Introduction
In recent years we have seen a fundamental
shift in interest among both language
scholars and educators away from the purely
formal study of linguistic structure, typically
employing constructed sentences and/or
written language models, with more
attention now being devoted to the analysis
of everyday language use in natural settings.
Much of the new development in language
study has been in the direction of discourse
analysis, with an explosion of work in
corpus linguistics, allowing for the analysis
of vast amounts of naturally occurring
spoken data (as opposed to constructed
sentences). In language education, there has
been a large-scale transition into a variety of
communicative approaches, with a much
stronger emphasis on oral language skills,
especially in spoken interaction. This shift
has focused a lot more attention on features
of the spoken language that were previously
ignored in both linguistic analysis and
foreign language teaching materials. Among
these features are some linguistic
expressions that are primarily found in face-to-face conversational interaction. Because
these expressions have no discernible
semantic content and are usually optional
elements in syntactic structure, they are
associated, not with independent linguistic
meaning, but with meaning in context and so
they are generally described as pragmatic
expressions.
In this paper, I will briefly describe the
range of types of pragmatic expressions,
including discourse markers (Well) and
pragmatic markers (you know), then focus
on one group of expressions called general
extenders. After looking at some English
examples, such as and stuff (like that), or
something (like that), and how they are used,
I will review some studies of comparable
forms in other languages and observations
by other scholars on the subtle differences in
their uses and the types of difficulties
associated with comparing any pragmatic
expressions cross-linguistically, or what we
might characterize as the problem of
determining pragmatic equivalence.
Following that, I will review some studies
from language learning, mainly involving
English as a foreign language (EFL), and
investigate ways in which we might be able
to foster pragmatic awareness of both the
first language (L1) and the second language
(L2) as a prerequisite for developing
pragmatic competence in that L2.
Pragmatic expressions
The study of pragmatic expressions owes its
development directly to the availability of
recording devices that allowed researchers
not only to capture everyday spoken
interaction, but also to transcribe it and
investigate it “on the page/screen” in ways
that were almost impossible while the data
was whizzing by “in the air”. From early
studies of the underlying structure of
conversational interaction (Sacks, Schegloff
& Jefferson, 1974), through investigations
of social meaning and how it is signaled
(Erman, 1987; Schiffrin, 1985; Schourup,
1985; Tannen, 1984), to the micro-analysis
of forms of language previously unexplored
(Channell, 1994; Overstreet, 2011), there
has been a steady stream of new findings
about the complex nature of spoken
interaction and the linguistic expressions
being used to hold it all together. Among
those linguistic expressions are a number of
forms that seemed to have no meaningful
role in linguistic communication and were
often viewed as “purely performance fillers”
(Channell, 1994, p. 120), but which, on
closer inspection, have been identified as
integral elements in how spoken discourse is
structured and made meaningful.
Some of the expressions that were treated as
simply meaningless interjections are actually
structuring devices within spoken
interaction. Forms such as Oh, Right, Now,
So, and Well, used at the beginning of a
speaker’s turn, are now recognized as
discourse markers (Schiffrin, 1987,
Schourup, 1999), each one signaling subtle
aspects of how the speaker is marking the
sequential connection between a previous
turn and what is about to be said. Other
pragmatic expressions have been identified
that can function in initial, medial or final
position, with different influences on the
interpretation of speaker’s meaning. These
are more generally known as pragmatic
markers and include you know, you see, I
mean and I think, which developed from
subject plus lexical verb combinations into
parenthetical adjuncts and finally into fixed
phrases that are used idiomatically (Aijmer
& Simon-Vandenbergen, 2011; Brinton,
1996). The range of different functions that
these pragmatic markers can fulfill in the
contemporary spoken language has only
recently been revealed. As a brief
illustration, consider these three different
functions of you know, as detailed in Erman
(2001). It can function as a text-organizing
device, used when introducing a supporting
example, as in (1), an interactive device,
when making a comprehension check, as in
(2), and a hedging device on the assumed
accuracy of reported information, as in (3).
(1) they did it in a completely, slapstick
farce way, you know, the the men who were
dressed up supposed to be women
(2) A: you’ve got to use one of them cap
things, not a swimming cap
B: Steam cap?
A: Yeah, well, you know, them white ones,
have you seen the plastic ones yeah?
(3) She said you’re, you’re nice, you’re
pretty, you know, whatever
The historical development of these
pragmatic expressions has revealed a regular
pathway of change through a number of
processes similar to grammaticalization
whereby phrases containing lexical items
such as the main verbs know and think lose
their propositional content and become
indicators of how speakers are presenting
themselves and their attitude to the message
and/or the addressee within face-to-face
interaction. As many of these newly
revealed functions are tied to issues of
politeness, cooperation, social solidarity,
attitudes, and evaluations, rather than
marking grammatical functions, this process
is now also discussed in terms of
pragmaticalization (Overstreet & Yule,
forthcoming). In addition to these types of
markers, which are syntactically
disconnected from the utterances in which
they appear, there is another group of
pragmatic expressions that are typically
attached to the end of phrases, clauses and
utterances. Among these are general
extenders such as and stuff (like that) and or
something (like that), which will be the main
focus of the rest of this study.
General extenders
General extenders have been the subject of a
fairly large number of studies in different
varieties of English, most of which are
reviewed in Pallacios Martínez (2011)
They can be divided into two types:
adjunctive forms, beginning with and, as in
(4), and disjunctive forms, beginning with
or, as in (5), both examples from the
Canadian English data of Tagliamonte and
Denis (2010, p. 337).
(4) So it was- it was pretty general, you
know, nice and quiet, never a lot of noise,
and stuff like that
(5) it was in- when- oh I think it was like,
grade seven or something
The disjunctive form or something is the
most frequently used version across
different varieties of English. It often
functions as a hedge on the accuracy of what
is being said, as illustrated in (5), and is
quickly learned by EFL students, allowing
them to mark some part of what they are
saying as “possibly not exactly correct” in
the same way as native speakers (NS) do it.
In contrast, their use of adjunctive general
extenders to express “there is more, but I
don’t need to say it” can vary much more,
and often in ways that don’t match typical
NS usage. In the following discussion, I will
focus more on the use of adjunctive general
extenders and offer, in Table 1, a list of the
four most frequent forms found in detailed
studies of Canadian English (Tagliamonte &
Denis, 2010), British English (Pichler &
Levey, 2011) and American English
(Overstreet & Yule, 1997).
In Table 1, the proliferation of the American
English form and stuff (like that) is clear,
supplanting and things (like that) in
Toronto, according to Tagliamonte (2011, p.
258), and gradually becoming more frequent
among younger middle class speakers of
British English, according to Cheshire
(2007, p. 187), who also notes that the
British form and that is associated more
with working class speech, especially among
male speakers. The expression and
everything (without the modifier like that)
has increased in frequency in all varieties,
possibly due to a new use as an intensifier
(Overstreet & Yule, 2002). The form and
blah, blah, blah is used in American English
to indicate that more could be said, but it has
a downgrading function, implying that the
“more” is of little value (Overstreet, 1999, p.
146). This form is not included in the
lengthy lists of forms in the reports
describing Canadian and British English.
Because general extenders are placed after
the items they modify, they are often found
at the end of clauses and hence of utterances
in English. This tendency is sometimes
overstated, as in definitions that describe
them as “sentence-final” (Tagliamonte,
2011, p. 258). While general extenders are
frequently attached to objects, that does not
necessarily place them at the end of
utterances, or even clauses, as shown in the
use of and everything in (6), nor does it rule
out the possibility of attaching to the subject,
as or something does in the same example,
from Aijmer (2002, p. 245).
(6) I got my coat and everything caught
under me and a young postman or
something got up and I thought ooh this is
grand
When we look at other languages, we see
general extenders in quite a wide range of
clause-internal positions. As described in
Overstreet (2005, p. 1849), there are several
structures in German that require a main
verb or the past participle of a verb to be at
the end of the clause, hence regularly
positioning the general extender inside the
clause, as in (7).
(7) Ich hab’ nun jetzt erstmal meine ganzen
Pflanzen da in die Erde gebracht und – und
– sehr viel Tulpen und Krokusse und so was
gesteckt
[I’ve just got all my plants there in
the ground and – and – lots of tulips and
crocuses and so on put in]
In Persian, with its basic SOV structure,
general extenders can readily attach to a
subject (8) or an object (9) in clause-internal
positions, as in these examples from
Parvaresh, Tavangar, Eslami Rasekh and
Izadi (2012, pp. 270-276).
(8) gæzâ væ inâ âli bud!
(the food and stuff was great!)
(9) je mântoji, jâ čizi mixâm bærdâræm
(I want to an overcoat or something buy)
In connection with the syntactic position of
general extenders, Parvaresh et al. (2012)
point out that none of their advanced EFL
students tried to use the Persian structure in
(9) in their English. Though they did transfer
some general extender types, to be discussed
later, they didn’t transfer Persian syntax.
This would seem to support the observation
(cf. Bouton, 1994) that structures involving
pragmatic expressions may be harder to
acquire than grammatical structures in an
L2.
Pragmatic expressions in EFL studies
When we employ a cross-linguistic
perspective to investigate pragmatic
expressions in the use of a foreign language,
we find a number of different explanations
offered for the patterns perceived. The most
general finding from this area of research is
that non-native speakers (NNS), even those
at an advanced level, typically use a more
limited number, as well as a more limited
range of pragmatic expressions than native
speakers (NS). While I will focus mainly on
studies in EFL situations in this discussion,
similar findings have been reported from
investigations where other languages are the
target. For example, in Dippold’s (2008)
study of the use of hedges in argumentative
discourse by advanced level British students
speaking German, the NNS used relatively
few hedges in comparison with the frequent
use by a comparable NS group. If NNS
groups are not using the types of pragmatic
expressions normally found in NS
performance, is there a simple way to
demonstrate their uses, and might there be a
way to help build awareness of these forms
and their functions within EFL studies?
The first and most obvious explanation for
the absence of L2 pragmatic expressions is
that the NNS don’t need them in most
situations where they use English. Often this
follows from the nature of the discourse
and/or the participants and may be indicative
of the way in which new varieties of English
evolve, as when it is used as a lingua franca
in interactions between NNS. When
speakers of two different languages use
English as their medium of communication,
there seems to be a very general absence of
the types of pragmatic expressions typically
found when native speakers interact. As
Murray (2012) has noted with regard to
spoken exchanges involving English as a
lingua franca, “discourse markers and
particles appear to be relatively scarce”
(2012, p. 321). It may be that such
encounters are treated as more
“transactional” by the participants, that is,
more concerned with communicating
referential meaning, “factual or
propositional information” (Brown & Yule,
1983, p. 2), and that expressions more
associated with social meaning in
“interactive” encounters are not included
when personal relationships between
speakers are not a primary issue. This type
of situation may have more in common with
written English discourse, which brings us
to another explanation of the patterns
observed in NNS use of pragmatic
expressions.
In a study comparing the use of English
general extenders in the speech of two
groups of university students, one consisting
of French L1, advanced level NNS and the
other a NS group at a British university, De
Cock (2004) found a highly systematic and
quite revealing pattern of usage. As shown
in Table 2, adapted from De Cock (2004, p.
237), there is a divergence in preferred
forms of adjunctive general extenders, with
the NNS group mostly relying on the
expressions and so on, etcetera, whereas the
forms and things (like that), and everything,
and stuff (like that) were favored by the NS
group. This split exactly parallels the
difference in distribution Overstreet and
Yule (1997) discovered between formal and
informal spoken language use. Formal
expressions such as and so on are more
common in academic English, both spoken
and written (Biber, Johansson, Leech,
Conrad & Finegan, 1999; Simpson, 2004),
which may have been the primary input
source for these NNS students. One might
speculate that it is the inclusion of such
formal expressions in their interactive
spoken language that accounts for the
impression that some advanced EFL
speakers “may sound rather bookish and
pedantic” (Channell, 1994, p. 21).
Reviewing the results in Table 2, we might
suspect that these NNS university students
have learned some English expressions to
serve the general extender function that will
inevitably make their speech sound more
formal than NS usage and contribute to “the
impression of detachment and formality they
may well give in informal situations” (De
Cock, 2004, p. 236). We should note that
French has a wide range of general
extenders, described in some detail by
Dubois (1992) as “extension particles”, with
different forms available for different
functions, both formal and informal, but the
students in De Cock’s (2004) study had
obviously not carried their L1 pragmatic
knowledge of general extenders over into
their understanding of how English
expressions are used with comparable
functions. As a result, in De Cock’s report,
“the findings suggest that the learners are
lacking in routinized ways of interacting and
building rapport with their interlocutors”
(2004, p. 243).
A similar conclusion was reached by Otu
and Zeyreck (2008) in their study of Turkish
learners of English when they investigated
how these NNS performed requests in
English. They found that the students
themselves had a sense of their unfamiliarity
with NS norms for performing the speech
acts appropriately. The researchers noted
that “most of these students refer to this
situation as knowing textbook English only,
and being totally blind of the rest of the
picture” (2008, p. 265).
Yet another explanation, and hardly a
surprising one, is emerging from other
recent studies that find pragmatic
expressions in NNS English that seem to be
derived from the L1. Eslami Rasekh and
Noora (2008) noted that “even highly
proficient learners often rely on their L1
strategies or conventions of form” (2008, p.
321) when they investigated request
strategies by Persian learners of English. In
another study of Persian learners and their
use of general extenders, Parvaresh et al.
(2012) came to the conclusion that “first
language norms influence the use of general
extenders by non-native speakers” (2012, p.
261). To take a specific case, both German
(und, und, und) and Persian (væ, væ, væ)
make use of a structure that is a possible
combination in English (and, and, and), but
one that is not typically found in everyday
uses of English. Despite the fact that the
expression is unlikely to be part of any NS
input, it is found in the English speech of
Persian NNS, as in (10), from Parvaresh et
al. (2012, p. 266).
(10) I have to study, I mean, memorize
things and and and
Other structures, such as and this and that,
which are very occasionally recorded in
English NS data (see Tagliamonte & Denis,
2010, p. 363), may be used more frequently
by a NNS group when a parallel structure
exists in the L1. Example (11) is from
Persian NNS data and is described as an
example of transfer from Persian (væ in, væun).
(11) A: No! I really love to be there
E: I love to be there and this and that
According to Parvaresh et al. (2012, p. 275),
this particular form is not a signal that
communicates the basic adjunctive general
extender meaning of “there is more”, but has
a particular interpersonal meaning and is
used by speakers in response to a comment
by another speaker. The comment is usually
repeated before the general extender, which
signals that the comment is being treated as
“offensive” in some way. In this case, it is
important to recognize that, although
structurally identical forms may exist in two
languages, they cannot be treated as
translation equivalents because their
functions are so different. The closest form
in American English with a comparable
function, though not an obvious lexical
equivalent, might be (or) whatever. In
Kleiner’s (1998) analysis, the general
extender (or) whatever can be used to mark
preceding material as “other-authored” and
to express “the speaker’s disaffiliation with
or opposition to that material” (1998, p.
602). Although this usage may have a
function similar to that conveyed by the
Persian expression in particular contexts,
there will almost certainly be socio-cultural
implications tied to negative commentary
and its effect on the participation framework
that are likely to differ cross-culturally.
Realizing this, we should always be careful
about assuming pragmatic equivalence
cross-linguistically, even when we think we
can identify a form (lexically or structurally
similar or not) that seems to have a parallel
function. As Koutlaki (2002) has pointed
out, in a comparison of the pragmatics of
some English and Persian speech acts, what
seems face-threatening in the act of making
an offer in one culture may actually be
considered face-enhancing in another.
However, if we can find forms with parallel
functions, which are used with comparable
frequency in similar situations, then we may
be able to advise students about what are,
and are not, good translation equivalents
(though not necessarily perfect pragmatic
equivalents).
There is a subtle trap waiting for EFL
learners because of the existence of what
appear to be cognate expressions, such as
German oder so and English or so. They
look like they would be direct translation
equivalents. However, the German
expression (oder so) was, by a wide margin,
the most common disjunctive general
extender in Terraschke’s (2007a) German
NS data, while the English expression (or
so) was used only once in her New Zealand
English NS data. Given this substantial
difference, we might suspect that the
similarity in form disguises a difference in
function and, indeed, we find that the
English form is highly restricted in its
collocations, accompanying only numbers
and time expressions. The German
expression is not subject to such narrow
restrictions and its wide range of functions
seems to be readily transferred, as in NNS
English examples such as (12), from
Terraschke (2007a, p. 94), where the speaker
is comparing two towns and uses the
English expression in a way not found
among English speakers.
(12) But well I’m, yeah, I believe that
there’s more to do or so
It is worth noting that, in this and many
similar cases, there is no indication that any
form of miscommunication took place and
hence no feedback is provided to the student
that an inappropriate pragmatic expression is
being used.
Fostering pragmatic awareness
A further explanation, and one that may
provide the best reason from a processing
point of view, is that the NNS have no idea
that there are pragmatic expressions in
language use. This doesn’t mean, of course,
that they don’t use pragmatic expressions
such as general extenders in their L1, but
that they are completely unaware that they
do so. As Overstreet (2000) reported from
her research in the 1990s, NS of American
English, including English language
teachers and professors of linguistics, not
only appeared to be unaware of the existence
of general extenders in their L1, but even
after being made aware of them, claimed
that they personally didn’t use such forms in
their speech (despite empirical evidence to
the contrary). One possible reason for this,
as Dines (1980) noted among speakers of
Australian English, is that some general
extenders are viewed as “stigmatized” in
some way and hence likely to be considered
inappropriate in the speech of educated
individuals. While a negative stylistic
perspective may indeed exist, it is perhaps
comparable to opinions on split infinitives
and ending sentences with prepositions in
English (cf. Yule, 2010, p. 85), forms that
may be condemned by prescriptivist
commentators, but are in widespread use
among all segments of the population.
General extenders are similarly used in
spoken interaction by virtually everyone
and, while some individuals may be more
frequent users of particular forms than
others, familiarity with the forms and
functions of general extenders is part of
adult NS pragmatic competence.
We cannot assume that L2 pragmatic
competence will develop by itself since it is
socio-culturally acquired and unlikely to be
part of any innate language acquisition
device. It is only through studies at the
metapragmatic level that we have become
aware of the phenomenon (cf. Overstreet,
2010). Consequently, given the socio-cultural limitations inherent in many EFL
learning contexts, there may be a need for a
more proactive approach to developing L2
pragmatic awareness. Schmidt (1993) has
argued that “noticing” has to take place in
order for pragmatic information to be
processed, at least in short term memory.
Indeed, as a number of studies have shown,
we can increase pragmatic awareness so that
L2 learners have an opportunity to develop
their own competence in the use of
pragmatic expressions. Kasper and Schmidt
(1996) and House (1996) provide examples
and reviews of studies where pragmatic
awareness was developed in different L2
contexts. In a similar vein, LoCastro (1997)
described how L2 pragmatic fluency was
improved in spoken English and Wishnoff
(2000) presented strong evidence that
raising L2 students’ awareness of hedges in
English academic writing resulted in
substantial improvement in their ability to
use those hedges appropriately in their own
L2 writing.
At a more fundamental level, especially in
EFL contexts, as Eslami-Rasekh (2005) has
argued, it may be more effective to begin by
raising students’ awareness of pragmatic
expressions in their L1 and encouraging a
comparison between L1 and L2 forms to
develop familiarity with similarities and
differences. The data provided by Parvaresh
et al. (2012) offers an opportunity to see
how one type of comparison might be
presented. In Table 3, the most frequently
used Persian L1 adjunctive general
extenders are listed alongside the most
frequent English L2 forms produced by
Persian EFL students.
The English L2 forms listed in Table 3 are
clearly not the same as the English L1 forms
listed earlier in Table 1. These students seem
to be using and everything with a similar
frequency to the NS in Table 1, but they are
using the more formal expression and so on
with greater frequency, more like the other
NNS in Table 2. The expression and other
things is not an impossible form in English,
but is extremely rare, so that it will
inevitably sound like an interlanguage form,
typical of neither the L1 nor the L2. These
students could be advised to omit the word
other in this expression to give it a more
target-like form.
The most intriguing L2 form and the most
frequent is and blah, blah, blah. In other
recent studies of the use of general extenders
by NNS (e.g. Fernandez & Yuldashev,
2011), this expression is not recorded at all.
There are no examples reported in recent
British English studies (e.g. Levey, 2012,
Pichler & Levey, 2011). As noted earlier,
this form is certainly used by NS of
American English, but in a quite restricted
way, with almost a pejorative meaning. It is
not clear if any of the Persian L1 forms in
Table 3 have similar negative implications,
but two of the most frequent items have
forms referring to “talks”. They are not just
signaling “there is more”, as with most
forms in Table 1. It is possible that the focus
of high frequency Persian general extenders
on “there is more talk (about something)”
has an influence on which English general
extender these students have chosen as the
best pragmatic equivalent. To help students
understand the effect of this choice, we
might present the examples from Table 1
alongside the forms in Table 3 as a way of
letting the students see for themselves that
their solution to the pragmatic equivalence
problem may be infelicitous on some
occasions and has the potential for
miscommunication if, when using and blah,
blah, blah, the speaker doesn’t actually want
to act as if all further information is being
downgraded.
Conclusion
I have suggested that it is possible to
increase students’ pragmatic awareness by
drawing attention to how pragmatic
expressions such as general extenders are
used in both the L1 and L2. This approach
would seem to be justified because of
reports that learners often adopt
inappropriate, or pragmatically non-equivalent forms, either because of
misperception of the typical functions of L2
forms or because of influence from L1
forms.
In order for this approach to work, however,
we need to pay more attention to the ways in
which pragmatic expressions are used in
both the L1 and L2 of particular groups of
learners, so that we have reliable
information on which to base our materials.
There are many signs that this goal can be
accomplished, as increasing numbers of
studies, particularly corpus-based
investigations, reveal patterns of language
use in spoken interaction that were
previously unnoticed. I have proposed that
general extenders represent a distinct and
easily identifiable group of pragmatic
expressions that lend themselves to cross-linguistic comparison and potentially allow
us to tease apart the subtle differences that
make cross-linguistic and cross-cultural
studies not only challenging, but
intellectually rewarding, and ultimately
beneficial for the development of better
understanding and greater tolerance among
people.
Notes
1 The label “general extender”, from
Overstreet and Yule (1997), has become the
most widely used technical term for this
range of forms. Among other labels that may
be encountered are “set-marking tags”
(Dines, 1980), “utterance-final tags”
(Aijmer, 1985), “list completers” (Jefferson,
1990), “vague category identifiers”
(Channel, 1994) and “coordination tags“
(Biber et al., 1999).
2 There have also been studies of general
extenders in languages other than English,
such as Brazilian Portuguese (Roth-Gordon,
2007), French (Dubois, 1992), German
(Overstreet, 2005), Lithuanian (Ruzaite,
2010), Persian (Parvaresh et al., 2012),
Spanish (Cortés Rodríguez, 2006) and
Swedish (Norrby & Winter, 2002). Their use
has also been studied among different
groups of learners of English as a foreign
language (EFL) whose first language is
French (DeCock, 2004), German
(Terraschke, 2007a, b, 2010; Terraschke &
Holmes, 2007), Norwegian (Hasselgren,
2002), Persian (Parvaresh et al., 2012) and
Swedish (Aijmer, 2004).
