{"id":1310,"date":"2024-12-03T16:33:59","date_gmt":"2024-12-03T14:33:59","guid":{"rendered":"https:\/\/linguistica.info\/b\/lei\/?page_id=1310"},"modified":"2025-07-02T08:35:12","modified_gmt":"2025-07-02T06:35:12","slug":"6-8-lexical-relations","status":"publish","type":"page","link":"https:\/\/linguistica.info\/b\/leiwp\/toc\/6-lexical-semantics\/6-8-lexical-relations\/","title":{"rendered":"6.8 Lexical relations"},"content":{"rendered":"

A traditional way of investigating the meaning of a word is to study the relationships between its meaning and the meanings of other words: which words have the same meaning (for example, mad<\/em> and crazy<\/em>), opposite meanings (for example, mad<\/em> and sane<\/em>), and so on. Note that such relations hold only between specific senses. For example, only one sense of mad<\/em> is a synonym of crazy<\/em>, while another sense is a synonym of angry<\/em>. That is why these relations are sometimes called sense relations<\/strong>, but the less specific term lexical relations<\/strong> is also used.<\/p>\n

In this section we discuss the most important sense relations: synonymy, several types of antonymy, hyponymy, and meronymy. We will define each of these relations in terms of relations between sentence meanings, because it is easier for speakers to make reliable judgments about sentences, where words are used in context, than about words in isolation. We will also mention types of linguistic evidence that we can use in order to identify each relation.<\/p>\n

These relations all have one thing in common: they hold between words that have some semantic similarity. If we are interested in the meaning of the word big<\/em>, it seems natural to look at its meaning relations with words like large<\/em>, small<\/em>, enormous<\/em>, etc. Comparing it with words like alive<\/em> or green<\/em> seems unlikely to be very enlightening. They also hold between words of the same word class.<\/p>\n

This does not mean that words belonging to different word classes cannot be compared, but the relationships that hold between such words are very different. For example, the words big<\/em> (adjective), grow<\/em> (verb), or size<\/em> (noun) all refer to the property \u2018size\u2019 \u2014 big<\/em> means something like \u2018having a large size\u2019 and grow<\/em> means something like \u2018become big\u2019. But these relationships are more similar to the relationship between simple and derived words \u2014 size<\/em> and big<\/em> are related in the same way as size<\/em> and sizeable<\/em>, and big<\/em> and grow<\/em> are related in the same way as large<\/em> and enlarge<\/em>.<\/p>\n

Synonyms<\/h2>\n

We often speak of synonyms as being words that \u201cmean the same thing\u201d. A more rigorous definition is that two words are synonymous (for a specific sense of each word) if substituting one word for the other does not change the meaning of a sentence. For example, we can substitute the word frightened<\/em> with scared<\/em> in the following sentence:<\/p>\n

\n
(1a)<\/div>\n
John frightened<\/em> the children.<\/em><\/div>\n<\/div>\n
\n
(1b)<\/div>\n
John scared<\/em> the children.<\/em><\/div>\n<\/div>\n

Since the two sentences have the same meaning, we can say that frighten<\/em> is a synonym of scare<\/em>.<\/p>\n

Complete synonymy is rare, however \u2014 some linguists would even say that it does not exist at all. Even for senses that seem truly equivalent in meaning, there are often differences in usage that suggest some difference between them. For example, tall<\/em> and high<\/em>, which both have the sense \u2018extends a long way from the bottom to the top when it is upright\u2019 as one of their senses, can be substituted with each other in (2a), but not in (2b):<\/p>\n

\n
(2a)<\/div>\n
Aylin\u2019s apartment building is tall\/high.<\/em><\/div>\n<\/div>\n
\n
(2b)<\/div>\n
Aylin is tall\/*high.<\/em><\/div>\n<\/div>\n

Antonyms<\/h2>\n

We often speak of antonyms<\/strong> as being words that mean the \u201copposite\u201d of each other. But what do we mean by \u201copposite\u201d? We obviously do not mean \u201cas different as possible\u201d. As noted above, the meaning of big<\/em> is very different from the meanings of alive<\/em> or yellow<\/em>, but neither of these words is an antonym of big<\/em>. When we say that big<\/em> is the opposite of small<\/em>, or that dead<\/em> is the opposite of alive<\/em>, we mean that they express different values of the same property or attribute. Big<\/em> and small<\/em> are different values of the property \u2018size\u2019, while dead<\/em> and alive<\/em> express different values of the property \u2018vitality\u2019. So two words which are antonyms actually share most aspects of their meaning, and differ only with respect to the value of one particular property.<\/p>\n

The term antonym actually covers several different sense relations, whose differences correspond to differences in the nature of the property or attribute whose values they express. We will look at four commonly recognized types of antonymy.<\/p>\n

Gradable antonymy<\/h3>\n

Gradable antonyms<\/strong> (also called scalar antonyms<\/strong>) express opposite ends of a scale. Examples are big<\/em> and small<\/em>, hot<\/em> and cold<\/em>, or easy<\/em> and difficult<\/em>. If we replace one member of a gradable antonym pair with the other in a sentence, then only one of the sentences can be true:<\/p>\n

\n
(3a)<\/div>\n
It is very hot today.<\/em><\/div>\n<\/div>\n
\n
(3b)<\/div>\n
It is very cold today.<\/em><\/div>\n<\/div>\n

If (3a) is true, then (3b) must be false and vice versa. However, they could both be false. It is possible (and natural) to say something like (3c):<\/p>\n

\n
(3c)<\/div>\n
It is neither very hot nor very cold today.<\/em><\/div>\n<\/div>\n

There are many values between the opposite ends of a scalar property that typically have corresponding intermediate words, for example, warm, tepid, cool<\/em> pick out points somewhere between hot<\/em> and cold<\/em> on the temperature scale.<\/p>\n

We can use gradable antonyms to talk about different degrees of the property they name:<\/p>\n

\n
(4a)<\/div>\n
hotter<\/em>\/colder<\/em>, easier<\/em>\/more difficult<\/em><\/div>\n<\/div>\n
\n
(4b)<\/div>\n
a little too hot<\/em>, a little too difficult<\/em><\/div>\n<\/div>\n
\n
(4c)<\/div>\n
not hot enough<\/em>, not difficult enough<\/em><\/div>\n<\/div>\n
\n
(4d)<\/div>\n
How hot is it today? How difficult is that exercise?<\/em><\/div>\n<\/div>\n
\n
(4e)<\/div>\n
It is very hot<\/em>, it is slightly difficult<\/em><\/div>\n<\/div>\n

Complementary antonymy<\/h3>\n

Complementary<\/strong> or simple<\/strong> antonym<\/strong> pairs exhaust the possible values that a particular property or attribute has, that is, they name discrete values. Examples are alive\/dead<\/em>, on\/off<\/em>, or inside\/outside<\/em>. As was the case with gradable antonyms, if we replace one member of a complementary antonym pair with the other in a sentence, then only one of the sentences can be true:<\/p>\n

\n
(5a)<\/div>\n
Johnny is alive.<\/em><\/div>\n<\/div>\n
\n
(5b)<\/div>\n
Johnny is dead.<\/em><\/div>\n<\/div>\n

However, unlike in the case of gradable antonyms, they cannot both be false: It sounds odd, for example, to say the following:<\/p>\n

\n
(5c)<\/div>\n
?? Johnny is not alive, but he is not dead either.<\/em><\/div>\n<\/div>\n

This is because there are no intermediate points between the opposites: a person is either alive<\/em> or dead<\/em>, a switch is either on<\/em> or off<\/em>, you are either inside<\/em> or outside<\/em> a particular room, etc.<\/p>\n

A significant challenge in identifying simple antonyms is the fact that speakers sometimes use them like gradable antonyms. For example, they might use (5c) if Johnny is a zombie \u2014 zombies are often described as being undead<\/em>, implying that they are not dead but not really alive either.<\/p>\n

However, the gradable use of simple antonyms is typically possible only if we do not use them strictly literally. They will also not behave like gradable antonyms in all linguistic contexts. For example, the following expressions seem natural at least in some contexts:<\/p>\n

half-dead, half-on, half-off<\/em>
\nmore dead than alive, more off than on<\/em><\/p>\n

But the following expressions seem odd under any circumstances:<\/p>\n

\n
(6a)<\/div>\n
?? deader\/more alive<\/em>, more off\/more on<\/em><\/div>\n<\/div>\n
\n
(6b)<\/div>\n
?? a little too dead, a little too off<\/em><\/div>\n<\/div>\n
\n
(6c)<\/div>\n
?? not dead enough<\/em>, not off enough<\/em><\/div>\n<\/div>\n
\n
(6d)<\/div>\n
?? How dead is Johnny?<\/em> How off is that switch?<\/em><\/div>\n<\/div>\n
\n
(6e)<\/div>\n
?? Johnny seems slightly dead<\/em>. The switch seems very off<\/em>.<\/div>\n<\/div>\n

As we saw earlier, for true gradable antonyms, all of these patterns are normally fully acceptable.<\/p>\n

Converse antonymy<\/h3>\n

Converse antonyms<\/strong> (or relational antonyms<\/strong>) refer to entities in an asymmetric relation, e.g. parent-child, above<\/em>–below<\/em>, buy-sell<\/em>. As in the case of gradable antonyms, if we replace one member of a converse pair with the other, we get sentences that cannot both be true but that can both be false:<\/p>\n

\n
(7a)<\/div>\n
Cansu and Emre are Aylin\u2019s parents.<\/em><\/div>\n<\/div>\n
\n
(7b)<\/div>\n
Cansu and Emre are Aylin\u2019s children.<\/em><\/div>\n<\/div>\n

What makes converse antonyms special, however, is that if we replace one word with the other and reverse the order of the arguments<\/em>, we produce sentences which do not mean the same thing, but which are true in the same situation:<\/p>\n

\n
(8a)<\/div>\n
Cansu and Emre are Aylin\u2019s parents.<\/em><\/div>\n<\/div>\n
\n
(8b)<\/div>\n
Aylin is Cansu and Emre\u2019s child.<\/em><\/div>\n<\/div>\n

Reverse antonymy<\/h3>\n

Reverse antonyms<\/strong> also refer to an asymmetric relation, but specifically, to processes that reverse each other’s outcomes. Examples are fill\/empty<\/em>, push\/pull<\/em>, break\/fix<\/em>, heat\/cool<\/em>. If we replace one member of such a pair with the other, we get sentences which, again, cannot both be true in the same situation but that can both be false:<\/p>\n

\n
(9a)<\/div>\n
Zoe filled the bathtub.<\/em><\/div>\n<\/div>\n
\n
(9b)<\/div>\n
Zoe emptied the bathtub.<\/em><\/div>\n<\/div>\n

Unlike with converse antonyms, we cannot reverse such sentences (*The bathtub filled Zoe<\/em>). One interesting property of reverse antonyms is that they often allow an interesting use of again<\/em>:<\/p>\n

\n
(9c)<\/div>\n
Zoe filled the bathtub and then emptied it again.<\/em><\/div>\n<\/div>\n

In such cases, again<\/em> does not mean that the action named by the second verb is repeated (repetitive reading), but rather that the situation is restored to its original state (restitutive reading).<\/p>\n

Hyponymy and taxonymy<\/h2>\n

When two words stand in a generic-specific relationship, we call this relation hyponymy<\/strong>. We refer to the more specific word (e.g. dog<\/em> or cat<\/em>) as the hyponym<\/strong> and to the more generic word (e.g. mammal<\/em>) as the superordinate<\/strong> or hyperonym<\/strong>. When you replace a hyponym with a hyperonym in a true sentence, the resulting sentence is also true:<\/p>\n

\n
(10a)<\/div>\n
Lassie is a dog.<\/em> (hyponym)<\/div>\n<\/div>\n
\n
(10b)<\/div>\n
Lassie is a mammal.<\/em> (hyperonym)<\/div>\n<\/div>\n
\n
(11a)<\/div>\n
Aylin is reading a paperback<\/em>. (hyponym)<\/div>\n<\/div>\n
\n
(11b)<\/div>\n
Aylin is reading a book.<\/em> (hyperonym)<\/div>\n<\/div>\n

However, if you replace a hyperonym with a hyponym in a true sentence, the resulting sentence may be true or false:<\/p>\n

\n
(12a)<\/div>\n
Lassie is a mammal.<\/em> (hyperonym)<\/div>\n<\/div>\n
\n
(12b)<\/div>\n
Lassie is a dog.<\/em> (hyponym, true)<\/div>\n<\/div>\n
\n
(12c)<\/div>\n
Lassie is a cat.<\/em> (hyponym, false)<\/div>\n<\/div>\n

This exemplifies a special kind of hyponymy: taxonymy<\/strong> (not to be confused with taxonomy<\/em>). Taxonymy is a relationship where it is natural to say that the referent of the hyponym is a kind of<\/em> or type of<\/em> the referent of the hyperonym, for example: A dog is a kind of mammal<\/em>. There are also hyponymic relations that are not taxonymic. For example, stallion<\/em> is a hyponym of horse, but it sounds odd to say (13):<\/p>\n

\n
(13)<\/div>\n
?? A stallion is a kind of horse.<\/em><\/div>\n<\/div>\n

Meronymy<\/h2>\n

A meronymy<\/strong> is a pair of words expressing a part-whole relationship. The word naming the part is called the meronym<\/strong>, the word naming the whole is called the holonym<\/strong>. For example, hand<\/em>, brain<\/em> and eye<\/em> are all meronyms of body<\/em>; door<\/em>, roof<\/em> and kitchen<\/em> are all meronyms of house<\/em>; etc.<\/p>\n

It is important to remember that when we study patterns of meronymy, we are studying the structure of the lexicon, i.e., relations between words and not between the things named by the words. One linguistic test for identifying meronymy is the naturalness of sentences of the form The parts of an X include the Y, the Z, …<\/em>, as in (14a-c):<\/p>\n

\n
(14a)<\/div>\n
The parts of an arm include the hand, the elbow, …<\/em><\/div>\n<\/div>\n
\n
(14b)<\/div>\n
The parts of a hand include the thumb, the index finger, …<\/em><\/div>\n<\/div>\n
\n
(14c)<\/div>\n
?? The parts of an arm include the thumb, the index finger, …<\/em><\/div>\n<\/div>\n

In terms of the referents of the words\u00a0arm<\/em>, hand<\/em> and finger<\/em>, (14c) is true: the arm includes the hand, thus it also includes the parts of the hand. However, the words arm<\/em>, hand<\/em> and thumb<\/em> \u2014 and words more generally \u2014 do not behave according to the real-world relations between their referents, but according to their sense, and sense relations do not always mirror reference relations.<\/p>\n

 <\/p>\n

\u2190<\/span> Previous section<\/a><\/span> Next section \u2192<\/span><\/a><\/span><\/p>\n

<\/p>\n

CC-BY-NC-SA 4.0, Adapted from Paul Kroeger, Analyzing Semantics<\/em>, shortened and restructured by Anatol Stefanowitsch with minor edits by Berit Johannsen.<\/p>\n","protected":false},"excerpt":{"rendered":"

A traditional way of investigating the meaning of a word is to study the relationships between its meaning and the meanings of other words: which words have the same meaning (for example, mad and crazy), opposite meanings (for example, mad and sane), and so on. Note that such relations hold only between specific senses. For […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":1268,"menu_order":8,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-1310","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1310","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/comments?post=1310"}],"version-history":[{"count":21,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1310\/revisions"}],"predecessor-version":[{"id":2190,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1310\/revisions\/2190"}],"up":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1268"}],"wp:attachment":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/media?parent=1310"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}