Language & Corpus
Chap 03 (pp.75-112): W. Teubert. from
Halliday, et al. (2004). Lexicology and
corpus linguistics: An introduction.
London: Continuum.
The pre-modern linguists
 Pre-modern linguistics in Europe was not concerned with
the productivity of language. From the Middle Ages well
into the nineteenth century, linguists were philologists,
which was, at the time, more or less synonymous with
 Their research was on 'dead' languages: Latin, Greek
and Hebrew. Their aim was not to produce new texts in
these languages; they wanted to understand the texts
we had inherited from ancient times.
 The rules they came up with were rules to help us make
sense of the sentences. The rules were meant to
describe what we were confronted with in the texts; they
were not designed to empower us to become competent
speakers of ancient Greek.
The Philologist
 The philologists may not have had a
scientific method. And yet we inherited
from them the academic editions of
classical and oriental texts we are still
using today, together with comprehensive
dictionaries, or rather glossaries, citing
each noteworthy occurrence of any word
embedded in its contexts and still
providing an irreplaceable aid in
understanding these texts.
The emergence of modern
 In the nineteenth century we find a novel interest in
languages, different from traditional philology. It was the
century when the enlightenment finally bore fruit and
nature began to be understood in terms of the laws of
 All the academic glamour now rested with the sciences.
 The hermeneutical approach to language was not
interested in immutable, eternal laws or rules. But that did
not necessarily mean that there weren't any. The first
domain of this new 'scientific' approach to language was
the study of relationships among languages. That became
the starting point of modern linguistics.
The modern linguistics
 To the linguists of the nineteenth century who
found patterns of changes in the Indo-European
languages, it seemed that the phonetic changes
these words underwent in the course of history
were governed by laws.
 The new linguists wanted to discover the laws of
phonetic change. It had become possible to
describe language in terms of rules; rules that
did not involve any decision-making on the part
of the linguists, rules that produced results that
had to be objectively correct once you accepted
the premises.
The modern linguistics
 The modern linguists who succeeded the
philologists saw themselves as scientists.
 They were not interested in the mental
processes linked to language but wanted to
investigate the structure of language, based on
analyses of texts, in order to understand the
language system behind it. They wanted to
describe a system of rules and means that
existed independently of its individual speakers
and its historical development.
What is grammar?
Winter’s and Sinclair’s view
 Utterance: a sequence of signs which represent the
content, which stand in place of the content.
The utterance 'a Martian scientist visits Earthlings' can
be said to represent an image, a photograph or a mental
image which is two- or even three-dimensional.
But the utterance is always a one dimensional string of
John Sinclair, citing the grammarian E. O. Winter that
'grammar is needed because you cannot say everything
at the same time'.
This is certainly the reason why all natural languages
need grammar, and perhaps also why these various
grammars can be described if not in identical, then in
very similar terms.
What’s grammar?
Chomsky’s view
 All humans share the same language
faculty, an innate faculty that regulates the
ways signs are to be organised so that
they become utterances. This is what is
called grammar.
 In Chomsky's view, the innate language
faculty shapes the grammar.
Language Universals
(A Chomskyan view)
 Chomsky sees the language organ as an
apparatus that gives limited options.
 Adjectives can precede the noun they
modify, or they can follow it. But all
languages have adjectives and nouns and
several other parts of speech.
Language Universals
(A Chomskyan view)
 Languages resemble each other because
their phonology, syntax, and morphology
can be described in the same - or at least
similar – terms.
 For mainstream linguists, languages are
all more or less the same, which may
follow different rules, but are made up of
the same entities and share many
Grammar and competence
(A Chomskyan view)
 Chomsky's revolution in linguistics is about the
generative power of rules, which do not describe
what is there but what is possible.
 This focus on the generative aspect of language
has changed the agenda of linguistics. The role
of linguistics is no longer to interpret what we
find in existing texts, but to describe the
language faculty, or, in abstract terms, the
competence of a speaker to produce new
grammatical sentences.
Lexical entry/item
 Each lexical entry consists of the word, an indication of the part of
speech it belongs to, and the syntactic and semantic properties it
 boy
 a noun, countable (hence there is a plural boys),
 it fits into a slot (i.e. a terminal element of the syntactic structure of a
given sentence), which asks for a word denoting a human being (such
as the subject and the object position of the verb love).
 'Big
love intelligent girls.
 adjective + noun + (transitive) verb + adjective + noun.
 Each noun and the verb exemplifies a slot into which we can insert a
suitable lexical element taken from the lexicon.
Similar but not the same
'This is a fake diamond.‘ Is fake a noun or an
 cf: big, which is gradable (big, bigger, biggest);
but not fake, faker, fakest
 ‘The house is big.”
 ‘The diamond is fake.'
 ‘The diamond is a fake.'
 fake as a noun that can be used as an adjective?
 fake as an adjective that can be used as a noun?
Grammar vs. Vocabulary
 We seem to have much less difficulty in learning
the syntax of a foreign language than its
 It is not always too difficult to construe
grammatically correct sentences in a second
language. But unless we are acquainted with it
very thoroughly, we will make mistakes when we
try to put our thoughts into words or to translate
a text from our native language.
Grammar vs. Vocabulary
 The meaning of words, as compared with the regularities
of phonetic change and sentence construction, is
generally fuzzy and vague, not only when we compare
one language with another, but also from a monolingual
 Words, single words, may be the ideal core units when it
comes to describing the working of grammar. But they
are much less the appropriate core units when we are
interested in meaning. Single words are commonly
ambiguous. Dictionaries capture this ambiguity by
assigning two or more word senses to a word.
Word meaning in context
 Usually we have no problem understanding what
a sentence means. This is because we do not
look at the words in isolation, but embedded in a
context. We read a word together with the words
to its left and to its right; we have no problem in
knowing what a word means.
 Ambiguity is a consequence of our misguided
belief that the single word is the unit of meaning.
Units of meaning are, by definition,
unambiguous; they have only one meaning.
While some words are units of meaning, many
are not.
 This enquiry into meaning makes the case that
meaning is an aspect of language and cannot be
found outside of it.
 It is entirely within the confines of the discourse
that we can find the answer to what a unit of
meaning means, be it a single word or, more
commonly, a collocation, i.e. the co-occurrence
of two or more words.
 A unit of meaning is a word (often called the
node or keyword) plus all those words within its
textual context that are needed to disambiguate
this word, to make it monosemous.
Compare fire and enemy fire
 Fire is therefore not a unit of meaning.
 The collocation enemy fire, meaning 'the
shooting of projectiles from weapons by
the enemy in an armed conflict‘, is (under
normal circumstances) monosemous, and
is therefore a unit of meaning.

Language & Corpus Linguistics