Page 136 - Contribution To Phenomenology

CONNECTIONISMAND PHENOMENOLOGY 129

structure they exhibit, then they appear to be closer to the mental than
the physical. Husserl's solution to this dilemma seems to be most often
that he thinks of them as sedimented beUefs (e.g., the 4th Cartesian
Meditation § 31 or Formal and Transcendental Logic §§ 42 ff.) that could
can be retrieved and made conscious again. Thus habitualities and
dispositions become a kind of active memory, non-conscious representa-
tions that can in principle always been retrieved again and brought to
direct consciousness.^^
Now this seems problematic to me in two respects: first of all, it is
not clear to me that it is indeed the case that sheer reflection alone is
ever capable of resurrecting all of the beliefs that are supposed to lie at
the bottom of these dispositions.^' Perhaps we can in some cases, but if
not in all of them, then the dilemma I just outlined still presents itself
for a number of them. And secondly, even if we could make all of them
expUcit, it is not clear that the best way to think of them is as sedi-
mented belief states, since it does not necessarily follow that they
therefore must have ever had the status of conscious behefs just because
we can now raise them to the level of consciousness.
Here is where I think connectionist models come in and are actually
quite compatible with Husserl's accounts of passive synthesis as governed
by structures of association. For as I mentioned in Section I of this

^^ This is not the only approach Husserl takes. In his analysis of the prepredicative
realm in Experience and Judgment (§§ 15-28), for instance, Husserl suggests that at the
most basic level, our cognitive life is permeated by "tendencies" that derive from our past
experiences apart from and prior to predicative judgment. To these tendencies or
"interests," which constitute our "horizonal consciousness," Husserl also gives the name
"habitualities." Interestingly enough, however, the phenomenological warrant Husserl
provides for positing such tendencies is not that they are directly discernable, but that they
can be discerned through an analysis of the role they play in the constitution of objects for
us in cognition. In one sense, then, they are even more problematic than habitualities
thought of in terms of stored judgments, since it is not clear in which sense they ever were
present as such to be stored and retrieved. One would have to read "sedimented" no
longer simply as "removed from the realm of active attention," but instead as "brought
about or caused" by past experience so that it would be a kind of belief which at no time
ever was consciously entertained. I think that this approach is phenomenologically
appropriate as a description of a large portion of our mental life, but that does not mean
the problems concerning its ontological status are any less problematic than those
connected with dispositions thought of as stored-away judgments.
^' One might take Husserl's move from mere "phenomenological description" to
analyses of "intentional implications" to be an attempt to come to terms with these kinds
of problems.