Lectures 1,2

LINGUISTICS, BRANCHES OF LINGUISTICS, PHILOLOGY, TRADITIONAL

GRAMMAR, LINGUISTIC DATA

Linguistics is sometimes called general linguistics to emphasize the fact that its principles
and techniques are for all languages, and not just for any particular language (e.g. English
or Polish). It is also sometimes called the science of language to emphasize the fact that its
methods are rather different from traditional methods of language study.
Thus, the object of the study of the general linguist is human language in all its forms and
manifestations. The general linguist is concerned with a scientific understanding of the
place of language in human life, and of the ways in which it is organized to fulfil the
needs it serves and the functions it performs.
In saying that linguistics is a science, we are actually implying that it deals with a specific
body of material, that is, spoken and written language, and that it employs research
methods that can be publicly described and justified by reference to a theory capable of
formulation.
Linguistics, as the science of language, in its operations and statements, is guided by the
following canons(principles) of science:

a. Exhaustiveness (or comprehensiveness), i.e. the adequate treatment of all the

relevant material. In other words, it should take account of all the available data.
b. Consistency, i.e. the absence of contradictions between different parts of the total
statements (consistent use of the employed terms and procedures).
c. Objectivity , i.e. another cardinal feature of scientificness. The questions scientists
ask, the conclusions they reach, and the evidence they cite must be capable of being
publicly observed and tested,they must be empirically verifiable. An empirical test is
one in which the examination of phenomena takes place under controlled ,
experimental conditions, the results being available to direct observation and
judgement, so that if the evidence is replicated , the same results and the same popular
judgements would be obtained. Put differently, the results are verifiable. The term
objectivity implies, among others, making use of procedures that would be accepted by
most, or all linguists as being valid, not needing separate justification each time they
are applied.

1
 One way of looking at objectivity is to see it in relation to its antonym,i.e. subjectivity.
Subjective investigation is classified as strictly unscientific by the criteria of standard
scientific methods. Speculation may be sophisticated and stimulating, but it is non-
science. Statements that cannot be tested, e.g. metaphysical explanations,(as these
have been called by B.Russell) , are often found in connection with the discussions on
the origins of language. For example, some people firmly believe that language was
invented by God, but, without providing some verifiable evidence, the issue must
remain outside the realm of science. If our data, methods and results cannot be
publicly observed(empirically verified) using standard techniques , then the scientist
(the linguist) has some justification for rejecting them.

d. Economy (or simplicity) whereby, all other things being equal, a shorter statement or
analysis (linguistic description) employing fewer terms is to be preferred over one
that is longer. But a word of caution will be in place here – Sometimes a simple
solution to one part of a language is possible, but only at the expense of a vastly more
complex account of some other part( or parts).

From what has been said above, it appears that linguistics is an empirical science, and
within the empirical sciences is a social science, in that the phenomena constituting its
subject-matter are part of the behaviour of human individuals in society, in interaction
with other individuals.
Linguistics differs from other sciences in that it both uses language and has language as
the object of its ultimate interest. For this very reason, linguistics has been described as
language turned back on itself or as language about language (i.e. metalanguage).

BRANCHES OF LINGUISTICS
There are various branches of linguistics distinguished according to the linguist’s focus
and range of interest.

Theoretical (general)linguistics
The aim of general(theoretical) linguistics is to establish general principles and concepts
for the study of all languages, to explain how language functions, and to describe its
structure,

2
Descriptive(synchronic ) linguistics
It is concerned with the description and analysis of the ways in which a language operates
and is used by a given community of speakers at a given time. This time may be the
present, and in the case of languages as yet unwritten or only recently given written form
it will inevitably be the present, as there is no other way of knowing any earlier stages of
them, though there are methods by which certain facts about such earlier stages may be
inferred. The time may equally well be the past, where adequate written records are
available, as in the case of the so-called dead languages like ancient Greek, Latin, Old
English, Old French, etc. What is important is that the descriptive study of a language
(and of any part of a language ), present or past, is concerned exclusively with that
language at the period involved, not with what may have preceded it or may have
followed it. Nor is the descriptive study of a particular language concerned with the
description of other languages at the same time.
Descriptive linguistics is the fundamental aspect of the study of language , as it underlies
and is presupposed by historical linguistics and comparative linguistics.

Historical (diachronic ) linguistics

The study of the developments in languages in course of time, of the ways in which
languages change from period to period, and of the causes and results of such changes,
both outside the languages and within them. This kind of language study must be based, as
indicated above, on descriptions of two or more stages of the continuous language series
being treated. Thus a study concerned with the development of English must be based on
earlier descriptions of the particular stages distinguished in its evolution (the stages may
be : Old English (700-1100/150), Middle English (1100/150-11450/1500), Early and Late
Modern English (1500-1800)).
Put in a nut shell, diachrony implies synchrony. A diachronic analysis of a language must
include not only the evolution of the structure of the language , but also an account of the
external forces i.e. political, economic, social and spiritual, which influence this evolution
in the course of time. (cf.J.Fisiak, An Outline History of English.Vol.1.External History,
1993:22).Thus, the internal history of English will include all the aspects of the
development of language structure, i.e. the evolution of phonology, grammar, vocabulary
and writing, while the external history of English deals with all non-structural (political,
economic, military and social) factors. which have exerted influence on the development
of the language and directed its course of development (op.cit.:23).

3
 Historical linguistics may be regarded as a special case of comparative linguistics (also
called historical-comparative linguistics), which is another branch of linguistics.

Historical-comparative linguistics
A major branch of linguistics , in which the primary concern is to make statements
comparing the characteristics of different languages
(dialects, varieties, etc.), or different historical states of a language. During the 19th c. the
concern of comparative analysis was exclusively historical, as scholars investigated the
relationships between such languages as Sanskrit, Greek, Latin, their hypothetical
antecedent (i.e. the Proto-language from which such languages developed) , and the
subsequent processes which led to the formation of language groups of the present day.
One of the basic methods used by comparative linguists was the comparative method, i.e.
the standard comparative philological technique of comparing a set of forms taken from
cognate languages in order to determine whether a historical relationship connects them.
It there was such a relationship , this analysis would then be used to deduce the
characteristics of the ancestor language from which they were assumed to have derived
( a process of ‘comparative’ or ‘internal’ reconstruction).

Applied linguistics
The primary concern of this branch of linguistics is the application of linguistic theories,
methods and findings to the elucidation of language problems which have arisen in other
areas of experience. The most well-developed branch of applied linguistics is the
teaching and learning of foreign languages, and sometimes the term is used as if this were
the only field involved . But there are several other fields of application, including the
linguistic analysis of speech pathology (clinical linguistics), the use of language in
mother-tongue education (educational linguistics) and developments in lexicography,
translation and stylistics. Applied linguistics uses information from sociology,
psychology, anthropology, and information theory as well as linguistics in order to
develop its own theoretical models of language and language use. There is an uncertain
boundary between applied linguistics and the various interdisciplinary branches of
linguistics, such as sociolinguistics and psycholinguistics, especially as several of the
latter’s concerns involve practical outcomes of a plainly ‘applied’ kind (e.g. planning a
national language policy). On the other hand, as these branches develop their own
theoretical foundations, the distinction between ‘pure’ and ‘applied’ is becoming more

4
apparent, and the characterization of research as being in ‘applied psycholinguistics’,
etc., is now more regularly encountered.

Contrastive linguistics (also contrastive analysis, contrastive studies)

The comparison of linguistic systems of two languages, e.g., the sound system or the
grammatical system. Contrastive analysis was developed and practiced in the 1950s and
1960s, as an application of structural linguistics to language teaching. It was claimed that
responsible for potential difficulties in foreign-language learning were structural
differences between the languages involved (This is known as contrastive analysis
hypothesis). However, as J.Fisiak (Papers and Studies in Contrastive Linguistics ,1972:
7)notices: “ From the beginning it has been accepted that the term ’contrastive studies’
should be used in a broader sense including both the studies of the differences and
similarities between two languages under comparison, for it is obvious that the ability to
differentiate also implies the ability to identify, i.e. differences and similarities are in
complementary distribution , and no complete characterization of one language vis-à-vis
another can be given without taking both these aspects into consideration.”

Fisiak (ibid.) proposes to recognize two basic types of contrastive studies:

1. GENERAL CONTRASTIVE STUDIES, which are a part of typological linguistics,

their aim being, among others, to construct an adequate model for the comparison of
two languages (including the formalization of such fundamental notions as
congruence, equivalence, correspondence,etc. ) to determine a method for
quantifying the diveregence and convergence of two languages or language
components, as perhaps, a new universal.
General contrastive studies are basic for specific theoretical contrastive studies (i.e.
Polish-English, Hungarian-English, etc.) which by using the model constructed by the
former should produce an exhaustive account of the differences and similarities
between a given pair of languages.
The relation between general and specific theoretical contrastive studies may be
considered as approximately parallel to the relation between universal grammar and
the grammars of particular languages.
2. GENERAL APPLIED CONTRASTIVE STUDIES belong to applied linguistics. They
should provide a proper model for the comparison of two languages for a specific

5
 purpose. (e.g. a simplification of the theoretical model for pedagogical purposes).
General applied contrastive studies should also provide methods for the prediction of
interference as well as for establishing the hierarchy of difficulty in learning the
categories in a foreign language,etc.
3. The results and methods of general applied contrastive studies and the findings of
specific theoretical contrastive studies should be utilized by specific applied
contrastive studies for a given pair of languages to facilitate the preparation of
teaching materials(e.g the appropriate selection, gradation and restrictions) the
construction of language tests and the choice of teaching strategies.

Typological linguistics
A branch of linguistics which studies the structural similarities between languages,
regardless of their history, as part of an attempt to establish a satisfactory classification ,
or typology , of languages. Typological comparison is thus distinguished from the
historical comparison of languages – the province of comparative philology and historical
linguistics.- and its groupings may not coincide with those set up by the historical
method. For example, in respect of the paucity of inflectional endings, English is closer to
Chinese than it is to Latin. One typological classification, proposed by the German scholar
Wilhelm von Humboldt (1768-1835) in the early 19th c. , established three main groups of
languages on structural grounds: isolating, agglutinative and fusional ; a fourth type,
polysynthetic, has sometimes been suggested. The morphological orientation of this
approach is, however, only one aspect of typological analysis, which can operate at all
linguistic levels (e.g. a phonological typology, in terms of consonant/vowel systems,
syllable structure, or suprasegmental patterns (e.g. ‘tone’ language, ‘click’ language).
It should be pointed out that typological similarities are not considered to be indicative of
genetic relatedness.

Psycholinguistics
A branch of linguistics which studies the correlation between linguistic behaviour and the
psychological processes thought to underlie that behaviour. There are two possible
directions of study. One may use language as a means of elucidating psychological
theories and processes (e.g. the role of language as it affects memory, perception,
attention,learning,etc.)and for this psychological linguistics is sometimes used.
Alternatively. One may investigate the effects of psychological constraints on the use of

6
language (e.g. how memory limitations affect speech production and comprehension). It
is the latter which has provided the main focus of interest in linguistics, where the subject
is basically seen as the study of the mental processes underlying the planning, production,
perception and comprehension of speech, and investigations typically proceed by
examining linguistic performance through small-scale experiments. The subject includes
a large number of research domains, e.g. child language acquisition, second language
acquisition, language processing, the relationship between linguistic and cognitive
universals, the study of reading, language pathology,etc.

Sociolinguistics
Sociolinguistics is a branch of linguistics which studies all aspects of the relationship
between language and society. Sociolinguist study such matters as the linguistic identity
of social groups, social attitudes to language, standard and non-standard forms of
language, the patterns and needs of national language use, social varieties and levels of
language, the social basis of multilingualism, , etc. An alternative name sometimes given
to the subject is the sociology of language. The study of dialects is sometimes seen as a
branch of sociolinguistics, and sometimes differentiated from it , under the heading of
dialectology. Included under sociolinguistics and sometimes referred to as macro-
linguistics are the study of language choice in bilingual and multilingual communities,
language planning, language attitudes,etc.

Other branches of linguistics include , among others, the following disciplines:

mathematical linguistics, statistical linguistics, ethnolinguistics, anthropological
linguistics, pragmalinguistics, ecolinguistics, and many others. For definitions of these
terms see D.Crystal, A Dictionary of Linguistics and Phonetics, 2011, J.C.Richards and
R.Schmidt, Longman Dictionary of Language Teaching and Applied Linguistics, 2002.

Philology
Philology is an earlier term for linguistics as the study of language that was practiced
before the 19th c. Nowadays philology and linguistics are treated as distinct types of study.
For example, in German tradition philology refers to the scholarly study of literary texts,
and more generally to the study of culture and civilization through literary texts.
According to the Polish scholar, Tadeusz Milewski (Introduction to the Study of
Language 1973: 37), philology is: “ the theory of defining and interpreting texts. And the

7
purpose of philology is, again to use the scholar’s own words, “ to reproduce the original
form of a text, to define the function of its elements in the process of communication, and
to establish the text’s history.”
The term ‘philology’ has been used differently at different periods and by different
‘schools of thought’. At present the term is also used in different senses. For example, in
Richard and Schmidt’s Longman Dictionary of Language Teaching and Applied
Linguistics ‘philology’ is treated as another term for’ comparative-historical linguistics’,
and David Crystal in his A Dictionary of Linguistics and Phonetics adopts the following
definition of philology: “ The traditional term for the study of language history, as carried
on by comparative philologists since the late 18th c. The study of literary texts is also
sometimes included within the term (though not in Britain), and is the study of texts as
part of cultural, political, etc., research. “

Traditional Grammars (the traditional approaches to language study)

The term ‘traditional grammar’ refers to the set of attitudes and procedures and
prescriptions characteristic of the prelinguistic era of language study, and especially of the
European school grammars of the eighteenth and nineteenth centuries. The emphasis on
such matters as correctness, linguistic purism, literary excellence, the use of Greek and
Latin models and the priority of the written language characterizes this approach. , and is
in contrast with the concern of linguistics for descriptive accuracy(criteria of analysis,
comprehensiveness, explicitness, objectivity, etc.). Traditional grammars were often
notional and prescriptive in their approach, i.e., they were based on the belief that there
are categories such as TENSE, MOOD,GENDER, CASE which are available to all
languages, although not all languages make full use of them. For example, a case system
is found in languages such as, e.g., German,Polish, Latin,Russian,etc., but not in
contemporary English. So traditional grammarians often stated that English has six cases,
because Latin, the linguistic model for them, has six cases. Well-known notional
definitions were of the NOUN as the name of ‘a person, place, or thing’, of the VERB as
‘a doing word’, of a SENTENCE as ‘ a complete thought’, and so on. Such definitions are
typical of notionally-oriented grammars. Linguists are critical of the notional approach to
language study in so far as the notions involved are incapable of systematic and
consistent account of linguistic facts, and replaces it replaces it with an emphasis on
formal criteria. Traditional grammars also attempted to lay down rules of correctness as
to how language should be used. The traditional grammars, using such criteria as purity,

8
logic, history or literary excellence, aimed to preserve imagined standards by insisting on
norms of usage and criticizing departures from these norms. And in this sense, they were
prescriptive. They included such recommendations as, e.g., I should be used after the verb
be , as in It is I; whom should be used as the relative pronoun in object function, as in the
man whom I saw, and so on. Patterns like It is me and the man who I saw, which are
widespread vernacular forms, were condemned by them as being incorrect usage.
It should be pointed out that linguists are critical of the prescriptivist approach,
emphasizing instead the importance of descriptively accurate studies of usage, and of the
need to take into account language variation in explaining attitudes to language.
Variation in a language may be related to region (regional variation) to social class
and/or educational background (sociolect ) or to the degree of formality of a situation in
which language is used (style).
Linguistic data
There are three kinds of data that linguists make use of in their studies of particular
languages:
a. The corpus (pl. corpora) of utterances, i.e. the spoken or written manifestations of a
language. This type of data is always restricted in the sense that the number of
utterances that have been produced , though it may be astronomically large, is always
finite. Therefore a corpus-based approach is one which restricts itself to the analysis
of a limited sample of a language in this way. This approach to language study was
represented by American structuralists, e.g. by C.C.Fries in his book The Structure of
English published in 1952. One of the fundamental theoretical assumptions made by
American structuralists insisted on limiting all linguistic investigations to what can be
directly observable only. Thus, for them, corpora of utterances were the only reliable
type of linguistic data.
b. The user’s linguistic judgements(intuitions) about his language.i.e., the elicited
reactions of the ‘naïve’ speakers of the language in question. Generative linguistic
theory, with his leading representative Noam Chomsky, goes beyond the corpus,
including as part of the raw data for the analysis the language user’s linguistic
judgements (intuitions) about his language. Chomsky(op.cit.:p.4) (Aspects of the
Theory of Syntax,) points out that a corpus can never illustrate a whole language, but
will only reflect a partial or selective picture. He says “ a record of of natural speech
will show numerous false starts, deviations from rules, , changes of plan in midcourse,
and so on.” These mistakes are due to many factors, such as poor memory, being

9
 distracted while we are speaking; but the important thing is that we have the ability to
to recognize them for what they are and discount them. And this implies that we have
learned a way of distinguishing mistakes from correct utterance, for discriminating
between sentences. One can therefore postulate that people, as language users, have
mastered a system of rules which allows them to sort out language in this way, and
this system of rules is something that is not contained within the corpus, but lies
outside it, in the mind of the speakers of the language. The study of this system of
rules is ultimately more important than the actual sentences themselves. This system
of rules, i.e. the speaker’s knowledge of his language(which Chomsky calls
competence) enables the speaker to produce an infinite number of sentences, and it
should be distinguished from the speaker’s actual use of the language in real-life
situations (called performance). So, from the generativist point of view, the ultimate
purpose of the linguist is the description of the native speaker’s linguistic competence,
his tacit knowledge of his language. Since linguistic competence is not directly
observable, being a mental system, we have to resort to the native speaker’s
judgements(intuitions) about his/her language.
c. The third type of data made use of by some linguists is introspection, i.e., the
intuitions, or just linguistic knowledge, of the analyst himself who either is a native
speaker of the language he happens to be concerned with , or at least has a reasonable
knowledge of the language.

Some American structuralists were highly distrustful of linguistic descriptions arrived

at on the basis of linguistic judgements(intuitions). Concerning the use of linguistic
intuitions as potential sources of data, W.F.Twaddell(On defining the phoneme, [in:]
Readings in Linguistics M.Joos(ed.),1935 ) stated what follows: “ The linguistic
processes of the mind as such are quite simply unobservable; and introspection about
linguistic processes is notoriously a fire in a wooden oven.” This statement happens to
be in full agreement with one of the fundamental theoretical assumptions of American
Structuralism which insisted on limiting all linguistic investigations to what can be
directly observable only. Thus, for those linguists, corpora of utterances were the only
reliable type of linguistic data.

10
 Glossary of technical terms (not explained in the lecture):

agglutinative (agglutinating) language

A language in which various affixes may be added to the stem of a word to add to its
meaning or to show its grammatical function.For example, in Swahili wametulipa ‘they
have paid us’ consists of :
wa me tu lipa
they + perfective marker+us + pay

Languages which are highly agglutinating include Finnish, Hungarian,Swahili, and

Turkish, although there is not a clear-ct distinction between agglutinating languages,
inflecting languages and isolating languages. Sometimes agglutinating languages and
inflecting languages are called synthetic languages.

ancestor language
An earlier language from which the later one is descended by the ordinary processes of
language change. For example, a Proto-Indo-European is an ancestor of English.

cognate languages
Genetically related languages, i.e. languages sharing a single common ancestor (see
ancestor language)

fusional language
Another term for inflecting language (see also agglutinating language)

generative theory
A cover term for a variety of linguistic theories that have the common goals of a/
providing an account of the formal properties of language, positing rules that specify
how to form all the grammatical sentences of a language an no ungrammatical ones (the
principle of descriptive adequacy). b/ explaining why grammars have the properties they
do and how children come to acquire them in such a short period of time (the principle of
explanatory adequacy ). The major versions of generative theory are all associated with
the pioneering work of the American linguist Noam Chomsky.

inflecting (inflectional)language
In this type of language, the function of words is determined by their form, i.e., case
endings, prefixes, phoneme alternation, etc. Thus, in Latin and Polish the determining
noun in a nominal group is distinguished by the ending in the genitive, e.g., Latin domus
part-is, Polish dom ojca; the subject of an intransitive sentence ends in the nominative,
Latin Marcus dormit, Polish Marek śpi. The subject in a transitive sentence is also
characterized by the nominative case ending, and the patient by the accusative case
ending, as in Lat. Marc-us necat leon-em , Polish Marek zabija lw-a (Mark is killing the
lion). The inflecting type of language includes archaic Indo-European Languages
(Sanskrit, Old Persian, Ancient Greek, Latin,etc.) and the modern Baltic and Slavic
languages with the exception of Bulgarian, the Semitic and Japhetic languages, and the
Uralic and Altaic languages.

isolating language (also analytic language)

11
A language in which word forms do not change, and in which grammatical functions are
marked by word order, and the use of function words. Languages which are highly isolating
include Chinese and Vietnamese.

polysynthetic(or incorporating) language

The term is opposed to synthetic and analytic type languages. Polysynthetic languages
demonstrate morphologically complex , long word forms, as in the constructions of many
American Indian languages, and encountered occasionally in English , in coinages such as
anti/dis/esta blish/ment/arian/ism/s ( antidisestablishmentarianisms ).

language maintenance
A term used in sociolinguistics, referring to the extent to which people continue to use a
language once they are part of a community in which another language has a dominant
position. For example immigrant groups may maintain their language, out of a sense of
language loyalty, despite the dominance of the language of their host country(as often
happened in the USA).

language loss (also language death)

A term referring to the situation which arises when a language ceases to be used by a
community.

language planning(or language engineering)

Making deliberate decisions about the form of a language, especially when this is done more
or less officially and on a large scale. A term used in sociolinguistics for a deliberate,
systematic and theory-based attempt to solve the communication problems of a community by
studying its various languages and dialects, and developing an official language policy.

language variation
This term refers to differences in pronunciation, grammar, or word choice within a language.
Variation in a language may be related to region (dialect, regional variation), to social class
and/or educational background (sociolect) or to the degree of formality of a situation in which
language is used(style).

macrolinguistics
Sociolinguistic research that deals with sociological and social psychological phenomena and
which studies language use in society as a whole, including the study of language main-
tenence and loss. (see language maintenance, language loss).

proto-language
An unrecorded and hypothetical language which is ancestral to one or more attested
languages and whose properties are deduced by some process of reconstruction, most often
comparative reconstruction, in characterizing the ancestor of a language family(e.g.Proto-
Indo-European).

purism
A term used pejoratively in linguistics to characterize a school of thought which sees a
language as needing preservation from the external processes which might infiltrate it and
thus make it change, e.g. the pressures exercised by other dialects and langauges (as in
loanwords) and the variations introduced by colloquial speech. The purist concern is

12
considered misplaced by linguists , who point to the inevitability of language change as
areflex of social, cultural and psychological development.

universal grammar
The term is used to identify the main aim of those linguists who hold that the ultimate purpose
of linguistics is to specify precisely the possible form of human grammar, and especially the
restrictions on the form such grammars can take. Universal Grammar is a theory which claims
to account for the grammatical competence of every adult no matter what language he or she
speaks.

13
Lectures 2, 3 :
Language, langue. competence, parole, performance, design features of language,
language functions

Generally speaking, we may say that language is the most highly developed form of
human communication. An act of communication is basically the transmission of some
kind of message from a source( sender) to a receiver. Both the sender and receiver are
both human, and the message may be transmitted either vocally, through the air(spoken
language) or graphically, by marks on paper (written language). Sometimes animal forms
of communication are referred to by the term ‘language’, but there is little in common
between human and animal forms of communication. The discipline that is interested in
the properties of animal communication systems is called ‘zoösemiotics’. No animal
system of communication displays the creative potential and complexity of structure
which are typical of human language. Since the ‘language’ of animals is genetically fixed
and not culturally transmitted, it is the same in a given species wherever in the world that
species occurs. This holds not only ‘ wherever’, but also ‘whenever’ whereas human
language not only varies from person to person and community to community, but also
from generation to generation. Linguists consider language as distinctively human
phenomenon.

Some definitions of language

“language is the expression of ideas by means of speech-sounds combined into words.

Words are combined into sentences, this combination answering that of ideas into
thoughts”( H.Sweet, A New English Grammar)

As can be seen, this definition of language reflects the representational (or cognitive)
function of language: language as a tool for the expression of the speaker’s thoughts.

According to another linguist,Otto Jespersen (Essentials of English Grammar),

“Language is nothing but a set of human habits, the purpose of which is to give expression
to thoughts and feelings, and especially to impart them to others”.

This definition of language draws attention to three functions of language: the phrase “to
give expression to thoughts” refers to the cognitive function, the phrase “ to give
expression to feelings” the emotive function, while the phrase “ to impart them to others”
to the communicative function, which, by the way, is considered by the author as the most
important function of language, which is implied by the adverb especially used in the
definition.

The American structuralist linguist, A.A.Hill (Introduction to Linguistic Structures) puts

forward the following definition of language:
“ Language is the primary and most highly elaborated form of human symbolic activity.
Its symbols are made up of sounds produced by the vocal apparatus, and they are arranged
in classes and patterns which make up a complex and symmetrical structure. The entities
of language are symbols, i.e. they have meaning but the connection between meaning and
symbol is arbitrary and socially controlled.” The most important terms appearing in Hill’s
definition of language seem to be:

14
symbols (“the entities of language are symbols”)
arbitrary (the relation between linguistic entities and their meanings/referents is not
necessary, essential, but arbitrary and conventional)
socially controlled (this phrase implies that language is a social phenomenon, in the sense
that it is common to all members of a given community, if it were not, communication
between the members would be impossible)
sounds ( this term implies that spoken language is chronologically prior to written
language)

What is also meant by ‘language’ is what F.de Saussure (Cours de linguistique

generale)called langue (distinguished from parole) and what N.Chomsky(Aspects of the
Theory of Syntax) calls linguistic competence (distinguished from performance) To use de
Saussure’s own words “Language is a form, not a substance.”

De Saussure’s dichotomy: la langue: la parole

De Saussure defines la langue as the system of linguistic signs and the relationships
holding between them. If two individuals speak the same language , they are said to share
the same langue.
La parole is identified with the products of the use of la langue (i.e. with actual
utterances).
The English term ‘language’ is often used in linguistic literature for de Saussure’s langue
and Noam Chomsky’s linguistic competence (or just competence).

Competence : performance

The distinction between competence and performance was first made by Noam
Chomsky (Aspects of the Theory of Syntax). Chomsky identifies competence with the
native speaker’s tacit knowledge of his language (not ABOUT his language!), and
performance is defined by him as the use of the speaker’s competence in particular
linguistic situations. Whereas for Saussure the repository of langue is the speech
community, for Chomsky the repository of competence is the ideal speaker/hearer.So
Saussure’s distinction is basically sociolinguistic, whereas Chomsky’s is basically
psycholinguistic.
It is the langue (linguistic competence) , i.e. the language system, which the linguist
describes (not parole or the speaker’s linguistic performance).
Competence , as our linguistic knowledge, can be linked to a set of linguistic skills that
enable us to use our language fluently. Thus, as fluent speakers of a language, we are
possessed of the following skills:

a/ the skill to tell grammatical from ungrammatical sentences;

b/ the ability to discover paraphrases of sentences (i.e. alternative versions of a sentence

or text without changing the meaning; the notion of paraphrase is based on semantic
equivalence). The phenomenon of paraphrase occurs on lexical,phrasal and sentence
levels of language. The following examples illustrate lexical paraphrase: aculist: eye-
doctor, conceal- hide, etc. Syntactic paraphrase is illustrated by the following semantical-
ly equivalent sentences: It’s obvious that he’s a liar: That he’s a liar is obvious.
c/ the skill to discover ambiguity

15
 Take,e.g., Latin Amor Dei where, depending on the context, Dei may be analysed as
Genetivus Obiecti, under which interpretation it functions as the receiver of love, or as
Genetivus Subiecti, under which interpretation Dei refers to the source of love (God love
people). To take one more example, They can fish, in which, again, depending on the
context , can and fish can be given various interpretations. Can may be taken to function
as either a modal auxiliary or as a full verb. Likewise, fish may be used as either a noun
or a verb.

Thus, ambiguity may be a function of either a constituent-structure or the distributional

classification of the ultimate sentence constituents. The example They can fish is
illustrative of the latter case. Ambiguity as the function of constituent-structure can be
illustrated by the phrase old men and women, where old may be taken to modify either
men ((old men) and women) or both men and women (old (men and women))

d/ the ability to find out about the internal structure of sentences. Consider the sentences:

1. He’s eager to please.

2. He’s easy to please.

Superficially 1 and 2 are identical sentences, but intuitively we know that they differ
crucially; while we can roughly paraphrase 2 by means of It’s easy to please him, we can-
not use it’s eager to please him as a paraphrase of 1, what’s more, this sequence is not
even a plausible English structure. In 2, he functions as the object of please, the pronoun
stands for the performer of the action described in the sentence.

Design (universal) features of language

Now we will be concerned with features of language that are assumed to occur in all known
languages.

1. Arbitrariness (or conventionality) This property of language is usually contrasted

with iconicity. Iconic signs are non-arbitrary signs, in that they are medium-
dependent. E.g. the English word cuckoo is iconic in the phonic , but not in the
graphic medium. The sound of the word resembles the sound made by by the animal
that the word refers to. Iconic signs, then, are signs where there is a resemblance
between the sign and what it refers to, that is, its referent ( for more on signs see
Lecture 3). What is, however, meant by ‘arbitrary’ with reference to language can be
explained by the following examples: In English we have a word tree, in Polish there
is a word drzewo, in Russian derevo, and in German Baum, etc.,etc.
No one of these words is more naturally appropriate than there are the other ones. This
shows that the relation between a word and what it refers to is not essential but
arbitrary, or conventional; the members of a given speech community simply agree to
call such and such an object by such and such a name. If the relation between a
linguistic form and what it stands for were not arbitrary, then a person who did not
know a given language, say, English, would easily guess the word if he /she knew the
meaning/referent , and vice versa.

There is a fair amount of arbitrariness in morphology and syntax. For example, the
same meaning/referent can be expressed by different linguistic expressions, as

16
 demonstrated by the following pairs of words: nonsensical: senseless, handsomer:
more handsome, etc. And in syntax: He came in and sat down (which sentence
reflects the order of the real events) , the same real-life situation can be also conveyed
by the sentence He sat down after he came in, where the order of the component
clauses is reversed.

There are voices claiming that the existence of so-called onomatopoeic words (e.g.
Polish bzz, kwa-kwa, kukuryku, miau, and the like) in language is evidence against its
arbitrary character. However, it is hard to agree with this stance, as, in actual fact,
onomatopoeic words are largely made to conform to the morphological systems of
particular languages, rather than being directly imitative of what they refer to. A
handy illustration of this is the existence of different-sounding words for the sound
made by the already invoked animal, the cock, in different languages. And thus, we
have Pol.kukuryku, French coquerico, German kikeriki, English cock-a-doodle-doo, all
referring to the sound produced by the cock.

To add to what I have already said, onomatopoeic words, imitative or echoic words
like, say, the English boom,splash, mutter, and the like, constitute only a small and
untypical minority in any language..

2. Duality To say that languages have the property of duality is equivalent to saying that
they have two levels of structural organization: phonological and grammatical, and
that the two levels are related, in that higher-level elements are composed of lower-
level elements (thus morphemes of phonemes, words of morphemes,etc.).
Duality makes it possible to distinguish a very large number of forms by combining a
relatively small number of lower-level elements in a variety of ways. For example,
various permutations of English phonemes (the lowest-level linguistic elements) yield
very large numbers of morphemes( next higher-level elements), while the number of
phonemes in English amounts to several dozen (both vowels and consonants), while
the number of morphemes goes up into hundreds.
Taken in conjunction with productivity, which is the next design feature of language
on our list, duality accounts for the fact that indefinitely many sentences can
beconstructed in any language out of a relatively restricted number of of phonological
segments. As N.Chomsky put it, “language makes infinite use of finite media.”
What’s more, duality is bound up with arbitrariness. If each phonological element in a
given word-form had to bear an identifiable relationship to some aspect of its
meaning, there would be severe constraints on combining phonological elements with
one another.

3. Productivity Productivity is still another property of language, such as enables us to

produce an infinitely large number of utterances, including potentially possible ones.
Productivity mustn’t be confused with creativity. All competent speakers of a
language are ‘productive’, but only some speakers , with special linguistic and other
abilities, can be ‘creative’. There are utterances whose novelty does not consist in the
fact that they have never been produced before, but in their originality of style, and it
is just for this kind of novelty that the term ‘creativity’ is most appropriate. Linguistic
creativity may manifest itself in the metaphorical use of words (not all speakers are
capable of creating metaphors!), or the unusual but stylistically affective combination,

17
of which a good illustration would be the utterance I saw him two cigarettes ago, and
similar utterances abounding with surprising metaphors.
Linguistic creativity, unlike linguistic productivity, is not rule-governed, and
consequently usually unpredictable.
The question that may be posed at this point is: How can a grammarian account for the
speaker’s ability to construct (and comprehend) an infinite number of utterances?
Well, grammarians have discovered that the grammars of particular languages
(particular grammars, e.g. English grammar) incorporate a number of rules which can
be reapplied indefinitely many times. Such rules are known as recursive rules in
linguistic literature. For example, the process by which an adjective comes to modify a
noun in English can be said to be ‘recursive’ in that we can, theoretically at least,insert
any number of adjectives in front of a noun. Consider some of the possibilities:
a tall boy, a dark tall boy, a nice dark tall boy, etc.

Coordination (or conjunction) is another linguistic device contributing to the infinity

of linguistic expressions of a certain type. Consider these types of coordination: phrase
coordination and clause coordination, illustrated respectively by: this prudent woman
and that naïve girl (where two phrases are coordinated: this prudent woman and that
naïve girl, yielding a larger and more complex phrase, more specifically a noun
phrase).
The structure of this coordinate phrase can be graphically presented like this:

NP

NP and NP

where and is a coordinating conjunction.

What follows is an example of ‘clause coordination’: John went to the cinema and
Mary went to the pub , and this can be graphically presented thus:

S and S

Clause embedding is still another grammatical device

contributing to linguistic productivity. For example,
expanding a simple (one-clause sentence) by means of the operation of clause-

18
 embedding , we can produce larger and more complex multi-clause sentences, as This
is the cat that ate the rat, This is the cat that ate the rat that lived in the house, This is
the cat that ate the rat that lived in the house that Jack built in the long ago, and so on
and on.
Like the other syntactic operations, clause-embedding can be potentially reapplied ad
infinitum.

4. Discreteness A design property of language (contrasting with the properties of other

semiotic systems), whereby the elements of a signal can be analysed as having
definable boundaries, with no gradation or continuity between them.If the elements are
discrete(separate), in the sense that the difference between them is absolute and not
subject to gradation, in terms of more or less, the system is said to be ‘discrete’.
Otherwise it is ‘continuous’, hence language is looked upon as a ‘discrete
combinatorial system’ A system lacking discreteness is said to be ‘ continuous’ or
non-discrete. The term ‘discreteness’ is especially used in phonetics and phonology to
refer to sounds which have relatively clear-cut boundaries. For example, in phonology
, the distinctive sound units,i.e. phonemes, are considered to be ‘discrete’ units. Thus,
the English word pin would consist of three such units (phonemes): /p/,/i/, and /n/.
Two word-forms, considered from the point of view of their form, are either
absolutely the same or absolutely different. For example, Polish słyszał and słyszał
are phonologically (and orthographically) the same , while słyszał and słyszałem are
different ( they are not’ more or less’ different, but just different).
Discreteness is rather unknown in animal systems of communication. For example,
bee-dancing is continuous rather than discrete and its productivity is dependent on this
fact. The bee body movements vary in intensity with a correspondingly continuous
variation of the source of nectar.

5. Semanticity This property implies that language allows us to convey any type of
meaning. There are various types of meaning: descriptive, social, expressive,etc.
(these will be the subject of one of our subsequent laecures). There are kinds of
meaning that can be conveyed only by human language, e.g. ‘descriptive’ (also called
‘symbolic or cognitive’) meaning. Linguistic expressions are used not only for social
communication, but also for the description of the extralinguistic reality, whereas
animal signals are only expressive and social.

6. Displacement Language also enables us to refer to (speak about) objects and events
that are remote in time and place from the time of speaking . That is to say, users of a
language are capable of speaking not only of what is going around them at the time of
speaking , but also about past and future events. Animal signals (calls) are generally
tied to specific situations, such as danger or hunger, and have nothing comparable to
‘displaced’ speech (unless, of course, this is artificially taught to them, as some
experiments with chimpanzees have tried to do).

7. Interchangeability By interchangeability we mean that any organism equipped for

the transmission of messages (i.e. encoding) in a system is also equipped to receive
(i.e. decode) messages in the system. As users of a given language, we are able both to
encode (speaker function) as well as decode (hearer/addressee function). Some animal
signals, by contrast, lack this property, e.g. female calls which are not shared by the
male members of the species.

19
8. Cultural transmission which is opposed to ‘genetic transmission’ (characteristic of
animal ‘languages’ ). The ability to use competently a given language is passed on
from one generation to the next by learning, not by instinct. People are born with
‘language faculty’ , but not with the ability to speak a particular language.

9. Learnability Closely related to cultural transmission is the property of language

called ‘learnability’. This property of language makes it possible for any human
individual of whatever race or intelligence to learn in childhood any language.
Noam Chomsky in one of his works goes even so far as to claim that human beings are
doomed to learn some language, that people are biologically equipped with, what the
linguist calls, ‘language acquisition device’. And in his Aspects of the Theory of
Syntax (p. 56) we come across these words:

“… a child cannot help constructing a particular sort of grammar to account for the
linguistic data he is exposed to any more than he can control his perception of solid
objects or his attention to line or angle.”

The linguistic data that the child is exposed to constitute the ‘input’ (dane wejściowe)
and the grammar underlying the data the ‘output (dane wyjściowe) of the LAD(i.e.
Language Acquisition Device). This can diagrammatically be shown thus:

INPUT (the primary linguistic data) → LAD → OUTPUT(grammar)

Language functions

In one of his major works Lives of the Poets, the 18th c. writer and lexicographer Dr
Samuel Johnson says that

“ … language is the dress of thought”, i.e., that language is primarily used to dress up
and send thoughts on their way, to give substance to thoughts.
Another famous 18th c. English writer Oliver Goldsmith said:
“ The true use of speech is not so much to express our wants as to conceal them.”
Of course, we must treat Oliver Goldsmith’s words as an artist’s view on language
functions.
As if alluding to Goldsmith’s words, the Danish philosopher Soren Kierkegard put it
this way:
“people use language not merely to conceal their thoughts, but to conceal the fact that
they have no thoughts.” This statement seems to imply a certain amount of irony and
even sarcasm. However, what we are really interested in are linguistic (scientific)
approaches to language and language’s functions.

The German philosopher and linguist Karl Ludwig Bühler spoke of three functions of
language:

1. representational function, 2. expressive function, 3. impressive function.

The representational function (elsewhere also called symbolic, cognitive) is

considered by Bühler to be the basic function of language. It consists in the use of

20
language as the expression of extralinguistic reality, which includes all actual and
fictional facts, i.e. products of our imagination. What’s more, language itself can be a
part of extralinguistic reality (language as the object of description).

The representational f. of language involves the existence of three elements: the

speaker, the linguistic sign, and extralinguistic reality. This function of language is
thus orientated towards extralinguistic reality.

2. The expressive function (also called emotive function) consists in the use of
language as the expression of the speaker’s own feelings and emotional states.
This function does not require the presence of the addressee. Expressions such,
e.g., as Ah! Oh! , various swearwords, and the like, tell us something about the
speaker’s emotional states at the time of utterance.
3. The impressive function (also called conative f.) consists in the use of language
to establish and maintain social roles: language is used to influence the addressee’s
behaviour, or just to get things done.

To this list of functions Roman Jakobson added three other functions: phatic, poetic
and metalinguistic.

4. The phatic function The function of language referred to in Bronisław

Malinowski’s works by the term ‘phatic communion’,which the scholar described
as “the type of speech in which ties of union are created by a mere exchange of
words”.
This use of language will include idle chatter, small talk, polite conversation about
anything and everything, and the like.

5. The poetic function (also called aesthetic f.) It is orientated towards the linguistic
text. It focuses on its phonological, grammatical and lexical properties.For
example, the poetic function of language manifests itself in puns (plays on words),
found in literary texts. The following pun My pony is a little hoarse, when spoken
is ambiguous, as it may mean either that the pony is a small animal or that the
pony has a throat infection.

6. The metalinguistic function We have to do with this function when a language is

used to describe a language, e.g., the statement ‘The word group the tall boy is a
noun phrase and functions as the subject of the sentence The tall boy is a big liar’
is a metalinguistic statement (here English is used to describe English).

7. The communicative function Neither Bühler nor Jakobson mention the

communicative function of language, a function distinguished by the French
linguist André Martinet, who considered it to be the most important language
function. Martinet viewed language, first and foremost, as an instrument of social
communication.This function involves four elements: the speaker, extralinguistic
reality, linguistic sign, and the addressee.It is an all-pervasive function, as it often
overlaps with the other functions. As it is, most utterances accomplish more than
just one function at the same time. For example, an utterance such as I’ m not
inviting the Smiths again, with appropriate intonation, signals an intended future
action (cognitive function), may show that the speaker does not like the Smiths

21
 (expressive function), it is part of a conversation in which the interlocutors share a
relationship that permits such expressions of dislike (impressive function ).

8. The performative function of language is another one distinguished by some

linguists (cf. Lewandowska-Tomaszczyk 1993: 18). It consists in the use of
language to perform certain, frequently ritualized, acts, as is the case with :

I pronounce you man and wife, or I name this ship the Athena

For language vs. speech see page 103.

Glossary of technical terms

constituent A basic term in grammatical analysis for a linguistic unit which is a fun-
ctional component of a larger structure. For example, the constituents
of the noun phrase a big boy are the indefinite article a, the adjective
big and the noun boy.

constituent-structure (phrase structure,syntactic structure) The constituent struc-

ture of a linguistic expression is a representation of the set of con-
stituents which the expression contains. The constituent-structure is
usually represented in the form of a labeled bracketing or a tree dia-
gram. For example, the following tree diagram represents the con-
stituent-structure of the sentence The man hit the ball:

NP VP

Det N Verb NP

the boy hit Det N

the ball

where the S(entence) consists of two constituents ; the NP(noun phrase: the boy), which
occurs to the left of the VP(verb phrase: hit the ball) ; this NP consists of Det (the) and
N(boy); VP consists of two constituents Verb (hit) and NP (the ball); and the NP to the

22
 right of Verb consists of two constituents Det(the) and N(ball). The string of elements
the+boy_hit+the+ball are so-called terminal elements.

referent A term used for the entity (object, state of affairs,etc.) in the extralin-
guistic world to which a linguistic expression relates. For example, the
referent of George Bush is George Bush himself.
lexeme The smallest unit in the meaning system of a language that can be dis-
tinguished from other similar units. A lexeme is an abstract unit. It can
occur in many different forms in actual spoken or written sentences and
is regarded as the same lexeme even when inflected. For example, all in-
flected forms such as give, gives, given, giving, gave belong to the one
lexeme GIVE.

lexicographer A person who writes and compiles a dictionary.

morpheme The smallest meaningful unit in a language. A morpheme cannot be di-

vided without altering or destroying its meaning. For example, the En-
glish word kind is a morpheme. If the d is removed, it changes to kin,
which has a different meaning. Some words consist of one morpheme,
e.g., kind or kin, others of more than one, e.g., un-kind, consisting of
the prefix un- and the stem kind.

onomatopoeic words The term refers to words that are considered by convention to be
imitative of nature, acoustically similar to the things to which they
refer. For example, the bow-wow of a dog, or the tick-tock of a clock,
etc. Languages differ in the range, choice and phonetic realizations of
onomatopoeic words. For example, an English dog goes bow-wow,
while a Polish one goes hau-hau, and so on.

phoneme The smallest unit of sound in a language which can distinguish two
words. For example, in English, the words pan and ban differ only in
their initial sound: pan begins with the phoneme /p/ and ban with the
phoneme /b/. The number of phonemes varies from one language to
another. English is often considered to have 44 phonemes, 24 con-
sonants and 20 vowels.

phonology The area of grammar concerned with how speech sounds function to
distinguish words in a language. The scope of phonology includes
how sounds are related, how they are combined to form syllables and
larger units, and how relationships between syllables are indicated by
features such as stress.

sign, symbol (for these terms see the lecture on Communication )

utterance When the linguist sets out to describe the grammar of a language on the
basis of a recorded corpus of linguistic material, he starts with a more

23
 ‘primitive’ notion than that of either the word or the sentence (by
‘primitive’ we mean here ‘undefined within the theory’ or ‘pre-
theoretical’).
This more primitive notion is that of utterance. The notion of utterance
is more primitive than those of ‘word,’ morpheme’ or ‘sentence’ in that
its application does not rest upon any technical definitions or postulates
of the science of language.
Z.S.Harris in his Methods of Structuralist Linguistics, 1951) states that:
“The utterance is any stretch of talk by one person before and after
which there is silence on the part of that person.” Thus, following
Harris’s definition, we could say that an utterance can consist of
single words, phrases, sentences,etc.
In Ferdinand de Saussure’s theory, utterances are stretches of parole
(for this term see lectures 3,4 ) produced by native speakers out of sen-
tences generated by the system of elements and rules which constitute
the langue( lectures 3,4).

word-form Word-forms are the various forms that lexemes assume in different syn-
tactic surroundings. For example, English give, gives, gave, giving,
and given are word-forms, i.e., different textual realizations of the
lexeme GIVE (for lexeme see above ).

24
Lecture 4
Non-verbal Communication

Communication refers to the transmission and reception of information (a message) between

a source and receiver using a signaling system: in linguistic contexts, source and receiver are
interpreted in human terms, the system involved is a language, and the notion of response to
the message becomes of crucial importance. In theory, communication is said to have taken
place if the information received is the same as that sent. Human communication may take
place using any of the available sensory modes (hearing,sight,etc.), and the differential study
of these modes, as used in communicative activity, is carried on by semiotics.
The total communication process comprises not only language (i.e. human verbal
communication) but also various kinds of non-linguistic signals (i..e., non-verbal
communication) such as, prosodic features, gestures, body movements, facial expressions, the
use of space, etc. ‘Verbal’ here implies that language is basically a matter of words.
Thus, the non-verbal component of language behaviour (human language communication)
consists of non-vocal as well as vocal elements.

human communication

verbal communication non-verbal communication

(by means of words) (paralinguistic system,
comprising vocal and non-vocal
elements )

The paralinguistic system is superimposed on the verbal system, in that it adds extra shades of
meaning to utterances, that is to say, it is complementary to the verbal system. Part of the
paralinguistic system of communication is constituted by such prosodic features as stress and
intonation.
It may be said that the domain of stress is the word, while its focus is one particular syllable
of the word. For example, if we put stress on the verb form seen in the utterance I haven’t
SEEN her, it might be taken to imply that, though I have not seen her, I have news of her
(that I happen to know her whereabouts).

25
Stress may serve to distinguish one word from another in some languages, including English.
Consider,e.g., the form produce, when taken as a noun, it is stressed on the initial syllable,
and when it is used as a verb, the stress falls on the second syllable. In such cases, word-
stress plays the same role as does the presence or absence of derivational suffixes, like, e.g.
English – ment , – al, etc. (thus stress can serve to distinguish derivationally related words):

develop + - ment > development

V N

Arrive + - al > arrival

V N

where V=verb, N=noun.

In some languages word-stress serves to distinguish inflectionally, rather than derivationally

related words. For example, Russian góroda ‘of the city’, as in the phrase krasota góroda
(the beauty of the city) , and gorodá ‘cities’, as in gorodá Rosji ‘cities of Russia’ , or Spanish
canto ‘I sing’ and canto ‘He/she sang’.

All this indicates that stress, like consonants and vowels, is an element of word-stress in
some languages.

Intonation , one of the two prosodic features mentioned earlier, may be generally described
as the pattern of rising and/or falling pitches (wysokość głosu) with which a sentence is
uttered. Pitch patterns make their distinctive contribution at clause/sentence level. Every
normal sentence is produced with a particular intonation pattern , which is determined partly
by the grammatical structure of the sentence and partly by the attitude of the speaker, which
may be ironical, surprised, sardonic, and the like. Stress and intonation are superimposed on
the strings of forms constituting the verbal component of communication.

The elements of the paralinguistic system play a supporting role in normal communication by
means of spoken language.

While the verbal system is more closely associated with the description of certain states of
affairs, the paralinguistic non-verbal component fulfils the social and expressive functions
of language. For example, uttered with appropriate prosodic features a sentence, like He’ll
do it tomorrow may be interpreted as a command, rather than a statement about some future
state of affairs.
However, what prosodic features cannot do is to change completely the descriptive meaning
(here a statement) , in the way that a replacement of one word for another can, e.g., He’s
clever and He’s foolish , where the adjective clever has been replaced by the adjective foolish.

Non-prosodic vocal elements of the non-verbal component comprise: volume scale (siła
głosu), which comprises loudness and softness, tempo scale (tempo mówienia), pitch scale
(ton), which has to do with tone of voice, which may be low or high, etc. The volume scale,
which comprises loudness and softness, can suggest to the listener that the actual meaning of
an utterance is different from its literal meaning. For example, oversoftness can be used to

26
invoke suspense in a story developed by the speaker, as in Do you think I’m stupid? , which,
depending on how we utter it, may function as a genuine question, and thus requiring an
answer, a very unlikely interpretation in normal language use, or as meaning ‘ Don’t think
that I am stupid’, i.e.’ I want you to know that I am not a fool.’

We often hear a remark , like It’s not what he said, but the way he said it, which testifies to
the recognition by the hearer of the significance of such vocal features.

All these examples show that such paralinguistic vocal features may contradict the meanings
conveyed by verbal components of utterances.

Let us now consider another paralinguistic feature or component , pitch (wysokość głosu),
which may be high or low. For example, we tend to use extrapitch, i.e. high tone of voice,
when addressing children, as in Now just what do you kids think you’re up to? . High pitch is

usually interpreted to indicate strain or excitement, whereas extralow pitch is a sign of

displeasure , disappointment or weariness.

The tempo scale, like the other scales, also varies: the speed, with which we speak, will vary
depending on a variety of factors, prominent among them being our communicative intentions
and the circumstances in which a given communicative act takes place.

The functions of paralinguistic components in normal language behaviour are of two types:
one type is what is called modulation and the other type is termed punctuation. By
modulation is meant the superimposing on an utterance a specific attitudinal colouring
indicative of the speaker’s involvement in what he is saying and his intention to impress or
convince his hearer (or his audience).

So far, I have provided you with some examples of the use of some of the paralinguistic vocal
components in this modulating function.
In the modulating function appear also non-vocal paralinguistic elements, such as body
movements, facial expressions, etc. For example, among the non-vocal paralinguistic
elements having a modulating function is the nodding of the head, with or without an
accompanying utterance, indicative of agreement. Interestingly, In Japanese, the nodding of
the head is indicative of disagreement, not agreement.
There are, of course, many other movements of head and hands as well as changes of facial
expression which modulate and punctuate utterances.

By the punctuation of an utterance is meant the marking of boundaries at the beginning and
end of an utterance and at various points within the utterance to emphasize particular
elements, or to segment the utterance into appropriate information units.

The discipline that studies the communicative functions of gestures, body movements and
facial expressions, relating to the use of language, is called kinesics (kinezyka). Another
discipline , proxemics (proksemika), is concerned with the speaker’s uses of space. It deals
with how people use the space between speakers and hearers in the process of
communication. There are appropriate distances for talking to friends, for communicationg
with strangers, for addressing one’s superiors,etc. The variations in interpersonal space are

27
often culture-specific. And can be analysed in terms of age, sex, intimacy, social role, and
other such factors.
As has been pointed out earlier, paralinguistic phenomena, both vocal as well as non-vocal.
Are an essential part of normal language behaviour. As D.Abercrombie (in”Paralanguage”
1968) put it,

“ We speak with our vocal organs, but we converse with our entire bodies.. Paralinguistic
phenomena… occur alongside spoken language, interact with it, and produce together with it
a total system of communication. The study of paralinguistic behaviour is part of the study of
conversation: the conversational use of spoken language cannot be properly understood unless
paralanguage elements are taken into account.”

The discipline that analyzes systems using signs and signals for the purpose of
communication (i.e. semiotic systems ) is called semiotics. The most important semiotic
system is human language, other semiotic systems comprise, Morse code, sign language,
traffic signals,etc.

Roman Jakobson’s model of communication

There have been proposed a number of models of communication, and among them that
suggested by R.Jakobson.
Jakobson’s model of communication assumes six factors, which must be present for an act of
communication to be possible. They are as shown below:

Context

Addresser Message Addressee

( encoder) (decoder)

Contact

Code
An addresser sends a message to an addressee. He recognizes that this message must refer to
something other than itself. This Jakobson calls CONTEXT.
By CONTACT the linguist means the physical channel and psychological connections
between addresser and addressee.
By a CODE is meant here a shared meaning system by which the message is structured.
Each of these six factors determines a different function of language. Thus, these factors
determine the following functions of human language:

1. Referential
2. Emotive
3. Conative
4. Phatic

28
 5. Poetic
6. Metalingual (or metalinguistic)

These functions are described in Lecture 3.

Categories of signs and their organization

C.S.Peirce distinguishes three categories of signs :

ICON In an icon, the sign resembles its object in some way; it looks or sounds like it.
For example, a photograph is an icon. Onomatopoeia is an attempt to make language iconic.

INDEX In an index, there is a direct link between a sign and its object: the two are
connected. For example, smoke is an index of fire (Consider the popular proverb: There is no
fire without smoke).

SYMBOL In a symbol, there is no connection between a sign and its object: a symbol
communicates only because people agree that it will stand for what it does.
For example, numbers are symbols; there is no reason why the shape 2 should refer to a pair
of objects. It is only by convention in our speech community or culture that it does.
Words could be given as another example of symbols. The arbitrary (conventional) nature of
words (linguistic signs) is the heart of human language.
F. de Saussure’s conception of linguistic sign

According to Ferdinand de Saussure ( Cours de Linguistique Générale ) , the sign is a

bipartite entity consisting of the signifier (Fr.signifiant) and the siginified (Fr.signifié). The
signifier is the marks on paper (letters), while the signified is the mental concept to which the
signifier refers. The mental concept is assumed to be broadly common to all members of a
given speech community.

sign

Signifier Signified
( physical (mental concept )
existence of
the sign)

29
The sign is then composed of the signifier and the signified. For example, the two letters OX,
if read as a word OX, function as a sign composed of the signifier( their very appearance) and
the mental concept ‘ oxness’, which, in this case, the writer , has of this particular animal.

Organization of signs

There are two ways in which signs are organized into CODES and SYSTEMS.

The first type of organization is by paradigms. A paradigm is a set of signs from which the
one to be used is chosen. For example, the set of shapes for road signs: square, round, triangle
shapes form a paradigm.

The second way of organizing signs is syntagmatic. A syntagm is the message into which
the chosen signs are combined. In language, we can say that the vocabulary is the paradigm,
and a sentence is a syntagm.
Thus, de Saussure distinguishes two types of relationships holding between linguistic signs
(words): paradigmatic and syntagmatic.

A paradigm, then, is a set from which a choice is made and only one unit of the set is chosen.
A simple example of this are the letters of the alphabet. These form the paradigm for written
language.

There are two basic properties of a paradigm:

1/ All the units of a paradigm must have something in common; they must share the
characteristics that determine their membership of that paradigm. For example, we must
know that, say, M is a letter and thus a member of the alphabetic paradigm, while 5 is not.

2/ Each unit must be clearly distinguished from the other units in the paradigm, which means
that we must be able to tell the difference between linguistic signs in a paradigm in terms of
both their signifiers and signifieds. The means by which we distinguishe one signifier from
another are called distinctive features.
Words constitute a paradigm. They are categorized into other, more specific paradigms,i.e.,
grammatical paradigms, such as NOUNS, VERBS, ADJECTIVES, etc.
Now consider:
u
bid
e
In the form bid the unit i is in a paradigmatic relationship with u and esince either of these
may be chosen to replace i ( bud, bed ). This is a relationship in absentia ( w nieobecności)
,i.e. between the element which is there i (in the form bid) and the units u and e , which are
not there in that particular form.

In the same word bid, the unit i is in a syntagmatic relationship with the preceding b and the
following unit d.

A sign enters into a syntagmatic relationship with other signs which precede and follow it
in a message. This is the a relationship in praesentia (w obecności), i.e., between the sign in
question and the preceding and following ones, which are all present in the message.

30
Let us take another example, a sentence, not a word, this time: Paul hit John. The verb form
hit is in a syntagmatic relationship with Paul and John, and Paul with hit.

To describe the extent to which the object determines the sign, Saussure employs such terms
as MOTIVATION and CONSTRAINT.
Thus, a HIGHLY MOTIVATED sign is a very iconic one; a photograph is a good example
of such a sign. It is more highly motivated than , say, a road sign. On the other hand, symbols
are entirely unmotivated.
In other words, the more motivated a sign is, the more its signifier (physical shape) is
CONSTRAINED by the signified (the object to which it refers). A photograph of a man is
highly motivated, for what the photograph looks like is determined mainly by what the man
himself looks like. A painted portrait is less motivated, more arbitrary.

And just one more thing. An important role in communication is played by CONVENTION.
Convention describes the rules by which signs work. It is necessary to the understanding of
any sign, not only words, i.e. symbolic signs.
29

Convention is the social dimension of signs; it is the agreement among the users about the
appropriate uses and responses to signs. Signs with no conventional dimension are purely
private and do not perform the communicative function!

Glossary (terms not described in the lecture)

derivational suffixes Suffixes used for the formation of new lexical items. For
example the English suffixes –al and –ment ( arriv-al,em-
ploy-ment). Derivational suffixes change the grammatical class
of morphemes to which they are attached.
descriptive (or
referential) meaning Also called denotative meaning, the term refers to that part of the
meaning of words or utterances which is stylistically neutral
objective meaning is sometimes used as its synonym).
(cf. expressive/connotative/affective meaning).

inflectional Related to inflection, which refers to a change made in the form

of a word in order to mark its syntactic function in the sentence.
Inflection is commonly marked by affixation ; e.g., the affix-
ation of - s to sing to form the ‘3rd person singular present tense’
of the lexeme SING. Inflection is often contrasted with derivation
within morphology.
intonation
(the pitch pattern, or
‘melody’ of a sentence) When we speak , we generally raise and lower the pitch of our
voice, forming pitch patterns. We also give some syllables in
our utterances a greater degree of loudness and change our speech
rhythm. These phenomena are called intonation. Intonation does
not happen at random but has definite patterns (see intonation
contour). Intonation is used to carry information over and above

31
 that expressed by words in the sentence.
intonation contour
also pitch pattern/con-
tour ) The pattern of pitch changes that occur across an utterance , often
accompanied by differences in loudness and speech rhythm.
Intonation contours may have grammatical functions. For
example, the word ready? - said with rising intonation- is a
question, but uttered with falling intonation is a statement(‘I’m
ready’). Some intonation contours/patterns are associated with
specific sentence types. For example, declarative sentences in
English typically have an abrupt pitch rise on the last stressed
Word of the sentence followed by a fall. And, to take one more
example, yes/no questions , like, say, Is language a social
phenomenon? Typically have a long gradual rise in pitch from
the beginning to the end of the sentence. (for pitch see below)

pitch when we listen to people speaking, we can hear some sounds or group
sounds in their speech to be relatively higher or lower than others.
This relative height of speech sounds as perceived by a listener is
called pitch. For example, in the Spanish question Listo? (‘ready’)
meaning ‘Are you ready?’, the second syllable –to will be heard as
having a higher pitch than the first syllable.

prosodic features These are sound characteristics which affect whole sequences of
syllables. They may involve, for example, the relative loudness or
duration of syllables, changes in the pitch of a speaker’s voice and the
choice of pitch level, i.e. the relative height of the pitch of a speaker’s
voice, as this is perceived by the listener. ( For example, for English
three levels of pitch have often been recognized: normal pitch level,
higher than normal level, lower than normal level. And these pitch
levels cannot be identified in absolute terms, since one person’s
high pitch will not be the same as another person’s high pitch. Thus
differences in pitch level are relative.

Lectures 5,6
Classifications of languages

Genetic classification

One of the ways to classify human languages is according to their ancestry, i.e., their common
origin. Languages which are genetically related have a common ancestor, i.e., they belong to a
single language family, the structure of which is typically exhibited in a family tree.
In historical linguistics, a language family is identified with a group of languages which all
derive from a single common ancestor, that is, they all started off at some time in the past as
no more than regional dialects of that ancestral language. The languages in a family are
therefore linked in a genetic relationship. The number of language families currently spoken
on the planet is around 300; further research will no doubt succeed in reducing this number
by combining some families into larger families. Speculative attempts at combining
recognized families into huge macro-families. The term ‘macro-family’ is applied to a

32
speculative and unestablished grouping of several established families, such as Austric,
Amerind and Nostratic. For example, Amerind refers to a proposed vast language family
joining all the languages of North and South America, except Eskimo-Aleut and Na-Dene.
However, the hypothesis, proposed by Joseph Greenberg(1987), is highly controversial, and is
thus rejected as unfounded by the great majority of specialists. The languages assigned to
Amerind constitute at least 150 distinct genetic families. Also deeply controversial are the
macro-families Austric and Nostratic.

As has been said in one of the earlier lectures, the branch of linguistics whose primary
concern is the classification of languages is called comparative linguistics, which is divided
into historical-comparative linguistics (concerned with genetic classification of languages)
and typological-comparative linguistics (concerned with the classification of languages on the
basis of presently existing common features). Thus a historical(genetic) classification of
languages is established according to the degree to which they have a common genesis , and
then analyses the processes of language evolution which led to this division of languages,
attempting to present, at least in part, the causes of particular evolutionary processes.

Established language families:

The Indo-European family

It is a vast language family occupying most of Europe and much of western and southern Asia
and formerly also occupying much of Anatolia and central Asia. Since the European
expansion, Indo-European(IE) languages now also predominate in the Americas, Australia
and New Zeland, and are prominent elsewhere. Today the family has more speakers than any
other family on earth. Indo-European languages are assumed to have had a common ancestor
language, called Proto Indo-European(PIE) , i.e. the hypothetical language reconstructed on
the basis of the large number of the living and recorded IE languages and whose properties
are deduced by some process reconstruction (see the glossary) or comparative
reconstruction.. PIE is unrecorded, since its speakers were illiterate, but it is usually
assumed to have been spoken about 6000 years ago (i.e. around 4000 BC). Where exactly
PIE was spoken is unknown and controversial. The issue has been debated for a century or
more , with no consensus in sight. The single most popular proposal is the Pontic steppes, but
the Near East, Anatolia, central Asia and the Balkans have all been defended,
while some linguists have argued for east central Europe, the Baltic coast, northern Europe,
and Egypt.

The IE family began to be recognized in the 18 th c.; Sir William Jones (1746-94), English
philologist and jurist, famous for translations from Greek and Oriental languages, is usually
credited with making the first public declaration of the family in 1786. Latin, Greek and
Sanskrit( see the glossary) were the first members to be recognized, followed in the next few
years by most of the other branches, though Albanian was not recognized as IE until 1854,
and Illyrian and Phrygian (extinct branches) only turned up in texts discovered around the
turn of the 20th century.

The branches of the IE family are as follows:

(1) the Indic group , including, in antiquity, the Vedic language and Sanskrit, which
derived from it, today including the various languages of Hindustan and Pakistan;

33
(2) the Iranian group, including present-day Parsian, Afgan, Ossetic;
(3) the Tokharian languages, which became extinct in the 8 th c. A.D., used in Central Asia
and Chinese Turkistan;
(4) the presently extinct Anatolian branch, the main representatives of which were
Hittitite and the Luvian language used in Asia Minor in the 2 nd millennium B.C., and
known from texts from the 14th and 13th centuries B.C. ;
(5) the Thraco-Armenian branch which, in antiquity included the Thracian dialects in the
Balkans and Phrygian in Asia Minor, and the only representative of which today is
Armenian;
(6) Greek, the oldest preserved form of which is the language of the texts written in so-
called Linear B -1450 to 1200 B.C.- and next, the language of Homer; the Macedonian
language was closely related to Greek;
(7) Albanian, the last remnant of the Daco-Mysian dialects used in antiquity in the land of
the Daco people and in the south-east part of the Balkans;
(8) the Illyrian and Messapian dialects in the Adriatic region;
(9) the Italic languages, including three sub-branches in antiquity (a) Latinic languages,
represented by Latin (today continued by the Romance languages), (b) Osco-Umbrian
dialects, (c) the Venetic language;
(10). the Celtic languages – today Irish, Scots, Welsh, and Breton;
(11). the Germanic languages – German,English, the Scandinavian languages, and
Gothic(now extinct);
(12). the Slavic languages - West Slavic(Polish, Lower Lusatian, Upper Lusatian, Czech,
Slovak), South_Slavic (Slovene, Serbo-Croatian,Macedonian, Bulgarian), East-Slavic
(Ukrainian,Russian,White-Russian);
(13). the Baltic languages- Lithuanian, Latvian, Old Prussian, which died out in the 17th c.

The hypothesis concerning the relation of all these languages is based on the fact that the
elements of their semantic and syntactic systems, word endings, vowel alternations,
grammatical morphemes, numerals, pronouns and basic vocabulary, can to a great extent
be traced to a single proto-source (= Proto-Indo European language) .

As has been indicated above, the scientific method, that allows to establish genetic
relationships among languages and hence to identify language families, is the
comparative reconstruction method. It constitutes a part of the comparative method,
i.e., that part in which the linguist, on the basis of systematic correspondences sets up a
system of proto-phonemes in the ancestral language being reconstructed and constructs
a proto-form for each word or morpheme assignable to that language as a sequence of
those proto-phonemes; where possible morphological paradigms and other grammatical
features are also reconstructed. Thus, the unrecorded language that emerges from this
process is a proto-language, i.e., the ancestor of the languages used in performing the
reconstruction.

In practicing reconstruction, linguists begin with the earliest actual data available for the
members of a language family, whether written or spoken , and attempt to ascertain what
earlier stages of the languages , or ultimately their common ancestor, might have been
like. For instance, below some words cognate with the English word ewe, from various
IE languages, are given ; by the method of linguistic reconstruction (comparative
reconstruction), these can be traced back to a projected ancestral form in PIE. This proto-
form is given below with a preceding asterisk, to indicate that there is no direct evidence

34
for it; the linguist has no texts and no speakers, and must rely on comparative
reconstruction using the daughter forms to hypothesise what the word would have been:

Proto-Indo European *owis

Lithuanian awis Greek ois

Luwian hawi Sanskrit avis

Latin ovis English ewe

Old Irish oi

There are several identifiable steps in the comparative method. First, we must satisfy
ourselves by inspection that there exists a prima facia case for genetic relatedness among
the languages of interest and hence cause to proceed further . When the languages are very
closely related, this may be easy; for example, no one looking at several Slavic languages
would fail to conclude quickly that they must be related. When the languages are more
distantly related, the step is more difficult. Mere lists of miscellaneous resemblances do
not constitute evidence, though it may be possible to extract evidence from such lists by
careful statistical methods. Similarly, one can attach no weight to shared typological
characteristics like: word order, ergativity, tones, vowel harmony, etc. The most
persuasive diagnostic evidence of relatedness is grammatical correspondences and
shared anomalies, i.e., features that are almost totally inexplicable by anything other than
common ancestry. Languages with little or no morphology cannot exhibit this kind of
evidence, and such languages are accordingly often more difficult to work with.

Secondly, once the linguist is satisfied with clear evidence of relatedness, he may proceed
to the next, second step, that is to say, to the identification of systematic correspondences
in sound between words and morphemes of similar meaning or function, and with the
identification of cognate sets . This second step is only possible when the languages
involved are moderately closely related. If we can find enough recurrent
correspondences, we can proceed to the third step, i.e., setting up a proto-phoneme for
each correspondence and assigning a provisional phonetic value to each proto-phoneme,
as represented by the symbol chosen. Once we have done this, we can go on to what has
been earlier described as comparative reconstruction.

Glossary of terms (not explained in the above text):

attested form A linguistic form which is securely recorded in writing or in recordings

of speech (and unattested form- A linguistic form in a language which is nowhere
recorded but which has been reconstructed or posited by linguists. Such a form is always
preceded by an asterisk).

asterisk The symbol ‘*’ , prefixed to a linguistic form for any of several purposes but
always indicating ‘ unattested’, ‘non-existent’, ‘impossible’. In historical studies, the

35
 asterisk marks , among others, a form which is nowhere recorded, but which has been
reconstructed by linguists, as e.g., when English head is traced back to an unrecorded
Proto-Germanic * haubudam , or English ewe (see above) traced back to the PIE *owis.

cognate (used as an adjective and a noun) A word in one language which is similar in
form and meaning to a word in another language because both languages are related. For
example, English brother and German Bruder. Sometimes words in two languages are
similar in form and meaning but are borrowings and not cognate forms, as e.g. kampuni in
the African language Swahili is a borrowing from English company(thus, these may be
said to be ‘false cognates’).

cognate set A set of cognate words or morphemes from several related languages,
sometimes particularly when these are being used as the basis for the reconstruction of an
ancestral form, or a set of ancestral forms.

ergative A term used for the case of the subject of a transitive verb when this
differs from the case of the subject of an intransitive verb. This term originally applied to
languages like Basque, in which the object of a transitive verb and the subject of an
intransitive verb are assigned the same case. However, by extension , the term has come to
be used to denote verbs like break , which occur both in structures like Someone broke
the window and in structures like The window broke; in both these sentence the window as
the object undergoing the action expressed by the verb break.

grammatical correspondences A systematic match in morphological systems between

languages. Grammatical correspondences, as a rule, constitute the most powerful evidence
that the languages exhibiting them are genetically related. A good example is the
pronominal paradigm shared by Algonquian ( also Algonkian, a major language family of
North America, covering the larger part of southern Canada and much of the northern and
northeastern USA. Among its better-known members are Arapaho, Blackfoot, Cheyenne,
Cree, Shawnee, and Kickapoo) with the distant Yurok and Wiyot languages of California.
In all these languages there are three pronominal stems – first person n-, second person k-,
and third person zero, and in all of them a –t- is inserted before a vowel initial stem. This
degree of grammatical correspondence is so great that it can hardly result from anything
but common ancestry, which led to the setting up of the Algic family of languages.

proto-form A hypothesized form from which cognates in various languages appear to

have descended. Proto-forms are usually asterisked in written linguistic texts. For
example, the PIE * owis, from which English ewe, Latin ovis, etc., are assumed to have
descended.

Sanskrit The most important of the ancient Indo-Aryan languages, recorded in the first
millennium B.C. and codified by the grammarian Pānini in the 5th or 4th century B.C. in the
form called classical Sanskrit. Sanskrit is still in the ritual of the Northern Buddhist Church.
Because of the antiquity of its written expression and the detailed descriptive analysis in the
Sutras of the Hindu grammarian Pānini, Sanskrit has been very important in the origin and
development of comparative Indo-European linguistics.

36
statistical methods Techniques employed for establishing general principles concerning the
statistical regularities governing the way words, sounds,etc., are used. Techniques used in
the studies concerned with the frequency and distribution of linguistic units and with
problems, such as stylistic distinctiveness, authorship identity, etc.

Typological comparison and classification of languages

Comparison and classification (or typology) of languages is the province of the branch of
linguistics called typological linguistics. Typological linguistics is distinguished from
historical linguistics which is concerned with the historical comparison of languages.
Typological comparison is based on general features of the systems and structures of
languages , and thus forms a wider process of classification of any observed phenomena
according to revealed similarities of form and structure. Its linguistic groups are set up
irrespective of historical language families, and may in part agree with them or cut across
their boundaries. Languages are classified according to the similarities of form they exhibit
with other languages at any level or levels(e.g. the phonetic level, phonological level,
grammatical level, etc.).

The phonetic level (phonetic typology)

At the phonetic level some groups of languages make use of similar ranges of sounds, while
others differ more widely from them in the types of sounds they employ. For instance, the
main articulatory positions, bilabial, dental or alveolar, and velar, and the main articulatory
processes, voice and voicelessness, plosion and friction, oral and nasal release, etc., are
employed in all, or in almost all, languages, but within these limits very marked phonetic
differences are found. Thus, languages may be grouped into classes on the basis of shared
phonetic features or types of articulation that play an important part in their phonological
systems while being only represented to a limited extent or altogether lacking in others. It has
been observed that that languages that occupy geographically contiguous areas , whether they
are related or not in descent, often share common phonetic features. To give just one
example, within the IE family , the feature of distinctive front rounded vowels, familiar in
French (in words like lune /lyn/, peu /po/, bæuf /bæf/ is found in neighbouring German,
Dutch, Swedish, Norwegian, and Danish, all Germanic languages, but not in English( which
is also a Germanic language), nor in the Romance languages, like Italian, Spanish or
Portuguese . A fact like this has been explained in part as the result of the spreading of
phonetic habits across genetic family boundaries as a result of population movements and the
effects of bilingualism.

The phonological level (phonological typology)

Of more significance appear to be typological classifications base not primarily on the

presence in languages of particular phonetic features, but on the different ways in which the
sounds and sound features are organized into phonological systems and syllable structures.
For example, the distinction between tone languages and non-tonal languages, according to
the functions assigned in them to difference of voice pitch represents a major typological
classification. In some languages, but notably in the languages of China pitch differences
help to distinguish one word from another and may be the only differentiating feature between
two or more words whose composition in terms of consonants and vowels is the same. Pitch
differences used in this way are called tones (also defined in one of our glossaries) and

37
therefore these languages are called tone languages. The different Chinese languages are the
most widely spoken group of tone languages.
One of the more noticeable differences between languages lies in the different types of
syllable structures permitted. Thus, while syllables of the type CV (consonant preceding
vowel) are found in every known language, English and German are distinguished from many
other languages by the degree of consonant clustering allowed in syllable initial and syllable
final position, and a language like Fijian is a language notable for the simplicity of its syllable
structures in terms of consonant and vowel elements occurring in the two positions and for its
paucity of consonant clusters. The possibilities of syllable structures in different languages
may be arranged on a scale, with those languages that permit the greatest degree of syllabic
complexity at one end, those with the least at the other. On such a scale, for example, some of
the Caucasian languages would come near one end, and, say, Japanese and some of the
Oceanic languages at the other end, with very simple syllabic structures.

In terms of prosodic phonology, languages may be classified according to the predominant

role played by particular types of prosodic features(defined in one of our earlier lectures).
Languages such as Irish, Russian and a number of other Slavic languages , in which prosodies
of palatalization and velarization play a prominent part in the distinction of lexical items and
different members of grammatical paradigms, suggest the lines on which such classifications
could well be made.

The grammatical level (grammatical typology)

At the grammatical level, languages may be classified according to the predominant

characteristics of their grammatical systems. In one such dimension, Chinese and several
other languages of South-east Asia stand at one extreme, in relying almost entirely on word
order and word class membership as the markers of syntactic relationships and sentence
structures, and are in contrast with languages, such as Latin and Ancient Greek, wherein word
order is relatively insignificant grammatically and syntactic relationships and sentence
structures are mostly expressed by morphological categories of concord and government,
exhibited by the word forms of different word classes. On this dimension, languages like
English,French, and German lie somewhere in the middle between the two extremes, with
English further towards the Chinese end of the scale than the other two, and French further
towards it than German.
Within the languages relying heavily on the categories of concord and government (e.g.
Latin,Ancient Greek, Sanskrit,Swahili, etc.) further classifications can be made according to
the types of categories involved (case,gender, number,tense,etc.).

The structural level (structural typology)

One typological classification of languages, proposed by the German linguist Wilhelm von
Humboldt (1768-1835), established three main groups of languages on structural grounds :
isolating (also analytic ), agglutinative and fusional (also inflecting). Humboldt’s
classification is more strictly a classification of the word structures of languages and of the
way in which phonological structures and grammatical structures are related in word forms.
Such features as isolation, agglutination( from the Latin agglūtino ‘I stick on to’, it involves
the systematic combining (fusion)of independent words into compounds without marked
change of form or loss of meaning) and fusion are usually to be found in most languages,

38
though in different proportions, and as a mode of typological classification the three types are
best regarded as directions towards which languages approach with greater or lesser purity.

Agglutinating languages

An agglutinating (also agglutinative ) language is one in which affixes may be added to the
stem of a word to add to its meaning or to show its grammatical function. For example, in
Swahili the word wametulipa meaning ‘they have paid us’ consists of :

wa me tu lipa

they + perfective marker + us + pay

Examples of highly agglutinating languages are Finnish, Hungarian, Swahili and Turkish.
There is no clear-cut distinction between agglutinating languages, inflecting languages and
isolating languages. Different languages display the feature of agglutination to a greater or
lesser degree. Sometimes agglutinating languages and fusional(inflecting) languages are
called synthetic languages.

Fusional (inflecting) languages

First let us define the term inflection, to which the adjective inflecting is related. Inflection
is a term used in morphology to refer to the process of adding an affix to a word or changing
it in some other way according to the rules of the grammar of a language. For example, in
English verbs are inflected for 3rd person singular: I work, he/she works and for past tense: I
worked. And most nouns are inflected for plural : horse- horses, book-books, etc. Thus a
word is said to inflect for 3rd person singular, past tense, plural, etc. In traditional grammatical
studies, the term accidence was used in this sense, as was the term flexion.
A fusional (inflecting) language is , thus, a language in which the form of a word changes to
show a change in meaning or grammatical function. See the above examples.
Fusional(inflecting) languages include Latin,Greek,Arabic,Polish,etc. In such languages the
inflectional forms of words may represent several morphological oppositions(for this term
see the glossary below).. For example, in Latin amo ‘ I love’ the form simultaneously
represents tense (present), active voice, first person singular (subject), indicative mood. This
fusing of properties has led to such languages being called fusional. As it is, languages
display the characteristic of inflection(fusion) to a greater or lesser degree.

Isolating (analytic) languages

The term isolating(analytic) characterizes a type of language in which all the words are
invariable and syntactic relationships are primarily shown by word order and the use of
function words. Languages which are highly isolating Vietnamese, Chinese and many
South-East Asian languages. For example in, Mandarin Chinese:

Juzi wô chi le
orange I eat (function word showing completion)
“I ate the orange”

Wo chi le juzi le

39
I eat (function word) orange (function word)

“ I have eaten an arrange”

In isolating languages, bound morphemes are rare, and words containing more than a single
morpheme are not thereby grammatically different in other respects from monomorphemic
words. Because the boundaries of syllables and morphemes largely coincide, these languages
are sometimes rather loosely referred to as monosyllabic. Words in such languages are
assigned to word classes on the basis of different syntactic functions, but they do not exhibit
any marks of such functions or class membership in their forms or morphemic structures.

English is in fact a fairly mixed type of language in respect of the three types, described
above, and each can be illustrated from English. For example, invariable words, such as
prepositions, conjunctions and many adverbs, are isolating in type. They exhibit no formal
paradigms, and in many cases they are monomorphemic (e.g. since, from, as, with, seldom,
now, then,etc.), and their grammatical status and class membership are entirely determined by
their syntactic relations with the rest of the sentences in which they appear, without formal
mark of these appearing in their own word structure. And morphologically complex words, in
which individual grammatical categories may be fairly easily assigned to morphemes strung
together serially in the structure of the word form, exemplify the process of agglutination;
un-god-li-ness, un-avoid-abl, etc., are examples of English agglutinative word structures.
And finally, words in which several grammatical categories are marked by word forms in
which it is difficult , if not impossible, to assign each category to a specific and serially
identifiable morphemic section are instances of fusional(inflecting) word structure. Thus, in
English often there is no clear distinction between the basic part of the word and the part
which shows a grammatical function such as number or tense, as in: mice _(=mouse+plural),
came (come+ past tense).

Polysynthetic (incorporating) languages

There is a fourth type of language that is sometimes introduced, viz., the polysynthetic or
incorporating type. Polysynthetic languages exhibit morphologically complex , long word
forms, containing numerous bound morphemes and including the subject, object and verb,
the translations of which would be represented by separate words in more familiar languages.
Such complex word structure occasionally occur in English, as,e.g., in : anti-dis-establish-
ment-arian-ism-s. Examples of such languages are Eskimo and some American-Indian
languages. Consider the Inupiaq Eskimo word kringakortuyok, which could be translated into
English as He has a big nose. It should be pointed out that some linguists are reluctant to
regard this tpe of language as distinct. For example, Robins (1967:334) has the following to
say in this matter:
“ The fourth class of language types (i.e. the polysynthetic group) , however, is of little use in
linguistic typology. Languages falling within it only differ from the agglutinative and fusional
types by carrying these processes to extremes and uniting within single grammatical words
what in most other languages one would find spread out among several words. Thus fourth
type introduces no qualitatively new feature of word structure; it only multiplies the numbers
of bound morphemes within its words. The muddled nature of this type is well illustrated by
the amusingly circular definition in the Shorter Oxford English Dictionary: ‘polysynthetic:
characterized by combining several words of a sentence into one word’.”.

40
Glossary (of terms undefined in the lecture text and earlier glossaries):

concord (also agreement )

A type of grammatical relation between two or more elements in a sentence, in which both or
all elements show a particular feature. For example, in English a 3 rd person singular subject
occurs with a singular verb, and a plural subject occurs with a plural verb (this is called
number concord): He walks- They walk. Concord may affect such grammatical categories as
CASE,GENDER, NUMBER,PERSON.

government A kind of grammatical relationship between two or more elements in a

sentence, in which the choice of one element causes the selection of a particular form of
another element. In traditional grammar, the term has typically been used to refer to the
relationship between verbs and nouns or between prepositions and nouns. For example, in
Polish the preposition z (with) governs , i.e. requires, the Instrumental Case: Paweł z
Andrzejem poszli do kina ( Paul and Andrew went to the cinema). A verb like oglądać(see)
governs the Accusative Case: Paweł oglądał ten film wiele razy (Paul has seen this film
many times). In the verb plus noun construction , the verb functions as the governer, and in
the preposition plus noun construction, it is the preposition that functions as governer.

monomorphemic word A word consisting of one morpheme, thus having no morphological

structure, e.g. English knight and kick are monomorphemic words.

opposition Two linguistic units may contrast with one another(be opposed to one another)
with respect to some feature. In such an opposition one of the terms is unmarked
(nienacechowany) and the other the marked (nacechowany) term. For example, English
consonants /b/ and /p/ are the two terms(człony) of the opposition based on the distinctive
phonological feature +/- voice; /p/ is the - voice(.i.e. voiceless) term of the opposition, /b/
the +voice (i.e., voiced) term. The term of an opposition which possesses a given feature (in
our example the +voice feature) is the marked term/member of the opposition
(nacechowany człon opozycji), while the other term is unmarked (nienacechowany człon
opozycji). Thus, in the following pairs of word-forms the plural forms are the marked terms
of the noun +/- opposition: day-days, book-books, etc.
The unmarked term of an opposition is more general in meaning and consequently has a wider
distribution (i.e., the number of context in which it can appear) than the marked form.
Consider the opposition /d/:/t/ in Polish. The marked member of the opposition /d/ never
occurs in word-final position, while the unmarked member /t/ can freely appear word-initially,
as well as word-finally; e.g., pot and tok. The division into marked and unmarked terms also
applies to morphological categories. For example, in the English system of moods the
Indicative Mood is unmarked, while the other moods are marked (Imperative Mood,
Subjunctive Mood). And in the Tense System, the Present Tense is the unmarked term. The
present tense , in contrast to Past Tense, can express various temporal relations: e.g., ‘present
time’, as in I work very hard, ‘ past time’, as in I was sitting at my table writing some letters.
Suddenly the wind bursts the window open and scatters all my letters on the floor, ‘ future
action’ , as in He starts for London tonight, and ‘universal truths’ , as in The earth rotates
round its axis. Light travels more quickly than sound.

41
paradigm A list or pattern showing the forms which a word can have in a grammatical
system. For example, in English:

singular plural

boy boys
boy’s boys’
(‘of the boy’) (‘of the boys’)

Paradigms may be used to show the different forms of a verb, as in Polish:

Singular Plural

ja mówię my mówimy
ty mówisz wy mówicie
on/ona mówi oni mówią

Paradigms typically show a word’s inflections (see inflection above).

word order An arrangement of words in a sentence. Languages often differ in their word
order. English is a language which heavily relies on word order as a mode of expressing
grammatical relationships within constructions. It is sometimes said that English is a language
with fixed word order, as opposed to a language like e.g., Polish or Latin, which are more
flexible in this respect, in which grammatical relations are signaled by inflections. Polish is a
fairly free word order language.

42
Lectures 7,8
Pidgins, Creoles, pidginization, creolization,word-formation inTok Pisin, decreolization,
pidgins,creoles and linguistic theory of change

Pidgins
Definition of a pidgin language: A pidgin is a contact language , developed in a situation
where different groups of people require some means of communication but lack any
common language.
A pidgin is nobody’s first language; all its speakers learn it as adults as a second or further
language, and all have native languages of their own.
Pidgins are typically restricted in their function, since they are used only in contact situations .
Along with the restriction in function goes a restriction in form.
Pidgins have at least three parent languages: one will typically be a European language, since
pidgins arose predominantly in the context of European colonization; this language is known
as superstrate (or superstratum). The superstrate is spoken by the dominant social group
(i.e. more socially prestigious in the community).
Two or more substrate ( or substratum) languages are usually involved, less prestigious and
often indigeneous languages (an indigeneous language = a language that is spoken by the
indigeneous (original) inhabitants of a country, e.g. Hawaiian and American Indian languages
in the USA and aboriginal languages in Australia).

Pidgins are not, as some would have it, ‘lazy, corrupted or debased’ forms of either the
superstrate or substrate languages, nor haphazard mixtures thereof. Pidgin languages have
rules of their own and they can be learned as languages in their own right. Pidgins not only
simplify the grammars of their superstrate and substrate languages, but also restructure them
to produce a new linguistic system.
Pidgin languages: Tok Pisin, New Guinea Pidgin, Bislama(Solomon Islands pidgin),Kriol,
Creoles

A pidgin becomes a creole when it is adopted as the native language of a speech community.
E.g., children may be born to parents who have no common language other then the pidgin.
Such children will be exposed to the pidgin as the initial linguistic input, and would acquire it
as a native language.
The creole has to fulfil many more functions than its pidgin ancestor, which was used only in
particular restricted situations( in the directive and referential functions) by speakers who had
a native language to fall back on; and creoles very quickly become expanded and elaborated
in all areas of their grammar.

Theories of origin of pidgins

There is considerable disagreement in the linguistic literature over how pidgin languages
arise, the five most popular theories probably being nautical jargon,
monogenesis,polygenesis, substratum/substrate theory and baby talk/foreigner talk.

43
Nautical Jargon Theory
This theory assumes that pidgins are derived from the lingua franca used by the crews of
ships, presumably through trading and other contacts. This hypothesis might explain some
lexical similarities among pidgins, but it is hard to extend this to cover the numerous
structural similarities. Furthermore, as Romaine (1988) notices, it is hardly surprising that
pidgins have nautical elements in their lexicon, since the majority are spoken near the sea.
Also other issues cast doubt on the credibility of this theory; for instance, such as the possible
way in which sailors would have transmitted their jargon to other people and also the matter
of whether the presence of nautical vocabulary items is indeed sufficient for assuming that the
jargon existed (cf. Singh, 2000: 44-45).

Substrate Theory

This theory is based on the division of languages into two groups: substrate and
superstrate.Those which belong to the former group are the indigeneous languages with
which the lexifier( i.e. a language which provides a large proportion of the lexical items
found in a pidgin or a creole) comes into contact during the formation of a pidgin (i.e.
pidginization). The latter group consists of the lexifiers, together with any other non-
indigeneoue languages which came into contact with the substarte . Thus the substrate theory
implies that “ the superstratum (=superstrate) or lexifier language contributes the vocabulary
to a pidgin , while its grammar comes directly from the substrate(s).” (cf.McMahon, 2000:
255).

Polygenetic Theory (independent parallel development)

This theory was formulated by R.A.Hall, jr.(1953) and it assumes that pidgins arose
independently. They are similar in structure because they are restructurings of similar
languages , with predominantly European superstrates and African substrates, which
developed in similar social and physical conditions. (McMahon, 2000).

Monogenesis with relexification

The essence of this theory is that all pidgins originated from one language. It is commonly
believed that this language was Sabir , the Mediterranean lingua franca, which was a 15th c.
proto-pidgin. It is suggested that it has evolved from the Portuguese language , which was
used in trading and as the language of the first expedition to India, West Africa and the Far
East. The monogenetic theory is associated with the name of Hesseling, who investigated the
influence of Malayo-Polynesian on the formation of Afrikaans, and in this research
emphasized the importance of answering the question how pidgins come into existence, and
thus, is regarded as the father of the theory. Romaine (1988:86) points out that the essence of
the monogenesis theory of pidgin origin is “.. . that all pidgins are genetically related to one
proto-pidgin. Thus, one accounts for their similarities by virtue of the fact that at some distant
point in time they had a common ancestor from which they are descended. And the
differences are accounted for by appealing to the process of relexification” ( The term
relexification appearing in this citation refers to the process of substituting the vocabulary of
one language with the lexical items of another. ).

Baby Talk/Foreigner Talk theories

44
The baby talk or foreigner talk theory both relate pidgin origin to second language
acquisition. It was Hesseling (1897/1933 in Romaine 1988:72), who “…put forward the view
that pidgins arose out of the imperfect learning of the model language on the part of the
slaves”. Another linguist, Schuhardt (1914 in Romaine 1988: 72) argued that the reduced
structure of pidgins came about as the result of a conscious effort at simplification by whites
in a master-slave relationship typical of colonial situations. Thus, the’ baby talk’ theory
characterizes the situation in which non-Europeans learned an imperfect version of the
superstrate language. And it is also argued that the European colonizers deliberately
simplified their native languages in order to make them easier for the substrate speakers to
acquire it (the ‘foreigner talk’ theory).
It has been sometimes stated that foreigner talk results not only from simplification but also
from other processes such as accommodation, imitation, or telegraphic condensation(see
below) (cf.Singh, 2000). Accommodation has several features which include changing the
tempo of speech into rather slow, making the sentences shorter and less complex, or
introducing pauses. For the sake of facilitation of comprehension, certain structures, usually
ungrammatical, may be imitated and reinforced. When deletion of lexical items, mainly
function words, takes place, we may speak of telegraphic condensation. The process is very
well described by the following statement by Schuchardt (quoted in Holm,2000:34):

“ For the master and the slave it was simply a matter of mutual comprehension. The master
stripped off from the European language everything that was peculiar to it, the slave
suppressed everything in it that was distinctive. They met on a middle ground… the white
man was the teacher of the black man. At first the black man mimicked him.”

Pidgin Structure

The term ‘pidgin’ is a cover term for languages at various points of a continuum. The first
phase involves the formation of jargons., which are extremely simple, ad-hoc systems of
communication, with very restricted structural/grammatical devices, great variation, and
sentences generally no more than one or two word long.

Jargons may in time acquire particular norms, become more complex, and extend into other
domains of usage, and as the result of this they become stable pidgins. It is argued that a
stable pidgin is not a product of a bilingual society, but that it occurs in a situation when
people who do not share a lexifier language are in need of a medium of communication.
(Mülhäusler, 1997: 138).
A pidgin language at this stage can be characterized by having both simple and complex
sentences and its grammatical structures are independent from the source languages
(Sebba,1997:105.). What’s more, its grammar develops gradually and is “…achieved at
different times in its different parts” (Mülhäusler, 1997:138).

Stable pidgins may expand, acquiring more morphology and more complex syntax, and a
characteristically faster tempo of speech. Such expanded pidgins are already partly
creolized in linguistic terms, although they are not yet properly creoles since they lack native
speakers.
Stable pidgins are used only in particular domains and fulfil certain restricted functions.
They are frequently used in the directive function (i.e., for getting people to perform particular
tasks) and referential function, which involves describing some situation to achieve a

45
particular end. They are rarely (if ever) used in the interactional function, which promotes
social cohesion. Neither are they used in the expressive, poetic, and metalinguistic
functions(for these functions see one of the earlier lectures).
All these restrictions mean that pidgins lack stylistic options, puns and metaphors, and have
few sociolinguistic markers , such , e.g., as politeness phenomena.

These restrictions in functions and contexts accompany a restriction on form, which itself
falls into two categories: pidgins are reduced, in that they have a reduced referential
capacity making certain meanings difficult or impossible to convey; and they are simplified
in that regularity in grammar is increased.

The lexicon of a pidgin is characteristically reduced by comparison with the superstrate and
substrate languages. Thus, normal languages have ca. 25-30,000 lexical items, while the
pidgin called Tok Pisin has about 500 items.
Each word in Tok Pisin has a wider range of meaning than is the case in the superstrate and
substrate, or covers a larger semantic domain. Words are often multifunctional, acting as
nouns, verbs and adjectives; and there is no compounding (for word-formation processes in
pidgins see lecture 9), so that the expression of complex ideas requires a good deal of
circumlocution and periphrasis. For example, liklik brum bilong klinim tit means
‘toothbrush’).

Pidgins rarely exhibit any inflexional morphology , so that no marking for GENDER, CASE,
NUMBER,TENSE, etc. occurs. Since pidgins lack inflexional morphology, words are
typically invariant with no allomorphy and no irregular forms; this makes the connection
between form and meaning as transparent as possible, and is of considerable help to non-
native second -language learners ( the minimal grammar and maximal reliance on the
vocabulary).

Besides, pidgins tend to have a fixed (invariable) word order, which is SVO (subject,
verb,object).

The phonology of pidgins is simple, with usually 5 vowels and no length distinction. The
consonantal system is also very small.

Creoles

Creolization is the linguistic inverse of pidginization (defined above), while pidginization

involves reduction and simplification, creolization is characterized by expansion and
elaboration. A pidgin becomes a creole when it acquires native speakers (i.e., when the
pidgin is their primary linguistic input (see above). Although the pidgin was adequate for
their parental generation, who used it only in specific contexts and otherwise had recourse to a
non-pidgin non-native language, it is too restricted a system for the children, who need a
native language to fulfil not only the referential and directive functions of language , but also
the expressive, pragmatic, poetic and metalinguistic functions, which pidgins do not perform.

Functional and linguistic expansion and elaboration typically go together: as the creole
becomes a native language of a community and is gradually extended into ever new domains,
its linguistic(structural) resources also increase.

46
An overview of creole morphology and syntax

In creolized Tok Pisin , speakers insert ol before a noun to mark plurality.

In Tok Pisin, often new affixes are the result of grammaticalization (term defined earlier);
e.g., Tok Pisin save ‘to know’ has become a habituality marker sa , it often coalesces with the
main verb and is analysed as a prefix.
Earlier circumlocutions give way to compounding in creoles. Thus Tok Pisin man bilong
save (=man belong know ‘expect’) is replaced in the creole by saveman.

Tok Pisin periphrastic causative constructions formed by using the auxiliary verb mekim, as
in mekim sam wara i^boir ‘ you make some water boil’ , which is replaced in the creole by a
shorter construction with the causative suffix –im on the main verb, as in You boilim wara.

While pidgins lack sentence embedding (see glossary below) and have mainly main clauses ,
constructions with embedded clauses tend to develop in creoles; e.g.:

Pidgin Tok Pisin: Mi no save. Ol i wokim dis pela haus.

Creole Tok Pisin: Mi no save olsem Ol i wokim dis pela haus.

(= I didn’t know that they built the house)

Creoles make no syntactic difference between statements and questions , although they have
question words which tend to bipartite; e.g., wa mek lierally ‘ what makes’ (=why), the initial
element being derived from the superstrate.
The passive construction is very rarely used in creoles.

Creoles have a system whereby a definite article is used for presupposed specific noun
phrases (NPs), which refer to a particular entity which the speaker assumes the hearer knows
about, and an indefinite article for asserted specific NPs, which have a particular referent
which the speaker is introducing to the hearer) . Thus, e.g.: mi tin e buki ‘I have the book’ ,
mi tin e bukinan ‘ I have the books’.(for more information on creole grammar see
Holm,J.A.2000. An Introduction to Pidgins and Creoles. Cambridge University Press).

Decreolization

In a situation when a creole language remains in constant and direct contact with its
superstrate the process of decreolization may take place. Bickerton,D.(1976) in his Pidgin
and Creole Studies observes that “ … a characteristic feature of this process is the emergence
of a linguistic continuum of varieties between the creole language and the standard language
which was the main contribution to the creole’s formation.” The so-called post-creole
continuum involves three stages: 1. basilect – the creole language, i.e., a linguistic variety (or
lect most remote from the the superstrate (prestige) language (the matrilect), 2. mesolect ,i.e.,
an intermediate linguistic variety (lect) falling between matrilect (also called acrolect) and
basilect. 3. The standard language to which the creole is corresponding.

Word-Formation in Tok Pisin

(based on Mülhäusler’ paper ‘The development of word-formation in Tok Pisin’ [in:] Folia
Linguitica, Acta Societas Linguisticae Europaeae 17/1-4, 1983) ( Tok Pisin – one of the
official languages of Papua New Guinea, now in the process of creaolization)

47
The jargon stage. Early pidgins did not possess a word-formation component, which was due
to two factors: a/ their lack of structural sophistication which prevented them from
intergrating borrowed lexical items; b/ their syntax was so primitive that even circumlocution
was barely possible.

Whenever complex lexical words were borrowed from a superstrate language, they were
treated as unanalyzed stems, thus:

Tok Pisin English

kolta coal-tar
trawsel tortoise shell
kaswel castor oil

The stabilization stage. The emergence of syntactic regularities in the stabilization stage
enabled the development of word-formation devices . The development of word-formation in
stabilized pidgins takes place in the context of reinterpreting and extending syntactic rules
rather than by way of generalizing borrowed patterns

The first evidence for pidgin speakers using language-internal devices to expand their lexicon
is circumlocution. For example, for the piano Papuan Pidgin speakers created things like
big fellow box, white fellow master fight him plenty too much, he cry (of the big box which the
white man beats so much that it screams) and also a creation like him big fella box, suppose
you fight him, he cry.

Later, the piano came to be called (in Pidgin English box bilong cry (’screaming box’ ).

Bilong

The transition from individual (private) circumlocution , like the examples given above, to
socially sanctioned circumlocutory phrases was made possible by the presence of the lexical
item bilong.
Concerning the use of bilong, Mülhäusler has this to say:

“ In view of the small size of the vocabulary it is often necessary to use circumlocutions for
new concepts, for example, for ‘half-moon’ small fellow moon, for ‘bed’ place bilong sleep,
for ‘to write,paint’ make bilong paper. Such neologisms are obviously ver dependent on the
individuality of the natives.”

Consider also the following examples with the use of the element bilong:

man bilong pekato ‘ a sinner’

man bilong les ‘ a lazy fellow’’
man bilong krikai ‘ a glutton’
man bilong kros ‘ a bad-tempered person’

48
Bilong has nothing to do with the original meaning it has in English, i.e., the superstrate
language, but indicates a perceptual field. It signals a number of relationships, and it can be
followed by nouns, verbs and adjectives.
It is believed that this kind of word-formation may have contributed to the development of
conversion (or zero-derivation).
Shortly after conversion a new word-formation device comes into existence, viz., the use of
the verbal causative affix – im, used to derive transitive verbs from nouns referring to
instruments, as in:

lock ‘lock,padlock’ - lokim ‘ to lock’

potograp ‘camera’ - potograpim ‘ to take a picture’
built ‘ glue’ - bulitim ‘to glue’

Even towards the end of the stabilization stage, the word-formation capacity of Tok Pisin
remains quite modest, and the lexicon is expanded mainly by borrowing from superstrate and
substrate languages. At this stage, there is no reduplication.

The expansion stage. This stage begins with Tok Pisin’s institutionalization as the inter-
tribal lingua franca in the late 1920s and its subsequent use in an increasing number of new
social domains and linguistic functions (apart from the directive and referential functions).

Expansion can be seen as a prolonged process characterized , at the level of word-formation,

by the following developments:
1. a shift from syntactic circumlocutions to lexical phrases and finally word-level lexical
items (i.e., single words become preferred as expressions of single concepts). For
example, aipas blind’, which has developed out of constructions like ai bilong mi I
pas ‘ (‘ my eyes are obstructed’), aipas being a compound word.
2. 2. A second characteristic of Tok Pisin’s expansion is the addition of new patterns to
its existing lexical grammar. Thus, word-level reduplication of words to indicate
intensity is observed around 1925, though it became a fairly productive process only
very recently.
At this stage, we also observe the disappearance of the use of the dummy verb mekim
to do’, followed by a noun to express various activities. Consider the following
formations:

mekim sigu ‘ to smoke’

mekim hos ‘to saddle a horse’
mekim tebol ‘ to lay a table’

( mekim was finally reduced to the affix – im)

3. Another characteristic of expanded Tok Pisin is that the increase in productivity of

some patterns appears to be a well-ordered development governed by the principle
that a new word-formation rule appears in the most natural environment first and
subsequently in less and less natural ones until it reaches full productivity, i.e. assumes
the nature of a categorical word-formation rule. Illustrative examples of this come
from the development of morphological causatives like bikim ‘to enlarge’ from
bik’big’, sarapim ‘to shut somebody up’ from sarap ‘quiet’, and the like.

49
Specialists have evidence that morphological causatives develop in a predictable sequence.
They appear to be most readily derived from stative intransitive verbs and least readily from
nouns. A natural developmental hierarchy of causatives looks as follows:

most common
Vintr. stative pinisim ‘to finish’ ( from pinis ‘ be finished’)
V(intr.)-stative pundaunim ‘to fell’ ( from pundaun’fall down’)

A bikim ‘to enlarge’ ( from bik ‘big’ )

Vtrans. dringim to make a drink’ (from dring ‘to drink’)

N no examples

least common

where V intr.stative= stative intransitive verb, Vintr.-stative= non-stative intransitive verb,

A= adjective, V trans.= transitive verb, N= noun

stative intr verbs > causatives (e.g., slip ‘to sleep, be horizontal’ – slipim ‘ to cause to lie
down’)
non-stative intr verbs > causatives (e.g., kamapim ‘to cause to come up’ - kamap ‘to come
up’)
adjectives > causatives (e.g., kolim’ to cause to be cool’ - kol cool’ )

The first example of a causative from a transitive verb was found in 1973:

Dokta i dringim sikman ‘ the doctor causes the patient to drink’

The appearance or non-appearance is also determined by a number of language-internal

restrictions on productivity. Thus:
a/ words ending in p- i cannot take the causativizer – im , which excludes forms like *
sambaiim’ to cause to help’ , and the like;
b/ the ideal word-length in expanded Tok Pisin is two syllables, hence the impossible *
wokabautim ‘ to cause to walk’. Thus, words with three or more syllables are increasingly
unacceptable.

c/ Word bases cannot undergo more than one morphological process at a time, which means
that no causatives can be derived from reduplicated verbs or adjectives, from compounds or
forms created by conversion. This restriction will then account for the impossible
*krukrugutim ‘to cause to be very crooked’.

Formation of agent nouns

Agent nouns provide another example of increase in morphological productivity in expanded

Tok Pisin. Take the following examples:

50
Didiman ‘ agricultural officer’, from the name of the founder of Botanical Gardens in Rabaul,
Dr. Bredeman.
Paiaman ‘ somebody who cares for the fire in a copra-dryer’
Busman ‘ bush dweller’

To signal ‘a person living in a certain place and the place itself’, a word stem referring to that
location is used, as in: Ostrelia ‘ an Australian, Australia’ , Siapan ‘a Japanese, Japan’,
Ailan ‘an islander, island’, etc.
There is a tendency now in the Tok Pisin of young speakers towards the use of the affix –
man for such purposes: e.g., yu lukim Amerikaman i draivim trak ‘ You see the Americans
drive trucks’.

Whereas in earlier Tok Pisin the same form appears in many meanings and functions , in its
later stages the principle of one form-one meaning is increasingly followed.

Just one more remark concerning agent formation; agents are also expressed by lexical
phrases of the form man bilong meaning ‘ sb who usually does what is referred to by the
verb.Examples of such phrases:

man bilong singaut ‘ noisy person, beggar’

man bilong slip ‘sleepy, lazy person’
man bilong stil thief’
man bilong toktok ‘ a talkative person’

and many, many similar others.

In Present-day Tok Pisin, however, in younger speakers’ language phrases are more and more
often replaced by compounds; e.g., man bilong mauspas > mauspasman ‘ mouth-obstructed
man, dumb person’.

Pidgins, creoles and linguistic theory of change

Pidginization and creolization have become the test case for any linguistic theory of change.
The question may be asked: Why does the historical linguist wish to concern himself with the
study of pidgins and creoles? There seem to be at least the following reasons for this: 1. These
processes provide conclusive proof against Leonard Bloomfield’s view, presented in his
magnum opus Language (1935), namely, that it is methodologically impossible to observe
linguistic change in progress. However, the point is that if we can actually see new languages
being born and trace their subsequent development, then we are certainly observing language
change. 2. Change also seems to proceed more quickly during pidginization and creolization
than elsewhere, so that the developments which can only be suggested for the history of
other languages can be directly observed or at least recovered in pidgins and creoles, without
resorting to apparent time-studies: this may allow the confirmation of various hypothesis on
the nature and spread of change. However, the processes found in pidginization and
creolization are only likely to be relevant to theories of change if they are of the same type as
changes which affect other languages.
P.Trudgill( On Dialect, 1983 ) tries to draw parallels between the changes characteristic of
pidginization and creolization and those found elsewhere. He suggests to divide changes into
two types: natural and non-natural changes. For him, natural changes are those changes

51
which are ‘liable to occur in all linguistic systems, at all times, without external stimulus’,
because of the inherent nature of linguistic systems themselves. Natural changes are likely to
include grammaticalization ( term defined in one of the earlier lectures), sound changes due to
ease of articulation (e.g. assimilation), increase in redundancy , i.e., double-marking of
categories , etc. These processes involve a general move from analytic to synthetic structure.

Non-natural changes – these on the whole involve reductions in morphologies (marking of

case, an increase in the use of prepositions, the reduction of conjugational classes for verbs,
and declensional classes for nouns, increased use of periphrases and reduced use of
inflectional forms , and the development of fixed word order). Thus, all these involve a shift
from synthetic to analytic structure .
Thus, Trudgill’s thesis is that languages which have undergone a good deal of linguistic
contact will exhibit more change , faster change, and specifically more ‘non-natural’ changes
than related languages which haven’t experienced such changes. The linguist claims that
pidgins, as high-contact languages, exhibit predominantly ‘non-natural’ changes, while
creoles show mainly ‘natural’ ones, which seems to be borne out by the available evidence.
As it is, pidgin languages do show a dramatic shift towards analytic structure and a reduction
to the point of loss of inflectional morphology, while creolization is characterized by
grammaticalization and increased redundancy which lead to a move towards synthetic
structure.

There are, however, some weaknesses in Trudgill’s position. For one thing, the distinction
between ‘natural’ and ‘non-natural’ changes is not quite clear, since a change which is already
in progress in a language may be reinforced by a model in another language with which it is
in contact; a particular change may be the result of language-internal phenomena on one
occasion, but motivated by contact , on another.
If we accept that certain types of processes are characteristic of both pidgins and high-contact
languages, while others operate frequently in creoles and low-contact languages, we can
regard pidginization and creaolization , not as different in kind from other processes of
language change, but as extreme cases of the hybridization that goes on in language all the
time. It is very probable that pidginization and creolization themselves may have formed part
of the history of languages which wouldn’t now be classified as pidgins or creoles. For
example, Middle English(ME) shows a large number of non-natural changes, involving the
loss of inflectional morphology. But if ME were really a creole, then we would have expected
it to have developed not only a relatively free word order, periphrastic constructions, but also
grammaticalized inflections to replace the ones lost earlier, which it hasn’t.
To sum up, pidginization and creolization share certain properties and features with other
cases of language change, but that does not make those cases pidgins and creoles: the
sociolinguistic context and the output are clearly different.

Glossary (terms undefined in the lecture text):

allomorphy Phonetic realization of a morpheme appearing in different contexts by means

of allomorphs.

allomorph Any of the different forms (textual realizations) of a morpheme. For example,
in Present-day English, the PLURAL morpheme is often shown in writing by adding – s to
the end of a word, e.g., cat-s. Sometimes this plural morpheme is pronounced as/z/, as in

52
dogs/dogz/, and sometimes as /iz/ as in glasses,and sometimes as /s/, as in books /buks/.
/s/,/z/,and /iz/ all have the same grammatical function in these examples, they all signal plural;
they are all allomorphs of the PLURAL morpheme.

accommodation The process in which, in a bilingual or multilingual community with no

dominant language , the languages converge towards each other, becoming more similar than
formerly, possibly to the point at which sentences in the different languages are morpheme-
by-morpheme glosses i.e.translations) of each other.

asserted Used for sentences (declarative sentences) which present information as

true, as opposed to those which ask questions(interrogative sentences), issue commands,
(imperative sentences)etc.

contact language Any speech variety which arises out of an instance of intense language
contact. Contact languages are highly variable in nature, but three types are commonly
distinguished : pidgins, creoles and mixed languages. Only the last two are mother tongues,
and they are sometimes called non-genetic languages. Examples of mixed languages: Media
Lengua (in Ecuador), Mbugu(or Ma’a ( in Tanzania), Michif (in the USA and Canada), etc.

conversion (also zero-derivation) The derivation of one lexeme from another without any
overt change in shape , e.g., the verb to father from the noun father.

dummy word A type of word which has no intrinsic semantic content, but which is
used to satisfy a structural requirement that a certain position in a structure be filled. For
example, the auxiliary do in a sentence such as Does he like pasta? Is said to be a dummy.
Likewise, the first occurrence of the pronoun there in the sentence There is nobody there is a
dummy (i.e. contentless) pronoun, as opposed to the sentence-final there(an adverb referring
to some place).

embedding A term used in GENERATIVE GRAMMAR to refer to the process

or construction where one sentence is included (embedded) in another. For example a relative
clause within a noun phrase is an example of embedding (the relative clause is included
within the noun phrase), as in The man who is watching us… .

function word Function words are words which have little meaning on their own,
but which signal grammatical relationships in and between sentences (they have grammatical
meaning). Function words include: conjunctions, prepositions, articles, etc. They are opposed

to content/lexical words , i.e., words which refer to things, qualities, states, and actions; they
have lexical meaning.

grammaticalization A term which refers to a process whereby a lexical word begins to

function as a grammatical/functional word. Consider,e.g., Polish będę in Jutro będę w
Warszawie, where będę is used as a lexical verb, and będę in Zaraz będę pisał list do
rodziców, where będę occurs as a future tense auxiliary.

(linguistic)input A term used in psycholinguistics to refer to the external linguistic

data available to speakers in the course of acquiring a language. The notion is especially
relevant to child language acquisition of a mother tongue.

53
invariant (invariable) words Those words which are used without any
morphological change, e.g., English under, but, them.By contrast, variable words are those
which inflect, e.g., house, houses, drink, drinks, drinking, etc.

morphological productivity In morphology, if a process is fully regular and actively

used in the creation of new words, this process is considered productive. For example, the
English suffix – ness, can be attached practically to any adjective to make a noun,
e.g.,happiness, emptiness, niceness, etc. And this can be contrasted with unproductive
suffixes, which as –th , are no longer used to form new words.

presupposed The information assumed (presupposed) by a speaker;that part

of a sentence which is assumed to be true by the speaker and the hearer, and which remains
as such when the sentence is negated; e.g., in one possible interpretation of this notion, the
sentence Where’s the salt? is said to presuppose that the salt is not present to the speaker, that
there is someone who the speaker thinks might know where the salt is, and so on. (see
asserted )

reduplication A kind of morphological process which involves reiteration of

the whole stem ( or root) , or just a part of it( the initial or final segment or syllable); e.g.,
Indonesian kitab ‘book’ and kitab-kitab ‘various books’ (complete reduplication),
Tagalog tawa ‘a laugh’and tatawa ‘one who will laugh’ (partial reduplication)

stem That part of a word to which an inflectional affix is or can be

attached. For example, in English the inflectional suffix – s can be added to the stem work to
form the plural word-form works, as used in e.g. the works of Hemingway.

54
Lecture 9,10
Dialect (v.language),accent, dialect levelling, dialectology, types of dialects,Standard
British English, Cockney

Language or dialect?

Some current definitions of ‘dialect’:

D.Crystal’s (2008) definition goes like this:

“ A regionally or socially distinctive variety of language, identified by a particular set of

words and grammatical structures. Spoken dialects are usually associated with a distinctive
pronunciation, or accent. Any language with a reasonably large number of speakers will
develop dialects, especially if there are geographical barriers separating groups of people
from each other, or if there are divisions of social class. One dialect may predominate as the
official or standard form of the language, and this is the variety which may come to be
written down.”

A more or less similar definition is formulated in Richards and Schmidt (2002):

A dialect is “… a variety of language, spoken in one part of a country (regional dialect), or
by people belonging to a particular social class (social dialect or sociolect ), which is
different in some words, grammar, and/or pronunciation form other forms of the same
language. A dialect is often associated with a particular accent. Sometimes a dialect gains
status and becomes the standard variety of a country.”

And in Trask(2000) we are provided with the following definition of dialect:

“ … any distinctive variety of a language spoken by some group of people. A particular style
of pronunciation is an accent; in the USA, an accent is considered to be one part of a dialect,
while in Britain accents are considered to be independent of dialects, which are characterized
by lexicon and grammar.”

As can be seen, the three definitions of dialect are basically similar and all of them employ the
term variety, which calls for explanation here. The way in which this term is defined in
Richards and Schmidt’s dictionary and Crystal’s dictionary indicates that it is more
comprehensive than the term dialect, as defined in the above definitions. Thus, in Richards
and Schmidt ‘variety’ (i.e. language variety) is defined as :

“ a term sometimes used instead of language, dialect, sociolect, pidgin, creole,etc., because it
is considered more neutral than such terms. It may also be used for different varieties of one
language, e.g. American English, Australian English, Indian English.”

55
Incidentally, ‘variety’ in this latter sense is used by A.H.Marckwardt and R.Quirk in their A
Common Language (1965), where they make a clear-cut distinction between ‘dialect’ and
(language)’variety’. Thus, for these linguists, British English and American English are
different varieties of English; there are distinctive dialects in the USA and distinctive dialects
in Britain.

In D.Crystal’s dictionary, under the headword ‘variety’ we read:

“ A term used in sociolinguistics and stylistics to refer to any system of linguistic expression
whose use is governed by situational variables. In some cases, the situational distinctiveness
of the language may be easily stated, as in many regional and occupational varieties (e.g.
London English, religious English); in other cases, as in studies of social class the varieties
are more difficult to define, involving the intersection of several variables (sex, age,
occupation, etc.) Several classifications of language varieties have been proposed, involving
such terms as DIALECT,REGISTER,MEDIUM,FIELD. For some sociolinguists, ‘variety’ is
given a more restricted definition, as one kind of situational, distinctive language – a
specialized type of language used within a dialect, e.g. for occupational purposes.”

From Crystal’s definition of ‘variety’ it follows that the term is also sometimes used in a
sense narrower than that of ‘dialect’.

Accent is another term that appears in the definitions of ‘dialect’ cited above.

Thus, in Trask’s definition , ‘accent’ is described as a particular style of pronunciation , and in

Crystal’s definition, it is stated that “… a dialect is associated with a particular accent”.
The dialectologist M.F.Wakelin (1977:1) says that accent “refers only to pronunciation, while
‘dialect’ refers to all the linguistic elements in one form of a language.”. Thus conceived of,
then, accent is a phonetic or phonological aspect of a dialect.

The same linguist, Wakelin(ibid.) writes that dialects are “ variant, but mutually intelligible,
forms (varieties, elsewhere-P.K.) of one language, whereas ‘language’ is assumed to imply a
form of speech not on the whole intelligible to other languages. “ In other words, it is usually
said that people speak different languages when they do not understand each other. For
example, the ‘vernaculas’ of Devon and Yorkshire are dialects, whereas those of France and
Spain are languages. As Wakelin also points out, “ It is, of course, recognized that there are
occasions when it is not clear whether a ‘dialect’ or ‘language’ is relevant: Old Norwegian
and Old English (the English language spoken in the period 700-1100 A.D.), probably being
mutually intelligible, must, under this definition of ‘dialect’ , be regarded as dialects, but
modern Norwegian is not intelligible to speakers of modern English, and thus in their later
stages these two dialects become two languages.” And, though Swedes, Norwegians and
Danes are generally able to understand each other, their separate histories, cultures, literatures
and political structures warrant Swedish, Norwegian and Danish being referred to as different
languages, not different dialects.

An interesting linguistic situation can be observed in China, where the so-called ‘dialects of
Chinese’ (e.g. Mandarin, Cantonese, etc.) are mutually intelligible in their spoken form. They
do, however, share the same written language, which is the main reason why one talks of
them as ‘dialects of Chinese’, not as separate languages (Crystal 2008).
As pointed out in Robins(1967: 34-5):

56
“ the differences between ‘languages’ and closely-related ‘ dialects’ are for the most part
political and cultural, rather than linguistic. From a strictly linguistic point of view, what are
customarily regarded as ‘languages’ (e.g. standard Latin, English,French, etc.) are merely
‘dialects’ which, by historical ‘accident’, have become politically or culturally important.”.
Studies of the history of the classical and modern languages show that dialects are not
‘imperfect and distorted ‘ versions of standard literary languages (which many traditional
linguists assumed to be the case), but that they developed more or less independently. “They
were no less systematic- they had their own regularities of grammatical structure,
pronunciation and vocabulary- and they were no less suitable as tools for communication in
the contexts in which they were used.”(Crystal, ibid.).

On the basis of the above considerations it may be concluded that the difference between
language and dialect is one of degree, rather than one of kind, “ a language being only one
stage of development further on from a dialect.” (Wakelin,ibid.).
Dialects are the products of linguistic change, and such change
(phonological,grammatical,lexical,semantic) is taking place all the time.

Dialectology , structural dialectology, perceptual dialectology

Dialectology (also called linguistic geography or dialect geography) is the systematic

study of all forms of dialect, but especially regional dialect . Traditional dialectology studies
began in the late 19thc. and have taken the form of detailed surveys using questionnaires and,
more recently, tape-recorded interviews. Regionally distinctive words (distinct in form, sense
and pronunciation) were the centre of attention , and collection of such words were plotted on
maps and compiled in a dialect atlas (also linguistic atlas), i.e., a book containing a large
number of dialect maps , each illustrating a particular linguistic variable within a certain
geographical area. If a number of distinctive items all emerged as belonging to a particular
area, then this would be the evidence for saying that a dialect existed.
It was often possible to show where one dialect ended and the next began by plotting the use
of such items , drawing lines around their limits of use (called isoglosses), and where a bundle
of such isoglosses fell together, postulating the existence of a dialect boundary.

Traditional dialectological methods of this kind have more recently been supplemented by
the methods of structural dialectology( the approach postulated by U.Weinreich in his ‘Is a
structural dialectology possible?’ Word 10,1954) , which tries to show the patterns of
relationship which link sets of forms from different dialects. The systems of structural
correspondence published by this approach are known as diasystems.

The branch of dialectology which is concerned with the way dialects , and individual dialect
features , are perceived by speakers within a speech community is called perceptual
dialectology. Real and imaginary linguistic differences , stereotypes of popular culture, local
strategies of identification , and other factors combine to generate a conception of individual
dialects , whose perceptual identities and boundaries may differ significantly from those
defined by objective dialect methods.

Regional/geographic/rural dialects and social/class dialects(sociolects), urban dialects

Urban dialect is a dialect which is typical of an urban area. Early dialectologists were
inclined to think that only rural areas preserved ‘ pure’ dialects, while urban speech was
merely ‘corruptions’ of standard speech and unworthy of study. Thanks to the investigations

57
of the sociologists, we now know that this is not so. First, urban speech is invariably
vernacular in form (i.e., the indigenous language or dialect of a speech community, e.g. the
vernacular of Liverpool, Berkshire, Jamaica, etc,), and cannot be regarded as derived from
standard varieties> Besides, there is a great deal of evidence that most linguistic changes
originate in urban communities and then spread slowly out over adjoining rural areas, so that
rural dialects, far from being ‘pure’ , are often no more than version of the urban speech of
several generations ago.

Social dialects/sociolects

These are linguistic varieties (or lects) defined on social (as opposed to regional) grounds,
e.g., correlating with a particular social class or occupational group.

Regional/rural/geographic dialects

Dialects which are characteristic of particular geographical areas (e.g. dialects within a
country, such as New England, Midland and Southern dialects in the USA, and the northern
and southern English dialects in Britain.

Dialect levelling

John Lyons ( Language and Linguistics, 1981) notices that language contact leads to diffusion
of phonological, grammatical and lexical variations through both geographical regions and
social classes of a given speech community. This process has been observed by other linguists
, too, and was called dialect levelling . Thus, Williams and Kerswill( cited in Foulkes
1999:13) define’ leveling’ as “ a process whereby differences between regional varieties are
reduced, features which make varieties distinctive disappear, and new features emerge and
are adopted by speakers over a wide geographical area. Kerswill(2001) says that this process
is composed of two stages running in parallel. The first stage affects the traditional rural
dialects of the country the speakers of which tend to “abandon these dialects in favour of a
type of English that is more like the urban speech of the local town or city. These more urban
ways of speaking have been labeled modern dialects or main stream dialects by P.
Trudgill(1998).[…]. The second stage affects these urbanized varieties of English
themselves” (Kerswill 2001, the World Wide Web : http://www.shunsley.eril.
net/armoore/lang/rp.htm ). They are subject to further leveling , but unlike traditional rural
dialects which change in their grammar, vocabulary and phonology, they typically undergo
phonological modifications only. What is common in the process of levelling is that speakers
move towards more prestigious variants of speech, thus adopt standard forms in preference to
the local ones. The process of this kind leads usually to the so-called dedialectalization (or
standardization (after Trudgill, cited in Foulkes 1999). Standardization does not have to
accompany dialect levelling. Certain recent findings suggest that it is perfectly normal for
dialects to converge without getting closer to the standard. It very often happens that that they
absorb features associated with no prestige whatsoever. In fact, what has come to play a
greater role in exercising influence on language variation are the non-standard varieties,
among which, according to Wells (cited in Foulkes 1999:11), London’s “working class
accent is today the most influential source of phonological innovation in England and perhaps
in the whole of the English-speaking world.” At this juncture, the conclusion might be drawn
that levelling might lead to homogenization of dialects (or accents) over the areas where
this process takes place. However, this should not happen, since these are just the most
locally-marked variants that tend to be avoided in preference for the more popular forms.

58
Various aspects of regional speech are still preserved because the driving force behind this
process is not the desire to disguise one’s geographical origins, but simply the urge to sound
more cosmopolitan. Therefore, while adopting certain non-local features speakers continue
using their regional forms signaling their loyalty to the local community by this means. An
important factor in the process of levelling is that the adopted features cannot be belong to
any other distinct well-defined variety, otherwise speakers would be considered disloyal to
their speech community.
The mechanism of levelling is strongly linked with social and geographical mobility of
speakers. Its socially related driving force is the social dynamics resulting from the divisions
of gender, age and class. That is to say, the social dynamics causes a shift in a social class
structure which, in recent years, has been leading British society towards ‘increasing
embourgeoisement ‘ (cf.Foulkes 1999:14). The geographical dynamics, in turn, is connected
with the movement of people from rural regions into city, , and, in the opposite direction, city-
dwellers into rural areas, both long-term (when people move out of city centres seeking
homes in the surrounding countryside) and short-term ( when they leave a city just for the
purpose of leisure). This kind of mobility results in an increased contact between speakers of
different dialects, a change in social networks (like dissolution of close-knit networks or
expansion of the range of individual personal network ties), and, eventually, in linguistic
change. There are voices which claim that interpersonal contact is not the only factor leading
to change. For instance, a significant role in the process of levelling is ascribed to the spoken
media, which expose listeners to both standard and non-standard varieties of language.
However, while it is true that such exposure can develop tolerance towards different forms of
speech and trigger imitation of certain features, it cannot enable speakers to use them
habitually and carry out everyday conversations in varieties heard on television or radio. In
order to acquire a language, personal interaction is essential. Milroy(2001:26), for example,
stress the significance of conversation, which is “ a channel in which language change is
normally passed.

Standard and nonstandard language/ variety/dialect

In Trask (2000) standard language/variety is defined as “ a highly codified and elaborated

variety of a language which is regarded by its speakers as the most appropriate (often the
only appropriate) variety suitable for educated discourse. Standard English , French and
Spanish largely developed out of the speech of politically preeminent regions. Standard
Italian and German were mainly the creations of prominent literary figures. Standard
Norwegian (of which there are two varieties) , Finnish, Basque and Turkish were mainly
constructed by conscious and deliberate language planning ( for this term see the glossary in
one of the earlier lectures) by governments or language academies. And Standard Arabic is
based on the language of the Koran.
A standard variety/language is generally used in the news media and literature, described in
dictionaries and grammars ,and taught in schools and taught to non-native speakers when they
learn the language as a foreign language.
The standard variety of American English is known as Standard American English and the
standard variety of British English is Standard British English.

Standard languages/dialects/varieties cut across regional differences, providing a unified

means of communication , and thus an institutionalized norm, i.e., a standard practice in
speech and writing.

59
 Linguistic varieties or dialects which do not conform to this norm are referred to as
nonstandard or substandard, though linguists prefer the term nonstandard as it is a more
neutral term than the term substandard.

Standard [British] English – a historical perspective

The first written standard of British English was set as far back in history as the tenth century.
It was then that the restoration of monasteries took place which, in turn, triggered the revival
of learning. The cradle of written standard was Winchester and there were two reasons for
that. First of all, it was a place where scholars were seriously involved in the study of
language.” Here they translate from Latin and try even to regulate the use of vocabulary”
writes (Gneuss 1972 in Fisiak 1993:58). The other important thing is that Winchester enjoyed
a great prestige within the English Church of that time, which helped to spread the written
standard over a rather vast area , from Canterbury to York. It was not for a long time, though,
that this standard was in use. The invasion of Normans in 1066 brought French to deal with
central administration and Latin to run business. The acceptance of the written form of these
languages as official ones resulted in a decline of the written standard Old English.
Winchester, too, lost its importance as a cultural centre. Because of the lack of a national
standard, local vernaculars gained on strength a great deal. They continued to be spoken and,
soon, there appeared their written records, as well.

It was four centuries later , when the new written standard English emerged and six centuries
later when the spoken standard was developed. In linguistic literature there are at least two
views on the source of Standard English. For example, L.Bloomfield (Language,1961) and
J.Lyons (Language and Linguistics, 1981) claim that modern Standard English has ‘grown
out of the provincial type of language that prevailed in the upper class of the urban centre that
became the capital of the unified nation” [i.e. London]. They are not the only linguists that
insist on the 14th c. London being the immediate source of Standard English
Thus, Baugh and Cable says explicitly that “ the history of Standard English is almost a
history of London” (Baugh and Cable 1978 in Fisiak 1993:83). According to Fisiak
(1993:84), however, “ what emerged in the 15th c. as a written standard was an artificial form
of English which cannot be localized in any single place and which was independent of any
spoken dialect.”
Fisiak also stresses that it was not London, but Westminster, which was not a part of London
at that time, where the standard dialect eventually came into being in the 15th c. Fisiak points
out that the written Standard English evolved out of two sources, viz., out of Wycliffite
manuscripts (based on the dialects of the Central Midland counties) and the writings of
London at the end of the 14th c., represented mainly by Chaucer, Gower and Hoccleve
(rendered in the language that varied considerably from the varieties used by Londoners ,
and that was generally based on the Central East Midland dialects). Standard English of the
15th c. is known as Chancery Standard , because it was commonly used by the administrative
body of the capital , that is Chancery in Westminster. It was there where most of the
government offices were situated and where clerks were taught quite a uniform language
which they used for preparing official documents and for manuscript copying. From this
place the new Standard was being spread. It reached different parts of the country not only
thanks to the documents that were sent from there, though. A very significant role in its
popularization played also both the educated clerks who could easily find a job in any place
of England and continue to advocate the Chancery Standard, and individual scribes who used
the Chancery practices in their private work. Later on, since 1476 onwards, J.Caxton, who
set up the press , helped circulation of this standard. It was easily accepted in different parts of

60
England for it did not reflect any spoken dialect. Moreover, it was based on the language of
the Central Midlands, which was the variety understood by a great variety of English
speakers. It had its own orthographic system, introduced by the Chancery scribes of the
Standard.
IT must be pointed out that although it was not London itself that gave rise to the Standard,
one cannot underestimate its role. After all, Westminster was situated within the area of the
capital city, which means that it exerted its influence on London, and vice versa. This mutual
influence must have had its reflection in Standard English. Besides, at that time London
was , and, of course, still is, the heart of England. Since it is commonly known that the
development of the capital (of any country, for that matter) as a social, political and
commercial centre is also connected with the flourishing of major linguistic trends, it
becomes obvious that the prestige of London is also responsible for Later expansion of the
Standard over a larger area. Thus, the role of the capital can be seen as vital in the
development of Standard English.
To the same degree as the standard form of English used to radiate from London, the
linguistic features still spread from this cultural centre of England. With the greatest force
they ‘invaded’ the immediate surroundings of the London area. Bloomfield (1961:485) writes
that “the neighbouring dialects may be so permeated with standard forms as to lose all their
individuality. We are told that within thirty miles off London there is no speech-form that
could be described as local dialect.” However, Standard English reaches much further than
that. All over the country the so-called dialect speakers learn to speak according to its rules.
Because the form of speech they eventually acquire differs slightly from the Standard, it
tends to be described as substandard, or more favourably as a coloured standard
( Bloomfield 1961). It has to be said, though, that the linguistic influence was not one-sided.
There has always been a tendency to absorb certain elements characteristic of adjacent
dialects. Standard English has been in constant change. And still undergoes the process of
modification.

Standard accent of English (Received Pronunciation=RP)

It is the type of British Standard English pronunciation which has been traditionally
considered the prestige variety and which shows little or no regional variation. It historically
derives from the prestige speech of the Court and the public schools. It has often been

popularly referred to as ‘BBC English’, because it was until recently the standard
pronunciation used by most British Broadcasting newsreaders. The BBC originally adopted
RP for its announcers because it was the form of pronunciation most likely to be nationally
understood, and to attract least regional criticism – hence the association of RP with the
phrase ‘ BBC English’. Like all other varieties of language RP has been subject to change
over time. These days , the BBC, and indeed educated speech at large, displays considerable
regional variation, and many modified forms of RP exist (modified RP). A far greater
number of people (especially the younger generations) today feel free to use a form of
regional dialect without seeming uneducated, and oddly enough, the BBC is partly
responsible here too. It has not only spread a familiarity with RP, it has also made people
more familiar with the different regional dialects that are heard throughout the British Isles,
which made more and more speakers more tolerant of them.

RP differs from Standard American English pronunciation in various ways. For example,
speakers of RP do not have an r sound before a consonant, while most Americans do, as in
farm /fa:m/ - /fa:rm/.

61
Cockney – a regional or social dialect, or a mixture of both?
Nay in the very metropolis two different models of pronunciation prevail ,
by which the inhabitants of one part of the town are distinguished from
those of the other. One is current in the city, and is called the cockney;
the other at the court-end, and is called the polite pronunciation.

Thomas Sheridan ( cited in Wakelin 1977: 41)

There is one more variety associated with London, namely Cockney. The degree of its
connection with the capital city of England is so great that sometimes it is called London
English. However, this name is not a very fortunate one, since it emphasizes Cockney’s
regional character and omits its social connotations. We mustn’t forget that, as Gimson
(quoted in Cruttenden 1994:85) writes , “ Cockney is as much a class dialect as a regional
one”. Gimson’s words may then serve as an answer to the question we posited above.
Cockney has roots in London and is usually heard among the members of London’s working
classes, who willingly use its characteristic nonstandard grammatical features, its distinctive
pronunciation and specific vocabulary, including so-called rhyming slang , i.e., unusual word
formations in which a word is replaced by a pair of words, the second of which rhymes with
the one replaced, e.g., apples and pears stands for ‘stairs’, read and write replaces ‘fight’,
etc. The speakers of Cockney are proud of the way they speak and like the idea of
emphasizing their group membership. Therefore, the covert prestige of Cockney seems to be
obvious. Although its covert prestige is not too high, it is sufficient enough for this dialect to
influence speech of the middle class of the area. Features of Cockney pronunciation is what
this class uses most readily, but it sometimes adopts specific components of its lexical set,
too. Still, Cockney continues to be popular among the members of the lower strata of society.
For that reason, and the fact that it clearly differs from the ‘court-end’ pronunciation,
Sheridan once said that Cockney “ bears marks, either of a provincial, rustic, pedantic or
mechanic education; and therefore has some degree of disgrace annexed to it.” (Wakelin
1977:41). Nonetheless, it does not mean that Cockney can be considered ‘worse’ than RP.
On the contrary, it is the working –class accent that has always had a major influence on the
development of the most prestigious one, hence the pronunciations of Cockney have often
shown the way in which RP would develop itself. Gimson (in Cruttenden 1994: 86) notes that
RP is still being affected by Cockney and that the “ type of regional RP which is heavily
influenced by Cockney is often referred to as Estuary English.
Like other urban and rural dialects and accents
, Cockney has the same ancestor as RP and Standard English, viz., Old English. Just like
them, it has undergone linguistic change and lost some of the traditional dialect
characteristics.
Cockney introduces lots of colour and creativity to the speech of London. Its speakers are
famous for ‘quickfire’ wit in a chat and so their conversational styles are regarded by
Londoners as amusing and interesting.

Cockney rhyming slang It originated in the East End of London in the middle of the 19 th c. It
a was a secret code used by traders, entertainers, and thieves to outwit a policeman or
stranger.
At the beginning it was popular among East-End pub-goers, but in recent times it is used by
many Londoners and speakers from other areas of the country.
Above we gave two examples of rhyming slang. Here is another one: butcher’s hook for
‘look’. When the rhyming slang is a phrase (as is the case with our examples), the rhyming
part is often dropped; thus,e.g., when a speaker wants to describe feet, he will say plates,

62
rather than plates of meat. Sometimes rhyming slang is shortened, which makes it often
difficult to follow. Following are a few examples of shortened rhyming slang:

daisies /daisies roots/ boots

minces / mince pies/ eyes
turtles / turtle doves/ gloves
rabbit / rabbit and pork/ talk
etc. etc.

The best examples of Cockney rhyming slang are those which combine a rhyme with irony,
wit or provocative social description: e.g. artful dodger rhymes with ‘lodger’, sorry an’ sad
describing ‘dad’.

Many rhymes are connected with names of people, e.g.,Lord Lovel and ‘shovel’, Jimmy
Skinner and ‘dinner’, Charlie Frisky and ‘whiskey’ , etc.

Some linguistic features characteristic of Cockney:

1/ initial /h/ is dropped in word-initial position , as in: ‘eart for heart, ‘ope for hope , etc.
2/ in words like taking, walking the final –g is dropped, so the Cockney forms
are,respectively, takin’, walkin’.
3/ The fricative /r/ used in words like, say, frosty and drink are replaced by/w/: so the
corresponding Cockney forms are fwawsty and dwink.
4/ The voiced and voiceless th are replaced by /v/ and /f/, as in fanks for thanks, fought for
thought , and wiv for with, brovver for brother, ovver for other.
5/ Cockney speakers use we was instead of we were, ain’t for isn’t.
6/ To express an action in progress, Cockney speakers use the construction go and Verb+-ing
or do and Verb+-ing, as in Don’t go moving my things!, I get messin’ about in my garden.
7/ Another feature is the tendency to use the accusative form in subject position , as in Me
and mi brovver was born’.
8/ The use of the indefinite article a before nouns beginning with a vowel, as in a orange, a
apple, etc.
9/ Multiple negation, as in Your gov’ner can’t do no good here, Sammy. (from Ch.Dickens’
The Pickwick Papers), I don’ t owe him nothing, I ain’t got no mother,etc.
10/ Use of present tense continuous instead of simple present tense: I’m wanting to tell you.
11/ For more examples of Cockney linguistic features see, among others, P.Wright, Cockney
Dialect and Slang. London: B.T.Batsford LTD, 1981 , Franklyn, J. The Cockney. A Survey of
London Life and Language. London: A.Deutsch, 1953, and elsewhere.

The Cockney dialect makes its appearance in the works of popular English writers.; e.g. in
Dickens’ The Pickwick Papers, Our Mutual Friend, and Bleak House. The following
passage taken from Dickens’ Bleak House will give you a flavour of the speech of the
Cockney character Joe, created by Dickens in the novel:

“They’re wot’s left , Mr Snagsby,’ says Joe, ‘ out of a sov’rin’ as was give me by a lady in a wale as sed she
was a servant and as come to my cossin’ one night and asked to be showed this ’ere ouse and the ouse wot
him as you giv the writin to died at, and the berrin-ground wot he’ berried in. She ses to me, she ses “are you
the boy at the Inkwhich ?” she ses. I ses, “yes”, I ses. She ses to me, she ses, “can you show me all the places?”
And she ses to me “do it”, and I dun it, and she give me a sov’rin and hooked it. “

63
The dramatist, G.B.Shaw, was an excellent observer of the Cockney dialect, who put himself
in line with the pejorative judgements on Cockney expressed by many linguists. In the play
Pygmalion , Henry Higgins, professor of phonetics, tries to teach a Cockney girl to speak with
an upper-class accent and trains her in etiquette. His task is very difficult, which Eliza
confirms by saying to her mother:

“ There’s menners f’yer! Te-oo banches ‘o voylets trod into the mad… Ow eez yee-ooa san, is ‘e? Well, fewd
dan y’de-ooty bawmz a mother should, eed now bettern to spawl a pore gel’s flahzn than ran awy athaht pyin.
Will ye-oo py me f’them?” (adopted from Wright 1881: 19).

Glossary

Estuary English A variety of British English supposedly originating in the counties

adjacent to the estuary of the River Thames, and thus displaying the influence of London
regional speech, especially in pronunciation; also called simply Estuary. The name is
somewhat misleading, in that the defining linguistic features (such as the increased use of

glottal stops and the vocalization of final /-l/ extend well beyond the river throughout much
of south-east England, among lower middle-class speakers , and have been around much
longer than the arrival of a new name suggests. It is to be distinguished from working-class
Cockney, lacking some of the salient characteristics of that accent , such ,e.g., as the fronting
of th to /f/.

Glottal stop A consonant that is produced by the momentary closing of the

glottis,i.e., the space between the vocal cords, thus trapping the airstream from the lungs
behind it, followed by a sudden release of the air as the glottis is opened. In some varieties of
British English , a glottal stop is used instead of a /t/ in words like bottle or matter.

64
Lecture 11
Theory of language; grammar, some types of grammar, constraints on grammars

The theory of language does at least the following:

a. It delimits the object of its interest, i.e. language (that is to say, language lies at the
centre of interest of the theoretical linguist);
b. It specifies the goals and methods for the description of the various aspect of
language;
c. It works out the theoretical equipment for description of linguistic facts (e.g.
phoneme, morpheme, word, clause, rule, etc.).

In linguistic literature, the term ‘grammar’ is used in at least two senses; in one sense, the
term ’grammar’ is used as a synonym of what transformational linguists call ‘native speaker’s
linguistic competence’, i.e., the native speaker’s knowledge of his language. On this
understanding of the term , grammar is a kind of a psychological entity, sometimes referred to
as ‘internal grammar’. In the other sense, the term’ grammar’ is equivalent to a description
of linguistic competence (of our linguistic knowledge),or what Saussure calls ‘langue’. It is a
description of the structure of a language and the way in which linguistic units such as words
and phrases are combined to produce sentences in the language. This kind of description may
or may not include the description of the sounds of a language (i.e. the phonological
component of the language). Grammarians adhering to this restricted sense of ‘grammar’
exclude from their investigation not only the phonological component but also the semantic
component of the language they set out to describe.

There are various types of grammar that can be distinguished. Among them:

Synchronic (descriptive) and diachronic (historical )grammars

Particular languages can be described synchronically, i.e. by looking at them at a particular

point in time, or diachronically (historically) , by taking into account their development from
one point in time to a later one. For example, S.Horobin & J.Smith’s An Introduction to
Middle English is an example of a synchronic description of the English language as it was
spoken in the Middle Ages (in the period 1100-1500), and B.Strang’s A History of English
and E.Closs-Traugott’s A History of English Syntax are diachronic descriptions of English
(Closs-Traugott’s book being restricted to the evolution of English syntax ).

The 19th century Neogrammarian School of Linguistics was exclusively concerned with the
diachronic aspect of language study. Thus, the 19th c. Neogrammarian linguist Herman Paul
put it quite bluntly, “ what is not historical in language study is not scientific”.

65
It was not until Ferdinand de Saussure, the Swiss scholar (1857-1913) that a clear distinction
had been made between these two points of view on linguistic study. Saussure (the author of
Cours de linguistique générale, 1916) is responsible for introducing the terms ‘diachronic
linguistics’ for historically oriented investigations and ‘ synchronic linguistics’ for
descriptive studies. De Saussure said: “ Synchronic is everything that has to do with the
static aspect, while diachronic is everything which is related to evolution.”

Thus synchrony and diachrony correspond to a language state and language evolution,
respectively. The strict separation of the two approaches to language study was followed by
the American Structuralist School of Linguistics, while the Prague School has favoured a
synthesis of the two.

Theoretical grammar This type of grammar goes beyond the study of individual
languages , using linguistic data as a means of developing theoretical insights into the nature
of language as such, and into the categories and processes needed for successful linguistic
analysis. Such insights include the notion of ‘grammatical categories’, ‘grammatical
meaning’, the study of ‘grammatical relations’, etc.

Scientific and pedagogical (teaching) grammars

All types of grammars use the concepts and terms worked out by the theory of language.
They make use of terms such , e.g., as ‘phoneme’, ‘morpheme’, ‘word’, ‘clause’,’
grammatical rule’, etc. in order to describe concrete linguistic facts.

Scientific grammars are grammars that are written for scientists and specialists and they are
theoretically homogeneous, i.e., they are grammars that are based on a single theory of
language. For example, D.T.Langendoen’s Essentials of English Grammar is based on a
The generative-transformational theory of language.

Pedagogical grammars are grammars whose primary objective is of purely didactic nature,
and they are often theoretically eclectic, i.e. they often make use of the findings of two or
more language theories.

A grammar, as a linguistic description, is a formal representation, a model of the native

speaker’s competence (his linguistic knowledge), and as such, it must explicitly describe
(characterize) native linguistic competence. It must, among other things, explicate the ability
of the speaker to produce and comprehend an infinite number of sentences. Our linguistic
competence then can be viewed as an infinite rule-governed creativity. Fluent speakers of a
language can produce and understand not only actual sentences , but also sentences they have
never come across.

One of the most outstanding American transformational linguist, Noam Chomsky, in his
Aspects of the Theory of Syntax , writes, among others, thus:

“ The normal use of language is innovative in the sense that much of what we say in the
course of normal language use is entirely new, not a repetition of anything we have heard
before, and not even similar in pattern to sentences that we have heard in the past.”

66
What’s more, our linguistic competence consists not only in our intuitions about sentence
well-formedness ( A sentence is well-formed if it can be generated by the rules of a grammar;
it is ill-formed if it cannot be.) and sentence structure (we kind of intuitively ‘know’ which
syntactic structures are correct and which are not, are ‘ill-formed’). This kind of knowledge
may be looked upon as our syntactic competence.

In addition to ‘syntactic competence’, we can also speak of our morphological and semantic
competences.

Morphological competence relates to our knowledge about the structure of words and our
ability to create entirely new words on the basis of the existing stock of linguistic elements.

Semantic competence has to do with our intuitions about sentence interpretation and
coreferential relations. In other words, we have intuitions about semantic well-formedness
(i.e., about which linguistic expressions are semantically coherent and which are not). Take,
for example the following complex sentences:

1. I thought that Mary was ill, but it turned out that she wasn’t.
2. * I realized that Mary was ill, but it turned out that she wasn’t.

In (2), the asterisk indicates anomaly, i.e. semantic ill-formedness.

(1) Is both syntactically and semantically perfectly normal, while (2) is semantically
anomalous (ill-formed). As fluent speakers of English , we intuitively know that there
is something wrong with (2), and the grammarian must somehow account for the
semantic anomaly implied in the sentence.
Notice that what is implied by the structure it turned out that she wasn’t is semantically
incompatible with what is being implied by the structure I realized that Mary was ill. While I
realized that Mary was ill implies ‘ Mary was ill’, the structure but it turned out that she
wasn’t implies ‘ she was not ill’.

Verbs like realize, regret, criticize, know, etc. are so-called factive verbs , which are verbs
that imply the truth of what is conveyed by their complement clauses. And verbs like think,
believe, hope, etc. are non-factive , they do not imply the truth of their complement clauses.

And let us now consider the following:

3. I wonder who the men expected to see them .

4. The man expected to see them.

In these sentences there are coreferential relations holding between some of the sentence
constituents. In (3), the men and them are coreferential, i.e. here the pronoun them refers
back to the phrase the men , and they both refer to the same individual in the extralinguistic
world. However, in (4), the men and them are non-coreferential , as each of them refers to
distinct individuals.

67
In (3) them functions as anaphora, and the function of them in (4) is deictic, here them refers
the speaker not to some earlier unit in the verbal context, but to real world entities.

The task of the grammarian trying to construct a grammar which models/ formally represents
this kind of infinite linguistic competence is, according to the adherents of the Generative
Transformational Theory , to devise a finite set of rules capable of specifying how to
construct and interpret an infinite set of sentences, and , generally, correct linguistic
expressions.

A generative-transformational grammar of a language , thus, aims to explicitly

describe/specify what the speaker ACTUALLY KNOWS, NOT WHAT HE/SHE MAY
REPORT ABOUT HIS/HER LANGUAGE. Linguistic competence, then, is not the fluent
speaker’s knowledge about his/her language, but his/her knowledge OF his/her language.

A typical generative-transformational grammar of a language will consist of the following

components:

a. Syntactic component (concerned with the speaker’s syntactic competence);

b. Morphological component ( concerned with the speaker’s morphological
competence);
c. Semantic component ( concerned with the speaker’s semantic competence ).

A generative-transformational grammar of a language will also contain the so-called

phonological component which aims to describe the patterns of distinctive sounds and their
functions in the system of the language. Traditional grammars were not concerned with the
phonological subsystem of a particular language, focusing their attention on the problems of
morphology and syntax.

Constraints on grammars

Grammars, as scientific descriptions of particular languages, must be subject to a number of

constraints, with the constraints being formulated by the general theory of language (by
language theoreticians).

Structure- Dependence Principle is one such constraint. This implies that all grammatical
rules (rule in linguistics refers to a formal statement of correspondence between linguistic
elements or structures.) operate on structures, not on individual words. By ‘structure’ is
meant a sequence of linguistic elements that are in a certain relationship to one another. For
example, one of the structures of a noun phrase in English may be ‘article+adjective+noun’,
as in an expensive car.

In order to see how this principle operates, let us consider the following examples:

1. Bill can help you.

2. Can Bill help you?

A hypothetical ‘structure independent rule’, let us call it Second-Word Preposing would fail
to account for inversion in direct questions in English. What the rule states is that what we
must do to account for direct questions in a language like English is that we move the second
word in a corresponding declarative sentence to the front of the sentence-initial word. It is

68
easy to see that although this hypothetical rule will work for structures like our two examples
above, it won’t work for a multitude of other structures, like, for instance, the following:

3. That man will help you.

When, in accordance with the Second-Word Preposing rule, we move the second word in
(3)to the front of the sentence, what we get as the result this operation is an anomalous
construction (4):

4. * man that will help you .

As can be observed, this hypothetical rule is based on a linear order of sentence elements
(words), not on syntactic structures. In (3), that man is a noun phrase(NP), which is a
structure of a specific type.

A rule which will adequately handle the facts implied in our examples, and potential similar
ones, must be a structure-dependent rule. The English grammar does incorporate such a rule,
called Subject NP –Auxiliary Inversion, which states :“ In order to form a question invert
the subject NP with the following auxiliary verb”. When applied to example (3) above, the
rule will generate the perfectly correct (5):

5. Will that man help you?

The SUBJECT NP- AUXILIARY INVERSION rule is structure-dependent in the sense that
you cannot apply it to a given sentence unless you know what the syntactic structure of the
sentence , i.e., what the grammatical categories and phrases the words in the sentence belong
to. Thus, you must know whether the sentence contains an NP, whether this NP functions as
the subject of the sentence, whether the sentence has an Auxiliary, and whether the NP is
immediately followed by the Auxiliary element.

Another constraint I would like to draw your attention to is called A-Over-A Principle,
which states that no constituent of, say, category A can be moved out (transferred from) a
larger containing constituent of category A (i.e., category of the same type).

Let us take for our illustration of how this principle actually operates the following
prepositional phrase (PP): out of this tunnel. The internal structure of this phrase can be
presented as follows:

PP [out PP[of this tunnel] PP]PP

The PP (complex PP) out of this tunnel is the ‘larger containing’ phrase , it contains the
‘smaller contained ‘ phrase of this tunnel, which is also a PP (i.e. a structure of the same type).
Notice that (8) is ungrammatical, since it is the effect of the violation of the A-Over- A
Principle:

6. The players emerged out of this tunnel an hour ago.

69
In English grammar, there is a syntactic rule, called topicalization, which moves a
constituent to sentence-initial position. But consider what happens when only a part of the
constituent is undergoes this movement, as in the ungrammatical (8):

7. Out of this tunnel the players emerged an hour ago.

In (7), the whole PP (out of this tunnel) has been moved, which will account for its correct
structure. And in the ungrammatical (8), only a part of the PP has been transferred to
sentence-initial position:

8. * of this tunnel the players emerged out an hour ago.

Coordinate-Structure Constraint is still another constraint to be discussed here.

First, let me explain what is meant by coordination in grammatical analysis. Grammarians
use the term’ coordination to refer to the process of linking linguistic units which are of
equivalent syntactic status, e.g., words, phrases, clauses, etc. Thus, coordinate clauses can be
illustrated in the sentence John walked and Mary ran (here two clauses are being coordinated
or conjoined). In the example and functions as a coordinating conjunction.

Coordinate structures are islands, in the sense, that an island is a construction out which no
subpart can be moved , though the whole island can be moved as one unit. For example, in
(9) , the poet and the car dealer, is a coordinate structure :

9. I really like the poet and the car dealer.

Only the whole structure (NP) , the poet and the car dealer, is subject to topicalization,
hence (10) is perfectly normal, while (11) is not:

10. The poet and the car dealer I like - .

11. * the car dealer I like the poet and - .

In (11), only a part (the car dealer ) of the coordinate phrase (the island) has been topicalized.

Another constraint, Sentential Subject Constraint, states that no constituent can be moved
out of a sentential subject , i.e. out of a clause which functions as the subject of another
clause. It must be said that, like coordinate constructions, sentential subjects are also
islands.
Let us take the following for our illustrative example:

12. For me to give up what would be a pity.

The sequence for me to give up what is a clause functioning as the subject in (12). Notice now
that if we move only what out of the clause for me to give up what , the effect of this
unlawful operation is the ungrammatical (13):

13. * what would for me to give up be a pity?

70
The STRUCTURE-DEPENDENCE PRINCIPLE,the A-OVER-A PRINCIPLE, and the
SENTENTIAL SUBJECT CONSTRAINT are treated by linguists as so-called absolute
linguistic universals, since they operate in all languages without exception (they are also
called nonstatistical universals).

The other types of linguistic universals distinguished by theoretical linguists are called
relative universals (also statistical universals). An example of this kind of universals is the
so-called Consistent serialization Principle, according to which all languages tend to place
modifying elements either consistently before or consistently after the modified elements
(called heads or denotata ).

For example, languages like English and Japanese, generally place their modifiers before
heads , they are, from this point of view, pre-modifying languages. French, with its general
tendency to place modifiers after heads, is a good example of a post-modifying language (
le chair noir’ black cat’, linguistique générale ‘general linguistics’, etc.).

The CONSISTENT SERIALIZATION PRINCIPLE represents a general word-order

tendency in the languages(i.e. in the grammars of the languages) of the world, there are many
exceptions to it. There are languages which represent mixed types , they put some modifiers
before, others after their heads . To this group of languages belong, among others, Greek and
Basque.

What has been referred to above as absolute and relative universals some other linguists
choose (e.g. Noam Chomsky) to call formal universals.

Glossary

grammatical category A class or a group of items which fulfil the same or similar
functions in a particular language. For example, CASE,TENSE,ASPECT, PERSON are
grammatical categories. Some linguists also refer to related groups of words such as nouns,
verbs, and adjectives as grammatical categories, but these groups of words are referred to in
traditional grammar, as parts of speech.

constituent A linguistic unit, usually in sentence analysis, which is part of a

larger construction . For example, the, tall, and man are constituents of the NP the tall man,
and The tall man , has read, and the book are constituents of the sentence The tall man has
read the book.

head The central part of a phrase. Other elements in the phrase are in
some grammatical or semantic relationship to the head. For example, the NP the fat lady in
the floral hat the noun lady is the head of the phrase, it is the central constituent of the
phrase.

modifier A word or group of words which gives further information

about, i.e., modifies another word (the head). Modifiers before the head are called
premodifiers e.g., expensive is a premodifier in an expensive camera; and modifiers after the

71
head are called postmodifiers, e.g. in the man with a white beard, with a white beard is a
postmodifier.

topicalization, topic Topic is what the discourse is about. Topics represent

given/old information. The topic of a sentence is the entity about which something is said.
Topicalization takes place when a constituent is moved to the front of a sentence , so that it
functions as topic, as, e.g., in The answer I’ ll give you in a minute. Topic is sometimes
referred to as ‘psychological subject’.

Lecture12,13
Meaning and its types, semantic
relations(synonymy,polysemy,homonymy,antonymy,hyponymy partonymy/meronymy);
semantic field(lexical field),componential analysis

Every word and every sentence in a language is associated with at least one meaning. The
branch of linguistics that is devoted to the study of conventional meanings of these units is
called semantics. Semantics which deals with word meanings is often referred to as lexical
semantics, while the study of sentence meaning is referred to as
compositional/propositional semantics.

Following are some of the kinds of meaning distinguished in linguistic literature:

denotational/denotative (also called conceptual, cognitive,referential, or ideational)

connotational/connotative meaning
affective meaning
stylistic meaning
sentence meaning
text meaning
thematic meaning
grammatical meaning
lexical meaning

Denotational m. It is the basic meaning of a word, i.e. what a word refers to. For example,
the verbs hide and conceal have the same denotational meaning, as they refer to the same
type of action in the extralinguistic world. Denotational meaning may be regarded as
the’central’ meaning or ‘core’ meaning of a lexical item.
The denotational meaning of a word can be represented by a set of its criterial/ defining
features. Thus, the criteria semantic features that constitute the denotational meaning of the
noun woman are: ADULT, FEMALE , HUMAN.

Connotational meaning The additional features that are associated with a given entity
/referent denoted by a given word. Such features are non-criterial and non-defining. For
example, features such as FRAIL, TALKATIVE, and OVEREMOTIONAL, and maybe a
few others, are connotational features traditionally associated with the referent of the English
word woman.
Connotation, then, has to do with the communicative value of a word , i.e., with meanings
which are socially acquired. Connotational meanings show people’s emotions and attitudes
towards what the word (or phrase) refers to. The connotational features associated with child

72
could be AFFECTIONATE, AMUSING, LOVABLE, SWEET, MISCHIEVOUS, NOISY,
and possibly a few others.
There are good connotations and bad connotations. For example, MOTHERLY would be
an example of a good connotation associated with woman , while some of the other features
mentioned earlier constitute bad connotations associated with woman.
Word connotations are culturally and socially determined, and, ulike denotations, they tend to
change from period to period. While denotational meaning is stable, connotational meaning
is unstable and peripheral. Thus, the word woman has a conceptual sense which is unlikely to
change over time. But if someone were to refer to an acquaintance as a real woman , we
would know that the phrase carried more meaning than just its basic conceptual meaning.

What is exactly connoted is then debatable and dependent on a number of cultural variables,
but in many respects connotation offers us a greater insight into speakers’ social attitudes
than does conceptual meaning.
Words are not simply bearers of neutral conceptual/denotational content, as we can see when
they are used to discuss sensitive areas of people’s social and cultural life. Notice that, e.g.,
the problem of race is rendered more difficult because words such as white, black, and
coloured are connotationally weighted(marked). At the same time connotation provides a
rich source for people who exploit the imaginative possibilities of language, for instance,
individuals like poets, or those writing advertising copy.
In spoken language, connotation is often signaled by intonation. A particular tone of voice, or
pattern of stress, are employed to indicate a whole penumbra of extra shades of meaning.

Affective meaning Language can also reflect the personal feelings of its speakers,
including their attitude towards something they are talking about. This kind of meaning is
called affective meaning (also emotive meaning) and it is related to the expressive function
of language. Thus, e.g., a linguistic expression can signal the speaker’s irritation, anger, joy,
etc., or a violently abusive attitude towards his addressee. Very good examples of
emotionally loaded linguistic expressions are interjections, i.e., such items as God damn it!,
Blast it! Get stuffed!, etc. These are very often slang words. Consider also: Oh!, Aha! Tut
tut!, expressing impatience, contempt, depending on the context of their use. The only
function of interjections is to express emotions and feelings.
Positive emotions and feelings are typically conveyed by diminutives and negative ones by
augmentatives. Examples of such expressions are, respectively: English doggie,
girlie,auntie,etc., and Polish babsko, dziadyga, nochal, etc.

The means conveying affective meaning , apart from vocal expressions, also include
intonation and tone of voice.
Affective meaning is rather a parasitic category in the sense that to express our feelings and
emotions we must rely on the mediation of other types of meaning:
denotational,connotational, and stylistic.
Emotional expression through style comes about when we adopt an impolite tone to express
displeasure, or when we adopt a casual tone to express friendliness.

In a meaning system, that part of the meaning which is covered by affective meaning is
sometimes referred to as connotative meaning.

Stylistic meaning It is the type of meaning which a linguistic expression conveys

about the social circumstances of its use. Thus, some expressions are recognized as being’

73
dialectal’, i.e., as telling us something about the geographical or social origin of the speaker.
Other features inform us about the social relationship between the speaker and the addressee.

There is also a scale of status of usage, ranging from ‘formal/literary’ English to

‘colloquial/familiar’ English, and eventually slang.
Examples: little( general,stylistically neutral), tiny (colloquial, informal), wee(coll. And
dialectal, used especially in the North of England and Scotland), horse (general, neutral), nag
(slang or informal), throw (general), cast (literary), chuck ( informal or even slang, for some
lexicographers), etc., etc.

The style dimension of status is especially important in distinguishing synonymous

expressions . In the following examples the difference in social status finds reflection in the
syntax and vocabulary of the sentences:

1. They chucked a stone at the cops, then did a bunk with the loot (do a bunk ‘da c
dyla/drapaka/zwiać , loot ‘ łup, szmal’, cop ‘glina,gliniarz’ )

And its stylistically formal version,

2. After casting a stone at the police, they absconded with the money.

(1) and (2) are cognitively, but not stylistically equivalent , synonymous sentences.

The branch of linguistics that studies stylistic variation is called stylistics. This discipline is
concerned with the linguistic choices that are available to the speaker/writer and the reasons
why particular forms and expressions are used rather than others.

Sentence meaning It is a sentence’s propositional content (put succinctly, proposition

is what a sentence is about the real or fictitious world. Thus a proposition is a logical entity,
while sentence is a linguistic/structural entity).
A sentence may convey one or more propositions in different contexts. And vice versa,
different sentences may convey one and the same proposition. For example, the sentence (3):

3. I’ll meet him tomorrow.

may be produced to convey different propositions depending on who the speaker is, who is
the referent of him , and the time of utterance

Text meaning A sequence of two sentences (say, S1 and S2) may make up a text. If S1
and S2 constitute a text, then we say that the two sentences form a logically coherent text.
For example,

4. I saw a man on the boat (=S1) and 5. A man was very tall (=S2)

Here S1 followed by S2 do not make up a text, since a man in (4=S1) and a man in (5=S2)
are not coreferential; they do not refer to one and the same individual. But consider now (6)
and (7):

74
 6. I saw a man on the boat. 7. The man was very tall.

Here a man and the man are coreferential, that’s why (6) and (7) constitute a (coherent) text.
The text derives its unity from the interaction between its linguistic structure and the situation
in which it is used and to which it refers. The property of being a text is referred to as
textuality/texture.
The textuality is provided by certain linguistic features present in the text and the features
inherent in the extralinguistic context , i.e., the socio-cultural context, world knowledge,etc.

The linguistic features guaranteeing the texture are markers of textual cohesion, they
underlie the coherence of the text.
Consider in this connection:

8. Apples are very healthy. 9. That’s why he eats apples regularly.

The textual cohesion between (8) and (9) here is marked by the repetition of the word
apples and the expression that’s why, specifying the reason.
An now, if we compare (10) and (11), we will see that there is no explicit linguistic link
between them:

10.I won’t go to Paris tomorrow. 11. French railways are on strike.

Despite this, the text made up by (10) and (11) is easily comprehensible, and thus textually
coherent, and this is owing to our knowledge of the world (and perhaps the tacit assumption
that the speaker intends to communicate to us something that logically makes sense, which
would be in agreement with what is called Grice’s Cooperative Principle.

Thematic meaning It is one part of the meaning of a sentence that is not part of its
propositional content. The following sentences convey the same propositional content, but
differ as to their thematic meaning:

12. I haven’t read this book.

13. This book I haven’t read.
14. It is this book that I haven’t read.
15. This book hasn’t been read by me.

For example in (12) I (=the subject) is the theme (information that is not new to the
addressee), and in (13) , it is this book which is the theme(old information).

Sentence elements can be divided into those that convey old information (theme) and those
that convey new information (rheme).

Consider now (16) and (17):

16. Janek zobaczył białego psa. 17. Białego psa zobaczył Janek.

In (16), Janek (=Theme), białego psa (=rheme). In (17), białego psa (=theme), Janek
(=rheme).

75
However, cognitively (16) and (17) are synonymous sentences.

Lexical and grammatical meanings

Lexical meaning corresponds to cognitive/referential meaning. By contrast, the distinction

between the subject and the object of a sentence, TENSE, ASPECT, NUMBER, VOICE,
DEFINITENESS, INDEFINITENESS, etc. are grammatical/structural meanings.

The total meaning of a sentence consists of lexical meanings of separate words plus various
grammatical/structural meanings. The linguistic devices that express grammatical meanings
constitute the grammar of a language. The grammatical meanings DEFINITE and
INDEFINITE are signaled in English by means of appropriate articles and pronouns (more
specifically, demonstrative pronouns), which are English grammatical, not lexical words.

SEMANTIC RELATIONS

Synonymy

Synonymy is a type of sense/meaning relation holding between linguistic expressions:

linguistic expressions which have the same meanings are synonyms. For two items to be
synonyms, it does not mean that they should be identical in meaning, i.e., interchangeable in
all contexts and with identical connotations.
John Lyons in his Introduction to Theoretical Linguistics speaks of two types of synonymy:
total synonymy and complete synonymy.
Thus, for the linguist, expressions x and y are totally synonymous, if they meet the following
criteria:

a. They are interchangeable in all contexts.

b. They have identical both cognitive and emotive meanings.

And x and y are completely synonymous, if they meet only the (b)condition. Thus, liberty:
freedom, hide : conceal are only completely synonymous.
Total synonymy is extremely rare in natural language, if it is possible at all. The German
linguist, W.Winter points out “ if forms differ, a priori, semantic equivalence cannot be
expected”.
Being fully aware of the intricacies and problems involved in finding a satisfactory definition
of synonymy, some linguists propose to restrict this concept to SIMILARITY IN RESPECT
OF COGNITIVE MEANING. Consider,e.g.:

1. On jest bardzo cierpliwy and 2. On ma wiele cierpliwości.

On the cognitive meaning level, these two sentences must be treated as synonymous.

Synonymy occurs at word, phrase, and sentence level. So, there are synonymous word,
phrases and sentences.

76
There are some linguists who suggest to treat synonymy as a relative concept. Under this
approach to synonymy, two linguistic expressions are regarded as synonymous with reference
to the solution of some linguistic problem X, and as non-synonymous with reference to the
solution of some linguistic problem Y, under the condition that these two expressions are
similar in meaning. For example, the modal items can, may and be allowed/permitted to,
under this approach, will be treated as synonymous as long as they denote NON-SPEAKER –
BASED PERMISSION, i.e., with somebody else than the speaker as the authority underlying
the permission. Thus the following sentence, as long as they express this modal concept, are
synonymous:

John can leave tomorrow/John may leave tomorrow/John is allowed/permitted to leave

tomorrow.

But when can and may denote SPEAKER-BASED PERMISSION (speaker as the source of
the permission,i.e., the permitter), they are not synonymous with be allowed/permitted to,
which always implies sb else than the speaker as permitter.

Linguistic studies of synonymy have emphasized the importance of context in deciding

whether a set of lexical expressions is synonymous, or not.

Polysemy (=multiple meaning)

The term refers to lexical items which convey a range of different meanings. The meanings
of a polysemous word are related to one another (i.e. one is derivable from another).
Let me illustrate this using the English modal verb can . In Present-day English this modal
item can be used as an expression of the following modal categories:

a. theoretical possibility, as in : Coffee can be grown in this country.

b. permission, as in: You can leave now (=I permit you to leave now)
c. ability , as in : He can lift a hundred weight ( =He’s able to…).

Underlying all these three modal categories is the notion of POSSIBILITY , so all these
modal notions (a,b,c) can be derived from this notion of possibility. Thus: He can lift a
hundred weight= It is possible for him to lift…, Coffee can be grown in this country= it is
possible for coffee to be grown in this country., etc.

The theoretical problem for the linguist is how to distinguish polysemy from homonymy (see
below). A number of criteria have been suggested, one of them we have already touched
upon, viz. the closeness of the relationship between the meanings involved. Another criterion
often resorted to by linguists is the etymology of lexical items ( see below, homonymy).

Homonymy

Synonymy is usually contrasted with homonymy, where several words with different
meanings have the same phonological form; for example, bank1, as in the bank of a river, and
bank2 ‘ He deposited all the money in his New York Bank (where bank ‘ the financial
institution where we deposit money’). Here, although there is one phonological form bank,
there are two lexical items .

77
The meanings of a homonym are totally unrelated, in contrast to a polysemous word whose
meanings, as has been shown above, are derivable one from another/are related.

There are two ways in which homonymy comes about. Firstly, related meanings of a once
polysemous word have drifted so far apart that there is no noticeable relationship between
them. This, e.g., the case with English pupil used in the senses: 1. a scholar, 2. iris of the eye.
Pupil, in both its senses, comes from the Latin pupillus/papilla, which is a diminutive of
pupus ‘child’. The pupil of the eye was so called, because of the tiny reflection of a human
being that can be observed in a person’s eye.
Alternatively, unrelated words , which were once phonologically distinct, in the course of
time have become phonologically identical; e.g., English die ‘to expire’ (verb) ‘ a cube
thrown in games of chance (Pol.kostka do gry)( noun).
In the first sense, when used as a verb, it derives from Old English verb diegan’to die’, and in
its use as a noun , it comes from Old French de ‘kostka do gry’, with which meaning it is still
used.
This shows how etymology comes to the linguist’s aid in his attempt to solve some of his
problems.

Hyponymy, meronymy/partonymy

Hyponymy is the relation which obtains between specific and general lexical items , such that
the former is ‘included’ in the latter, i.e., is a hyponym of the latter; e.g. cat is a hyponym of
animal, flower-pot is a hyponym of pot, etc. In each case there is a superordinate term
(called hypernym or hyperonym ) with reference to which the subordinate term can be
defined, as is the usual practice in dictionary definitions (e.g.,’ a flower-pot is a kind of pot’).

Meronymy/partonymy

Meronymy is the relationship which obtains between ‘parts’ and ‘wholes’, such as
wheel:car, knee:leg, etc. ‘ X is a part of Y’, i.e., X is a meronym of Y. The study of the
relationship between parts and wholes is called mereology.
The distinction between hyponymy and meronymy is clear enough. A knee is not a kind of
leg, but a part of leg; so phrases like knees and other parts of a leg are nonsensical.

Antonymy(oppositeness of meaning)

It is one of the most important semantic relations. It may be exemplified by the items big and
small. A characteristic property of this class of words is that they are gradable, and grading
implies comparison; e.g., X is bigger/smaller than Y. Some linguists use the term antonym to
mean only gradable pairs.
There are several types of antonyms:

polar antonyms (e.g. long:short) These are typically evaluatively neutral and objectively
descriptive.

overlapping antonyms These all have an evaluative polarity as part of their meaning’ one
term is commendatory , e.g., good, pretty,honest,etc., and the other term is deprecatory , e.g.,
bad, plain, dishonest,etc.

78
equipollent antonyms The number of such antonyms is highly restricted and they refer
to subjective sensations and emotions (e.g., hot;cold, happy:sad) or evaluations based on
subjective reactions rather than ‘objective’ standards (e.g., nice:nasty, pleasant;unpleasant)

Sense(intension) v. reference(extension)

One must distinguish between the sense of a linguistic expression and the reference of the
expression.

By sense (or meaning) of a word we mean its place in a system of relationships in which the
word contracts with other words in the vocabulary (of a language). This is de Saussure’s
definition of word meaning.

Reference is the class of objects or events to which a given linguistic expression , be it a word
or a larger unit, refers. For example, the phrase 1. hotel “Cracovia” in Kraków refers to a
definite object in Kraków. But an expression may refer neither to an object nor to an event ,
which can be illustrated by the following the lake situated in Katowice in the Koszutka
section, which has an empty reference (pusty zakres), since there is no such lake in
Katowice.
The same reference may be assigned by means of several , or even many , linguistic
expressions; e.g., information about Hotel “Cracovia” in Kraków may be given , apart from
1 above, also by The new hotel situated opposite the national Museum or The hotel situated
at the corner of 3rd May street and Trzech Wieszczow Lane in Kraków. These convey distinct
senses (meanings), but have the same reference.

To determine whether linguistic expressions have the same reference, we must know not
only the language in which the statements have been made (as in our examples above), but
also have sufficient knowledge about the given class of objects or events in the
extralinguistic world.
If knowledge of the language alone is sufficient to determine whether expressions refer to the
same class of objects, then we say that such expressions have the same
sense/intension/meaning. Take,e.g., the following:

1. Adam Mickiewicz is the author of Pan Tadeusz.

2. Adam Mickiewicz wrote Pan Tadeusz.

For anyone knowing English there is no doubt about the fact that (1) and (2) must be both
either true or false. We know on the basis of our knowledge of English that it is impossible
for one of the sentences to be true and the other to be false.

Translational semantics, Referential semantics

Semantic research on sense is called translational semantics, while semantic research on

reference is called referential semantics.

The object of translational semantics is primarily the relations of meanings that hold between
individual linguistic expressions,
Referential semantics has as the object of its research the relationship between linguistic
expressions and the objects or events to which they refer.

79
The term sense is used by some linguists for what others would describe as cognitive or
descriptive meaning. In this connection consider:

3. The Morning Star is the Evening Star.

The expressions the Morning Star and the Evening Star have the same reference since they
each refer to the same object, here, the planet, but they cannot be said to have the same sense,
for if they did, (3) would be tautologous, as is no doubt the following:

4. The Morning Star is the Morning Star.

But (3), unlike (4), is potentially informative, in the sense that it can make the hearer aware of
some fact of which he wasn’t previously aware. It follows that the Morning Star and the
Evening Star are not synonymous, i.e., they do not have the same sense.

Since synonymy is a relation which holds between two or more linguistic expressions, it is a
matter of sense, not of reference. Identical reference is a necessary , but not sufficient,
condition of synonymy.

( The Morning Star in Collins Cobuild Dictionary: ‘the planet Venus, which can be seen in
the sky just after sunrise; The Evening Star in Webster’s New World Dic of AmEnglish: ‘ the
bright planet Venus that can be seen in the western sky soon after sunset’)

Componential analysis (lexical decomposition), semantic field(lexical field)

The kind of analysis that helps to economically represent the systematic relations that obtain
between lexical items is called Componential Analysis(CA). CA is a semantic theory which
developed from a technique for the analysis of kinship vocabulary devised by American
anthropologists in the 1950s. This approach is used not only in semantics but also in
phonology (eg. by the linguists of the Prague School). In this kind of analysis the meanings
of lexical items are analysed not as unitary concepts but as complexes made up of components
of meaning which are themselves semantic primitives. For example, the distinction between
murder and kill may be stated explicitly and economically if murder is analysed as having a
meaning (sense) which is a complex component representing INTENTION, CAUSATION,
and DEATH, and kill as having a complex comprising only the semantic components
‘causation’ and ‘death’. Unlike killing someone, murdering someone is always intentional.
Or consider the items give and take , which can be shown to be distinct by virtue of their
contrasting complexes of components representing CAUSATION, CHANGE and
OWNERSHIP.
The main value of CA lies in the economy of statement of the relationships holding between
lexical items (e.g. words denoting kinship, colour terms, etc., constituting semantic fields ).
Semantic field is the organization of related words and expressions into a system which
shows their relationship to one another. For example, kinship terms such as father,
mother,brother, uncle, aunt belong to a semantic(or lexical) field whose relevant features
include generation, sex, membership of the father’s or mother’s side of the family, etc. The
absence of a word in a particular place in a semantic field of a language is called a lexical
gap. For example, in English there is no singular noun that covers both cow and bull as
horse covers stallion and mare. To give one more example, the words denoting colour are
also often cited as an example of a semantic field: the precise meaning of a colour word can

80
be understood only by placing it in relation to the other terms (words) which occur with it in
demarcating the colour spectrum.
CA enables us to account for relations such as hyponymy and antonymy and
incompatibility. Thus, e.g., using the method of CA, a lexical item P can be defined as a
hyponym of Q if all the features of Q are contained in the feature specification of P. For
example, spinster is a hyponym of woman because it contains all the features of woman as
part of its specification, and spinster is incompatible with bachelor by virtue of the contrast
of sex specification and with wife by virtue of the marital specification.
If we consider now pairs of words like cold-hot, young-old, etc., these pairs differ from the
following: alive-dead, married-single, etc. in that unlike the latter, they are gradable pairs,
hence the possible colder, hotter, younger,older and the impossible *aliver, *deader,etc.
Thus, the CA analysis shows that lexical items do not just stand in one relationship to one
another, but that each stands in relationships to many other items . Illustrations of this we
have provided above, but consider also spinster, which is a hyponym of woman, and woman
itself has other hyponyms; e.g., sculptress, waitress, and is itself a hyponym of adult. All
these interrelated items can be said to form networks of relations (or lexical fields).

This is not to say that the CA approach is without its weaknesses, but for the discussion of
these I would like to refer the discerning student to John Lyons’s book Semantics, vol.I.

Glossary
augmentative A morphological construction which is used to denote something of
large size, e.g., Polish chłopisko, konisko, etc. Very often, though not always, augmentatives
are interpreted pragmatically as being bad or nasty in some sense.

diminutive A morphological costruction which denotes something small, as for

example, Polish piesek, żabka, młoteczek, etc. In many languages, though not universally,
diminutives are interpreted pragmatically as referring to something nice or pleasant.

cohesion(Pol.spójność formalna,kohezja) The cohesion of a text is the explicit marking of

its coherence (for this term see below) by means of cohesive links (see below), e.g.
pronouns (he, they, my etc.) and word repetitions.

coherence (Pol.spójność właściwa,koherencja) Coherence is based on the conceptual links

between various entities referred to in the text(for text , see one of the earlier paragraphs)
between the various events evoked. The former is referred to as referential coherence, the
second as relational coherence. Typical referential expressions are pronouns and full noun
phrase (e.g. the woman next door ) . The reference may be to something outside the text or to
other concepts and referents in the text. A coherence relation is that aspect of the
interpretation of the text that is additional to the interpretation of the sentences or clauses in
isolation. For example, The unicorn died because it was lonely the coherence relation is
explicitly signaled by because. Here the second clause beginning with because specifies the
cause for the death of the unicorn.

Grice’s Conversational Principle A term derived from the work of the philosopher
H.P.Grice and frequently used in linguistics as part of the study of conversational structure.
The principle states that speakers try to cooperate with each other when communicating: they
will attempt to be informative, truthful ,relevant and clear (maxims ‘quantity’, ‘quality’,
‘relation’, and ‘manner’, respectively; (The Principle consists of these maxims)). And
listeners/addressees will normally assume that the speaker is following these maxims.

81
tautology An apparently meaningless expression in which one word is defined as
itself, as, e.g., in Boys will be boys. If such tautological expressions are used in a
conversation , clearly the speaker intends to communicate more than is said. It is argued that
the tautological statement Boys will be boys is inferred to mean something like ‘ That’s the
kind of unruly behaviour you would expect from boys (Levinson,S.C. Pragmatics,
1983:125)

Lecture 14
Language Contact and Borrowing

In one sense, by language contact we mean the use of more than one language in the same
place and at the same time. This definition of language contact draws our attention to some
communication between speakers of different languages but does not imply fluent
bilingualism or multilingualism, usually treated as a prerequisite, i.e., necessary condition.
This definition of language contact is not free of some ambiguity. First, ‘language’ can be
substituted by ‘dialect’, since, as has been indicated in one of the earlier lectures, the
boundary between language and dialect is vague. Given enough time and the right social
circumstances, dialects will turn into separate languages; during the transition period, from
dialect into language, it is difficult to differentiate between the language and the dialect. In
this perspective, the categories of dialect borrowing, i.e., borrowing from another dialect,
and cultural borrowing , i.e., borrowing from another language, are closely related.
Secondly, language contact does not necessarily involve face-to-face interactions between
groups of speakers. For example, in the past , religious languages like Latin spread mainly
via written texts, and there was often no direct or active contact between the speakers
involved. Nowadays, many people come into contact with a different language through the
radio, television or the Internet, so the media have, to some extent, substituted people in the
traditional language contact situation.

Language contact may be also referred to as a situation when speakers of one language are
exposed to another language spoken elsewhere. Under such circumstances, they may,
deliberately or unconsciously, introduce into their language elements of another language.
In all speech communities, language contact has some social consequences. Sometimes they
are advantageous, e.g., when two languages coexist peacefully (in countries which have two
or more official languages, like Switzerland). On the other hand, the dominant position of one
language can even lead to the extinction of the ‘weaker’ language. In numerous cases, a
strong position of one language led to the ‘death’/extinction of the vanishing language , as
was, for example, the case with hundreds of indigenous languages of African and American
tribes.

In the specialist literature, linguists have referred to ‘upper’/dominant language or

superstratum/superstrate, in contrast to ‘lower’/subordinate language or
substratum/substrate.(cf.L.Bloomfield, Language, 1933).

The assumptions that borrowings always show a superiority of the nation from whose
language they are transferred is not quite valid, particularly when the relations between the
languages involved are those of adstratum/adstrate, i.e., of equal prestige.

82
It is not just lexical items that are subject to borrowing, although they are no doubt imported
most easily; in fact, all aspects of language structure can be transferred from one language to
another, given the right social and linguistic circumstances.
Though in most cases one-way or mutual influences on languages are confined to
lexical(cultural) borrowing, we can find transfer in all areas of language structure:
phonology, morphology and syntax.

Borrowing: reasons for borrowing

Linguists claim that two linguistic factors are responsible for lexical borrowing:
a. The prestige motive
b. Necessity (the need-filling motive)

If speakers of a language take over new cultural items, such, e.g., as new technologies,
religious or political concepts or references to foreign fauna(animal) or flora(plants), there is
an obvious need for the relevant vocabulary. In this way, English borrowed a number of
Russian words referring to traditional Russian realia; for example, muzhik, roublr, troika,
vodka, tsar, tsarina (also spelled czarina), Kalashnikov, the quite recent perestrojka, and
many others.

On the other hand, prestige is taken to be a stimulus that implies heavy borrowing and it
reflects the higher status of the donor/lending language (język-dawca).

The term borrowing is used in linguistic literature in two senses: in one sense, it refers to the
transfer of linguistic features of any kind from one language to another as a result of contact.
In the other sense, the term borrowing(zapożyczenie,pożyczka) refers to a linguistic form
taken over by one language or dialect from another; such borrowings are usually called loan
words, for example garage, chagrin, bonhomie are loan words of French origin.
Less commonly sounds and grammatical structures may be borrowed, e.g. the pronunciation
of the above loan words with French or quasi-French accent (see below for structural
borrowing).

Some more basic terms :

Borrowing/recipient language (język-biorca) The language to which a loanword is taken.

Loanblend/hybrid A loanword/borrowing in which one element is imported and one
substituted , e.g., refusnik E.refuse + Russian –nik, Sovietize Rus. Soviet+ E.-ize.
Loanshift/loan translation/calque (kalka językowa) A borrowing in which direct
translation of the constituent elements of a foreign word takes place, e.g., superman (from
German Übermensch), mother tongue (from Latin lingua maternal ), drapacz chmur (from
English skyscraper), etc.
Loanword 1. A borrowing with full morphemic importation. 2. A fully assimilated
lexical item taken over from a donor language into a recipient/receiving language.

Lexical /cultural borrowing

As has been hinted at, borrowing relies on bilingualism or multilingualism, and these
necessarily mean language contact. In contact situations, linguistic elements can be
transferred from one language to another. Sheer necessity (need-filling motive) seems to be

83
the most common motive for borrowing. Speakers may have to refer to some unfamiliar
object or concept for which they have no expression in their own native language.
Using ready-made designations is more economical than describing things afresh. Following
are some examples of cultural borrowings/loan words in English:

apartheid (from Afrikaans,’apartness’, one of the two official languages of the Republic of
South Africa)
perestrojka, sputnik, troika (troika ‘ a small Russian carriage drawn by a team of three horses
abreast’, also ‘ a group of three persons, especially political leaders’)
pyjamas (from Hindi)

Cultural/lexical borrowing is often bidirectional; e.g., Polish has borrowed from Russian,
which in turn borrowed from Polish (e.g. vodka is originally Polish, not Russian, as is
generally believed (in this connection cf. K.Polański’s paper ‘ From the history of the word
vodka’ in Studia Anglica Posnaniensia 4, 1972 ).

However, usually , especially in cases of close language contact, the direction is from the
more prestigious to the less prestigious language.
After the Norman Conquest , and especially in the 13 th and 14th centuries, with Central French
as the major source of French loans, there was a great influx of French vocabulary into
English, mainly connected with the administration, warfare, the arts and the Church. These
loanwords had prestigious connotations. Borrowed and native English words occupied
different registers, i.e. belonged to different styles. The French words were literary and
formal, and the native ones were ordinary words. Consider for example:

Ask-question-interrogate
Leech-doctor-physician

In contemporary English dictionaries the native English leech is qualified as ‘old use’. Here
interrogate and physician are more literary and formal than their native English synonyms.
Because of the prestige of French (due to its political and social status in England), such
words were taken by some people to be a sign of education or social superiority, marking
them off from the common herd.

Approximately 50 new words were being introduced into English every year around 1500,
but about 350! per year by 1600 . This widespread borrowing resulted from the new
technology of printing which involved translation and widespread dissemination of classical
literature.

Opponents of the new loanwords (loans) called them inkhorn terms to indicate their literary
character, and said that they were obscure and difficult to English speakers.. So they
suggested to avoid them as much as possible and use native words in their place. For
example, Ralph Lever, an Elizabethan language purist, published as textbook under the title
the Art of Reason, Rightly Termed ‘Witcraft. In the Forespeache (i.e., ‘preface’ ) to the work,
the author argues that new technical terms should be made by forming compounds from
existing English monosyllabic words, not by borrowing from Greek or Latin. His main
argument is comprehensibility, as he explains, the meaning of such a compound is self-
evident, whereas classical words are obscure. Examples of some of the form that he coined

84
are: witcraft for logic, endsay for conclusion , naysay for negation , saywhat for definition,
etc. However, Lever’s coinages did not catch on.
When loans enter the more prestigious language , these often have derogatory connotational
meanings.
Some words are more borrowable than others: thus, basic vocabulary is borrowed almost
always in situations where neither of the languages concerned is perceived as more
prestigious than the other. For example, English borrowed a great deal of basic vocabulary
from Old Norse and Middle Norse, when the languages were in close contact (In the North of
England, which was settled by the Viking invaders). Items like skin,sky, window, get,take, etc.
are of Scandinavian origin. Since Old English and Old Norse were most likely mutually
intelligible at that time, we have a case of dialect borrowing (borrowing from another
dialect).

Adoption/importation and adaptation/nativization/assimilation

The bilingual speaker has a choice, he can either borrow a word in its donor-language form,
preserving features unknown to the recipient language (and this is called
adoption/importation), or he may adapt/nativize the loan, trying to adapt it to the linguistic
patterns of the recipient language ( and this is called adaptation/assimilation/nativization).

Speakers tend to produce more adoptions when they try to impress their listeners, or when
the vocabulary involved has particularly prestigious connotations, as, e.g., French loans in
the area of food, wine and cooking currently do in English.

If the speakers of the recipient/lending language are familiar with the donor language, they are
less likely to adapt/nativize the words borrowed from it.

Adaptation takes place at the phonological, morphological and semantic levels of the recipient
language.

Phonological adaptation (or phonic substitution ) takes place when speakers attempt to
replace a foreign sound by a sound from his native language. Speakers usually substitute the
closest possible sound from their native language. For example, Polish lacks the dental
fricative /Ø/ , so it tends to replace it by either/t/ or /f/ (compare E. thriller pronounced in
Polish as /triler/).

Phonic substitution will also depend on spelling.

English loanwords in a language like Polish must be assigned grammatical gender. This is
adaptation on the morphological level of Polish. In his discussion of this process, J.Fisiak
(Biuletyn Polskiego Towarzystwa Językowego 33, 1975) mentions the following criteria as
governing gender assignment : 1. Sex, 2. The form of the loanword, 3. Semantic association
with a displaced native equivalent, 4. Suffixal analogy, 5. Graphic analogy, 6. Homonymity,
and 7.semantic categories. For Polish, the following criteria appear to be sufficient:
phonological/graphical and semantic criteris.
The phonological/graphical criteria have to do with the structure of the termination of a loan.
89

Polish has three genders:

85
Masculine Feminine Neuter

consonant -a (krowa) -e (słońce)

-y,-i(chorąży) consonant (noc) - o (płuco)
- ę (kurczę)
- a (włóczęga) - um (muzeum)

Polish masculine nouns include a large number of items ending in a consonant, therefore
English loans ending in a consonant are usually treated as masculine; e.g., aut, brydż,
chuligan,bar, etc.

English loanwords ending in –ist end in Polish in –ista, e.g., Methodist, conformist
(Pol.metodysta, konformista).

Using semantic criteria, we treat dingo as masculine, despite its ending in –o, since it is a kind
of a dog and the word pies in Polish masculine gender.

Some loans ending in a consonant or in – are assigned feminine gender on account of their
specification to sex; these include miss. Lady, call-girl, and the like.

A few borrowings are feminine in Polish owing to semantic relations with native words, e.g.,
whisky, which is semantically related to the Polish word wódka.

Another example of a similar kind is provided by city , assigned neuter gender through its
semantic association with Polish miasto, which is neuter gender.

Structural borrowing

In structural borrowing the components of language subject to the process are: phonology,
morphology and syntax. These are much less borrowable than the lexicon.

Structural loans are only possible between two very similar language systems. It seems to be
the case that the structural elements borrowed are typically those that fit typologically into
the recipient language.

As for the borrowing of morphological material, it is easier to borrow derivational affixes i.e.,
elements such as –able, --ity, etc. than inflectional ones,i.e., such as –s or –ing (in present
participle forms) ones.

And thus, the derivational morpheme (suffix) –able comes from French> Middle English
borrowed quite a number of French words ending in –able, e.g., measurable, reasonable,
comfortable and a lot of similar formations ending in – able. Some time later these French
loans were reanalyzed and allowed the suffix to attach to other Romance words.And now –
able attaches also to non-Romance native English words.
Other examples of structural borrowings come from Old Norse/Scandinavian(the language of
the Vikings).These include functional words (i.e., closed-system items) such as pronouns,e.g.,
they,them,their,she, prepositions, conjunctions and adverbs: the same(pron.), fro (advb.), used
today only in the phrase to and fro, till (conjunct.and prep.), though (conjunct.and
advb.),which the Old Norse equivalent of OE þeah, with, which replaced the OE mid. The Old

86
Norse use of at is a sign of the infinitive to be seen in the English ado (at-do) (Baugh
2002:100).
A certain number of inflectional elements peculiar to the Northumbrian dialect are attributed
to Scandinavian influence, among them the – s of the third person singular present indicative
of verbs, and the participial ending – and ( bindand) , corresponding to – end and –ind in the
Midlands and the South, now replaced by – ing.
O.Jespersen (1948:76-77) notes that the omission of the relative pronoun in relative clauses
(rare in Old English) and the retention or omission of the conjunction that are in conformity
with Danish usage.

The phoneme / / is an example of a borrowing of an element belonging to the phonological

component(plane) of language. First of all, there was no such phoneme in Old English , it
arose from medial combinations of /z/ and /j/ in words of French origin such as measure,
treasure; then its distribution extended by French loans like rouge, beige, with final / /.
(cf.modern English beige ,pronounced as [bei /.

And just one more example from the phonological plane. Old English had no /v/ phoneme. In
OE, /v/ was an allophone of the phoneme /f/ which occurred in an intervocalic position
(between two vowels), as in OE lufu [luvu], or between a vowel and a voiced consonant, as in
hæfde[hævde]. Thus, the /v/ phoneme arose under the influence of French loans beginning
with /v/, as in variety.

In OE, there were only two tenses: present tense and preterite (past tense); there was no
future tense construction with shall/will. The OE antecedants of these verbs were used as full
verbs with lexical meanings; OE sculan meant ‘have as one’s duty,be under obligation’, and
willan meant ‘to will, to wish’. In OE future time reference was expressed by the present
tense indicative form of a verb, accompanied by adverbs when necessary. For example, in
King Alfred’s translation of Pope Gregory’s Pastoral Care , the Latin future is rendered 61
times by the OE present indicative, 7 times by the past optative mood form, and 6 times by
shall plus infinitive and 9 by will plus infinitive. But here shall and will do not serve merely
as auxiliaries to form the future construction, but appear in their lexical senses.

The development of the periphrastic future construction shall/will plus infinitive, as it is

claimed, was due to the Scandinavian influence on Old English and Middle English . In Old
and Middle Scandinavian futurity was expressed by an auxiliary verb plus infinitive. (cf.
Kirsch,M. 1959. Scandinavian influence on English syntax.[in:] Publications of the Modern
Language Association of America, vol.74 )

Glossary

distribution The sum of the contexts in which a linguistic element may appear, e.g., the
distribution of the Polish /b/ phoneme does not include the word-final position, but only the
initial and medial positions. Three types of distribution of linguistic elementsare usually
distinguished:
a. Complementary d. (dystrybucja uzupełniająca), b. contrastive d. (d.kontrastująca),
and c. free variation d. swobodna)
If 2 linguistic units have no contexts in common, they are in complementary distribution.
For example, in English the word initial /p/ (as in pat) is aspirated (for this term see below),

87
but if it appears before a consonant (as in primary), it is unaspirated. Text units that are in
complementary distribution are variants of the same linguistic unit, here the phoneme /p/.
If two linguistic units share some of their environments, and if the use of either of them
involves change of meaning. They are in contrastive distribution. For example, /t/ and /d/ in
bat and bad or – ed and – ing in bombed and bombing . Elements that are contrastive
distribution belong to different units of the language.
When two or more units occur in the same position without any apparent change of meaning,
they are said to be in free variation. For example, who and whom in the English sentence The
man who/whom I saw. Such variations are now often considered as social variations or
stylistic variations.

aspiration/aspirated/unaspirated A puff of air (acoustically a period of voicelessness)

after the release of articulation . For example, in English the stop consonants /p.t,k/ are
aspirated when they are syllable initial , as in initial sounds of pie, tie, kite. When these
phonemes , consonants here, are preceded by, say, /s/ , as e.g. in span. stairs, skate, there is no
puff of air and these sounds are unaspirated.

gender A grammatical category used for the analysis of word-classes displaying such
contrasts as masculine, feminine, and neuter, animate and inanimate , etc. In linguistics, a
distinction is made between natural gender, where items refer to the sex of real world
entities and grammatical gender, which has nothing to do with sex, but which has an
important role in signalling grammatical relationships between words in a sentence (e.g.,
adjectives agreeing with nouns, as in Polish wysoki mężczyzna, wysoka kobieta, wysokie
drzewo)

periphrasis/periphrastic In grammatical analysis, the term refers to the use of separate

words instead of inflections to express the same grammatical relationship. For example, in
English, the comparison of adjectives involves both inflection (happier,happiest) and
periphrasis ( more happy, most happy – periphrastic forms/constructions). And compare
the Polish napiszę and będę pisał, the latter being an example of a Polish periphrastic
construction.

reanalysis/restructuring In the study of language change, a development which alters

the structure or function of a linguistic form. For example, when two words coalesce as a
compound, their separate identities need to be reanalyzed/restructured as a whole; thus,
hair noun+ cut verb > haircut noun.

88
Lecture 15
Language change, its types and causes

Language Change is the central object of study in historical linguistics. Every living language
is continually changing; indeed, it is always in the middle of a number of changes. Historical
linguists seek both to explain the long sequences of changes which have occurred in the past
in particular languages and to extract general principles of language change.

Among the causes of language change are mentioned such factors as:

1. The principle of least effort (or economy-of-effort principle/least effort principle),

i.e., any of several related but distinct putative principles of language structure and
change, according to which languages tend to change in such a way as to minimize the
effort involved in speaking. A version of the economy-of-effort principle is one of
the principles advanced by Georg Zipf , according to which “ A segment of high
frequency tends to become phonologically simple (or unmarked; for unmarked see
one of the earlier glossaries)” ( Collinge,N.E. 1985. The Laws of Indo-European.
Amsterdam:John Benjamins, pp.256-258).

2. Analogy ( which we will deal with at more length in one of the following parts of this
lecture)

3. For some linguists, the main cause of language change is the fact that each generation
learns its language anew and makes its own hypotheses about the grammar. The
child’s grammar differs from that of the adult. (McMahon, 1996)

4. Language variation, i.e., the existence of competing linguistic forms within a single
speech community or language. Linguistic variation applies not only to two or more
speakers, it also occurs within the speech of one individual (we speak of one’s own
idiolect). To a less or greater degree, all speakers adjust their speech to the social
situation. Being social creatures, speakers are more or less conscious of the
functional value of of the different varieties of speech and vary their styles according
to situational context.

The linguists of the past often tended to regard a language as essentially homogeneous
and invariant at any point in time, and they accordingly interpreted language change
as a more or less abrupt shift from one steady state to another. In this view variation
was regarded as peripheral at best and as a nuisance at worst. However, in the 1960s
linguists came to realize that variation is the normal state of a language , and that the

89
 absence of variation would be pathological. Currently almost all studies of language
change recognize the central role of language variation; in short, variation is seen as
the vehicle of all change.

Given all the variations in the language, children will tend to develop a grammar which at
many points differs from that of old speakers. Children have the capacity to restructure the
language they learn, i.e., to reanalyze it and thus create new sets of patterns. For example,
the development of the English perfective constructions is assumed to have appeared as the
result of restructuring the earlier (Old English) possessive construction with have.
Thus,

OE He hæþ hinne gebundenne > Modern English He has bound him

He –has –him- in the state of being
bound

OE habban ‘possess’

When enough speakers restructure in similar ways, radical changes in the language occur>
It is assumed that the ability to reanalyze/restructure whole grammatical patterns and then
generalize the analysis consistently to all kinds of contexts is not characteristic of speakers
after the language maturation point. So, according to this view, adult speakers may innovate,
but they do not restructure the language (McMahon 1996:41). In earlier English, both forms
be and have were used as perfect aspect auxiliaries. At first be combined with past participles
of mutative verbs (e.g., verbs like go, move, ride,arrive,turn, etc.) and have went with other
verbs. Since the perfective construction with have was the by far predominant pattern, in
course of time it ousted the perfective construction with be ( a relic of this old perfective
construction is preserved in He’s gone).

Types of language change: simplification and elaboration

Simplification There are two kinds of simplification: loss of a linguistic pattern or a

grammatical rule and generalization. For example, in Old English, adjectives agreed with
the nouns they modified in number, case and gender ; thus: sumu tid ‘some time’ (Nom.
fem.sg.), sum monn ‘some man’ (Nom.masc.sg.). Recall also the history of the English
perfective construction. (loss of the perf.aspect be variant).

Generalization is the other kind of simplification and it consists in the extension of

patterns ,which at an earlier stage had limited use, to more and more contexts. In Old English
there were many suffixes which were used to form plurals of nouns, the suffix – s was one of
them. Other OE pluralizing suffixes were: - a, - an, - u, etc. Plurals of nouns in OE could
also be formed by mutation (i.e., by the change of the root vowel of a word, as in boc:bec,
cū:cy, , etc. In course of time, the – s became the only productive pluralizing suffix.

One of the principal conditions of simplification is the existence of variable rules and
patterns in the grammar of a language. E.g. when speakers had a choice between have and
be perfective constructions, or a choice among the various pluralizing suffixes.

Another example of simplification from the history of the English language is reduction of
multiple negation. In earlier English, including Shakespeare’s English, constructions like, for

90
example, I don’t want no toys no more were perfectly normal. Nowadays, at least in Standard
English, only one plural marker is possible.

Elaboration

Elaboration involves increased complexity. The simplest kind of elaboration is the addition
of a pattern or rule, as when a new auxiliary do came into the language in Middle English.
In OE, questions were formed by simple inversion, as in Slog he þā leode…? And consider
the following from Shakespeare’s A Midsummer Night’s Dream:

Demetrius: Why rebuke you him that loves you so?

Hermia: Lysander, whereto tends all this?

( each one involving inversion of the subject and the verb)

Elaboration occurs in language also when restrictions are imposed on a pattern or rule,
making the conditions for the use of the pattern or rule more complex. For example, the
restriction on the relative clause reduction that has come to operate in the 20 th century , and
which no longer permits us (that is, Standard English speakers) to say things like He’s the
man hijacked the plane. This construction, without the relative pronoun who, is quite normal
in substandard varieties of English, e.g., in Cockney).

Linguists claim that simplification is brought about almost exclusively by children (see
McMahon 1996, chapters on language change). The adult speaker normally elaborates. Since
it is restructuring by children which brings about major changes , and since restructuring
almost always involves simplification, simplification is regarded as the main type of change
(Note that this statement would be in agreement with the economy of effort principle).

However, it must be pointed out that what is simplification in one part of the grammar of a
language may sometimes result in elaboration in another. For example, loss of case
inflection on nouns in English was a simplification as far as the formation of nouns was
concerned. However, it was accompanied by elaborations in the use of prepositions instead of
cases and increased constraints on word order. In this connection, compare OE þa scipu þāra
Dena, which was replaced by the later the ships of the Danes . Notice that this kind of change
involves segmentalization, a process whereby a noun+case pattern is replaced by a
preposition+noun phrase (of the Danes ), a process in which inflection is replaced by a word:
-ra by of in our illustrative example.

The opposite of segmentalization is desegmentalization, which involves replacement of a

phrase by one-word unit, as was the case with English by cause , which developed into
today’s because.

Semantic change

Meaning changes relatively quickly and easily. For example, the word hierarchy was first
used in English for the medieval classification of angels into various ranks. In the 17 th century
its meaning was extended to ‘ the ranking of clergymen’, and later to any system of grading.
The first element of the word derives from Greek hieros meaning ‘supernatural, powerful,

91
holy’ or ‘sacred’. And the second element of the word comes from Greek archos meaning
’ruler, leader’.
Just as the meaning of hierarchy developed from angels to priesthood, so too it was extended
from the secular government to the classification of a group of people with regard to ability
or economic or social standing, and finally, to the arrangement of objects or values in a
graded series. In this connection, consider the following example which refers back to the
early meanings of hierarchy:

In your issue of the Daily Telegraph March 3,1976 you refer to the Soviet Communist
Hierarchy. I eagerly await your publication of a photograph of Mr Brezhnev wearing a cape
…- or even wings”.

Meaning is intrinsically linked with culture, which implies that to understand a change in the
meaning of a word , we may require a good grasp of of the socio-cultural situation within a
speech-community. To illustrate this, let us consider the etymology of the English word
money .English money and mint happen to come from the same root as Latin monere
meaning ‘ to warn, admonish, advise’, because in Ancient Rome money was made in the
precincts of the temple of Juno Moneta, i.e., Juno the Admonisher. Thus the proper name has
been transferred to the product owing to historical accident.
Although there is a lot of truth in saying that words have their own histories, it is at the same
time true that semantic changes are to a great extent regular. It is worth quoting on this
occasion Otto Jespersen’s words:

“[…] there are universal laws of thought which are reflected in the laws of change of meaning
[…] even if the science of meaning (i.e. semantics –P.K.) has not yet made much advance
towards discovering them”. (O.Jespersen, 1946. Mankind, Nation, and Individual from a
Linguistic Point of View. London: George Allen and Unwin, p.212).

Considering the current progress in semantic studies, we must say that Jespersen’s words have
become somewhat out of date. Those words were uttered almost seventy years ago, but today
the methods of linguistic research are much more advanced so that those dealing with
semantic problems nowadays are much better prepared for their tasks than their
predecessors.

For a semantic change to occur there must be fulfilled a number of conditions:

a/ the existence of polysemy

Words are typically polysemic (or polysemous); most words convey a number of meanings
or cover a whole range of shades of meaning.

Words can lose or gain meanings relatively easily, and they do not have to lose an earlier
meaning to acquire a new one. Each word has one central meaning and various marginal
meanings, and semantic change takes place when speakers stop using the central meaning
and reinterpret a marginal meaning as the central one. Take, for example, the word
slowness, which ousted earlier sloth; sloth was related to the adjective slow, just as truth still

92
is to the adjective true, but sloth is now retained in its previously marginal/peripheral sense
‘laziness’.

Alternatively, a loanword may take over the central meaning of a native word, which may
then become obsolete or be retained with an altered marginal sense. In this connection,
consider:

Old English(OE) loanwords 20thc.

deor ‘animal’ animal deer (Pol.jeleń,sarna,łoś)

(from Latin via French) (restricted to a specific
species of animal)
fugol ‘bird’ bird fowl (Pol.drób)
(etymology unknown)

steorþan ‘to die’ die starve ‘die of hunger’ (cf. modern German
(of Scandinavian origin) sterben ‘to die)

b/ another condition has to do with the fact that language is transmitted discontinuously, i.e.,
and , as has already been indicated, children do not receive a ready-made grammar from
their parents , but construct a grammar for themselves on the basis of available linguistic
data, i.e. the utterances produced by their parents.
Children may, therefore, learn imperfectly, or make wrong inferences which, with time,
modify or change the language they acquire as their native tongue. This process may be
responsible , e.g., for the change of OE (ge)bed , originally meaning prayer, to its Present-day
English sense ‘ small wooden or other balls on a string’. If an adult using a rosary explains to
a child that he/she is counting his/her beads, we have an ambiguous context; the adult intends
to say that he is saying his prayers, but the child sees only the accompanying concrete
activity involving the movement of the little balls which make up the rosary. Thus bead
consequently changes its original sense from ‘prayer’ to ‘a small wooden ball’
(Pol.paciorek).

And consider the now obsolete phrase to tell one’s beads meaning ‘ to say one’s prayers while
counting beads on a rosary’. One of the OE meanings of the verb tellan was ‘to count’, and
its other meanings were ‘ to account, esteem’. The verb comes from the same source as the
modern German zahlen.
Thus:

OE bed prayer’ > Present-day English bead ‘ a ball of wood, glass’

c/ Also, semantic change is sanctioned by the doctrine of the arbitrariness of the linguistic
sign.
Words are linguistic signs (see one of our earlier lectures) which are symbols (unlike icons
and indexes. And, as we now know, in a symbol there is no connection or resemblance
between sign (signifier)) and object(signified). A symbol conveys information only because

93
people agree that it will stand for what it does. The nature of linguistic signs is thus
arbitrary/conventional.
Earlier it has been said that a linguistic sign is bipartite , which means that it is made up of a
form (called signifier) and content/meaning (called a signified), and these two components of
a sign are arbitrarily linked;

form/signifier

content/meaning

Before the idea of arbitrariness

became current in linguistics,
semantic change could not be really studied at all.
Arbitrariness now allows to regard the signifier and the signified as essentially independent;
either may, therefore, change with time, i.e., form(signifier) independently of
content/meaning(the signified), and vice versa.

For example, consider onomatopoeia. In the case of onomatopoeia, where a non-arbitrary

link exists between form and content, there tends to be greater stability, which is to say that
onomatopoeic expressions tend to resist both sound change and semantic change.
Onomatopoeic expressions comprise, among others, items such as the following: oh, ach, brr,
wow, ruff-ruff, tut-tut, tick-tock, etc., etc.

Types of semantic change: extension/generalization and restriction/specialization/

narrowing

Semantic change may consist in extending or narrowing the range of meanings denoted by
words.

Restriction/specialization/narrowing of meaning involves an increase in the information

conveyed, since a restricted form is applicable to fewer situations, but tells us more about
each of them. The scope/extension (Pol.zakres/ekstensja) of the word has been restricted but
its meaning (sense/intension) (Pol.treść/intensja) has been enriched with an additional
feature.
Good illustrations of this type of semantic change are the earlier quoted examples from Old
English : deor, fugol, steorþan.

94
OE steorþan ‘ to die’ and its modern English counterpart starve ‘ die of hunger’, but also ‘ be
extremely hungry’, and dialectally, ‘ be very cold’ .
OE mete ‘ had the meaning ‘ food’ , and it now means ‘ meat food’ or just ‘ meat’.
In Middle English(ME) , voyage (derived from French) meant ‘ a journey’, and now its
meaning is restricted to ‘ a journey by sea’.

To take just one more example, the ME antecedent of today’s girl /gurgle referred to young
people, both boys and girls, and its present-day counterpart has the meaning ‘ female child,
young woman’.

Extension /generalization of meaning results from a generalization from the specific case to
the class of which the specific case is a member. Extension here means an extension of
possible referents (Pol.desygnaty wyrazu) or contexts of occurrence.
As our illustration of this type of semantic change , let us take the French lexical item panier
‘basket’ (Pol. kosz,koszyk), which comes from the Latin panarium ‘ bread- basket’, which
itself derived from the Latin noun panis ‘ bread’. When, with time, the connexion with bread
disappeared, panier could be applied to more objects than before and thus be used in a wide
range of contexts.

Good examples of extension provide English words which have historically derived from
proper names. For example, bedlam and quisling were originally proper names, and now they
are used to refer to a class of objects.
Bedlam was originally the name of a specific hospital (lunatic asylum) (St.Mary of
Bethlehem) in London. Now the word refers to any chaotic situation (Pol. bałagan,
harmider), and ‘asylum for old people’ is now old use.

In the 16th c., an inmate of this asylum came to be called a bedlam (a use not noted in PDE).

Bedlam, the hospital, was infamous for its brutality and disorder. By the latter half of the 17th
century the word bedlam had begun to be used in a generic way for any lunatic asylum
(equivalent to the colloquial madhouse).

Consider the following example:

(Year 1713) Our house is a sort of Bedlam and nothing in order.

Quisling (Pol.zdrajca) is of Norwegian origin. A Norwegian by the name Vidkun Quisling

was known for his collaborating with the Nazis during World War Two. Today quisling
refers to anyone who collaborates with an enemy.

Also trade names are known to be used as common names , as is the case, e.g., with Kleenex,
now a common name for a facial tissue.

To take one more example of an extension of a word’s meaning, consider OE foda’animal

fodder’, which has evolved into modern English food with the meaning ‘solid food, all sorts
of solid nourishments’. In this meaning(i.e. ‘solid food’), food replaced the OE mete ‘food’,
which itself was restricted to the sense of ‘meat’.
Apart from mete, OE had a word like flæsc referring to animal tissue in general (Pol.mięso).
In course of time , the meaning of mete has been restricted to animal flesh and that of OE
flæsc to the tissue of humans (Pol. ciało). But compare the use of modern English flesh in the
phrase fruit flesh (Pol.miąższ).

95
This shows that a change of meaning in one word in a language often leads to a shift of
meaning in other words that are somehow semantically related, i.e., belong to the same
semantic field (see one of our earlier lectures).

Amelioration and pejoration

Other types of semantic change are amelioration (Pol.melioryzacja) and pejoration (In
M.Szymczak’s Słownik Języka Polskiego I have found the adjective pejoratywny, but no
word like pejoryzacja).

Amelioration involves an improvement in assigned value. Ameliorative words are thus

words conveying positive meanings. For example, OE cniht ‘boy, attendant,servant’ had less
positive connotations than its modern English counterpart knight (cf. German Knecht, which
preserves the original sense’ parobek,giermek’, and metaphorically, ‘służalec’.

To take another example, English sophisticated now means’ worldly-wise, intellectually

appealing, cultured’ as compared to its earlier sense ‘ artificial’.

By pejoration we mean the development of a less favourable meaning or unpleasant

connotations of a word. It is a well known fact that the attitude of speakers to particular
words may also change as the value assigned to the referents of words changes. Pejoration is
thus often due to social prejudice.

Illustrative examples : French cretin shifted from meaning ‘ christian’ to ‘ stupid’, a change
which is sometimes attributed to the praiseworthy Christian habit of turning the other cheek ,
when attacked or insulted.
And, the earlier English word sely meant ‘ blessed’, its modern E counterpart silly means
‘foolish’.

Causes of semantic changes

Andre Meillet in his book Linguistique historique et linguistique générale (Historical

linguistics and general linguistics) mentions three types of causes:
linguistic, historical and social

Linguistic causes are language-internal (structural/systemic), which means that they are
linguistically conditioned exclusively. The best example of linguistically conditioned
semantic change is grammaticalization , i.e. a process whereby a lexical word acquires the
function of a grammatical/functional word. For example, OE verb habban has acquired the
function of the perfect aspect auxiliary verb, in which it occurs in the sentence He has seen
the film many times.

Old French had two negative particles: non and ne. Initially, non appeared in stressed and ne
in unstressed positions. Examples:

Elle ne dort pas. (‘She’s not sleeping)

Vous venez ou non? (‘Are you coming or not’)

96
In early French, ne began to be reinforced by the particles point and pas , which had been
independent nouns. ( pas from Latin passus ‘a step’), as in Je ne vais pas (‘I’m not going’).
And point (from Latin punctum ‘ place, a spot’, as in Je ne’ai point de café ( ‘I have no
coffee’).
These French particles originally carried an emphatic , reinforcing meaning something like ‘
a damn/a shit/ a fig’ , as these are used in English (e.g. I don’t give a damn/I don’t care a
fig/one iota).

Such negative reinforcers appear to be popular emphatic strategies in language.

These French particles (point, pas) later lost their emphatic meaning and marked the end of
the scope of negation, before they became semantically negative themselves, i.e., before they
began to function as negative markers.

Historical causes

Historical causes are language-external, in that they involve a change in the material culture
of speech communities. Referents of words frequently change due to technological
innovation, but the name(form) remains the same. Consider,e.g., Latin carrus’four-wheeled
vehicle, chariot’, which has given English car ‘automobile’.
And Eskimo umiaq originally meant ‘ 18-foot sealskin boat’, and now the word refers to any
kind of boat or even ship.
English pen, which historically goes back to Latin penna’feather’ , now denotes any writing
implement, and its original meaning was ‘ quill pen’ (Pol.gęsie pióro), which in those days
was the only tool used for writing.

Social causes

A word tends to acquire a new meaning due to its use by a particular social group, or a word
used in a specific sense by some group comes into common currency with an extended
meaning. Examples:

English lure in its present-day meaning ‘to attract’ derives from the field of falconry , where
it referred originally to the bundle of feathers with which the falconer tried to attract his
hawk.
English bishop , from the Greek episkopos ‘overseer’; its religious sense results from its use
within Christian community.
And,finally, Latin trachere’pull’ gave a corresponding French word with the sense ‘milk’, as
the result of its restriction to farming contexts.

More on this subject you can find in G.Hughes. 1988. Words in time: a social history of the
English vocabulary. Oxford: Blackwells.

Psychological causes

To the above list of causes of semantic change S.Ullmann ( The Principles of Semantics,1957,
Glasgow:Jackson,Son&Co., Oxford: Basil Blackwell) adds a fourth one, viz., a psychological
category of causation.

Psychological factors responsible for semantic change have largely to do with taboo and
euphemism.

97
Often religious concepts , dangerous animals and acts and objects thought of as unpleasant
or distasteful become taboo (sth which religion or custom regards as forbidden). Their names
cannot be spoken, and euphemisms ( a polite ,pleasant word/expression used to refer to sth
which people find unpleasant, forbidden, etc. ) are substituted , thus causing a semantic
change in the euphemistic expression.
For example, English bear originally denoted ‘brown’ , as its Lithuanian cognate still does.
Compare also Russian medv’ed ‘honey eater’, which is its original meaning.

Also, the Devil cannot in some speech communities be referred to directly, which would
account for the wealth of euphemistic expressions. Consider some scots euphemisms for the
evil spirit: Auld Nick, Clutie, etc. The same applies to God referred to by numerous
euphemisms meaning ‘master’(e.g. Pol. Pan Niebios).

Some sources of euphemisms

One of the sources is war and violence, which provide a fertile ground for all kinds of
euphemisms; e.g., liquidation for murder, intervention/military operations for war, etc.
Such euphemistic expressions are assumed to distance the reader/addressee from what is
really happening.
Another rich source of euphemisms is so-called Political Correctness (PC) , i.e. behaviour or
belief that reflect the attitudes and language that are typical of people who hold liberal or
left-wing views. Illustrative examples of euphemisms related to PC are provided in the
previous paragraph. Other examples: refuse collector/garbage engineer instead of the earlier
and ordinary dustbin man, and rodent operative instead of rat catcher (which, by the way,
are typically American English expressions).

Handicaps of various kinds are also euphemized: visually challenged instead of blind,
financially challenged instead of poor, differently abled instead of disabled , and so on and
on.

Before I abandon the subject of causes of semantic changes for good, let me discuss, if only
very briefly, one more type of cause, viz. semantic change as the result of similarity and
contiguity in the meanings of two words.
Thus, meaning change may result from a similarity in the meanings of two words, leading to
Metaphor, and from a contiguity of meaning , leading to metonymy (metonymy is a
fundamental part of our conceptual system whereby we take one well-understood aspect of
something to represent/stand for the thing as a whole; part of a whole).
For example, consider The buses are on strike , where the object , the buses are used for the
users of the buses. Or, for example, where Warsaw stands for the Polish government.

While metonymy is based on the part-for-the whole relation, metaphor involves similarity of
meanings; here an imagined link is established between two concepts, leading to the transfer
from one to the other. For example, people may be mulish, owlish, catty, and the like, if their
behaviour is similar to some characteristic of the animal concerned.

Metaphors often shift meanings from concrete to abstract, as is the case with E. grasp (verb),
whose original meaning was ‘ to take hold of something physically’, but now the verb also
means ‘to take hold of something mentally, to understand something’. With the latter,
metaphorical, sense the word appears in He grasped my argumentation.

98
This time , no need for a glossary, as ,I believe, I managed to explain all the technical terms
that appear in Lecture 15.

Language vs. speech

We say that people whom we don’t understand speak a language different from ours.
Language manifests itself in speech, it is an indispensible component of speech. It is that
which is permanent, social and abstract in speech. It corresponds to Saussure’s language and
Chomsky’s competence. (see lectures 2,3).
Speech is composed of other elements besides language. Unlike language, speech is concrete
and individual. It is communication by sound between two individuals (speaker and
addressee/receiver). Thus defined, speech involves several phases: one of them is the process
of speaking on the part of the participants in communication. Thus, speaking is the primary ,
fundamental phase of speech. The second phase is the receiver’s comprehension of the
speaker’s words. The third phase is the text . i.e., a product of speaking , which exists
independently of speaking, in the sense that texts may be preserved, they can be either
memorized or written. Finally, the fourth stage is language. Although texts are reproducible,
they cannot themselves serve as a direct basis for new texts. It is language, i.e., the system of
words and rules abstracted from existing texts, which constitutes the tool enabling speakers of
a language to produce new texts.

Language, preserved in memorized or written texts, can be abstracted and isolated. Since
language exists in the minds of speakers, it is supraindividual, it is a generalized system of
rules/norms of communication to which all speakers must conform if communication is to be
possible at all. Besides, and as has already been indicated, language is abstract in contrast to
concrete processes of speaking and comprehension, and in contrast to text, which is also a
concrete product, i.e., a physically existing entity.
Language is abstract in that neither its words nor grammatical rules or norms refer to specific
concrete phenomena , but rather refer to the general functions of relating these classes to one
another.
The total language system is composed of a number of sub-systems, of which the most basic
one is the phonological sub-system, the other sub-systems being morphological, syntactic ans
semantic ones.

Lectures: 16,17

LINGUISTIC SCHOOLS/THEORIES

I. American Structural School (behaviourists,Bloomfieldians and Neo-

Bloomfieldians/distributionalists)
II. Generative Transformational Theory

99
Lecture 16
American Structuralism

The behaviouristic theory in psychology and the immense interest in native American
languages had important implications for the theory of language. Behaviouristic psychology
claimed that all differences between people are environmentally determined and that all our
behaviour is reducible to our responses to external stimuli. The American psychologist,
J.B.Watson, one of the founders of behaviourism in America, defined stimulus as any object
in the general environment or any change in the psychological condition of the animal. The
term response is defined by the psychologist as “ that system of organic activity that we see
emphasized in any kind of activity, such as building a skyscraper, drawing plans, having
babies, writing books, and the like, which is the consequence of of some stimulus”
( J.B.Watson,, 1929 The Battle of Behaviorism , p. 19).
Behaviourists considered that since behaviour is subject to precise and experimental
analysis, it should be taken as the basis for all psychological investigations.
As native American languages had been a real challenge to descriptive linguistics, so
‘practical needs’, ranging from the teaching and learning foreign languages to the
requirements of communication engineering, stimulated the development of research in the
direction of synchronic description instead of diachronic studies.
The strength of descriptive synchronic interests can be partly explained by the necessity of
describing native American languages which had no written tradition which would
immediately pose the problem of the history of particular languages.

American behavioural structuralists rejected speculation (introspection) and mentalistic

concepts from language studies. They believed speculation (introspection) to be
scientifically useless, since it ruled out empirical control. So they said that the linguist
should deal with observables only. One of the most significant consequences of this approach
to language study was the elimination of meaning as a tool of linguistc analysis (this position
was in particular followed by the Neo-Bloomfieldians).
Leonard Bloomfield, who made the greatest contribution to American structuralism,
assumed that meaning was not a part of language structure, but it was outside of it, just as
sound substance is outside phonemics(phonology). The linguist defined the meaning of a
linguistic form as “ the situation in which the speaker utters it and the response it calls forth in
the hearer.” ( Bloomfield , Language). This can be visualized thus:

A B C

speaker’s situation speech hearer’s situation

A and C are ‘real’ in the sense that they are reduced to perceptible physiological events.
Thus, meaning is not in B, i.e. the utterance, but in part of the real world , A and C. The real
events of A and C are the proper domain of other sciences : e.g., physiology, sociology, etc..
In this way, as can be seen, meaning is limited to just external or referential meaning.

Some linguists, quite wrongfully, accuse Bloomfield of entirely eliminating meaning from
linguistic description. In fact, Bloomfield’s Language contains a whole chapter on meaning.
Meaning and form are equal partners in Bloomfield’s assumption that “ “in every speech-
community some utterances are alike in form and meaning”. For the writer, a linguistic form

100
is a correlation of units of form with units of meaning. Neo-Bloomfieldians (also referred to
as distributionalists) , that is, linguists like Z.S.Harris,Ch.Hockett, B.Bloch.GL.Trager.
A.A.Hill and C.C.Fries, left no place for meaning (semantic criteria) in their theory of
language. Partly responsible for this stance towards linguistic description was Bloomfield’s
statement to the effect that “ The study of language can be conducted without special
assumptions only so long as we pay no attention to the meaning of what is spoken
(Bloomfield, Language 1933: 75). Neo-Bloomfieldians made the sound features of an
utterance the foundation of their theory and practice. One of the eminent Neo-
Bloomfieldians, B.Bloch, completely eliminated semantic criteria, pointing out that “ In our
wording we shall avoid all semantic and psychological criteria. The implication is that such
criteria play no part , or at least need not play one, in the theoretical foundations of
phonemics. The basic assumptions that underlie phonemics can be stated without any
mention of mind or meaning.” (B.Bloch, ‘ A set of postulates for phonemic analysis’ in
Language 24, 1948: 35-36).
Accordingly, the Neo-Bloomfieldians claimed that the machinery needed to describe
linguistic data consisted of two steps: a/ the setting up of linguistic elements , b/ the
statement of the distribution of these elements relative to each other.
In order to achieve objectivism and preciseness in linguistic description , the linguist must
limit himself to the analysis of the purely physical aspect of language only. Bloomfield’s
attitude towards meaning led him to the antimentalistic position. Bloomfield refused the
mentalistic theory that ascribed the variability of human behaviour to the interference of
some non-physical factor, a spirit, or will or mind, which did not follow the cause-and-effect
sequences of the material world.

In linguistics, the influence of mentalistic theory, according to which mental states and
processes exist independently of their manifestations of behaviour, and can explain
behaviour, is most marked in the work of Noam Chomsky, especially in his concepts of
competence and innateness (the innateness hypothesis) , and in his general views of the
relationship between language and mind .
Chomsky stresses the role of language as a key means to the investigation of the human
mind. The view that linguistics can be profitably seen as a branch of cognitive psychology is
argued especially in Chomsky’s Language and Mind (1968).

The two basic methods used by the structuralists in linguistic description were substitution
and distribution (hence linguists who followed these methods came to be known as
distributionalists). One of the basic assumptions of distributionalism was that linguistic
description was considered as a set of operations (substitution and distribution), which , if
applied to the description of a text in a proper order, enables us to ‘discover’ the grammar
underlying the text (this is called grammar discovery procedure).

Segmentation consists in the analysis of utterances in terms of their basic units. At the
phonological level the segmentation of a portion of a text may be carried out at various
points. For example, if we compare the phoneme sequence bit with the sequence bid, we can
see that the segment bi- is shared by both the sequences, and thus it is separable from the rest
of the text. Now, comparing bi- and ba- (in bat ), we can further separate b and i as two
independent segments.
Distributionalists assumed that theoretically this kind of segmentation of linguistic elements
was feasible without recourse to the meanings of the elements.
Segmentation allows to single out basic units of the text, but it fails to indicate which of
these units are identical, i.e., which of them are variants of the same linguistic unit( e.g.,

101
which textual units are allophones of the same phoneme). To perform this task, the linguist
resorts to distributional analysis. Three types of distribution of textual elements were
distinguished (distribution= the sum of the contexts in which a linguistic unit can appear,e.g.,
the distribution of the Polish /b/ phoneme does not include the word-final position, but only
the initial and medial positions.).

a. Complementary distribution (dystrybucja uzupełniająca)

If two textual units have no contexts in common, they occur in complementary

distribution. For example, in English, the word-initial /p/ (as in pat) is aspirated [pæt] , but
if it appears before a consonant it is unaspirated (as in play). Text units that appear in
complementary distribution are variants of the same basic linguistic unit (here the
allophones of the phoneme /p/.

b/ contrastive distribution (dystrybucja kontrastowa)

If two elements share some of their environments, and if the use of either of them
involves change of meaning , they occur in contrastive distribution. For example,
English /t/ and /d/ in the pair: bat: bad , or -ed and -ing in: bombed:bombing.
Elements that occur in contrastive distribution belong to different units of language.

c/ free variation (wymiana swobodna)

Different units of the text are allowed to appear in identical environments with no
change of meaning. For example, in French /r/ may be front trilling (przedniojęzykowe
wibrujące) or uvular (języczkowe). Or take the English word economy , which can be
pronounced with the initial /i/ or /e/ vowels, with no change of meaning.
American distributionalists distinguished three levels in linguistic analysis: phonological,
morphological and syntactic. These levels are ordered hierarchically, with phonology as the
ultimate level:

SYNTAX

MORPHOLOGY

PHONOLOGY

102
Each linguistic level has its own basic units. The units of a higher level are composed of the
units of the level immediately below it. Thus, morphemes are made up of phonemes, words
are composed of morphemes, and syntactic constructions are composed of words. This
assumption determined the order of the stages distinguished in linguistic analysis. Any
linguistic analysis proceeded from phonology through morphology to syntax. Level mixing
was entirely out of the question, that is, the linguist was not allowed to make use of
morphological data in his analysis of the other two levels.
Distributionalists defined linguistic units as classes of the units of the text. To be exhaustive,
a description of a language structure must include setting up the basic units of each of the
three levels, establishing classes of these units, and finally formulating the laws according to
which the units may combine. All these, as was pointed out above, were achieved by means
of the segmentation of the text and the distributional analysis of the units arrived at as the
result of the segmentation operation.
Distributionalists pointed out that ,theoretically, it was possible to carry out the segmentation
operation at all the levels of language analysis without appealing to semantic criteria.
However, these attempts proved unsuccessful, since in, practice, linguists used distributional
criteria side by side semantic criteria, the latter having been introduced through “the back
door” as practical procedures or operational short-cuts.

Immediate Constituent Analysis (ICA) (analiza składników bezpośrednich)

ICA is another technique used by American structuralists to establish classes of linguistic

elements and ways of their combinability. The term itself was introduced by Leonard
Bloomfield in his Language, where we come across this statement:
“ Any English-speaking person who concerns himself with the matter, is sure to tell us that
the immediate constituents (ICs) (składniki bezpośrednie) of Poor John ran away are the
two forms: Poor John and ran away ; that each of these is, in turn, a complex form; that the
ICs of Poor John are Poor and John, and that the ICs of ran away are ran and away. “
Notice the parallelism between ICA and the traditional grammar procedure parsing (analiza
składniowa, rozbiór gramatyczny zdania), parsing sentences into subject and predicate , and
each of these into words and phrases and clauses of various types. Our example here might
be described as a simple sentence whose subject is a noun phrase (NP) made up of the noun
John modified by the adjective poor, and whose predicate is a verb phrase (VP) consisting of
the verb ran and its modifier away.
From this sort of analysis it is clear that sentences are not just linear sequences of elements,
but are made up of layers of ICs, each lower-level being part of a higher-level constituent.
This analysis of our example can be presented in the following two ways:

1. [ (Poor John) (ran away)] (this notation is called bracketting nawiasowanie)

2. By means of a tree diagram:

103
 a

b c

poor john ran away

This shows that a sentence is not just a string of morphemes or ultimate constituents (
poor.John.ran,away) in a fixed order. The ultimate constituents are grouped together in a
hierarchy of subdivisions. Take, for example, the following:

Little Tommy Tucker sings for his supper

1 2 3 4 5 6 7 8 9

This example sentence consists of 9 morphemes: little, Tom- , -y, Tucker, sing-, -s,
for,his,supper. Yet the important cut in the sentence is not into nine but into two : subject and
predicate. Immediate constituents are the direct components of a sentence, i.e., the forms
that directly go to make up that which is under analysis. Thus, IC is not any particular kind of
form, but a form standing in a particular relationship with another. There are one-constituent
forms (E.g., Fire!) and discontinuous constituents , which can be illustrated with the
following example:

Don’t you think it would be nice?

where the brace indicates that the discontinuous parts ( don’t and think ) of the predicate
function at the same level.

Three periods of development in the theory of constituent structure may be distinguished:

1. Bloomfield did little more than introduce the notion and explain it by means of
examples.

104
 2. His followers formulated the principles of constituent analysis in greater detail and
replaced Bloomfield’s somewhat vague reference to taking account of the meanings
with explicitly distributional criteria.
3. Finally, in the 1950s the theory of constituent structure has been formalized and
subjected to mathematical study by Noam Chomsky and other generative linguists
who gave considerable attention to the nature of rules required to generate sentences
with the appropriate constituent structure ( N.Chomsky, Syntactic Structures).

Bloomfield’s conception of language

Being under the influence of the behaviouristic theory, Bloomfield defined language as a
form of human behaviour describable in terms of stimuli and responses. His is thus a
mechanistic conception of language, which may be visualized thus:

S r… s R

where S(the stimulus) calls forth the individual’s linguistic, not practical, response , r , which
functions as a substitute stimulus for the other individual participating in the speech
situation(s), which induces him to act (R). Thus, in Bloomfield’s formula S, the stimulus,
induces the individual to react linguistically (r), which, for the hearer, acts as a substitute
stimulus (s), which makes him act practically. The practical actions (S) and( R) are non-
linguistic in character in opposition to( r) and (s), which are strictly linguistic.

Bloomfield illustrated his definition of language with a story about Jack and Jill. Jill sees an
apple on the tree and shows signs of being hungry. The feeling of being hungry makes Jill
ask Jack to get her the apple. Jill’s feeling hungry functions here as (S), i.e., as the stimulus,
which makes Jill act (r) , that is, ask Jack to get the apple for her. Jill’s request functions as as
(s) , the substitute stimulus, which prompts Jack to get the apple for Jill (R).
Today we know that this conception of language does not hold much water. Our use of
language is not necessarily dependent upon any stimuli. What’s more, the existence of certain
stimuli does not always make us use our language. For example, the sight of a lamp does not
always result in our uttering the word referring to that particular object. In short, language is
by far a more complex phenomenon than Bloomfield and some of his followers were willing
to admit.

Concluding, it may be pointed out that the chief contributions of American structuralists to the
general theory of language bore a methodological rather than strictly theoretical character, as
they aimed at working out clear and objective methods of analysis. They did not concern
themselves with explanations concerning linguistic data, but limited themselves to merely
stating facts. They were content with recording and organizing the data and devising
appropriate terminologies. That is to say, a linguistic analysis was regarded by them as a set
of procedures (grammar discovery procedure) independent of the structure of the languages
under consideration; the application of these procedures in a fixed order was to automatically
lead to the discovery of the grammar of a given language. Not without a reason, then, Jurij
Apresjan referred to American structuralism (also referred to as American descriptivism) as
“a prescription for a description”. No wonder, then, that American descriptivism is looked
upon by some linguists as rather a method of description than a linguistic theory. Since
American descriptivists were exclusively concerned with the classification(taxonomy) of

105
linguistic elements, this linguistic school is sometimes also referred to as taxonomic
linguistics.

The limitations of taxonomic approach in linguistic analysis has been emphasized by

generative linguists (like Noam Chomsky1957, 1965, and others ), who have criticized the
overreliance of structuralist (taxonomic) linguistics on procedures of segmentation and
classification.

The generative theory of language is the subject of our next lecture, lecture 17.

Glossary:

the innateness hypothesis (or nativist hypothesis) This hypothesis argues that that the
rapid and complex development of children’s ‘grammatical competence’ can be explained
only by the hypothesis that they are born with an innate knowledge of at least some of the
universal structural principles of human language.

106
Lecture 17
Generative Transformational Theory (also referred to as Generative Transformational
Grammar = GTG)

The Generative transformational theory developed in the USA in the second half of the 1950s
as a reaction to the American behaviouristic structuralism, which due to its narrowing down
of its research field (see Lecture 16), failed to provide acceptable answers to the vital
questions posed by many other language researchers of that time, including Noam Chomsky,
the founder of the generative transformational theory/grammar.
The criticisms leveled at American structuralist studies concerned, among others, their
inability and lack of desire to deal with the problem of meaning and the limitation of their
analysis to corpora of utterances ( which are always finite and not devoid of deficiencies).
The earliest version/model of GTG was presented by Noam Chomsky in his Syntactic
Structures , published in 1957, referred to in relevant linguistic literature as Syntactic
Structures Model/version of GTG.

Syntactic Structures Model of CTG

In Syntactic structures the definition of GTG is as follows: GTG is “a device which

generates all and only grammatical sentences of a language”.

There are a number of terms in the definition that need explanation. First of all, this approach
is firmly oriented towards the sentence. Second, the definition emphasizes that only
grammatical sentences of a language are to be explained in the grammar, i.e., sentences
which we intuitively accept as being well-formed (grammatical). Another point which arises
out of the definition is connected with the word all. It is an important principle that a
grammar should be able to account not only for the sentences occurring in a given corpus of
data (i.e. utterances), but should be able to account equally well for all possible grammatical
sentences in a language (of which the grammar is a description). It is assumed that the
number of sentences that any grammar has in principle to account for is infinite. But the rules
that constitute the grammar of a language must be finite, otherwise they would be
unlearnable. So what has to be done is to relate the finite system of rules to the infinite set of
possibilities by projecting the former on the latter. And this is done through the notion of
recursion (or recursiveness) , which has to do with grammatical rules which may be applied
more than once in generating a particular sentence. Recursion will be dealt with in greater
detail in one of the subsequent sections of this lecture.
Now, as the term ‘device’ is used in the definition, it is a totally abstract concept , referring to
the theoretical properties of a given system (system of rules here) , not its physical
manifestations.
In the Syntactic Structures model of GTG, Chomsky adopts the Immediate Constituent
Analysis (ICA) (for this kind of analysis see Lecture 16) to provide information about the
constituent structure of sentences. The difference with the earlier technique is that he does
not just draw a diagram to illustrate the structures involved, but formalizes the analytic
divisions into a system of ordered rules. This Chomsky refers to as phrase structure
grammar.
There are, however, parallels between Chomsky’s early version of GTG and Bloomfield’s
approach, which I will deal with after the presentation of Chomsky’s model.
Below I present the organization of GTG, as it is propounded in N.Chomsky’s Syntactic
Structures.

107
Chomsky’s 1957 model of TG consists of two basic components:

I. SYNTACTIC COMPONENT consisting of:

a/ Phrase structure rules (PS –rules)

and
b/ Transformational rules (T-rules)

II. MORPHOPHONEMIC COMPONENT (generating phonemic representations)

The phrase structure sub-component of the syntactic component, consisting solely of PS-rules,
specifies the hierarchical structure of a sentence, the linear sequence of its constituents , and
indirectly some types of syntactic relations. PS-rules (or rewrite rules- reguły przepisywania,
or formation/substitution rules- reguły formowania/substytucji) have the shape:
XAY XBY, where the single solid arrow means “ rewrite,substitute” (przepisz,zastąp),
hence this type of syntactic rules are also called rewrite/substitution rules.
By a ‘rule’ here we understand a kind of instruction, and X and Y are variables (zmienne),
which may stand for any linguistic unit including zero as a possible value.
For example, the PS-rule generating a sentence (S) may assume the following shape:

S NP + VP , where the sign “+ “ symbolizes concatenation (konkatenacja, operacja

łącząca symbole w rządki (strings)).

Concatenation must not be identified with ordinary mathematical addition. In mathematics, 2

and 3 always makes 5, irrespective of their actual arrangement (3 and 2 also makes 5).
However, the same is not true of PS-rules, where the ordering/arrangement of elements is of
crucial importance. Thus, for example, we cannot have a rule like S AUX +NP+VP, as
such a rule would generate a syntactic structure that would not be a (correct) English
sentence.
At this point the term generate needs explanation. As it is used in generative grammars, the
term must not be taken to mean ‘produce’, it should be taken as being synonymous with terms
like enumerate, define, and assign.
The PS-rules of the Syntactic Component are context-free rules reguły bezkontekstowe),
which means that this type of rules apply regardless of context ; they are all of the type
“Rewrite A as B” , no further conditions being specified.

The PS-part of the grammar might contain the following rules (this part of the grammar
contains also rules that rewrite lexical category symbols as concrete lexical items, e.g. rules
6,7, 9, and 11):
1. S → NP+VP
2. NP → NPsg
. NPpl
3. VP → Aux+ MV
4. NPsg → T+N+O
5. NPpl → T+N+s
6. T → the, an/a

The braces (Rules 2 and 11) indicate choice;e.g., NP rewritten as either NPsg or NPpl)
7. N → man, ball,door etc.
8. MVerb→ V+NP

108
 9. V → hit, take,open etc.
10. Aux → Tn(M) (have+en) (be+ ing) (Tn=Tense:Pres,Pas; M=modal verb)
11. Tn → Pres(ent)
Pas (Pas for past tense)
12. M → will, can, may, shall, must,etc.

PS-rules are indicated by a single solid arrow.

The result of the application of these rules are phrase markers(PMs) (on PMs see p.14) and
terminal/underlying strings (rządki terminalne, wyjściowe):

# the+man+O+Pres+may+have+en+open+the+door+O# - the string generated by the PS-

rules 1-12 (and see p.10)

The rules of the transformational subcomponent of the Syntactic component are

transformational rules (T-rules), indicated by a double solid arrow.

In contrast to PS-rules, T-rules apply under strictly specifiable conditions.

Every transformation consists of two parts:

I. Structural Index(SI) or Structural Analysis (SA)

II. Structural Change (SC)

SI specifies the conditions under which a given transformation(T) applies. SC specifies the
operations that must be carried out plus the effects of these operations. T’s operate on phrase
markers , not on single symbols (i.e. structural representations of phrases or sentences in
terms of labeled bracketing or tree diagram, assigned by the rules of the grammar). T’s apply
under strictly specifiable conditions, and they must be marked (for each language) as either
obligatory or optional (facultative). Besides, T’s apply in a fixed order. Consider, for
example, the Reflexive Transformational rule (Trefl)(which introduces the reflexive pronouns
into a sentence) and the Imperative Transformational rule (Timp)(which derives imperative
sentences). The application of Trefl must precede the application of Timp. Thus:
a simplified kernel string: #you+shave+you# , Trefl replaces the second occurrence of you
by an appropriate reflexive pronoun, here yourself, yielding the derived structure You shave
yourself.Trefl operates on structures involving two identical/coreferential NPs occurring in the
same simple clause (like in our example You shave you) Now, if we reversed the order of
application of these two Ts, the conditions for application of Trefl would not be met, since
Timp deletes the you which functions as subject. The deletion of the subject you would
render the application of Trefl impossible.
In Chomsky’s Syntactic Structures version of GTG, the relationships among sentences are
seen as transforms or transformations of each other ; e.g. the relationships among simple
active declarative sentences (e.g. He went to the cinema), negative sentences (e.g. He didn’t
go to the cinema) and questions (e.g. Did he go to the cinema? ) . Relationships like these are
accounted for by appropriate transformations. For Chomsky, transformations are rules of the
grammar that have the effect of changing one phrase marker into another through addition,
deletion and permutation of elements. In the model of GTG being presented here, Optional
T’s have the effect of changing the meaning of a sentence; these rules are then meaning –
changing rules. Thus, in this model of GTG, the differences between various types of

109
meaning are accounted for by means of optional T-rules (For example, the Passive
Transformational Rule is treated as optional). Obligatory T’s are meaning-preserving, i.e.
they do not affect the meaning of the sentences to which they have been applied (e.g. the
Affix , Flip- Flop Transformation .
Suppose our PS-rules (1-12) generated the following terminal string:

#the+man+O+Pres+may+have+en+open+the +door+O#

The Flip-Flop T rearranges the elements Aff(=affix) and V (here may) : V+Aff:
Pres+may may+Pres ( a morphophonemic rule converts this string, the output of the Flip-
Flop T into the phonemic shape may. The Affix/Flip-Flop is an obligatory transformation in
English.
Let us now consider the operation of the optional Passive Transformation( which converts
active sentences into corresponding passive sentences). This T-rule operates on phrase
markers underlying active sentences involving the use of a transitive verb and two non-
coreferential NPs (noun phrases), one of which functions as the subject and the other as the
object of the sentence. These conditions are specified in the SI of The passive T (T pass):
Tpass
SI: NP1+ Aux+Vtr+NP2 (NP1= NP2 )
SC: X1 - X2 –X3-X4 X4-X2+be+en+X3-by+X1

The hrase marker generated by the PS-rules, underlying the sentence The boy could see the
door:

NP1 VP

Aux MV

Tn M Vtr NP2
T N

T N

the boy pas can see the door

110
the terminal string : the boy pas can see the door

As can be seen, the phrase marker(PM) satisfies the conditions specified in the SI of Tpass.
Tpass applied to the above PM, derives the following structure:
11

NP2 VP

Aux MV

Tn M (be en) Vtr PP

the door
P N
Pas can P
see by the boy

The Tpass rearranges NP1 and NP2, and introduces into the underlying PM the grammatical
elements: be (the ‘passive’be ) , en (past participle morpheme) and by. It is thus a
transformation which involves two basic operations, viz., permutation (rearrangement) and
addition (introduction of new elements).
Transformations are not claimed to refer to any actual mental processes that take place in the
head of the speaker at the moment of speaking. They are rather formal operations that help us
relate two or more different PMs.
In the grammatical model under consideration here, the application of the PS-rules and
obligatory transformations, such as,e.g., the Affix/Flip-Flop transformation, yields the so-
called kernel sentences (zdania jądrowe). All other types of sentences (called derived
sentences- zdania derywowane) are the result of the operation of PS-rules and optional
transformations (e.g. the Tpass ).
Depending on whether a T-rule operates on one or more PMs, Chomsky distinguishes
between singulary (pojedyncze) and conjoining (generalized – zbiorcze ) Ts. In the SS
version of GTG, generalized Ts function as recursive devices allowing to generate an infinite
set of sentences.

Now, let us move to the MORPHOPHONEMIC COMPONENT of the early generative

grammar.

111
This component consists of rewrite rules which change strings of morphemes into strings of
phonemes (their phonetic representations). They operate on the output of transformational
rules( e.g., the string hate +Pas,the result of the operation of Taffix, is rewritten, by a
morphophonemic rule, as hated, open+ -en as opened , etc. ).
The relations between the particular components of the model of GTG here under
consideration may be visualized thus:

PS-rules

obligatory T-rules

kernel sentences

optional T-rules

derived sentences

Morphophonemic rules

Semantic representations

To conclude, the great influence of Bloomfield’s structuralism on Chomsky in his early model
of GTG is only too obvious. Thus, in Syntactic Structures:

a. There is no place for a semantic component (semantics was relegated from linguistic
study);
b. Syntax is assigned the central role in the grammar;
c. The introduction of the morphophonemic component (analogously to the phonological
components of structuralist grammars).

Like Bloomfield, Chomsky rejects meaning from formal description. On page 93 of his
Syntactic Structures he points out that ; “There is no aspect of linguistic study more subject to
confusion and more in need of clear and careful formulation than that which deals with the
points of connection between syntax and semantics.” Chomsky(op.cit.:100) believes that only
a purely formal basis can provide a firm and productive foundation for the construction of a
grammatical category.” According to Chomsky, then, syntax can and should be studied

112
 independently of semantics. The author’s dependence on the ICA (immediate constituent
analysis) has been already mentioned.

At this point I would like to refer you to what J.Searle has to say in this matter:

“ …I believe, the desire to keep syntax autonomous springs from a more profound
philosophical commitment: a man , for Chomsky, is essentially a syntactic animal. The
structure of his brain determines the structure of his syntax, and for this reason the study of
syntax is one of the keys to the study of the human mind.” (J.Searle, ‘Chomsky’s revolution in
linguistics’ in On Noam Chomsky: Critical Essays, p.15. For Chomsky, language is defined
by syntactical structure , not by the use of the structure in communication, and syntactical
structure is determined by innate properties of the human mind (ibid., 17).

And now just a few words about how grammaticalness is viewed in Chomsky’s Syntactic
Structures. Chomsky holds that a semantically based definition of grammaticalness is futile.
Grammaticalness is not attributed to a sentence by virtue of the fact that the sentence is
recognized to be meaningful. In Chomsky’s view, the study of grammatical structure is
quite independent of meaning.

Phrase Markers: their nature and function

PMs represent the constituent structure (another term for phrase structure = a formal
representation of the set of constituents that a linguistic expression contains) of sentences. A
PM is a graph consisting of nodes (węzły) connected by branches (gałęzie). The nodes at
the bottom of a PM are called terminal nodes ; the other nodes are non-terminal. In a
labeled PM (oznaczony znacznik frazowy) each node carries a label (etykietkę). Terminal
nodes are filled with lexical items.
A pair of nodes in a PM can be related by one of two types of relations: a/ dominance, b/
precedence (poprzedzanie).

When we say that one node , say, X dominates another node, say, Y, it means that X occurs
higher in the PM (also tree diagram) than Y, as,e.g., in :

NP1 Aux VP

V NP2
Det N

D N

this
The S nodema
(the highest node)will
dominates the other nodes in thewrite
tree; and the VP node
n and NP. D,N, V are terminal nodes. At a certain stage in the derivation
dominates V a letter
of the

113
sentence This man will write a book terminal nodes get replaced by lexical items, by so –
called Lexical Insertion Rule ( in the Aspects version of GTG).

Immediate domination. A node immediately(directly) dominates another node if it is the

next highest node in the tree and is connected to the other node by a single branch. Thus, e.g.,
in the above tree diagram (PM) the node VP directly dominates both V and NP , and S
directly dominates the left-hand NP, Aux. and VP.

The relation of precedence. One node precedes another if it occurs to the left of the other
node on the printed page. For example, in the tree diagram above the left-hand NP1 precedes
Aux,VP, V and NP. The NP immediately precedes Aux, but not VP, since Aux intervenes
between the NP and VP.

The relation of dominance and immediate dominance are used to define the terms
constituent and immediate constituent.
Thus:

1/ a set of nodes form a constituent iff (if and only if) they are exhaustively dominated by a
common node , i.e., if they all branch out of a single node (e.g. D and N form a constituent,
here an NP, since they both branch out of NP)
2/ X is an immediate constituent of Y iff X is immediately dominated by Y (e.g. NP1,
Aux,and VP are immediate constituents of S, since S immediately dominates each one of
them).

But consider the following PM:

VP

V PP

P NP

which represents the structure of, say, stay in the hotel . Here V and P do not form a
constituent although they are both dominated by VP ; however, they are not exhaustively
dominated by VP , since VP also dominates NP. But the P-NP sequence IS a constituent ,
since P and NP are exhaustively dominated by the PP node ; this PP node branches out only
into P and NP.

Recursion in Grammar

Every grammar which purports to be descriptively adequate (In Generative Grammar, the
term refers to an account of the native speaker’s linguistic competence, and not merely to an

114
account of a corpus of data - observational adequacy) must have built into it a mechanism
which allows to generate an infinite set of well-formed sentences. This mechanism helps the
grammarian to account for the fact that fluent speakers of a language have the ability to
produce and properly interpret an infinite set of sentences.

In the Syntactic Structures model of transformational grammar it is

The generalized(conjoining)transformational rules that function as the recursive device,
while in the Aspects (or Standard) version of GTG(to be presented below), this function is
performed by a small number of categorial PS-rules (i.e. rules which expand a category into
other categories (e.g. S is expanded as NP +Aux+VP by Rule 1 below) (in generative
grammar, a concept such as ‘sentence’, ‘noun’, ‘phrase’, ‘verb’, etc. These are shown by
category symbols such as S,NP,V,etc.).
Recursive rules are syntactic rules which may be reapplied more than once , and they apply
to one another’s output or to their own output. For example Rules 1-3 below consist a small
subset of rules which will generate an indefinite number of sentences of the following type:

Mary knows Peter/Mary knows that Peter knows Bill/ Mary knows that Peter knows that Bill
knows Frank, etc., etc.

1. S → NP+Aux+VP

NP
2. VP → V+
S

3. NP → (D) + N

The recursion consists in the fact that Rule2 generates as part of its output the category
symbol S, which in turn can serve as input to Rule1, whose output includes the symbol VP,
which can serve as input to Rule2; in this way, Rules 1 ad 2 form a recursive subsystem of
categorial PS-rules of the syntactic component of the grammar.
Rule 1 specifies that we can form a sentence (S) out of an NP, Aux, and VP; Rule 2 that we
can form a VP (verb phrase) out of a V(verb) followed by an NP (noun phrase), or a V
followed by an S, and Rule 3 states that we can form an NP out of a D(determiner) an N
(noun). ( Determiner is a word which is used with a noun and which limits the meaning of
the noun in some way; e.g., the following word categories can be used as determiners:
articles , as in ‘ The car has been stolen’, demonstratives (this,that,etc.), ‘ This/that car has
been stolen’, quantifiers (some, many,etc.), Some cars have been stolen’, numerals (first,
two,etc.) ‘ Three cars have been stolen’). This is formally represented by the following tree
diagram:

115
 S

NP Aux VP

V NP
If we take the other option in Rule 2 and expand VP into V followed by S, we generate the
following structure:

NP Aux VP

V S

Rules 1 and 2 may be reapplied indefinitely many times. To generate the structure Mary
knows (that)Peter knows (that) Bill knows Paul , Rule 2 must be applied three times:

NP Aux VP

V S
N
NP Aux VP

V S

N NP Aux VP

N V NP

Mary Pres know Peter Pres know Bill Pres know Paul

116
The inserted clauses (that Peter knows, that Bill knows Paul) are object nominal clauses
formally introduced by the complementizer that,which, being semantically empty, may be
omitted (complementizer is a clause-introducing word, such as that, as in the above example,
or if in I wonder if this is the right road, or for ,as in They’re keen for you to come ).

The syntactic structure realized by our example and formally represented by the above PM is
a case of S-Embedding /Insertion ; a clause is embedded (zanurzony) into a clause. In this
way, simple structures are expanded into more complex structures.

Another source of syntactic recursion is connected with relative clauses. Consider the
examples:

1. He wanted to call the nurse who he had met (restrictive relative clause)
2. My uncle, who you know very well, went abroad. (non-restrictive rel.clause)

The non-restrictive rel.clause in 2 is not part of the clause with My uncle as subject , such
relative clauses function as parenthetical comments (zdania wtrącone, parenteza).
Restrictive rel.clauses are adnominal complements , as they function as noun modifiers . For
example, in the model of GTG followed by Jacobs and Rosenbaum in their English
Transformational Grammar (1968), the categorial rule responsible for generating sentences
with restrictive rel.clauses is of the following shape:

NP NP + S

Relative clauses require that the NP in the embedded clause be identical with the NP to the
left of the relative clause. Thus the Rule: NP NP+S (recursive rule) and the rule
expanding S are able to generate relative clauses like the following: This is the cat that ate
the rat:

NP VP

V NP

NP S

NP VP

V NP

This is the cat the cat ate the rat

117
CONJUNCTION -another source of syntactic recursion. It is a process that operates on
sentences, phrases and words.

Sentence conjunction: S → S and S and S, … etc.

Or S → S and S

The rule will generate sentences such as John is a writer and Barbara is an actress and….

We cannot conjoin any two sentences or phrases or word categories, as is demonstrated by:

3. He can play the piano and she can swim

4. Did he go to the opera or did he go to the cinema?

which are quite normal, while 5-7 are anomalous:

5. * What’s your hobby and my car is broken down.

6. * You can go hunting and come with us.
7. *Come with us and do you like fish?

We can conjoin two or more declarative sentence, two or more imperative sentences, but not
distinct types of sentences, which would account for the grammatical and ungrammatical
sentences above. Only categories of the same type may be conjoined (NP with an NP,VP with
VP, etc. ). For example, the rule for NP conjunction will look as follows: NP NP- and –
NP, for PP,etc., PP PP-and- PP, etc.

Recursive processes occur also in morphology. By way of illustration , let us consider

endocentric compounds (in this kind of a compound, the modifier is subordinate to its head,
and the whole composite word is paraphraseable as “ a kind of – “; e.g. flower pot is a kind of
pot; it is a hyponym of pot, its meaning is included in that of pot.
Due to recursion, endocentric compounds can reach unnatural lengths, subject only to
performance ,i.e., extragrammatical limitations. In this connection, consider the examples:

vacuum cleaner(simple compound)

vacuum cleaner repair
vacuum cleaner repair shop
vacuum cleaner repair shop supplies
………………………..

Such expressions are only conceptually restricted; language provides expressions for available
concepts.
Also some derivational processes are recursive:

Make/re-make/re-remake, etc., theory/meta-theory/ meta-meta-theory

Such expressions would certainly be formed if there were a need felt by speakers for such
constructions. However, in English not all types of prefixation are recursive. For example,
negative prefixation does not seem to be recursive, which would take care of the impossible

118
formation *ununlikely. Generally speaking, combining prefixes is not a very productive
method of word-derivation, which is certainly true of English, and maybe some other
languages.

The Aspects/Standard Model of GTG

The Standard model of transformational grammar, as presented by Chomsky in his Aspects

of the Theory of Syntax (1965) (hence this model is also referred to as the Aspects model of
GTG), consists of three components:

I. Syntactic Component:
A. Base:
1. Categorial PS-rules
2. Lexicon

B. Transformational rules (Ts)

II. Semantic Component

III. Phonological Component

The Syntactic Component is the only generative component of the grammar , since it is here
that the meanings of the sentences generated by the grammar are determined (established).
The other two components are interpretive, which means that the rules they contain in no way
affect the meanings of sentences, they provide only semantic and phonological interpretations
to the sentences generated by the grammar.

A new, semantic component was introduced into the grammar to account for the fact that
speakers are able not only to produce syntactically well-formed sentences, but also to
comprehend them. Thus, an adequate grammar of a language must provide for both sentence-
generation as well as sentence-interpretation. In other words, and quoting the words of Katz
and Fodor (‘The structure of a semantic theory ‘ Language 39, 1963 )” a complete linguistic
description must contain an account of meaning”.

It will be recalled that in the Syntactic Structures model grammar was defined as a sentence-
generating device, i.e., a device which generates all and only grammatical sentences. In the
Standard model, grammar is viewed as a theory of linguistic competence whose two basic
aspects are, as indicated above, sentence-generation and sentence-interpretation.
The following diagram shows the organization of this model:

Base Component Transformational Component

S

Initial Semantic Component Phonological Component

element
Meaning Sound

119
Generally, for the followers of the Standard/Aspects model, grammar is a system of rules
relating meaning of each sentence it generates to the physical manifestation of the sentence in
the medium of sound. Thus:

Grammar (a system of grammatical rules)

MEANING SOUND

Deep Structure, surface structure

The notions of deep structure and surface structure are unknown to the early model of
GTG. In the Standard model, deep or underlying structures are the structures generated by
the categorial rules of the Base. The rules of the Base define grammatical categories,
grammatical functions, and determine an abstract underlying order of the sentences being
generated. They are, thus, abstract syntactic, not semantic, objects determining the meanings
of sentences. They should not be identified with sentence meanings. The meaning of each
sentence is derived from its deep structure by means of the rules of semantic interpretation ,
and the phonological interpretation of each sentence is derived from its surface structure by
means of phonological rules.
All sentences have both deep structure and surface structure,with the former determining
the meaning of a sentence and the latter relating to its actual phonetic representation. This
can be graphically presented thus:

DEEP STRUCTURE → MEANING → SURFACE STRUCTURE → SOUND

The two levels of linguistic structure, the deep structure(DS) level and the surface
structure(SS) level, are interrelated by a system of transformational rules (reguły
transformacyjne,przekształcania).

Base PS-rules → DSs → T-rules → SSs

Transformations operate on deep structures, or, better, on phrase markers.

The Lexicon, which is a part of the Base of the Syntactic Component, is a list of all lexical
items in a language together with a specification of their idiosyncratic syntactic, semantic and
morphological features. These lexical items are inserted in specified positions in
base/underlying structures by appropriate rules of so-called lexical insertion (wstawienie
leksykalne) (lexical insertion rules).
Categorial PS-rules generate derivations terminating with strings(=preterminal strings) that
consist of grammatical formatives (in GTG, formative = the minimal grammatical unit in a
language; e.g., in The drivers started the engines, the formatives are: the -s plural morpheme,
the – ed past tense morpheme. A terminal string is formed from a preterminal string by
insertion of lexical items in accordance with the following lexical rule:

If Q is a complex symbol of a preterminal stringand (D,C) is a lexical entry, where C is not

distinct from Q, then Q can be replaced by D.(C=complex symbol,a collection of syntactic

120
features ; see the glossary following the lecture) ( D=phonological distinctive feature matrix
“spelling “ of a lexical entry.
Lexical Insertion rules of the Base attach words under appropriate word-level category
nodes (węzły kategorialne), e.g., inserting man under the category N in the underlying
structure generated by the categorial rules of the Base:

NP VP

D
N

+Human
-Proper

+Male

Lexical Insertion: +Adult

man

In order to account for the co-occurrence relations within the sentence, lexical entries of the
Lexicon contain two types of features:
strict subcategorization features (reguły ścisłej subkategoryzacji) which specify whether or
not a given lexical item can occur in a given syntactic environment (specify the
complements that a given item takes): E.g.:

break: +[_NP]

elapse: -[_NP]

Violation of these features results in ungrammaticality (syntactic illformedness). The

ungrammaticality of the following is the result of violation of specific sub-categorization
features:

• John broke last year (and the correct John broke the vase)
• Two weeks elapsed ( and the correct Two weeks elapsed since his arrest)

The other type of features are selection(al) features (cechy selekcyjne) which indicate more
specific demands of the context:

121
surprise: + [__NP animate ]

For the verb surprise to be used correctly, it must be followed by an animate object NP (i.e.
NP referring to an animate entity ). This explains the ungrammatical * John surprised the
theory. Consider also the anomalous * The books elapsed (The verb elapse requires that its
subject NP be abstract, while in this example it is plus concrete (so, a selection features has
been violated here). Or consider * Sincerity may admire the boy, whose
deviance/unacceptability results from the illegitimate collocation of sincerity and admire
(admire occurs only with plus human subject NPs).

Violation of selection features is the source of sentence unacceptability

Basic rules: NP VP

D N Aux MV

Lexical
Insertion : V NP

the boy will help John

Diagrams like the above one are called tree diagrams or phrase markers , which are of two
types: base/underlying PMs and derived PMs. Underlying PMs represent deep structures
(DSs)of sentences and , as has already been stated, they are generated by the categorial PS-
rules of the Base; derived PMs are the products of T-rules.

In the Aspects model, T-rules have no impact on the meanings of sentences , they are
meaning-preserving (semantycznie obojętne). All meaning distinctions (e.g. the difference
between an active and a corresponding passive sentence) are handled at the level of deep
structure . If we assume that an active sentence and its corresponding passive sentence are
semantically distinct, then we must assign to each of them a distinct deep structures.
If we assume that the active sentence and its passive counterpart are synonymous, then we
must assign to them the same deep structures; they will be derived from the same deep
structure. The superficial difference between them will be accounted for by means of the
passive transformation.

Chomsky’s deep structures are assumed to have a universal character.

122
Some arguments for postulating two levels (deep structure and surface structure level)
of sentence organization:

To begin with, let us consider the following sentences:

1. Jack is easy to please.

2. Jack is eager to please.

Notice that on the surface (i.e. formally) (1) and (2) are very much alike, in (1) the superficial
subject Jack and the verb to please stand in the Object-Verb relation (Jack is the object of
pleasing), thus, underlying (1) is the structure ‘ It is easy to please Jack’, which is the actual
meaning of (1), while in (2), the noun Jack and the verb to please are in the Subject-Verb
relation, Jack appears here as the agent of the action of pleasing (Jack pleases someone).
Certain types of grammars ( a good example of such grammars are American
structuralist/descriptive grammars), being restricted to purely surface structure phenomena
(i.e. to what can be seen with the naked eye), are not able to account for semantic differences
like these implied in (1) and (2), i.e., for sentences that are on the surface (formally) similar
but semantically distinct. And let us now consider the following four sentences:

3. Mary sent a letter to Henry.

4. Mary sent Henry a letter.
5. Henry was sent a letter by Mary.
6. A letter was sent to Henry by Mary.

Sentences (3) –(6) are cognitively (basically) semantically equivalent sentences in that they
describe the same type of a real life situation , a sending situation involving three participants
: the Sender (Mary), the Object (a letter), and the Sendee (Henry). The fact that (3)-(6) are
synonymous sentences is accounted for by the Standard model of GTG by assigning to each
of them a common/identical deep structure (or Phrase Marker), which can be represented by
the following tree diagram:

NP Aux VP

N Tn V NP PP

D N P NP

Mary Pas send a letter to Henry

123
This diagram is a graphic representation of the deep structure of the sentences under
consideration here.
Sentences (3)-(6) would thus be derived from one common deep structure:
D
S

SS1 SS SS3 SS4

where DS= deep structure, SS=surface structure.

SS1-SS4 are related to the same DS (deep structure), as shown in the above diagram.

The syntactic phenomenon where one DS is realized by a number of SSs is called syntactic
split.
And now consider the sentence Flying planes can be dangerous. This sentence allows for two
interpretations: under one interpretation, it means ‘ It is dangerous to fly planes’, and other the
other possible interpretation, it means ‘ Planes that fly are dangerous ‘ (under this
interpretation the sequence flying planes constitutes a noun phrase (NP) functioning as the
subject of the sentence). So here we have a situation where two distinct DSs are realized by
one SS. This syntactic phenomenon is called syntactic merger.

In an American structuralist grammar , sentences like (3)-(6) would be assigned 4 distinct

descriptions , so the synonymy relation holding between them would be left unaccounted for.
As has been shown above, a generative transformational grammar accounts for this kind of
synonymy by assigning to the sentences an identical deep structure.

Evidence for the existence of two levels of syntactic organization from morphological
facts ( argument based on case marking facts )

In the system of English pronouns a morphological distinction is made between Nominative

and Accusative forms of personal and relative pronouns. The rule governing the use of
pronominal forms requires that a Nominative cas form occur before the verb form, while the
Accusative form follow the verb form, thus:

He admires Jane. Jane admires him.

Nom.Case Acc. Case

And consider now:

7. Who/*whom are you sure admires Jane?

8. Whom are you sure Jane admires?

The asterisk indicates an incorrect form.

124
Again, limiting ourselves to purely surface structure phenomena (and thus following in the
footsteps of American structuralists/distributionalists), we would fail to explain the facts
implied in (7) and (8), since in both sentences the pronominal forms (who, whom)precede the
verb form. So, according to the Case Marking rule, (8) would have to be recognized as
ungrammatical, which, you must admit, it is not.
To adequately account these facts , the transformational grammarian assumes that underlying
(7) and (8) are, respectively:
7a. You are sure who/*whom admires Jane.
8a. You are sure Jane admires whom.

Then, an appropriate T-rules must be applied to the structures (7a) and (8a) in order to arrive
at the correct : Who are you sure admires Jane? , Whom are you sure Jane admires?

Summing up, in the standard model of GTG, deep structure provides a natural explanation
for the fact that sound and meaning are not directly related; that the phonetic surface of
linguistic symbols incorporates no intrinsic semantic properties. The separation of underlying
and superficial syntax recognizes this fact by proposing that surface structure embodies all
information relevant to the phonetic representation of sentences, and deep structure contains
all information necessary to the semantic representation of sentences. The transformational
route between deep and surface structure may be considered the means whereby the
‘pairing’ of sounds and meanings is achieved.
There is no place for ambiguities in the deep structure. Ambiguity is treated as a feature
pertaining to the surface structure only. Recall our example Flying planes can be dangerous
given above. Like this surface structure, ambiguous is also the structure The chicken is too hot
to eat. This sentence is the surface structure realization of the two deep structures, which
may be roughly paraphrased thus:
a/ the chicken (the animal) is too hot to be able to eat;
b/ the chicken (the meat dish) is too hot for someone to eat it

In the early sixties, linguists came to realize that a grammar which consists of syntax and
phonology alone cannot be a full description of the native speaker’s linguistic competence.
A native speaker not only knows how to distinguish grammatical from ungrammatical
sentences and correct from incorrect pronunciations, he also knows what is a possible
interpretation for a sentence. The structures generated by the Syntactic Structures model of
grammar are purely formal and thus ‘meaningless’.
To take account of the speaker’s knowledge of the meanings of sentences , an adequate
analysis of a language should contain a semantic component (otherwise the structures
generated by the grammar must remain uninterpreted. In his Aspects, Chomsky holds that the
meaning of sentences should be subjected to the same kind of analysis as their syntactic
structure, and that semantics should be included as an integral part of the grammatical
analysis(description ) of languages. The grammar is now viewed as a device which relates
meanings (represented by means of deep structures or base PMs) to surface structures.
For Chomsky , transformations are rules of the grammar that have the effect of changing one
PM into another PM through an addition, deletion and permutation. With the exception of a
few optional/stylistic transformations, all other transformations are obligatory and meaning-
preserving (i.e., they have no impact on the meanings of sentences).

Glossary:

125
lexical entry, Lexicon A term used in grammatical description to refer to the accumulated
structural information concerning a lexical item as formally located in a lexicon/dictionary.
The lexicon has a special status in GTG; it refers to the component containing all the
information about the structural properties of the lexical items in a language, their
specification semantically, syntactically and phonologically. In the Standard/Aspects model,
these properties are formalized as features, and put in square brackets; e.g., word-class
assignments
include NOUN [+N], any item in the lexicon specified by [+N] can be attached to the node N
in a PM (phrase marker), etc.

acceptable, grammatical The judgement by the native speaker of a speech variety that a
certain linguistic expression is possible in their variety. The linguistic expression could be a
sentence, a word, a particular syntactic structure (e.g. phrase), etc. A linguistic expression
that is acceptable (dopuszczalny) to one group of speakers, or language variety, need not be
acceptable to another. For example, speakers of some varieties of English accept an
expression such as I want for him to come, but speakers of other varieties would not accept it
and use instead I want him to come. In generative transformational grammars acceptability is
distinguished from grammaticality. In these grammars, a given structure is grammatical as
long as it is generated by the rules of the grammar. A good example would be The teacher
who the man who the children saw pointed out is a cousin of Joan’s, which is grammatical
because it can be generated by the rules of the grammar, but some speakers would regard it as
rather unacceptable because of its complex structure which makes it difficult for a listener to
understand easily.

Recommended Literature:

Bach,E.1966. An Introduction to Transformational Grammars.

Chomsky,N. 1957. Syntactic Structures.
Chomsky, N. 1965. Aspects of the Theory of Syntax.
Fisiak, J. 1975. Wstęp do współczesnych teorii lingwistycznych.
Kakietek, P. 1989. ‘ American structuralism and Transformational-Generative Grammar’.
Linguistica Silesiana 10. pp. 31-40.
Lepschy, G.C. 1970. A Survey of Structural Linguistics.
Lyons, J.1970. Chomsky.

126