Distinctions & Relations
On the previous page, we discussed how basic features can help us to categorise words with regard to their meaning. However, in order to be able to distinguish between words that may occur at the same level within a hierarchy, i.e. potential co-hyponyms, we also need further ways of expressing any other potential relationships that might exist between them, or explain how any rankings or groupings/orderings may have been come about. The latter may either be based on some type of natural ordering, or imposed through the use of artificial categorisation schemes. The descriptive symbols used in the sections below based on Leech (21981).
Distinctions
Polysemy & Homonymy
As we’ve already seen repeatedly before, word forms can have multiple meanings associated with them. In other words, they can be polysemous (Gr. πολλοί, /pɒlˈlɪ/ = many + σήμα = sign). Polysemy – in the traditional sense – only refers to word forms where these meanings are assumed to be related to one another, and word forms that only look, and possibly also sound, the same (i.e. are homographs, as well as homophones), but are (apparently) unrelated to one another, are seen as homonyms (Gr. όμοιος /ˈɔːmɪɒs/ = alike, same; όνομα /ˈɒnɒmɐ/ = name). However, many linguists theses days tend to avoid making the distinction between homonymy and polysemy, and always use the latter term to cover all cases where a word form has different meanings. We’ll adopt that convention here, too, unless there’s an absolutly compelling reason for assuming completely different meanings, and having to say so.
We can essentially distinguish between two different types of polysemy. In the first one, which we’ll call ‘lexical’ polysemy, there’s no change in/difference of word class, simply a difference in meaning, e.g.:
- flight: ‘air travel’ ⇔ ‘act of fleeing’
- ear: ‘hearing organ’ ⇔ ‘top part of a grain’
- bat: ‘flying nocturnal animal’ ⇔ ‘instrument for hitting a ball’
- club: ‘organisation/location’ ⇔ ‘instrument for hitting/beating’
- bachelor: ‘unmarried man’ ⇔ ‘university degree’
- fan: ‘ventilator’ ⇔ ‘extremely fond person/follower’
In the other type, which we’ve repeatedly encountered when we discussed zero derivation, there’s always a change of word class involved, and we’ll thus refer to it as ‘grammatical’ polysemy, as in e.g.:
- book: ‘reading material’ (N) ⇔ ‘reserve a ticket’ (V)
- make: ‘create’ (V) ⇔ ‘brand’ (N)
- turn: ‘move around/sideways’ (V) ⇔ ‘position/point in an ordered sequence’ (N)
- swallow: ‘ingest food’ (V) ⇔ ‘migrating bird’ (N)
- hand: ‘body part’ (N) ⇔ ‘give/pass something’ (V)
- Look through the examples of polysemy above and try to identify which ones may share a common core of meaning, i.e. can be seen as true polysemes.
- Justify your decisions by explaining the potential (or absent) meaning relationships further.
- Think of a few more examples yourself.
Synonymy & Antonymy
As we’ve already seen before, too, it may be useful to try and establish the exact meaning of a word (form) either by contrasting it with another word that is similar in meaning (i.e. a synonym), or one that represents (more or less) exactly the opposite (i.e. an antonym).
When you first learnt about synonyms in school, you were probably told that they’re pairs of words that express the same meaning. However, this type of total synonymy is rarely achievable, apart from maybe if two different dialects use two different words to refer to exactly the same concept. In most cases, though, we have to assume partial synonymy because the exact meaning usually depends on the exact context and/or domain of use, and the meanings may differ very subtly in some cases.
There are a number of different reasons for why synonyms exist, some of which are illustrated in the table below:
| Motivation | Example(s) |
|---|---|
| common vs. learned/borrowed | buy ⇔ purchase; begin ⇔ commence |
| common vs. technical/scientific | water ⇔ H2O; car ⇔ automobile |
| ‘dialectal’/historical | autumn ⇔ fall; boot ⇔ trunk; lorry ⇔ truck |
| ‘grammatical’ | start ⇔ begin; finish ⇔ end |
At least some of the above reasons can be traced back to the influences of other languages at various stages of the development of the English language, as we’ve seen on the first page of this course.
- Open the COCA website.
- For every synonym pair above, enter both alternatives, separated by a forward slash (e.g. buy/purchase) into the box next to the label ‘WORD(S)’, and click on .
- Select the options for both items in the window on the right-hand side, and click on . This should open a window displaying the results in the bottom half of that frame.
- Look through the examples and check to see whether the context the examples occur in fits the descriptions/explanations given in the table. Also check to see whether the label for the text category and source give you any indication.
- If one or both of the words can be grammatically polysemous, observe the differences in behaviour/context, too.
- Always keep a record of your observations in the text box and save the results once you’re finished.
Antonymy occurs when two words express opposite meanings. As we’ve seen before, this can be achieved by adding a negative prefix, or using a completely unrelated word form. There are two different sub-types of antonymy, polarity and in- or converseness. In polarity, two words express opposite ends of a scale, relative to a norm:
- cool (↑) ⇔ uncool (↓)
- high (↑) ⇔ low (↓)
- hard (↑) ⇔ soft (↓)
- hard (↑) ⇔ easy (↓)
- up (↑) ⇔ down (↓)
- rich (↑) ⇔ poor (↓)
In converseness, the antonyms express opposite ‘directions’ or ’relations’, but with an absolute end point:
- buy (∆) ⇔ sell (∇)
- above (∆) ⇔ below (∇)
- possible (∆) ⇔ necessary (∇)
The relative nature of polar antonyms can – at least for adjectives – be tested by checking whether the word can be modified by the adverb relatively.
- Find further examples of antonyms, including some from potentially grammatically polysemous word classes.
- Each time, test whether they are polar or inverse. Justify your decision, especially if they’re not adjectives.
- Also think about about where the norm may come from in polar antonyms.
Gradability
As we’ve seen above in the polar antonyms, contrasting values for adjectives in a system are often expressed relative to a given norm. In some cases, these contrasts can only expressed through two absolute contrasting values, as in the binary distinctions we used in the description of basic semantic features earlier. In this case, we refer to them as being non-gradable, e.g.
- dead (-) ⇔ alive (+)
- ill (-) ⇔ well (+)
If, in contrast, there are intermediate values within the categories or scales involved, then the properties described are relative, rather than absolute, and the items become gradable, e.g.:
- good → better → best
- big (↑) ⇔ small (↓)
For inherently gradable adjectives, their gradability is expressed morphologically through the comparative and superlative morphemes {er} & {est}. In contrast, non-gradable adjectives, because they generally represent absolute properties, cannot be graded morphologically, apart from maybe in highly idiomatic constructions, such as deader than a doornail. Thus, if we want to ‘coerce’ them into becoming relative, we need to modify them using adverbs or adverbial phrases, e.g. half dead, slightly ill, a little pregnant, etc.
- Again, using the COCA corpus we used above, look for any special ways in which non-gradable adjectives may be graded through using adverbs.
- In order to do so, type [r*], followed by a space, and a non-gradable adjective, into the box.
- Click on the radio button for KWIC.
- Click .
- From the context list on the top right, select the most unusual/interesting ones and click again.
- Explore the options to see whether they fit or contradict your expectations, and especially whether you may also find some idiomatic expressions like the one I’ve listed above.
- Try the same thing again by looking for other adjectives preceding the non/gradable ones. To do so, simply change [r*] to [j*].
Markedness
In previous exercises, we’ve repeatedly seen that all words come with a certian set of expectations attached to them, and if we use them in a way that somehow infringes one of the expectations, we can create a new or special meaning. Thus, in using a (normally) sensory verb like see, which is by nature non-dynamic (i.e. momentary or static), in a dynamic way by attaching an ing-form, we either created a special effect of indicating that we assumed that there was something wrong with our senses (e.g. I think, I’m seeing things/ghosts.), or, if used with an animate, human direct object, to indicate that someone is currently in a relationship (e.g. She’s seeing someone.). Similarly, when we used non-count nouns, such as water, in their plural form, we quantified them in an unusal way, either to say that we’re really referring to countable containers of them (e.g. Two waters, please.), or we used them with a more (domain-)specific, specialist meaning (e.g. A country’s territorial waters.). And, every time we use a word in such an unusual/unexpected form, we create a marked impression.
In gradable or polar adjectives, usually one item at one end of the spectrum is a marked one, while the other is either neutral or positive, e.g.
- unmarked: tall, busy, well
- marked: short, idle, ill
A marked term often has negative connotations, i.e. evokes negative or ‘unfavourable’ associations, so it’s generally avoided in positive or neutral constructions, and the marked construction sounds at the very least striking, but possibly even ironic, e.g.:
- he/she’s five foot tall – *?he/she’s five foot short
- are you busy/well? – ?are you idle/ill?
Try to identify further examples of markedness. Hint: it might be easiest to look for items that are similar to the examples given above first...
Relationships
Relationships always involve a kind of ‘comparison’ between the meanings expressed by two words. Unlike the distinctions we discussed above, though, they don’t necessarily express a direct contrast between items, but instead some other type of natural or artificial connection, including perhaps a certain kind of ranking, between two words.
Directionality
In unidirectional/converse relationships, one item is usually the ‘opposite’ to the other, so the situation is similar to antonymy. Sometimes, at least in nouns, there’s also a kind of implicit hierarchical structure involved, where one noun may signal a higher status.
- parent (→) ⇔ (←) child
- boss (→) ⇔ (←) employee
- teacher (→) ⇔ (←) student
- back (←) ⇔ (→) forth (???)
In bidirectional relationships both items are always the same to one another, so the issue of a hierarchy doesn’t arise:
- sibling (<>)
- cousin (<>)
Are the pairs of words brother & sister, and mother & son uni- or bi-directional to one another? Justify your decision, also stating whether there are any potential (assumed) hierarchies involved.
(Multiple) Taxonomies
In multiple taxonomies, there’s usually an ‘abstract’, scientific term grouping together similar items. This term is usually a top-level hypernym and its subordinate members are the direct hyponyms. There is no natural ordering between the co-hyponyms, but there may be culturally ‘imposed’ values or rankings of quality or preference attached to some items. Examples for such multiple taxonomies are:
- METAL: gold (*), copper (†), iron (§), mercury (.), etc.
- DRINK: water (*), juice (†), wine (§), beer (.), whisky (¶), etc.
With metals, we can clearly se what I said above about cultural values attached to items, as in most cultures gold would usually be ranked higher than silver or copper, and the latter again higher than iron. The category of drinks, on the other hand, generally exhibits less of cultural ranking, and it’s more due to personal preference whether one might beer over wine or whisky if one drinks alcohol at all, or juice over water if one doesn’t.
Explicit Hierarchies
Explicit hierarchies, on the other hand, exhibit a clear and ordered ranking, expressed as a chain of multiple linked elements, such as in gradable adjectives. Rankings of this type may be based on relatively clear and ‘natural’ units, as they are often found in measurements, such as in inch → foot → yard → mile, or constitute ‘artificially imposed’ categories/judgements, as in winner → runner-up → last.
Sources & Further Reading:
Leech, G. (1981). Semantics: the Study of Meaning (2nd ed.). London: Penguin.
Taylor, J. (1995). Linguistic Categorization: Prototypes in Linguistic Theory (2nd ed.). Oxford: OUP.