70
UNIVERSITATEA TRASILVAIA DIN BRASOV FACULTATEA DE LITERE FORMA DE INVATAMANT: ID OPTIOAL COURSE I EGLISH SEMATICS (A compilation) Course coordinator: ELENA BUJA

Semantica Curs Ansamblu 2007

Embed Size (px)

DESCRIPTION

Semantica -curs facultate

Citation preview

Page 1: Semantica Curs Ansamblu 2007

UNIVERSITATEA TRA�SILVA�IA DIN BRASOV FACULTATEA DE LITERE FORMA DE INVATAMANT: ID

OPTIO�AL COURSE I�

E�GLISH SEMA�TICS (A compilation)

Course coordinator: ELENA BUJA

Page 2: Semantica Curs Ansamblu 2007

2

Contents Chapter 1: Introduction to semantics: a short history of semantics p. 3 Chapter 2: The problem of meaning p. 10 Chapter 3: Motivation of meaning p. 15 Chapter 4: Seven types of meaning p. 20 Chapter 5: The componential analysis of meaning p. 29 Chapter 6: The distributional analysis of meaning p. 39 Chapter 7: Ambiguity and vagueness p. 48 Exercises p. 57 Bibliography p. 65

Page 3: Semantica Curs Ansamblu 2007

3

Chapter 1. A short history of semantics 1.1. Why study semantics? Semantics (as the study of meaning) is central to the study of communication; and as

communication becomes more and more a crucial factor in social organization, the need to

understand it becomes more and more pressing.

Semantics is also at the centre of human mind – thought processes, cognition, conceptualization

– all these are strongly connected to the way in which we classify and convey our experience of

the world through language.

Semantics can be defined as a branch of linguistics; it is an area of study parallel to, and

interacting with syntax and phonology. While syntax and phonology study the structure of

expressive possibilities in language, semantics studies the meaning that can be expressed. Nearly

all linguists have accepted a linguistic model in which semantics is at one end and phonetics at

the other, with grammar somewhere in the middle. However, until recently, semantics has been

the ‘Cinderella’ of linguistics, a branch that had been abandoned to philosophers and

anthropologists. But in the past20 –25 years there has been a swing away from the view that

semantics is a messy, unstructured intellectual no-man’s-land on the fringes of linguistics, and

little by little it has acquired a central position in linguistic studies. The concentration on

semantics has come not only from linguists, but from logicians, too. Consequently, in semantics

we witness an unusual convergence of disciplines; the techniques and investigations of

philosophy and cognitive psychology, in particular, have helped to lay a more solid foundation

for linguistic studies.

1.2. A short history of semantics

Although semantics is consider a rather young branch of linguistics, interest in today’s problems

of semantics was alive already in ancient times.

Antiquity

In ancient Greece, philosophers dealt with the problem of the way in which words acquired their

meaning. One of the questions they tried to find an answer to was the following: Why is a thing

called by a given name? The answers provided made the Greek philosophers divide into two

‘parties’: on the one hand we have the adepts of the physei theory, and on the other hand the

adepts of the thesei theory. Let us now briefly present these two points of view.

Page 4: Semantica Curs Ansamblu 2007

4

a) The physei theory. Some philosophers considered that the names of things were arrived at

naturally, that they were somehow conditioned by the natural properties of the things themselves.

An example provided by them is that of the letter rho [ρ] which seems apt to express motion,

since the tongue moves rapidly in its production; hence, its occurrence in such words as rhoein

‘to flow’. Other sounds like [s], [f] and [ks], which require greater breath-effort in production,

seem suitable to appear in words like kseon ‘shaking’. Despite the inadvertences of such

correlations, the adepts of the physei theory kept on believing that it is the physical nature of

sounds in a name that can tell use something about its meaning.

b) The thesei theory. Some other philosophers held the opposite view, namely that names are

given to things arbitrarily through convention.

The physei – thesei controversy was discussed by various philosophers of the time, one of the

most representative one being Plato. He wrote a dialogue entitled Cratylus in which the two

discussants are Cratylus, the partisan of the physei theory, and Hermogenes, the defendant of the

thesei point of view. The two positions are debated by Socrates, who in an attempt to mediate

between the two discussants, points out an interesting fact, i.e. that there are two types of names:

simple names and compound names, which are divisible into smaller constituent elements and

analysable into the meaning of these constituents.

Two other dialogues by Plato, Theatetus and Sophists mark an important step in the

development of semantics. In them, he dealt with problems such as the relation between

THOUGHT, LANGUAGE and the OUTSIDE WORLD. Language is defined as the expression

of one’s thought be means of onomata (the name of the performer) and rhemata (the name

defining the action).

THOUGHT

LANGUAGE OUTSIDE WORLD

onomata rhemata (performer) (action) LOGOS

Page 5: Semantica Curs Ansamblu 2007

5

Identifying onoma and rhema as the constituents of LOGOS, Plato opened the way for analysing

the sentence in terms which are partly linguistic and partly pertaining to logic. He was dealing

therefore with the meaning of utterances rather than the meaning of individual words.

Another philosopher of Antiquity who had a contribution to the birth of semantics was Aristotle.

His works (Organon, Rhetorics, Poetics) mark a major contribution to language study in general,

and to semantics, in particular. He approached language from the point of view of a logician and

was interested in the following issues:

- What is there to know about the world?

- How men know it?

- How they express this knowledge in language.

He also identified the lexical level of language analysis the aim of which was to study the

meaning of words either in isolation or in syntactic constructions. This marks his own

contribution to semantics.

The Middle Ages

During the Middle Ages an important contribution to linguistics and semantics was brought by a

group of philosophers called the Modistae because of their writings entitled On the Modes of

Signification. These writings were some kind of speculative grammars in which semantics

considerations held an important position. The Modistae adopted the thesei point of view of the

ancient philosophers and their efforts were directed towards pointing out the ways in which we

can know things and the various ways of signifying things.

Throughout Antiquity and the Middle Ages, and almost until the 19th century, almost everything

that came to be known about meaning in language was the result of philosophic speculation and

logical reasoning. Philosophy and logic were the two important sciences which left their strong

impact on the study of linguistic meaning.

The 19th

century

In the 19th century, semantics became an independent branch of linguistics. The fist works which

dealt with the study of semantic problems as we understand them today date as far as this century.

Ch. C. Reisig, a German linguist, was the first to formulate the object of study of the new science

of meaning, which he called semasiology = a historical science studying the principles governing

the evolution of meaning. Towards the end of the century, more exactly in 1877, M. Brèal, a

Page 6: Semantica Curs Ansamblu 2007

6

French linguist, published an important book Essay de sémantique, which in many ways marks

the birth date of semantics as a modern linguistic discipline. Brèal did not only provide the name

of the new science, which became general in use, but also defined more clearly its subject matter.

The theoretical sources of semantics as outlined by Brèal are:

-classical logic

- rhetorics, and

- psychology.

In following the various changes in the meaning of words, interest is focussed on identifying

certain general laws governing these changes. Some of these laws are arrived at by recourse to

the categories of logic:

-extension of meaning: eg. persona (Lat.) = face mask worn by actors → characters in a play →

a man, somebody →a person (feminine or masculine).

- narrowing of meaning: e.g. the River may be used by a Londoner to refer to the Thames, or by

a French person to refer to the Seine.

- transfer of meaning (metaphor): e.g. the camel = the ship of the desert.

Other changes are due to a psychological approach:

-degradation of meaning: e.g. silly (Old English) = happy, poor, innocent → helpless→ stupid.

-elevation of meaning: e.g. minister = a servant→ an important public official.

In the 19th century, the study of meaning was considerably enhanced by the writing of

dictionaries. Lexicography played a significant role in the development of semantics. The

dictionaries represent collections of a huge volume of data concerning the meaning of words and

the changes in their meaning throughout the history of languages. Some dictionaries resorted to

the alphabetical order of listing lexical items. Others, however, tried to arrange words according

to some more natural ties existing among them. It is in this manner that ‘notional’ or

‘conceptual’ dictionaries appeared, having a tremendous importance for the progress of semantic

studies.

The 20th

century

Within the process of development of the young linguistic discipline, the 1921 – 1931 decade

has a particular significance. This period was marked by the publication of three important books.

In 1931, Jost Trier published his ‘Der Deutsche Wortschatz im Sinnbezirk des Verstandes’ (The

German vocabulary related to the semantic field of understanding). Analysing the meaning of a

set of lexical elements related to one another by their content, and thus belonging to a semantic

Page 7: Semantica Curs Ansamblu 2007

7

field, Trier reached the conclusion that they were structurally organised within this field in such

a manner, that the significative value of each element was determined by the position which it

occupied within the respective field. It was for the first time that words were no longer

approached in isolation, but analysed in terms of their position within a larger ensemble – the

semantic field – which, in turn, is integrated together with other fields into an even larger one –

the lexicon. Unfortunately, at that time Trier’s valuable ideas did not enjoy broad circulation.

The second important publication of the decade mentioned above was that of Gustav Stern,

‘Meaning and Change of Meaning’. This was an ambitious attempt at examining the component

factors of meaning and of determining the causes and directions of changes of meaning. Stern

postulated several classifications and principles which are essential in answering one of the most

important problems of semantics, namely: WHY and HOW does change of meaning occur in

linguistic forms?

The third important book, published in 1923, was co-authored by C.K. Ogden and J.A. Richards

and was entitled ‘The Meaning of Meaning’ This was a book which dealt with different accepted

definitions of the word meaning, not only in linguistics, but in other disciplines, as well.

In the following decades a period of crisis in semantics appeared. Meaning was completely

ignored in linguistics, particularly due to the position adopted by Leonard Bloomfield1, who

considered that the study of meaning was outside the scope of (semantics) linguistics proper. He

believed that the study of meaning should be the object other sciences, such as philosophy,

psychology, sociology and anthropology. Reference to semantics was only made in extremis,

when the various linguistic theories were not able to integrate the complexity of linguistic events

within a unitary system. Thus, semantics was considered some kind of ‘dumping place’, a vast

container in which all language facts that were difficult to formulate could be disposed of.

In the last years of the ‘60s, semantics was given again due attention. Nowadays, the various

linguistic theories admit that no language description can be regarded as being complete without

including facts of meaning in its analysis. A specific feature of modern research in linguistics is

the ever growing interest in the problem of meaning.

1 Leonard Bloomfield (1887-1949), American linguist who had a strong formative influence on structural linguistics.

Page 8: Semantica Curs Ansamblu 2007

8

1.3. Definition and object of semantics

In linguistic terminology, the word semantics is used to designate the science of word-meaning.

However, the term has acquired a number of senses in contemporary science and, at the same

time, it has sometimes been replaced by other terms that cover the same area of study (i.e. the

study of meaning), such as semasiology and semiotics. But, there is a clear distinction between

semantics and the other two terms and branches of linguistics.

The term semasiology (Introduced by Ch. Reisig) should be kept for a more restricted usage, in

opposition to onomasiology. Semasiology stands for the study of meaning starting from the

signifier (i.e. the acoustic/graphic image) of a linguistic sign and examining the possible

signified attached to it. Onomasiology, on the other hand, accounts for the opposite direction of

study, i.e. from the signified to the various signifiers that may stand for it. Therefore we need to

linger for a while on the terminology used in the study of meaning and to point out the main

concern in the science devoted to the study of meaning.

On particular meaning of the term semantics is used to designate a new science, general

semantics, which is the psychological and pedagogical doctrine founded by Alfred Korzybsky

(1933), aiming at the general improvement of human beings by better training them in the use of

words. General semantics points out that there is a dialectic interdependence between language

and thought in the sense that language does not merely serve to express thought, but takes an

active part in the moulding of thought.

In the more general science of SEMIOTICS (the study of signs), the term semantics is used in 2

senses:

a) theoretical (pure) semantics, whish aims at formulating an abstract theory of meaning in the

process of cognition and therefore belongs to logic;

b) empirical (linguistic) semantics, which studies meaning in natural languages, i.e. the

relationship between linguistic signs and their meanings.

Of the two types of semantics, it is the latter that falls within the scope of linguistics. However,

even within the narrower field of linguistics, the term is far from having a unique and commonly

agreed upon usage. Differences in usage refer mostly to the definition and object of study.

The most commonly agreed upon definition of semantics remains the one given by M. Brèal as

the science of the meanings of words and of the changes in their meaning. With this definition,

Page 9: Semantica Curs Ansamblu 2007

9

semantics is included under lexicology, the more general science of words, being its most

important branch. In this capacity, as a sub-branch of lexicology, semantics is concerned with:

- the identification, definition and evolution of the meaning of words;

- the uncovering of the multiple relations established among words;

- the possibility of analysing the meaning of words into component elements of meaning,

which are shared by a set of words in various combinations, characteristic of each item;

- the analysis of those lexical items which are larger than just one word (i.e. compounds) into

meanings which are not simply the sum of the meanings of the component words.

With the advent of generative, grammar, emphasis has been switched from the meaning of

words to the meaning of sentences. Semantic analysis will be required:

- to explain how sentences are understood by the speakers of a language;

- to explain the relation existing among sentences, namely:

• why certain sentences are anomalous, though grammatically correct: e.g. Colourless

green ideas sleep furiously.

• why other sentences are semantically ambiguous, since they admit several

interpretations: e.g. Flying planes can be dangerous.

•other sentences are synonymous or paraphrases of each other: e.g. She would water her

plants every day/She used to water her plants every day.

Still, much of the information required to provide an answer to these problems is carried by the

lexical items themselves.

Page 10: Semantica Curs Ansamblu 2007

10

Chapter 2. The problem of meaning (explanations of the word ‘meaning’)

One classical problem of semantics is that of what we mean when we refer to the meaning a

word or a sentence has. In order to answer this question we have to give a coherent account of

meaning.

There are 3 main ways in which linguists and philosophers have attempted to construct

explanations of meaning in natural languages:

a) by defining the nature of word meaning;

b) by defining the nature of sentence meaning;

c) by explaining the process of communication.

In the first case, lexical/word meaning is taken as the construct (i.e. something constructed by

mental synthesis) in terms of which sentence meaning and communication can be explained.

In the second situation, it is sentence meaning which is taken as basic, with words characterised

in terms of the systematic contribution they make to sentence meaning.

In the third, both sentence and word meaning are explained in terms of the ways in which

sentences and words are used in the act of communication.

The existence of these 3 types has an explanation. In the first place, there is a relation between

words (on the one hand) and objects and actions (on the other hand), and one of the tasks of

semantics is to explain this relation.

Similarly, sentences are used to describe events, beliefs, opinions, and again it is the task of

semantics to explain the nature of the relation between sentences and the states of affairs those

sentences describe.

Finally, since language is the vehicle by means of which we communicate, it is arguable that the

interpretation of language should be explained in terms of its role in communication.

Moreover, these 3 aspects of meaning, i.e. word meaning, sentence meaning and communication

are reflected in different uses of the word MEAN.

Corresponding to a) word meaning: SPINSTER means ‘unmarried woman’;

Page 11: Semantica Curs Ansamblu 2007

11

Corresponding to b) sentence meaning: JAMES MURDERED BILL means that ‘someone

called James deliberately killed someone called Bill.’

In these 2 uses, the word MEAN has a meaning approximating the verb ‘to indicate/to show’.

But the word MEAN is used in a different sense in the following conversation between speakers

A and B, a sense which corresponds to explanation c) what he means to say:

A: Are you going to bed soon?

B: What do you mean?

A: I mean that I’m tired, and the sooner you go to bed, the sooner I can.

Consider the following example:

I mean to be here tomorrow (MEAN= ‘to intend’)

In the last two examples, MEAN is attributed to speakers and has the same meaning as the

expression ‘to intend’.

Thus we have at least 3 possible starting points from which to construct an explanation of

meaning + the signification of words, the interpretation of sentences, or what a speaker is

intending to convey in the act of communication. Most traditional explanations of meaning

constitute an attempt at explaining meaning in terms of the naming relation that holds between a

word and its object.

2.1. Definitions of meaning

The two best-known theories of meaning are:

A) the ‘sign’ theory of F. de Saussure (1922);

B) the ‘semiotic triangle’ of Odgen and Richards (1931).

A) The simplest and most obvious concept of meaning is to regard it as a bipolar relation

between two independent sides of a linguistic sign, i.e. between EXPRESSION (Saussure’s

signifier, more strictly a sound image) and CONTENT (Saussure’s signified or concept).

According to Saussure, these two elements of the LS are linked by a psychological ‘associative’

bond. Both the noises we make and the objects in the world that we talk about are mirrored in

some way by the conceptual entities.

This, in fact, can be said of any sign, irrespective of the semiotic system to which it belongs. One

would be tempted to equate meaning with content.

Page 12: Semantica Curs Ansamblu 2007

12

B) Ogden and Richards (1931) saw the relationship as a triangle, pointing out that there are at

least 3 factors involved in any symbolic act:

- the symbol itself, in our case the material aspect – phonic or graphic – of the LS;

- the thought or reference , which stands for the mental content that accompanies the

occurrence of the symbol in the minds of both speaker and listener;

- the object itself (the referent) designated by the symbol.

THOUGHT (‘reference’) SYMBOL (‘word’) (‘thing’) REFERENT

Figure 1. Ogden and Richards’ ‘triangle of signification’

While the relations symbol – reference and reference – referent are direct & causal ones, in that

the symbol expresses or symbolises the reference, which, in turn, refers to the referent, the

relation symbol – referent is an imputed, indirect one.

Of the three sides of the triangle, only the right-hand side can be left out in a linguistic account

of meaning (of interest to psychologists and philosophers). Linguists have directed their attention

towards the other two sides. Hence two basic types of meaning definitions: REFERENTIAL

definitions, which analyse meaning in terms of the relationship symbol-object, and

CONCEPTUAL definitions, that regard the relation, and CONCEPTUAL definitions, that regard

the relation symbol – thought.

REFERENTIAL definitions of meaning TENET: meaning is taken to be outside the word itself; meaning = an extralinguistic entity.

Shortcomings:

1) these definitions reduce the LS to its material (phonic/graphic) aspect alone. This would mean

leaving meaning to a large extent undefined, for it is debatable to what extent the characteristic

traits of an object as an extralinguistic reality, are identical with the distinctive features of lexical

meaning as a linguistic reality. (e.g. woman [+ A, +H, +AD, -Male]; but woman [ ]+/- long hair,

big mouth, etc.

Page 13: Semantica Curs Ansamblu 2007

13

2) not all words have a referent in the outside world: e.g. ghost, unicorn, elf. The reverse is also

true, namely that there are words with an identical referent, but with clearly different meanings:

the Morning Star and the Evening Star (common referent; planet Venus).

CONCEPTUAL definitions of meaning - meaning defined in terms of the other possible relation in the basic triangle – the relation

symbol – thought. In other words, it is proposed to define meaning in terms of the notion, the

concept, or the mental image of the object/situation in reality, as reflected in man’s mind, which

can be transferred from the mind of the speaker to the mind of the hearer by embodying them in

the form of one language or another.

The identification of meaning with concept will not help us to answer the Q: What is meaning?,

unless the term concept is clearly defined. As it is commonly employed, the term is too vague, or

too general to be taken as a foundation stone in the conceptual theory of meaning. What is there

common among the concepts associated with the following words: the, for, I, first, year, little,

school, boy, development, name? In some cases, we might reasonably say that the associated

concept is a visual image of some kind. But we cannot surely maintain this view with respect to

words like the, for or name. Even for the cases for which it is plausible to think of concepts as

visual images, this creates more problems than it solves. Take, for example the word boy. I can

visualise a boy in my mind’s eye, but I do not do so every time I utter the word boy. More

reasonably would be to say that I relate my utterance of the word boy to some more abstract

concept. But this would not be of much help – what is this abstract concept, what is his age,

colour of eyes/hair, shape of face, height, etc?

Mental images associated with a certain word by different people, are variable and full of detail.

Very often there is little or nothing that is common to these detailed and very personal images.

And yet, we still wish to say that, in general, people use words with more or less the same

meaning. There is no evidence to suggest that the visual images that we can call up, voluntarily

or involuntarily, in association with particular words are an essential part of the meaning of those

words.

Semantic studies have used mainly such conceptual definitions of meaning, taking it for granted

that for a correct understanding of meaning it is necessary to relate it to that reflection in our

mind of the general characteristics of objects and phenomena. Any study of meaning which takes

into account the close relationship between language and thought cannot ignore this aspect of

Page 14: Semantica Curs Ansamblu 2007

14

language meaning. On the other hand, complete identification of meaning with concept or notion

is not possible either. This would mean depriving meaning of any objective foundation.

Furthermore, languages provide whole categories of words (proper names for one, but also

prepositions, conjunctions, etc.). It has been argued that even in the case of notional words, the

notion (concept) may be regarded as being both wider and narrower than meaning. The former

has a universal character, the latter a specific character. The meaning of a word can be defined

only within a given language.

In practice we all know that a word has to have a meaning. Knowing the meaning of a word

means that we can do a number of things: we can use it properly and we can explain it to others

in terms of paraphrase or synonyms. But it does not follow from this that there is an entity that is

MEANING, or a whole group of entities that ARE the meaning of words.

From all this we could conclude that the problem of semantics is not to search for an elusive

entity called MEANING, but rather an attempt to understand how it is that words and sentences

can ‘mean’ at all, or how they can be meaningful.

Page 15: Semantica Curs Ansamblu 2007

15

Chapter 3: Motivation of meaning

The physei – thesei controversy of the antique philosophers concerning the way in which words

acquire their names has been settled in favour of the physei theory by modern linguists.

Ferdinand de Saussure’s statement that the linguistic sign is arbitrary, in the sense that there is

no direct relationship between the sound sequence (the signifier) and the idea expressed by it (the

signified) was taken for granted in the study of language. However, the discussion on the

arbitrary character of the linguistic sign in the late thirties and early forties proved that the

problem is not as simple as it might seem. To begin with, as Benveniste (1940) pointed out, to a

certain extent ‘the relationship between the signifier and the signified is not arbitrary, on the

contrary, it is a necessary one.’

What Benveniste had in mind was the fact that there must be a necessary signified attached to a

signifier in order for the linguistic sign (LS) to discharge its function. But in addition to this

obviously correct statement, there are numerous words in all languages in which a special

correlation may be said to exist between meaning and sound. These words include in the first

place INTERJECTIONS and ONOMATOPOEIA, which are somehow imitative of non-

linguistic sounds, as well as those instances in which it can be said that some sounds are

somehow associated with certain meanings, in the sense that they suggest them. This latter

aspect is known as phonetic symbolism.

In addition to these cases, which are marginal in the language, there is also another instance in

which the meaning of words may be said to be related to their forms, i.e. the possibility of

analysing the LSs by reference to the smaller meaningful elements of which they are made up.

Indeed, complex (derivative and compound) words can be analysed from the point of view of

meaning of their constituent morphemes.

It is obvious that while the general principle remains valid, i.e. that there is no inherent reason

why a given concept should be paired to a given string of sounds, it is the linguist’s task to

examine those instances when it is possible to sat something about the meaning of the LS by

reference to its sound and grammatical structure, in other words it is necessary to assess the

extent to which there is some motivation in the case of at least a number of words in the

language.

Page 16: Semantica Curs Ansamblu 2007

16

Motivation is viewed as a manifestation of the theoretical category of condition. Linguistic signs

are said to be more or less motivated to the extent to which their inner organisation is not

altogether accidental.

There are two main types of linguistic motivation postulated by F. de Saussure: absolute and

relative motivation.

1. Absolute motivation includes linguistic signs whose sound structure reproduces certain

features of their content. There is a quasi-physical resemblance between their signifier and the

signified .There are several classes of linguistic signs which can be said to be absolutely

motivated:

a) Interjections. It would be wrong to consider that interjections somehow depict exactly the

physiological and psychological states they express. The fact that interjections differ in sound

from one language to another is the best proof of it. Compare, for instance, the Romanian

interjections au!, aoleu!, vai! and the English ouch!, which may be used in similar situations.

b) Onomatopoeia. The same holds true for onomatopoeic words. Despite the relative similarity in

the basic phonetic substance of words meant to imitate the sounds of animals or other sounds and

noises, their phonological structure follows the rules of patterns and arrangement characteristic

of each separate language. Compare the following examples:

Romanian English zbirr whizz zdrang twang poc stud pleosc plop ham-ham bow-wow

An even more illustrative example, picked from a French magazine – EĆHOS No.67/1992, is

represented by the sound produced by the cock:

French: cocorico, English: cock-a-doodle-doo, German: kikeriki, Dutch: kukeluku, Italian:

chicchirichi, Spanish: quiquiriqui, Portuguese: cocoroco, Romanian: cucurigu, Greek:

koukoupikou, Finish: kukkokkiekuu, Sweedish/Norwegian: kuckeliku, Danish: kykliky, Polish:

kukuryky, Russian: kykapeky.

c) Phonetic symbolism. This is based on the assumption that certain sounds may be associated

with particular ideas or meanings, because they somehow seem to share some attributes usually

associated with the respective referents. The problem of phonetic symbolism has been debated in

Page 17: Semantica Curs Ansamblu 2007

17

linguistics and psychology, and numerous experiments have been made without reaching very

conclusive results.

Otto Jesperson (1920) paid particular attention to the phonetic motivation of words and tried to

give the character of law to certain sound and meaning associations. On the basis of ample

evidence provided by a great variety of languages, he maintained that the front, close vowel [i]

suggests the idea of smallness, rapidity and weakness.

English: little, slim, kid, bit, flip, tip, pinch, flicker, twinkle German: bisschen French: petit Italian: piccolo Romanian: mic, pic Of course, one can find counter examples – e.g. big – but on the whole, it doesn’t seem

unreasonable to argue that a given sound, or sequence of sounds is associated to a given meaning.

Eduard Sapir (1929) maintained that a contrast can be established between [I] and [a] in point of

the size of referents in the names of which they appear, so that words containing [a] usually have

referents of larger size, e.g. Rom. mic vs. mare, Engl. little vs. large.

A similar finding was reported by Newman (1933) who found a systematic relationship between

the vowels arranged along the close-open axis and the scale of sizes arranged from small to large,

in the sense that larger sizes are associated with larger oral cavity characteristic of open vowels.

At the same time he pointed out that very often the initial cluster of consonants gives an

indication of meaning of a rather special kind. Thus, words beginning with:

- sl- are ‘slippery’ in some way: slide, slip, slush, sluice, sludge; or are merely pejorative:

slut, slang, sly, sloppy;

- sk- refer to surface or superficiality: skate, skim, skid, skin

- sn-/fl- suggest rapid movement: snap, flicker.

On the other hand, almost all the words ending in –ump refer to some kind of roundish mass:

plump, chump, rump, hump, stump.

But one should not generalise too far. Not every word with these phonological characteristics

will have the meaning suggested.

Page 18: Semantica Curs Ansamblu 2007

18

2. Relative motivation. In the case of the relatively motivated LSs, it is not the sounds, which

somehow evoke the meaning; whatever can be guessed about the meaning of such words is a

result of the analysis of the smaller linguistic signs, which are included in them. Relative

motivation involves a much larger number of words in the language than absolute motivation.

There are 3 types of relative motivation:

a) Motivation by derivation. An analysis of the use of derivational means to create the new

words in the language will reveal its importance for the vocabulary of a language. The prefix

{in-}, phonologically realised in various ways (i.e. im-, ir-, il-) and meaning either not or in/into,

appears in at least 2,000 English words: inside, impossible, enclose, embrace, irregular, etc.

Similarly, the Latin capere (to take) appears in a great number of English words: capture,

captivity, capsule, caption, reception, except, principal, etc.

Brown (1964) tried to give keys to the meanings of over 14,000 words which can be analysed in

terms of combinations between 20 prefixes and 14 roots.

Words Prefix Meaning Root Meaning detain de- ‘away, down’ tenere ‘hold, have’ intermittent inter- ‘between, among’ mittere ‘send’ offer ob- ‘against’ ferre ‘bear, carry’ mistranscribe mis- ‘wrong’ scribere ‘write’ trans ‘across, beyond

These examples alone are sufficient to indicate the importance of relative motivation for the

analysis of meaning. It could be wrong to assume that the derivational form of words does not

suggest to a speaker of English anything about its meaning.

b) Motivation by compounding. In the case of compounds, there is a problem with what St.

Ullmann (1962) called transparent and opaque words.

Transparent words are those words whose meanings can be determined from the meaning of

their parts (e.g. chopper, doorman, greenhouse or blackboard), while opaque words are those for

which this is impossible, i.e. one cannot guess their meanings (e.g. blackleg, legman).

The comparison with other languages, German in particular, is interesting:

English: thimble glove linguistics → opaque

German: Fingerhut Handschuh Sprachwissenchaft →transparent

Page 19: Semantica Curs Ansamblu 2007

19

This suggests not only that one word may be seen as consisting of several bits of meaning, but

also that the number of bits is arbitrary.

c) Derived compounds represent a special class of words created on the basis of two word-

formation rules, i.e. composition and derivation.

e.g. heavy smoker is not a smoker who is heavy

artificial florist amusing examples; they are ambiguous.

criminal lawyer

It is obvious therefore that the lexicon of a language presents items which differ in the degree in

which their meaning can be said to be motivated. While some are opaque, i.e. their sounds give

no indication of their meaning, others are more or less transparent, in that one can arrive at some

idea of their meaning by recourse to their phonetic shape or to their derivational structure or to

some semantic relations that can be established with other words in the language.

Synchronically, any language contains both words with a motivated sound structure and words

with unmotivated sound structure. Each of these types has advantages and disadvantages. In

the case of unmotivated words the advantage lies in the fact that they are not redundant and their

meanings are more clearly revealed and grasped identically bay all speakers. On the other hand,

these words are more difficult to memorize. If there were only such words in a language, it

would mean having a large number of basic words (roots). Such relations as polysemy and

homonymy would not exist.

In the case of motivated (analysable) words, the advantage lies in that they are easy to memorise,

the meaning is found out without any difficulty; on the basis of a small number of words we can

easily form others. The disadvantage is sometimes determined by the length of the sound

structure, by the faulty grasping of the meaning because of some false morphological analyses.

We also have to take into account the danger of false etymologies (sensible vs. sensitive).

Page 20: Semantica Curs Ansamblu 2007

20

Chapter 4: Seven types of meaning

We have seen that some linguists would like semantics to pursue the study of meaning in a wide

sense of ‘all that is communicated by language’; others limit it in practice to the study of logical

or conceptual meaning. But, by carefully distinguishing types of meaning, we can show how

they all fit in the total composite effect of linguistic communication, and to show how methods

of study appropriate to one type may not be appropriate to another.

Geoffrey Leech (1990) breaks down meaning in its widest sense into seven different ingredients,

giving primary importance to logical meaning or conceptual meaning. The other types are:

- connotative meaning; - social meaning; - affective meaning; - reflected meaning; - collocative meaning and - thematic meaning.

A) CONCEPTUAL MEANING C.M. (sometimes called ‘denotative’ or ‘cognitive’ meaning) is widely assumed to be the central

factor in linguistic communication, and it can be shown to be essential to the functioning of

language in a way that other types of meaning are not. Still, we cannot say that CM is the most

important element of every act of linguistic communication. The chief reason why Leech assigns

priority to CM is that it has a complex organisation of a kind which may be compared with

similar organisation on the syntactic and phonological levels of language. In his opinion there are

two structural principles that seem to lie at the basis of all linguistic patterning:

-the principle of CONTRASTIVENESS;

-the principle of STRUCTURE.

Contrastive features underlie the classification of sounds in phonology, for example, in that any

label we apply to a sound defines it positively, by what features it possesses, but also negatively,

by what features it does not possess. Thus, the phonetic symbol /b/ can be described as (+bilabial,

+voiced, +plosive, - nasal). We assume that the distinctive sounds or phonemes of a language are

identifiable in terms of binary contrast. In a similar way, the conceptual meanings of a language

can be studied in terms of contrastive features, so that, for example, the meaning of the word

woman could be specified as [+H, -Male, +Adult], as distinct from, say boy, which could be

defined as [+H, +M, - Ad].

Page 21: Semantica Curs Ansamblu 2007

21

The principle of structure is the principle by which larger linguistic units are built up of smaller

units. Looking at the issue from the opposite point of view, we may say that it is the principle by

which we are able to analyse a sentence syntactically into the constituent parts, moving from its

immediate constituents through a hierarchy of sub-divisions, to its ultimate constituents. This

aspect of organisation of language is often given visual display in a tree-diagram,

S

Subject Predicate

Det Noun Vb Complement

Det. Noun

no man is an island

or it can be represented by bracketing: {(No) (man)} { [(is)] [(an) (island)]}.

It has long been taken for granted that the syntax of a language is to be handled in such terms,

but it is now widely accepted that the semantics of natural languages has its own counterpart of

syntactic structure.

The two principles of contrastiveness and constituent structure represent the way language is

organised on what linguists have termed the PARADIGMATIC (selectional) and the

SYNTAGMATIC (combinatory) axes of linguistic structure. Leech mentions a third generally

acknowledged principle of linguistic organisation, according to which a piece of language is

structured simultaneously on more than one level. At least the following three levels seem to be

necessary for a full account of the linguistic competence by which we are able to generate and

understand utterances:

Listener

PHONOLOGY (A)

encodes SYNTAX (B) decodes

SEMANTICS (C)

Speaker

This means that for the analysis of any sentence, we need to establish a phonological, a syntactic

and a semantic representation, as well as the stages by which one level of representation can be

derived from another. The aim of conceptual semantics is to provide, for any given interpretation

Page 22: Semantica Curs Ansamblu 2007

22

of a sentence, a configuration of abstract symbols, which is its semantic representation, and

which shows exactly what we need to know if we are to distinguish that particular meaning from

all other possible sentence meanings in the language, and to match that meaning with the right

syntactic and phonological expression.

From this account it follows that conceptual meaning is an essential part of what language is,

such that one can scarcely define language without referring to it. A language which

communicated by other means than by conceptual meaning (e.g. a language which

communicates solely by means of expletive words like Oh!, Ah!, Oho!, Alas!) would not be a

language at all in the sense in which we apply that term to the tongues of men.

B) CONNOTATIVE MEANING

More of what is distinctive about conceptual meaning will appear when we contrast it with

connotative meaning. Connotative meaning is the communicative value an expression has by

virtue of what it refers to, over and above its purely conceptual content. To a large extent, the

notion of ‘reference’ overlaps with ‘conceptual meaning’. If the word woman is defined

conceptually by 3 features [+Human, -Male, +Adult], then the 3 properties must provide a

criterion for the correct use of the word. These contrastive features, translated into ‘real world’

terms, become attributes of the referent. But there is a multitude of additional, non-criterial

properties that we expect a referent of ‘woman’ to posses. They include not only physical

characteristics (biped, having a womb and breasts), but also psychological and social properties

(gregarious, subject to maternal instinct), and may extend to features, which are merely typical of

womanhood (capable of speech, experienced in cookery, skirt/dress wearing). Moreover,

connotative meaning can embrace the ‘putative properties’ of the referent, due to the viewpoint

adopted by an individual, or a group of people. So, in the past, ‘woman’ has been given such

attributes as [frail, prone to tears, emotional, inconstant], as the dominant male has been pleased

to impose on her.

Connotations vary from age to age, from society to society (see the status of women in Iraq vs.

America), from individual to individual within the same speech community.

In talking about connotation, we are actually speaking about the ‘real world’ experience one

associates with an expression when he uses or hears it. Connotation is somehow incidental to

language rather than an essential part of it.

Page 23: Semantica Curs Ansamblu 2007

23

Connotations refer to the associations that words have for us. Psychologists have long been

aware that in addition to ‘naming things’. Words carry overtones of meaning which colour our

reactions to them. Words can conjure up associations that may affect our attitude and our

response to an utterance which contains them.

In an early word-association experiment carried out by Kent and Rosenoff in 1910, 1000 people

were given the stimulus word chair and were asked to write the words which first came to their

minds in connection with it. Nearly 60% of the sample answered with table, seat, furniture, sit or

sitting. 107 of the subjects gave as their first response words which seem to indicate some kind

of evaluatory association: comfort, convenience, rest, idle, pleasure.

If people react to lexical items in this way, with value judgements, sensations of like and dislike

and so on, then it is important to know both in what ways they respond to words and the degree

to which it may influence them in their reaction to speakers who use them. A much more direct

way of getting such connotative or associative meaning is to ask speakers to rate words on scales

such as good/bad, pleasant-unpleasant, strong-weak, fast-slow, etc. Even where such evaluations

appear bizarre or inappropriate, subjects manage to perform this task remarkably well, and find it

possible to indicate how ‘rough’, ‘tasty’, or ‘hot’ they perceive a word such as sin. (Graddol et

all, 1994:103-105) introdus in 11.11.2009

♠Connotative meaning is not specific to language, but is shared by other communicative

systems, such as visual art and music. Whatever connotations the word baby has can be invoked

by a drawing/photo of a baby, or even by an imitation of a baby’s cry. The overlap between

linguistic and visual connotations is particularly obvious in advertising, where words are often

the lesser partners of illustrations in the task of conferring the product a halo of favourable

associations.

♠ A second fact, which shows that connotative meaning is peripheral, as compared to conceptual

meaning, is that connotations are relatively unstable, i.e. they vary considerably according to

culture, historic period, and the experience of the individual.

♠ Thirdly, connotative meaning is indeterminate and open-ended in a sense in which conceptual

meaning is not. Connotative meaning is open-ended in the same way as our knowledge and

beliefs about the universe are open-ended; any characteristic of the referent, identified

subjectively or objectively, may contribute to the connotative meaning of the expression which

denotes it. In contrast, it is generally taken as fundamental to semantic theory that the conceptual

Page 24: Semantica Curs Ansamblu 2007

24

meaning of a word or sentence can be codified in terms of a limited set of symbols, and that the

semantic representation of a sentence can be specified by means of a finite number of rules.

C) SOCIAL MEANING

It is a meaning that a piece of language conveys about the social circumstances of its use. In part

we decode the social meaning of a text through our recognition of different dimensions and

levels of style within the same language. We recognise some words or pronunciations as being

dialectal, i.e. as telling something of the geographical or social origin of the speaker.

The following dimensions of socio-stylistic variation have been recognised:

DIALECT (the language of a geographical region or of a social class)

TIME (the language of the 18th century vs. the lg. of the 20th century)

PROVINCE/REGISTER (language of a particular field of activity: law, science, advertising)

STATUS (polite, colloquial, slang)

MODALITY (language of lectures, jokes)

SINGULARITY (the style of Dickens, Hemingway, Agatha Christie).

The list indicates something of the range of style differentiation possible within a single

language:

e.g. steed (poetic), horse (general, neutral), nag (slang), gee-gee (baby language).

The style dimension of ‘status’ is particularly important in distinguishing synonymous

expressions:

e.g. a) They chucked a stone at the cops, and then did a bunk with the loot – could be

uttered by criminals

b) They cast a stone at the police, and then absconded with the money – might be

said by the chief inspector in making the official report. Both could be describing the same

happening, and their common ground of conceptual meaning is evident.

In a more local sense, social meaning can include what has been called the ILLOCUTIONARY

FORCE of an utterance: for example, whether it is to be interpreted as a request, an assertion, an

apology, a threat, etc. The function an utterance performs in this respect may only be indirectly

related to its conceptual meaning.

e.g. I haven’t got a knife - has the form and meaning of an assertion, and yet in social

reality (if said to a waiter in a restaurant) it can readily take on the force of a request such as

Please bring me a knife.

Page 25: Semantica Curs Ansamblu 2007

25

D) AFFECTIVE MEANING reflects the personal feelings of the speaker, including his attitude

to the listener or his attitude to something he is talking about. Affective meaning is often

explicitly conveyed through the conceptual or connotative content of the words used. Someone

who is addressed:

- You are a vicious tyrant and I hate you for that

is left in little doubt as to the feeling of the speaker towards him. But there are less direct ways of

disclosing our attitude than this, for example, by scaling our remarks according to politeness.

With the object of getting people to be quiet, we might say the following:

- I’m terribly sorry to interrupt you, but I wonder if you would be so kind as to lower your

voice a little.

- Will you belt up!

Factors such as intonation and voice timbre are also important here. The impression of politeness

in the first sentence can be reversed by a tone of biting sarcasm; the second sentence can be

turned into a playful remark between intimates if delivered with the intonation of a mild request.

Affective meaning is largely a parasitic category in the sense that, to express our emotions, we

rely upon the mediation of other categories of meaning, i.e. conceptual, connotative or stylistic.

On the other hand, there are elements of language (chiefly interjections like Aha!, Yippee!)

whose chief function is to express emotion. When we use these, we communicate feelings and

attitudes without the mediation of any other kind of semantic function.

E) REFLECTED MEANING & COLLOCATIVE MEANING

Two further, though less important types of meaning involve an interconnection on the lexical

level of language.

Reflected meaning is the meaning that arises in cases of words with multiple conceptual

meanings, when one sense of a word forms part of our response to another sense. On hearing in a

church service the synonymous expressions The Comforter and The Holy Ghost, both referring

to the Third Person of the Trinity, the reaction of some people may be conditioned by the every-

day non-religious meanings of ‘comfort’ and ‘ghost’. The Comforter sounds warm and

comforting (although in the religious context it means ‘the strengthener or supporter’), while the

Holy Ghost sounds awesome.

Page 26: Semantica Curs Ansamblu 2007

26

One sense of a word seems to rub off another sense in this way, only when it has a dominant

suggestive power, either through relative frequency and familiarity (as in the case of The Holy

Ghost), or through the strength of the association.

The cases where reflected meaning intrudes through the sheer strength of emotive suggestion are

most strikingly illustrated by words which have a taboo meaning. Since their popularization in

senses connected with the physiology of sex, it has become increasingly difficult to use terms

like intercourse, ejaculation, erection and rubber in ‘innocent’ sense without conjuring up their

sexual associations. This process of taboo contamination has accounted in the past for the dying-

out of the non-taboo senses of a word: cook, in its farmyard sense, is replaced by rooster due to

the influence of the taboo use of the former.

Collocative meaning consists of the associations acquired on account of the meanings of words

which tend to occur in its environment. Pretty and handsome share common ground in the

meaning of ‘good-looking’, but may be distinguished by the range of nouns with which they are

likely to co-occur or collocate.

girl boy boy man woman car flower vessel pretty garden handsome overcoat colour airliner village typewriter etc. etc. The ranges may overlap: handsome woman and pretty woman are both acceptable, although they

suggest a different kind of attractiveness because of the collocative associations of the two

adjectives.

Some words are collocationally restricted, in that they occur only in conjunction with other

words:

rancid -bacon addled -eggs wander/stroll (cows may wander, but not -butter -brains stroll) This does not seem to be a matter of their meaning, but of the company they keep.

Not all differences in potential co-occurrence need to be explained as collocative meaning: some

may be due to stylistic differences, others to conceptual differences:

Page 27: Semantica Curs Ansamblu 2007

27

e.g. stylistic difference: The knight mounted his steed

Little Tommy got on his gee-gee.

conceptual difference: The donkey ate hay

The donkey ate silence.

Only when explanation in terms of other categories of meaning does not apply do we need to

invoke the special category of collocative meaning: on the other levels, generalizations can be

made, while collocative meaning is simply an idiosyncratic property of individual words.

Languages differ in the collocational ranges of their words. In English, we distinguish between

between wiping our nose, brushing our teeth and polishing our shoes, whereas in German the

term putzen can be used for all these activities (Stork and Widdowson, 1974, quoted in Graddol

et all, 1994:110).

Cliches are born when a word occurs very frequently in a particular collocation. The word may

may then lose some of its semantic ‘force’, because we become used to thinking of the phrase as

a single unit. Some examples from everyday English are ‘last but not least’ and ‘the more the

merrier’. Different professions tend to coin their own clichés: estate agents, for example, talk of

‘a wealth of exposed beams’ and ‘tastefully decorated throughout’.

Habitual collocations often reflect social conventions and social attitudes: the collocations of the

words pretty and handsome, for example, indicate that we categorize good looks for men and

women separately. This is one way that language can perpetuate social divisions. (Graddol et al,

1994:110)

ASSOCIATIVE MEANING: a summary term

Reflected meaning and collocative meaning, affective meaning and social meaning, all these

have more in common with connotative meaning than with conceptual meaning: they all have

the same open-ended, variable character, and lend themselves to analysis in terms of scales or

ranges. They can all be brought together under the heading ASSOCIATIVE MEANING. We can

contrast them with conceptual meaning because conceptual meaning seems to require the

postulation of intricate mental structures, which are specific to language and the human species.

F) THEMATIC MEANING is the meaning of what is communicated by the way in which a

speaker or writer organizes the message, in terms of ordering, focus and emphasis. It is often

felt, for example, than an active sentence such as

Page 28: Semantica Curs Ansamblu 2007

28

(1) Mrs. Betty Smith donated the first prize WHAT did Mrs. Smith donate?

has a different meaning from its passive equivalent:

(2) The first prize was donated by Mrs. Betty Smith. WHO was the first prize donated by?

although in conceptual content they seem to be the same.

Certainly these have different communicative values in that they suggest different contexts: the

active sentence seems to answer an implicit question: What…?, while the passive sentence

seems to answer an implicit question: Who was the first prize donated by?. This means that (1),

in contrast to (2), suggests that we know who Mrs. B. Smiths is (perhaps through a previous

mention). The same truth-conditions, however, apply to each: it would be impossible to find a

situation in which (1) was an accurate report, while (2) was not, or vice-versa.

Thematic meaning is mainly a matter of choice between alternative grammatical constructions,

as in:

(3) A man is waiting in the hall. (4) There’s a man waiting in the hall. (5) They stopped at the end of the corridor Q; Where did they stop? (6) At the end of the corridor, they stopped. Q: What did they do at the end of the

corridor? (7) I like Danish cheese. What do I like? (8) Danish cheese I like best. Which cheese do I like best? (9) It’s Danish cheese that I like best.

In other cases, it is the stress or intonation, rather than grammatical construction that highlights

information in one part of a sentence.

(10) I asked the captain Mr. Brown.

(11) John said his girlfriend is a student.

(12) Bill uses an electric razor vs. Bill uses an electric razor.

The effect is to focus attention on that word which contains the new information, against a

background of what is already assumed to be known (i.e. that Bill uses a razor).

Demarcation problems As far as the 7 types of meaning are concerned, there are always problems of ‘demarcation’ and

of separating conceptual meaning from the more peripheral categories. The difficulty in

delimiting conceptual from connotative meaning is parallel in other borderline areas, such as

that between conceptual and socio-stylistic meaning.

Page 29: Semantica Curs Ansamblu 2007

29

(13) He stuck the key in his pocket. (colloquial) (14) He put the key in his pocket. (neutral) We could argue that the two sentences are conceptually synonymous, and that the difference

between the two is a matter of style. On the other hand, we could maintain that the shift in style

is combined with a conceptual difference: that to stick in a context such as (13) has a more

precise denotation than to put in (14), and could be roughly defined as ‘to put carelessly and

quickly’. There is support for the second explanation in the slight oddity of the following

sentences:

(13a) *He stuck the key slowly in his pocket.

(13b) *He stuck the key carefully in his pocket.

As a second illustration of the demarcation problems, we may take as a case on the border

between conceptual and collocative meaning that of the verbs smile and grin. Do these words

have different conceptual meanings, or is it just that the range of expressions with which they

habitually combine is different.

(15) The Queen smiled graciously as she shook hands with her guests. (16) The gargoyles grinned hideously from the walls of the castle. Few would hesitate over which of the two words to use in these sentences.

The question is whether such differences in collocation spring from different conceptual and

connotative content: whether, for example, a grin can be defined as a broader, toothier and more

potentially hostile expression than a smile, and is more likely to be found on the face of a

gargoyle than that of a queen. This is a particularly complex case in that the differences between

social and affective meaning are also clearly implicated.

INTENDED and INTERPRETED MEANING

Intended meaning is that which is in the mind of the speaker when he is framing his message,

and the interpreted meaning is that which is conveyed to the mind of the hearer/listener when he

receives the message.

Meaning, in its broader sense, is equated with the ‘communicative effect’, and communication

usually means transfer of information from a source (A) to a target (B). On this basis, one might

argue that communication has only taken place if we know that what was in mind (A) has been

transferred to mind (B). It is natural, then, that studies of meaning (particularly in philosophy)

should have devoted much attention to the question of the relation between meaning, intention

and interpretation. In spite of this, a linguist may feel entitled to ignore the difference between

Page 30: Semantica Curs Ansamblu 2007

30

the intention of a message and its effect, because he is interested in studying the communication

system itself. He is interested in studying the semantic aspect of the language, which we may

assume to be common to the minds (A) and (B), and this includes incidentally studying

ambiguities and other aspects of language which can cause miscommunication. But the

important point is that meaning, for semantics, is neutral between ‘speaker’s meaning’ and

‘hearer’s meaning’. And this is surely justifiable, since only through knowing the neutral

potentialities of the medium of communication itself can we investigate differences between

what a person intends to convey and what he actually conveys.

Chapter 5: The componential analysis of meaning

Page 31: Semantica Curs Ansamblu 2007

31

The most significant development in semantic theory based on the isomorphism between

language expression and content is the componential analysis. Componential analysis assumes

that all meanings can be analysed into distinctive semantic features called semes/semantic

components/semantic primitives. This form of analysis was used in particular by anthropologists

seeking to give an account of kinship terminology in various cultures. Given the items:

1. father 5. uncle

2. mother 6. aunt

3. son 7. nephew

4. daughter 8. niece

one can arrive at a number of semantic features (semes) by examining the relations obtainable

among them. Thus, by opposing 1, 3, 5, 7 to 2, 4, 6, 8, the feature of sex will be uncovered: a=

male, -a= female.

Opposing 1,2,3,4 to 5,6,7,8 we will come to another semantic feature, namely the line distinction,

which can be further analysed into b= direct line, -b= indirect line.

By a similar procedure which opposes 1,2,5,6 to 3,4,7,8 one arrives at the generation seme, with

2 possible distinctions: c=older generation, -c=younger generation.

Each element can now be represented as an individual grouping of these 3 semes:

-father: a, b, c uncle: a,-b, c

-mother: -a, b, c aunt:-a,-b, c

-son: a, b,-c nephew: a,-b,-c

-daughter; -a, b, -c niece: -a, -b, -c.

The analysis of meaning into semes leads therefore to a relatively small number of semantic

features which are to be encountered in a large number of meanings.

The semes are arrived at by comparing various lexical items in the language. Thus, the set of

English words denoting furniture intended for sitting are defined in a dictionary as follows:

stool: made of wood or metal; having 3-4 legs; short or tall; plain or upholstered.

chair: sharing all the descriptive features of stool, but also provided with a back support.

armchair: a chair which in addition to a back support is also provided with a support for arms.

bench: sharing all the descriptive features of chair, but much wider, so as to seat more people.

Page 32: Semantica Curs Ansamblu 2007

32

sofa: upholstered bench.

The semantic features which are encountered in the meanings of all these items can be

represented as follows:

lexical item

semantic

feature

for sitting with back

support

with arm

support

for more

people

upholstered

stool x - - - 0

chair x x - - 0

armchair x x X - 0

bench x x X x -

sofa x x X x x

The diagram indicates that all items share at least one feature in common (for sitting) and all

differ by at least one feature. In a number of cases, marked by 0, the presence or absence of a

given feature is not relevant.

The isomorphism between expression and content can be carried even further. Just as on the

expression level, an archiphoneme can be identified as the result of neutralization of a distinctive

phonological opposition, on the content level, too, an archilexeme will result from neutralization

of a lexemic opposition. In the case of the term mentioned before, the archilexeme function may

be taken over by the more general of them, namely CHAIR, or it can find an expression in

another lexical item, namely SEAT. Neutralization is a very frequent phenomenon in the field of

lexis. Such terms as man, horse, sheep, etc. can be used for any member of the class of being

they designate, irrespective of the sex distinction, which, however, becomes operant whenever it

is required.

If we take the componential definitions of adult, child and man, these will illustrate that not all

semantic oppositions relevant to a given semantic field need to be operant in a given definition

within that field: child and adult are not specified for [sex], man (with the meaning of ‘human

being’) is unspecified both for [sex] and [adulthood], and the adjective female is unspecified for

both [species] and [adulthood]. We could represent this neutralization of oppositions by the

symbol “0”. The more fully expressed definitions would then be:

man +HUMAN (0 MALE) (0 ADULT)

adult +HUMAN (0 MALE) (+ ADULT)

Page 33: Semantica Curs Ansamblu 2007

33

child +HUMAN (0 MALE) (- ADULT)

female (0 HUMAN (- MALE) (0 ADULT)

The delimitation of semes can also be profitably pursued by comparing the meaning of what are

considered to be equivalent lexical items in a number of languages. It becomes obvious that

features which are relevant for one language may be irrelevant for another one. A well-known

example was provided by L. Hjelmslev (1961) by comparing the terms for siblings in several

languages

lexical items HU�GARIA� E�GLISH MALAYSIA�

older brother bàtya

younger brother öccs Brother

older sister nène

younger sister hug Sister

Sudará

English, too, has a term which may be taken to be the equivalent of sudará, namely sibling, but

its status in language is altogether different. Also in some dialects of Romanian there are distinct

terms of address for older brother (nene) and older sister (lele, ţaţă), but none for the younger

siblings.

Provided the analysis is applied to an extensive number of lexical items, and provided it involves

a great number of languages, preferably unrelated, a universal inventory of semantic primitives

could be arrived at. These universal semantic features would then be part of a metalanguage to

be used in the semantic description of all languages.

5.1. The relation between lexical items and semantic components It is essential to clarify the relation between semantic component and lexical item. The question

we have to ask is: What is the relation between [ADULT] and ‘adult’, between [MARRIED] and

‘married’, etc?

It is generally claimed by those linguists who postulate the existence of semantic components

that the components in terms of which lexical meaning is represented are not themselves lexical

items, part of the language being described, but are part if the metalanguage2, the language or the

theoretical vocabulary set up to describe all languages. Components such as [MALE], [ADULT], 2 Metalanguage= a higher-level language for describing an object of study – in this case the object is itself language.

Page 34: Semantica Curs Ansamblu 2007

34

NEVER MARRIED] are therefore universal constructs in terms of which such lexical items as

‘bachelor’ in English can be characterised. Items such as ‘adult’ which happen to correspond to a

single semantic component [ADULT] have a unitary semantic characterisation. Now it is

particularly clear in these cases that if these semantic components are the basis on which lexical

meaning is to be explained, there must be some characterisation of [ADULT] independent of the

lexical item ‘adult’, for otherwise we have given no explanation at all. Moreover, it will not

suffice at this point merely to say that to give the meaning of a lexical item is to specify the

contribution which that item makes to the meaning of a sentence in which it occurs.

Therefore, it seems that in addition to the vocabulary of semantic components expressing

generalizations between lexical items, we need to have a principled way of relating these abstract

components to the properties or individuals that they describe.

The idea that word-meanings are composed of primitive universal semantic elements was taken

up and further developed by M. Bierwisch and R. Jackendoff. The latter assumes that the

primitive semantic elements designate some kind of a mental construct or concept. Still, his

theory fails to show how these concepts connect with what we talk about.

Dowty, a generative semanticist) used this kind of decomposition to analyse different aspectual

classes within which predicates in natural languages seem to fall. He classifies verb-phrase into:

states, activities and telic eventualities.

a) States are like snapshots of the world at a given instant; they lack culmination or end-point;

their subject is not perceived as an agent, but as an experiencer.

- states typically cannot be used in the progressive: * He is drunk

- states are odd in the imperative: *Be drunk!

- states are bad in sentences of the form ‘it took NP an hour (minute, year) to VP: *It took him

an hour to be drunk.

b) Activities lack culmination or end-point; they have a subject doing something; they cannot be

viewed as instantaneous snapshots of the world.

- they can be used in the progressive: John is drinking.

- they can be used in the imperative: Drink more!

c) Telic eventualities have a natural end-point or culmination (John fell asleep). They are of two

types, namely (1) achievements and (2) accomplishments.

-they are generally good in the progressive: John was falling asleep.

-they are generally good in the imperative: Fall asleep!

Page 35: Semantica Curs Ansamblu 2007

35

-they are good in the construction ‘it took NP an hour (minute, year) to VP’: It took John an hour

to fall asleep.

What is interesting to notice is that the interaction of a predicate with certain phrase may shift the

aspectual class of that predicate. Thus, walk classifies as an activity, since it does not have an

end-point. But once we attach a culmination point (for instance to school), the activity turns into

a telic event. In other words: activity + culmination point = telic event (walk to school).

The question that arises is how we should account for these changes? Dowty suggested the

introduction of 3 aspectual operators: DO, BECOME and CAUSE. He suggests that verbs are

defined from basic stative predicates in terms of these operators. The semantic properties of such

operators determine the properties of various aspectual classes. If we assume we are endowed

with basic stative predicates and the combinatorial apparatus of logic and the 3 operators,

predicates will form the 3 classes mentioned above (i.e. states, activities and telic events.)

Let us briefly present the three operators.

DO shows a binary relation between individuals and properties. A formula like DO (j, Motion)

translates as John moves. Assumption: MOTION is an abstract property, such that MOTION (x)

means that (x) is not static. Interpretation: something that John does causes him to be non-static.

Thus, DO is a relation between a property and an individual that says that the individual has the

property as a consequence of his doing something.

ACTIVITIES = STATES + DO

BECOME is a one-place operator3. A formula like BECOME (┐ALIVE) translates as ‘to die’

(to become not alive).

CAUSE is a two-place operator

Achievements and BECOME Achievements and accomplishments are both events. An event takes place at a time if one state

(the initial state) is replaced by a second state (the final state). An example of an event would be

the opening of the window: it consists of a change from a state when the window is closed to a

state when the window is open.

3 Operators are verbs and adjectives that receive arguments in order to form syntactic structures. The operators can take one, two or three arguments

Page 36: Semantica Curs Ansamblu 2007

36

Any event can be defined as a change of state, where the two states (the initial and the final ones)

are of a particular form – one state is the negation of the other. An event is a change from state

p to state q, where p = ┐q.

Dowty hypothesizes that all achievements have a semantic structure consisting of BECOME +

an embedded clause.

Example: realise = to become [to know something which one didn’t know earlier]

forget: to come [not to know what one knew earlier]

arrive at: to come[to be at a place]

A sentence like The soup cooled can be translated into the following formula: BECOME [the

soup ┐warm]

Accomplishments and CAUSE Accomplishments are also events, i.e. changes of states. But they are ‘topic-marked’, i.e. an

accomplishment conceptualizes not only the change of state and therefore the attainment of a

certain goal, but also the causing factor, that which produces the change. Accomplishments are

causative predications. Compare the following sentences:

a) The door is closed. STATE The soup is warm.

b) The door opened. EVE�T (achievement) The soup cooled.

BECOME [the door is ┐closed] BECOME [the soup is ┐warm]

c) John opens the door. EVE�T (accomplishment) Mother cooled the soup.

[John DOES something that CAUSES [Mother DOES something that the

door to BECOME ┐closed] CAUSES the soup to BECOME ┐warm].

The examples show that accomplishments are conceptually more complex than achievements,

involving the operator CAUSE alongside of BECOME (verbs like to open and to cool are

inchoative verbs, i.e. they specify a change of state).

There is some evidence for the bi-sentential nature of CAUSE. The most obvious motivation of

CAUSE as a bi-sentential operator is a semantic one. Accomplishments are characterised as

being semantically bipartite in a way in which activities and achievements are not, i.e. they

involve both the notion of an activity and a change of state which comes into being as a result of

that activity. The sentence John killed Harry entails that John performed some act and Harry

came to be dead as a result of that act.

Page 37: Semantica Curs Ansamblu 2007

37

Further evidence that Dowty brings in support of his decompositional analysis involves data

about the interaction of certain adverbials (again, for the third time) with the meaning of non-

stative verbs.

non-ambiguous: a) John is drunk again – implies that John was drunk before.

b) John is drunk for the third time – implies that there were at least three

occasions on which John was drunk.

ambiguous: c) John opened the door again – could be interpreted in two ways:

-John CAUSED again to become open

-John caused the door to BECOME open again.

This ambiguity is determined by the way in which the adverb again combines with the two

operators that are constituent elements of accomplishments. Dowty suggests that stative

predicates like in sentence (a) above are simple, and hence adverbial modification cannot yield

any ambiguity. On the other hand, telic eventualities are complex, and an adverb can be

interpreted either with the becoming or with the causing of the resulting state or with the

resulting state itself.

All in all, Dowty’s analysis of predicates is a refined form of the structural componential

analysis.

5.2. Criticism and problems of the componential analysis (CA) of meaning

Componential analysis, as a theory of word meaning, is controversial. While many have found it

a useful and revealing technique for demonstrating the relations of meaning between words,

others have criticised it. Let us consider some of the criticisms that have been made.

1. It is often said that CA accounts for only some parts of a language’s vocabulary (specifically

those parts which are neatly organised. If we restrict ourselves to too simple a concept of CA,

that is certainly true. But CA can be enriched to deal with a much wider range of vocabulary.

Even so, the principle that CA could deal with the whole vocabulary of a language is felt to be

implausible. To account for the immense open-ended expressive power of language we would

need an infinite number of oppositions. This objection is accepted, but should not throw out CA,

because this analysis can be fitted into a more powerful model of meaning. In this respect,

semantic features need not be regarded as ‘atomic contrastive elements’, as Geoffrey Leech

(1990) defines them, because they may have an internal structure of their own, i.e. semantic

features can be derived from configurations of other features. This recursive power of feature

Page 38: Semantica Curs Ansamblu 2007

38

creation is particularly important when considering METAPHOR and other kinds of meaning

transfer.

2. CA allows us to provide definitions for words in terms of a few components of meaning. Thus,

a boar is [PORCINE], [MALE], [ADULT]. Still, there are gaps in the system, in that there are

not terms to distinguish between the male, female and young with giraffes, elephants or

rhinoceroses. Often the distinction is made by using a term taken from another set in conjugation

with the generic one:

e.g. bull elephant, cow elephant, elephant calf.

3. CA attempts to treat components (as far as possible) in terms of ‘binary opposites’, e.g. male

vs. female, animate vs. inanimate. Notationally, there is an advantage in such binary terms in

that we can choose only one as the label and distinguish this in terms of + and -. But this works

well only where there is a clear distinction; often, there is indeterminacy, as with tar or porridge

in relation to [SOLID] vs. [LIQUID].

4. It is often objected that CA suffers from a ‘vicious circle’ in that it merely explains one set of

symbols (e.g. English words) by another set of symbols (which often turn out to be English

words), e.g..[ADULT] and ‘adult’.

5. CA postulates universal features of meaning, and therefore relies upon the strong assumption

that the same semantic features are found in all languages. But as we have seen in the example

provided by Hjelmslev, these semantic features are not operant in all languages.

6. It has also been claimed that CA is unexplanatory in that it does not provide for interpretation

of semantic features in terms of real-world properties and objects they refer to. For example,

[ADULT] remains an abstract, uninterpreted symbol unless we can actually specify what adults

are like, i.e. how we decide when the feature +ADULT refers to something. To expect CA to

provide an interpretation in this sense is to expect it to provide a theory not only of MEANING

but of REFERENCE, or, to put it in other words, not only of conceptual meaning, but also of

connotative meaning as well. But CA cannot have this wider goal; it is meant to explain word

sense, not the encyclopaedic knowledge which must enter into a theory of reference.

7. Words such as colour terms and days of the week cannot be analysed as differing from each

other by the simple presence or absence of a particular semantic feature.

Page 39: Semantica Curs Ansamblu 2007

39

8. It is not necessarily particularly revealing simply to list the semantic components of a word,

and even if it does seem that this can be done, we may need to say how the components are

combined. An example is the word kill, which can be analysed into the components [cause],

[become], [-alive] but for which we need, in addition, to specify that the components are ordered

in a hierarchical relationship: [cause]→[become] →[alive] and to specify the participants (for

example X[cause]Y→Y[become] →[-alive]).

9. A further difficulty is that, although it is economical to represent semantic features as being

present or absent, it is not necessarily helpful to give the meaning of a word in terms of what it is

not, rather than in terms of what it is. The meaning of the word girl, has been analysed in terms

of the features [-adult], [-male], which do, indeed, distinguish it from boy [-adul], [+male], but

which do not tell us a great deal about the meaning of girl. (Graddol et al, 1994)

Despite all these shortcomings, CA remains one of the most advanced theories in semantics. It

has been used to bring out the logical relations that are associated with words. Thus, by marling

‘man’ as [MALE] and ‘pregnant’ as [-MALE], we can rule out a phrase such as ‘pregnant man’.

Page 40: Semantica Curs Ansamblu 2007

40

Chapter 6: The distributional analysis of meaning

One of the main shortcomings of traditional semantics consists in the fact that individual words

were analysed in almost complete isolation from other items in the lexicon of a given language.

It was believed and taken for granted that words have one or more meanings which remain stable,

representing the contribution they make to any utterance in which they occur. In the case of

polysemantic words, it is the linguistic context which solves the problem of which of the

multiple meanings of the word is used in any given utterance. But traditional semantics could not

account for any relation between a given word and its actual use in different utterances. Its main

concern was to find explanations for the meanings of words, to find their origins or to follow the

changes in their meanings. On the other hand, dictionaries only listed words with as many

meanings as possible.

One of the main developments in modern semantics, the context theory of meaning, focuses on

the close relationship that can be established between the meanings of a given lexical item and

its actual use in speech. But what do we understand by CONTEXT? It is not limited to linguistic

environment alone, but it includes all other non-verbal elements which are also relevant for

communication.

In the case of meaning analysis, it is important to make the distinction between CONTEXT,

standing for the entire non-verbal environment which is linguistically relevant, and CO-TEXT,

which stands for the linguistic environment proper, for the items in the text which play a role in

specifying the meaning of a given lexical item. Modern linguistics has paid attention to the

analysis of meaning in co-textual rather than contextual terms.

Page 41: Semantica Curs Ansamblu 2007

41

Context is very important; it has a big influence upon vocabulary. This influence can be

attributed to the fact that each lexical item is characterised by suppleness and original vagueness

which makes it adequate for use in the most various situations. This makes it possible for

language to describe the most refined distinctions of reality. The result is an interplay between

lexical items and context, which is very important in defining meaning.

There are two directions in which CO-TEXT proves useful in the analysis of meaning:

a) the distribution of the same lexical item in structurally differing linguistic co-texts;

b) the distribution of different lexical items in identical or structurally similar co-texts.

We usually discuss only the first dimension, i.e. the interplay between the distribution of a

lexical item in various utterances and the meanings actualised by each of these co-texts.

Consequently, we shall analyse meaning along the syntagmatic axis.

The distributional analysis of meaning (DAM) is based on the assumption that starting from

a grammatical description of language, it is possible to engage in a semantic analysis of

lexical items in terms of their distribution. Our task is to establish some correlations between

these distributional models and the various meaning of the lexical items employed in them.

If we analyse at random some polysemantic lexical items, we will find a biunique

correlation between meaning and distribution which can be stated in the following terms:

while it is difficult to maintain that all semantic distinctions will manifest themselves in

distributional difference, it can be assumed with greater certainty that to each distributional

difference corresponds an essential semantic one.

The meaning of lexical items can be defined in two different ways along the syntagmatic axis:

a) Differences in meaning will be marked by differences in syntactic structure. This rule

applies (in the case of English, a highly analytical language) for clarifying the ambiguity of

the multiple grammatical functions of a given lexical form. The syntactic structure alone will

indicate which meaning of the lexical item HAWK is meant in the following examples:

1) a. The hawk killed the chicken. (N = bird of prey)

b. They hawk haberdashery in the outskirts of the town. (V = to sell goods from house

to house)

The syntactic structure can also disambiguated the meanings of a lexical item when used in

the same grammatical function:

Page 42: Semantica Curs Ansamblu 2007

42

2. a. The objection still stands. (Adv = even to this time)

b. The child stood still. (Adv = quiet, without movement or sound)

c. Her brother is taller still (Adv = yet, even).

b) Differences in meaning will be determined not only by the syntactic structure of the

utterances in which the respective lexical items appear, but also by the semantic markers of

the items they usually collocate with. Consequently, we are dealing with a semantic

differentiation of meaning. Examples of two major word-classes (nouns and adjectives) will

illustrate this process of meaning delimitation by reference to co-occurring elements in the

text.

MASTER

a) He was the master of the dog

master1 = owner of’ marked by a group of words designating persons, animals or things that

can be owned, such as slave, dog, house, ship, etc.

b) Master Heathcliff left early in the morning.

master2 = title of respect for boys; associated with proper names designating young

unmarried men.

c) Our class-master was a rather old man.

Master3 = head of; marked by a class of words such as school, station, designating

institutions that imply a collective activity which needs coordination by a head.

HIGH (a polysemantic adjective whose main meanings are differentiated by means of

collocation)

a) The Tower of London is not very high.

High1 = extended upwards; this meaning is revealed in connection with a group of words

designating objects that are measured vertically, such as tower, tree, building, mountain.

b) One day, a high official came to my place.

High2 = important in rank, position or significance; this meaning is revealed in association

with words such as official, authority, treason, etc.

c) The driver changed into high gear.

Prices are very high nowadays.

Yesterday the wind was high.

High3 = intense, extreme; association with words like speed, gear, wind, temperature, price,

cost, level, favour, opinion, spirits, etc.

d) He drove fast along the highway.

Page 43: Semantica Curs Ansamblu 2007

43

High4 = main; this meaning is revealed in conjugation with such words as way, road, street.

e) She seems to speak in a high tone.

High5 = shrill; this meaning is revealed by a relatively small number of words such as voice,

tone, key.

In order to set up the distributional formulae of various lexical items with a view to analysing

their syntagmatically conditioned meanings, we will make use of the grammatical categories

provided by grammar. In order to account for the combinatory values, nouns have been divided

into structurally defined subclasses. This classification has proved to be useful for the analysis of

the combinatory valences of the English verbs and adjectives.

NP

A (animate) Non-A (inanimate)

H (human) Non-H (non-human) C (concrete) Non-C (abstract)

M (male) F (female) Count Non-Count

(he) (she) (it, they) (it)

In what is to come, we shall try to illustrate the correlation between distribution and meaning by

means of members of two lexical classes, i.e. verbs and adjective. The meaning relations existing

between members of these classes on the one hand, and members of the class of nouns of the

other, were known to lexicographers who used them extensively in describing properly the many

shades of meaning characteristic of polysemantic lexical items.

Lexical item Lexical class

Distributional formula Example sentence

Meaning

H +fail+H John failed his parents

disappoint

H+fail+V They failed to climb the mountain

Not succeed

FAIL

V

Non-C+fail+H Luck failed him Abandon, leave H+entertain+H We often

entertain our friends

Play host to

ENTERTAIN

V H+entertain+Non-C She has

entertained the consider

Page 44: Semantica Curs Ansamblu 2007

44

idea for weeks H+examine+NP He examined

my eyes Get

information/knowledge about

EXAMINE

V H+examine+H The teacher

examined his students

Test knowledge of

H+feel+NP The boy felt the edge of the

blade The dog felt the

burglars

Become aware through senses

FEEL

V

H+feel+S He felt that it was his duty to

marry her

consider

C+Be+flat The ground is flat

Smooth, even, level

H+Vt+H+flat They struck him flat

To the ground

FLAT

ADJ

NP+Vi+flat Life seems flat uninteresting C+be+green The fruit is

green Not ripe

GREEN

ADJ H+be+green He is still green Not mature, untrained,

inexperienced

THE DISTRIBUTIO�AL A�ALYSIS OF THE STRUCTURAL MEA�I�G OF THE

VERB ‘TO MAKE’

The distributional models which are significant for the specification of various meanings of a

lexical item represent the various syntactic constructions in which this item may occur. In

English, for a lexical item belonging to the class of verbs, we have to consider the following

distributional formulae:

N+V+N N+V+Prep+N2

N+V+ADJ N+V+ADV

N+V+V

However, it is important to make sure that the syntactic constructions we compare are really

different. It would be useless to expect two different meanings from two distributional formulae

of the kind:

N1+V+N2 He wrote a letter.

N2+be+Past Part.+by+N1 The letter was written by him.

These two sentences are related transformationally. Consequently they will belong to one single

distributional formula in the set provided by grammar for the purpose of the distributional

analysis of the structural meaning of lexical items.

Page 45: Semantica Curs Ansamblu 2007

45

Transformational considerations are also useful to account for more refined shades of meaning

within the general meaning supplied by a certain distributional formula. Moreover, to account for

combinatory valences, sometimes we need a more detailed sub-grouping within a certain

grammatical category. For the specification of some restricted meanings, some semantic criteria

different from the semantic features revealed by the componential analysis have to be used.

The way in which transformational criteria and more refined semantic criteria are introduced into

the distributional analysis of meaning can be illustrated by an attempt to specify the meanings of

a highly polysemantic lexical item: MAKE. The aim of the analysis is to arrive at a set of

relevant distributional formulae which specify the various meanings of the word.

I. As a first distributional formula, we shall postulate N+ MAKE+N, without assigning any

specific meaning of make to it.

The syntactic sub-pattern N+V+N+N will present 2 variants for the verb MAKE:

a) H+MAKE+N1+N2 – where N1 and N2 belong to different subclasses of the class of nouns

E.g. Mother made her daughter a dress (N1 is animate, N2 is inanimate)

b) H+MAKE+N1+N2 – where N1 and N2 belong to the same class of nouns, i.e. they are both

animate or non-animate.

e.g. John made his son a musician (N1 and N2 are animate)

The burglar made the metal rod a weapon (N1 and N2 are inanimate)

Both constructions are transformationally related to constructions of the type:

N+MAKE+N1+N2 → N+MAKE+N2+ Prep+N1

Mother made her daughter a dress → Mother made a dress for her daughter.

Father made his son a musician → Father made a musician of his son.

The two patterns are identical except for the prepositions. However, a clear-cut distinction will

appear when further transformational relations are investigated (i.e. if we consider the result of

the action).

N+MAKE+N1+N2 → N1 be N2

→ N1 have N2

Father made his son a musician → His son is a musician.

Father made his son a toy. → His son has a toy.

Page 46: Semantica Curs Ansamblu 2007

46

One further distinction appears when deletion is applied:

N+MAKE+N1+N2 → N+MAKE+N2 (N1 is deleted)

Father made his son a toy. → Father made [√ ] a toy.

Father made his son a musician. → *Father made [ √ ] a musician. (unacceptable)

II. A second distributional formula we will consider is N+V+N+Adverbial particle4

e.g. I try to make matters up (compensate)

You must make yourself over (you must change your manner)

It is important to note the following things in connection to this formula:

- the presence of the noun in front of the adverbial particle is obligatory’

- only a limited number of adverbial particles (out, up, over) can appear in this formula

III. A third formula would be N+MAKE+N+Prep+N. In order to assess its significance for the

meaning of the verb MAKE, we shall distinguish 2 variants:

- sequences that allow the transformation N+MAKE+N1+Prep+N2 → N+MAKE+N2+N1

e.g. He made a man of you → He made you a man

Frost makes ice from water. → Frost makes water ice.

- sequences that do not allow the above mentioned transformation:

e.g. He made a speech at the meeting.

The fact that this last sentence belongs to a separate pattern than the sentences above is proved

structurally, in that it allows the following transformation:

N+MAKE+N1+Prep+N2 → N+Vn+Prep+N (where Vn is a verb derived from a noun).

He made a speech at the meeting→ He spoke at the meeting.

Such sequences can be treated as extended �+MAKE+�, since the additional Prep+N is not

relevant in most of the cases. Consequently, we can conclude that neither variant of

N+MAKE+N+Prep+N can constitute a significant distributional formula relevant to the meaning

of MAKE, since both can be relegated either to the pattern N+MAKE+N1+N2 (N1=N2) or to the

more general pattern N+MAKE+N, which will be discussed in more details later.

IV. The fourth distributional formula we shall consider is N+MAKE+N+ADJ. This resembles in

a way the N+MAKE+N+N pattern, more specifically its variant which allows the N1 be N2

transformation.

e.g. She made her a lady. → She is a lady.

She made him happy. → He is happy.

4 An adverbial particle is an adverbial that can also be used as a preposition.

Page 47: Semantica Curs Ansamblu 2007

47

If we consider sequences of the type N+MAKE+N+ADJ+ (that) +S (He made it clear that he

would fight to the end), we can see that the pre-adjectival N position is filled by the pronoun it

which anticipates the post adjectival subordinate clause. The relation is: it = that+S. On this basis

we can postulate that we are dealing here with a variant of N+MAKE+N+ADJ pattern, and

consequently will have to consider the latter an independent distributional formula for the

meanings of MAKE.

V. The formula N+MAKE+N+V is of particular interest. This syntactic construction cannot be

reduced to the basic N+MAKE+N pattern, because a close interrelationship obtains between

MAKE and the other verb in the sequence.

e.g. He made her cry.

He made my blood run cold.

It follows that this distributional formula (with possible additional sequences after V) should be

considered a separate basic pattern constituting one of the general meanings of the verb MAKE.

VI. N+MAKE+ADV (off, away, in) is another pattern that needs consideration.

e.g. He made away with the money.

The important thing about this pattern is the fact that the sequence is itself self-sufficient for the

specification of the meanings of MAKE. Consequently, it can be considered a basic pattern.

The patterns established so far do not represent simple distributions but rather sets of

distributions within a certain more general type. The meaning each of these patterns reveals is

very general and unspecified. For instance, the only meaning we can specify at this stage for the

first pattern N+MAKE+N is very wide and vague; it is actually related to the general meaning of

the transitivity of verbs. For further specification of meaning within a certain pattern, other

possible grammatical operations have to be applied to the postulated distributional formula.

The basic pattern N+MAKE+N will be discussed now in greater detail. We shall try to arrive at

more specific meaning of the verb MAKE by testing the reaction of the pattern to various

grammatical operations. Let us first consider the transformations that are allowed by it:

a) N1+MAKE+N2 → N2+MAKE+N1

e.g. The boy made a mistake → A mistake was made by the boy.

b) N1+MAKE+N2→ N1+MAKE+(N2) (a transformation that changes the MAKE+N2 sequence

into a verb derived from N2)

Page 48: Semantica Curs Ansamblu 2007

48

e.g. He made an analysis of the situation → He analysed the situation.

c) N1+MAKE+N2 → N1 +BE+N2 (a transformation that allows the replacement of MAKE by BE)

e.g. She makes a good cook. → She is a good cook.

But there is no actual N+MAKE+N sequence which will allow all three transformations. Some

will allow only one (transformation © above):

e.g. He will make a good doctor → He will be a good doctor.

Others will allow two transformations (transformations (b) and (a) above:

e.g. John made an analysis of the situation → John analysed the situation.

→An analysis of the situation was made by John.

Still others do not allow any transformation:

e.g. His argument made sense.

As a result, the general pattern N+MAKE+N will be subdivided into several sub-groupings,

according to their reaction to the set of transformations (a, b, c). Each sub-grouping will be

assigned a specific meaning of MAKE. Each f these groups can be further analysed in terms of

various semantic markers of the verb MAKE.

This has been an exemplification of the way one can postulate the relevant distributional

formulae for a specific lexical item. Even if such an analysis were pursued further, it could not

cover all the meanings of MAKE. The possibilities of distributional analysis of meaning are

limited! There are a number of very specific meanings which can be probably solved at a

different level – the idiomatic or phraseological level.

CRITICISM OF THE DISTRIBUTIONAL ANALYSIS OF MEANING

The attempt to state meaning in this way is not satisfactory.

First, it does not deal with what is usually meant by meaning. Secondly, it is difficult to see how

such an approach could do more than just indicate sameness and difference of meaning (in terms

of sameness and difference of distribution). It is not at all clear how it could say what meaning is,

except by listing all the environments in which an element occurs.

Thirdly, and most importantly, it is surely obvious that to define meaning in terms of distribution

is very largely to put the cart before the horse: words have different distributional patterns

because they have different meanings.

Page 49: Semantica Curs Ansamblu 2007

49

Chapter 7: Ambiguity and vagueness

It may seem that we have little to say about ambiguity, thinking that it is a clear-cut phenomenon.

But it is astonishing how ambiguous most words and sentences are. Most common words are

ambiguous; if you look in a good dictionary, for any common word you will usually find two or

three meanings, or maybe twenty or thirty, if it is really common. If combined together in a

sentence, these ambiguities multiply together to make things worse. Even the syntactic structures

are themselves ambiguous, as in:

(1) A small girls’ school.

(2) The chicken is ready to eat.

Without looking at the ambiguities of words themselves, the first example may be either [a small

school for girls] or [a school for small girls], while the second is likely to mean [the chicken is

ready for someone to eat], but it could also mean [the chicken is ready to eat something]. It is

probably the first because we normally eat chicken and not the reverse, and anyway chicken are

always ready to eat, so the second reading is pointless. If we change ‘chicken’ for ‘alligator’,

however, the other is more likely, so it is not just the structure of the sentence that makes one or

another more reasonable.

One of the goals of a semantic theory is to describe and explain ambiguities in words and

sentences. The semantic rules the linguist sets up must state correctly for each language which

words and sentences have more than one meaning. But there is a problem, namely to decide what

counts as ambiguity. While there are some clear cases (as in We saw her duck), there are many

cases where it is not at all clear whether the word, phrase or sentence in question is ambiguous or

not. If we take a word like GOOD, opinions about its ambiguity are different. Some may think

that it is not ambiguous. But if we consider the sentence:

(3) She has good legs,

we shall come to several interpretations:

a) she has healthy legs (no varicous veins, no broken or badly mended bones, no weak

ankles);

b) she has beautiful legs;

c) she has legs which function well (as an athlete’s or a gymnast’s, or if the object referred

to is a horse, her legs may be understood to function well from the point of view of

racing).

So we have to grant that the word GOOD may be used in sentences with different interpretations,

where the difference lies solely in the basis of evaluation the word GOOD has been used to make.

Page 50: Semantica Curs Ansamblu 2007

50

The question remains: do we wish to say that the meaning of the word GOOD differs as the basis

of evaluation differs, or do we say that this word corresponds to one single lexical item, whose

meaning is common to all these different bases of evaluation? If we think of the word in isolation,

there might be a tendency to consider it as a single lexical item with a single interpretation,

however hard that interpretation may be to state. But if we think of sentences containing GOOD,

then there is a conflicting tendency to see sentences as ambiguous.

The problem is compounded when we look at other phrases containing GOOD. A good student

describes either someone who behaves well, or someone who works well, or even someone who

works haphazardly but shows a high level of ability. A good film is either one which gives

enjoyment or one which is thought to be of lasting value. What has to be decided is whether the

meaning of GOOD is homogenous and neutral between all these different specifications, or

whether GOOD has different meaning according to its use in describing different things. In more

general terms, this presents an example of the difficulty of distinguishing ambiguity from

vagueness.

7.1. Ambiguity

A word or a sentence is ambiguous when it has more than one sense. We can crudely classify the

sorts of ambiguity found in sentences as follows:

1. Pure syntactic ambiguity – is the ambiguity in which the variant readings of a sentence

involve identical lexical units; the ambiguity is thus necessarily a matter of the way elements are

grouped together.

(4) Old men and women - can be interpreted either as [[old men] and women] or as [old [men

and women]].

2. Quasi-syntactic ambiguity requires careful consideration because there may be a temptation to

diagnose it as cases of lexical ambiguity. This is because there is no straightforward syntactic

explanation of the ambiguity. Not only are the lexical units identical for the two interpretations,

but they are identically grouped, too. Consider the following examples:

(5) The astronaut entered the atmosphere again. The sentence can have two readings:

i) The astronaut entered the atmosphere for (at least) the second time.

ii) The astronaut returned to the atmosphere after what could have been his first trip into the

outer space.

Page 51: Semantica Curs Ansamblu 2007

51

This ambiguity can be accounted for without the need either for 2 different elements ENTER, or

2 different elements AGAIN, if we regard the meaning of ENTER as being constituted out of

more elementary semantic entities which are related quasi-syntactically:

Enter= [come to be] [in]

The two readings can be represented as follows:

i) [(come to be) (in)] [again]

ii) [come to be] [(in) (again)].

Consider now another example:

(6) A red pencil. This may read as ‘a pencil painted red’ or ‘a pencil writing red’.

3. Lexico-syntactic ambiguity is the ambiguity determined both by the syntactic structure of the

sentence as well as by the multiple meanings of the words contained in them.

(7) We saw her duck. This can be paraphrased as [we saw her lower her head] or [we saw the

duck belonging to her]

(8) I saw the door open.

4. Pure lexical ambiguity is determined by the multiple meanings of words.

(9) He reached the bank (either the institution where you deposit your money or the margin of a

river).

(10) What is his position? This can be interpreted either as someone’s location or someone’s

opinion.

A more complex definition of an ambiguous sentence says that a sentence is ambiguous if it has

two (or more) paraphrases that are not themselves paraphrases of each other. If we consider the

lexico-syntactic ambiguous sentence (7) above, we realize that its two readings are not

paraphrases of each other. In the case of words and phrases, we say that they are ambiguous if

they have two or more synonyms that are not themselves synonymous of each other. For

example, COACH is synonymous with trainer and with bus, but these two are not synonymous

of each other, so COACH can trigger ambiguity.

In the case of ambiguous words, a distinction is sometimes made between polysemy and

homonymy. This distinction has basically to do with the closeness, or relatedness of the senses of

the words. A case of homonymy is one of an ambiguous word whose different senses are far

apart from each other and not obviously related to each other in any way. Cases of homonymy

seem very definitely to be matters of mere accident or coincidence. Thus, the word MUG means

Page 52: Semantica Curs Ansamblu 2007

52

both ‘a drinking vessel’ and ‘a gullible person’, but there is no conceptual connection between its

two meanings.

A case of polysemy is one where a word has several very closely related senses. Thus, MOUTH

may be used in connection with a river and an animal. The two senses are clearly related by the

concepts of an opening from the interior of some solid mass to the outside, and of a place of

issue at the end of some long narrow channel.

In practice it is impossible to draw a clear line between polysemy and homonymy.

It is not always possible to find an exactly synonymous phrase for a given word. Where exact

synonyms are not available, it is possible to indicate different senses of a word by giving

different environments in which the word may be used. The word GRASS has two senses which

are indicated by the following environments:

11) Please keep off the grass.

12) The informer grassed on his partners-in-crime

In many cases a word used in one sense belongs to one part of speech, and used in another sense

it belongs to a different part of speech. Thus, LONG in the sense of ‘yearn’ is a verb, and in the

sense of ‘not short’ is an adjective.

Ambiguity is a major problem for semantics; how can a person understand the one an only one

and almost invariably the correct meaning that a sentence has in a given context? It clearly

involves the CONTEXT and the KNWLEDGE of the listener, but how it happens is not yet

known for sure. It must be a fairly simple process, because we do it so swiftly that we are usually

completely unaware of any other possible way to understand a sentence, and we do it so

accurately that we seldom ever make a mistake in understanding.

We also have to mention anther situation that must not be confused with ambiguity, namely

referential versatility. We say that a phrase is referentially versatile if it can be used to refer to a

wide range of different things or persons. Thus, the pronoun SHE can be used to refer to any

female person. On a given occasion SHE might be used to refer to Mary, on another occasion to

Lucy, but this does not mean that SHE is ambiguous, because although it is used to refer to

different people. This is not a matter of difference in sense.

Page 53: Semantica Curs Ansamblu 2007

53

7.2. VAGUE�ESS

In order to see the extent of the problem of distinguishing ambiguity from vagueness, let us

consider the different types of vagueness which languages present us with. Broadly speaking,

there are 4 types of vagueness, though they are related to each other.

1) Referential vagueness appears when the meaning of a lexical item is, in principle, clear

enough, but it may be hard to decide whether or not the item can be applied to certain objects.

Let us take for example the lexical items CITY and TOWN. Presumably we can roughly agree

that a city is a place where a large collection of people live, and it is made up of a large number

of houses, whereas a town is simply any place where a collection of people live, made up of a

certain number of houses. Towns may be small or large, but cities are big by definition (just as

villages are small by definition). Now, even if we can agree that the meanings of the items need

to have a specification along these lines, we shall certainly find difficulty in individual cases in

deciding whether or not some place is a city or a town. It will not do to specify as part of the

meaning of the item that a city must contain a minimum number of inhabitants, for we can talk

of Roman cities where the number might scarcely exceed that of a present-day village. There are

many examples of this kind of vagueness. When is a mountain not a mountain but merely a hill?

When is a forest not a forest but a wood? What crucially distinguishes a house from a cottage?

2) Indeterminacy of the meaning of a lexical item or phrase. In this situation, the interpretation

seems itself quite intangible and indeterminate. Perhaps the most extreme example of this in

English is the possessive construction. Thus, John’s book can be used to describe:

- the book that John wrote;

-the book that John owns;

-the book that he has been reading;

-the book that he was carrying when he came into the room, etc.

A phrase like John’s train can describe:

-the train he normally goes to work on;

-the train he is going to catch;

-the train he drives;

-the train he is a guardsman of;

-the train he designed, etc.

Page 54: Semantica Curs Ansamblu 2007

54

In the face of this variety, it seems clear that we can say little about the meaning of possessive

constructions other than that there must be some relation of association between the ‘possessor’

and the ‘possessed’. The meaning is otherwise quite indeterminate. We could introduce the word

GOOD into this class, since its meaning seems to be variable.

3. Lack of specification in the meaning of a lexical item is the situation in which the meaning,

though in principle quite clear, is very general. The simplest example of lack o specification is an

item like NEIGHBOUR which is unspecified for sex, race, age, etc. Perhaps less obvious

examples are verbs such as GO and DO which both have clearly specifiable meaning and yet

cover a wide variety of actions, since this meaning is so general. The sentence He went to the

station can be used to describe actions as dissimilar as walking, running, going on a bicycle,

going on a motorbike or in a car, to mention but a few, for GO is quite unspecified as to the kind

of the action. It simply has the meaning of directional motion.

I’ve done the sitting-room can be used to imply that she/he has dusted, cleaned, painted, laid the

floor in the room, depending on whether the speaker is a cleaner, a painter, a floor-layer, etc.

Despite this, the meaning of the item DO is not itself indeterminate. The expression to do some

object (as in to do the dishes, to do the cupboard) means to carry out some action involving that

object. But what the action is, is quite unspecified.

4. Disjunction in the specification of meaning is a type of vagueness that involves a rather

different kind of lack of specification: the cases where the meaning of an item involves the

disjunction of different interpretations. It may seem that the distinction between disjunct

specification within a single lexical item and cases of ambiguity characterised by discrete lexical

items remains quite unclear. Let us take as an example the verb RUN which is a word with

multiple interpretations:

13) He ran onto the field. The referents of the subjects are animate and have

14) He ran the race for Hampshire. → good legs.

15) The ball ran onto the field. → implication of fast directional motion.

16) The car is running well. → motion

17) The road runs from Manchester to Birmingham. → direction

18) He ran the motor show. → motion is generalized and transferred to an event in time.

Each of them provides a potentially distinct lexical item. If we say that RUN in these sentences

is the same lexical item, then our lexical entry for RUN will have to characterise the verb as

Page 55: Semantica Curs Ansamblu 2007

55

having a disjunct specification of meaning: either it has a meaning corresponding to its use in 13

and 14 above (where the referent is animate and has legs), or it has the meaning corresponding to

its use in 15, where presumably the implication of fast directional movement is retained, but not

any of the precise physical characteristics of running, or it has the meaning corresponding to its

use in 16 where only the implication of motion is retained, and so on. We thus have a 5-part

disjunction to characterise the meaning of RUN.

If on the other hand we take the alternative of specifying RUN as five-way ambiguous, we shall

have distinct lexical items, each with a distinct semantic representation. Characterised in this

informal way, it might seem that despite the conflict in claims as to what constitutes a lexical

item, there is little to choose between these alternatives, since both allow the word RUN to have

more than one interpretation. However, they lead to rather different predictions about sentence

ambiguity.

Disjunction within a single lexical item leads to a prediction that where more than one of these

disjunctions can be interpreted, then such interpretations should be possible simultaneously.

This is not so for most cases of multiple meaning. However, there are a few cases of lexical

items which have this property. To see the validity of this kind of characterisation, consider what

is perhaps the central example, i.e. OR.

19) The applicants for the job either had a first-class degree or some teaching experience

20) All competitors must either be male or wear a one-piece swimming costume.

In each degree of these cases, the implication that OR contributes to the sentence as a whole is

that one of the two conjuncts is true. In (19) the applicants are implied to have had either a firs-

class degree but no teaching experience, or teaching experience but not a first-class degree, or

possibly both, i.e. there is an interpretation in which both implications can hold simultaneously.

In (20) the implication is similar. The utterer of such a sentence would certainly not be excluding

the possibility of both conjuncts being true, for this would imply that a male competitor had

either to wear nothing or a two-piece costume! On the contrary, the sentence allows the

following types of competitors: male competitors (whether wearing a one-piece costume or not)

and non-male competitors wearing a one-piece costume.

An ambiguity test

In order to be certain of separating cases of vagueness from cases of ambiguity, we need a test

which distinguishes clear cases of vagueness and ambiguity, and which will give us some basis

Page 56: Semantica Curs Ansamblu 2007

56

for deciding on the less clear cases. Let us look first at the preliminary definition of ambiguity: a

sentence is ambiguous if it can be true in quite different circumstances. But this would predict

that in all cases where the meaning is unspecified, the sentence in question would be in as many

ways ambiguous as the contrasting circumstances which that unspecified meaning allowed the

sentence to be true in. There is an alternative, equivalent formulation of this definition: that a

sentence is ambiguous if it can be simultaneously true and false, relative to the same state of

affairs. But this alternative characterisation is no more helpful in unclear cases. For, suppose we

are uncertain whether a given sentence is ambiguous with respect to some contrast, or merely

unspecified as to that contrast.

Let us take, for example, the following sentence:

21) John killed Bill.

The question is: is this sentence ambiguous, one interpretation being that John killed Bill by

accident, the other interpretation being that John killed Bill intentionally? Or is the sentence

merely unspecified as to whether the action is intentional or not. The sentence can be used to

describe these two rather different kinds of events. In other words, it can be true in these two

different sets of circumstances. In fact, we distinguish lexically between MURDER and

MANSLAUGHTER which are specified as to intentionality on the part of the agent. Suppose

one linguist claims that sentence (21) has no specification in its meaning as to whether the action

implied is intentional or not. He would anticipate that the sentence would be true both in

circumstances in which the action was unintentional and in circumstances in which the action

was intentional. He would agree that the question of intentionality is simply not relevant to the

assessment of the truth of the sentence. But suppose also that some other linguist disagrees,

claiming that the sentence is ambiguous between an intentional interpretation and an

unintentional one. This linguist will say that in circumstances where the action was unintentional,

the sentence is true on the unintentional interpretation, and simultaneously, false on the

intentional interpretation. In circumstances where the action was intentional, the same linguist

would say, conversely, that the sentence is true on the intentional interpretation, and false on the

unintentional interpretation. Since, he would argue, this sentence meets the requirement that a

sentence be ambiguous if it is simultaneously true and false relative to the same state of affairs, it

must be ambiguous. Now these linguists have reached an impasse. The characterisation of

ambiguity as the simultaneous assignment to a sentence of the values TRUE and FALSE has not

provided a criterion for deciding unclear cases; it merely accentuates the point of disagreement.

Page 57: Semantica Curs Ansamblu 2007

57

For a more helpful way of distinguishing sentences which are ambiguous from those which are

not, we have to turn to anaphoric processes, processes which refer back to an earlier part of the

sentence. One example of this is the expression to do so, too. This is used where the action

described has already been specified and is being referred to again. Consider the following

sentence:

22) John hit Bill and Jason did so, too.

This implies that both John and Jason hit Bill. In more linguistic terms, the use of the expression

to do so, too demands identity of meaning of the two verb phrases in question. Now, this

provides us with a test for ambiguity. If some verb phrase is two-ways ambiguous, then we can

predict that when it is conjoined to a do so or other verb-phrase pro-form expressions (so did X,

X did/has/will/is, too), the entire sentence will be two-ways ambiguous. Whichever

interpretation is implied the do so expression must be identical to that interpretation. More

formally, a sentence which is two-ways ambiguous must be given two semantic representations

to characterise its two meanings. Since a do so expression or any other verb phrase pro-form

demands identity of meaning, a two-ways ambiguous sentence together with such an expression

can only be two-ways ambiguous. For example, we predict that the sentence:

23) John saw her duck and Will did so, too

is only two ways ambiguous. Either it means that John saw the duck which belonged to her and

Will also saw the duck which belonged to her, or it means that John saw her quickly lower her

head and Will also saw her lower her head. What we predict that it cannot mean is that John saw

the duck which belonged to her, and Will saw her quickly lower her head, because in such a case

the meaning of the two verb-phrases would not be identical. And so it is. Except as a pun (a pun

depends for its effect, at least in part, on breaking the convention that where a sentence or a word

is ambiguous, only one interpretation may be conveyed at any one time), there is no possibility

of such crossed interpretations when the verb-phrase do so is added to be ambiguous.

In the case of an unspecified or vague verb-phrase, we have a contrary prediction. Do so

expressions require identity of meaning, and where the meaning in question is unspecified with

respect to some contrast, there is no reason to expect that non-identical interpretations are

excluded. Consider the following example:

24) John is my neighbour and Sue is too.

Page 58: Semantica Curs Ansamblu 2007

58

The sentence does not imply that because Sue is also my neighbour she must have all the

properties that John has – being, say, a six-foot West Indian male. Or to take our previous

example of do the room (The painter has done the sitting room and the carpet man has too), this

does not imply that the carpet-man must also have painted the sitting room. On the contrary, the

natural interpretation is that different actions have been carried out.

If the expression to do so the sitting-room were said to be ambiguous as the actions it described

differed, we could predict that such different interpretations of the expression could not be

conveyed in the example above. But they can; and they can because the expression to do so the

sitting-room is not ambiguous, but merely unspecified.

To conclude, AMBIGUITY is a property of sentences: it is a ONE - MANY relation between

syntax and sense. It results from diversity of specific rules. VAGUE:ESS results when general

rules invite the ‘reading-in’ of indefinitely variable specific information.

EXERCISES I� SEMA�TICS �o. 1 1. This exercise is a class activity. Read the following paragraph:

The hunter crept through the leaves. The leaves had fallen. The leaves were dry. The hunter was tired. The hunter had a gun. The gun was new. The hunter saw a deer. The deer had antlers. A tree partly hid the antlers. The deer was beautiful. The hunter shot at the deer. The hunter missed. The shot frightened the deer. The deer bounded away.

Without changing important words or the meaning, rewrite the paragraph so as to avoid the many short, choppy sentences. Then compare the re-written versions prepared by the different members of the class. Are the paragraphs alike? If not, describe the differences and account for the fact that passages that appear in such varying forms have the same meaning. 2. Assume you are a native speaker of English. As such, you have an internalized knowledge of the language - call it a 'native-speaker intuition' if you will. For example, you can recognize a grammatical English sentence, you can interpret a sentence, you can perceive ambiguity, and you can determine when strings are synonymous. Examine the following groups of sentences. What can you tell about each group? a. 1. The bus station is near the bank. 2. The soldiers were told to stop marching on the parade ground.

Page 59: Semantica Curs Ansamblu 2007

59

3. The chicken is ready to eat. b. 1. That student continually sleeps in class. 2. Student in class continually that sleeps. 3. In class that student continually sleeps. c. 1. The Toronto Blue Jays beat the Atlanta Braves in the World Series. 2. The ones that the Toronto Blue Jays beat in the World Series were the Atlanta Braves. 3. The Atlanta Braves were beaten by the Toronto Blue Jays in the World Series. d. 1. Sam asked the students to build a display. 2. Sam promised the students to build a display. 3. Sam told the students to build a display. Present your conclusions. 3. The semantic differential was developed originally by psychologists as a method for semantic differentiation and determination of the connotations of words, and was made well known, especially by Charles Osgood. Joseph S. Kess suggests marking your 'impressions of a word on the following seven-point scale according to whether you view it as being, for example, extremely good, very good, good, neutral, bad, very bad, or extremely bad. Thus, if you had no feeling one way or the other about a word, you would mark the middle slot, indicating neutrality. Take for example, the word mother, and mark it according to the way in which the word strikes you as being meaningful on the following sample 'semantic differential' mother good ------ ;--------;-------; -------; -------; --------;-------; bad kind -------; -------; -------; -------; -------; -------; -------; cruel weak-------; -------; -------; -------; -------; -------; -------; strong beautiful -------; -------; -------; -------; -------; -------; -------; ugly nice -------; -------; -------; -------; -------; -------; -------; awful active -------; -------; -------; -------; -------; -------; -------; passive positive -------; -------; -------; -------; -------; -------; -------; negative heavenly -------; -------; -------; --------; -------; -------; -------; hellish reputable-------; -------; -------; -------; -------; --------; -------; disreputable large -------; --------; -------; -------; -------; --------; -------; small Different informants will inevitably use different intuitive criteria when making judgments about the same words. It follows, therefore, that a high degree of subjectivity is inherent in the semantic differential, which, paradoxically, is both its principal strength and its major limitation. Select two words, and ask two people to fill out a rating scale similar to Kess's. After tabulating the results, see what conclusions you can draw. For example, you could compare reactions to the words sick and ill, or woman and lady. If this exercise is done as a class project, then class members themselves should fill out rating scales. After the results are tabulated, discuss the results and draw conclusions about the meanings of the words that were evaluated. �o. 2: MOTIVATIO� OF MEA�I�G The lexicon of a language presents items which differ in the degree to which their meaning can be said to be motivated: while some are opaque, i.e. their sounds give no indication of their meaning, others are more or less transparent, in the sense that one can arrive at some idea of their meaning by recourse to their phonetic shape or derivational structure or to some semantic relations which can be established with other words in the language.

Page 60: Semantica Curs Ansamblu 2007

60

Practice: 1) Relying on your "basic" knowledge of Latin, try to analyze the extent to which the meaning on the following words is motivated (the first word is analyzed for you): Word Prefix Common meaning Root Common meaning percept pre- 'before' capere 'take, seize' offer insist aspect uncomplicated reduplication 2) What is the meaning of the prefixes in the following words (relative motivation by derivation): im-possible= inter-mittent= de-tain= mis-pronounce= trans-mutable= in-disposed= over-sufficient= 3) Idioms are word combinations that vary structurally and semantically.

Structurally, idioms are classified into: - free - semi-free/ semi-fixed categories

- fixed categories A) Analyze the following groups of idiomatic phrases and state the structural category they belong to. Motivate your choice. a)

To cut a poor Figure to make little fine

to come of age to b) to show the white feather to kick the bucket c) to smoke a pipe to have two cigarettes to buy several cigars to borrow a lot of pot to fill to light to re-light to puff away to put away ........................ Idioms are free to vary within certain limits. The degree of limitation of such a variation is given by the quantum of extensionality (QE). The QE shows the number of lexical items which

Page 61: Semantica Curs Ansamblu 2007

61

introduced in the idiom in the place of its elements will generate an equal number of idioms, in which the elements which have both been substituted retain their basic lexical meaning. Example: for the idiom under a) above, the QE is 2x1x3x1=6. B) Determine the QE for the other idioms above (b, and c) Semantically, idioms are classified into opaque and transparent.

C) Look at the following table! FORM\MEANING OPAQUE TRANSPARENT Fixed 1) 2)

A B 1) full structural integrity (no insertion & no substitution) 2) QE= 0

Semi-fixed 1) 2)

C D 1) some degree of structural integrity (limited substitution and insertion) 2) QE= very low

Free 1) 2)

E 1) no structural integrity (any substitution and any insertion is possible) 2) QE= extremely high

Give examples of idiomatic phrases for each of the above cases (A, B, C, D, E). 4) Give synonyms to the following idioms. Which have a transparent and which an opaque meaning? a. kick the bucket= b. out of the blue= c. get butterflies in one’s stomach= d. show the white feather= e. make a fool of oneself= f. change horses while crossing the stream= g. die in one’s boots= 5) Compound words What can you say with respect to the motivation of meaning of the following compounds? Bluebottle, blackbird, greenfingers, snow-white, blackleg, green belt, catgut, dead slow, hard-boiled, open-handed, cool-headed, straight-laced. 6) Here are some English onomatopoeic words. What or who does the following: bangs; howls; blares; plops; chimes; roars clicks; shrieks; crackles; squeaks; creaks; squelches; crunches; taps; gasps; thumps; groans; �o. 3: TYPES OF MEA�I�G A) CONCEPTUAL MEANING Linguistics as a supraordinate science studies the communication system. But communication implies interest in form and meaning. Everything in the world has form and meaning. Any change in form entails a change in meaning, analyzed at various levels.

Page 62: Semantica Curs Ansamblu 2007

62

Find examples of your own in which: a) a vowel change brings about a change in the meaning of words; b) a consonant change brings about a change in meaning; c) the change of a cluster of consonants determines a change in meaning (of larger constructions than a word): d) the change of stress determines a change in the meaning of the words; e) the change of stress and intonation .......................... B) CONNOTATIVE MEANING Find connotations for the following words: Cigarettes, baby, president, teacher, C) SOCIAL MEANING Analyze the following excerpts and state the approximate time when they were written and the register (province) they belong to: 'I returned and saw under the sun, that the race is not to the swift, nor the battle to the strong,

neither yet the bread to the wise, nor yet riches to men of understanding, nor yet favour to men of

skill; but time and chance happeneth to all'

'Objective consideration of contemporary phenomena compels the conclusion that success or

failure in competitive activities exhibits no tendency to be commensurate with innate capacity,

but that a considerable element of the unpredictable must invariably be taken into account' D) REFLECTED MEANING Give examples of words that used to have a 'neutral' meaning and that nowadays are sexually loaded. E) THEMATIC MEANING State the difference in meaning between: 1. a) Bill is addicted to morphine b) It is morphine that Bill is addicted to. 2. a) I like Danish cheese best. b) Danish cheese I like best c) It's Danish cheese that I like best. 3. a) They stopped at the end of the corridor. b) At the end of the corridor, they stopped. Do the same thing for the following examples in which the capitalized words represent the stressed words. 4. a) Harriet tickled Mavis in the CONSERVATORY. b) Harriet tickled MAVIS in the conservatory. c) Harriet TICKLED Mavis in the conservatory. d) HARRIET tickled Mavis in the conservatory. F) COLLOCATIONS Many intensifying adverbs and adjectives go together, e.g. absolutely fascinating, totally

destroyed. It is not easy to know which can go together. a) Sometimes there is a logical link between the adverb and the adjective. Match an adverb in column A with an adjective in column B: A B

Page 63: Semantica Curs Ansamblu 2007

63

fully disappointed deeply moved bitterly simple perfectly disturbed informed offended

b) Certain intensifying adverbs can only be used with gradable adjectives. Which of the adverbs in column A and the adjectives in column B go together? A B relieved important simple greatly annoyed terribly impressed disappointed offended c) Certain adjectives (called limit adjectives) already have very strong meanings. For example, exhausted means very tired. To intensify a limit adjective we need an extreme adverb. This is why we cannot say *very exhausted. Which of the adverbs in column A and the adjectives in column B go together? Sometimes more than one adverb is possible. A B delighted disgusted absolutely convinced totally appalled utterly determined quite obvious amazed untrue thrilled �o. 4: COMPO�E�TIAL A�ALYSIS 1) Make a componential analysis of the following words which belong to the semantic field of ‘human race”: man= woman= boy= girl= child= 2) According to the componential formulae, which is the antonym of woman? 3) Make a componential analysis of the word man in the following sentences:

a) Man is mortal. b) A man hit me.

What is the semantic relationship between the two meanings? 4) Make a componential analysis of the following sets of words and discuss the semantic relationship among the items:

Page 64: Semantica Curs Ansamblu 2007

64

a) horse c) dog stallion bitch

mare puppy colt filly d) cock hen b) sheep chicken

ram ewe lamb �o. 5: DISTRIBUTIO�AL A�ALYSIS Write the distributional formulae and the meanings of the lexical items in the following examples (Hornby Disctionary): Lexical item Lex. category Distrib. formula Example foster They fostered the child

The lady has always fostered the poor. He fosters high hopes. The nephew fostered revenge.

frame He framed the theory. They framed him. He framed his picture. fire He fired the workers. She fired the servant. They fired the building. graduate He graduated at Cambridge.

The school graduates 100 students each year. They graduated him with honours. He graduated the stick.

grasp He grasped my hand. He grasped the idea. greet He greeted us joyfully. The newcomers greeted them.

The sound greeted us from afar. hang They hung their coats. They are going to hang him. grave The situation is grave. The patient’s illness is grave. He is a grave man.

Page 65: Semantica Curs Ansamblu 2007

65

Her grave was under a willow tree. feel The boy felt the edge of the blade. The dog felt the burglars. He felt it was his duty to marry her. �o. 6. AMBIGUITY 1. Say what is wrong with the following sentences and rewrite them so that the intended meaning is clear:

a) Wearing a striped T-shirt, we thought he was rather casually dressed for an interview. b) Having been designed as a racing bike, I found it very smooth and easy to handle. c) Not looking where he was going, the car almost hit him as he was crossing the road. d) The letter presented no particular problems to translate, having learned German at school. e) Not being starving hungry, the huge helping of paella was more than I could manage.

2. Identify the types of ambiguity encountered in the following sentences: a) This is a good tool b) They are good students. c) The soup is good. d) Jane has never seen a triangle. e) That is a new plane. f) The American history teacher is quite nice. g) Mrs. Powel can’t bear children. h) Old men and women enjoy going to the theatre. i) Visiting professors can be boring. 3. Ambiguity in jokes. a) Artillery commander: ‘Fire at will!’ Recruit: ‘Where’s Will?’ b) Airhostess: ‘You’ll have to change twice, in Frankfurt and in New York before you go to Los Angeles.’ Nick (a cowboy): ‘Goodness me! And I’ve only got the clothes I am standing up in.’ c) Hungarian customer: ‘Waiter, have you got frog’s legs?’ Waiter: ‘No sir, I always walk like this.’ d) An old lady has just bought a postage stamp. ‘Must I stick it on myself?’, she asked. ‘Definitely not, madam’, replied the postal clerk. It will accomplish more if you stick it on the envelope.’ e) Three cars were reported stolen by the London police.

Page 66: Semantica Curs Ansamblu 2007

66

BIBLIOGAPHY

1. Chitoran, D. (1973). Elements of English Structural Semantics. Ed. didactica si pedagogica,

Bucuresti. 2. Cruse, D.A. (1986). Lexical Semantics. CUP 3. Gamut, L.T.F. (1991). Logic, Language and Meaning. Chigaco University Press 4. Hjelmslev, L. (1961), Some Reflections on Practice and Theory in Structural Semantics,

Copenhague 5. Hurford, J.R and Heasley, B. (1983). Semantics . A Coursebook. CUP 6. Jesperson, O. (1920) Language, London 7. Kempson, R, (1992). Semantic Theory. CUP 8. Leech, G. (1990). Semantics. Penguin Books 9. Lyons, J. (1977). Semantics. CUP 10. Ogden, C.K. & Richards, I.A. (1923), The Meaning of Meaning, London: Kegan Paul 11. Palmer, F.R. (1976). Semantics. A �ew Outline. CUP 12. Ullmann, S. (1962), Semantics: An Introduction to the Science of meaning, Oxford; Basil

Blackwell

Page 67: Semantica Curs Ansamblu 2007

67

Key to exercises: Unit 2. THE PROBLEM OF MEA�I�G 1) (One possibility) A tired hunter crept through fallen, dry leaves. He had a new gun. He saw a beautiful deer whose antlers were partly hidden by a tree. The hunter shot at the deer, but missed it. Frightened by the shot, the deer bounded away. By comparing your own versions of the text with those of your group-mates’, you will see that each of you used different ways of linking clauses together, by focusing on different aspects. 2) a. Sentences in this group are all ambiguous. b. A change in word-order brings about ungrammatical sentences. c. A change in word-order highlights a certain aspect of meaning (thematic meaning: see unit 3) d. A change of lexical items in the same paradigm brings about different meanings of the same sentence. 3) The semantic differential: this exercise should be done on your own and the results will be discussed at the tutorial. Unit 3. MOTIVATIO� OF MEA�I�G 1) Word Prefix Common meaning Root Common meaning percept pre- 'before' capere 'take, seize' offer ob- ‘against’ ferre ‘bear, carry’ insist in- ‘into’ stare ‘stand’ aspect ad- ‘to, towards’ specere ‘see’ uncomplicated un- ‘not’ con- ‘together with’ plicare ‘fold’ reduplication re- ‘again’ du- ‘two’ plicare ‘fold’ 2) im-possible = ‘not’ inter-mittent = ‘between’ de-tain = ‘away from’ mis-pronounce = ‘badly’ trans-mutable = ‘across’ in-disposed = ‘not’ over-sufficient = ‘too much’ 3) Idioms A) a) Semi-fixed idioms b) Fixed category (no substitution and no extension are possible) c) Free category (substitution and extension are possible) B) Determine the QE for the other idioms above (b, and c) b) QE= 1; c) QE= 9x4x4=144)

Page 68: Semantica Curs Ansamblu 2007

68

C) FORM\MEANING OPAQUE TRANSPARENT Fixed

A To show the white feather

B To take little action

1) full structural integrity (no insertion & no substitution) 2) QE= 0

Semi-fixed

C To change (swap) horses while crossing the stream

D To make (deliver) a speech

1) some degree of structural integrity (limited substitution and insertion) 2) QE= very low

Free

E To drink coffee /tea/ milk

1) no structural integrity (any substitution and any insertion is possible) 2) QE= extremely high

4) a. kick the bucket = to die (opaque) b. out of the blue = unexpectedly (opaque) c. get butterflies in one’s stomach = to be nervous (transparent) d. show the white feather = to be a coward (opaque) e. make a fool of oneself = to make yourself seem stupid by behaving in a silly or embarrassing way (transparent) f. change horses while crossing the stream = to make changes at the worst possible time) (transparent) g. die in one’s boots = to die young (opaque) 5) Compound words Bluebottle – not motivated; green belt - motivated Blackbird - not motivated; catgut – not motivated Greenfingers - motivated, dead slow - motivated Snow-white - motivated, hard-boiled – not motivated Blackleg – not motivated; open-handed - motivated Cool-headed - motivated straight-laced – not motivated 6) Onomatopoeia

- you bang a drum/hammer; - fire crackles - a radio blares - doors/floorboard creak - bells chime - crunch of footsteps on gravel/snow - click of fingers/camera - you gasp with amazement/shock - you groan if you are in pain - a wolf howls - plop of stone in a deep pond - roar of a lion - children shriek with excitement - mice squeak - squelch of shoes in deep mud; - thump of feet.

Unit 4: TYPES OF MEA�I�G A) CONCEPTUAL MEANING a) a vowel change brings about a change in the meaning of words: tip →top;

Page 69: Semantica Curs Ansamblu 2007

69

b) a consonant change brings about a change in meaning: top → pop; c) the change of a cluster of consonants determines a change in meaning (of larger constructions than a word): our dear old queen vs. our queer old dean d) the change of stress determines a change in the meaning of the words; I hate mankind (humanity); I hate mankind (men) e) the change of stress and intonation (But) …. That is your car, Õisn’t it? (near certainty) vs.

(Oh, so) that is your car, Ãisn’t it? (need for confirmation). B) CONNOTATIVE MEANING Cigarettes – harmful, expensive, staining your fingers and teeth. Baby – small, innocent, sweet-smelling, plump President – incompetent/competent, honest/dihonest, etc. Teacher – strict/permissive, fair/corrupted, etc C) SOCIAL MEANING Time: First text: Ecclesiastes (approximately 17th/18th century) Second text: G. Orwell ‘Politics in the English Language’ (20th century) Province: Language of the bible Language of journals. D) REFLECTED MEANING Give examples of words that used to have a 'neutral' meaning and that nowadays are sexually loaded. Cock (bird), rubber (material/eraser), ejaculation (utterance) E) THEMATIC MEANING 1. a) Bill is addicted to morphine b) It is morphine that Bill is addicted to. – the speaker is presupposing that Bill is addicted to something and asserting that that something is morphine (SPEAKER PRESUPPOSITION VS. ASSERTION) The meaning of a sentence is said to be divided between the part that a speaker asserts and the part that he presupposes, or assumes to be true. 2. a) I like Danish cheese best. (They answer different questions): a. What do I like? b) Danish cheese I like best b. Which cheese? c) It's Danish cheese that I like best. 3. a) They stopped at the end of the corridor. Where did they stop? b) At the end of the corridor, they stopped. What did they do at the end……? F) COLLOCATIONS a) bitterly disappointed; deeply moved; perfectly simple; deeply disturbed; fully informed; deeply offended. b) greatly relieved, terribly important, terribly simple, greatly/terribly annoyed; greatly/terribly impressed; greatly/terribly offended; terribly disappointed. c) absolutely delighted; absolutely/utterly/quite disgusted; absolutely/ totally/ quite/utterly convinced; utterly appalled; absolutely/ totally/ quite obvious; absolutely/ quite/utterly determined; absolutely/utterly amazed; totally/quite untrue; absolutely/utterly thrilled. NB. Absolutely can go with nearly all limit adjectives. Other collocations have to be learned.

Page 70: Semantica Curs Ansamblu 2007

70

Unit 5: COMPO�E�TIAL A�ALYSIS 1) man = [+ Animate], [+Human], [+Male], [+Adult] woman =[+ Animate], [+Human], [-Male], [+Adult] boy =[+ Animate], [+Human], [+Male], [-Adult] girl =[+ Animate], [+Human], [-Male], [-Adult] child =[+ Animate], [+Human], [+/-Male], [-Adult] 2) Woman (and boy) 3) Make a componential analysis of the word man in the following sentences: a. Man is mortal. [+ Animate], [+Human], [+/-Male], [+/-Adult] b. A man hit me. [+ Animate], [+Human], [+Male], [+Adult] The semantic relation between the two meanings of man is hyponymy (meaning inclusion): the meaning of man (human being) is included in the meaning of man (adult male human being). Unit 6: DISTRIBUTIO�AL A�ALYSIS Lexical item Lex. category Distrib. formula Example foster verb H+foster+N They fostered the child

The lady has always fostered the poor.

H+foster+NonC He fosters high hopes. The nephew fostered revenge.

frame verb H+frame+NonC He framed the theory. H+frame+H They framed him. H+frame+ He framed his picture. grave adjective NonA+Be+grave The situation is grave. The patient’s illness is grave. H+Be+grave He is a grave man. �oun grave +Be + loc.adv. Her grave was under a willow tree. Unit 7. AMBIGUITY 2. Identify the types of ambiguity encountered in the following sentences: a) This is a good tool b) They are good students. Referential versatility c) The soup is good. d) Jane has never seen a triangle (lexical). e) That is a new plane (lexical). f) The American history teacher is quite nice (syntactic). g) Mrs. Powel can’t bear children (lexical). h) Old men and women enjoy going to the theatre.(syntactic) i) Visiting professors can be boring (lexico-syntactic).