UNIT OF PERCEPTION (UNIT OF PROCESSING)
المؤلف:
John Field
المصدر:
Psycholinguistics
الجزء والصفحة:
P314
2025-10-22
37
UNIT OF PERCEPTION (UNIT OF PROCESSING)
A unit into which the raw features of speech are automatically ‘packaged’ at an early pre-lexical stage– i.e. before any match is sought with words in the lexicon.
Traditional accounts of language processing tend to assume that the listener identifies phonemes from phonetic cues in the speech signal. However, this is difficult to square with the fact that the features which make up individual phonemes are highly variable in connected speech. There is also evidence that our awareness of the phoneme may only be achieved as a result of alphabetic literacy.
One proposed alternative is that the signal is processed exactly as it is received, without being analysed into any intermediate units. In these direct access models, phonetic features (+ nasals, + consonantal etc.) are the evidence that is used for making word matches.
Another proposal is that there is a unit of processing which is larger than the phoneme, and therefore less variable. Suggestions have included the syllable, the demi-syllable and the diphone (a combination of two adjoining phonemes). The advantage of the syllable or demi syllable is that there is a restricted number in any language (the estimates for French are 6000 and 2000 respectively). There is also evidence that infants become aware of the syllable as a unit in speech at an early stage of linguistic development. However, although the syllable is more robust than the phoneme, it is still subject to variation through accommodation (assimilation, elision, resyllabification etc.). More importantly, there are languages such as English where the syllable boundary is not always clearly marked: the sound /m/ in a word such as LEMON is ambisyllabic in that it does not clearly form either the end of the first syllable or the beginning of the second but seems to participate in both. This makes it difficult to divide some English words into clear syllabic units.
The rival claims of phonemes and syllables as ‘units of perception’ have been hotly argued. An early study obtained faster detection times for syllables than for phonemes, suggesting that words were primarily analysed syllabically and only afterwards broken into phonemes. However, this finding was later widely questioned. A subsequent study demonstrated faster results for disyllabic words than for syllables, suggesting that the size of the unit was a factor in the results achieved. Further studies followed, some supporting the syllable and some not. But a criticism that is levelled against many of them is that they provide evidence of how subjects identify a syllable in connected speech rather than how (at an earlier and more automatic stage) they perceive it.
A more sophisticated method explored the assumption that, if the syllable were indeed the basic unit of perception, the same time would be taken to recognise the difference between PID and PIT (third phoneme divergent) as the difference between PID and BID (first phoneme divergent). The results did not support this hypothesis, and the researchers concluded that, if speech is ‘packaged’ by English listeners, it is into units smaller than the syllable. An alternative view holds that the listener has several units of analysis available to them. The unit that is employed is determined by the nature of the task they are performing and by the attentional demands that the task makes upon them.
See also: Lexical segmentation, Phonological representation, Speech perception: phoneme variation
Further reading: Norris and Cutler (1988); Nygaard and Pisoni (1995)
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
