Lexical Relations

1 Introduction

There is no doubt, also - and especially - among experts, that our mental vocabulary is highly organised. There are a lot of relations between the single words of a language and the meanings of these words, respectively. Among linguists, these relations are called “semantic relations”, “sense relations” or “lexical relations”. These semantic relations can be analysed and described for the most part, and in the following, the most important ones of these relations are to be presented.

In order to give a short, critical description of the state of the art, it must be said that there are lots of research projects on this topic. However, this paper can only include some of them. Literature which was used can be found under point six, “List Of Works Cited”. Project delimitations have only been made as far as detail is concerned. Since this paper is only a very short piece of research, the authors have confined themselves not to go into too much detail, but rather try to give a good survey of the topic.

Lexical relations can be roughly divided into:

Types of ambiguity (polysemy, homonymy)

Types of congruence (synonymy, hyponymy, meronymy, relations of contrast)

Lexical fields

2 Types Of Ambiguity

2.1 Polysemy And Homonymy

Polysemy can be defined as “a term used in semantic analysis to refer to a lexical item which has a range of different meanings” (Crystal 1997, 297). Crystal gives as example for polysemy the lexical item “plain”, which has the different meanings “clear”, “unadorned”, “obvious”, etc.(ibid. Crystal).

Now, the problem that arises for linguists is how to distinguish polysemy from another type of ambiguity, from homonymy. Crystal defines homonymy as “a term used in semantic analysis to refer to [two or more] lexical items which [happen to] have the same form, but differ in meaning” (Crystal 1997, 185). Crystal’s examples here are “bear” and “ear”. “Bear” can define an animal or can have the meaning of “to carry”, “ear” can refer to the human body or to corn (ibid. Crystal).

In these examples, homonymy covers both spoken and written forms, but it is also possible that the identity of two lexemes is within a single medium, in which case linguists would speak of partial homonymy or heteronymy (ibid. Crystal). One can distinguish two types of partial homonymy:

Homography: two lexical items have the same written form, but differ in pronunciation (an example would be the two lexical items of “lead”, one pronounced [li:d] and meaning “to be in front”, the other pronounced [led] and defining a special kind of metal).

Homophony: two lexical items have the same pronunciation, but differ in spelling

(e.g. the two lexical items “led” and “lead”, both of which are pronounced [led], the first being the past tense of “to lead”, the latter again defining a special kind of metal).

Thus, polysemy and homonymy can be distinguished from each other by the existing or missing relatedness between the meanings which can be allocated to one phonological form. What is the core of the matter, is the question to what extent this notion of “relatedness” can be specified. In other words: how can “relatedness” be defined? If a clear and precise definition could be provided, the whole problem would be solved, for then the phenomenon of phonological forms whose relatedness can be proved would be called “homonymy”, whose relatedness cannot be proved would be called “polysemy”. However, as happens so often in the field of semantics, one cannot provide a clear and indisputable definition of the term “relatedness”. There are two basic approaches to this matter, one offered by historic semantics, the other by synchronic semantics.

2.2 “Relatedness” As Seen By Historic Semantics

Historic semantics interprets the notion “relatedness” mostly genetically and therefore speaks of polysemy if a lexeme with different meanings bears the same etymological roots (Kastovsky 1982, 121). Examples are “game” with the two meanings “wild animals” and “playful activity” or “funny” meaning either “strange” or “amusing”. Both examples show lexemes whose different meanings have the same etymological roots and are thus interpreted as polysemy by historic semantics.

The opposite of polysemy, homonymy, is said to occur if words of different etymological origin have happened to develop the same phonological form in time (ibid. Kastovsky). A good example provide the two different lexemes of the phonological form “meal”. “Meal 1” has the meaning of “food” and roots from the Old English word “mæl”, while “meal 2” means “powder” and has its roots in the Old English word “melo”.

2.3 “Relatedness” As Seen By Synchronic Semantics

For synchronic semantics, it can only be relevant if a competent speaker of the language can establish a natural connection between the various meanings of a lexeme. This natural connection can be missing even though there are equal etymological roots, for example if we take “sole”, meaning either “part of a shoe” or “kind of fish”. Another example is “pupil”, meaning either “part of the eye” or “student”. These are examples where the meanings of the lexeme have the same etymological origins, however, only few people will claim that a shoe and a fish bear a natural connection, the same applies for eye and student. So, synchronic semantics says that this is homonymy.

Vice versa, also words of different etymological origins can semantically approach each other, so that their relation to each other can be interpreted as polysemy from a synchronic point of view (Kastovsky 1982, 121 – 122). A good example is the phonological form “ear”, which has the two Old English forms “ear” and “eare”, the first meaning the organ used for hearing, the latter the upper part of the corn plant. In a synchronic semantic approach, however, one may argue that the ear referring to corn is the metaphorical ear of the corn plant, that it resembles a human ear, and thus this phenomenon cannot be homonymy, but must be polysemy.

2.4 Psychological Relatedness

Some linguists (Leech, Lehrer, Nida, Schifko, Ullmann cited according to Kastovsky 1982, 122) have replaced the criterion of etymological relatedness by “psychological relatedness”, which happens “if a user of the language is able to postulate a connection between them [i.e. two lexical meanings] by lexical rules, e.g. by the rule of metaphorical transfer”(Leech 1974, 229, cited according to Kastovsky 1982, 122).

However, this utterance of Leech is not a clear definition of polysemy either, since no two speakers of a language have the same opinion of which lexical meanings have a connection and which have not. The safest position in this difficult matter is probably Lyons’, who says that the distinction between homonymy and polysemy is in the end always indeterminate and arbitrary (Lyons 1968, 406, cited according to Kastovsky 1982, 122).

For structural semantics, however, this problem is of minor importance, since the whole matter is upon semantic oppositions, no matter if the different meanings belong to one or more lexemes, as long as it is clear that these really are different meanings.

[...]