Thanks to all who left comments on my previous post. This discussion has been very useful and led me to adjust my views. Here’s how I would formulate the issues now:
(1) ‘Phenotype’ is determined jointly by (i) genetically stored information, (ii) culturally stored information, and (iii) the environment. It doesn’t make sense to speak separately of ‘phenotype resulting from genetic influences’ or ‘phenotype resulting from cultural influences.’ Culture can affect the morphology (foot binding in China) and skin coloration (tattoos). Genes affect cultural behaviors (e.g., political leanings to liberalism versus conservatism). Different traits are affected by different mixtures of genes, culture, and environment, but there are no sharp boundaries.
(2) Genetic information stored within an organism is its genotype, but it is also important to know how much genetic variability there is in a population.
(3) Cultural information can be stored in a variety of media. Initially it was just inside people’s heads, later texts and images became very important, and today (or in the very near future) most of cultural information will reside in the electronic form. It doesn’t seem to matter how it is stored. People who want to call cultural information ‘cultural genotype’ are welcome to it, but I prefer not to do it, because:
(i) culture, unlike genes, can be stored in a variety of media
(ii) what’s important is not cultural information stored in a single person (the most direct analog of the genotype), but the collective store of culture. So, if you want to push the analogy, human groups have cultural genotype, not human individuals
(iii) there are several other differences between genetic and cultural kinds of information that make this analogy not very useful (as detailed in my previous post)
Well, strictly speaking the concept of collective store of cultural information would match more with the concept of gene pool than with the concept of genotype. In any case I agree: there is no need to take too seriously the analogy between genes and memes (or whatever you want to call them).
In my opinion there doesn´t seem to be a cultural analog to the biological genotype. As you already pointed out, the internal representations of the memes (whatever they look like) aren´t tranferred. Only behaviours (or they codifications) are tranferred. The very concept of genotype wouldn´t be useful if it didn´t account for fragments of “hidden” information that are silently and unadvertedly transmitted through generations, but that can suddenly turn into phenotype in the future. That is mainly a consequence of the sexual reproduction and nothing makes us think that there is an analog of this hidden information transmited culturally… unless we think of books printed without anyone reading them, or paying any attention to their contents, until it is necessary.
On the other hand, there are two (and possibly more) ways of understanding phenotype: The first is phenotype at the individual level. This is the phenotype that was explored by Mendel and the level we are familiar with. This level is about general traits like the color of the eyes. The other way of approaching phenotype is at the single cell level; Through differentiation mechanisms the different cell lineages specialize in different tasks. That is: different cell lineages express different sets of genes in the form of proteins, self-regulation, etc.
Speaking in general, phenotype is the partial expression of the genetic information stored. These two approaches fit the definition but are extremely different because one of them (the individual level) has a sexual drive, molded to create variation in order to enhance competition, an the other (the single cell level) has a very different drive, molded to create variation which enhances cooperation. We could say that the individual level is kind of an “epiphenomenon” of the single cell level, but that probably wouldn´t be accurate. Could we say that “what the genotype is for the single cell, the gene pool is for the individual”? I don´t know. Does that make any sense to you?
It is not unconceivable that there could be phenotypes at higher levels, like the group level. In fact that is exactly what happens with humans and social insects. Different individuals specialize in different tasks, what reminds me of the single cell level of phenotype I decribed earlier. Of course this wolud lead to differences at the group level which could account for a kind of group analog to the individual level of phenotype.
I have no doubt that cutural transmision has a key role in this group level phenotype in humans.
Linguistics may have its own take on this topic. Going all the way back to the early days of the discipline, Ferdinand Saussure drew a distinction between ‘parole’ and ‘langue.’ Literally these mean “speech” and “language,” but the real distinction, according to Saussure, is that ‘parole’ refers to the speech acts produced by individuals, while ‘langue’ refers to language as a shared system of signs. For example, the particular tone of voice that tells you who is speaking, and the particular sentences that the speaker produces (unless they’re rote), are part of ‘parole’ but the words and the grammatical rules they use are part of ‘langue.’ ‘Parole’ is ephemeral, ‘langue’ is what gets replicated. ‘Parole,’ I suggest, is the linguistic analog of phenotype, while ‘langue’ is the analog of genotype (memotype?). (And ‘parole’ is also affected by genotype: whether you have a Y chromosome affects how deep your voice is.) The human mind is probably actually designed to reproduce some fairly abstract features of other speakers’ language use, in order to give speakers a shared code to work in, while leaving people free to decide for themselves what messages to send in that code.
Mendel advanced the study of heredity by going beyond the old observation that children resemble parents, giving evidence that the distribution of visible traits was governed by the transmission of invisible factors. In this sense, linguistics has been operating at a Mendelian level for some time. It’s an open question how far similar distinctions can be drawn in cultural domains outside linguistics.
Doug, this is a very good point. In general, because we know so much more about linguistics and linguistic change, it is one aspect of culture where we can really hone our approaches and models to cultural evolution in general.
Starting from the biological approach, I may see things a bit simplistic.
Genotype, as strongly controlled by DNA, has built in for the individual much of its initial ability to survive ( the bird going for the wiggling worm, the baby suckling and the sounds it makes as the start of speech). Genotype as DNA controls growth (e.g for people ears keep growing larger all through life but bones don’t) and how to interact with the environment and other individuals.
This is then modified by the interactions with the environment (the black worm contains a toxin that stings-leave it alone), and individuals (dad looks after me and helps find food -mum ignores me and runs away -well at least for emus).
At the species level the genotype does similar, especially in the interactions between individuals, helping determine whether the species survives, grows and reproduces. The DNA breeds true to species as a very strong way of maintaining the species integrity.
What we see and hear (from the outside?) is phenotype as a complex interaction between genotype and the environment, other individuals and culture. Take a species and spread it from the beach up the rivers out to the desert and up to the mountains, and soon there will be differences in the phenotype.
Even the language will change (waves and tides to fishing but snow skiing etc in the mountains) so I would put language as phenotype and speech as genotype opposite to Doug above.
Given enough time and survival of the fittest/natural selection wil encourage divergence of the DNA (genotype), into varieties and even new species.
So into culture and what part does cultural genotype play? Surely it is the rules, traditions and behaviours that help (or even control?) how a culture survives, grows and competes with other cultures (even how it reproduces). These rules, traditions and behaviours were present before written records and language in the form of stories, songs and dance learnt verbatim and passed on to the next elders.
However cultural information is cultural phenotype, including (and reacting to) cultural genotype. In particular in cultures with written records (and now electronic records) only part, perhaps only a very small part, is likely to be cultural genotype. Say in the USA, the Constitution and Amendments, the legal system and laws are an important part of the cultural genotype (trying to maintain the culture), but most of the couple of teraBytes of “information” pumped onto the internet daily is utterly insignificant.
History (cliodynamics?) may change a small part to important?
Let me provide a standard linguistic example to flesh out my previous post:
There are three pronunciations of the regular plural (and possessive) in English, -s, -z, and -iz. Think cats, dogz, and horsiz. But these are not three memes. Rather there is really a single underlying representation of the plural, with the actual pronunciation varying depending on the preceding phoneme. We can even specify that the underlying representation of the plural takes the form -z. Why do we think this? Consider that a vowel in English can be followed either by -s, as in “grace,” or by -z, as in “graze.” If the underlying plural took the form of -s rather than -z, then we would expect the plural of “gray” to be pronounced like “grace” rather than like “graze.” This implies that when we say “cats,” we are not copying other people’s pronunciation of “cats,” but calling up a -z, and then modifying it to avoid following the unvoiced -t- with a voiced -z, a no-no in English. (For a more thorough treatment, see Steven Pinker’s “Words and Rules.”)
In fact, many linguists believe that some underlying phonological representations are never pronounced, but always modified on the way to producing speech. In other words, there is strong indirect evidence for the existence of unpronounced representations, just as there was strong indirect evidence for recessive genes, even before people knew much about DNA.
So I take it that there are different kinds of memes. Some are supported by specialized psychological machinery dedicated to copying selected abstract features of other people’s behavior. Knowing how this machinery works will help a lot in understanding constraints on what sorts of memes will or won’t be copied. Other memes depend on external representations, and sorting out the psychology is less important: I want to know that my scribe, or my photocopier, can copy a document, but I probably don’t need to know much about the psychology/internal machinery that lets them do it. The constraints on what size and color paper my copier can handle are broader and less interesting than the constraints on what languages people can learn.