j0unuti~

ELSEVIER Journal of Pragmatics 29 (1998) 531-544

Telling it like it isn't - exploring an instance of

pragmatic ambivalence in supervisory discourse
Ruth W a j n r y b
41 Wiley Street, Waverly, NSW 2024, Australia

Received 18 August 1995; revised version 20 June 1997

Abstract
Those engaged in teacher supervision have always known, in an experiential sense, that the
delivery of personal feedback to teachers about their teaching is a face-threatening event.
Researchers into supervisory discourse have recently begun to monitor these processes and
track the face-oriented constraints in supervisory language. This article explores an instance
of pragmatic ambivalence in supervisory discourse in order to show how the naturally 'slip-
pery' quality of language enables supervisors to transact difficult supervisory messages while
at the same time avoiding face threat to the teacher. But the path is a precarious one and not
without risks.

I. Introduction

Critical feedback, typically following an observed lesson, is known to be a diffi-

cult and fragile speech event. From the teacher's point of view, it is face-threaten-
ing; and from the supervisor's point of view, it is uncomfortable because delivering
face threats to another trespasses the delicate, mutually constraining, tacit under-
standings that bind social beings in face-to-face social relationships (Goffman,
1972). As a result, the situation is potentially one of 'dysphoria' (Goffman, 1961:
42); and demands a large amount of emotional energy and socio-strategic skill to
enable the participants to traverse the event successfully (Brown and Levinson,
1987).
The pressure that face concerns places on participants is reflected in the language
used to negotiate the encounter. Specifically, supervisors frequently resort to a high
degree of mitigation (Wajnryb, 1994) to ease them through unenviable tasks. A
major form of mitigation is extreme linguistic obliqueness which functions interac-
tionally as pragmatic ambivalence. In short, in delivering unpalatable messages, it
helps to blur the edges of your intention so as to create a safer environment for the
negotiation of meaning.

0378-2166/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved

PII S 0 3 7 8 - 2 1 6 6 ( 9 7 ) 0 0 0 7 6 - 3
532 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

2. Indeterminacy of meaning

Before exploring pragmatic ambivalence in some depth, it is worth distinguishing

it from other kinds of indeterminacy. In fact there are three major types of meaning
indeterminacy. These are semantic ambiguity; pragmatic ambiguity; and pragmatic
ambivalence. Semantic ambiguity refers to utterances where the proposition is
ambiguous despite the unambiguous illocutionary force indicating devices (IFIDs).
According to Thomas (1987), grammatical (or semantic) ambiguity relates to ambi-
guities in the surface structure of an utterance, leading to the possibility of assigning
more than one referential sense to an utterance: e.g. "some British tourists have
found biting flies a problem" (Thomas, 1987: 6). Apart from some notable excep-
tions (such as poetry and jokes), grammatical ambiguity is unintended.
On the other hand, pragmatic ambiguity, which may have a grammatical origin,
occurs when a speaker intends only one force, but another force is possible (Thomas,
1987: 9):

A: Have you got something to drink?

B: Yes, thanks
A: No, not do you want something - I want something

Here, B computes A's question as an offer while it is intended as a request. What

both ambiguities have in common is the following: only one meaning is intended;
the ambiguity is an accidental performance error; and the speaker is often not aware
of the availability of another meaning until it is pointed out (Thomas, 1987).
In contrast, pragmatic ambivalence (Brown and Levinson, 1987; Leech, 1977) -
which may perhaps be considered a special case of ordinary pragmatic ambiguity -
involves an utterance which is deliberately indeterminate, wherein a speaker
expresses his or her meaning in such a way as to allow a number of illocutionary
meanings to be possible. "Is that the phone?" may be a straight question or a request
to the hearer to answer it (Thomas 1987: 9). Indeed, as Leech points out, "the
rhetoric of speech acts often encourages ambivalence", with utterances like "would
you like to come in and sit down" hovering on the uncertain boundary of invitation,
request and directive (1977: 99). Of course, contextual variables allow for various
interpretations, and what may be an invitation in one situation is a directive in
another. The central point to be made about pragmatic ambivalence, however, is that
the speaker's intention is unclear, and deliberately so. The exploitation lies in the
deliberativeness: both speaker and hearer understand that more than one interpreta-
tion is possible, and both understand that the motivational origins of ambivalence lie
outside the linguistic system (Bavelas et al., 1990b; Thomas, 1987).
Pragmatic ambivalence is a normal, pervasive and necessary ingredient of human
communication. It facilitates the negotiability of interactions to the point that Pocock
states that "communication rests upon ambiguity" (1973: 33). Because it is so hard
to know with complete certainty what is 'inside the head' of one's interlocutor, we
need language to have a 'slippery' quality: the facility to ease us in and out of our
apparent meanings, to give us deniability (Bavelas et al., 1990b: 14), to help us
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 533

avoid causing offence or socio-conversational breakdown. Pragmatic ambivalence

provides this facility and allows us to use language with expediency. In the terms of
Bavelas et al. (1990a,b), it gives us the notion of 'equivocation' that lies in the
murky waters somewhere between truth and falsehood.

3. Pragmatic ambivalence

Let us look a little closer at the notion of pragmatic ambivalence. An utterance is

ambivalent when the hearer cannot be certain of its intended force, as this force is
not derivable from the sense or the context or a combination of sense and context.
Speaker intention is indeterminate and not retrievable in the way it is from less indi-
rect utterances. When A says to B, looking at an open window, "isn't it cold in
here?", B may readily compute the remark as a request to walk to the window and
close it. Such a computation of meaning derives from the particular nexus of text and
context; and participants daily engage in dozens of such computations without a sec-
ond thought.
Ambivalent utterances are the opposite pole on the continuum to explicit utter-
ances whose IFIDs readily signal speaker intention (Searle, 1969; Thomas, 1984). ~
In one conference that was recorded and transcribed, the supervisor commented on
the fact that the student teacher had used ideas in his teaching that had been sug-
gested in a recent lecture (given by X) to the trainees:

(1) I noticed you trying to use some of X's suggestions.

This could mean:

(a) exactly what it says and no more, i.e. a neutral observation of fact;
(b) praise: it is good that you are trying to use suggestions being made on the
course; this shows that you are willing to try out new things;
(c) criticism: I noticed you trying but failing to succeed in using these suggestions;
(d) a combination of praise and criticism: it is good that you are trying to use these
suggestions but you still need to work at it.

It emerges in the unfolding discourse (see below) that the intention is the last, (d);
and it takes a few turns to reach it, during which the student teacher tries to make the

The continuum may be representedthus:

explicit/ ~ conventional ~ implicit --4 pragmatic
direct indirectness indirectness ambivalence
(cf. Searle (1975), (invites
using fixed inferences made
linguistic by computation
conventions) of text and
context)
534 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

meaning less imprecise and more determinate, initially by being self-critical. The
receiver of the bad news message here interprets the cues to guess the nature of the
bad news, a process which conforms with Schegloff's thesis (1988) about how such
interactions work.

(2) S: well I noticed you trying to use some of X's suggestions

T: I'm glad you used the word 'trying' ((both laugh))
S: what, you didn't feel it was successful?
T: ah, no, it's difficult, it's not my voice on the tape, so what they're hearing on
the tape and what I'm saying back to them, I don't know, it's probably a dif-
ferent stress [mm] 2 and the intonation is probably slightly different
S: had you listened to the tape and practised it beforehand?
T: I listened to it a couple of times
S: and practised?
T: I'd been trying! ((both laugh))
S: I thought, I thought it worked well ah but it wasn't always consistent

4. Motivation for indeterminacy

The motivation for pragmatic ambivalence, lying outside the linguistic system, is
strategic (Leech, 1977): it is employed deliberately for a goal-oriented purpose. A
number of explanations are possible, and these are discussed separately below. How-
ever, it may be that more than one motivation for pragmatic ambivalence exists at
any one time and that multiple concurrent motivations co-exist to varying degrees in
various supervisors in various situations to varying degrees of conscious awareness.
A key one is the 'clash of goals' explanation: the need to say something face-
risky but at the same time the need to minimise risk to face. The conflict that super-
visors feel in relation to their roles (helper/assessor) is palpably evident at times of
giving criticism where there is the tension between supporting morale and enabling
better instruction. Going off-record by clothing criticisms ambivalently allows both
goals to be pursued. Bavelas et al. contend that equivocation is a means of "saying
nothing while saying something" (1990b: 57).
Another explanation is the need for an 'out': where the speaker, uncertain of the
terrain he or she is traversing, is apprehensive about the reaction an intended utter-
ance might provoke. Ambivalence enables 'deniability' and therefore ensures an
'out' if this is what turns out to be required (Weisner, 1974): it allows one to
'retract' or 'amend' what turns out not to "meet with hearer acceptance" (McLaugh-
lin, 1984: 144). Especially in cases of supervisors encountering a resistant teacher,
or inexperienced supervisors lacking confidence in the role, or supervisors troubled
by the contradictions of the role, it helps to be able to 'back down' a little. This is

2 The text inside square brackets refers to phatic responses, though not of the nature of a full turn, by
the listener to the speaker.
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 535

only feasible if the initial criticism has been expressed ambivalently. Supervisors
quickly leam the wisdom of leaving a path open for retreat.
In one conference, the supervisor felt 3 that the teacher had been quite sceptical of
the points of criticism he had been making; he retreated somewhat up the path of
deniability with

(3) Don't get me wrong, it's not like um it was a great problem.

He was able to do this because the criticism he had voiced earlier had not been
directly explicit. Thus, indirectness served him well by providing him with the
manoeuvrability of "pragmatic space" (Leech, 1977: 23).
In another instance, a supervisor created this pragmatic space through deixis, by
distancing her concems through tense switching. She repeatedly prefaced her nega-
tive comments - or face-threatening acts (FTAs) - with:

(4) I was worried that you ...

The ambivalence derives from the fact that it is never clear whether she meant "I
was worried but am no longer" or "I was worried and continue to be worried". In
the following extract, she makes a startling (albeit unwittingly illuminative) admis-
sion (as italicised):

(5) S: yeah um did you have any particular things in mind that you were trying you
said you were trying to get, get them to tell you what was polite [mm], like
did you have any particular things that you were hoping that they might say?
T: oh just, not really not, just [nod] just general politeness tags [nod] whatever
would come out would be [nod] dealt with, but no particular sort of set words
or phrases [nod]
S: yeah that was just, that was really quite an honest question [mm] because
you know either you did or you didn't um ...

A third explanation for ambivalence is the desire by the supervisor to elicit rather
than impose: to have the criticism come from the teacher rather than from them-
selves. Supervisors tend to feel that gentle, leading questions (which often are prag-
matically ambivalent) can steer the student teacher toward the point of self-aware-
ness, which, they argue, is more effective than when it is other-imposed. This is
consistent with research that shows that suggestive supervision is preferred over
directive supervision (Pajak and Glickman, 1984). Thus the following examples:

(6) Do you think the kids like the book?

(7) Did you have any particular things that you were hoping they might say?
(8) Do you remember the question you asked?
(9) Were you happy with the language analysis?

3 These views were expressed in a subsequent interview.

536 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

Other motives may (co-)exist. There is the impact of "instrumental rationality"

(Thomas, 1987: 37) where a speaker is assured by previous experience of the expe-
dience of indirectness. Also, Lakoff (1990) points out that indirectness may suggest
intimacy by signalling shared understanding. There is, too, the sheer fun of "verbal
playfulness" (Lakoff, 1990: 171), either for its own sake or promoted by the wish to
compliment one hearer's intelligence.

5. Investigation of an instance of pragmatic ambivalence

To unpack the thinking that underlies the 'decision' to be pragmatically ambiva-

lent, an instance of pragmatic ambivalence that seemed typically motivated by face
concems within a supervisory dialogue was selected. The supervisor and the teacher
were interviewed separately about their language and motivations a few days after
the event. This procedure, referred to as triangulation by Cicourel and Jermings
(1974, cited by Adelman, 1981a) is used by Adelman (1981a) as a means of elicit-
ing "honest accounts" from key protagonists in a research project (198 l b: 78). As
with the present study, such triangulation "resuscitates intended meanings (as well
as) the significance of talk for the talkers" (Adelman, 1981b: 7). Speech acts are
seen as "incomplete", requiring interpretation to fulfil meaning in context (Adel-
man, 1981a: 80)
Stimulated recall, based on transcripts of the supervisory dialogue, was used to
elicit informants' comments on the language. This blend of post-event introspection
based on the informant-initiated data (the transcripts) makes the study more reliable
than introspection used alone (Nunan, 1992a: 132).
Introspective methods of gaining access to invisible processes such as decision-
making are fraught with difficulty (Linde, 1988; Nunan, 1992a,b). The first is the
question of conscious and sub-conscious strategies. A major difficulty was that in
order that informants might shed some light on the thinking/decision-making that
underpinned their language, they have to reflect on processes that in large part had
occurred below the level of conscious awareness, i.e. they have to retrieve intention.
A second difficulty is the threat to validity presented in the time lag between the
event and the report.
Despite these difficulties, introspective methods have the advantage of being able
to offer rich insights into otherwise largely inaccessible processes, and therefore
were considered a useful instrument. In their favour is the fact that criticism has an
affective 'cargo' which takes it out of the routine and lends it salience. The fact that
it was probably not experienced 'mindlessly' - and may have involved rehearsal and
planning, therefore making it effortful - offers greater promise of retrievability
(Fairhurst et al., 1984; Kiesler and Sproull, 1982; Tracy et al., 1987).
Some background information is necessary. The teacher, who had had about three
years of classroom experience, was new to Australia, having trained in England and
worked in South America. The supervisor's impression was that the teacher was still
in the process of adjusting to the new client-students (young Asian adults); and that
her lesson-planning, especially choice of materials and design of tasks, was not quite
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 537

appropriate. This judgement had been formed in the staffroom and was confirmed
when he watched her lesson, which he felt reflected her current stage of cross-cul-
tural adjustment. Overlaid on this impression was his sense that she herself was
unaware of what he privately called her 'culture shock'. Furthermore, he was sensi-
tive to the fact that she saw herself as successful and competent, and not in need of
induction or supervision. He was also aware that teachers w i t h some experience
behind them tend to be less willing to accept criticism than students or neophytes
(Cervi, 1991). The fact that she was new to Australia, to the school, and to this type
of client did not shake her apparent confidence in herself and her teaching. Thus,
over and above the usual reticence that accompanies the delivery of criticism, was
his awareness of how these additional elements might impact on the conference. It is
little wonder, then, that, cautiously, he poised his language on the indeterminate bor-
der of innocent inquiry and critical remark. He asked her:

(10) Are you still adjusting to teaching Latin American students? 4

The teacher's reaction to this indicates that she has chosen to interpret the illocu-
tionary force as innocent inquiry - an open, information-seeking question carrying
no negative judgement:

(11) No, I don't think so, well not in the sense that I'm really I'm still not always
comparing.

On interview, the supervisor stated that he had expressed himself ambivalently out
of fear of offending her and then having to deal with the resultant unpleasantness. He
recognised that the topic (cross-cultural difficulty) was a delicate one and that he had
to tread warily in this terrain. His guarded approach was born of his assessment of
how she might react to this topic.
That his assessment was absolutely correct emerged on interview with the teacher.
She admitted that the question (even put to her ambivalently) raised shackles: "I
didn't like it, I thought it was irrelevant". It turned out that his choice of strategy -
extreme indirectness - was well-founded. On interview, the teacher said that had the
supervisor expressed himself "more baldly", she would have "taken offence" and
reacted verbally. Indignantly, she said: "I would have said 'you don't know what
you're talking about!' ". It was clear she was offended at the very notion that she
needed time to make necessary cultural adjustments.
It emerged during the separate interviews that a great deal was happening in the
invisible pragmatic space afforded by the ambivalence of the remark. Indeed, it
seems that the interaction is happening on two concurrent levels of operation: the
utterance level and the computing-of-intention level. The participants, at any given
moment, are engaged in a surface-level interaction; but because of the indeterminacy
afforded by pragmatic ambivalence, they are also aware of a juggling of illocution-

4 Thisquestion, unpacked, means: "Are you still making the adjustment from having Latin American
students?"
538 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

ary options happening between the lines or behind the scenes or at any case at a level
that is not readily able to be pinned down. The conversation has a certain 'bi-partite'
quality. Certainly, as Nofsinger shows (1983), interlocutors orient themselves to
utterances as symbolic acts - part of a goal-oriented tactical plan which goes beyond
the utterance itself and takes into account stretches of talk.
In asking the question "are you still adjusting to teaching Latin American stu-
dents?" the supervisor clearly has to prepare himself for a number of possible
responses, some of which are outlined here:

(a) She might interpret it as an open question, in which case he can leave it at that
for the moment, content that he has 'planted the seed', confident that he can
return to the topic and raise the matter at a later time.
(b) She may react to it as an open question, as in (a); but at the same time recognise
the ambivalence as a signal of his wish to be negatively polite in respect to a
FFA. This is probably Gibbs' "authorised inference" (1987: 580) - "a speaker's
intention to produce an effect in the listener by means of the hearer's recognition
of that intention" (Gibbs, 1987: 580); what Grice (1968) called 'm-intention'.
(c) She might react to it as a negative criticism, in which case he can either pursue
the matter and support his case by presenting evidence; or he can retreat through
denial, saying that she has misinterpreted him and clarifying his supposed non-
critical intentions.
(d) She might challenge him to own up to his illocutionary force, with a question
such as "what do you mean by saying that?" to which he has a number of
options (see (c) above).
(e) She might react verbally to the remark as if it were an 'innocent' question, while
actually knowing/sensing that something else is intended; in this case she might
'play innocent' until the supervisor goes more public (perhaps with an IFID); or
she may wait until such time as she feels ready to respond to the criticism; and
all the while that she is considering these options, she knows that he probably
knows that she is doing so, as well the fact that she knows that he knows this)

While these thought processes are happening, the conversation continues at a rel-
atively innocent level. The supervisor, gaining heart from the fact that his first tenta-
tive foray into this difficult terrain has been successful (i.e. yielded no locutionary
defensiveness), moves out of the ambivalence slightly, by adjusting his indirectness
up one notch on the gradient and trying out a strategy of implicit indirectness, which
requires the hearer to recover speaker meaning inferentially, using cues both within
the utterance and within the context:

(12) I just imagined that for example that vocabulary activity with a (Spanish-speak-
ing) class I can see them all jumping around and having a lot of fun.

5 In fact, she responded according to (e). Note that (b) differs from (e) in being compliant and defer-
ential whereas (e) is a response in which the listener is simmering with emotion just beneath the calm,
surface, on-line response.
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 539

To this the teacher responds, again at a surface (i.e. locutionary) level:

(13) Well it depends, well they work together much more readily I think that's the
thing, they work together.

Still moving cautiously, but gaining confidence from the teacher's concession -
which itself forces some complicity - that there is some difference between the two
types of students, he continues, again with implicit indirectness:

(14) ... whereas our sort of classes need to be wanned up [mm] I think that you
know to me that while it was a very well constructed and interesting um way of
presenting the vocabulary I didn't see it as warmer [no, no?] and I don't I did-
n't see that as warming up in any way [no] and I think that our students really,
particularly in that first lesson of the day and particularly with the lower levels,
they really need wanning up you know.

All the while, the teacher is listening and producing phatic acknowledgment c u e s 6
that encourage him to continue. This cautious progress - stepping warily and gin-
gerly from pragmatic ambivalence toward gradually increasing explicitness - is a
recurring pattern in the data on which this research is based.

6. The pursuit of goals

A pragmatic analysis such as this exposes the staging of degrees of indirectness.

In addition, it highlights the fact that the conversation is driven by two concurrent
yet discordant imperatives. There is, in the first place, a social imperative that carries
the interaction along in a face-attending manner, with interactants "working to cre-
ate social order out of potential conflict" (Good, 1979: 166). This appearance of
mutual agreement and amicability suffices - for face, as the word suggests, is largely
a matter of surface appearances rather than deeply-felt sincerity. There is an enor-
mous pressure on participants to maintain the flow, to avoid what Weisner calls
'flow-breaking responses' (1974: 727). Indeed, she posits this as a Grice-like maxim
- 'maintain smooth flow' - derived as are the other maxims from the overarching
Co-operative Principle.
Surface-level harmony can operate even while a great deal of less-than-amicable
thinking is going on in the pragmatic space around what is said. The teacher feels
discomfort and the supervisor knows he is treading on thin ice; yet the quest for
social harmony compels them both to pursue an amicable route even while they both

6 Evidence of phatic material came from listening to the tapes on which the tapescripts are based.
During these interviews, the tapescripts were used but the tapes were available as a back-up in the event
that either T or S denied having said anything that appears on the transcript. However this did not
happen. Initial direct use of tapes was avoided as it was considered potentially more threatening to the
interviewees.
540 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

are entertaining thoughts of illocutionary forces other than the ones to which they
appear to be responding in their surface utterances.
However, there is a second imperative. As well as the pressure to follow the con-
versation, participants are also driven onward by their own goal-oriented need,
which may or may not be in harmony with the locutionary level, socially-driven
goal. In an instance, such as the one analysed here, where there is a discordance
between the two imperatives, the attendance to the social imperative is still predom-
inant. It carries the conversation forward (Weisner, 1974); and therefore gives the
participants time to consider their own next move. Also, by keeping the wheels of
the conversation oiled, it fuels hope that the individual's own goals may yet be met,
something that would not happen if the communication broke down prematurely. So,
even while the social imperative and the goal-driven imperative may be essentially
discordant, the strategy for addressing both is the same - maintaining (at least for the
moment) flow and amicability at the locutionary level. The pragmatic space in which
these imperatives jostle and compete is beautifully afforded by the slipperiness of
pragmatic ambivalence.

7. The risk of pragmatic ambivalence

Indirectness allows interactants to navigate a socially-acceptable course in pur-

suance of their own goals. In this way it serves supervisors strategically in expedit-
ing their oft-conflicting roles of helper and assessor, allowing them to resolve the
competing tensions of clarity and politeness. Yet, as "no failsafe algorithm exists for
verbal communication" (Gibbs, 1987: 569), a central issue here is the level of risk
involved - i.e. the likelihood of pragmatic failure. In Thomas' terms, there is a risk
of "cross-cultural pragmatic failure" (1983: 91), defined as any communication
between two people in any particular domain where a gap in their shared background
assumptions lead to some degree of miscommunication (Thomas, 1983).
Dascal and Berenstein (1987) define success in communication as measurable by
the addressee's ability to reach the pragmatic interpretation being conveyed by the
speaker. The more indirect the communicative act, the more complex the listener's
task in creating their own "construal of the speaker's meaning" (Dascal and Beren-
stein, 1987: 141). In this regard, the listener faces a bundle of unknowns, termed
'layers of significance' which must be unearthed and clarified. They cite Fillmore's
terms (1976) in which the addressee has the duty to answer four questions about the
speaker's meaning: 'what did he [sic] say? what was he talking about? why did he
bother to say it? why did he say it in the way that he did?' (1987: 140). The first
two are essentially semantic questions, the last two pragmatic, and the fourth,
rhetoric. Dascal and Berenstein add a fifth 'why did he [sic] choose m e as an inter-
locutor for this particular subject matter?' (1987: 142). They posit that there are
"two distinct modalities of pragmatic interpretation" - comprehending and grasping
(1987: 148). While Fillmore's four questions have to do with the addressee's mode
of comprehending, the last question requires a mode of grasping, i.e. recognition of
which game rules are governing the interaction.
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 541

Success in communication, therefore, is a two-way street; and the onus has much
to do with how the hearer handles their 'side'. The more indirect the speaker's lan-
guage, the more work is required of the hearer; and the less assured the speaker may
be of end product success. There are, in fact, two risks. The first has to do with the
intended perlocutionary effect (reduced offence through a softened FTA), which ulti-
mately is "in the hands (or ears) of the hearer" (Fraser and Nolen, 1981: 96). Of this
risk, Fraser writes:
"To mitigateis to bring about a consequentpsychologicaleffectwithinthe hearer,and successin doing
this is problematicat best ... ultimatelyit is up to the hearer to determinewhetheror not an unwelcome
effecthas been softened."(Fraser, 1980: 349)

The second risk is whether the trade-off in face has in fact worked; or whether the
one goal has succeeded at the cost of the other - in other words, how 'intact' is the
message? As M~iseide shows in his analysis of doctor-patient interaction, "delivered
messages require a receiver in order to have any communicative function" (1981:
161). Pocock, similarly, cites communicative constraints: "all verbalised action is
mediated.... In performing a verbalised act of power, I enter upon a polity of shared
power" (1973: 33).
Blum-Kulka (1987) would argue that a third risk lies in a point on the gradient of
indirectness where 'indirect' becomes 'impolite'. She defines politeness as "the
interactional balance achieved between two needs: the need for pragmatic clarity
and the need to avoid coerciveness" (1987: 131). Thus, an overly indirect speaker
may impose on the hearer by the sheer weight of the cognitive burden involved in
the inferencing of indirectness.
Fairhurst et al. (1984) may argue that there is a further risk. Because people in
authority positions have difficulty asking purely informational questions (Goody,
1978), subordinates may interpret apparently fact-finding questions as blame-loaded,
thus prompting defensiveness which itself may curb or distort the flow of informa-
tion (Gibb, 1969).
That the process is fraught with peril is supported by evidence from experimental
social psychology, which contributes to our understanding of how~ people interpret
each other in face-to-face interaction. For example, Swann and Read (1981) investi-
gated the influences of people's self-conceptions on how they solicit and respond to
feedback during social interaction. The findings suggest that there is a systematic
tendency for people to solicit self-verifying feedback. The conclusion is that within
interaction people operate as "active agents, who, after fashioning images of them-
selves, behave in ways that tend to bring their social environments into harmony
with these images" (1981: 1127). A later study (Swann, 1987) investigated the
negotiation of identity that happens at the point of interaction; and uncovered a set
of "self-verification strategies" (1987: 1039) by which people process feedback in
ways that make the responses seem more supportive than in fact they are.
Such evidence suggests that teacher-supervisees are not passive recipients of
imposed messages, but rather co-constructors of their own social reality. It is part of
the contention of this research that lack of supervisory explicitness runs the danger
of fuelling mis-conceptions: it leaves interpretation relatively open to chance, and by
542 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

so doing, perhaps lubricates the wheels of meaning construal by which teachers,

given the option, would rather not take up a negative appraisal. This would facilitate
a process by which central pedagogical issues that should be on the supervisory
agenda may be skirted, camouflaged, down-weighted, and ultimately ignored.
Apart from the issue of contentious topic and conflict avoidance, the supervision
literature reports ample support for the notion that messages delivered indirectly are
risky. For example, Pajak and Seyfarth refer to habits of 'intellectualising' and
obliqueness by means of which supervisors diminish teacher defensiveness but
increase the chance "that the teacher will fail to understand exactly what the super-
visor is criticising" (1983: 22). Further, they write:

"Supervisors who dilute their messages ... can come across as unclear, and they leave teachers confused
about how concerned they really ought to be about the problem ... too much indirectness can result in a
loss of clarity. Being direct is not only clearer; it is also frequently more humane." (Pajak and Seyfarth,
1983: 22-23)

This is corroborated by Cervi's findings on feedback in which he asserts that

supervisors often erroneously conclude that a point of criticism they have made has
been understood whereas, in fact, "the student teacher may not have heeded this as
criticism" (Cervi, 1991: 109).
Thus, the issue of message vulnerability and therefore risk of failure is a very real
one. When supervisors deliver criticisms in a less than explicit way, student teachers
are required to make certain intermediary inferences in order to understand the bTA
qua criticism and recognise the operant issues. While the human mind is astounding
in what it can compute at the delicate interface of form and context, there is no doubt
that journeying along the gradient of indirectness away from the pole of explicitness
is a journey in the direction of pragmatic risk. As can be seen in the analysis of the
'Latin American' incident, an assumption may be "manifest" to a person to varying
degrees (Leech and Thomas, 1988: 52). There it was shown that although the text of
the interaction skirted the issues, the teacher computed the criticism qua criticism
but chose not to respond to it at this level. (One may well question the value of this
type of awareness within a supervisory context: after all, if it is not firmly 'on the
agenda', it is being tangentially addressed at best.) Even beyond this, however, there
may also be a point at which the teacher may not compute the criticism qua criticism
at all - at any level. At such a point, where it might well be said that the teacher has
missed the message, one may reasonably speak of 'pragmatic failure'.

References

Adelman, Clem, 1981a. On first hearing. In: Clem Adeiman, ed., Uttering, muttering: Collecting, using
and reporting talk for social and educational research, 78-97. London: Grant McIntyre.
Adelman, Clem, 1981b. Introduction. In: Clem Adeiman, ed., Uttering, muttering: Collecting, using and
reporting talk for social and educational research, 1-12. London: Grant McIntyre.
Bavelas, Janet Beavin, Alex Black, Nicole Chovil and Jennifer Mullett, 1990a. Truths, lies and equivo-
cations: The effects of conflicting goals on discourse. In: Karen Tracy and Nikolas Coupland, eds.,
Multiple goals in discourse, 135-161. Philadelphia, PA: Multilingual Matters.
 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544 543

Bavelas, Janet Beavin, Alex Black, Nicole Chovil and Jennifer Mullett, 1990b. Equivocal communica-
tion. Newbury Park, CA: Sage.
Blum-Kulka, Shoshana, 1987. Indirectness and politeness in requests: Same or different? Journal of
Pragmatics 11 : 131-146.
Brown, Penny and Stephen Levinson, 1987. Politeness: Some universals in language usage. Cambridge:
Cambridge University Press.
Cervi, David, 1991. A study of the practical component of a short, pre-service teacher education program
for adult TESOL, with a special focus on feedback interaction. Unpublished M.A.-thesis, Macquarie
University, Sydney.
Cicourel, Aaron and K.H. Jennings, 1974. Language use and school performance. New York: Academic
Press.
Dascal, Marcelo and I. Berenstein, 1987. Two modes of understanding: Comprehending and grasping.
Language and Communication 7: 139-150.
Fairhurst, G.T., S.T. Green and B.K. Snavely, 1984. Face support in controlling poor performance.
Human Communication Research 11 : 272-295.
Fillmore, Charles J., 1976. Topics in lexical semantics. In: Roger W. Cole, ed., Current issues in lin-
guistic theory, 76-138. Bloomington, IN: Indiana University Press.
Fraser, Bruce, 1980. Conversational mitigation. Journal of Pragmatics 4: 341-350.
Fraser, Bruce and W. Nolen, 1981. The association of deference with linguistic form. International Jour-
nal of the Sociology of Language 27: 93-109.
Gibb, John, 1969. Defensive communication. Journal of Communication 11(3): 141-148.
Gibbs, Raymond W., 1987. Mutual knowledge and the psychology of conversational inference. Journal
of Pragmatics 11 : 561-588.
Goffman, Erving, 1961. Encounters: Two studies in the sociology of interaction. Indianapolis, IN:
Bobbs-MerriU.
Goffman, Erving, 1972. On face-work: An analysis of ritual elements in social interaction. In: J. Laver
and S. Hutcheson, eds., Communication in face to face interaction, 319-346. Harmondsworth, UK:
Penguin.
Good, Colin, 1979. Language as social reality: Negotiating conversation. Journal of Pragmatics 3:
151-167.
Goody, Esther N., 1978. Towards a theory of questions. In: E.N. Goody, ed., Questions and politeness,
17--43. Cambridge: Cambridge University Press.
Grice, H.P., 1968. Utterer's meaning, sentence-meaning and word meaning. Foundations of Language 4:
225-242.
Kiesler, S. and L. Sproull, 1982. Managerial response to changing environments: Perspectives on prob-
lem sensing from social cognition. Administrative Science Quarterly 27: 548-570.
Lakoff, Robin, 1990. Talking power: The politics of language. New York: Basic Books.
Leech, Geoffrey N., 1977. Language and tact. Linguistics Agency, University of Trier, Series A, Paper
No. 46.
Leech, Geoffrey N. and Jenny Thomas, 1988. Pragmatics: The state of the art. Lancaster Papers in Lin-
guistics 48.
Linde, Charlotte, 1988. The quantitative study of communicative success: Politeness and accidents in
aviation discourse. Language in Society 17: 375-399.
Mhseide, Per, 1981. Sincerity may frighten the patient: Medical dilemmas in patient care. Journal of
Pragmatics 5: 145-167.
McLaughlin, Margaret, 1984. Conversation: How talk is organised. Beverly Hills, CA: Sage.
Nofsinger, Robert, 1983. Tactical coherence in courtroom conversation. In: Robert Craig and Karen
Tracy, eds., Conversational coherence: Form, structure and strategy, 243-258. Beverly Hills, CA:
Sage.
Nunan, David, 1992a. Research methods in language learning. New York: Cambridge University
Press.
Nunan, David, 1992b. The teacher as decision-maker. In: John Flowerdew, Mark Brock and Sophie
Hsia, eds., Perspectives on second language teacher education. Proceedings from the Conference on
Teacher Education in Second Language Teaching, 135-165. 17-19 April 1991, Hong Kong.
544 R. Wajnryb / Journal of Pragmatics 29 (1998) 531-544

Pajak, Edward F. and Carl D. Glickman, 1984. Teachers' discrimination between information and con-
trol in response to videotaped simulated supervisory conferences. Paper presented at the Annual Meet-
ing of the American Educational Research Association. 23-27 April 1984, New Orleans. Educational
Information Resources Centre (ERIC), ED 243 842.
Pajak, Edward F. and John T. Seyfarth, 1983. Authentic supervision reconciles the irreconcilables. Edu-
cational Leadership 40: 20-23.
Pocock, J.G.A., 1973. Verbalising a political act: Towards a politics of speech. Political Theory 1:
27-45.
Schegloff, Emanuel A., 1988. On an actual virtual servo-mechanism for guessing bad news: A single
case conjecture. Social Problems 35: 442-457.
Searle, John, 1969. Speech acts: An essay in the philosophy of language. Cambridge: Cambridge Uni-
versity Press.
Searle, John, 1975. Indirect speech acts. In: P. Cole and J.L. Morgan, eds., Syntax and semantics, Vol.
3: Speech acts, 30-57. New York: Academic Press.
Swann, William B., 1987. Identity negotiation: Where two roads meet. Journal of Personality and Social
Psychology 53: 1038-1051.
Swann, William B. and Stephen J. Read, 1981. Acquiring self-knowledge: The search for feedback that
fits. Journal of Personality and Social Psychology 41:1119-1128.
Thomas, Jenny, 1983. Cross-cultural pragmatic failure. Applied Linguistics 4:91-112.
Thomas, Jenny, 1984. Cross-cultural discourse as 'unequal encounter': Towards a pragmatic analysis.
Applied Linguistics 5: 226-235.
Thomas, Jenny, 1987. Complex illocutionary acts and the analysis of discourse. Lancaster Papers in Lin-
guistics II. Department of Linguistics, University of Lancaster.
Tracy, Karen, Donna Van Dusen and Susan Robinson, 1987. 'Good' and 'bad' criticism: A descriptive
analysis. Journal of Communication 37: 46-59.
Wajnryb, Ruth, 1994. The pragmatics of feedback: A study of mitigation in the supervisory discourse of
TESOL teacher educators. Unpublished Ph.D. thesis. Macquarie University, Sydney.
Weisner, Ann, 1974. Deliberate ambiguity. In: Papers from the Tenth Regional Meeting of the Chicago
Linguistics Society, 723-731, 19-21 April 1974, Chicago, IL.