UNDERSPECIFICATION
المؤلف:
John Field
المصدر:
Psycholinguistics
الجزء والصفحة:
P312
2025-10-22
36
UNDERSPECIFICATION
A theory that it is not necessary for all the characteristics of a linguistic unit to be stored in the mind, but only those which serve to uniquely characterise it. The concept is especially applied to the phonological representation which enables a language user to recognise the sounds of a language.
Early generative phonology represented the phonemes of a language in binary form, indicating the presence or absence of certain acoustic features (þ/ consonantal, þ/ nasal etc.). A basic underspecification could restrict this to the þ features (those that are actually present). But among those features would still be some which are redundant: for example, if a consonant is nasal, it must also be voiced. It is therefore possible to reduce the specification further by eliminating all features which can be inferred from others by means of a rule.
Underspecification of this kind provides an economical account of how we store phonology in the mind. By reducing the importance of non-essential cues, it also streamlines the process of recognising phonemes and thus compensates to some extent for the fact that phonemes vary greatly from one context to another in terms of the phonetic features that compose them.
At syllable level, underspecification enables the listener to ignore any information in the acoustic signal which is not contrastive. For example, it has been shown that distinctions between nasalised and non-nasalised vowels are not represented phonologically for an English listener since they do not serve to differentiate words. However, they do form part of the phonological representation of a Bengali speaker since Bengali possesses vowels of both types.
See also: Phonological representation
Further reading: Crystal (1997); Goldsmith (1995: 13–18)
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
