Models of Mechanisms: The Case of the Replicator Dynamics - Mechanism and Causality in Biology and Economics - page 92

Fig. 5.6 Fliegende Bla¨tter
(Oct. 23, 1892, p. 147, Nr.
2465)
organisms. For example, it omits representations of how organisms reproduce and
instead describes the stage as a general process of reproduction in all its possible
forms. So if the BRD is an abstract representation of a class of mechanisms, why
should the general RD not be an even more abstract representation?
Furthermore, the fact that both the BRD and the learning models use the RD for
their representational tasks seems to provide evidence that indeed there is an
abstract mechanism instantiated both in the more concrete biological and social
mechanisms and that this more abstract mechanism is represented by the RD model.
Because the RD contains those features shared by the BRD and the learning models,
it might seem plausible to conclude that the RD represents that abstract causal
structure shared by the biological and the social mechanisms.
Against this appearance, I will now argue that the general RD model is not a
representation of an abstract mechanism, instantiated by both the biological and the
social mechanisms. Rather, the way the two disciplines 'fill in' the RD model in
order to represent their respective mechanisms differs considerably. Users of the
RD model, when filling it in, make systematically different kinds of idealisations,
depending on whether it is interpreted in economics or in biology. This leaves little
to be shared between the respective represented mechanisms - little that could be
represented by a single RD, however interpreted. Instead, the RD model faces an
idealisation gap : it can be interpreted either biologically or in one of the learning
senses, but it cannot be interpreted to capture the essence of all, because there is
little essence to capture. To clarify my argument, let me illustrate it with a joke.
The joke's not mine - it was published 120 years ago in the Fliegende Bla¨tter ,a
German satirical weekly. Most philosophers know its subject, the duck-rabbit, from
Wittgenstein's discussion of aspectual perception or from Kuhn's discussion of a
paradigm shift. What those discussions ignore is the way the joke was posed, as
shown in Fig. 5.6 . The German headline reads 'which animals are most similar?',
and the answer is 'rabbit and duck'.