Language

UNIT 2 LANGUAGE PROCESSING

(COMPREHENSION AND
LANGUAGE EXPRESSION)
“Language is a process of free creation; its laws and principles are fixed, but
the manner in which the principles of generation are used is free and infinitely
varied. Even the interpretation and use of words involves a process of free
creation.”
– Noam Chomsky
Structure
2.0 Introduction
2.1 Objectives
2.2 Functions of Language
2.3 Structure of Language
2.3.1 Basic Units of Language: Phonemes and Morphemes
2.3.2 Higher Levels of Linguistic Analysis
2.3.3 Phase Structure of Sentences
2.3.4 Structure and Deep Structure in Sentences
2.4 Processes in Language
2.4.1 Production of Language
2.4.2 Speech Perception and Comprehension
2.5 Language Development
2.6 Comprehensive Model of Language Processing
2.6.1 Kintsch’s Model of Comprehension
2.6.2 Propositional Representation of Text and Reading
2.7 Let Us Sum Up
2.8 Unit End Questions
2.9 Suggested Readings and References

2.0 INTRODUCTION
One form of knowledge shared by all human societies is the knowledge of
language. Language is the principal means by which we acquire and express
knowledge; thus, the study of how language is used is a central concern of
cognitive psychology. In the previous chapter, we studied the modes of language
acquisition; here we will do an overview of the research on the processes of
language involving the comprehension and expression of language.

Language, like other cognitive functions, is a (dynamic) process. Such an

enterprise requires that we examine the microstructure of the entire process as it
occurs in real time. It is only through the careful examination of the temporal
course of mental operations involved in the various levels of analysis underlying
speech that we can hope to discover its nature.

How do we understand language, given its multifaceted encoding? One approach

to this question centers on the psychological processes involved in speech
22
perception (Hickok & Poeppel, 2000). It also considers how listeners deal with Language Processing
(Comprehension and
the peculiarities resulting from the acoustic (related to sound) transmission of Language Expression)
language. A, second, more linguistically oriented approach focuses on descriptions
of the grammatical structure of language. Finally, a third approach examines the
psycholinguistic processes involved in language comprehension at the discourse
macro-level of analysis. All three approaches overlap to some degree and offer
interesting insights into the nature of language, its use, and understanding.

2.1 OBJECTIVES
After reading this unit, you will be able to:
• Examine the basic features involved in the language processing;
• Explain the expression and comprehension of language;
• Elucidate the functions of Language;
• Analyse the structure of Language;
• Explain the processes in Language; and
• Describe Walter Kintsch’s model of comprehension.

2.2 FUNCTIONS OF LANGUAGE

Language serves many functions, which are all related to the fundamental process
of communication. Perhaps most important is that language conveys meaning
and is part of almost all kinds of social interaction. Language conveys intentions,
motives, feelings, and beliefs. Language is used to issue requests and commands;
and is also used to teach and to convey information. Language is useful because
it can represent ideas and events that are not tied to present. You can also describe
abstract ideas, such as beauty and justice, as well as concrete objects of everyday
experience. Thus, language is symbolic, in that speech sounds and utterances
stand for or represent various objects, ideas, and events.

Regardless of whether we are considering spoken language, written language, or

sign language, there are three elements of language expression and human
communication that have been identified as operating in the speaker-listener
situation: speech acts, propositional content, and thematic structure. A brief
description from the analysis by Clark & Clark (1977) is as follows:
i) Speech Acts: Speakers normally intend to have some influence on their
listeners. To do so, speakers get the listeners to recognise the speakers’
intentions. Indeed, failure to recognise these intentions can result in awkward
situations. Speech-act theory holds that all utterances can be classified as to
the type of speech act they represent. For example, speech acts may make
assertions, make verbal commitments, convey thanks, give a warning, or
issue a command. Typical examples of speech acts including the following:
“I insist that you turn down the volume on the stereo” (a command); “What
are your plans for weekend?” (a question); “I promise to pay you tomorrow”
(a verbal commitment), symbolise ordering, questioning, committing etc,
which are common direct speech acts.
Searle (1969) pointed out that some speech acts are indirect. When your
mother asks if you live in a barn, a guest in your house asks if you are chilly,
23
Language they are conveying information about their desires, but in a rather indirect,
nonliteral way. The meaning of any particular speech act, including whether
it is direct or indirect, will depend on the context in which it is uttered
(Gibbs, 1986), as well as its content.
ii) Propositional Content: The second element of communication concerns
the propositional content of a sentence. In communication, speakers want
to convey certain ideas, and to do this, they must be sure that they are
understood. Thus, the content around a speech act is very important. As a
general rule, the propositional content of a sentence is used to describe certain
states or events; it can be part of other propositions. For example, the sentence
“The bright student received an A in Mathematics” expresses two separate
propositions: “the student is bright” and “the student received an A in
Mathematics.” Combined into a single sentence, the propositions convey
what the speaker intends to convey. There is experimental evidence that we
represent as propositions. For example, the more propositions contained in
a sentence, the longer the time required to read the sentence (van Dijk &
Kintsch, 1983)(discussed in detail in last section).

iii) Thematic Structure: The third component in communication is thematic

structure. To communicate effectively, good speakers pay careful attention
to their listeners. Good speakers have to judge what listeners do and do
know, keep track of where they are leading their listeners, and regularly
examine any assumptions about the listeners’ knowledge of the topic being
discussed. In short, the speaker must be able to make reasonably accurate
judgments of the listener’s current level of understanding. All of these
features are present in good teachers, entertaining and effective storytellers,
and interesting conversationalists.

2.3 STRUCTURE OF LANGUAGE

A theoretical intervention about the process which leads to the understanding of
an utterance in communication should involve two aspects. Firstly, the aspects
of language linked to the recognition of the form of the utterance itself (phonology,
morphology, and syntax); secondly, questions about how the meaning of what is
understood can be defined, which are linked to semantics and pragmatics of the
communication process. These two aspects cannot be separated, and in order to
analyse the process of language, both are to be taken into consideration. Thus, to
understand the language processes, it is fundamental to understand the basic
structure of language first.

As should be evident by now, language can be divided into three basic parts,
each with its own structure and rules: phonology, syntax (grammar), and
semantics. The first of these, phonology, concerns the rules for pronunciation of
speech sounds. The second aspect of language, syntax, deals with the way words
combine to form sentences. And semantics focuses on the meaning of words and
sentences.

2.3.1 Basic Units of Language: Phonemes and Morphemes

All languages are made of basic sounds called phonemes. Adult human beings
can produce approximately 100 phonemes, and the English language is made up
24
of about 45 phonemes. Languages vary in the number of phonemes, ranging Language Processing
(Comprehension and
from as few as 15 to as many as 85. One reason why it is difficult for many Language Expression)
Americans to learn foreign languages is that different phonemes are used. For
instance, Germanic and Slavic languages contain phonemes never used in the
English language. (Phonemes and morphemes have already been defined in the
previous chapter).

2.3.2 Higher Levels of Linguistic Analysis

The study of speech sounds which make up a language is called phonology, and
the study of how these sounds combine to produce morphemes is called
morphology. However, psychologists are frequently interested in a more global
analysis of language than is provided by phonology and morphology.
Psychological investigations of language typically adopt words, phrases,
sentences, or prose, rather than more elementary speech sounds, as the most
fundamental unit of analysis.

There are several levels at which these higher-order analyses can be made.
1) First, one could analyse the lexical content of a sentence or of some other
unit of language production. When a lexical analysis is performed, the
question is simply, what words are used, and how many times they are used
in this sample of language? Information gained from lexical analysis of
language, such as that by Thorndike and Lorge, has proved to be very useful
in predicting the ease with which different words can be learned in laboratory
situations.
2) At another level of linguistic analysis, the syntactic content of language text
may be investigated. In the study of syntax, interest is focused on the
arrangement or ordering of words to form phrases and sentences. The
question asked in this type of analysis is, how is this phase (or sentence)
structured? Psychologists and linguists interested in syntactic theory have
attempted to specify rules that account for the productivity of language
(Chomsky, 1985). The set of rules indicating how the elements of the
language may be combined to make intelligible sentences is referred to as a
grammar. Although a large number of different grammars have been
proposed, there is little agreement about the necessary features of an adequate
grammar.
3) Another level of analysis of language is the one that considers the semantic
content or meaning of passage. This perspective on language results in the
asking of questions such as the following: What does the passage
communicate? What is the meaning of this particular sentence?
Word meaning is a function of the interaction between word features and
the extent to which they match those belonging to certain prototypical and
nonprototypical contexts (Lakoff, 1987). Here, both feature theory and
prototype theory are seen as important.
The critical role of semantics is not under question and has been clearly
demonstrated in a number of psychological investigations. In general, current
views of semantics and comprehension view the listener (or reader) as an
active participant who formulates hypotheses about subsequent input based
on context (both verbal and situational), on knowledge of constraints in the
language, and on knowledge of the world. This is in contrast to the more
25
Language passive view of the comprehended as someone who waits for the input before
acting upon it.

2.3.3 Phase Structure of Sentences

In order to understand language in an adult, it is necessary to examine the structure
of sentences. At one level of analysis, a sentence can be regarded simply as a
string of phonemes. At another level, a sentence can be regarded as series of
morphemes, which are grouping of phonemes. From this viewpoint, however,
the sentence is viewed as a string of words. Linguists have found it more useful
to describe a sentence in terms of phrases, which are grouping of words.

Analysis of a sentence into its various phrases describes the phrase structure of
a sentence. A sentence is viewed as composed of two basic phrases, a noun
phrase and a verb phrase, which in turn are composed of subcomponents.

Figure 2.1 shows the phrase structure of a simple sentence, “The boy ate an
apple.” The noun phrase is composed of a determiner and a noun, and the verb
phase is composed of a verb and noun phrase; the latter noun phrase is also
composed of a determiner and a noun. Pause in speech usually reflect underlying
phrase structure. For example, we are most likely to say, “The boy…ate…an
apple,” pausing ever so briefly after boy and ate. We are not likely to say,
“The….boy ate…an apple,” or “The… boy ate an….apple,” grouping boy, ate,
and an. While in normal speech a speaker may search and grope for a particular
word and, thus, alter the pauses, the listener still tends to understand the message.

Sentence

Noun Phrase Verb Phrase

Determiner Verb Nour Phrase

Noun
Determiner Noun

boy ate
The an apple

Fig. 2.1: Phrase structure in a sentence represented by a tree diagram (adapted from Hunt
& Ellis, 2006)

2.3.4 Structure and Deep Structure in Sentences

The surface structure is the organisation that describes the sequences of phrases
in a sentence as it is actually spoken (or read) and reflects the phonological
realisation of the complex, underlying linguistic structure.
Deep structure, in contrast, refers to the underlying structure that includes the
26 relevant string of linguistic units, the grammatical requirements for lexical (word)
selection, and the grammatical relations between words in sentences. The deep Language Processing
(Comprehension and
structure of a sentence, thus, specifies the derivations of both its surface structure Language Expression)
and meaning.

Consider the sentences, “Rohit threw the ball” and “The ball was thrown by
Rohit.” Both sentences convey the same meaning despite the fact that they sound
different. Hence their deep structure is same. But consider the sentence “The
lamb is ready to eat,” which can have two meanings. The lamb may serve as
food to be eaten, or as an animal, the lamb is prepared to eat food. Thus, the deep
structure can vary within the same sentence, depending on what meaning the
speaker wishes to convey. Evaluate the meaning of the following ambiguous
sentences” “Visiting relatives can be a nuisance,” “the corrupt police can’t stop
drinking.”

Sentences with essentially a single deep structure and two or more surface
structures are synonymous. Sentences with different deep structures and the same
surface structure are ambiguous. Thus, important problem remaining concerns
the theoretical rules by which the deep structure of a sentence comes to be realised
in a particular surface structure. Rules for the specification of this linkage process,
called transformational rules, have been developed by Noam Chomsky
(Chomsky, 1965, 1975) and other linguists.

Transformational rules have clear implications about what features of sentences

human beings do store in memory. If the sentence is simple, then features of the
surface structure may be stored. As sentences become more complex, what is
thought to be stored is some underlying base structure, or schema, plus one or
more “footnotes” that serve as rules necessary to regenerate the sentence in its
original surface form. Thus, what is stored is some coded representation of the
complex sentence.

Information contained in a linguistic message tends to be comprehended, and

sometimes is remembered, in syntactically defined chunks, although semantically
based chunking also may be used, depending on the demands placed upon the
listener and the nature of the material (Marschark, 1979). Thus, the phrase
structure of a sentence appears to play an important organisational role in language
processing at a very basic level (Ferreira & Clifton, 1986).
Self Assessment Questions
1) Explain language in your words.
...............................................................................................................
...............................................................................................................
...............................................................................................................
...............................................................................................................
2) What are the three functions of language?
...............................................................................................................
...............................................................................................................
...............................................................................................................
...............................................................................................................
27
Language
3) Explain surface structure and deep structure.
...............................................................................................................
...............................................................................................................
...............................................................................................................
...............................................................................................................
...............................................................................................................

2.4 PROCESSES IN LANGUAGE

In this section we will examine some basic processes in language. The focus is
on three processes: (i) production of language, (ii) speech perception and
comprehension, and (iii) language development.

2.4.1 Production of Language

The beginning of a dialogue is usually the production of speech by one of the
participants, although a gesture or other sign may initiate such an interaction and
have its origins in a similar verbal plan (McNeil, 1985). But, before uttering a
sentence or manually expressing any information, the speaker must do some
planning based on the intended effect the utterance is to have on the listener;
based on the speaker’s knowledge of the listener’s scope of understanding
(example, is the listener familiar with the topic?); and based on the syntactic,
semantic, and pragmatic (or social) form that the production and its desired
effects requires. Thus, speaking is very much an instrumental act, which is to
say that speakers talk in order to produce an effect of some kind.
The process of speaking is basically concerned with planning and execution.
But just how is speech planned and executed? Clark & Clark (1977) described a
rough outline of this process, which involves five steps. The first step for speakers
is to decide on the kind of discourse to be initiated, which is the issue of discourse
plans. Do they want to engage in a conversation, to describe an event, to give
instructions, or to regale a friend with a humorous story? Each type of discourse
has a particular structure, and speakers must plan their utterances to fit that
structure.
For example, if you are telling a joke, you first describe the setting or context,
then describe the sequence of events, and end with the punch line. If you fail to
follow this structure, you obviously will not be an effective joke teller. If you
give away the joke by accidently telling the punch line before the appropriate
time, you will defeat your purpose. Similarly, instructions and conversations
have an orderly structure.
One set of guidelines that speakers and listeners seem to follow to foster good
communication during a conversation has been described by Grice (1975) and
others (e.g., Levelt, 1989). These “Gricean Maxims” are:
Quantity: Avoid running off at the mouth.
Quality: Don’t lie or stretch the truth.
Relation: Avoid making statements irrelevant to the topic of conversation.

28
Manner: Avoid vague or ambiguous statements.
Failure to follow these maxims often results in a conversational implicature. Language Processing
(Comprehension and
For example, imagine that you are reviewing an applicant’s letter of Language Expression)
recommendation for a highly technical job, and the letter reads as follows:

I am writing a letter on behalf of Gopal Bhatnagar. Gopal dresses very well and
has a charming wife. He also drives a nice automobile and sings in his church’s
choir. Thank you.

Would you hire Gopal based on this letter? Probably not. Clearly, the content of
this letter violates the Relation Maxim. Because of this, the letter writer has
conversationally implied that John is not the person for the job. Speakers (and
letter writers) usually adhere to these Gricean Maxims; but, as this example
demonstrates, it is quite informative when they don’t.

Planning discourse is planning at global level. The second stage of speech

production involves planning of sentences, the components of discourse. Once
the nature of discourse is decided, specific sentences that will accomplish the
objective must then be selected. The speech act, the propositional content, and
the thematic structure need to be determined. The order in which sentences are
produced and the type of information to be conveyed must be thought about. For
example, suppose you are describing your new house. You might first describe
the location; next, you might describe the overall type of house; then you might
proceed to describe the floor plan and arrangement of rooms and, finally, give
sfics of each room. Notice that there is a structure that involves going from
global, or general information, to progressively more specific details.

The third phase of speech production deals with constituent plans of the sentence.
Once a sentence is decided on, its components must then be planned. The
appropriate words, phrases, and so forth must be picked out and put in the right
order. These first three phases describe three levels of planning. At the most
general level, planning is directed towards the type of discourse. At the next
level, planning concerns the type of sentence to be uttered. At the third level,
planning deals with specific components of the sentence.

An interesting feature of slips of the tongue is that they point out regularities in
the planning stages of productions. For example, slips are seldom “illegal”
combinations of sounds for the language; morphemes tend to slip as entire units
(Clark & Clark, 1977). Some classics slips are known as “bloopers” in the world
of radio and television. Some bloopers are fairly obvious. For example, an
announcer for the ‘Friendly Homemaker Program’ said, “And now we present
our homely friendmaker. Another example is a remark of the commentator
covering visit of the king and queen of England: “When they arrive, you will
hear a 21 son salute.”

The fourth phase of speech production deals with what is called the articulatory
program. This concerns the plans for the execution of speech, which is a
coordinated sequence of muscular contractions in and about the mouth. And the
final phase of speech production is articulation itself. This is the actual output of
speech. Interested readers are referred to Clark & Clark (1977) and Levelt (1989)
for a detailed discussion of planning and execution of speech.

29
Language
2.4.2 Speech Perception and Comprehension
To understand speech is crucial to human communication. Hence, speech
perception is fundamental to language use in our day to day life. We are able to
perceive speech with amazing rapidity. On the one hand, we can perceive as
many as fifty phonemes per second in a language in which we are fluent (Foulke
& Sticht, 1969). On the other hand, we can perceive only about two thirds of a
single phonemes per second of nonspeech sounds (Warren et al., 1969). This is
why foreign languages are difficult to understand when we hear them. Even if
we can read them, the sounds of their letters and letter combinations may be
different from the sounds corresponding to the same letters and letter combinations
in our native language.

The comprehension of speech begins with the perception of raw speech sounds.
Comprehension starts where speech production ends. Speakers produce a stream
of sounds that arrive at the listener’s ears; then, listeners are able to analyse the
sound patterns and to comprehend them. Speech perception is not, however, the
simple identification of sounds.
It involves the complex processes of encoding and comprehension. In other words,
interpretative processes, meaning, contextual influences, and the like play
important roles in speech perception. Thus, the transformation from raw speech
sounds to propositions in memory is a complex process. The physical signal that
reaches the ear consists of rapid vibrations of air. While the sounds of speech
correlate with particular component frequencies, there is no direct one-to-one
correspondence between the sounds of speech and the perception of listeners.
Recognition of words is very much dependent on context, explanations, and
knowledge. For example, a hungry child can interpret the question “Have you
washed your hands for dinner?” as a call to come directly to dinner (i.e., as
indirect speech act rather than a direct question).
The role of context also can be easily seen in complete sentences in which context
allows words to be inferred quite easily. For example, the sentence “The young
girl was awakened by her frightening d….” allows listeners to infer dream. There
is no need to think about what the word might be; it just seems to pop out
automatically.
A similar context effect was studied in the laboratory of Warren (Warren &
Obusek, 1971) using phonemes. Subjects were read sentences that had a single
speech sound obscured. For example, the sentence “The state governors met
with the respective legislatures convening in the capital city” had the first s in
legislatures masked by a coughing sound. The experimenter then asked the
subjects to identify where the cough had occurred. [The results indicated that
subjects somehow “restored” the missing s sound and were unable to locate the
interjected cough.]
The phenomenon, appropriately called phonemic restoration, has been shown to
be even more likely when more than a single word can result from the restoration
(example: “_egion” can become either “legion” or “region”), indicating an active
word-searching process in speech perception (Samuel, 1987).
Many people have the impression that the words they hear are distinct, separate
combinations of sounds, but this impression in not correct. Cole (1979, 1980)
30
and other speech researchers have demonstrated that words usually run together Language Processing
(Comprehension and
as sound patterns. This is seen by use of a spectrograph, an electronic device for Language Expression)
measuring the variations in energy expended when a person talks. Moreover, it
is often the case that a single word cannot be recognised correctly when it is
taken out of its sentence context. This was shown some years ago by Pollack and
Pickett (1963), who played different segments of a normal conversation for
subjects. When the subjects heard just one word from the conversation, it was
often incomprehensible. Without the context of the meaningful sentence, the
single word could not be understood.

More generally, an important feature of speech perception is that speech is not

comprehended simply on the basis of the sounds per se. Rather, speech is
comprehended on the basis of many additional factors (e.g., intentions, context,
and expectations) from which an interpretation of what the speaker says is
constructed. (Tyler & Marslen-Wilson, 1986; see Paivio & Begg, 1981, for a
review).

One main approach equates processes of speech perception with processes of

auditory perception of other sounds. These kinds of theories emphasises either
template-matching or feature-detection processes. Such theories postulate that
there are distant stages of neural processing. In one stage speech sounds are
analysed into their components. In another stage these components are analysed
for patterns and matched to a prototype or template (Kuhl, 1991; Massaro, 1987).
One theory of this kind is the phonetic refinement theory (Pisoni et al.,1985),
which says that we start with an analysis of auditory sensations and shift to
higher level processing. A similar theoretical idea is embodied by the TRACE
model (McClleland & Elman, 1986). According to this model, speech perception
begins with three levels of feature detection: the level of acoustic features, the
level of phonemes, and the level of words. According to this theory, speech
perception is highly interactive. Lower levels affect higher levels and vice versa.

One attribute of these theories have in common is that they all require decision-
making process above and beyond feature detection or template matching. Thus,
the speech we perceive may differ from the speech sounds that actually reach
our ears. The reason is that cognitive and contextual factors influence our
perception of sense signal. For example, the phonemic-restoration effect involves
integrating what we know with what we hear when we perceive speech (Samuel,
1981; Warren, 1970).

2.5 LANGUAGE DEVELOPMENT

Language acquisition and development follows a fairly orderly course. (This
order has been discussed in detail in the earlier chapter on Language Acquisition).
Here we will focus on development of semantics which ultimately help the child
in comprehending the language.

Making speech sounds is only the first step in acquiring language. The sounds
must come to represent objects, symbols, and events in the child’s environment
and they must acquire meaning for the child.

Children are familiar with many aspects of their environment before they learn
to speak. Their parents, toys, pets, siblings, and household objects are familiar
31
Language stimuli. At this early stage of language development, their task is one of learning
to associate particular environment stimuli with particular sound symbols and
responses. For example, they must learn to associate the sight of mother with the
sound of Mama.

Only when such associations are acquired, the speech sound come to represent
or symbolise a specific object or event for the child. These associations are only
a part of language development. Thus, the development of meaning begins with
the acquisition of associations between objects/events and speech sounds.

One popular view of the acquisition of word meaning is that children learn
semantic features and then attempt to apply an original word that includes the
features to objects that share those features. For example, a child may learn the
word ball and then overgeneralise it to other round objects such as moon and
orange.

Gradually, the child begins to construct more complex sentences that take on the
characteristics of adult language. This is an enormously challenging task (Brown,
1973). What the child learns are sets of grammatical, semantic, and pragmatic
rules for constructing sentences. Usually, children are unable to verbalise the
rules, but their linguistic performance indicates that they do possess linguistic
competence, the knowledge necessary to produce all and only those situations of
a given language.

Indeed, many adults who speak grammatically acceptable English are unable to
specify the rules they use. But these rules allow us to generate the almost infinite
number of sentences. One of the best pieces of evidence for learning syntactic
rules is the phenomenon of overgeneralisation. For example, children learn to
say went correctly, apparently by rote, then learn the rule of forming the past
tense by adding ed, and then incorrectly as goed. They later learn the exception
to the rule and go back saying went. Similar overgeneralisations occur in deaf
child’s acquisition of sign language.

This brief description only begins to sketch some of the complexities of language
development. What is clear is that young children have an enormously complex
task in learning to speak, read, and use language in a meaningful fashion. The
fact that human beings can acquire and use language emerges as a remarkable
achievement.

Finally, relating language development to the earlier discussion on speech acts,

there have also been some interesting findings. For example, it appears that
younger children view the meaning of “I Promise” differently than do older
children and adults. According to philosopher Sourly (1969), certain conditions
must be present for a sincere promise to be made.

One condition is that the person making the commitment actually intends to
carry out the promised action. A second condition is that it is apparent that the
person to whom the promise is directed desires the action to be carried out. In a
recent study, Bernicot & Laval (1996) report that 3-year-olds have difficulty
understanding only the second condition. But, by age 10, children evaluate both
conditions equivalently well in determining the outcome of a scenario (concerning
the occurrence of the promised activity), where these conditions were
manipulated.
32
These findings indicate that the meaning of “I Promise” is quite different, Language Processing
(Comprehension and
depending on the age of the child to whom it is said. Specifically the understanding Language Expression)
of the contextual circumstances underlying the making of a verbal commitment
increments with age. Initially, children are primarily concerned with whether the
promised activity simply occurred; as they grow older, they begin to grasp the
intentions of the speaker in evaluating the likely outcome of that activity
(Astington, 1988).

2.6 COMPREHENSIVE MODEL OF LANGUAGE

PROCESSING
This chapter has progressed systematically from the simple linguistic entities
(phonemes and morphemes), to syntax and grammar, to speech perception and
comprehension. One might wonder, whether there are any comprehensive theories
of language. In fact there are many. One by Kintsch is particularly significant
because it incorporates many bits of wisdom from earlier studies and, at the
same time contains a model of the mind. Let’s discuss now the principal
components of the most influential, extensive and comprehensive model of
language processing by Kintsch and van Dijk (Kintsch, 1974, 1979, 1988, 1990;
Kintsch & van Dijk, 1978) briefly.

2.6.1 Kintsch’s Model of Comprehension

This model of comprehension is more than a system that deals with the way
textual information is understood. It is a theory that cuts across many topics in
cognitive psychology, including memory and comprehension of the written and
spoken language. Comprehension is dependent on two disparate sources that are
similar to top-down and bottom-up processing. [Borrowed from computer
language, bottom-up processing is cognitive processing initiated by the
components of a stimulus pattern which, when summed, lead to recognition of
the whole configuration; whereas top-down processing is hypothesis-driven
recognition of the whole stimulus configuration, which leads to the recognition
of component parts.] At the highest level is the goal schema, which decides what
material is relevant. At the opposite extreme of the model is the text.

The model is based on a proposition. A proposition is an abstraction, and, as

such, it is difficult to define concretely. However, some characteristics of
propositions can be identified: they are abstractions based on observations (such
as reading text material or listening to a speaker); they are retained in memory
and follow the laws governing memory processes; and, in Kintsch’s system,
they consist of a predicate and one or more arguments. Predicates corresponds to
verbs, adjectives, adverbs, or connectives in the words a person reads or hears.
This is called the surface structure, a term already discussed in previous sections.
Arguments correspond to nouns, noun phrases, or clauses. The model is illustrated
with the following little story:

The Swazi tribe was at war with a neighbouring tribe because of a dispute over
some cattle. Among the warriors were two unmarried men, Kakra and his younger
brother Gum. Kakra was killed in battle.

The first sentence is divided into five groups: the swazi tribe…..was at war
with….a neighbouring tribe…. because….a dispute over some cattle. According
33
Language to coherence analysis of this sentence, only first three of the factors are in working
memory. The predicate “was at war with” is considered the most important part
of this sentence insofar as comprehension of the story is concerned. The other
parts are clustered around it.

A significant feature of the model is that the initial processing of the text is
assumed to take place in Short Term Memory (STM), which we know has limited
capacity. Because of this constraint, only a portion of the propositions is held in
memory. With the reading of the second sentence, some of propositions from the
first sentence are still vital in STM. The reader tries to connect the old and new
propositions but finds no match between them.

Failing to find a match between the propositions in STM, the reader searches
Long Term Memory (LTM) for a possible match. This search in the LTM is
called reinstatement search and is one reason that text material may be hard to
read. In the example, the lack of a match between propositions in the first and
second sentence requires the reader to construct a new network for the ideas and
to attempt to relate the two sentences.

One inference that the reader makes is that the two men were members of the
Swazi tribe, a reasonable conclusion even though that the fact is not stated directly.
With the reading of more sentences, the semantic network begins to get more
complicated and interrelated. The reading of the sentence “Among the warriors
were…., Kakra and…….” Retains in memory the names of the men, which can
easily be related to the information in the last sentence “Kakra was killed in
Battle.”

2.6.2 Propositional Representation of Text and Reading

As stated before the model of comprehension holds that the underlying unit of
memory for text material is the proposition. Additionally, the model predicts
that sentences of greater propositional complexity are more difficult to
comprehend than sentences with simple propositional structure, even if the surface
complexity of the two sentences is about the same. Kintsch and Keenan (1973)
designed an experiment to test this prediction. A sample from many experiments
is reported here.

Subjects were asked to read ten sentences, all of which had about the same number
of words but varied greatly in the number of propositions. Some sentences had
as few as four propositions, and others had as many as nine. For example, read
the following two sentences:
Romulus, the legendary founder of Rome, took the women of Sabine by force.
Cleopatra’s downfall lay in her foolish trust in the fickle political figures of the
Roman world.
Which sentence was more difficult to read? If you are like the subjects in Kintsch’s
and Keenan’s experiment, you had more difficulty with the sentence about
Cleopatra than the sentence about Romulus. Even though the surface complexity
of the two sentences is about the same, they differ markedly in the number of
propositions and the macrostructures that are required to interconnect the
propositions.

34
In the Kintsch’s and Keenan’s experiment, subjects were presented with sentences Language Processing
(Comprehension and
similar to those just discussed above by means of slides. The subjects were asked Language Expression)
to read each sentence and then to write it. They could then advance the slides and
see the next sentence. Time taken in reading each sentence was noted. The authors
found an extraordinarily consistent relationship between the number of
propositions and the time required to read the sentences.

2.7 LET US SUM UP

In all, the approaches that have been taken in exploring many of the issues involved
in the perceptual analysis of language and lexical processing merely scratch the
surface of the complexity of both theory and fact that must be developed to
provide a sufficient characterisation of the cognitive system. Language processing
requires a multidisciplinary examination. To conclude, we can say that, like many
other cognitive process, language processing is a very dynamic and complex
process. No single method, function or theory can explain the process of language
completely in itself; only a comprehensive approach should be appropriate and
applied for the comprehension of language.

2.8 UNIT END QUESTIONS

1) Note the various experimental tasks that have been used to study language
comprehension. Have you run into any of them before in this course?
2) It is intuitively obvious that context facilitates word interpretation, but how
can it interfere with interpretation?
3) What is the role of context and expectations in the interpretation of speech?
How has the influence of context been studied experimentally?
4) What are several major features of language development?
5) Compare and contrast the role of speech perception, syntax and semantics
in the development and understanding of language.
6) What are the different processes involved in language comprehension?
7) The exposition of Kintsch’s model is necessarily abstract and therefore
difficult to comprehend. Preserve in your reinstated searches! See if you
can use it to deal with a new example of text selected from another course.
8) What factors are included in Kintsch’s model? How does the reader enter
into this model?
9) Give an example of a humorous violation of one of Grice’s four maxims of
successful conversation.

2.9 SUGGESTED READINGS AND REFERENCES

Clark, H. H. (1996). Using Language. Cambridge: Cambridge University Press

Clark, H., & Clark, E. (1977). Psychology and Language: An Introduction to

Psycholinguistics. New York: Harcourt Brace Jovanovich.

Garrod, S. & Pickering, M.J. (2001). Language Processing. East Sussex:

Psychology Press.
35
Language References
Gleitman, L. R., Liberman, M., & Osherson, D. N. (Eds.)(2000). An Invitation
to Cognitive Science, 2nd Ed. Volume 1: Language. Cambridge, MA: MIT Press.

Hunt, R. R., & Ellis, H.C. (2006). Fundamentals of Cognitive Psychology. New
Delhi: Tata McGraw Hill.

Jurafsky, D., & Martin, J.H. (2009). Speech and Language Processing: An
Introduction to Natural Language Processing, Computational Linguistics, and
Speech Recognition. NJ: Prentice Hall.

Just, M.A., & Carpenter, P.A. (1987). The Psychology of Reading and Language
Comprehension. Boston: Allyn and Bacon.

Reed, S.K. (2010). Cognition: Theories and Applications. London: Cengage.

Solso, R.L. (2006). Cognitive Psychology. New Delhi: Pearson Education.

Sternberg, R.J. (2009). Applied Cognitive Psychology: Perceiving, Learning,

and Remembering. London: Cengage.

36