The Case for Commentary

The article `Why the Thought Contagion Metaphor is Retarding the Progress of Memetics' argues that the field of memetics took a wrong turning in the early 1980s, with the result that all subsequent memetic theorising has driven down the same blind alley. The wrong turning was to substitute the earlier, flexible definition of a meme (which in the article I term Dawkins A) with a far narrower and more specific definition as a unit of information in a mind or brain (termed Dawkins B).

Since science generally abhors vagueness and thrives on precise defiintions, such a move naturally seemed like a step forward. In this case, however, it was counter-productive, since:

a) Direct observation was relegated to a subsidiary role, while inference and speculation took centre stage. The derivation of internal memes from behaviour relies on acts of the most naive inference. This is not the same situation as that of a classical geneticist of the early 20^th century, since genes were not naively inferred from phenotypes, but were indirectly demonstrable as segregation ratios, patterns of independent assortment or linkage. There is no equivalent phenomenology in memetics. Internal memes are not even indirectly demonstrable. Their existence cannot be independently verified in any way.

b) Quantification became impossible. We cannot quantify any internal mental unit, and quantification is as essential as precise definitions (if not more so). While it is possible to quantify behaviour, either roughly in the field or more precisely in the laboratory, such quantification of behaviour does not translate into quantification of internal memes, unless the internal meme and the behaviour exist in a one-to-one correspondence. There is no evidence for this.

c) The theory lacks explanatory power. To say that behaviour can be explained in terms of something which cannot be observed is unhelpful. In any case, there is also no evidence that replication of behaviour is necessarily correlated with replication of internal mental states. We have no grounds for believing that thoughts are contagious, and such a view trivialises psychology.

d) Lynch's calculus of mnemon conjugations is incompatible with any present thinking in either cognitive sciences or linguistics. The picture of the mind as a stack of `awareness of statement x' and `belief in statement x' would require the mind to store language in a manner reminiscent of a computer RAM. Such possibilities were effectively excluded by Chomsky in the early 1950s.

I should therefore like to ask the commentators if they consider `internal' memetics to be still tenable in the light of these criticisms. If so, how do they believe it should be constitued as an experimental science (as opposed to a pseudoscience dealing in postulated unobservable entities, where neither verification nor falsification are possible)? If the commentators agree with me that `internal' memetics is untenable, do they tend towards my quasi- behaviourist version of `external' memetics, or have they a third alternative?