M

w
H

HWiiilliil

[HIM

HHIIEH

‘l
1

40).;
Iooo
mc04>

 

SYMRGMATM} MED PAPJXD§GMM§C RELANQNS
tN LEXIS

Ther' :3 fart fix: Segree of M A
M CH GAN STATE. UNIVERSHY
t..JMHRU l. CfiiTuRAN

1. 3‘15 7.?

_ _, Inc-v -*.._

{J ernAnY

_. I , ‘ .
1.! L131» cm, by

;.. __._
~ "‘9' -

  
  

 

SYNTAGMATIC AND PARADIGMATIC

RELATIONS IN LEXIS

By

Dumitru I. Chitoran

A THESIS

Submitted to
Michigan State University
in partial fulfilment of the requirements
for the degree of

MASTER OF ARTS

Department of Linguistics and Oriental and
African Languages

1967

 

TABLE OF CONTENTS

1. IntrOduCtiOn O0..0..OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO...

 

1.1. The Present Situation of Semantics as a

Branch of Linguistics .....................................
1.2. The Problem of Meaning ....................................
1.}. Object and Aim of the Thesis ..............................

2. The Manifestation of Syntagmatic and Paradigmatic

 

Relations in LeXiS .OOOOOOOOOOOOO0.0.00000COOOOOOOOOOOOOOOOOO

 

2.1. COllocation and Set .0..0...COOOOIOOOOOOOOOOOOOOCOOOOOCOOOC
2.2. COllocation a.nd Mea‘niHSOOOOOOOOOO0.00000COOOOOOOOO00.0.00...
2.5. COllocation and COlligation O..0OOOOOOOOOOOOOOOO00.0.0000...

3. Distributional AnalySiS Of Meaningr oooooooooooooooooooooooooo

 

5.1. The Relation between Grammar and Lexicon ..................
5.2. Method of Approach and Terminology ........................
3.5. Distribution and Meaning ..................................
5.h. Distributional Analysis of the Structural

Meanings of the Verb 'Make' ................................
5.5. More Refined Meaning Distinctions Revealed

by Grammatical Operations .................................

h. Paradigmatic Relations ......................................

 

h.2. The Semantic Field of English Factitive Verbs .............

5. componential AnalISiS 0.0.0.....00.0..0...OOOOOCOOOOOOOOOOOOOO

 

5.1. Semantic Components of a Lexical Item .....................
5.2. Semantic Components of a Lexical Field ....................

6. COIlClqu—Ons .OOOOOOOIOOOOOOOO0.00000000000000000000.0.0000...

 

BIBLIOGRAPHY OOIOOOOOOOO0.00000000000000000000000000000000000000C

Page

23
2A

26
28

52
39

42

42
43

51

51
52

52

5h

1. Introduction

 

1.1. Thegpresent situation of semantics as a branch of linguistics.

 

Recent develOpments in linguistic research prove that the crisis
of semantics, the Open neglect of meaning, is already a thing of the past.
The increasing number of books and articles in specialized periodicals
which deal with problems of semantics, as well as the diversity of
semantic theories and approaches point to a state of effervescence, of
constant searching, the result of which will undoubtedly be genuine
progress in semantics, this still very young branch of linguistics.

Among the most important gains in modern semantics I would include
the following:

a) It is becoming increasingly clear that the meaning of a linguistic
form can no longer be regarded as an indivisible entity; on the contrary,
it is desirable to consider meaning as a bundle of semantic distinctive
features, not unlike the phonetic distinctive features used in phonology
to denote the component elements of the phoneme.

b) Every serious semantic analysis along structural lines seems to
indicate the existence of some complex interlocking and overlapping
systems of interrelationships among the units of the lexicon. The nature
of these relations is at present very difficult to grasp fully and
consequently to formalize, but significant steps have already been made

along these lines. Equally Open is the question of whether the above

mentioned system of relationships extends over the entire lexicon,
resulting in a systematic organization of the whole vocabulary.

c) There exists an isomorphism between the two levels of language
defined in glossematicslz the Expression level and the Content level.
This makes possible the application of similar methods and techniques
of analysis, and furthermore ensures a unified character of the general
linguistic theory.2

It would be very difficult to assess the value of each new theory
of semantics that has been put forward in the last decade or so. Given
the complex nature of meaning, it can hardly be the case that any one
particular linguistic theory can lay claim to final truth. Indeed, it
would be more plausible to believe that many of these theories put
forward both valid and unvalid generalizations, both tenable and untenable
assertions, thus contributing simultaneously both to clarifying and
obscuring the problem of meaning - the most debated problem not only in

linguistics but also in a whole range of related sciences.

1.2. The Problem of Meaning.

 

Any real progress in semantics is conditioned by an understanding
of meaning, the object of its analysis. Given the very complex nature
of meaning, it is very difficult and practically impossible to give a
complete definition of meaning without reference to logic, psychology
and the theory of knowledge, in addition to linguistics prOper.

None of the various definitions that have been given to meaning is

adequate for linguistics or for any other science dealing with meaning.

 

lHjelmsev, L.: (1961)

2Saumjan, s. K.: (1962)

All of them are both too wide for the particular needs of a specific
science and too narrow to cover the complexity of meaning, with its
diversity of elements and variable factors. One way out of the difficulty
might be to give separate definitions of meaning suitable for the needs

of each particular science. This is in fact what is often done. Psychology,
logic, anthrOpology, literary criticism, information theory, etc., give
different accounts of meaning. What one discipline regards as disputable
and in need of further exploration, other disciplines take for granted
and use as a basis for their entire theoretical edifice. This situation
is, to my mind, highly characteristic at present of a group of related
disciplines which try, following their own methodologies and theoretical
apparata, to account for one and the same phenomenon: human linguistic
activity.

Linguistics however, unlike the other kindred sciences seems to
have been concerned from the very beginning with matters which at closer
analysis would seem to fall beyond its actual boundaries. Hence, the
persistent concern of modern linguistics to limit itself to matters that
are "prOperly" linguistic and leave other matters to the apprOpriate
sciences. It has been discovered however, that the related sciences
(psychology, logic, semiotics, etc.) are themselves reluctant or, even
if willing, unable to COpe with the complex problem of meaning. Linguistics
in its turn faces a situation requiring either complete dismissal of
meaning - which is what most of descriptive linguists did for a while -
or, the continuation of the traditional concern for 'total' meaning.

Semiotics and information theory have brought a significant
contribution to the clarification of many problems related to meaning,

and linguistics has benefitted from it. The following ideas about

linguistic meaning seem to be common today in modern linguistics; of
them, some are complementary, while others are mutually exclusive:

- Meaning, as a property of linguistic signs, cannot be identified
with the notion, the concept or the object for which linguistic signs
stand; meaning is essentially a relation (Ogden, Richards, Stern,
Smirnickij, Ullmann) between the physical (sound), and the ideal (mental)
aspects of the linguistic sign; between the 'signifiant' and the
'signifie', to use de Saussure's terminology.

- The meaning of linguistic forms can be determined by the position
they occupy in the syntagmatic chain. This notion, derived from
Wittgenstein, has been applied to linguistics especially by Firth,

Harris, Wells, Apresjan etc.

- The meaning of linguistic forms can be determined by the position
they occupy in the system of equivalent linguistic forms - in the
paradigmatic series to which they belong. The most ambitious theory
that has ever been put forward to account for meaning in terms of paradigms,
is Jost Trier's theory of the semantic field. The more recent approaches,
such as Osgood's theory of semantic space, componential analysis, the
thesaurus approach etc., are further develOpments of Trier's ideas.

- Meaning has to be conceived as a complex entity, decomposable into
simpler elements. By performing such an operation, and also by taking
into account the gnoseological, logical and psychological factors
involved, the following aspects of meaning can be delimited: a) Structural
meaning - the relation between a sign and other signs. The relation is a

double one, as it extends along two axes:

 

1Apresj’an, Ju: (1965)

5

- along the syntagmatic axis - accounting for syntactic meaning
- along the paradigmatic axis - accounting for differential
meaning, for "la valeur" in Saussurean terms.

b) Significative meaning - the relation between signs and 'significata'.
c) Denotative meaning - the relation between signs and 'denotata'. (b)
and (c) are studied extensively by philOSOphers (e.g. Carnap, Russel,
Morris, Church). According to Charles Morris, significative meaning is
identical with extensional meaning; denotative meaning is identical with
intensional meaning. For linguistics, denotative and significative
meaning represent together what is traditionally called lexical meaning.
d) Pragmatic meaning - the nonlinguistic reaction to a linguistic stimulus.
It includes also the emotional and expressive (connotational) aSpects of

meaning.

1.5. Object and Aim Of the Thesis.

 

In terms of the types of meaning discussed under (1.2.), this thesis

will deal with structural meaning, which is considered tO be the prOper

 

Object of study of structural semantics.l Based on the underlying
isomorphism Of the two levels Of language - Expression and Content - an
attempt will be made to employ the two fundamental types Of linguistic
relations - syntagmatic and paradigmatic - as a starting point for

constructing a general theory of lexis.2

 

lApresjan, Ju: (1965)

The isomorphism Of the two levels Of language is far from being
unanimously accepted (cf. the reports and discussions at the 8th
International Congress Of Linguists in Oslo, on the tOpic "TO what extent
can meaning be said to be structured"); Opinions differ from complete
agreement: Saumjan, S. K.: (1962), Prieto, L.: (1961), Coseriu, E.
(1965), etc., to partial agreement: Lamb (196A, 1966), or total rejection:
Wells (1958), etc.

In the first part of the thesis, examples are given to illustrate
the manifestation Of syntagmatic and paradigmatic relations within
lexis, while the second part is devoted to an attempt to use them for
the distributional analysis of meaning, the establishment of semantic
fields according to Objective criteria and the assessment Of the possi—
bilities that such an approach Opens to a componential analysis of

meaning.

2. The manifestation of syntagmatic and paradigmatic relations in lexis.

Following de Saussure, Hjelmslev and more recently Saumjan and
Lamb, the primary Object Of structural linguistics may be taken to be
the network of language relations, while the terms of these relations
are considered to be secondary.1

The relational frame Of language has two axes - the syntagmatic
and the paradigmatic. The linguistic elements situated on the syntagmatic
axis find themselves in a 'both - and' relation; they coexist within the
Spoken or written chain; they find themselves - tO use Martinet's
terminology2 in 'contrast' with one another. This type Of relation is
directly observable in the utterance.

On the paradigmatic axis, linguistic elements exclude each other
in a given utterance, they are in an 'either - or' relation or, again
in Martinet's terms, they are in 'Opposition'. Unlike contrastive
relations, the Opposition relations are not directly observable in the

utterance.

 

lSaumjan, S. K.: (1962)

2Martinet, A.: (1961)

Any linguistic element situated in a certain position along the
syntagmatic axis may then be described by means of a vector of two
parameters - the syntagmatic and the paradigmatic. The nature Of the
two types of relations has been studied more extensively with relation
to phonology and grammar. If there is indeed an isomorphism between
expression and content, as postulated by glossematics, it is plausible
to assume that the same fundamental types of relations manifest them-

. O O O I 1
selves in an identical manner in the lex1con as well.

2.1. Collocation and Set

 

On the basis of these two types of relations one could postulate
two theoretical categories necessary for the description of lexis. We
could, of course, refer to them as "lexical syntagm" (for the first
type) and "lexical paradigm" (for the second), the two terms revealing
in this way the nature Of the relations they are called on to define.
However, both are overburdened terms in contemporary linguistics, with
primarily grammatical pertinence, and consequently we are reluctant to
use them as basic categories of a theory Of lexis. Instead I shall use
the terms suggested by M.A.K. Halliday: collocation and set. Collocation
corresponds to syntagmatic relations; set corresponds to paradigmatic ones.
The term collocation - seemingly used only by British linguists —
is not a new term in linguistics but the meanings given to it differ

greatly. In most cases collocation has been understood as being synonymous

 

1The necessity for such studies has been repeatedly pointed out;
suffice it to give the following quotation: "It is apparent...that
syntagmatic and paradigmatic relations are nearly as important in
dealing with the lexicon Of a language as they are at the levels Of
grammar and phonology". Robins, R. H.: (1965) p. 75.

with syntagm (another undefined notion), or with any patterned grouping
or arrangement Of words.

Another meaning given to the term - to be found mostly among those
who follow J. R. Firth's theory of language1 — limits collocation to
the selectivity displayed by certain lexical items in combining with
certain other lexical items in order to form utterances. Hence the notion
"expectation of collocation" and "collocability" which are used to denote
the combinatory valences Of lexical items. Thus for MitChell2 an element
"ditch water" does not occur apart from "dull". Consequently "dull as
ditch water" is a collocation. The idea is useful for the assessment Of
the real meanings of words. J. R. Firth asserted that one part of the
meaning of lexical items is their collocability with certain other
lexical items. "One of the meanings of 'night' is its collocability
with 'dark', and of 'dark', of course, its collocability with 'night'."3
There is nothing wrong with this assumption, based on the Wittgensteinean
definition Of meaning as the sum total of its uses in languages,h were
it not for the practical impossibility Of dealing with meaning in these

5

terms exclusively. But that the combinatory valences of words, their
collocability with other words, is part Of their meaning, cannot be

doubted in any possible way. Thus in a dictionary6 under the entry

 

lFirth, J. R.: (1957).
2Mitchell, T. F.: (1958) p. 103.
3Firth, J. R.: (1957)

hWittgenstein, L.: (1961)
SChomsky, N.: (1966)

6The following dictionaries were used in discussing various words in
this thesis:

-The Shorter Oxford Dictionary (3 o D) C. T. Onions ed, Oxford
University Press, London 1961.

'maiden' whose basic, original meaning 'girl' is Specified as Obsolete

and archaic, the item is said to be used almost exclusively in collocations

such as:

a) maiden aunt where the general meaning is "unmarried"
maiden lady (maiden name = name of a woman before
maiden sister marriage)

maiden name

b) maiden ship In all these collocations the general
maiden voyage meaning of "maiden" is "inaugural".
maiden speech

maiden horse

Evidently the meanings of a) and b) are based on a metaphor, but the
actual disappearance of the original meaning from colloquial use makes
them part and parcel Of the collocations rather than of the metaphorical

extension.

2.2. Collocation and meaning.

 

One method Of analysing meaning in terms Of collocations is to
present them as textually quantifiable by virtue Of a law expressing the
probability that "there will occur at 'n' removes (a distance Of 'n'

lexical items) from an item 'L x', the items a,b,c..."l

 

-Webster's New World Dictionary of the American Language,
College Edition, The World Publishing Company, New York
196A.

-A. S. Hornby, et a1 :' A Learner's Dictionary of Current
English, Oxford University Press, London 1948.

lHalliday, M.A.K.: (1961) p. 276.

10

The items with which Lx may be collocated (a,b,c...) form a 'set',
an Open grouping of lexical items mutually exclusive in the given
utterance. In a set we usually find members Of the same grammatical
class. Thus, from the class "adjective", all those designating colours
can be used in the collocation:

"The ....... dress suits her best".

Similarly, from the class "noun" all those denoting 'food' can be used
in the collocation:

"We had .... today at .... " (The last position can, in turn,
be filled by any member Of a set grouping together nouns denoting the
meals Of the day).

It does not follow Of course, that members Of other sets are not allowed
in these collocations. The frame: 'We had .... today at lunch' allows
equally well members of a very large lexical set including 'human' nouns:

We had my sister today at lunch

the mayor

 

a musician
However, with elements Of the new set, the verb 'have' acquires a different
meaning than it has in: 'We had meat today at lunch'.
On the other hand, it is necessary to point out that sets are not arrived
at entirely in a mechanical way. In (2.5.1.) and also in (h.l.) I
describe the manner in which sets can be delimited by an interplay of
semantic and grammatical factors.

Graphically, the collocation and the set can be represented in the

following manner:

 

lHalliday, M.A.K.: (1961) p. 276.

 

11

L x Ll'°°°"°°L2""°'°°L5""°'°°Ln a more probable
(a lexical (other lexical items, each b
item) representing a 'remove'
from L x) c
V/

 

n less probable

- less delicate 3) more delicate

 

As can be seen from the chart the greater the number of removes
from Lx to the collocated lexical item, the more 'delicate' is the
collocation. In their turn, the terms of a set can be arranged in
relation to a given collocation along a gradient fnom "more probable"
to "less probable".

One very simple way Of deciding on the appurtenance of various
lexical items to one lexical set, has been pointed out by Firth: "In
the study of vocabulary elements it is important to group together those
forms which form a series. Such series for example might be kinship
terms, parts of body, terms Of orientation in time and space, numerals,
calendrical terms, names of social units, prOper names of persons as
well as places, in short all lexical elements which exhibit structure".1
His name for such sets was 'Ordered series of words' and their relation
to the various possible associations within lexis mentioned, by other
linguists starting with Saussure is evident. One ultimate goal of such
a grouping is to somehow reflect the way the elements of lexis are stored
in our minds. In the second part of this thesis, a method will be
suggested to delimit sets in an objective manner based on structural

criteria.

 

1Firth, J. R.: (1957)

12

The other important thing about collocation is the close inter-
relationship which Obtains between the collocated lexical items. This
interrelationship manifests itself as the Specification Of the meaning
of one lexical item by another lexical item. The two terms of a collo-
cation may be called determining (d) and determined (D), in which case
collocation appears as a larger linguistic unit, above the word, the
primary function of which is to specify the various meanings of lexical

items.

2.5. Collocation and Colligation.

 

Collocation can be described as the habitual association of a
lexical item with certain other lexical items in utterances, so that a
constant meaning relation Obtains between the collocated lexical items.
Both lexical items are considered in their individual capacity, no
attention being paid to the grammatical classes to which they belong.

When grammatical factors are taken into consideration, a different
type of structure will be revealed. The term collocation in lexis will
be paired by another term 'colligation' in grammar. In the following
incomplete utterance: "John + verb + Mary for...." colligationally we
can fill in the blank with terms in several grammatical classes. The
number of terms in any of these classes is potentially unlimited.
Collocationally however, a choice Of a member of the class 'verb' will
limit considerably the choice of the items that can fill in the blank.
If the verb 'take' is chosen for instance, we can have the blank filled
in mainly with members of two sets of nominals:

'John takes Mary for Joan'

'John takes Mary for a walk'

13

'Joan' is a member Of the subclass of human nouns (substitutable by
'he', 'she', 'they').

'Walk is a member in a more restricted set which includes nominals
denoting human activities, related mainly to the kind Of leisure
implying movement (trip, holiday to Spain, etc.). As is only to be
expected, when the lexical item 'take' is collocated with elements Of
the first set it has one meaning 'mistake for'; when it is collocated
with elements in the second set, its meaning is 'to bring with, to cause

to come with'.

2.5.1. We shall try to exemplify these assertions by means Of a
highly polysemantic word: "blind". In all combinations that follow,
"blind" will be the determined part of the collocation the item
surrounding it, being its determiners. The analysis will try to
determine:

a) whether the meaning of 'blind' can be said to remain essentially
the same when its determiners are members Of the same set (or ordered
series Of words),

b) whether the change Of meaning expected whenever we go from one set
to another actually takes place,

c) whether there is a prOpensity for 'blind' to combine with some lexical
items rather than with others, in short, whether there really exists
an expectation of collocation.

The lexical item "blind" is combined with elements in 6 sets, each

characterised by a common semantic feature:

14

(l) blind person (2) blind hand
man type
girl print
boy track
bird patch
wolf
horse F
(5) blind alley (A) blind action (5) blind side i
tube passion part i
pipe procedure L
gut choice
shaft love (6) blind poppy
fury daffodill

The elements in set (1) denote persons and beings possessing eyes with
which they can see; those in set (2) denote objects which by their
nature should be distinct, visible; and places which should be lighted
and visible. Lexical items in set (5) stand for tubular or long linear
Objects which usually have two ends, while those in (H) stand for actions
and emotions. Finally, the lexical items grouped under (5) denote
components of human personality, and those under (6) are names of
flowers and plants. With each of the six sets there is definitely a
change in meaning of 'blind'. Thus in collocations with lexical items
Of the first set, the meaning of 'blind' can be defined as: 'not able
to see, having completely or partially lost sight'. With terms in the
second set it means 'invisible', 'dark', 'hardly visible'. With members

of the third set, there is again a new meaning: 'filled at one Of the

15

ends', 'lacking one of the orifices'. The same thing happens with set
number four, where the new meaning is 'not well judged', 'unreasonable',
'the result of feeling or temper rather than Of thought', while with
the fifth set the meaning is 'weak, vulnerable', and the meaning for
the sixth set is 'without buds or terminal flowers'.

We can see that at least points a) and b) seem to be corroborated
by the evidence found in a dictionary. Moreover, the meaning seems to
remain the same when we paraphrase grammatical transformation to various
collocations. Thus there will be no change of meaning for 'blind' if

we change: 'I saw a blind man' into 'The man I saw was blind'.

2.5.2. However, one major Objection can be raised to the scheme.
Such possible uses of the same lexical item in collocations such as:
'He is blind to his son's defects', or 'Help the blind', or 'blind man's
buff' and 'The room had the blinds pulled down', etc., are left
unaccounted for.

The way to analyse them is to turn to grammar, that is to make
use Of grammatical distinctions which will prove relevant to the meanings
of 'blind'.

A sentence like 'He has always been blind to his son's defects'
displays a different grammatical structure; it does not allow the
passive transformation: 'His son's defects have always been blind to
him'. One solution is to regard as a Single lexical item the whole of
'be blind tO' with a set of determining items naming generally 'qualities
and defects in human character'. The lexical items in the set appear
in utterances accompanied by various grammatical modifiers - this will

not change the meaning of the lexical item 'be blind to'.

v .I' a. F,

16

Ex: 'He has always been blind to faults
his faults
his son's faults

As in the case Of paraphrase this seems to go against Halliday's
assumption that the distance between two collocated lexical items is
relevant in the sense that each additional remove means an increase in
delicacy.1

The lexical item 'blind' in 'Help the blind', can be defined by
means of colligation. It will be found in utterances such as:
'Help the blind.‘
'The blind need our sympathy.‘
The lexical item is invariably accompanied by a determiner. Now, we
need no longer Speak of sets, we define 'blind' with this particular
meaning in terms of the new grammatical category to which it belongs.
Its nominal nature is revealed both by function and position. We shall
also incorporate here, the meaning of 'blind' in collocations such as:

blind hospital

asylum
alphabet

which can be paraphrased by the more explicit form:

hospital

asylum for the blind

alphabet

 

1In fact, the real distance between the collocated lexical items
remains the same. I believe that in relation to the concept Of 'delicacy'
in lexis, a distinction should be made between '1inear ordering' and
'structural ordering' Of lexical items.

17

The conversion into a noun of the lexical item 'blind' is complete (as
indicated by the use Of the plural marker) in utterances such as: "The
room had the blinds pulled down." "There were no blinds at the windows

Of the house", etc. "The horse has blinds" (in British English) as

Shown by its making use Of the plural Sign. All such uses Of the lexical
item 'blind' can be covered by one general meaning: object that obstructs
light or sight. Further distinctiOns: (a) 'blind' + 'room', 'house',
'castle', etc, and (b) 'blind' + 'horse' will reveal two more meanings:
(a) 'screen for window made of woven material and mounted on a roller'
and, (b) 'blinker'.

In analysing 'blind man's buff', we start from the definition Of
the entire collocation as it is found in a dictionary ('a game in which
one player is blindfolded and tries to catch and identify any one Of
the others, who on their part push him about' (cf. The Shorter Oxford
Dictionary). The meaning of 'blind' is Of course 'not able to see' but
its relevance for the utterance as a whole is lost. That is why the only
possible way of analysing the entire collocation is to consider it as
being a fixed one in which the interdependence between the component
terms is total. The three items are related syntagmatically, but no
sets correspond to any of them. Now, it is the whole Of 'blind man'
which is to be understood as the determined part (D) Of the collocation;
and 'buff' as the determining one (d). Graphically this can be represented

as:

 

blind - man's - buff

 

 

D d

 

 

In a dictionary we find another meaning given to the same linguistic

18

form 'blind man': a post Office Official who deals with blind letters.
(Of. The Shorter Oxford Dictionary). To solve the problem we have again
to consider 'blind man' as one lexical item. Man, here, is a term in a
restricted set, (including 'reader' and 'Official'), but none of them
Operates as determiners, specifiers of 'blind' with the meaning postulated
above. The determiners should be looked for in the immediate vicinity
of the linguistic form namely lexical items forming a set denoting Objects
related tO the postal service. Purely linguistic factors (items in the
same text or chain Of speech) will prove insufficient to reveal the exact
meaning: 'Who's the blind man in the post Office?' may be interpreted
either 'Who's the official in this post office dealing with blind letters'
or 'What is the name Of this man who cannot see and who is now in this
post Office?'. In such situations it is the extra—textual factors, the
actual speech situation which will intervene to make communication
possible.1

Contextual factors do in fact intervene in all acts of speech
confering the particular shade of meaning to linguistic forms or to
utterances as a whole which corresponds best to the actual message to be
communicated. There is however a difference between this latter function
of context, when, to an already existing meaning, new shades are added,
and the situation when context is the only element that acts as determiner
of the meaning of a certain linguistic form. (see 2.5.5)

Another lexical item "sick" (as used in British English) may be

said to have two general meanings, again identifiable by collocation and

 

1A distinction has been suggested between 'co-text', which is the
linguistic context prOper (items in the text), and 'context', which
stands for the situational factors that prove to be linguistically
relevant. cf. Catford, J. C.: (1965) p. 51.

l9

colligation. Meaning one: 'suffering from illness of any kind' - 'ill',

'unwell' - is revealed in collocations Of this type:

1) Sick man (where the set is represented by lexical items denoting
person persons and beings or parts of their body)
child
finger

2) sick room (where the set is represented by lexical items denoting
bed spaces or objects occupied or used by ill persons)
berth
bay
cookery

5) sick allowance (where the set is represented by lexical items
leave denoting facilities created for the sick)
club

A fourth set will be represented by a limited number Of verbs

to get sick

to report sick
Meaning two of the same lexical item: 'having an inclination to vomit'
or 'actually vomiting' can be defined by means of colligation. When-
ever 'sick' will be found in a predicative position after such verbs as:
to be, to feel, to seem, etc., (verbs Of state), it will have meaning
number two. From the point of view of style, one could bring in addi-
tional information by specifying that, while meaning number one, although
the primary, basic meanining Of 'sick' is preserved only in official or
Special language and in the above mentioned collocation, which are very

limited and almost fixed, meaning number two has become the most important

 

 

20

(in British English) and has subsequently developed another figurative
meaning: 'mentally affected' or 'weak', as in the collocations: 'to
be sick (and tired) of...‘ (The same can be said regarding the meanings

of the derived verbal form 'to sicken').

2.5.5.‘ This is one more example of the difficulties encountered
in attempting to analyse meaning in the usual way: from a primary (basic)
meaning, numberous 'secondary' meanings are develOped by means of:
radiation, concatenation, extension of meaning, narrowing Of meaning,
degradation of meaning, elevation of meaning, etc. Thus in his 'Course
in English Lexicology' Leon Levitschi presents the meanings of the
lexical item 'leg' by means of a graphical scheme:l

leg: meaning 1 (one of the parts of the body by means of which
men and animals stand and walk)

2 meaning 2 (the part Of the garment covering the leg)

1 leg<:;—fi>5 meaning 5 (a bar or pole used as a support or prop)

1

5

h meaning 4 (one Of the supports Of a piece of furniture)
meaning 5 (one Of the branches of a fork or jointed Objects)

All meanings are to be understood as having develOped historically
from the first basic meaning. Meaning four, for instance, is treated
as a metaphor - a lexical metaphor. But if we are to consider the
existance of a general (basic) meaning of the lexical item 'leg', why
not consider it to be 'thing that supports body or Objects' which will
incorporate all meanings (except 2 and 5). In that way, meaning four
will no longer be a metaphor. A lexical item, when taken in isolation

may have a very vague meaning which will be clarified by its use in the

 

lLevitschi, L.: (1965) p. 72

21

utterance, and it is precisely the unclear character of the meaning of
a lexical item that makes possible its varied use in language.

Psycholinguists use the term 'nucleus meaning' to denote this

 

vague meaning displayed by lexical items even when uttered in isolation.
The various significations of a lexical item are organized around its
nucleus meaning and form an inner system of the respective lexical item -

its microstructure.2 Thus the nucleus meaningof 'leg' is 'One of a set

 

of supports on which a body (or Object) can stand'; its various meanings
(those to be found in the dictionary - including the meaning 'limb')

are actualized by uses in various utterances. There are no well estab-
lished criteria according to which we could class certain meanings as
basic and others as secondary. A certain hierarchy could be established
within the microstructure Of a lexical item by taking into account the
amount and importance of contextual or co-textual factors needed for the
evocation of the respective meaning. The multiple meanings Of a lexical
item can be arranged along a gradient from meaning 1 - that meaning for
whose evocation the least context is required - to meanings 2, 5, 4...n,
each requiring an increasing degree of contextualization for their
evocation. The meanings of a lexical item for whose evocation less
context is required are also its most frequent ones. Thus for 'blind',
the obvious meaning on tOp of the scale is, of course, 'unable to see'.
According to the Thorndike semantic frequency count published in West's
"General Service List of English Words" it is also the most frequent one

covering 42% of its total uses. Second (25%) is the 'figurative' meaning

 

1Rossetti, A.: (1945) p. 54

2Slama-Cazacu, T.: (1957)

22

'not able to grasp, unreasonable'.

In the pair:

'You have a ladder on your right stocking'

'You have a ladder on your right',
it is the amount of context required by the lexical item 'ladder' that
differentiates its meaning. The collocational Spread of the lexical
item 'ladder' when meaning 'a flaw in a stocking, due to one or more
slipped threads' (in British English) is limited. Here the lexical item
'stocking' acts as a 'key' in revealing this particular meaning. On the
contrary the collocational spread of the same lexical item with its
other meaning is much wider. The more frequent meanings, those which
need the least context for their evocation, have also a large collocational
Spread and the items with which they are collocated form very diverse
sets. The lower down we go on the scale, the more limited is the number
of terms Of collocable sets, the closer the interrelationship between
the determined and determining terms in the collocation until we reach
the fixed ones, where one lexical item is collocated with one single
other lexical item in order to reveal a specific meaning.

The meanings of a lexical item form a microstructure Of the
respective lexical item, and are incorporated as such into the general
system Of language. They are part of "la langue"; they are formally
definable by means Of the relations the respective lexical item enters
into in the language. They have to be kept apart from the minute shades
of meaning acquired in actual uses Of the lexical item in various speech
situations. These uses are, of course, based both on the nucleus

meaning and on one Of the multiple significations of the lexical item,

25

but they are no longer facts of "la langue"; they are facts of "la
parole". It is again a point where language (la langue) and speech
(la parole) should be kept apart in the analysis.1

Indeed, not unlike phonetic analysis, semantic analysis may
result in an infinite number of distinctions, unless structural consid-
erations are introduced to distinguish what is essential from what is
not essential on the semantic plane Of language.

There remains however, the problem of idiomatic meaning of lexical
items. Their status as 'faits de langue' cannot be questioned. The
collocational and the colligational approach to meaning as described in
this thesis seems to encounter several difficulties in analyzing
idiomatic meaning. For this reason, it is advisable to suggest that
idiomatic meaning requires a separate level of analysis - the idiomatic
or phraseological. However, this is not a final position. It may very
well be that the refinement of the distributional technique of analysis
(Of. 5.5.) and especially the introduction Of the notion of "fixed

collocation" (see p. 16) will cover idiomatic meaning too.

5. Distributional Analysis Of Meaning

 

The analysis of syntagmatic and paradigmatic relations has been
carried on at a low level of abstraction. Its main purpose has been to
prove the existence of relations between individual lexical items. The
technique Of analysis has not differed markedly from what is known as
contextual analysis which puts its main emphasis on the dependence of the
meaning of linguistic forms on the environment in which they occur. It

has been pointed out by many linguists however, that it would be

 

1Zvegincev, V. A.: (1962)

24

impossible to list the occurrences Of a lexical item in all possible
environments since they are practically unlimited.1

What is required is to perform the analysis at a higher level Of
abstraction. Consequently, we shall be concerned with lexical items not
as individual units, but as members of grammatical classes. In other
words, collocational considerations will be coupled with colligational
ones.

In colligational terms, we shall be interested in distributional
formulae, assuming that to each linguistic unit used in a given formula
there corresponds a certain meaning of the respective unit. Once a
distributional formula has been established, we next analyse the
combinatory valences Of a lexical item within such a formula, with
various members of the grammatical classes characteristic of the
respective formula.2 This latter aspect of the analysis is reminiscent

of the projection rules in the semantic theory of Katz and Fodor.

5.1. The relation between grammar and semantics.

The close interdependence between grammar and semantics is a
characteristic principle of present day research in linguistics. Gen-
erative grammarians regard semantics as a component of the general

5

theory of language which continues and completes syntax.

 

lUllmann, St: (1962)
2Apresjan, Ju. D: (1966)

3Chomsky, N: (1965)
Postal, P. and Katz, J. J.: (1965)

‘9

25

British linguists also adOpt a similar position asserting that
the analysis Of lexis begins where grammar has exhausted its own possi-
bilities of analysis.1

Understanding the real nature of the relation between grammar and
semantics is one of the most important problems in linguistics today.
By assigning mutually exclusive domains of language to grammar and
semantics, most modern theories of language explicitly attempt to enlarge
the task of grammar so as to account for as much of language as possible,
and to leave as little as possible to semantic analysis.

An alternative approach has been suggested, among others, by
U. Weinreichz. Characteristic of this latter approach is the emphasis
laid on the simultaneous interaction and deep interpenetration between
grammar and semantics. A similar approach is followed in this thesis.
It is my Opinion that while grammarians should exhaust the possibilities
of grammatical analysis before turning to lexical analysis, it is also
the task Of the students of lexis to advance into grammar with their
statements of meaning. This was done by many in the form of "Grammars
of Words"3. Lexicographical practice has reached a similar conclusion
and, indeed, many lexicographers have come to wonder where their task

ends and the grammarian's begins.

 

lHalliday, M.A.K.: (1961)
Lyons, J: (1965)

2Weinreich, U: (1966)
3Palmer, H. E: (1958)

hLevitschi, L: (1957)

26

5.2. Method of approach and terminology.

The principle adopted in this paper is the following: Starting
from a grammatical description of language, it is possible to engage
in a relevant analysis of lexis in distributional terms. Grammar will
provide a list of patterns, of distributional models characteristic of
a certain language. The task of the student of lexis is to establish
some correlation between distributional models and the various meanings
of lexical items.

This principle, originating with Harris, has been practiced quite
extensively by Apresjanl. In fact, it has strong foundations in Older
lexicographic practice.

The second task is to account for what I call (following Ollson
and Firthian Linguistics) the combinatory valences of words. (This
concept also resembles Chomsky's notions of complex symbols and strict
subcategorization in generative grammar).3

The classes of words - the grammatical categories according to

Chomsky, in terms of which the various distributional formulae will be

set up, are the following:

N = noun or any nominal construction
V = verb or any verbal construction
A = adjective or any adjectival construction

Adv = adverb or any adverbial construction

Prep = preposition

 

lApresjan, Ju. D: (1962 a) (1962 b) (1965)
2Hornby, A. S: (1948)

3Chomsky, N: (1965)

27

Det determiner

Aux = auxiliary of modal verbs
In the course of the analysis (cf. 4.2.) it has been discovered
that it is more significant to express distributional formulae in terms

of functional categories, in which case the following terminology will

be used:
Subj = subject
Pred = predicate

D. O. = direct Object
Ind. O. = indirect object
Obj. Compl. = Object complement

In order to account for combinatory valences nouns have been
divided into structurally defined subclasses. The subclassification
presented below has been found useful for the analysis of the combinatory

valences Of English verbs.

 

 

 

 

A(animate) Non a(inanimate)
H(human) Non H(nonhuman) C(concrete) Non C
(abstract)
(he, she, who) (it, what)
M(male) F(female) Count Non Count
(countable) (non
(he) (She) countable)
(it, they) (it)

 

 

 

 

 

 

Nouns are first Of all divided into 'Animate' and 'Inanimate'; animate
nouns are further subdivided into 'Human' (substitutable by : he, she,

who) and 'Nonhuman' (substitutable by: it, what). Inanimate nouns are

28

in turn divided into 'concrete' and 'non-concrete' (abstract). There
are further relevant distinctions Of Human nouns into male and female,
as well as Of concrete nouns into Countable and Noncountable.

In order to arrive at distributional models which will prove
Significant for the specification of the various meanings of a lexical
item it is necessary to analyse that item's occurance in various syntactic
constructions. For a lexical item belonging to the class 'verb', for
instance, such distributional formulae will have to be considered:

'N+V+N'; 'N+V+A'; 'N+V+V'; 'N+V+Prep+N'; 'N+V+adv'; etc.

5.5. Distribution and Meaning.

 

The analysis of polysemantic lexical items will reveal a biunique
correlation between meaning and distribution.

NUmerous examples can be supplied, from any dictionary of English:1

 

Lexical Item Gram. Class Distrib. Formula Examples Meaning.
I
fail V H+fail+H John failed
his parents
'disapoint'
They failed us
H+fail+to+V He failed to
come
'not to
succede'
She never
fails to
tease him
NonC+fail+H Luck failed
him 'abandon'
'leave'
Courage often
1 fails them

 

 

 

 

 

lHornby, A. S: (1948) was used for these exemples.

entertain ‘

examine

feel

foster

frame

fire

 

 

H entertain H

H entertain NonC

H examine N

Hemmhmfi

An+feel+N

H+feel+(that)+S

H+foster+H

H+foster+NonC

H+frame+NonC

H+frame+H

H+frame+C

H+fire+H

 

We often enter-1
tain friends.

She has enter-
tained the idea
for months.

He examined my
eyes. They
are now exam-
ining the data.

The teacher
examined the
pupils.

The boy felt
the edge Of the
blade. The dog
felt the hare.

He felt that it
was his duty to

go.

They fostered
the child.

The lady has
always fostered
the poor.

He fosters high
hOpes.

The nephew

fostered revenge.

He framed the
theory. They
framed the plant

They framed him,
He framed the
picture.

He fired the

workers. She
fired the

 

servant.

29

'play host
to'

'consider'
'get infor-

mation (knowl-
edge) about'

'test knowl—
edge of'

'become
aware through
senses'

'consider'

'nurse'

'care for'

'hold in
mind'

'make, form'

'plot
against'

'enclose in
a frame'

'dismiss'

graduate

grasp

greet

hang

flat

 

 

 

H+fire+C

50

{They fired the set fire

{bUildj—ngo

to'

l
H+graduate+Prep NEHe graduated ftake a degree

N+graduate+H

H+graduate+C

H+grasp+C

H+grasp+NonC

H+greet+H

NonA+greet+H

H+hang+NonA

H+hang+H

C+Be+flat

iat Cambridge.
i
iHelen grad-

1
The school
graduates 100
‘students each
iyear.

He graduated
the stick.

He graSped
my hand.

He grasped
the idea.

He greeted us
joy fully o

The newcomers
greeted them
roughly.

The sound
greeted us
from afar.

The view
greeted us
magnificently.

They hung
their coats.

They hung him.
The ground is
flat.

The table is
flat.

 

They graduated
him with honors

4
:
n

I

or diploma'

{uated from Yale;
! I

 

"give a
gdegree'

i
l
I
I

'mark for
purposes Of
measurement'

'seize with
hand f irmly'

'understand'
'show feelings

at meeting
somebody'

'impinge
senses'

'suspend'
'put to death
by hanging'

'smooth, even
and level'

 

grave

green

 

 

)

H+Vt+N+flat

N V. flat
1

H+Be+grave

NonA+Be+grave

End products of
plants+Be+green

H+Be+green

He struck him
flat.

The whole
forest was
felled flat.

Life seems
flat.

Beer tastes
flat.

He is a grave
man.

The situation
is grave.

The patient's
illness is

grave.

The fruit is
green.

He is still

 

L green.

 

51

'to the
ground'

'uninteresting'

'serious,
thoughtful'

'important,
serious'

'not ripe or
seasoned'

'untrained,
unexperienced'

The nature Of semantic relation is such that not all semantic

distinctions will manifest themselves in syntactic differences, although

we can assume with greater certainty that to each syntactic difference

. . 1
corresponds an essential semantic one.

refined methods of analysis.

It is for this reason that we need to introduce further and more

As regards the distributional formulae it is necessary in the first

place that the syntactic constructions we compare are really contrast.

It is for instance futile to expect two different meanings from two

distributional formulae of the kind:

 

Apresjan, Ju. D.:

(1965)

52

'N1 V N2' and 'N2+Be+past+V+by Nl'
'He read the book' 'The book was read by him'

Since the two sentences are transformationally related, they will
consequently belong to one single distributional formula in the set
provided by grammar for the purposes of the distributional analysis of
the structural meaning of lexical items.

Transformational considerations will also be useful for distinguishing
more refined Shades of meaning within the general meaning supplied by a
certain distributional formula.

Similarly, to account for combinatory valences, Often, more detailed
divisions and subgroupings within a certain grammatical category will be
required. In subgrouping the class "noun" we have used a fairly simple
classification based on a structural criterion (substitution). However,
for the specification of some restricted meanings, pure semantic

criteria, not unlike the semantic features revealed by componential

analysis will have to be used.

5.4. The Distributional Analysis of the Structural Meanings of the

Verb 'Make'.

 

The way transformational and more refined semantic criteria are
introduced into the distributional analysis of meaning can be better
illustrated by an attempt to discern among the meanings of a highly
polysemantic lexical item. The verb 'make' has been chosen for this
purpose. What is required is, as mentioned above, to arrive at a set
of relevant distributional formulae which specify the various meanings

of 'make'.

55

We shall proceed to a detailed analysis Of the syntactic con-
structions in which it can occur, and then select the significant ones
from those that prove to be transformationally related and therefore
nonsignificant for the purposes Of meaning Specification.

The various verbal syntactic constructions characteristic of
English verbs have been mentioned earlier: 'N V N; 'N V N N'; 'N V A';

'N V Prep'; 'N V Prep N'; 'N V V'; 'N V N Prep N'; 'N V Prep Prep N'; etc. 5

5.4.1. The sequence N V N , characteristic Of English transitive

 

verbs poses no problems as to its significance for the specification of

~w4

one of the structural meanings Of 'make'. It will pose problems,

however, in further specifying of shades Of meaning, and this will be
discussed later. (See 5.5.) For the time being we shall simply postulate
'N make N' as the first distributional formula, without assigning any

specific meaning Of 'make' to it.

5.4.2. The syntactic pattern 'N V N N' will present two variants
for the verb 'make':

a. N make N1 N2 - where Nland N2 belong to different subclasses of

the class N (We shall symbolize that by Nl % N2).

Ex. The father made his son a toy N1 is animate; N2 is inanimate.

The boy made the dog a leash

b. N made N1 N2 - where N1 and N2 belong to the same subclass of N,

they are both animate or non—inanimate. (We shall

symbolize it by N = N2).

1
Ex. The father made his son a teacher
War made John a brute

The boy made the stone a weapon.

54

Both constructions are transformationally related to constructions Of

the type:

 

I ' \
N make N1 N2 ___________ ,) N make N2 Prep Nl

Ex. The father made his son a toy :=fi> The father made a toy for his son

 

The father made his son a teacherweThe father made a teacher of his son

Except for the fact that different prepositions are used in the transforms,

 

 

 

 

“a:

the two patterns go together. However, a clearcut distinction will appear A J
I! "

when further transformational relations are investigated: 3

I a _ ___.__1 3 B

1‘: make N1 N2 2 N1 e N2 ‘

Ex. The father made his son a teacherngis son is a teacher ’

The father made his son a toy ;__> *His son is a toy

One further distinction appears when a deletion is applied:

 

N make N1 N2 '> N make N

The father made his son a toy :2:=::::$>The father made a toy

 

 

The father made his son a teacher .___;>*The father made a teacher1

We can solve the pnablem by relegating the first variant 'N make NlNZ'
(Nl # N2) to the first pattern (N make N) since it allows the delition
Of N1, and preserves only 'N make N N ' (N1 = N2) as a second significant

1 2

distributional formula for the meanings of 'make'.

5.4.5. Next sequence to be considered is 'N V N Adv. particle'2
Ex. Try to make matters up

You must make yourself over

 

1It is true that such transformations are possible with a limited

set of the subgroup H.
The king made the warrior a knight The king made a knight, but a new
meaning of 'make' is achieved here.

2Adverbial particle is an adverbial that can also be used as a
preposition (cf. Hornby, A. S.: (1948) p. ix)

55

It is necessary to note the following things in relation to this
distributional formula:

a) the presence of N in front of the adverbial particle is obligatory;
b) only a limited number of items ('Out', 'up', 'Over') can appear in
the formula.

'N make V N Adv. Part', will be postulated as the third distributional

formula for the specification of the meanings of 'make'.

5.4.4. In analyzing a sequence Of the type 'N make N Prep N' in
order to assess its Significance for the meanings Of make, we Shall
distinguish two variants:

a) Sequences that allow the transformation:
N make N1 Prep N2“::—>N make N2 N1

He made a man of you ———> He made you a man

 

 

He made a toy for the boyu__g>He made the boy a toy
Frost makes ice from water::::;>Frost makes water ice
which have been discussed earlier under (2)
b) Sequences that do not allow the above mentioned transformation:
Ex. He made a speech at the meeting
The minister made a statement to the press this morning
That they belong to a separate pattern is proved structurally by the

fact that they allow the following transformation:

 

N make N P N ————j>N V(n) P N (N(n) = verb derived from a
noun)

 

He made a speech at the meeting.___;>He spoke at the meeting.
But for our particular purposes, namely the specification of the meanings

Of 'make', such sequences can be treated as extended 'N make N' sequences,

 

56

Since the additional 'P N' does not prove to be relevant in most
cases.

On this basis, we can conclude that neither variant of 'N make
NPN' can constitute a significant distributional formula relevant to
the meaning of 'make', since both can be relegated either to the pattern
'N make N N (N1=N2)', or to the more general pattern 'N make N' which

will be discussed in a more detailed manner under (5.5.).

5.4.5. The syntactic pattern 'N make N A'
Ex. He made her happy; The news made me sad ressembles in a way the
'N make N N' pattern (more specifically its variant that allows the
'N Be N ' transformation)

1 2
He made her a lady ____;>She is a lady

 

He made her happy ::==:$>She is happy

But for reasons to be explained below, it is not advisable to dismiss
this pattern as irrelevant.

If we consider sequences of the type:

'N make N adj (that) S' - He made it clear that he will fight to the end
it can be noticed that the preadjecti'al N position is usually filled by
the pronoun 'it' which 'anticipates' the post adjectival subordinate
clause. The relation is 'it = (that) 8'. On these grounds we can
postulate that we are dealing here with a variant of the 'N make N A'
pattern, and consequently will have to consider the latter an inde-

pendent distrubutional formula for the meanings of make.

 

To be more specific, this is true for all cases and all distri—
butional formulae when the 'P N' sequence is an adverbial (of manner,
substitution item: there; or time, substitution item: then).

 

57

In related extended sequences: 'N make N A P N' - 'He made that
clear in the House' - it again appears, like with other syntactic
patterns, that the 'P N' segment functioning as an adverbial, does not
play a significant role in specifying the meanings of 'make'. All
sequences discussed under 5.4.5. can therefore, be grouped together

under one single basic type: 'N make N A'.

5.4.6. Of particular interest is the sequence 'N make N V'. This
syntactic construction can not be reduced to the basic 'N make N',
because a close interrelationship Obtains between 'make' and the other
verb in the sequence.

Ex: He made her cry

I found it difficult to make John accept it

You make my blood run cold
It follows that 'N make N V' (with possible additional sequences after
V) should be considered a separate basic pattern constituting one of the

general meanings of 'make'.

5.4.7. In the sequence 'N make A', the element A proves relevant
for the specification of the meaning of 'make', so that the pattern
will be considered a basic one:

Ex. He is sure to make good

To this pattern we shall add also sequences of the type 'N make APN' —
I made certain about that: He made sure of her, since, as in other
instances, the appended 'P N' sequence does not prove relevant for the

meanings of 'make'.

 

-319

58

5.4.8. The sequence Of the type 'N make A N' will also be
considered a separate pattern, including the variant 'N make A (that)
S' (that is the replacements Of a post adjectival noun by a nominalizing
clause).

Ex. This attempt has made clear his real intentions.

5.4.9. Many of the uses of 'make' occur in the distributional
formula 'N make Prep'.
One variant is represented by combinations of 'make' with spatial
prepositions (Off, away, in).
Ex: The travellers made in before dusk; He made away with the treasure.
The important thing about this pattern is the fact that the sequence
'make + Prep' is self-sufficient as far as the specification of the
meanings of 'make' is concerned. On these grounds the sequence will be

set as a basic pattern.

5.4.10. Sequences: 'N make Prep N' are of two kinds.
- Some allow the transformation:
N make P N N make N P (with a subsequent transformation
which regulates the realization of 'N' as a pronoun):
He made up the story He made it up.
The prepositions are actually adverbial particles.
- Some others do not allow this transformation; moreover an immediate

constituant analysis reveals a structure:

 

 

[N makEJ P N! as Opposed to {N;_makeg_Pl_Nj , characteristic of

 

the other variant.1

 

lAgain, P in the sequence 'N make P' is an adverbial particle
(also in 5.4.11.)

59

Ex: She made quickly for the door.
We made into the forest.
This particular variant of the distribution 'N make P N' is another

basic pattern Specifying yet another general meaning of 'make'.

5.4.11. Sequences Of the type 'N make P P N' also belong to two

large groups:
- Some allow the immediate constituent analysis [—N make P_7 + {-P N_7
which leads to the pattern discussed under (5.4.9.).
Ex: He made away with the treasure.
- Some others do not allow this immediate constituant analysis, and
moreover, diSplay a closer interrelationship between the sequences
'make P+PN':

We must make up for the last time

He is trying to make up to me now.
It is necessary again to postulate this latter sequence as a separate

basic pattern.

5.5. More Refined Meaning Distinctions Revealed bngrammatical Operations.

 

The patterns established thus 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 consequently
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, actually it is akin to the general meaning Of transitivity of

verbs.

40

For the further Specification of meaning within a certain pattern
it is necessary to increase our accuracy of analysis by applying further
possible grammatical Operations to the postulated distributional formula.

In the following pages I shall discuss the basic pattern 'N make
N' and attempt to arrive at more specific meanings of 'make' by testing
the reaction of the pattern to various grammatical Operations.

Let us consider first the transformations that are allowed by it:

(a) 'N make NZ':::;:'N2 Be made by N ' (the active-passive transformation)

1 1
Ex: The boy made a mistake-==$>A mistake was made by the boy.

(b) 'N make N '::=$IJ\I )' (a transformation which changes the
1 2 1(N , , . .
2 make N sequence into a verb derived
from N2?

Ex: He made an analysis Of the situationz=$>He analyzed the situation

(6) 'N1 make N2':=::;>'Nl Be N2' (a transformation that allows the
replacement of 'make' by 'be')

Ex: 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:

He will make a fine lawyer :==:>He will be a fine lawyer (transformation

c); others allow two:

John made an analysis of the situation:=:;An analysis Of the situation

was made by John (transformation a) and 22:9»John analyzed the situation

(transformation b), while still others will not allow any Of the three

transformations:

Ex: The argument made sense.

As a result the general basic pattern 'N make N' will be subdivided

into several subgroupings according to their reaction to the set of

41

three transformations (a,b,c). Each subgrouping will be assigned a
specific meaning Of 'make'.

This will not exhaust the possibilities of analysis.

Each Of these subgroups can be further analysed in terms of
various semantic makers of N, by the technique of combinatorial analysis.

If we take one of the subgroups arrived at by virtue of the fact
that it allows only the third transformation (N make N====$ N Be N),
further distinctions can be made and consequently more specified meanings
can be arrived at.

Similarly, all other subgroups can be differentiated further into
more specified meanings by making use of the above mentioned criteria.

A more detailed analysis will eventually result in even finer
delimitations Of the meanings of 'make'. The question that arises now
is, what is more important in meaning specifications: the distribution
of the respective lexical item, the transformations the construction Of
which it is a part may undergo, or the semantic factors Of the adjacent
Ns?

There can be no doubt about the primary importance Of the former
two factors. This can be very easily proved by the fact that we get
different meanings of 'make' by collocating it with one and the same N
in various constructions:

"He made 60 dollars" - allows the passive transformations;
"Your award makes 60 dollars" - does not; it allows the N make N:;>N Be N
tranformation.

'Make' definitely diSplays different meanings in the two utterances

and they are primarily the result of their characteristic distribution

42

formulaa.and their possible transformations rather than of the lexical

items with which 'make' collocates in the two utterances.

This has been in no way an attempt to analyse exhaustively the
distributional meanings of make, but rather an exemplification of the
way one can postulate the relevant distributional formulae for a Specific
lexical item. But even if such an anlysis were pursued further, it
could not cover all the meanings of 'make'. The possibilities of
distributional analysis of meaning are limited, there is a number Of
very specific meanings which will probably be resolved at a different

level - the idiomatic or phraseological.

4. Paradigpatic Relations.

 

4.1. Delimiting semantic fields.

 

The distributional analysis of meaning along the syntagmatic axis
is completed by a similar analysis along the paradigmatic axis. In
other words, instead of being interested in the distributional formulae
of a given lexical item which prove relevant to its meanings, we shall
direct our attention to the various lexical items that could fell one
and the same position in a given distributional formula.

The claim is that the elements Of a lexical set fill a certain
position in a distributional formula primarily on account Of the gram-
matical category to which they belong, but also because they share in
common some semantic feature.

The substitutional occurrence of various lexical items in a
certain distributional formula is regarded, therefore, as an indicator

of some semantic relationship among them.

45

Following Halliday (1961) (i.e. his concept of lexical set) and
especially Apresjan (1966), I prOpose to use distributional models as
Objective criteria for delimiting semantic fields. By suggesting an
interplay between collocational and syntactic criteria in establishing
semantic fields I attempt a synthesis Of Halliday's and Apresjan's
approaches.

For the exemplification of structurally delimited semantic fields,
I Shall use again lexical items belonging to the grammatical class Verb.
The reason for this choice is that English verbs seem easier to handle

in terms of the nominal constructions with which they collocate.2

4.2. The Semantic Field of English Factitive Verbs.

 

Let us consider the semantic relations Obtaining among various
lexical items Of the class verb that may occur in the following distri-
butional formulae: 'N V N N'3

For reasons that will become immediately apparent, it is necessary
at this stage to reformulate the distributional model in terms Of
functional categories.

Indeed, as already indicated when discussing the meaning of the

verb 'make', a general distributional formula of the type: 'N V N N',

 

1Semantic fields isolated in this manner are of course, different
from the conceptual fields prOposed in the past by Trier, Bally, Wartburg.

I have in mind, especially the notion of strict subcategorization,
Chomsky, N (1965)

3It is given as verbal pattern no. 8 in Hornby, A.S: (1954). It
is necessary to point out that the 25 verbal patterns postulated by
Hornby are highly tentative and are accepted here as such. Also, there
is a gross inconsistency in Hornby's use of grammatical and functional
categories which has been avoided in my thesis.

m:

if left at the level of grammatical categories, will not distinguish
between the two types of structure exhibited in the following sentences:
We chose him a new suit.
We chose him our leader.
A rewriting Of the formula in terms of functional categories will
reveal two separate distributional patterns:
(a) 'Subj + Pred + Ind. O. + D. O.‘

Ex: We chose him a new suit.
(b) 'Subj + Pred + D. O. + Object Complement'

Ex: We chose him our leader.

There are several structural considerations that lead to this

distinction:

- A sequence Ind. O. + D. O. can be transformed into:

'D. O. for Ind. O.‘
to

Ex: The boy left him a note ==% The boy left a note for him.
On the other hand the sequence Dir. O. + Be + Obj. Compl. (A secondary
transformation will change the case of the N functioning as direct
Object into nominative).

Ex: We chose him our leader ‘——%>He is our leader.

 

- A relation of complete agreement in number and gender Obtains between

'D. O. and Obj. Compl.‘ which is not necessary for the sequence 'Ind.O. +

Ex: We chose the boy our leader

(he) (he)

45

We chose the boy two new suits
(he) (it, they)
(sg) (pl)
- A sentence including the Ind. O. + D. 0. sequence allows two passive
transformations:
We left the man a note 1. The man was left a note (by us)
2. A note was left for the man (by us)
while a sentence which includes the Dir. O+Obj.Compl. sequence allows
a single transformation
We left the man a moral wreck The man was left a moral wreck
Similarly it would be impossible to differentiate between each type
of the 'N V N N' constructions discussed above, and one exemplified
by this sentence:
I saw my brother, the doctor. - in which the second noun serves
as an appositional attribute for the direct Object. There are structural
criteria for this distinction too; namely the ability to replace the
appositional noun (or noun phrase) by a relative clause:
I saw my brother, the doctor I saw my brother who is a doctor
By using functional categories and also by Specifying the semantic
markers of noun that account for the combinatorial valencies of the
verbs to be discussed, the distributional formula can be finally
represented as:
Subj + Predicate + D. O. + Obj. Compl.
(human) human
non human

Ex: They christened the baby Mary.

46

We found the house a mess.
The class elected John chairman.
The boss appointed him sales-manager.
Let us consider in detail the verbs that can fill in the 'predicate'
position in the formula: 'Subject+Predicate+D.O.+Object Complement'

A list of such verbs is given below:

A B C D
appoint call think keep
designate name consider find
nominate nickname see leave
name dub fancy
make label envisage
crown Christen imagine
proclaim baptize judge
entitle prove
style prefer
elect like
choose want

find

Such verbs are usually called 'factitive' and a dictionary definition
will refer to them as 'making, calling or thinking something to be of a
certain character' (Webster's New World Dictionary of the American
Language, College Edition) in addition to defining them grammatically
as'verbs that take an object complement'. Such double definitions -
semantic and grammatical - are not rare. They point out the real

possibilities that exist to find structural criteria which define many

semantic subgroupings within various grammatical classes.

1+7

There can be no doubt about the existence of a semantic feature
shared by all factitive verbs, however vague and undefined it may seem
to us at first sight. The term 'factitive' used to define the verbs
seems to me as good as any other and will be preserved here to cover

their common semantic feature.

4.5.1. Almost intuitively however, one can see the possibilities
Of making further distinctions within the established group of factitive
verbs. I have divided them into four subgroups, Of which three (A,B,C)
are based on the dictionary definition mentioned above: "making
(subgroup A), calling (subgroup B), or thinking (subgroup C) something
to be of a certain character". Subgroup D is 1 set aside on semantic
grounds too. The division is only tentative; it can be maintained only
if sufficient structural Or collocational, see 5.5.4) considerations
support its validity.

Subgroup A is characterized by the fact that the noun that
functions as object complement is not preceded by an article.

Ex: He appointed his brother manager.
We elected John chairman.
The court crowned the young prince king.

A second characteristic of verbs in subgroup A is the fact that
they allow 'as' or 'for' (both meaning 'in the capacity of') to precede
the Object complement.

Ex: They will certainly choose a young man for chairman this time.
Their best move was to appoint John as general manager.
All verbs in subgroup A share in common an additional semantic feature

(this time in Opposition to verbs in the other subgroups We can call

48

this semantic feature 'designation' (Human X acquires status Y) and

 

call the verbs designative.

4.5.2. The verbs in subgroupsB and C are Opposed to the ones in
subgroup A, by virtue of the fact that they demand the presence Of an
article in front Of the Object complement and do not allow 'as' or 'for'
to precede it.

Ex: We call him a real friend
We labeled him a traitor.
(Note: when the Object complement is a proper noun, no article is
employed)
Ex: We called him Jack
The pastor christened the baby Helen.).

In order to distinguish subgroups B and C from one another, it
is necessary to look for some other criteria. This can be done by
testing the reaction of verbs in both subgroups to other grammatical
Operations. Thus, verbs in subgroup C allow the insertion of 'to be'
in front Of the Object complement, while verbs in B do not.

Ex: We all thought John (to be) a real friend.
I find his conversation (to be) a delightful eXperience.

In addition to this, verbs in subgroup C are distinguished by the
fact that they allow the substitution Of the noun Object complement by
an adjective, or past participle.

Ex: I find his conversation delightful.

We proved him false.
Enough evidence has been gathered therefore to consider verbs in

list C as a separate subgroup of the factitive verbs. The semantic

49

feature they share in common could be termed 'evaluation' and the verbs

 

can be refered to an evaluative.

4.5.5. Verbs in list A appear now as defined negatively (by
their negative reaction to the grammatical Operations applied to verbs
in subgroup C). Their common semantic feature could be termed

'gppellation' and the verbs can be called 'appelative'.

 

4.5.4. Three verbs: 'keep', 'find' and 'leave' also appear in
the distributional formula of factitive verbs.

Ex: They kept him prisoner for a long time.

We found the place a complete mess

Structurally, they partly belong tOgether with the 'evaluative' group

Of factitive group of factitive verbs in that they allow a substitution

Of the noun—Object complement by an adjective, for instance:

Ex: They kept it safe

We found the place dirty and left it very clean.}

Two possibilities are Open for their analysis:

-tO stick to their structural classification within the subgroup of
evaluative verbs (possibly as an even smaller subset), in utter dis-
regard Of the fact that they do not seem to possess the semantic
feature 'evaluation';

-to apply collocational criteria (cf. 2) and establish them as a
separate subgroup D of factitive verbs sharing in common a semantic

feature which could be called 'situational'. I have chosen this second

 

alternative.

50

It is in applying such additional criteria that I depart from
Apresjan who finds it absolutely necessary to hold to syntactic

considerations exclusively.

4.4.5. Further distinctions are possible within each of the four
subgroups delimited thus far. It is possible for instance, to maintain
that within designative verbs, 'crown', 'proclaim', entitle' form an
even smaller subject, 'elect and choose' another one, etc. Similarly,
within appelative verbs 'baptize and christen' are in many ways distinct
from the others.

But the distinctions are no longer arrived at by the application
Of structural criteria; they are now a function of the "collocational
range" Of the reSpective verbs, as defined in part (2) Of this thesis.
That is, there is a propensity for collocation with a certain set of
lexical items (christian names) that 'baptize' and 'christen' share,
which differentiates them from the other appelative verbs.

It is possible that in this way we could fill the gap between
structural meaning and significative and denotative meaning, since it
seems to me that collocational considerations are more pertinent to the
two latter aspects Of meaning rather than to the first.

It seems, for instance, that further semantic features could be
abstracted in such a manner: appoint, designate, make name, nominate
(on the one hand) will be Opposed to 'elect and choose' as being
antonymous in meaning (inclusion vs exclusion of election process in

Human X acquiring status Y).

51

5. Componential Analysis.

The Operations performed under (5), appear as a successive
delimitation of semantic features shared in common by increasingly more
refined groups and subgroups Of lexical items. This Opens the way for
a type of componential analysis both Of the meaning of semantic fields

as wholes, and of individual lexical items.

5.1. Semantic Components of a Lexical Item.

 

A lexical item 'designate', for instance, could be analysed into
the following semantic components:
designate v / trans+-{human action+ z-factitive+(designative+1ack of

election procedureZZ}-/
which reads: the semantic characteristics of the verb designate are:
transitive, human, factitive, designative, indicative of lack Of election
procedure.

With each grammatical Operation employed to isolate a structural
meaning of a lexical item we define also a semantic feature which it
shares in common with some other lexical items of the same grammatical
class.

The more general semantic features will be arrived at by means of
grammatical Operations, while more specific ones will be revealed by
collocational considerations. This sort of division seems to correspond
to an actual inner organization of meaning. Fodor and Katz too, in
their semantic theory prOpose a distinction between semantic markers and
distinguishers, although as it has been pointed out it is not always very

clear within the framework of their theory how to draw a line between the

two categories of semantic features.

52

5.2. Semantic Components of a Lexical Field.

 

It is Of course possible to apply a componential analysis technique
to whole semantic fields and thus arrive at what would appear as
equivalents Of "conceptual dictionaries" of the Roget Thesaurus type.

The significant difference between the two lies in the fact that unlike
conceptual dictionaries, they will be based on structural considerations,
that is on the network Of syntagmatic and paradigmatic relations that
criss-cross the vocabulary Of a language.

It is possible for instance to set up classes of verbs within
larger classes of verbs until the whole grammatical class of English
verbs will be exhausted.

The first division will be into transitive and intransitive verbs
on the grounds of the reaction to the passive transformation in a
distributional formula of the 'N V N' type.

Further, it is possible to distinguish within transitive verbs a
comprehensive class of verbs indicating 'human actions'. Within this
larger group, the semantic group of factitive verbs will occupy a certain
position. In its turn the semantic field of factitive verbs is divisible

into smaller subsets.

6. Conclusions.

 

It would be premature to draw any definitive conclusions regarding
the approach to structural semantics outlined in this thesis, in view
of such a limited analysis. There are many theoretical problems (i.e.
the problem of idiomatic meaning, etc.) that still need to be clarified,
and above all, there is the obvious task Of applying the theoretical

apparatus to a more comprehensive analysis Of lexis.

55

Several strong arguments in favour Of the approach can however
be made even at this stage.

The first is of a theoretical nature. By postulating syntagmatic
and paradigmatic relations as the basis for a theory of lexis, a unity
within a general theory Of language is achieved. The approach also
insistgvthe isomorphism between grammar and semantics, as well as their
close interrelationship and deep interpenetration.

As for arguments of a practical nature, some have already been
exemplified in this thesis. The relational approach to semantics proves
very useful in analyzing the structural meanings of lexical items, in

delimiting semantic fields, and on this basis it Opens interesting

possibilities for a componential analysis Of meaning.

BIBLIOGRAPHY

Ahmanova, O. S., Nekotorje VOprosy semantidekogo analiza slova (Problems
of the Semantic Analysis of Word), Vestnik M.G.U., 2, 1957.

Ahmanova, O. S., Vinogradov, V.V., Ivanov, V.V., O nekotorje voprosax i
zadaEax lekcikologij. (Some Problems and Tasks of Lexicology),
Voprosy Jazykoznanija, 5, 1956.

Apresjan, Ju. D., O ponjatiax i metodax structurnoj leksikolOgij.
(Notions and Methods Of Structural Linguistics) in Problemy
strukturnoj lingvistiki, Moskow, 1962.

Apresjan, Ju. D., Sovremennye metody izucenija znaEenij i nekotorye
problemy strukturnoj lingvistiki. (Present Methods of Meaning
Analysis and Some Problems of Structural Linguistics) in Problemy
strukturnoj lingvistiki, Moscow, 1965.

Apresian, Ju. D., Analyse distributionelle des significations et champs
sémantiques structurés, Langages, 1, Paris, 1966.

Baldinger, K., Die Semasiologie. Berlin, 1957.

Bazell, C. E., Meaning and the Morpheme. Word, 8, 1962.

Bendix, H. E., Componential Analysis of General Vocabulary. The Semantic
Structure of a Set of Verbs in English, Hindi and Japanese,
International Journal of American Linguistics. Vol 52, NO 2,

April 1966 part II, publication 41.

Benveniste, E., Problemes sémantiques de la reconsrtuction, "Word",
V01. 10, 1102’}, 1951+.

Bloomfield, L., Language, New York, 1955.
Breal, M., Semantics: Studies in the science Of Meaning, New York, 1964.

Budagov, R. A., Introducere in ptiinta limbii (An Introduction to the
study of language), Bucuresti, 1960.

Carnap, R., Bar-Hillel,Y., "Semantic Information", British Journal Of
the PhiloSOphy Of Science, 4, 1955.

55

Catford, J. C., A Linguistic Theory of Translation, Oxford Univ. Press,
London 1965. 'hagagn at W' , in Wu, W, m
Basin. 1333

Chitoran, D., Colocatie si sens - O contributie 1a studiul sensului
elementelotlexicale, Studii Si Cercetari de Linguistica, XVI,
Bucurepti, 1965.

Chomsky, N., Syntactic Structures (J.L., Ser, Min., n.4), Mouton and Co.
La Haye.

Chomsky, N., Aspects of the theory Of Syntax, M.I.T. Press, Cambridge,
Mass., 1965.

Chomsky, N., Topics in the Theory Of Generative Grammar, in Current
Trends in Linguistics, v.III, Thomas Sebeokied; Mouton and Co.,

La Haye, 1966.

Coseriu, E., Pour une semantique diachronique structurale, in Travaux
de linguistique et de literature II, 1, Strassbourg, 1964.

Coteanu, 1., In legatura cu sistemul vocabularului. Probleme de
lingvistica generala, vol. II, Bucuresti, 1960.

Dixon, R., W., On Formal and Contextual Meaning, Acta Linguistica
Academiae Scientarum Hungaricae, 14, 1964.

Dubois, J., Esquisse dun dictionnaire structural, Etudes de linguistique
appliquée, l, 1962.

Fodor, J. A., Katz, J. J., The Structure of Language, Readings in the
PhilOSOphy of Language, Prentice Hall, Inc. Englewood Cliffs,
New Jersey, 1964.

Firth, J. R., Papers in Linguistics, London, 1957.

Fries, C. C., Meaning and Linguistic Analysis, Language, vol. 50, 1954.

Goodenough, W. H., Componential Analysis and the Study of Meaning,
Language, 52, 1956.

Graur, A., Studii de lingvistica generala (varianta noua), Bucuresti,

1960.

Greimas, A. J., Sémantique structurale. Recherche de methode, Larousse,
Paris, 1966.

Guiraud, P., Les Champs morpho—semantiques. Bulletin de la Societe de
Languistique de Paris, III, 1956.

Guiraud, P., La Sémantique (Que sais-je? 655), Paris, 1962.

56

Halliday, M.A.K., Categories of the Theory of Grammar, "Word", vol. 17-
3-1961 0

Hallig, R. and Wartburg, W. von, Begriffsystem als Grundlage der

Lexicographie, Berlin, 1952.
Harris, Z., Methods in Structural Linguistics, Chicago, 1951.

Hjelmslev, L., Dans quelle mesure les significations des mots peuvent—
elles etre considerées comme formant une structure. Reports for
the eighth International Congress of Linguistics vol. II. Oslo,

1957-

Hjelmslev, L., Prolegomena to a Theory of Language, The University Of
Wisconsin Press, Madison, 1961.

Hjelmslev, L., Some Reflections on Practice and Theory in Structural
Semantics, in Language and Society, COpenhague, 1961.

Hocket, Ch. F., A Course in Modern Linguistics. New York, 1958.

Hornby, A. S., (et a0: A Learner's Dictionary of Current English
0UP, London, 1948

Hornby, A. S.: A Guide to patterns and usage in English, 0UP, London,

1954

\

Jakobson, R., Essais de linguistique genErale (Arguments, 14), Les
Editions de Minuit, Paris, 1965.

Jakobson, R., Halle, M., Fundamentals of Language, Mouton and Co., The
Hague, 1957.

Katz, J. J. and Fodor, J. A., The Structure Of Semantic theory, Language,
59, 1965-

Katz, J. J. and Postal, P. MP., Semantic Interpretation of Idioms and
Sentences Containing Them, Quarterly Progress Report (M.I.T.),

70, 1965.

Katz, J. J. and Postal, P. M., An Integrated Theory of Linguistic
Description, The M.I.T. Press, Cambridge, Mass., 1964.

Kurylowicz, E., La notion de l'isomorphisme. COpenhague, 1949.

Kurylowicz, J., Zametki o Znachenij slova, (Notes on Word Meaning),
VOprosy jazykoznanija, 5, 1955.

Lamb, Sydney M., Outline Of Stratificational Grammar, Georgetown Univ.
Press, Washington D. C., 1966

57

Lamb, Sydney M., The Sememic Approach to Structural Semantics, in
Transcultural Studies in Cognition, American AnthrOpologist 66,

n.5.

Levitzki, L., Bilingual Dictionaries: Suggestions, Melanges linguistiques,
Bucarest, 1957.

Levitzki, L., A Course in English Lexicology, Bucharest, 1965.

Longacre, R. E., Items in Context, Their Bearing on Translation Theory -
Language vol 54 no 4, 1958.

Lounsbury, E. G., The Structural Analysis of Kinship Semantics, Proceedings
of the IXth International Congress Of Linguists, The Hague, Mouton,

1965.

Luria, A. R. and Vinogradova, O. S., An Objective Investigation Of the
Dynamics Of Semantic Systems, British Journal of Psychology 50,

1959-

Lyons, J., Structural Semantics. An Analysis of part of the Vocabulary
of Plato. Publications of the Philological Society, 20, Oxford,
1965.

Martinet, A., Elements de linguistique genérale, Paris, 1961.

Meillet, A., Review of J. Trier "Der deutsche Wortschatz", in Bulletin
de la Societe de linguistique de Paris, vol. 52, f.2, 1951.

Mitchell, T. F., Syntagmatic Relations in Linguistic Analysis, in
Translations of the Philological Society, 1958.

Nida E. A., A System for the Description of Semantic Elements, Word 7,

1951.
Nida E, A., Towards a theory Of TranSlation, Leiden, Brill, 1964.
Ohman, S., Theories of the Linguistic Field. "Word", vol. 9.n.2, 1955.
Ollson, Yngre, On the Syntax of the English Verb GOteborg, 1961.

Odgood, Ch., Suci, G. J., Tannembaum, P. M., The Measurement of Meaning,
Urbana, 1957. (Univ. of Illinois Press).

Palmer, H. E., A Grammar of English Words, London, 1958.

Pottier, B., Recherches sur l'analyse sémantique en linguistique et en
traduction automatique, Universite de Nancy, 1965.

Pottier, B., Vers une sémantique moderne, Travaux de linguistique et de
litterature, l, 1964.

58

Prieto, L. J., Rapport paradigmatique et rapport syntagmatique sur 1e
plan du contenu, "Omagiu lui I.Iorjan", Bucuresti, 1958.

Prieto, L. J., Principes de Noologic, Mouton and Co., The Hague, 196%.

Problems in Lexicography, ed. by F. Householder and Sol Saporta (IJAL,
28, 2; publ. n.21, Baltimore, 1962.

Problemy formalizacij semantiki, Ier MGPIIJ, Moscow, 1964.

Proceedings of the IXth International Congress of Linguists, Mouton and
Co., The Hague, 1964.

Quine, W. V., Word and Object, Cambridge, Mass., 1960.

Quirk, R., The Use of English, London, 1962.

Riffaterre, M., Stylistic Contxt, Word, 16, 1960.

Robins, R. H., General Linguistics: An Introductory Survey, London, 1964.
Rosetti, A., Le mot. Copenhague - Bucuresti, 1943.

Sapir, E., Language, New York, 1921.

Shaumjan, K., Strukturnaja lingvistika kak immanentaja teorija jazyka.
Moskva, 1962.

The Shorter Oxford Dictionary, C. T. Onions ed., London, 1961
Saussure, F. de, Cours de linguistique generale, II edition, Paris, 1922.
Schaff, A., Introduction to Semantics, Pergamon Press, New York, 1962.

Slama- Cazacu, T., La "structuration" dynamique des significations, in
Mellanges linguistiques, Bucarest, 1957

Slama—Cazacu, T., Langage et Contexte, Mouton and Co, La Haye, 1961.
Smirnickij, A. I., Znacenije slova, VOprosy iazykoznanija, n.2, 1955.
Smirnickij, A. I., Leksikologija anglijskogo jazyka, Moskva, 1956.

Stati, S., Contenu et expression, symetrie et assymetrie, Cahiers de
linguistique appliquée, Bucarest, 1962.

Stern, G., Meaning and Change of Meaning, Goteborg, 1931.

Tabory, R., Semantics, Generative Grammar and Computers, Linguistics, 16,

1965.

U1
\0

Trier, J., Der deutsche Wortschatz im Sinnbezirk des Verstands, Heidelberg,

1931.

Ullmann, St., Historical Semantics and the Structure of the Vocabulary,
"Miscelanea Homenaje a A. Martinet", I, La Laguna Univers. Press,

1957.

Ullmann, St., Semantics: An Introduction to the Science of Meaning,
Oxford, 1962.

Vinogradov, V. V., Slovo i znacenije kak predmet leksikologiceskogo
izucenija, (Word and Meaning as Objects of Lexic010gica1 Analysis),
Leningrad, 1945.

Vinogradov, V. V., Osnovnije tipy leksiéeskix znacenij slova (Basic Types
of Lexical Meanings of Words), Voprosy jazjkoznanija, 5, 1953.

Wartburg, W. von, Problemes et Methodes de la linguistique. Zeitschrift
fur deutsch Philologie, n. 57.

Webster's New World Dictionary of the American Language, College Edition,
New York, 1964.

Weinreich, U., Travels through Semantis Space. "Word", vol. 14, n.2, 3,

1958.

Weinreich, U., Lexicographic Definition in Descriptive Semantics, in
Problems in Lexicography, 1962.

Weinreich, U., On the Semantic Structure of Language, in Universals of
Language ed. by J. Greenberg, Cambridge, Mass., 1963.

Weinreich, U., Explorations in Semantic Theory, in, Current Trends in
Linguistics, vol.III, 1966.

Weisgerber, L., Die Sprachfelder in der geistigen Erschliessung der Welt.,
in "Festschrift fur J. Trier", Meisenheim, 1954.

Weisgerber, L., Der Begriff des Wortens. "Corolla Linguistica. Festschrift
fur F. Sommer", Wiesbaden, 1955.

Wells, R., Meaning and Use. "Word", vol.10, n.2-5, 1954.

Wells, R., To What Extent Can Meaning Be Said to Be Structured? Proceedings
of the Eighth International Congress of Linguists, Oslo, 1958.

West, Michael, A General Service List of English Words with Semantic
Frequencies and a Supplementary Word-List for the Writing of POpular

Science and Technology, New York, 1953.

Wittgenstein, L., Investigations phiIOSOphiques, Gallimard, Paris, 1961.

6O

Whorf, B. L., Language, Thought and Reality, New York, 1956.

Ziff, P., Semantic Analysis, Cornell Univ. Press., Ithaca, New York,

1960.
Zveginéev, V. A., Semasiologija, Moskva, 1957.
ZveginEev, V. A., Meaning as a Fact of ”Langue" and as a Fact of "Parole",

Proceedings of the 9th International Congress of Linguistics,
Massachuetts, U.S.A., 1962.

              

   

     

 

NST TIVEUN

”111111111111 11111

 

 

 

 

 

 

11111111111111111111 111111155

30