Machine Translation: Theoretical and Methodological Issues. Sergei Nirenburg, ed. Cambridge University Press, New York, 1987. 350 pp. $49.50 HB, $17.95 PB.

Machine translation was proposed by Warren Weaver in 1947 for the newly developed computer. The proposal was pursued because language—a system of symbols usually called signs—seemed manipulable in the same manner as the system of number symbols. During World War II, moreover, translation was a major concern for military and diplomatic texts, and later for scientific purposes. Federal agencies later began to fund research geared toward developing "fully automatic high-quality machine translation." Their expectations, however, were unrealistic.

There were massive problems. Linguistics was a new science, not yet recognized as an independent field meriting university departments. Grammars and dictionaries were totally inadequate. Software was developed for the much simpler number system; it achieved limited success on the cumbersome pre-transistorized hardware. Then, after a decade of research, funding was abruptly...

Interested in reading more?

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!