REPLICATORS AND REPLICATION IN A SELECTIVE CONTEXT

A central concept to the meme-gene analogy is replication. A replicator is firstly defined as a unit that has largely the same structure before and after a copying or replication process. Some errors may be made in the process, but too many errors are not allowed for a copying process to count as a replication process. The second essential characteristic is that it is part of a lineage. Genes are called replicators because they are copied over and over. Over and over means that a replicator is first copied from a mould, then in a next replication event it is used as a mould itself, from which a new copy is produced, that again will function as a mould in the next event, and so on. This happens in replication processes like cell division (mitosis and meiosis) or in other processes like virus replication. The mould-copy-mould/copy connection is an historical physical connection. The historical sequence mould-copy/mould-copy is called a (genealogical) lineage.

Analogous to gene replication, cultural traits are for instance copied over and over again from human to human (Boyd and Richerson, 1989 ). In the social context memes (as replicators) form lineages, for example particular central theories (like the Darwinian theory of evolution by natural selection, or Newton's laws [Dennett, 1991 , p 201]) are passed on from human to human, and from generation to generation. As mentioned, replication in biology is attached to the central selective concept of variation (since variation [2] results from the many replications of genes). David Hull ( 1980; 1982; 1988a, b ) has described evolution by natural selection by means of concepts that are designed to describe evolution by selection in general, and which coincide with Dawkins' concept replicator. Therefore, I find it useful to use his concepts to examine the meme-gene analogy.
Hull's critical concepts are replicators, interactors, and lineages. This leaves me to elaborate on the concept interactor. Hull's actual definitions of the concepts are appended in an endnote [I] .
A gene or replicator lineage, which changes or variates because of replication errors, might be a single lineage, or a spliced one. It would be a single lineage if every mould is only copied to one copy. For instance if an organism only reproduces itself ones (asexually), transmitting only one copy of its genes to a new organism, and this organism does the same, we have a single lineage. We can picture this situation as mould-copy/mould-copy or m-c/m-c.
However, one mould can also serve for the production of several copies. For instance in the case where organisms (asexually) produce a large progeny. If these several copies differ, the different moulds will be the beginning of a 'spliced' lineage, which has several replicator 'lines'. This spliced lineage now contains variation, within one generation, since the different lines have become different by copying errors. We can picture this situation as

m-c1/m-c/m-c

-c2/m-c/m-c
-c3/m-c/m-c
-c4/m-c/m-c
where there is only one generation that has a large progeny.

In a selective context, the organisms represented by the lines (one to four) can compete or interact in some competitive way, resulting in what Hull would call (natural) selection. To describe (natural) selection we not only need replicators and lineages, but in addition, but also interactors. Interactors are wholes, like spermcells, organisms, species, or genes, that interact in a selective context (Brandon, 1988 ; Hull, 1988a ).

Note that the most obvious replicators in biology, physical genes [3] , are not directly subject to selective forces. Selection takes place on characteristics (the phenotype) of interactors.
There can be many steps or processes between a replicator and the characteristics of interactors (organisms, species, spermcells, etc.). Genes are messages, symbols or commands that contain information that prescribes the building of interactors including their behaviour, and other characteristics upon which selection can take place. In the biological literature on replication, organisms as the most obvious examples of interactors, are sometimes also called the vehicles of genes. Authors that do not emphasize the many steps between the genotype and the phenotype incline to see organisms as vehicles steered by genes, where those that do are find the concept interactor more apt (Hull, 1988b ).

A similar selective context is often presupposed in memetic evolution. Like genes, memes form lineages, and these memetic replicators can interact in a selective context. For example, in science, a theory can be seen as a meme (Hull, 1988a ). Theories can compete for the description of phenomena. In policy making, particular instruments, such as license systems, that were previously used in one policy field are copied to other policy fields. If a choice has to be made between instruments, or between amounts of money used for different instruments, we have a selective situation. Thus in both science and policy there is a selective context.

We can minimally distinguish between two approaches to such selective situations that differ in what is selected for. Dawkins ( 1989 ) and Dennett ( 1991 ) emphasize 'braintime', or 'entry in as many minds as possible' as the major selective factor, where Hull emphasizes selection for cases where memes compete for being 'true', or 'useful'. There is considerable difference between the two approaches, because memes that are untrue, might still be very successful in entering many minds. Take for instance the idea of Lamarckian evolution, that is often replicated in scientific publications for the sole purpose to show that it is wrong (Dawkins, 1982 ; Dennett, 1991 ; Hull, 1982 ).

Above I have characterized A) the concept replicator, B) that replicators (by definition) form lineages C) that replicators can interact in selective events, in which case they, or their vehicles, can be called interactors.

next part

Back to hypertext essay on memes

Back to Memes Back to Evolution Back to Hans-Cees