than having the sort of non-inferential access to their own thinkings
necessary for those thoughts to count as conscious ones.
(Gazzaniga – 1992, 1994 – thinks that this is the only kind of access to
our own thoughts which we modern humans enjoy. He believes that there
is a specialist theory-of-mind system – which he dubsthe interpreter–
located in the left frontal lobe, whose business it is to construct, by
self-interpretation, a meta-narrative about the course of the subject’s own
mental life. Our view is that in that case there is no such thing as conscious
thinking. We return to this point below.)
With the arrival of language, we humans would then have been capable
of entertaining imaged sentences of natural language, in ‘inner speech’, to
whose forms and contents we would have had non-inferential access, by
virtue of their availability to our mind-reading faculty – hence qualifying
them as ‘conscious’ (see chapter 9). If these sentences had then somehow
come to occupy the causal roles distinctive of thought, then this would
have meant both that we became capable of conscious conceptual thinking
for theWrst time and that natural language sentences would have been
constitutive of such thinkings.
At issue here is the question of the causal role of ‘inner speech’. The
phenomenonof inner speech is not in doubt. Nor should be the claim that
we have non-inferential access to the events in inner speech, hence allowing
them to qualify as ‘conscious’. The question is whether or not sentences in
inner speechthemselvesoccupy the causal roles which are distinctive of the
thoughts which those same sentences express. One view is that they do. On
this account, it isbecauseI token in auditory imagination the sentence,
‘The world is getting warmer, so I must use less fuel’, for example, that I
may thereafter be found walking rather than driving to work. The other
view is that they do not. On this account the thought itself is carried by a
sentence of Mentalese, say, and it is the Mentalese vehicle which has the
further eVects in cognition and action distinctive of a thought with that
content. But even on this weaker view the imaged sentence need not be
epiphenomenal – it may, for example, play a role in enhancing memory,
thus rendering possible sequences of thought which would otherwise be
too complex to entertain (JackendoV, 1997; Clark, 1998; Varley, 1998).
It is by no means easy to adjudicate between these two views of the
causal role of ‘inner speech’. But one point in support of the ‘thinking in
language’ hypothesis, is that it may turn out to be a presupposition of our
belief that we do engage inconsciouspropositional thinking at all. If
‘inner speech’ does not count as thinking, then probably we only ever
know of our own thoughts by swift self-interpretation. This conditional
conclusion is strongly supported by a rich body of data coming out of the
social psychology literature, where it has been found that there are many
218 Forms of representation