Dialogue

http://journals.cambridge.org/DIA

Additional services for Dialogue:

Email alerts: Click here

Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here

Knowledge and the Flow of Information Fred I.

Dretske Cambridge, MA: MIT Press, 1981. Pp. xiv,
273. \$18.50 (U.S.)

Douglas Odegard

Dialogue / Volume 21 / Issue 04 / December 1982, pp 778 - 779

DOI: 10.1017/S0012217300023969, Published online: 05 May 2010

Link to this article: http://journals.cambridge.org/abstract_S0012217300023969

How to cite this article:

Douglas Odegard (1982). Dialogue, 21, pp 778-779 doi:10.1017/
S0012217300023969

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/DIA, IP address: 61.129.42.15 on 01 May 2015

778 Dialogue

Knowledge and the Flow of Information

FRED I. DRETSKE
Cambridge, MA: MIT Press, 1981. Pp. xiv, 273. $18.50 (U.S.)
Dretske's goal is to help construct a bridge between philosophy and the cognitive
and computer sciences, by developing an information—theoretic account of
what we refer to when we talk of knowledge, perception, and belief. The aim is
not conceptual analysis, but extensional equivalence. He has an underlying hope
of doing for epistemology what identity theorists have tried to do for philosophy
of mind, viz. fashion a pro-physicalist vocabulary that can be used to describe
the selected territory—although his surface objective is simply to improve our
philosophical understanding by tapping what for him is a wealth of relevant
material in the sciences.
Part 1 lays out some key technical concepts in a way that presupposes no
special background. Parts 2 and 3 apply the technical apparatus in a study of
knowledge, sensation, concepts, meaning, and belief. There are also extensive
Notes to each chapter, containing useful substantive as well as bibliographical
material. My comments will be restricted to some of his views on knowledge and
sensory experience.
He disavows ajustificationist account of knowledge partly on the ground that
the concept of justification operates as a primitive, and he is (rightly) uneasy
about having to make heavy use of an intuitive concept of justification. But the
primitive concept in Chisholm is "being more reasonable than" and in Pollock it
is "prima facie justified", and the uses of those concepts is not left entirely to the
whim of intuition. Also, Peter Klein, in Certainty (also published in 1981),
supplies a partial model of justification that may help to reduce the reliance on
intuition. At any rate, we needn't assume that the task of finding an adequate
analysis ofjustification is any less imposing than that of carving out a whole new
information-theoretic approach to knowledge. Although the new venture should
be encouraged, it need not be regarded as vital to the survival of epistemology.
He explains perceptual knowledge as belief caused by, or causally sustained
by, the receipt of information and uses this explanation to solve a number of
familiar epistemological problems. For instance, he can cope with Gettier-type
cases more easily than ajustificationist, since, unlike justification, information
cannot be acquired or transferred when the proposition involved is false. The
information-theoretic account has a clear advantage in this respect.
He wants to escape having to claim knowledge of each conjunct in a conjunc-
tion while also claiming knowledge of the conjunction's negation. He does this
by denying, in the case of a large fair lottery, that he has information, concerning
each particular losing ticket, that it will lose. Indeed, this is his only escape
route, since he accepts the conjunctive principle that knowing each of several
conjuncts involves knowing their conjunction as well. But the principle is
dangerous. It means that I cannot list a large number of things that I in fact
happen to know and consistently disclaim knowledge of their conjunction,
which seems an odd result. Justificationism has an advantage in this respect,
since the accumulation of evidence that justifies each individual claim need not
be strong enough to justify the conjunctive claim.
He replies to a variety of skeptical arguments. For instance, he resists Unger's
attempt to exploit the absolute character of knowledge by pointing out that other
 Book ReviewslComptes rendus 779

relevant absolutes do exist: e.g. a wallet can be absolutely empty even if it

contains dust particles. He also criticizes forms of skepticism that mistakenly
assume that knowledge requires both information from a source and information
about the channel through which the first-order information is delivered. His
discussion here is structurally similar to positions that deny that knowing re-
quires knowing that the conditions of knowledge are satisfied.
The moves against skepticism are generally effective, but they raise an in-
teresting methodological question. Suppose that Dretske has successfully de-
fended the position that beliefs are sometimes causally related in a suitable way
to the receipt of information. Does this mean that standard skeptical objections
fail to establish that, in the relevant ordinary sense of "knowledge", we do not
have knowledge on those occasions? Skeptics can, I think, concede that he has
protected the information-theoretic claims and still maintain that ordinary
knowledge claims carry implications that we are not entitled to accept on such
occasions. If the skeptics are right, we rarely, if ever, have knowledge in an
ordinary sense, even though we often satisfy conditions for being technically
informed. In that case, Dretske loses his sought-after match between ordinary
knowledge claims and technical information-theoretic claims. In defending the
latter claims against skepticism, he should therefore do more to establish that
analogous replies are available on behalf of ordinary knowledge claims.
He draws a firm distinction between sensory experience and cognitive
mechanisms, explaining the former as an analogue coding device and the latter
as digital coding devices. He sees no need for sensory "images", even in cases
where psychologists standardly talk of persisting images. He says things like,
"What persists is a structure in which incoming information about a pictorial
array is coded in preparation for its cognitive utilization" (150). Although the
word "structure" may be perfectly adequate for the purposes of cognitive
psychology, it hardly achieves much ontological penetration. Consequently,
more needs to be said before the adequacy of his metaphysics can be assessed.
He is properly critical of attempts to identify the object of sensory experience
in causal terms. If someone rings a doorbell by pushing a button, Dretske thinks
that the button-pushing has relevantly the same causal standing as the bell-
ringing; yet normally we hear the bell-ringing and not the button being pushed.
He tries to accommodate such cases by identifying the object of sensory experi-
ence as an object the properties of which are represented in a primary way. Since
receiving information about the button depends upon receiving information
about the bell and not conversely, only the bell's ringing is represented in a
primary way: hence we hear only the bell. Yet in Berkeley's coach example, it
seems that only the sounds are represented in a primary way,, since receiving
information about the coach depends upon receiving information about the
sounds and not conversely. Yet it also seems that we hear both the sounds and
the arriving coach. Further adjustments therefore seem necessary.
The book is clearly written, with a lively style and no pretensions. It probably
will be most useful to scientists who are already familiar with the terminology
and who seek a philosophical overview. Yet because Dretske is a good philoso-
pher, there is also much for philosophers to digest. Given the uncertainty of
philosophical conclusions, the book cannot contain information in the technical
sense. Nevertheless, it is instructive in an ordinary sense.
DOUGLAS ODEGARD University of Guelph