To produce speech, the speaker expels air from the lungs through the vocal tract, and modulates the airflow by changing the vocal tract’s shape and moving the tongue against the lips, teeth, alveolar ridge (behind the upper teeth) and palate, thus creating speech sounds. Voiced sounds (vowels, many consonants—for example [d], [l], [m] and [v]) involve audible vibration produced by the vocal folds in the LARYNX; unvoiced sounds (other consonants—for example [t] and [f]) are produced without such vibration. The speech sounds are produced in a smooth continuous manner which has the effect that the articulation of a sound overlaps with the articulation of adjacent sounds (known as co-articulation). One result of this is that the same PHONEME can be produced differently in different phonetic contexts (compare the [k] in ‘key’ and ‘caw’). The production of sounds is the final stage of a process that begins with the planning of a message to be spoken, and continues via the selection of words and syntactic structure to convey this message, the determination of the appropriate phonological structure (including prosodic form) of the chosen sequence of words, and the specification of the precise phonetic form to be articulated. The selection of words from the mental LEXICON in speech production proceeds in two stages: first conceptual representations are mapped onto the meaning representation of words, then the phonological forms linked with these meaning representations are activated. The construction of syntactic form is in part dependent on the precise words chosen and in part on the conceptual content of the message. Speech production sometimes goes wrong, and such SLIPS OF THE TONGUE can occur at any level of production: choosing the concept (‘Pass the dog’ instead of ‘pass the salt’ when the speaker is looking at the dog), selecting the words (instead of ‘Sunday’ the semantically related ‘Monday’ or the phonologically related ‘summer’), constructing the syntax (‘the boys who I saw the boys’), determining the phonological form (‘Lunday sunch’ for ‘Sunday lunch’); and many more. Slips can thus provide evidence about the processes and stages involved in speaking.
During speech production speakers monitor their output and may repair infelicitous or faulty utterances; this monitoring can also apply to any aspect of the output—style, content, words, syntactic form, pronunciation. What aspects of the output are monitored depends however on where the speaker’s attention is directed (a non-native speaker may concentrate on careful monitoring of pronunciation, a witness in court may concentrate on correct choice of formal style, etc.)
Fowler C.A. (1995) Speech production. In Handbook of Perception and Cognition, ed. E.C.Carterette & M.P.Friedman, vol. 11, Speech, Language and Communication, ed. J.L.Miller and P.D.Eimas, pp. 29–61. Academic Press: New York.
Levelt W.J.M. (1989) Speaking: From Conception to Articulation, MIT Press: Cambridge MA.
ANNE CUTLER
This is the complete article, containing 463 words
(approx. 2 pages at 300 words per page).
