Professional Documents
Culture Documents
1 Introduction 1
1.1 Purpose of the study 2
1.2 Organization 3
2 Theoretical Foundations 4
2.1 Vague language: what is it and why do we need it? 5
2.2 Vague language in different settings 11
2.3 Vague language and education 15
2.3.1 Vague language in the classroom 15
2.3.2 Vague language in language learning and teaching 17
2.3.3 Vague language and learner language 19
2.4 Theoretical frameworks 23
2.4.1 Vague language and the cooperative principle 23
2.4.2 Vague language and Relevance Theory 26
2.4.3 The elasticity of vague language 29
2.5 Concluding remarks 31
3 Methodology 33
3.1 Three approaches 33
3.1.1 Quantitative approach 33
3.1.2 Qualitative approach 35
3.1.3 Mixed methods approach 38
3.2 Naturally occurring data 41
3.3 Data 42
3.4 Data analysis 44
3.5 Concluding remarks 46
4 Lexical Analysis of Vague Language 47
4.1 Subjectivizers 47
4.1.1 I think 48
4.1.2 I guess 53
v
vi Contents
References 200
Figures
Tables
viii
List of Figures and Tables ix
x
Transcription Conventions
[ S1:] Overlap
<SU-f, SU-m> Unknown speaker with gender identified
<P: 04> Pause of 4 seconds
… Pause of 2–3 seconds
<LAUGH> Current speaker laughs
<S8 LAUGH> Speaker 8 laughs
<SS LAUGH> Two or more speakers laugh
(xx) Indecipherable
xi
List of Abbreviations
xii
1
Introduction
Attention has been, and is still being, drawn to vague language use in set-
tings such as work-related interactions (Koester 2007), healthcare (Adolphs,
Atkins & Harvey 2007), and courtrooms (Cotterill 2007), but suffers a
dearth of research in academic contexts. While Ruzaitė (2007) carried out
a useful study in educational settings, the scope of her research is limited
to quantifiers and approximators. This research is one of the few investi-
gations of vague language in English language learning classes with stu-
dents from two vastly different socio-cultural and linguistic backgrounds,
Chinese and Persian-speaking learners of English, compared with L1s.1
1
In this study L1 refers to American English speakers, L2 to Chinese and Persian
learners of English. The equivalent terms NS (native speaker) and NNS (non-
native speaker) are kept if used in the work of other researchers.
Introduction 3
a. How is vague language realized among L1s, the Chinese and the
Persian speakers?
b. How frequently is vague language used, and what are the most flu-
ently used lexical items? How does frequency differ among the three
groups?
c. How and why is vague language strategically mobilized? What dis-
crepancies are discerned among the three groups?
d. What cultural and linguistic factors underlie interlinguistic and
intercultural diversity in vague language use?
1.2 Organization
Conrad & Finegan 1999). The reasons that support this claim are that
in spoken discourse, interlocutors share contextual clues such as facial
expressions, which may not exist in written discourse, and less preci-
sion is required in spoken forms than written discourse (Cook 1989).
Speakers have access to intonation (Brazil 1997), which can help them
clarify what they mean by what they say. None of these reasons for
using vague language in speech implies that in written mode it should
be ignored or neglected (Hyland 1998). Myers (1996) points out that
research into academic discourse demonstrates that writers too use
vague language.
Vague language is elastic. Zhang (2011) states that its elasticity allows
it to stretch in many directions, as required by the interlocutor for
effective communication. The elasticity of vague language is a func-
tion of its lack of specific connotation, so that its interpretation relies
on context and communicative purpose. This allows language users to
make strategic use of vague language to mitigate, emphasize, evade and
so on (Zhang 2011).
robins
eagles
chickens, ducks, geese
penguins, pelicans
bats
In this, some kinds of bird are more birdy than other kinds: robins are
regarded as typical of birds; eagles are less typical than robins; chickens,
ducks and geese less typical than eagles; penguins and pelicans less typi-
cal than chickens, ducks and geese, and bats are counted as hardly a
bird at all. Such a hierarchy is in line with the ‘prototype theory’ (Rosch
1973). The meaning of a vague term, such as birdiness, could be appro-
priately represented by a ranking exemplified by a prototype.
Vague language is defined by Carter and McCarthy (2006, p. 928) as
‘words or phrases with very general meanings which deliberately refer
to people and things in a non-specific, imprecise way’. Williamson
(1994, p. 4869) states that ‘Used as a technical term, “vague” is not
pejorative. Indeed, vagueness is a desirable feature of natural languages.
Vague words often suffice for the purpose in hand, and too much
precision can lead to time wasting and inflexibility’. In the same way,
vague language in this study is used without any negative connotation;
instead it is considered to be an important part of everyday language.
Various terms similar to vague language have been used in linguistic
studies, the most common being indirectness and inexplicitness. These
are not the same. Cheng and Warren (2003) propose a classification in
an attempt to clarify the confusion caused by similar terms, arguing that
‘indirect language’ involves an inferential process through which mean-
ing is created, while the hearer has access to language and context. This
term embodies paradigms such as conversational implicatures (Grice
1975), illocutionary acts (Austin 1962), indirect speech acts (Searle 1968)
and pre-sequences (Levinson 1983). Inexplicitness refers to cases such as
Theoretical Foundations 7
items of reference, that and it, which cannot stand on their own. They
are independent of context, but once used in a specific context they gain
a certain meaning: in other words, meaning is created through ‘joint
construction’ (Cheng & Warren 2003, p. 397) by the participants in the
context in which the words are used. Substitution, deixis, and reference
all fall into the category of inexplicitness. Inexplicitness is considered
a characteristic of a native speaker’s conversation, ‘the degree to which
linguistic behaviour is reliant on context to convey meaning’ (Cheng &
Warren 1999, p. 295). It emerges when a speaker chooses to use ellipsis
and substitution, deictic conversation and reference, relying on context
to convey the intended meaning. Vagueness differs from both indirect-
ness and inexplicitness in that even when used within a specific context,
its property of vagueness is retained. Vague language remains vague in
context, rather than becoming precise. As a linguistic phenomenon,
it is associated with fuzziness, imprecision, and indefiniteness (Ruzaitė
2007; Zhang 1998). Janicki (2002) opts for the phrase ‘incomprehensible
language’ as a broad term which embodies vague language, defining it as
‘words, expressions, formulations, idioms, texts, and the like which are
easy to misunderstand, which are hard to understand, or not possible to
understand at all’ (p. 215), and claims that ‘incomprehensible language’
appears consistently rather than sporadically in conversation. However,
this concept is not equivalent to vague language, which does not disrupt
ease of communication.
Channell (1994) presents cogent evidence that in order for communi-
cation to be effective and successful, speakers need to use vague words
and expressions at an appropriate level: in other words, they need to
be appropriately inexplicit. An important element in what constitutes
a speaker’s communicative competence is the use of vague language,
which is contextually appropriate and understandable. Channell also
states that vagueness in language is not a matter of badness or goodness,
but one of appropriateness, serving a variety of purposes:
Not all vague words are equally vague: the boundaries of conceptual cat-
egories manifested through vague words are vague to different degrees,
with some more vague or more context-dependent than others (Ruzaitė
2007).
Basing their classification on degree of vagueness, linguists have
proposed different categories of vague language. An early version
presented by Crystal and Davy (1975) described it as 1. placeholders,
2. summarizing lexical items, 3. vague generic terms and collective nouns,
4. approximate quantities, 5. words with suffixes. A more recent classi-
fication proposed by Channell (1994) offered 1. quantifiers, 2. approxi-
mators, 3. placeholders, 4. vague references to categories. As with other
terms, there is some overlap and discrepancy in the terms offered by
linguists: what is called ‘vague references to categories’ by Channell is
called ‘general extenders’ by Overstreet (1995) and ‘summarising lexical
items’ by Crystal and Davy (1975), while Stenström (1944) uses the term
hedges to refer to approximators.
Vague language functions as a hedging strategy in making a claim.
Prince, Frader and Bosk (1982) state that ‘plausibility shields’ (p. 90), as
they call vague expressions, function to protect speakers from a full or
personal commitment regarding the truth condition of an utterance.
Similarly, Brown and Levinson (1987) claim that an application of
vague language can maintain the negative face of the interactants and
smooth the progress of a difficult conversation. Myers (1989) argues
that to be on the safe side when making claims with regard to new
research findings, authors of scientific writing employ hedges to indi-
cate a potential lack of certainty. Erev, Wallston and Neal (1991) likewise
find that vague communication in tasks which demand cooperation
between group members reinforces the sense of cooperation. Hamilton
and Minoe (1998, p. 6) maintain that ‘imprecise language can facilitate
a polite exchange between source and receiver. A precisely worded mes-
sage might come across as too personal, threatening a receiver’s self-
esteem. Vague language allows the preservation of face’. Along the same
line, Metsä-Ketelä (2006, p. 123) argues that ‘when it comes to interac-
tion, vague language functions as a marker of politeness and unreserved
atmosphere’. Given these many necessary functions, the investigation
of vagueness is as important as the study of preciseness.
There have been two studies of how vague language originates in
communication. The first focuses on language itself. Ullman (1962,
p. 118) refers to a number of factors of the origin of vague language
including ‘(a) generic character of words; (b) meaning is never homo-
geneous (that is, it is context-bound); (c) lack of clear-cut boundaries
Theoretical Foundations 9
in the non-linguistic world; (d) lack of familiarity with what the words
stand for’. He assumes of factor (a) that a word refers to a broad term,
one that is not a single entity but a class of items or events that have
some elements in common. In (b), meaning needs to be interpreted
with reference to context: that is, the context specifies meaning. In (c),
the world that the word represents is vague by nature. To consider how
this works, ‘ask oneself when a hill becomes large enough to qualify as
a mountain, or at what precise age a girl starts to be correctly referred
to as a woman’ (Channell 1994, p. 7). Factor (d) refers to uncertainty of
what is being talked about.
The second focus takes a psychological perspective. Deese (1974)
maintains that vagueness exists in the structure of ideas rather than
in the language system, arising from the ideas that express language
rather than from the language itself. Crystal and Davy (1975, p. 11)
put forward four reasons for vagueness, or ‘lack of precision’ as they
call it:
Jucker, Smith and Lüdge (2003, p. 1765) believe that the most obvious
reason for vague language use is ‘uncertainty at the time of speaking.
Sometimes speakers lack information about a given quantity, quality
or identity. They, therefore, cannot be more precise even if they want
to’. Cutting (2007, pp. 3–16) claims that if speakers are exhausted or in
a hurry so that they can’t find the right word, a vague expression may
yield implications that are contextually more relevant for the hearer
than the exact words are.
Regardless of the lack of an agreed definition, various classifications
of vague language have been proposed. Walsh, O’Keeffe and McCarthy
(2008) divide vague categories into lexical and non-lexical types. Lexical
categories, or what Channell (1994, p. 123) calls ‘common categories’,
are items which have a graded structure and which can have a proto-
type, as we have seen with ‘bird’. Non-lexical categories, ‘vague cat-
egory markers’ (VCM) (Walsh et al. 2008) or ‘vague category identifiers’
(Channell 1994), refer to ad hoc items, which are the by-products of
interaction. Examples include vague tags such as (cloth) and that kind of
10 Communicating through Vague Language
thing, (money) and things like that, which allow the listener to infer what
is meant by the speaker.
All languages, whether solely spoken or comprising both spoken and
written modes, use a variety of components to express vagueness, and
those used in speaking outnumber those used in writing (Metsä-Ketelä
2006). The rich verbal vagueness arises from the existence of semantic
vagueness, and also from concepts lacking clear definitions, thereby
necessitating imprecision. Take the distinction between a hill and a
mountain as an example: since there does not exist a clear-cut bor-
derline splitting these two concepts, a distinction between when a sea
becomes an ocean or a kitten becomes a cat seems highly unlikely. The
second factor contributing to imprecision in language is the existence of
phenomena such as metaphor, ellipsis, euphemism and pronoun refer-
ences (Metsä-Ketelä 2006). These two factors are two examples among
many other contributing factors for vagueness in language.
Precision can sometimes create inappropriateness, which a vague
expression can avoid (Jucker et al. 2003). Tannen (1989) claims an
increase of precision may have adverse effects. In other words, inap-
propriate use of detail can be boring, as can be witnessed when young
people listen to old people give more information about their daily rou-
tine than the hearers need to know. This can become insulting if used
in criticism. On the other hand, a precise statement can sometimes be
fuzzier than a vague statement. In his psychologically oriented study on
precision and vagueness in what he calls a preciseness paradox, Teigen
(1990) concludes that precise language suits any circumstance involv-
ing discussion in the past or present tense, but vague language is more
appropriate for any kind of future prediction. Past and present involve
more precision, but using precise language for the future can prompt
skepticism. For example, ‘Tomorrow at two o’clock the mountain will
erupt and 5,000 people will be killed’ vs. ‘That mountain is going to
explode any day now’. The latter statement is more believable than the
former one, because the latter resorts to vague language rather than pre-
cise language. Similarly, Moxey and Sanford (1993) state that ‘it would
appear that if one is looking for reasons to have faith in a proposition,
then specificity suggests expertise, which in turn meets that criterion. In
contrast, if one is looking for reasons to be sceptical, then precision may
signal suspicion’ (p. 16). However, Teigen rejects the trade-off between
confidence and skepticism, arguing that some features that consolidate
confidence can also contribute to doubts. It seems that the relation-
ship between confidence and precision or doubt and imprecision is not
straightforward, rather multi-faceted.
Theoretical Foundations 11
indicate that medical science does not yet have a thorough understand-
ing of such diseases. This demonstrates that the use of vague language in
such cases is an appropriate acknowledgment of their limitations,
rather than an indication of undesirable imperfection. Adolphs et al.’s
(2007, p. 74) analysis of vague language in UK’s National Health Service
direct phone-ins shows that it also mitigates the force of requests for
embarrassing information, or orders to undergo unpleasant regimens
of treatment. Since the consultations are conducted on the phone, the
discourse involves a higher level of vague language use, than in face-
to-face patient–nurse interactions, which demand less. Vague language
helps the invisible participants feel socially closer and, while giving
the patient a clear idea of ‘the serious nature’ of the topic, helps keep
the atmosphere relaxed. As Adolph et al.’s findings illustrate, vague lan-
guage serves specific and significant roles, evidence that it has special
communicative purposes in medical settings.
Factors such as speakers’ linguistic and cultural background also play
determining roles. In a study of native and non-native speakers, Cheng
and Warren (2001) find that rather than creating confusion and misun-
derstanding, vague language can, in a broad sense, enhance friendliness
and reinforce a ‘cooperative tone of exchange’, and create informality
in conversation. It also has ‘more specific uses, normally classification,
compensating for a lack of vocabulary (as an accommodation strategy
and as an avoidance strategy), compensating for a lack of knowledge,
politeness and finally “self-protection”’ (p. 86). An instance of accom-
modation strategy is when native speakers adjust their language to a
non-native speaker’s level by using short sentences, simple structures
and commonly used words: ‘varying the degree of specificity is one way
in which NS accommodate NNS, and whether this means using more or
less vague language will depend on contextual factors such as the NS’s
perception of the NNS’s linguistic ability’ (p. 94). This demonstrates
that vague language can be used to create a friendly atmosphere when
it is needed.
Another favorable function of vague language is that by enhancing
mutual tolerance between interlocutors, it makes the audience an active
participant in the process of communication. It, therefore, serves as a
hearer-involvement device (Ruzaitė 2007), using vague expressions to
urge active and attentive participation from the interlocutor to con-
struct the meaning of the message.
The study of vague language has a wide application in various lan-
guages. There have been important works based on Chinese language.
Wu’s (1979) seminal work introduced Zadeh’s (1965) fuzzy set theory
Theoretical Foundations 13
two-way portals. For the teacher, they can open a door to what is
likely shared knowledge for this phase of the lesson and create a
shared space around this commonage. For the learner, they open a
door to a space where it is safe to take risks. Tentative propositions
can be marked using VCMs and loss of face is avoided. In this mode,
they engage cooperative listenership on the part of peers which also
facilitates learning.
(Walsh et al. 2008, p. 26)
In the last stage, the classroom context mode, the local context deter-
mines the management of turns and topics and there are abundant
opportunities for communication, which allows students as much time
and space for interaction as they need. The teacher is mainly a listener
and promotes interaction. The goal here is for students to extend dia-
logue and discussion, so they are given the chance to express themselves
and participate in the discussion and give long responses. A vague cat-
egory marker at this stage behaves similarly to the way it functions in
daily conversation, since ‘it acts as an “involvement device” ensuring
listener participation and prompting equity and understanding’ (Walsh
et al. 2008, p. 25). While Walsh et al.’s study is narrow in scope as only
vague category markers are investigated, it sheds light on the role of
vague language in educational settings in that L2 learners’ vague lan-
guage use is similar to the use in day-to-day conversations.
Although it seems that English for Special Purposes includes scientific
writings made up of a series of objective statements regarding facts,
vague expressions are frequent in scientific journals and play significant
roles in academic writing (Hyland 1996, 1998). In academic discourse,
writers may be able to express a proposition with more precision by
using vague language, as quantifying the world precisely is almost
impossible. To present information as accurately as possible, a writer
Theoretical Foundations 17
1
This study has a different view in terms of ‘underuse’ or ‘overuse’. The position
of this study is that as far as the frequency of language use is concerned, the
norm of L1s is not necessarily the norm that L2s have to adhere to. For various
reasons L2 speakers may use certain types of expression more or less, and this
is a legitimate discrepancy between L1 and L2 speakers rather than a matter of
‘overuse’ or ‘underuse’. For a detailed discussion of this issue, please see Zhang
and Sabet (in press).
20 Communicating through Vague Language
a. Quality:
As far as I know, …
b. Quantity:
As you probably already know, …
c. Relation:
Oh, by the way, …
d. Manner:
This may be a bit tedious, but … (pp. 26–27)
Relation:
John: What’s the time?
Mary: The museum hasn’t opened yet.
[Implicature]: It’s at least before whenever the museum normally
opens.
This seems to flout the Maxim of Manner, in that it does not provide
a clear, straightforward answer to the question. Huang argues that
the Maxim of Relevance here leads to conversational implicature.
Mary’s answer is relevant, assuming Mary is cooperatively answering
John’s question; so ‘we can infer that while Mary is not in a position
to provide a straightforward answer, nevertheless she thinks that the
museum’s not being open yet might help John to get a partial answer’
(Huang 2007, p. 29).
Theoretical Foundations 25
Maxim of Quality:
Abdullah: Mumar is engaged to three different girls.
Susan: Oh, he’s a real monogamist.
[conversational implicature] Mumar is no monogamist at all.
The listener will know that the statement is not true, and assumes that
the maxim of quality is being violated to maintain the cooperative
principle. The hearer will also tend to assume that the speaker means
something quite different from what has actually been said, something
beyond the literal meaning of the statement: ‘the ironic meaning’
(Huang 2007). Vague language use does not necessarily flout the maxim
of quantity because it is often used for succinctness, saying no more and
no less than the context requires, as often there is no need to say more
(Zhang 2005).
Conversational implicature is classically a type of figurative language
(Sperber & Wilson 1986a, p. 155), in that the speaker violates the coop-
erative principle on the assumption that the hearer is able to understand
the implied meaning, because ‘When the literal interpretation is inap-
propriate, the appropriate figurative interpretation somehow comes to
the hearer’s mind’. Channell (1994, p. 33) mentions that when she is
asked about the time to be home from work and she does not know, the
most truthful answer would be ‘about six o’clock’ (a vague expression).
She assumes that from this answer her hearer will infer that she cannot
say exactly: that is, she is trying to be as truthful as possible (Maxim of
Quality) and is, to the best of her ability; but her answer violates the
maxim of manner in that she does not clearly address the question.
Grice’s framework has met with some criticism. Davies (2007, p. 2310)
has concerns that when speakers flout conversational maxims and cre-
ate conversational implicatures, conveying their intentions in a veiled
manner, it makes the task of the hearer more difficult. It would seem
that the cooperative principle is not about making the task of the hearer
straightforward; it can be quite the reverse. However, Davies’s concern
is a non-issue in the case of vague language use, as a not-so-precise
26 Communicating through Vague Language
relevance theory principles are not like rules to be followed and obeyed
in communication, but ‘are an automatic reflex of the human men-
tal capacity that works without the communicators having any overt
knowledge of it’ (Huang 2007, p. 202).
In the framework of relevance theory, pragmatics is regarded as a
single notion of relevance, manifested in two principles: the cognitive
principle of relevance and the communicative principle of relevance.
As visualized by Sperber and Wilson (1995), relevance is a measure
consisting of cognitive effect and processing effort. Cognitive effect
refers to the interaction of a new input and a set of assumptions already
existing in a cognitive system, and processing effort addresses the effort
spent by that cognitive system to produce an appropriate interpretation
of incoming information ‘geared to the maximization of relevance’
(Sperber & Wilson 1995, p. 252). The relevance of an input to the
person is a matter of balance between cognitive effects (benefit) and
processing efforts (cost):
a. Other things being equal, the greater the positive effects achieved
by processing an input, the greater the relevance of the input to the
individual at that time.
b. Other things being equal, the greater the processing effort expended,
the lower the relevance of the input to the individual at that time.
(p. 252)
The optimum is to achieve the greatest cognitive effect using the least
processing effort. As for the communicative principle of relevance
(ostensive-inferential communication), it is not true that in the process
of communication we are absorbing every possible input and scan-
ning it for relevance, because this would make communication quite
difficult. Relevance theory assumes that communication contains two
kinds of information: that which the speaker wishes to transmit, and
that which conveys the speaker’s intent to inform the audience of this.
In other words, ostension and inference are the two poles of communi-
cation. Ostension informs the communicator’s point of view and infer-
ence informs the audience’s perspective; the communicative principle
of relevance is met when the ostensive stimulus is relevant enough to
be worth the audience’s processing effort, and is the most compatible
with the communicator’s abilities and preferences (Sperber & Wilson
1995, p. 254).
Cheng and Warren (2001, p. 93) assert that ‘Since vague language
seems to be easier to process and makes fewer demands on the listener,
28 Communicating through Vague Language
it is probably also the case that a speaker may choose to use a greater
amount of vague language to make the discourse easier for the hearer(s)
to understand’. A speaker will skip technical words or ‘specialized lan-
guage’ that the listener doesn’t know and employ a simplified language
instead.
Jucker et al. (2003), who adopt relevance theory in their study, state
that vague language enables speakers to maintain fluency when some
information is not immediately available to them. When this happens,
the speakers ‘may then decide that the processing cost of accessing it,
and the cost to fluency, are not warranted in terms of any benefits to
be gained by precision. However, speakers may choose vague expres-
sions even when they could have stated their utterances more precisely.
A vague utterance may be more efficient in the sense that it yields the
same contextual assumptions for lower processing cost’ (Jucker et al.
2003, p. 1765).
Zhang (2005) argues that relevance theory rightly supports the non-
numerical approach to vague language, but that the numerically and
semantically oriented approach has its own merits. She argues that
Zadeh’s (1965) fuzzy set theory and relevance theory have convergent
and divergent principles: both give priority to optimality, relying on
what is best or most suitable in a situation, but the former adopts a
quantitative approach, emphasizing mainly the semantic aspect of
meaning with numerical values, whereas relevance theory insists on a
cognitively oriented approach to interpretation, which is useful because
‘sometimes we don’t know or cannot agree on the exact numerical
value for fuzzy expressions’ (p. 80). More importantly, there are occa-
sions when we know the numerical information but prefer to use vague
expressions for such reasons as withholding information or safeguard-
ing ourselves. Zhang asserts that what determines the realization of
optimal relevance is not the option for fuzzy or non-fuzzy forms of
language, but the communicator’s perception of the relevance of the
utterance in question.
A similar concept to vague language proposed in relevance theory is
loose talk. Sperber and Wilson (1986a) maintain that loose talk is a type
of non-literal use ‘based on resemblance relations among representa-
tions, and involve interpretive rather than descriptive dimensions of
language use’ (p. 164). When someone loosely understands a proposi-
tion or concept, this does not mean that the concept or proposition is
vague; nor does it indicate that the proposition comes with a guarantee
of approximate truth. ‘Instead, certain of its logical and contextual
implications are taken to be accompanied by regular guarantees of
Theoretical Foundations 29
These four maxims serve various pragmatic functions. They are not an
exhaustive list but should cover a majority of vague language uses. They
are interconnected by the fact that all can be used strategically to serve
functions that precise language is unable to perform. Each has a differ-
ent priority. For example, ‘go just-right’ stresses the approximated infor-
mation and ‘go subjective’ concentrates on the speaker’s attitudinal and
stance information. Zhang also states that vague language elasticity has
the following three major characteristics:
The review of literature in this chapter has looked into the different
paradigms of vague language research, which provide the required
theoretical foundations for this study. Vague language can be looked at
from different perspectives, including Grice’s cooperative principle and
Sperber and Wilson’s relevance theory. Zhang’s framework of elasticity
offers a more promising approach for this study.
32 Communicating through Vague Language
Quantitative Qualitative
This study falls under the first of these categories, with the quantita-
tive section dealing with lexical analysis in the form of the frequency
of occurrence of vague language. This was conducted before the quali-
tative study of the functional use of vague expressions across three
groups. However, quantitative and qualitative methods should not be
40 Communicating through Vague Language
completely separate: at the very least the qualitative work in this study
should validate the findings of the quantitative analysis. This study
also fits in with the last type of design, as the qualitative analysis is
dominant, mainly because the data were collected through a qualitative
approach.
A concept closely associated with mixed methods approach is trian-
gulation. Creswell (2008, p. 553) describes this as an investigator being
able to ‘improve his investigation by collecting and integrating vari-
ous kinds of data on the same phenomenon’; the three points of the
triangle are ‘the two sources of the data and the phenomenon’. The
idea of triangulating data sources as a way to create convergence across
quantitative and qualitative methods was first developed by Jick (1979).
Merging the two and using the results to better understand a research
question.
Creswell (2008) claims that the strength of the triangulation design
is that it takes advantage of the strengths of each type of data collec-
tion. However, its popularity cannot disguise potential problems. As it
involves both quantitative and qualitative data, it demands hard work
and a high level of expertise in both methods. The researcher may
find that the results of one approach contradict the other, requiring
the collection of fresh data, which may be difficult to obtain. Another
challenge is how to integrate the two different data sets so that they are
comparable (Creswell 2008). In any case, integrating the results of very
different methods to answer research questions is challenging, and ‘like
the triangulation design, the simultaneous data collection of quantita-
tive and qualitative data may be labor intensive for a single researcher’
(Creswell 2008, p. 559).
The other mixed methods design is the embedded design. The simi-
larity between this and triangulation is that both involve concurrent
quantitative and qualitative data collection. The difference is that for
the embedded design, one form of data makes up the primary source
while the other counts as a supportive source. In other words, the
researcher gives more weight to one and counts the other as com-
plementary evidence (Creswell & Plano Clark 2007). In an embedded
design the researcher can adopt a one-phase or a two-phase approach,
using quantitative and qualitative data to answer different research
questions (Hanson et al. 2005). In this design one method is considered
a component of another, answering questions which the primary source
has failed to address adequately. Creswell and Clark (2007) state that
the strength of the embedded design is that it gives the researcher the
opportunity to take advantage of two methods in a single study. It can
Methodology 41
There are two means of gathering natural data: compiling field notes of
real life data and tape-recording. With regard to the first means, Tran
(2006, p. 3) states, ‘In this ethnographic method, researchers observe
real-life interactions and take notes of natural data on the communi-
cative acts in focus’. The latter involves audio- or videotaping social
interactions as in conversation analysis, to capture communicative acts
in progress.
The most salient advantage of tape-recording is that the data is natu-
ral and represents discourse features. It frees the researcher from note-
taking, so there is no need to rely on memory or selective attention; as
a result the reliability of data collection rises. Tape-recording allows the
researcher to replay the data and improve transcription; more impor-
tantly, it preserves the sequence of talk.
Despite these advantages, this method has met with some criticism.
The degree of the researcher’s control of social variables concerning the
interlocutors is low. Beebe and Cummings (1996, p. 81) note that ‘many
studies of natural speech have not given us scientifically collected sam-
ples that represent the speech of any identifiable group of speakers.
They do not give us situational control’; thus, conversation analysis
may yield unsystematic data.
The researcher may run the risk of not getting enough data regarding
the communicative acts under investigation after recording authentic
interactions for a long period. Another disadvantage is that an audio
recording may harm the confidentiality of the respondents, if not han-
dled carefully. As their exact words and voices are recorded, they run the
risk of being identified or their secrets being disclosed. The participants
42 Communicating through Vague Language
may also feel nervous if there are cameras around, and act unnaturally.
The remedy to this problem is to keep the recording device as unobtrusive
as possible. The equipment need not be concealed, but should be located
where it does not distract respondents and its presence is not confronting.
Vague language is an indispensable part of natural speech, so the
possibility of not finding vague expressions in interactions through
conversation analysis is relatively low. The data collected for this study
offered a rich resource of vague language, and the application of a sim-
plified version of the transcription conventions of conversation analysis
provided a good platform for its analysis.
3.3 Data
were maximally comparable and valid. These measures mean that the
reliability of the data collection in this study is high, and the use of
video-recording has provided the researchers with the opportunity to
improve the accuracy of transcription.
The data in this study was analyzed according to the categories shown
in Tables 3.2 and 3.3, which present forms of vague language realization.
As Table 3.2 shows, in this study vague language is analyzed in five
lexical categories: subjectivizers, possibility indicators, vague quanti-
fiers, vague intensifiers, and placeholders. The term ‘placeholder’
is borrowed from Channell (1994), but with much wider scope in
this study. The lexical investigation is at micro level, particularly
corresponding to the first two research questions (a. and b.) listed
in Section 1.1: the realization and frequency of vague language in
the data, and how its form and distribution differ among the three
groups.
The lexical analysis was conducted using Wordsmith Tools (Mike Scott
2010 version 6.0, for example concordance function) for the Chinese
and Persian speakers and L1 data. This program was used to acquire
information regarding the types and frequency of vague expressions
used and to provide information on the most and the least used vague
expressions and vague language clusters (vague expressions + other
collocated words). A Chi-square test was applied to statistically examine
the significance of differences among the three groups.
While the lexical level analysis in Table 3.2 is a micro-study, the
pragmatic analysis set out in Table 3.3 is a macro-analysis. This study
investigates three major functions of vague language: mitigation, giving
This chapter presents the lexical analysis of the five vague language
categories by examining the frequency of vague expressions and their
clusters. A cluster refers to the grouping of a vague expression with
adjacent expressions.
4.1 Subjectivizers
Note: All tables are ranked according to the frequency of the item in L1 interaction, and this
table is adapted from Zhang and Sabet (in press: 7).
distribution. The Chinese were again very different. They unevenly used
the four subjectivizers, concentrating disproportionately on I think (732
tokens, 99 per cent of the total subjectivizers used). The remaining three
items combined made up only 1 per cent, indicating an extremely focused
preference. It seems that the Chinese group required nothing but I think
to express a stance or attitude. While L1s and the Persian speakers also
preferred I think, they demonstrated similarity in the ranking order of the
other subjectivizers, which the Chinese barely used.
4.1.1 I think1
As Table 4.1 shows, the most remarkable difference in the use of indi-
vidual subjectivizers among the three groups emerges in the most
frequently used expression, I think. The Chinese showed a strong prefer-
ence for it, while the other two groups used it less consistently. The L1s
used it four and a half times (161 tokens) less and the Persian speakers
three and a half times (207 tokens) less than the Chinese (732 tokens).
The heavy use of I think may be associated with L2 learners’ inadequate
language proficiency and their discourse management, such as delay-
ing, searching for words and so on.
The three groups differed in positioning I think in a clause, par-
ticularly in clause-initial position and as a turn-initiating device. The
Chinese used I think in clause-initial position, 620 (85 per cent) times,
the Persian speakers 165 (80 per cent) times, and the L1s 122 (76 per
cent) times. The Chinese preferred I think in clause-initial position
roughly five times more than L1s, and four times more than the Persian
speakers, who used it about 35 per cent more than the L1s did. I think
in the turn-initial position was used 256 times by the Chinese, 89
by the Persian speakers, and 52 by the L1s. The L2 groups both used
1
A version of the data and some data analysis in this section has been published
in Zhang and Sabet (in press).
Lexical Analysis of Vague Language 49
2
Ch = Chinese data, 4 = session number, 254 = speaking turn count, S1 = Speaker 1.
50 Communicating through Vague Language
over I think I is that they tended to pay more attention to the hearer
and were less authoritative. The L1s preferred I think I over I think we,
indicating that they tended to be more self-oriented and assertive.
The data reveal two opposing clusters involving I think: I think that
is vs. I think that + subject. I think that in this study is divided into two
categories according to the function of that: the first is where that serves
as the subject (pronoun) of the sentence, followed by is as the verb; the
second is where that serves as a conjunction. The first usage is demon-
strated below.
(4.2): This is a discussion between two L1 participants over two turns.
They are talking about whether newspapers should be granted access to
people’s photos.
In (4.2) that is a subject. The Persian speakers used this cluster only
twice (1 per cent), the Chinese 16 times (2 per cent), and the L1 group
19 times (12 per cent), a much higher percentage than either L2 group.
In this case, the two L2 groups were similar to each other but different
from the L1, providing a case of L1 vs. L2.
The second use is illustrated in (4.3), where the function of that in
I think that shifts from a subject to a conjunction introducing a clause.
(4.3): This is a discussion between two Persian-speaking participants
over two turns. They are talking about problems that people experience.
S5: … Actually by ‘we’ I mean all the people living in the world. You
know I think that people in the world suffer from spiritual crisis.
(P: 4: 48)
S1: There might be a special crisis in the world. What is happening in
Iran? (P: 4:49)
In this excerpt, that brings in a clause in the form of I think that + subject,
performing a different role from that of I think that is. The Persian speak-
ers used this pattern 16 times (8 per cent), more than either the Chinese
(eight times, 1 per cent) or the L1s (four times, 2 per cent). When the
function of that shifted, the preferences of the Persian speakers and L1s
swapped, so that the Persian speakers used it most and the L1s least.
Lexical Analysis of Vague Language 51
S7: Ok, first of all, I must say the culture. (P: 6:442)
S3: I think it is not cultural. I think it is not cultural, whereas the L1s
prefers. (P: 6:443)
S2: The press has already put heat on ’em they’re putting heat on
themselves. I don’t think it’s necessary. (L1: 1:61)
S1: Well they’re obviously not having enough heat put on them
because it keeps happening. [S2: I think (xx)] I mean for at least it’s
possible. Yes? (L1: 1:62)
The I don’t think pattern occurred 22 times (14 per cent) in the L1 data
and 19 times in the Chinese (3 per cent), while the Persian speakers made
use of this cluster six times (3 per cent). An interesting phenomenon is
that the Chinese used both patterns (I don’t think plus I think + not) at
similar rates, making any attempt to explain the patterns in relation to
L1 and L2 distinction murky (see Zhang & Sabet (in press) for a more
detailed discussion of the two patterns).
The occurrence of the conjunction but followed by I think expresses
a contrast. The Chinese data offered 33 occurrences (5 per cent), more
than the Persian speakers (23, 11 per cent). The least use of I think but
was among the L1s, seven times (4 per cent). Another cluster of I think
is discourse marker + I think: for example, I mean I think. No such clus-
ter appeared in the two L2 data at all, but the L1s used it seven times
(4 per cent). The other cluster is you know I think. This time the L1 group
did not use the cluster, while the Persian speakers used it nine times
(4 per cent) and the Chinese twice. The Persian speakers did not use
I mean I think, but used you know I think; the L1s did not use you know
I think but used I mean I think. These two discourse marker + I think
52 Communicating through Vague Language
clusters were used for discourse management, for instance to repair the
speaker’s speech. You know I think also performed other functions, such
as drawing the attention of the hearer, eliciting the hearer’s agreement,
and creating informality. The Chinese were not inclined to use either
expression. The evidence of the use of these clusters suggests that the
L1 and Persian groups tend to go opposite ways, while the Chinese take
the middle ground.
In this study, for tables where a total does not appear in the last row,
one item may be classified under more than one cluster. Thus the total
of a column may sum to more than 100 per cent.
The results in Table 4.2 show two patterns emerging: L1s vs. L2s, and
L1 vs. the Persian speakers, with the Chinese in the middle. L1s and L2s
preferred opposite I think clusters, as shown in items 1, 2, 3, 5, 8, 10,
and 14. For example, in item 14, L1s used I mean I think seven times
(4 per cent), but no use occurred in any of the L2 conversations. On the
other hand, both L2 groups used I think (item 1) in clause-initial posi-
tion more than the L1s. Another interesting phenomenon is that the
L1s contrast with the Persian speakers, as in items 6, 7, 9, and 13. For
example, the cluster in item 7, ‘I think that is’, was used most by the L1
group and least by the Persian. In item 13, the situation is reversed. In
both cases the Chinese group sit in the middle, using each particular
item less than one group but more than the other. The remaining three
cases in Table 4.2 (items 4, 11, and 12) were used most by the Chinese
group.
Table 4.2 illustrates that the L2 groups preferred to put I think in the
beginning of a clause or a speaking turn more than the L1 group, and
also to put I think in the end of a clause where it takes a mitigating func-
tion. The L1 group used more I think I but less I think we; the Persian
speakers did the reverse. I think we may suggest a sense of collaboration.
The L1 group solely used I don’t think, while the Persian speakers and
to a lesser degree the Chinese showed a strong preference for I think +
negative sentence—a less authoritative replacement for I don’t think.
Based on the evidence of the use patterns of items 3, 4, 5, 8, and 9,
it appears that the Persian group was more tentative and cooperative,
and less authoritative, than the L1 group. The evidence points to the
Chinese group as having elements of both traits: more certain than the
Persian speakers but less affirmative than the L1s. The Persian group
also preferred the formal usage of I think that + subject more than the L1
and Chinese groups, but both L2 groups stayed away from the structure
I think that is. This is supported by other evidence: the Persian speakers
used more but I think than the L1 group to gently move to a contrastive
argument; and used more you know I think to create intimacy before
expressing an opposing viewpoint. The Persian speakers stand out in
their use of I think clusters, in that they tended to use them to express
disagreement or contrast without confrontation (see Chapter 6 and
Zhang & Sabet (in press) for more discussion of the issues raised in this
paragraph).
Similarities as well as discrepancies were revealed among the three
groups, but with more distinctive differences between L1 and L2, par-
ticularly between the L1 and Persian speakers.
4.1.2 I guess
Compared with I think, I guess seems to be more informal and to
suggest a greater degree of tentativeness. This was the Persian and L1
speakers’ second most frequently used subjectivizer (41 tokens, 15 per
cent; 23 tokens, 11 per cent) in this study. It was also the least common
for the Chinese, with a frequency of one: a huge contrast with their
preference for I think.
I guess was used by the Persian speakers in clause-initial position 15 times
(37 per cent), far ahead of the L1s who used it five times (22 per cent).
54 Communicating through Vague Language
It was used in clause-final position seven times (17 per cent) by the Persian
speakers, and only twice by the L1, so the Persian speakers again used it
most; the Chinese provided only one occurrence. The use of I guess after
conjunctions for both the Persian speakers and L1s were almost equal (five
for L1, six for Persian), demonstrating a similarity in combining I guess
with conjunctions; however, the overall distribution of I guess reveals that
substantial intergroup disagreement exists in terms of both frequency and
pattern of use.
S1: …. now if you still wanna enter in, I don’t know maybe this
is gonna, be what set six set seven set set eight, I don’t know
where it’s gonna end, but let’s say it ends at set thirteen? And ….
(L1: 3:53)
S11: Where would you put the parenthesis in the second line?
(L1: 3:54)
In (4.6), the first instance of I don’t know seems a shield, implying the
speaker is unsure about where it may end. This uncertainty is reinforced
by the use of maybe, suggesting that the speaker has some vague idea
about where it may end, in the range of set six to set eight. However,
the speaker is not entirely sure: I don’t know and maybe hedge this sug-
gested end. The second I don’t know appears to be literal: at this point
the speaker declares no knowledge of the issue in question.
(4.7): This is a discussion between two L1 participants over two turns.
They are referring to morals as an example of how a large society is
structured.
S1: … um, it’s not immoral, to cut class well maybe it—I mean you
could ma- maybe someone could make an argument like, I don’t
Lexical Analysis of Vague Language 55
know your parents paid all this money and, you made a promise
to them to go to class and so it’s immoral to cut class but, um, let’s
say uh, let’s say for the sake of argument you know. (L1: 2:68)
S2: The grade. (L1: 2:69)
S2: Meredith I don’t know who you are, or where you were sitting.
(L1: 3:556)
S24: I’m Meredith. (L1: 3:557)
4.1.4 I believe
It is commonly perceived that among the four subjectivizers listed in
Table 4.1, I believe is the most formal and the least used. This could be
because the three data sets in this study are of spoken and face-to-face
interactions, so a formal word would be somewhat less appropriate than
an informal word. Both the L1 and the Persian speakers used I believe
eight times (4 per cent and 3 per cent, respectively), showing some con-
sistency. The Chinese used it three times. This trend is in line with the
common perception.
Further examination reveals that the Persian speakers used I believe
five times in clause-initial position and three times in clause-medial
56 Communicating through Vague Language
600
500
400
300
161 207
200
100 41
23 13 26 8 8
1 5 3
0
I think I guess I don’t know I believe
The Chinese also used I believe less frequently than the other two
groups.
Figure 4.2 illustrates the percentage distribution of the subjectivizers
by the three groups. A striking similarity lies in the percentile rank-
ing of the Persian speakers and L1s (in descending order: I think,
I guess, I don’t know and I believe); the Chinese group demonstrates a
totally different pattern. The percentage values of the Persian
and L1 data are quite close, but as the Chinese did not use I guess
and I believe in their interactions, their entire use of subjectivizers
differed substantially. I think comprises 73 per cent of the Persian
speakers usage and 79 per cent of L1, but the Chinese group’s heavy
use of I think reached 99 per cent of the subjectivizers used in their
interactions. Of the other three subjectivizers, the Persian speakers
used more I guess and I don’t know than the L1s: 15 per cent vs. 11 per
cent and 9 per cent vs. 6 per cent. For I believe, the situation was
reversed, with the L1 group using this more than the Persian speakers
(4 per cent vs. 3 per cent).
What is shown from the summaries in Figure 4.1 (raw frequency)
and Figure 4.2 (percentage frequency) is that there is a discrepancy in
L1 Speaker CSLE
4% 0% 1%
6% 0%
PSLE
3%
9%
I think
15% I guess
I don't know
73% I believe
4.2.1 Maybe
A similarity among the three groups in using maybe is that all three
preferred it over any other possibility indicator. As Table 4.3 shows,
Chinese used maybe 312 (82 per cent) times, the highest frequency,
about twice as much as the Persian speakers (156 times, 81 per cent) and
five times as much as the L1 group (64 times, 26 per cent). Maybe is the
only item in the category of possibility indicators that the Chinese and
the Persian speakers used more often than the L1s. Except for maybe,
the L1 group used all other categories more than the L2s, including may,
might, probably, and possible.
There were various positions where maybe was placed. In clause-initial
position, it occurred 128 times in the Chinese data, 85 times in Persian
and only 11 times in L1 data. It is obvious that L2 speakers used it about
ten times more than L1s. Turn-initial maybe also was used differently
by the three groups, with the pattern following that of clause-initial
maybe: the Chinese used the most (59), the Persian speakers second-
most (44), and again the L1s least (five); again the L2 speakers’ use was
approximately ten times that of the L1s.
The discrepancy was smaller in the cluster of conj + maybe. The L2
groups used it similarly (24 and 23, respectively), and the L1 group
13 times. This time, the difference between L1 and L2 groups was double,
much smaller than with maybe in clause and turn initial positions. The
pattern emerging is that the Chinese preferred all three maybe positions
and clusters the most, the L1 preferred them the least and the Persian
speakers sat in the middle. The similarity between the two L2 groups
was bigger than between L1 and L2s.
The Persian speakers used 55 per cent of maybe in clause-initial posi-
tion, but only 15 per cent for conj + maybe. The Chinese used 41 per cent
of maybe in clause-initial position, but only 8 per cent for the cluster
conj + maybe. In contrast, the L1 group used 20 per cent conj + maybe
but only 8 per cent maybe in turn-initial position. The gaps for each
group between the most and the least preferred are: 40 per cent for the
Persian speakers, 33 per cent for the Chinese, and only 12 per cent for
the L1s. This again supports the trend of L1 using vague expressions
more evenly than L2 speakers, the same trend encountered in their use
of subjectivizers, as shown in Section 4.1.
The following excerpts show how maybe was used. Both the Chinese
and Persian groups frequently placed maybe at the beginning of a sen-
tence, and in situations when taking over from another interlocutor,
maybe was employed as a turn-taking device.
60 Communicating through Vague Language
S3: You are forcing. Maybe now she doesn’t have anything to say.
(P: 4:150)
S1: Maybe, we are respecting you ladies. (P: 4:151)
S8: So you don’t, you don’t care about her past? (P: 6:781)
S2: No, no. But I am talking about the effects, Okay? Maybe the effects
will continue. Now we are going to start talking, ok? Because …
(P: 6:782)
S1: Because they have the potential, maybe not necessarily in the
ERIC database. (L1: 3:275)
In (4.12), the speaker uses maybe not to tactically exclude certain things;
the cluster makes the speaker’s tone softer to establish empathy with
the hearer. The use of maybe not is part of hedging taking place in the
utterance.
62 Communicating through Vague Language
The data shows that maybe was placed before an ellipsis or similar
structure. This might be linked to the nature of maybe, which expresses
a somewhat weak possibility; thus, omitted utterances seem natural
companions. The ellipsis can enhance the effect of maybe. The cluster
of maybe + ellipsis (or the like) occurred almost equally among Persian
speakers (17 times) and L1s (18 times). The Chinese used it 31 times,
almost as much as the other groups combined.
Other maybe clusters emerged in the data, most from the Chinese
group. I think maybe appeared in the Chinese data 20 times and there
were six uses of maybe I think (swapping the order of maybe and I think);
but none of these was used by Persian or L1 speakers. The Chinese
also employed maybe I/we will 13 times, but the other two groups did
not use any. Another cluster, maybe it is, was used by Persian speakers
20 times and by the Chinese 17 times; none from the L1 group used it.
The trend observed in this group of maybe clusters is that the Chinese
were the leading users while the L1s were not at all interested in using
the clusters. The data exhibits a gap between L1 and L2 speakers’ lin-
guistic behaviors in the use of maybe and its clusters.
Table 4.4 shows that the Chinese used all maybe positions and clusters
the most, except maybe + negation and maybe it is. The L1 group used all
maybe positions and clusters the least, except, maybe + ellipsis. In most
cases the Persian speakers used maybe positions and clusters more than
the L1 group but less than the Chinese, holding the middle ground.
There are three exceptional cases, though: Persian speakers were the top
users of maybe + negation and maybe it is, and in the case of maybe +
ellipsis, they were the least users.
As maybe was used by the Chinese the most, it is natural that they
also chose to use the most maybe clusters. The opposite of the Chinese
was the L1 group, who not only used maybe far fewer times in their
interactions but also used the fewest maybe clusters. In the case of
maybe usage, the Persian speakers sat in between the Chinese and L1
groups in terms of the frequency of use of maybe and its clusters. Given
that the Chinese used much more maybe and clusters than the other
two groups, and if we assume that speakers who use more maybe tend to
be more tentative, then the Chinese seem more tentative than the L1
and the Persian speakers, the L1 group looks less tentative than either
L2 group, and the Persian speakers may be less tentative than the
Chinese but more tentative than the L1s.
4.2.2 May
According to Table 4.3, in this study, may appeared as the second most
common vague expression in the category of possibility indicators.
Unlike maybe, used most by the Chinese and least by the L1s, may was
most frequently used by the L1s with the Chinese the second most fre-
quent users. Against the trend shown in the maybe data where the L1s
used the least, this time the L1 group showed a strong tendency to use
may (56 times, 24 per cent), which had the highest frequency among
the three groups. The Chinese were not far behind, using it 50 times
(13 per cent) but much less than their use of maybe (312 times). The
Persian speakers used may only about one third of the other two groups
(15 times, 8 per cent), but used maybe (156 times) reasonably frequently.
The negative form of may was consistently overlooked by the two
L2 speaker groups, who used it once each, meaning that Chinese and
Persian speakers very rarely used may not in their classroom interac-
tions. The L1s showed a frequency of nine for may not. Similar to the
pattern shown in their use of all categories of possibility indicators
(shown in Table 4.3), the L1s used may and may not clusters more
diversely than the other two groups.
Distribution of may co-occurring with verbs in the data was exempli-
fied by two clusters: may be and may have. The L1s were most frequent
users of both, using may be 14 times and may have nine. The Chinese
64 Communicating through Vague Language
used may be 11 times and may have six. These two clusters appeared
once each in the Persian data, showing that the Persian speakers had
little tendency to use either cluster. There were some less frequent may
clusters, including may say, may make and may wanna/want: of these the
L1 used four times each of the three clusters, and the other two groups
used them rarely if at all.
Pronoun (subject position) + maybe phrases such as you may, it may,
we may and they may revealed some patterns as well. The L1 group used
you may 18 times, the Chinese 12, and the Persian speakers seven. The
ordering was the same for it may: the L1 group used it nine times, the
Chinese three and the Persian speakers not at all. The pattern was differ-
ent for we may, in that this time the L1s did not use it but the Chinese
used it five times and the Persian speakers three. The Chinese also used
they may five times, and the other groups twice each. The cluster there
may be occurred in the L1 data six times, but was absent from L2 data.
By and large, the L1s used somewhat more may in combination with
pronouns, the Persian speakers had little interest in using this kind of
may cluster, and the Chinese sat in the middle. There were some infre-
quent uses of pronoun + may clusters, such as I may, which the L1 group
used four times, the Chinese two, and the Persian speakers never.
As Table 4.5 shows, L1s used may and all the may clusters the most,
except in two cases. In contrast, the Persian speakers used the least may
and clusters. This time it was the Chinese turn to be in the middle: in
most cases they employed may and its clusters less than the L1s but
more than the Persian group. The cluster that all the three groups used
the most was you may. The pattern shown in Table 4.5 is a part of a big-
ger picture that shows the L1s used possibility indicators more evenly
4.2.3 Might
The third commonly used possibility indicator, might occurred with a
frequency of 56 in the L1 data, outweighing its use by Persian (13 times)
and Chinese (ten times) speakers. L1 data for might displays consist-
ency with regard to may and might, in that participants used all three
vague expressions at similar rates (maybe 64, may 56, and might 56).
This consistency is not present in the Chinese and Persian data: for the
Chinese the ordering is maybe 312, may 50, and might 10, with a huge
difference of about 30 times more maybe than might; for the Persian
speakers the discrepancy is smaller but still substantial (maybe 156,
may 15, and might 13), with about ten times more maybe than might.
The L2 groups preferred maybe much more than the other items in the
category of possibility indicators, possibly due to the constraints of their
limited knowledge and proficiency in speaking English.
Table 4.6 shows that the L1s used might clusters more than L2 speak-
ers across the board. In addition to the frequent might clusters dis-
played, there are also infrequent ones, including it might, we might, he
might, and she might.
The investigation of subjects used before might reveals that more
than half of the 31 occurrences of might came before subject pronouns
in the L1 data, and two-thirds (eight) of the Persian and half (five) of
the Chinese. The L1s showed a far greater inclination to use subject
pronouns with might than the Chinese or Persian participants, but as
percentages, the performances of all three groups fall within almost the
same range.
The L1s used six different kinds of subject pronoun in their talk,
although some occurred infrequently. The most common subject
pronoun for this group was you (12 times), followed by it (six), they
(five), I and we each three times, and two occurrences of he. The Persian
speakers’ data demonstrates a preference for combining might with
only four different subject pronouns: you, I, he and she. As they rarely
used might, however, clusters of subject pronouns with might in the
Persian speakers’ data were infrequent. She might (which doesn’t occur
in the L1 data) ranked first in the Persian data with four uses, followed
by two I and one each of you and he. It seems unusual that no Persian
speakers used might with a plural subject pronoun: the one you might
that appeared was singular. By comparison, one third of the subject
pronouns before might in the L1 data were we might and they might. The
Chinese showed reluctance in using might with third person singular
subject pronouns such as he, she, and it, preferring to use this possibil-
ity indicator with a more limited number of subject pronouns than the
other groups.
A pattern observed in the data is the use of subject + might, which
was used five out of 13 times by the Persian speakers. Five out of
56 cases in the L1 data occurred at the beginning of utterances, acting as
turn initiators, but the Chinese never used might with a turn-initiating
subject in their interactions. Besides appearing at the beginning of the
clause, these occurrences turn up at the beginning of the speaking turn.
Despite the same raw frequency of use in the Persian speakers and the
L1s data, the pattern has a substantially different proportional value,
and once it is converted into a percentage, it becomes evident that a
mere 9 per cent of the overall might in the L1 data occurred in subject +
might structure, while in the Persian speakers’ data the proportion is
more than four times as much, 39 per cent.
The most frequent verb combined with might by all three groups is be,
but with different frequency. Might be appears six times (60 per cent) in
the Chinese data, four (31 per cent) in the Persian, but 20 times (36 per
cent) in the L1. Might have appeared five times in the L1 data, only once
in the Chinese and not at all in the Persian. Other less frequent might +
verb combinations include might want and might say.
the Persian speakers used only about 1 per cent of probably each. This
implies that the L1s were more diverse in their use of possibility indica-
tors in the classroom context.
Possible was the least or second least favorite by all three groups in
terms of the category of possibility indicators, as shown in Table 4.3. As
with probably, L1s made the most use of possible: in fact, of the five pos-
sibility indicators apart from maybe, the L1s used all four more than the
other groups. In contrast to the probably trend, the Chinese used possible
twice, less than the Persian speakers’ five.
A diversity of possible clusters appears in the L1 data, such as the last
possible date, the best way possible, and if at all possible. Of the five sen-
tences containing possible used by Persian speakers, two were negative,
two interrogative and one a statement. The L1s produced one negative
and one interrogative, with the rest statements. The Chinese used pos-
sible in one positive and one negative sentence each. Of all five occur-
rences of possible by Persian speakers, three appeared in clause-final
position, a signal that the speaker was going to hand over to another
speaker: this kind of possible acts as both a sentence closer and a turn
closer. In the L1 data, six instances of possible occurred in the final posi-
tion but only one acted as a turn closer, the other five being followed
by another clause by the same speaker. Of the 14 other sentences that
contained the non-final position possible, four were collocated with
that. The Chinese used one out of two possible clusters as a turn closer
in clause-final position, but the other, a mid-clause possible, was not
followed by that.
Figure 4.3 demonstrates that the Chinese used possibility indica-
tors the most, Persian speakers the least, and L1s in between. Ruzaitė
(2007, p. 158) states that ‘maybe or perhaps suggest a lower degree of
the speaker’s commitment to the truth of the claim and make the
claim less categorical’. If we agree, then the total number of tokens
suggests that the Chinese are more tentative and less committed to
their claims than the other two groups, and the L1s are less tenta-
tive and more committed than the Chinese but more tentative and
less committed than the Persian speakers. The other trend shown
in Figure 4.3 is that the Chinese almost exclusively concentrated on
maybe and may; the same pattern exists among the Persian speakers,
but usage was heavily skewed to maybe. The contrastive trend is that
the L1 group spread their uses more evenly across the five expressions
in the category. This phenomenon could be because as native speak-
ers of English, the L1s were better able to use the language resources
available to them than were L2s.
68 Communicating through Vague Language
250
200
156
150
100
64
56 50 56 42
50
15 20
10 13 5 1 2 5
0
Maybe May Might Probably Possible
L1 Speaker CSLE
1%
3% 1%
8%
Maybe Maybe
26% May 13% May
18%
Might Might
24% Probably Probably
24% Possible Possible
82%
PSLE
1%
3%
7%
8% Maybe
May
Might
Probably
81% Possible
but may appears about four times as often as might in the Chinese data.
Probably, at 1 per cent in each L2 speaker group, compares with 18 per
cent in the L1 data. Possible at 3 per cent constitutes the second last
item of possibility indicator for the Persian group, but is the least used
item in the Chinese and L1 data. The overall picture is that the usage of
possibility indicators by the Chinese and Persian speakers is more or less
similar, whereas the L1s demonstrate a very different trend.
The L1 group again tended to use vague quantifiers evenly (except some),
and the L2 groups did not—although the Chinese had a relatively lower
degree of unevenness. The study of subjectivizers and possibility indicators
found that only the first items, I think and maybe, were heavily used, but the
Chinese demonstrate a different pattern in their use of vague quantifiers,
revealing a preference for using all five terms appearing first in Table 4.7,
some, much, many, a lot of, and most, more than the other groups. The
Chinese used vague quantifiers relatively more evenly than they did with
subjectivizers and possibility indicators.
The Persian speakers used some heavily, but other than that, they
used four other items in a relatively close range: many, much, most, and
lots of. The overall frequency of quantifiers by Persian speakers and
L1s is roughly the same, 435 and 423, but there are 741 instances in
the Chinese data. Statistical analysis reveals significant differences in
the use of vague quantifiers by the three groups, p < 0.05 (χ² = 211.976,
d.f.18).
There is some consistency in the use of the first five items in Table 4.7;
all three groups employed some more than the other four items. The
ranking of the five top items is almost consistent across all groups,
with the exception of a lot of in the Persian data. Of the remaining five
items, consistency is limited: for instance, L1s were the most frequent
users of (a) few, a little, lots of and a lot, whereas the Persian speakers
favored lots of.
Lexical Analysis of Vague Language 71
4.3.1 Some
The most commonly used vague quantifier for all three groups, some,
occurs in the Chinese interactions 264 times, in the Persian 229, and
in the L1s 173 times. Some makes up 53 per cent of all vague quanti-
fiers in the Persian data, 41 per cent in the L1 and 36 per cent in the
Chinese. Some was used predominantly by all three groups, which sup-
ports Ruzaitė’s (2007) finding that some was the most frequent of vague
quantifier in British and American academic discourse. This study too
finds that some is not only a favorite of L1s of English, but also of L2s.
What is strikingly revealed in this study is that L2 speakers used far
more some than L1s.
Of the 229 occurrences of some in the Persian data, some + noun clus-
ters appear 160 (70 per cent) times; 144 (55 per cent) out of 264 in the
Chinese, and 74 (43 per cent) out of 173 in the L1 data. The Chinese
located about half of their uses of some with nouns, and Persian speakers
more than two thirds; the L1s used less of this some + noun pattern. Of
the two forms some + noun and some + adjective + noun, the latter was
much less preferred by all three groups. The nouns in some + noun were
both countable and uncountable. In the L1 data, most nouns in some +
noun were countable: 50 (68 per cent) countable and 24 (32 per cent)
uncountable. This trend was even more prominent in the L2 groups:
the Chinese used 129 (90 per cent) countable nouns and 15 (11 per
cent) uncountable, the Persian speakers 153 (96 per cent) countable and
a mere seven (4 per cent) uncountable, revealing an almost exclusive
preference for a countable noun plus some. This distribution is in line
with the patterns that have emerged above, that L1s tend to use various
types of vague language more evenly, and the L2 groups concentrate on
a fewer number.
in clause-initial position four times, the Chinese just once, and the L1s
not at all.
The examination of words co-occurring before some of reveals that
many of the L1s placed some of after a vague expression (mostly a sub-
jectivizer) or a discourse marker such as I mean, okay:
(4.13): This is a discussion between two L1 participants over two
turns. They are talking about the application of a search engine.
In (4.13), S1 uses the discourse marker I mean three times when recom-
mending something to S4. ‘I mean some of these are’, where I mean
appears to mitigate the force of the recommendation, making it less
imposing. Some here is a vague quantifier providing the right amount
of information. Some of is used by the speakers to reinforce their uncer-
tainty; this function of some of is missing in the Persian data and was
used by the Chinese once.
As Table 4.8 shows, the groups preferred different some clusters. The
Persian speakers made use of some people 32 times, three times more than
the Chinese and ten times more than the L1s. The Chinese used some other
16 times, over three times more than the other two groups separately.
L1s used some kind of (14 times), some of you (nine times), some of the,
some things (eight times, each), some of these (six times), and give me some
(five times); none of these clusters appeared in L2 speakers’ interactions.
The Persian speakers used there are some (ten times), some problems and
some rules (seven times each), some of them and in some tribes (six times
each), some people who and for instance some (five times each); other two
groups use none of these except there are some, used by the Chinese five
times. The Chinese use some students (16 times), to do some (ten times),
I think some (eight times), some, some (seven times), you have some and
some money (six times each), go to some, some other things, and some experi-
ence (five times each); the other two groups use none of these.
74 Communicating through Vague Language
Table 4.8 shows that the occurrence of and + some was a favorite of the
Chinese and to a lesser degree of the L1s, with 16 and seven occurrences
respectively. The Persian speakers, on the contrary, use and + some only
twice. Other less frequent clusters of conjunction + some include but
some, used six times by the Persian contingent, once by the Chinese
and never by the L1s). Or some was used by the Chinese five times, three
times by the Persian speakers and twice by the L1s. The total number of
conjunctions used before some by L1 speakers was nine and by Persian
speakers 11: similar results, as opposed to Chinese group’s 22 times.
The Chinese used the pattern the most and L1s the least, just below the
Persian speakers.
The highest frequency of some occurs in the Chinese data (264 times).
The Persian data is not far behind (229 times). Despite using some less
than the Chinese and Persian speakers, the L1s used a more diverse,
wider range of some clusters.
4.3.2 Much
The booster much is often used as an emphasizing marker to enhance a
speaker’s claim. The Chinese contingent used much 106 times, twice as
often as the L1s who used it 53 times and more than twice as often as
the Persian speakers who used it 40 times. This indicates that the
Chinese were more inclined to strengthen their tone of speech than
the other two groups. Compared with the use of much, much of was
used very infrequently: four times by L1s, twice by Chinese, and none
by Persian speakers. Other more frequent much clusters are listed in
Table 4.9.
As shown in Table 4.9, very much is used most frequently by the
Chinese: 48 times, as against two or four times by the other two groups.
The L1 and Persian speakers preferred so much over very much and, to a
lesser degree, over too much. The ranking of booster preferences for the
Chinese is very much, so much, and too much: different from the L1 and
Persian speakers’ so much, too much and very much. The question form of
much is how much; the Persian speakers used how much nine times, the
other groups a half or a third as often. Of its use by the Persian speakers,
five occurred at the beginning of a turn; it occurred as a turn initiator
only once in both the L1 and Chinese interactions.
The Chinese showed an inclination for the heavy use of intensi-
fiers (so, very, and too) before much, a trend that was also apparent
when much appeared in clause-final position. The emergence of much
in clause-final position, modifying a verb, had roughly the same fre-
quency in the Persian and L1 data, but the Chinese showed much heav-
ier use of this: much appears six times in clause-final position in the
Persian data and five times in the L1, but 39 times in the Chinese. Two
other much clusters were found positioned in the end of a clause: so
much and very much. Again, as shown in Table 4.9, the Chinese were
the most frequent users: 27 very much and ten so much; the other
groups rarely used either.
The much clusters discussed above are constructs ending with much;
other types begin with much. For example, much + more occurs fre-
quently in the L1 data but not in L2, which implies that L2 speakers fail
to use much + comparative adjective to intensify their classroom inter-
action. The Chinese differed from other groups in using ‘very much +
conjunction’: nine times for very much and, four times for very much but,
three times for very much because. This shows that the Chinese preferred
more much clusters than the other two groups.
4.3.3 Many
Many is a typical vague quantifier conveying an unspecified but above-
average amount. It can also function as a booster, as does much. Many
occurs 163 times in the Chinese, 47 times in the Persian and 46 times in
the L1 data. There is little difference between L1s and the Persian speak-
ers in their use of many, but the Chinese use is about 3.5 times higher,
showing a strong preference for it in this cohort.
As shown in Table 4.10, the question form of how many was used
17 times by the L1, four by the Chinese and only twice by the Persian
speakers. So many was used 22 times by the Chinese, 14 by the Persian
and six by the L1 group. The Chinese also used many people over four
times more than the other groups. The L1s did not use conjunctions
76 Communicating through Vague Language
before many in their interaction, but the L2s did, the Chinese twice as
often as the Persian speakers. The three groups had a many construct in
common; ‘many + adjective + noun’, which was used five times by the
L1, 23 by the Chinese and seven by the Persian speakers. Unlike some,
which occurs in clause-initial position (as shown in Table 4.8), what
is noticeable is that many in clause-initial position is a very rare case.
Neither the Persian speakers nor the L1s puts many at the beginning of a
clause, and it occurs this way in the Chinese data only three times. The
analysis of the clause-final position use of many or many + noun reveals
that the Chinese and the Persian speakers, with 12 and 19 examples,
inclined to use many or many + noun this way; the constructs occurred
only twice in L1 data. The Persian and L1 speakers also performed simi-
larly in under-using many of, which occurred only twice in the Persian
data and once in L1, but is relatively more frequent in the Chinese, with
seven uses.
Persian speakers were the most frequent users of many or many +
noun in clause-final position and the L1s the most frequent users of
how many. All remaining items in Table 4.10 occurred mostly in the
Chinese data, indicating that the Chinese group were not only the most
frequent users of many, but also of the majority of the many clusters
discussed here.
There were also some many clusters exclusively used by one group. For
instance, the Chinese used there are many 22 times, have many 13 times,
see many, many years eight times, many places seven times, and many
opportunities and many students five times. As many occurred six times,
all in the L1 data, and so many different with a frequency of seven, and
many problems with five occurrences, appeared in the Persian data only.
Despite the extensive use of many in the Chinese interactions, the
Persian and Chinese speakers showed similarity in terms of total
Lexical Analysis of Vague Language 77
4.3.4 A lot of
Like other vague quantifiers discussed so far, the Chinese demonstrated
dominant use of a lot of, using it 85 times, as opposed to 22 times by
the Persian speakers and 39 times by the L1s: four times as often as the
Persian speakers and twice that of the L1s.
The data show that a lot of tends to cluster with countable nouns:
47 times for the Chinese, the most frequent users, 25 times for the L1s
and 14 for the Persian speakers. There is consistency among the groups
in that in all of them, more than half the cases of a lot of are a lot of +
countable noun. For a lot of + uncountable noun, which was much less
in evidence overall, the Chinese again used the pattern most with
19 instances, followed by the L1s (seven) and Persian speakers (six).
Other patterns include a lot of + adjective: the Chinese used this eight
times and the L1s five, but the Persian speakers never.
Some different a lot of clusters emerged among the groups. The
Chinese used there are a lot of eight times and learn a lot of things five, but
neither appeared in the L1 or Persian data. A lot of money was used ten
times and a lot of time four times by the Chinese, but not by L1s or the
Persian speakers. A lot of things occurred 19 times in the Chinese data,
only twice in the L1 and once in the Persian data. The cluster a lot of
people occurred in L1 data six times, three times in the Chinese and only
once in the Persian speakers’ data. This is in conflict with what Drave
(2002) finds in his research on the use of vague language by two cultur-
ally different groups: that the most common phrase of native speakers
of English is a lot of, while a lot of people is used mostly by native speak-
ers of Cantonese conversing in English. The results are in reverse in this
study: a lot of was used mostly by the Chinese, and the L1s used more
a lot of people than the L2 groups. This difference may be attributable to
the different nature of the discourses in the two studies.
4.3.5 Most
The Chinese showed a stronger tendency to use most than the other
groups: 86 times for the Chinese, more than twice as much as the
78 Communicating through Vague Language
S1: What’d we say? On the farm, two intervening words. Two-N farm.
So instead of using, W I use N. In fact most of the constructions
that you see out there, that use, a W, could often be turned around
to be a, two-N. So. (L1: 3:339)
S21: Hairy animals. (L1: 3:340)
S4: Something like this. But nowadays because I think, huh, most of
the house especially in towns, in cities, such as big cities like Tehran
and the other cities are, the house doesn’t have any. (P: 6:603)
S2: They are like flats. (P: 6:604)
S2: I think, maybe, most of you have seen the film Scrappy. (Ch: 7: 64)
S7: Yeah. (Ch: 7: 65)
S2: In this film a lot of students always make troubles to the teacher
and sometimes they put some glue, on the, the chair and the
teacher sits on it. I think that’s terrible. (Ch: 7: 66)
Lexical Analysis of Vague Language 79
4.3.7 A little
Following (a) few, a little was again more commonly used by the L1s (20
times) than by either L2 group: 11 times by the Chinese and nine by
the Persian speakers. The L1s used this quantifier twice as often as the
other two groups separately.
Around half of the sentences (nine) containing a little in the L1 data
are followed by a comparative adjective, whereas only a third of the
sentences (three) in the Persian data follow the same pattern. There was
only one example from the Chinese, indicating that this pattern may
have occurred accidentally.
4.3.8 Lots of
Lots of was used 34 times by the Persian, 16 by the L1s, and 11 by
the Chinese. This is the first vague quantifier employed more by the
Persian speakers than by the other two groups, and they used lots of
more than twice as often as the L1s and three times more often than
the Chinese.
The occurrence of words before lots of reveals some patterns. For
instance, while there are lots of occurs five times in the Persian data, the
L1s did not use this cluster in their interaction and the Chinese used it
only once. By contrast, the L1s used there is lots of five times, the Persian
speakers once and the Chinese not at all. This difference between L1
and L2 speakers may be attributed to a lack of knowledge of the phrase’s
use on the part of L2 speakers.
There are clusters consisting of elements occurring after lots of. For
example, while the L1 and the Chinese participants did not use lots of
things, this cluster occurred 11 times among the Persian speakers.
Of the total of 34 lots of + noun in the Persian data, 25 involved
countable nouns, as did 12 out of 16 by the L1s and five out of 11
by the Chinese. This finding shows that the Persian and L1 groups
preferred lots of + countable noun more than the Chinese group, who
preferred lots of + uncountable noun instead, although only slightly.
The uncountable nouns found in the data are typically money for the
L1s, and knowledge and time for the Persian speakers.
4.3.9 A lot
A lot occurs 16 times in the L1 data, ranking ahead of the second
favorite, lots of. A lot occurs half as often in the Chinese data, with the
frequency of eight; it appears nine times in the Persian data. A lot is
infrequently used, at less than 5 per cent of all vague quantifiers.
Lexical Analysis of Vague Language 81
Of the nine occurrences of a lot in the Persian data, three are in clause-
final position. There are four occurrences of a lot in clause-final position
in the Chinese but only one in the L1 data, showing that this group is
much less interested in this construction. There is no consistent pattern
of words used before a lot in any of the data sets, but words appearing
after a lot include comparative adjectives, which the L1s used six times,
and three of which are multi-syllabic adjectives placed after more, two
uses of the irregular comparative adjective better, and two infinitives.
More than half of a lot in the Persian data are followed by conjunctions,
two for reason, two for contrast and one for addition. It seems that the
L1s’ use of a lot is mainly associated with comparisons, but when used
by the Persian speakers, it appears to create the need for a new utterance
through conjunctions. As half of the Chinese use of this phrase occurred
in clause-final position, no meaningful pattern emerged for words
following a lot.
4.3.10 Majority
Majority appears least in all data sets. There were a few majority used by
the Persian speakers and L1s, but the Chinese interactions were devoid
of this vague quantifier. Despite the relative similarity in the use of
majority by the two groups, they present a different pattern from major-
ity. All six occurrences of majority by the L1s are followed by of; none
of the seven occurrences in the Persian data do. Three of the majority in
the Persian data occur in clause-final position.
Figure 4.5 shows the frequency ranking of vague quantifiers as: some
(moderate quantifier); much, many, a lot of, and most (bigger quantifi-
ers); (a) few, and a little (smaller quantifiers); lots of, a lot, and majority
(bigger quantifiers). The pattern is from most frequent to least frequent:
moderate—bigger—smaller—bigger. The result is somewhat similar to
Drave’s (2002) finding that a lot and many are the most common vague
quantifiers.
In terms of overall frequency, the vague quantifiers conveying a
smaller quantity (some, a few, a little, and few) are slightly fewer than
the vague quantifiers conveying a larger quantity (for example many
and lots of). The 1599 quantifiers occurring in this data comprise
734 smaller-amount quantifiers (46 per cent) and 865 bigger-amount
quantifiers (54 per cent). This indicates that the participants use vague
quantifiers to mitigate (for example a little) or emphasize (for example
a lot) at relatively similar rates.
The overall frequency of quantifiers by the three groups suggests that
this vague language category occurred almost evenly in the classroom
82 Communicating through Vague Language
300
264
250 229
200 173
163
150
106
100 85 86
53
40 46 47 39 38 34
50 22 33 21 20 20 16 9
7 0 119 11 8 6 7
0
0
f)
ew
tle
of
ity
to
lo
of
an
uc
(o
or
lit
ts
t(
)f
A
lo
M
M
e
Lo
aj
os
A
(A
m
M
M
So
interactions of two groups: the Persian speakers (435) and the L1s
(423). In contrast, the Chinese group used them 74 times: almost twice
as often. Statistical analysis of the occurrences of this category among
the three groups proves significantly different, and close investigation
of some subcategories also reveals discrepancies between the groups at
relatively similar frequencies. Of the ten vague quantifiers studied, the
first five shown in Figure 4.5 were consistently and most heavily used
by the Chinese; this was not the case for the second five, which were
used infrequently, and mainly by L1s (except lots of).
While the discrepancy in the use of the first five items by the three
groups is large, the same does not hold for the second five items, where
the differences are of a smaller scale. Some emerges as the most com-
mon vague quantifier among all three groups, occurring 264 times in
the Chinese interactions, 229 times in the Persian, and 173 times
in the L1. The second most frequent item, much, with 106 instances in
the Chinese data, was used twice as often as by the L1s (53 times), and
even more often than the Persian speakers (40 times). On the whole,
there is consistency in the pattern of small quantity quantifiers in that
(a) few and a little were both used more frequently by the L1s within
the limited frequency range presented by the three groups, while there
were fluctuations in the use of quantifiers expressing large quantities.
Lexical Analysis of Vague Language 83
relatively more evenly than the two L2 groups. The L2 speakers’ use
was skewed toward five of the nine quantifiers, and the others were
used minimally; the unevenness is not so evident in the L1 group. This
difference between L1 and L2 groups echoes and repeats what has been
revealed in their use of subjectivizers and possibility indicators, and the
trend continues in this category as well.
The two L2 speaker groups show a closer ranking distribution, as
shown in Figure 4.6. For instance, the ranking order of the first four
items is the same between the two groups: some, much, many, and a lot
of. More than half of the quantifiers used by the Persian speakers con-
sisted of some; this proportion in the Chinese and the Persian speakers
covered two items (some and much). At the lower end of scale, the total
percentage of the last five items is only 5 per cent for the Chinese group,
while it is 14 per cent and 19 per cent for the Persian speakers and the
L1s, respectively.
the L1 group stood in the middle. The total number of vague intensi-
fiers for the Chinese was 2.2 times that of the L1s and 2.7 times that of
the Persian speakers. This indicates that the Chinese participants tended
to emphasize more than the others, and the Persian speakers tended to
emphasize the least. The most preferred item was very for both L2 groups,
but the L1 group preferred really. Only one item was predominantly
used by the Persian speakers: actually. Chi-square test reveals a signifi-
cant difference in the use of vague intensifiers among the three groups,
p < 0.05 (χ² = 321.263, d.f.10).
4.4.1 Really
Not all occurrences of really in the data are vague intensifiers, as some
carry out functions that do not emphasize, as shown in (4.17) below.
(4.17): This is a discussion between three L1 participants over seven
turns. They are talking about the result of a search using a search engine.
and the Persian speakers 19, less than half the Chinese tally. Of the five
items, only four occur in the Persian and L1 data, but the Chinese used
them all. The four items used by the Persian speakers were distributed
relatively evenly, but the Chinese concentrated heavily on two items
only and the L1 on three. The Chinese preferred the first person sin-
gular and plural, I really and we really, while the other groups showed
no such preference: the L1 group did not use the first person, and
third person was used only once; the Persian speakers did not use
the third person plural.
Really occurs quite differently in negative sentences. The frequency is
quite high in the L1 data, with 39 instances (22 per cent), lower in the
L2 speaker groups (seven and 4 per cent for the Chinese, one and 1 per
cent for the Persian). The L1s prefer really in negation much more than
the L2s, inclining to use really to highlight the force of a negation.
As shown in Table 4.12, really occurs five times in clause-initial posi-
tion in the Persian data, and twice each in the L1 and Chinese. Nine
per cent of the Persian speakers’ intensifiers were positioned at the
beginning of a clause, much more than the 1 per cent of the other two
groups. An example from the L1 group, where really starts a sentence,
appears in (4.18).
(4.18): This is a turn from an L1 participant talking about studying
speech.
S1: I know this wasn’t gonna happ- help work sooner or later. Junior,
high, school. Okay? Makes sense. Really makes sense. –Really
makes sense. Junior high school. I know I’m gonna find that con-
struction. Senior high school. (L1: 3: 346)
S1: And the song won a big award and even, 11 very familiar awards.
Really amazing, amazing, and other five songs in this original
song are also very beautiful. (Ch: 6: 71)
S2: What’s the name of this album? (Ch: 6: 72)
88 Communicating through Vague Language
In terms of lexical clusters with really, apart from the subject pronoun +
really, other types of cluster are barely present in any of the three groups.
One rare occurrence is the use of really want: the Chinese used this nine
times, the L1s five, and the Persian speakers four. The Chinese used
really like eight times and really need five times, the L1s used them twice
each, but no examples appeared in the Persian data. Really good is used
by the L1s nine times and three times by the Chinese, really important by
the Chinese seven times and four times by the L1s, neither are found in
the Persian data. Some really lexical clusters are used by one group only:
from the L1 group: really start eight times, really interested 11, that’s
really nine, and it doesn’t really five times. From the Chinese: really hard
six times, really I think six, I really want five, and really like to five times.
A trend here seems to be that the L1 and Chinese speakers are similar
in their use of lexical clusters of really, while the Persian speakers do not
use these clusters as often.
The findings demonstrate that the L1 and Chinese cohorts are similar
in many ways and are quite different from the Persian speaking group.
Table 4.12 reveals that the L1 and Chinese groups make more use of
really and its clusters, while the Persian speakers use them less often and
more diversely.
4.4.2 Very
The second most commonly used vague intensifier by the L1s, very, was
the most common vague intensifier used by the Chinese and Persian
speakers. With 498 uses, the Chinese were the leading users of very,
followed by the Persian speakers at 108 occurrences, and trailed by the
L1 group with 79 instances. More than half (56 per cent) of the vague
intensifiers in the Chinese data include very, one third (32 per cent) in
the Persian data, and one fifth (20 per cent) in the L1. It seems that very
is more popular with the L2 groups than the L1s.
An examination of different sentence positions in which very occurs
indicates that while the L1s avoided using this quantifier in clause-
initial position, the Chinese and the Persian speakers used it four and
six times respectively, in this position.
(4.20) This is a discussion between two Persian participants over three
turns. They are talking about the younger generation.
S7: I think when he or she knows me, he respects me. (P: 6:111)
S3: Ahuh, the others? (P: 6:112)
S1: I think it is not very important. (P: 6:113)
S1: And so what it does, and the system is doing this not a human
being a computer system. and the computer can do this very
quickly and very efficiently. Computer system starts to number,
each word in the field and preface, the n- word, position informa-
tion or the word position number, by the name of the field so in
this case we have. (L1: 3:268)
SS: Stop word (L1: 3:269)
S1: It’s a, called a stop word and what is a stop word? Anyone know?
(L1: 3:270)
As shown in Table 4.13, the most frequent users of the cluster be + very
were the Chinese (19 instances), far ahead of the other groups (four
times by the L1 and two by the Persian speakers). Despite the differ-
ence in raw numbers, the percentage values are 6, 4, and 2 per cent,
closer than the difference in raw numbers. In the case of is + very,
the Chinese produced 145 instances, the Persian speakers 32, but the
L1s only three. In raw numbers, the Chinese used it five times more
than the Persian speakers and almost 50 times more than the L1s.
Lexical Analysis of Vague Language 91
The Chinese and Persian percentages are similar (about 30 per cent), but
very different from the L1 group (4 per cent). The situation is similar
to the case of are + very and was + very: where the L2s showed a similar
frequency (by percentage). While the Chinese and Persian speakers
were different in raw frequency, that difference was much smaller than
the difference between them and the L1s. These findings suggest that
there is more diversity between the L1 and L2 groups than within the
L2 groups.
Despite using very the least, L1s revealed some usages that were
uncommon in the L2 data. For instance, the duplication of very was
used in (4:24) below.
(4.24): This is a discussion between two L1 participants over two
turns. They are talking about how culture can affect morals.
S1: And um, that’s actually something there’s a um, a really famous,
philosopher who, who wrote something very, very similar, to
that so, just in case you’re interested, um, Alasdair MacIntyre.
I don’t know jus– … (L1: 2:45)
S11: But definitely I think like something as extreme as a child
killing someone, um, some are more, like are more obvious
than others and haven’t been known to like, work, like certain
actions. (L1: 2:46)
followed by the Persian speakers. The L1 group are the least frequent
users, and do not even use two of the clusters. Of the five clusters, the
Chinese used it is very the most: 49 instances and 10 per cent of the
very instances in their data. The Persian speakers used it is very 12 times,
about five times less than the Chinese. The two items they used a bit
more were very important and very good (15 and 14 times respectively).
The Chinese also used it’s very (a shortened version of it is very) 32 times,
but no instances of this were presented by the other groups.
Some very clusters are absent from one or two groups, suggesting that
they were less popular. For example, very busy and very difficult were used
by the two L2 groups only. Very busy occurred five times in the Chinese
discourse and four times in the Persian; very difficult was used 11 times
by the Chinese and four by the Persian speakers; the cluster a very good
was used eight times by the Chinese and five by the L1s, but not by the
Persian speakers. Some very clusters were used by the Chinese only: there
are very (27 instances), very happy (12), will be very, very beautiful (11), he
is very, very famous (10), this is very (7), very hot, very easy, very convenient,
that’s very (6), very fast, and very young (5). The L1 group used the very
least five times, but it did not appear in any L2 data.
All the adjectives placed after very by the L1s were positive (impor-
tant, good, and interesting), while the Persian speakers used positive and
negative adjectives evenly: three positive adjectives (important, good and
interesting), and three negative (busy, difficult and hard). The Chinese
used more positive than negative adjectives. The trend emerging here
is that very tends to cluster with positive adjectives, and negative adjec-
tives are less often grouped with very.
There is an inconsistency in the patterning of very, and all other
vague intensifiers, among the three groups. The use of vague intensi-
fiers as well as most individual lexical items in the Chinese interactions
outnumbered those of the L1 and Persian speakers, indicating that the
Chinese group tended to enhance their speech more than the others.
4.4.3 Actually
Actually ranks fourth and is one of most frequent intensifiers. Unlike the
trend in very, this time the Persian speakers used it most (73 instances,
22 per cent), the L1 group 67 times (17 per cent), the Chinese least,
with 24 instances (3 per cent), about three times fewer than either other
group. While actually constitutes around one fifth of the overall intensi-
fiers used by the L1 and Persian speakers, the Chinese used it only as
3 per cent of their intensifiers in similar contexts. The Persian speakers
and L1s behaved alike to a large extent, in this case.
Lexical Analysis of Vague Language 93
S1: I would have to take a look at your search to actually tell you
exactly what went on … my guess is, and I’m, I’m this is just my
hunch, um, that again you may have used an or, instead of an and.
That’s a possibility. Or … (L1: 3:197)
S9: Physical disabilities, and then not-ing out mental retardation (dis-
cussing it back). (L1: 3:198)
S2: For instance, imagine that you understand that she has a boyfriend
what would you do. (P: 6:722)
S1: Well, you know, honestly, what would I do? Just shake but the
most important thing about boyfriends these days in Iran is
I think, actually, I am sorry … you know, actually I talked about
this matter with my daughters. I’ve got three daughters. I had to
talk about these things with them. (P: 6:723)
4.4.4 So
The Chinese used actually least often, but so most often. Their 154 uses
mean that so was used almost twice as often by the Chinese as by the
Persian speakers (75 instances) and roughly four times as often as by the
L1s (40 instances).
The three groups were relatively consistent in using so in negative sen-
tences. Seven instances each occur in the L1 and Persian data, and nine in
the Chinese. Another cluster is so + adjective, where adjectives are of three
types: positive, neutral, and negative. For the L1 group, the frequency
was 11 for positive, one for neutral, and 11 for negative; for the Chinese
group, 83 for positive, 40 for neutral, and 31 for negative; for the Persian
group, 20 for positive, six for neutral, and 20 for negative. There is some
inconsistency between the Chinese and the other groups, but at the same
time there is a consistent pattern between the Persian and the L1 speakers,
who used negative adjectives as often as positive adjectives, and neutral
adjectives less often. Another consistency lies in the use of neutral adjec-
tives, which happened least often in all groups: one instance by the L1s,
five by the Chinese, and six by the Persian speakers.
Lexical Analysis of Vague Language 95
There are two frequent lexical clusters in the data: so much and so
many. The L1 and Persian speakers used so much almost evenly, with
eight and nine occurrences, respectively. The Chinese, on the other
hand, used 19 so much, more than twice as often. Like so much, the
Chinese used so many most frequently (17 occurrences), and the L1s
with five occurrences were still the least frequent users of this cluster.
The 15 occurrences of so many by the Persian speakers brought them
closer to the Chinese. In addition to so much and so many, other less
frequent lexical clusters included so difficult, and you are so.
4.4.5 Too
The intensifier too was used most by the Chinese, with 34 instances; the
L1 group presented 24 instances and the Persian speakers 15. Contrary
to so which was used least often by the L1s, the Persian speakers used
too the least commonly, at only half the rate of the Chinese. The three
groups were similar in their use of this term from a percentage perspec-
tive, at about 5 per cent of the total vague intensifiers in the data.
Too occurred in both positive and negative sentences by the L1s and
the Chinese, while the Persian speakers used it in positive sentences
only. There are two instances of too by the Chinese occurring in sen-
tences containing never, but the Persian speakers showed no inclination
to use too in negative sentences.
An example of too used in a negative form is given in (4.27) below. In
turn 128, S1 uses not too soon, where too increases the degree of soon.
(4.27): This is a discussion between two L1 participants over four
turns. They are talking about how state systems will treat religious
beliefs.
SU-m: That’s really interesting we’ll have to see. Probably won’t happen
in this lifetime. (L1: 2:127)
S1: I don’t think so, not too soon anyway, but we’ll see I mean, you
know. (L1: 2:128)
SU-m: No, it’s a slow process. (L1: 2:129)
S1: It is. (L1: 2:130)
The L1s used three times of emphatic just + too, a phrase absent from
the Chinese and Persian data. In excerpt (4.28) below, the word young is
intensified by double boosters, just and too.
(4.28): This is a discussion between two L1 participants over two
turns. They are talking about why young people are more likely to steal
things.
96 Communicating through Vague Language
S1: Yeah that’s, yeah, I mean, yeah rebellious, or, or not we, we could
call it socialized which basically just means they’re not accustomed
to society’s norms yet. Yeah. (L1: 2:13)
S2: Or they’re just too young to understand the uh, consequences of,
stealing. (L1: 2:14)
In addition to the discussion of what occurs before too in the data, the
analysis of what occurs after too shows a partial similarity between the
L1 and Persian speakers. The first trend can be seen in the occurrence of
too much: four instances from the L1 and Persian speakers each, 14
from the Chinese. The L1 and Persian groups behave differently, how-
ever, in what too much is paired with: the L1s used too much twice at
the end of a clause, and twice in the middle, once after a verb and the
other before a noun; the Persian speakers placed all four instances of
too much before nouns, one of which is grammatically incorrect. The
Chinese used two tokens of too much at the end of a clause, and the
rest before nouns.
4.4.6 Quite
As with too, quite occurs with patterns that are different across the three
groups. Quite was used 16 times by the L1s, but by the Persian speakers
four times only. The Chinese used it nine times, about half as often as
the L1s. The percentage of quite constitutes a small portion of vague
intensifiers in each group: 1 per cent by the Chinese and Persian speak-
ers, and 4 per cent by the L1s.
As in the case of too, the Persian speakers avoided quite in negative
clauses. The same trend is found in the Chinese sample as well, but half
of the clauses containing quite by the L1s are negative. Quite + adjec-
tive occurs seven times in the L1 data, three in the Persian and six in
the Chinese. In addition, the cluster of quite + adverb occurs four times
in the L1 samples, but the Persian speakers used it only once and the
Chinese cohort never used quite before an adverb.
The L1s used other quite clusters, including quite a bit, quite a while,
quite sure, and quite easy, but the Persian speakers were limited to quite +
adjective and quite + adverb. The Chinese were slightly more diverse,
using quite sure, quite easy, quite + a + adjective + noun, and quite +
preposition in their interactions, making them somewhat similar to the
L1s in the use of quite clusters.
Overall, as Figure 4.7 shows, vague intensifiers form a category with
diversity among the three groups. The discrepancies go beyond simple
frequency of individual items, which are reinforced when patterns of
Lexical Analysis of Vague Language 97
498
500
400
300
174
200
164 154
79 108
75
100 73
58 67
40 34
24 24 15 16
9 4
0
Really Very Actually So Too Quite
use and clustering behavior are taken into consideration. The Chinese
participants, with 883 occurrences, showed a stronger tendency to use
these intensifiers than the L1s with 400 occurrences and the Persian
speakers with 333. The groups had more similar rates of use for too and
quite. Despite this, the difference in the application of vague intensifiers
by the three groups is found to be statistically significant.
In terms of the frequency of individual items, the Chinese used three
categories the most often, very, so, and too, with an extremely high
use of very. The most significant frequency difference among the three
groups lies in this term, which the Chinese used approximately five
times as often as the Persian speakers and six times as often as the L1s.
The L1s used only two items the most often: really and quite (the first
and the last items in Table 4.7); the Persian speakers used only one item
the most often, actually.
Comparison of the frequency of the six vague intensifiers shows
that the L1 and Persian speakers behaved similarly in using four items,
very, actually, so and too. The Chinese and L1 groups were similar
in their use of two items, really and quite. The Chinese and Persian
speakers do not share any similarity in the use of any item. What
can be drawn from this is that as far as the frequency distribution
98 Communicating through Vague Language
L1 speaker CSLE
4% 1%
4%
6% Really Really
10% Very 17% 19% Very
Actually 3% Actually
17% 43% So So
Too Too
56%
20% Quite Quite
PSLE
5% 1%
17% Really
23% Very
Actually
32% So
22% Too
Quite
4.5 Placeholders
4.5.1 Something
Something was used most by the Persian speakers (177 instances), almost
twice as often as the second most frequent users, the L1 group with 94
instances; the Chinese were the least frequent users with 80 instances.
The percentage distribution reveals that the difference among the
three groups is not as significant as the frequency suggests. Something
constitutes around one-third of the total placeholders used by each
group, and all three were quite consistent in terms of the percentage of
something in relation to the total number of placeholders, ranging from
28 per cent to 37 per cent.
As shown in Table 4.15, the Persian speakers most frequently put
something in clause-final position. The 47 occurrences in their data
(27 per cent of the total use of something) are more than three times
those of the L1s with 14 occurrences (15 per cent) and more than six
times those of the Chinese with eight tokens (10 per cent).
There is a phenomenon in the data: a discrepancy between something +
conjunction vs. conjunction + something. In terms of something +
conjunction (something precedes the conjunction), the highest
frequency is found in the Persian data where conjunctions after some-
thing appear nine times; this cluster occurs four times in the L1 data
and three in the Chinese. The clusters consist of something + but, and
and or for the L1s and Chinese, while the Persian speakers combined
something with and and or only. This suggests that the Persian speakers
prefer something + conjunction to express addition (and) and choice
(or), but not contrast (but).
S5: Inevitable means something you cannot escape from. (P: 4:42)
102 Communicating through Vague Language
The L1s do not use something you in their interaction, instead opting for
clusters such as something that, exemplified in excerpt (4.30).
(4.30): This is an utterance by an L1 participant who is referring to a
photo.
Item 3 in Table 4.15, something that, is one of the clusters most fre-
quently used by the Persian speakers (15); the L1s used it ten times and
the Chinese never. Analysis of the function of that after something shows
that the Persian speakers used that as a conjunction or a subject almost
evenly: seven and eight times, respectively. However, L1 use of that
after something reveals uneven distribution, the speakers preferring that
as a conjunction (seven uses) almost twice as often as that as a subject
(three uses).
Item 7, something + infinitive, appears most frequently in the L1 data
(eight times), while the Chinese and the Persian speakers use it three
and two times respectively, clearly not a favorite for L2s. Item 9 is some-
thing else: it was again most often used by the Persian speakers (15),
while the Chinese used it around half as often (seven) and the L1s least
(three). The Persian speakers used a third of this cluster after the verb
do or does, the Chinese only once; instead, they used it five times after
conjunctions such as and and or.
Some clusters were not used by all three groups. For example, be some-
thing was used 18 times by the Persian speakers, a third of that by the
L1s (six occasions), but not by the Chinese. Nor did the Chinese offer
any examples of something that or something wrong. Two clusters absent
from the L1 data are something you and to do something. Some clusters
were used exclusively by a single group: the Persian speakers provided
seven instances of you do something; six of it is something and something
for eating; five of attention to something, to say something, buy something,
and know something; the Chinese were alone in using learn something
(six instances). This suggests that the three groups differed in their pref-
erences in using something, demonstrated by the absence of particular
something clusters from certain groups.
4.5.2 Things
As with something, things was used the most by the Persian speakers, but
unlike something, where the Chinese were the least users, it was the L1s
that used things least often. The Persian speakers used this placeholder
Lexical Analysis of Vague Language 103
108 times, exactly twice as often as the L1s (54 instances); the Chinese
used it 85 times, not far from the Persian speakers’ use. Figures in
percentages show that the contribution of things to the formation of
placeholders is around one-fifth in the Persian and L1 data, but one-
third in the Chinese data.
As shown in Table 4.16, things in clause-final position occurs most
in the Persian data (37 times), while it occurs 28 times in the Chinese
and eight times in the L1 data. The cluster makes up about 33 per cent
of things in both L2 groups, and less than half of that (15 per cent) in
the L1 cohort. This seems to be a case of L1 against L2. Around half
the clause-final position things in each group occurred as turn-shifting
devices: 16 by the Persian speakers, 12 by the Chinese, and three by
the L1s. The same trend continues in the use of things in negative and
interrogative sentences, where again the L1s were the least users and the
Persian speakers the greatest. The Persian speakers used things in nega-
tion eight times, the Chinese five, and the L1s never. The L1s used things
in interrogative sentences twice, the Persian speakers six times and the
Chinese four. There is consistency here in that the Chinese and Persian
speakers both used things at a relatively similar rate and ranking when
things was in clause-final position, negative or interrogative sentences,
while the L1s tended to use these less often, coming last in the rankings.
Items 2–6 in Table 4.16 are clusters in the form of lexical item +
things. The frequency of use among the L1s was the opposite of that of
the Persian speakers. A form that was used most by the L1s tended to
be least used by the Persian speakers; an exception is the case of kinds
of things, where the Chinese group was the least user and the Persian
the greatest. Of the five items, four (kinds of things, different things, other
things, and new things) were used most by the Persian speakers, and all
except kinds of things were used the least by the L1s. For the remaining
one, some things, the positions reverse: this time the L1 group was the
greatest user and the Persian cohort, with no uses, the least. This part of
data depicts an opposing tendency between the L1 and Persian speakers.
The Chinese, in the middle, use all five clusters, the L1 group three and
the Persian speakers four.
Some clusters take the form of conjunction + things. Item 7 in Table
4.16 shows the cluster things + and. The most frequent users were the
Chinese, who provided 12 instances; the Persian speakers were the sec-
ond most frequent users with seven instances, and the least frequent
users were the L1s with two instances: a sixth of the frequency of the
Chinese. The contrastive conjunction cluster of things + but was used
by the Persian speakers six times and once by the Chinese, but not by
the L1s. The cluster things + because is used three times by the Persian
speakers and once each by the Chinese and L1s. The cluster things + so
was used exclusively by the Chinese, twice; it did not occur in the other
two groups. The Persian speakers and the Chinese were somewhat simi-
lar in their clustering of things with conjunctions; the L1 group showed
a different behavior.
The Persian speakers frequently used things that and things. OK? The
former was also used frequently by the L1s, but was less frequently
employed by the Chinese. The total occurrence of things that is 23 in the
Persian data and ten in the L1, but only twice in the Chinese. It serves
in the Persian data as either a conjunction or the subject of sentence,
almost equally (12 and 11 times). The L1s preferred that as a conjunc-
tion (ten times), but also used it as a subject (six occurrences). Of the
two occurrences of that after things by the Chinese, one served as a
conjunction and one as a subject.
Some things clusters are used exclusively by one group. The Persian
speakers provided 15 instances of the things that, ten of lots of things; six of
of things but; five each of things that they, things that I, and and other things.
The L1 group offered five instances of of the things. The Chinese gave
19 instances of a lot of things and five of learn a lot of things. These suggest
that despite the closeness between the Persian and the Chinese speakers
in their total use of things, here the Chinese and L1s are closer, in that
they use far fewer things clusters than the Persian speakers.
4.5.3 Thing
The only placeholder that turns up with a close occurrence across
all three groups is thing. The Chinese, with 82 uses, employed it the
most often, but it was only slightly less common in the Persian (74)
Lexical Analysis of Vague Language 105
by the L1s—about half the Persian use. The Chinese frequency was three
(4 per cent), indicating thing that was not as popular with them. When
that is placed after thing, it serves as the subject of the sentence or a con-
junction: these uses were evenly represented in the L1 interactions (five of
each), but it was used more as a subject than a conjunction by the Persian
speakers (14 instances as a subject, eight as a conjunction). Two of the
three uses of that by the Chinese performed as conjunctions.
In Table 4.17, among seven items (items 5–11), there is relative con-
sistency among the three groups for the first four items (one thing, the
same thing, the only thing, and another thing): the frequencies are all small
and close to each other. This consistency disappears in the last three
items. For item 9, the thing, while the L1 and the Chinese are exactly
the same (four instances each), the Persian speakers stand out by using
the cluster 17 times, about one fourth of the total thing they use. For
item 10, good thing, this time the L1 and Persian speakers are exactly
the same (twice each), but the Chinese use it nine times, four times
more than either group. The last item, 11, most important thing, was
not very different in frequency of use by the L2 groups (11 and eight
instances respectively), but is absent from the L1 data. There are even
more diverse phenomena in the data. Some clusters only occur in one of
the three groups: for instance, the Chinese use interesting thing six times
and most interesting five, but neither group uses them; and the Persian
speakers use thing that I five times, but it does not appear in the data of
the other groups.
4.5.4 Anything
As with all the placeholders examined so far except thing, the Persian
speakers were the most frequent users of anything, employing it twice
as often as the L1s (48 vs. 24) and four times as often as the Chinese
(48 vs. 12). In all groups, the percentage rate was 10 per cent or less of
the total occurrence of anything (Table 4.18).
4.5.5 Someone
The occurrence of someone is rather like that of anything, with the
Persian speakers again the most frequent users and the Chinese the least
frequent. Someone appears up 31 times in the Persian data, slightly less
(28 times) in the L1 and half as often (13 times) in the Chinese. The L1
and Persian speakers are alike in using someone at a similar frequency.
108 Communicating through Vague Language
The percentage distributions for all three groups are all less than 10 per
cent: 9 per cent for the L1, 5 per cent for the Chinese and 6 per cent for
the Persian speakers.
Someone is rarely found in clause-initial or clause-final position in
this data. The former is missing in the Chinese data, but occurs once in
the L1 data and twice in the Persian. The Persian group used someone
in clause-final position twice, the other two groups did not use it at all.
It seems that the three groups demonstrate consistency in positioning
someone: the L1 group used conjunction + someone seven times, the
Persian speakers three and the Chinese two. There are four kinds of
conjunctions in the L1s data: if to express condition; and to express
addition; because to express reason; and or to refer to choice. The Persian
speakers use only two, if and and, while the Chinese use the conditional
if. The only conjunction appearing in all three groups is if. The L1s
chose to use a more diverse combination of conjunctions + someone,
while the two L2 groups were less inclined to do so.
Group differences appear in the use of prepositions (about, for, to, and
with) occurring before someone: the Persian speakers used this type of
cluster the most often with 11 occurrences, the Chinese twice and the
L1 speakers once. Somebody + conjunction occurs only once in the L1s
and Chinese interactions respectively, both being someone and; and the
Persian speakers used someone + conjunction three times, with one use
of each and, if and or.
4.5.6 Somebody
As one of the less frequent placeholders, somebody is used slightly less
than someone. The highest frequency of somebody (23) was among the
Persian speakers; the L1s offered 20 examples, and the Chinese, with
11 occurrences, employed someone about half as often as the other two
groups. The percentages reveal a small proportion of the data consisted
of somebody, under 7 per cent for each group.
The participants showed little inclination to use somebody in clause-
initial and final positions. The only interesting trend in this regard was
that this placeholder occurred in clause-initial position and in clause-
final position with the same frequency in each group: that is, somebody
occurs twice in the final position and twice in the initial position in
the Persian data, and once in each position in the Chinese and L1 data.
There are occasional examples of a conjunction occurring before
somebody: the Persian speakers used and once and if three times, and the
Chinese and L1s even less: and was employed once by the Chinese and
or once by the L1 speakers. The pattern of conjunctions occurring after
Lexical Analysis of Vague Language 109
somebody is more or less similar: the Persian speakers used and twice,
and the L1s used and as the only conjunction after someone; the only
conjunction used by the Chinese was or.
The L1s chose who as one of the two relative pronouns after somebody,
while the Persian speakers preferred that. Relative pronouns were not
used in the Chinese interactions.
4.5.7 Anybody
The least frequently used placeholder in the data is anybody. The Persian
speakers employed anybody 17 times, and it appeared half as often (nine
times) in the L1 interactions and only three times in the Chinese data.
The percentage analysis, however, levels off these differences, and all
occur as less than 4 per cent.
Anybody occurs in clause-initial position as a turn-taking device,
three times each in the Persian and L1 data, while the Chinese used
all three tokens of this placeholder in clause-mid position. The Persian
speakers used anybody in interrogative clauses seven times, in four of
which the question mark appears immediately after anybody. The L1s
presented only three interrogatives including anybody, and none in
which anybody comes directly before the question mark. The Persian
speakers used anybody else four times, two in negative sentences and
two in questions.
gs
ne
ng
y
od
od
in
in
hi
eo
hi
Th
eb
yb
Th
et
yt
m
An
An
m
m
So
So
So
L1 speaker CSLE
4%
3% 1%
7% 4%
Something Something
4%
Thing Thing
8% Things
31% Things
Someone 28% Someone
9%
30%
Anything Anything
18% Somebody Somebody
24% 29%
Anybody Anybody
PSLE
4%
5%
Something
Thing
10%
37% Things
6% Someone
Anything
23% 15% Somebody
Anybody
all placeholders: 73 per cent by the L1s and 75 per cent by the Persian
speakers. Eighty-seven per cent of the overall placeholders used by the
Chinese consist of the same three items, which means that the Chinese
used placeholders in a more concentrated way or less uniformly than
the other groups.
The remaining one quarter of placeholders in the L1 and the Persian
data is composed of the other four items, someone, anything, somebody,
and anybody, with similar distribution, the only difference being in the
different ranking of someone and anything. While anything (at 10 per
cent) is more frequent than someone in the Persian data, someone is more
frequent than anything in the L1 data.
Despite the fact that the Chinese performed differently from the
other groups, there is a substantial consistency in its rates of use for the
three least frequent items in the data: someone, anything, and somebody,
each make up 4 per cent of the placeholders.
(29 per cent) and in the Persian by placeholders (28 per cent). These per-
centages are similar in quantity but different in kind. It seems that the
Chinese were more emphatic than the other two groups, using substan-
tially more vague intensifiers. In the same way, the smallest proportions
refer to a different category in each group: the smallest in the L1 data
is subjectivizers (13 per cent), in the Chinese placeholders (9 per cent)
and in the Persian possibility indicators (11 per cent). The proportions
within each of these largest and smallest groups show a rough consist-
ency across the three sets of participants.
5
Pragmatic Functions of
Vague Language
The data show that participants aligned their vague language use with
their communicative needs, stretching vague language to smooth the
flow of communication. The growing body of literature on vague lan-
guage acknowledges the pervasiveness of this often underestimated
feature of natural language in communication (Channell 1994; Cutting
2007; Jucker et al. 2003; Ruzaitė 2007; Zhang 2015). The present
chapter will deal with a functional analysis of vague language in the
three data sets.
Channell (1994, p. 194) provides the following categories of vague
language functions: giving the right amount of information, deliber-
ately withholding information, using language persuasively, lexical
gaps, lacking specific information, displacement, self-protection, power
and politeness, informality and atmosphere, and women’s language.
Drave (2002, p. 26) believes ‘The major function of vague language is
to tailor conversational contributions to the perceived informational
needs of the other participant(s) so as to maintain and enhance the
on-going relationship’. The major functions of vague language are
listed as:
5.1 Mitigation
5.1.1 Self-protection
One of the common uses of vague language is as a tool with which
speakers can protect themselves (self-protection, self-defensive). This
tool ‘attends to the face needs of the speaker’ or serves ‘other-protective
face-work’ (Trappes-Lomax 2007, p. 135). It can also help to protect
the face of the listener or a third party. There seem to be existing areas
of overlap between self-protection and politeness. Discussed under the
term shield by Prince et al. (1982), the uncertainty is found to result
from two different sources: plausibility shield, to express doubt, and
attribution shield, through which the speaker attributes a belief to some-
one else. Below are three examples showing how self-protection works
among the three groups of participants in this study.
(5.1): This is a discussion between two L1 participants over nine turns.
They are talking about online search and database during a tutorial.
while you were online. Now with this awful, Telnet, animal, you
can only really look back a few screens, and I just find it very,
disconcerting. I I wish I could look back, all the way to when
I began, and D S is really the only thing we have, to accom-
modate, that browsing backwards, in your search. <PAUSE:04>
yeah? (L1: 3:194)
S3: I think, in one word, the most important thing about teaching
is, huh, teaching, teach, teach students how to be a man, yeah.
(Ch: 7: 143)
S9: I have an opinion. Huh, I want, I think if you want to be a good
teacher, you have to be three Ps. < SS: LAUGH> (Ch: 7: 144)
S2: Be patient, profession? (Ch: 7: 145)
S9: Performance. (Ch: 7: 146)
S4: Performance, yeah. (Ch: 7: 147)
S6: Passion. (Ch: 7: 148)
S2: Three Ps. Patience, passion, and performance. (Ch: 7: 149)
S1: What about you? (Ch: 7: 150)
S8: Yes, patience is an important factor. Many students pay more
attention to you, in your class. Therefore your class is efficient
I think. I think this is very important. (Ch: 7:151)
S7: I think, huh, I don’t think passion is a very important thing
because I think passion is a temporary thing which cannot exist
long. I think we should choose what we like, and what we love.
Yes, it is the most important. (Ch: 7: 152)
S5: To be a teacher is maybe very easy, maybe not, maybe very dif-
ficult. It is all up to you and, huh, hope you have a bright future.
(Ch: 7: 153)
Pragmatic Functions of Vague Language 117
S6: What kind of job do you think, the, the teachers, what kind of
job? Very honourable or very boring, simply as, huh, as stable job?
Huh. (Ch: 7: 154)
S5: Different people have different answers. (Ch: 7: 155)
S6: What about you? (Ch: 7: 156)
In Excerpt (5.2) the discussion begins with a moral view of the teaching
job by S3 in turn 143, and proceeds with S9 mentioning three criteria,
the three Ps required for being a good teacher. Following S8’s confir-
mation of one of the required factors in turn 151, S7 through I don’t
think in 152 softly expresses disagreement with the criterion passion as
a required element, and suggests love as a replacement. In turn 153, S5
changes the direction of the conversation to whether or not teaching is
an easy job. She first claims it is very easy to be a teacher, but quickly
uses a vague expression maybe in the negative form, then another maybe
to present an opposing possibility that teaching may be very difficult.
Vague language is used here to protect the speaker against opposing
views. The speaker’s hedged view is reinforced by ‘it is all up to you’ in
the same turn, implying that the speaker is going to attribute the valid-
ity (truthfulness) of her statement to the would-be teacher’s discretion.
S5 uses vague language to protect herself against responsibility for her
utterance.
(5.3): This is a discussion between two Persian speakers over five
turns. They are talking about democracy and freedom.
S3: Don’t say ‘but’ look I know if you want to say something, you get
punished in this place, Ok? The knowledge that you see, the peo-
ple that you see, it is accepted. What you see to happen. But over
there, there is no excuse. When you say to yourself this is supposed
to be the land of freedom, these people are supposed to know
everything. Ok? Then you see abuses there … come and see what
they do with, come and see what the story is over there. Find out
about McDonald’s story, find out about Rockefeller, what he did or
he raised all the crisis in South because there was competition and
in the north there was all his, he raised the High Street Times, ok
guys. Find out about John F Kennedy. Then you can see, hey, this
is not Rafsanjani. Yes. (P: 1: 499)
S2: But. (P: 1: 500)
S3: But. (P: 1: 501)
S2: But, you know this level of intellectuality if I am right, you know
here if we have the maximum, one hundred, ok? I think in Iran
118 Communicating through Vague Language
5.1.2 Politeness
When people speak, there are certain linguistic choices they make,
which indicate the social relationship perceived to exist between the
interlocutors. One such phenomenon can arise in expressing polite-
ness. ‘Politeness as it is understood in linguistics involves more than the
common-sense notion of politeness as the conventionalized observance
of certain social norms which spell out the appropriate ways of thank-
ing or greeting’ (Nikula 1996, p. 92). Politeness is mainly associated
with Brown and Levinson’s (1978, 1987, 1994) politeness theory, which
stands on the pillar called ‘face’ raised by Goffman (1967).
Face is defined as ‘the public self-image that every member wants
to claim for himself’ (Brown & Levinson 1978, p. 66). Each individual
needs to look after the face of others in case they need their face to be
maintained: in other words, interactants need to mutually look after
each other’s face. Face in Brown and Levinson’s politeness theory is
divided into two separate but related categories, positive face and negative
face. The former deals with the individual’s desire to be liked and appre-
ciated by others while the latter ‘concerns a person’s want to be unim-
peded and free from imposition’ (Tracy 1990, p. 210). Positive politeness
Pragmatic Functions of Vague Language 119
S12: Well if you can’t identify whose body it is it’s not really so much
an invasion of privacy. (L1: 1:447)
S1: Mhm. Okay, lots of times, a picture of a body uh, you don’t see
the, the face, [S12: yeah] you know, either it’s covered or just the
angle, s- you just see a, you know, a form. [S12: yeah.] um, so
there’s the question of the distance. What about uh, related to
that is how it’s played, right the play of the photo what, what
does the book say about that? I think that’s in there, isn’t it? On
the checklist? See it on the checklist or in one of the case studies.
is there anything about how the photograph is played questions
they ask about how it’s played? (L1: 1:448)
S13: Didn’t it say something like if it was, pl- like, under the fold it’s
not in the checklist but didn’t it say something like it’s, not as
harmful? (L1: 1:449)
S1: Yeah you’re right um, I think there’s another checklist, after the
first checklist. Um, page two-twelve. (L1: 1:450)
S12: Oh that’s right (L1: 1:451)
S1: This was the checklist, [S13: oh] after the um. [S5: oh, yeah] the
campus tragedy. (L1: 1:452)
120 Communicating through Vague Language
5.1.3 Downtoning
Downtoners or detensifiers (Hübler 1983) are what Prince et al. (1982)
call ‘adaptors’. Blum-Kulk et al. (1989) define downtowners as ‘sentential
or propositional modifiers which are used by a speaker in order to mod-
ulate the impact his/her request is likely to have on the hearer’ (p. 284).
They include words such as a bit, a little, and a little bit. Jucker et al.
(2003) state the vague terms ‘introduce vagueness into a proposition or
increase the degree of vagueness of an utterance’ (p. 1746). They also
claim that downtowners are used when speakers find that an available
word does not adequately cover the meaning they have in mind.
Pearson (1998, p. 103) states downtowners often consist of adverbi-
als (for instance just), modal words (for example can), and non-factive
predicators (for example one way of defining a … is). Wu et al. (2010)
maintain that mitigators such as probably and maybe may follow I think,
giving a much stronger downtoning function to it. Ruzaitė (2007,
p. 94) adds that ‘The quantifiers (a) little, a bit and a little bit minimize
the force of verbs and downtone the intensity of adjectives’. Below
124 Communicating through Vague Language
S1: Or your, okay. Well I’ll take a look at e- you know, it’s good_ did
you realize then when you were online when you did that or after
you signed off and you, stewed about it for a while? (L1: 3:192)
S14: After I stewed a little while. <SS: LAUGH> (L1: 3:193)
S1: Okay and, this is sometimes typical, when you’re first learning
how to search. That’s why I try to say don’t revise online, because
sometimes it does take you a little while to sort of figure out, what
exactly did I go wrong, where I wound up with, zero or I wound
up with, thirty thousand, you know sometimes, when you’re
online it’s just too hard to assess that. [S14: yeah unfortunately]
yeah I sort of like the old days, I guess maybe I’m just one of
those, analog print people, when we had these long printouts,
and you could look through that long printout while you were
online. Now with this awful, Telnet, animal, you can only really
look back a few screens, and I just find it very, disconcerting. I,
I wish I could look back, all the way to when I began, and DS is
really the only thing we have, to accommodate, that browsing
backwards, in your search. <PAUSE:04> yeah? (L1: 3:194)
S2: I want to introduce some singers to you just like, huh, Britney. She
has some songs like ‘Everytime’. It is so slow and beautiful. Once
I wanted to train my listening, and it is good to train my listening.
And if you want to improve a high level, you may, you may.
Of course, it is just my suggestion, choose some rap. (Ch: 6: 20)
S1: Rap? (Ch: 6: 21)
S2: It is a little slow, a slow rap, not so quick. I found it just like
Brittany’s Circus, although it is very fast. I like ‘New (xx). It is very
fast. The speed is very fast but (xx) I don’t know how many times
you have, you have heard it. Huh, you will feel it is not slow. Huh,
it is not fast at all and you can hear at your work clearly. I think
it is also a good way. It is up to you what kind of music you like.
(Ch: 6: 22)
S3: Yeah, I think it is, it is a way to enjoy life and some days I told me
that, huh, you, it’s necessary for everybody to learn to sing a song
very well and, huh, only I like ‘Terry Sif’. Yeah, I think her songs
are very beautiful. (Ch: 6: 23)
S7: Ok, I think there is something about this country. You know for
sure I, I am agree with the previous regime, Shah and the King,
huh, 100 per cent, hey, I think they were wrong and I guess even
now but there is, huh, a little chance that if we changed ourselves
at that time, we changed ourselves, we could improve because they
changed the rules. (P: 1: 440)
S3: Infrastructure. (P: 1: 441)
126 Communicating through Vague Language
S7: Yup, but there is some basis for building structures, building, build-
ing a house. You know, you cannot build a two-story building on
some weak basis. (P: 1: 442)
In Excerpt (5.9) S7 expresses his overall agreement with what the previ-
ous regime, the Kingdom, did. He even highlights it with the expression
one hundred per cent. S7 in turn 440 uses the downtoner a little before
chance to underline his view that even changing themselves would not
have been effective, because of the changes in rules that occurred. In
turn 442, S7 gives the example of building a house to show that for
anything to work, some foundation work is needed, just as a building
needs to have a strong foundation. To intensify the point, the speaker
uses two expressions: a two storey building and the weak basis. The latter
is presented in a weaker form, modified by some.
5.2.1 Quantification
Vague language performs the function of quantification when the
speaker realizes that precision is not necessary. This category is exempli-
fied by expressions such as some, many, much, few, little, several, a lot of,
plenty of, and large amounts of. Channell (1994) believes non-numerical
quantifiers help create implicature and thus avoid breaking the maxim
of quantity. Below are examples of how vague language performs quan-
tification functions for the three groups of participants in this study.
(5.10): This is a discussion between two L1s over eight turns. They are
talking about ERIC (Education Resources Information Centre) and web
searching.
through to the end. They’re gonna look at the first ten depending
on, you know what their needs are. (L1: 3:141)
S1: This is actually a re- real important point. Why should it not mat-
ter in this case-searching Dialog? (L1: 3:142)
S15: Cuz everything, should be, as good, as the beginning searches.
(L1: 3:143)
S1: Right. There is no ranking, using Dialog at least, this classic,
Dialog or Dialog Classic that we’re using. A Dialog does have
another system called I believe, Freestyle. Which does do some
ranking, okay? But I think it only gives you the first fifty or, what-
ever. Oh we may get to that or you may wanna do that for one of
you s- uh search reports. Yeah? (L1: 3:144)
S15: So is that okay? (L1: 3:145)
S1: That’s it? Okay, good what facets did you have? (L1: 3:146)
S15: I had three facets, I did pregnancy, um, teenagers and dropping
out. [S1: okay] and then I expanded under each of those and
I didn’t, have as many under pregnancy, um, I limited it to, preg-
nancy slash D F and pregnant students. Um may (L1: 3:147)
S1: Okay, I, I would suggest to you, that there are some more [S15:
okay] like unwed mothers, and early childhood or early parent-
hood or whatever it was, that there some others to use. [S15:
okay] and, since there are so few, I would add more. [S15: okay]
on the other hand you got sixty citations so it didn’t necessarily
hurt but if you wanna be more comprehensive, you wanna add
more descriptors. Any other experiences on the search? Yeah.
(L1: 3:148)
of facets she could have used. This can be viewed in the examples she
gives through like or even emphasizes by or whatever it was, which indi-
cate that she is reinforcing the quantification by giving examples and
using or whatever.
(5.11): This is a discussion between three Chinese participants
over three turns. They are discussing public transport in Beijing and
Shanghai.
In Excerpt (5.12), when asked about the destiny of the pilots, S3 uses
some twice to express a vague quantity in turn 180, assuming this quan-
tification, rather than an exact number, can best fit the context. S3 may
or may not know the exact number, but apparently thinks that it is not
important here, and some serves the purpose better; otherwise, in the
next turn, S4 would have asked for clarification of the vague quantity
expressed by some. That did not happen, meaning that the vague infor-
mation was accepted by the listener with no fuss.
5.2.2 Emphasizing
Most intensifiers perform an emphasizing function. Intensifiers have
been referred to as boosters (Holmes 1990; Hyland 2000). Boosters are
mostly studied in comparison with hedges (Bradac, Mulan & Thompson
1995; Holmes 1990; Hyland 2000). Investigating the functions of boost-
ers in writing, Hyland states that ‘Boosters like clearly, obviously and of
course allow writers to express conviction and to mark their involve-
ment and solidarity with an audience’ (2000, p. 179).
Besides the discussion of intensifiers (boosters)/ hedges dichotomy
available in the literature, gender-related examinations of the applica-
tion of these forms make a substantial contribution to the study of
vague language (Bradac, Mulan & Thompson 1995; Holmes 1990).
Holmes’ work shows significant gender differences in the use of dif-
ferent boosters. Bradac, Mulan and Thomson (1995) demonstrate that
women show more consistency in using intensifiers than men. This has
been supported by other studies as well (McMillan et al. 1977; Mulac &
Lundell 1986; Mulac, Lundell & Bradac 1986; Mulac et al. 1988).
Wright and Hosman (1983) claim the overuse of intensifiers by female
speakers brings more interactiveness to their side in communication.
Even the context of communication has been claimed to contribute to
differences in the language used: Bradac et al. (1995) note that women
130 Communicating through Vague Language
use more intensifiers when talking to women but more hedges when
talking to men.
Ruzaitė (2007) maintains intensification can also be expressed
through quantifiers. For instance, multal quantifiers can emphasize a
large quantity or long period of time. The other possible ways to add
emphasis, Ruzaitė states, are repeating the same quantifier (lots and lots)
or placing an intensifying premodifier (really) in front of a quantifier.
Below are examples of how intensifiers are used in the data to fulfill
emphasizing functions. The examples are not gender-specific but are
viewed from culturally and linguistically distinct perspectives.
(5.13): This is a discussion between five L1 participants over four
turns. They are talking about the death of a race car driver and the safety
rules applying to race car driving.
S3: Is there any posi- possible cause of death? I mean, if you s- you,
I saw it on T-V. So, you see just, crash. I mean and he was one of
the top, race car drivers right? And so I mean it. (L1: 1:38)
S1: So o- I mean obviously [S3: why do they need them] his death was
caused by the crash so that’s not the question [S7: they’re b-] the
ques- the question is what aspect of the crash specifically caused
his death? (L1: 1:39)
S6: There were a lot of questions about the type of restraints um, what
s- [S2: (like how to change)] like there’re so many different seat
belts and there’s certain ones, yeah. (L1: 1:40)
S1: Like how did it start? What happened to him at the moment of,
impact? Which part of his body, made contact with. (L1: 1:41)
things, but in smaller cities maybe we have, huh, less such activi-
ties. (Ch: 4:120)
S3: My friends in Jason University, in holidays, they are only at home
five days, so after that they will go back to school. They were study
hard, huh. For instance, TOFO and do some experiment, huh, do
many experiences about school. So it is very, they are very busy
and, and. They are, life is very full. (Ch: 4:121)
S4: I think our university life is boring. It is too boring. It is really bor-
ing. Just study and, huh, study, study. We have no to, we have no
chance to. Ok, I want to be a volunteer and to, huh. (Ch: 4:122)
S5: We have no chance. (Ch: 4:123)
S3: As our parents had, for example, the problems that were between,
the problems that were between my parents with their parents,
huh, were more than our problems, ok? And in the future our
problem will have less I think. (P: 7:322)
S1: This generation is getting more aggressive. (P: 7:323)
S2: Actually, I think your problem will be worse, not less. (P: 7:324)
S5: Exactly. (P: 7:325)
S2: You know nobody can know everything. He is right. Lack of
information can cause problems. I believe personally being a par-
ent comes with a great responsibility. You have to have really,
really. You have to be really talented to be a parent and you have
to have lots of responsibility. For example, if you want to have a
child, you have to think over everything. You have to know a bit
of psychology. You have to know a bit of, I don’t know, whatever.
(P: 7:326)
S1: Sociology. (P: 7:327)
132 Communicating through Vague Language
S2: Sociology, yes. You have to be into politics. You have to know lots
of things to be parent. Some people think it really really ideal a
child is coming, is growing up, is as easy as that but it is not this, it
is not this. I mean he is right. If his parents know about computer,
maybe they would encourage him to do it. I mean my daughter
was. Sorry if I keep examples of my children. I am just speaking of
my experience. (P: 7:328)
In Excerpt (5.15), S3 is of the opinion that the gap between the new
generations will be narrower, based on observations of his parents’ and
his generations. He is disagreed with by other interlocutors. S1 uses more
aggressive in turn 323 and is disagreed with by S2, with ‘will be worse
not less’ in turn 324. In turn 326, as an actual parent, S2 gives details
of requirements to be a good parent. She adopts an emphatic approach,
trying to address the problems facing parents and the responsibilities
parents should feel. The first factor she counts, talented, is emphasized
by the intensifier really; then she refers to the broad concept responsibil-
ity again and uses a quantifier, lots of, for emphasis. So she first empha-
sizes by using really and then uses a different phrase, lots of. This shift
between intensifier and quantifier for emphatic purposes is also seen
in turn 328: S2 uses the quantifier lots of first, to refer to what has to
be known to be a parent, then resorts to an intensifier to highlight the
emphatic tone: really really ideal.
5.2.3 Possibility
Prince et al. (1982) examine the possibility function under both approx-
imators and shields. In this study, it serves to express different degrees
of possibility.
(5.16): This is a discussion between six L1s over ten turns. They are
talking about how social control works in the society and why people
do what they do.
S1: Sure. I mean, I I mean, social control is obviously, not perfect,
so um <PAUSE WHILE WRITING ON BOARD> so yo- so young
people. Um what does that tell us about young people, um,
if young people are more likely to say, steal something? You
might wanna talk to her after class just to, find out what she’s
doing and, whether you wanna participate in it. (L1: 2:8)
SU-m: that’s you. (L1: 2:9)
S3: Me? Okay. <SS: LAUGH> (L1: 2:10)
Pragmatic Functions of Vague Language 133
S1: Um, okay so what does that, what does that tell us already, if if
young people, are more likely to steal things than, than o- um,
older people? Yeah. (L1: 2:11)
S4: Either that they’re, more rebellious, or maybe, or just that
they’re not as accustomed to, society’s norms yet. (L1: 2:12)
S1: Yeah that’s, yeah, I mean, yeah rebellious, or, or not we, we
could call it socialized, which basically just means they’re not
accustomed to society’s norms yet. Yeah. (L1: 2:13)
S2: Or they’re just too young to understand the uh, consequences
of, stealing (L1: 2:14)
S1: Oh that’s yeah, and this is important um, the consequences …
somebody else have something they wanted to say? (L1: 2:15)
S5: There also like, isn’t like as many consequences for them.
(L1: 2:16)
S1: Yeah, that’s true, I mean, um why do you think that there are
not as many consequences? I mean why do, why do um, why
does society_ why is our society set up so there won’t be as
many consequences? (L1: 2:17)
S2: My pen pal will come to see me, who is doing a course in Oriental
studies. I will go along with her to Japan. (Ch: 1: 168)
S1: Wow, so it, sounds really interesting. How do you get along?
(Ch: 1: 169)
S2: We’ll probably use one of those very fast and poor trains to get
there and then go by taxi or on foot. (Ch: 1: 170)
134 Communicating through Vague Language
S1: Oh, it’s a good choice and how do you go around? (Ch: 1: 171)
S2: On foot or by boat? (Ch: 1: 172)
S1: And, huh, what do you want to wear? (Ch: 1: 173)
S2: Huh, it depends on time of the year. I would want to go there
in spring for the cherry blossom, so probably just jeans and a
sweatshirt. I’d make sure I had a clean pair of, pair of, socks or and
or some slippers shoes because I think you have to take off them
when you visit the temples. (Ch: 1: 174)
S1: Yes, it is a good idea. What would you buy? (Ch: 1: 175)
S2: Nothing. Tourist things. I might buy an electronic gadget like a
calculator. They’re supposed to be cheap in there. (Ch: 1: 176)
S1: Huh, what would you eat and drink? (Ch: 1: 177)
S2: I’d look for ‘Western Food’ and probably end up eating at
McDonalds. She can’t stand raw fish and she doesn’t like rice
much, either. (Ch: 1: 178)
S7: It’s easier when you write something or speak with someone; talk
with someone on the phone. It’s easier for you to talk some issues
or problems that you cannot tell them face-to-face. (P: 2:75)
S1: Ok. Sometimes writing is much better, huh, when you cannot
speak easily and speak some (P: 2:76)
S8: Problems and a– (P: 2: 77)
S9: But, it isn’t common, you know, the writing. (P: 2:78)
S1: Ok, why in our daily life, sometimes we want to speak with each
other, instead of saying directly, ok? (P: 2:79)
S9: Yes. (P: 2:80)
S1: We say to our partner. Ok. Go home I will call you, yes? Why? (P:
2:81)
S2: Because we are are ashamed of. (P: 2:82)
S8: Because by phone we can talk together easily. (P: 2:83)
S7: Maybe we want plenty of time to speak in a better condition, in.
(P: 2:84)
S1: In a more relaxed situation? (P: 2:85)
S7: Yes. (P: 2:86)
S5: Maybe we need sometimes to prepare ourselves to say that. (P:
2:87)
S1: So these kinds of instruments help us. (P: 2:88)
5.2.4 Uncertainty
Lack of information brings about uncertainty, associated with Grice’s
Maxim of Quality which is stated as ‘Do not say that for which you
lack sufficient evidence’ (1975, p. 46). Channell (1994) discusses uncer-
tainty under displacement, which occurs mostly when talking about
past and future events, adding that there are instances that go beyond
tense constraints, as uncertainty exists in the present as well. Channell
points out that vague language is resorted to when the speaker feels
stressed by the uncertainty of the subject, lack of knowledge/vocabu-
lary or an unequal relationship between participants. Examples of how
uncertainty is expressed through vague language across the three data
sets are discussed below.
(5.19): This is a discussion between two L1 participants over nine
turns. They are talking about an editorial on a crime.
S1: Hm. There’re some more examples, from, the case studies yeah?
(L1: 1:466)
S12: Um the, the, the kid that got shot outside the convenience store
[S1: mhm] the editor, um, wrote wrote a, column was it the same
day of the paper? I’m not sure, but he wrote a column explaining
why they, [S1: yeah] why they ran it cuz at first he didn’t wanna
run it. (L1: 1:467)
S1: You can actually see it, y- you can see the column, [S12: yeah]
tu- turn the page. (L1: 1:468)
S12: Yeah, oh. (L1: 1:469)
S1: Well it’s in there somewhere I don’t know. (L1: 1:470)
S12: Yeah yeah it’s right here. Oh wait, (L1: 1:471)
S1: Th- (L1: 1:472)
S12: No no no, that’s not it. (L1: 1:473)
S1: I know that it’s in there somewhere. I just saw it. Anybody r- (L1:
1:474)
In Excerpt (5.19), S1 asks about more examples and S12 refers to one
that was published in the paper, but there is uncertainty as to where
in the newspaper the column appears. In turn 468, S1 tries to help S12
locate it and asks him to turn the page. In turn 470, the vague term
Pragmatic Functions of Vague Language 137
S5: Shean, what are you going to do when you grow up? (Ch: 7: 9)
S1: Huh, when I was a small child, I always wanted to be a teacher.
Maybe because teaching is the only profession I’ve seen and
I’ve had this dream for about, huh, many years. But and I chose
university teacher when I was in college. But, now, when I, huh,
when I, huh, graduate from this university, I already have doubt
that if I really want to be a teacher in the future, maybe, I will
choose another job for me in a later time. What about you?
(Ch: 7: 10)
In Excerpt (5.20), when asked about her future job, S1 says that there is
a gap between her childhood dream job and the job she might pursue
in the future. She highlights the transition from her childhood job
to a partially different position by pointing to her career at college.
When it comes to her current favourite job she is unsure, saying ‘I
already have doubt if I really want to be a teacher in the future’ in
turn 10. The uncertainty is reinforced by the following vague word
maybe, where she explicitly talks about choosing a different job at a
later stage in life.
(5.21): This is a discussion between five Persian speakers over twelve
turns. They are talking about the reasons for cave paintings.
S1: And go to the past, old time. Why, for instance, in cave now we
found some pictures? (P: 2:103)
S3: Yes. (P: 2:104)
S1: Yes? You know what is the aim of this kind of pictures? In the
cave? (P: 2:105)
S5: It’s I think … (P: 2:106)
S1: By that writing they want to communicate with the next genera-
tion. (P: 2:107)
138 Communicating through Vague Language
In Excerpt (5.21) S1, who raises the question of the reasons for cave draw-
ings, tries to answer it in turn 107. The answer is agreed with by S7 in
turn 108, but in turn 109, S1 says Yes?, expecting more detailed replies.
S1’s strategy proves to be effective, as S5 proceeds with a reply (religious
reason) that contains some uncertainty (in the exploratory vague term
I think). In turn 111, S1 tries to elicit more responses by making points
such as ‘feeling alone’ as a reason for cave drawing. As the discussion pro-
ceeds, more reasons come up, and more uncertainty is revealed in replies
expressed through maybe by S7 in turn 113.
5.3.1 Repair
The study of speech repair has been significantly influenced by the
work of Schegloff, Jefferson and Sacks (1977), which introduces repair
sequences as segments. The sequence involves a repair-initiating turn
followed by a coming turn that creates the outcome of the repair,
referred to as execution of repair by Rutter (2008). Either participant
in the conversation can produce these turns. Schegloff et al. (1977)
divide repair sequences into four types, summarized by Rutter (2008,
p. 36) as
S2: Yes, I agree with you. I’ll try my best. (Ch: 2:60)
S5: I am going to be educated further, huh, because during these
years, my second degree is Chinese, so I want to be an editor, so
I want to go to a newspaper office. Maybe, I want to be a teacher
because you know, as a teacher the happiest thing is giving the
knowledge you have learned to the, to your students and you,
and you see them grow happily. So this is my goals. (Ch: 2:61)
S3: I think I will find a job first. Maybe, it’s very hard but just like
Yang Fan [one of the students, noted by the author]I think huh the
four years study really cost my parents so much and I really want
to support myself. So if you, I can, I think, I will find a job. I did,
I don’t know what kind of job I will get, so I have many plans.
142 Communicating through Vague Language
S3: That is the price that you have to pay for your own dignity.
(P: 1:360)
S2: You. (P: 1:361)
S12: Alan, it is like that when you go to the nature. Ok? And you see
everyone threw out their garbage in the environment. Ok? Huh,
you said that I want to change myself, I would not do that. Huh,
maybe the others don’t do that, but I actually it is you are saying
it is about yourself, that’s you did it. (P: 1:362)
S3: Said it in an example. (P: 1:363)
S12: And maybe, and maybe. The others when they see you. (P: 1:364)
S2: They learn it. (P: 1:365)
S12: Yeah, they learn it. (P: 1:366)
In Excerpt (5.24), S12 points out, in turn 62, some common problems
faced quite often in the society, and tries to present a simple example
that demonstrates how changing oneself can contribute to the evolu-
tion of a society. The speaker makes a false start where he refers to a
Pragmatic Functions of Vague Language 143
5.3.2 Hesitation
Hesitation markers are defined as ‘a set of tools with certain time dura-
tion that are used to solve oral discourse generation and reproduction
problems and that can be both retrospective (for example correction
of a produced discourse piece) and perspective (for example planning
problems of the coming discourse piece)’ (Khurshudyan 2006, p. 1).
Stubbe and Holmes (1995) believe that discourse markers provide
speakers with verbal planning time, but this function has rarely been
attributed to vague expressions. Wiese (1984) claims hesitation may
appear in different forms such as filled pauses (for instance uh, mhm),
repetitions, corrections, and drawls. Focusing on I think by Chinese
learners of English, Wu et al. (2010) find that EFL learners use this
marker to show difficulty in finding the right word to use next. The
following examples present instances of hesitations occurring among
the three groups.
(5.25): This is a discussion between four L1 participants over 13 turns.
They are talking about why a photo of an accident was chosen from
among other photos to be published in a newspaper.
S17: If they had to, sh- show one, of the a- from the accident scene.
(L1: 1:515)
S1: And why would you prefer that? (L1: 1:516)
S17: M- I don’t know. <SU-f: LAUGH>. (L1: 1:517)
S1: Okay. (L1: 1:518)
S17: Cuz it doesn’t, I don’t know I just, think seeing dead bodies in
a newspaper I don’t (L1: 1:519)
S1: I mean it, it bothers you. (L1: 1:520)
S17: Yeah. (L1: 1:521)
S1: Okay. Yeah? (L1: 1:522)
S3: I, I mean I agree that it it’s bothersome and that would be help-
ful but then if you you know after the explanation was made if
you look at, um the letters that they received um, I was kind of,
shocked to think that the reader would I mean that some of these
would be, <CLEARS THROAT> excuse me sent into the newspaper
about, if you think about basically, you know if you think about
how many lives were saved from this or if somebody thinks of
this image, it’s it’s it is upsetting and it is powerful and that’s why
144 Communicating through Vague Language
In Excerpt (5.25), S17 expresses her preference for a particular photo but
once asked the reason for the preference by S1 in turn 516, proceeds
with a quick answer, I don’t know in turn 517. Although S1 implies this
answer is sufficient, S17 decides to elaborate on the reason for her pref-
erence in turn 519, but still finds the reason elusive. She uses the same
answer as before, I don’t know, but this time it functions as a device to
help her cope with her hesitation. In other words, this device allows her
to buy more time to think of an answer. There seem to be other markers
which highlight the state of hesitation by S17, like I just, think right after
I don’t know and even the final position I don’t, when the speaker meant
to say I don’t know again but was interrupted by S1.
S3 in turn 523 expresses agreement that publishing the photo may be
bothersome, but raises the advantages it has brought by making refer-
ence to the letters the newspaper received. Her attempt to specifically
point out one of the advantages raises the need to use some devices
to handle the hesitation brought up by memory lapse. She resorts to
expressions basically and you know to buy time to think and arrive at the
implication of this advantage, which is many lives saved.
(5.26): This is a discussion between five Chinese participants over
eleven turns. They are talking about Disneyland in different countries.
S1: Have they said we will not publish the photo? Or have they just
said we don’t intend or so … (L1: 1:66)
S12: (xx) We don’t intend to. (L1: 1:67)
S5: Right. (L1: 1:68)
S1: That’s different from saying we won’t, won’t do. (L1: 1:69)
S12: Right. We don’t intend to well what if something else comes up?
(L1: 1:70)
S5: Right the wording was is has no intention of publishing photos.
(L1: 1:71)
S1: No intention of publishing. (L1: 1:72)
S5: It was their attorney who said the photos are important because
they might reveal what caused, [S1: mhm] Earnhardt’s death. [S1:
mhm] and then the other th- as I said before the other thing is that
that, in Florida, granting public access to autopsy photographs is
permissible if it’s not part of a criminal investigation. [S1: mhm] so
the other feather in their cap is the fact that if it was another state
then we we might not be having this argument. (L1: 1:73)
S1: I guess another question would be has this happened in the past?
And it, has there been an autopsy photo that, news media have
obtained access to? You don’t happen to know do you? (L1: 1:74)
S5: Uh <SS: LAUGH> (L1: 1:75)
S1: Not not that you’re the expert on this uh, (L1: 1:76)
S2: Thinking back to who’s died recently no. (L1: 1:77)
Pragmatic Functions of Vague Language 147
In Excerpt (5.28), S12 is probably the one who talked to the people in
charge of the journal and is passing the answer to others. Or he might
have read it in the journal, or heard it from a representative on TV. S5
continues the discussion by pointing out the importance of photos as
stated by the attorney, and also to the requirements for the photo to be
published. S1 in turn 74 tries to take a turn by using the vague expres-
sion I guess to make a new point as to whether autopsy photos have
previously been made available to the news media.
(5.29): This is a discussion between four Chinese participants over
fifteen turns. They are talking about how culture can contribute to dif-
ferences in communication.
S5: Ok, you are talking about the psych, you know, actually the physi-
cal problem children will face. (P: 7:70)
The pattern drawn from the data shows that the expressions used most
freely in multiple positions in the clause (such as subjectivizers) are used
more often by the L2 groups, especially the Chinese. With subjectivizers,
the differences between the three groups are statistically significant
(p < 0.05, χ² = 410.347, d.f.2), with the Chinese using the greatest and L1
using the least. Subjectivizers are employed most often by the Chinese,
and the L1s are found to be the least frequent users of this category.
As with subjectivizers, possibility indicators are significantly different
among the three groups (p < 0.05 χ² + 71.755, d.f.2), with the Chinese
using them most often and the Persian speakers least.
Subjectivizers can appear in clause-initial, clause-mid or clause-
final position. This is where the elasticity of vague language can arm
154 Communicating through Vague Language
CSLE
PSLE
L1S
Possibility indicators
CSLE
L1S
PSLE
Vague quantifiers
CSLE
PSLE
L1S
Vague intensifiers
CSLE
L1S
PSLE
Placeholders
PSLE
L1S
CSLE
Figure 6.1 Positions of the three groups in the use of five lexical categories
(6.2): This is a discussion between two Persian speakers over two turns.
They are talking about how social injustice emerges.
S5: You must hit everybody who violates the others’ rights. You know
I think social injustice is because of some people opinion. Because
they think they are better than the other race, racism. (P: 6:1259)
S4: Ahuh. (P: 6:1260)
156 Communicating through Vague Language
In Excerpts (6.1) and (6.2), I think in each sentence covers the entire
sentence. In (6.1) it applies to the whole sentence following but, and in
(6.2) it refers to the entire sentence regarding social injustice. In con-
trast, in Excerpts (6.3) and (6.4) below, which exemplify the emphatic
nature of intensifiers, vagueness is locally linked to the segment which
immediately precedes or follows it rather than to the entire sentence.
(6.3): This is a discussion between two L1 participants over two turns.
They are talking about a photo in a newspaper.
S20: Yeah, I just grabbed this from the Michigan Daily which I thought
was really interesting that this is a normally a colour, uh daily, [S1:
mhm] and they got Ellerbe here in black and white. (L1: 1:735)
S1: Wonder why. (L1: 1:736)
S1: Well yeah and there’s, <SS: LAUGH> there’s another, there’s another
question. This is a really good, a really good point that you bring
up. Um, I mean we say that, and maybe you would I mean I’m not
saying you wouldn’t, but um there’s lots of people in the world, take
Bill Gates you know for an example, who um, make lots of money,
have lots of money or other people who inherit lots of money. They
don’t have to work, but they do. (L1: 2:82)
S16: It makes ’em happy. (L1: 2:83)
distribution between four items in the L1 data: maybe (26 per cent),
may (24 per cent), might (24 per cent), and probably (18 per cent). The
values in the Persian and Chinese data reveal huge inconsistencies.
Maybe, with a frequency of 81 and 82 per cent respectively, indicates
that the L2 speaker groups inclined heavily toward this possibility
indicator. Maybe was used by the Chinese the most, while at the other
extreme the L1 group utilized far fewer maybe and the least maybe
clusters. The Persian speakers sat between the Chinese and L1 groups
in their frequency of use of maybe and its clusters. If we assume that
speakers who use maybe tend to be more tentative, then the Chinese
are more tentative than the Persian and L1 speakers, the Persian speak-
ers less tentative than the Chinese but more than the L1s, and the L1s
least tentative of all.
The Persian speakers preferred all placeholder items the most apart
from thing, favoured by the Chinese. While placeholders (with 478
tokens, 28 per cent) were the most frequent vague expressions of the
Persian speakers, the Chinese used this category as the least common
group of vague words (286 tokens, 9 per cent); the L1 group used
301 tokens (19 per cent). Placeholders seem to play a more promi-
nent role in the Persian speakers’ interactions than in the other two
groups’.
In the data, some clustering patterns appear that are uncommon in
English. The unusual patterns appearing in L2 speaker data may have
originated from insufficient exposure to English, which drove the par-
ticipants either to use their own creativity in using clusters or to borrow
relevant knowledge from their own L1. For instance, we may, which
does not appear in the L1 data, is used by the Chinese and the Persian
speakers to show that the speaker is attempting to say something indi-
rectly with a tentative tone. Specifically, we in this context is used by the
Persian speakers to show politeness by creating some kind of intimacy,
avoiding the explicit and direct disagreement and warning.
The Chinese and Persian speakers have a cultural preference
for indirectness in making a request or expressing disagreement
(Gudykunst et al. 1996; Zarei & Mansoori 2007). Additionally, we is
the subject pronoun a Persian speaker usually uses instead of I, to
show politeness. This is confirmed in the data, which shows that the
Persian speakers do not use I may; but the cluster occurs frequently in
the L1s’ discussions.
(6.5): This is a discussion between two Persian speakers over three
turns. They are talking about why it is thought that nuclear power is
needed in a country.
158 Communicating through Vague Language
In Excerpt (6.5), S6 does not agree with S8 but is trying to indirectly say
so by using we to involve the interlocutor in the negative reply. This
also indicates a listener-oriented approach by the Persian speakers (see
Section 6.9 for further discussion). S8 starts turn 39 with and and huh,
two hesitation markers signalling the time needed to find the most suit-
able indirect strategy to express disagreement.
The clusters of may not and maybe + not demonstrate discrepancies
between the L1 and L2 speaker groups. While may not and might not are
found to be used more frequently by the L1s, the Persian speakers show
an inclination to use maybe + not. As explained in Section 6.5, which
deals with the influences of a first language on a second, this trend
seems to have emerged from the closeness of maybe to the structure
mainly used to express possibility in the Persian language: that is, maybe
or maybe + negative are structures which the Persian speakers are accus-
tomed to using in their L1 language and the similarity of the structures
between the two languages leads the Persian speakers to prefer maybe
not to may not and might not.
The most different patterns are observed in Table 4:14, the distribu-
tion of placeholders. Two major discrepancies are the large overall
number of uses by the Persian speakers, and the inconsistency in
using clusters of placeholders across the three groups. For instance,
something clusters occur in the Persian data 36 times, in the Chinese
eight times, but never in the L1s data. The clusters with anything show
only one type shared among the three groups. The Persian data pro-
vides the highest number of anything clusters, the Chinese the least.
There are six types of anything clusters in the Persian data, with a total
of 32 occurrences, while the 14 occurrences in the L1 data are of four
types, only two of which are in common with the Persian speakers.
The Chinese data reveal only two types of anything clusters over seven
occurrences.
There are two reasons for using placeholders: they appear to help
speakers cope with no need or inability to speak specifically. They have
the potential to be used almost like a discourse marker to enhance com-
munication. Placeholders are used by L1s when there is no need to speak
specifically, or to compensate for the lack of a word or the inability to
remember the appropriate word; as the Chinese and Persian speakers
General Discussion 159
S5: You are talking about parents who are not young enough to know
about computers. When there is a … there … (P: 6:877)
S2: My parents don’t. My parents don’t can’t work with computer,
too, either. Huh, they don’t know lots of things. They didn’t.
They know a lot of things but about these kinds of things that are
I don’t know common in nowadays. They, most of them time they
don’t agree with that with these, ok? But I try to have a relation-
ship with them, that it solves our problems, ok? For example, I do
my works, I do my, huh, I got my ways, ok? And they just, huh,
watch me send, huh, I don’t know how I can say. (P: 6:878)
S1: This is their way. (P: 6:879)
S4: It is very useful for his or her future, for his own development.
I think that is what a teacher should do and we can also make
friends with our students that are quite interesting. And to be a
teacher is not just about teaching, just about giving, just about
giving the same lessons. If we do just this, that might be a little
bit boring. Yeah, that’s what I think. (Ch: 7: 57)
The occurrence ratio between a vague expression and the total word
count in the data, namely the density of vague language, may present
another perspective as to how these expressions occur differently across
the three groups of participant.
As Table 6.2 shows, on average, there is one subjectivizer per 250
words in the L1 data, whereas this vague expression occurs more fre-
quently, once per 182 words, in the Persian data. The Chinese data
show a more concentrated use at once per 69 words. Possibility indica-
tors occur once per 270 words in the Persian data, fewer than in the L1
once per 216 words. As with subjectivizers, the number of uses in the
Chinese data is the greatest, at one per 135 words.
General Discussion 161
Vague quantifiers occur once per 122 words in the L1 data, slightly
less often than in the Persian at once per 118. The Chinese data again
demonstrate high use, once per 69 words (the same density as subjectiv-
izers). Vague intensifiers appear most often in the Chinese data, once
per 58 words, and least in the Persian at once per 154. L1s used vague
intensifiers once per 129 words. The Chinese used fewest placeholders
at once per 179 words, followed by the L1s at once per 171 and the
Persian speakers at once per 107. This density analysis acts as a supple-
ment to the frequency discussion presented so far.
The Chinese speakers used 741 vague words acting as subjectivizers,
732 of them I think: 99 per cent of their total use of subjectivizers. Both
the Persian and L1 speakers used this subjectivizer item the most as
well, but at slightly lower concentrations (73 per cent and 79 per cent
respectively).
I think may be heavily used by the Chinese because it is being
employed as a discourse marker, appearing where a discourse marker
might seem more appropriate. This trend supports the pattern that
emerged in Wu et al.’s (2010) study, in which Chinese speakers were
found to use I think as a filler. Such usage may meet a speaker’s need for
a filler, achieved by stretching a piece of vague language (Zhang 2011).
Other reasons for the popularity of I think may lie in the need to elasti-
cize vague language, as a turn-taking device, or as a cognitive processing
focus (see sections 6.7 and 6.9 for details). Like subjectivizers, vague
intensifiers have the most concentrated distributions. Very makes up
more than half of the vague words in this category in the Chinese data,
with the remaining use spreading over five other items; the L1 group
concentrates mostly on really.
The most even distribution of possibility indicators is observed
among the L1s, with the first three items in Table 4.3, maybe, may and
162 Communicating through Vague Language
Individuals tend to transfer the forms and meanings, and the distri-
bution of forms and meanings, of their native language and culture
to a foreign language and culture—both productively when attempt-
ing to speak the language and to act in the culture, and receptively
when attempting to grasp and understand the language and the
culture as practised by natives (Lado, p. 2).
This can be seen in the use of possibility indicators by the two L2 groups
in this study. The Persian speakers seem to have been influenced by
the transfer of maybe from Persian, leading to less frequent use of other
similar items, may and might. These generally appear as expressions
indicating politeness rather than possibility, imitating the function of
the equivalent words in Persian. The Chinese, by contrast, use may at
a frequency close to that of the L1s, because in Chinese yexu (may) can
indicate possibility as well.
General Discussion 163
(2004) reports that the two words fall in the lower rank in the modal
auxiliary list by the L1s, might sixth and may eighth as the penultimate
item (indicating that even L1 English speakers do not make much use of
these terms or English L1 speakers can’t get the terms right), while can,
should and must occupy higher ranks.
Maybe is the only possibility indicator the L2 groups use much more
than the L1s. It occurs in clause-initial position 85 times, or to 55 per
cent of total possibility indicators for the Persian speakers, 128 times
and 41 per cent for the Chinese, and 11 times and to 17 per cent for
the L1 group. The popularity of this item may lie in its versatility, as it
has the potential to be used in different positions in a clause. Both the
Persian speakers and the Chinese use around half of the overall tokens
of maybe in their clause-initial position: the Persian speakers use possi-
bility indicator items at the beginning of clauses in their L1 and maybe
seems as the most compatible with this pattern. As it is quite common
to use an equivalent of maybe at the beginning of the clause in Persian,
this tendency is also transferred to English. The situation is similar in
Chinese, where maybe can be used in clause-initial position, especially
in spoken exchanges. The tendency to use maybe in English structures
that resemble L1 structures, and the availability of this possibility
marker in both L1 and L2, facilitate the transfer of this item from one
language to another and accounts for the heavy presence of maybe and
the related low appearance of other possibility markers in the L2 data.
(6.10) These are two turns by an L1 describing a photo.
S4: People think they should be very brief in everything they, for
example, when they are in a line, they try to go earlier, to for
example buy something or when they are in taxi lines, they try to
go to the taxi sooner than the others. (P: 6:149)
S2: Ok. (P: 6:150)
S4: I think that it is wrong. (P: 6:151)
and five by the Chinese. The preference for countable nouns among
the Persian speakers appears in this study in the form of 11 times of
some + countable noun; the other groups do not have examples of this
pattern.
The low frequency of lots of (11) and high frequency of a lot of (85)
in the Chinese data may be because the Chinese language is insensi-
tive to the distinction between countable and mass nouns. The Persian
speakers use there are lots of five times, while the L1 data does not show
any such combination and the Chinese use it only once. By contrast,
the L1s use there is lots of five times, but this form is non-existent in
the Persian data. The proportion of countable nouns to mass after lots
of reveals that the L1s and the Chinese use them indiscriminately,
whereas the Persian speakers prefer countable nouns to mass nouns by
three to one. This difference may originate from the system of plural
and singular nouns in Persian: a study by Sharifian and Lotfi (2003) on
mass–count distinction finds that Persian speakers’ concept of mass–
count nouns allows them to use some mass nouns as count nouns
(plural) in certain contexts, and that the mass–count distinction across
different languages arises from ‘underlying discrepancies in conceptu-
alizing experience that is being coded in linguistic expression’ (p. 229).
What this implies is that the English, Chinese and Persian mass–count
systems may be different, and that Persian speakers may have devel-
oped a more flexible count nouns system that they prefer to use over
mass nouns. This may explain why the Persian speakers in this study
preferred to use count nouns with quantifiers that otherwise apply to
both mass and count nouns. Due to such discrepancies in mass–count
distinctions, quantifiers are sometimes used quite differently between
L1 and L2 speakers of English.
L1 influence manifests in the use of vague intensifiers as well. The
occurrence of so and too across the three groups of participants displays
two overall trends: the first is that there is consistently heavy use of
these items by the Chinese, but a different pattern is revealed between
the Persian and L1 speakers. The Persian speakers, with 15 uses, were the
least frequent users of too, which the L1s, with 24 tokens, used nearly
twice as often. However, the Persian speakers presented 75 occurrences
of so, while the L1s’ 40 uses made them the least common users. The L1s
seem to resort to other intensifiers to compensate for the low frequency
of so in their data.
(6.13): This is a discussion between two Persian speakers over two
turns. They are talking about discrimination in their society.
General Discussion 167
S1: So it’s an exception to the rule. Case study thirty-five photo digi-
tal cover-up very important. This issue is extremely important.
I’m sorry they don’t have more case studies, like this because …
(L1: 1:685)
1
Persian word meaning ‘someone from a privileged family’.
168 Communicating through Vague Language
In Excerpt (6.15), the L1s use too small to emphasize the smallness of the
room. In Excerpt (6.16) below, S5 uses so busy to confirm S2’s statement,
which employs too busy. The frequency of so + adjective in the Persian
data brings up the number of occurrences of so.
(6.16): This is a discussion between two Persian speakers over two
turns. They are talking about how stressful study is for a student who
is also a parent.
(6.17): This is a discussion between two Persian speakers over two turns.
They are talking about discrimination in their society.
S2: Yes? And how can we improve our culture? For instance practic-
ing? What? What should we do? Training? (P: 6:434)
S5: But you know I think our culture is very hard to change because
one of … (P: 6: 435)
S3: No, very easy to change. (P: 6:436)
Excerpts (6.18) and (6.19) demonstrate that the Persian speakers tended
to use very rather than too to intensify some adjectives, as is the case
for the Chinese as well. The Chinese use very heavily as many as 498
occurrences, and the Persian speakers use 108, the least user is the L1
group with 79 occurrences. It is possible that many of very in L2 data
could be replaced by too.
The influence of a first language contributes to discrepancies in the
choice of vague intensifiers when speaking in English. Too does not
exist in Persian but so and very do, and the Persian speakers found
them more comfortable to use where L1s used too. The same applies
to the use of quantifiers, explaining why the Persian speakers used
neither a few nor few. The influence of L1 can sometimes result in
avoiding, or using only rarely, other items: for Persian speakers, these
include such concepts as countable or mass nouns, which are not
distinguished in Persian. It may not be appropriate to conclude that
Persian speakers are unable to use them appropriately, but fairer to
argue that learners of English have not yet reached a level at which
mastery of such cross-language differences in the use of vague items
has been achieved.
Cultural protocols emerge in the way the speaker uses vague language in
interaction and can reveal particular ways in which speakers vary their
use of vague language (Terraschke & Holmes 2007). One such mani-
festation of cultural protocols can be identified in the pattern I think
that with that functioning as a complementizer, a phrase more widely
used by the Persian speakers than the Chinese or L1s. Persian speakers
prefer not to express a proposition directly, especially when there is
uncertainty about it (Abdollahzadeh 2003). That reinforces the speaker’s
doubt, expressed by I think, or indicates that the speaker is going to
170 Communicating through Vague Language
(6.21): This is a discussion between two Persian speakers over two turns.
They are talking about the issue of spirit.
S6: Actually, huh, it is more than here we can say. You don’t, you
say that it is not but I think that it is. My question is that are all
these common with human structure? I mean structure of spirit.
Huh? … (P: 4:110)
S3: Is it good for us? (P: 4:111)
(6.22)
W1: This is a very beautiful dress. Did you sew it yourself?
W2: Yes.
W1: Well done. What an artist.
W2: It’s not as skilful as your sewing.
W1: Thanks, but it’s not true. You are a professional who has surpassed
me. I have become old.
W2: You are welcome. I take my hat off to you. (pp. 100–101)
(6.23)
W1: Your hair is very nice.
W2: It is by chance, this time it became like this.
W1: No, you’re beautiful so anything suits you.
W2: Beauty comes from your eyes. (pp. 102–103)
172 Communicating through Vague Language
S1: She doesn’t know anything how to turn, how to switch the com-
puter on. (P: 7:213)
S2: I am worse than your mom, dear. Because before I bought a com-
puter for my children. First of all, I went to computer classes.
You’ve got that. (P: 7:214)
S1: You know that, you know what you should do but what can my
mother do about it. You went to university, you know that, but my
mother is, has left school. My mother had a child when she was
17. (P: 6:215)
S2: And I had it, actually, when I was 18. (P: 7:216)
Their claim is supported by Chan (2013), who argues that Chinese tend
to use indirect and circular styles when they interact, as these allow
them to avoid direct confrontation. They use various strategies such
as contrary-to-face-value (CTFV), or use vague expressions. Ma (1996,
p. 258) defines CTFV as ‘any communication in which what is said is
the opposite of, or different from, what the speaker believes to be true
or what he or she is “logically” expected to say’. This phenomenon can
be realized in ‘yes’ for ‘no’, or vice versa. Another strategy to achieve the
same goal is to remain vague and avoid a direct statement. This norm
in Chinese politeness practice seems to encourage vague language use
in their second-language patterns.
In Excerpt (6.25), the Chinese speaker uses three very in a row to indi-
cate extreme heat; it is possible that the speaker is not familiar with the
alternatives too or really. This is observed of the Persian speakers as well.
Excerpt (6.26) shows that in a context similar to Excerpt (6.25), the L1s
uses too instead.
The L2 speakers in this study made a wide use of different vague
words, but sometimes failed to make appropriate use of vague lan-
guage because of their lack of expertise in English. For example, both
the Persian and Chinese speakers showed that they had not yet acquired
all the applications of the placeholder anything in English, mainly limit-
ing its use to negative sentences. The L1s used it more dynamically, in
negative, interrogative and affirmative phrases. Its elasticity has not yet
General Discussion 175
(6.28): This is a discussion between two Persian speakers over two turns.
They are talking about the fact that computer games can play signifi-
cant roles in children’s futures.
S1: … I believe in them, Ok? I don’t think they are. I don’t know they
should be deleted. (P: 6:516)
S5: Yes, of course not. Not deleted. (P: 6:517)
In Excerpts (6.27) and (6.28), I don’t think so and I don’t think they are
are formulaic, explicitly taught in ELT books and used as the result of
explicit instructions.
The data indicate that the Persian speakers prefer I think + negative
statement. I don’t think seems to be a more processed result of the utter-
ance and demands more cognitive load, thus the Persian speakers use
176 Communicating through Vague Language
S3: Ok. I think, huh, for developing, all the things return to the per-
sonal culture I think. You know if you want to develop, huh, all
the people one by one should start from themselves and (huh)
and to have the culture to be, for being a developed country and,
huh, find the capacity of it. You know I think there are lots of
problems in, in even this class, in this small society that we have.
For instance, he says that <LAUGH>. (P: 1: 270)
S7: Giving ideas. (P: 1:271)
S2: Let’s say, let’s say that in the process we cheat, ok? Who is cheating
again? We are cheating us, people again. (P: 1: 112)
S4: But I think that people that are cheating, they should be separated
from people that are living and are, huh. (P: 1: 113)
S2: Maybe that’s the biggest problem in Iran. Maybe that is one of the
biggest problems. That crime does pay in Iran because there is no
way of (xx) people. Ok. (P: 1: 114)
The Chinese go beyond this and use I think twice or even three times, as
can be seen in the examples below, to handle memory load.
(6.31): This is a discussion between two Chinese participants over two
turns. They are talking about what they would like to be in the future.
S6: … There exist too much difference between Chinese and, huh,
Western culture. For some more, I think, huh, I think we can, we
can, huh, we can enjoy the nature, huh, relax ourselves. And we,
as we know, tourist can earn more money and, huh, strengthen,
strengthen our body. Thank you so much. (Ch: 6: 13)
S1: But many just said that it is just too funny to be real. Yeah, I think,
I think, huh I think it is right to some degree because every career
needs much, much effort and, and, and we have limited energy as
well as limited time. So if I have to choose one as my career, I still
want to be a good learner because linguistics is all, is really, really,
very useful. (Ch: 6: 14)
times. In both cases the speakers seem to use the repetition of I think to
buy time to think about their next move; the pause marker huh buys
them even more time. There are more such repetitions of I think in the
Chinese data than in the Persian or L1.
Cognitive processing can lead to the rise of the frequency of a vague
word like maybe by the Chinese. Like I think, the Chinese use maybe as
a discourse management device to search for words.
(6.33): This is a discussion between three Chinese participants over
three turns. They are talking about job security in China.
S2: I think every person must have strong position to change them-
selves. I must try to change myself and I improve my culture.
General Discussion 179
S10: I don’t think it’s really a difference I think that, in bands tribes
and chiefdoms, they had to do that, to get people to follow ’em.
And, support ’em but if they could’ve, like just kept it all to them-
selves they would’ve. (L1: 2:40)
S1: Maybe so. I mean I’m not saying that’s not true but it was a, it’s
a standard of that society right? Yeah. (L1: 2:41)
Excerpt (6.37) shows that the Persian speakers use something you to
address the listener. The language the Persian speakers use in English
shows traces of getting the listener involved in the interaction.
A flexible middle position is taken by the Chinese, with one trend
in common with the Persian speakers and one with the L1s. The L2
groups share a preference to include the listener in what they are talk-
ing about, manifested in what occurs after I think: both the Persian and
Chinese speakers use we rather than I. This creates intimacy before any
disagreement is expressed, and so can be used to maintain face of the
other participants. The listener-oriented approach also occurs in sole-
authored papers by the Chinese authors, which use we to involve the
reader in the communication process. The equal proportions of I think
we (6 per cent) in the Persian and Chinese data, against 2 per cent by the
L1s, highlight this difference. At the same time, the Chinese also share
a pattern with the L1s, using I think I at similar proportions to the L1s
(nine and eight respectively), against 2 per cent by the Persian speakers.
I may similarly accounts for 7 per cent of the overall occurrence of
may by L1 speakers and 4 per cent by Chinese, but none by Persian
speakers, who, on the other hand, accounted for half the overall occur-
rences of you may, which involves the listener, as opposed to about one-
third of the Chinese and L1 uses.
Another pattern of the same kind is observed in the subject pronouns
occurring after actually. While the Persian speakers are inclined to use
actually we to involve the listener, the L1s show no interest in using this
cluster, and the Chinese use it only once. The presence of actually I in
interactions of the L1s and Chinese suggests that these groups prefer a
speaker-dominated approach.
them from using hedges (Mauranen 1997). This seems to be true for the
Persian group, who used the fewest possibility indicators, but not for
the Chinese group whose overall frequency of possibility indicators is
actually more than the L1s’.
Elasticity also emerges in the employment of I think in the sense that
both L2 groups use it as a strategy to take turns more often than the
L1s. While the L1s allocated the turn-taking task to this vague expres-
sion only 17 per cent, more than a quarter of the overall occurrences
of I think at the beginning of the Chinese and Persian sentences acted
as a turn-taking device. Both L2 groups used it in clause-final position
as well, whereas the L1s did not show much interest in ending a clause
with I think.
Similarly, elasticity is manifested in the use of I believe in clause-initial
and medial positions by the L2 speakers, while all tokens of I believe in
clause-initial position were in the L1 data. The L2 speakers use I believe
as a versatile tool for communication, as in medial position, it might
take the position of some discourse markers, and function as such. The
fact that the L2 speakers used vague expressions more elastically than
the L1s is because the inherent vagueness and consequent elasticity of
such words allow them to achieve the goal the speakers had set, despite
gaps in their vocabulary that might otherwise have prevented clear
communication.
Elasticity in the use of vague language can be manifested in quanti-
fiers where all instances of some of by the L1s occurred in clause-medial
position, but the L2 speakers preferred it in initial or medial position. As
with the subjectivizers discussed previously, the L2 groups gave the key
role of turn-initiator to some of: in Excerpt (6.41), for example, S2 uses
the turn-taking some of to take over from S6 in the interaction.
(6.41): This is a discussion between three Persian speakers over three
turns. They are talking about a nuclear bomb dropped on Hiroshima.
The Persian speakers stretched something more than the other groups,
using it highly frequently in clause-final position. This also happened
with their placement of conjunctions after something, which was done
almost twice as often by the Persian speakers as by the other two groups.
General Discussion 185
While the L1s did not use things in negative sentences, the L2 groups
made strategic uses of this placeholder in such contexts—and not con-
fined to negative sentences: the use of things in interrogative sentences
by the L2 groups is also proportionately more common than by the L1s.
Although the Persian and L1 speakers used thing relatively evenly, 74
and 72 times, the L2 speakers used it twice as often in clause-final posi-
tion. This occurred in the Chinese interactions as well, where clause-
final position thing turned up 30 times, while the Persian speakers used
it only 14 times and L1 speakers only seven times.
The data show that speakers, especially L2 speakers, consistently
attempt to utilize the elasticity of vague language by digging into the
most flexible areas of its use and stretching it to fit their communica-
tive needs. This can be illustrated by the following contrasts between
the overall frequency continuum of a vague language category and the
elastic continuum of a relevant category member (Figures 6.2 and 6.3).
L1s
PSLE
CSLE
The least frequent
PSLE
L1s
L1s
CSLE
CSLE
L1s
S2: But, you know this level of intellectuality if I am right, you know
here if we have the maximum, one hundred, ok? I think in Iran it
is 20. I think, I don’t know European countries maybe it is 30 but
it is much more. Ok? Maybe it is 50. I agree. Maybe it is the same
as here but there are differences. (P: 1: 502)
Questions such as how often is often or how many is many have been
the concern of the study of vague language. As an integral part of the
language, vague language in this study refers to inexact expressions
that are elastically used to contribute to effective communication. This
study, one of the few looking at vague language in terms of its elasticity,
has investigated how it can meet the communication needs of L2 learn-
ers of English, as compared with L1s, providing insights into its use in
intercultural contexts, including English language teaching.
With the fast-growing body of literature on vague language in recent
years, the inadequacy of studies of this feature of natural language in
ELT has become more conspicuous, and more research in this field
has been inspired. This study is a small step toward a more adequate
account of vague language in the context of academic settings, with
special attention to L1 versus L2 use of English.
The results show greater tendencies for vague language use by the
learner groups, and all three groups showed statistically different perfor-
mances. The Chinese were the most frequent users of vague language, but
with an uneven distribution of items in each vague category. Usability
of a vague expression in multiple positions is found to contribute to its
functionality, which results in the large frequency of vague expressions.
The most intriguing finding of this study is that the elastic feature
of vague language allows speakers to stretch it to satisfy their com-
municative needs. The most versatile vague categories (for example
subjectivizers) and items were most preferred by the L2 groups, as these
met their diverse communicative needs. Elasticity allows vague words
to stretch and provide the speaker with opportunities to make strategic
use of these expressions to enrich communication. For example, vague
language can mitigate, show politeness or solidarity, or maintain face.
189
190 Communicating through Vague Language
7.1 Conclusions
vague language. Based on the use of maybe and if it correlates with the
tentativeness, then the Chinese seem to be more tentative than the L1
and the Persian speakers, the L1 group looks less tentative than the two
L2 groups, and the Persian speakers are less tentative than the Chinese
but more tentative than the L1s.
One of the striking findings of this study is that the Persian speak-
ers adopted a listener-oriented approach, against the speaker-oriented
approach of the L1s; the Chinese took a middle position. That is,
Persian speakers were less authoritative than the L1s whose speaker-
dominated approach was evident in the more assertive language used.
The less assertive language of the Persian speakers was manifested in the
frequent application of but I think to softly express disagreement and
indicate contrast, and in the use of I think + negative clause to mitigate
negativity. This was further reinforced by the dominant use of I think
we and you know I think, used to establish intimacy and create a sense of
cooperation. Supporting evidence is the absence of I mean I think in the
Persian data. The L1 group, on the other hand, used the cluster I don’t
think + positive clause, which indicates a self-assertive attitude. The data
reveal a negative correlation: when the speaker’s view is foregrounded,
the focus on the negativity of the proposition is backgrounded; when
the negativity of the proposition is foregrounded, the speaker’s view is
backgrounded.
Elasticity refers to the usability of a category or item in multiple
positions: its diverse functionality. This feature provides speakers with
the opportunity to achieve communication needs appropriately and
effectively. With its elastic nature, vague language can be stretched and
enhance communication. The most versatile category and items were
consistently the two most popular among the L2 speakers in the class-
room interactions under study here. For example, subjectivizers, which
can occur in different positions in a sentence, are found to be the most
popular category with the Chinese and the second most popular with
the Persian speakers.
Elasticity contributes to the frequency of vague language in L2 speak-
ers’ interactions. Versatile items are much preferred by L2 speakers, and
elasticity seems to be processed by them depending on their ease of use
and the potential to meet communicative needs; if an item in one lan-
guage has a recognizably similar function in another, or if it has been
internalized sufficiently to be used effectively in L2 interactions. The
potential to meet communicative needs is judged according to how an
item can convey an intended meaning and enhance communication. It
appears that English learners need to talk about things more than L1s,
192 Communicating through Vague Language
versatile continua. None of the lexical items in this study was the most
frequently used by L1s; the L2 speakers used the most vague words.
They also showed a more concentrated use in each category of vague
language, such as I think in subjectivizers, for their versatility. This is
where the maximum potential of vague language elasticity is fulfilled:
to address the speakers’ needs and their goals in communication.
What this implies is that in addition to the overall vague language
continuum which indicates vague language elasticity, there exists an
intra-category continuum (in terms of different rankings) within each
group arranged according to the item’s versatility for that group: the
more versatile a vague word, the more frequently it occurs. The L2 data
reveal that the elasticity of vague language leads to the versatility of a
vague item, and the versatility of the item contributes to the frequency
of its use.
The elasticity of vague language emerges as a result of the uneven
distribution of vague items in each vague category that the L2 speakers
employed in communication. By contrast, the distribution by the L1s
downplayed the continua in the data. One reason why there is no or
little evidence showing L2 speakers having difficulty in communicating
might lie in the fact that they are able to resort to the elasticity of vague
language to compensate for any inadequacies; therefore, the overall
heavier use of vague language by the L2 speaker groups does not mean
they choose to be vague, but that they resort to vague language as a
versatile and reliable tool to secure enhanced communication.
This study confirms the interconnection between the linguistic reali-
zations of a vague lexical item and the particular functions it can serve.
Although there does not seem to be a correspondence between each
vague item and a function, this study reveals a continuum of particular
functions in relation to the linguistic realization of a vague item. The
function of vague language is determined by the context involved.
The findings of this research shed light on the different linguistic
behaviors of L1 and L2 groups, especially in improving the pragmatic
competence of EFL learners. They also promote the need for vague lan-
guage teaching in English. Despite the fast-growing evolution of English
as a global language whereby a uniformity of discourse verities emerges,
some discrepancies are inseparable from discourse communities. That
is, some discrepancies remain, inherited from the cultural and linguistic
norms of different communities. As a result, differences such as vague
language use need to be highlighted among discourse communities
with the aim of fostering better communication among speakers, espe-
cially between different L1 or L2 groups.
194 Communicating through Vague Language
7.2 Limitations
7.3 Implications
Given the significant differences found among the three groups (L1,
Chinese, and Persian speakers), an area for further research is an investi-
gation of vague language within each group, which will neutralize such
factors as linguistic and cultural differences and determine whether
individual psychological factors such as personality-type cause differ-
ences in the frequency of vague language categories or vague language
patterns.
Another area open for investigation is the examination of other vague
categories such as approximators and general extenders among the
same three language groups, to provide a more comprehensive view of
the conceptual dimension of vague language study across the groups.
The data for the present study comprises spoken language in class-
room interactions, but other studies might focus on written discourse
to determine if the mode of discourse is a source of differences in the
frequency and the patterns of vague language use: in other words, a
comparison of spoken and written language can provide a more detailed
Conclusions and Implications 197
account of the use of vague language. Other future research might also
investigate whether different topics can affect the frequency or elasticity
of vague language in both L1 and L2 contexts.
With the significance of vague language for communicative compe-
tence, shown in this study and acknowledged in the literature, there
is no question but that it should be included in the curriculum and
included in language pedagogy. Future research can also address a ques-
tion one step further: whether the explicit teaching of vague language
can lead to enhanced mastery of this important feature of language.
Appendix I: Consent Form for the
Director and Teachers
This consent form will be held for a period of five years.
Title: An investigation of vague language use in academic settings
• I agree to give access to the researcher for recording the classes in my language
centre.
• Students and teachers may participate in the above study if they so wish.
• I acknowledge that the nature of the study and the recording procedure has
been explained to my satisfaction by the researcher and my consent is given
voluntarily. I have had the opportunity to ask questions and have them
answered.
• I understand that the data will be stored in a secure place to safeguard
confidentiality.
• I understand that the data will be stored for any possible future research.
Signature: ---------------------------------------
Name: ------------------------------------
(Please print clearly)
Date: -------------------------------------
Contact number/E-mail: ---------------------------------------
APPROVED BY CURTIN UNIVERSITY HUMAN RESEARCH ETHICS COMMITTEE
FOR……………..YEARS ON………(DATE)……………….,
REFERENCE NUMBER………………………..
198
Appendix II: Consent Form for
Participants
This consent form will be held for a period of five years.
Title: An investigation of vague language use in academic settings
Signature: ----------------------------
Name: -------------------------------
(Please print clearly)
Date: ---------------------------------
Contact Number/E-mail: ------------------------------
APPROVED BY CURTIN UNIVERSITY HUMAN RESEARCH ETHICS COMMITTEE
FOR……………..YEARS ON……… (DATE)……………….,
REFERENCE NUMBER……………………….
199
References
200
References 201
Biber, D., & Barbieri, F. (2007). Lexical bundles in university spoken and written
registers. English for Specific Purposes, 26, 263–286.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman
grammar of spoken and written English. London: Pearson Education.
Billig, M. (1999). Conversation analysis and the claims of naivety. Discourse and
Society, 10(4), 572–576.
Blum-Kulka, S., House, J., & Kasper, G. (1989). Cross-cultural pragmatics: Requests
and apologies. Norwood: Ablex.
Bradac, J., Mulac, A., & Thompson, S. (1995). Men’s and women’s use of intensi-
fiers and hedges in problem-solving interaction: Molar and molecular analysis.
Research on Language and Social Interaction, 28(2), 93–116.
Brazil, D. (1997). The communicative role of intonation in English. Cambridge:
Cambridge University Press.
Brown, P., & Levinson, S. C. (1978). Universals in language usage: Politeness phe-
nomena. In E. N. Goody (ed.), Questions andpoliteness. Cambridge: Cambridge
University Press.
Brown, P., & Levinson, S. C. (1987). Politeness: Some universals in language usage.
Cambridge: Cambridge University Press.
Brown, P., & Levinson, S. C. (1994). Immanuel Kant among the Tenejapans:
Anthropology as empirical philosophy. Ethos, 22(1), 3–41.
Carter, R., & McCarthy, M. (2006). Cambridge grammar of English: A comprehensive
guide. Spoken and written English grammar and usage. Cambridge: Cambridge
University Press.
Chan, A. Y. (2013). Analysis of Chinese speakers’ indirect and contrary-to-
face value responses to survey interview questions. In Y. Pan & D. Z. Kàdàr
(eds.), Chinese discourse and interaction (pp. 175–198). Sheffield: Equinox
Publishing Ltd.
Channell, J. (1994). Vague language. Oxford: Oxford University Press.
Chen, W. Z., & Wu, S. X. (2002). Fanchou yu Mohu Yuyi Yanjiu (in Chinese) [A
Study on Category and Semantic Fuzziness]. Fujian (China): Fujian Renmin
Press.
Cheng, W. (2007). The use of vague language across spoken genres in an
intercultural Hong Kong Corpus. In J. Cutting (ed.), Vague language explored
(pp. 161–181). Basingstoke: Palgrave Macmillan.
Cheng, W., & Warren, M. (1999). Inexplicitness:What is it and should we be
teaching it? Applied Linguistics, 20(3), 293–315.
Cheng, W., & Warren., M. (2001). The use of vague language in intercultural
conversations in Hong Kong. English World-Wide, 22(1), 81–104.
Cheng, W., & Warren, M. (2003). Indirectness, inexplicitness and vagueness
made clearer. Pragmatics, 13(13), 381–400.
Cheshire, J. (2007). Discourse variation, grammaticalisation, and stuff like that.
Journal of Sociolinguistics, 11(2), 155–193.
Chow, K. F. (2010). A quasi-self-dual interpretation of negative raising academic area:
Formal semantics. Paper presented at the 3rd Postgraduate Research Symposium:
Language and Cultural Studies in the Pearl River Delta, Hong Kong.
Conrad, S., & Biber, D. (2000). Adverbial marking of stance in speech and writ-
ing. In S. Hunston & G. Thompson (eds.), Evaluation in text: Authorial stance
and the construction of discourse (pp. 56–73). New York: Oxford University
Press.
202 References
Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (eds.), Syntax
and Semantics (Vol. 3: Speech acts, pp. 41–58). New York: Academic Press.
Grice, H. P. (1989). Studies in the way of words. Cambridge: Harvard University
Press.
Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., &
Heyman, S. (1996). The influence of cultural individualism-collectivism, self
construals, and individual values on communication styles across cultures.
Human Communication Research, 22(4), 510–543.
Hamilton, M. A., & Minoe, P. J. (1998). A framework for understanding equivoca-
tion. Journal of Language and Social Psychology, 17, 3–35.
Hanson, W. E., Creswell, J. W., Plano Clark, V. L., Petska, K. P., & Creswell, J. D.
(2005). Mixed methods research designs in counseling psychology. Journal of
Counseling Psychology, 52(2), 224–235.
Heider, E. R. (1973). On the internal structure of perceptual and semantic catego-
ries. In T. E. Moore (ed.), Cognitive development and the acquisition of language
(pp. 111–144). New York: Academic Press.
Holmes, J. (1986). Functions of you know in women’s and men’s speech. Language
in Society, 15(1), 1–22.
Holmes, J. (1988). Doubt and uncertainty in ESL textbooks. Applied Linguistics,
9(1), 21–44.
Holmes, J. (1990). Apologies in New Zealand English. Language in Society, 19,
155–199.
Holmes, J., & Brown, D. (1987). Teachers and students learning about compli-
ments. TESOL Quarterly, 21(3), 523–546.
Hu, Z., Brown, D., & Brown, L. (1982). Some linguistic differences in the
written English of Chinese and Australian students. Language Learning and
Communication, 1, 39–49.
Huang, Y. (2007). Pragmatics. Oxford: Oxford University Press.
Hübler, A. (1983). Understatements and hedges in English. Amsterdam: John
Benjamins.
Hutchby, I., & Wooffitt, R. (1998). Conversation analysis: Principles, practices and
applications. Oxford: Polity Press.
Hutchby, I., & Wooffitt, R. (2008). Conversation analysis (2nd ed.). Cambridge:
Polity Press.
Hyland, K., & Milton, J. (1997). Qualification and certainty in L1 and L2
students’ writing. Journal of Second Language Writing, 6(2), 183–205.
Hyland, K. (1994). Hedging in academic writing and EAP textbooks. English for
Specific Purposes, 13(3), 239–256.
Hyland, K. (1996). Nurturing hedges in the ESP curriculum. System, 24(4),
477–490.
Hyland, K. (1998). Hedging in scientific research articles. Amsterdam: John
Benjamins.
Hyland, K. (2000). Hedges, boosters and lexical invisibility: Noticing modifiers in
academic texts. Language Awareness, 9(4), 179–197.
James, A. R. (1983). Compromisers in English: A cross-disciplinary approach to
their interpersonal significance. Journal of Pragmatics, 7(2), 191–206.
Janicki, K. (2002). A hindrance to communication: The use of difficult and
incomprehensible language. International Journal of Applied Linguistics, 12(2),
194–217.
204 References
McMillan, J. R., Clifton, A. K., McGrath, D., & Gale, W. S. (1977). Women’s lan-
guage: Uncertainty or interpersonal sensitivity and emotionality. Sex Roles, 3,
545–559.
Metsä-Ketelä, M. (2006). Words are more or less superfluous: The case of more
or less in academic Lingua Franca English. Nordic Journal of English Studies,
5(2), 117–143. Retrieved from http://ojs.ub.gu.se/ojs/index.php/njes/article/
view/72/76
Metsä-Ketelä, M. (2012). Frequencies of vague expressions in English as an aca-
demic lingua franca. Journal of English as a Lingua Franca, 1(2), 263–285.
Morse, J. M. (1991). Approaches to qualitative-quantitative methodological trian-
gulation. Nursing Research, 40, 120–123.
Moxey, L. M., & Sanford, A. J. (1993). Communicating quantities: A psychological
perspective. Hove: Lawrence Erlbaum.
Mulac, A., & Lundell, T. L. (1986). Linguistic contributors to the gender-linked
language effect. Journal of Language and Social Psychology, 5, 81–101.
Mulac, A., Lundell, T. L., & Bradac, J. (1986). Male/female language differences
and attributional consequences in a public speaking situation: Toward an
explanation of the gender-linked language effect. Communication Monographs,
53, 115–129.
Mulac, A., Wiemann, J. M., Widenmann, S. J., & Gibson, T. W. (1988). Male/
female language differences and effects in same-sex and mixed-sex dyads: The
gender-linked language effect. Communication Monographs, 55, 315–335.
Myers, G. (1989). The pragmatics of politeness in scientific articles. Applied
Linguistics, 10, 1–35.
Myers, G. (1996). Strategic vagueness in academic writing. In E. Ventola &
A. Mauranen (eds.), Academic writing:Intercultural and textual issues (pp. 1–18).
Amsterdam: John Benjamins.
Neuman, W. L. (1997). Social research methods: Qualitative and quantitative
approaches. Boston: Allyn & Bacon.
Niezgoda, K., & Rover, C. (2001). Pragmatic and grammatical awareness: A func-
tion of the learning environment. In K. R. Rose & G. Kasper (eds.), Pragmatics
in language teaching (pp. 63–79). Cambridge: Cambridge University Press.
Nikula, T. (1996). Pragmatic force modifiers: A study in interlanguage pragmatics.
Jyvaskyla: University of Jyvaskyla.
O’Shea, M. (2000). Cultural shock: Iran. Portland: Graphic Arts Publishing Company.
Ohta, A. S. (2001). Second language acquisition processes in the classroom: Learning
Japanese. Mahwah: Lawrence Erlbaum Associates.
Overstreet, M. (1995). The form and function of general extenders in English interna-
tional discourse. (Unpublished PhD thesis). The University of Hawaii, The USA.
Overstreet, M. (2005). And stuff und so: Investigating pragmatic expressions in
English and German. Journal of Pragmatics, 37, 1845–1864.
Parvaresh, V., Tavangar, M., & Eslami Rasekh, A. (2010). General extenders in
Persian discourse: Frequency and grammatical distribution. Cross-Cultural
Communication, 6(3), 18.
Parvaresh, V., Tavangar, M., Eslami Rasekh, A., & Izadi, D. (2012). About his
friend, how good she is, and this and that: General extenders in native Persian
and non-native English discourse. Journal of Pragmatics, 44, 261–279.
Parvaresh, V., & Tayebi, T. (2014). Vaguely speaking in Persian. Discourse
Processes, 51(7), 565–600.
206 References
A go hypothetical 31
aberou 171 go just-right 30
accommodation strategy 12 go subjective 31
adaptors 123
approximation 45 H
attribution shield 115 hesitation markers 143
B I
boosters 129 imprecision 4
incomprehensible language 7
C indirect language 6
classical set theory 5 indirectness 7
cluster 47 inference 27
cognitive effort 27
cognitive load 175 L
common categories 9 Lexical categories 9
contrary-to-face-value (CTFV) 173 lexical density 17
conversation analysis 35, 36 listener-oriented approach 158,
conversational implicature 25 180, 181
cooperative principle 23, 24 local vagueness 155, 156
cultural protocols 169 loose talk 28, 29
D
M
detensifiers 123
managerial mode 15
discourse management 55, 153
material mode 15
discourse markers 143
mitigating 114
downtoning 45
mitigation 45, 114
downtowners 123
mixed methods 38, 39
mode 15
E
modifying devices 18
elasticity 5, 29, 30, 181, 182, 186, 191
multal quantifiers 130
embedded design 39
emphasising 114
N
empirical science 34
negative face 118
epistemic phrases 47
Non-lexical categories 9
equivalent status design 39
F O
face 118 Ostension 27
fuzzy set theory 5
P
G Parallel/simultaneous studies 39
global vagueness 155 placeholder 44, 99, 158
go general 30 plausibility shield 8, 115
209
210 Subject Index
A Crozet, C. 2
Abdollahzadeh, E. 14, 169 Crystal, D. 8, 9
Adolphs, S. 2, 11 Cummings, M. C. 41
Aijmer, K. 49, 119 Cutting, J. 1, 9, 19, 20, 26, 28,
Alcon, E. 18, 19 69, 113
Allison, D 22,
Alwood, J 146 D
Aristotle 1 Davies, B. L. 25
Atai, M. R. 163 Davy, D. 8, 9
Atkins, S. 2 De Cock, S. 19
Austin, J. L 6 Deese, J. 9
Drave, N. 20, 77, 81, 113, 152
B Dudley-Evans, T. 22
Bachman, L. F. 18
Badovi-Harlig, K. 18 E
Behnam, B. 160 Erev, I. 8
Beighmohammadi, A. 14 Eslami Rasekh, A. 13, 171
Biber, D. 4, 23, 49
Billig, M. 37 F
Blum-Kulka 44 Finegan, E. 5
Bosk, C. 8 Firestone, W. 33
Bradac, J. 129 Flick, U. 34
Brazil, D. 5, 23 Frader, J. 8
Brown, P. 8, 18, 22, 118, 119 Franken, N. 29
C G
Cai, Z. 22 Grice, H. P. 6, 23, 24, 25, 26, 31,
Carter, R. 6 26, 136
Chan, A. Y. 173 Griffin, R. 18
Channell, J. 1, 4, 7, 8, 9, 11, 20, 25, Gudykunst, W. B. 157
26, 44, 45, 69, 99, 113, 119, 126,
136, 152 H
Chen, W. Z. 13 Hanson, H. F. 40
Cheng, W. 3, 4, 6, 7, 11, 12, 17, 18, Harvey, K. 2
19, 20, 27, 45, 152 Heider, E. R. 6
Cheshire, J. 13 Holmes, J. 18, 22, 129,
Chow, K. F. 176 143, 169
Connor, U. 22 Hosman, L. A. 129
Conrad, S. 5, 49 Huang, Y. 24, 25, 26
Cook, G. 5, 23, 33 Hübler, A. 123
Cotterill, J. 2, 11 Hutchby, I. 35, 36, 37
Creswell, J. W. 33, 34, 35, 38, 39, Hyland, K. 5, 16, 17, 21, 22, 23,
40, 41 68, 129
211
212 Author Index
I Niezgoda, K. 2
Izadi, D. 13 Nikula, T. 19, 21, 118, 119, 190
Niroomand, M. 160
J
James, A. R. 18 O
Janicki, K. 7 O’Keeffe, A. 9
Jefferson, G. 36, 139 O’Shea, M. 171
Jick, T. D. 40 Ohta, A. S. 2
Johansson, S. 4 Overstreet, M. 8, 13
Johnson, M. 1
Jucker, A. H. 9, 10, 23, 28, 29, 113, P
114, 123, 138 Parvaresh, V. 13, 14
Patton, M. Q. 38
K Pearson, J. 123
Kaltenböck, G. 49 Plano Clark, V. L. 40, 41
Kärkkäinen, E. 47, 48 Plato 1
Khurshudyan, V . 45, 143 Pomerantz, A. 18
Koester, A. 2 Prince, E. F. 8, 11, 44, 45, 114, 115,
Koutlaki, S. A. 171 123, 132, 139
L R
Lado, R. 162 Reichardt, C. S. 33
Lakoff, G. 1, 5, 11 Remero Trillo, J. 180
Leech, G. 4 Ringbom, H. 20, 189
Levinson, S. C. 6, 8, 18, 26 Roomer, U. 164
Liddicoat, A. J. 2, 36, 37 Rosch, E. H. 6
Lotfi, A. R. 166, 170 Rossman, G. B. 38
Lüdge, T. 9 Rowland, T. 15
Lundell, T. L. 129 Russell, B. 5
Rutter, B. 139
M Ruzaitė, J. 2, 4, 7, 8, 11, 12, 14, 44,
Ma, R. 181 45, 47, 67, 69, 71, 83, 113, 114,
Mansoori, S. 157, 172 119, 123, 130, 138, 139
Martinez-Flor, A. 19 Röver, C. 2
Martinovski, B. 114
Mauranen, A. 184 S
McCarthy, M. 6, 9, 23 Sabet, P. 47, 49, 51, 53, 175,
McMillan, J. R. 129 178, 179
Metsä-Ketelä, M. 8, 10, 21, 152 Sacks, H. 36, 139
Milton, J. 21, 22, 23, 68 Sadr, L. 163
Minoe, P. J. 8 Sanford, A. 10
Morse, J. M. 39 Schegloff, E. A. 36, 139
Moxey, L. M. 10 Searle, J. R. 6
Mulac, A. 129 Sharifian, F. 166, 170, 171
Mulan, A. 129 Simon-Vandenbergen, A. M. 49
Myers, G. 5, 8, 17 Smith, A. 9
Soter, A. 22
N Sperber, D. 23, 25, 26, 27, 28, 31
Neal, m. M. 8 Stenström, A. B. 49
Neuman, W. L. 35 Sweller, J. 175
Author Index 213
T Wang, J. 21
Tannen, D. 10 Warren, M. 3, 4, 6, 7, 12, 17,
Tarnyikova, J. 4 18, 19, 27, 152
Tavangar, M. 13 Wiese, R. 143
Tayebi, T. 14 Williamson, T. 6
Teigen, K. H. 10 Wilson, D. 23, 25, 26, 27, 28,
Terraschke, A. 42, 169 29, 31
Thomas, J. 18 Wolfson, N. 18
Thompson, S. 129 Wooffitt, R. 36, 37
Tracy, K. 118 Wright, H. J. 129
Tran, G. Q. 41 Wu, Y. 13, 22, 123, 142,
Trappes-Lomax, H. 115 143, 161
Travers, R. M. W. 34
Z
U Zadeh, L. A. 5, 6, 12, 14, 28
Ullman, S. 8 Zarei, G. R. 157, 172
Zhang, G. 1, 7, 13, 23, 25,
W 28, 31, 44, 45, 49, 51, 113,
Wallston, T. S. 8 152, 155, 161, 175, 178,
Walsh, S. 9, 15, 16 181, 188