11.   Writing allows private communication with the computer that is not possible with speech recognition.

12.   One of the major areas of research in speech recognition concerns the ways in which general linguistic knowledge can compensate for such errorful or ambiguous acoustic input.

13.   These experiments show that the recognition strategies suggested by structural analysis of large lexicons do not, in their present form, extend to continuous speech recognition.

14.   HARPY was primarily concerned with achieving near real-time speech recognition for a specific, constrained task.

15.   Psychological models of speech recognition at first concentrated on the way in which sentential context could speed up the process of word recognition in fluent speech.

16.   I will also discuss the main attributes of the chart parsing process with reference to the dynamics of speech recognition.

17.   As we saw earlier, the designers of TRACE would say that these approaches are based on some fundamental misconceptions about the nature of speech recognition.

18.   In earlier chapters we discussed the possibility that poor acoustic-phonetic processing was primarily to blame for delays or errors in automatic speech recognition.

19.   It is interesting to compare this work with that of Lee on Hidden Markov Modelling of speech recognition.

20.   Our research has practical applications in such areas as speech recognition, text critiquing, and speech synthesis.