What is a Phoneme?
The chief areas of grammar are phonetics, phonology, morphology, syntax, and semantics. Simplifying matters slightly, linguistic handbooks say that phonetics deals with the physical aspect of sounds, while phonology deals with the psychological aspect. But, in fact, phonological descriptions deal with both aspects. In other words, there is always a phonetic (or physical) aspect on a phonological description.
Phonology is the science of speech patterns. More precisely, phonology studies the phonological processes or the grammatical similarities of the sounds, that is, how sounds combine to make morphemes and words. Hence, the dominant aim of phonology is to provide a phonemic representation of morphemes and a series of processes that adequately express the phonological generalizations of a language.
Sounds are the product of human anatomy, the vocal tract: the lips, the tongue, the velum, and the larynx. Within the field of phonetics, there are two basic areas: a) articulatory phonetics, the study of how speech sounds are articulated; and b) acoustic phonetics, the study of the physical similarities of the sounds.
Let us now consider that speech sounds have structure and function in languages.
Each language or dialect has its own rare set of sounds, its sound system. Languages use sounds in very different ways: the sound inventories may be dissimilar; the sounds may occur in different orders, and the processes or rules that affect sounds may be different. however there are characteristics which are shared to all languages (true linguistic universals). In fleeting, language inventories of sounds and the phonological processes that occur in languages are limited in various ways.
applicable here is the sharp distinction between a sound or a phonetic component (phone) and a phonological component (phoneme). A phoneme is one of the basic recondite sound units which every accent of a language possesses. Of course, as noted earlier, phonemic segments in a language represent a physical phonetic reality. And all phonological systems use the same alphabetical signs for phonemes or phones. The discrete segments or phones are transcribed within square brackets [p], and phonemes are transcribed within forward slashes /p/.
The basic function of sounds is to convey meaning; differences in sound are related to differences in meaning in a given language. Trubetzkoy says that phonemes are discriminative elements. But how can we work out the inventory of phonemes (not phones) in a language? We can classify sounds based on the possibility of their appearing in the same structural ecosystem. consequently, to determine which sounds belong in the same class (or phoneme) we look for minimal pairs. For example, take the following words in Finnish: /takka/ (‘fireplace’), /taakka/ (‘burden’), and /taka/ (‘back’). We have here two minimal pairs, /takka/ and /taakka/; /takka/ and /taka/. What this method is that in the first pair the vowels, /a/ and /aa/, combine with the same surroundings consonants (/t_kk/) and in the second pair the consonants, /kk/ and /k/, combine with the same ordern (/ta_a/). These sounds relate contrastively to each other because they appear in the same ecosystem. No doubt, in these examples Finnish uses short and long sounds to discriminate different words. The differences are functionally meaningful at the level of the information formation.
However, it is not always possible to find minimal pairs. In this the case, it is necessary to rely on “near-minimal pairs”, whenever we can assume that the second difference is not likely to have an influence on the main component observed.
We turn now to the affirmation above that sounds may be part of a class. It is interesting to see that a single phoneme need not always get the same phonetic realization. In more concrete terms, there is only one phoneme, but it turns up in two different phonetic “shapes”. What follows from this is that these sounds do not change the meaning when we make a substitution. We say that they are allophones (from the Greek information allos, other), that is, variant forms of a phoneme. To take a simple case, in Brazilian Portuguese there are allophones of a single /l/ phoneme: the l sounds of such words as lápis (‘pencil’) and mal (‘bad’). The first l (clear l) is articulated much further forward in the mouth than the second (dark l). These sounds are found in mutually exclusive environments and are not contrastive. They are contextual variants or combinatory variants of a single phoneme. When two sounds are found in different environments, this is termed as complementary dispensing. consequently two phones are then stated to one single phoneme. This may be stated in terms of nearby segments, syllable, morpheme, etc. We can express the Portuguese examples above in terms of morpheme border (#) and nearby segments, pre-vocalic and post-vocalic, to know: (#_V); (V_#).
however, it is also possible that two phones may appear in the same context without a change in meaning. And the consequence of this is termed free variation; they are optional variants or free variants. To see how this works, let us look back to the issue of the /l/ in Brazilian Portuguese. We have seen that the clear l occurs in a pre-vocalic position. So it is a little surprising to find out that old gauchos in South Brazil use the clear l in a post-vocalic position at the ending of words as especial (‘special’). In view of this, we can assume that in South Brazil the clear l is a free variant, that is, it is not associated with a contextual position. Surely this difference has no effect on the establishing of phonemic contrast. We consequently have found that this kind of variation is not conditioned by context (complementary dispensing), but is optional (free variation). In summary, there are at the minimum two types of variants: contextual and free.
Finally, observe that there is a distinction in a language between processes which require phonological information and those which require nonphonological information. It follows that phonetic contrasts should not be employed to cover up the former information. To clarify further, notice that in Finnish the endings (case inflections) are not attached mechanically to the words according to general patterns. This is because words undergo sound alternations (changes). One of the most important is termed as consonant gradation, since it affects the long and short stops p, t and k. There are two types of gradation: a) quantitative gradation: long consonants alternate with the corresponding short consonants, and b) qualitative gradation: short consonants generally alternate with other consonants. Let us turn to an example of b): p alternates with v, despite the fact that /p/ and /v/ are separate phonemes. Following a general rule, /p/ is realized as [v] in closed syllables (that is, syllables ending in a consonant). For example, /halpa/ (‘cheap’) has the genitive /halpan/ = [halvan]. This is then the consequence of a general phonological course of action which weakens intervocalic stops. It is obvious that saying that we have two phonemes allomorphs /halpa/ – /halva/ would conceal the generalization.