Linguistics For Beginners - Basic Concepts

Tariq Rahman
te d
d i
e e
b
to
e rsoOxford University Press is a department of the University of Oxford.
e V It furthers the University’s objective of excellence in research, scholarship,
i tl and education by publishing worldwide in
T Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Turkey Ukraine Vietnam
Oxford is a registered trademark of Oxford University Press
in the UK and in certain other countries
Published in Pakistan by Oxford University Press
© Oxford University Press 2010
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
First published 2010
All rights reserved. No part of this publication may be reproduced, translated,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press.
Enquiries concerning reproduction should be sent to
Oxford University Press at the address below.
You must not circulate this work in any other form and you
must impose this same condition on any acquirer.
ISBN 978-0-19-547925-6
For sale in Pakistan and not for export therefrom.
Typeset in Adobe Garamond Pro

Printed in Pakistan by
________________, Karachi.
Published by
Ameena Saiyid, Oxford University Press
No. 38, Sector 15, Korangi Industrial Area, PO Box 8214,
Karachi-74900, Pakistan.
Contents
page
Preface vii
1. Introduction 1
2. A Brief History of Linguistics 8
3. The Study of Sounds 24
4. The Study of Sound Patterns 44
5. How Words are Made 53
6. The Study of Meaning 62
7. Introduction to Chomsky’s Theories 72
8. Transformational Generative Grammar 80
9. More about Grammar 101
10. Government and Binding Theory 111
11. Universal Grammar 117
12. Anthropological Linguistics 127
13. Sociolinguistics 140
14. Language Politics 162
15. Language Death 171
16. Educational Linguistics 176
17. Writing in South Asia 182
18. Notes for the Reader 198
19. Bibliography and Further Reading 199
Preface
T his book is written to introduce South Asian students to

linguistics. It is written in simple English, and the explanation
is so clear and detailed that no previous knowledge of the subject
is necessary to understand the basic concepts of linguistics. Indeed,
the aim of this book is to teach linguistics to beginners who have
never studied it and who do not have a teacher. Thus the book
explains concepts in the kind of language which a teacher would
use when explaining concepts to undergraduates or high school
students. It is meant for the bachelors’ level but may be used in
the ordinary two-year masters’ level in some South Asian
universities which comes after a two-year BA and which does not
have specialized courses to begin with. As this is not a specialized
book on linguistics, there is a little of most of the concepts which
South Asian universities require.
Of course, in order to make the beginners familiar with the most
important aspects of a course in linguistics, there is much more of
such core subjects as sounds, the production of sounds in the
mouth, the description of sounds, the way words and sentences are
constructed, and how there are rules governing these constructions.
All these are areas of what may be called microlinguistics. But
there are also important aspects of how language is used in society,
how we create meaning, and how we express reality through
language. This is called macrolinguistics and is often more
interesting for people studying sociology, anthropology, computer
sciences, politics, and history etc. The book does introduce some
of these areas but, due to their importance, it gives more space to
concepts without knowing which nobody will call you a linguist.
Linguistics—the scientific study of language—is taught in most
of the major universities of the world. It helps us understand what
human languages are and how they produce meaning. As languages
are used in societies there is much that knowledge of languages and
viii Preface
how they are used in society can tell us about society itself. We can
even understand some of the ways the brain itself functions by
understanding how it produces the sentences of a language. In
short, linguistics contributes towards the study of culture,
behaviour, and thinking. People who work in artificial intelligence
and robotics also use linguistics to provide them insights into how
language is processed in the brain.
There are excellent departments of linguistics in India and a
number of linguists of world stature. There are also departments
of linguistics in Nepal as well as Sri Lanka. In Pakistan, courses of
linguistics were introduced from 1987 onwards by the author and
now there are degree courses in linguistics in several universities,
though most of the courses they offer owe their origin to English
Language Teaching (ELT). However, even so there are many
students who study linguistics at the masters’ level in Pakistan.
Unfortunately, there are very few books of an introductory nature
for them in Pakistan.
The book which I used when I was teaching myself linguistics in
1987 (incidentally I had a PhD in English literature and was full
professor then, but I had lost interest in literature for both research
and teaching) was An Introduction to Language (Fromkin and
Rodman 1974). In 1990, while doing a short course on language
and the mind in the University of Cambridge, I met a really gifted
lecturer, Dr Jean Aitchison, whose books were so clear that I would
recommend them for anybody who is learning linguistics. A few
months later in 1991, at the University of Sana’a, I met a colleague
from India whom our family called Deedi but part of whose name
Dr Chanda stays in my memory. She was trying to write a simple
introductory book for students of English Language Teaching
(ELT). I do not know if that book ever got written but the idea of
a book which could teach beginners linguistics even without the
mediation of teachers stayed in my mind. A few years later, while
writing Language and Politics in Pakistan (OUP 1996; available in
India in the Orient Blackswan edition of 2007), I wrote a small
book called An Introduction to Linguistics (Lahore: Vanguard, 1987)
which is the basis of this revised and updated edition. This,
however, is a new book as almost everything has been changed in
Preface ix
it, though it does overlap with the earlier work in some places. I
thank Najam Sethi, an old friend and publisher, who published An
Introduction to Linguistics. I also thank Ms Ketaki Bose who
encouraged me to rewrite it so as to produce a new edition for
Indian students. Since I wanted the book to be available to
Pakistani and South Asian students as well as others, I wanted it
to be published in Pakistan in addition to India. I thank Oxford
University Press for making this wish come true. I am especially
grateful to Ms Ameena Sayid, Managing Director of Oxford
University Press in Pakistan, for encouraging me to submit the
manuscript. I also thank Mr Samuel Ray, Senior Editor, Oxford
University Press for having taken pains to make the book see the
light of the day. But for the work he and his colleagues put in the
book would not have been before the readers today.
I realize there are many such introductory books, including some
recent additions (Matthew 2003). However, I flatter myself that
none of them cover as many aspects of linguistics and in as simple
language as this one.
I have kept South Asian names and cultural references with a view
to engaging the attention of students of all SAARC countries.
However, being familiar more with the culture and publications of
Pakistan and India, I apologize for there being more references to
these two countries. I hope students in Pakistan and India, as well
as other South Asian countries, will find this book useful. As the
book is written in clear English and the jargon of linguistics is
introduced to the reader with adequate explanation, it can also be
read by the lay person. In fact, one of the reasons for writing it is
to introduce the ordinary reader to linguistics.
This book has been divided into the following chapters. Chapter
1 defines what linguistics is, and Chapter 2 traces out the history
of its development. Chapter 3 and 4 are about the sounds which
combine together to make up a language (phonetics). Chapter 5
explains the way words are constructed (morphology). Chapter 6
is about the meaning of words or semantics. This is followed by
five major chapters—7, 8, 9, 10, and 11—on the theories of Noam
Chomsky, i.e. transformational generative grammar and the
x Preface
theories born as it developed, such as the government binding

theory and the minimalist approach. As these theories require a
higher standard of knowledge than is assumed at this introductory
level, the treatment is simplified and the more complex
developments of recent years have not been touched upon.
Chapters 12 and 13 are on the social and cultural aspects of
language which will be useful for students of anthropology and
sociology, especially those interested in gender and language death.
Chapter 14 is on language politics. It will be useful for those who
are interested in the relationship between language, identity, and
politics, and especially for those who write dictionaries and
grammars of languages or make language policies. Chapter 15, on
language death, is meant to create an understanding of one of the
consequences of modernization and globalization, i.e. that these
processes entail the disappearance of aspects of our culture and
identity. Our sense of belonging is under an assault the like of
which has probably never been experienced before. Chapter 16 is
on educational linguistics, i.e. the teaching of languages and the
political, educational, psychological, and economic perspective in
which this teaching takes place. Chapter 17 is on writing in South
Asia but there are some introductory sections on the history of
writing in the world also. These chapters connect the study of
language with the study of globalization, politics, language policy,
public policy, and economics. In short, such awareness may bring
about a deeper understanding of what human knowledge is, and
how it can help the human race to live in peace and harmony. In
chapter 18, while bidding good bye to the reader, a few hints are
given as to what paths may be followed if one pursues linguistics
as a field of study. The bibliography gives details of books which
have been used in preparing this brief introductory guide. However,
it also serves as a list of books for further reading. They will serve
to give details of aspects of linguistics which have merely been
touched upon in the text.
Tariq Rahman PhD
Islamabad
2010
1
Introduction
What is linguistics?
D o you know how many languages exist in South Asia? If not,

go to the Ethnologue which is published by the Summer
Institute of Linguistics. The latest online edition (2005) gives the
following information.
India 438 (out of which 13 are extinct)
Nepal 126 (out of which 3 are extinct)
Pakistan 72
Bhutan 24
Bangladesh 39
Maldives 01
Source: Gordon 2005. Online version retrieved on 1 August 2008.
The world has about 6,912 languages. But it is not only numbers
which tell us how important language—just the fact that we
humans can convey meaning and write them down in symbols—
really is. Without language we would have no conversation, no
songs, no stories, no jokes, and no civilization as we understand
the term. Indeed, perhaps we would be like intelligent monkeys
but would we be human? One doubts that we would. Language
distinguishes us from animals.
Just observe how we behave. People talk more than they do
anything else. They generally talk when there is someone to hear
them but sometimes even where there is nobody to do so. In
literate societies words are also written and read. Almost everybody
reads signs on the road, in shops, and on the TV, but there are also
people who spend most of their waking hours writing or reading.
2 Linguistics for Beginners
All this is language in action. Such an important thing—human

language—is an object of interest in itself. Those who are interested
in its study are called linguists and the scientific study of language
is called linguistics.
What do we study in linguistics?
All human languages are made up of sounds arranged in such ways

that they communicate meaning or express feelings or attitude. The
sounds come first. In fact, when writing had not been invented,
the sounds of languages existed. The study of sounds—i.e. how
they are produced and described—is called phonetics. The study
of how these sounds are arranged in a language, i.e. the sound-
pattern of a language, is called phonology.
These sounds combine to make small units of meaning which are
called morphemes. For example, boy, girl, -ish, and -s are
morphemes. They express meaning on their own or in combination
with other morphemes. Thus we can make the new words boyish,
girlish, girls, and boys with these morphemes. The study of how
words are made (i.e. the shapes of words) is called morphology.
When words combine together in sentences, they do so according
to certain rules. These rules are the rules of syntax. If we want to
know how sentences are formed in a language, we study its
syntax.
Now that we have got our sentences, we should know what they
mean. The study of meaning is called semantics.
Hence, these are the four main aspects of the study of a language:
(1) phonetics and phonology (2) morphology (3) syntax (4)
semantics.
Some people also add another study—that of language in real life
situations. This is called pragmatics. Pragmatics deals with the way
people actually use words in daily conversation. For instance, if
someone asks me what time it is and I say the news have not started
yet, that person will know what time it is only if he or she knows
Introduction 3
that the news start at such and such time. The study of implications,
the use of titles, words of politeness and rudeness etc. come under
pragmatics.
Let us arrange these aspects—or levels as they are called (Crystal
1988: 82) in an order:
LANGUAGE
STRUCTURE USE
PRAGMATICS
SOUND GRAMMAR MEANING
SEMANTICS
PHONETICS PHONOLOGY
MORPHOLOGY SYNTAX
The linguist Leonard Bloomfield (1887–1949), one of the pioneers

of linguistics, called these levels the structure of language
(Bloomfield 1933). He recommended starting from phonetics and
going on to phonology, morphology, syntax, and semantics. This
kind of study would tell us a lot about the structure of a language.
Such an approach to language—i.e. seeing it as a structure—is
called structuralism.
Language as system
Every structure is a system in itself. If you change one part of it

the other parts change in order to adjust themselves. The first
linguist to point this out was Ferdinand de Saussure (1857–1913)
whom we shall discuss in more detail later. Saussure gave the name
of la langue to the whole system of a language.
The knowledge of the whole system is in the mind of the native

speaker. Without such knowledge sounds cannot be produced nor
meanings understood.
Saussure’s other concept, parole, involves ‘the combinations by
which the speaker uses the code of the linguistic system’ (Saussure
1916: 14). In other words it refers to what the speaker actually says,
i.e. the sounds one produces. Now, of course, one may be in a
hurry or under some pressure which may cause one to grope for
words or stutter. Thus parole is often faulty, incoherent or
inadequate whereas la langue is not. It cannot be because it is the
knowledge of how the whole system of a language works. It is
stored up in the brain but not all of it is used perfectly all the time
(For Saussure’s ideas read Culler 1976).
If we want to know how the system works then we need not look
at the actual performance of individuals. We can look at the laws
by which sounds combine together (phonology); the way units of
meaning combine to create words (morphology); and how words
are placed one after the other to create sentences (syntax). Such
studies of abstract laws or rules of the system of a language would
be the study of la langue. This is what theoretical linguists do.
Saussure’s concept of parole, which refers to utterances of
individuals, can also be called the performance of those individuals.
Noam Chomsky (1928–), perhaps the greatest linguist in existence;
used the term performance roughly for parole. For Saussure’s
concept of la langue, Chomsky used the term competence. As
competence is the knowledge of the system of a language, it is in
the mind of a speaker of that language. Theoretical linguists are
interested in this system in the mind but it is beyond the scope of
this introductory guide.
So far we have talked about linguistic theory, the focus of the
science of language. Some people call it microlinguistics because
it deals with only one aspect, the description of language, and not
with the other things language is connected with. These other
things can be society, culture, biology, mathematics, literature,
Introduction 5
politics, and so on. The subject dealing with the relationship of

language with society is called sociolinguistics. If the emphasis is
on culture, worldview, and preliterate societies, the subject is called
anthropological linguistics. Those who are interested in language-
learning and how the mind processes language study
psycholinguistics and neurolinguistics. The latter is very
interesting for biologists and medical scientists because it studies
speech defects such as aphasia and the involvement of the brain in
producing, understanding, and storing language. A pioneering
book on the biological foundations of language by Lenneberg
(1967) might repay reading. The brain has areas which store words
and the rules which produce sentences. Most of us use the left side
of the brain for linguistic work. We have about 12 billion neurons
and they connect with each other as if electrical currents are passing
through them. These days positron emission tomography (PET)
can tell us which part of the brain is working. The principle is
simple. If radioactive particles are given to a person in some form
of sugar and the person does linguistic work, the part of the brain
which consumes more energy will emit more particles. Thus we
can find out which part of the brain does what kind of work.
Unfortunately the scholarly work on this subject is a bit technical
in nature. Yet one book which can be recommended for beginners,
despite the fact that it has been superseded by new research (which
is again a bit too difficult), is Peter Russell’s book called The Brain
Book (1979). An advanced book on aspects of the mind-brain and
language—which I recommend because it has been edited, among
others, by a Pakistani linguist called Anjum Saleemi—is In Search
of a Language for the Mind-Brain (Saleemi et. al. 2005). This book
takes the reader into the highly interesting realm of philosophy as
well as neurolinguistics and brain science.
Apart from biologists, computer experts, and mathematicians too
are interested in linguistics. Thus there are branches of linguistics
called computational and mathematical linguistics. Computational
linguistics has given us the gift of seeing our South Asian scripts
on the computer screen. We now have Urdu written in the
nastaleeq script as well as Pashto and Sindhi in the naskh script on
everything including e-mail in Pakistan. In India we have many

scripts including the devanagari script. To make the rules which
put them there is a very complicated bit of computer science as
well as linguistic knowledge. Recently insights from linguistics have
been used by those who study the mind in general and artificial
intelligence in particular. There are new intellectual ventures and
philosophical linguistics, which is even older than the works of
the Greek philosophers Plato and Aristotle on the subject, is being
examined afresh. Highly complicated questions about how language
helps people to construct reality are being debated. Style, including
the use of style in literature, is being examined in stylistics. All
these subjects may be called macrolinguistics, i.e. the study of
language in relation to a number of other subjects. The following
chart helps clarify what has been explained above.
LINGUISTICS
Microlinguistics Macrolinguistics
Phonetics Anthropological
Language linguistics
Phonology
Pragmatics Planning
Morphology (Policies)
Psycholinguistics Mathematical
Semantics linguistics
Syntax Philosophical
linguistics
Sociolinguistics Computational
Introduction
linguistics
Historical
Stylistics linguistics
Neurolinguistics
In this book we will not cover all these subjects. However, an effort will be made to introduce you to microlinguistics and some aspects of language
in relation to society and culture.
7
2
A Brief History of Linguistics
E very society, and indeed almost every individual, has myths

about language. In general these are myths about origin: that
such and such language was created first by God or gods. Such
myths are not testable, and because everybody does not believe in
the same myth, they remain in the realm of faith-specific or
culture-specific beliefs. In order to pursue scientific linguistics, we
should aim at universality (i.e. our results should be valid for the
whole world) and objectivity (i.e. everyone should reasonably
believe our conclusions). An example of such an undertaking is
Guy Deutscher’s book on the evolution of language (Deutscher
2005). The book can be understood by those who know basic
linguistics. The arguments in the book are summed up below.
Basically, Deutscher argues that human languages have a complex
structure but it was not crafted by some genius or a committee of
geniuses in a short time. Instead, it arose through natural forces
which keep changing all the time. Of course, it is the people who
speak the language who change it. And why do they do that?
Possibly because, among other things, they may want to borrow
things from other cultures for which they have no words. A recent
example is the borrowing of English words in our South Asian
languages, especially in Urdu and Hindi. But there were also
borrowings from other languages such as Portuguese (did you know
that peon, almirah, and even achar is from that language?), and of
course, the borrowings from Persian and Arabic are too many to
be listed. However, it is not only the words which change. The
pronunciation also changes since we may cut short a few sounds
(economy) or decide to give a powerful impact to our words
A Brief History of Linguistics 9
(expressiveness). When we do the latter we may combine words

together to create new words (like fantabulous!). Since our minds
crave for order we keep creating patterns and so language keeps
changing. The author sums his thesis admirably as follows:
All that was required at the ‘me Tarzan’ stage were words for physical
objects and actions (as well as two pointing words), and a few natural
principles for ordering them. From these basic materials, the natural
forces of change could have fashioned the structure of language in all
its prolix splendour (Deutscher 2005: p.262).
Of course, nobody knows definitely whether there ever was a stage

when people used only a few words like the Tarzan comics do
(I Tarzan; you Jane). This is a hypothesis but it should be accepted
as being true for the time being till we get a better explanation.
This is how scientific research proceeds. Nothing is sacred in
science, but whatever theory is accepted must explain things for
the moment. Incidentally, this theory is very similar to the theory
of the development of genes in the bodies of plants, animals, and
humans. For a very clear account, read Richard Dawkins’ book
called The Selfish Gene (1978). He too says that all these
complicated forms of life grew as genes learned to produce
photocopies of themselves in order to survive. But let us get back
to language.
The first language
If language arose with the Homo Sapiens (sapient= wise) then it

was born about 150,000 years ago probably in Africa, though it is
possible that several humanoid groups arose in places within years
of each other. This has been calculated on present-day variation of
a form of DNA passed by the female line. The rate at which it
mutates can be calculated and that gives us this approximate figure
(Corballis 2002: 42). But we still do not know what this early
humanoid spoke. There is, indeed, a theory that the earliest form
of communication was the gesture. Michael Corballis has argued
this in a highly interesting book (Corballis 2002).
Another question which people ask is whether there is a natural or

‘original’ language of humanity. The brief answer is that we do not
know whether there is any such thing, and if there is, we do not
know which one it was or is. People have carried out cruel
experiments to find out what this language is. For instance, the
Mughal king Akbar put children in a house where only deaf-mutes
served them. They were not allowed to hear any human being
talking. These children could not speak when they grew up.
Children brought up by wolves and other animals (called feral
children) do not know any human languages but do try to make
sounds of animals. In short, human beings probably do not have
any original or natural language. They merely learn the language
which they hear around them.
Are languages equal?
Yet another myth is that some languages are sweet while others are
rough. Linguists do not classify languages in this manner because
‘sweet’ and ‘rough’ are subjective ideas. If we are talking about
sounds we find the sounds called sweet in one language also present
in a language which is considered rough. If, however, we are talking
about intonation (the way we stress certain parts of words and our
voice goes up and down) then we need to test this intonation
pattern on a big sample of people in order to be able to claim that
the terms ‘sweet’ or ‘rough’ are related to properties of speech
itself.
Most people also agree on calling some languages superior and
others inferior. They say that one language is more ‘logical’ or has
more words than another which lacks these characteristics. Modern
linguists hold the view that all languages are equal; none is superior
to another. This does not mean that some of them are not better
suited for a certain function. English, for instance, is better suited
for space research or computers than a language which does not
have the technical terms for such kind of work. This, however, is
functional suitability not structural superiority. As far as the
structure—phonemes, morphemes, syntax etc.—is concerned, we
find it difficult to set up a standard of excellence and then measure

languages according to it. It may be true that certain sounds or
patterns of stress may sound musical or pleasing to everyone but
no scientific research has proved that any language is objectively
‘sweet’ or ‘beautiful’ for all human beings. Thus there are no
scientific grounds for saying that any language is intrinsically (i.e.
structurally) inferior or superior to any other language. As for
functional suitability, any language can be changed to perform any
function. For example, if we want Punjabi to be used for
sophisticated scientific research, we can invent or borrow words to
express new meanings. That is what the speakers of English did
and that is what all language planners do everywhere when they
want to use a language for science and technology.
Why then do people hold such strong myths about languages? In
general these myths are expressions of their views about the world.
If they look down upon someone or a certain culture, they also
tend to look down upon that person or cultures’ language. In South
Asia city dwellers look down upon villagers. Thus they also look
down upon villagers’ languages. They regard them as inferior and
even assert the view that they do not have a grammar. As you
know, it is impossible to speak a language if its speakers do not
have grammatical rules in their minds. Villagers do have the
grammar of their languages in their minds but it may not be a
written grammar. It may be unwritten not because it is inferior,
but because villagers neither have power nor money, and without
these things one cannot promote one’s language or publish books
in it.
The persistence of these myths shows us that one must put one’s
prejudices and beliefs aside when one makes a scientific study of
language. If we look at the history of linguistics, we find that it is
only recently that people became scientific in this field. This does
not mean that there were no eminent linguists in earlier eras. There
were many, and this chapter will introduce you to their thoughts.
Linguistics in antiquity
Most Western works on the history of linguistics begin from Plato

(427–347 bc). However, it was in South Asia—to be precise,
somewhere around Taxila in modern Pakistan—that the Sanskritic
grammarian Panini lived between the seventh and fifth centuries
bc, i.e. earlier than Plato. There might have been linguists even
earlier than Panini but there is no known written work left
behind.
Panini, however, left behind 4000 sutras (verses) on Sanskrit. His
eight books, Ashtadhyayi, contain descriptions of Sanskrit grammar.
Panini may be said to be the pioneer of phonetics because he tells
us how sounds are produced, i.e. where the tongue moves in the
mouth.
Plato, the great Greek philosopher, was more interested in
semantics than phonetics. In his book Cratylus, some characters
argue about the nature of language. One of them, named
Hermogenes, argues that language is a product of social convention.
This means that we use the word dog for a four legged animal of a
certain kind only because of an arbitrary convention. We can agree
to call it kutta, sug, kulb, kutto, spai or chien. No matter what we
call it, the sounds we produce refer to this animal. There is no
necessary relationship between the sounds and the object. This is
known as the theory of the arbitrariness of language or the
conventionalist view. It is also known as the theory of language
being an agreed upon convention, an arbitrary code.
The opposite view, argued by Cratylus, is that words and the
objects they refer to have an intrinsic relationship; that there is
some essential core of meaning in the sounds we produce which
refers only to certain objects in the world.
Socrates, the mouthpiece of Plato, seems to incline to this view
though no final conclusion is reached. Later, the Greek philosopher
Aristotle (384–322 bc) also agreed with this view in his easy called
De Interpretatione. This is also known as the essentialist or
naturalistic view.
The debate between the conventionalist and naturalistic views

persisted throughout the medieval ages. In fact Ferdinand de
Saussure, whom we shall discuss below, also contributed to it.
Linguistics in the Middle Ages

From the time of the Romans, linguists became more interested in
grammar that anything else. Marcus Terentius Varro (127–116
bc), a Roman grammarian, wrote a book on Latin grammar called
De Lingua Latina. The Romans were interested in the art of
speaking forcefully and grammar was part of it.
This emphasis on speaking and writing (style) continued in Europe.
In the medieval universities—Padua, Paris, Oxford, Cambridge
etc.—grammar, dialectic and rhetoric were taught. As Latin was
the language of scholars, it was considered superior. Thus when
European scholars wrote grammars of modern European languages,
such as that of English, they made these grammars like Latin
grammar.
In South Asia too grammar became an artificial subject full of
complicated rules and special terms. Modern Urdu grammars are
even now full of Persian and Arabic terms which one has to
memorize. Hindi grammars also have terms borrowed from
Sanskrit which one has to memorize. These grammars do not
describe what native speakers already know. They merely give
artificial rules telling you to do this or not to do that. This is
prescription and such grammars are called prescriptive grammars.
This prescriptive tradition is part of schooling in the teaching of
all languages, including English, in South Asia.
Comparative linguistics
In the nineteenth century linguistics progressed most in central

Europe. European (especially German) universities were research
oriented and their main object of research was comparing the
words of one language with another. This method is also known
as philology (philo means love and logos is word, i.e. love of words).
As the meanings of words change, these scholars had to study the

origin of words (i.e. etymology) also. That is why they are known
as historical and comparative linguists.
The main contribution of these philologists was the discovery that
languages can be grouped in families. The languages of Europe and
South Asia fall into the Indo-European family of languages. This
was established by comparing words which were so similar that it
seemed probable that they were derived from one word. This
means that there was one parent language which gave birth to other
languages as people moved away from their original homeland.
One of the main contributors to this kind of research was Sir
William Jones (1746–1794). He was a judge in India and he made
the Asiatic Society of Bengal. He established that Sanskrit is so
similar to Latin and Greek that it could not be by chance. Instead,
he argues, these three languages must have descended from one
common ancestor.
Following this theory, the German linguist Franz Bopp wrote a
book called On the Conjugation System of the Sanskrit Language, in
comparison with those of the Greek, Latin, Persian, and Germanic
Languages (1816). In 1818, Rasmus Rask, a Danish linguist,
published his book Investigation on the Origin of Old Norse or
Icelandic Languages. Later Jakob Grimm (of the fairy tales fame)
also contributed to this field. Karl Brugmann (1897–1916)
contributed his two volumes on the sounds and forms to the
Outline of Comparative Indo-European Grammar in five volumes
published between 1886 to 1893. This was an authoritative work
and the idea of there being an Indo-European language became
common.
The reasoning of comparative linguists can be understood by
looking at Table 1 and 2. Both give words common between
several European and Asian languages.
Table 1
Concept Languages
Pashto Latin Greek Sanskrit French English Persian Punjabi
father plar pater pater pita pere father pidar peo
three drae tres treis tri trios three sae tin
Table 2
Languages
S.No. Persian Urdu/Hindi Punjabi Sindhi Pashto French Arabic
1. yak ek ik hik yau un vahid
2. du do do ba dwa deux athnan
3. sae teen tin tre drae trois salasa
4. chahar char char char salor quatre arba
5. punj panch punj punj pinzao cinq khamsa
6. shish che che cha shpug six sitta
7. huft saat sut sut owo sept saba
8. husht aath uth uth ato huit samania
9. noh nao no nava nava neuf tisa
10 daeh das dah dah las dix ashara
In Table 2, notice that Arabic sounds for the same number are
often completely different, whereas those in the other languages are
similar. This is because Arabic belongs to a different language
family, the Semitic family, whereas all the other languages
mentioned above belong to the Indo-European language family.
August Schleicher (1821–1868), a German scholar, introduced the
concept of a family tree theory. According to the family tree theory,
languages of the same family descend from a common ancestor.
This kind of tree was not constructed at once. Scholars spent years
of painstaking research to provide evidence which made such
constructions possible. The family tree in Figure 1 is for Proto
Indo-European (PIE), i.e. our idea of what the original Indo-
European language was.
Besides the Indo-European family of languages, there are many

other families too. Table 3 gives the names of some major language
families.
Table 3
Language Family Languages
Semitic Arabic, Hebrew, Amharic
Altaic Turkish, Mongol
Sino-Tibetan Mandarin Chinese, Wu Chinese of Shanghai, Languages of Tibet
and Southeast Asia
Dravidian Tamil, Telugu, Kannada, Malayalam
Bantu African languages
Apart from these language families, there are other families too as
mentioned above. Besides these there are some language isolates,
i.e. languages which cannot be assigned to any family. One example
of such an isolate is the Burushaski language spoken in Hunza,
Yasin, and parts of Gilgit in Pakistan. For details of about 1000
languages listed according to their numbers of speakers, see Crystal
(1988: pp.436-444).
Figure 1
The Family Tree for Proto Indo-European (PIE) Languages
Anglo-Frisian English
Frisian
Germanic Dutch
Dutch-German Afrikaans
German
Yiddish
Danish
Swedish
Scandinavian Bokmal
Nynorsh
Breton Icelandic
Celtic Welsh Faroese
Irish and Scottish
Gaelie
Portuguese and creoles
Oscan and Umbrian Castilian and creoles
(dead languages) Catalan
Italic Occitan
French and creoles
Latin
Italian
Sardinian
Hellenic Demotic Rhaeto-Romanic
PRIMITIVE Albanian Katharevousa Rumanian
INDO-EUROPEAN Armenian Old forms: old Persian, Avestan and Pahlavi
Tocharish Kurdish
Iranian Persian
Pashto
Baluchi
Tadzhik
Indo-Iranian
Old forms: Sanskrit and Pali
Hindi
Urdu
Indic Bengali Gujarati
Assamese Sindhi
Oriya Punjabi
Marathi Dardic
Nepali Sinhalese
Rajasthani Bihari
Lithuanian Bhili Pahari etc.
Baltic
Latvian
Polish
Balto-Slavonic (West Slavonic) Czech
Slovak
Serbo-Croatian
Slavonic (South Slavonic) Slovene
Bulgarian and Macedonia
Russian
(East Slavonic) Ukrainian
Byelorussian
Twentieth century linguistics
Modern linguistics was born mainly out of the work of Ferdinand

de Saussure who has already been mentioned. He was born in
Geneva in 1857 in a family of educated people some of whom were
eminent in their lifetime. In 1876 while studying in Leipzig,
Ferdinand de Saussure studied comparative linguistics which was
fashionable in those days. Like other linguists, Saussure too started
his career with historical-comparative linguistics in the philological
tradition. His book was originally written in French and the
English translation of the title is Memoir on the Primitive System of
Vowels in the Indo-European Languages was published in 1879.
However, Saussure was not satisfied with it. On 4 January 1894,
he wrote in a letter to Antoine Meillet,
‘The utter inadequacy of current terminology, the need to reform it,

and in order to do that, to demonstrate what sort of object language
is, continually spoils my pleasure in philology.’ (Culler:1976: p.15).
Such a terminology he made himself, though he did not publish it

in his lifetime. In 1891, Saussure was offered a chair (i.e. a
professorship) in Indo-European linguistics in the University of
Geneva. It was here that his students were sufficiently impressed
by his lectures to take down notes. After his death in 1913, two of
his students collected the notes and in 1916 published a book
entitled Cours de Linguistique Generale (A Course in General
Linguistics). This is the book which created modern linguistics by
presenting the following theories.
(1) Linguistic signs are arbitrary
If you remember the conventionalist–naturalistic debate from

Plato’s Cratylus mentioned earlier, you will have no difficulty in
understanding Saussure’s theory. Briefly, Saussure agreed with the
conventionalist view. He said that language was merely an arbitrary
system in which any series of sounds could mean anything
arbitrarily.
The term linguistic sign (or simply sign) means the sounds which
refer to some object (material or non-material). For example, dog
is a series of sounds. It is a linguistic sign. In English it refers to a
four-footed animal of the canine family. Thus, it signifies an
animal. The animal it signifies is called signifie, while the sounds,
or words, can be called signifiant. (See Table 4).
Table 4
Language Signifiant Signifie
English dog An animal of the canine family
French chien ”
Arabic kulb ”
Urdu/Hindi kutta ”
Persian sug ”
Pashto spai ”
We see that the sounds which signify (signifiant) are different but
the thing which is signified (signifie) is the same. Thus, argued
Saussure, the relationship between the linguistic sign (signifiant)
and the objects of the world (signifie) is arbitrary. We could have
chosen any series of sounds to mean anything. There is nothing in
the nature of sounds themselves that makes them fit for particular
things. You might object that there are words which sound like the
thing they describe such as the noises of animals. These are called
onomatopoeic words. But even here different languages have
different sounds for noises. (See Table 5).
Table 5
Barking of a Dog
Urdu/Hindi bhau bhau
English bow wow
The onomatopoeic words in Malay are as follows:

Onomatopoeic Lexicon Meaning
debuk The sound of slapping or boxing.
debum The sound of heavy things falling.
debung The sound of drumming.
debur The sound of waves breaking on the shore.
debus The sound of flapping wings.
Source: Jhy Wae 2005: Vol. 2, 796-797.
Barring these partial exceptions, Saussure’s theory is correct. This

theory is very important because it led to the study of linguistic
signs in relation to other things and eventually to the relationships
themselves. These relationships can be expressed in mathematical
language, which makes modern linguistics very precise and
scientific (for details see Culler 1976:p. 19–23).
(2) Synchrony versus diachrony
Saussure distinguished between synchronic and diachronic studies

of human language. Chronos means time in Latin; syn means at the
same time while dia refers to the passing of time. Thus synchronic
linguistics is the study of the language system at this time while
diachronic linguistics is the study of this system in history.
In Saussure’s time diachronic linguistics was emphasized. Nowadays
synchronic linguistics is in fashion. It is felt that if we want to
know about how a language system actually operates or how it is
stored in the human mind, we should study that system as it is.
This has made people resort to descriptions of language systems as
they are at present.
(3) Syntagmatic versus paradigmatic
Consider the following sentence.

(1) He is like a horse
Here the words which follow each other in a linear direction (i.e.
a straight horizontal line) are He + is + like + a + horse. This
relationship between the words is called a syntagmatic
relationship.
Now let us look at the same sentence and what we can put in place
of the words in it:
(1a) He is like a horse
(2) She was unlike the cow
(3) I am like one monkey
(4) You are like a boy
(5) We are unlike most cats
You can see that he can be replaced by words of the same kind.
Similarly other words can be replaced by words of the same kind.
In the slot of the verb ‘to exist’ (is), for instance, we can only put
in other forms of that verb (i.e. am, was, are etc). We cannot put
in words of another kind. These slots represent places which can
be filled in by words related to each other. This relationship is
called the paradigmatic relationship.
Thus out of the paradigm of pronouns we can choose any: he, she,
it, I, you, we etc. Similarly we can choose words for other slots
also.
(4) Langue versus parole
As this concept has been explained earlier, there is no need to spend

more time on it. Briefly, la langue is the system of a language as a
whole while parole are individual utterances.
Structuralism
Ferdinand de Saussure’s work influenced the development of
structuralism. Briefly, language is a system. It is a structure with its
various parts connected with each other. In Europe, a well known
school of linguistics is called the Prague school. It was represented
by Nikolay Sergeyvich Trubetskoy (d. 1965), and it carried on the

structuralist tradition in linguistics. He also lectured in Paris, and
therefore influenced French linguists like Meillet.
In America, the anthropological linguists Franz Boas (d. 1942),
Edward Sapir (d. 1939), and Benjamin Lee Whorf were
structuralists. However, their work will be dealt with in the context
of anthropological linguistics. The only American structuralist
linguist who will be mentioned here is Leonard Bloomfield
(d. 1949). In 1924, Bloomfield reviewed Saussure’s book which
influenced the development of linguistics in America. Chomsky’s
concepts of competence and performance are very close to
Saussure’s distinction between langue and parole.
Nowadays the ideas of Saussure are being promoted by the Cahiers
Ferdinand de Saussure and there is also an Institute Ferdinand de
Saussure. Some fresh material by Ferdinand de Saussure has also
been published and interest in his work has been revived (Sanders
2005: 931).
Structuralism not only dominated linguistics but also came to
dominate sociology, anthropology, and philosophy. We will study
its application by Claude Levi Strauss to anthropology later. Here
it may be mentioned that Emile Durkheim and Marcel Mauss,
both eminent sociologists, used the ideas of structuralism to study
society. Just as language is a system in which there are binary
oppositions, so society too was classified according to binary
oppositions. Thus a group of people could be ‘bride-givers’ (those
who give their daughters in marriage) while those who are given
the brides may be called ‘bride-takers’. They, in turn, may give
their own girls in marriage to another group. In short various
aspects of a culture are connected with each other in such ways that
they cannot be fully understood in isolation. Such were the insights
which affected all the social sciences greatly as a consequence of the
ideas of the linguists Wilhelm von Humboldt (1767–1835) and
Ferdinand de Saussure.
Humboldt was a German linguist who worked on the Basque

language which is spoken in Spain. However, he is better known
for his theory that there is some correspondence or similarity
between the character of the speakers of a language and the
structure of the language. This theory leads to racist stereotyping
and it has long been abandoned. However, his work in philology
did influence other linguists and he is considered an important
figure in nineteenth century linguistics.
Bloomfield was also strongly influenced by the work of the
psychologist Wilhem Wandt who believed in behaviourism, i.e.
the study of behaviour. This meant that reference to mental
categories was eliminated or reduced to a minimum. Since the
mind itself could not be observed, it was considered unscientific to
talk about what went on inside it. Behaviour, on the other hand,
could be observed. Thus human behaviour was the focus of the
psychologist’s study. In linguistics, this meant studying speech as
if it was the response to some stimulus. Stimulus means what we
put in and response means what comes out or what happens. For
instance, if we apply the stimulus of extreme heat to an animal’s
body, the response of the animal will be to move its body away so
as to escape the pain. Bloomfield felt that meaning too could be
seen as response to some kind of stimulus. Bloomfield’s book
Language (1933) dominated American linguistics almost till the
time of Chomsky. In this book, however, we shall mostly be
concerned with Chomskian linguistics. However, let us first
understand some of the basic concepts of theoretical linguistics.
For this we turn to phonetics in the next chapter.
3
The Study of Sounds
T he study of sounds in human speech is called phonetics. It

has the following branches:
(1) Articulatory phonetics
(2) Acoustic phonetics
(3) Auditory phonetics
Articulation means producing sounds from the mouth. Thus, this
branch of phonetics deals with the production of sounds, their
description, and their transcription (i.e. writing them with the help
of symbols).
Acoustic phonetics is the study of the physical properties of the
sounds themselves. There are machines which give bands of
frequencies for sounds. Other features of sounds can also be
studied.
Auditory phonetics is the study of the reception of the sound by
the ear and its processing in the brain.
In other words we study (1) how sounds are produced, (2) how
they travel from the mouth to the ear, and (3) how they are
perceived by the ear and the brain. In this book we shall only study
articulatory phonetics.
The production of speech sounds
Let us first look at the mouth where speech sounds are produced.
Diagram 1 will help you locate the main articulators, i.e. organs
which help us in producing sounds.
The Study of Sounds 25
Diagram 1
Mouth
Sounds are made by the air striking against the articulators. This
air comes in the form of a stream from the lungs. This is known
as the pulmonary stream. It is also known as the egressive stream,
i.e. that stream of air which goes out of the mouth.
Sounds can also be produced by taking in air into the mouth. This
is called the ingressive air stream. There are sounds in Sindhi and
Siraiki which are produced by the ingressive air stream.
(1) Larynx
The larynx is a box-like space which allows the air from the lungs
to pass through it.
(2) Vocal cords
In the beginning of the larynx are vocal folds, or vocal cords. They
are two bands of muscles lying opposite to each other (see
Diagram 2).
(3) Pharynx
It is a tube which begins above the larynx (i.e. the passage which
goes down to the lungs). It is approximately 7cm long in women
and 8cm in men. At the top it divides into two. One part goes into
the nose and the other in the back of the mouth.
Diagram 2
When the air passes through, the vocal cords can open and close
very fast. This is called vibration and it can be from 70 to 1000
times per second. This vibration is called voicing. If you say /z/
and keep your tongue in the same position saying zzzzz you will
hear a buzzing sound. This is the vibration of the vocal cords. You
can place your fingers on your throat and feel this vibration. If you
want to switch it off say ssssss and it will cease. Now you know
that /z/ is a voiced sound and /s/ is an unvoiced sound. Other
sounds are also distinguished like this.
(4) The nasal cavity
Let us look at the space in the nose. It is called the nasal cavity. It
has air in it which can produce a kind of humming noise called
resonance.
The nasal cavity can be made larger if the soft palate is lowered.
As the soft palate is a movable bit of flesh it can divide or connect
the nasal and the oral cavities. If it is lowered the volume of air
which will vibrate will be more. Now if the mouth is blocked and
all this air passes out of the nose we get what is called a nasal effect.
This is known as nasalization.
If your nose is blocked at the back of the nasal cavity, you will not
be able to nasalize your sounds. If, however, your soft palate is
permanently lowered, nasalization will increase.
(5) The oral cavity
Most of the articulators are in the cavity of the mouth or the oral
cavity. Let us discuss them one by one.
(a) The lips

The lips can close together stopping the air. This action creates
sounds which are called stops or plosives. They are called
plosives because the air is stopped for a very short time and then
goes out or in, i.e. explodes or implodes. Two such sounds are
/p/ and /b/. Say them and feel how your lips close together.
Speakers of Urdu–Hindi do not use implosive sounds in their
languages. They are produced by sucking the air inwards. Try
to say /p/ and pull the air in. You will produce an implosive
sound. If you cannot do it, request a Sindhi friend to say /b/ as
used for a goat in that language. You will hear an implosive /b/.
The lips can also be rounded or brought close together without
meeting each other. These actions too produce sounds which we
will describe later.
(b) The teeth

The tongue can be behind the teeth or between them for certain
sounds. The lower lip meets the upper teeth if we produce the
sounds /f/ and /v/.
(c) Alveolar ridge

Just behind the upper teeth is a slightly raised longish ridge
called the alveolar ridge. Some sounds are produced when the
tongue touches or comes near this ridge.
(d) Hard palate

It is the hard sloping place behind the alveolar ridge. If you take
your tongue back from the alveolar ridge you can feel it.
(e) Soft palate

It is a soft fleshy place behind the hard palate. It is also called
the velum and sounds produced by the tongue touching it or
coming near it are called velar sounds.
(f) Uvula
It is a triangular piece of flesh hanging down from the roof of
the mouth. You can see it in a mirror if you open your mouth
keeping your tongue down. Some sounds are produced when it
comes in contact with the tongue.
(g) The tongue

The tongue can move up and down and turn in the mouth. By
doing so, it comes in contact with other parts of the mouth and
helps in producing sounds. The tongue is divided in the
following parts (see Diagram 3):
(i) tip (ii) blade (iii) front (iv) middle
(v) back (vi) root
Remember that the front of the tongue is not its tip. Behind the
tip (or apex) is the blade and what is called the front is behind
that. In fact, the front is almost towards the middle of the tongue.
When the tongue is at rest the tip touches the teeth, the blade lies
below the alveolar ridge, the front is below the hard palate, and the
back is below the soft palate. The root joins the tongue with the
back of the mouth.
Some books use the following adjectives:

apical to do with the tip of the tongue.
dorsal to do with the dorsem, i.e. front and back of the tongue.
lingual to do with the tongue (lingua is Latin for tongue).
laminal to do with the blade of the tongue.
Diagram 3
There are differences in terminology in linguistics and you are

advised to read dictionaries of linguistics if you come across
unfamiliar terms.
(h) The glottis

This is the place between the vocal cords. All glottal sounds—
sounds coming deep from the throat—come from it.
The classification of sounds
We classify sounds by describing where they are produced in the

mouth and how they are produced. The first is known as the place
of articulation and the second the manner of articulation. The
place of articulation may be the lips, the lips and the teeth, and so
on. The manner of articulation means how the air stream comes
out of the mouth. It can be stopped and then released, it can be
released with friction, the lips can be rounded or opened, the vocal
cords can vibrate or not, and a number of other things can
happen.
Let us start from some of the commonly used sounds of English

and South Asia languages which are articulated by the lips and go
on to those which come out of the glottis. These sounds will be
written in the symbols of the International Phonetic Association
(IPA).
The IPA was founded in Paris in 1886 by school teachers who
wanted a notation for the pronunciation of foreign languages in
their schools. It was based on the Roman alphabet because the
people who made it were Europeans and this was the alphabet they
used. However, since they wanted exactly one symbol to represent
one sound they put marks (diacritical marks) on the symbols and
brought some from the Greek alphabet to represent the sounds of
many languages. Academics started working on the IPA with a view
to understanding pronunciation in general as well as the
pronunciation of individual languages.
One prominent scholar who has influenced British phonetics is
Henry Sweet (1845–1912). Sweet was a don (faculty members at
the universities of Oxford and Cambridge were called dons) at
Oxford from 1901 to 1912. He was interested in spelling reform,
old English, as well as phonetics. His History of English Sounds
(1874) as well as Handbook of Phonetics (1877) influenced a whole
generation.
Before we start a discussion on phonetics, let us review a few
definitions. A sound is known as phonema in Greek. From this
comes the technical term phoneme which means the smallest unit
of sound of a language. Actually it is an abstract concept because
we cannot pronounce most pure phonemes. If we try to pronounce
a pure consonant we cannot do so. We have to add a vowel either
before or after it. However, we have to talk of phonemes, units of
sounds, in order to describe them.
Now let us define types of sounds. A consonant is a phoneme
which is produced with some sort of obstruction in the air stream.
A vowel, on the other hand, is produced without obstruction in
the vocal tract.
Consonants
Now let us classify the consonants of English and some South
Asian languages:
Bilabials
Bilabial stops Labia means lips. /p/ and /b/ are sounds
made by the lips. The place of articulation
is the lips. The manner of articulation is
that the lips are brought together, the air
stream is stopped and then released with
an explosion. Thus, /p/ and /b/ are called
stops or plosives.
The vocal cords do not vibrate in /p/ but
they do vibrate in /b/. Thus /p/ is a
voiceless bilabial stop and /b/ is a voiced
bilabial stop. If the air stream is taken in
the stops are known as ingressive stops
and written as follows: /ƥ/ and /ɓ/
Bilabial nasal /m/ is also produced by the meeting of the
two lips, but in this case the velum is
lowered and the air comes out of the nose.
The manner of articulation is nasal and
/m/ is a bilabial nasal.
Semi vowel /w/ is produced by the lips rounding
together. The manner in which it is
articulated is very similar to vowels. The
air comes from the lungs and flows out
with very little obstruction except that
which is offered by the rounded lips. That
is why this sound is known as a semi-
vowel or approximant.
Labio-dentals
Labio-dental fricatives /f/ and /v/ are articulated when the upper
teeth come in contact with the lower lip.
The manner in which they are articulated
is that the air keeps flowing out with
friction. That is why they are known as
fricatives: /f/ is unvoiced and /v/ is
voiced.
Dentals
Inter-dental Written as /q/ and /ð/, they are fricatives
articulated by the tip of the tongue
coming between the teeth. They are
fricatives so the air flows out with friction.
/q/ is unvoiced and /ð/ is voiced. These
phonemes exist in English and Arabic, but
not in Urdu, Hindi, Punjabi, Sindhi,
Pashto, Bengali, Gujrati etc. They do,
however, occur in Eastern Balochi. Thus
most South Asian speakers of English use
/th/ for /q/ as in think and /d/ for /ð/ as
in then.
Dental stops When the blade of the tongue touches the
back of the upper teeth, we produce
dental phonemes. The air stops and is
released. Hence these are dental stops or
plosives. To show that they are dental, a
dental diacritical mark is put below them.
/t /̪ is unvoiced and dental /d̪/ is voiced.
These are the ‫ ت‬and the ‫ د‬in the Perso-
Arabic (Urdu) script.
They do not exist in English but do exist
in French, Spanish, and Italian. They are
very common in South Asian languages;
e.g. teen in Urdu means three, and deen

means religion.
Alveolars
Alveolar stops /t/ and /d/ are articulated by the blade or
the tongue coming in contact with the
alveolar ridge. As they are stops, the air
stream stops before being released. /t/ is
unvoiced and /d/ is voiced.
Retroflex The place of articulation remains the same
but the manner is different. For retroflex
sounds, the tip of the tongue goes up and
curls back. These sounds are common in
South Asian languages and distinguish
speakers from Pakistan, India, Bangladesh,
Sri Lanka, and Nepal from people of
other countries. They are written in the
IPA as /ʈ / and /ɖ / but for the sake of
convenience we have represented them
with a dot beneath them as /ṭ/ and /ḍ/.
It is interesting to observe that British
English does not have retroflex phonemes
so that their alveolar /t/ and /d/ are not
the same as our retroflex equivalents. We
do, however, pronounce ten and den with
retroflex stops. They, in turn, pronounce
our words beginning with retroflex stops
in their own way, i.e. with alveolar ones.
The difference is most obvious when the
English, the Arabs or the Iranians try to
pronounce words with retroflex /ɽ/. They
produce /r/ and not /ɽ/.
Alveolar fricatives Written as /s/ and /z/, these are articulated
by bringing the blade of the tongue close
to the alveolar ridge. Being fricatives, the
air goes out with friction. /s/ is unvoiced

and /z/ is voiced.
Alveolar nasal Written as /n/, it is articulated by bringing
the blade of the tongue close to the
alveolar ridge. As the mouth is closed and
the velum is lowered, the air is forced to
go out from the nose. This makes it a
nasal sound. It is voiced.
Lateral Written as /l/, it is articulated by putting
the blade of the tongue on the alveolar
ridge. It is called a lateral because the
tongue is pushed in a lateral direction, i.e.
the left side of the mouth. You can verify
this by saying ‘lllll’ and sucking in air.
Only the right side of the tongue will be
cooled by the incoming air proving that
the left side is close to the mouth.
Palato-alveolar
Fricatives The middle of the tongue makes a hump
which comes near the hard palate only
slightly behind the alveolar ridge. The air
goes out with great friction and a strident
sound of sh is produced. This is the
unvoiced version of the phoneme which
is written as /ʃ/ while the voiced version
is written as /ʒ/. The first is used in she
while the second is in vision.
Palatal
Behind the alveolar ridge is the hard
palate. Many sounds are produced with
the tongue coming in contact with it.
However, the one which is significant
because it is used in English as well as
South Asian language is /j/.
Palatal semi vowel Written as /j/ in the IPA, this phoneme

or approximant occurs in the beginning of yes. Remember
it is not the first phoneme of judge which
is written as /ʤ/.
Velar
Velar stops or plosives Behind the hard palate is the soft palate
called velum. The sounds produced with
the tongue coming in contact with (or
very near) it are called velar sounds. The
ones we need to know are /k/ and /g/.
The tongue meets the velum and the air
is stopped before it is released. /k/ is
unvoiced and /g/ is voiced.
Velar nasal The only velar nasal sound is ng written
as /ŋ/. It is the sound produced by ng in
finger and singer etc. In English and South
Asian languages, it occurs only in the
middle or end of words. In some other
languages it occurs in the beginning of
words.
Velar fricatives The tongue comes near the velum and the
air keeps flowing out with friction. These
sounds do not occur in modern English
but they are used in Urdu, Pashto, Persian,
Arabic, and many other languages.
The unvoiced velar fricative is written as
/x/ and is the first phoneme of the word
khan. In Hindi it takes the form of /kh/
though some people do use /x/.
The voiced one, written as /ɣ/ is the
phoneme depicted by gh in Afghan. If you
want to hear the exact sounds of these
symbols ask a speaker of one of the above
mentioned languages to pronounce these
words.
Uvulars
These sounds are produced with the back
of the tongue coming in contact with the
uvula. There are no uvulars in English but
they do exist in Persian and Arabic.
Uvular stops The sound produced by stopping the air
with the back of the tongue coming in
contact with the uvula is a uvular stop.
The unvoiced one is written as /q/ and
the voiced one as /ɢ/.
/q/ is used in Persianised Urdu which was spoken in parts of UP,
such as Lucknow and Delhi. In Pakistani Urdu /k/ is used in its
place. In Hindi too it is not an indigenous phoneme but it is used
by some Hindi-speakers.
Pharyngeals
Fricatives Sounds produced in the pharynx. The
only one which you may come across is
the fricative written as /ħ/. It is the Arabic
pronunciation of <‫>ﺡ‬. You may hear
Arabic-speakers (or those who know the
Arabic pronunciation) use this sound
when they say Rahman. It sounds almost
like Rakhman but it is actually /ħ/.
The voiced pharyngeal fricative is written
as /ʕ/. The sound of <‫ >ع‬in Arabic is
represented by it. The air is released with
friction deep in the throat and the vocal
cords vibrate. South Asians generally use
a vowel instead of this sound which is not
part of their sound inventory.
Glottals
The most commonly used glottal fricative
is /h/ which is described as a voiceless
glottal central fricative. It is actually a

voiceless vowel with some friction in the
glottis. There is a voiced version also
written as /ɦ/. There is also a glottal stop
written as /ʔ/. The air stops in the glottis
when this is produced, and is released
with an explosion.
The information given above can be summarized in the chart of
consonantal sounds given in Figure 2. For more details about
consonants read other books on phonetics such as O’Connor
(1973), Ladefoged (1982), and Roach (1983).
Figure 2
THE INTERNATIONAL PHONETIC ALPHABET
CONSONANTAL PHONEMES
PLACE OF ARTICULATION
Retroflex
Alveolar
Pharyn-
alveolar
Bilabial
Glottal
dentals
Uvular
Dental
Palatal
Labio
Velar
Post
geal
M Plosive pb td ʈɖ cɈ kɡ qɢ ʔ
A
N Nasal m ɱ n ɳ ɲ ŋ N
N
E Trill B r R
R
Tap or Flap ɾ ɽ
O
Fricative φβ fv θð sz ʃʒ ʂʐ çʝ xɣ χʁ ħʕ hɦ
F
Lateral
A fricative ɬɮ
R
T Approxi- ʋ ɹ ɻ j ɰ
I mant
C Lateral
U approxi- 1 ɭ ʎ ʟ
L
mant
A
T Ejective
I stop p’ t’ ʈ’ c’ k’ q’
O
N Implosive ƥɓ ƭɗ ƈʄ ƙɠ ʠʛ
Where symbols appear in pairs, the one to the right represents a voiced consonant.
Vowels
As in the case of the consonants, vowels too will have to be
classified with reference to the position of articulators in the
mouth. Vowels are sounds in which the air is not stopped in the
mouth. Thus it does not stop and explode (as in the case of
fricatives). The tongue does move in the mouth while producing
vowels, and the lips expand and contract. Vowels are determined
by the tongue and the lips. The following three dimensions are
used in the classification of vowels:
1. Height
The tongue can be made to go up or down in the mouth. If it is
up, i.e. near the roof of the mouth, the vowel will be a high vowel.
If it is low, i.e. near the lower jaw, the vowel will be low. If it is
in the middle, the vowel will be a mid vowel. This is also called
closeness. When the tongue is close to the roof of the mouth, i.e.
high, the vowel is a closed vowel.
2. Frontness
The tongue forms an arch or hump when a vowel is pronounced.
If this arch is made in the front of the mouth, the vowel is called
a front vowel; if in the middle, it is a central vowel; and if in the
back, it is a back vowel.
3. Roundness
If the lips are rounded, the vowel is a round vowel. If they are in
the neutral position, it is neutral, and if they are spread, the vowel
is a spread vowel.
Thus, the point of articulation of a vowel is the highest point on
the arch of the tongue. The position of this point on the height
and frontness dimension gives us the description of the vowel. To
this we add information about lip rounding to get a complete
classification of the vowel. The following chart summarizes this
information:
FRONTNESS
FRONT CENTRAL BACK
H HIGH
E (or close)
I Space in
MID
G the mouth
H LOW
T (or open)
The tongue can move in the space in the mouth forming an arch
in front, in the centre, and at the back. It can either be low or in
the middle or high in the mouth. Let us see some vowels and
classify them according to where the arch is formed.
/i/ While saying ee as in beet or seat (written as /i/) the tongue
is up near the roof the mouth. The arch is made in the front
of the mouth. Thus it is a high front vowel.
/u/ When you say oo as in boot or suit (written as /u/) the tongue
is high but the arch is made in the back of the mouth. Thus,
it is a high back vowel.
/ɑ/ When you say /ɑ/ as in father the arch is still made at the
back of the mouth but now the tongue is down in the mouth.
This is a low back vowel.
You may not be able to feel the movement of the tongue in the
beginning but if you say /i/ and then /æ/ you will feel your lower
jaw moving down. This means your tongue is going down in the
mouth.
While saying /u/ or /ɑ/ you will notice the tongue contracting.
This will be because it will be arched at the back of the mouth.
Produce these sounds while noticing your tongue and its movement,
and you will understand the meaning of height and frontness.
Cardinal vowels
The space of the mouth has an irregular shape but, for the sake of
simplicity, let us imagine it to be a rectangle as given below. Let
us also imagine the position the tongue can occupy and where the
arch can be made. There are, of course, many such positions but
the extreme ones are as follows. We call them the cardinal positions
and the vowels produced there are called the cardinal vowels.
Diagram 4
Front Central Back
Close і У ɨ ʉ ɯ u
IY Ʊ
e ø ǝ ɵ ɤ o
Close-mid
ǝ
Open-mid ɛ œ ɜ ɞ ʌ ɔ
æ ɐ
Open a ɶ ɑ ɒ
The first eight cardinal vowels which you will generally come across
are given below:
1. /i/ high front (ee as in beet, seat, keep)
2. /e/ half high front (as in Urdu-Hindi /pet/, i.e. stomach)
3. /ɛ/ half low front (The English vowel in egg is close to it)
4. /a/ low front unrounded vowel (/æ/as in man is used
in its place)
5. /ɑ/ low back unrounded vowel as in arms.
6. /ɔ/ half low back (as in British English saw /sɔ:/
7. /o/ half high back (as in Hindi–Urdu i.e. people /log/
or Scottish /bot/ for boat).
8. /u/ high back (oo as in boot, loot, root etc. are close to
it. The German long /u/ is even closer).
These are the first eight cardinal vowels. The other eight are
produced by changing the shape of the lips while everything else
remains the same. Thus, if we round the lips for the first four
cardinal vowels we get /У/, /ø/, /œ/, /ɶ/. If we open the lips for
cardinal 5 /a/ we get cardinal 13 which is /ɒ/ as in dog. If we open
the lips (unround them) for the other three back vowels we get /ʌ/,
/ɤ/1, and /ɯ/ which are cardinals 14, 15, and 16. Diagram 4
illustrates this clearly.
The cardinal vowels are extreme positions used to measure other
vowels. They do not exist—or at least all of them do not—in their
pure forms in most languages but the vowels which are like them
are given the symbols designed for them. The vowel sounds in the
brackets are like cardinal vowels. However, they are actually
somewhat different than the sounds we use.
The vowels used in English and most other languages used in
Pakistan are given below. You will notice that they are not all
cardinal vowels or even near them. Thus new symbols are required
for them. However, linguists often use the same symbols as they
use for cardinal vowels. This can be a source of confusion so you
should note the similarities and the differences.
/ı/ Front high vowel but lower than cardinal vowel 1. As in hit,
did, bid etc. Note that the /ı/ does not have a dot on it.
/e/ As in bed. Note that we use this symbol for cardinal vowel 2
though this vowel is between cardinal vowels 2 and 3. Some
people use the symbol /ɛ/ for it.
/æ/ Front low. It is slightly higher than cardinal vowel 4. As in
sad, bad.
/ʌ/ a central back unrounded vowel near cardinal vowel 14 as in
English cup, but.
/ɑ/ A short low back rounded vowel as in dog, hot etc. Cardinal
vowel 13.
1
Remember that the symbol of the uvular voiced fricative is a bigger /ɣ/ than
that for cardinal vowel 15 which is a small /ɤ/
/Ʊ/ A short semi-high back rounded vowel between cardinal

vowel 7 and 8. As in put, look etc.
/ǝ/ Mid central unrounded vowel as in English again. It is
sometimes undistinguishable in speech from /ʌ/. It is called
schwa. For details see Pullum and Ladusaw (1986: 44).
Diphthongs
Diphthongs may be defined as combinations of two vowels. The
tongue shifts from pronouncing one to the other and in representing
them we combine the symbols of the two vowels. The following
diphthongs are used in English:
/ıǝ/ as in beard
/eǝ/ as in scarce
/ʊǝ/ as in tour
/eı/ as in pain, paid
/aı/ as in tide, time
/ɔı/ as in voice, void
/ǝʊ/ as in home, road
/aʊ/ as in gown, house
It should be noted that for /eı/ South Asian speakers generally use
cardinal vowel 2 /e/. For example, in Urdu-Hindi we say /pe:n/
and not /pein/ for pain (Remember /e/ is lengthened here. It is the
IPA cardinal vowel 2, not the English vowel /e/).
Similarly we use /o/ for /ǝʊ/ Thus we say /hom/ and /lod/ for
/hǝʊm/ and /lǝʊd/ when we pronounce the English words home
and load.
Transcription
Now that we have learned the symbols for both vowels and
consonants, we can write down the sounds of any language. This
is called transcription. Transcription is of two kinds:
(1) Broad or phonemic (2) Narrow or phonetic

In broad transcription we just write down the main phoneme
generally used by speakers. We do not bother about its varieties or
the minor changes the speaker may have made when pronouncing
it. This kind of transcription is in slashes (/ /).
In narrow transcription the exact sounds are produced. This
transcription is put in square brackets, i.e. [ ].
Let us see both kinds of transcriptions of the words pen, ten, and
car.
Broad
/pen/
/ten/
/kar/
Here the actual way in which British or Pakistani/Indian speakers
pronounce these words is not given, just the phonemes in them are
represented.
Narrow
South Asians British
[pen] [phen]
[ten] [then]
[kar] [kha]
Notice that the South Asian speakers use the retroflex /ʈ / while the
British use alveolar /t/. Similarly, we use /p/, /k/, and /t/ instead
/ph/, /kh/, and /th/.
Narrow transcription is more difficult than broad transcription
because it is not easy to catch and represent every little difference
a speaker may give to his or her production of sounds.
Anthropological linguists and sociolinguists should learn to
transcribe sounds as they are spoken.
4
The Study of Sound Patterns
T he study of the arrangement or the combination of sounds in

a language is called phonology. As such an arrangement may
be called a pattern, phonology may be defined as the study of the
sound-pattern of a language.
Distinctive features
To study phonology we shall have to study sounds in a little more

detail. We defined phonemes as the units of sounds earlier but even
phonemes can be seen in terms of certain oppositions. For instance,
they may be voiced (vocal cords will vibrate) or unvoiced (vocal
cords will not vibrate); they may be consonantal (air will be
obstructed) or vocalic (air will not be obstructed); and so on. These
oppositions are called binary oppositions. Either a characteristic
is there or it is not there. When it is there we may use the sign of
+, and when it is not there we may use the sign of –. Thus we say
that a phoneme may be either + voice (i.e. voiced) or – voice
(unvoiced) and so on.
These features or characteristics of phonemes distinguish them
from each other. Thus they are called distinctive features.
According to Roman Jokobson and Morris Halle, two great
linguists, there are ‘twelve basic oppositions, out of which each
language makes its own selection’ (Jakobson and Halle 1956: 40).
Let us have a look at some of the major binary oppositions in
languages.
The Study of Sound Patterns 45
1. Consonantal/non-consonantal [+ cons]
Consonants, or consonantal phonemes, are produced with the air
stream being obstructed in the mouth.
2. Vocalic/non-vocalic [+ voc]
Vocalics are produced without the air stream being obstructed in
the mouth. However, there are some sounds which are not pure
vowels but have some of the characteristics of vowels.
3. Voiced/unvoiced [+ voice]
In voiced sounds the vocal cords vibrate, in unvoiced ones they do
not.
4. Anterior/posterior [+ ant]
Anterior sounds are produced in the front part of the mouth, i.e.
at or in front of the alveolar ridge. Posterior sounds are produced
at the back of the mouth.
5. Coronal/non-coronal [+ cor]
Coronal sounds are produced by raising the blade of the tongue
towards the teeth or the hard palate. Non-coronals are produced
without this action.
6. Nasal/oral [+ nas]
Nasal sounds are produced by the air passing through the nose
while the lips are closed. In oral sounds the air passes through the
mouth.
7. Continuant/stop [+ cont]
Sounds in which the air stream continues to flow are called
continuants. In stops (or plosives) the air stops and then explodes.
In continuants like /f/ or /v/ the air goes out with friction, because
it is impeded but it is never stopped as in stops like /p/ or /b/.
8. Strident/non-strident [+ strid]
In strident sounds the air strikes two surfaces and a lot of noise is
produced. In non-strident sounds the noise is less. Some of the
stridents are /s/, /z/, /f/, and /v/ which produce high-intensity noise
or friction.
9. Low/non-low [+ low]
In low sounds the tongue is low—right at the bottom—of the
mouth. The tongue is in this position in all low vowels and in
pharyngeals. All other sounds are non-low sounds.
10. Tense/lax [+ tense]

Only vowels are classified as tense or lax. It is said that in tense
vowels the muscles of the tongue are more tense. The body of the
tongue or its root gives the air less space to flow than in lax vowels.
The root of the tongue may be drawn forward and its body may
be raised. This is called advanced tongue root [+ ATR]. Some
linguists consider +ATR same as + tense. It should be noted that
tense vowels are longer than lax ones. For instance, /i/ is long and
/ı/ is short so /i/ is + tense and /ı/ is – tense.
11. High/non-high [+ high]

The body of the tongue is high in the mouth (i.e. towards the
palate) for high sounds. For non-high ones it is not.
12. Back/non-back [+ back]

When the hump made by the tongue is in the back part of the
mouth, the sounds produced are called + back sounds. When it is
not, we produce other sounds, i.e. – back sounds.
Now we have the means of classifying all the sounds of all languages
through these distinctive features. The +/– or binary oppositions
give us features which will allow us to understand the rules
according to which sounds combine in different languages. Let us
first, however, classify some of the sounds of English here. This
classification (Figure 3) is different from the one used by Jacobson

and Halle (1956).
FIGURE 3
CHART OF DISTINCTIVE FEATURES
p b f v m t d θ ð s z n tʃ dʒ ʃ ʒ k
Cons + + + + + + + + + + + + + + + + +
Voc – – – – – – – – – – – – – – – – –
Cont – – + + – – – + + + + – – – + + –
Cor – – – – – + + + + + + + + + + + –
Voice – + – + + – + – + – + + – + – + –
Nasal – – – – + – – – – – – + – – – – –
Strid – – + + – – – – – + + – + + + + –
Ant + + + + + – – + + + + + – – – – –
g l r w j h ŋ x ɣ q ʕ ħ
Cons + + + – – – + + + + – –
Voc – + + – – – – – – – – –
Cont – + + + + + – + + – + +
Cor – + + – – – – – – – – –
Voice + + + + + – + – + – + –
Nasal – – – – – – + – – – – –
Strid – – – – – – – – – – – –
Ant – + + + – – – – – – – –
Lat – + – – – – – – – – – –
Some linguists use other features for classification. For instance,

they contrast sounds like /s/ with /z/ on the basis of the pressure
exerted by the lungs. For /s/ the lungs push harder and the sound
produced is sharper and stronger. It is called a fortis sound. In
contrast /z/ is a lenis sound. Yet another way is to classify sounds
is according to the part of the tongue which is used to produce
them, i.e. apical (tip of the tongue); dorsal (tongue-back
articulation). We may also classify them as being labial (produced
by the lips or non-labial [±labial]).
Let us now classify some sounds with reference to these classificatory

criteria:
p b m t d n k g ŋ
stop + + + + + + + + +
nasal – – + – – + – – +
fortis + – + – + –
lenis – + – + – +
labial + + + – – – – – –
apical – – – + + + – – –
dorsal – – – – – – + + +
Source: O’ Connor 1973: 204
With the help of these distinctive features we can describe sounds

without even using IPA symbols. Even more significant is the fact
that we can write down laws or rules about the combination of
sounds in a language or its sound pattern. These laws or rules are
called phonological rules.
Phonological rules
Let us now write down some phonological rules. They are written
in the form of certain conventional symbols which are similar to
those used in mathematics. You must learn the following
symbols:
becomes
ø nothing (or zero)
# word boundary
#– word-initial
– # word-final
/ in the environment of
[ ] Square brackets enclose the distinctive features of a
phoneme.
— in this place.
Most of the symbols are self evident but the symbols for word
boundary and ‘in the environment of ’ need explanation.
Words end and begin with phonemes. Both these phonemes are
boundaries. When we say word-initial we mean that which comes
before the first phoneme of a word begins. When we say word-final
we mean after the last phoneme of the word ends. Consider the
word below:
d i g /d ı g/
The phoneme /d/ is in the word-initial position and /g/ is in the
word-final position. All that comes before /d/ is word-initial and is
written as #–. All that comes after /g/, i.e. word-finally is written
as –# (i.e. at the boundary after the last phoneme ends).
# dıg# #– dıg–#
Now look at the symbol of ‘in the environment of’ which is /. It
means in such and such conditions; when the following given
conditions exist; when these sounds exist in this given pattern or
occur in this manner. Then the sounds themselves are written in
the sequence in which they occur.
Let us now write down a rule. First let us have an observation (an
event, a phenomenon) which the rule will explain.
Observation: Speakers of Urdu-Hindi put an /ı/ before words like
school, stool, spool, speak, street etc.
British English Urdu-Hindi Speakers
/skul/ /ıskul/
/stul/ /ıstul/
/spul/ /ıspul/
/spik/ /ıspik/
/strit/ /ıstrit/
Objective: To make a phonological rule to express the above, i.e.
to find out what rule the speakers of Urdu-Hindi have in their
minds which makes them put an /ı/ before the words given
above.
Method: Notice that the first phoneme is the same in all the words,
i.e. /s/. The second is different, i.e. /k/, /t/, /p/ etc. let us write
down the distinctive features of all of them:
s k t p
Cons + + + +
Voc – – – –
Voice – – – –
Ant + – + +
Cor + – + –
Cont + – – –
Nasal – – – –
Strid + – – –
We should now look for other examples of phonemes of the same

kind and write them down in our list.
You will observe that the only relevant difference is that the first
phoneme is always cont + while the second one is cont– (i.e. a
stop).
This means that Urdu-Hindi speakers add /ı/ in the word-initial
position to all words where the consonantal cluster begins with a
+ cont while the second phoneme is a –cont.
Let us put it down through conventional symbols:
Ø /ı/
This means that ‘nothing becomes /ı/’. In other words the speaker
puts /ı/ or pronounces /ı/ out of nothing. There was no /ı/ but the
speaker pronounces it or inserts it out of nowhere.
The environment is that after the words begin (# –) there is a cont
followed by – cont consonantal sound. In such conditions the
speaker will pronounce an /ı/ before the word begins. The exact
place where this /ı/ will appear is given by the dash (–) sign.
The rule is given as follows:
ø /ı/ ∕ # -[+cont][- cont]
But now take the pronunciation of the word sphere /s f ı ə r/.
Hindi-Urdu speakers pronounce it as /ı s f ı ə r/. In other words
our rule does not cover all the possible cases. We need to refine it
further. This refinement is as under.
Ø /ı/ ∕ # – [+cons] [+cons]
This means that a vowel is inserted by the speakers of Urdu and
Hindi before consonantal clusters if such clusters occur in the
beginning of words.
In the same words Punjabi speakers put the vowel /ə/ between the
first and the second phoneme of such a consonantal cluster. You
can work out the rule for these speakers, and for details of such
rules in Pakistani and Indian English, see Rahman (1990: 31–
32).
Allophones
Besides the sounds given in the chart, most South Asian languages
also have aspirates. These are /bh/, /ph/, /dh/, /th/, /dh/, /th/, /kh/,
and /gh/ etc. they are used in Urdu-Hindi words like bhalu (bear),
phool (flower) etc.
In English some phonemes such as /p/, /k/, and /t/ are aspirated
in the beginning of words but the aspiration in this case is less than
in the case of Hindi-Urdu. Another important point is that in
English the meaning remains the same even if you do not aspirate
the phonemes e.g. king has the same meaning whether you
pronounce it as /kıŋ/ or /khıŋ/ In Urdu-Hindi, Punjabi, and Sindhi
the meaning changes according to aspiration, e.g. /pul/ means a
bridge in Punjabi but /phul/ means a flower.
In English /k/ and /kh/, and /p/ and /ph/ are allophones. In Hindi-
Urdu /k/ and /kh/ or /p/ and /ph/ are different phonemes. If two
sounds create two meanings, they are different phonemes in that
language. The words they make are minimal pairs. If they do not
create different meanings, they are allophones.
Phonology gives the rules according to which sounds are organized
in a language and you will find it a highly interesting study. It is
far more technical than this brief chapter can tell you so be
prepared to devote serious attention to it if you want to master
it.
5
How Words are Made
T o study how words are made is called morphology. Morphe

means shape in Greek and logos refers to knowledge, the study
of something etc. Thus morphology is the study of the shapes of
words, i.e. how words are formed in a language. Words are the
building blocks of the house of language. For a modest dwelling in
the English language one needs about 13,000 words, for a big
house about 60,000, and if you want a palace then you would need
over 120,000 of them (Akamajian et. al. 2001: 11). But people
who learn many words, whether to read the works of Shakespeare
or play scrabble, do not learn how these words are formed. Your
aim as a student of linguistics is to find out precisely this.
Morpheme
A morpheme is a unit of meaning which makes words. Just as a
phoneme is a unit of sound in a language, we use the term
morpheme for those parts of words which mean something on their
own or when they are used with other morphemes.
Look at the following examples:
Words Morphemes making the words
boy -boy
man -man
unman -un -man
unmanly -un -man -ly
unmanliness -un -man -li -ness
boyish -boy -ish
Morphemes like –boy and –man, which have meanings on their

own, are called FREE morphemes. They can stand on their own
without being combined with other morphemes.
Morphemes like -ish, -un, -ly (or -li) have no meaning on their
own. They do indicate some concept but are not used alone. They
are called bound morphemes as they are used with other
morphemes. They are therefore bound with other morphemes to
give meaning.
As you can see, long words are made up of free and bound
morphemes. There is a well known game in which inordinately
long words are made up in this way. Here is an example:
HYPERPOLYSYLLABICSESQUIPEDALIANISM
Here are the morphemes used for making the word: poly means
many and hyper means great or big. Sesquipedal means being one
and a half feet long. It is itself made up of other morphemes, i.e.
ped (foot). You can find out the meaning of the other morphemes
and come to the conclusion that the word means: THE HABIT
OF USING ONE AND A HALF FEET LONG WORDS!
Classification of morphemes
Let us see how linguists classify morphemes. Besides the major

categories of free and bound morphemes given above, linguists use
the following classification:
Morphemes
Bound Free
Derivational Inflectional [never Lexical Functional

[will derive a changes grammatical [lexical items [do, ko in
new word] category of word] boy, girl] Urdu]
Pure derivational Affixes [Generally do not

[New word is in a change grammatical
different grammatical category of new word]
class]
Suffixes Infixes Prefixes

[Used after a word] [Used between a word] [Used before a word]
How Words are Made 55
Free morphemes
Free morphemes are lexical items or commonly used words like

boy, girl, horse etc. They mean something on their own but cannot
be broken down into lesser units of meaning.
The functional free morphemes can be used on their own but they
serve some function. They may give a relationship or refer to
something or be used as fillers.
Let us look at the sentences below.
(1) The door is in the house.
(2) Open the door.
(3) /dervaze ko kholo/.
(4) /bua kholo/.
(5) /dervaza kholo/.
(6) Do open the door.
In sentence (1) in serves the function of indicating a location. In
sentence (2) the indicates a particular door. In the Urdu-Hindi
sentence (3) ko is a filler. It has no equivalent in English. The
Punjabi sentence (4) as well as the Urdu-Hindi one (5) say the
same thing without using ko. Similarly do in sentence (6) is an
empty filler. The sentence would give the sense if do is omitted but
the urgency or the emphasis would be missing. So fillers are not
always empty of all meaning. They may also convey emotion and
attitude. All these morphemes are called functional free
morphemes.
Other examples of functional morphemes are:
Articles the, a
Demonstratives this, that
Conjunctions and, or
Quantifiers all, most, some, few
Prepositions to, from, at, with
Bound morphemes
Derivational morphemes derive, or make, a new word. The pure

derivational morphemes derive a word which is always in a different
grammatical class from the original one. Let us take the common
morphemes -able and -ish and add them to words.
eat + able = eatable [eat is a verb but eatable is an adjective or a
noun]
boy + ish = boyish [boy is a noun but boyish is an adjective]
Here the newly derived words are in a different grammatical
category from the original ones.
Affixes too derive new words but generally do not change the
grammatical categories of the original ones. To affix means to
attach. If the affix is attached before the word it is called a prefix.
If it is attached after a word it is called a suffix. If it is attached
inside a word, it is called an infix. English and most Pakistani
languages do not have infixes. However, Brahvi—a language of
Balochistan—has infixes.
Here are some suffixes, prefixes, and infixes:
a-, auto-, re-
a + moral = amoral [not concerned with morality.
Both are adjectives]
auto + biography = autobiography [both are nouns]
re + print = reprint [both are verbs]
/pǝ/ in Brahvi is an infix meaning not. Its use is as follows:
/xǝlniŋ/ to beat
/xǝlpǝniŋ/ not to beat.
In these cases the grammatical class of the new words remains
unchanged.
Inflectional morphemes never change the grammatical category of

the words to which they are attached. They are added according to
the grammatical rules of the language and are used mostly in
inflectional languages (i.e. languages in which meaning is expressed
by changing some morphemes after the main roots of words).
The past tense morpheme –ed in English is used in many verbs:
jump + ed = jumped; dance + ed = danced; and save + ed = saved.
However, there are other ways of expressing the past in English
also. Latin, on the other hand, is much more inflectional than
English.
The morpheme –’s (as in the girl’s) behaves both like an inflectional
and a derivational morpheme. Thus there are problems with all
classifications which must be kept in mind.
Neologism
The process of creating new words is called neologism. This is
sometimes a political activity in the sense that the morphemes (or
words) chosen are selected to emphasize identity (ethnic, religious
or nationalistic) rather than comprehensibility. For instance, we
find borrowings from Sanskrit in modern Hindi, and from Persian
and Arabic in modern Urdu. These do not make those who use
these styles of language intelligible to ordinary people but that is
not the aim of those who have put in these words. Their aim is to
emphasize identity. Sanskritized Hindi is symbolic of the Hindu
identity while Persianized Urdu stands for the Muslim identity.
Indeed, one major reason why these two styles of one language (old
Hindi or Hindvi) drifted apart from the 18th century onwards is
because of this politicized neologism. If you want details of this
process see Amrit Rai’s excellent book A House Divided (1994). If
you want to read about neologism in Pakistan by those who express
their ethnic identities through coining words rooted in the ancient
forms of their languages see the present author’s books Rahman
(1996) and Rahman (1999).
But not all neologism is politically inspired. Adolescent slang in

America uses the word geek for an ‘intellectual’. This word will be
accepted in the language one day as new words often are. In other
cases, new things are created and we need words for them.
Sometimes old words are given new meanings and sometimes new
ones are created. For example:
Radar radio detecting and ranging
Laser light amplification by stimulated emission of radiation
URL uniform resource locator
DOS disc operating system
HTML hypertext markup language
Words are also made by blending them such as brunch (breakfast
and lunch), motel (motor and hotel), and edutainment (education and
entertainment), or they may be borrowed from another language.
Kindergarten from German means the garden of children (i.e
children’s class); croissant is from French; avocado and guava are
both from the Aztec language, and so on. Of course Hindi-Urdu
have a lot of borrowing from Persian, Arabic, Turkish, Portuguese,
and English on an Indic base which has a large Sanskritic base
mixed with some other languages of South Asia.
Allomorphs (or morphs)
The morpheme is written in a fixed way. However, its actual

pronunciation is not the same in all situations. These different, or
variant, forms of morphemes are known as allomorphs or
morphs.
Let us take the English plural morpheme (which is inflectional) -s.
It is written as -s in all contexts but can be pronounced /s/, /z/ or
/ız/ according to the sounds which come before it. See the words
given below:
Words Plural morpheme Actual Pronunciation

(1) cat -s /kæts/
(2) dog -s /dɒgz/
(3) bus -s /bʌsız/
(4) sheep -s /ʃip/
As you can see, the morpheme is conventionally written as -s in all

cases. However, it has three allomorphs /s/, /z/, and /ız/. They are
used according to the rules of pronunciation, or phonological rules,
in our minds. When the word ends on an unvoiced phoneme such
as /t/, the allomorph /s/, which is unvoiced, is added. When the
word ends on a voiced phoneme such as /g/, the voiced allomorph
/z/ is added. When the ending is a sibilant, the kind of fricative
in which the hissing noise is very sharp, the allomorph used is /ız/.
When the singular and the plural is the same as in sheep, we may
say that a zero allomorph has been used, i.e. it has no
pronunciation.
These rules are part of the language system and speakers are not
consciously aware of them. Because we want to study language in
a systematic manner, we create names for these different ways of
pronouncing an ending which is written in the same way.
It is because we want to create uniform rules that the concept of
the zero morph is used. It is merely another way of saying that
there is an allomorph which has no pronunciation. This allows us
to add the morpheme -s before all the words of English no matter
how they are actually pronounced.
This desire for using one conventional symbol for pluralization (i.e.
-s in English) can lead to complications. For instance, look at child
– children; mouse – mice; ox – oxen; man – men and foot – feet. We
may say:
child + s = children
mouse + s = mice
ox + s = oxen
man + s = men
foot + s = feet
This creates regularity as far as our use of the conventional symbol
for the morpheme -s is concerned. But the pronunciation is so
different that we have to give special rules of how the allomorphs
are derived. The study of the sounds of morphemes involves, as we
have seen, both phonological and morphological rules. It is
therefore called morphophonemics or morphophonology. In
such cases, complicated morphophonemic rules are used to
represent the derivation of words.
Morphophoneme
Some linguists believe that it is useful to think of the variant
pronunciations of a morpheme as a morphophoneme. Taking the
example of the pluralization morpheme in English again:
Morpheme Allomorphs Morphophoneme
/s/
/z/
-s /ız/ /z/
/ø/
The morphophoneme is an abstract concept like the morpheme
itself. The actual phonemes which give the plural sounds are the
allomorphs. However, the morphophoneme can be used as a
symbol. It can be written to indicate the class of all the allomorphs.
Then we can write the rules which will give its actual phonetic
value in different environments (i.e. the way it is pronounced after
a voiced ending or an unvoiced ending etc.)
To conclude, morphology is an important aspect of a language. It

can be seen as a process or an arrangement. If it is seen as a process,
a form is taken as a basic or underlying form, and other forms are
derived from it according to certain processes. If it is taken as an
arrangement of items (morphemes), the rules for this arrangement
are studied. Nowadays, most linguists use both these concepts to
study the shapes of words in different languages. For more details,
read books on morphology such as Matthews (1974).
6
The Study of Meaning
T he study of meaning is called semantics. Semaino means ‘to

mean’ or ‘to signify’ in Greek, and –ics, like –logos, is used to
indicate that this is a branch of learning. Since meaning is so
crucial, not only linguists but philosophers are also interested in it.
In fact philosophers have been much more concerned with it in
the past than linguists.
Our main concern in this introductory book is with the way
linguists study meaning. As it is a complex subject we shall attempt
nothing more than to explain some of the terms used in the subject.
After that we shall touch upon those aspects of it which will be of
use for further study.
Some basic terms
Most people talk of the meaning of words. However, the term

word is rather inexact. Let us examine the concept die. It can be
expressed a follows:
(1) die He died.
(2) to lose life He lost his life.
(3) to pass away He passed away.
(4) to kick the bucket He kicked the bucket.
As you can observe, not one word but two are used in examples
(2) and (3), i.e. lose and life; pass and away. In example (4) three
words are used kick, the, bucket. Moreover, die and died are the
same concept but expressed in two variant forms.
The Study of Meaning 63
In short, the underlying concept is one but more than one word,
as well as variants of a word, are used for expressing it. If we want
to talk about the underlying concept alone, we should use some
other term and not the term word for it. Linguists use the term
lexeme to mean a unit of meaning. Some people use the term
lexical items for the same concept. In this book, the term lexeme
is used for a unit of meaning. Remember the term morpheme does
not mean a single concept. It is that which makes what we call
words. Thus the idiom pass away has words made up of several
morphemes but it expresses only one concept—‘to die’ (to cease to
remain alive).
The other terms we shall come across are: synonymy, antonymy,
hyponymy, polysemy, homonymy etc.
Synonymy refers to the sameness of meaning. A dictionary
generally gives you synonyms. For instance, the synonyms of
beautiful are pretty, handsome, good looking, personable, beauteous,
comely, pulchritudinous etc. However, if we use them without
reference to what or whom we are talking about, we shall create
misunderstandings. See the sentences in columns A and B.
A B
My horse is personable. My horse is beautiful.
This man is pretty. This man is handsome.
That boy is beauteous. That boy is good looking.
The heroine is pulchritudinous. The heroine is pretty.
The scene is comely. The scene is beautiful.
The sentences in column A do not use the synonyms in their

precise shade of meaning. For instance, if a man or boy is called
pretty, it would indicate that they are effeminate. Pulchritude is used
by philosophers, and if we use it for a woman, we would appear
unnatural.
However, the underlying concept is the same and the precise use
is determined by other factors.
Antonymy refers to opposites. The antonyms of good is bad, fat is

thin, beautiful is ugly, and so on.
Hyponymy refers to the notion of inclusion. For example, when
we say animal we include tigers, cats, and horses under this main
category. The word at the top (i.e. animal) is the superordinate
term, whereas the words under it are co-hyponyms.
Animal
bird fish insect tiger human beings
Remember that your world view will be reflected in the way your
language classifies the world. People who believe that animals and
human beings belong to different categories will not put human
under the superordinate term animal. For them living creatures will
be the superordinate term:
Living creatures
animal human plant
man woman
Societies may see fire as living and plants as dead. Thus, reality is
something we construct and language helps us in expressing it.
These are ideas which you will find in the chapters on
anthropological linguistics etc.
Polysemy refers to a lexeme having more than one meaning (poly
= many). Thus chip in English means a piece of fried potato, an
electronic device, and a piece of wood. We use the same sounds to
refer to different concepts.
Homonymy refers to two or more lexemes having the same shape.
Bank, for example, is a verb meaning ‘to rely upon’. It is also a
place where money is kept as well as the side of a river on which
one can walk. The context makes it clear in the example given
below:
I bank upon the Habib Bank situated on the left bank of the river
Ravi to give me a loan.
Homophone refers to words having the same sound but different
spelling (threw and through).
Homograph refers to words having the same spelling but different
sounds. Wind means air movement and is a noun. It also means to
turn into one direction; to finish, i.e. wind the clock; wind up your
work etc. The pronunciation is different.
wind (n) — /w ı n d/
wind (v) — /w aı n d/
Apart from these words you will also encounter words like deixis.
These are lexemes which can be understood only with reference to
the speaker’s position in space or time. Actually deixis comes from
a Greek word for pointing. Thus deictic forms point out. Personal
deictics are you and I. They point out people. Spatial deictics are
this/that; here/there; come/go; bring/take. They point out where the
speaker is in space in relation to others. What is near the speaker
is here, what is far is there, and so on. Temporal deixis refer to
time. They are now, yesterday, and tomorrow, as well as then and
now. They also mark tenses. They tell you how time is perceived
by the speaker.
Semantic components
You have seen how the concept of binary opposition helped us to

divide a sound into its features, i.e. if it was a consonant we called
it + cons – voc. If it was voiced we called it + voice, and so on. In
the same way, some lexemes too can be seen as being made up of
binary distinctions.
In short, a boy would be:
+ human
+ animate
boy
– adult
+ male
but what about a dead boy? We would have to make the term
animate minus i.e. – animate. As David Crystal points out,
meaning is so complex that such an analysis is not always possible
(1988: 107). However, this concept is used in componential
analysis which you will study later.
The problems of semantics
Although we have explained some of all the terms given above, we

do not have a semantic theory on which everyone can agree. Part
of the problem is that the term meaning is far from clear. In fact
C. K. Ogden and I. A. Richards published a book in 1923 which
had the title The Meaning of Meaning. It had about twenty two
definitions of meaning (1923: 186–87) but even this list is not
exhaustive. Much later Chomsky said that ‘meaning’ tends to be
used as a catch-all term to include every aspect of language that we
know very little about (1957: 103).
Later Katz and Fodor (1964) tried to present a semantic theory
which could fit in with Chomsky’s concept of generative grammar.
According to them a native speaker has a dictionary of lexical items
of his or her language in the mind and a system of rules as to where
they will be used. These rules tell the speaker where to put in a
word. They are called projection rules. For instance, unless we are
talking about the world of make-believe, we cannot say:
* The stone talked to me.
The sentence is syntactically perfect but in the world we live in
stones do not talk. Thus the noun stone has different properties
from the noun girl (who can talk). This is how the semantic
properties of nouns and verbs are projected onto the syntax and,
in a sense, pre-determine the structure of a sentence.
This is how Katz and Fodor define the entry for the word play:
1. To move swiftly,
erratically; to flutter; to
verb dart to and fro; to vibrate.
2. To perform an activity for

verb pleasure.
intr
play To perform in the character
or part of; to act.
The stage representation of a

noun drama; a performance.
Every item is marked for its syntactic category i.e. is it a noun, verb,
preposition etc. These categories are called grammatical markers.
The sense it conveys, or the semantic label we put it under, is called
the semantic marker. The way its meaning, or what it refers to, is
distinguished from its other senses is called its distinguisher. Here
is the entry for bachelor.
bachelor
human noun animal
(male)
(male) [male or female who has
the lowest academic
degree of a university] [young male seal when
without a mate in the
[who has breeding season]
never married]
[young knight serving
under the standard of
another knight]
Adapted from Katz and Fodor 1963: 496.

Here, the grammatical marker is noun, the semantic markers are

human, animal, male etc., and the distinguishers are the meanings
given in square brackets.
Initially, as you have seen, Chomsky concentrated on syntax. But
later it became clear that semantics has a close relationship with
syntax. The theory of generative semantics denies that syntax is
separate from semantics. Thus, concepts determine syntax at the
deep level and are not inserted in a given structure later. These
theories are by no means final. Research on language is going on
and new theories are being proposed. If you are interested in the
subject you should keep abreast with them.
Meaning in philosophy
Although meaning has been discussed since time immemorial by

philosophers—see Plato’s Cratylus—it became a focus of philosophy
in the twentieth century. The German mathematician Gottlob
Frege and the English philosopher Bertrand Russell were both
interested in it. However, their major interest was in a logical
language (such as the symbols used in mathematics) and not in
natural human languages.
The major problem of philosophical linguists was whether the
words we use refer to something. The word Islamabad, for instance,
refers to a city in Pakistan. But all words do not have such referents
in the world. For instance, what do words like love, ask, consistent,
justice etc. refer to in the world? We have some idea or thought in
our mind, and our word (called symbol below) refers to the ‘thing’
or ‘concept’ (referent) through our mind. Ogden and Richards
gave the following triangle for this concept:
Thought
Symbol— — — — — — — — — — — Referent
(Source: Ogden and Richards 1923: 99)

Philosophers also tried to understand meaning with reference to

truth value i.e. is a sentence true or not? The Polish-born
mathematician Alfred Tarski defined Truth according to the
following formula:
X is true iff P (iff means if and only if).
X = sentence (i.e. description or statement).
iff = if and only if.
P = a variable telling us about something in the world.
X P
Coal is black (iff coal is black)
But in natural languages curses, blessings, orders, requests may
neither be true nor false.
Some philosophers, therefore, talk about the use of words. The
philosopher Ludwig Wittgenstein called language a game and
compared a word to a piece used in a game of chess. This will be
explained in a little more detail below.
First, however, let us briefly describe the life of Ludwig Wittgenstein
whose theories had such a profound influence on some aspects of
linguistics. Born in Vienna on 26 April 1889, he studied
mathematics and natural sciences in Vienna from 1903–1906.
From 1906 to 1911 he studied Mechanical Engineering in Berlin
and Manchester
The most important event of his life from the point of view of his
philosophical career is his arrival in 1911 to Cambridge to attend
Bertrand Russell’s lectures in mathematical logic. But Wittgenstein
soon got bored of Cambridge and also developed differences with
Russell, so he moved to Skjolden, Norway in order to pursue his
reflections on the philosophy of language. However, when the First
World War started in 1914 he abandoned philosophy to join the
Austrian army thus becoming Britain’s enemy. In 1918 as a
prisoner of war he kept working on his book later published as
Tractatus (Wittgenstein 1922) which he sent to Russell.
Being the eccentric he was, he now resolved to become a school

teacher and that too in a primary school in rural Austria. This
position he filled from 1920 till 1925 when he was thrown out for
having boxed a pupil’s ears till he bled. For some time he was an
equally unsuccessful architect too but this too came to nothing.
Oddly enough it was not wealth he was after because he refused
his share in his father’s considerable business fortune giving
everything to his sisters. Even more unexpectedly this failed school
teacher now wrote to his Cambridge friends that he was coming
over to the University and that they should provide him with
means of subsistence. They did, and in 1929 Wittgenstein settled
down in Trinity College Cambridge and submitted Tractatus for a
doctorate. The account of his telling Russell and G.E. Moore, the
occupant of the chair of philosophy at Cambridge, is so hilarious
that you should not miss reading it in his biography. In 1939 he
succeeded Moore as the Professor of Philosophy at the University
of Cambridge and it was only because he himself resigned in 1947
that he ceased to hold that exalted academic position. Wittgenstein
never married and had no house of his own, but he was lucky in
his friends because till his death on 29 April 1951 he never lacked
a roof and somebody to take care of him (Malcolm 1958).
It was this man who influenced linguistics, especially pragmatics,
through his books. In Tractatus his main focus is on the truth or
falsity of sentences. He says that everyday language is an impure
vehicle of thought. The basic assumption is that reality does exist
and thought is the interpretation of this reality which is expressed
through language. Thus, ‘the limits of my language are the limits
of my world’ (Tractatus 5.6). The ideal language would be logical
symbolism. However, in his later book Philosophical Investigations
published after his death in 1953, he relinquishes the search for an
ideal logical language.
In Investigations he looks at words in everyday use. Language-
games refers to rules for playing with words as we have rules for
games. When we sing we play a language game. The theory of
language-games mentioned above can be understood by looking at
the way we use a word. According to Wittgenstein ‘the meaning

of a word lies in its use’. Speakers and readers use words according
to the situation which has its conventions, rules, and norms of
appropriate behaviour. For instance, if we meet acquaintances we
ask them how they are. They reply that they are fine. The game is
to adhere to this ritual question and answer. The words in this
game mean only an exchange of noises to convey to the other
person that we acknowledge their presence. If someone actually
replies truly to the query then it would be deviation from the rules
of the game.
The idea of language-games forms the basis of pragmatics. G. Ryle,
P. F. Strawson, and J. L. Austin, who developed the idea of speech
acts, were all influenced by Wittgenstein (Battaner-Moro 2005:
Vol 2, 1179–1181).
According to John L. Austin’s theory of speech acts, saying
something was doing something. All requests, commands, promises,
warnings, declarations have illocutionary force—that which
pertains to some form or aspect of action.
This brief account of the philosopher’s concern with meaning is
not meant to be exhaustive. On the contrary we have not even
touched upon most of the theories in this field. Another fascinating
aspect of meaning is that it tells you a lot about your world
view—the way you see reality. This is an aspect of it which will be
taken up again. Meanwhile, if you want to learn more about
semantics you should refer to the books given in the bibliography.
To begin with read Leech’s book entitled Semantics (1974).
7
Introduction to Chomsky’s Theories
N oam Chomsky was born on 07 December 1928. His father

was a professor of Hebrew and Jewish Education at Grate
College, Philadelphia. Chomsky was initiated into linguistics by
Zellig Harris, (who was teaching at the University of Pennsylvania)
and completed his PhD from the same university in 1955. Since
1961 he is a full professor at the Massachusetts Institute of
Technology, one of America’s leading research universities. Since
1966, he has been Ferrari Ward professor of Linguistics, and in
1967 he was made Institute Professor of Linguistics and Philosophy
at MIT. He is now considered the most distinguished and original
linguist in the world. He is also famous as a radical intellectual.
Chomsky became a radical in politics with a left-wing orientation
during his youth. He is now famous as an articulate and vitriolic
critic of America’s neo-imperialist policies. Although brought up
in a religious Jewish family, he is also a critic of Israel. His books
on international relations, media, and philosophy are just as erudite
and scholarly as those on linguistics.
He has received many honours, including doctorates from all
around the world. In 1988 he received the Kyoto Prize in basic
science which is given for subjects not covered by the Nobel prize
committee. He has also delivered the John Locke Lectures at
Oxford University in 1969 and the Bertrand Russell Memorial
lectures at the University of Cambridge. He has also lectured in
India and Pakistan on political subjects (Lasnik and Otero 2005).
This chapter introduces some basic ideas of Chomsky in very
simple words.
Introduction to Chomsky’s Theories 73
Grammar
The rules governing the way words are placed after each other to
create sentences are dealt with in syntax. In the following three
chapters, we will mainly refer to Noam Chomsky’s theories about
syntax in particular and linguistics in general. Let us begin with a
few definitions.
First, let us define grammar. You have been studying grammar at
school. It consisted of certain rules which were meant to make you
write and speak a language correctly. Your teachers told you what
to do. They prescribed rules such as (1) Do not say ‘It is me’. Say
‘It is I’ and (2) Do not end a sentence with a proposition. This is
called prescriptive grammar.
Some rules taught by teachers are not necessary at all. For instance,
‘It is I’ is considered pedantic and people normally use ‘It is me’.
Moreover, excellent writers of English end sentences with
prepositions. These rules come from Latin grammar. As the
language of the scholars of Europe was Latin, they were so
impressed by it that they thought other languages too should follow
these rules (see Palmer 1971: 15-27). You will observe that in Urdu
many people follow the rules of Persian and Arabic. For example,
the plural of ustad (teacher) in Urdu is ustad. But some people use
the pedantic usataza which follows an Arabic pluralization rule.
Likewise modern Hindi grammar follows Sanskritic rules.
We are not talking about prescriptive grammar in this book. Most
linguists, including Chomsky, whose theories will be discussed in
this book, are interested in descriptive grammar. Like all scholars
their real interest is in how people form sentences: What rules are
there in the minds of native-speakers which enable them to create
sentences?
Like all scientists, our job is merely to describe these rules. We are
not concerned with anybody’s notions of good and bad language.
Thus all varieties of language, including the varieties of language
spoken by uneducated people, are part of our study. All languages
have grammar. In fact without having some rules for producing
sentences, they could not have been produced at all. Thus it is

meaningless to say that such and such language does not have a
grammar.
Now look at the following two sentences of English.
(1) I went home
This arrangement or order is known as syntax, and some rule in
our brain/mind tells us that the correct order is as in (1) and not
as in the following sentence.
(2) *Went home I
Obviously (1) is a sentence in English, and your mind, can only
tell you that it is correct (or well-formed) if you are a native speaker
of English or know English quite well. Given any utterance in your
own mother tongue, you will be able to judge whether it is well-
formed or not. This judgement will be largely based on intuition.
Thus we have to depend on the judgements of the native speaker
of a language to determine the ‘well-formed-ness’ or otherwise of
sentences/utterances. Since we take sentences/utterances produced
by native speakers as data when we attempt to describe a language,
it is clear that our grammar is not meant to teach anyone how to
write correctly. Our aim here is to understand how the mind works
when it creates language. This can be understood by an analogy
with computers which are fed with instructions which are processed
in order to give answers. In the same way, the human brain also
appears to be acting under certain instructions which we shall call
rules or principles. These rules process words and parts of words
to produce speech (utterances) or writing (sentences). Chomsky is
mainly interested in these rules and principles and feels that we can
understand how the mind actually works if we can find out how
it produces language. This means that Chomsky’s approach to
language is mentalist, i.e. it is based upon understanding mental
phenomena. Indeed, one of the aims of his theories is to understand
the mind itself. Chomsky himself says:
Personally I am primarily intrigued by the possibility of learning

something, from the study of language, that will bring to light inherent
properties of the human mind (Chomsky, 1972: 103).
Chomsky’s other aims
To understand the mind, however, is the ultimate goal. To attain

this goal, Chomsky wants to understand what language is. Now
language obviously has a structure, that is, some kind of systematic
connection between the ways words combine together to create
meaning. Clearly you cannot put words together in any random
order, nor can you produce any sounds you like if you want to
make sense. The whole language is like a system or structure, and
one of Chomsky’s aims is to understand this structure. In other
words, he wants to develop a theory of language structure. Such
a theory seeks to answer questions such as:
• What are natural, or human, languages?
• Are they different from artificial languages, such as those used
in computers?
• Are there any common or universal common features, and is it
possible to construct a Universal Grammar?
Along with these questions, Chomsky is also interested in the
theory of language acquisition, i.e. how small children learn the
language they hear most. Both these theories will help us understand
how the brain/mind works when it learns or produces human
language.
Particular grammar
The goals given above cannot be reached unless we have a great

deal of data at our disposal. Our data in this case is obtained from
human languages, such as English, French, Urdu, Persian, and so
on. If we describe these languages fully we shall understand the
structure of particular languages. This might help us to understand
the structure of language in general. These particular descriptions
are also called grammars. These grammars should cover phonology,
morphology, syntax, and semantics. However, most people only
talk about syntax in grammar. Notwithstanding this fact, the ideal
grammar of particular languages will describe the phonology,
morphology, syntax, and semantics of that language.
Competence and performance
Native speakers of a language, however, carry the grammar of their

particular language in their minds. They know the structure of
their language, although of course they are rarely consciously aware
of the extent of their knowledge, nor can they describe it adequately.
However, the knowledge which they have, which is assumed to
cover the whole system of the language, is called competence. Thus
particular grammars are merely reflections of the rules and
principles, as well as the lexemes of the people who speak it.
In practice however, people do not use the complete system of the
language they are speaking. They use only some words in a certain
combination, whereas, of course, they could use many other words
in longer or shorter sentences. Often, they also pause and break
sentences off in the middle. This actual use of language is quite
different from the speaker’s potential competence; it is called
performance.
Let us now summarise the two definitions in Chomsky’s own
words given in his famous book Aspects of the Theory of Syntax
(1965): Competence is ‘the speaker-hearer’s knowledge of his
language’ and performance is ‘the actual use of language in
concrete situations’ (Chomsky 1965: 4).
Language acquisition device (LAD)

We have been talking about rules and principles over the last few
pages. But what exactly are these? In this section we shall try to
answer this question. Remember that we mentioned that producing
a sentence in a language may be compared to the process of getting
results out of a computer. The computer is programmed and
arranges items according to the instructions given to it. It is
equipped to process information in a certain way. Suppose the
human brain too has some device which can help it to process
language; what would it do? This device, which Chomsky called
the language acquisition device, (or LAD for short), would have
the capacity to arrange the lexical units of any human language
according to some general universal principles. Then, there would

be variation according to the language which we are using. These
variations would not be haphazard, but would be controlled or
governed by certain rules. This would make the LAD’s work easier
since it would already have universal principles ‘programmed’ into
it, and all it would have to do would be to learn the rules of
different particular languages. Since this principle of economy is
seen in all natural processes, Chomsky maintains that this may also
be how the human brain works (for an easy discussion of these
theories, see Lyons 1970: 122).
Rules and principles
The principles common to all languages (or universal principles)

are general patterns of language use. We can say that all languages
must be governed by certain rules and must have a structure. They
cannot be haphazard. The rules of a language determine how it is
produced. Without rules no sentence, however simple, can be
produced, so that we can say that language is always rule-governed.
Consider the following simple sentences:
(1) Neha ran to the school.
(2) Neha school bhagi.
Sentence (1) is in English and (2) is in Hindi-Urdu. Now we
simply cannot put these sentences in any order we like. We also
cannot give a general rule that one particular word, for example,
school, will always come at the beginning of the sentence. We shall
have to consider the structure of the sentence. This means that we
must consider whether a word is a noun, verb, a preposition, and
so on. These are syntactic categories, and structure depends upon
them. In other words, no matter which language we use, any rules
which produce well-formed sentences in that language must be
structure-dependent. Rules will apply to the structure of the
language which is determined by nouns, verbs, adjectives, and so
on—that is, by syntactic categories. This structure-dependence is
therefore a universal principle. It applies to all human languages.
Let us now come to the rules of English and Hindi-Urdu, as

demonstrated in sentences (1) and (2). It is clear that in (1) the
verb ran comes before the place where the subject (Neha) went. In
Hindi-Urdu however, the verb ran (bhagi) comes after school.
Moreover, in English the verb ran is not marked for gender;
whether Neha is a boy or a girl, we shall use only ran. But in
Hindi-Urdu it is clear that we are talking about a girl because we
put the morpheme -i at the end. If we had used -a, then the subject
would have been masculine. This adding of -a or -i is an example
of inflection, and if we are describing the morphological rules of
Hindi-Urdu we will have to give the information that the gender
governs the morpheme used for inflection in verbs. This means that
universal principles remain the same, but the rules vary from one
language to another. Our task is to discover the rules of particular
languages as well as the universal principles of all languages. In this
way we can try to find the answer to questions such as ‘What is
language?’ and ‘How does the brain work when we produce and
understand it?’
Different kinds of rules
We know that languages have the following dimensions: (a)

phonological, (b) morphological, (c) syntactic, and (d) semantic.
Thus we would need to take account of at least these four types of
rules in order to describe a language.
(a) Phonological rules specify how words and combinations of
words are pronounced.
(b) Morphological rules specify how words are built up of small
units of meaning, i.e. morphemes.
(c) Syntactic rules specify how words combine to form
sentences.
(d) Semantic rules tell us the meaning of words and combinations
of words.
This chapter has defined certain key concepts in modern
grammatical theory as developed by Chomsky and his associates.
Through these definitions and their explanations, we hope that you

will have grasped the preliminary concept that grammar is the
reflection of all that the native speaker knows about his or her
language. Through describing the particular grammars of different
languages, we will be able to postulate a universal grammar, which
will give us the principles common to all languages, all of which
are rule-governed, and these rules are phonological, morphological,
syntactic, and semantic. The aim of the Chomskyan linguist is to
discover these rules with a view to understanding the nature of
language itself, and in the final analysis, the properties of the
human mind, which seems to be especially equipped to understand
and produce language.
8
Transformational Generative Grammar
T he term ‘transformational generative grammar’ is used to refer

to Noam Chomsky’s theories about syntax which have been
touched upon in the last chapter. These theories were first put
forward in a book entitled Syntactic Structures which was published
in 1957. If you want to know more about Chomsky’s ideas on this
as well as other aspects of his thought read the French linguist
Mitsou Ronat’s interviews of Chomsky in a book entitled On
Language (Chomsky 2002). However, here is a summary of these
views. Basically, Chomsky tried to find certain rules which would
create well-formed sentences of a language. A sentence was defined
as being ‘well-formed’ if the native speakers of that particular
language felt that it was correct or grammatical. As Chomsky
wanted to create rules of a general kind about language he had to
use what is seen by many as mathematical terms. For instance, he
defined a language as follows:
I will consider a language to be a set (finite or infinite) of sentences

(1957: 13).
But he also said that it was not mathematics he was doing. For
instance, he told Mitsou Ronat:
It turns out that the way to ‘speak precisely’ is by formalization, but

it would not be correct to consider that as mathematics. For example,
some variety of recursive function theory provides the means, in
principle, to express linguistic rules. But up to that point, this is
formalization, not mathematics.
(Chomsky 2002: 125)
Transformational Generative Grammar 81
What he meant by his definition of language was that any language,

whether English or Urdu or Hindi or any other you can think of,
was made up of a number of well-formed sentences. These
sentences could be infinite, since it is possible to keep making new
sentences. However, at any given time (for example, at the moment
you are reading this), the number of sentences is actually finite.
That is, at this moment they can actually be counted, though of
course more sentences will be produced after the fixed time up to
which they are counted. According to Chomsky, his grammar is
generative, since it can generate or create an infinite number of
sentences. It is called transformational since a basic or simple
sentence like:
(1) I read the book
can be changed or transformed into a number of sentences with
either the same meaning:
(2) The book is being read by me.
or with different meanings:
(3) Do I read the book?
(4) I read the book, don’t I?
(5) I do not read the book.
What we have done is that we have transformed the basic sentence
by adding words, deleting words, and, above all, by the movement
of words. These changes also take place through specific rules, and
we call these rules transformational rules. Thus grammar generates
and transforms sentences. It can therefore be called transformational
generative grammar, or TG for short.
Properties of transformational grammar (TG)

TG has the following properties:
(a) It will generate only the well-formed or grammatically
correct sentences of a language. It will not generate a
sentence (that is to say, a syntactic structure) which is ill-

formed or incorrect, since it is meant to create those rules
and principles which are in the mind or brain of a native
speaker. These rules and principles are by definition about
that form of a language which native speakers consider
correct or well formed.
(b) This grammar will have recursive rules. This property of
recursiveness is the capacity of a rule to be applied again and
again in order to generate an infinite set of values. In this
case, by ‘values’ we mean new combinations of words which
are grammatically correct. In other words, recursive rules
allow us to generate an infinite number of well-formed
sentences of a language. Let us see how recursive rules work.
Consider the sentence:
(6) This is the boy.
Now it can be made longer by adding
(7) who rode the horse.
But the rule which added (7) can also be applied to (8), and we
can add
(8) that chased the fox.
and again
(9) that chased the jackal.
But if we keep adding new sentences beginning with ‘who’ or
‘that’, we will never finish the sentence, even if we run out of words
and have to begin with boy or horse again. This kind of change is
called embedding. In this, we put in (or embed) a smaller sentence
or several small sentences in a large sentence. But recursive rules
can work in other ways too. One could even keep adding and to
create an endless list. The point to remember is that recursiveness
accounts for the endless creativity of language. Because we can
apply a rule again and again, we can create an infinite number of
sentences out of finite number of words (that is to say, out of a list

of finite lexical items).
Deep and surface structure
In sentences (1) and (2), you notice that the first sentence is
simpler or more basic than the second. Let us reproduce them:
(1) I read the book.
(2) The book was read by me.
According to Chomsky, sentences may be present in the mind/
brain at two levels. Simple, basic or kernel sentences consisting of
ideas or rough meanings, must be present at a deep level. The way
these sentences are formed at that level is known as deep structure.
However, we cannot actually ‘see’ sentences in the brain at that
level, we only hear spoken sentences or see written ones at the
surface level. The structure of these sentences is known as surface
structure.
At this point you must understand clearly that all we can actually
see or hear is the surface structure. The deep structure (D-structure)
is an abstract concept, i.e. a theoretical concept that cannot actually
be seen or touched or felt. But why do we talk of D-structure or
D-level, then? Is it only to make syntax more difficult? The answer
is that we formulate certain hypotheses to explain certain facts that
we can observe. A hypothesis is just a guess, but it is the guess of
an expert in that particular branch of knowledge. Thus it has a
possibility of being correct, and even if it proves to be wrong, it
helps researchers find out more about the facts being investigated.
The deep level is assumed since we think that the brain/mind
would first assemble certain concepts, that is to say, a basic
sentence. This would be operated upon by rules and changed into
more complex sentences. We think this would be more economic
for the brain or mind. However, we could be wrong. In science
any theory can be proved to be wrong. At the moment, however,
we assume that sentences are formed at some deep level in the mind
or brain. They are very basic, and can be called kernel sentences at
that level. These kernel sentences are then operated upon by certain
rules so as to produce surface sentences. The following simple
diagram illustrates this:
DEEP STRUCTURE
Kernel sentences
Transformational rules
SURFACE STRUCTURES
Surface sentences
This diagram is very simple and therefore ignores certain
complications which we shall come across during the course of this
study.
Symbols used in writing rules
A symbol represents something else. For example, the national flag

stands for the honour and unity of a nation. In the language of
science there is a need for symbols. We can choose any letter, for
example, a, b, > (is greater than) or + (plus) as a symbol of
anything. In linguistics too, we use symbols when we write down
rules. The writing down of rules is known as formalization, that
is to say, explaining something in the form of a precise formula.
Since Chomsky was greatly interested in mathematics, he expresses
rules in a quasi-mathematical manner. Here are some of the
symbols and the basic rules which can produce the sentences of
English:
S = sentence (e.g. I walk home)

N = noun (e.g. Tania, he, they, boys…)
V = verb (e.g. walk, talk, smile…)
Adv = adverb (e.g. quickly, slowly, clearly…)
Adj = adjective (e.g. good, green, pretty…)
P = preposition (e.g. on, in, to…)
Art = article (a, the)
As you know, the terms ‘noun’, ‘verb’, and so on, are all syntactic
categories. You must have read about these basic categories when
you studied the parts of speech at school. Now we shall introduce
you to some terms which you probably did not meet at school.
These are:
NP = noun phrase
VP = verb phrase
Adv P = adverbial phrase
Adj P = adjectival phrase
PP = prepositional phrase
The noun phrase
A noun phrase is also a syntactic category, like a noun. It can have

more than one kind of noun (that is, a proper name, a common
name, or a pronoun); it can have several nouns joined together; it
can have nouns modified by articles; or some other words which
can determine the nature or the quality of the noun, or specify it
more accurately. A noun phrase can also have an adjective
qualifying the noun. The following are all noun phrases:
With only one Sameer NP has only PN

proper noun (PN)
With a common boy NP has only CN
noun (CN)
With a pronoun he NP has only
(ProN) ProN
With two proper Iqbal and Sonia NP has PN and
names PN
With noun and a boy, the boy, this NP has
determiner boy, that girl, those determiner +N
boys
With adjective good boys, a bad boy, NP has
qualifying the noun the bad boy determiner + adj
+N
We can see that the noun takes the same place as the noun phrase
does. They belong to the same syntactic category and can fill the
same slot in a sentence. For example:
(1) SAMEER ran home NP = PN
(2) THE BOY ran home NP = CN
(3) HE ran home NP = ProN
(4) THAT BOY ran home NP = Determiner + N
(5) THAT BAD BOY ran home NP = Determiner + Adj + N
If we do not use the term noun phrase (NP), we cannot express
the fact that proper nouns, common nouns, pronouns, and nouns
specified exactly or modified by adjectives (and so on) belong to
the same syntactic category. That is why we use the term noun
phrase even when we have only a single noun, as well as when we
have many nouns or a modified noun. The definition of an NP is
that its key constituent is a noun. We call this key constituent a
head. If there is anything else in that phrase, it will tell us
something more about that noun, and is said to be dominated by

it.
(a) NP (b) NP (c) NP
N det N det adj N
Sameer a boy a good boy
(a), (b), and (c) are all NP’s. In all of them the most important
constituent is a noun, and, if there is anything else, it tells us more
about that noun.
Other syntactic categories
Just as an NP has a noun at its head, a verb phrase (VP) has a verb
as its head. In the sentence Sameer ran home, ‘ran home’ is a VP.
The verb which is at its head is ran. This VP can contain a number
of other things, for example, it could have a noun (home), an
adverb, (quickly), or a prepositional phrase (to his mother) after it.
But all other phrases will come under the head or will be dominated
by the head.
For example:
(6) Sameer ran.
(7) Sameer ran home.
(8) Sameer ran quickly.
(9) Sameer ran home to his mother.
In all the above examples ran (that is, the verb) is the most
important thing. Everything else is just additional information
telling us more about the verb. Thus everything else will be
dominated by the verb. Thus a VP will certainly contain a verb at
its head, although it can contain much more under that head.
Similarly, in a prepositional phrase (PP) the preposition dominates;
in an adverbial phrase (Adv P) the adverb dominates; and so on.
Phrase structure rules
As we have seen, phrases have definite structures. For example, an

NP consists of at least a noun. It can have a determiner, an
adjectival phrase, and so on. Similarly, a sentence also has a
structure. If we can find a formula, or a formal way of writing how
these structures are made or generated, we can write down the
phrase-structure rules (or PS rules) of a language. Let us therefore
write down the PS rules of English.
S(entence) NP + VP
NP {art, (adj), N, PN}
VP V + NP + (PP) + (Adv)
PP Prep + NP
N {boy, Sameer, dog…}
N {Ashwariya, Shabana, Rome…}
Det {a, the, that, this, these…}
Adj {pretty, bad, ugly…}
V {run, wrote, will think…}
Prep {with, to, near…}
Adv {quickly, recently, yesterday…}
In the PS rules the following symbols have been used:

consists of
+ and or followed by (This is omitted by some people).
{ } curly brackets contain elements out of which only one
should be selected.
( ) brackets contain constituents which may be omitted.
Now let us generate the sentence: Sameer ran.

As it is a sentence, it contains an NP followed by a VP. Let us make
a labelled tree diagram of it:
Tree diagram
A tree diagram is only a convenient way of showing how the

different constituents are arranged in a sentence. It tells us which
categorical constituent dominates which, and gives us a pictorial
model of the sentence.
S NP + VP
NP = N (i.e. Sameer)
VP = V (i.e. ran)
S
o < — — — — — sentence node

NP VP
NP node — — — — > o o < — — — — — VP node
| |
N V
N node — — — — > o o < — — — — — V node
| |
Sameer ran
It is clear here that the S node dominates all the other nodes. It
dominates the NP and the VP. It is just another way of saying that
the sentence is a larger syntactic category which contains a noun
phrase and a verb phrase. Similarly the NP node and the VP node
dominate all the other things which come under them. Consider
the tree diagram for the following sentences:
(10) Geeti ran very fast.
(11) Geeti ran to the lady.
NP VP
N V
Adv P
Intensifier Adv
Geeti ran very fast
Here, fast is an adverb. It tells us more about the verb ran, or

modifies it. Thus it comes under an adverb phrase (adv P). But we
have another modifier of fast which is very; this is an intensifier.
The Adv P node will dominate the intensifier also. However, all
that the Adv P does is to tell us something about the verb ran, thus
the VP node dominates the Adv P node.
S
NP VP
PP
NP
N V P Det N
Geeti ran to the lady
Here, as you can see, the VP node dominates the PP node, which
in turn dominates the second NP node. Since certain nodes
dominate other nodes in the tree diagram, we say that is
hierarchically arranged.
Other ways of showing syntactic structures
Making a tree diagram is a convenient way of showing the syntactic

structure of a sentence. There are, however, some other ways of
giving the same information. For instance, we can use brackets.
Below the brackets we can write down the name of the category
we have enclosed in the brackets. This is called ‘labelling’, and
labelled brackets can give the same information as the labelled tree
diagram, see below:
(12) Sheela ran.
First step: Put the noun and verb in brackets and label them.
(Sheela) (ran)
N V
Second step: Put N inside bigger brackets labelled NP and V inside
a bracket labelled VP.
( (Sheela) ) ( (ran) )
NP N VP V
hird step: put both the NP and the VP inside a larger bracket
T
labelled S.
( ( (Sheela) ) ) ( ( (ran) ) )
S NP N VP V
As you can see, this method uses many brackets and can be
confusing. It is, however, useful if you want to save space. The tree
diagram is generally simpler to use, but it takes more space.
We have learned so far that there are phrase structure rules which
determine the structures of phrases, i.e. the head of an NP is N,
the head of VP is V, and the head of AP is A (adverb). Suppose
we want to reduce the number of phrase structure rules—how can
we do this? Obviously, we will have to integrate all the rules under
one rule. But why should we want to do this? Because it is scientific
to have as few rules as possible (the ‘minimalist’ approach).
Let us start with VPs. We know that VPs have V and other items
(NP, PP, PP…) under it. The maximal projection is VP but under
it there are projections of V labelled as V-bar written as ⊽ or V’.

Consider the sentence:
(13) I will run down the garden
(14) I am sure you will do so
Now do so substitutes run down the garden. This constituent phrase
is also a small verb phrase. But, being under a big VP, it is a VP-
bar or V’.
Before V’ there may be a specifier.
(15) We have all run down the garden.
All is a quantifier. It occupies the specifier position [spec, VP]. It
combines with V’ to form a V” projection. The V” is the same as
VP.
In short the structure of V” is as follows:
V” → spec – V’
V’ → V…
V” is the maximal projection of V and this is what we called VP
earlier.
Now let us extend this kind of structure to other phrases. First, to
NPs.
(16) The meeting of the artists in the studio went well.
This is how we used to parse such a sentence.
NP
det N PP PP
the meeting of the artists in the studio
Suppose we say now:

(17) The one at night did not go well.

We have substituted one for the meeting of the artists in the studio.
Now let us use the idea of the specifier. As you can see the specifies
a particular meeting. Thus we can parse it as follows:
N” (NP)
N'
Spec
The N' PP
in the studio
N PP
meeting of the artists
Nowadays, since the linguist Abney’s work (Abney 1987), the

determiner is called the head of what used to be called NPs. Thus
NPs are called DPs and this particular re-naming is called the DP-
hypothesis. The reasons for this change are the same as the other
changes we are discussing, i.e. it makes analysis more satisfactory.
However, we will not go into the reasons put forward by linguists
for bringing about this change. You can read them later once you
understand syntax better. The gist of the matter is that NPs can be
called
N” → Spec; N’
N’ → N; XP
We can propose a similar model for adjective phrases, propositional
phrases etc. This is called the X-bar theory and it can be summed
up as
X” → Spec; X’
X’ → N; X; YP
X stands for N, V, A or P. so now we have a general format instead
of PS rules for each of these nodes.
We can sum up the X-bar theory saying that the lexical head of a
projection is a zero projection (X0). Complements combine with X
to form X’ projections as do adjuncts. The topmost or maximal
projection is formed with a specifier. When the specifier position
is not filled in by a word it is considered as filled in by zero.
Now let us go ahead to consider sentences (S) as projections. In
our earlier analysis the S was on its own. In this new analysis we
will propose that there is a complementizer position before S.
Comp positions are clearly obvious in sentences beginning with
wh-words (whether, what, when etc.).
(18) Whether you will do it is doubtful.
(19) What you will do is doubtful.
(20) When you will do it is doubtful.
In all the sentences above the position before you is occupied by
wh-words. The general structure is:
S'
S
Comp
wh-words
NP Aux VP
you will
Now S’ and S are called exocentric categories not endocentric

ones. This means that they do not get projected from heads like
N, V, etc. They are composed of several units next to each other.
This means we need a head for them. If we get such a head we will
have a schema common to the other categories we have been
advocating above. Let us call this inflection (INFL or I for short).
Now let us see what the I slot can be filled in with. First let us see
whether we can argue that all sentence nodes are under INFL.
Auxiliary nodes
These are nodes dominated by AUX. This means it will have tense
(past, present) and modals to mark the future (shall, will etc.).
But remember that verbs inflect for person and number also. This
is less for verbs in English, but French and Farsi have many forms.
Let us compare Urdu-Hindi forms with English ones.
speaking = bolna
S.no Male Female Male Female

1 Mai bola Mai boli I speak I speak
2 Tum bole Tum boleen You speak You speak
3 Tu bola Tu boli You speak You speak
4 Vo bola Vo boli He speaks She speaks
5 Hum bole Tum bole We speak (pl) We speak (pl)
6 Aap bole Aap boleen You speak (pl) You speak (pl)
7 Vo bole Vo boleen They speak (pl) They speak (pl)
As you can see in Urdu-Hindi the verb is inflected –i, een, -e but
the pronouns are the same. Moreover, there are three ways to refer
to the second person aap (politest), tum (intimate; also less polite
than aap), and tu (impolite or very intimate). English has only one
word you instead of these three. In English, moreover the pronouns
for male and female are different for the third person (he and she)
but not in the other instances. As for the verb, it does not inflect
for gender at all. As for the tense it has only two forms.
In Farsi the verb inflects for person, singular and plural but not for
gender. Moreover the pronoun also does not indicate gender.
Man amadam I came (singular both male and female)
To amedi You came (singular: m and f)
Oo amad He/she came
Ma amadaem We came
Shuma amadeed You came (plural as well as polite: m and f)

Anha amadand They came (m and f)
In short, there are six forms of the inflection of the verb and yet
we need not clarify whether we are talking about a boy or a girl.
To be grammatically correct agreement rules should be followed.
If one says I then we should say speak; if we say he then we should
use the form speaks and so on. No matter how many forms that
particular language may have, the agreement (called AGR) feature
is necessary. The inflection, we sum up, has tense, agreement, and
modals. So tensed clauses have +Tense + AGR
and we can see that
sentences are either tensed or infinitives. The following are tensed
sentences.
(21) I will run in the garden.
(22) I am running in the garden
(23) I run in the garden
The INFL contains:
Sentence (21) contains will (modal indicating future intention).
Sentence (22) contain am and –ing i.e. the verb to be and an
indicator of an action continuing at present.
Sentence (23) contains past tense of the verb run.
Now let us take a sentence without tense. Such a sentence is as
follows:
(24) I intend to run in the garden.
As you can see, to is the infinitival marker. The infinitive form of
the verb introduced or marked by to does not have tense. This too
is in INFL. To has neither tense nor agreement. We can call it
–Tense
– AGR
.
So finite clauses have [+tense] and [+AGR], and infinitival clauses

have [-tense] and [-AGR]. All clauses then have [±tense] and
[±AGR] in the INFL.
So if we assume that INFL is the head of S, then S will be a
projection of INFL phrase (or IP). Now, using our previous schema
of using specifier, we can re-write S as:
I” → Spec; I’
I’ → I; VP
Let us parse a sentence of the S → NP + VP under this schema.
(25) I will drive the car.
I”
I'
NP VP
I
V'
V NP
I will drive the car
Now let us go back to complementizers. As you can see the

complementizers that and if go with finite clauses and for goes with
infinitival ones. As for whether, it can choose both types of
clauses.
If there is no complementizer, we do not say that the slot does not
exist at all. We say that the slot is there but it is empty. You will
understand why we do this. It is because our schema remains the
same but in some cases we have zero and in others we have some
words.
Complementizer phrases (CP) may be seen to head complementizers
as in the other schemes.
(26)
CP
C' IP
I'
C
NP I VP
I will run in the garden
In this case the CP is empty but we can fill it in with whether:

(27) Whether I will run in the garden is doubtful.
Thus it is better to assume that there is a CP position but it is not
always filled in. Of course, in the light of our previous schema,
there will be a specifier too. So the structure will be:
CP
Spec
C'
C IP
If you have the question word will as in:

(28) Will I run in the garden?
Then will moves to the specifier position. In such a case whether
will not be this node. We cannot have:
(29) * will whether I run in the garden.
This is because the node can have only one word not more than
one. But we can have this sentence:
(30) When will I run in the garden?
The tree of this is as follows:
C”
C'
IP
I'
Spec C NP I VP
When will I empty run in the garden
The auxiliary moves from one head to another. This is called head-
to head movement. We will not, however, go into the details of
these movements here.
This means that question words move to [spec CP]. In main clause
questions, they precede the inverted auxiliary. In subordinate
clauses, however, wh-words do not occur with the complementizers
that or whether. This is true for English as far as the words quoted
above are concerned. However, the principle is a general one. We
have it in our heads and we modify it according to the language
we speak.
The question is, what kind of structure do we have in the mind?
Answers to such questions come from theory, i.e. we speculate that
the structure will be like this. Then we compare it to another
model. Then, whichever model gives the minimum complexity
must be adopted. Let us look at some diagrams.
(31) I am walking
Let us have all the models of branching which are possible here.
Here are the possibilities:
(31a)
I am walking
(31b)
I am walking
(31c)
I am walking
Now let us have a theory which allows only binary branching.

(31d)
I am walking
If we have a principle in mind it will be: binary branching is

allowed. All others are ruled out. This principle is easy—it is a
minimalist principle—rather than some principle which allows all
kinds of branching. The kind of structural principles we have been
talking about are part of what is called minimalism in grammar.
An easy introduction to minimalist grammar is Andrew Randford’s
Syntax (1997). You are advised to read it along with more recent
works.
9
More about Grammar
I n the last chapter, you have been introduced to transformational

generative grammar. In this chapter, let us try to understand
more grammar. Let us take up the simplest definition of grammar:
it is a device which can create a well-formed sentence in a human
language. Let us now take a simple sentence:
(1) The girl followed the boy.
Now let us forget the phrase structure (PS) rules we read earlier
and try to guess how this sentence was created. The simplest guess
(or hypothesis) is that words have been placed one after the
other:
The + girl + followed + the + boy.
This is a linear model because words come one after another in a
straight line. It can also be called a chain model, since words are
connected to each other like links in a chain. It is a static model
since nothing can move from one place to another. It is also called
a finite state grammar since it is static, and therefore cannot change
state. This simple model of grammar can be created by imagining
the flow of an electric current from one point to another, creating
a sentence as it goes. For example:
o-----------o------------o-------------------o-----------o------------o
A The B girl C followed D the E boy F
Now, if you start the ‘current’ at B, and stop it at C, the sentence

will be
(2) * girl followed

Similarly, if you want to create more than one sentence you will
have to create loops so that the current can be passed through one
circuit or the other as you like.
Let us try to generate the sentences given below by this model of
grammar:
(3) The girl followed the boy.
(4) The boy followed the girl.
(5) The boys followed the girls.
The models will be as follows:

o-----------o------------o-------------------o-----------o------------o
For (3)
o-----------o------------o-------------------o-----------o------------o
A The B boy C followed D the E girl F
For (4)
o-----------o------------o-------------------o-----------o------------o
A The B boys C followed D the E girls F
For (5)
But it would be extremely uneconomical to use a new model for

every sentence. Let us see what is common in these models and try
to make only one model which will produce all these sentences.
This model will be:
boy girl
o-----------o---boys---------o---------------o-----------o---girls--------o
More complex sentences would require even more loops and the
grammar would not be efficient because it would create sentences
we do not want: for example, (6) The boys followed the boy, (7)
More about Grammar 103
The boy followed the boy, (8) The girl followed the girls, and
so on.
Non-linear models
Because of these defects, Chomsky abandoned this linear model of

grammar in Syntactic Structures (1957). As the title indicates, he
concentrated on syntax, i.e. the arrangement of words in a
sentence, and stated that this arrangement was subject to rules. We
have already looked at phrase structure (PS) rules, which, he said,
generated basic or deep structures. These deep structures were then
subjected to change by transformational rules (T-rules) which
created surface structures. In Aspects of the Theory of Syntax (1965),
Chomsky changed his grammatical model to look something like
this:
BASE component
PS rules Lexicon
DEEP STRUCTURE
T-rules
SURFACE STRUCTURE
Semantic component Phonological component
To put it very simply, the idea is that the base contains PS (phrase
structure) rules which create a kind of skeleton for the syntactic
categories (NP, VP, and so on) to fit in one after, or under, the
other. Then comes the lexicon, which is a dictionary or collection
of words. Every word is specified in such a way that only that word
can be fitted into the PS rules correctly. This creates a kind of deep
structure. Now T-rules are applied, which change the sentence so
that it comes in a changed form to the surface level. Now it takes
on meaning, and it is the semantic component which gives it that
meaning. It can be spoken, or written in symbols which refer to

spoken sounds, which is the phonological component. This is a
simple model of TG which is presented in Chomsky’s books
mentioned above. The theory behind this model is slightly more
complex, but it has been simplified here to make it understandable.
This theory is called the standard theory.
As you can see, the PS rules can generate sentences in a fixed order.
If this were the only way of generating sentences, we would need
an infinite number of PS rules to generate an infinite number of
sentences. That would not be feasible. Thus, according to
Chomsky, only the simplest sentences are generated through PS
rules. Even these kernel sentences are generated at a deep level.
These are then changed through movements of constituents. This
movement transforms sentences into different kinds of sentences.
That is why they are called transformational rules. To illustrate
transformation through movement, consider the following
sentences:
(9) The boy followed the dog.
The following sentence means the same thing:
(10) The dog was followed by the boy.
These second sentence (10) must be a transformed form of the
basic sentence (9) at the deep level. Some transformational rules
(T-rules) have produced the second, passive structure. Let us
number the NPs and VPs:
The boy followed the dog

NP1=1 VP = 2 NP2 = 3
The structure is 1 + 2 + 3. This is called the structural description
(or SD).
Now we need a rule which will bring 3 in place of 1, and 1 in place
of 3. We also need to put in ‘was’ and ‘by’. The transformation is
shown by a double arrow ( ). This change is called the
structural change (or SC).
So, SD = 1 + 2 + 3
SC = 3 + was + 2 + by + 1
1+2+3 3 + was + 2 + by + 1
(10a) The dog was followed by the boy
As you can see, this T-rule has moved constituents and inserted
‘was’ and ‘by’ in this case. The passive transformation is a little
more complex, but it has been kept very simple here so that you
can understand the concept of the T-rule better.
Grammar as a machine which generates sentences
To sum up, you can imagine grammar as a machine or a device for

creating sentences. It is in the mind of the speaker and represents
his or her knowledge of the language. It contains rules for creating
simple sentences at the deep level which are then transformed by
transformational rules into more complex sentences.
Now let us see what kind of arguments can be advanced in favour
of their existence. One of these arguments is that in some sentences
movement seems to have occurred, i.e. it appears that some
constituents have been moved from their original places. Let us
take an example:
(11) I put the book on the table.
Now, ‘put’ is a verb which is followed by an NP, that is, something
which can be put somewhere. This NP is followed by a PP
(prepositional phrase), beginning with in, on, inside, outside, and
so on—that is, something which tells us where the thing has been
put. In other words, for a verb like ‘put’, the lexicon must specify
what constituents follow it.
put = V (followed by) NP, PP
Thus if we omit either the NP or the PP after ‘put’, we will
produce a grammatically ill-formed sentence. For example:
(12) * I put on the table. (Here the NP following the V has been
left out).
(13) * I put the book. (Here the PP following the NP after the V
has been left out).
Now consider the following sentence:
(14) What did I put [----] on the table?
In this sentence, ‘put’ is not followed by an NP. However, we have
just stated that ‘put’ is subcategorized in the lexicon as being a verb
followed by an NP and a PP. These subcategorisation facts are
constant. Thus, what has happened is that the NP after put has
moved away from its original position at some stage. But the very
fact that it has moved away must mean that there was an original
position for it to move away from. Thus, we assume that there is
a deep level at which it was there, and that certain T-rules have
moved it away. Some T-rule also brings ‘what’ to the beginning of
the sentence. In short, rules move parts of phrases from one place
to another in sentences. We will try to understand this in more
detail later.
Transformations
We said earlier that T-rules move, add or delete constituents in a
sentence. First, we must find out if the sentence has got the kind
of structure to which rules can apply. This is its structural
description (SD). If this structure is such that these rules can apply,
then we can use the rules. Let us now study only one
transformation—the passive transformation.
If you understand this T-rule, you will not only learn more about
T-rules, but also about argumentation in linguistics. Let us then
take the passive transformation. We will see how a sentence, which
states a proposition, changes into the passive form. But before we
do that, we will go back to our PS rules. We have covered some of
them, but we have not covered the expansion of the auxiliary.
AUX Tense {present}
This rule simply means that a verb, for example walk, must have
a marking for tense. If it is in the past tense, (i.e. + past), we will
read it as walked; this addition of the morpheme –ed inflects the
verb for the past. If it is in the present tense (i.e. + present), we
will read it as walks.
Other helping verbs (or auxiliaries) can be verbs, such as will. This
will give us the future tense too. Such words are called modals and
their list includes will, shall, can, would, must, may, and so on. So
another expansion of AUX will be.
Aux Modal
But, whereas all verbs must be marked for tenses, modals are not
necessary in all phrases. As modals are optional, they are put into
brackets. Our rule should be read as follows:
Aux Tense (Modal)
But verbs can also be in the progressive aspect or the perfect phase.
For instance:
(15) He is reading.
is in the progressive aspect. It shows that an action is in progress
(or in the progressive phase). Now take the sentence given below:
(16) They have been reading.
Here the perfect is shown by have and the -en shows the past
participle ending. In other words, to show that a verb can be in
either the progressive or perfect form, we need two optional
elements in the AUX. Speakers of English have chosen the
following formulas to show these two elements:
Progressive (be + ing)
Perfect (have + en)
These formulas are the only ways of showing that in the progressive
form, there is the verb to be and the ending –ing. In the perfect
form, there is have or has and the ending –en. The formula is
somewhat misleading, since the past participle of ‘read’ is not
formed by adding –en. Here the past participle is merely, read

(/red/). The –en in the formula therefore simply means ‘make this
into the past participle, whatever realization that might have’. This
realization is determined by other rules which give the exact
pronunciation of every morpheme. These are called phonemic
rules. Let us now add another optional expansion of AUX to our
list:
AUX (be + ing) (have + en)
The full expansion of AUX will be:
AUX Tense (Modal) (have + en) (be + ing)
The passive transformation
Let us now come to the passive transformation and look at one

specific example:
(17) The man may have opened the door.
The passive form will be:
(18) The door may have been opened by the man.
The structural description (SD) of (17) is as follows:
NP1 + AUX + V + NP2
The T-rule given below, which makes the passive transformation,
can be applied to this SD:
NP2 + AUX + be + en + V + by + NP1
Notice that be + en and by have been added to this rule. Let us
now expand this as follows:
NP 1 and (have + en) + by are added in certain places.
This creates a deep level sentence which will be brought to the
surface level.
D-Level: the door present may + have + en be + en open by the
man
Applying Auxiliary Inversion Rule (Auxiliary is inverted. It is also

called affix hopping): The door may + present have + be + en open
+ en by the man
Applying the morpho-phonemic rule (i.e. the way we pronounce
the past of open, etc.), we do not say openen, we say opened. We
call the present form of have simply have and the past form of the
verb to be, been. So we have:
(18a) The door may + have + been + opened + by the man
i.e.
(18b) Surface Level: The door may have been opened by the
man.
This might seem to be a very artificial and roundabout way of
deriving a passive sentence. However, the principle is quite
significant. It is that complex sentences are not created like simple
sentences. Simple sentences, or kernel sentences, are created at the
deep level and they are changed to complex sentences by
transformation rules.
To read Radford’s book you do not need any more knowledge of
syntax than you already possess after having read this book. The
conclusion of his detailed treatment of V movement, I movement,
NP movement, WH movement etc. is the intriguing rule, alpha
movement: Move alpha (where alpha is a category variable, i.e.
designates any random category you care to choose) (Radford 1988:
537).
This rule is intriguing because it appears to be a recipe for syntactic
anarchy. Can we really move any category anywhere in a sentence?
The answer is that this is a rule of Universal Grammar, whereas in
every language there are particular rules which prevent certain
kinds of movements. These restrictions are the ones which linguists
have to find out and write down. For instance, the structure must
be preserved. The rule for this is written as follows:
Structure-preserving principle: Major grammatical transformations

are either root or structure-preserving operations (Radford 1988:
539).
A root transformation is one that applies to a root clause, i.e. the
main or independent clause. A non-root transformation is one in
which the transformation is such that it can be generated by the
grammar in that position.
In short, if we extend this principle further, we come to the
conclusion that all transformations are structure-preserving. This
would make sense even if you do not go into the rather complicated
arguments of linguists, because sentences would not make sense
unless there were some rules to tell us how we can change them.
In order to understand how such arguments are given, you should
read the detailed chapter on alpha movement in Radford (1988:
527-584).
Next, you may read Yule (1985), Lyons (1970), and Radford
(1988). If you want to read more about transformations and
Chomsky’s theories, the next step would be to read Radford
(1997), before going on to Hageman (1991) and more recent
works.
10
Government and Binding Theory
I n order to understand how the mind puts words in a sentence

in order, we will consider some more theories. One of them is
the government and binding theory, which is based on Chomsky’s
work of the 1980s. However, the simplified account given in this
chapter is based upon Liliane Haegeman’s excellent and accessible
explanation of the theory (Haegeman 1991). Let us begin by
looking at the idea of dominance.
Dominance
If we say that a node dominates another node, we mean that it
occurs higher up the tree than the latter, and is connected to it by
an unbroken set of lines. A more technical definition is as
follows:
Node A dominates node B if and only if (iff) A is higher up in the
tree than B and if you can trace a line from A to B going only
downwards (Adapted from Haegeman 1991: 122).
In the sentence below:
(1) I will leave my horse.
One can represent this in the tree-form as follows:
Node A S
Node B NP VP Node B
V NP
I will leave my horse

Here Node A (S) dominates node B (NP), as it is higher up the

tree-than NP and because one can trace a line from S to NP going
only downwards. S also dominates VP like it dominates NP with
no node in between. This is immediate dominance. Here both NP
and VP are sister nodes.
This gives us a definition of government.
Government: A governs B if
(a) A is a governor
(b) A and B are sisters (Adapted from Haegeman 1991: 123).
Governors are heads and if x is a head and it governs Y, then x
head-governs Y.
Another aspect of governance is the notion of C – command. The
rule of C-command is:
Node A C-commands node B iff
(a) A does not dominate B and B does not dominate A.
(b) The first branching node dominating A also dominates B.
Consider the following tree:
S
B A
NP VP
If we want to know which node VP C-commands, we should move

up till we reach the first branching node and then we come down
the NP node. Then NP is C-commanded by VP. Similarly NP
C-commands VP. Government, by this definition, turns out to be
a matter of ‘mutual c-command’—something which politicians
would never agree with! Let us put this new definition in the
following form:
Government (revised definition): A governs B iff
Government and Binding Theory 113
(a) A is a governor.
(b) A C-commands B and B C-commands A.
Refining this further, there are barriers to government. Chomsky
has written about them in Chomsky (1986 b: 8). To put it simply,
if there is a node in between the two nodes, then it is a barrier to
government. Let us explain this further.
Government
If the node X is fully projected, this is called maximal projection.
The term M-command is used when X is a maximal projection.
Let us define government again:
A C-commands B iff A does not dominate B, and every X that
dominates A also dominates B.
Here X is not the first branching node (this would be strict C-
command). But now X is the maximal projection. Then we say
that A M-commands B. Now we turn to Chomsky’s definition of
government.
A governs B iff A M-commands B and no barrier intervenes
between A and B. Maximal projections are barriers to
government.
Governors are heads (Chomsky 1986:8).
Let us now turn to binding. It is also based on Chomsky’s work of
the 1980s. It is based on a simple notion: how are references to an
NP interpreted? For example, how do we interpret a sentence
like
(2) My uncle hurt himself.
Of course, the reflexive pronoun ‘himself’ is bound or indexed to
‘my uncle’, but how do we understand this? Or rather, how do we
think the mind must have processed this idea?
First, quite clearly, the mind must be indexing the two versions of
the nouns ‘my uncle’ (NP1) and ‘himself’ (NP2). In this case NP1
is the antecedent of ‘himself’. We can say they are co-indexed and

give them the same number so that we know they refer to the same
person.
(2a) My uncle1 hurt himself1.
The two noun phrases are co-indexed or bound with each other.
The antecedent is the binder of the reflexive.
There are barriers to binding also. For instance, look at the
following sentence:
(3) My uncle thinks that I have hurt himself.
Here the reflexive (‘himself’) is not locally bound. The antecedent
(‘uncle’) is too far away from it and it is not in the same clause.
Thus, it cannot bind the reflexive. To bind the reflexive, the
antecedent must be in the same clause. This is known as the clause-
mate condition. Let us state it in more detail.
Principle of reflexive interpretation
A reflexive must be bound by a clause-mate antecedent. Binding

can then be defined as follows:
A binds B iff
(a) A c-commands B.
(b) A and B are co-indexed.
Binding by antecedents is known as A-binding.
Besides reflexives, we have reciprocals which are bound to their
antecedents. The following sentence illustrates this.
(4) My two unclesί like each otherί.
Here, ‘my two uncles’ί is the antecedent of ‘each other’ί. The term
anaphor is used to refer to both reflexives and reciprocals. NPs,
then, are of three types:
Government and Binding Theory 115
(1) R-expressions. These have independent reference. They do

not need an antecedent
(2) Anaphors
(3) Pronouns
If we say:
(5) Uncle likes her.
Here ‘uncle’ is an R-expression. It is free. And so is the pronoun
‘her’.
We have already given examples of anaphors and reflexives
above.
Binding theory
Let us sum up the Binding theory.

First Principle: An anaphor must be bound in its governing
category.
Second Principle: A pronoun must be free in its governing
category.
Third Principle: An R-expression must be free everywhere.
But the governing category is not so easy to decide. Let us discuss
it in a little more detail.
We said earlier that a reflexive must be bound by a clause-mate
antecedent. But suppose there is a sentence such as:
(6) My uncle believes himself to be the best.
Here ‘my uncle believes’ is one clause and ‘himself to be the best’
is another clause. Thus, the reflexive is not a clause-mate of the
antecedent. Yet, the sentence is grammatically correct. This means
that our earlier rule was wrong.
The new rule is as follows:

A reflexive must be bound inside a clause that contains its
governor.
According to this rule, the reflexive need not have an antecedent
in the same clause. If the governor of the antecedent and that of
the reflexive are the same, then the reflexive will be bound to its
antecedent.
(7) My uncle ί believes [himselfί to be the best]
NP VP
As we can see the verb ‘believes’ case-marks the reflexive ‘himself’.
This means that the reflexive is governed by the verb ‘believe’ and
so we can extend its domain. By this extension, we bind ‘my uncle’
with ‘himself’.
The theory we have outlined above helps us understand how
sentences are processed in the brain. These are theories and they
keep changing because this is how science moves forward.
However, the account given above is very simple. If you want to
study the theory in detail, read Liliane Haegeman’s (1991) account
of it.
11
Universal Grammar
U niversal grammar (UG) is defined by Chomsky as follows:
[It is]…the system of principles, conditions, and rules that are

elements or properties of all human languages—the essence of human
language.
(Chomsky 1976: 29)
This definition is fairly simple and does not need any further
explanation. In fact, we have examined some of the rules and
principles of one particular language, English, in the previous
chapters. Just as English has rules and principles, so do all other
languages. The linguist’s task is to find out if all languages have
some basic principles which are universal. These principles would
be shared in common and may be called universal grammar.
The principle of structure-dependency
We have touched upon the idea of structure-dependency earlier.

We said that the fact that languages have structure and movement
proves that we must be able to understand this structure. For
example, consider the sentences below:
(1) I gave sweets to him.
(2) Who gave the sweets to him?
(3) Main ne us ko mithai di. [Urdu-Hindi]
(4) Kis ne us ko mithai di? [Urdu-Hindi]
In both questions (2) and (4), the word which introduces the
question (‘who’ and kis) comes at the beginning of the sentence.
This is not a random process, but comes from wh-movement,

which depends upon the structure of these languages. Wh-
movement takes place in other languages too. In fact, we can say
that it is a universal feature of human languages and all languages
are structure-dependent. Since human languages are structure-
dependent, we give rules like.
S NP + VP
These rules have been refined and changed as you have read in the
discussion of the x-bar theory given earlier. However, for the sake
of simplicity, we will use the older rules.
In this rule, we mention categories (i.e. NP and VP) and we give
their relationship in the structure of a sentence. If there was no
such principle, we would have given some other kind of rule, for
example, ‘Place the words in alphabetical order’. But that will not
allow us to generate grammatical sentences. It is essential, therefore,
to understand that structure-dependency is the only theory that
helps us to understand how languages function.
The head parameter
Parameter may be defined as a variable which may be given a series

of values. In other words, it is something which can change in
value, but which does refer to a ‘heading’ or measure under which
similar things can be placed. For example, a tailor’s parameters
would consist of headings such as length of leg, length of arm,
width of shoulder, and so on. All of these parameters would vary
from client to client, but all of his clients would possess these
measurements.
Head, as we saw earlier, means the ‘head of a phrase’. For instance,
in the NP a girl, the head is the N (= girl ). In talking about the
head parameter, we are concerned with the place of the head in
various languages. The shared or universal fact is that all languages
have phrases with heads, i.e. NP, VP, PP, and so on, but the head
may be either at the beginning of the phrase or at its end. Thus,
the head parameter varies from language to language, from head-
Universal Grammar 119
left, to head-right in structure. Let us look at the placing of the

head in Urdu-Hindi and English below.
(5) The picture is good. [English]
NP VP
(6) Tasveer acchi hai. [Urdu-Hindi]
NP VP
S S
NP VP NP VP
det N V Adj P N Adj P V
Adj Adj
The picture is good Tasveer acchi hai
In (5), the VP = ‘is good’ and the V = ‘is’. This V comes at the
head which is on the left. In the Urdu-Hindi sentence (6), the VP
= ‘acchi hai’ and V = ‘hai’; that is to say, the head comes on the
right.
Similarly, according to Cook (1988: 7), Japanese is also a language
in which the head comes at the end of the VP. This feature is
common to Urdu-Hindi, Punjabi, and so on, but it is not shared
with English. Thus, in the choice of placing the head in a phrase
(i.e. filling the head parameter), UG gives us two possibilities: the
head can be first or last in the phrase. Individual languages give
specific rules as to whether it should be put first or last in that
particular language. People who learn that language learn that
specific rule.
The projection principle
We have touched upon this principle in the chapter on semantics,

but let us study this in more detail. We know that every word, or
lexical item, has properties which decide whether it is suitable or

unsuitable for playing a certain role within the syntax of any given
sentence. Thus, we can give the general principle that:
The properties of lexical entries are projected onto the syntax of
the sentence.
To explain this further, let us take the example of the verb put,
which we used earlier. As we saw earlier, put is always followed by
an NP and a PP of location.
(7) I put my car in the garage.
Here put is followed by the NP (= my car) and the PP (in the
garage). Just saying
(8) *I put.
(9) *I put the car.
or
(10) *I put in the garage.
would be ungrammatical. Thus, we can sub-categorise put as
follows:
PUT (= V) NP PP {loc}
This is to say, put is a verb which must be followed by an NP and
a PP of location.
In the same way, like is a verb which must be followed by an NP
because you must like someone or something.
(11) I like mangoes.
whereas just saying
(12) *I like.
would be wrong. Thus, we can give the following information for
like:
LIKE (= V) NP
But if you say

(13) I smile.
(14) I go.
you will be correct. Smile and go do not need an NP to follow
them. They are called intransitive verbs. Now let us turn to verbs,
both transitive and intransitive, and talk about the way they are
projected onto the syntax.
Verbs are said to have arguments. The term argument comes from
logic but it is used in linguistics in the following way. Let us
consider the verb defeat.
1. The Zombo defeats the Zombi.
In this ‘defeats’ has two arguments: the noun ‘the Zombo’ and the
noun ‘the Zombi’. In traditional grammar, you could parse the
sentence into subject and predicate. The predicate defines some
relation between the two nouns. This is a two-place predicate or a
two-argument predicate.
Now if we say
(16) I shouted.
then ‘shouted’ is a one-place or one-argument predicate.
In traditional grammar, as you will recall, ‘defeats’ is a transitive
verb and ‘shouted’ is an intransitive verb. Transitive verbs are
followed by a noun or a clause. Intransitive verbs can stand on their
own. They do not need to be followed by nouns or clauses etc. For
example:
(17) * I am fond
(18) I am fond of horses.
The verb ‘fond’ needs to be followed by something, i.e.
PP
(of NP
((horses)))
In addition to these two, traditional grammar recognizes

ditransitive verbs. These require a direct object and an indirect
object. The verb give, for instance, is ditransitive.
(19) I will give my pen to you.
In this example ‘my pen’ is the direct object and ‘to you’ is the
indirect object. We can represent this as: give (NP ((PP))).
We can also have another NP instead of the PP. For instance:
(20) I will give you my pen.
which can be written as: give (NP ((det + NP)))
This consideration of argument structure leads us to what is called
theta theory (or θ-theory). Let us consider the arguments of the
verb kill.
(21) The snake killed the mouse.
Kill has two arguments: ‘the snake’ (NP1) and ‘the mouse’ (NP2).
The snake is the agent because it has agency or power to take an
action (that of killing). The mouse (object) undergoes the action
of being killed and is the patient of this activity.
In short, one has to look for the meaning of verbs in order to
understand how they are related to their arguments. These relations
are called thematic roles or theta roles (θ-roles). It is the meaning
of the verb which gives sense to its arguments. This is called
marking the arguments. In simple words, this marking, or theta-
marking (θ-marking), determines whether the argument will be
agent, patient, or something else.
Some thematic roles (θ-roles) you will come across are:
a. Agent (or actor): The one who initiates the action (for
example, the action of killing in (21) above).
b. Patient: The one who undergoes or suffers the action (for
example, the mouse who is killed in (21) above).
c. Theme: The person, thing, phenomenon, state of being etc.

moved by the action (for example, same as above in this
case).
d. Experiencer: That which experiences some state, psycho
logical or otherwise, expressed by the predicate (for example,
same as above in this case).
e. Beneficiary (or benefactive): The entity which benefits
from the action expressed by the predicate. The above
example has no beneficiary but the following one does:
(22) The snake gave the mouse to the cat.
Here ‘the cat’ is the beneficiary. Of course, ‘the snake’ is the
‘agent’ and the ‘mouse’ is the ‘theme’.
f. Goal: The entity towards which the activity expressed by
the predicate is directed. For example:
(23) The mouse fell near the cat.
In this case the ‘mouse’ is the theme and the goal is ‘near
the cat’.
g. Source: This refers to the entity from which something is
moved.
(24) The cat brought the mouse from the snake
Here ‘the snake’ is the source from which ‘the mouse’ is
moved.
h. Location: This refers to the place in which the action or
state expressed by the predicate is located. For instance:
(25) The mouse is in the mouth of the cat .
THEME LOCATION
Thematic roles (θ-roles) are assigned by the meaning of a verb. In
short, the information given in a word (lexical information) is
represented in the way the sentence is constructed (syntactical
representation). This is called the projection principle which is

expressed as:
Lexical information is syntactically represented. Further, each
argument is assigned one and only one theta role.
Thus, words themselves determine where they can be placed in the
syntax. Thus, when we put the words from the lexicon into the
syntax to create the structure of sentences, those words fit in
according to their lexical properties. That is why we say that lexical
entities project their properties onto the syntax.
Types of universals
We have given only three universal principles so far. The question

that arises now is whether a principle which is not found in all
human languages can be called a universal principle of grammar or
not. For instance, movement, examples of which have been given
in English and Hindi-Urdu, does not exist in the same form in all
languages. In Japanese, question words (what, how, when, etc.),
are not moved to the beginning of the sentence as in English.
However, some other kind of change (or movement) does take
place. As far as this syntactic movement is concerned, we can call
it a parameter which varies from language to language. It is still
universal, though all languages may not have it. Since it is universal,
there cannot be a human language which will allow movement to
take place without structure-dependency. If movement takes place,
it will be structure-dependent, otherwise it may not occur at all.
This means that we can classify languages according to which
principles of UG they share. These are called implicational,
statistical, or Greenbergian universals: Cook gives the example
of the presence of relative clauses in sentences (1988: 18-19).
Relative clauses are clauses which are introduced by that or who.
Data collected for a number of languages suggests that there is an
order according to which languages take relative clauses. They can
have only subject relative clauses, i.e. the subject of the clause is
related to the noun.
(15) Nelson Mandela is the man who remained in jail.

The relative clause ‘remained in jail’ and its subject ‘the man’ are
related to the noun ‘Nelson Mandela’. All languages permit this
kind of relative clause. Some languages also permit relative clauses
in which the object relates to the noun, and so on. This means that
languages can be arranged in a descending scale according to the
kind of relative clause they take:
Subject relative clause All languages
Object relative clause Most languages
Indirect object relative clause Many languages
Object of preposition relative clause Some languages
Genitive relative clause Few languages
Object of comparison relative clause Very few languages
This represents a hierarchical model in which we can arrange
languages on the basis of one variable—in this case according to
what kind of relative clause they can take. It is also called the
accessibility hierarchy. Note that this can also be represented as
follows:
Subject > Object > Indirect Object > Object of preposition >
Genitive > Object of comparison. (Here ‘>’ means ‘greater than’
or ‘more than’). This means that all languages allow subject relative
clauses, many allow object ones, and so on.
The language faculty
As we have seen before, language is structure-dependent. But there

are areas of the mind which are not structure-dependent. Thus, we
are able to think about things such as poetry and mathematics.
Children can learn how to walk. These activities seem to be
different from learning a language. The kind of logic used in
mathematics is different from the kind of principles used in
language. Thus it may be said that the human mind possesses the
faculty of acquiring language, and that this is a unique faculty, and
is not just part of normal intelligence as such. As we have seen
before, Chomsky calls this faculty the Language Acquisition Device
(LAD). Now let us see what light Universal Grammar sheds on

LAD.
When we learn a language, we use the LAD, which, according to
Chomsky and his associates, uses the principles of UG. But these
are general principles, and every language has its own particular
rules. The rules merely determine the choices with reference to the
values set out in the parameters. For instance, we have the UG
principle that there is a verb (V) at the head of a VP. Then we have
the choice that the verb (V) will either be at the left or at the right
of the VP. The rules of Urdu-Hindi say that the head of the VP
will be on the right, and the rules of English say that it will be on
the left. Thus, when children acquire a language, they make use of
the principles of UG resident in the LAD in their brains, and then
set the values of the universal parameters according to their
language. Thus, they learn the rules of their own language.
The basic thing to remember is the general principle of UG: We
all possess certain basic rules for processing human languages in
our brain at birth. When we hear a particular language, we set the
‘switches’ of our brain according to the particular rules of that
language. Since the basic rules are already in the brain, language
learning becomes easy for human beings.
12
Anthropological Linguistics
T he study of culture through language; the relationship between

language and culture; and how world view is related to
language are some of the concerns of anthropological linguistics
(or linguistic anthropology).
Franz Boas, the famous American anthropologist who studied
Native Americans (or Red Indians as they were called earlier),
wrote an article suggesting that anthropologists should know the
languages of the people they study (Boas in Hymes 1964: 15–26).
The obvious reason is that if s/he2 depends on an interpreter, s/he
will not be able to correct the bias of the interpreter or any
particular view which the interpreter might have acquired from the
people whose language s/he is interpreting. Another reason, which
is crucial, is that language expresses thought:
It has been claimed that the conciseness and clearness of thought of a
people depend to a great extent upon their language (ibid 17).
Boas thought that European languages have more terms for abstract
concepts so that they can express abstract, philosophical ideas more
easily than primitive languages. Nowadays, linguists do not agree
with such views because thought is too complex a phenomenon to
be explained by any one factor. Anyway linguistics became a part
of ethnology—the study of peoples’ culture, world view, and
behaviour (i.e. cultural and social anthropology).
Language was part of the study of culture even before Boas laid
emphasis upon it. Sir Edward Tyler, a British anthropologist, gave
2
S /he means she or he. It has been used for the generic masculine pronoun he
because feminists have raised the point that the use of ‘he’ alone ignores them.
importance to it in his books Primitive Culture (1871) and

Anthropology (1881). R. G. Latham, another Englishman, had
already written on the importance of language. Later, the famous
British anthropologist Malinowski wrote ‘Meaning in Primitive
Languages’ (1923) and Coral Gardens and Their Magic Part 2
(1935). In both works, he was concerned with the differences
between civilized and primitive languages. (See his supplement to
Ogden and Richards’ The Meaning of Meaning, 1923). As we have
seen, Ferdinand de Saussure influenced thinking in both sociology
and anthropology. As discussed earlier, Saussure differentiated
between la langue (the system of the languages, which is an abstract
thing) and parole (concrete utterances or instances of use of the
system). We shall study later how the anthropologist Claude Levi-
Strauss used Saussure’s ideas.
In the United States, much work was done on the American Indian
languages, not only in order to understand the languages themselves
but also to find out about the culture of the native tribes (ethnology,
ethno-semantics) through language. Franz Boas was a pioneer in
the field. He edited the Handbook of American Indian Languages
(1911) and pioneered courses at Columbia University. Edward
Sapir, another American linguist, was Boas’s student. Other
eminent linguists like Alfred Kroeber and Bloomfield were
influenced by him.
Edward Sapir compared languages and his contribution to theories
about the influence of language on thought (the Sapir-Whorf
hypothesis) is well known. Benjamin Lee Whorf is also associated
with this theory. Anthropological linguistics is still developing is
the USA, though the Chomskyan tradition is better known.
We have seen how structuralism came to dominate linguistics. The
structuralists emphasized upon the system—an abstract entity—but
this emphasis was not accepted by some socio- and anthropological
linguists. For Hymes, it was the speech act (parole) which was
important (1964: 63).
Anthropological Linguistics 129
Malinowski saw language as a mode of action, not an instrument

of mere reflection (1923: 312). His approach is called functional.
The function of an institution is the part it plays in fulfilling the
needs of individuals. Language maintains social solidarity. Even idle
chatter or meaningless formulaic utterances keep the channels of
communication open. This is called phatic communion and its
purpose is to bind people together.
Society can also be seen from the point of view of interacting
persons or groups. This is interactionism, and the ethnography of
speaking—how people actually use language in society—is in this
tradition.
The aims of anthropological linguistics
The aims of anthropological linguistics, as described by Dell

Hymes are as follows:
1. Evaluation of differences and similarities among languages
(especially exotic ones).
2. The significance of linguistic patterns for the basic outlook
of a people (i.e. their world view).
3. The relation between a people’s vocabulary and their other
interests.
4. How speaking enters into norms of interaction among
persons.
5. How the motives of play and art are manifested
linguistically.
6. The relation between levels or varieties of speech on the one
hand and types of community and their boundaries on the
other.
7. How social factors enter into linguistic change.
8. Modes of classifying and interpreting resemblances among
languages, especially their historical import (1964: xxiv).
Thus, linguists came to focus upon worldview, ‘acts of speech’—

how one uses language in the social context—or the ‘ethnography
of speaking’.
Levi Strauss’s use of linguistic concepts
Let us now see how Levi Strauss used some of the concepts of
linguistics to study culture.
One of Levi Strauss’s papers called ‘Structural Analysis in
Linguistics and in Anthropology’ gives us an insight into his
methods. He says that the concepts of phonology can be used to
study social systems such as kinship phenomena. As we have seen,
the relations between terms are very important in phonology. The
features voiced and voiceless (+ voice and –voice) are in binary
opposition. They distinguish many phonemes such as /p/ and /b/
etc. Similarly, two elements may be in complementary distribution,
i.e. they are the same thing but they take different forms. Thus,
allophones are the same thing. /p/ has two allophones, /p/ and /ph/.
We use /p/ in the end or middle of a word, and /ph/ in the
beginning, but they do not change the meaning.
These concepts of binary opposition and complementary
distribution etc. may be used in describing social systems such as
kinship relations. One of the ways in which this is done is
componential analysis, which will be touched upon later. Here I
shall concern myself with Levi Strauss’s study of the avunculate.
By the term avunculate, Levi Strauss meant the concept of the
relationship between uncles and nephews/nieces. An uncle can be
one’s (i.e. the ego’s) father’s brother or mother’s brother. There
may be a social system according to which this uncle’s relations
with his nephews/nieces may be strict, or informal and affectionate.
The uncle’s attitude as part of a system, and not only because of
his temperament, is what Levi Strauss is interested in.
To understand this, we can take the example of the institution of
the relationship between a newly married bridegroom and his wife’s
sisters in Pakistan and many parts of India. The man (doolah or
jeeja ji) can joke with his wife’s sisters and girl cousins (salian) and
vice versa. However, he is supposed to be strictly formal with his
parents-in-law. This joking relationship is allowed by society and
in some cases a day is fixed for it. On this day, the girls and the
groom indulge in playful ragging which sometimes leads to
frivolous horseplay (this is called chauthi, i.e. the event of the
fourth day in some communities).
This kind of system was studied by a number of anthropologists
including Radcliffe-Brown and Malinowski. It is this which Levi
Strauss takes up for analysis using concepts borrowed from
phonology.
If the maternal uncle is feared and exercises authority on his
nephew, the relations between them are strict. When this happens,
the relations between fathers and sons are familiar and affectionate.
If the relations of fathers and sons are strict, those between nephews
and maternal uncles are familiar.
In short if one pair of relations is known, the other can be deduced.
According to Levi-Strauss:
The two groups of attitudes form, then, as the phonologist would
say, two pairs of oppositions (Levi Strauss 1945: 45).
Thus +/– can be used to refer to this relationship.
But this is not all. These relations can further be seen in relation
to the relations between men and their wives, and brothers with
their sisters. Here, too, there are binary oppositions.
Societies which allow tender relations between husband and wife
(i.e. the public exhibition of intimacy), insist upon strict relations
between brother and sister. The reverse is also true: if brother and
sister are intimate and tender, husband and wife are supposed to
be formal and strict in public.
Now, if the relations between uncle/nephew and brother/sister are
tender (+ positive), then the relations between father/son and
husband/wife will be strict (- negative). Roughly speaking, this is
the position in some areas of Pakistan and parts of India.
This can be depicted in a chart as follows:

Wife – Husband
Sister Father
+ –
Brother Son
Uncle + Nephew
The law is:

The relation between maternal uncle (mamun) and nephew is to
the relation between brother and sister, as the relation between
father and son is to the relation between husband and wife.
Thus if mu + n + then b + sis +
f+s – then h + w –
(mu = maternal uncle; n=nephew; b=brother; sis=sister; f=father;
h=husband; s=son; w=wife)
Componential analysis
Kinship systems can be understood in general through the

application of linguistic concepts. A kinship system contains, in
essence, a certain relationship. The relationship can be by marriage,
by descent or by consanguinity (being brother and sister or cousins
etc.).
The idea of using binary oppositions to contrast meaning is the
basis of componential analysis. We have already come across this
idea in semantics. Let us now see how it is used in understanding

aspects of culture.
Goodenough and Lounsbury, for example, have constructed studies
of kinship systems using componential analysis (Goodenough
1956; Lounsbury 1964).
Before we deal with componential analysis, let us understand the
vocabulary, or metalanguage, used in it. Here are some basic
terms:
(1) Designata: It is a linguistic expression which designates a
class of images or concepts. The singular is designation and
the plural is designata.
(2) Significata: The significata are the necessary or definitive
attributes of this designated class. They are the properties of
the things which form that class. Other things which do not
have these properties are excluded
(3) Denotata: The denotata are the things we refer to.
There are other attached or associated meanings of the linguistic
expression called connotata, but componential analysis is not
concerned with them.
Let us take the series of sounds ‘dog’. This linguistic expression
designates a class of images. The attributes of the members of this
class are that they are animals with four feet, a long snout with
teeth, and a tail. They are domesticated and they bark. They are
members of the canine family, which wolves also belong to.
However, they are different from wolves, jackals etc. They are also
different from cats and horses etc. These are the significata of dogs.
Thus, the significata are the necessary attributes of a dog. So when
we use the word ‘dog’, it denotes animals having the attributes
mentioned above as significata. All animals which belong to the
canine family, which bark, and are domesticated are called dogs.
The term ‘dog’ also connotes faithfulness or a low despicable nature
and may be used for human beings. That is its connotation.
Let us now list the denotata of ‘uncle’. If we take all kinds of people
called uncle. They are:
mother’s brother
father’s brother
mother’s sister’s husband
father’s sister’s husband etc.
Let us now abstract the essential distinctive features:
Uncle + male
+ senior [i.e. of a senior generation]
+ 2 [i.e. 2 degrees of genealogical distance
from ego]
– lineal [i.e. one has not descended from him]
– marriage tie [i.e. ego is not connected by marriage ties,
but ego’s parents may be].
As you can see, if any of these terms are changed, we get someone
else who cannot be called ‘uncle’. The word or words used for
‘uncle’ may be absent or may mean different things but we can find
out what a term actually refers to.
Before going ahead, however, let us explain the abbreviation we
will use in this analysis. The capital letters are used as follows:
F = father, M = mother, B = brother, S = sister, H = husband,
W = Wife
The small letters stand for:
s = son, d = daughter
Now let us see how these symbols and used in componential
analysis:
The anthropologist Goodenough wrote about the American Indian
(lroquois) society in which one’s sex determined whether one’s
brother’s daughter was to be called ‘niece’ or ‘daughter’. This is
very different from the Indo-European languages, but it is a way
of classifying reality which becomes clear in terms of componential

analysis. The formulas are:
(1) For senior generation kin:
Is there a sex equivalence between ego’s linking parent and
alter?
(2) For junior kin:
Is there a sex equivalence between ego and alter’s linking
parent?
(3) Same kin of the same generation:
Is there a sex equivalence between ego’s linking parent and
alter’s linking parent?
In these formulas, the key terms are defined as follows:
sex equivalence = having the same sex. For example, nephews
and uncles have the same male sex.
ego = the person from whom a relationship is
traced.
alter = the person who is referred to in the kinship
term.
linking parent = the parent who provides a link. For instance,
in father’s brother as well as father’s brother’s
son etc. the linking parent is father.
Let us take the senior generation first. The question whether the
term father can be used for more than one male of this generation
can be determined by answering the question: Is there a sex
equivalence between ego’s linking parent and alter?
Let us take a boy as ego. The relation will be traced from him. Let
the alter be his father (i.e. the biological mate of his mother). The
linking parent too is the father. As we can see the linking parent
and the alter are the same person. We can say that there is a sex
equivalence between the linking parent and alter (both being male).
Thus, the boy will call his real father as father.
A girl too will call her father as father since the linking parent and
alter is the same person in this case too.
Now let us see how many other relatives are linked by the linking
parent, i.e. father. These are:
F B Father’s brother
F Mss Father’s mother’s son’s son
F F Bs Father’s father’s brother’s son
F M Bs Father’s mother’s brother’s son
F Fss Father’s father’s son’s son
F F F Bss Father’s father’s father’s brother’s son’s son etc.
All these males are linked by the male linking parent to the ego
(the boy). Since there is a sex equivalence between the linking
parent (father) and the alter (FB etc.), all these males will be called
father by him.
Now let us take the terms for junior kin. The question is who shall
be called son, daughter, nephew or niece. The question to be asked
is whether there is sex-equivalence or non-equivalence between ego
and alter’s linking parent.
Now suppose the ego is a man. The alter is a boy of the younger
generation (< generation + male – adult). This man is linked to the
boy by his brother—the boy being his brother’s son (Bs). The ego
(man) and alter’s linking parent (his brother) are both males. Thus
there is sex-equivalence between both. Hence, the boy will be called
his son.
Now consider the same man again. Let the alter again be a boy of
the younger generation. However, this boy is the son of his sister
(Ss). The ego (man) and alter’s linking parent (his sister) are of
different sexes. Thus, there is no sex-equivalence between both.
Hence, the boy will be called his nephew and not his son.
Now let us take the case of a woman, i.e. ego is female. The alter
is her own biological son. The ego (woman) and the alter’s linking
parent (her own self) are of the same sex. Thus, there is sex
equivalence between the two and her son will be called her son.
However, her brother’s son becomes a nephew, while her sister’s
son is again her son.
The terms are given as follows:
For male ego:
son : s, Bs, MSss etc.
For female ego:
son : s, Ss, MSds etc.
Let us take the last case now—that of the same generation (i.e. =
generation). Here, the question to be considered is whether there
is a sex equivalence between ego’s linking parent and alter’s linking
parent.
If the ego is a boy and the alter is a boy who is his father’s brother’s
son (FBs), the linking parent is the father. As the ego’s linking
parent (his father) and the alter’s linking parent (the alter’s father)
are of the same sex, he will be called brother. If the alter was his
father’s sister’s son, then he would have been called cousin and not
brother, because the linking parents would be of different sexes.
Lounsbury’s use of the term ‘parallel’

in componential analysis
According to Lounsbury, we can simplify our formulas even further

by using the term + parallel instead of sex equivalence. The idea
on which + parallel is based upon is called complementary
distribution. It is often used in phonology. The theory is that if
two or more linguistic elements are (1) similar in form, and (2)
non-contrastive in function, they are variant forms of the same
element.
Thus, as we saw in the chapter on morphology, the morpheme –s
is pronounced as /s/, /z/, and /ız/ in cats, dogs, and buses respectively.
These are variant forms of the same morpheme in complementary
distribution. This means only one can occur in a given situation.
They are not distributed at random, but are distributed according

to fixed laws. Thus, after the unvoiced ending /t/, only the
unvoiced variant /s/ occurs. After the voiced ending /g/, the voiced
ending /z/ occurs. After the hissing sound (sibilant) /s/, only /ız/
occurs. They complement each other being the same thing, but
their distribution follows certain laws.
In the same way, the sex equivalences too are really the same thing
but it has three kinds of distributions: (1) for senior generation (>),
(2) for junior generation (<), and (3) for the same generation (=).
If we call it + parallel, we are using one term for something which
is in fact a single phenomenon. By + parallel we mean equivalence
of sex and by—parallel we mean non-equivalence of sex. But the
formula is a little more specific:
+ parallel means equivalence of sex between the two kin of
the generation above ego or alter (whichever is
junior).
– parallel means non-equivalence of sex between the two kin
of the generation above ego or alter (whichever is
junior).
As we have made a single formula to cover all the relationships
considered earlier, it appears somewhat complex. However, it is
quite clear if we work it out.
Let us take only one example: father
It is > generation + parallel + male
Let us concentrate only on + parallel. According to the definition,
there should be equivalence of sex between the two kin of the
generation above ego or alter (whichever is junior).
Let us try out: F, F B, F M Ss, F F Bs only
ego = a boy
alter = his biological father
The junior is ego in this case (the boy). The two kin of the
generation above the ego are the father and the linking parent who
is also the father. Thus, the father will be called father.
Now take FB. The junior is again the ego. The two kin of the
generation about the ego are the linking parent (his father) and the
father’s brother. Since they are of the same sex (+ parallel), FB will
be called father.
Now take FMSs. The junior is the ego, who is the boy. The two
kin of the generation above the ego are the linking parent (his
father) and the father’s mother’s sister’s son. They are of the same
sex. Hence, the alter will be called his father, and so on and so
forth. If you work out all the formulas carefully, you will
understand them. Some of them are given below:
mother : kin > generation + parallel – male
uncle : kin > generation – parallel + male
son : kin < generation + parallel + male
cousin : kin = generation – parallel
sister : kin = generation + parallel – male
(Adapted from Leech 1974: 246).
Kinship analysis has a mind teasing quality but that is because of
the number of relationships involved. In any case, most societies
do not have such complex formulas of kinship. The point to
remember is that linguistic theories can be used to understand
kinship systems.
13
Sociolinguistics
L anguages are spoken by human beings who live in societies.

The study of language in relation to society is called
sociolinguistics. This is a vast subject and the present chapter
cannot even introduce all aspects of it, let alone do justice to them.
This is because there are so many subjects one can investigate that
only a very brief list of some important ones is possible here.
Among these subjects are: the varieties of language; language and
world view; language and gender; linguistic politeness; and identity
and language planning. Some of these subjects, like the relationship
of language with identity and language planning, are so vast that
they are covered under a different sub-heading of their own. Ethnic
identity and aspects of language planning are covered under
language politics in this book. Subjects not covered here are the
language of schooling and issues of the medium of instruction
which are focused upon in educational linguistics. The other issue
not dealt with here is the analysis of conversation and the
ethnography of communication. Ethnography of communication
is also part of anthropological linguistics. Indeed, there are overlaps
between anthropological and sociolinguistics because both focus
upon human beings living in interaction with each other.
Let us then understand some basic concerns of sociolinguistics.
First, do human languages vary? That is, when we say ‘the English
language’ do we mean just one set of words in a single pronunciation
and a grammar which does not vary at all? If not, then what is
English? Let us understand the theory of varieties of language
before we answer this question.
Sociolinguistics 141
Varieties of language
What we call languages are merely abstractions. Going back to our

example of English, have you ever seen it? Does it exist? Well,
dictionaries exist and books of grammar exist but the spoken
language came first and these written products came much later.
When it was a spoken language only, speakers existed and what
they spoke could be called English if other speakers felt they were
using the same sounds for the same objects. But, while doing so,
speakers ignored the minor differences in the way other speakers
pronounced words or used grammatical rules. If these differences
are minor—such as the pet words a person might use—we ignore
them or call them personal idiosyncrasies. Thus every individual
uses a different (however minor the differences are) form of a
language which we call his or her idiolect.
But sometimes we can understand the other speakers while being
aware of quite clear differences in pronunciation, meanings, and
even grammar. These are called the varieties of the same language.
The varieties can be classified as below:
Language
According to Use According to User
Register
Idiolect
Diglossia Dialect Sociolect
Non-Native Varieties
Variety
The variety of a language according to its use is dependent upon

its function and is called register. If you use it to teach medicine,
you will put in it the special language of medicine (its technical
jargon), derived ultimately from Latin and Greek. Thus, a fat
person will be called ‘obese’ and inability to sleep will be diagnosed
as ‘insomnia’.
Similarly there is a theoretical language—called a metalanguage—
for all academic subjects as well as different professions. Thus, in
the language of the British Indian army ‘RV’ was used for
‘rendezvous’. But ‘rendezvous’ means meeting place, while in the
register of the British Indian army it meant the place and the time
of the meeting.
Remember that the register is meant for a certain use. It is not used
outside that particular subject or profession. Thus, only a pompous
or incompetent person will write and speak to everyone in the
jargon of his profession or specialization. The best public lecturers
are those who can hold people’s attention by not using the register
of their specialization.
A special situation arises when a language has a specific high
variety (H) which is used in all formal domains, while ordinary
conversation is in the low variety (L). This concept was called
diglossia by the linguist Charles Ferguson (in Hymes 1964:
429–439). Ferguson referred to the use of Arabic in the Arab
world, Modern Greek, Swiss German, and Haitian Creole. He
made the point that the H variety is nobody’s mother tongue nor
is it used in ordinary conversation by anybody. It is learnt in
schools and used in the class rooms, offices, courts, places of
worship etc. The H variety has got great prestige because it is the
language of classical literature and religious texts. Moreover, it
unites people at least at the formal, written level. However, the L
variety is the real living language of the common people. It should,
however, be remembered that L is not made up of one language
but has many varieties. We now take up the case of varieties
according to the user.
As mentioned earlier, the idiolect is the special variety used by an
individual. The dialect is the variety of a language used by a
geographical area. Thus, the Pashto of South Waziristan, which is
different from that of Peshawar, may be called a dialect of
Pashto.
The important thing about dialects is that they are mutually
intelligible. People who speak one dialect can understand the
speakers of another dialect. It is, in fact, a variety of the same
language and not a different language.
However, there are two views about dialects: (i) the traditional or
popular view, and (ii) the modern linguistic view.
(i) The traditional view

According to this view, the standard language is the one which is
used by educated people while writing books, lecturing, and in
formal conversation. It is also used in the domains of power such
as higher education, administration, judiciary, armed forces, and
politics.
The dialects are spoken by uneducated people. They are
stigmatized—i.e. seen in a negative way—by educated people. They
are considered rough (uncouth), rustic, sub-standard, and inferior.
They are considered intrinsically inappropriate for education and
literature etc. Some people consider them distorted forms of the
standard language, which is called the ‘real’, ‘actual’ or ‘pure’
language. They are also area-bound because they are used only by
the people of a certain area. For instance, Maldivian is used in the
Maldives islands.
The traditional view is hierarchical, i.e. people tend to regard some
languages, or rather varieties of them, as superior to the others.
They develop contempt for the speakers of these stigmatized
varieties. Indeed, the fact is that they already have contempt for
the speakers and this is transferred to the language they speak. Here
is an example of such an attitude.
Slovenly and careless of speech, these Gullahs seized upon the peasant
English used by some of the early settlers and the white servants of the
wealthier colonists, wrapped their clumsy tongues about it as well as
they could, and, enriched with certain expressive African words, it
issued through their flat noses and thick lips as so workable a form of
speech that it was gradually adopted by the other slaves and became
in time the accepted Negro speech of the lower districts of South
Carolina and Georgia.
Gonzales 1922:10. Quoted from McDavid 1980:78
The Gullahs were the African-Americans who were brought as

slaves to the American cotton-growing states of the south. The
writer’s contempt for the Africans is obvious and it is not for the
language alone. It is also for their ‘thick lips’ and ‘flat noses’ and
their dark complexion. Moreover, there is also contempt for the
peasants from England and white servants, i.e. all the subordinated,
working class groups from all societies. Arguably, it is this contempt
for these groups which extends to their speech.
This example is from America of the 1920s but we in modern
South Asia consider several of our languages inferior and make fun
of them. Think about which language(s) you consider superior and
which you think of as inferior. Then sit down and try to understand
whether the language, or variety, you consider inferior is spoken
by people who do not have a high social status in the eyes of the
community you belong to.
One major objection to the traditional view of languages is that it
is related to bias and encourages prejudiced attitudes in society. Let
us now come to the other view.
(ii) The modern linguistic view

According to this view, the standard variety of a language is also a
dialect. Thus all dialects are basically varieties which are mutually
comprehensible. A language is, then, merely the sum total of all its
dialects. This can be represented as follows:
L = d1 + d2 + d3 + ———————— dn
(where L = language and d1…dn are its dialects. The number may
vary, so we write ‘n’ which may represent any number of
dialects.)
As you will observe, dialects are real in the sense that people
actually speak them. The language (L) is an abstraction because
nobody speaks it. Everybody speaks a dialect. If someone speaks
what we call the ‘standard language’, even that is a dialect of the
abstract entity called ‘the language’.
If you look at the historical development of languages, you will

discover that the standard variety was developed by language
planning. The decision to make one dialect the standard was a
political one—it happened to be the language of those who had
power and money or for some other such non-linguistic reason.
Thus, the language of the upper class of English around London
was standardized. Standardization means fixing spellings,
pronunciation etc. and writing dictionaries, grammars etc. Thus,
languages are standardized for non-linguistic reasons.
Another point you should keep in mind is that all dialects are
equal. Their structure is similar and there are no scientific reasons
for calling one of them superior or inferior. However, the standard
variety is meant to serve many functions—being used in education,
law, commerce etc.—and has a larger vocabulary than other
dialects. This means that the standard variety is functionally more
useful for the purpose of modern life. However, if any other dialect
had been standardized and modernized, it too would have been
functionally just as useful.
As linguists, we say that everybody speaks or writes a dialect or
variety, and by language we mean the sum total of all mutually
intelligible dialects or varieties of a language. If we want to use the
term ‘dialect’ in its traditional sense, we may call it a non-
standardized variety of a language.
You should also remember that for non-linguistic reasons—ethnic
identity, nationalism etc.—mutually intelligible forms of speech
may also be called languages. Thus Hindi and Urdu are mutually
intelligible (except in their extreme Sanskritized or Persianized
varieties) but are called different languages. The languages of
Sweden and Norway etc. are also quite similar but are called
languages for nationalistic reasons. This may not be strictly correct
in linguistic theory, but it is an established practice.
Now are we in a position to identify the English language? Well,
but we come to the strange conclusion that there is no such thing
as English at all. There are several varieties which are mutually
intelligible which we add together to call their sum total—English.

There is, for instance, British Standard English, American Standard
English, Australian Standard English, South African Standard
English, New Zealandian Standard English, and Canadian Standard
English. These are the standardized varieties of native-speakers’
Englishes. But then within Great Britain, there are several non-
standardized varieties (dialects) such as Cockney, Scottish English,
Yorkshire’s English etc. In the USA too, there are different varieties
of English as the book Varieties of American English (McDavid
1980) testifies. And varieties are not associated with region only;
they are also associated with class. In England, the upper classes
speak what is called RP (Received Pronunciation). In America, the
English of African Americans is called ebonics or African Vernacular
English (AVE), and it is different from the English of middle class,
white Americans. But the story does not end with native-speakers
only. English is the second language of the educated elites of South
Asia also. They use it with varying degrees of competence, but rules
of their own first languages interfere with the way they speak and
write English. These are non-native varieties of the language. In
short, the idea of the English language makes sense only if we call
it the sum total of all its varieties. But if it is the sum total of all
its varieties, then there is no one monolith or entity which we can
call English. In short, languages do not exist; only their varieties
exist. Languages are an abstract concept; varieties are concrete
systems which can be studied.
Language and world view
One of most interesting theories about culture and language came

from Benjamin Lee Whorf (1897–1941) and Edward Sapir (1884–
1939). The theory which is associated with their name is called the
Whorfian hypothesis or the Sapir-Whorf hypothesis. It is also
called the theory of linguistic relativity. It has two forms: (a)
strong linguistic determinism, and (b) weak linguistic relativity.
Form (a) asserts that language completely determines thoughts.
Thus, your world view is so strongly determined by your language
that you live in a different universe from speakers of other

languages.
Form (b) argues that language makes the expression of some
thoughts easier than others. Thus, we can understand other cultures
but may find it difficult to express the concepts of these cultures
in our language.
According to some linguists—(Hill 1988: 15)—Sapir and Whorf
did not give the strong form at all. Some of their statements were
a bit exaggerated and people thought they had given the strong
form. As for linguistic relativity, there have been many experiments
to confirm to what extent human languages are variable and how
we can relate this variability to culture. The basic idea is to find
out whether language reflects the world view of the culture which
created it and uses it. This is, in fact, the cultural determinism of
language. Then there is the question of finding out whether a
language makes us think in a certain way, i.e. determines our world
view? The second question, at least in its extreme form, has been
ruled out. Our experiments will then confirm or disconfirm only
the weak form of linguistic determinism.
One of the major ways in which linguistic relativity was tested was
by comparing colour terms in different languages. Berlin and Kay
(1969) gave 329 coloured squares to speakers of 98 languages with
instructions to arrange them into the basic colours of their
languages. By a basic colour they meant something which was not
a mixed colour like ‘reddish-brown’ or ‘olive’. They found that
colours were found distributed in languages according to certain
rules. If a language had only two colour terms, they would always
be black and white. If a third existed, it had to be red. The next
category of languages had a choice between yellow or either blue
or green (Berlin and Kay called it ‘grue’). The next category would
have blue and green. Then brown was added and the last category
had 11 colour terms and these were purple, pink, orange, and grey.
This phenomenon was called implicational universals by the two
researchers. The idea was that colour terms is a universal
phenomenon and is predictable, so that if there are three terms you
can be sure that this implies they would be black, white, and red,
and not blue, purple, and orange. This implies that the human
mind is programmed to think in this manner and that the strong
version of the Sapir-Whorf hypothesis is wrong. Our perception is
not determined by our language alone. It is also determined by
some common human ways or abilities of understanding reality
(cognitive universals). However, speakers do categorize what they
see in colour terms which they possess. Human beings see the same
colours but refer to them the words of classification they possess.
This supports the weaker version of the Sapir-Whorf hypothesis
that language does influence perception.
Language and colour terms
If a language does not have a name for a colour, its speakers

consider it the shade of another colour for which there was a name.
In Pashto, for instance, the word sheen is used for green and blue.
This does not mean, however, that green and blue are seen as
exactly the same. They are seen as different shades of a colour
which comes under the broader category of sheen. Sometimes, one
could refer to grass to give the idea of greenness or to the sky for
blueness. However, the way the colour spectrum (i.e. the outside
world) is cut up is somewhat different from a person whose
language has different terms for blue and green. Likewise, we
consider scarlet, pink, shocking pink, purple, and other shades of
redness as red. However, another culture might have a language
with words which regard them not as shades of red but different
colours. Then the speakers of this language will define pink and
scarlet clothes not as shades of red but different colours.
The order is as follows:
purple
white green
< red < < blue < brown < pink
or
orange
black yellow
grey
The sign < means that the colour (s) towards which the tip of the
arrow points will be there in a language containing the colour on
the broad end of the arrow. Thus, a language which has brown will
certainly have blue and the other colours on its left.
Let us order them according to the chart of Berlin and Kay:
Type of Number Colours Example
Language of Terms
1 Two white, black Jale (New Guinea highlands)
2 Three white, black, red Tiv (Nigeria)
3 Four white, black, red, green Hanunoo (Philipines)
4 Four white, black, red, yellow Ibo (Nigeria)
5 Five white, black, red, green, Tzeltal (Mexico), Pashto
yellow (Pakistan and Afghanistan)
6 Six white, black, red, green, Plains Tamil (India)
yellow, blue
7 Seven white, black, red, green, Urdu and Hindi (South Asia)
yellow, blue, brown
8 Eight, white, black, red, green, English
Nine, Ten yellow, blue, brown,
or Eleven purple, and/or pink, and/or
orange, and/or grey
Adapted from Leech 1974: 235
As you can see, types 3 and 4 are different only in one respect. In
type 3, green is the fourth colour, whereas in type 4 it is yellow.
Both are four-term languages. Similarly, there can be no ordering
in purple, pink, orange, and grey. Any set of colours can be there
in any language.
Berlin and Kay also gave an evolutionary hypothesis: that languages
evolve from having fewer terms for colours to having more. The
idea is that as society becomes more complex, it needs to
differentiate colours more precisely so that what were considered
shades of a basic colour are given the status of colours. Thus, new
terms for colours are added to the vocabulary.
Can you think of additions of colour terms in Punjabi, Tamil,
Singhalese, Nepalese, or any other language of South Asia? Arrange
the colour terms in these languages and see if you confirm or
disconfirm Berlin and Kays’ hypotheses.
Kinship terms and world view
Let us see whether our world view is reflected in our kinship terms.
In English, we have just the term ‘cousin’. If we want to be precise,
we can call girls ‘female cousins’ and boys ‘male cousins’. See the
English sentence below.
(1) I met my cousin. (English)
Sentence (1) dies not tell you anything whether this cousin was
male or female and related from father’s or mother’s side. Now see
sentence (2).
(2) Main khala zad bahen se mila. (Urdu)
Literal translation: I met my mother’s sisters’ daughter (who is like
a sister for me).
Meaning: I met my mother’s sister’s daughter.
Sentence (2) makes it clear that the speaker met a female related
to him in such and such a way. This only implies that language,
such as the possession of colour terms or kinship terms, facilitates
the expression of certain thoughts. It does not make it impossible
for us to express them but we may not have terms corresponding
to what we have in our mind. This is what makes translation such
a difficult thing. This supports the weak version of linguistic
determinism but not the strong one.
What is especially significant is that we can find out something
about a culture through its language. We can find out about its
world view for instance. If there are many terms for ‘cousin’, as
there are in Siraiki, it must be a culture where the family, or
kinship, is given great importance. In other words, it would not be
a modern individualistic (Western) kind of culture.
Language and gender
Sex is a biological reality, i.e. human beings are mostly born as male
and female. Some very few individuals may be born as having the
physical features of both sexes. Yet others may be psychologically
inclined to feel like females, even when they have male sexual
organs and vice versa. However, these are exceptions. Gender, on
the other hand, is how sexual differences are played out in societies.
It refers to the way a female or male is expected to behave. In our
traditional South Asian cultures, there was even a niche for
eunuchs, hermaphrodites, and transvestites or people who, for any
reason, physical or psychological, did not conform to the expected
gender roles of males or females. In short, sex is biologically
determined, but gender is constructed by us human beings in our
societies. Why we say we construct it is because when we bring up
our children, we tell them that boys do not keep long hair and girls
get their ears pierced to put earrings in them. We say men do not
wear pink clothes and girls do not go about whistling at boys. In
other words, we socialize our children into the expected gender
roles we want them to accept and play out in our societies.
As anyone with any historical knowledge will tell you, boys and
men did have long hair in Europe up till the eighteenth century.
In South Asia, Sikh men do not cut their hair so they also have
long hair. Some Muslim Pashto-speaking tribesmen in Waziristan
have long hair. They also put kohl in their eyes, which is thought
to be something effeminate by people living in the city. In the past,
in some parts of the world men did get their ears pierced and one
can see a number of faqirs with ear rings in South Asian cities.
Some sadhus and malangs also wear bangles and rings for good
measure. We also know that purple was the colour of kings, and
some bishops and cardinals sometimes wear shades of red even
now. In other words, these stereotypes of masculinity which we
take to be fixed are only recent phenomena. They are European
constructs of the 19th century and not universal or natural in the
sense that not all human beings believe them to be true now nor
did they so in the past.
The gender roles which we enforce upon children are accepted as
being natural by them, but the fact that they differ from society to
society and time to time suggests that they are largely constructed.
According to the anthropologist Margaret Mead, who studied
tribes of New Guinea between 1931 and 1933, the gender roles of
men and women there were different from Western societies. The
Arapesh men, for instance were cooperative, unaggressive, and
caring while Mundugumor women were aggressive and ruthless. In
the Tchambuli tribe, the sex roles were as follows:
We found a genuine reversal of the sex-attitudes of our own culture,

with the woman the dominant, impersonal, managing partner; the man
the less responsible and emotionally dependent person.
Mead 1935:190
In short, the men were like our stereotype for women. As she put
it in the preface for the 1950 edition, the men ‘are catty, wear curls,
and go shopping, while the women are energetic, managerial,
unadorned partners’ (Mead 1935: Preface to the 1950 edition).
Those who have read tales (dastans) will remember that the men
in these stories are romantic lovers who faint away often only by
looking at the beauty of the beloved. They cry easily as powerful
emotions grip them and they are apt to woo their mistresses
through long speeches in verse quite as much as performing daring
feats for them. The rakes (bankas) of colonial Lucknow wore very
fine clothes and went about scented as, indeed, did the dandies of
Paris in the 18th century and London of the decadent 1890s which
is called the fin de siecle (end of the century). In fact, wearing some
kinds of scents is considered auspicious by Muslims and men put
on itar on Eid and when they visit the tombs of the great sufi
masters like the Dargah of Sheikh Nizamuddin Auliah in Delhi or
that of Sheikh Bahauddin Zakariya in Multan.
Gender, then, is constructed. Language plays a role in constructing
gender. In our male-dominated societies, this role is negative as far
as women are concerned. Following is a list of devices that are used
in English for registering bias against women. They are taken from
Lia Litosseliti (2006: 14–15). This list can be used for our South
Asian languages as the examples given below indicate.
1. Sex specification: The sex is specified when one mentions a

trade, profession or role. The English term ‘authoress’ is given as
an example of this. We now use the more neutral ‘author’ without
specifying the sex of the person in question.
2. Gratuitous modifiers: These are said to diminish prestige and
draw attention to the sex. We often use the term ‘lady doctor’ in
our hospitals so that when just the word ‘doctor’ is used it is
assumed that we are talking about a man. This makes women feel
that while it is normal for a man to be a doctor for them it is so
unusual that their sex is specified. Similarly the term male nurse is
used since just nurse refers to a female nurse.
3. Lexical gaps or under-lexicalization: Not having enough words
to describe the same kind of behaviour in men and women. For
instance, both men and women can be promiscuous. Yet, the terms
for women of this kind are far more than those for men.
4. Semantic derogation: A word when used for a woman gets
degraded while equivalent terms for men do not. Just look at the
term ‘mistress’ which is the equivalent of ‘master’. But mistress has
come to mean a woman kept for sexual purposes, whereas master
still means one who owns or is in control of something.
Incidentally, this happens to terms for low-prestige professions also.
For instance, the term ‘mehtar’ means exalted in Persian. But when
it came to be used for sweepers in parts of north India, it came to
be degraded. The term ‘jamadar’, who was an official with
considerable prestige, also came to be used for sweepers and was so
degraded that the army officials called jamadars were embarrassed
by it and the term ‘subedar’ has been used for them since many
years in Pakistan.
5. Asymmetrically gendered language items: Single words used
for women without similar words for men. For instance, the title
‘Mrs’ was used for married women, but there was nothing to show
that a man was married. Similarly, ‘Miss’ was an unmarried woman
but an unmarried man, in common with a married one, was
referred to as ‘Mr’. This was objected to by feminists because it
meant that women were identified in relation to men (i.e. through

marriage with them) but men were not.
Other such items are ‘chairman’, ‘policeman’, ‘fireman’ etc. These
refer to the male gender and now they have been changed to refer
to females also, such as ‘chairperson’.
6. Connotations of Words: A word may mean one thing literally
but have negative implications. The first is its denotative meaning
and the second its connotative meaning. For instance, the word
‘hot’ means having a high temperature. However, its connotation
is that the person is sexually attractive. For a girl to be called ‘hot’
will be taken as an insult in most situations in South Asia. Even
the terms ‘girl’ and ‘girlish’ may imply immaturity rather than just
femaleness.
Basically, this bias is called sexist language, i.e. language which
discriminates against women on account of the gender roles
expected from them and the stereotypes these have given rise to.
Recently, feminists have started opposing such kind of words
because they depict women as sex objects or as mothers and
nurturers. They also trivialize and judge women. This means that
they do not take women seriously and pass judgments upon them
related to their gender, rather than related to what they have or
have not done. For instance, if men talk loudly, they are called
friendly, but if women do so, they are stigmatized as non-serious
chatterers.
An attempt to create an unbiased language, at least as far as English
is concerned, is the opposition to androcentric generics (andro =
man; centric = centred upon; and generics are general terms such as
‘mankind’, ‘manpower’ etc.) The feminist argument is that such
generics refer to men on the assumption that women are
automatically included. However, this makes women appear as
special cases or makes them invisible. Thus, the words which they
substitute instead are ‘humanity’, ‘personnel’ or ‘workforce’ etc.
This campaign has succeeded in so far that scientific journals issue
guidelines not to use sexist language. However, the ideological
forces which subordinate women, as scholars have pointed out are

changing rather slowly (Cooper 1989: 14–21).
Most languages have ways to show gender. For instance, in English
if we say:
(1) She went.
(2) He went.
(3) It went.
The pronouns she and he refer to the female and male gender
respectively and it refers to neutral. In Urdu-Hindi, the sentences
would be like this:
(4) Vo gaya.
(5) Vo gayi.
and we cannot have the equivalent of (3) because everything has a
grammatical gender in Urdu-Hindi. So if we say:
(6) Garmi gayi. (The summer went)
then garmi has a feminine grammatical gender, while jara /dj a ɽa/
is masculine. So if we want to say The winter went, we would
say:
(7) Jara gaya.
However, another word for winter is sardi, and it is feminine. The
sentence would then be:
(8) Sardi gayi.
Now consider Persian.
(9) Oo raft (She/he went)
In this language, neither the inflexion of the verb nor the pronoun
indicates the gender. In short, different languages indicate gender
differently, but the relationship between this and the way women
are actually treated in society is too complex to be determined by
language alone. Just because Persian refers to man and women in
the same manner does not mean they are treated exactly alike in
society.
Another aspect of gender and language is the study of Women’s
Language (WL). The classic study of this is by Robin Lakoff who
says it has certain special features such as tag-questions (isn’t it?),
rising tone on declaratives, words like ‘charming’ etc. (Lakoff
1975). The idea was that women are weak as a collectivity, and
therefore use the language of persuasion rather than that of
authority and strength. Later some scholars suggested that women
are more cooperative as conversationalists and care for the ego of
the other person (Coates 1996; Holmes 1995; Tannen 1990).
However, in Western sociolinguistic writings, these things are not
seen as virtues. They are associated with women being weaker and
subordinate. This makes many women want to speak more
aggressively like they think men do.
It seems that the kind of style we associate with the traditional
upper class culture of Lucknow of a century ago had some of the
features of WL. It was considered polite and gentlemanly, not
effeminate or womanly. This area needs more research.
However, some features of WL do stand out in South Asia at least
among traditional sub-cultures. Evidence of this comes from a
genre of Urdu poetry called rekhti (not rekhta which is another
word for Urdu itself). This poetry was written by men in order to
enjoy themselves at the expense of women. Thus they present
stereotypes, and not real women. However, real women of this
period in certain social circles did use the words reproduced
below:
hariyan used to show disapproval for a man or thing
zanakhi female friend
naoj to show disapproval
mardua man (pejorative use)
ooi oh!
hae oh!
The WL shown in Urdu writings is not politer than the language

of men. It does, however, have special words used only by women
and fewer Persian and Arabic words than men of the same social
classes used. That is why the reformers of that period such as Altaf
Hussain Hali and Maulana Ashraf Ali Thanvi urged that women
should be taught standard Urdu, the style used by educated
gentlemen, rather than the begmati zaban (the language of ladies)
(Minault 1998: 69 and Minault 1984). In short, WL was not seen
by Indian Muslim reformers as a necessary attribute of gender.
They saw it as a bad habit which education and training could
change.
While these stigmatized and much ridiculed markers of WL in
north India have disappeared, some special features remain even
now. In Pakistan and India, women use ‘Uf’, ‘Hae’, ‘Hae main mar
gayi’, ‘Hae Allah’, ‘Hae Ram’, and so on.
Uf Oh
Hae Oh
Hae main mar gayi ! Oh! I am dead!
Hae Allah ! Oh God! (Muslim women)
Hae Ram ! Oh God (Hindu women)
Hae Guru Ji ! Oh God (Sikh women)
Such expressions tend to differentiate the sexes and, more
significant in terms of gender roles, they are interpreted as
indicators of lack of power and courage. They are used in distressful
situations which are supposed to be faced with stoicism by men.
Because of these associations, these expressions are not used as
much by Westernized women as they are by those who live in
traditional sub-cultures.
Another aspect of language about women in parts of South Asia is
that women are concealed from unrelated men. This segregation,
whether in the form of taking the veil or modest segregation from
men, is part of traditional sub-cultures in this part of the world.
This is expressed in language by the circumlocutions for ‘wife’
which exist in Urdu, Punjabi, Pashto, and English as used in

Pakistan:
Urdu ahle khana; ghar vale; andrun-e-khana
Punjabi kar vale; karon; bachche
English family; ladies (not lady)
Pashto kor valan; kor
Verbal politeness is expressed by males by not mentioning the
other’s young female relatives. For instance, one man asking
another about his wife’s health will either not make the enquiry at
all or, if he does so, it will be in some such form (as evidenced in
Pakistan):
(1) Family theek hai? (Is the ‘family’ all right?)
(2) Ghar vale theek hain? (Are the people of the house fine?)
(3) Hamari bahen theek hain? (Is my sister all right?)
(4) Aap ki Mrs theek hain? (Is your Mrs all right?).
Language and gender can be studied in relation to the media, the
workplace, the classroom, and in discourses of various kind. The
scope of research in this field is tremendous because South Asian
societies are in flux and there is so little research on such issues.
Linguistic politeness
Politeness is expressed through language as well as non-verbal

expressions such as smiling, bowing, bending the body, folding
hands, inclining the head downwards, and looking pleased. These
gestures are called body language, and they express attitudes and
emotions of various kinds. In this section, however, we will give a
brief introduction only to linguistic politeness, not to non-verbal
expressions of it.
The classical study of linguistic politeness is Brown and Gilman’s
famous work on the pronouns of power and solidarity (1960).
Basically, they argue that there are two basic forms of pronouns:
tu (T) forms and vos (V) forms. Both occur in Latin and the former
stands for tu/ tum in Urdu-Hindi while the latter is the equivalent
of aap. They claim that the tu form was used for all till the Roman
emperor became so exalted above other people that the plural form
(vos) came to be used for him to show his special status. When
people use the T form with each other, they show solidarity or
intimacy; when they show the V form they are showing respect but
also distance. When one uses T but receives V in response, then
the user of T is less powerful in relation to the user of V. This is
non-reciprocal use of T and V. In our hierarchical societies,
children use V forms for parents and elders, and receive T forms
instead. Servants also use the V form and receive the T form from
employers and so on.
As democracy and egalitarianism increased in Western societies, the
non-reciprocal use of V and T became less fashionable. Nowadays,
English has lost the T form (‘ye’) and ‘you’ is used by everybody.
In French, however, the T and V forms remain, but the T form is
used among friends to signal intimacy and equality. It is mostly
reciprocal. In the few instances where it is not reciprocal, it is
confined to the family circle with parents using T and receiving V
from their children. However, in public one uses V for the working
classes as well as younger people.
Besides T and V, the use of honorifics and titles with names and
first names themselves are part of linguistic politeness. The
phenomenon of using first name (FN) versus title and last name
(TLN) has been studied with respect to American English. This
study relates the use of FN with informality (Brown and Ford
1961; Erwin-Tripp 1972). Another study analyses the address
systems in several languages, suggesting the use of T/V and such
other devices for expressing linguistic politeness as ways of
expressing power. However, while the basic reasons for choosing
one set of linguistic markers rather than another remain the same
(i.e. the desire to show intimacy or differences of power etc.), the
fashions for doing so keep changing (Braun 1988). But now the
question arises as to what is politeness.
Scholars have answered this in terms of ‘adequacy’ i.e. a form of
address which is in accordance with the rules of the community in
question is considered adequate or polite (Braun and Schubert

1988:49). Brown and Levinson define politeness in terms of ‘face
needs’. The want to be unimpeded is ‘negative face’, while that of
being approved is ‘positive face’ (Brown and Levinson 1987: 63).
My own definition of politeness is:
Politeness is that which is considered gratifying by the addressee. It is

a matter of subjective interpretation and is not necessarily bound to
verbal or non-verbal signals without reference to context.
Rahman 1999:191
This definition emphasizes the subjectivity of the interaction. This

means that if young people meet friends who greet each other with
shrieks and abusive words, it will not be rudeness because it will
mean that the friends are really intimate with each other. Also, if
you meet your professor and address him by his first name when
he belongs to a generation when young people used the honorific
Ji or Sahib after the name and a title (Professor or Doctor etc.)
before it, he will be right to interpret it as rudeness and insensitivity.
On the other hand, a professor in a Western country who wants
his students to address him by his first name will interpret this as
an indicator of being at ease or relaxed with him.
In my article on this subject, I have presented the hypothesis that
a small English-using elite in South Asian cities is using only the
first name for older people without being sensitive to their norms
of politeness. They are doing it in the name of equality, informality,
and ease of interaction. However, they are doing it from a position
of power because they are Westernized and highly educated, and
serve in institutions (NGOs, Think Tanks, corporate sector etc.)
which can afford to be indifferent to indigenous norms of behaviour
(Rahman 1999: 183–223).
Yet another aspect of the relation between linguistic politeness and
power is that terms of prestige change according to the power of
the community which uses them. For instance, when the British
started ruling South Asia, the older titles and forms of address
which carried respect earlier, became less prestigious, and the

British forms came to be endowed with prestige. A list of these
terms is given below:
Indian And British Terms In The North Western
Provinces in the 1850s
Hindus Muslims British
Lalaji Mianji Master Sahib
(Baniya or Kaesth)
Nai Pandhe Mullahji Teacher
(barber teacher)
Panditji Munshiji Sir
Guruji Sheikh Sahib Miss (for females)
(Sunni)
Maharajji Mir Sahib (Shia) Madam (for females)
Bajpeiji Mirza Sahib –
(Mughal)
Parohatji Khan Sahib –
(Pathan)
Daoji – –
Vaed Ji Hakeem Ji Doctor Sahib
– Qazi Ji Judge/Justice Sahib
– Darogha Ji Thanedar Sahib
Police Officer
SHO
Diwan Wazir Minister
Source: Reid 1852: 81-82 and other sources.
The study of the relationship of society with language is fascinating.

The field is so vast that it will require years of research to be able
to understand it. This chapter was meant to introduce you to the
basic concepts and to stimulate your appetite for further research
of your own.
14
Language Politics
P olitics is defined as the pursuit of power. Power is needed by

human beings in order to do things. First, to secure food,
security, shelter, and produce children. Secondly, to go beyond
those necessities of survival itself to obtain pleasures such as
beautiful things, prestige symbols (such as posh cars), and
pleasurable experiences. If you have power, you can make your life
pleasant. Moreover, you can make your village, town, and country
more comfortable. For instance, if you get political power by
getting elected as a municipality head, you can construct roads and
schools for your area. If you become the prime minister, you can
improve the lives of millions of people. However, people tend to
blame the pursuit of power, because people do it at the expense of
others and do not help anyone but themselves or their relatives and
friends. Let us remember, however, that one cannot do any good
to anyone if one does not have power. Moreover, weak people are
bullied and made miserable. Power enables us to preserve our
pleasures. It is for pleasure that one seeks power. But, of course, it
should be sought within rules, otherwise we end up fighting wars
which reduce everybody’s pleasure. Indeed, modern weapons are
so destructive and poisonous that if we fight for power, we could
kill the human race itself.
Language and power
But what has language got to do with power? Well, it connects

with power in so many ways that one is surprised. First, we use a
language or languages in our government, judiciary, administration,
education, commerce, media, films, and the corporate sector. If we
Language Politics 163
know these languages, we can get jobs in these domains, but if we

do not, we will be left out. Now these domains of work are called
domains of power because modern states and the private sector
exercise power through them. The languages of power are the
languages used in them. As you can appreciate, the major language
of power in South Asia is English. If you know English very well,
you are much likely to enter the domains of power. If you know
the other languages used in the lower domains of power, you will
get jobs but not very good ones. If you do not know any of the
languages used in any of the domains of power, you will probably
not even be reading this book. If, therefore, you or your parents
are keen that you learn English, then it is because they want to
empower you. And, if whole groups want to empower themselves,
they learn the language of power. As Muslim rulers, including the
Moghuls, used Persian, this was the language which the people of
India learned if they wanted to work in the domains of power. The
Kaesth class among the Hindus became very good at Persian and
have left behind works of literature in it. When the British started
ruling India, the Indians learned English. Raja Ram Mohan Roy,
for instance, requested the British to teach English to the Indians.
Similarly, Nawab Abdul Latif and Sir Syed Ahmad Khan wanted
Muslims to learn English because, in their view, this was the best
way of empowering themselves.
As we have mentioned earlier, power is also expressed through
pronouns such as the different forms of ‘you’ (aap, tum, and tu),
and it is part of gender as expressed through language. Let us now
take language planning (LP) as one way of using language or
manipulating language for political aims, i.e. power-oriented aims.
The term ‘language planning’ was popularized by the famous
linguist Einar Haugen. He defined it as follows:
The activity of preparing a normative orthography, grammar, and

dictionary for the guidance of writers in a non-homogeneous speech
community (1959: 8).
In short, language planners make alphabets, write dictionaries, and

tell people what rules to follow in writing or speaking a language.
Some people divide the activities of language planners under
different headings. These are: (1) status planning, (2) acquisition
planning, and (3) corpus planning.
Status planning
The term means determining the status of a language. Status means

whether it will be used as a national language, a provincial language,
the language of primary education, the language of higher
education, the language of administration, and so on. Such
decisions are generally taken by governments and enforced by
governmental officials. These officials, whether civilian or military,
make up the apparatus of the state.
Hence, we can say that status planning is the same as language
policy. This policy is enforced by the state. If someone opposes it,
the state may use force to counter this opposition. Sometimes,
however, the state compromises and the policy is changed.
Acquisition planning
‘Acquisition’ means acquiring or learning a language. If a language

is taught or used in the TV and radio etc., then people acquire it.
If someone makes a policy to increase the number of people who
acquire a language, then it is known as acquisition planning.
Nowadays, this phenomenon is also known as language shift (LS).
Sometimes, people learn the language which is required for getting
good jobs, social prestige, and political power so well that they
forget their own language. In some cases, a groups’ own language
is completely forgotten. This phenomenon is known as language
death. In Ireland, Irish Gaelic, which was the language of Ireland
before the British conquest of the island, died. People shifted to
English. In Wales, too, a similar thing happened. However, in
Wales, some people tried to revive the use of Welsh once again.
This phenomenon is called reversing language shift (RLS). We

will study this phenomenon in detail in the next chapter.
Some people consider acquisition planning a part of status
planning. Their argument is that the decisions about using a
language for education and in the media is, indeed, status planning.
Since this increases the number of people who use these languages,
status planning is the same thing as acquisition planning.
Corpus planning
Corpus planning means making planned changes in the language

itself. For instance, we may want to change the script of a language,
introduce new words, fix the spellings of words, or eliminate some
words. The two major activities of corpus planners are (1)
standardization and (2) modernization.
Standardization means fixing a standard. In pre-modern
communities there are no fixed norms or standards. Some people
speak one variety of a language while others speak another one. In
fact, the language changes after a few miles. Even if the language
is written, the spellings are not fixed. Other rules of writing—where
to put commas, where to put full stops etc.—are also changeable.
When a society gets modernized, it needs a fixed standard. The
printing press prints millions of written pages which are read by
many people and it would be expensive to have different spellings
in different places. Moreover, it would create confusion in an
industrialized society where the printed word is used all over the
country. Thus, one variety of the language is standardized while
others are not. Spellings are fixed and the rules of punctuation etc.
are decided. These standard linguistic norms are written in
dictionaries, books of grammar, and manuals. They are taught in
schools and other training institutions. In short, the whole
education system of the modern state tries to eliminate diversity
while creating uniformity.
The other major activity of corpus planners is modernization. As
we can see, standardization too is part of modernization. Thus,
modernization of vocabulary can also be called neologism or lexical

modernization. It refers to creating new terms for concepts or
things which did not exist in the language. As you read earlier,
every language is meant to express the reality of the culture in
which it was created. The Siraiki and Arabic languages have many
words for ‘camel’ because camels were important for the people
who spoke these languages. The societies which make aircrafts, cars,
rockets, and computers also made words for them. They change
the meanings of existing words or borrow words from other
languages to express new ideas. This is lexical modernization or
neologism.
There are other activities of corpus planning too—such as purging
a language of some words, changing the script, making a new
script, and so on. If you want to learn about it in detail, read
Cooper (1989).
Language planning and politics
Apparently, all the activities described above are about language.

However, a little reflection will tell you that they are also connected
with politics. The Westernized elite in both India and Pakistan use
English. They know English so well that it is to their advantage to
keep using it in the domains of power.
In France, the French Academy made Parisian French the standard
language for the whole of France. The political advantage of this
move was that all the other varieties of French as well as the other
languages spoken in France became sub-standard rustic dialects.
Nobody could attain a high social position if they knew only these
languages and not the standard French. Secondly, the French ruling
elite could claim that the whole of France was united by the French
language. So, on the one hand the ruling elite consolidated its rule
all over France by claiming that everyone spoke only French, while
on the other hand, since the village dialects were considered
inferior, the elite kept the masses away from power. Of course some
clever students could learn standard French but this was not easy,
especially until the twentieth century, because earlier school

education was not so easily available.
Ethnic politics and language
The policies of the ruling elite are sometimes challenged by those

who are not in power. The challengers are sometimes called the
proto-elite. They are generally led by intellectuals and educated
people. These people want a symbol for mobilizing support.
Sometimes this symbol is religion, while in most instances it is
language. Language gives a sense of unity; of being one. It is one
of the things which give identity. Those who oppose the state use
language to form a pressure group. Some people feel that if their
own language is used in the domains of power, they will be at an
advantage. This is how ethnicity or ethno-nationalism is created.
Some people feel that ethnicity is a throwback to the past. They
argue that pre-modern people used to define themselves in ethnic
and religious terms, while modern people are nationalists. This
theory is called the primordialist theory of ethnicity, and it claims
that language is a primordial bond which is asserted because it has
never been lost. Other people argue that it is modernity which
makes people assert their ethnicity. There is a lot of competition
for jobs and power in the modern state. If a group feels that it is
not getting its proper share of goods, services, and power, it feels
wronged. Language may be a symbol used by the leaders of the
group to mobilize people and create a sense of common identity.
These mobilized people become a pressure group. In short
language-based ethnicity is a product of modern conditions. This
is the instrumentalist theory of ethnicity.
It is true that modern conditions create the desire to assert
ethnicity, but people in language movements may have both
primordialist (or sentimental) and instrumentalist (or rational)
motivations. People may feel that their culture—of which language
is a symbol—is being taken away from them. This is a sentimental
reason. They may also feel that they will be left behind another
group in the competition for power and money. These are

instrumentalist reasons.
If you look closely at the language-based ethnic movements in
Pakistan and India, you may be able to explain them through these
theories (for India see Phadnis 1989; for Pakistan see Rahman
1996).
Corpus planning and politics
It has been mentioned earlier that people may be emotionally

attached to their script. They may regard it as part of their identity
and may be unwilling to change it. It should also be mentioned
that when new words are formed, the morphemes which make
them are also borrowed according to one’s political views. Thus,
when modern Hindi was developed, Hindu language planners put
in words of Sanskrit origin in it in order to make it a symbol of
Hindu identity. The official policy in Pakistan is to emphasize the
Islamic and Pakistani identity. Hence, the new terms made in
Pakistani Urdu are borrowed from Arabic and Persian. For
instance, ‘white paper’ could have been translated as sufaid kaghaz
in Urdu. However, it has been translated as qartas abiaz (qartas =
document, paper; abiaz = white—both from Arabic). On the
contrary, the ethno-nationalist language planners use morphemes
of their indigenous languages to express their native roots. Here are
examples of new words created by ethno-nationalists in Pakistan:
PUNJABI
Word commonly used Punjabi Coinage Meaning

lafz akhar word
salana varhe var yearly
bhejna ghalna to send (a letter)
sailab harh flood
taqreeb ikath gathering
PASHTO
Word commonly used Pashto coinage Meaning

sumreen dub mako submarine
vai jaaz alvatak aeroplane
profesar pohand professor
dastkhat las lik signature
rasmulkhat lik dud script
BALUCHI
Word commonly used Baluchi coinage Meaning

mauzu bangapp subject
lavz, labz gal word
lahja, boli galwar dialect
kitab wanagi book
film tamur film
SINDHI
Word commonly used Sindhi coinage Meaning

brake hawai brake air brake
horn narghat air born
automatic khudkar automatic
seedha hath sajir right hand drive
ulta hath kabir left hand drive
SIRAIKI
Word commonly used Siraiki coinage Meaning

ilm bhum knowledge
alim bhum val intellectual,
erudite person
lisaniyat boli bhum linguistics
radio sur vaja radio
atom bomb suth gola nuclear bomb
To conclude, Language Planning (LP) is related to politics. That

is why controversies about languages, scripts, and words are not
only about these things but about things much deeper than these.
The real dispute is about way of life, way of thinking, and who is
going to get what, and not only about language.
15
Language Death
T he average life of human beings on this planet is said to be

between 5 to 10 million years. While it is not known exactly
when human languages were invented, all known societies have
always used them. It is estimated that there were 10,000 to 15,000
languages in prehistoric times. Languages have always died when
their speakers have died or started using other languages. Indeed,
even very powerful languages, which spread over large areas of the
world, have faded out and now we owe it to linguistic historians
like Nicholas Ostler to tell us something about their rise and
decline (Ostler 2005 and 2007). But nowadays, the pace of the
disappearance of human languages has increased. We still have
about 6,000 plus languages but we are losing them day by day. You
must have noticed in the first chapter where the languages of South
Asia are listed, some extinct languages are also mentioned. Indeed,
the Summer Institute of Linguistics, whose figures I have quoted,
does not mention any extinct language for Pakistan. However,
some researchers have confirmed that at least one Pakistani
language (Badeshi) is dead, another (Domaaki) is about to die, and
three (Chilliso, Gowro, and Ushojo) are getting so weakened that
they will die unless something is done to strengthen them (Rahman
2006: 81–82). In short, we may actually have more languages on
the way to extinction than the reports on languages tell us.
Do languages die or are they killed?
What is ‘language death’? As previously mentioned, it occurs when

the last speaker of a language dies. This happens when the speakers
have shifted to another language. That is why the phenomenon is
also called language shift (Fishman 1991). Other people argue that
people do not shift from their mother-tongue to another language,
unless there is pressure upon them to do so. It may not be that they
are forbidden to speak their own language (though this has
happened in history), but it does happen that there is no education
and no jobs in their language. Or, as is usually the case, their
language is looked down upon and treated with contempt. When
this happens people learn the language which is used in jobs, the
education system, media, commerce, and respectable company. This
means that the language does not die, it is ‘killed’, i.e. conditions
are created in which people have no benefit from learning it and
start considering it inferior themselves. That is why Tove Skutnabb-
Kangas, a linguist who is a great supporter of language rights, calls
this phenomenon ‘linguistic genocide’ (Skutnabb-Kangas 2000:
312). As the very group whose language is being killed wants to
learn the killer language because it is so useful to empower them as
individuals and as a group, it is also possible to see this as ‘linguistic
suicide’. However, the usual term for this phenomenon is language
death and that is what we will use in this chapter.
Modernity and language death
Why languages are dying so fast is because of the impact of

modernity and globalization. Modernization makes societies mobile
and educated. They leave their villages and crowd into the cities.
Here, they generally acquire jobs where they are exposed to one
language which is either the national language or some other
dominant language of power. The schools teach it; the media uses
it; and they themselves have to learn it in order to continue living
the new life they are trying to adjust into. Globalization has
increased all these trends. The means of communication have
improved and people are much more mobile than they were ever
before. People also travel abroad and are exposed to the Internet
and TV far more than ever before. Thus, more and more people
are shifting to the dominant languages and leaving their own
language behind in the village—and eventually, even the village
Language Death 173
gets swamped by so many modern things that it no longer is the

village they remember from their childhood.
In South Asia, English, Hindi, and Urdu are the major languages
which put pressure on the smaller languages. In other parts of the
world, too, there are many dominant languages which threaten
minor and weak languages. This is now a major concern of linguists
who have started writing on the phenomenon of language death.
In the year 2000 alone, there were several major books on the
subject (Two fairly easy ones to understand are Nettle & Romaine
2000; Crystal 2000). The UNESCO published a position paper
supporting education in minority languages in 2003 and a
conference on multilingual education was held in Bangkok on this
subject in November 2003. The second major conference on the
same subject was held in July 2008 in Bangkok, where a large
number of participants described how minority languages were
being used to teach little children. This is one way of saving a
language from dying out.
Can we do something to save languages from dying?
Should we save languages from becoming extinct at all? After all,

there is a view that having many languages is costly, divisive, and
inconvenient. One has to learn them and translation costs a lot of
money. Besides, one needs reading material, teachers, and jobs in
these languages; otherwise they could not be learned. As such,
according to this point of view, it is better to become monolingual.
Those who advocate this in a multilingual country often support
the official national language and deny ethnic identity as well as
the languages on which it is based. Those who support a world
language favour English monolingualism or English-only policies.
They are often found expressing antagonism to the tradition and
literature of all other than the English-speaking peoples or their
colonial clones in the former colonies. One reason for not favouring
the domination of any one language is precisely this: it leads to
contempt for others and helps in the domination of weak and
vulnerable groups.
David Crystal has given the following five major reasons for
preserving the linguistic diversity of the world:
1. We need diversity.
2. Languages express identity.
3. Languages are depositories of knowledge.
4. Languages contribute to the sum of human knowledge.
5. Languages are interesting in themselves. (Crystal 2000:27–67)
Arguments in favour of linguistic diversity take other forms also.
These have been presented with reference to ethics, philosophy,
and psychology by scholars. You can read Skutnabb-Kangas (2000:
249–284) for a summary of their views.
You will understand, of course, that states do not value diversity
or the assertion of identity because they are afraid that, carried to
an excess, both might break them up. The fact is that this excess
occurs only when the rulers suppress the minorities and the
suppression of their languages is part of it. For instance, the
Kurdish language is suppressed and punished in Turkey, but this
has not made the Kurds become reconciled to Turkish rule. Indeed,
this has only hardened their resistance to Turkish domination.
Some linguists are now trying to write the grammars and
dictionaries of languages which are threatened with extinction.
There is also a book by a famous sociolinguist called Can Threatened
Languages be Saved? (Fishman 2001). (This is a sequel to an earlier
book called Reversing Language Shift, Fishman 1991). Fishman
discusses a number of strategies for saving languages. Ten years
later this is what the author had to say about this issue:
So can threatened languages be saved? This question now has an

informed, though uncertain, answer: Yes, more of them can be saved
than has been the case in the past, but only by following careful
strategies that focus on priorities and on strong linkages to them, and
only if the true complexity of local human identity, linguistic
competence and global interdependence are fully recognized.
Fishman 2001:481
Language Death 175
One of the ways of saving languages is to educate children in them.

That way, they understand concepts better and, what is more
important, come to respect their language and their identity. But,
of course, the teachers and the other grown ups they come across—
the role models—must show respect for their languages and
culture. If they do not, the pupils will regard their language as
nothing more than an additional and unwelcome burden. That is
where language policy is connected with educational policy, and
both are connected with social psychology and the art of teaching
itself.
Some people suggest that those who specialize in language death
and reversing language shift should study educational linguistics as
well as writing grammars in disappearing languages. In fact, this
whole branch of learning is sometimes called preventive linguistics,
i.e. that branch of linguistics which will help us to prevent the
death of languages. Such a person is a practitioner and a language
activist as well as a scholarly linguist. If this is what you want to
be, South Asia is an excellent place for you.
16
Educational Linguistics
E ducational linguistics refers to the use of language in education.

This is a very vast subject indeed. While discussing language
planning (LP), we referred to acquisition planning (Cooper 1989:
157–163). Acquisition planning is not simply language teaching,
but it is the most important part of it. In addition to teaching, one
can plan to increase the number of people who know one’s
language through other means: radio, television, films, songs,
sports, games, and so on. But, of course, teaching—especially in
schools—is the most important way of spreading a language. But
teaching a language is by no means a simple matter. It is so
complicated that there are branches of applied linguistics dealing
with foreign language teaching (FLT), second language teaching
(SLT), the teaching of classical or dead languages, and so on. In
the South Asian context, most people going to college must have
heard of subjects like ELT (English Language Teaching), ESP
(English for Special Purposes), and TEFL (Teaching of English as
a Foreign Language) etc. In Pakistan, the Society of Pakistan
English Language Teachers (SPELT) was created with the
pioneering leadership of Zakia Sarwar in 1984 (Rahman 2002:
316).
Debates in educational linguistics
There are several debates in the field of educational linguistics.

First, there is no clear consensus on what a ‘second’ and a ‘foreign’
language is. German is obviously a foreign language in Pakistan, as
it is not used in any domain. If you want to study it, you have to
go to those specialized institutions which offer courses in it. But
Educational Linguistics 177
English is used in the media, educational institutions, the officer

corps of the armed forces, superior judiciary, high-level bureaucracy,
and so on. Many people actually use it spontaneously in private
conversation because they are very fluent in it. Is it then a ‘second’
rather than a ‘foreign’ language? The consensus of most ELT
theoreticians is that English is, indeed, a second language in most
of South Asia and it should be taught through the techniques called
TESL (Teaching of English as a Second Language). This is what
has been happening in Pakistan and India since the last quarter of
a century. However, the fact remains that our societies are deeply
divided and the class exposed to English is only found in the urban
areas and even then only at the upper socio-economic levels. After
all, it is only the rich and the powerful who are exposed to English
at home and in the school, and not the ordinary people for whom
maybe Urdu or Hindi is a second language and English is a foreign
language. However valid such kind of questions may be, the
consensus of educational linguists is that if a country uses a foreign
language in many domains, then it is a second, and not a foreign,
language. So, in the final analysis, we settle these things by
convention and consensus, even if there are nagging doubts. This
does not mean that anything is finally settled, because the scientific
spirit requires us to keep investigating, keep asking questions, and
keep up our inquisitiveness.
Do we learn other languages better

if we start with our own?
Another question is whether we learn other languages better if we

start learning them as early as possible or graduate to them after
having learned our first languages (mother tongues)? English-
medium schools in most cities of South Asia tell us proudly that
they do not allow the children to speak the ‘vernaculars’. Even
small children are forced to speak in English and this is considered
a great distinction. The logic is that they must be immersed in
English—the target language—from early childhood, otherwise
they will never master it. As for the mother-tongue, why they will
acquire it anyway, since it is all around them. This apparently

commonsensical approach is refuted by many linguists, including
the researchers at the United Nations Educational Scientific and
Cultural Organization (UNESCO), which says in its latest
report:
1. UNESCO supports mother tongue instruction as a means of
improving educational quality by building upon the knowledge
and experience of the learners and teachers.
2. UNESCO supports bilingual and/or multilingual education at
all levels of education as a means of promoting both social and
gender equality and as a key element of linguistically diverse
societies.
3. UNESCO supports language as an essential component of
inter-cultural education in order to encourage understanding
between different population groups and ensure respect for
fundamental rights (UNESCO 2003).
There is plenty of evidence at least about one thing: children do
not learn their languages very well if they begin with a foreign
language. South Asian children exposed to English from childhood
not only do not know their own languages except at the most
elementary spoken level, but also express contempt for them. This
is regrettable since the languages of this subcontinent have some of
the classics of world literature in them. But you find many English-
medium students actively ignoring, shunning, and denigrating the
significance of the rich literatures of our part of the world.
The other part of the claim, i.e. that children learn the target
language better if they are taught no other language but that from
early childhood, is not true either. Well, of course there are people
in India and Pakistan who speak English very well. Their writing
skills in that language is as good as the best writers of English
anywhere in the world and they produce literature in that language
(see Rahman 1991). Yet this does not prove that they would not
have excelled in English if they had first learned their first languages
at school just as well as they learned English. Indeed, in a study of
language learning, Tove Skutnabb-Kangas argued that it is a part

of linguistic imperialism to teach another language at the expense
of one’s first language. This is subtractive bilingualism (i.e. it takes
away or subtracts a language), whereas it is only additive
bilingualism (i.e. that which adds a language without taking away
one’s first language) which is welcome. She equates this kind of
imperialism with racism and argues that it infringes upon the rights
of children to know and respect their languages. She states:
Promoting diversity is an accepted principle in declarations. Still

Western language policies have to a large extent been based on false
either-or thinking (you need to choose between languages, you
cannot have both this language and that language and maybe others
too). It has promoted subtractive rather than additive language
learning and subtractive spread of English and other dominant
languages: the learning of a dominant language has been presented as
necessarily happening at the cost of a dominated language, instead of
in addition to it. Diversity is killed.
The central question is whether it is easier to learn only English in

order to be good in it, or is it better to learn one’s first language
along with English or even before one learns English. The authors
argue for multilingual education (MLE) for social justice. They say
that when only a few dominant languages are taught and all others
excluded from the school system, there is injustice. People are
marginalized and feel inferior and the rate of literacy increases very
slowly. Even more to the point the psychological cost of this
exercise in discrimination is appalling. In the words of the
authors:
Why then are mother tongues neglected despite persuasive evidence to

the contrary? As the voiceless minorities suffer the sinister exclusion of
mother tongues, the silent elites enjoy the pre-eminence of dominant
languages such as English. In the post-colonial world, ‘the killer
languages’, including English, thrive at the cost of other languages, and
in many countries the myth of English medium superiority is
propagated to the detriment of the poor and the marginalized. English
and other ‘killer languages’ set in motion a hierarchical pecking order
of languages which severely disadvantages the other languages, those

of the indigenous peoples and minorities, in particular. And yet, it does
not have to be so. In a true multilingual system, all languages can have
their legitimate place: mother tongues, languages of regional, national,
and wider communication.
Mohanty 2009:7
Look at the first sentence. Yes, there is evidence that if the mother
tongue is taught first, other languages, such as English, are learned
efficiently. The idea that we must start with English is not
supported by research. However, research in the social sciences,
including language-learning, is very complex. Human beings are
complex themselves and circumstances keep changing in unknown
ways, so you will find people making contrary claims in the
research literature. My own view, however, is that power is a very
important component. If a community is powerful and its language
is prestigious, it is learned more easily by everyone concerned. But
if the community is less powerful and the language is despised or
marginalized, then it is more difficult to learn in school situations.
Human beings are impressed with power and glamour and are
simply more self-motivated to learn a prestigious language. This
does not mean, however, that a less prestigious language should not
be taught. Indeed, if anything, it should be given more time, and
more reading material as well as radio and TV programmes should
be produced to sustain it. Ours is a multilingual world and if it is
to be a just one, then we must care for language rights, diversity,
and pluralism. For this, we must teach all languages in an additive
manner, beginning with the first language as far as possible.
The solution?
This is obviously a political and economic issue. After all, someone

has to decide which first languages are to be taught and who will
pay for the reading material and the teachers etc. These are
ultimately political questions, as my books on the politics of
language learning in Pakistan (Rahman 1999: 88–102; Rahman
2002) suggest.
In a nutshell, if English is taught well to a few elitist children, they

will be qualified for the best jobs in most South Asian countries.
But when this happens, most other children will find English a
sieve or obstacle which forces them to remain at the bottom, unless
they are very bright or work much harder than their English-
medium counterparts. Thus, the class boundaries are maintained
by English, while it appears that we are talking of merit and not
class advantage. But what goes unstated is that English is a kind of
cultural capital (a concept given by the French sociologist Bourdieu
1991: 55), which the upper classes possess and which they use to
dominate the society culturally and intellectually. Thus, educational
linguistics connects at some level with political linguistics and
language policy. This is one of the great beauties of macrolinguistics—
it provides new insights into politics, economy, society, culture,
and education.
17
Writing in South Asia
D o you know that South Asia is home to very ancient writing

systems? The Taxila Museum houses a number of specimens
in very ancient scripts. One of these is the Aramaic script which
grew out of the Phoenician script and was very influential for over
a thousand years. This will be discussed later in more detail.
The inscription is described as follows (unfortunately, the words
‘Aramaic’ and ‘emperor’ have been misspelt here):
Writing in South Asia 183
Let us, however, introduce you to the basics of writing before we

discuss the scripts now used in the world. As you will agree, writing
is very important in the modern world. It creates a permanent
record (i.e. memory) in a certain language by conveying ideas
through visible marks. These marks combine together to create
meaning according to various principles. Let us look at these
principles below.
Marks carrying meaning
If the visible marks convey meaning directly, we call the system in

which they are used a semasiographic system. This system is used
even now in pictures. For instance, a road sign has the picture of
a running child. This conveys the meaning that children may be
running across the road and the drivers should slow down. Of
course the picture leaves a lot to the driver’s knowledge about the
world. Someone from Mars may not know exactly what the picture
means and whether any action is required or not. Some instances
of semasiography are given below:
The basics of
Blissymbolics
Blissymbolics is a graphic, meaning-based communication system. Some
of the symbols are pictographs. They look like the things they represent.
house men women person face
You will notice that semasiographic systems are rather limited. You
can hardly convey difficult messages through them. Perhaps that is
why glottographic systems of writing developed (Glotta is tongue
and graphein is writing in Greek). These systems provide visible
representations of speech, i.e. sounds which make sense.
Glottographic systems are used for all kinds of languages, all the
way from Urdu and Hindi to Chinese. However, the writing

system of different languages are based on different principles.
Some glottographic systems are called logographic. As logos means
sense, logic, idea or concept, it appears as if sense or idea is directly
represented in writing. This is the general view about the Chinese
language. However, it is not true. Let us go into some details about
this language. Words consist of units of sounds (phonemes) as well
as units of meaning (morphemes). If we represent morphemes, then
we will have a phonographic and segmental system of writing. A
segment, as you will recall, is a vowel or consonant, i.e. any
phoneme.
The writing system of the Chinese language is predominantly
logographic. In this language, morphemes are represented by
separate symbols or marks. These marks are also called graphemes.
In Chinese, as it happens, a word is generally a morpheme. Thus,
we can say that almost every word will be represented by separate
graphemes. This means that we would need to learn several
thousand different shapes of graphemes to be able to communicate
in written Chinese. This is obviously a disadvantage for anyone
who is learning Chinese.
The advantage is that the written script unites the speakers of the
various dialects of Chinese. In fact, these dialects are so different
that they are not always mutually intelligible. This means that they
are like different languages (if you remember, dialects are defined
as being mutually intelligible). But, because the symbols in writing
do not represent pronunciation, the same symbols are used. Thus,
a word is written in the same way, although it is pronounced
differently in different dialects of Chinese.
Marks carrying sounds
The phonographic principle is the one to which we are accustomed

in our South Asian languages and English. The idea is that every
sound should be represented by a grapheme—a visible mark. These
graphemes make up the alphabet (the word comes from alpha and
beta, which in the Greek script stood for ‘a’ and ‘b’). Ideally, every
sound (i.e. segment) should be represented by one and only one
grapheme. In fact, this generally does not happen. English is quite
notorious for having only five graphemes (the letters a, e, i, o, and
u) to represent over sixteen vowels and many diphthongs. It also
has graphemes which give different sounds at different places. Look
at <c> and <k>. They stand for the phoneme /k/ in the words king
and cut. But in cent the letter <c> stands for the sound /s/ so that
cent is pronounced as /s e n t/. Similarly, <g> has two sounds and
<x> represents a combinations of three sounds (/e k s/) at places as
in the word rex which is pronounced as /r e k s/.
Urdu and Hindi use the Arabic and the Devanagari scripts
respectively. Both are phonographic, but there are letters which are
not pronounced. Urdu also has some letters borrowed from Arabic
which are pronounced in the same way. For instance, the phoneme
/z/ has the following graphemes in Urdu <‫>ﺽ‬, <‫>ﻅ‬, <‫ >ﺫ‬and <‫>ﺯ‬.
Hindi too has instances of several graphemes for the same
phoneme.
On the other hand, the Urdu script, as well as all scripts based on
the Arabic system of writing, lack graphemes for vowels. The
Devanagari script too lacks some vowels, though it does have more
symbols for them than English. The reasons for this will be
discussed later. Another problem in both the Devanagari and the
Arabic scripts is that the graphemes take on different shapes when
they are joined together. In other words, the grapheme <‫ >ﻡ‬in
Urdu which represents the phoneme /m/ takes the shape <‫ >ﻣ‬in
the initial position, while it becomes <‫ >ﻡ‬in the end of words.
The script of Urdu

The script of Urdu is based on the script of Arabic, but it is written
somewhat differently. The Arabic script in which most Arabic
books are written in South Asia is called naskh while the script in
which most Urdu books are written is called the nastaleeq. Nastaleeq
is the script of Persian which is a modified form of the Arabic
script. It can, therefore, be called the Perso-Arabic script. In

Pakistan, Punjabi, Siraiki, Hindko, Balochi, and Brahvi use the
nastaleeq script. The naskh is used for writing Sindhi and Pashto.
However, all these languages have modified the graphemes of the
Arabic script to represent the phonemes of their languages. This
process is also going on in the case of Burushaski, Shina, Wakhi,
Balti, Khowar, and other minor languages in Pakistan.
Besides representing the segmental features of languages, the non-
segmental features of languages such as stress, tone, and rhythm
can also be represented by diacritical marks.
Other scripts and the principles they are based upon
Apart from representing segments, a script can also have graphemes

for syllables. For instance, the word only has two syllables on and
ly. We can have only two graphemes (instead of the four we have
in English) to represent this word. Such scripts are called syllabic
scripts.
A third kind of phonographic script is called featural. Featural
scripts are somewhat unique. Features are the + – signs we use for
the qualities of sounds. A sound could be a consonant and we call
it + consonantal, – vocalic. It could be voiced and we call it + voice,
and so on and so forth. Now, if there is a writing system which has
graphemes showing these or other features of sounds, it would be
a featural script.
There is, in fact, such a script. It was made in Korea in the fifteenth
century and is called Hangul (Coulmas 1989; Sampson 1985).
There were 28 graphemes which relate to the phonetic features of
sounds. Nowadays, there are only 24 (four became obsolete), and
they represent the sounds of the language classified as continuants,
stops, vowels, and so on.
Now arrange all these systems of writing in a simple diagram:

WRITING
Semasiographic Glottographic
(not writing proper)
Phonographic
Logographic
Based on Morphemic Syllabic Segmental Featural

word (non- (based on
existent) morpheme)
Adapted from Sampson 1985:3
Semasiography, the representation of sense directly through

pictures or signs, is not writing proper. Logographic systems based
on a word (where a word means a combination of several
morphemes) are not known so they are only a hypothetical
possibility. Chinese, as you will recall, is based on morphemes
which may also represent words, i.e. words of one morpheme (See
Coulmas 1989 for Chinese).
The history of writing
All writing does not come from one common ancestor. However,
it is possible that segmental phonographic scripts—such as those
of Urdu, Punjabi, English, Persian, and Sanskrit etc.—do have one
source. Let us, however, look at the earliest forms of writing before
coming to this interesting theory.
Evidence suggests that the Sumerian script is probably the oldest
script known to us. Sumer flourished in the Middle East (roughly
in Iraq) about 4500 to about 1750 years before the birth of Christ.
Some scholars go so far as to narrow the search for the originator
of writing to the city of Uruk in modern Iraq in about 3300 bc
(Robinson 2007: 11). Nobody can be sure of this, but one thing
is certain and that is that clay tablets, or tokens, with marks made
by a wedge-shaped stick have been found from the ruins of these
ancient dwellings of our ancestors. Are these ‘tokens’ for counting
lists of goods? Probably yes. Certain marks were made on clay
probably to record a list of items or something connected with

commerce and administration. The stick used for making these
marks made wedge-like shapes. Later, European scholars, who used
Latin words, called these shapes ‘cuneiform’ (cunei = wedge in
Latin). Some of these shapes represented pictures initially, i.e. they
were pictograms. Later, sounds also came to be represented. Here
is the name of Hammarabi, the man who gave the first written
code of law, to the human race:
ha + am + mu + ra + bi
fish + wild bull + ear + ? + innkeeper
The point is that only the sound of fish (ha) and wild bull (am),
ear (mu), and innkeeper (bi) were considered, not the meaning.
Thus, certain marks or shapes (graphemes) came to stand for
certain sounds and a phonographic—but only partly phonographic—
system or writing was born.
The Armaic script, based upon the Phonecian script, was used in
South Asia for over 1000 years. It was the official script of a
number of ancient empires whose names will come later in this
chapter. Some people say it was brought to ancient India by traders
who came from Egypt, Turkey, and the Middle East (Robinson
2007: 172). But Aramaic was not the name of only a script; it was
also the name of a language. Indeed, it was an important Semitic
language (the family to which Arabic and Hebrew belong) and was
probably the mother tongue of Jesus Christ. It was one of the
earliest influences of the Middle East on India. But India did have
its own scripts even then. These will be described later.
Hieroglyphics and the scripts of the Middle East

The Egyptian system of writing, called hieroglyphics by modern
scholars, was used around 3000 bc in Egypt. It is also made up of
pictures, i.e. ideograms as well as phonograms. It was a very complex
system which is being deciphered even now. Some of the shapes of
the hieroglyphs were borrowed by the inventors of the Semitic
alphabet who used them to represent speech sounds. This was the
making of the alphabet as we know it to which we turn now.
Early Greek
Phoenician
Hebrew Arabic
Symbols
Modern
Name
Name
Form
Form
, ‫א‬ aleph ‫ا‬ ‘alif
b ‫ב‬ bēth ‫ب‬ bā’
g ‫ג‬ gimel ‫ج‬ ǧīm
d ‫ד‬ dāleth ‫د‬ dāl
h ‫ה‬ hē ‫ه‬ ḥā’
w ‫ו‬ waw ‫و‬ wāw
z ‫ז‬ zayin ‫ز‬ zāy
h ‫ח‬ heth ‫ح‬ ḥā’
t ‫ט‬ teth ‫ط‬ ṭā’
y ‫י‬ yod ‫ي‬ yā’
k ‫כ‬ kāph ‫ك‬ kāf
l ‫ל‬ lāmed ‫ل‬ lām
m ‫מ‬ mēm ‫م‬ mīm
n ‫נ‬ nūn ‫ن‬ nūn
s ‫ס‬ samekh ‫س‬ sīn
‘ ‫ע‬ ayin ‫ع‬ ‘ayn
p ‫פ‬ pe
s ‫צ‬ sade ‫ص‬ ṣād
o ‫ק‬ qoph ‫ق‬ qāf
r ‫ר‬ reš ‫ر‬ rā’
sh/s ‫ש‬ šin ‫ش‬ šīn
t ‫ת‬ taw ‫ت‬ ṭā
The Middle East is the cradle of the segmental phonographic script

proper. Phoenician, Hebrew, and Arabic represented consonants by
shapes which exist in modified forms in most of the languages of
Europe and Asia. These languages had vowel sounds but did not
have any graphemes to represent them. For instance, in Arabic the

consonants are written and the native speakers pronounce the vowels
without actually writing them. That is why South Asian Muslims
find it difficult to read the Quran printed in Arab countries. Thus,
the languages which based their scripts on the Middle Eastern scripts
also did not have shapes for vowels. Some of them did invent shapes
for vowels but they were less than the number of vowels.
The scripts of Europe

The Greeks borrowed their alphabet from the Middle East. The
Romans in turn borrowed theirs from the Greeks, and all European
scripts are derived from the Greek or the Roman. Arabic is, of
course, of Middle Eastern origin as is Hebrew which is used in
Israel. But Persian, Urdu, pre-modern Turkish, Punjabi, Siraiki,
Hindko, Balochi, Brahvi, Pashto, Sindhi, Kashmiri, and other
Muslim languages are also based on the Arabic script. Thus, we see
that in the matter of script, European and Muslim Asian languages
are related. They share a common ancestor. Remember, however,
that these languages do not belong to one language family. Arabic
and Hebrew are Semitic languages; Turkish is Altaic; while English,
Urdu, Sindhi, Hindi, Bengali, Pashto, Persian etc. belong to the
Indo-European family of languages. This script is important for us
in South Asia because the subcontinent is home to so many
Muslims. A number of languages used by them, notably Urdu, is
written in this script. Let as now turn to the most widely used scripts
in South Asia—those belonging to the Brahmi family of scripts.
The scripts of South Asia from before the arrival of

the Perso-Arabic scripts
Manuscripts in Brahmi and Sanskrit languages have been found in
Jaulian near Taxila. They are inscribed on burnt birch bark and are
from the fourth and fifth century. Sanskrit existed in the
Subcontinent from about 1000 bc. However, it was not always
used for worldly purposes. A contemporary scholar of the language,
Sheldon Pollock, tells us:
From around the beginning of the first millennium bce when the
earliest form of Sanskrit appeared in South Asia, until around the
beginning of the first millennium ce, Sanskrit functioned as a
communicative medium that was restricted both in terms of who was
permitted to make use of the language and which purposes the
language could subserve (Pollock 2006: 39).
The following are the graphemes of the ancient Brahmi script, the
mother of most of scripts used in India and other countries
today.
Source: http://www.ancientscripts.com/brahm.html
Retrieved on 7 Jan 2010.
The Brahmi and the Kharosthi scripts, both shown above, are the
oldest deciphered scripts of ancient India. An undeciphered
script—if it is a script at all—is the Indus Valley script which is
found in Mohenjo Daro and Harappa. The Brahmi script is written
from left to right. It is the parent of most of the Indian scripts
including the Devanagari script. It has also given rise to the
Burmese, Thai, Lao, Khmer, and several other scripts (Bright 1966:
373–74). Kharosthi was restricted to the areas now in Pakistan and
parts of Afghanistan. It was written from right to left and Ashoka’s
edicts in this part of the world are in this script.
This leaves the development of the Devanagari script out. Some
people think it too came across the sea from the Middle East. Most
of the scholars, however, believe it is derived from the Brahmi
script which has been mentioned above. The north Indian branch
of the Brahmi script is called the Gupta branch, while the south
Indian one is the Grantha branch. This script is used for writing
Gujrati which is used both in Pakistan and India. The Devanagari
script which is used for writing Hindi, Marathi, and Nepali is a
huge script considering the population of the people using these
languages. The scripts used for writing Hindi, Punjabi, Bengali,
and Oriya in India are based on the Gupta system. The variant
used for writing Punjabi is called Gurmukhi in India. The script
is considered sacred by the Sikhs since it is said to have come out
of the mouth (mukh) of the saintly teacher (guru). In Pakistan, it
is written in the Perso-Arabic script which is also called Shahmukhi
in the context of Punjabi, i.e. that which was created by the state
(shah means ruler).The scripts of Tamil, Telugu, Malayalam, and
Kannada are based on the Grantha system.
The most popular script of India and Nepal

The Devanagari script with which students in India and Nepal are
familiar but which the Hindi cinema and TV shows are showing
to Pakistanis also, is based on articulatory phonetics. First come
the primary vowels both short and long. These are followed by the
secondary vowels, i.e. diphthongs or two vowels joined together.
Then there are velar, palatal, retroflex, dental, and labial sounds.
To make the chart more organized, the unaspirated (without the
little ‘h’ sound) is followed by the aspirated sound. The following
chart illustrates this (The abbreviations are: Dev = Devnagari; Guj
= Gujrati; Pun = Punjabi; Ben = Bengali; Ori = Oriya; Tel =
Telugo; Kan = Kannada; Tam = Tamil; Mal = Malyalam; Sin =
Singhalese).
Consonants
Figure 4 gives the origin and relationship of writing systems with

each other.
Figure 4 196
Proto-Semitic alphabet
Phoenician alphabet Aramaic alphabet
Classical Greek Etruscan Square Hebrew Nubatean
Modern Hebrew Arabic
Cyrillic Modern Greek Gaelic Gothic Roman
Naskh Nastaleeq
Russian
English
Eastern
European French Modern Sindhi Pashto
Linguistics for Beginners
languages Arabic
Spanish
Other Persian
Western Urdu Brahmi is put
European Punjabi under this tree
languages Balochi in Voigtlander
Siraiki and Lewis
Brahvi (1990: 72)
Scripts and culture
The invention of writing is perhaps the most significant

development in the history of humanity. Had there been no
writing, systematic knowledge would not have increased. That
means that there would have been no science and the modern
industrial revolution would not have taken place. It also means that
there would have been no nuclear, chemical or biological weapons
which now threaten to wipe out humanity. Whether the
development of writing is good or bad is for you to decide in the
light of these diverse facts.
One notable thing about scripts is that they are part of culture,
history, and identity. Thus, Muslims may not want to leave the
script in which Arabic is written even if there are advantages in
using some other script. Similarly, Hindus and Sikhs may not want
to leave their scripts. Matters of scripts, therefore, are not
straightforward questions of rationality or utility as some people
wrongly assume. They are questions of politics, economics, and
identity. Moreover, scripts have centuries of a groups’ literary and
cultural treasures—songs, stories, riddles, aphorisms, literature, and
other cultural goods. People do not want to kill them by losing the
ability to read the script in which they are written. That is why
they keep struggling to preserve their scripts, even if it does not
appear rational to do so. Questions of changing the script are often
very difficult to decide. The linguist should keep these extra-
linguistic matters in mind when questions of making new scripts
for unwritten languages or modifying existing scripts are
disputed.
Notes for the Reader
T he purpose of this introductory book was to introduce you to

basic concepts and the technical terms used in linguistics. This
should enable you to read books of linguistics on your own.
As for the theoretical aspects of linguistics, no attempt has been
made to introduce you to theories of linguistics other than those
of Chomsky. If you want to read about other theories, you can
refer to Robins’ chapter on the subject (1964: 278–288). You
should read Halliday’s systemic grammar (Halliday 1961),
Perlmutter’s relational grammar (Blake 1990), Pike’s tagmemics
(Cook 1969), Lamb’s stratificational grammar, and other such
theories.
If your interest is in the relationship of language with society, you
can read sociolinguistics and anthropological linguistics. If you like
computers, you may be interested in computational linguistics. If
you have a flair for biology, you may like to focus on neurolinguistics
or animal communication etc. Those who are interested in
philosophy may like to study philosophical linguistics, which also
has a mathematical side to it. Linguistics is not only about language
itself—though linguistic theory is—but about so many aspects of
reality that everyone can enjoy it. Even political scientists can
investigate how language is related to the struggle for power in the
world.
A brief bibliography has been provided in the end but this is not
enough. If you are interested in further research, you should look
into the bibliographies of the books suggested at the end. They will
lead you on into the infinite world of the linguistic sciences.
Bibliography and Further Reading
Abney, S. P. ‘The English noun phrase in its sentential aspect’, Ph.D dissertation.
Cambridge, MA: MIT, 1987.
Aitichison, Jean. 1976. The Articulate Mammal: An Introduction to Psycholinguistics.
London & New York: Routledge, 4e, 1998.
1972. Linguistics. Sevenoaks, Kent: Hodder and Stoughton. 3rd ed.,
1988. (First Edition entitled General Linguistics).
Akmajian, Adrian; Demers, Richard A., Farmer, Ann K., and Harnish, Robert
M., eds. Linguistics: An Introduction to Language and Communication. 2001.
MIT. (New Delhi: Prentice Hall of India Pvt. Ltd, 2006).
Battaner-Moro, Elena. ‘Wittgenstein’ in Strazny, 2005: 1179–1181.

Berlin, B., and Kay, P., Basic Colour Terms. Berkeley: University of California
Press, 1969.
Bernstein, Richard J., The Restructuring of Social and Political Theory. Oxford:
Basil Blackwell, 1976.
Blake, Barry, Relational Grammar. New York: Routledge, 1990.
Bloomfield, Leonard. Language. New York: Holt, Rinehart & Winston, 1933.
Bourdieu, Pierre. Language and Symbolic Power. Cambridge: Polity Press, 1991.
Translated from the French by Gino Raymond and Matthew Adamson.
Braun, Friedrike and Schubert, Kloss. ‘When Polite Forms are Impolite, or what
Politeness Really Is?’ in Braun 1988: 45–67.
Braun, Friedrike. Terms of Address: Problems of Patterns and Usage in Various
Languages and Cultures. Berlin and New York: Mouton de Gruyter,1988.
Daniels, Peter T, and Bright, William. The World’s Writing Systems. New York:
Oxford University Press, 1996.
Brown, Penelope, and Levinson, Stephen. Politeness: Some Universals in Language
Use. Cambridge: Cambridge University Press, 1987.
Brown, Roger, and Ford, Marguerite. 1961. ‘Address in American English’.
Journal of Abnormal and Social Psychology 62: 375–85.
Chomsky, Noam, Syntactic Structures. Mouton: The Hague, 1957.

Aspects of the Theory of Syntax. Cambridge, Mass: MIT Press, 1965.
1972.
Language and Mind. New York: Harcourt Brace Jovanovich, 1968.
Reflections on Language. New York: Pantheon, 1975.
Barriers. Cambridge, Mass: MIT Press, 1986.
200 Bibliography and Further Reading
On Language. Chomsky’s Classic Works Language and Responsibility and

Reflections on Language in One Volume. Translated from the French by
Mitsou Ronat: New Delhi Penguin Books, 2003 edition.
Coates, J., Women Talk: Conversation between Women Friends. Oxford: Blackwell,
1996.
Cook, V., Chomsky’s Universal Grammar: An Introduction. London: Blackwell,
1988.
Cook, W. A., Introduction to Tagmemic Analysis. Washington: Holt, Rinehart &
Winston, 1969.
Cooper, Robert L., Language Planning and Social Change. Cambridge: Cambridge
University Press,1989.
Corballis, Michael C., From Hand to Mouth: The Origins of Languages. Princeton
and Oxford: Princeton University Press, 2002.
Coulmas, Florian. The Writing Systems of the World. Oxford: Basil Blackwell,
1989.
Crystal, David. Cambridge Encyclopaedia of Language. Cambridge: Cambridge
University Press, 1988.
Crystal, David. Language Death. Cambridge: Cambridge University Press, 2000.
Culler, Jonathan. Saussure. Glasgow: Fontana/Collins, 1976.
Deutscher, Guy. The Unfolding of Language: An Evolutionary Tour of Mankind’s

Greatest Invention. New York: Henry Holt and Company, 2005.
Erwin-Tripp, Susan. ‘On Sociolinguistic Rules: Alternation and Co-occurrence’

in Gumperz, J. J. and Hymes, D., eds. Directions in Sociolinguistics. New York:
Holt, Rinehart & Winston, 1972.
Ferguson, Charles H., 1959. ‘Diglossia’. In Hymes 1964: 429–439. Originally

published in Word 15: 325–340.
Fishman, Joshua. ed. Can Threatened Languages be Saved? Clevedon: Multilingual
Matters, 2001.
Fishman, Joshua. Reversing Language Shift: Theoretical and Empirical Foundations
of Assistance to Threatened Languages. Clevedon: Multilingual Matters, 1991.
Fromkin, Victoria, and Rodman, Robert. Introduction to Language. New York:
Holt, Rinehart and Winston, Inc, 1974.
Gonzales, Ambrose E., The Black Border. Columbia, S.C: The State Co., 1922.
Goodenough, W. H., 1956. ‘Componential Analysis and the Study of Meaning’,
Language. 32: 195–216.
Gordon, Raymond G. Jr., ed. 2005. Ethnologue: Languages of the World. 15e.
Dallas,Tx: SIL International. Online version: http://www.ethnologue.com/.
Haegeman, Liliane. Introduction to Government and Binding Theory. Oxford:

Blackwell, 1991.
Bibliography and Further Reading 201
Halliday, M. A. K., 1961. ‘Categories of the Theory of Grammar’, Word 17:

241–192.
Haugen, Einar. 1959. ‘Planning for a Standard language in Modern Norway’,
Anthropological Linguistics 1; 3: 8–21.
Hill, Jane H., 1988. ‘Language, Culture and World View’ in Newmeyer (ed)
1988: 14–36.
Holmes, J., Women, Men and Politeness. London: Longman, 1995.
Hymes, Dell. ed. Language in Culture and Society: A Reader in Linguistics and
Anthropology. New York: Harper and Row, 1964.
Jakobson, Roman, and Halle, Morris. Fundamentals of Language. The Hague:

Mouton, 1956.
Katz, Jerrold J. and Fodor, Jerry. A. 1964. ‘The Structure of a Semantic Theory’.
In Katz and Fodor (eds). The Structure of Language: Readings in the Philosophy
of Language. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. 1964. pp.
479–518. Originally printed in Language 39 (Apr–Jun 1963), 170–210.
Ladefoged, P., A Course in Phonetics. 2e. New York/Harcourt Brace Jovanovich.

1982.
Lakoff, R., Language and Women’s Place. New York: Harper and Row. 1975.
Lasnik, Howard and Otero, Carlos. 2005. ‘Chomsky, Noam’. In Strazny, Vol. 1:
pp. 205-208.
Leech. Geoffrey. Semantics. Harmondsworth: Penguin Books Ltd. 1974.
Lenneberg, Eric H., Biological Foundations of Language. New York: John Wiley
& Sons. 1967.
Levi-Strauss, Claude. 1945. ‘Structural Analysis in Linguistics and Anthropology’.
in Hymes 1964: 40–52.
Litosseliti, Lia. Gender and Language: Theory and Practice. London: Hodder
Education. 2006.
Lounsbury, F. G., ‘A Formal Account of Crow and Omaha-type Kinship
Terminologies: in Goodenough, W. H (ed). 1964. Explorations in Cultural
Anthropology. New York: Mc Graw Hill.
Lyons, John. 1970. Chomsky. London: Fontana Press, 15th impression, 1985.
Malcolm, Norman. Ludwig Wittgenstein: A Memoir. Oxford: Oxford University

Press.1958.
Matthews, Peter Hugoe. Morphology. Cambridge: Cambridge University Press.
1974.
Linguistics: A Very Short Introduction. Karachi: Oxford University Press.
2003.
McDavid, Raven I. (Jr)., Varieties of American English. Anwar S. Dil (ed).
Stanford, CA: Stanford University Press. 1980.
202 Bibliography and Further Reading
Mead, Margaret. 1935. Sex and Temperament in Three Primitive Societies. New
York: Mentor Book, 1950 edition.
Minault, Gail. 1984. ‘Begmati Zuban: Women’s Language and Culture in
Nineteenth-Century Delhi’. India International Centre Quarterly. 11: 2 (June):
155–170.
Secluded Scholars: Women’s Education and Muslim Social Reform in
Colonial India. Delhi: Oxford University Press. 1988.
Mohanty, Ajit K., Minati Panda, Robert Phillipson, and Tove Skutnabb-Kangas
(eds). Multilingual Education for Social Justice: Globalising the Local. Delhi:
Orient Blackswan. 2009.
Mohanty, Ajit K., 2009. ‘Multilingual education: A Bridge too Far’. In Mohanty
et.al. 2009: 5–17.
Nettle, David and Romaine, Suzanne. Vanishing Voices: The Extinction of the
Worlds’ Languages. New York: Oxford University Press. 2000.
Newmeyer, Fredereck (ed). Linguistics: The Cambridge Survey. (4 volumes)
Cambridge: Cambridge University Press. 1988.
O’ Connor, J. D., Phonetics. Harmondsworth: Penguin Books. 1973.

Ogden, C. K, and Richards, L. A., 1923. The Meaning of Meaning. London:
Routledge and Kegan Paul, 1952.
Ostler, Nicholas. Empires of the Word: A Language History of the World. London:
Harper-Collins Publishers. 2005.
AD Infinitum: A Biography of Latin. New York: Walker and Company.
2007.
Palmer, Frank. 1971. Grammar. Harmondsworth: Penguin Books. New edition

1983.
Phadnis, Urmila. Ethnicity and Nation Building in South Asia. New Delhi &
London: Sage Publications. 1989.
Pollock, Sheldon. The Language of the Gods in the World of Men: Sanskrit, Culture,
and Power in Pre-modern India. New Delhi: Permanent Black. 2006.
Pullum, Geoffrey K, and Ladusaw, William A. Phonetic Symbol Guide. Chicago:
The University of Chicago Press. 1986.
Radford, Andrew. Transformational Grammar: A First Course. Cambridge:

Cambridge University Press. 1988.
Syntax: A Minimalist Introduction. Cambridge: Cambridge University
Press. 1997.
Rahman, Tariq. Pakistani English: The Linguistics Description of a Non-Native
Variety of English. Islamabad: National Institute of Pakistan Studies, Quaid-i-
Azam University. 1990.
Language and Politics in Pakistan. Pakistani edition, Karachi: Oxford
University Press. 1996. Indian edition, Delhi: Orient Longmans, 2007.
Bibliography and Further Reading 203
Language, Education, and Culture. Karachi: Oxford University Press.

1999.
Language, Ideology, and Power: Language Learning Among the Muslims
of Pakistan and North India. Karachi: Oxford University Press. 2002.
2006. ‘Urdu as an Islamic Language’. The Annual of Urdu Studies
101–119.
Roach, Peter. 1983. English Phonetics and Phonology: A Practical Course.
Cambridge University Press, 1991 edition.
Robins, R. H. 1964. General Linguistics: An Introductory Survey. Longman, 3e.
1980.
Robinson, Andrew. 1995. The Story of Writing: Alphabets, Hieroglyphs and
Pictograms. London: Thames and Hudson Ltd, 2007.
Russell, Peter. The Brain Book. London and Henley: Routledge & Kegan Paul.
1979.
Saleemi, Anjum P., Ocke-Schwen Bohn and Albert Gjedde (eds). In Search of a
Language for the Mind-Brain. Aarhus, Denmark: Aarhus University Press.
2005.
Sampson, Geoffrey. 1985. Writing Systems. London: Hutchinson, 1987.
Sanders, Carol. 2005. ‘Saussure, Ferdinand de’. In Strazny. 2005. Vol. 2: 930–
932.
Saussure, Ferdinand de. 1916. A Course in General Linguistics. Trans. from French
by Wade Baskim. London: Peter Owen, 1960.
Sew, Jhy Wae. ‘Onomatapoeia’ in Strazny. 2005: 796–797.
Skutnabb-Kangas, Tove. Linguistic Genocide in Education or Worldwide Diversity
and Human Rights London: Erlbaum. 2000.
Strazny, Philipp. (ed). Encyclopedia of Linguistics. (2 vols.) New York & Abingdon
(UK): Fitzroy Dearborn. 2005.
Tannen, D. You Just Don’t Understand: Women and Men in Conversation. New
York: William Morrow. 1990.
UNESCO. Position Paper: Teaching in the Mother Tongue. Paris: UNESCO.

2003.
Voigtlander, Katherine and Lewis, Karen. 1990. The Alphabet Makers. California:
Summer Institute of Linguistics, 1991.
Wittgenstein, Ludwig. 1922 and 1961. Tractatus Logico-Philosophicus. London:

Routledge and Kegan Raul. 1993 reprint.
Yule, George. The Study of Language: An Introduction. Cambridge: Cambridge

University Press.1985.
Index
A Basque language, 23
Behaviourism, 23
accessibility hierarchy, 125 Bengali, 32, 190, 192, 193
acquisition planning, 164, 165, 176 Berlin, B. and Kay, P., 147, 149
adverbial phrase (Adv P), 85, 87 Bilabial sounds, 31, 37
affix hopping, 109 Bilingual education, 178
Akbar, 10 Binary branching, 100
Allomorphs, 58, 59, 60 Binary oppositions, 22, 44, 46, 131
Allophones, 51, 52, 130 Biological foundations of language, 5
Alpha movement, 109, 110 Bloomfield, Leonard, 3, 22, 23, 128
Alveolar sounds, 28, 33, 37, 43, 45 Boas, Franz, 22, 127, 128
American Indian language, 128 Bopp, Franz, 14
Anthropological linguistics, 5, 7, 22, Bourdieu, P., 181
64, 128, 129, 140, 198 Brahmi, 190, 191, 192
Antonymy, 63, 64 Brahvi, 56, 186, 190
Aphasia, 5 Brugmann, Karl, 14
Apical sound, 29, 47, 48 Burushaski, 16, 186
Arabic
grammar, 8, 13, 15, 19, 32, 35, 36, C
73
system of writing, 8, 166, 168, 185, Chilliso, 171
186, 188, 189, 190, 192, 197 Chinese, 16, 184, 187
Aramaic, 183, 188 Chomsky, Noam, ix, 4, 23, 66, 68,
Arbitrariness of language, 12 72–81, 83–4, 103–4, 113, 117,
Aristotle, 6, 12 125–6
articulation, 24, 29, 31, 33, 37, 38, 47 Chronos, 20
Ashtadhyayi (Panini), 12 cognitive universals, 148
Asiatic Society of Bengal, 12 comparative linguistics, 13, 14, 18
Aspects of the Theory of Syntax (Noam competence and performance, 22, 76
Chomsky), 76, 103, 199 complementary distribution, 130, 137,
Austin, John L., 71 138
Aztec, 58 complementizer phrases, (CP) 97
componential analysis, 66, 130, 132–
B 35, 137
computational and mathematical
Badeshi, 171 linguistics, 5
Balochi, 32, 186, 190 connotata, 133
Index 205
connotations of words, 154 ethnography of communication, 140

consonants, 31, 37, 38, 42, 45, 189, ethnology, 127–8
193 ethnic politics and language, 167
conventionalist and naturalistic debate, etymology, 14
13, 18
corpus planning and politics, 168 F
Cratylus (Plato), 12, 18, 68
Crystal, David, 3, 16, 66, 173–4 family tree theory, 15
Cultural determinism, 147 feral children, 10
Culture-specific beliefs, 8 Ferguson, Charles, 142
Foreign language teaching (FLT), 176
D Formalization, 80, 84
Fortis sounds, 47–8
De Lingua Latina (Marcus Terentius Frege, Gottlob, 68
Varro), 13 French, 15, 18, 19, 22, 32, 58, 75, 80,
dental sounds, 193 95, 159, 160, 181
descriptive grammar, 73 Fricatives, 33
Devanagari script, 6, 185, 192 Functional suitability of language, 10,
diachrony, 20 11
diacritical marks, 30, 186
diglossia, 142, 200 G
diphthongs, 42, 185, 192
Domaaki, 171 Gender construction and role of
dorsal sound, 29, 47–8 language, 150–2
Durkheim, Emile, 22 gender roles and language, 151–2, 154,
157
E German, 13–15, 23, 40, 58, 68, 142,
176
Educational linguistics, x, 140, 175–6, glides (semi-vowels), 195
181 glottal sounds, 29
embedding, 82 glottographic system of writing, 183–4
English government and binding theory, 111
Dominant language of education, Gowro, 171
179 grammar
Grammar, 75 linear model, 101, 103
morphology, ix, 2–4, 53, 61, 75, transformational generative (TG),
137 ix, 80, 81, 101
production of speech sounds, 24 universal, 75, 79, 109, 117, 118,
in South Asia, language of power, 124, 126
163 types, 124
English for Special Purposes (ESP), Grantha system of writing, 192
176 graphemes, 184–6, 188–9, 191
English Language Teaching (ELT), viii, Greek, 6, 12, 14, 15, 30, 53, 62, 65,
176–7 141–2, 183, 185, 190
206 Index
Grimm, Jakob, 14 Institute Ferdinand de Saussure, 22

Gujrati, 32, 192, 193 Interactionism, 129
Gupta system of writing, 192 International Phonetic Association
Gurmukhi, 192 (IPA), 30, 33, 35, 42, 48
intonation, 10
H Italian, 10
Haegeman, Liliane, 111–12 J

Hali, Altaf Hussain, 157
Halle, Morris, 44, 47 Jakobson, Roman, 44
Hammarabi, 188 Japanese, 119, 124
Handbook of American Indian Languages Jones, William, 14
(Franz Boas), 128
Harris, Zellig, 72 K
Haugen, Einar, 163
Hebrew, 16, 72, 188–90 Kashmiri, 190
Hermogenes, 12 Khan, Syed Ahmad, 163
hieroglyphs (Egyptian system of Kharosthi, 192
writing), 188 Kinship systems and linguistic concepts,
Hindi 132, 133, 139
grammar, 73 Kinship terms and world view, 150
morphemes, 168 Kroeber, Alfred, 128
writing, 192 Kurdish, 174
Hindko, 186, 190
Homo sapiens, 9 L
homograph, 65
homonymy, 63–4 labelling, 91
homophone, 65 labial, 47, 193
human brain, 74, 76–7 labio-dental sounds, 32
Humboldt, Wilhelm von, 22–3 language
Hymes, Dell, 127–9, 142 acquisition device (LAD), 76, 126
Hyponymy, 63–4 and culture, relation, 127
death, x, 164, 171–3, 175
I faculty, 125
and gender, 140, 150, 158
identity and language planning, 140 and identity, x, 140
implicational universals, 147 learning, 5, 125–6, 164, 179, 180
Indo-European family of language, 14, as a system, 20, 59
16, 190 and world view, 140, 146
Indus Valley script, 192 planning (LP), 163, 170, 176
infixes, 56 politics, x, 140, 162
inflection, 78, 94, 96 rights, 172, 180
inflectional languages, 57 shift, 164, 165, 172, 174, 175
ingressive air stream, 25 varieties, 143–6, 166
Index 207
langue (morphophonemics), 60
versus parole, 4, 21–2, 128 mother tongue, 74, 142, 172, 177–8,
larynx, 25–6 180, 188
Latham, R.G., 128 multilingual education, 173, 178, 179
Latif, Nawab Abdul, 163 myths, 8, 11
Latin, 13–15, 20, 29, 57, 73, 141, 158,
188 N
lenis sounds, 47–8
Levi-Strauss, Claude, 128, 131 nasal cavity, 26–7
lexeme, 63–4 nasal sounds, 45
lexicon, 20, 103, 105–6, 124 naskh and nastaleeq, 5, 185–6
linguistic neologism, 57–8, 166
diversity, 174 neurolinguistics, 5, 198
politeness, 140, 158–60 noun phrase (NP), 85–91, 93, 97, 103,
relativity, 146–7 105, 106, 108, 109, 112–13, 116,
linguistics 118–22
in antiquity, 12
brief history, 8–23 O
in the middle ages, 13
lips, role in speech sound, 27, 30, 31 onomatopoeic words, 19, 20
logographics oral cavity, 27
glottographic systems, 183–4 original language of humanity, 10
Lounsbury, F.G., 133, 137 Ostler, Nicholas, 171
M P
macrolinguistics, vii, 6, 181 palatal, 34–5, 37, 193

Malinowski, 128–9, 131 palato-alveolar sound, 34
Mauss, Marcel, 22 Panini, 12
Meaning Paradigmatic relationship, 21
in philosophy, 68 parole, 4, 21–2, 128
medium of instruction and schooling, Pashto, 5, 15, 19, 32, 35, 142, 148,
68 149, 151, 158, 169, 186, 190
minimalism, 100 performance, 4, 22, 67, 76
modernity and globalization, 172 Persian, 8, 13–15, 17, 19, 35–6, 57–8,
Moore, G.E., 70 73, 75, 153, 155, 157, 163, 168,
morphemes 185, 187, 190
classification, 54 Perso-Arabic script, 186, 192
bound, 54, 56 prescriptive grammar, 73
functional, 55 pharyngeals, 36, 46
inflectional, 54, 57–8 pharynx, 26, 36
morphological rules, 60, 78 phatic communion, 129
morphology, Ix, 2–4, 53, 61, 75, 137 philology, 13, 18, 23
morphophonology
208 Index
Philosophical Investigations (Ludwig received pronunciation (RP), 146

Wittgenstein), 70 recursiveness, 82
Phoenician, 182, 189 register, 141–2
phonemes, 10, 30, 32–3, 37, 43–5, relative clauses, 124–5
49–52, 60, 130, 184, 186 retroflex sounds, 33
phonetics (study of sounds), ix, 2, 3, Reversing Language Shift (RLS), 165
12, 23, 24, 30, 37, 192 Romans, 13, 190
phonograms, 188 Ronat, Mitsou, 80
phonographic script, 186, 189 Roy, Raja Ram Mohan, 163
phonological rules, 48, 59, 78 rule and principles, 74, 76–7, 82, 117
phonology (sound patterns), 2–4, 44, Russell, Bertrand, 68–70, 72
52, 75, 130, 131, 137 Russel, Peter, 203
phrase structure (PS) rules, 88, 93, Ryle, G., 71
101–04, 106
Plato, 6, 12 S
pluralization morphemes, 60
politeness Sanskrit, 12–15, 17, 57, 168, 187,
politics of language, 158 190–91
Pollock, Sheldon, 190–91 Sapir, Edward, 22, 128, 146, 148
polysemy, 63–4 See also Whorf, Benjamin Lee, 22,
Portuguese, 8, 58 128, 146, 147–8
positron emission tomography (PET), 5 Sarwar, Zakia, 176
posterior sounds, 45 Saussure, Ferdinand de, 3, 4, 13, 18,
pragmatics, 2, 3, 70–1 19, 20, 22, 128
Prague School of Linguistics, 21 Schleicher, August, 15
Prefixes, 56 schwa, 42
prepositional phrase (PP), 85, 87–8, scripts, 5, 6, 170, 182–8, 190–2, 197
90–1, 105–6, 118, 120, 122 See also writing, ix, x, 2, 13, 24, 74,
preventive linguistics, 175 84, 88, 143, 145, 164–5, 173, 175,
projection principle, 119, 124 178, 182, 183–4, 186, 188, 191–2,
pronunciation of foreign languages, 30 197
properties, 10, 24, 66–7, 74, 79, 81, second language teaching (SLT), 176
117, 120, 124, 133 segmental and non-segmental features,
proto Indo-European (PIE) family tree, 186
15, 17 semantics (study of meanings), ix, 2, 3,
pulmonary (egressive) stream, 25 12, 62, 66, 68, 71, 75, 119, 128,
Punjabi, 11, 15, 32, 51, 55, 119, 149, 133
158, 168, 186, 187, 190, 192 component, 103
problems, 66
R rules, 78
semasiographic systems, 183
Radcliffe-Brown, 131 semi vowel, 31, 35
Radford, Andrew, 109, 110 Semitic family, 15
Rask, Rasmus, 14
Index 209
sentence, 20–1, 55, 67, 69, 73–4, 76– synonymy, 63

7, 80–3, 85–9, 91–2, 94, 96, 97–8, syntactic rules, 78
101–09, 111, 113–15, 117–21, Syntactic Structures (Noam Chomsky),
123–4, 150, 155, 180 80, 103, 199
sexual differences, 151 syntactic structures; tree diagram, 91
Sign, 19, 44, 51, 148, 183 syntagmatic versus paradigmatic
Sindhi, 5, 15, 25, 27, 32, 51, 169, 186, relationship, 20
190, 196 syntax, 2–4, 10, 67–8, 73–6, 80, 83,
Siraiki, 25, 150, 166, 186, 190 93, 100, 103, 109, 120–21, 124
Skutnabb-Kangas, Tove, 172, 174,
179 T
social factors and linguistic change,
129 Tarski, Alfred, 69
Society of Pakistan English Language Teaching of English as a Foreign
Teachers (SPELT), 176 Language (TEFL), 176
sociolinguistics, 5, 140, 198 Teaching of English as a Second
Socrates, 12 Language (TESL), 177
sound Teeth, speech sound, role in, 27–9, 32,
classification, 29 45
patterns, 44 Thanvi, Maulana Ashraf Ali, 157
physical properties, 24 Thematic (theta) roles, 122, 124
South Asian language Theoretical language, 141
Production of speech sounds, 24 Tongue, speech sound, role in, 12,
Spanish, 32 26–9, 32–6, 38–9, 40, 42, 45–7
speech defects, 5 Tractatus (Ludwig Wittgenstein), 69,
speech sounds, production, 24 70
standardization, 145, 165 transcription, 24, 42–3
stops (plosives), 27, 31–2, 35, 45 transformational generative (TG)
Strawson, P.F., 71 grammar, 80, 81, 104
strident sounds, 46 transformational rules (T-rules), 103–
structural change (SC), 104 4, 106
structural description (SD), 104, 106, passive transformation, 105–6,
108 108
structuralism Trubetskoy, Nikolay Sergeyvich, 22
structure, deep and surface, 83 Turkish, 16, 58, 174, 190
structure dependency principle, Tyler, Sir Edward, 127
117, 118, 124
structure preserving principle, 110 U
style, 6, 13, 156–7
suffix, 56 United Nations Educational Scientific
Sumerian script, 187 and Cultural Organization
Sweet, Henry, 30 (UNESCO), 173, 178
Symbols used in writing rules, 84 Urdu
synchrony versus diachrony, 20 grammar, 13, 75, 77–8, 95, 126
210 Index
morphemes W
script, 32, 185
Ushojo, 171 Wandt, Wilhem, 23
Uvular sounds, 36–7 Western language policies, 179
Whorf, Benjamin Lee, 22, 128, 146–8
V Wittgenstein, Ludwig, 69, 70–1
women’s language (WL), 156–7
Varro, Marcus Terentius, 13 writing
velar sounds, 28, 35 history, x, 187
verb phrase (VP), 85, 87–9, 90–2, 97, in South Asia, 182–197
103–4, 112, 116, 118–9, 126
vibration, 26 X
vocal cords, 25–6, 29, 31, 36, 44–5
voicing, 26 X-bar theory, 93, 94, 118
vowels
classification, 38, 47
cardinal, 39, 40–2
dipthongs, 42, 192

Linguistics For Beginners - Basic Concepts

Uploaded by

Copyright:

Available Formats

Linguistics For Beginners - Basic Concepts

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Linguistics For Beginners - Basic Concepts

Uploaded by

Copyright:

Available Formats

Tariq Rahman

For sale in Pakistan and not for export therefrom.

Typeset in Adobe Garamond Pro

T his book is written to introduce South Asian students to

theories born as it developed, such as the government binding

D o you know how many languages exist in South Asia? If not,

All this is language in action. Such an important thing—human

What do we study in linguistics?

All human languages are made up of sounds arranged in such ways

SOUND GRAMMAR MEANING

The linguist Leonard Bloomfield (1887–1949), one of the pioneers

Every structure is a system in itself. If you change one part of it

The knowledge of the whole system is in the mind of the native

politics, and so on. The subject dealing with the relationship of

everything including e-mail in Pakistan. In India we have many

E very society, and indeed almost every individual, has myths

(expressiveness). When we do the latter we may combine words

Of course, nobody knows definitely whether there ever was a stage

The first language

If language arose with the Homo Sapiens (sapient= wise) then it

Another question which people ask is whether there is a natural or

Are languages equal?

find it difficult to set up a standard of excellence and then measure

Most Western works on the history of linguistics begin from Plato

The debate between the conventionalist and naturalistic views

Linguistics in the Middle Ages

In the nineteenth century linguistics progressed most in central

As the meanings of words change, these scholars had to study the

Besides the Indo-European family of languages, there are many

Twentieth century linguistics

Modern linguistics was born mainly out of the work of Ferdinand

‘The utter inadequacy of current terminology, the need to reform it,

Such a terminology he made himself, though he did not publish it

(1) Linguistic signs are arbitrary

If you remember the conventionalist–naturalistic debate from

The onomatopoeic words in Malay are as follows:

Barring these partial exceptions, Saussure’s theory is correct. This

(2) Synchrony versus diachrony

Saussure distinguished between synchronic and diachronic studies

(3) Syntagmatic versus paradigmatic

Consider the following sentence.

(4) Langue versus parole

As this concept has been explained earlier, there is no need to spend

by Nikolay Sergeyvich Trubetskoy (d. 1965), and it carried on the

Humboldt was a German linguist who worked on the Basque

T he study of sounds in human speech is called phonetics. It

The production of speech sounds

(2) Vocal cords

(4) The nasal cavity

(5) The oral cavity

(a) The lips

(b) The teeth

(c) Alveolar ridge

(d) Hard palate

(e) Soft palate

(g) The tongue

Some books use the following adjectives:

There are differences in terminology in linguistics and you are

(h) The glottis