Formalisations of Evolutionary Biology 495highly abstract and are far removed from the empirical phenomena to which they
will be applied. As I have argued [Thompson, 1983; 1985; 1986; 1987; 1988a;
1989], this feature is one of the major strengths of this conception since it accords
with the actual scientific practice of using a multiple number of auxiliary theories
(from mathematics and other domains of science) in the application of a particular
scientific theory to phenomena.
Suppes suggests two reasons why the syntactic account is so widely and strongly
held, despite what he argues are logical and practical weaknesses with it. First,
philosopher’s examples of scientific theories are usually fairly simple and, there-
fore, easily able to be given a linguistic formulation. Not surprisingly, most exam-
ples used to explicate and defend the syntactic account are drawn from Newto-
nian mechanics — and from a reasonably simple and sketchy account of it. Also
not surprisingly, advocates of the semantic account discuss complex theories such
as quantum mechanics (see [van Fraassen, 1972]), learning theory (see [Suppes,
1962]), and evolutionary theory (see [Beatty, 1980a; 1980b; Lloyd, 1983; 1984;
1986; 1987; Thompson, 1983b; 1985; 1986; 1987; 1988a]). Second, compared to
the view that theories are axiomatic-deductive structures formulated in first-order
predicate logic and partially interpreted by correspondence rules, far more math-
ematical sophistication is required to formulate a theory on the semantic account
and to characterise explanation, prediction, and validity.
During the late 1960’s and the 1970’s the semantic account was consolidated
and extended by a number of philosophers from a variety of perspectives (see,
for example: [Suppe, 1967; 1972a; 1972b; 1974; 1976; van Fraassen, 1970; 1972;
1980; Sneed, 1971; Stegmuller, 1976]). Despite this coalescing of the account,
there were, and continue to be, important differences of motivation and structure
between the views of these advocates. Suppe, for example, is a scientific realist
(see [Suppe, 1988]), whereas van Fraassen is a constructive empiricist (see, [van
Fraassen, 1980]). During the 1980’s John Beatty, Elisabeth Lloyd and I extended
and applied the semantic account to biology and, in particular, evolutionary theory
and genetics (see [Beatty, 1980a; 1980b; Lloyd, 1983; 1984; 1986; 1987; Thompson,
1983b; 1985; 1986; 1987; 1988a; 1988b; 1989]).^17
One of the major features of theories on the semantic account is that the class of
models, which specify directly in mathematical English the behaviour of a system,
is an extra-linguistic, highly abstract entity which is most often quite removed from
the phenomena to which it is intended to apply. For example, laws do not describe
the behaviour of objects in the world, they specify the nature and behaviour of an
abstract system.
The application of the model(s) to a particular empirical system requires the
extra-theoretical assertion that the model(s) and the phenomena to which they are
intended to apply are isomorphic (in algebraic contexts such as set theory, groups
and rings) or homeomorphic (in topological contexts). This is an extra-theoretical
(^17) Frederick Suppe applied the semantic account to aspects of taxonomy in the 1970a. But this
seems to be an isolated application to biology prior to the early 1980s (see Suppe [1989]).