one individual, each of which conforms to the norms of rationality, but
which fail to interact (1982a). Now we do not deny that there may be many
diVerent parts and levels to human cognition – indeed we will make
frequent use of this idea ourselves. But we do think it highly implausible
that we should need to postulate divided persons or minds in order to
explain irrationality. Rather, in many cases a belief can simplyslip our
minds– remaining real, but failing in the particular circumstances to
become active in reasoning.
2.2 DennettDennett’s position (developed in his 1971, 1981, 1987, 1988a, 1991b) is a
diYcult one to grasp. Its front end appears unashamedlyinstrumentalist,
but he adds a rider which appears to cancel out the anti-realism of his
picture. We think his position is unstable, and that he really cannot have it
both ways.
Presenting the boldly instrumentalist front endWrst, Dennett (1981)
declares: ‘What it isto be a true believer is to be anintentional system,a
system whose behaviour is reliably and voluminously predicted via the
intentional strategy.’ He appears to be maintaining that ascription of
beliefs, desires, and so on, is produced by the adoption of a particular
predictive/explanatory stance – a stance justiWed by nothing other than its
predictive success and practical utility. He introduces theintentional stance
by contrasting it with two others – thephysical stanceand thedesign stance.
The physical strategy uses knowledge of laws of physics (and/or chemis-
try), combined with details of physical states and constitution, to predict
outcomes. This may seem the most fundamental and scientiWcally well-
founded approach. But it will rarely be the most convenient, and often is
not feasible when trying to cope with a system of any degree of complexity.
Thedesign strategypredicts that something will behaveas it was designed
to behave. This strategy is obviously useful in relation to artefacts like
motor cars, computers, and alarm clocks; and also to functional biological
systems like hearts, livers, pistils, and stamens, which have acquired a
design through evolution.
Theintentional stanceis a further option (and one might wonder why
there should be just three):
Here is how it works:Wrst you decide to treat the object whose behaviour is to be
predicted as a rational agent; then youWgure out what beliefs the agent ought to
have, given its place in the world and its purpose. Then youWgure out what desires
it ought to have, on the same considerations, andWnally you predict that this
rational agent will act to further its goals in the light of its beliefs. (Dennett, 1981,
p.57)
Two varieties of anti-realism 29