adulthood. A. Smith (1966) describes a case of a man who underwent left hemispherectomy at age 48. The surgery left him nearly totally speechless:

"E. C.'s attempts to reply to questions immediately after operation were totally unsuccessful. He would open his mouth and utter isolated words, and after apparently struggling to organize words for meaningful speech, recognized his inability and would utter expletives or short emotional phrases (e.g. 'Goddamit'). Expletive and curses were well articulated and clearly understandable. However, he could not repeat single words on command or communicate in propositional speech until 10 weeks post-operative." Expressive speech showed some development in E. C., but Smith reported that his creative language was still "severely impaired" 8 months after the operation.

A patient studied by Whitaker (1971) suffering from nonfluent aphasia exhibited an interesting sort of automatic speech behavior: he responded to nearly every question or attempt at conversation with the utterance "What'cha gonno do right now? yea yea". Whitaker points out to the student of neurolinguistics that one cannot use such utterances as data in determining a patient's true linguistic competence "... on the basis of (this one) utterance, it would be rather farfetched to assume that L. S. (the patient) had retained the WH-question transformation and correct pronominal reference. ..." (pp. 145-146).

Thought to be related to automatic speech are "ictal speech mechanism", words or utterances spoken by psychomotor epileptics during, before, or immediately after seizures. Ictal speech mechanisms consist of stereotyped expressions ("I beg your pardon") and emotional utterances which are spoken out of context. As in the case with AS associated with the right side of the brain as well as the left, Serafatinides and Falconer (1963), in agreement with other studies, found that "of 15 patients with truly ictal speech automatisms 4 were operated on the left side and 11 on the right or recessive side" (p. 345).

The neurolinguistic evidence, then, points to the fact that automatic speech is neurologically different from creative language in that it is localized on both sides of the brain, as opposed to just the left hemisphere, and can be preserved in cases of aphasia. If AS is related to routines and patterns, then routines and patterns may have a

85