Vous êtes sur la page 1sur 12


A Preview

What is human language? What does it mean to "know" a language? To answer
these questions, it is first necessary to understand the resources that a language
makes available to its native speakers, those who have acquired it as children in a
natural setting. The scope and diversity of human thought and experience place
great demands on language. Because communication is not restricted to a fixed set
of topics, language must do something more than provide a package of ready-
made messages. It must enable us to produce and understand new words, phrases,
and sentences as the need arises. In short, human language must be creative
allowing novelty and innovation in response to new experiences, situations, and
Underlying the creative aspect of language is an intricate mental system that
defines the boundaries within which innovation can take place. The operation of
this system can be illustrated by a relatively simple phenomenon in English: the
process that creates verbs (roughly, words naming actions) from nouns (roughly,
words naming things). (Table 1)

Table 1.1 Nouns used as verbs

Noun use Verb use
leave the boat on the beach beach the boat
keep the airplane on the ground ground the airplane
crush the aspirin into powder powder the aspirin
stab the man with a knife knife the man
catch the fish with a spear spear the fish
make the child an orphan orphan the child

However, there are also constraints on this freedom. For instance, a new verb is
rarely coined if a word with the intended meaning already exists.
Although we may say carton the eggs to mean `put the eggs in the carton', we
do not say hospital the patient to mean `put the patient in the hospital'. This is
presumably because the well-established verb hospitalize already has the meaning
that the new form would have.
There are also narrow constraints on the meaning and use of particular
subclasses of these verbs. One such constraint involves verbs that are created from
time expressions such as summer and holiday.

2. a) Julia summered in Paris.

b) Kent wintered in Mexico.
c) Martine holidayed in France.
d) They honeymooned in Hawaii.

While the sentences in 2 are all acceptable, not all time expressions can be used in
this way. (Throughout this book an asterisk is used to indicate that a sentence is

3. a) *Jerome midnighted in the streets.

b) *Andrea nooned at the restaurant.
c) *Philip one o'clocked at the airport.

These examples show that when a verb is created from a time expression, it must
be given a very specific interpretationroughly paraphrasable as `to be
somewhere for the period of time X'. Thus, to summer in Paris is `to be in Paris for
the summer,' to holiday in France is `to be in France for the holidays', and so on.
Since noon and midnight express points in time rather than periods of time, they
cannot be used to create verbs of this new class.
Constraints are essential to the viability of the creative process. If well-
established words were constantly being replaced by new creations, the vocabulary
of English would be so unstable that communication could be jeopardized. A
similar danger would arise if there were no constraints on the meaning of new
words. If winter in Hawaii could mean `make it snow in Hawaii' or `wish it were
winter in Hawaii' or any other arbitrary thing, the production and interpretation of
new forms would be chaotic and would subvert rather than enrich communication.
This rule-governed creativity characterizes all levels of language, including the
way in which sounds are combined to form words. The forms in 4, for instance,
are recognizable as possible names for new products or inventions in English, [or
the second file for inventions with Spanish designations]

4. a) Eng. prasp Sp. rufro

b ) Eng flib Sp. blerco
c) Eng.traf Sp. mirienes

Such forms contrast with the patterns in 5, which simply do not have the "sound"
of English words. [or for the case, also imposible forms in Spanish]

5. a) psarp
b) bfli

c) ftra

The contrast shows that our subconscious knowledge of English includes a set of
constraints on possible sequences of sounds.
Still other constraints determine how new words can be created from already
existing forms with the help of special endings. Imagine, for example, that you
learn that there is a word soleme (used perhaps for a newly discovered atomic
particle). As a speaker of English, you then automatically know that something
with the properties of a soleme can be called solemic. You also know that to make
something solemic is to solemicize it, and you call this process solemicization. [For
Spanish, if you know the worrd froco (meaning ligue) you may suppose there is a
verb frocar, and other nouns like frocador, or frocamiento, frocazn).
Nowhere is the ability to deal with novel utterances in accordance with rules
more obvious than in the production and comprehension of sentences. Apart from
a few fixed expressions and greetings, much of what you say, hear, and read in the
course of a day consists of sentences that are novel to you. In conversations,
lectures, newscasts, and textbooks you are regularly exposed to novel
combinations of words, the expression of unfamiliar ideas, and the presentation of
new information. Such is the case with the sentences you have just read. While
each of these sentences is no doubt perfectly comprehensible to you, it is
extremely unlikely that you have ever seen any of them before.
This ability to deal with novel utterances does not ensure that you can
understand or use any imaginable combination of words. You would not ordinarily
say a sentence such as 6a, although 6b would be perfectly acceptable.

6. a) *He brought a chair in order to sit on.

b) He brought a chair to sit on.

Or, to take another example, 7a is well formed-if bizarre-but 7b is gibberish.

7. a) The pink kangaroo hopped over the talking lamp.

b) *Pink the the talking hopped kangaroo lamp over.
As with other aspects of language, your ability to produce and comprehend
sentences is subject to limitations.


As we have seen, speakers of a language are able to produce and understand an
unlimited number of utterances, including many that are novel and unfamiliar.

This ability, which is often called linguistic competence, constitutes the central
subject matter of linguistics and of this book.
In investigating linguistic competence, linguists focus on the mental
system that allows human beings to form and interpret the words and sentences
of their language. This system is called a grammar. For the purposes of this book,
we will divide the grammar into the following components.

Table 1.2 The components of a grammar

Component Responsibility
Phonetics the articulation and perception of speech sounds
Phonology he patterning of speech sounds
Morphology word formation
Syntax sentence formation
Semantics the interpretation of words and sentences

Linguists use the term grammar in a rather special and technical way. Because this
usage may be unfamiliar, we will devote some time to considering several
properties of the system that linguists call a grammar.

GENERALITY: All language have a Grammar

One of the most fundamental claims of modern linguistic analysis is that all
languages have a grammar. This can be verified by considering a few simple a
facts. Since all languages are spoken, they must have phonetic and phonological
systems; since they all have words and sentences, they also must have a
morphology and a syntax; and since these words and sentences have systematic
meanings, there obviously must be semantic principles as well. As these are the
very things that make up a grammar, it follows that all human languages have this
type of system.
It is not unusual to hear the remark that some language-Acadian French,
Navaho, or Chinese-"has no grammar." (This is especially common in the case of
languages that are not written or have not yet been analyzed by Western scholars
[like some indigenous languages of America, or Australia, or New Guinea].
Unfamiliar languages sometimes appear to an untrained observer to have no
grammar simply because their grammatical systems are different from those of
better-known languages. In Walbiri (an aboriginal language of Australia), for
example, the relative ordering of words is so free that the English sentence The
two dogs now see several kangaroos could be translated by the equivalent of any
of the following sentences.

8. a) Dogs two now see kangaroos several.

b) See now dogs two kangaroos several.
c) See now kangaroos several dogs two.
d) Kangaroos several now dogs two see.
e) Kangaroos several now see dogs two.

Whereas Walbiri may not restrict the order of words in the way English does, its
grammar imposes other types of requirements. For example, in the sentence types
we are considering, Walbiri speakers must place the ending lu on the word for
`dogs' to indicate that it names the animals that do the seeing rather than the
animals that are seen. In English, by contrast, this information is conveyed by
placing two dogs in front of the verb and several kangaroos after it.
Rather than showing that Walbiri has no grammar, such differences simply
demonstrate that it has a grammar unlike that of English in certain respects. This
important point is applicable to all differences among languages: although no two
languages have exactly the same grammar, there are no languages without a
A similar point can be made about different varieties of the same language. As
you are probably already aware, English is the language of many different
communities around the world. The particular variety of English found within each
of these communities has its own characteristic pronunciation, vocabulary, and
sentence patterns. This is just another way of saying that each variety of English
has its own grammar. Just as it is impossible to have a language without a
grammar, so no variety of language could exist if it did not have a grammar.

EQUALITY. All grammars are equal

Whenever there is more than one variety of a particular language, questions arise
as to whether one is somehow better or more correct than another. From the point
of view of modern linguistics, it makes no more sense to say that one
variety of English is better than another than it does to say that the grammar of
English is better (or worse) than the grammar of Thai.
All languages and all varieties of a particular language have grammars that
enable their speakers to express any proposition that the human mind can produce.
In terms of this all-important criterion, then, all varieties of language are
absolutely equal as instruments of communication and thought. The goal of
contemporary linguistic analysis is not to rank languages on some imaginary scale
of superiority. Rather, linguists seek to understand the nature of the grammatical
systems that allow people to speak and understand a language.

This same point is sometimes made by noting that linguistics is descriptive, not
prescriptive. This means that linguists seek to describe human linguistic ability
and knowledge, not to prescribe one system in preference to another. A parallel
point of view is adopted in other scientific disciplines as well. The first concern of
all scientists is to describe and explain the facts that they observe, not to change
Even though it rejects prescriptivism, modern linguistic analysis does not deny
the importance of clear expression in writing and speech. Such skills are quite
rightly an object of concern among educators. However, the difficulties that arise
in these areas typically result from the inconsistent or careless use of one's
linguistic knowledge, not from any inherent flaw in the grammar itself.
Linguists also acknowledge that certain patterns (I seen that, They was there,
He didn't do nothing, He ain't here) may be restricted to particular socioeconomic
groups within the English-speaking community. As discussed in more detail in
chapter 12, the use of these patterns may therefore have negative social
consequences: it may be harder to win a scholarship, to get a job, to be accepted in
certain circles, and so forth. From a purely linguistic point of view, however, there
is absolutely nothing wrong with grammars that permit such structures. Like
grammars for other variants of English (and other languages), they permit their
users to express and understand the same unlimited range of thoughts and ideas.

CHANGEABILITY. Grammars change over time

It is a well-established fact that the grammars of all languages are constantly
changing. Some of these changes are relatively minor and occur very quickly (for
example, the addition of new words such as glasnost, yuppie, fax, cursor, and
attrit to the vocabulary of English). Other changes have a more dramatic effect on
the overall form of the language and typically take place over a long period of
time. The formation of negative structures in English has undergone this type of
change. Prior to 1200, English formed negative constructions by placing ne before
the verb and a variant of not after it.

9. a) Ic ne seye not. (`I don't say.')

b) He ne speketh nawt. (`He does not speak.')

By 1400 or thereabouts, ne was used infrequently and not (or nawt) typically
occurred by itself after the verb.

10. a) I seye not the wordes.

b) We saw nawt the knyghtes.

It was not until several centuries later that English developed its current practice of
allowing not to occur after only certain types of verbs (such as do, have, will, and
so on).
11. a) I will not say the words. (versus *I will say not the words.)
b) He did not see the knights. (versus *He saw not the knights.)

These modifications illustrate the extent to which grammars can change over time.
The structures exemplified in 10 are archaic by today's standards and those in 9
sound completely foreign to most speakers of modern English.
Through the centuries, individuals and organizations who believe that certain
varieties of language are better than others have frequently expressed concern over
what they perceive to be the deterioration of English. In 1710, for example, the
writer Jonathan Swift (author of Gulliver's Travels) lamented "the continual
Corruption of our English Tongue." Among the corruptions to which Swift
objected were contractions such as he's for he is, although he had no objection to
Its for It is.
In the nineteenth century, Edward S. Gould, a columnist for the New York
Evening Post, published a book entitled Good English; or, Popular Errors in
Language, in which he accused newspaper writers and authors of "sensation
novels" of ruining the language by introducing "spurious words" like jeopardize,
leniency, and underhanded. To this day, the tradition of prescriptive concern about
the use of certain words continues in the work of such popular writers as Edwin
Newman and John Simon, who form a kind of self-appointed language police.
Linguists reject the view that languages attain a state of perfection at some
point in their history and that subsequent changes lead to deterioration and
corruption. As noted above, there are simply no grounds for claiming that one
system of grammar is somehow superior to another. There is therefore no reason to
think that language change can or will undermine the adequacy of English (or any
other language) as a medium of communication.

UNIVERSALITY: Grammars are alike in basic ways

There are many differences among languages, as even a superficial examination of
their sound patterns, vocabularies, and word order reveals. But this does not mean
that there are no limits on the type of grammars that human beings can acquire
and use. Quite to the contrary, current research suggests that there are important
grammatical principles and tendencies shared by all human languages.
One such principle involves the manner in which sentences are negated. With
unlimited variation, one would expect the equivalent of English not to occur in
different positions within the sentence in different languages. Thus, we might

predict that each of the following possibilities should occur with roughly equal

12. a) Not Pat is here.

b) Pat not is here.
c) Pat is not here.
d) Pat is here not.

As it happens, the first and fourth patterns are very rare. In virtually all languages,
negative elements such as not either immediately precede or immediately follow
the verb.
The relative ordering of other elements is also subject to constraints. To see
this, we need only consider the six logically possible orders for a simple three-
word statement such as Canadians like hockey.

13. a) Canadians like hockey.

b) Canadians hockey like.
c) Like Canadians hockey.
d) Like hockey Canadians.
e) Hockey like Canadians.
f) Hockey Canadians like.

Interestingly, more than 95 percent of the world's languages adopt one of the first
three orders for basic statements. Only a handful of languages use any of the last
three orders as basic. This once again reflects the existence of constraints and
preferences that limit variation among languages.
These are not isolated examples. As later chapters will show, some grammatical
categories and principles are universal. And where there is variation (as in the case
of word order), there is typically a very limited set of options. Contrary to first
appearances, then, the set of grammars learned and used by human beings is
limited in significant ways.

TACITNESS: grammatical knowledge is subconscious

Because the use of language to communicate presupposes a grammar, it follows
that all speakers of a language must have knowledge of its grammar. However, this
knowledge differs from knowledge of arithmetic, traffic safety, and other subjects
that are taught at home or in school. Unlike these other types of knowledge,
grammatical knowledge is acquired without the help of instruction when one is
still a child and it remains largely subconscious through out life. As an example of

this, consider your pronunciation of the past tense ending written as ed in the
following words.

14. a) hunted
b) slipped
c) buzzed

Notice that whereas you say id in hunted, you say t in slipped and d in buzzed.
Moreover, if you heard the new verb flib, you would form the past tense as flibbed
and pronounce the ending as d. Although it is unlikely that you have ever been
aware of this phenomenon before now, you make these distinctions automatically
if you are a native speaker of English. This is because the grammatical subsystem
regulating this aspect of speech was acquired when you were a child and now
exists subconsciously in your mind.
Even more subtle phonological patterning can be found in language, as the
following contrasts help illustrate.

15. pint *paynk

fiend *fiemp
locked *lockf
wronged *wrongv
next *nexk
glimpse *glimpk

The words in the left-hand column obey an obscure constraint on the selection of
consonant sequences in word-final position: when a vowel is long and fol lowed
by two consonants (pint) or when a vowel is short and followed by three consonant
sounds (next, pronounced `nekst'), the final consonant must always be one made
with the tongue tip raised. (The consonants t, d, s, and z are made in this manner,
but consonants such as p, f, v, and k are not.) Words that do not adhere to this
phonological constraint (the right-hand column) are unacceptable to speakers of
English. Even linguists have to dig deeply to uncover such patterning, but in
everyday language use, we routinely make decisions about the acceptability of
forms based on subconscious knowledge of such constraints.
Consider one final example. Speakers of English know that there are certain
structures in which the word he can refer to each member of a group or to a single
individual outside that group.

16. Each boy who the woman interviewed thinks that he is a genius.

Sentence 16 can mean either that each boy in the group that the woman
interviewed thinks that he himself is a genius or that each boy thinks that a
particular person not mentioned in the sentence (say, the teacher) is a genius.
However, only one of these interpretations is possible in the following sentence.

17. The woman who each boy interviewed thinks that he is a genius.

In 17, he can refer only to someone not mentioned in the sentence. In contrast with
what happens in sentence 16, he cannot refer to each individual in the group
designated by the phrase each boy. Since speakers are able to make this contrast,
they must have knowledge of the relevant grammatical principle even though they
are not consciously aware of it.

Linguists use the term grammar to refer to a subconscious linguistic system of a
particular type. Consisting of several components (phonetics, phonology,
morphology, syntax, and semantics), a grammar makes possible the production
and comprehension of a potentially unlimited number of utterances. Because no
language can exist without a grammar and no one can use a language without
knowledge of its grammar, the study of grammatical systems has come to be the
focus of contemporary linguistic analysis.
As noted above, the grammatical knowledge needed to use and understand
language is acquired without the benefit of instruction and is for the most part
subconscious. Since we therefore cannot investigate grammar by simply recalling
prior training or by self-consultation, the study of human linguistic systems
requires considerable effort and ingenuity. As is the case in all science, information
about facts that can be observed (the pronunciation of words, the interpretation of
sentences, and so on) must be used to draw inferences about the sometimes
invisible mechanisms (atoms, cells, or grammars, as the case may be) that are
ultimately responsible for these phenomena. A good deal of this book is concerned
with the findings of this research and with what they tell us about the nature and
use of human language.

As far as can be determined, the languages spoken in the world today cannot be
traced to a common source. Rather, they seem to belong to a number of distinct
families whose histories can be traced back no more than a few thousand years.
Archaeological evidence suggests that language existed prior to that time for

perhaps as long as 100,000 years, but virtually nothing is known about this period
of linguistic prehistory or about how language originated in the first place.
There is every reason to believe, though, that humans have a special capacity
for language that is not shared by other creatures. The evolutionary adaptation of
certain physiological mechanisms for linguistic ends has occurred only in humans.
The so-called speech organs (the lungs, larynx, tongue, teeth, lips, palate, and
nasal passages) did not originally evolve for speech; rather, they were-and still are-
directly concerned with ensuring the physical survival of the organism. But each
nonlinguistic use of these organs is paralleled by a linguistic use unique to
humans. Table 1.3 compares the linguistic uses of the major speech organs with
their primary survival functions in humans and other mammals.
In humans, these organs have all become highly specialized for linguistic ends.
The vocal folds, for example, are more muscular and less fatty in humans than in
nonhuman primates such as chimpanzees and gorillas.

Table 1.3 Dual functions of the speech organs

Organ Survival function Speech function
Lungs to exchange CO2, oxygen to supply air for speech
Vocal folds to seal over passage to lungs to produce voice for speech sounds
Tongue to move food back to throat to articulate vowels and consonants
Lips to searl oral cavity to articulate vowels and consonants
Teeth to break up food to provide place of articulation for

Because of a highly developed network of neural pathways, they also respond

more precisely to commands from the brain. The same extensive set of neural
pathways allows a high degree of control over other speech organs, such as the
tongue, palate, and lips. Such control exceeds anything found in even our closest
primate relatives.
There are additional indications of the evolution of linguistic vocalization.
Unlike the breathing of survival respiration, speech breathing shows higher lung
pressure and a longer exhalation time than respiration. Abdominal muscles that are
not normally employed for respiration are brought into play in a systematic and
refined manner in order to maintain the air pressure needed for speech. Again, a
specialized, extensive set of neurological controls exclusive to humans makes this
type of breathing possible.
In other words, the human capacity for speech is superimposed on already
existing biological structures. Evolution has produced a refinement both in degree
and in kind through a long interplay between the demands of language and the
development of the human speech-producing apparatus.

We know considerably less about the evolutionary specialization for nonvocal
aspects of language such as word formation, sentence formation, and the
interpretation of meaning. Nonetheless, it is clear that some sort of evolutionary
specialization must have occurred. As we will see in Chapter 9, specific parts of
the brain are associated with each of these linguistic activities. Moreover, the brain
areas in question have no counterparts in other species. These facts suggest that the
human brain is specially structured for language, and that species with different
types of brains will not be able to acquire or use the types of grammars associated
with human language. After devoting most of this book to the study of
grammatical phenomena in human language, we will, in Chapter 14, return to the
question of whether comparable linguistic systems occur in other species.

Human language is characterized by rule-governed creativity. Speakers of a
language possess a grammar, a mental system of elements and rules that allows
them to form and interpret familiar and novel utterances. The grammar governs the
articulation, perception, and patterning of speech sounds, the formation of words
and sentences, and the interpretation of utterances. Contrary to popular belief, all
languages have grammars that are roughly equal in complexity and are acquired
subconsciously by their speakers. The existence of such linguistic systems in
humans is the product of unique anatomical and cognitive specialization.