Synonymy
In discussing synonymy, the relation of meaning identity, an initial dis tinction needs to be drawn between lexical synonymy (synonymy between individual lexemes) and phrasal synonymy (synonymy between expressions consisting of more than one lexeme). We will only be concerned here with lexical synonymy, assuming that phrasal synonymy can mostly be derived from the synonymy of the phrases’ component lexemes (considered in their associated grammatical structures).
Meaning identity (synonymy) is a part of the metalinguistic stock-in trade of ordinary speakers of English: we often refer to words as ‘having the same meaning’. However, we usually restrict our statement of the synonymy of two words (or phrases) to the utterance level:
When questions of sameness of meaning arise for unsophisticated speakers, no appeal is made to an abstract entity of ‘meaning’: a given word or phrase is accepted as having the same meaning as another word or phrase if its substitution for the other in the given context yields an utterance which they will accept as having the same meaning as the first utterance. (Lyons 1968: 75)
Speakers do not, that is, characteristically seem to base their judgements of synonymy on a ‘bottom-up’ analysis of the meaning of each of the words involved, concluding that words are synonymous if their separately established meanings are identical. Instead, a top-down procedure often seems to be at work: the fact that two expressions have the same contextual effect is what justifies labelling the substituted words as synonyms in that context
. Lexical synonymy has been variously defined in the semantics literature. The general definition of ‘identity of meaning’ is mostly accepted (Cruse 2002a: 486 however defines it as ‘identity/similarity’ of meaning), and it is the one we adopt here. Within this definition, however, there are a number of different terminological conventions. Of course, what is important in such cases is not to decide which of the different possible uses of a technical term like ‘synonym’ is better (or, even less, correct), but simply to defi ne what is meant by the label in question and to use it consistently and without ambiguity.
For some authors synonymy is a context-bound phenomenon, two words being synonyms in a certain given context, whereas for others it is context-free: if two words are synonymous they are identical in meaning in all contexts. The question of synonymy and grammatical context is another on which disagreements exist. Thus, two words are synonymous for some authors if, like likely and probable in (20), they have the same meaning, even if they show a different set of grammatical cooccurrence possibilities – here, the possibility of raising the subject of the complement clause in (20a) to the subject of the main clause in (20b), which exists for likely but not for probable:
For other authors, however, both identity of meaning and identity of grammatical properties are required (see Hudson et al. 1996 for many other examples).
Another important distinction is between synonymy of words and synonymy of senses. Sense-synonymy is the synonymy of some, but not all, the senses of a word. Thus, pupil is arguably synonymous with student with respect to one of its senses (‘person being instructed by a teacher’); but with respect to the sense ‘centre of the eye’ the two words are, of course, non-synonymous. Pupil and student are thus not lexical synonyms, but they are synonymous with respect to one of their senses. Similarly, Murphy (2003: 30) demonstrates that the pair baggage/luggage are synonymous with respect to the sense ‘bags’ but not with respect to the metaphorical sense ‘emotional encumbrances’:
Recognizing sense-synonymy as a category implies viewing meaning identity not as a binary property of two words, but as a graded one: the more senses two words share, the more synonymous they are.
The limiting case of sense-synonymy is word-synonymy, which is the situation in which two words share all their senses. Typically, lexical synonyms are taken to be mutually intersubstitutable in every environment, with each synonym being equally normal in each environment (Cruse 2002a: 488; see box). The clearest examples of word synonymy are trivial ones, where there are alternative pronunciations for what is, in fact, intuitively a single lexeme, such as (n)either (pronounced ['(n)i:ðә] or ['(n)aiðә]) and eco nomics (pronounced [i:kә'nÅmiks] or [εkә'nÅmiks]). In both these cases the same meaning is indisputably involved, but it is not clear that two words should be recognized. Some more interesting possible examples would be the pairs Islamic and Muslim, Peking and Beijing or Bombay and Mumbai.
QUESTION Do these examples survive the test of mutual intersubstitut ability?
Ullmann (1972: 141–142) points out that one of the few places where full word synonymy seems reasonably common is technical vocabulary, giving as example the fact that in medicine inflammation of the blind gut can be synonymously referred to as either caecitis or typhlitis.
However, as Ullmann also notes (1972: 142), word-synonymy ‘runs counter to our whole way of looking at language. When we see different words we instinctively assume that there must also be some difference in meaning.’ Consistently with Ullmann’s point, genuine lexical synonyms which are not, unlike the examples just given, proper nouns or adjectives prove extremely hard to find. Once their combinatorial environments have been fully explored, proposed lexical synonyms often prove not to be such. For example, Bolinger (1976, discussed by Murphy 2003: 164) showed that everybody and everyone are not lexical synonyms since they are not mutually substitutable in every context:
Similarly, almost and nearly fail the test, as demonstrated by (23):
QUESTION Can you find any lexical synonyms in any language you know? Are they really substitutable for each other in every environment?
Very often, the difference between lexical synonyms is not one of denotation but of connotation: the associations and emotional values of a word (see 1.4.2). Thus, the lexemes doctor and quack both arguably share the definition ‘medical practitioner’, and would be substitutable in every con text but for the fact that they differ in the neutral and pejorative connotations attaching to each respectively. Other examples would be lunch and luncheon and fag and cigarette.
QUESTION Consider the pairs of nouns prize/award, couch/sofa, and coronary/heart-attack. Are any of these synonyms? If so, what kind?
So far, we have concentrated on the place of synonymy within the paradigmatic language system and largely ignored its place in language use. An initial observation on this subject is that, at least in many varieties of educated English written discourse, it is considered good style to avoid the repetition of identical words in nearby contexts. As a result, near-synonyms are often enlisted as equivalents, without any semantic difference between the equivalent terms being intended. The surrounding context thus endows the equivalent words with the temporary status of synonyms, a status which is in no way permanent, and which may be subsequently revoked so that the formerly equivalent terms can be brought into a relation of contrast. Lyons (1968: 80) generalizes this conclusion to all lexical relations: ‘any meaning relations that are established are established for particular contexts or sets of contexts, and not for the totality of the language’. Thus, red has different possible antonyms depending on whether the context is wine (where the antonym is white), traffic lights (green), or accounts (black).
Second, and lastly, we’ll turn to a particularly interesting case of absolute lexical synonymy which has been observed widely in the Aboriginal societies of Australia (Alpher and Nash 1999). In most of these societies, an individual’s name would not be used after their death. Furthermore, in many of them, words which sounded similar to that individual’s name were also prohibited. This practice would clearly present many inconveniences if there were not some way of replacing the banned vocabulary. The usual practice, resting on the widespread multilingualism that was a standard feature of traditional Aboriginal society in Australia, was to adopt the translational equivalent of the prohibited word from a neighbouring language, and to use it until the old word became reusable (an interval of time which differed according to a number of variables). This process of temporary lexical replacement has resulted in Aboriginal languages possessing a wide range of absolute lexical synonyms. In Warlpiri, for example, a particularly well studied Australian language for which a large corpus of citations exists, facilitating semantic and lexical study, we could give perhaps hundreds of examples of absolute synonyms which appear to be completely equivalent and interchangeable in all contexts. The noun karnta ‘woman’, for instance, has at least the nouns mardukuja and rduju as absolute synonyms; ‘dog’ is translated synonymously by jarntu and maliki; waku ‘arm’ has the absolute synonym nginyanyka; and marlu ‘red kangaroo’ has jurrkapanji, wawirri and yawarrangi. Not all of these cases of synonymy are necessarily due to bereavement-induced borrowing: there may be a higher general tolerance of synonyms in Warlpiri than in familiar European languages. While it is possible that the synonymy of some of these examples may not survive the scrutiny of deeper lexicographical investigation, the number of candidates for synonymy in Warlpiri constitutes a striking exception to the pattern observed widely in European languages, which is that a loan-word synonym of an indigenous expression typically develops some semantic difference from the native word. This was the case with the words beef, veal and mutton, all borrowed into English from French, originally synonyms of cow, calf and sheep, but subsequently specialized to refer simply to the edible flesh of these animals.
QUESTION English has many pairs of near synonyms consisting of a native (Germanic) form and later Latin one. The verbs begin-commence and end-terminate are good examples. How many more can you find? How syn onymous are they?
