Table of Contents
2 Main Part
2.1.2 Types of Synonymy
2.1.3 Distinctions among loose synonyms
2.2 Synonymy in Dictionaries
2.2.1 Synonymy as a Defining Style
2.2.2 Synonymy in General Dictionaries
2.3 Synonym Dictionaries and Thesauruses
2.3.1 Types of Synonym Dictionaries
2.3.2 Webster’s New Dictionary of Synonyms
2.3.3 Roget’s Thesaurus of English Words and Phrases
2.3.4 Rodale’s Synonym Finder
2.4 History of Synonym Dictionaries
In the course of the semester, we have dealt with a number of linguistic problems and their relationship to lexicography. Since one main aim of dictionaries is to define words or, in other terms, to explain the meanings of lexical units, lexicography is very closely linked to the linguistic branch of semantics. On the one hand, dictionaries describe the semantic relation holding between words and the extra-linguistic world. This relationship is called reference or denotation. On the other hand, dictionaries deal with the semantic relations between the words of a particular language themselves. These relations are called lexical relations or sense relations. Both denotation and sense relations contribute to defining the meaning of a lexeme. The lexical relation which is probably the most important one with regard to lexicography is synonymy. On the following pages, I will therefore define the term synonymy, discriminate different types of synonymy and discuss some ways in which synonyms may be differentiated. Afterwards, I will try to show which role synonymy plays in general dictionaries. After discussing different types of specialized synonym dictionaries, I will finish with a brief survey of the history of synonym dictionaries.
2 Main Part
In general, the term synonymy, which derives from Greek (syn + nymy = “same + name”) is very briefly defined as “sameness of meaning” (Palmer 1981: 88). Expressions are therefore synonymous, if they “have the same meaning” (Jackson 1988: 65). Palmer also states that synonymy can be defined as “symmetric hyponymy”, which means that “if mavis and thrush are synonymous, we can say […] that all mavises are thrushes and all thrushes are mavises” (1981: 88). This kind of definition, however, seems to be rather imperfect, because the sense relation of hyponymy mainly applies to nouns, especially to concrete nouns. Moreover, the definitions of various linguists differ in the degree of the meaning similarity. Whereas Hartmann and James define synonymy as “the sense relation obtaining between the members of a pair or group of words or phrases whose meanings are similar” (2001: 135), Lyons is convinced that only “identity, not merely similarity, of meaning” has to be the criterion of synonymy (1995: 60). Furthermore, it is arguable whether the sense relation of synonymy is restricted to lexemes. Lyons, for example, attaches importance to the “possibility that lexically simple expressions may have the same meaning as lexically complex expressions” (1995: 60). For Palmer, by contrast, “bull” and “male adult bovine animal”, an expression which is related to componential analysis, are not synonymous because no one would use the latter expression in natural speech. This means that no one would say “There is a male adult bovine animal in the field” and that the two expressions are therefore not interchangeable (cf. Palmer 1981: 93). Furthermore, there is no consensus among linguists about the conception which underlies the term meaning, even though synonymy is defined as “having the same meaning” in the majority of cases. On the one hand, there are linguists who restrict the notion of synonymy to the identity of what Lyons calls “descriptive meaning” (1995: 63). According to Palmer, on the other hand, it is not useful to separate the descriptive meaning, which he calls “basic” or “cognitive” meaning, from the “emotive” or “evaluative” meanings of words, which can be equated with the connotations that a word has. As an example, he mentions the synonym pairs politician and statesman, hide and conceal and liberty and freedom, the first member always implying approval, the second one disapproval. He is convinced that “the meaning of words is not simply a matter of objective facts”, that “a great deal of it is subjective” and that the cognitive meaning of a word can therefore not be separated from its emotive or evaluative meaning (cf. Palmer 1981: 90-91). In contrast to Palmer, Webster’s New Dictionary of Synonyms, which will be discussed below and which explains its notion of synonymy in the front matter of the dictionary, draws – like Lyons – a clear distinction between the denotation or reference of a word and the implications that a word has. As a consequence, two words may evoke the same associations or ideas in a language user and yet not be synonymous. In Webster’s New Dictionary of Synonyms, a synonym is therefore always “one of two or more words in the English language which have the same or very nearly the same essential meaning” (Webster’s New Dictionary of Synonyms 1973: 24a). Consequently, the most important criterion for synonymy is the common denotation of words, which means that synonyms must have the same or very nearly the same definition. All in all, the definition of synonymy is not as simple as it seems at first glance. The greatest problem of defining synonymy as “sameness of meaning” is the underlying definition of meaning, which is a complex problem of semantics.
2.1.2 Types of Synonymy
As I have already mentioned, the degree of the meaning similarity among synonyms may vary, which means that some members of synonym pairs are closer in meaning than others. Linguists agree that strict synonymy, which is sometimes also called complete, absolute, or total synonymy is impossible or at least almost impossible. Even if two lexemes have the same denotation, they can always be differentiated in many other ways, for example in terms of connotation, formality or currency. For this reason, synonyms are hardly ever “interchangeable in all contexts”, which is the criterion for strict synonymy according to Jackson (1988: 66), who also gives three arguments against strict synonymy. Firstly, a language does not need two words which are interchangeable in all contexts. As a consequence, strict synonyms are unlikely to survive in a language because of economic reasons. Secondly, if strict synonyms happen to occur in a language, it is likely that one member of the synonym pair alters its meaning a little bit and starts to be used in a different context or on a different style level than the other member. An example of this phenomenon is the word mutton, which had exactly the same meaning as the English word sheep when it was borrowed from French. Nowadays, it only denotes the meat of the animal, but no longer the animal itself. Thirdly, it may also happen that one member will become obsolete and fall out of use (cf. Jackson 1988: 66-67).
Another type of synonymy that is much more frequent than strict synonymy and more important with regard to lexicography is loose synonymy. Loose synonymy may also be called relative, quasi -, pseudo - or near -synonymy. Although the definitions for this phenomenon vary to a certain degree, loose synonymy is often used for words which have the same (or very nearly the same) denotation, but are not interchangeable in every context like strict synonyms. So we speak of “words that can substitute for each other in a wide range of contexts but not necessarily absolutely” (Jackson 1988: 67) because they differ in a number of ways which will be discussed below. Examples for loose synonyms are freedom and liberty, malodorous and stinking, begin and start or sometimes and occasionally.
There is a third kind of synonymy which is sometimes called partial synonymy. This term means that only “part of the meanings of two (or more) words are the same”, so there is only “overlap in meaning but not complete identity of meaning” (Jackson 1988: 73). The lexemes mature, ripe and adult, for example, all have the common core meaning “fully developed”, but each of them has an additional meaning, too. Ripe, for example, also infers “readiness for use or enjoyment”, especially in relation to fruits (Webster’s New Dictionary of Synonyms 1973: 527). Of course, the transition from strict to loose synonymy and from loose to partial synonymy is gradual. It is especially difficult to decide whether the similarity of meaning is sufficient to categorise words as loose synonyms or if there is only an overlap in meaning which is characteristic of partial synonymy.
2.1.3 Distinctions among loose synonyms
As already mentioned above, loose synonyms are not interchangeable in every context because they differ in at least one of the following respects. Firstly, loose synonyms may belong to different dialects. There are, for example, many pairs of synonyms in British and American English such as lift and elevator, pavement and sidewalk, flat and apartment or biscuit and candy. Of course, there are also synonym pairs between other national varieties of English (including Canadian English, Australian English and so on). This differentiation, however, only applies if we regard these national varieties as dialects of one common language. Otherwise, the status of such synonyms is “no different from the translation-equivalents of, say, English and French” (Palmer 1981: 89). Apart from synonym pairs between national varieties of English, there are also synonyms between dialects of a national variety. The Collins English Dictionary, for example, tells us that sandwich is the standard dialect synonym for the regional dialect word butty. Secondly, many synonyms differ stylistically, which means that one word is more formal than the other one. In many cases, the loan word borrowed from French or Latin (for example ascend or commence) is the formal equivalent of the more colloquial Anglo-Saxon word (for example climb or begin). A third aspect by which synonyms are differentiated is technicality. This means that experts in particular professions, trades, sports or hobbies often use technical vocabulary, which is also called jargon, to be as precise as possible or to express their membership of a certain group or domain.
 This section follows Jackson (1988: 68-74) where no other source is indicated.