EN G L I SH G R A MM A R 1 –U n it 1 – H a nd o ut # 1

History of Grammatical S t u d y1
Copyright © Juan Luis Stamboni, 2000 / 2004

1. T r a d i t i o n a l G r a m m a r  5th century BC – early 20th century

1.1. Antiquity
Early grammatical study appears to have gone hand in hand with efforts to understand archaic writings.2 Thus,
grammar was originally tied to societies with long-standing written traditions. The earliest extant grammar is that of
the Sanskrit language of ancient India, compiled during the 5th century B.C. by the Indian grammarian Panini, who
described and analysed the sounds and words of Sanskrit, showing how words are formed and what parts of
words carry meaning. Ultimately, the grammars of Panini and other Hindu scholars helped in the interpretation of
Hindu religious literature written in Sanskrit.
The Arabs are believed to have begun the grammatical study of their language before medieval times; and the
Jews completed a Hebrew lexicon during the 10th century A.D.. They also produced a study of the language of the
Old Testament.
The study of grammar in Europe began with the ancient Greeks, who engaged in philosophical speculation
about languages and described language structure. They introduced, for example, the concept of grammatical
categories. In the 3rd century B.C., the Greek grammarian Dionysius Thrax wrote the Art of Grammar, upon which
many later Greek, Latin, and other European grammars were based. The Greek grammatical tradition was passed
on to the Romans, who translated the Greek names for the parts of speech and grammatical endings into Latin;
many of these terms are still found in modern grammars. The Roman grammarians Donatus (5 th century A.D.) and
Priscian (7th century A.D.) wrote Latin grammars based upon Thrax’s. The Greeks and Romans did not, however,
compare languages with one another.
Greece:
 Plato made the distinction between NOUN and VERB
 Aristotle founded LOGIC
 The Stoics studied CASE and made the distinction between FORM and MEANING
 The Alexandrians attached more importance to writing than to speech and set up standards of correctness,
among which, the classical fallacy. They regarded literature as the most worthy expression of language.
 Dionysius Thrax, who recognised 8 parts of speech, and Apollonius Dyscolus wrote the first (Greek)
grammars of the Western World.
Rome:
 Varro differentiated DERIVATION from INFLECTION, and studied TIME and ASPECT.
 In the 1st century B.C., Julius Caesar wrote a linguistic treatise On Analogy, which is now lost.
 Donatus and Priscian wrote Latin grammars, which were used throughout the Middle Ages.
1.2. Middle Ages and Renaissance
With the spread of Christianity during the early Middle Ages, and the translation of the Bible into the languages
of the new Christians, written literatures began to develop among previously nonliterate peoples. The medieval
European scholars generally knew, in addition to their own languages and Latin, the languages of their nearest
neighbours. This access to several languages set scholars to thinking about how languages might be related to
one another and, therefore, compared. Eventually, the development of printing in the 15th century further
encouraged the comparison of languages.
Medieval Europe:
Earlier Middle Ages: translations of the Bible. Latin as the vehicle of learning
Later Middle Ages: speculative grammar. The question of UNIVERSALS. Invention of the printing press
Renaissance:
15th–16th centuries: consolidation of European vernacular languages. The Classical languages as literary
languages. Circulation of printed material. Language surveys.
1.3. Comparative Linguistics
Medieval and Renaissance learning led, on the one hand, to 16 th and 17th century surveys of all the then-known
languages in an attempt to determine which language might be the oldest. Following this line of study, in the early
18th century the German philosopher Gottfried Wilhelm Leibniz suggested that most European, Asian, and
Egyptian languages might have a common ancestor and derive from the same original language, which was

1
In the reading comprehension exercise, students are asked to account for the meaning of the underlined words. Underlined book titles are, of
course, excepted.
2
The study of written records, the establishment of their authenticity and correctness, and the determination of their meaning constitute an
approach to language which is known as Philology (from Greek philos, “loving”; logos, “speech”). In the 19th century, the term philology often
referred more narrowly to the study of linguistics. In the 20th century, philology has been used in literature, historical linguistics, and other areas
of study in order to reconstruct the texts of imperfect or mutilated manuscripts and inscriptions.
1
referred to as Indo-European. Leibniz's postulation, which was later proved to be partly correct and partly incorrect,
stimulated the beginnings of the field of comparative philology or comparative linguistics.
Toward the end of the 18th century, Sir William Jones, an English scholar, observed that Sanskrit bore
similarities to Greek and Latin, and proposed that the three languages might have developed from a common
source. Language scholars in the early 19th century took this hypothesis much further. Jacob Grimm, a German
philologist, and Rasmus Christian Rask, a Danish philologist, noted that when the sounds of one language
corresponded in a regular pattern to similar sounds in related words in another language, the correspondences
were consistent.3
By the late 19th century much analysis had been done on sound correspondences. A group of European
language scholars known as the neogrammarians put forth the theory that sound correspondences between
related languages were regular and that any exceptions to these phonetic rules could develop only from
borrowings from another language. This method of comparing related words in different languages to discover the
existence of regular sound changes became known as the comparative method. It served as a tool in establishing
language families, that is, groups of related languages. Using the comparative method, linguists posited an Indo-
European family composed of numerous subfamilies, or branches. It is to this family that English, one of the
Germanic languages, belongs.
Comparative grammar became the dominant approach to linguistic science in the 19th century, when scholars
developed systematic analyses of parts of speech, mostly built on the earlier analyses of Sanskrit. The early
Sanskrit grammar of Panini was a valuable guide in the compilation of grammars of the languages of Europe,
Egypt, and Asia. This writing of grammars of related languages, using Panini's work as a guide, is known as Indo-
European grammar, a method of comparing and relating the forms of speech in numerous languages.
The description of regular sound correspondences also made it possible to compare different forms of a given
language as spoken in different regions and by different groups of people. This field, known as dialectology, may
focus on differences in sounds, grammatical construction, vocabulary, or all three.
On the other hand, the revival of classical learning in the Renaissance laid the foundation for a misguided
attempt by grammarians to fit all languages into the structure of Greek and Latin. Considering Latin the most
logical of languages and/or admitting that the structure of Latin embodies universally valid canons of logic, 17th and
18th century grammarians tried to discover logical rules of syntax and usage which were almost invariably
prescriptive. The methodology used in the ancient Greece and Rome was adopted by English grammarians such
as Robert Lowth, Joseph Priestly, George Campbell and Lindley Murray during the 18 th and 19th centuries. Their
rules for English, based on Latin models, dictated precisely the usage to be followed by all English speakers and
writers.
17th century: Empiricism versus Rationalism. The [French] Grammar of Port Royal (1660)
18th century: German Romanticism. Humboldt: language and thought. Samuel Johnson: English dictionaries.
English grammars.
19th century: Comparative Philology: historical comparative linguistics. English Grammars: Robert Lowth; Joseph
Priestley. The school tradition: Nesfield (England). Lindley Murray (parsing); Clark & Alonzo Reed, Brainerd Kellogg
(analysis) (USA). The scholarly tradition: Poutsma; Kruisinga; Jespersen; Zandvoort (Europe). Curme, Long (USA)

2. S t r u c t u r a l G r a m m a r  1920’s – 1970’s
The 19th century background, within which the early 20th century scholars grew up, was characterised by the
overlaping of all the above-mentioned concerns. Not until the early 20th century did grammarians begin to describe
languages on purely linguistic terms. During the 19th century, linguists who were also university scholars or
professors of English in Germany and in the Netherlands shaped a grammar for specialists in the English
language. This 'European scholarly tradition' culminated in three great grammars produced in the first part of the
20th century by H. Poutsma (1909), E. Kruisinga (1909) and the Danish linguist Otto Jespersen, who published A
Modern English Grammar, in four parts, between 1909 and 1931, and The Philosophy of Grammar (1924).
Poutsma and Kruisinga stand in a central position in Traditional Grammar; their works are conservative, very
detailed and thoroughly documented. But Jespersen made considerable innovations. He based his treatment of
syntax almost completely on meaning and classified words into ranks on the basis of their syntactic hierarchy.
Jespersen thought grammar should be studied by examining living speech rather than by analysing written
documents, but he wanted to ascertain what principles are common to the grammars of all languages, both at the
present time (synchronic approach) and throughout history (diachronic approach). Jespersen, as one of the
earliest descriptive linguists, developed precise and rigorous methods to describe the formal structural units in the
spoken aspect of language.
Significantly, the key figure in the change from 19th to 20th century attitudes was the Swiss linguist Ferdinand
de Saussure (1857-1913), who first made himself known to scholarship through an important contribution to Indo-

3
Grimm’s Law (Jacob Grimm, 1822) describes the pattern of two stages of sound changes, known as the German consonant shift and the High
German consonant shift. Grimm's law is important in that it demonstrates the development from the old Germanic languages of more recent
languages such as English, Dutch, and Low German. It also shows that changes in a language and in groups of languages come about
gradually and not as a result of random word changes. Grimm based his research on the 1818 treatise of Rasmus C. Rask, a discussion of the
origin of Old Norse. Grimm's work was augmented by the explanations of stress shifts formulated by the Danish philologist Karl Adolf Verner.
2
European comparative linguistics.4 Though he published little himself, Saussure’s lectures on linguistics in the
early 20th century so impressed his pupils in Geneva that in 1916 they published his posthumous Cours de
linguistique générale as far as they could reconstruct it from their own lecture notes and such materials as survived
in Saussure’s handwriting.5 Saussure drew on a restricted range of languages, mostly the familiar languages of
Europe; but his influence on 20th century linguistics, which he could be said to have inaugurated, is unsurpassed.
The publication of the Cours has been likened to a ‘Copernican revolution’ in the subject. Saussure’s teaching
came in an age when the comparative and historical study of language had reached a resting place in the works of
the neogrammarians (Poutsma, Kruisinga, Jespersen). Saussure’s ideas revolved around three main concerns:
1. He formalized and made explicit the two fundamental and indispensable dimensions of linguistic study:
synchronic, in which languages are treated as self-contained systems of communication at a particular time
in history; and diachronic, in which the changes undergone by languages in the course of time are treated
historically. Earlier linguists had either taken this dichotomy for granted or simply ignored it. Saussure’s main
achievement was to establish the distinction between these two axes or dimensions of linguistics. The
synchronic or descriptive axis had methods and principles of its own which were different from those
belonging to the diachronic or historical dimension.
2. He drew the distinction between langue and parole. Saussure understood langue as an abstract system of
language structure, and parole as the actual instances of speech, the data which can be directly analised by
linguists. Parole is immediately accessible, however, the linguists’ proper object of study is the langue, the
knowledge that all members of a speech community share about what is grammatically acceptable in their
language. Langue is therefore understood as the lexicon, grammar, and phonology implanted in each
individual by his upbringing in society. It is on the basis of langue that a native speaker of a given language
speaks and understands his or her language.
3. He rejected the traditional concept of language (i.e. grammar) as a mere nomenclature or an aggregate of
self-sufficient entities and proposed that any particular langue (French, Spanish, Arabic, Chinese or Swahili)
be described synchronically as a system of interrelated elements at various levels: lexical or morphological
(word), grammatical or syntactic (sentence), phonological (sound), and semantic (meaning).

Phonetics is the study of all speech sounds and the ways in which they are produced.
Phonology is the study and identification of the meaningful sounds of a language.
Morphology is concerned with the units (morphemes) that carry meaning in a language.
Syntax refers to the relations among word elements in a sentence.

Thus, from Saussure’s point of view, linguistic terms must be defined relatively to each other, not absolutely.
That is, a langue is a structure made up of elements identified and characterised by their place in the whole
system of language. The elements of language should not be identified by their mere substantial composition
(meaning). The interrelations which serve to identify these elements take place at the two dimensions of
synchronic linguistic structure, i.e. syntagmatic and paradigmatic:
a) The syntagmatic dimension represents the linear succession of elements typical of human utterances (syntax).
b) The paradigmatic or associative dimension represents systems of contrastive elements or categories.
Hjelmslev, one of Saussure’s followers, defines form as the interrelations of elements, and believes that the
analysis of semantics and grammar must be independent of extra-linguistic existential criteria (i.e. substance), and
that phonological analysis must be independent of phonetic criteria. Hjelmslev believes that relations between
elements, not the elements themselves, are the object of a science, and only by keeping this strictly to the fore
can the Saussurean ideal of an autonomous linguistics, not dependent on any other discipline, be realised.
Saussure’s statement of the structural approach to language underlies virtually the whole of modern linguistics,
and justifies his claim on the independence of linguistics as a subject of study in its own right. Although the term
‘structuralism’ may have different interpretations, few linguists would nowadays disclaim structural thinking in their work.
 In Europe, STRUCTURE meant a whole made up of parts which were related to each other in specific ways. 6
Structure was the opposite of an aggregate or a pile. Language is a system of signs and each sign has a
value in the system. The linguistic sign is made up of both expression and content (form and substance, i.e.
meaning). Saussure’s dichotomies were:
syn c hr oni c dia ch ro ni c
ling ui s tic s ling ui s tic s
la ng u e pa r ol e
sig nifi é sig nifi a n t
su bs ta n ce v ers u for m
s

4
De Saussure. Mémoire sur le système primitif des voyelles dans les langues indo-européenes, Leipzig, 1879.
5
De Saussure, 1949, preface to first edition; P. Godel, Les sources manuscrites du Cours de linguistique générale de F. de Saussure, Paris, 1957.
6
Other early 20th century linguistic schools in Europe were Copenhagen (Glossematics) and Prague (Functional Linguistics). British Structural
Grammar: Halliday (Scale and Category Theory): SCALES: rank, exponence, delicacy / CATEGORIES: unit, structure, class and system
3
 syn t a g ma ti c pa rad ig ma ti c
re la tio ns re la tio ns
de s cri p ti v e pr e s cri p ti ve
gra m ma r gra m ma r
civ ili se d l a n gua ge s pr i mi tiv e
lang ua ge s
Towards 1911, the German-American anthropologist Franz Boas and his colleague, the American anthropologist
and linguist Edward Sapir, attempted a description of hitherto unrecorded native North American languages; 7
which were non-Indo-European languages with no written records. In their work, they saw grammar as a
description of how human speech is organised in a language, postulating that a descriptive grammar should
describe the relationships of speech elements in words and sentences. They challenged traditional methods and
techniques of linguistic description which were based on written texts, and formulated methods for identifying the
distinctive, meaningful sounds of a language (i.e. phonemes), and the minimal units of sound combination that
carry meaning, which they called morphemes. They gathered data from native speakers and analysed the
components of their speech, organising the data into separate hierarchical levels of language: phonology,
morphology, and syntax. Boas's work formed the basis of various types of American descriptive grammar study.
Given impetus by the fresh perspective of Boas, the approach to grammar known as descriptive linguistics became
dominant in the U.S.A. during the first half of the 20th century.
Building on the work of these descriptive linguists, the American linguist Leonard Bloomfield proposed a
behaviouristic analysis of language, avoiding semantic considerations as much as possible. He insisted that
linguistics should concentrate its attention on formal analysis by means of objectively describable operations and
concepts. This system of language analysis is known as American structuralism. A structuralist grammar,
according to Bloomfield, should describe the relationships which underlie all instances of speech in a particular
language. Therefore, while descriptive grammar described the elements of transcribed speech, American
structuralism concentrated on the utterances of speech. The two fundamental units of description of American
structuralism were the phoneme, the minimal distinctive phonetic unit of linguistic utterances, and the morpheme,
the minimal unit of grammatical structure. Sentence structure was set out in terms of immediate constituent
analysis, in which the morphemes were linked together in trees, representing constructions of ascending size and
complexity. Bloomfield also emphasised techniques to be used to discover the sounds and grammatical structure
of unrecorded languages. Because of the status of his book Language8 as a students’ textbook and his deliberate
concentration on methodology, Bloomfield’s interpretation of linguistics predominated in the attitude and outlook of
most American linguists during the 1930’s and 1940’s.
 In America, STRUCTURE meant form, the surface level of language. Meaning was left out of account. 9
Language FORM was characterized by five grammatical signals:
 i nf l ec t i on
 der iv at i o n
 s truc tu re w or ds
 wor d or d er
 s tres s a n d i nt o na t io n
In concentrating on linguistic form, rather than meaning, American linguistics neglected the study of syntax and
emphasized and developed the study of phonology and morphology.

S T R U C T U R AL G R AM M AR S : In Europe: Ferdinand de Saussure (1916) Cours de Linguistique Générale

In America: Edward Sapir (1929) Language
Leonard Bloomfield (1933) Language

Both schools of Structural Grammar were a reaction against Comparative Philology, which often treated
language in a piecemeal fashion, studying the history of words in isolation. But since the mid 20th century, Modern
Comparative Linguistics has been concerned with establishing language families in areas such as North and
South America, New Guinea, and Africa. In such regions it has only recently become possible to gather the vast
amounts of data that are needed to reconstruct the former stages of current languages and thereby to trace family
relationships. Modern comparative linguistics is also involved in a search for linguistic universals. Interest in the
typological characteristics of languages has been renewed, and linguists are now comparing languages with
regard to their syntactic structures and grammatical categories (such as gender languages versus nongender
languages, and languages with subjects versus languages with topics). The American linguist Joseph Greenberg
has shown that languages that share a basic word order (such as subject-verb-object or object-verb-subject or
object-subject-verb) also share other features of structure. Linguistics, from the early 20 th century onwards,
encompasses studies from both the diachronic and synchronic points of view, whereas 19th century language
studies usually focused on a diachronic approach.

7
Boas, Franz et al. (1911) Handbook of American Indian Languages
8
Bloomfield, Leonard (1933) Language
9
Other important structuralist linguists in the USA are Hockett, Zelig Harris, Pike, and Lamb.
4
 The field of psycholinguistics merges overlapping interests from the studies of both psychology and
linguistics. It is concerned with cognitive, perceptual and acquisition processes, e.g. speech perception, oral and
written language comprehension, neurolinguistics (i.e. language and the brain), language acquisition and aphasia.
Sociolinguistics is the study of how language functions in society. Sociolinguistic studies attempt to describe
the human ability to use the rules of speech appropriately in different situations. Sociolinguists believe that the
mechanism of language change can be understood by studying the social forces that motivate using different
forms in different circumstances.

3. F u n c t i o n a l G r a m m a r  1930’s / 1970’s – nowadays10

Another form of linguistics flourished in Prague in the 1930’s. The Prague school looked outside the structure
of particular languages and attempted to explain the relation between what is spoken and the context. The Prague
school linguists stressed the function of elements within a language and emphasised that the description of a
language must include how messages are put across. In the area of phonology, their concept of distinctive
features, which divides sounds into their component articulatory and acoustic elements, has been highly regarded
and adopted by other schools of language analysis.
As from the 1970’s, various Functionalist trends have developed both in Britain and the USA. Following the
European functional tradition of the Prague school and borrowing some abstract principles from Hjelmslev and
Firth’s system-structure theory, M.A.K. Halliday put forth his Systemic Functional Linguistics in the mid-1980’s.
This model is presented as a theory of language as choice in the form of a system network. A particular language,
or any part of it, is seen as a resource for making meaning by choosing. The grammatical structures are merely the
output of this network; they just realise the sets of features that are chosen for specific communicative purposes.
The use of the term choice does not necessarily imply a conscious process of selection by the speaker. What a
functional analysis aims to uncover is the reasons why the speaker produces a particular wording rather than any
other in a particular context. Thus, Functional grammar sets out to investigate what the range of relevant choices
is, both in the kinds of meanings that can be expressed and in the kinds of wordings that can be used to express
these meanings; and to match these two sets of choices. Therefore, meaning choices are identified by: (1)
looking outwards at the context, thus determining what contextual factors make one set of meanings more
appropriate or likely to be expressed than another; (2) identifying the linguistic options (i.e. the lexical and
structural possibilities that the language system offers for use); and (3) exploring the meanings that each option
expresses. Thus, Halliday’s approach sets out to relate in a systematic way functions to wordings, that is to say,
the connection between certain meanings and certain structures, so as to reduce the analysis to a minimal number
of relevant functions.
Functions are grouped into three basic kinds of meanings or metafunctions: Experiential, Interpersonal and
Textual. The first is related to the way human beings represent the world and organise it in their minds, their
experience of the world includes description of events (and states) and of the entities involved in them. The second
refers to the manner in which human beings use language to interact and relate with each other, establishing and
maintaining relations, influencing other people’s behaviour, expressing their viewpoints and eliciting and changing
others’. The Textual Metafunction deals with the organisation of the message in ways which indicate how they fit in
with the other messages around them and with the wider context in which the communicative events take place.
Thus, Functionalism refuses to recognize strict theoretical or methodological boundaries between the syntax
and the realms of semantics, pragmatics, and discourse. According to Scott DeLancey, human language is not
simply a device for presenting and pointing to interesting objects and events in the world, it is, rather, a set of tools
for communicating our experience, and its structure is fundamentally informed by the structure of our experience
and our cultural models of experience.
Halliday argues that the basic opposition in grammars of the second half of the 20th century is between those
that are syntagmatic in orientation (i.e. Structural and Generative “formal” grammars, with their root in logic), and
those that are primarily paradigmatic (i.e. Functional grammars, with their root in rhetoric and ethnography). The
former take regular syntactic relations as the foundation of language, having structure as the organising concept
and using special devices to relate one structure to another, whereas the latter consider that describing a language
consists in relating it to everything else, i.e. the cultural and situational contexts and communicative purposes and
choices.

4. G e n e r a t i v e G r a m m a r  1957 – nowadays
By the mid-20th century, the American linguist Noam Chomsky, who had studied structural linguistics, was
seeking a way to analyse the syntax of English in a structural grammar. This effort led him to see grammar as a
theory of language structure rather than a description of actual sentences. His idea of grammar is that it is a
device for producing and understanding sentences in any and all languages. He further proposed that linguistics
should go beyond describing the structure of languages and it should provide an explanation of how sentences
in any language are interpreted and understood.
Chomsky believed that this process could be accounted for by a universal human grammar (that is, a model

10
This section is based upon Halliday, M.A.K. (1985). An Introduction to Functional Grammar. Edward Arnold / Hodder & Stoughton, London;
Thompson, G. (1996). Introducing Functional Grammar. Bristol: Arnold; and DeLancey, Scott (2001) “Lectures on Functional Syntax”. University
of Oregon, Department of Linguistics. Eugene, OR 97403-1290 – U.S.A.
5
or theory of linguistic knowledge or competence). Chomsky’s universalist theories are related to the ideas of
those 18th and early 19th century grammarians who urged that grammar be considered a part of logic: the key to
analysing thought. For instance, 19th century universal grammarians such as the British philosopher John Stuart
Mill, writing as late as 1867, believed rules of grammar to be language forms that correspond to universal thought
forms. [See the section ”Knowledge of Language” for a more detailed introduction to Chomskyan linguistics]

T RADITIONAL AND S TRUCTURAL APPROACHES TO G RAMMAR

Traditional grammar is notional and prescriptive, whereas structural grammar is formal and descriptive. A
grammar is said to be notional when it bases its study of language on the assumption that there exist universal
grammatical categories which are, in turn, based on meaning. And it is said to be prescriptive when it ‘prescribes’
how the members of a speech community ought to speak and write their language. Traditional grammar is
represented by the basically Aristotelian orientation toward the nature of language, as exemplified in the work of
the ancient Greeks and Romans, the speculative work of the medievals, and the prescriptive approach of 18 th
century grammarians. Structural grammar is said to be formal11 because it claims to describe the structure of
language on its own terms, basing its study on the formal elements of language (i.e. phonological, morphological
and syntactic) rather than on their meaning. And it is said to be descriptive12 because it ‘describes’ how the
members of a speech community actually speak and write. Structural grammar is empirical, it makes exactness a
methodological requirement and insits that its statements be verifiable and refutable.
The cardinal differences between the classifications and rules of traditional and structural grammars are those
that distinguish humanistic studies, on the one hand, from empirical sciences, on the other:
 Whereas the categories of traditional grammar are to a large extent determined by their characteristic “modes
of signifying”, structural grammar devotes little or no attention to semantics13 on account of the fact that there
is doubt whether meaning can be studied as objectively and as rigorously as phonology and grammar.
Furthermore, the problem of meaning seems to be of greater concern to philosophy, logic and psychology, and
perhaps also to other disciplines such as anthropology and sociology, than it is to linguistics.
 By defining classes and assigning rules for language based on meaning, traditional grammar proceeds
subjectively, explaining how the different elements of language are related to the individual who uses the
language. Structural linguistics, on the other hand, sees language as a structure of interrelated components,
and, in defining classes and assigning rules in language, it proceeds objectively, showing how the different
elements of language are related to one another.
 Traditional grammar often confuses levels of analysis, whereas structural grammars distinguish various levels
of analysis, for example morphological and syntactic.
 While traditional grammars appear to assign the reason why certain language features occur, and how they
must behave, formal structural grammars approach language empirically, by merely stating the observable
facts of language, without attempting an explanation for nonlinguistic correspondences.
 Because the Greek investigation of language started with logic, traditional grammar has unthinkingly taken the
declarative sentence as ‘basic’. The parts of speech are defined, therefore, according to their function in that
sentence type alone. Other sentence types are ‘explained’ as deviant forms of the declarative sentence, from
which they are said to be derived by a conscious psychological process. From a structural point of view, this
method is inadequate for the full description of any language, ancient or modern. Structural linguistics has
studied all utterances on the same terms. The declarative sentence is taken as basic on the norm of frequency
and descriptive convenience and not as a consequence of a priori psychological assumptions.
 Traditional grammar has a cultural history which links it to a fundamentally Aristotelian psychological theory,
with the semantic doctrines of the medievals and the individual preferences of 18 th century authors, but
structural linguistics has prescinded from disputed psychological, logical or metaphysical systems.
 Structuralists are accused by traditional grammarians of dehumanising language by equating it with any formal
signal system; and of so stressing the differences between languages as to obscure the preponderance of
similarities, which are more impressive on comparison.
 Traditional grammarians are accused by structuralists of giving pseudo-explanations of carefully selected
sentences, artificially manufactured to fit their rules; of applying a mixture of semantic, morphological and
syntactic criteria in no fixed order; and of producing endless grammars without explaining their methodological
presuppositions.
 Structuralists claim that traditional grammar does not adequately distinguish lexical, morphological and
syntactic meaning; and its hazy distinction of morphology and syntax results in an inadequate notion of
“modification”, and in questionable criteria for the classification of the “parts of speech”.
11
In the field of linguistics, the word ‘formal’ can be used in various senses. Within the Saussurean theoretical framework, form represents the
analizable elements of langue and is opposed to substance, which refers to the dimension of meaning. Within the Chomskyan theoretical
framework, formal (i.e. definable in theoretical terms) is opposed to informal (i.e. resulting from intuitive or uncontrolled observation)
12
This particular interpretation of the adjective ‘descriptive’ corresponds more closely to American Structuralism than to Saussurean or
European Structuralism.
13
Though grammarians have been interested in the meaning of words since the earliest times, the term semantics is of relatively recent origin,
being coined in the late 19th century from a Greek verb meaning “to signify”.
6
 Although traditional grammar is the most widespread and best-understood method of discussing Indo-
European languages within the Western world, the categories it has developed are inadequate for the study of
non-European languages; and all those usages of language which are not amenable to its rules are
ungrammatical in some imputable sense.
Since the mid-1950’s many linguists have been dissatisfied with the conception of linguistics as a mere
taxonomic or descriptive discipline. It has been proposed that a satisfactory general linguistic theory must provide
the basis for an explanation as well as an accurate description of language. Such a proposal would require the
inclusion of meaning considerations in the descriptive and explanatory mechanism. The most recent developments
in linguistics have largely returned to the traditional goals of grammatical work, but with the rigour of the formal,
structural methods developed by many linguists over a period of many years.

K N OW L E D G E O F L A N G U A G E
With regard to the acquisition of knowledge it was widely held in previous centuries that the mind is not "so
much to be filled therewith from without, like a vessel, as to be kindled and awaked" 14; “The growth of knowledge
rather resembles the growth of fruit, however external causes may in some degree cooperate, it is the internal
vigour and virtue of the tree that must ripen the juices to their just maturity” 15 These statements can be taken to
mean that knowledge grows because of genetic instructions under the triggering and shaping effects of
environmental factors. Applied to language, this essentially Platonistic conception would suggest that knowledge of
a particular language grows and matures along a course that is in part intrinsically determined. These ideas
were generally regarded with disapproval in the mainstream of linguistic research by the late 19 th century and on
up to the 1950’s. In part, this negative attitude developed under the impact of empiricism and later behaviourist
and operationalist doctrines.
Before the 1960’s, the study of child language was dominated mainly by the ‘behaviourist’ approach to
language and learning16. According to this approach, language is not a mental phenomenon: it is behaviour. Like
other forms of human behaviour, it is learnt by a process of habit-formation. Language acquisition was understood,
then, as a case of overlearning; as a habit system, which was determined by available evidence. Production and
interpretation of new forms was taken to be a matter of analogy.
The main components of this habit formation process are:
1. The child imitates the sounds and patterns which he hears around him.
2. People recognise the child’s attempts as being similar to the adult models and reinforce (reward) the
sounds, by approval or some other desirable reaction.
3. In order to obtain more of these rewards, the child repeats the sounds and patterns, so that these
become habits.
4. In this way the child’s verbal behaviour is conditioned (or ‘shaped’) until the habits coincide with the
adult models.
The habit-formation process is essentially the same as when a pigeon’s behaviour is shaped, so that it pecks at
the correct discs in order to obtain food. Within this framework, the child’s own utterances were not seen as
possessing a system in their own right. They were seen as a faulty version of adult speech. The ‘mistakes’ were
simply the result of imperfect learning.
The behaviourist view of first language ‘learning’ was strongly challenged from the 1960’s onwards, especially
under the influence of Noam Chomsky’s linguistic theories and cognitive psychology. These are some of the
arguments which have convinced most researchers of the inadequacies of the behaviourist approach:
 Underlying the actual behaviour that we observe (performance), there is knowledge of a complex
system of rules (competence), which enables speakers to create and understand an infinite number of
sentences, most of which they have never encountered before. This creativity would not be possible if
we had to rely on individual bits of learnt behaviour.
 What children acquire, then, is an abstract knowledge of rules (competence), which they are not
exposed to, and which is often reflected very indirectly in the actual surface structure of the speech 17.
Information about deep relationships could not be acquired simply by observing and imitating verbal
behaviour.
 Although children are exposed to different actual speech, they arrive at the same underlying rules as
other children in their community. The evidence also suggests that they pass through similar
sequences in acquiring these rules.

14
R. Cudworth (1617-1688)
15
James Harris (1709-1780)
16
B.F. Skinner (1957) Verbal Behaviour
17
For example, the surface structure of John is easy to please looks identical to that of John is eager to please, yet their deep structure is
completely different: in the first, it is a question of other people pleasing John, whereas in the second, it is John himself who wants to do the
pleasing.
7
  The process of language acquisition is perhaps more complex than any learning task that most
human beings undertake. Yet it occurs at a very early age and with exceptional speed. Again, this
cannot be explained by habit-formation alone.
From the point of view of Generative linguistics, then, humans do not acquire language by applying generalised
learning mechanisms with greater efficiency or scope than other organisms. There is, instead, a property of the
mind that differentiates humans from animals; a distinct language faculty with specific structure and properties.
How does Generative theory reach this conclusion? Through mere observation of the fact that, although young
children have brief, personal and limited contacts with the world which surrounds them, they have nevertheless
internalised all the basic structures of their language by the age of between three and half and five. How can they
know so much on the basis of such limited evidence? The problem that arises is one of poverty of stimulus.
Many examples have been given to illustrate the problem of poverty of evidence. A familiar example is the
structure-dependence of rules, the fact that, without instruction or direct evidence, children use computationally
complex structure-dependent rules rather than computationally simple rules 18. Furthermore, children do not make
errors about the interpretation of sentences past a certain stage of development.
The difference of perception concerning the problem of language acquisition –overlearning versus poverty of
evidence– reflects very clearly the effect of the shift of focus which inaugurated the study of Generative Grammar.

Language as a system of knowledge

Cognitive systems result from the interaction of experience and the organism’s method of dealing with it (e.g.
analytic mechanisms, maturation, and cognitive growth). Although our cognitive systems reflect our experience, a
careful specification of the properties of these systems on the one hand, and of the experience that led to their
formation, on the other, shows that the two are separated by a considerable gap. In order to account for the
specificity and the richness of the cognitive systems that arise in the individual on the basis of the limited
information available, we should determine the innate endowment that serves to bridge the gap between
experience and the knowledge attained.
The study of human language is particularly interesting in this regard. In the first place, it is a true species
property and one central to human thought and understanding. Furthermore, it is possible for us to characterise
the system of knowledge attained -knowledge of English, of Spanish, etc.- and to determine the evidence that was
available to the child who gained this knowledge. We are thus in good position to ascertain the nature of the
biological endowment that constitutes the human language faculty, the innate component of the mind that yields
knowledge of language when presented with linguistic experience, that converts experience to a system of
knowledge. If we can discover something about the principles that enter into the construction of this particular
cognitive system, if we can specify the principles of the language faculty, we can progress toward a solution to the
problem of poverty of stimulus.
Generative Grammar is concerned with those aspects of form and meaning that are determined by the
language faculty, which is understood to be a particular component of the human mind. The nature of this faculty is
the subject matter of a general theory of linguistic structure that aims to discover the framework of principles and
elements common to all human languages: this theory is called UG (universal grammar). UG may be regarded as
a characterisation of the genetically determined language faculty. One may think of this faculty as a language
acquisition device, an innate component of the human mind that yields a particular language through interaction
with presented experience, a device that converts experience into a system of knowledge attained: knowledge of
one or another language.

How is knowledge of language acquired? The answer to this question is, as we have seen, given by a
specification of the biologically endowed language faculty of human beings, or UG, along with an account of the
ways in which the principles of UG interact with experience to yield a particular language. UG is a theory of the
initial state of the language faculty, prior to any linguistic experience.

What constitutes knowledge of language? The answer to this question is given by a generative grammar,
which is a theory concerned with the state of the mind of a person who knows a particular language. Traditional
and structuralist grammars did not deal with this question, the former because of its implicit reliance on the
unanalysed intelligence of the speaker/hearer, and the latter because of its narrowness in scope.
The concerns of traditional and generative grammar are, in a certain sense, complementary: a good
traditional or pedagogical grammar provides a full list of paradigms, examples of regular and irregular
constructions, exceptions, and observations at various levels of detail and generality about the form and meaning
of expressions, but it does not examine the question of how the speaker of a language uses such information to
attain the knowledge that is used to form and interpret new expressions, or the question of the nature and
elements of this knowledge. One could describe such a grammar as a structured and organised version of the
data presented to a child learning a language, with some general commentary and often insightful observations.

18
For further development of this topic, see Andrew Radford (1997). Syntactic Theory And The Structure Of English. A Minimalist Approach.
CUP. Chapter 1, section 1.6
8
 Structuralist theories did concern themselves with analytic procedures for deriving aspects of grammar from
data, as in the theories of Harris, Bloch, Trubetzkoy and others, but primarily in the areas of phonology and
morphology. The procedures suggested were inadequate and, in any event, could not provide an answer to the
question of how knowledge of language is acquired; or determine what a comprehensive account of the knowledge
of the speaker/hearer would involve.
Generative Grammar, in contrast, is concerned primarily with the knowledge of language attained by the
speaker/hearer, and the principles and procedures put to use in order to attain this knowledge.

In the mid-1950’s, a research programme was inaugurated to investigate the adequacy of certain proposals
which had been advanced in order to provide an answer to the question “What constitutes knowledge of
language?”. This program was one of the strands that led to the development of the cognitive sciences, sharing
with other approaches the belief that certain aspects of the mind can be usually construed, on the model of
computational systems, of rules that form and modify representations, and that are put to use in interpretation
and action. This research programme has since been running its course, along a number of different paths. The
study of generative grammar represented a significant shift of focus in the approach to problems of language. The
shift of focus was from behaviour or the products of behaviour to states of the mind that enter into behaviour. If one
chooses to focus attention on this latter topic, the central concern becomes knowledge of language: its nature,
origins, and use.