CHAPTER 5.

Varieties of Discourse
(D. Crystal. Cambridge Encyclopedia of the English Language.
2019. Pp. 300-xii)
Prereading task.
What is discourse?
What do we refer to as discourse?
What is pragmatics?
How do you understand ‘Mixed Medium’?
What is dialogue and monologue?
How do you imagine New Textual World?

Read the text.

1. Language in Use and Paradigms of Enquiry
There is a major qualitative difference between studying the components of
English structure and studying the domains of English use. The structural
properties of the language are many and complex, but at least they are finite and
fairly easy to identify: there are only so many sounds, letters, and grammatical
constructions, and although there is a huge vocabulary, at least the units are
determinate and manageable. None of this applies when we begin to investigate the
way English is used: we are faced immediately with a bewildering array of
situations, in which the features of spoken or written language appear in an
apparently unlimited number of combinations and variations. Sometimes the result
is a use of English, or of a feature of English, which is highly distinctive and easily
explained, as is often encountered in regional dialects and in some of the more
institutionalized areas of language use, such as religion and law. Rather more
often, we are faced with usage that is subtle and indeterminate, and which demands
detailed and lengthy analysis before we can reach an understanding of what its
purpose is and how it works, as is often found in social dialects and in some of the
more creative areas of language use, such as humour and literature.
Recent years have seen considerable progress in the study of language in
use, and the emergence of several paradigms of enquiry, as people probe the topic
from different points of view. Some linguists favour a ‘bottom up’ approach,
studying the way sentences combine into larger units of discourse, and focusing on
the role played by specific features of language in facilitating successful
interaction. In this approach, whole books might be written on the communicative
role of a tiny aspect of language (such as the use of you know in conversations).
Other linguists work ‘top down’, beginning with a broadly defined category – such
as an area of knowledge (science, politics), a social situation (gender, class), or a
communicative genre (poetry, joke) – and examining the range of linguistic
features which are found within it. Every conceivable kind of academic enquiry
can be found, such as heavily illustrated descriptions of data samples, meticulous
statistical or experimental analyses of individuals and groups, ambitious
taxonomies, and highly abstract theoretical outlines.
The various branches of linguistics that investigate the topic, such as
sociolinguistics, stylistics, discourse analysis, pragmatics, and textlinguistics,
present a remarkable range of methodologies and emphases.

2. Microlinguistic Studies
There is only one way to establish the exact function of the various elements
which contribute to the organization of discourse, and that is to subject a
substantial amount of linguistic data to a microlinguistic analysis. In the case of
spoken discourse, a recording of reasonable acoustic quality needs to be made,
then transcribed with maximum attention to detail, paying particular attention to its
pauses, interruptions, false starts, hesitations, and other such features. Ideally, a
full prosodic transcription should be included, though the level of specialized
training required to hear prosodic effects accurately means that this is not always a
practical option. Each instance of a particular item of interest is noted – the word
well, the hesitation noise er, the clause you see – and its context examined to
establish what role it may be playing at that point in the discourse. An immediate
intuitive response to the item can be sharpened by manipulating the data in various
ways, such as omitting the item to see how this affects the meaning or acceptability
of the utterance, or contrasting it with another item. By comparing a large number
of instances, the aim is to arrive at an informative classification of uses, and to
develop a theory of the organization of discourse which can then be tested against
other kinds of utterance.
Discourse direction. Some items play a particular role in controlling the
direction of movement within a discourse. They signal such broad organizational
features as topic identification, change, and exemplification, and such logical
relationships as topic contrast and reinforcement. It is never possible to present a
truly simple account, as several items have ‘fuzzy’ meaning, and analytical
categories (such as evaluating and summarizing) are not always easy to apply
consistently. Nonetheless, several studies have provided useful first
approximations.
One such approach focuses on lexical phrases, recognizing eight types of
what it calls macro-organizers. These are seen to operate at two levels: global
features determine the overall shape of the discourse; local features mark changes
of direction operating in a more restricted way.
Global macro-organizers.
Topic markers: let’s look at X; what do you think of X?; have you heard
about X?; let me start with X.
Topic shifters: by the way; let’s move on to Y; that reminds me of Y; this is
off the subject, but Y.
Summarizers: in a nutshell; that’s about it; in effect; to cut a long story short;
what I’m trying to say is Z.
Local macro-organizers.
Exemplifiers: in other words; for instance; to give an example; it’s like A.
 Relators: nonetheless; however; and also; it has to do with B; it’s the same
with B.
Evaluators: I think/don’t think that C; as far as I can see; seems to me; I’m
absolutely certain.
Qualifiers: the catch is; it depends on D; that’s true but D; this doesn’t mean
that D Asides: where was I?; I guess that’s beside the point; I’m getting ahead of
myself here.
The following extract from a meeting between a teacher (T) and a student
(S) to talk about a thesis shows the use of several of these organizational features.
This exchange is typical of a discourse where serious business is being transacted,
and where care needs to be taken (notwithstanding the informal style) with the
manipulation of meaning. There are far fewer such features in the more loosely
structured language of everyday conversation. (After J. R. Nattinger & J. S.
DeCarrico, 1992.)
 Please. The most interesting property about please (apart from its ‘magical’
social role in persuading others to cooperate) is its discourse function. From a
structural point of view, please is unique.
• It is not easily assigned to any word class: grammars tend to call it an
adverb, but it is like no other adverb. It cannot, for example, be modified by very:
we can say very kindly, but not *very please.
 • It can act as a minor sentence in its own right: A: Would you like some tea?
B: Please.
• When functioning within a sentence, it is not constrained by the syntactic
type. It may occur with statements (I’d like some pudding, please), questions (May
I have some pudding, please?), commands (Give me some pudding, please), and
moodless clauses (Pudding, please).
• It has no easily stateable dictionary meaning. If we were to try to define the
‘meaning’ of please at all, it would have to be in terms of what it does – to
persuade someone to do something.
Are there any sentences, then, where we may not say please? That there are
many such constraints can be seen from the following:
• He ate more pudding, please.
• I promise you can have some more pudding, please.
• Would you like some more pudding, please?
• Do you want to come to a party, please?
• Give me more pudding or I’ll hit you, please.
• I think you’re beautiful, please.
These sentences are, respectively, a narrative statement, a promise, an offer,
an invitation, a threat, and a compliment. Please cannot be used with such
sentences, but only with those which are interpretable as a request. The point is not
simply a matter of common sense, as can be seen from the errors made by non-
native learners of English, who often produce sentences similar to the unacceptable
ones above. In short, please is an item whose function is entirely defined by its role
in discourse – and moreover, in discourse of a very particular kind (requesting).
More than any other word in English, it is a discourse-identifying feature. (After
M. Stubbs, 1983, Ch. 4.)

3. Pragmatics
Pragmatics is the study of the choices we make when we use language, the
reasons for those choices, and the effects that those choices convey. When linguists
started to investigate pragmatics, they approached it from different angles. Some
focused on the choices themselves; some on the intentions and beliefs behind the
choices; some on the results of the choices. Special attention was paid to the way
context contributes to meaning, to the way utterances implement actions
(performatives), and to the assumptions about the world (the presuppositions) that
people make when they use and interpret utterances. Textbook introductions to
pragmatics can look very different, as a result, but they all have one aim: to explain
the choices that are made.
The name pragmatics is no more than a specialised application of the
everyday use of the word. The OED sums it up in its opening definition: ‘Dealing
with matters in accordance with practical rather than theoretical considerations or
general principles’. If we say that Mary is pragmatic, we mean she aims at what is
achievable rather than ideal. She is matter-of-fact, practical, and down-to-earth,
choosing to behave in one way in situation X and perhaps in a different way in
situation Y. She is not (to give its opposite) dogmatic.
The concept of choice is critical, and this is what fosters a new perspective
when phonology, graphology, grammar, vocabulary, and discourse are studied
from a pragmatic point of view. Why do we choose one word rather than another,
or a particular tone of voice, punctuation, spelling, grammatical construction, or
pattern of discourse? The subject is of immediate interest to teachers, because it
reflects the kind of decision-making their pupils are involved in every day: a
correction is basically the offering of an alternative choice. Indeed, the entire
notion of ‘knowledge about language’ can be thought of as making children aware
of the choices that are available to them when they speak and write, and
understanding the reasons for the choices others have made when they listen and
read.

3.1. Some Pragmatic Choices

Punctuation. Our choice of punctuation is going to be chiefly guided by
semantic and pragmatic considerations. Normally, it will be semantics: we will aim
to make our meaning clear to the reader. But there are occasions when pragmatic
factors take precedence: we can decide to use or not use a mark because it looks
beautiful/ugly, because it is easier/more difficult to write/type/ text, because it is
available/unavailable in a chosen font, because it takes up more/less space on a
page, or simply because we were taught that way (without necessarily knowing
why). In particular, the ‘look’ of the page can become a priority in guiding our
choices of which punctuation marks to use. This turns out to be a major factor in
literary writing and in publishing. Publishing houses and style-guides opt for a
solution, partly in the interests of aesthetics and consistency, partly to present a
clear identity, and they expect their users to follow it.
Punctuation has always been a matter of trends in fashion. No one has ever
been able to define a set of rules which explain all uses of all marks. Practice varies
so much between formal and informal writing, Britain and America, page and
screen, men and women, publishers, authors, and generations. The best we can do
is identify norms, plot trends, emphasize the need for personal consistency, and be
very cautious indeed about making generalizations. Hardly any rules of
punctuation are followed by all of the people all of the time. A pragmatic approach
respects this reality, and tries to come to terms with it. Being pragmatic doesn’t
mean ‘anything goes’ – ignoring the rules that do exist. It means respecting all
linguistic realities, whether rule-based or not.
Grammar.
Which pronoun should we use when writing? The pragmatic choices
include: – The use of I, to convey a personal involvement: I pointed out earlier...
– The use of one, to convey impersonal involvement: One pointed out
earlier...
– The use of we, to avoid an egotistical impression: We pointed out earlier...
– The use of we, to suggest that writer and reader are involved in a joint enterprise
(the ‘authorial we’):
As we saw earlier...
 – The use of you, which introduces a distance between writer and reader,
and sometimes a hint of abruptness:
As you saw earlier...
The use of you, meaning ‘anyone’, as an informal alternative to one:
You can see the beach...
– Avoiding all of these options, by using a passive:
It was pointed out earlier...
Once a decision is made, other pragmatic factors come into play, such as the
need to be consistent. Most people would find sequences such as the following
unpalatable:
We started the climb at six o’clock. You could see the summit clearly, and
we thought it wouldn’t take more than an hour to reach it. But one forgot about the
mist, and I was surprised to find we were still a long way off at seven/
As other styles are explored, other techniques come to light, such as the
replacement of a pronoun by some other construction. When a noun phrase
replaces I or you, for instance, the effect is one of self-conscious formality:
– This reviewer has to say he has never seen such a terrible production.
– Readers of this column will not be surprised to learn...
We need a pragmatic perspective to understand how pronouns are used in
developing a pleasing personal or professional style

Conversational Maxims
The philosopher H. P. Grice is well known in pragmatics research for his
four maxims of conversation that underlie the efficient cooperative use of
language.
Maxim of quality. Try to make your contribution one that is true,
specifically: Do not say what you believe to be false. Do not say that for which you
lack adequate evidence.
Maxim of relevance. Make your contributions relevant.
 Maxim of quantity. Make your contribution as informative as is required for
the current purposes of the exchange. Do not make your contribution more
informative than is required.
Maxim of manner. Be perspicuous, and specifically: Avoid obscurity. Avoid
ambiguity. Be brief. Be orderly.
The point of an analysis of this kind is not to suggest that we always behave
exactly according to the principles; common experience shows that we do not. But
we do seem to tacitly recognize their role as a perspective or orientation within
which actual utterances can be judged (though the Internet presents some
interesting exceptions). For example, people who tell lies or make false claims can
be challenged; if they talk too much they can be told (in so many words) to shut
up; if they say something irrelevant, they can be asked to stick to the point; and if
they fail to make themselves clear, they can be requested to say it again. The fact
that we do all of these things indicates that we are bearing these maxims in mind.

4. Conversational Maxims by H. P. Grice

The philosopher H. P. Grice is well known in pragmatics research for his
four maxims of conversation that underlie the efficient cooperative use of
language.
Maxim of quality
Try to make your contribution one that is true, specifically:
Do not say what you believe to be false.
Do not say that for which you lack adequate evidence.
Maxim of relevance
Make your contributions relevant.
Maxim of quantity
Make your contribution as informative as is required for the current purposes
of the exchange.
Do not make your contribution more informative than is required.
Maxim of manner
 Be perspicuous, and specifically:
Avoid obscurity.
Avoid ambiguity.
Be brief.
Be orderly.
The point of an analysis of this kind is not to suggest that we always behave
exactly according to the principles; common experience shows that we do not. But
we do seem to tacitly recognize their role as a perspective or orientation within
which actual utterances can be judged (though the Internet presents some
interesting exceptions). For example, people who tell lies or make false claims can
be challenged; if they talk too much they can be told (in so many words) to shut
up; if they say something irrelevant, they can be asked to stick to the point; and if
they fail to make themselves clear, they can be requested to say it again. The fact
that we do all of these things indicates that we are bearing these maxims in mind.

5. Cognitive Approaches
Cognitive linguistics explores the relationship between language and mind,
focusing on the meaning expressed by the levels of linguistic structure (phonology,
graphology, grammar, vocabulary) and investigating how this affects the ways in
which we think about the world, experience it, and socially interact. The study of
the way metaphors condition our thinking forms an important element in this fresh
approach to the study of language.
The Metaphorical Mind. There is an ancient tradition, dating from classical
times, that sees metaphor as a literary or rhetorical phenomenon, and this is how
most people first encounter the topic in school when they are introduced to ‘figures
of speech’. An effective metaphor describes one thing in terms of another, and in a
literary context, especially in poetry, it is typically something unusual,
incongruous, dramatic, striking, surprising, shocking... conveying fresh insight.
But there is another, deeper notion of metaphor that is so fundamental to our
everyday way of talking and thinking that we hardly notice its figurative status at
all, and it is this notion that has become a central topic within cognitive approaches
to language, especially following the publication of George Lakoff and Mark
Johnson’s Metaphors We Live By (1980).
In this approach, metaphors are seen not simply as linguistic constructs –
creative lexical collocations – but as expressions that structure our perception and
understanding of the world, and influence the way we act and interact within it.
They are mechanisms of thought and action, shaping our view of life and our
behaviour. If we conceive of an argument as a war – one of Lakoff and Johnson’s
most discussed examples – we will describe it using such terms as attack, weak
point, defend, on target, shoot down, gain ground, defeat, lose, surrender, win, and
the interaction is likely to be much more confrontational than if we conceive of it
as a quest or journey (explore, examine, find out, step-by-step, a stage further...),
as a building (foundation, construct, shaky, knock holes...), or as a cooperative
endeavour (exchange, contribution, consensus, complement...).
‘Argument as war’ metaphors have been in the language a long time, so they
resist change; but some cognitive theorists have begun to explore alternatives.
What might be a less violent way of depicting an argument? We might explore the
potential of argument as an art form, for example.
Instead of ‘shooting down’ an argument, we might ‘erase’ it.
Instead of saying that an argument was someone’s ‘best weapon’, we might
say it was their ‘best canvas’.
Instead of ‘shooting holes’ in an argument, we might ‘paint over its errors’.
Hobbies could be a fruitful source, such as knitting. ‘Shoot down’? Unravel.
‘Shoot holes’? Drop stitches.
These analogies make the same point, but in a non-confrontational way,
thereby giving encouragement to people who think of arguments as unpleasant
situations, and offering opportunities to those who look for collaborative outcomes.
There are innumerable possibilities. Instead of an argument being ‘strong’ or
‘weak’, it might be ‘pretty’ or ‘ugly’, ‘large’ or ‘small’, ‘sweet’ or ‘bitter’, and so
on. In each case, a different mental model is being suggested, and a different set of
behavioural outcomes.
From Metaphors to Mindsets. Metaphors are not exact parallels.
Arguments are not actually wars. There are many words from warfare that are not
usually used in relation to arguments (campaign, crossfire, besiege, front line,
prisoner-of-war, blockade...), and several war words are also found in non-military
settings, such as chess or football (defend, position, strategy...). But the metaphor
is a powerful one, used in many everyday contexts, such as politics, sport, and
disease (as in the battle against cancer), and influencing the way we perceive
situations. A newspaper that reports a political demonstration as a ‘battle’ is seeing
it in a very different way from one that reports it as a ‘game’, and the consequences
are much more likely to form a climate that motivates violent intervention.
The arrival of immigrants in a country can be reported in very different ways
in the media. A story that reports them as ‘invaders’, or as ‘swarming’ or ‘flooding
in’, reflects a negative mindset that is likely to generate in readers an increased
antagonism towards refugees and asylum seekers; one that describes them in
positive terms, such as ‘new blood’, ‘breath of fresh air’, or ‘welcome channels of
irrigation’ is likely to promote a more sympathetic reaction. There is a clear link
with the ‘snarling’ and ‘purring’ words of some earlier semanticists. In such cases,
metaphors do much more than express underlying cognitive operations; they
reflect and reinforce ideologies. This is the focus of the developing field of critical
metaphor analysis, where scholars interpret the metaphors people use to uncover
the ideologies that may lie behind them.

6. Testing Cognitive Effects

The cognitive approach to metaphor is based on certain assumptions about
lexical usage. If ARGUMENT is WAR is a valid equation, it will be reflected in
the way people actually use warfare terms in this context – not just in the media,
but in the language as a whole. This is not something that individuals can easily do
intuitively, but it is a task that can be facilitated using large corpora containing
diverse textual categories, where warfare terms can be scrutinized to see whether
their collocations do indeed occur with high frequency in relation to arguments.
Corpus-counts are only a first empirical step, though, for frequency does not
account for everything that causes a term to stand out in our minds – to be
psychologically salient. An alternative, experimental paradigm of enquiry is
needed to test the effects of choosing one metaphorical frame rather than another.
One such experiment examined the way the London riots of 2011 were
reported in the media, focusing on the use of the metaphor of FIRE. Fire metaphors
are often found in discourses where the underlying issues are to do with power,
authority, and legitimacy. In the headlines and reports of the riots, repeated use
was made of such expressions as fan the flames, igniting, raging, inflaming, and
engulfing:
A riot that engulfed north London was sparked when a teenage girl threw a
rock at police, it was claimed last night (Daily Star 08.08.11)
Rioting has spread across London on a third night of violence, with unrest
flaring in other English cities (BBC News 09.08.11)
Civil disorder is being metaphorically construed as a damaging fire.
Metaphors are not exact parallels; they draw attention to certain similarities
between situations and hide certain differences. In this context, the fire metaphor
highlights a single cause (sparking) and shows the protests as a destructive force,
thereby taking our attention away from the underlying conditions that led to the
event, and the human motives that caused individuals to participate. And there is a
further implication: if we control fire with water, why not people with water
cannon?
Is this last interpretation valid? Is it the case that the presence of fire
metaphors in describing protests would facilitate support for police use of water
cannon? One study tested this hypothesis by presenting people with two versions
of an online news report about rioting in a fictitious city. One version used a fire
metaphor, the other did not:
 PROTESTS OVERWHELM CITY Protests have overwhelmed the city of
Sudfield. The protests started in the morning but quickly extended throughout the
city where they continued to occur all day. Eventually the police used water
cannon to disperse the protesters.
PROTESTS ENGULF CITY Protests have engulfed the city of Sudfield. The
protests ignited in the morning but quickly spread throughout the city where they
continued to rage all day. Eventually the police used water cannon to disperse the
protesters.
The influence of accompanying photographs was included in the
experiment, with one picture showing a burning car and another showing a vehicle
being vandalized. Participants experienced one of three conditions:
• They saw a version in which there was no fire in the picture and no fire
metaphors in the text.
• They saw a version in which there was fire in the picture and no fire
metaphors in the text.
• They saw a version in which there was fire in the picture and fire
metaphors in the text.
Participants read the text once, and were then asked two questions:
• How logical was the decision to use water cannon at the protests in
Sudfield?
• How justifiable was the decision to use water cannon at the protests in
Sudfield?
The results were striking. Participants presented with the picture containing
fire were far more likely to consider police use of water cannon as legitimate,
regardless of the metaphor. When they saw a non-fire picture, the potential effect
of the metaphor was overridden, and they were no more likely to support the use of
water canon. But when they were shown the two versions of the text without any
accompanying pictures, the fire metaphors produced a similar effect to the picture
of fire. People evidently allow their reactions to be influenced first and foremost by
the pictures they see. But if there are no pictures, then the metaphorical properties
of the text do strongly operate.
The results of this experiment applied regardless of the political orientation
of the participants, though there were hints that conservatives were least
susceptible to textual influence and liberals most susceptible. The researcher
concluded:
This suggests that conservatives and independents are more fixed in their
attitudes and opinions towards political protests and policing, subscribing to a
discourse of deviance, while liberals formulate their views on a more context
dependent basis taking into account local information supplied by texts.
Case studies of this kind suggest that media representations in both language
and image can influence public opinion on matters of policing. And they clearly
show the importance of adopting a multimodal semiotic perspective – one in which
the pictorial medium is viewed as operating along with the medium of written text
in eliciting a cognitive response. When it comes to matters of cognitive
interpretation, evidently, the analysis of linguistic levels is not enough. (After C.
Hart, 2018.)
Metaphorical frames are conceptual structures representing particular areas
of knowledge and experience. One analysis distinguishes core elements, essential
to the meaning of a frame, and non-core elements, which are not unique to a frame
but may be a salient feature of it. The convention has developed to identify frames
and elements in small capitals.
The fire frame
Entails frames: NATURAL PHENOMENON, HEAT, LIGHT
Elaborated in frames (chiefly for UK citizens): Great Fire of London
Metaphorical frame relations: ANGER, DESIRE, CONFLICT
Core frame elements (with examples): CAUSE (spark), INITIATION
(ignite, light, break out, erupt), ENDURANCE (burn, rage), SUBSTANCE
(engulf, consume), TRANSMISSION (spread), SUSTENANCE (fuel, fan the
flames, inflame), CESSATION (extinguish, put out, burn out)
 Non-core frame elements: TIME, PLACE, MANNER (intensely, violently,
rapidly), INSTRUMENT/MEANS (petrol bomb, fire-hose), MATERIALS (wood)
Associated scripts/scenarios: EMERGENCY, DANGER, WARMTH,
SAFETY (From C. Hart, 2018.).

7. Texts and Varieties

The notion of ‘using English’ involves much more than using our knowledge
of linguistic structure to create and interpret sequences of sentences and
conversational interactions. It also involves being aware of the range of situations
in which English can be used in a distinctive and predictable way, and of the
possibilities available to us when we wish to produce or respond to creative uses of
the language.
These situations are enormously varied, and not always easy to define, but
we can begin to make sense of them by looking at the communicative products, or
texts, with which they are associated, and at the linguistic features which define
these texts’ identity. Prayers, posters, roadsigns, lectures, sports commentaries,
novels, speeches, interviews, and recipes are all texts, by this account. They each
have a particular communicative purpose – easier to state in relation to a road-sign
than a novel, perhaps, but a communicative purpose nonetheless. They are also
relatively self-contained units of discourse, whether spoken or written, and each to
some degree has a definable linguistic identity.
Qualifications such as ‘relatively’ and ‘to some degree’ are important,
because not all texts have boundaries which are equally easy to identify, or use
language which is equally distinctive: for example, the boundaries of written texts
are typically more determinate than spoken ones, and (within speech) most
sermons have a much clearer beginning and end than most conversations. But
when investigating uses of English, it is usually practicable to work with a
hierarchy of the following notions: situations give rise to texts, and texts make use
of sets of distinctive linguistic features. A particular set of these features,
representing a category of text, is known as a variety. To take just one example: O
living God is a distinctive feature (a vocative with O) of a prayer text which is
found in a religious situation: it is therefore a feature of the variety of religious
language.
Sociolinguistic and Stylistic Features. The features which identify a
variety are not features of the language as a whole, occurring anywhere the
language is spoken or written, in all possible social situations. Variety features
depend on the presence of certain factors in the social situation. Classifications of
these factors vary, but it is possible to group them into two general types, which
give rise to what are here called sociolinguistic and stylistic features.

Sociolinguistic features relate to very broad situational constraints on

language use, and chiefly identify the regional and social varieties of the language
(e.g. Canadian, Cockney, upper-class, educated). They are relatively permanent,
background features of the spoken or written language, over which we have
relatively little conscious control. We tend not to change our regional or class way
of speaking as we go about our daily business, and usually do not even realize that
it is there.
Stylistic features relate to constraints on language use that are much more
narrowly constrained, and identify personal preferences in usage (poetry, humour)
or the varieties associated with occupational groups (lecturers, lawyers,
journalists). They are relatively temporary features of our spoken or written
language, over which we do have some degree of conscious control. We often
adopt different group uses of language as we go through our day (e.g. family, job,
religion, sports), and frequently change our speaking or writing style to make a
particular effect (as when we put on an accent while telling a story, or play with
language in an informal letter).

8. Speech and Writing

Spoken and written language display a number of important differences,
over and above the obvious distinction in physical form – that speech uses the
medium of ‘phonic substance’, typically air-pressure movements produced by the
vocal organs, whereas writing uses the medium of ‘graphic substance’, typically
marks on a surface made by a hand using an implement. These differences are
chiefly to do with language use, arising out of the fact that speakers and writers are
operating in fundamentally different communicative situations. But there are also
several differences in language structure: the grammar and vocabulary of speech is
by no means the same as that of writing, nor do the contrasts available in
phonology correspond to those available in graphology.
Writing is sometimes thought to be little more than ‘speech written down’.
Speech, correspondingly, is often judged by its closeness to writing. Neither
position is valid. The two mediums, though historically related, function as
independent methods of communication. There are few circumstances where we
are faced with a genuine choice between speaking or writing.
Normally, whenever two people are in earshot, they speak to each other.
Only very special circumstances – wicked children passing secret messages in
class; partners who are ‘not talking’ to each other; a jury foreman passing a verdict
to a court official; someone who cannot speak or hear (and who is unable to use
sign language) – would motivate the enormous trouble of writing down what we
wish to ‘say’. Conversely, people who are separated by distance in space or time,
and who lack electronic means of communication (or the money to use them), have
no alternative but to write to each other.
Moreover, the status of the two mediums is not the same. Written
formulations, such as contracts, are usually required to make agreements legally
binding. Historical documents, ancient inscriptions, original manuscripts, first
editions, sacred writings, and other such material are given a kind of respect which
is rarely accorded to speech (though archives of recorded sound are beginning to
introduce a balance). Above all, written English provides the standard that society
values, and its relative permanence and worldwide circulation have given it a very
special place within the life of the community.
Two Electronic Exceptions to the Rule. Speech is normally interactive –
but not when talking to a telephone answering machine, where we have to produce
a monologue while pretending it is a dialogue. This is not something which comes
easily to most people – though abilities improve with practice. Writing is not an
interactive medium in the same way as speech, because of the delay in getting the
written message to the reader; and in many kinds of writing there is little
expectation of a reply (none at all, pace the other sense of ‘medium’, when the
writer is dead). But the advent of electronic mail and online forums has altered the
time parameters dramatically. Questions and answers fly around the world now
which are very similar to those that would be used if the participants were talking
to each other.

8.1. Differences Between Speech and Writing

Speech is time-bound, dynamic, transient. It is part of an interaction in
which both participants are usually present, and the speaker has a particular
addressee (or several addressees) in mind.
• There is no time-lag between production and reception, unless one is deliberately
introduced by the recipient. The spontaneity and speed of most speech exchanges
make it difficult to engage in complex advance planning. The pressure to think
while talking promotes looser construction, repetition, rephrasing, and comment
clauses. Intonation and pause divide long utterances into manageable chunks, but
sentence boundaries are often unclear.
• Because participants are typically in face-to-face interaction, they can rely on
such extralinguistic cues as facial expression and gesture to aid meaning
(feedback). The lexicon of speech is often characteristically vague, using words
which refer directly to the situation (deictic expressions, such as that one, in here,
right now).
• Many words and constructions are characteristic of (especially informal) speech.
Lengthy coordinate sentences are normal, and are often of considerable
complexity. There is nonsense vocabulary, obscenity, and slang, some of which
does not appear in writing, or occurs there only as graphic euphemism (e.g. f***).
• Speech is very suited to social or ‘phatic’ functions, such as passing the time of
day, or any situation where casual and unplanned discourse is desirable. It is also
good at expressing social relationships, and personal opinions and attitudes, due to
the vast range of nuances which can be expressed by the prosody and
accompanying non-verbal features.
• There is an opportunity to rethink an utterance while it is in progress (starting
again, adding a qualification). However, errors, once spoken, cannot be withdrawn
(the one exception is when a sound engineer performs wonders of auditory plastic
surgery on a digital recording of nonfluent speech); the speaker must live with the
consequences. Interruptions and overlapping speech are normal and highly audible.
• Unique features of speech include most of the prosody. The many nuances of
intonation, as well as contrasts of loudness, tempo, rhythm, and other tones of
voice cannot be written down with much efficiency.
 Writing is space-bound, static, permanent. It is the result of a situation in
which the writer is usually distant from the reader, and often does not know who
the reader is going to be (except in a very vague sense, as in poetry).
• Writers must anticipate the effects of the time-lag between production and
reception, and the problems posed by having their language read and interpreted by
many recipients in diverse settings. Writing allows repeated reading and close
analysis, and promotes the development of careful organization and compact
expression, with often intricate sentence structure. Units of discourse (sentences,
paragraphs) are usually easy to identify through punctuation and layout.
• Lack of visual contact means that participants cannot rely on context to make
their meaning clear; nor is there any immediate feedback. Most writing therefore
avoids the use of deictic expressions, which are likely to be ambiguous.
• Some words and constructions are characteristic of writing, such as multiple
instances of subordination in the same sentence, elaborately balanced syntactic
patterns, and the long (often multi-page) sentences found in some legal documents.
Certain items of vocabulary are never spoken, such as the longer names of
chemical compounds.
• Writing is very suited to the recording of facts and the communication of ideas,
and to tasks of memory and learning. Written records are easier to keep and scan;
tables demonstrate relationships between things; notes and lists provide
mnemonics; and text can be read at speeds which suit a person’s ability to learn.
• Errors and other perceived inadequacies in our writing can be eliminated in later
drafts without the reader ever knowing they were there. Interruptions, if they have
occurred while writing, are also invisible in the final product.
• Unique features of writing include pages, lines, capitalization, spatial
organization, and several aspects of punctuation. Only a very few graphic
conventions relate to prosody, such as question marks and underlining for
emphasis. Several written genres (e.g. timetables, graphs, complex formulae)
cannot be read aloud efficiently, but have to be assimilated visually.
 9. Mixed Medium
The distinction between the medium of speech and the medium of writing at
first sight seems clear-cut: either things are written or they are spoken. In practice,
the situation is considerably more complex. When we choose to use either one of
these mediums, the reason for our choice may require us to bear in mind the
existence of the other, and that then influences the nature of the language we use.
The figure below summarizes the chief alternatives which are likely to produce
distinctive styles of spoken or written English, and the text illustrates some typical
situations under each heading.
Speech Writing Mixing
If we choose to speak, we If we choose to write, we There remain a few situations
may intend our utterance to normally intend that what we where speaking and writing
be heard immediately. This is have written should be read; are mutually dependent: the
the normal state of affairs. and the norm, at least since language used is partly made
But there are several late classical times, has been up of speaking/listening
interesting alternatives. for the recipient to read activities and partly of
• We may intend our silently. Here too there are reading/writing activities, in
utterance to be heard at a later several alternatives. proportions that are
point in time, as when we use • We may choose to write sometimes difficult to
an answerphone. with the intention that what disentangle. There are three
• We may intend that what we we have written should be chief possibilities, depending
say should not be heard, as read aloud. If so, we must on the nature of the
when we speak sotto voce make a further choice. We addressee.
(‘under our breath’). There may write in such a way that • We may address ourselves
are of course two further our end-product, when read in this mixed way, as when
options here: the genuine aloud, will sound like written we compile a shopping list
sotto voce, which our listener language. It will be relatively simultaneously questioning
does not hear, but which formal and controlled. Those ourselves about what we want
nonetheless makes us feel who prepare the text for radio while writing down some of
better for having said it; and news-readers fall into this what we say.
the pseudo sotto voce, which category. Alternatively, we • We may address a single
we intend our listener to hear may write in such a way that listener, as when people work
(usually for jocular purposes). the end product will not together in a co-authorship
Unintentionally overheard sound scripted, as in those situation, jointly poring over
sotto voce can lead to trouble who write material for radio a text (an academic paper, a
for the speaker, though this and television drama. The sitcom script) and each
depends on non-linguistic latter are not always contributing suggestions to it.
factors (such as the relative successful, of course. • We may address a group of
physical build of speaker and • We may choose to write listeners, as when a teacher is
listener). with the intention that only using the blackboard, keeping
• We may intend our some of what we have written up a running commentary to a
utterance to be written down. should be read aloud, the rest class while doing so. In such
If so, there are two further being ignored. An example of cases, an audio recording
possibilities: we may leave this rather unusual situation would tell only half the story,
the task of representing what can be found in a radio as would a photograph of the
we say to the listener, thus channel’s continuity studio, written work. Both mediums
speaking in a relatively where information of jointly work together to
‘natural’ way (as in some potential interest to the produce a successful use of
magazine interviews or police listener (e.g. about the language.
statements); or we may speak weather, traffic delays) is
‘carefully’, instructing the continually coming in on a
writer to ignore non-fluencies television screen or being
and errors (as in letter passed to the presenter in note
dictation). form. The presenter selects
what there is time to
incorporate into the running
order of the programme. The
material arrives in a variety of
styles, often highly elliptical,
reflecting the ongoing rush of
the live broadcast situation.
 10. Monologue and Dialogue
A factor which fundamentally influences the linguistic character of a use of
language is the number of participants involved in the activity. Theoretically, the
distinction is clear-cut: there is monologue, in which only one person is involved in
the linguistic act, and there is dialogue, in which (typically) two people are
involved. We would also expect there to be a close correspondence with the two
categories of medium: monologue is associated with the activities of writing and
reading, and dialogue with speaking and listening. As with so many of the
theoretical distinctions presented in this book, the outline is broadly correct, but
there are several cases where the distinctions become blurred or overlap, and it is
these which provide some of the most interesting examples of the way we use
language.
We can see how some of these cases arise by paying careful attention to
definition. Monologue does not mean that a person is alone, as is typical of most
authorial writing – the ‘lonely profession’, as it has been called. It refers rather to
an activity in which the language producer does not expect a response, even though
an audience may be present (and even though that audience may, from time to
time, respond, as in the heckling which can accompany a political speech). In a
monologue, the language is conceived as a self-contained presentation. By
contrast, it is of the essence of dialogue that the participants expect each other to
respond, and it contains many linguistic features which enable this to happen (most
obviously, question forms). The interesting cases, accordingly, are those where the
situation imposes special demands or constraints upon the speaker/writer, and
interferes with the normal expectations of response.

10.1. Unexpected Features of Dialogue

When we investigate how dialogues actually work, as found in recordings of
natural speech, we are often in for a surprise. We are used to seeing dialogue in
contexts where the language has been carefully crafted, such as the script of a play
or the conversations in a language teaching textbook. Such dialogues may be very
effective for their purpose, but they are usually a long way from what can happen
in everyday conversation. The stereotype is that people speak in complete
sentences, taking well-defined turns, carefully listening to each other, and
producing balanced amounts of speech. The reality is that people often share in the
sentences they produce, interrupt each other, do not pay attention to everything that
is said, and produce a discourse where the contributions of the participants are
wildly asymmetrical. Yet all of this nonetheless produces a perfectly normal,
successful conversation.

10.2. Monologue Variations

There are two possible situations in which someone may choose to engage in
a monologue, whether spoken or written: there may be an audience present, or
there may be no audience. In each case there are several interesting variations
which lead to linguistically distinctive texts. Few have received in-depth stylistic
investigation.
Audience present. With an audience present, the likelihood is that the
medium will be speech, and interruptability provides an interesting basis for
classification. Many spoken monologues presented to an audience are in principle
uninterruptable (other than by non-linguistic responses, such as applause).
Examples include a very formal speech, a lecture, and a sermon (in conservative
religious traditions). On the other hand, there are several such situations which do
permit interruption. The preachers facing many US black congregations are
reinforced in their rhetoric by responses from their listeners, and often adopt a
questioning style in order to elicit them. Political speeches, likewise, regularly play
to the audience in this way.
An interesting category is the case of an audience which is present but in no
position to respond (a ‘pseudo-audience’). Examples of these situations include the
dentist who carries on a conversation (even including questions!) while the
listener’s mouth is full of dental equipment, and the adult talking to a prelinguistic
infant (or the mother talking to the baby in her womb). It is a moot point whether
such events are best described as monologue or dialogue.
Audience absent. Leaving aside the case of literary expression, which can
be defended as either monologue or dialogue, the notion that there could be
monologue without an audience present at first seems somewhat unusual. Why
should we say anything at all, if there is no one to hear what we say? Why write
anything, if there is no one there to read it? Both speech and writing, however,
provide interesting cases where monologic activities do take place.
Speech activities. There is little scientific data on the point, but evidently
people do speak to themselves. The author has it on good authority that academics
have been known to talk through solutions to their problems while alone (e.g. in
the bath). There is also the common case of another kind of pseudo-audience – this
time, where no human being is present – though it is debatable whether such uses
might not better be called ‘pseudo-dialogues’. There are, for example, people who
talk to plants (and who are ready to give reasons for doing so). There are also
people who talk to their car – often to condemn it for malfunctioning. Indeed,
virtually any object can be addressed as if it were a person. ‘Aren’t you nice?’,
someone in a department store was overheard to say to a dress. (Whether we
include animals within this category, or in the same category as the infants
mentioned above, presents a further topic for debate.)
 Writing activities. Here too we have the unusual possibility of addressing
ourselves. The diary is the classic instance. Other examples include making notes
while preparing a talk, and note-taking while listening to a talk being given by
someone else. That notes are written for the benefit only of the note-taker is
evident if ever we try to use another person’s material – a situation which will be
familiar to any student who has missed a class and tried to catch up in this way.
The notetaker’s selection of information will have been made with reference to
what the writer already knows, and this, along with the elliptical style that comes
with writing under a time constraint, limits the possibilities of shared coherence.
Pseudo-audiences for monologic writing activities are also rather unusual.
Written examination answers are probably the clearest instances. There has been a
dialogue in one direction (the examiner has asked the student a question), but the
reply is a monologue (for the student has no expectation of a response – except
indirectly, in the form of a grade). Some party-games also provide pseudo-
audiences for written language. In one such game, participants each write a
sentence about someone else in the room and drop their contributions
(anonymously) into a hat. The sentences are then pulled out randomly in pairs and
placed in a sequence. In a children’s party, the enjoyment comes from the
juxtaposition of incongruous activities (such as Michael has got a new rabbit –
Jane’s feeling hungry). In adult parties, rather more risqué incongruities can
transpire.
Lastly, there are cases where we can write as if an audience is present,
because we know that at a later stage one will be. Activities here include preparing
a handout for a talk, writing an essay for a tutor, or indexing a book. Indexing has
sometimes been described as a task where the compiler is trying to anticipate every
possible query about content which future readers of the book might have. Indexers
are in effect trying to provide answers to a host of unasked questions – an
interesting reversal of communicative priorities. They therefore need to work as if
their audience is present – though, without knowing who this audience will be, and
without receiving any feedback as to whether their judgments have been
successful, the task is a difficult one, requiring exceptional communicative
commitment.

10.3. Dialogue Variations

One way of classifying dialogues is to examine their symmetry – to see
whether the participants are co-equally involved. There may also be variation in
the timing of the language contributions relative to each other. The norm is for
there to be two participants, who speak in sequence (but with a certain amount of
expected overlapping. However, several types of dialogic situation depart from this
norm in interesting ways.
Symmetrical dialogue. It is possible for people to use language
simultaneously, giving the impression of dialogue, but probably with little meeting
of minds. If two groups of protesters, both carrying placards expressing their
views, were to confront each other, the juxtaposition of written texts would
produce a kind of dialogue, but one in which all ‘utterances’ were on display at the
same time.
In speech, any simultaneity is likely to be unintelligible – but this does not
stop it happening, as is regularly heard in public political confrontations. Dinner
parties also bring up some interesting cases, where a person might end up
contributing to two conversations at the same time – introducing remarks into each
in sequence, but listening to both at once. It is something which succeeds only
when one is either very sober, or very drunk.
A further variation is for a dialogue to depend on a third party, or
intermediary. A common example is in foreign-language interpreting and
translation, where A has to communicate with B via C. Within a single language,
there are also well-known situations where one person (or group) communicates
with another via an ‘official spokesperson’, or (in an apparently rather different
domain) a ventriloquist’s dummy.
An interesting variant is for a dialogue to be generated using the utterances
of a third party. A loud-voiced person in a restaurant may cause a couple at another
table to provide their own responses (sotto voce) to what is said, creating their own
ongoing dialogue stimulated by the outsider’s utterances. In François Truffaut’s
Oscar-winning film, Day for Night (La Nuit Americaine, 1973), two of the film
crew are seen passing a television set which happens to be showing a quiz game
about films. They stop to watch, and try to answer the questions ahead of the
participants on the screen. They talk to each other while reflecting on their
answers, engaging in a dialogue which is, once again, dependent on a third party.
Asymmetrical dialogue. These are the most unusual dialogues of all, as
they take place with only one person apparently present. The qualification
‘apparently’ is important, because of course what happens is that the participant is
imagining someone else to be present. In some cases, the missing person is the one
who should start the conversation – as in the case of a seance, where people sit
waiting for someone to talk to them. In other cases, the missing person is the one
who should respond, as when we call uncertainly into the darkness ‘Is there anyone
there?’, and hope that we really are engaged in a monologue after all.
Letters to the press or a radio station perhaps also fall into the asymmetrical
category, given that there is only a remote chance that they will be used. In such
cases we are trying to make a contribution to a dialogue over which someone else
has control. If we do manage to get our contribution published or read out, there is
no way of knowing whom we shall end up ‘talking to’. Editors and programme
presenters are adept in making dramatic juxtapositions of letters in this way.
 11.New Textual Worlds
The study of discourse has become much more challenging since the arrival
of the Internet. The notion of a ‘text’, spoken or written, as described in this
chapter – a letter, interview, advertisement, sports commentary, news broadcast,
shopping list, textbook, newspaper editorial, lecture, prayer, road sign, novel,
poem – is identifiable and determinate. All satisfy five criteria:
• They have definable physical boundaries, either spatial (as in letters and books)
or temporal (as in broadcasts and interviews) or mixed medium (as in lecturing
karaoke using PowerPoint™).
• They are created at a specific point in time.
• Once created, they are static and permanent.
• Each text has a single authorial or presenting voice (even in cases of multiple
authorship of books and papers).
• The authorship is either known or can be established (except in some historical
contexts). It is a stable, familiar, comfortable world. And what the Internet has
done is remove the stability, familiarity, and comfort.
 E-mail, Websites, Hypertext links, Securities, Cookies, Incompletes,
Refreshing, Threads, Forums.
The traditional notion of ‘text’ cannot cope with these questions. It looks as
if some broader, more inclusive notion is going to be needed. What we see in these
examples are aggregates of functional elements, communicative acts, which
interact in various ways in different Internet outputs. No standard terminology has
yet evolved, and the issues remain questions for future research.

Literature
James R. Nattinger and Jeanette S. DeCarrico. Lexical Phrases and
Language Teaching. Oxford: Oxford University Press, 1992. Pp. xvi + 218.
 Stubbs, M. (1983). Discourse Analysis: The Sociolinguistic Analysis of
Natural Language. Chicago, IL: The University of Chicago Press. 
The Rules of Co-operative Conversation (from H. P. Grice, “The Logic of
Conversation”, 1975)
George Lakoff and Mark Johnson ‘Metaphors We Live By’ (1980)
Hart, C. (2018).  Event-frames affect blame assignment and perception of
aggression: An experimental case study in CDA.  Applied Linguistics 39 (3): 400-
421.