Seeing Clearly through Fuzzy Speech

SEEING CLEARLY THROUGH FUZZY SPEECH William C.Stokoe The years of searching inthe dark for atruth that one feels but cannot express, the intense desire and the alternations of confidence and misgiving until one breaks through to the clarity and understanding are known only to him who has himself experienced them. -Einstein Speech is a wonderful capability that only our species possesses . Those who have made speech the focus of their research have added much to human understanding. But others, who seem to suppose that speech and language are two names or aspects of the same thing, have often confounded knowledge. Speech is physical-sound produced by the voice and received by the sense of hearing, but language is mental; it begins in neural activity and integrates sensory and motor systems-plus knowledge, intellect, will, desire, and much else-into thought and social interaction . Failure to see the separation between language and speech has resulted in wasted intellectual effort, because ever since the time of Aristotle, most students of speech and language have attempted to apply crisp categorical logic to something that defies such precise categorization. Speech is by nature imprecise. Not only do speakers of different languages speak differently; speakers of the same language have various ways of making the sounds of their own language. The term "dialect" labels the speech of speakers from different regions or classes or with different first languages. This kind of difference is often artistically exploited: An actor learns to speak in the dialect used by the character portrayed; a writer uses odd spelling and phrasing to represent the speech habits of a region or class. Individual speech differs at least as much as individual fingerprints . The expert impersonators on "Saturday Night Live" almost make us believe that we are listening to the prominent personality whose utterances are being turned into high comedy. @1994 Linstok Press See note inside cover February 17, 1994 SLS 82 If it is true that one cannot step twice into the same river, it is equally true that one cannot repeat with complete precision something one has just spoken. Even a single speaker does not speak exactly the same way in different circumstances. Individual differences are real as well. Despite this wide variation in the speech of individuals, groups, and whole nations, students of language have tried for centuries to describe spoken languages with rigorous logic, proposing (some perhaps hoping that they are imposing) laws or rules. They may recognize that variation occurs and that context, the knowledge speakers possess, is important; but they ignore much more often than they attempt to address it the imprecision of speech-and language. One result is that contending disciplines have developed to counter the rigorously logical focus of linguistics and language description. One ofthese is sociolinguistics , the study of what speakers actually say. An anthropological approach to describing how social variables influence speakers' behavior has been called "the ethnography of speaking." And recently , cognitive linguists are beginning, from the fact of variation and uncertainty, to look for the nature of language in the human ability to know, remember, form concepts, imagine, and make decisions. Much of today's "standard" linguistic description goes back to speech, not language, and to Roman Jakobson's belief that "we perceive each phoneme categorically." For example, speakers of English recognize as different words that have a d or an n in the same position (dote vs. note; cad vs. can; buddy vs. bunny) but ignore the difference between the actual sound ofd in words like dig anddog, or the sound of n in near and not. Audiologists and phoneticians have shown that the actual acoustic difference between d and n may be no more than a few microseconds in the length of vocal fold vibration. And yet linguists insist that d and n belong to two distinct crisp sets they call phonemes. (A less obvious example is the difference between 1and m. On many mornings I listen to a radio announcer reading the news at a Washington radio station. Whether he or others speak his name on the program it always sounds like "Bill Redman;" but when I saw his photograph captioned "Bill Redlin" in the station...