Dretske, Fred (1932–)

views updated

DRETSKE, FRED
(1932–)

Born in 1932, Fred Dretske received his PhD from the University of Minnesota. He is emeritus professor of philosophy at Stanford University and professor of philosophy at Duke University. Since the early 1970s Dretske's work has been at the center of a number of key disputes in epistemology and the philosophies of perception, mind, and consciousness. Despite their range, two basic motivations unify Dretske's writings: the need to understand the mind in relation to its environment and a steadfast naturalistic outlook on the mind and its operations.

In Seeing and Knowing (1969), Dretske emphasized a form of perception that he labeled "nonepistemic seeing." This is an direct relation between perceiver and object not involving any particular conceptualization of the perceived object nor requiring any particular beliefs about it. Dretske argued that the concept of nonepistemic seeing is fundamental to understanding perception and the place of the mind within the world. Without it we have no way of understanding how we can all experience the same world despite having widely divergent concepts and beliefs. Via the notion of nonepistemic seeing, we can strip away our cognitive interpretive faculties and be left with the content of perception: the objects of the world we perceive.

Attention to nonepistemic seeing also undercuts the old idea that seeing involves "direct acquaintance" with some mysterious mental object, from whose incorrigibly known features we can only infer the existence of the external world.

We might naturally ask, What is the basic enabling feature of nonepistemic seeing? The answer is that there is an internal state of the perceiver that "carries the information" about the seen object. In Knowledge and the Flow of Information (1981), Dretske developed a sophisticated, elaborate, and technical account of information and its role in knowledge, thought, and perception. Building on his earlier epistemological work, Dretske analyzed knowledge in terms of informationally caused beliefs. To take one of Dretske's famous examples (from 1970), someone at a zoo knows that there is a zebra in front of him if that very information is causing his belief. Whether the appropriate information is available depends on the context of its occurrence, since information is a function of the relevant alternative messages that a signal could deliver. If there are lots of cleverly disguised mules about, his belief may not be caused by the information that there is a zebra in front of him (since the presence of that information may depend upon how much the perceiver knows about how zebras look), and thus he may not know that there is a zebra in front of him.

Dretske's account has an infamous consequence: the denial of inferential knowledge via known entailments. If our subject knows that these (the creatures in front of him) are zebras and that it follows from x 's being a zebra that x is not a disguised mule, then it would seem he could infer that these are not mules and hence know this. But how could he know this when he is utterly unable to distinguish a painted mule from a zebra? Dretske asserted that someone could know that something is a zebra without knowing that it is not a painted mule. While the mechanics of information allow this "paradox," the general issue remains highly contentious.

How can information or content play a causal role in the world? This is a key issue for Dretske's project of naturalizing the mind, or as Dretske puts it, baking "a mental cake with physical yeast and flour." Crudely put, the problem is that all behavior appears to have purely physical explanations that need appeal not to any information but only to local causes. We know how charge, momentum, and gravity cause events; informational causation seems to be something else altogether and quite mysterious.

In Explaining Behavior (1988), Dretske addressed this problem via a distinction between "triggering" and "structuring" causes. If C is an efficient or local cause of M, it is a triggering cause. The structuring cause of M is "the processes which explain why C causes M " (p. 91). In particular, the structuring causes of behavior are the historical processes that institute the triggering causal links between information-carrying mental states and behavior. Two aspects of this sort of explanation must be distinguished. The first comprises the historical processes—evolution, learning, or design—by which some internal state comes to have an "indicator function." The second is the deployment of the indicator to modify behavior because of what is indicated. Dretske maintains that, while a great many states serve to carry information of one sort or another and while these states certainly do enter into causal relations, only learning can bring about systems in which the carried information causally explains why these states cause the behavior they do. Only in learning do "we see meanings … doing some real work in shaping behavior" ("Dretske's Replies," p. 201).

The emphasis on learning leads to obvious difficulties. It seems to imply that innate mental states cannot explain behavior (perhaps cannot even cause behavior and maybe cannot even exist). In Naturalizing the Mind (1995), Dretske, elaborating his view, allowed that evolutionary processes can produce representational mental states that do not depend on learning for their efficacy. There he distinguished systemic and acquired representational states. The former are the experiential qualities of experience. Their content is nonconceptual and fixed by biology. Systemic representation underpins nonepistemic perception. And it enables acquired representations, a form of which constitute beliefs and the other propositional attitudes. This distinction allows for a more nuanced theory of mind and forms the basis for an ambitious representational theory of consciousness. In Naturalizing the Mind, Dretske also develops an intriguing theory of introspection in which our self-knowledge involves a special application of mentalistic concepts to our own experience.

Dretske continues to claim that representation is essentially linked to the external environment. In his theory of consciousness, the experiential nature of mental states depends on their representational properties (and all conscious states, including such "pure" sensations as pain or tickles, are conceived of as representational). While promising a complete naturalization of the most troublesome feature of the mind, representational properties have a downside. Since representational properties are determined and constituted by relations with the environment, Dretske's views have the consequence that a newly created duplicate of a person would utterly lack consciousness. Many find this less than plausible.

Be that as it may, Dretske presents an elegantly unified and comprehensive theory of mind that makes our mental lives fully causal in an entirely naturalistic way. His views, in their clarity, argumentative care, and intellectual honesty, exemplify the best features of modern analytic philosophy.