Thursday, August 23, 2012

Polysemy and homonymy

Not only do different words have different meanings; but also the same word may have different meanings. This is polysemy; such a word is polysemic (flight: 'passing through the air', 'power of flying', 'air journey', 'unit of the Air Force', 'volley')
 
Problems with this concept:
1. We cannot clearly distinguish whether two meanings are the same or different and, therefore, determine exactly how many meanings a word has.
  • The verb eat has three meanings (the 'literal' sense of taking food and the derived meanings of 'use up' and 'corrode'). But we can also distinguish between eating meat and eating soup; the former with a knife and fork and the latter with a spoon. The problem, however is to decide whether this represents a distinct mean of eat. The moral is not to look for all possible differences of meaning, but to look for sameness of meaning.
2. If one form has several meanings, it is not always clear whether this is an example of polysemy (one word with several meanings) or of HOMONYMY (several words with the same shape).
  • The dictionary treats flight as a single (polysemic) word, but recognises five words (i.e. homonyms) for mail (armour, post, halfpenny, payment and spot). We do not make the same distinctions in writing and speech. Thus Iead (metal) and lead (dog’s lead) are homography (spelt in the same way, but pronounced differently), while site and sight, rite and right are homophony (spelt differently but pronounced in the same way).
The question is whether a single form with two meaning is one word with different meanings (polysemy) or two different words with the same shape (homonymy).
1. Dic­tionaries usually base their decision upon etymology. If identical forms have different origins, they are treated as homonymous and given separate entries. If they have one origin, they are treated as polysemic and given a single entry. However, the history of a language does not always reflect its present state (e.g. pupil (=student) and "pupil of the eye' are a pair of unrelated words today (homonymy). Historically, they have the same origin (polysemy). On the other side, the word "ear" used of the "ear" of corn seems to be an example of metaphor, polysemy. Yet, they are historically unrelated. Even the difference of spelling does not always indicate a difference of origin. Even today's homophones may be derived from the same original form, polysemy (e.g. metal/mettle, flower/flower). The dictionary maker treats these as different words because they are spelt differently, and he needs to keep words in their alphabetical position.
2. It is reasonable to suggest that where the differences are regular, we have polysemy rather than homonymy. One of the most familiar kinds of relationships between meanings is that of "metaphor" where a word appears to have both a "literal" meaning and one or more transferred meanings (e.g. parts of the body "hands and faces of a clock"; "the foot of a bed or a mountain"; "the leg of a chair"; "the tongue of a shoe"). However, metaphor is haphazard. Other languages do not show that "foot" is appropriate to mountain. or eye to needle as in French. Moreover, in English "eye" is used with a variety of other meanings, e.g. the centre of a hurricane or a spring of water, yet it is not used for the centre of a flower, though this might seem to be a reasonable candidate for the extension of meaning.
Other kinds of transference: many adjectives may be used either literally for the quality referred to or with the transferred meaning of being the source of the quality (a person may be sad and book may be sad). The language recognises the difference of meaning in that we cannot say "John is as sad as the book he was reading". This is called ZEUGMA. Here, one word co-occurs with two other words and these two each require the first to have a different meaning, and this the language does not allow. Similarly, many nouns have a concrete and an abstract sense (the score of the symphony is on the table/the score of the symphony is difficult to follow). Again, we cannot say "the score is on the table and difficult to follow."
However, it is not always easy to decide whether a relationship is regular or not. English has intransitive and transitive uses of verbs (The door opened; I opened the door/The bell rang; I rang the bell). Slightly different are the basic and causative forms of verbs (He marched/walked/ran vs. He marched/walked/ran the dog/children). It is reasonable not to recognise homonymous pairs here (two verbs open/ran/march). Yet, the meaning relations are not wholly regular ("run the children" does not mean "cause the children to run"). There is no intransitive (*the man wounded) and no causative (*He swam them).
3. We can base our distinction on the core meaning. This is possible with metaphor "transferred" meanings (sad and score). But it is difficult to decide whether there is any core meaning in some other words (it is obvious why "key" is used with the door or translation, but it is not with the keys of a piano and, therefore, not at all clear that this is an example of polysemy. In a historical viewpoint, this problem has arisen as words change their meaning. Thus "arrive" is derived from Latin "rip" (a shore) and originally meant (reach shore).
4. We can use the test of ambiguity. (I went to the bank) is ambiguous since "bank" can mean either river bank or a financial institution. But decisions are not always easy. "Kill" is used to refer either to murdering or killing accidently. ("John killed Bill" may or may not seem ambiguous). One test of ambiguity is the "co-ordination test." "John and Bill went to the bank" cannot be taken to mean that one went to the river and the other to the financial institution. A particular version of this is the "do so" test. We cannot say "John went to the bank and so did Bill" with the two meanings of the bank. But judgments about co-ordination depend upon judgments about sameness of meaning, and the doubtful cases remain. The coordination test will force us to make too many distinctions. Ambiguity is not even sufficient to establish homonymy. ("He ran the race for Hampshire" may mean either that he was a competitor or that he organized the race. But since the two meanings of ran here are related in terms of causativity like those of walk and march, we cannot say that we have two lexical items.
A word that is polysemic will have a variety of synonyms each corresponding to one of its meanings. It will often have a set of antonyms. "Fair" may be used with (1) hair (2) skin (3) weather (4) sky (5) judgment (6) tackle. The obvious antonyms would be (1) dark (2) dark (3) foul (4) cloudy (5) unfair (6) foul. We might be tempted to say that where the antonyms are the same we have polysemy, and that difference of antonyms implies homonymy. "Fair" with "hair" and "skin" have the same antonym (dark), and so do with "weather" and "tackle" (foul). But this will suggest that "fair" with "weather" is more like "fair" with "tackle" than "fair" with "sky". Intuitively, "sky" is more related to "weather" and "tackle" to judgment.
Multiplicity of meaning is not confined to words of the dictionary, but it is a very general characteristic of language. It is also found with grammatical elements – the English past tense has two different meanings. So do some suffixes; in- usually means "not", but this is not so in "inflammable." There is also similar ambiguity in syntax (The old men and women / Visiting relatives can be a nuisance). Both can be analysed differently in syntax with accompanying difference of meaning.
Related Articles