systems.

He sought a term to refer to small pieces of heritable information that would apply equally well to all living systems - and came up with the term "replicator".

One good thing about this term is that it independently proclaims that it refers to a process involving copying.

However, by the time it arrived, the "gene" term had become well established. The study of genes is called "genetics". A collection of genes in an organism is called a "genome". A particular collection of genes is called a "genotype". The study of lineages is called "genealogy". Phenotype changes which are not based on genes are called "epigenetic" - and so on. Biology is saturated with gene-talk.

Basically, the term "replicator" arrived on the scene much too late.

If it had been there from the beginning, genetics might have been called "replicatics", genomes might have been called "replicatomes", genotypes might have been called "replicatypes", genealogy might have been called "replicatology" - and so on.

However, that's not what happened. So - for the sake of consistency - rather than abandon the terminology of "genetics", "genomes", "genotypes" and "geneaolgy" - I think that we should go back to calling small pieces of heritable information what they always used to be called - namely "genes".

My second objection

The term "replicator" does not normally just mean "copy" - it has strong connotations of high-fidelity copying.

If you look at the etymology of the word, it starts with the term "replica" - and a "replica" is a copy that is almost indistinguishable from the original.

David Hull acknowledges this in his definition of replicator - in "Science as a Process" - as follows:

replicator: an entity that passes on its structure largely intact in successive generations

These connotations of high-fidelity are inappropriate for an entity that acts as the foundation stone for evolutionary biology. We know - from information theory - that information can survive in relatively poor-fidelity copying systems - provided that they employ compensating redundancy and error correction mechanisms. This insight dates back to John von-Neumann's paper: "Probabilistic logics and synthesis of reliable organisms from unreliable components."

For adaptive evolution, information needs to persist with high fidelity - but such persistence can be implemented even in noisy data-transmission systems. High-fidelity copying is absolutely not necessary.

This terminological issue has caused real confusion among working scientists. For example, see: the sections entitled "Memes may not be replicators" and "Replicators are not necessary for cumulative adaptive evolution" in the Richerson and Boyd paper in my references.

Of course, you could define the term "replicate" to avoid such connotations - but I think that it is fair to say that these connotations are there in the normal English language usage - and that this is an argument against using the term.

It's not too late

I think that - between them - these objections to the concept of a "replicator" as the foundation-stone of evolutionary biology are terminal.

When new forms of inheritance were discovered, rather than attempting to sweep away the existing terminology, what should have happened is that the term "gene" should have been expanded slightly - to inculde all types of heritable information transfer - not just those based on DNA.

I think that it is not too late to restore the term "gene" to its rightful place at the heart of biology.

Here's a diagram to show how Dawkins and Hull would classify various entities in evolutionary biology:

Here replicators are the fundamental unit of inheritance in evolution.

...and here is how I think we ought to do things:

In this second diagram, genes are the fundamental unit of inheritance.

What about those who claim that the term "gene" is already too overloaded - or say that their claim on the term is better - or that it has historical priority? My response would generally be to tell them where to go. "Gene" is evolutionary biology's term, first and foremost.

If a "gene" is defined to be a small chunk of heritable material, what name should we give to small chunks of nucleic acid?

Those are genes too, of course, and I would normally simply refer to them as such - but if you really wanted a term to refer specifically to nucleic acid chunks while excluding other forms of inheritance - for some reason or another - you could call them "nucleic genes". Similarly with "DNA genes".

Replicator terminology messes up genetic terminology

Currently many dictionaries define "genes" as being:

A hereditary unit consisting of a sequence of DNA that occupies a specific location on a chromosome and determines a particular characteristic in an organism.

Artificial Life enthusiasts have long derided definitions of life, genes and biology that reference the details of biochemistry - as carbon chauvanism.

The belief that genes are necessarily made of DNA is castigated as being nucleic-acid centrism.

I agree completely. The idea that genes are necessarily made out of DNA represents out-dated thinking - which biologists have got to get over - in my view.

We should not define fundamental biological terms based on the historical accidents of our own biochemistry! What if we create synthetic life? What if we want to look at our early ancestors? What if we encounter aliens?

Such terminology is quite an embarassment to philosophers of biology - whose job it is to straighten such messes out.


  1. Tim Tyler - Informational genetics;
  2. Richard Dawkins - The Selfish Gene
  3. Richard Dawkins - The Extended Phenotype
  4. David Hull - Science as a Process
  5. R. Boyd and P. J. Richerson - Memes: Universal Acid or a Better Mouse Trap;
  6. David Hull and John Wilkins - Replication;
  7. John von-Neumann - Probabilistic logics and synthesis of reliable organisms from unreliable components

Tim Tyler | Contact |