Wednesday 1 February 2012

Are memes discrete?

One issue that critics of memetics have perenially got their knickers into a twist over is the issue of whether memes are "discrete" or not.

Examples

Tim Lewens questions the meme-gene relationship, saying in his book Darwin (2006):
Genetic units are discrete particles; culture is not composed of discrete units
Mesoudi (2011) writes:
Memetics makes the neo-Darwinian assumption that culture can be divided up into discrete units that are inherited in a particulate fashion, like genes.
Luigi Luca Cavalli-Sforza and Marcus W. Feldman (1981) wrote in Cultural transmission and evolution: a quantitative approach:
Specific units, such as memes were intended to represent, have meaning when there is essential discontinuity between categories.

Response

If you check with the definition of 'meme' in a dictionary there is no mention of 'discreteness'.

Memetic transmission is, in fact, about as discrete as genetic transmission is. Just like genes in sexual organisms, memes can be divided into sections with different paths into the future at practically any point. Cultural information can be sliced and divided into discrete pieces. It is just a fact that any source of information can be so divided up. Such divisions typically happen during cultural transmission - with different pieces of culture taking different paths to immortality or oblivion. Here's Cloak (1975) on the topic of cultural discreteness during transmission:

I am rapidly coming to believe that much, if not most, of culture is acquired in tiny unrelated snippets, specific behavioral propensities culturally transmitted from one generation to another with remarkable fidelity. The fidelity and ease with which these "corpuscles of culture" are transmitted and acquired is possible only because the organisms in question are phylogenetically adapted for transmitting and acquiring cultural corpuscles.

Culture is typically divided into discrete packages during transmission - but these "packages" may not necessarily be memes, sometimes they may be huge memeplexes. So: the discreteness of culture during most transmission processes doesn't tell us too much about the discreetness of memes - since transmission processes often transmit huge chunks of information.

Another place where there's scope for confusion is meme frequency analysis. Many empirical studies of cultural epidemiology need to count the occurrences of cultural elements - to chart their frequencies, or to create phylomemetic trees. Some way of dividing cultural elements into discrete bins is consequently invented for this purpose. The culture is "digitised". The resulting memes which are then used in meme frequency analysis are necessarily discrete entities. This may contribute to the confusion over the discrete nature of memes. Memes used in meme frequency analysis are indeed discrete entities. This is true of any approach that uses frequency analysis to study and understand culture, and is not specific to memetics. This approach is absolutely fine - and it does not represent some kind of problem.

Then there's the (different) issue of whether culture is like a bar of chocolate - with "discrete" preferred fracture points. To be sure much culture is indeed like this. If you look at quotations, for example, many of them start with a capital letter and end with a full stop. However, culture is not necessarily always discrete in this sense.

Genes are less discrete than most memes are - in this sense. Occasionally there are preferred points of division (e.g. see restriction enzymes), but usually there are not - and the divisions that take place during meiosis can take place practically anywhere.

Once the fact that memes are not necessarily always discrete in the "bar of chocolate" sense is explained, critics claim that this breaks the meme-gene analogy - since, they claim, genes are discrete in that way - with start and stop codons. However, start and stop codons aren't really a part of genetics. They have more to do with developmental biology than genetics - since they are involved with how DNA-genes are expressed. If we get into cultural developmental processes, then some memes are expressed in obvious discrete chunks and some are not - which is indeed rather different from how most DNA-genes are expressed. However development is a rather different field from memetics. It simply doesn't seem appropriate to invoke developmental concepts to criticise memetics. Development is treated as a black box in memetics, just as it is in genetics.

The term phoneme provides another neat example of the use of the "-eme" suffix to denote a unit which divides a continuous stream into discrete parts. Few would debate the scientific usefulness of the concept of a "phoneme" - but, equally, few would claim that their starts and ends are necessarily going to be clearly defined down to the microsecond in any given audio stream. The starts and ends of phonemes are fuzzy - and sometimes it isn't even clear whether you have one phoneme - or two.

Hull (2001, p.120) addresses the "particulate" criticism, as follows:

Some authors argue that no general analysis of selection process equally applicable to biological and conceptual evolution is possible because genes are "particulate" while the units in conceptual replication are highly variable and far from discrete. In point of fact neither biological nor conceptual replicators are all that particulate. In both cases, the relative size of the entities that function either as replicators or as interactors is highly variable and their boundaries sometimes quite fuzzy.
Blackmore is among those who have tried to sort this issue out:
The population approach, they say, does not imply that cultural evolution is analogous to genetic evolution; nor does it depend upon “discrete, faithfully replicating, genelike bits of information.” I quite agree, but then so would Dawkins and most other memeticists.
Then there's the issue of whether memes are "discrete" - when they are inside brains. Here's Blackmore (2006) on the topic:

First, Mesoudi et al. claim that “A common assumption of memetics is that cultural knowledge is stored in brains as discrete packages of semantic information” (target article, sect. 3.5.2, para. 1). I disagree. This was not assumed by Dawkins (1976) when he invented the term “meme” thirty years ago this year, nor by Dennett (1991; 1995), nor by me (Blackmore 1999; 2001). Aunger (2002) does take this view, but otherwise it is mostly the critics of memetics who do so – aiding their attempts to demolish memetics.
Lastly, it's possible to argue that genes are discrete - in the sense that they are always be divided into discrete nucleotides sequences - while memes have no equivalent to nucleotides - and so are less "discrete" - in this sense.

This is reading the relationship between genes and memes a bit too closely - in my opinion. The intended idea is that memes are like genes - in that they convey heritable information down the generations - not that there's some kind of "cultural equivalent" to nucleotides!

Genes are, in fact, not as discrete as all that. The breakdown in "discrete" conceptions of the gene was covered by Dawkins in 1982:

Stent makes the more important point that my unit is not precisely delimited in the way that the cistron is. Well, perhaps I should say “in the way that the cistron once seemed to be”, for the recent discovery of “embedded” cistrons in virus ФX174, and of “exons” surrounding “introns” must be causing a little discomfort to anyone who likes his units rigid.
Mesoudi, Whiten and Laland (2006) (who Susan quotes above) go on to explain that genes are not all that discrete either:

However, the same putative “criticism” could equally be levelled at modern concepts of the gene (Laland & Brown 2002). As documented by Portin (1993; 2002), the concept of the gene has undergone significant changes through the past 150 years. The classical view, held from the time of Mendel (1866) until the 1930s, saw the gene as an indivisible unit of transmission, recombination, mutation, and function. That is, a gene is a unit of information that is transmitted whole, within which no recombination occurs, which mutates independently from other genes, and which produces a single molecular product (as captured by James Watson’s famous canon, “DNA makes RNA makes protein”). This simple and dated gene concept seems to be the view of the gene held by many social anthropologists who are critical of memetics. Advances in genetics since the 1930s, however, have shown this unitary gene concept to be inadequate and overly restrictive. Further reconceptualisation began in the 1970s following the discoveries of such phenomena as overlapping genes, where the same stretch of DNA codes for more than one protein; movable genes, DNA sequences that move around the genome; and nested genes, which reside inside other genes. Such revised conceptions have continued in the wake of modern discoveries, such as alternative splicing, nuclear and messenger RNA editing, cellular protein modification, and genomic imprinting.
We now know from that "discrete" conceptions of genes have the added complications of methylation, and histone protein wrapping to deal with. Basically the idea that genes are particularly "discrete" is - in a lot of ways - out of date science.

Genes also exist as sequences of information inside computers, and there, like memes, you can slice and dice them however you choose. Genes likely exist inside aliens on other planets - and the claim that they are always discrete thus seems like dubious and unproven assertion to me. If you think that genes are sequences of nucleic acid base pairs, you have a conception of genes and genetics that copes poorly with our distant ancestors, our distant descendants, and aliens - and need to expand your conception of what the term "gene" means.

So: the answer to the question of: "are memes discrete?" depends on what you mean by the question. Memes are often defined as being some kind of fragment of heritable cultural information. In that respect, they are the same as Boyd and Richerson's "cultural variants". Pieces of information are discrete - so memes are discrete. However, that isn't intended to imply that culture necessarily has natural "fracture lines" - that allow it to be naturally divided up at specific locations - in the same way that many chocolate bars have scored fracture lines. Memes can be divided practically anywhere during the transmission process. That's much the same as with DNA genes - which can be divided practically anywhere during meiosis.

I hope this article sorts out any confusion on the topic. Feel free to let me know if anything is unclear.

No comments:

Post a Comment