Author(s): Dr John B Corbett
1.0 Sense Relations
Over the last couple of weeks, we’ve been looking at the semantic relations involved in what is called reference, the relationships of words to the world. We saw that the word is a combination of form and concept, and that the form relates to phenomena in the world (or more specifically, to referents in the universe of discourse) via mental concepts, which seem to be fuzzy categories centred around what are called prototypes. The study of the so-called semiotic triangle – the relations between form and concept, between concept and referent, and (indirectly) between form and referent make up one branch of the discipline of semantics. The Sapir-Whorf hypothesis makes futher claims about the relationship between language, concepts and the world, arguing that our thoughts are determined by the language or languages we speak, and that this linguistic model is imposed upon the world. Last week we considered arguments for and against this theory. This week we are going to look at another branch of semantics. We’ll be focusing on word-meaning, but instead of looking at the relations between words and the world, we’ll be looking at the relations between words and other words. That is, instead of exploring reference, we’ll be looking at sense relations.
There are various types of sense relation. Traditionally, semantics looked at synonymy and homonymy. Synonymy is the relation between two or more words of similar meaning, let’s say car and automobile, or sofa and couch. These words are synonyms – you probably have heard that word before.
Homonymy is the reverse situation, in which one word-form has two or more very different meanings, for example, the bank of a river, and the bank that gives or denies you overdraft facilities. Bank (1) and bank (2) are homonyms. There are special cases of homonymy – sometimes the words sound the same but look different: for example, the tea that you drink versus the tee that you hit a golf-ball from. These are homophones. Alternatively, words may look identical but sound different – thus you might weep a single tear if you tear your best jacket. These are homographs.
As an intellectual pursuit, homonymy is of minor interest, and not just because it’s difficult to pronounce. Synonymy and homonymy originally grabbed scholars’ attention because there was a fashion, mainly in the 17th Century, to find the perfect language. In the perfect language, some scholars argued, there would be a one-to-one relationship between word-form and concept – messy things like synonyms and homonyms would be outlawed. Anyway, unless you are interested in contributing to the language planning of Esperanto, or Klingon, homonyms do not have much lasting attraction.
Synonyms do, though. We’ll be considering three types of sense relation in some detail today: synonymy, antonymy and hyponomy. These sense relations are more interesting than homonymy, because they do give us considerable insight into the way meanings are generally structured in language. The first of these at least should be reasonably familiar: synonymy, as we have seen, is the relation that holds between words of similar meaning; antonymy is the relation that holds between words of opposite meaning; and hyponomy is the relationship that holds between different members of a category (eg the relations between words like apple, pear, orange and the more general term fruit).
If I give these brief definitions then the sense relations do not sound too problematical – some words mean the same thing, others mean opposite things, others are co-members of a given category. So what? But in language nothing is ever quite that simple, as you have probably already seen. So I want to look in turn at these three sense relations — synonymy, antonymy and hyponymy — and discuss what makes them a little more problematical and therefore a little more interesting.
Let’s start with the most familiar one of the three – synomymy. What does it mean to say that words have the same or similar meaning? If you want to look up the synonyms a word has, then you go to a Thesaurus, a reference book sometimes known as a reverse-dictionary because it is classified by meanings, not alphabetically, by word-forms. Word-processing programs also now have thesauruses (or thesauri) as standard features. They are interesting to look at. The most famous thesaurus is Roget’s, named after Peter Mark Roget, a Londoner born of Swiss parents in 1779. Young Peter went to Edinburgh University, graduated as a doctor, practised in Manchester and London, and in 1852 he published a Thesaurus of English Words and Phrases Classified and Arranged so as to Facilitate the Expression of Ideas and Assist in Literary Composition. The 19th Century was not the century of snappy titles. Soon this cumbersome work was known simply as Roget’s thesaurus, and it was so popular that on Peter’s death, in 1869 at the age of 90, it became a kind of family business: his son, John Lewis Roget revised and expanded it, and his grandson later did the same. The Roget you see today will be one of many different editions down the decades. If you look at one recent copy, Roget’s
International Thesaurus 4th edn, revised by Robert Chapman and published in 1984 by Harper Collins, you will see among the many synonyms given for writer the following:
Pen or pencil driver or pusher [slang]
Inkslinger or Ink-spiller [both slang]
Knight of the plume or pen or quill [informal]
Chirographer [i.e. a person whose job it is to write]
Do all these words really ‘mean the same thing’? Obviously not, although we might readily grant that there is a connection between them – they all have to do with the activity of writing. We might say, then, that they are ‘loosely synonymous’ or that they belong to the same semantic field.
The acknowledgement of ‘loose synonymy’ leads us to a question: does ‘strict’ synonymy ever exist? Are there words in the language that have exactly the same meaning? What about nouns like ‘couch’, ‘settee’ and ‘sofa’? Is there a difference in meaning between these words?
This question has been the matter of some debate amongst linguists. Some like Stephen Ullman argue that strict synonymy does not exist because no two words are ever completely inter-changeable. For some people, ‘couch’, ‘settee’ and ‘sofa’ are different in terms of formality – though you might get into arguments about which are the formal and informal terms. If you design furniture, you might have a technical definition that distinguishes the three – though as we saw in our earlier discussion of tomatoes, avocados and cucumbers, those well-known fruits, scientific categorisations do not always correspond to linguistic concepts. Certainly, if you look at English as a whole, the distribution of the three terms is different – you can insult someone by calling them a ‘couch potato’ but you would look rather silly if you screamed that they were nothing but a ‘sofa potato’ or a ‘settee potato’. So, according to Stephen Ullman, strict synonymy just doesn’t exist: there are always differences of register (i.e. formality and informality), dialect, or distribution of usage to distinguish words and meanings.
Other linguists disagree. John Lyons argues rather cleverly that despite the reservations of people like Ullman, strict synonymy is possible. You just have to narrow your focus and agree that because two words are strictly synonymous in some contexts, they need not be synonymous in others. Lyons devised a formula for strict synonymy that goes something like this:
Take two sentences, S1 and S2, which differ only in one word – x is substituted by y. If S1 and S2 are identical in meaning, then x and y are strict synonyms in that context.
So, if you agree that The chair is broken, so could you use the sofa? is identical in meaning to The chair is broken, so could you use the settee/couch? then (in those contexts) the words are strictly synonymous. However, if you agree that the sentences She works as a writer and She works as a clerk or She works as a calligrapher mean different things, then writer, clerk and calligrapher are not strictly synonymous (again, in these contexts).
Synonymy is interesting. It makes us think closely about the relationships between different words of similar meaning, and gives us insight into the way we use them – why the casting couch but not the casting sofa (is it just the alliteration?). What would be the effect of putting on your passport: Profession: Knight of the Plume? Would you really say to the Immigration Officer that it’s an informal synonym for ‘writer’?
Synonymy can actually be described as a special case of hyponymy, as we shall see. Hyponomy is a more recent term in semantics than synonymy – it refers to the sense relation that holds between classes and their members. So, spanner, screwdriver, hammer, drill are all members of the class, tool. In technical terms, spanner, screwdriver, drill, etc are all co-hyponyms and tool is the superordinate term.
If we want to apply logic to this sense relation, then we can devise a formula along these lines:
Take two sentences, S1 and S2, which differ only in one word – x is substituted by y. If S1 implies S2 but S2 does not imply S1, then y is a hyponym of x in that context.
Let’s take the example of hammer and its superordinate term, tool. A sentence like Could you pass me that hammer? implies Could you pass me that tool? However, if you were to say, Could you pass me that tool? You might not necessarily have a hammer in mind. You might have what Americans are beginning to call a whole nother kind of tool in mind.
Unless you look closely into it, hyponymy might appear a kind of trivial subject. But it has its hidden depths. Two points can be made about hyponomy here. First, recall again the discussion we had about whether tomatoes and avocados are vegetables or fruit. To summarise, we said that English-speakers generally conceive of them as fruit, although technically (scientifically, and in some other languages) they are classified as fruit. In hyponymy, then, we would categorise them as vegetables, because here we are interested in the structure of mental concepts, not in the structure of botanical species. Thus tomato and avocado would be a co-hyponym with carrot in English, while avocado would be a co-hyponym with strawberry or passion fruit in Portuguese. Hyponomy can give insights into the way different cultures structure reality.
Secondly, as I said a few minutes ago, synonymy can actually be seen as a special case of hyponomy. All we need to do is alter our formula for hyponymy slightly and we come up with another formula for synonymy:
Take two sentences, S1 and S2, which differ only in one word – x is substituted by y. If S1 implies S2 and S2 also implies S1, then y is a synonym of x in that context.
Try it out with ‘couch’ and ‘sofa': If She was lying on the sofa implies She was lying on the couch and She was lying on the couch also implies She was lying on the sofa then ‘couch’ and ‘sofa’ are synonyms in that context. Again, for some people, with different mental concepts of sofa and couch, they may not be synomyms, but for many people they probably are.
The main point to grasp here is that logical formulae involving implications are being used to clarify and define semantic terms and relations. Much work in semantics became very formulaic and algebraic for much of the first half of the twentieth century. After the 1950’s things loosened up a bit, as we shall see later in the course.
After synonymy and hyponymy, antonomy should be a cinch, a pushover, a trifle, a thing of naught, mere child’s play, a piece of cake, duck soup. But it isn’t.
Let’s try a little audience participation, to cheer things up. I’ll give you a word, and you give me its opposite, its antonym. So, when I say hot you say cold, and so on. Got the picture? Ok, let’s go.
Single (Divorced?) (Separated?)
Winter? (Autumn?) (Spring?)
It should be clear from this little activity, that not all words have opposites, or at least, not all words have opposites in the same way. In semantics, antonyms are just one aspect of what is more broadly referred to as incompatibility of meanings. I want to conclude today’s lecture by looking briefly at different ways in which word meanings can be incompatible.
Technically, antonomy is restricted to those words like hot/cold, big/small which are often gradable adjectives. The ‘opposites’ can be seen as occupying space at diametrically-opposite ends of a scale, thus big/small, huge/tiny.
Complementarity applies to two-term sets of words where one excludes the other. Thus if you are married, it implies that you are not single; and if you are single, it implies you are not married. Other states fall into the general semantic field, but if you are divorced or separated you can argue that you are single once more. The same kind of relation holds for up and down.
Relations of converseness are said to apply to words like buy/sell; give/receive; husband/wife. Again implications are used to clarify this relationship. If you buy something from me, then it implies that I sell something to you. If x is y’s husband, it implies that y is x’s wife. Like complementarity, converseness relates to two-term sets of words.
The most general type of ‘oppositeness’ is simply incompatibility. This is a little like complementarity but pertains to sets of words greater than two. Seasons of the year fall into a four-word set, although we tend to perceive summer/winter, spring/autumn as opposites. As numbers in the set increase (to twelve with months of the year) it is more difficult to assign a single opposite to each word. But we can say that a sentence like ‘It’s January’ implies that it isn’t February/March/April/etc. If it is one, then it is not one of a set of others.
Antonymy, incompatibility, or oppositeness, then, is a sense relation which again tells us a lot about the ways in which vocabulary is structured in English. So far we have been looking at present-day English and assuming that meanings, though complex, are fairly stable. Next week we shall look at how the meanings of words change and what happens when they do.
This work is protected by copyright. All rights reserved.
- University of Glasgow: Copyright © 2004 The University of Glasgow. All rights reserved