From Content and Consciousness. London: Routledge, 1969.
Dennett argues that mental phenomena cannot be explained in ’sub-personal’
(i.e., neurophysiological) terms, because the mental is an autonomous domain
with its own type of explanation. Because of this, questions about mental
phenomena should not be answered in sub-personal terms, despite the obvious
fact that mental phenomena depend upon sup-personal phenomena in some
important sense. Dennett thinks this because he cannot see how a sub-personal,
mechanical explanation can apply to sensations such as pain. For a mechanical
explanation to apply, mental sensations would have to be analyzable qualities.
But Dennett believes that they are unanalyzable, and thus that mechanical
explanations are irrelevant.
Dennett considers three questions about pain that a mechanical explanation
might attempt to answer: How does a person distinguish pains from other
sensations? How does a person locate pain? Why do we abhor pain? In each
case, Dennett attempts to show that the mechanical explanation fails.
In the ﬁrst place, a mechanical explanation of our ability to distinguish
sensations from one another presupposes that the distinguishing is a personal
activity. But Dennett does not think this is right: if distinguishing were an
activity, then we could appeal to the qualities of the sensations in question in
order to diﬀerentiate them. But to appeal to qualities is to identity the quality
in question. Identifying a quality, in turn, requires that we either describe the
quality, or otherwise that we ostend it. Since we can do neither of these things
, it follows that we cannot identify the qualities of sensations and thus cannot
appeal to them in distinguishing sensations. But if we are not doing this, then
we are not engaging in an activity at all, and no mechanical explanation is
relevant to what is happening. Our ability to distinguish sensations is thus a
brute fact not in any further need of explanation.
Dennett thinks the same conclusion must be reached when we come to the
question of our ability to locate pains. A person is not doing anything when she
locates a pain, since doing something would require an appeal to describable
qualities of the pain. Whatever her brain is doing when she is locating the pain,
she is not doing anything, for if she were it would be a process, and such a
process is impossible.
And thus the question of why we abhor pain has the same answer: we simply
do abhor pain, and to explain this brute fact in mechanical terms is to provide
a non-explanation: a mechanical explanation will cite unanalyzable qualities
of the pain which we abhor, but of course this says merely that there is some
contingent quality about pain that we do not like, for some reason. The reason
cannot be explicated because pain is unanalyzable; therefore, the mechanical
explanation is really no explanation at all.
Dennett concludes with an example meant to diﬀerentiate the personal and
sub-personal levels of explanation in action. When a person touches a hot
stove, the personal explanation is merely that she had a sensation of pain in a
speciﬁc place, and that sensation caused her to withdraw her hand. No further
explanation is possible, and these facets of the explanation are mere brute facts.
If someone wishes that the situation should be explained further, then one must
leave the personal level of explanation and thus leave talk of ’pain’ behind in
order to enter the domain of sub-personal explanation, where a full account
of the person’s brain, nervous system, and movements can be given. Such an
explanation will be highly technical and will involve talk of aﬀerent-eﬀerent
connections and ’pain behavior.’ But these terms cannot be identiﬁed with the
feeling of pain. ’Pain’ is non-referential; it is thus the philosopher’s job to ﬁgure
out how ’pain talk’ relates to neural events or talk about neural events.
Primary Argument: Mechanical explanations are not appropriate for mental phenomenona.
1. A mechanical explanation of sensations ought to explain either how we distinguish sensations, how we locate them, or why we have certain reactions to certain sensations.
2. Mechanical explanations cannot account for our ability to distinguish sensations from one another.
S1. Mechanical explanations require that distinguishing be an activ-
S2. If distinguishing were an activity, then we would be able to appeal
to qualities of the sensations in question.
S3. Therefore, if a mechanical explanation of distinguishing is to be
viable, we must be able to appeal to qualities of sensations. [S1,S2]
S4. To be able to appeal to the qualities of a sensation, we must be
able to identify such qualites.
S5. To identify such qualities, we must be able either to describe
them or to ostend them.
S6. Therefore, if a mechanical explanation of distinguishing is to be
viable, we must be able to identify qualities of sensations. [S3,S4]
S7. Therefore, if a mechanical explanation of distinguishing is to be
viable, we must be able to either describe or ostend such qualities. [S6,S5]
S8. We can neither describe nor ostend the qualities of sensations.
S9. Therefore, mechanical explanations of distinguishing are not vi-
3. Mechanical explanations cannot account for our ability to locate sensa-
S1. Whatever the brain is doing, we locate sensations without engag-
ing in a personal activity.
S2. Mechanical explanations presuppose a personal activity is taking
S3. Therefore, a mechanical explanation is not appropriate for our
ability to locate sensations.
4. Mechanical explanations cannot account for why we abhor pain.
S1. A mechanical explanation of our avoidance of pain would require
pain to be an analyzable sensation.
S2. Pain is unanalyzable.
S3. Therefore, mechanical explanations are not appropriate with re-
gard to our abhorence of pain. [S2,S1]
5. Therefore, mechanical explanations cannot account for our ability to dis-
tinguish sensations, nor for our ability to locate them, nor our reactions
to them. [2-4]
6. Therefore, mechanical explanations are not appropriate for explaining sen-
sations, and thus mental phenomena in general. [5,1]