Professional Documents
Culture Documents
Chapter 1
Chapter 1
What is Lexicology?
Chapter Overview
1.1 Lexicology defined 2
Morphology 3
Semantics 5
Etymology 7
Lexicography 9
1.2 Lexicology as a level of language analysis 10
Lexicology and phonology 10
Lexicology and syntax 12
1.3 The structure of English vocabulary 14
The word and its associative field 14
Lexical fields 16
Word families 19
Word classes 21
1.4 Summary 23
Exercise 1/1
Examine the following definitions of ‘lexicology’. What do they
agree on as the scope of lexicology? And where do they disagree?
1.1.1 Morphology
Morphology is the study of morphemes and their arrangements in forming
words. Morphemes are the smallest meaningful units which may constitute
words or parts of words. They are ‘smallest’ or ‘minimal’ in the sense that they
cannot be broken down further on the basis of meaning, as Katamba (2005:
29) puts it: ‘morphemes are the atoms with which words are built’. They are
‘meaningful’ because we can specify the kind of relationship they have with
the non-linguistic world.
Consider the following items: cat, child, with, sleeping, armchairs, farmer.
A close examination shows that cat, child and with cannot be analysed
further into meaningful units. However, sleeping, armchairs and farmer can
be analysed as ‘sleep + ing’, ‘arm + chair + s’ and ‘farm + er’. The items cat,
child, with, sleep, -ing, arm, chair, -s, farm and -er are all morphemes. Some
are simple words such as cat, child, with, sleep, arm, chair and farm, while
others are only parts of words such as -ing, -s and -er. But both types meet
our definition of morpheme. On the one hand, they are minimal, since they
cannot be broken down into further meaningful units; on the other hand,
they are meaningful, because we can establish a stable relationship between
each item and the non-linguistic world of experience. For example, the
references of cat, farm and chair can be explained by pointing or acting out
the meaning as in ‘This is a chair’, ‘That is a farm’, ‘It is a domestic animal that
goes “meow, meow” ’. The meaning of with may be given as ‘in company of ’,
‘in antagonism to’; that of -s as ‘plural’; while that of -er may be expressed as
follows: ‘-er combines with the preceding lexical item to designate things or
persons with a function describable in terms of the meaning of the preceding
morpheme’. For example, the meaning of -er in farmer and dreamer is
describable in terms of those of farm and dream with which the morpheme
-er is combined.
Morphemes that can occur alone as individual words are ‘free’ morphemes.
Those that can occur only with another morpheme are ‘bound’ morphemes.
Thus, the morphemes ‘cat’, ‘chair’ and ‘farm’ are free, while ‘-ing’, ‘-s’ and ‘-er’
are bound, indicated by the hyphen (-). Any concrete realization of a
morpheme in a given utterance is called a ‘morph’. Hence, the forms cat,
chair, farm, -ing, -s and -er are all morphs. Morphs should not be confused
with syllables. The basic difference between the two is that while morphs are
manifestations of morphemes and represent a specific meaning, syllables are
parts of words which are isolated only on the basis of pronunciation.
An examination of a number of morphs may show that two or more
morphs may vary slightly and still have the same meaning. For example, the
indefinite article may be realized either as a or as an, depending on the
sound (not the letter) at the beginning of the following word. Morphs which
are different representations of the same morpheme are referred to as
‘allomorphs’ of that morpheme (from Greek allo, ‘other’, and morph, ‘form’).
For example:
The last pair of words deserves some comment. Its members begin with u
and o, which are classified as vowel letters. However, while union begins with
the same sound as yes which is treated as a consonant, onion begins with the
same sound as onwards, which is a vowel; hence ‘a union’ vs. ‘an onion’.
The use of ‘vs.’ (versus) highlights the point that where the allomorph an
occurs, its counterpart a cannot occur and vice versa. They are therefore
mutually exclusive and are said to be in complementary distribution. It should
be pointed out that, as a descriptive term, ‘distribution’ refers to the total set of
distinct linguistic contexts in which a given form occurs, sometimes under
different morphological shapes. For example, the distribution of the indefinite
article described above may be defined as: a before consonant sounds (e.g. a
battle) and an before vowel sounds (e.g. an apple).
1.1.2 Semantics
Semantics is generally defined as the study of meaning. Its aim is therefore
to explain and describe meaning in natural languages. The term ‘meaning’ is
used here in the ordinary, non-technical sense, without reference to any
particular theoretical framework. Most linguists agree that meaning
pervades the whole of language. However, they are not always unanimous on
the terms to be used in the discussion of semantics. For our purpose in this
book, we adopt the terminology presented and the theoretical distinctions
made by Jackson (1988: 244–7) in his brief treatment of semantics.
To highlight the pervasive nature of meaning, Jackson states that if we are
to talk about semantics at all, then we should identify several kinds of
semantics: pragmatic semantics, which studies the meaning of utterances in
context; sentence semantics, which handles the meaning of sentences as well
as meaning relations between sentences; lexical semantics, which deals with
the meaning of words and the meaning relations that are internal to the
vocabulary of a language. Semantics is usually approached from one of two
They may be considered meaningless in the sense that a coin cannot be both
‘a pound’ and ‘a euro’ at the same time. Similarly, it may be argued that an
object cannot be ‘a doll’ and ‘a bomb’, just as the same object cannot be
simultaneously ‘a walking stick’ and ‘a gun’. But with a bit of imagination, one
can think of contexts where such utterances, and others like them, can be
considered acceptable. For example, in a discussion of exchange rates, a
pound coin may be said to have the value of a euro, and so this statement
would be true, or someone could have mistaken one for the other and is
being corrected; in a film, an actor could be carrying a doll or a walking stick
which in fact could be deadly weapons such as a bomb or a gun. To paraphrase
Leech (1969: 13), the ‘effective message’ in all such utterances is ‘What appears
as an “x” is in fact a “y” ’.
There are other types of meaningless utterance that may be acceptable for
various reasons. Some may involve ‘slips of the tongue’, ‘typographical errors’,
‘sarcasms’, ‘different figures of speech’, etc. Others may be considered
deviations from the norm of the language under study. Still others may have
different origins or justifications. For example, if a person who has a bad cold
and a completely blocked nasal cavity says ‘It’s dice beeting you’, after he/she
has just been introduced to someone, this utterance may be considered
meaningless, strictly speaking. However, the ‘effective message’ it conveys in
this context would be something like ‘It’s nice meeting you but I have a bad
cold’.
The important point here is that there are several factors that contribute
to the meaningfulness and the acceptability of utterances. As opposed to
utterances that are meaningless but acceptable, others are meaningful but
unacceptable. The latter category includes assertions that are false because of
our knowledge of the real world, rather than for purely semantic reasons.
Consider the following:
We may use different criteria to account for such utterances. For example,
they may be explained by logical argument to highlight the contradictions,
inconsistencies or incompatibilities in the message. From a syntactic point
of view, such utterances are treated as errors in predication, meaning that the
subject or object noun phrases are syntactically unsuitable to the
corresponding verb phrases. Hence, the subjects crocodiles and the basket are
syntactically unsuitable to the verb phrases can fly and ate, respectively. Such
examples point to the fact that all of syntax, semantics and lexicology
contribute to a comprehensive study of language.
1.1.3 Etymology
A third field which should be of particular interest in lexicological studies is
etymology, which may be defined as the study of the whole history of words,
not just of their origin (cf. Durkin 2015). The term ‘etymology’ was coined by
the Stoics, a group of Greek philosophers and logicians who flourished from
about the beginning of the fourth century BCE. They noticed a lack of
regularity in the correspondence between the forms of the language and
their respective contents. In other words, they found no necessary connection
between the sounds of the language on the one hand and the things for
which the sounds stood on the other. Since they were convinced that
language should be regularly related to its content, they undertook to
discover the original forms called the ‘etyma’ (roots) in order to establish the
regular correspondence between language and reality. This was the beginning
of the study known today as etymology.
One of the difficulties faced by etymological studies is that some words
are not etymologically related to ancient forms. It is therefore difficult to
establish and indicate their origins. Consequently, the forms from which
such words are said to derive can only be produced by analogy. Another
difficulty is that while it is possible to specify the exact time when some
terms entered the language, for example through borrowing, it is clearly
impossible to say exactly when a form was dropped, especially since words
can disappear from use for various reasons.
The most crucial difficulty faced with etymological studies is that there
can be no ‘true’ or ‘original’ meaning, since human language stretches too far
back in history. To paraphrase an example given by Palmer (1981: 11), one
may be tempted to say that from the etymological point of view the adjective
nice really means ‘precise’ as in ‘a nice distinction’. But a study of its history
shows that the word once meant ‘silly’ (Latin nescius, ‘ignorant’), and earlier,
it must have been related to ne, ‘not’ and se, probably meaning ‘cut’. The form
se is also used in the modern English words scissors and shears. But at this
level of analysis, one would still be left with the Latin items ne, ‘not’ and se,
‘cut’, the origins of which are still unknown. In other words, no matter how
far back one goes in history, one cannot expect to reach the beginning of
time. So, the Stoics’ quest has proved fruitless.
As suggested in our definition, etymological information goes beyond the
origin of the word. It also makes reference to cognates (i.e. words related in
form) in other languages. Furthermore, in the case of borrowed words, it
gives the source language, together with the date when the borrowing took
place. Finally, it supplies any other information on the previous history of
the word. In dictionary entries, such information is contained traditionally
in square brackets. The amount of detail provided in etymologies varies
from one dictionary to another. But in spite of its potentially wide range of
coverage, etymological information is generally scanty in most monolingual
dictionaries. In fact, it is totally absent from both bilingual and learners’
dictionaries, presumably on the grounds that it is not helpful to language
learners. However, as pointed out by Jackson (1988: 175), it could be argued
that ‘knowledge of etymology may help some learners to understand and
retain new vocabulary items’.
Before we close this brief discussion of etymology, we should mention the
expression ‘folk etymology’. It is a historical process whereby speakers who
cannot analyse an obscure form replace it with a different form which is
Exercise 1/2
Etymology can be a fascinating area of study. Look up the
etymology of bridegroom in your dictionary. Does it mention the
folk etymology, i.e. the change from ‘gome’ to ‘groom’?
Now look up the etymology of adder. Is there any folk
etymology at work here?
Finally, look up the etymology of snake. What does your
dictionary say is the origin? And does it give cognates, i.e. related
words in other languages?
1.1.4 Lexicography
The fourth and last field which is of special interest in lexicological studies is
lexicography, which has already been defined as a special technique, the
writing and compilation of dictionaries. This definition may be considered
rather restrictive. In its widest sense, lexicography may also refer to the
principles that underlie the process of compiling and editing dictionaries.
Some of those principles are clearly lexical or lexicological in nature, while
others stem from the specific domain of book production and marketing.
But lexicographical compilation may be considered as derived from
lexicological theory (Jackson 1988: 248). It is in this sense that lexicography
can be regarded as ‘applied lexicology’.
However, it should be acknowledged that it is only in recent years that the
link between lexicography and linguistics has been clearly established
(Béjoint 2010: Chapter 8). For example, the accuracy and consistency in the
transcription of words and the adoption of a ‘descriptive’ as opposed to a
‘prescriptive’ approach to lexicography are direct applications of linguistic
principles. It may be argued that initially lexicography developed its own
principles and tradition independently of linguistics in general; but this is
no longer the case. In fact, since current dictionaries are compiled mainly by
lexicographers who have been trained in linguistics, one should expect a
more direct and more substantial input from lexicology. However, it should
be pointed out that lexicology is not the only branch of linguistics which
provides an input to lexicography. Clearly, morphology, syntax and
phonology do. And sociolinguistics, too, contributes, not only in the study
and selection of the language variety to be used in the dictionary, but also in
the inclusion of information on style and registers. For a detailed discussion
of lexicography as applied lexicology, see Chapter 8.
cases, the phonological difference does not involve discrete sound units but
‘suprasegmental’ or ‘prosodic’ features such as stress; and yet, even such
differences are enough to differentiate otherwise identical items; e.g. ex’port
(verb), vs. ’export (noun). Note that the symbol (’) is placed immediately
before the syllable which receives primary stress.
Compounds provide another good example to show the relevance of
phonology in lexicology. At first sight, the process of compounding may be
viewed as a simple juxtaposition of two words. Thus, green and house may be
put together to form greenhouse, ‘a glass house for growing plants’. But such
an analysis would be superficial, since the same items can be put together in
the same order to produce green house, ‘a house that is green’. The major
difference between the two utterances is a matter of stress, which is a
phonological feature. But this feature is enough to distinguish compounds
from noun phrases containing the same words. Compare the stress pattern
of the compound nouns in (a) and the corresponding adjective plus noun
constructions in (b).
As a general rule, the primary stress falls on the first word of the compound as
in ’whiteboard (a white smooth surface used in schools and elsewhere for
writing on). The same rule applies to the rest of the words in (a). However, in
the noun phrases in (b) like ,white ’board (as opposed to any board that is
painted green, red or yellow) both words can potentially receive stress. All the
examples in (b) are phrases, not words. Consequently, they function as units at
the syntactic level. They are therefore an indication that stress, which is a
phonological feature, has a direct influence on lexicology and syntax. The
relation between lexicology and syntax is further explored in 1.2.2. This is a
clear illustration of the interdependence of phonology, lexicology and syntax.
However, since language is so complex that it cannot be studied all together,
we must consider each level as if it were autonomous. But it must always be
remembered that such an approach is only methodological and does not
always reflect the way language operates when it is used by its speakers.
At first sight, it may be thought that when judged in terms of how deviant
they are, lexical restrictions are generally not a matter of well-established rules
but of tendencies. In other words, it may be assumed that if asked whether or
not a given lexical association is acceptable, one cannot answer by a categorical
‘yes/no’; one is more likely to give an answer of the nature ‘more/less’, or ‘it
depends on the context’. Such an assumption is, however, an oversimplification.
In syntax as well as in lexicology, there are cases of deviation which may be
answered by ‘yes/no’, and others that can be answered only by ‘more/ less’,
though a ‘yes/no’ answer is more likely in syntax than in lexicology. For example,
a sentence such as ‘Sophisticated mice prefer to eat red elephants’, though
undoubtedly syntactic is lexically doubtful because it does not correspond to
our experience of the world. Judging from our present knowledge of the
natural world, the acceptability of this sentence is not a matter of ‘more/less’ but
of a categorical ‘no’. However, a sentence such as ‘The flower gracefully walked
away’ may seem odd in the sense that ‘flowers’ are not normally associated with
‘walking’, but with some imagination we can picture a context in which this
sentence, which is already acceptable syntactically, is also acceptable lexically.
Similarly, some sentences are clearly ungrammatical, while others are
clearly well-formed syntactically. For example, a sentence such as ‘Did it he
and I’ is clearly deviant only on syntactic grounds and it could be corrected
simply by changing the word order into ‘He and I did it’. But there are also
marginal cases such as ‘Give it to whomever wants it’ versus ‘Give it to
whoever wants it’, where English speakers are not unanimous as to which
alternative is grammatical. As a final observation, it should be pointed out
that some sentences, such as ‘Did it John and the table’, are deviant on both
syntactic and lexical grounds.
Exercise 1/3
Consider the following ‘deviant’ sentences. In which of them would an
alternative selection of words (lexis) make an improvement, and in
which does the arrangement of the words (syntax) need to be adjusted?
differ from each other in form, meaning or both. Such relations are referred
to as ‘paradigmatic’. They are called relations ‘in absentia’, because the terms
involved consist of a word present in the utterance and others that are not
actually in the same utterance but that are substitutable for it in that context.
For example, ‘difficult’ is paradigmatically related with ‘easy’, ‘funny’, ‘silly’, etc.
in expressions such as ‘an easy question’, ‘a funny question’, ‘a silly question’,
etc. Similarly, ‘question’ is paradigmatically related with ‘problem’, ‘word’, etc.
in expressions like ‘a difficult problem’, ‘a difficult word’, etc.
Interesting as the notion of paradigmatic relations may be, it has one
major difficulty: every other word in the language either resembles or fails to
resemble in form or meaning any given word. This difficulty notwithstanding,
the notion has proved useful in language teaching and in the arrangement of
words in a thesaurus. It has also been used to establish cross-references in
the definitions of words in dictionaries.
This important notion was first introduced by Saussure under the
label ‘associative relations’. But the term ‘paradigmatic’ has been substituted
at the suggestion of the Danish linguist Hjelmslev (1963). Another notion
which was also introduced by Saussure was that of ‘syntagmatic’ relations.
They are called relations ‘in praesentia’ since the words involved are actually
co-occurrent items. For example, the word difficult is syntagmatically
related with the article a and the noun question in the expression ‘a difficult
question’.
Paradigmatic and syntagmatic relations may be represented in a diagram
as in Figure 1.2. This shows that every word may be considered in terms of
two dimensions or axes of structure: the ‘horizontal’ or syntagmatic and the
‘vertical’ or paradigmatic. It is precisely in terms of syntagmatic and
paradigmatic relations that the meaning of English words can be determined.
However, our interest in constructions larger than the word is secondary,
though see 5.6 on collocation.
Field theory was first put forward by a number of German and Swiss
scholars in the 1920s and 1930s. However, according to Lyons (1977: 250) its
origin can be traced back at least to the middle of the nineteenth century
and more generally to the ideas of Humboldt and Herder. We shall not
attempt a comprehensive treatment of field theory, nor shall we try to review
the considerable body of descriptive work based on it. Instead, we shall
discuss briefly how the theory may be used to account for the general
organization of English vocabulary (see also 5.5.3).
According to lexical field theory, the vocabulary of a language is essentially
a dynamic and well-integrated system of lexemes structured by relationships
of meaning. The system is changing continuously by the interaction of
various forces such as the disappearance of previously existing lexemes, or
the broadening or narrowing of the meaning of some lexemes. The system is
mainly characterized by the general–particular and part–whole relationships,
which hold not only between individual lexemes and the lexical field within
which they are best interpreted, but also between specific lexical fields and
the vocabulary as a whole. One of the early theorists, Jost Trier, puts it like
this: ‘Fields are living realities intermediate between individual words and
the totality of the vocabulary; as parts of a whole, they share with words the
property of being integrated in a large structure and with the vocabulary the
property of being structured in terms of smaller units’ (quoted in Lyons
1977: 253). For example, the lexical field of ‘colour terms’ includes the
lexemes: black, white, red, green, yellow, blue, orange, etc. Similarly, the
lexical field of colour terms, together with those of kinship terms, military
ranks, vehicles, among others, are only parts of the whole English vocabulary.
Furthermore, the general lexeme red, for instance, may in turn be considered
a lexical field (or subfield) within which the particular lexemes scarlet,
crimson, vermillion, etc. may best be interpreted. These characteristics of
field theory may be illustrated by Figures 1.3a and 1.3b.
As we would expect, the lexicologist who attempts to assign all the words
in English to lexical fields is bound to face a number of difficulties. According
to Crystal (2019: 167) these difficulties are of three kinds. First, some lexemes
tend to belong to fields that are vague or difficult to define. For instance, it is
not obvious to which field the lexemes noise or difficult should be assigned.
Secondly, some may validly be assigned to more than one field, e.g. orange,
either to the field of ‘fruit’ or to that of ‘colour’; tomato as ‘fruit’ or ‘vegetable’.
The last difficulty concerns the best way to define a lexical field in relation to
the other fields on the one hand, and its constituent lexemes on the other. To
use Crystal’s examples (2019: 167), one may ask such questions as the
used to form new words; -ly is still highly productive in deriving adverbs from
adjectives (stubbornly, speculatively). The third relates to the predictability of
the meaning of the affix; -ness is only used to form nouns from adjectives,
with the meaning ‘quality of ’ (craziness, tiredness), whereas -ist has a number
of possible meanings. The remaining criteria concern regularity of spelling
and pronunciation (of the base and affix) and regularity of the function of an
affix in terms of the word class of the base to which it attaches.
Using these criteria, Bauer and Nation (1993) establish seven levels of
family relationship. At the first level, each word form is regarded as a different
word; so, there is no family. The second level groups words that have a
common base but variant inflectional suffixes (plural and possessive for
nouns; present and past tense, and present and past participle for verbs;
comparative and superlative for adjectives), e.g.
At the third level are added words formed by the addition of ‘the most frequent
and regular derivational affixes’, which are established on the basis of an
analysis of a computer corpus (see Chapter 7), viz. -able, -er, -ish, -less, -ly, -ness,
-th, -y, non-, un-. At level four are added forms with ‘frequent, orthographically
regular affixes’: -al, -ation, -ess, -ful, -ism, -ist, -ity, -ize, -ment, -ous, in-. At level
five come forms derived with some fifty ‘regular but infrequent affixes’, e.g. -ary,
-let, anti-, sub-. The sixth level has forms derived by ‘frequent but irregular
affixes’: -able, -ee, -ic, -ify, -ion, -ist, -ition, -ive, -th, -y, pre-, re-. Lastly, at level
seven are included words formed using classical (Latin and Greek) roots and
affixes, e.g. bibliography, astronaut and the common prefixes ab-, ad-, com-, de-,
dis-, ex-, sub-. As an illustration, we borrow Bauer and Nation’s example of the
verb develop, which conveniently has examples at all of the levels two to six:
Exercise 1/4
For each of the following sets of words, say what the principle is
that groups them into a set. Is the principle one of a common
meaning, related forms of a lexeme, or something else (please
specify)?
The words on, in and under have the same function and express some kind
of locational relationship between the book and the cupboard. Since they
behave the same way in the sentence, they belong to the same word class,
which we call ‘preposition’. The notion of word class is therefore useful
because it allows us to make general and economical statements about the
way the words of a language behave. However, this notion is more useful in
syntax (1.2.2) than in lexicology.
Traditional lexicology has been concerned mostly with open classes, but
more recent descriptions, such as Jackson (1988) and Lipka (1990), attempt to
discuss all classes of words. A close examination of the major word classes
shows that they all have central and peripheral members and that they overlap.
The boundaries between classes are therefore fuzzy. Moreover, a word may
belong to more than one word class, e.g. round is an adjective in ‘a round stone’,
a preposition in ‘round the corner’, an adverb in ‘they all gathered round’, a noun
in ‘you can buy the next round’, and a verb in ‘when we round the next bend’.
Exercise 1/5
(a) Closed word classes have a restricted membership: list as
many prepositions as you can think of in modern English and
then compare your list with those in a grammar book.
(b) Attempt a definition of ‘adjective’. Think of the kinds of general
meaning adjectives have, their possible forms (inflections), and
where they typically occur in sentence structure.
1.4 Summary
This chapter has given us at least a partial answer to the question,
‘What is lexicology?’ In so doing, it has established the distinction
between lexicology as a level of language analysis and related
fields such as phonology, morphology, syntax, semantics,
etymology and lexicography. It has also examined the various
attempts made to account for the structure of English vocabulary
as a whole. Among other things, the chapter has revealed the
central position of the notion of ‘word’ in lexicology, which we
take up in Chapter 2.