7

CHAPTER TWO
RELATED PRAGMATIC LITERATURE

2.1 The Notion of Implicature

As its cognate ‘implication’, Mey (1993:99) believes that the
word implicature1 is derived from the verb ‘to imply’ which has an
everyday sense in which we usually do imply something other than
we actually say by means of our utterance, hence what is implied
means, in etymology, something which is folded into something else
within the utterance and has to be unfolded for a better understanding
of that utterance. Thus, an implicature, as Yule (1996a:36) asserts, is
a primary example of more being communicated than is said. In
presenting the expression:

1. A hamburger is a hamburger.

as a reply to a question asked by one’s colleague, in the middle of

their lunch hour, about how he or she liked the hamburger, the
hearer first has to assume that the speaker is being co-operative and
that he intends to communicate something. Being asked to evaluate
the hamburger, the speaker has responded in a way which shows
that he or she has no evaluation whatsoever and this has only one
implicature that he or she has no opinion, whether good or bad, to
express. According to Jaszczolt (2002:210), Akmajian et al.
 8

(2001:398), and Bach and Harnish (1979:xi), additional implicatures,

depending on other aspects of the context, might be inferred by the
hearer whose job is to work out the implicature intended in a
certain context on the basis of what is already known. This
suggests that an implicature, being a proposition implied by the
utterance of a sentence in a context, is an inference which the
hearer draws from an utterance and perceives as being intended by
the speaker, i.e., the speaker, in the sense of meaning to
communicate something by an utterance, must intend to produce,
by that utterance, some effect in the hearer like a belief or an
action, by means of the recognition of that speaker’s intention,
hence it is a communicative intention whose fulfilment, being one
of its peculiar features, consists of its recognition. Another essential
feature which Grice discovers concerning these communicative
intentions as long as they are intended to be recognized is that they
are “open” or “overt” and not hidden or deceptive2. After all, the
hearer is not only invited to merely recognize meaning but also to
construct it since conversation, as conceived by Jaszczolt
(1999:206), is:

a process of assumption creation performed

in-between the speaker and the hearer,
assigning to the hearer a more responsible role
than just recovering the already existing
assumptions.

As for the distinction between saying and implicating, Grice

considers it to be so fundamental that he uses the expression ‘what is
said’ to refer to the truth-conditional content of utterances, i.e., to
what is explicitly or literally said but he uses the expression ‘what is
implicated’ to refer to everything that is overtly communicated by an
 9

utterance but is not part of what is said. Hence, one important

contribution made by the notion of implicature is that it promises to
fill the gap which is sometimes so substantial between what is
literally said and what is conveyed by giving some account of how, at
least, large portions of speech are effectively conveyed; a case in
which it is not possible for any semantic theory to provide more than
a small part of an account of how we communicate using language.
Such gaps between the literal meaning of a sentence and what it is
used to convey are so common in everyday life. An example is the
negation in:

2. Can’t you shut up?

which must be interpreted ironically; the result of this ironic negation

is to bias the interpretations towards a positive reply. In addition to
irony, there are other perlocutionary effects like sarcasm,
exaggeration, metaphor, politeness, etc. which are all derived from
the literal meaning of an utterance by means of inference and special
principles (Levinson, 1983:98; Hofmann, 1993:273; and Aijmer,
1996:126; Ifantidou, 2001:40).

2.2 Grice’s Theory of Implicature

The notion of implicature, unlike many other topics in
pragmatics, has no extended roots in history. There has been,
however, some considerable speculation about the utility of this
notion provided by the key ideas Grice proposed in the William
James lectures delivered at Harvard in 1967. But before reviewing
the suggestions that Grice put forward about this topic, it is important
to consider first his other major theory of meaning. To Levinson
 10

(1983:16), Grice’s formulation concerning his theory of meaning

reads as follows:

3. S meant -nn z by uttering U if and only if:

(i) S intended U to cause some effect z in
recipient H
(ii) S intended (i) to be achieved simply by H
recognizing that intention (i)3

This is indeed a characterization of non-natural meaning or meaning

-nn (equivalent to the notion of intentional communication) which
Grice presents so as to distinguish it from what he calls natural
meaning expressed in a sentence like Those black clouds mean rain.
What this characterization states is that communication consists of
the ‘speaker’ who intends to cause the ‘hearer’ to think or do
something, just by getting him to recognize that this thought or action
is caused by the speaker’s intention to do so. We find then that
communication, according to the above definition, is a complex kind
of intention that the speaker achieves or satisfies by getting the hearer
to recognize it.
Grice presents his theory of meaning -nn to explain how
interesting discrepancies there can be between speaker-meaning
(Grice’s meaning -nn) and sentence-meaning which explain how it is
possible for people to mean more than, or something different from,
what the sentence they utter actually expresses or to communicate
more than just the content of the sentence as is the case with (4)
below:

4. A: Smith doesn’t seem to have a girlfriend

these days.
B: He has been paying a lot of visits to New
York lately.
 11

where B’s reply clearly implies that Smith may have a girlfriend in
New York. Jaszczolt (2002:207) argues that no semantic theory
would enable us to conclude that the speaker B communicated his or
her knowledge or suspicion that Smith has a girlfriend in New York,
hence no semantic theory would treat B’s response as being relevant
and informative. This leads us to assume that utterances may carry
within themselves two meanings: one is sentence-based meaning
defined by semantics, and some additional meaning which is left for
pragmatics to deal with. In addition to these indirect implications of
utterances, there are certain aspects of meaning that cannot be
accounted for by truth-conditional semantics like sentential
connectives, definite descriptions and quantifiers which seem to be
used in English in a way that is not adequately captured by this
semantic analysis, and also referring expressions, like pronouns and
demonstrative noun phrases, which heavily rely on context to identify
the referent. Thus, one possible way to deal with such phenomena is
to try to find a pragmatic explanation of them and then merge this
explanation with the semantic analysis.
Levinson (1983:18) adopts Grice’s theory of meaning -nn so
as to limit the scope of meaning he talks about in his definition of
pragmatics (Ibid.:12) which reads as follows: “Pragmatics is the
study of all those aspects of meaning not captured in a semantic
theory.” He then includes only those phenomena for pragmatics to
deal with like ironic, metaphoric and indirect implications of what we
say (elements 5, 6 and 7 in table (1) below) and excludes the
unintended inferences that intuitively have no part to play in a theory
of communication (like the aspects 1, 2, 3 and 4 below) which must
be left for the semantic theory to deal with since this theory is
 12

concerned with all the conventional content of an utterance’s

significance.

Table (1) Elements of the Communicational

Content of an Utterance

1. truth-conditions or entailments
2. conventional implicatures
3. presuppositions
4. felicity conditions
5. conversational implicature-generalized
6. conversational implicature-particularized
7. inferences based on conversational structure

It should be borne in mind, however, that conventional

implicatures, presuppositions, and even perhaps aspects of
illocutionary force, though they are conventional, cannot be
accounted for within the semantic theory on the assumption of
truth-conditional semantics, hence they must be left to pragmatics to
deal with since they are non-truth-conditional elements of
sentence-meaning.
Especially vexing is the relationship between speaker’s
meaning and implicature; therefore, there has been some
consideration, throughout this section, of Grice’s theory of meaning
to let readers of this piece of work know how utterer’s meaning
and utterance-type meaning are related to saying and implying.
For Harnish (1976:332), it is worth looking at one strong
supposition that “what one implicates (as well as what one says) is
always a species of what was meant, differences being traceable to
 13

different ways in which various intentions are to function and

different reasons the audience is intended to have.” It becomes
clear, then, that Grice’s theory of meaning -nn has a connection of
an important kind to his theory of implicature. As Levinson
(1983:101) writes:

If, …, Grice’s theory of meaning -nn is

construed as a theory of communication, it has
the interesting consequence that it gives an
account of how communication might be
achieved in the absence of any conventional
means for expressing the intended message. A
corollary is that it provides an account of how
more can be communicated, in his rather strict
sense of non-naturally meant, than what is
actually said. Obviously, we can, given an
utterance, often derive a number of inferences
from it; but not all those inferences may have
been communicative in Grice’s sense, i.e.,
intended to be recognized as having been
intended. The kind of inferences that are called
implicatures are always of this special intended
kind, and the theory of implicature sketches one
way in which such inferences, of a
non-conversational sort, can be conveyed while
meeting the criterion of communicated
messages sketched in Grice’s theory of
meaning -nn.

Grice develops the concept of implicature in his second theory which

is essentially a theory about how people use language. This theory is
based on his suggestion that the conduct of conversation is guided by
a set of overarching assumptions which arise, it seems, from basic
rational considerations formulated in a way to serve as guidelines for
the efficient and effective use of language in conversation to further
co-operative ends. As guidelines of this sort underlying the efficient
co-operative use of language, Grice proposes four basic
 14

conversational maxims which jointly express a general co-operative

principle, henceforth CP. (See section 2.6).

2.3 Implicature, Impliciture and Explicature

Recently, some additional terms have been suggested with the
aim of illustrating other kinds of pragmatic inferences for which the
notion of implicature is said to fail to account, these are: impliciture
and explicature.

2.3.1 Implicature vs. Impliciture

In explaining the difference between implicature and
impliciture, Bach (1994:126) comments: “In implicature one says and
communicates one thing and thereby communicates something else in
addition. Impliciture, however, is a matter of saying something but
communicating something else instead, something closely related to
what is said.” Hence, unlike Grice, Bach treats (5b) below:

5. a. I have had breakfast.

b. I have had breakfast [today].

as a case of impliciture rather than implicature because he thinks that

it has too close a relation with the sentence uttered.

2.3.2 Impliciture vs. Explicature

There are two different ways for the speaker to communicate
something without making it fully explicit: completion and
expansion. In this respect, Bach (Ibid.) writes: “…, part of what is
communicated is only implicit in what is explicitly expressed, either
because the utterance is semantically underdeterminate and
 15

completion is required or because what is being communicated is an

expanded version of the proposition expressed.” Thus, Bach
considers the pragmatic material one needs to add to (5a) above to
get (5b) an impliciture recovered by a process of expansion
because the utterance expresses a complete proposition but it does
not coincide with the proposition that the speaker means. However,
he does not include that material within what is said, i.e., the
explicature because doing so would be to blur the difference
between what is explicit and what is implicit in utterance
meaning.

2.3.3 Implicature vs. Explicature

Implicature and explicature are two main propositions which
form the communication. Agerri and Korta (2001:166) believe that
an implicature is an assumption which, when communicated by the
utterance, is not explicit but implicit. It is what the speaker conveys
implicitly, or in common parlance, what is implied. Furthermore, an
implicature, in Gricean tradition, has been assumed to be any
grammatically determined aspect of utterance meaning apart from
disambiguation and reference assignment. By contrast, an
assumption is an explicature only in case, as Sperber and Wilson
(1986:182) maintain, “it is a development of a logical form encoded
by the utterance.” Grice assumes that the explicature given by core
semantics is just the literal meaning of the words uttered by the
speaker. Kearns (2000:271), however, thinks that there is almost
always a gap between the literal meaning of an utterance and the
explicature and that there are three pragmatic processes involved in
the recovery of the explicature, these are: disambiguation of
ambiguous expressions, reference assignment to variables, and
 16

interpretation of indexical expressions. These are indeed the three

main areas which have long been acknowledged by traditional
semantics and “in which contextual information”, in Kearns’ opinion
(Ibid.), “is needed to establish a proposition for which truth
conditions can be given.”
Disambiguation of ambiguous expressions is a
straight-forward process. Both lexical and structural ambiguity can
often be resolved by pragmatic considerations in order to identify
which meaning the speaker actually expresses by uttering his
sentence. An example of a lexical ambiguous expression is the
phrase that creamy duck in:
6. They’ve got that creamy duck on special at
Forresters.

which may either refer to their special dish if Forresters is a

restaurant or to cream-coloured strong cotton twill if Forresters is a
fabric store. The quantifier scope ambiguity in the sentence:

7. Everyone should bring a pencil.

is an example of structural ambiguity which is resolved by common

sense in favour of the surface order of the quantifier, and so we
understand that each one should bring a special pencil.
Reference assignment to variables includes personal pronouns
and simple tense time references. Here, again the context plays a role
in determining the reference of some expression. In this respect,
Kearns (Ibid.:272) writes: “the referent for the variable is identified
through a linguistic context, but the connection between the variable
(pronoun or time variable) and the expression it depends on is still
 17

established pragmatically, not by any automatic syntactic or semantic

rule.” Let’s consider the sentences below:

8. a. The plumber came and he’ll send a quote.

b. Everyone came to dinner last night and
John got drunk.
In (8a) the pronoun he has the same referent as that of its antecedent,
the plumber, and in (8b) the time at which John got drunk is the same
as that of the party mentioned in the first conjunct; in both cases,
then, the reference is established by some expression earlier in the
sentence.
Indexical expressions also depend for their interpretation on
the context of utterance in some systematic way. There are, in fact,
certain key points which, as stated by Kearns (Ibid.), form the centre
of an utterance and on which the interpretation of deictic expressions
depends, these are: the identity of the speaker, the identity of the
hearer(s), the time and place of the utterance, and the position of the
speaker. Thus, you refers to the hearer(s), yesterday refers to the day
before the day of the utterance, now refers to the time of the
utterance. Likewise, the contrast between the deictics this and these,
that and those is interpreted in terms of nearness to or distance from
the speaker, and even the gesture which accompanies the
demonstrative uses of these deictics is a feature of the immediate
context of utterance. Tense also commonly causes the hearer to make
use of wider contextual information for the identification of the
specific time a speaker is talking about, hence, according to Kearns
(Ibid.:273), “Tense is also indexical, as past, present and future are
calculated as earlier than, overlapping with, or later than the time of
utterance.”
 18

It is worth noting that the disambiguation of ambiguous

expressions, the assignment of reference to variables, and the
interpretation of indexical expressions are the operations that
constitute the first pragmatic level giving rise to the explicature.
These operations differ from those that constitute the second
pragmatic level leading up to the implicature. The distinction
between the explicature (what is actually said) and the implicature
(what is additionally implicated or inferred) depends to a
considerable extent on the nature of difference in operations involved
in the first and second pragmatic levels. We understand, then, that
the five components of interpretation as identified by Kearns
(Ibid.:274) are:

1. The literal meaning of the words uttered.

2. First Pragmatic Level:
reference assignment to variables, including
tense; disambiguation of ambiguous
expressions, including names; interpretation
of indexical expressions, giving the
explicature as output.
3. The explicature, the main truth condition for
what was said.
4. Second Pragmatic Level:
further inferences taking the explicature as
output.
5. The implicature(s).

2.4 Types of Implicature

Implicature is a general cover term which Grice intends to
contrast with what is said or expressed by the truth conditions of
expressions and to cover all the kinds of pragmatic
(non-truth-conditional) inference discernible. Grice distinguishes
 19

two different sorts of implicature: conversational implicatures and

conventional implicatures.
Conversational implicatures, as defined by Crystal
(1980:172), “refer to the implications which can be deduced from the
FORM of an UTTERANCE, on the basis of certain CO-OPERATIVE
PRINCIPLES which govern the efficiency and normal
ACCEPTABILITY of conversations.” In other words, meaning is
conveyed through an utterance in a manner that cannot be garnered
simply from the meaning of the words in the utterance. (See section
2.5).
Conventional implicatures, on the other hand, “are
independent of context, to the extent that this is the case, is part
and parcel of their conventionality” (King, 1997:1). These
implicatures arise solely because of conventional features of the
words the speaker employs in an utterance. They are, in other
words, part of the conventional or lexical meaning of words, hence
they are, according to Potts (1999:2), commitments the speaker of
an utterance makes by virtue of the meaning of these words.
Bach (1999:370) also views conventional implicatures in the same
way when he says:

A proposition is a conventional implicature of

an utterance just in case (a) the speaker
(speaking seriously) is committed to the truth
of the proposition, (b) which proposition that is
depends upon the (or a) conventional meaning
of some particular linguistic device in the
utterance, but (c) the falsity of that proposition
is compatible with the truth of the utterance.

Conventional implicatures are carried by a wide range of

various indicators like and, but, if, so, etc. whose prime function is
 20

“to combine viewpoints into structure” (NØlke, 1992:197) or

“‘bracket discourse’, that is, to mark relations between sequentially
dependent units of discourse” (Traugott, 1995:5). In short, they
indicate the relationship between an utterance and the prior
discourse. Traugott (Ibid.) adds that these indicators “have major
pragmatic functions. They express speaker attitude to what has
gone before, what follows, the discourse situation and so forth.”
They are, then, all primarily pragmatic, or non-truth-functional,
which may explain why they have been largely ignored until the last
few years.

2.4.1 Conversational vs. Conventional Implicatures

As for the difference between these two varieties of
implicatures, Lyons (1977:593) writes:

The difference between them is not always

clear-cut in particular cases. In principle,
however, the difference seems to be that,
whereas a conventional implicature depends
upon something additional to what is
truth-conditional in the normal (i.e.,
conventional meaning of words), a
conversational implicature derives from a set of
more general conditions which determine the
proper conduct of conversation.

Conversational implicatures are crucially not decoded but

calculated on the basis of pragmatic or social principles including the
CP and its four maxims, hence they are pragmatic or social. By
contrast, conventional implicatures are linguistically decoded, hence
they are semantic. They form an entirely different kind of non-truth
conditional inferences which Grice envisages as being not derived
 21

from super-ordinate pragmatic principles like the maxims but are

simply attached by convention to particular lexical items or
expressions like but, therefore, even, yet, etc. (Levinson, 1983:127
and Ifantidou, 2001:53).
Conversational implicatures also differ from conventional
ones in that:

what is conversationally implicated is

somewhat indeterminate: what is potentially
conversationally implicated is a large (perhaps
indefinitely large) disjunction of propositions
…, and the speaker is in a position to rationally
deny that a particular one was intended. The
conventional implicature of a linguistic
expression, on the other hand, is quite specific,
and so not cancellable; you cannot deny that
you meant them without involving yourself in a
contradiction (Green, 1989:94).

In addition to cancellability (or defeasibility), conversational

implicatures are expected to contrast with conventional implicatures
on the basis of three other major distinguishing properties which the
former exhibit, these are: non-detachability, calculability, and
non-conventionality. (See section 2.5.1). In this respect, Levinson
(1983:128) asserts that:

conventional implicatures will be

non-cancellable because they do not rely on
defeasible assumptions about the nature of the
context; they will be detachable because they
depend on the particular linguistic items used
…; they will not be calculated using pragmatic
principles and contextual knowledge, but rather
given by convention …; they may be expected
therefore to have a relatively determinate
content or meaning; and there will be no
expectation of a universal tendency for
 22

languages to associate the same conventional

implicatures with expressions with certain truth
conditions.

Finally, it is worth mentioning that what is implied, in the

case of conversational implicatures, varies according to the context
whereas in the case of conventional implicatures the same implicature
is always conveyed regardless of the context in which the utterance
occurs. However, both types have in common the property that an
additional level of meaning, beyond the semantic meaning of the
words uttered, is generated in both cases (Thomas, 1995:57).

2.5 Conversational Implicatures

Conversational implicatures are so called because what is
implied (or in Grice’s term ‘implicated’) is implicated on the basis of
the fact that the speaker and hearer are co-operatively contributing to
a conversation. This suggests that the basic assumption in
conversation is that, unless otherwise indicated, the participants are
adhering to the CP and its maxims. (See section 2.6). Allan (1999:2)
argues that conversational implicatures also depend upon common
ground between speaker and hearer who are mutually aware that their
interlocutor is normally an intelligent being. Thus, the speaker does
not need all the time to spell out things which are quite clear to the
sensory receptors of the hearer or which the hearer can easily infer on
the basis of (1) knowing the conventions of language use, and (2)
using the knowledge that we normally develop from birth as we
experience the world around us.
Following Grice, one may distinguish two categories of
conversational implicatures: generalized and particularized
implicatures. “Generalized implicatures are ones which do not
 23

require a particular context in order to be inferred” whereas

“Particularized implicatures are valid only in a particular context”
(Marcu, 1994:4-5).
What these definitions suggest is that the difference
between these two types of implicatures is only a matter of degree
of dependence on context, not a categorial one. This leads Peccei
(1999:38) to assume that particularized conversational implicatures
“require not only general knowledge but also knowledge which
is particular or ‘local’ to the speaker and the hearer, and often to
the physical context of the utterance as well.” Consider this
example:
9. Mat : Want some fudge brownies?
Chris: There must be 20,000 calories there.
(‘No’)

Not only do the speaker and hearer, in this example, need some
general knowledge that food with a high number of calories makes
people put on weight but they also need to share the knowledge that
Chris is trying to lose weight, otherwise the implicature would be
(‘yes’) if they both knew that he is trying to gain weight.
Particularized implicatures, then, are inferences which require this
kind of shared knowledge between the speaker and hearer, and indeed
this is one reason why the present study is particularly interested in
this type of implicatures.
Jaszczolt (2002:217) also believes that the difference between
these two types of implicatures lies in the closeness of generalized
implicatures to the semantic content of an utterance as is the case
with, for example, the meaning of the connective ‘and’ from which
we extend the sequential sense of ‘and then’. By contrast, a sentence
like:
 24

10. Tom has wooden ears.

would generate the particularized implicature that:

11. Tom doesn’t appreciate classical music so

we should not invite him to the concert.

which is rather remote from the semantic content of sentence (10)

above.
Finally, it is worth noting that the second and most important
reason why the present study is especially interested in particularized
implicatures is that their generation comes as a result of observing the
maxim of relevance which is the main focus of this study since
relevance among utterances is, as Levinson (1983:127),
accomplished only with respect to the particular topic or issue at
hand. (See section 2.7).

2.5.1 Properties of Conversational Implicatures

To understand the nature of conversational implicatures and
have some sound ways of distinguishing them from other kinds of
semantic and pragmatic inferences, it is crucial that we list the
properties which Grice attributes to these implicatures, these
properties are: cancellability (or defeasibility), non-detachability,
calculability, and non-conventionality.
In addition to these, implicatures are the property of
utterances but not of sentences, thus one further property that
implicatures possess is that they change; that is, the same words give
rise to different implicatures on different occasions. A clear example
is the expression ‘How old are you’ which Thomas (1995:80-1) uses
 25

in three different situations just to show the different implicatures that

this same expression can carry:

12. A young boy is talking to a colleague of his

father:
A: It’s my birthday today.
B: Many happy returns. How old are
you?

Here, ‘How old are you?’ is a direct request for information.

13. Speaker A is talking to his son who is

annoying his little brother by doing things
that make the latter cry in the end:
Father : How old are you, George?
George: I’m eighteen, father.
Father : I know how old you are, you
fool.

In this example, ‘How old are you?’ generates the implicature that the
son is too old to do any silly behaviour.

14. A psychiatrist is talking to a woman patient:

A: Where do you work?
B: In a hospital, but my husband
won’t let me work.
A: How old are you?
B: I’m thirty-nine.

By using the expression ‘How old are you?’ the psychiatrist in

this situation wants the patient to consider whether, at the age of
thirty-nine, she isn’t old enough to decide for herself to go to work or
not.
Also, conversational implicatures seem to be the only kind of
pragmatic or semantic inferences that are reinforceable. Yule
(1996a:44) thinks that it is quite easy for a speaker to reinforce the
 26

implicature by adding further information without a sense of

anomalous redundancy as in:

15. You’ve won five dollars, that’s four more

than one.

2.5.1.1 Cancellability (or Defeasibility)

The surest test for distinguishing between what a speaker’s
words mean (i.e., semantic meaning) and what they actually imply
(i.e., implied meaning) is the notion of defeasibility which means that
the implicature can be cancelled by further information added to deny
that implicature without contradicting the utterance which implicates
it (Gazdar, 1979:40; Thomas, 1995:85; Yule, 1996a:44; and Allan,
1999:2). In this respect, Platts (1979:76-7) remarks: “The
implication from p to q is cancellable if and only if we can add a
further clause ‘withholding commitment from what would otherwise
be implied’, ‘without annulling the original assertion’.” To explicate
this property, let’s consider the following example offered by
Jaszczolt (2002:213):

16. a. Max has three children.

b. Max has exactly three children.
c. Max has three children, if not four.

Here, (16a) conversationally implicates (16b) but the added clause ‘if
not four’ in (16c) cancels this implicature.

2.5.1.2 Non-detachability
Some aspects of meaning are semantic which, by a process of
relexicalization or reformulation, can be changed by using some
 27

closely related word or phrase to replace the original one but without
the supposedly unpleasant connotation (Thomas, 1995:78). This is
referred to as the detachability of an implicature which Platts
(1979:76) defines as follows: “The implication from p to q is
detachable if and only if there is some other form of words which can
be used to state or assert ‘just’ what p ‘might be used to assert’, but
which does not carry the implication that q.” Below is an example
produced by Thomas (1995:79) to illustrate the point:

17. A radio interview with Jim Morgan, of the

management of Railtrack Southwest,
during the eighth week of a
signal-workers’ strike:
Int : Some will obviously, and
indeed already have, accused
you of trying to bribe staff to
cross the picket lines.
J. M.: Well, it’s not a bribe to staff to
cross picket lines, it’s an
offer.

Here, J. M. has relexicalized the offending lexical item ‘bribe’ by

replacing it with ‘offer’ which is a synonymous word that does not
have the negative connotation associated with ‘bribe’ and thus has
removed the unpleasantness which is part of the meaning of ‘bribe’.
This, however, should not be the case with an implicature, for the
implicature, as Thomas (Ibid.) explains, remains the same no matter
how much you reword an utterance and regardless of whether or not
the chosen lexical item has negative or positive connotations as when
one looks at a fat cat and says:

18.
Underfed/Fail/Puny/Skinny/Delicate/Light
-on-his feet/Slimline, isn’t he?
 28

where all the adjectives carry the implicature that the cat is fat. In
short, an implicature must be non-detachable which means that “it
must not be possible to substitute some other expression in the
sentence that lacks the implicature in question but which otherwise
means much the same thing” (Gazdar, 1979:40).

2.5.1.3 Calculability
As mentioned earlier, the same words may convey very
different implicatures on different occasions. It is the job of the
listener to calculate the implicature intended via inference. Thus,
calculability means ‘inferrability’ on the basis of the CP and its four
conversational maxims (Thomas, 1995:82; Yule, 1996a:44; Jaszczolt,
2002:213).
For Grice, the calculability requirement is fundamental for
distinguishing the conventional implicatures of an utterance which
are arbitrarily stipulated from its conversational implicatures which,
as stated above, one can recover by an inference process. As cited in
Wilson and Sperber (forthcoming:46), Grice (1975:50) remarks: ‘The
presence of a conversational implicature must be capable of being
worked out; for even if it can in fact be intuitively grasped, unless the
intuition is replaceable by an argument, the implicature (if present at
all) will not count as a conversational implicature.’

2.5.1.4 Non-conventionality
For an implicature to be conversational rather than
conventional, Gazdar (1979:40) asserts that it “must not be part of the
meaning of the expression to which it attaches. That is it must not be
given in the lexicon or specified as the meaning-changing effect of
 29

some syntactic operation.” That conversational implicatures are not

part of the meaning that has to be learned ad hoc is referred to as
non-conventionality. Conversational implicatures are natural and
predictable unlike the meaning of, for example, cat or therefore
which is conventional, arbitrary, and unpredictable, (Jaszczolt,
2002:213).

2.6 The Co-operative Principle and its Maxims

According to Grice, people involved in a conversation
acknowledge a kind of tacit agreement to co-operate conversationally
towards mutual ends. One aspect of the more general idea that
people will co-operate with each other is the concept that there is an
expected amount of information provided in conversation. In short,
conversation is a co-operative activity and is governed by certain
general principles of human rational co-operative behaviour that help
to account for the relation between (semantic) sense and (pragmatic)
force (or in Grice’s own terms between ‘what is said’ and ‘what is
implicated’), this kind of explanation is especially important to solve
puzzles which arise in a truth-based approach to semantics (Leech ,
1980:90; Leech and Short, 1981:295; Leech,1983:80; and Yule,
1996a:36-7). Thus, to work out the derivation of a given implicature
from a given sense one can make use of the contextual background
information as well as the maxims of rational conversational
behaviour which Grice puts under the heading of the CP. As quoted
in Ifantidou (2001:40), Grice (1989:26) defines the CP as a "‘rough
general principle which participants will be expected (ceteris paribus)
to observe, namely: Make your contribution such as is required, at the
stage at which it occurs, by the accepted purpose or direction of the
talk-exchange in which you are engaged.’" The CP and its maxims
 30

have the central role of explaining how it is possible for speakers to

communicate more than they actually say. This communication can
be characterized as being both rational and co-operative which are, in
fact, two fundamental assumptions underlying the CP and its
maxims. Ifantidou (2001:41) proceeds to say:

According to Grice, speakers are expected to

observe the maxims and hearers to assume, in
turn, that speakers are doing so. His idea was that
even if speakers appear to be violating a particular
maxim at the level of what is said, hearers will
provide any assumption needed to make the
utterance satisfy the maxims, or at least the CP.
They will, in other words, fill in the thought
behind the utterance so that the assumption that
the speaker is both rational and co-operative is
maintained.

This leads to the general assumption that the interpretation of

utterances is a collaborative enterprise in which a speaker and a
hearer are engaged in some shared goal guided by the CP and its
maxims. Consider the following example:

19. The speaker has accidentally locked herself

out of her house. It is winter, the middle of
the night and she is stark naked:
A: Do you want a coat?
B: No, I really want to stand out here
in the freezing cold with no
clothes on.

Here, B’s reply is both untrue and uncooperative, however, A will

look for an alternative interpretation assuming that, in spite of
appearances, B is observing the CP and has provided the appropriate
response to his question (Thomas, 1995:63 and Aronoff and
Rees-Miller, 2001:401 ).
 31

We conclude that there are certain conventions which

language interchange, like other social activities, requires
participants to mutually recognize. Among these conventions are
the maxims of the CP which, as viewed by Allan (1999:2) and
Aronoff and Rees-Miller (2001:402), are neither to be taken as
rules that dictate behaviour nor laws to be obeyed but, rather,
reference points for language interchange underpinning
collaborative exchange of information just like the points of
the compass which are conventional reference points for
identifying locations on the surface of the earth. Here follow these
maxims4:

2.6.1 Quantity
The maxim of Quantity relates to the quantity of information
to be provided, and under it fall two maxims:

1. Make your contribution as informative as is

required.
2. Do not make your contribution more
informative than is required.

Both Levinson (1983:106) and Kearns (2000:258) maintain that the

first clause or sub-maxim of Quantity which enjoins the speaker to
provide the full information is the important one, for it is identified as
the basis for generating a wide range of implicatures known as scalar
implicatures5.

2.6.2 Quality
This is the maxim of, as Kearns (Ibid.:257) calls it,
Truthfulness, for it commits the respondent when asked to give as
 32

much as he knows, rather than giving explicit but untrue

instructions ‘Try to make your contribution one that is true’. Under
this super maxim fall two more specific maxims which ask that
respondent not to give false route directions and also not to say that
of which he might be ignorant, these two sub-maxims read as
follows:

1. Do not say what you believe to be false.

2. Do not say that for which you lack adequate
evidence.
It is in fact this last case that leads Ervinn-tripp (1987:47) to suggest
that “the maxim appears to be overridden by another principle which
is “Don’t appear to be stupid or ill-mannered”.”

2.6.3 Relation
Under this category Grice places a single maxim, namely, ‘Be
relevant’. As this is the topic with which the present study is
particularly concerned, it will later be considered in some detail. (See
section 2.7 below).

2.6.4 Manner
Grice propounds the maxim of Manner including the general
maxim ‘Be perspicuous’ under which falls a general bunch of items,
these are:

1. Avoid obscurity of expression.

2. Avoid ambiguity.
3. Be brief (avoid unnecessary prolixity).
4. Be orderly.

Gazdar (1979:43-4) views each of these sub-maxims in the following

way: sub-maxim (1) instructs speakers and hearers to interpret each
 33

other using the same language6 or the intersection of their perspective

languages or idiolects. Sub-maxim (2) informs participants in a
conversation not to be ambiguous i.e., not to use ambiguous
expressions. In case they hear or use an ambiguous expression, they
must not treat it as simultaneously having several readings which
means that they must assign one and only one reading to it.
Sub-maxim (3) quantifies over the length of expressions at some
level of interpretation, and hence it instructs speakers, given two
potentially synonymous expressions A and B such that B is longer
than A, to choose A rather than B. It also instructs hearers on hearing
B to interpret it in a way that is different from A, for if B means the
same as A, then A would have been used instead. The last
sub-maxim ‘Be orderly’ can be explained along the following lines: if
A and B are two expressions in a sentence Q where A and B, in that
order, have distinct extensions and are members of some set Σ, then
upon hearing the expression A^ before ^B Q implies that the event
expressed by A precedes the event expressed by B.
In conversation, these maxims are to be taken as unstated
assumptions. Thus, we assume that normal human beings who
converse with each other have agreed to be co-operative; that is, they
will tell each other the truth, they will only provide information
assumed to be new and relevant to the listener, and they will only
request information that they sincerely want to have. This
assumption that people, in a communicative act (typically talking but
also writing), obey the CP and its maxims provides information about
the utterance itself so the utterance can be taken to be relevant, true,
clear, and informative; in other words, the CP together with its
maxims specify what participants in a conversation have to do so that
they can converse in a maximally efficient, rational and co-operative
 34

activity (Bates, 1976:27; Levinson, 1983:102; Yule, 1996a:37 and

Kearns, 2000:255).
On the basis of the assumption mentioned above, the hearer is
in fact led to draw a certain inference when the speaker deliberately
phrases his utterance in such a way that the content of that utterance
is implicated by the speaker. Inferences generated by the maxims are,
by definition, conversational implicatures7 and come about in two
distinct ways: either by observing the maxims in a fairly direct way
or by overtly and blatantly not following some maxim, in order to
exploit it for communicative purposes.

2.6.5 Observing the Maxims

In case of observing the maxims, Levinson (1983:104) shows
that the speaker “may nevertheless rely on the addressee to amplify
what he says by some straightforward inferences based on the
assumption that the speaker is following the maxims.” In this
example:

20. A: (to passer by): I’ve just run out of petrol.

B: Oh, there’s a garage just around the
corner.

B’s utterance is taken as an implicature that A may get some petrol

there. Levinson (Ibid.) calls inferences which do not require
particular contextual conditions in order to be inferred standard
implicatures (Grice’s term for this subset of inferences is generalized
implicatures). Thomas (1995:64) remarks that observing all the
maxims is the least interesting case as in the example below:

21. Husband: Where are the car keys?

Wife : They’re on the table in the hall.
 35

Where the wife, no more and no less, has said precisely what she
meant, she “has answered clearly (Manner) truthfully (Quality), has
given just the right amount of information (Quantity) and has
directly addressed her husband’s goal in asking the question
(Relation)” ; therefore, she has generated no implicature; that is,
there is no additional level of meaning since there is no distinction
to be recognized in this case between what she says and what she
means.
2.6.6 Non-observance of the Maxims
Grice suggests that this code of conversation whereby a
speaker observes all the maxims rarely holds constant across a given
sample of real dialogue. He adds that so long as participants in a
talk-exchange each assumes that the other is adhering to the CP,
meanings that are conveyed without being said follow as inferences
from the fact that some particular maxim appears not to be being
fulfilled. In other words, speakers sometimes contribute additional
information when they use the set of standard rules in such a way that
their deviations from the code will be recognized as violations, and
when this happens the listener must perceive the difference between
what the speaker says and what he means by what he says, the
particular meaning deduced for the latter being the implicature. In
brief, speakers may blatantly fail to observe a maxim having the
intention to prompt the hearer to look for a meaning which is
different from, or in addition to, the meaning expressed in their
sentences (Bates, 1976:27; Leech and Short, 1981:295; Green,
1989:88-91; and Thomas, 1995:65).
Grice lists four ways in which a participant in a talk-exchange
may fail to observe a maxim: when he or she may (1) violate a
maxim, (2) infringe a maxim, (3) opt out of observing a maxim, and
 36

(4) flout a maxim. By far, (4) is the most important category of

failing to observe a maxim, for it is the one that generates an
implicature8.

2.6.6.1 Violating a Maxim

Thomas (1995:72) observes that the term ‘violate’ is
incorrectly used by many commentators to cover all forms of
non-observance of the maxims. This leads Grice in his first
published paper on conversational co-operation (1975) to define
‘violation’ very specifically as the quiet and unostentatious
non-observance of a maxim. In this respect, Peccei (1999:27) says:
“Violations are ‘quiet’ in the sense that it is not obvious at the time of
the utterance that the speaker has deliberately lied, supplied
insufficient information, or been ambiguous, irrelevant or hard to
understand.” In Grice’s analysis, quietly violating a maxim is distinct
from openly flouting a maxim in that the former, though it might
hamper communication, does not lead to implicatures. To illustrate
this point, let’s consider the following situation:

22. Speech therapist: So you like ice-cream.

What are your favourite
flavours?
Child with a prag-: Hamburger … fish and chips.
matic disorder

Here, the child fails to observe the maxim of relevance because it

hasn’t come to his mind that favourite flavours would be interpreted
as ‘favourite flavours of ice-cream’, not as ‘favourite flavours in
general’. Peccei (Ibid.:29) views this case as a violation rather than a
flouting of the maxim because the child’s irrelevance is not
deliberate.
 37

2.6.6.2 Infringing a Maxim

A speaker who fails to observe a maxim because of imperfect
linguistic performance is said to infringe the maxim, i.e., the
non-observance occurs not because of any desire on the part of the
speaker to deceive or generate an implicature but, in Thomas’
opinion (1995:74), “because the speaker has an imperfect command
of the language (a young child or a foreign learner, because the
speaker’s performance is impaired in some way (nervousness,
drunkenness, excitement), because of some cognitive impairment, or
simply because the speaker is constitutionally incapable of speaking
clearly, to the point, etc.”

2.6.6.3 Opting Out of a Maxim

Thomas (Ibid.:74) points out that a speaker may opt out by
indicating plainly that he is unwilling to co-operate in the way
required by the maxim. One can frequently find examples of opting
out in public life, when legal or ethnic reasons may perhaps prevent
the speaker from replying in the way normally expected.

2.6.6.4 Flouting a Maxim

A flout occurs when a speaker blatantly fails to observe a
maxim at the level of what is said, not with any intention of
deceiving or misleading the hearer who is entitled, in this case, to
assume that the maxims (or the CP) are being observed at the level of
what is implicated. For example, the speaker in saying a sentence
like:
23. What lovely weather we’re having.
 38

on a dreadful day is not telling the truth; therefore, the hearer, in such
cases, must operate on the assumption that there is a general intention
by the speaker to conform to the CP as well as the assumption that
there is a conversational rule against blatant flashhoods and use a
conversational implicature to work out what the speaker meant from
what he said, and hence to discover that the sentence in the context
above serves as an ironic statement.
Sometimes, a speaker may be faced with a clash between
maxims. Thus, the speaker may have to sacrifice one maxim to the
other if two are in conflict because he or she is unable to conform to
both of them at once as in the following example:

24. A: Where does C live?

B: Somewhere in the south of France.

where B has breached the maxim of Quantity (say as much as

necessary) on the grounds of his awareness that to be more
informative would be to breach the maxim of Quality (have evidence
for what you say), hence B answers in a way which shows that he
does not know which town C lives in (Harnish, 1976:343-4; Bates,
1976:27; Garnham, 1985:107; and Green, 1989:89 ; Thomas,
1995:65).

2.7 The Maxim of Relation

‘Be relevant’ is not only a maxim for the speaker, but it is also
the principle on which the listener will base his interpretation of each
sentence. Thus, if the speaker says something irrelevant, the listener
will try to interpret it in a relevant way so that it adds as little as
possible. The listener, however, is liable to get the wrong meaning if
that sentence is ambiguous at all. As cited in Holdcroft (1987:481),
 39

Grice (1975:46) admits that his terse formulation of the maxim

involves difficult problems in that it offers no guide either to what an
expression is required to be relevant to, or to how what is relevant
can change during a conversation. Harnish (1976:341), however,
believes that this maxim is “so central and important in
conversational implicature that it is not clear that it belongs on equal
footing with the rest. I suspect that maxims are (at least partially)
ordered with respect to weight, etc. and that relevance is at the top,
controlling most of the others.”
Holdcroft (1987:481) postulates that the kind of relevance
Grice has in mind is relevance of topic. He believes, however, that
all Grice’s examples, though he explicitly declares his interest in the
notion of topical relevance, in fact incline towards that conception of
dovetailing which he (Ibid.:482) defines as “the determination of the
appropriateness of an utterance to its predecessor in a context.”
Thus, he (Ibid.:481) writes:

Now it is by no means clear that the notion of

topical relevance is the same as that of making
one’s contribution an appropriate one, given the
exchange’s aims and what has already been
said—for even if we are both talking on the
same topic, the question remains of how my
contribution dovetails with yours.

In example (20) considered earlier, it is taken for granted the kind of

response that A’s utterance calls for; that is, a context is assumed to
exist. If this is so, then there is no difficulty in interpreting the
response in that context, hence B’s response that there is a garage
around the corner is perfectly reasonable if the participants are
talking on a street in a town. It is worth noting, however, that A, in
this case can see B’s response as being relevant only if he thinks that
 40

the so-called topic framework, which, as stated by Brown and Yule

(1983:79), “consists of elements derivable from the physical context
and from the discourse domain of any discourse fragment,” can be
expanded to encompass a garage around the corner; which in turn
involves an expansion of the pool of discourse referents.
Furthermore, A can also see B’s response as being relevant only if he
takes into account the credibility of certain propositions which have
not been asserted in the exchange. Holdcroft (1987:482) remarks that
the emphasis on the notion of credibility at this point is important
since it involves more than just consistency. Consider the following
example given by Grice:

25. A: How is Smith getting on in his new job?

B: Oh quite well, I think, he likes his
colleagues, and he hasn’t been to prison
yet.

Here, if Smith is a known pillar of probity, then though the

assumption that he is dishonest is in consistency with everything that
has been asserted, it would not be sufficient for that assumption to
make B’s remark relevant—in this case some other credible
assumption would have to be made for that purpose.
Similar to Grice’s topical relevance is the interpretation
offered by Leech (1983:94) to the maxim of Relation ‘Be relevant’.
Leech treats the relevance of one utterance to another as part of a
broader sense of relevance which is the relevance of an utterance to
its speech situation. In this broader sense, Leech defines relevance as
follows: “‘An utterance U is relevant to a speech situation if U can be
interpreted as contributing to the conversational goal(s) of S or h.’”
Such conversational goals may cover both social goals as well as
 41

personal goals. This is explicated in Smith and Wilson’s example

(1979:175) given below:

26. A: Where’s my box of chocolates?

B: The children were in your room this
morning.

Here, A’s remark conveys his personal, illocutionary goal in finding

out where the box of chocolates is. B’s remark, though he may not
know the answer to A’s question, can be made relevant to the
question on the assumption that B, in his reply, will nevertheless
enable A to find out the answer by implicating at least that the
children may know where the chocolates are if not they may have
eaten them. Another way in which B’s reply can be made relevant to
A’s question is by interpreting it as contributing to the social goal
(e.g., observing politeness). The argument for saying that the
politeness principle also plays a role works like this: B’s preference
to give an indirect reply rather than a more direct one like The
children may have eaten them may be most likely motivated by polite
reticence in referring to a possibly sinful act made by children. In
other words, B has made A to come to the impolite conclusion by
giving a seemingly innocent statement concerning the whereabouts of
the children instead of accusing them directly. In both cases, B’s
reply is shown to be motivated by the CP; that is, it is intended to be
relevant to A’s conversational goals. In fact, in co-operative and
socially motivated conversation, one participant may normally adopt
to some extent the assumed goal or goals of the other. However, this
is not always the case as in this additional example provided by
Smith and Wilson (Ibid.:174):

27. A: Where’s my box of chocolates?

 42

B: I’ve got a train to catch.

Here, B’s reply is not very co-operative since it does not satisfy A’s
quest for his chocolates but it can be made relevant only if it is
understood as an explanation of B’s inability to answer A’s question.
It only enables B to conclude the conversation without being too
much impolite, hence its contribution to conversational goals is rather
negative. In fact, it contributes to B’s rather than A’s goal.
Wilson and Sperber (forthcoming:57) state that the speaker must
observe the maxim of relevance if he wants his utterance to convey
some relevant information. Sometimes the speaker may only have
the most general grounds for doing so; that is, he may provide
information that he only thinks will be relevant to the hearer having
no idea what specific implications this information will have for the
hearer. At other times, however, the speaker may have a much more
specific idea of the sort of context that will be brought to bear and the
sort of conclusions the hearer might derive. Wilson and Sperber
(Ibid.) believe that it is such situations that give rise to implicature
where a speaker’s remark appears to be irrelevant at first sight but
is then fully understood only on the basis of assuming that it is
relevant as in examples (26) and (27) above. Levinson (1983:107)
asserts that “such inferences are fundamental to our sense of
coherence in discourse: if the implicatures were not constructed on
the basis of the assumption of relevance, many adjacent utterances in
conversation would appear quite unconnected.” In addition to the
importance of the assumption of relevance in making and
understanding implicatures, it is also an all-pervading consideration
in understanding the basic content of what a speaker actually said.
Kearns (2000:261) illustrates this point in the following example:
 43

28. A: I’m sick to death of going to the

laundromat.
B: The man should be coming tomorrow.

Unlike an implicature, this content is a more explicit version of what

B actually said. Hence, B’s remark which seems apparently
irrelevant at first sight can then be understood to mean that the
repairman is coming to mend the washing machine the next day.
Thus, relevant to A’s remark is the possible implicature that A need
not go to the laundromat again. A’s remark, then, introduces the
laundry problem as a topic showing that the washing machine in A
and B’s flat has broken down and this is the general background
context. It is worth nothing that the role of relevance in
understanding what is actually said will be discussed in Chapter
Three. In brief,

the universal goal in cognition is to acquire

relevant information, and the more relevant the
better … a speaker who thinks it worth
speaking at all will try to make his utterance as
relevant as possible. A hearer should therefore
bring to the processing of every utterance the
standing assumption that the speaker has tried
to be as relevant as possible in the
circumstances. It is this assumption that we
call the principle of relevance. Wilson and
Sperber (forthcoming:57).

2.7.1 Flouts Exploiting the Maxim of Relation

Green (1989:97) assumes that what is apparently irrelevant
is, in fact, relevant as the basis for understanding correctly what is
said. He also believes that most of the implicatures we commonly
make arise from cases where the maxim of relevance looks like it has
 44

been disregarded; that is, cases which involve apparent violations of

the maxim ‘Be relevant’. A typical example of such cases is the use
of a question as a response to a question as shown below:

29. Cecilia: Do you love me more than you

love food, Gerald?
Gerald: Do chickens have lips?

Gerald’s apparently irrelevant question as a response to Cecilia’s

question makes it clear that he loves food above all else since the true
answer to his question implicates not only ‘no’ but also ‘as you ought
to know’. Cecilia realizes that the true answer to Gerald’s question is
obviously negative; therefore, she infers (as intended) that the true
answer to her question is likewise obviously negative. Rhetorical
questions as well as cliched ones (and their positive counterparts like
Is the Pope Catholic?) also work in the same way.
There are also cases where implicatures arise from
conspicuously disregarding the maxim of relevance. Green (Ibid.:98)
notes that in such cases what someone is trying to do is to tell the
other conversant(s) indirectly that what was being said could put its
speaker in an awkward situation (or worse); therefore, he or she
may change the subject abruptly to implicate ‘Don’t say any more
about it. What you said was a faux pas’ which is safer, quicker or
more effective than to openly try to persuade the other conversant(s)
of it. To illustrate this, Harnish (1976:347) gives the following
example:

30. A: Mrs. X is an old bag.

B: The weather has been quite delightful
this summer, hasn’t it?
 45

In the appropriate circumstances where B’s utterance is not relevant

to A’s remark, he has flouted the maxim of Relation to implicate that
they cannot discuss A’s remark or that A has committed a social
gaffe. In the example above B’s utterance can then be said to mean
‘hey, watch out, her nephew is standing right behind you’. Changing
the subject abruptly is not the only way by which the maxim of
Relation is exploited, sometimes a speaker may overtly fail to address
the other person’s goal in asking a question. Thomas’ example
(1995:70-1) below is illustrative:

31. I finished working on my face. I grabbed

my bag and a coat. I told Mother I was
going out … She asked where I was going.
I repeated myself. ‘Out’.

Here, Olivia, the speaker, does answer her mother truthfully, clearly,
etc., her response, however, does not satisfy her mother’s question,
for the mother can see that the daughter is going out, so what the
mother wants to know is where she is going; therefore, Olivia has
failed to address her mother’s goal in asking the question. It is also
possible to say that Olivia has given less information than the
situation demands, hence the example above can also be taken as a
flout of the maxim of Quantity giving rise to the connection that
Smith and Wilson (1979:176) suggest between relevance and
informativeness. In this respect, they write: “Relevance seems to
involve a special type of informativeness. Intuitively speaking, one
remark is relevant to another if the two combine to yield new
information which was not derivable from either in isolation.” This
leads them (Ibid.:177) to propose the following informal definition of
relevance: “A remark P is relevant to a remark Q if P and Q together
 46

with background knowledge, yield new information not derivable

from either P or Q, together with background knowledge alone.”
 47

NOTES TO CHAPTER TWO

1
As cited in Ifantidou (2001:53), Grice (1989:86), however,
uses the term ‘implicature’ so as to avoid having to make choices
between such terms as ‘imply’, ‘suggest’, ‘mean’, and ‘indicate’,
hence, implicature, according to him, is a blanket word since it covers
all these terms.
2
Hofmann (1993:273), however, takes a different view when
he argues that speaker’s intentions are, more often than not, hidden
behind the literal meaning or, in other words, between the lines of
what is said as when a teacher says ‘My, you’re early today!” to a
twenty-minute-late student.
3
It must be noted here that S stands for speaker; H for hearer;
‘uttering U’ for utterance of a linguistic token which may either be a
sentence part, sentence, or string of sentences or sentence parts; and Z
for some belief or volition invoked in H.
4
For Jaszczolt (2002:211), all maxims of Grice’s generate
implicatures since they all provide inferences which go beyond the
literal content of the utterances. Kearns (2000:261), however, has a
different view, for he believes that not all maxims have equal
importance in giving rise to implicatures and that most of the work is
done by Informativeness and Relevance; that is, by the two maxims
of Quantity and Relation which are clearly quite interdependent in
that if you are to obey the maxim of Quantity, you should know
exactly how much information is required, and surely the information
that is required is relevant to the current purposes whereas the
information that is not required is irrelevant to them. That’s why
“Grice expressed some doubt regarding the necessity of the second
maxim of Quantity, pointing to the possibility of subsuming it under
the maxim of Relevance” (Green, 1989:89).
5
The term scalar implicatures was coined by Laurence Horn
(1972) who points out that those are highly systematic Quantity
implicatures associated with certain scalar items where a stronger
item entails (in an appropriate sentence-frame) the weaker item, and
 48

the use of the weaker item implicates that the stronger item doesn’t
hold. As such, scalar implicatures are a special type of generalized
implicatures where the inference is made when the speaker chooses a
word from a scale which has the general form of an ordered set
(indicated by angled brackets) of linguistic expressions denoting
quantities or degrees of attributes which can be graded on some scale
of values from the lowest to the highest indicating informative
weakness and strength as shown below:

Scale of quantity : (strong) < all, most, some > (weak)

Scale of frequency: ” < always, often, sometimes > ”
Scale of coldness : ” < freezing, cold, cool > ”
Scale of likelihood: ” < certainly, probably, possibly > ”

Having such a scale of values at his/her disposal, the speaker in fact

chooses the one that is both truthful (the maxim of Quality) and
optimally informative (the maxim of Quantity) following the CP
(Peccei, 1999:37; Yule, 1996a:41; and Levinson, 1983:133).
6
Language, in such cases, is defined by reference to the
lexicon, set of syntactic rules, rules of semantic interpretation, and so
forth.
7
Garnham (1985:107) asserts that such implicatures cannot be
explained by a semantic theory whose principal concerns are
entailment and truth conditions, for what is conversationally
implicated is not logically entailed by what is said. Thus, it is
important to contrast the term implicature with terms like logical
implication, entailment, and logical consequence which are generally
used to refer to inferences derived solely from the logical or semantic
content. After all, one should not think of implicatures as semantic
inferences, but rather inferences drawn, in Levinson’s opinion
(1983:104), on the basis of “both the content of what has been said
and some specific assumptions about the co-operative nature of
ordinary verbal interaction.”
8
Bach and Harnish (1979:17) think, however, that violation
and clash also produce implicatures capable of being worked out, but
with flouting a maxim, the implicatures generated are intended to be
recognized as intended.