We have earlier that when the substitution of one sound segment of another result in adifferent word, we have sufficient evidence that the two sounds represent two different phonemes. Note, however, that two different forms may be identical in meaning because of the existence of free variation. This was shown by the fact tahat some speakers pronounce the first sound of the word economics as [e] and other as [i]. These two strings of sounds are not a minimal pair; the substitution of [e] for [i] or vice versa does not produce another word. Similarly, some speaker pronounce basic [beɪsɪk] and other as [bæsɪk]. Such pair do not tell us whether [i, e, eɪ, æ] represent phonemes in the language.
Homonyms
or homophones also show that two words of different meanings may have identical
forms; that is, they may pronounce exactly alike. Thus [soʊl] can nean ‘sole’
or ‘soul’. ‘sole’ itselft has three meanings ‘shoe bottom’ ‘fish’ and ‘only’
Thus
giving [soʊl] at least four meanings to one form. The sentence ‘Greta Garbo ate
her cottage chese with relish [relɪʃ]’ could mean she ate with ‘gusto’ or with
a particular kind of sauce.
The
determining fact that is both a change in form (pronunciation) and a chang in
meaning. When both changes occur, we know that substituted sound segments
represent different phonemes.
The
phonology of a language includes rules that relate the phonemic representations
of word to their phonetic representations. The phonemic representation need
only include the non-predictable distinctive features of the string of phonemes
that that represent the words. The phonetic representation includes all the
linguistically relevent phonetic aspect of the sounds. The phonetic
representation does not include all the physical properties of the sounds of an
utterance, since the physical signal may vary in many ways that have to do with
the phonological system. The absolute pitch of the sounds, or whether the
utterance is spoken slowly or fast, or whether the speaker shouts or whispers,
is not linguistically significant. The phonetic transcription is thus also an
abstraction from the physical signal; it includes the non-variant phonetic
aspects of the utterances, those features that remain relatively the same form
speaker to speaker and form one time to another.
Minimal
pairs and complementary distribution of phonetic units are helpful clues in the
attempt to discover the inventory of a phonemes in a language. Most linguists
agree, however, that by themselves they do not determine the phonemic
representation of utterances; the phonemic representation is a further level of
abstraction, as will be shown below in the discussion of phonological rules.
Phonemic
representation do vary, though, in their degree of abstraction. Some record a
‘surface’ transcription using phonemic inventory; other attempt to capture the
‘deeper’ knowledge that the speaker has about the processes of the phonology.
