Introduction
In Discourses of Counselling (Silverman:1997), I devote two chapters to the link between different forms of advice-giving and advice-reception in HIV counselling. In particular, in Chapter Seven of that book, I look at how, in certain conversational environments, resistance to advice may assume quite spectacular proportions. This chapter on advice seeks to advance the argument by spelling out a device used by counsellors to stabilise advice-giving.
This chapter is organized in two related sections:
1) It summarises what we have learnt so far about the interactional difficulties that can arise in personalised advice-giving.
2) It identifies an interactional solution for advice-givers and advice-recipients to the problem of the non-uptake of advice.
The bulk of the data discussed in this chapter is drawn from 21 pre-HIV-test counselling interviews in a U.S. city (US2). Some data from a U.K. testing centre (UK1) is introduced for comparative purposes.
Slightly different procedures are used at each centre. In the U.S. clinic, Ps fill in a questionnaire prior to the interview and Cs refer to Ps answers during counselling. In the U.K. clinic, Ps are usually seen 'blind' - although some come with referral note from a doctor at the clinic.
Categorization and Advice-Giving
In professional-client encounters, even apparently straightforward information-delivery can imply a categorization which the speaker might not want to imply. For instance, in the extract below from a British HIV-test counselling interview, the counsellor (C) is outlining her agenda to the client (P):
Ex 1 (UK1) [SW-1.8A]
C: we also need to go over with you what happens (.) when someone gets a positive test result, P: Yes. C: but plea:se remember we have to do this with .h everyone who's tested .hh[h and we're not saying= P: [Ri:ght. C: =that we think you know you're in (.) any greater ris[k (of uhm of a) positive result. .hhhhhh= P: [No ( )
Notice how C attends to the implications of her 'need to go over with you what happens (.) when someone gets a positive test result'. P is told that this is an activity that arises from the category 'someone having an HIV-test' and not from the category 'likely to be infected with HIV'. Through this postscript, C neatly marks and manages the implications of her agenda-statement which imply that P is 'someone (who might) get(s) a positive test result'. Notice also how P collaborates with this (re)categorization by his overlapping acknowledgments on lines 6 and 9.
Asking a particular question of someone can also be rich in implications. For instance, by asking the question, you imply that the question uses categories which are relevant to the person at whom such a question is directed. Such an issue arises when Cs ask Ps about their possible drug-use. Note the preface (line 1) and the postscript (line 4) to the question here:
Ex 2 (UK1) [SW-1.8A]
C: er: I have to ask you this have you ever injected drugs. P: No. C: Because they're the sort of highest ris:k (.)
As with questioning, advice-giving can imply particular categories through which the advice-recipient is being defined. For instance, if a counsellor advises me 'to be sure to have safer sex in future', the implication may be that I have not been having safer sex in the past.
A second problem, which is peculiar to advice-giving, is that it seems to require strong uptake from its recipient if the advice-giving is going to persist over many turns. For instance, as I noted in Chapter Six [Silverman 1996], Heritage and Sefi (1992) argue that the absence of clear uptake markers (like 'oh really' or 'I think you're right' or 'how do you mean?') are strongly implicative of a client's resistance to the advice offered.
Throughout our corpus of examples, counsellors exit quickly from personalised advice-giving when patients offer only minimal response tokens or when they display overt resistance. By contrast, according to Jefferson and Lee (1981), advice is most likely to be well received where it is 'initiated by the troubles-teller' and emerges:
'as the logical outcome of a diagnosis offered by the troubles-recipient and concurred in by the troubles-teller; i.e. the advice is sequentially appropriate and the talk is interactionally "synchronous"' (408)
This suggests the possibility that, where these features are absent, i.e. advice is delivered 'prematurely' and/or without a 'trouble' being presented by the client, we are likely to find the advice rejected.
Heritage and Sefi's (1992) account of health visitor-mother interactions entirely fits this explanation of advice-rejections. Unless the client seeks advice, she will usually resist advice which is not recipient designed to an elicited 'problem' or sequentially appropriate.
Ex 3 below shows how advice reception is organized in these circumstances, i.e. the health visitor (HV) has not based her advice (lines 1-3 and 5-6) on any 'diagnosis' or problem provided by the mother (M) or father (F):
Ex 3 [Heritage & Sefi (31) [4A1:14-15]]
HV: but it is recommended that if possible (0.2) all of them are better than: (0.7) i- it's better to have them all than (0.5) uhm: (0.7) no:t,= F: =( [ ) ( (to baby)) HV: [whooping cough can be a killer in the baby under one. (1.0) HV: Uh:m (1.2) but it m- (0.2) maybe you'd like to have a think about it and= M: Mm hm HV: uh:m talk it over with the doctor (1.0) HV: when you see him at clinic M: Mm hm.
Heritage and Sefi focus on the 1.0 second post-advice pause at line 8. They argue that HV's subsequent downgrading of her advice (line 9 onwards) shows that she monitors the parents' failure to use the 1.0 second slot as indicating resistance to it.
Following Jefferson and Lee (1981), Heritage and Sefi (1992) show how a more favourable environment can be created for advice-giving by the establishment of an agreed 'problem' being experienced by the potential advice-recipient. As I show in Silverman (1996, Chapters Six and Eight), advice is much more likely to be well received when it is addressed to a client 'problem' elicited by a series of questions and a request for specification.
Given the sort of difficulties in obtaining uptake that we have been examining, the data suggest a number of strategies
that professionals can use to stabilise advice-giving. For instance, in Ex 4 below, implied advice about 'safer sex' can be packaged in the form of a question about a client's personal dispositions. C has already learned that P does not currently use condoms with his partner (not transcribed). However, it is noticeable that C abstains from delivering explicit advice on the topic. Instead, he asks a question about condoms which sets up a perspective-display sequence (PDS)(see Maynard:1991) in which a client is invited to offer their own perspective on the situation:
Ex 4 (US2) [B39 B1]
C: now, as far as you're- as far as safer sex is concerned, hh because you have been in a twelve or whatever month relationship with this person, how would you feel about putting the condoms back on? P: I'd prefer not to actually, um C: and how does she feel? P: well the same way (0.3) C: awr:ighdy. (1.0) C: here's what I would recommend then (0.1) obviously: (0.6) you folks have been (1.4) not (0.7) using protection (0.5) for a year now. (2.0) C: ((tch)) wha' do you use for birth control?
This is a fascinating extract in the way in which C attempts to move towards a recipient-designed advice-sequence. Notice how C bases his question about condoms (lines 3-4) on information that he has already gleaned from his client.
However, C's question is highly implicative of a piece of advice (i.e. 'I think you should put the condoms back on'). P shows he monitors the question the same way using 'actually' to mark his answer as the dispreferred one.
Nonetheless, C still abstains from explicit advice-giving. At line 6, he extends the PDS by asking a question about P's partner. When this produces the same answer (marked as dispreferred by 'well'), C simply acknowledges that he has heard what P is saying ('awr:ighdy.').
Although C now starts to deliver personalised advice (line 11), this is then aborted by a further question from C ('wha' do you use for birth control?'). If C wants to advise the use of condoms, his questions have created a highly unfavourable environment. To achieve uptake of that advice, he will need to align P much more with his perspective.
This is how the interview continues:
Ex 5 (US2) [B39 B1] (continued)
C: ((tch)) wha' do you use for birth control?
(1.6)
P: pi:lls.
((C asks more questions and then delivers information
about the likelihood of passing HIV in a monogamous
relationship))
C: .hhh the moral of this story is that we
really don't know: (1.2) whe:n or why: (1.1)
an HIV infection is spread (0.3) between a man
and a woman. (1.3) .h sometimes it happens and
sometimes it doesn't. (1.1) .hh so that if: (1.1)
you: (0.5) are in fact HIV po:sidive and we- we
don't know that yet do we?
(1.0)
C: .hh that although you have had maybe hundreds
of sexual relationships: (0.5) with your girlfriend.
(1.2) .h she may still be HIV negadive.
(2.2)
C: okay:? (0.6) .h and that if you were to put on
the condoms for the next two weeks: (1.0) .h you
would (0.2) sto:p (0.6) any po:ssiblidy (0.8) of
infecting her. (0.2) see what I'm saying?
(0.2)
P: I understand, [yeah.]
C: [.hhh ] so ho:w would you
feel about using co:ndoms for the next couple
of weeks:?
P: =('at 'ould) definately make sense:.=
C: nkay,
At the end of this extract, notice how P finally provides a marked acknowledgment of C's implicit advice (=('at 'ould) definately make sense', indicating that his answer is preferred by his early entry and upgrade ('definately') . However C's final question about condom use ('ho:w would you feel about using co:ndoms for the next couple of weeks:?') has been delayed through many turns and is now limited to the period during which the test result is unknown (to 'the next couple of weeks').
Following Heritage and Sefi (1992), this extract shows how a sequence whereby the parties establish a shared alignment to a problem is likely to be associated with client uptake of the advice. It also underlines Maynard's (1991,1992) point that, where a PDS delivers a 'misaligned' P perspective, further questions may align the client to the professional's perspective.
Following Silverman (1996), the analysis also shows how a question can be heard as advice, without the need for clearly-marked advice to be delivered. The advantage of this strategy is two-fold. First, because questions are part of an adjacency-pair of question and answer (Sacks, Schegloff and Jefferson:1974), some sort of recognizable answer can be expected. Second, resistance by the answerer to the advice implied in the question need not be threatening to further pursuit of the topic precisely because it is ambiguous whether advice has been given in the first place.Thus we have identified a further means through which professionals can manage the problematic character of constituting themselves as advice-givers.
In this chapter, however, I am mainly concerned with another interactional solution to the instability of advice-giving. As with framing advice as a question, it plays with the ambiguity of how a minimal client response might be heard. In this case, the issue is advice which is packaged to allow multiple responses to be managed. The point here, as noted earlier, is that personalised advice-giving requires far stronger uptake to stabilise itself than does information-delivery. So, if a set of turns are hearable as perhaps information-delivery, then they can follow one another without any difficulty, given only an occasional response-token from the other party. I refer to this as an Advice as Information Sequence or AIS.
The following data from a British clinic shows how the AIS works. It involves 'advice' about pregnancy to a female client.
Ex 6 (UK1) [SW-2.8A]
C: (0.8) Uh:m the other a:spect that I must cover with you as far as the fact that you're fema:le .hhh is: (.) with (.) any females who're having the test if the test is positive .hhh we do tend to advi:se against becoming pregnant. (0.4) P: Mm [hm C: [Uh::m having said that (0.8) the statistics are showing about a one in six risk to baby.=t- It depends on whether: that person is an optimist or a pessimist you know as to how they look on that. .hhhh So if a woman really wants a chi:ld (.) then we advise her becoming pregnant sooner (.) rather than leaving it. (.) P: Mm hm=
In Ex 6, C begins by aligning P to the next topic by invoking P's gender, allowing C to appeal to one of the category-bound activities associated with that gender (i.e. 'becoming pregnant', line 5). This activity is embedded in what is hearable as a piece of advice bearing on a fundamental matter ('we do tend to advi:se against becoming pregnant.')
However, note P's very limited uptake of the advice. After a 0.4 second pause, she simply produces the utterance 'mm hm'. This sort of response token works only as a continuer; it does not indicate any uptake of the advice. Yet C continues her flow of talk in an undisturbed way, overlapping with P's response token and going on to produce a modified piece of advice on the same topic. Once again, this elicits only a response token from P.
Here is, presumably, an important body of advice. Although it receives very limited (unmarked) acknowledgments, possibly indicating passive resistance (Heritage and Sefi:1992), no interactional difficulties are observed.
How can this be? For instance, in Ex 7 below, we see how a client's 2.7 second pause after advice is given was highly implicative of resistance to advice about condoms:
Ex 7 (US2) [B42A1]
C: .hhh um (0.7) works >bedder< if you use 'em all the ti:me. (2.7) P: it's the heat of the mo:ment sometimes.= C: =tch .hh (0.5) n'kay, and then we'll ta:lk about tha:t, too:.
Here, P's delayed response (lines 7-8) is followed by C extracting himself from further immediate discussion of the topic by an agenda-statement. Given this, how can advice-giving proceed smoothly with only minimal response-tokens from P as in Ex 6 above?
I suggest that the character of the communication in Ex 6 is hearably much more ambiguous than in Ex 14. First, there is not a complete fit between the categories used in the 'advice' and the advice-recipient. Although P is identified as 'female' (and thus someone who may become pregnant), she has not had her HIV-test yet and, indeed, may not even decide to have one. Thus, when C invokes the category 'the test is positive' (line 4), this links her advice to a category that P does not now (and may never) occupy. This means that it is hearable as the advice C 'would give' if certain things were to happen to the client (i.e. as conditional advice). Such a tentative formulation of a situation which might be appropriate to someone but not necessarily the client has been identified in HIV counselling as a 'proposal of the situation' or POTS device (Kinnell & Maynard:1996).
Given the implied non-relevance of the 'advice' to P's present situation, what C says about pregnancy is hearable not as personalised advice all but as information-about-the-kind-of-advice-we-give-to-people-in-this-clinic. This way of hearing C's utterance is further strengthened by her preface to it: namely 'we do tend to advi:se', where the use of 'we' and 'tend' imply a general policy rather than recipient-designed, personal advice for this client. So what C says about pregnancy is doubly hearable as information about advice rather than unambiguous personal advice.
This has crucial implications for uptake. Given that what C is saying is hearable as information, P need only offer minimal response tokens to maintain the format. Moreover, apparent 'advice' about a highly sensitive topic can proceed in a way that manages the delicacy of what is being said.
So the AIS works by shielding both C and P from the implications of non-uptake of advice, given that such non-uptake is highly likely where the 'advice' has been delivered without any attempt by C to discover P's perspective about a presumably sensitive matter. In this environment, C's response tokens need not be heard as the implicit resistance that Heritage and Sefi (1992) suggest. On the contrary, they are hearable as sustaining information-about-the-kind-of-advice-we-give-to-people-in-this-clinic. Hence, through the AIS, everything can proceed smoothly and C can safely complete her counselling agenda.
The remainder of this chapter is devoted to showing how the AIS works in the U.S. data. We begin with an extract where the American counsellor is addressing the same possibility - a positive HIV test. Here, however, with a male client, C is discussing the medical needs of someone who tests positive:
Ex 8 (US2) [B38 B1]
((Early on in the interview; C has just emphasized the importance of early detection))
(0.6) C: a::nd (1.5) u::m some- okay, if a person (.) should happen to test >an'ibody< po:sitive .hhh we would strongly encourage them to seek out a a physician (.) particularly physicians who work in infectious disease .hhh u:m because they seem to: (1.0) uh:: be m:ore on top of what's happening in the world of: viral infections and things like that. (0.4) P: mm hmm= C: =and medications .hhhh we would also encourage a person certainly to take a look at some of their behavior patterns .hhh u:m encouraging them to: .hhh look at things that might have in fact (0.5) caused them to (feel) infected in the first place whether that be sexual practices that were not particularly safe? (0.7) o:r (0.6) u:m (0.5) drug use: (0.4) *or (0.7) um* (1.0) well those two basically (0.4) *I m'n not much you can do about transfusions.* (1.5)
Let us look at C's first turn. As in the British extract, C's advice commences with an 'if' given in regard to a category (testing antibody positive) with only a possible relevance to the advice-recipient. This is what Kinnell and Maynard (1996) call a Proposal of the Situation or POTS sequence. Moreover, once again, the 'advice' is not personalised but addressed to 'a person' (line 2) and 'them' (line 4), in the same manner as C's use of 'any females' in the previous extract. Similarly, just as the British counsellor says 'we tend to advise', her American colleague says 'we would strongly encourage' (lines 3-4).
Once more, we have a C packaging what she is saying as, at best, ambiguous advice-giving, hearable as information-about-the-advice-that-would-be-given. The evidence that both parties monitor C's talk in this way is found in the lack of local problems created by P's minimal uptake - no response at the candidate completion point of the 'advice' during the micro-pause after 'seek out a physician' (line 4) and the presence of only a response token on line 9.
Now C takes a second turn without any noticeable turbulence in the design of her utterance. In this second turn, she introduces fresh topics (safer sex, drug use, transfusions) but they are all referred back to the category 'testing positive' and packaged via the same pro-terms ('person', 'their' and 'them'). Despite the presence of two slots where P might have taken a turn (after 'drug use', line 16 and after 'transfusions', line 18), not even a response token is contributed.
These last two sequences have shown that the AIS is a powerful device which manages the potentially difficult interactional problems of advice-giving and advice-reception about presumably delicate topics. It constitutes the professional as a mere reporter on the-advice-we-give-in-this-clinic rather than as a potentially intrusive personal advice-giver. It allows the client to be defined as an acceptably passive recipient of information about the kinds of things that other people get told (or that she may get told in future). Finally, and most significantly, it overcomes the potentially damaging local implications of minimal client uptake which would arise if Cs' could be heard to be giving clear-cut advice.
So far, we have been only dealing with advice about a category which P does not currently occupy ('positive person') packaged as a proposal of the situation. Let us thus turn to examples of advice on categories which may be more relevant to P, albeit often constituted in a conditional way. We will focus on the topics of frequency of HIV-testing and condom use. Is the AIS found here and, if so, how does it work?
We begin with two examples of 'advice' about the frequency of HIV-testing. In Ex 9, this is packaged as 'the recommendation is':
Ex 9 (US2) [B35 A2]
C: THE RECMMENdation is: (0.2) uh: fo:r people: (0.8) who have been been at risk any time= P: =mmhmm= C: =regardless of: (0.4) uh::m (0.6) their (1.2) sexual ac[tividy] whether it being- going from you know= P: [mmhmmm] C: =z:ero to very active .hhh ah the recommendation is still that everyone be retested once a yea:r= P: =yeah (0.4) C: until there is actually a test for the virus itself= P: =right (1.6) P: (tch)okay= C: =okay^ (.) any other questions?=
As already noted, from this extract onwards we are dealing with advice based on categories which are made directly relevant to clients. Here this relevance to the client is asserted by C's reference to 'people: (0.8) who have been at risk any time=' since Cs can usually assume that clients present themselves for an HIV-test because they feel that they have been 'at risk'.
Nonetheless, C does not address his 'advice' directly to P but to 'people' (line 1) and 'everyone' (line 9). Moreover, the author of this 'advice' is impersonal and hearable as some official body ('the recommendation is', line 1, repeated on line 8).
This way of setting up advice follows Peyrot (1987) who noted how psychotherapists made propositions concerning 'some people' without personalising them to their client. Peyrot calls this an 'oblique reference'. As such, it may take the form of a reference to 'people' (as in Ex 9, line 1) or to a 'you' which may be heard as an indefinite 'one'.
Such advice-giving by 'oblique reference' plays with the ambiguity between advice and information-about-the-advice-that-we-give-here-in-the-clinic. Thus, as an AIS, RTs will satisfactorily maintain the format.
In Ex 9, at the first hearable completion point of C's 'advice', 'retested once a year' (line 7), P provides such a response token ('yeah', line 10). When C adds another piece of information ('test for the virus itself'), P offers 'right' and adds '(tch) okay' (line 15) when C does not take up the 1.6 second slot for a turn on line 14.
Given the ambiguity of the format, built into an AIS, P's response tokens are hearable as merely indicating recipiency to C's flow of information rather than as resistance to advice. It should be emphasised, however, that there is no automatic association between an AIS and minimal client uptake, like response tokens. Once a turn (or series of turns) is hearable as 'information', then a whole range of responses, from response tokens to newsworthiness tokens, satisfactorily maintain the format. Information-receivers minimally need to indicate recipiency, however they can do many other things as well without challenging the format.
In Ex 10 below, also concerned with a recommendation about repeat testing, we find an extended response from the client:
Ex 10 (US2) [32:1]
C: (.hh)so: ((cough cough)) if have had any possible exposure we do recommend that you:: be testedt (.h)um:: every six months fo:r at least ah a year t-to: (.) two years (1.0) P: no: I don't think that I- (2.0) I have but I- (1.0) (h) it's jus something I feel bedder about= C: =mmhmm=
Here C packages his 'advice' slightly differently to the previous extract. Although, he still uses an 'impersonal' voice ('we do recommend', line 2), the recommendation is now
addressed rather more directly to P (C says 'you' - line 2 -rather than 'people') albeit made conditional by C's reference to 'if you have had possible exposure'.
P monitors the talk in this way, questioning whether the category 'any possible exposure' (lines 1-2) applies to him:
'I don't think that I- (2.0) I have' (line 6). However, P now characterises his motivation in a way that potentially aligns himself to C's 'advice': 'it's jus something I feel bedder
about' (line 7).
Even if it turns out that P is resisting the 'recommendation' of follow-up testing, this need not to challenge the communication format established here. For, in an 'information' sequence, an information-recipient can choose whether to hear the information as personally relevant. Conversely, in an unambiguous advice-sequence, unless the advice-recipient shows that he hears the advice as personally relevant, the very advice-giving format is challenged. The beauty of the AIS is that it allows both parties to steer an untroubled course through the potentially choppy waters of advice-giving.
Advice-giving about the topic of condom use offers further examples of the AIS. In Ex 11, notice how C formats a 'recommendation' about condoms:
Ex 11 (US2) [B32 A1]
(10.6)
C: we recommend tha'chu use co:ndoms
(1.0)
C: ah:: if you are not su:re about your
partner's: (0.2) status.
(.)
Although C's 'u' (line 1) apparently addresses his advice personally to P, C constructs what she is saying as 'official' advice ('we recommend tha'chu use co:ndoms'), though once more made conditional by an 'if' format. As such, it is hearable as information-about-the-advice-we-give-in-this-clinic'. Hence his 'u' can be heard as an 'indefinite reference' and the sequence can work as an AIS. Now the absence of any uptake from P in the possible turn-transition points on lines 3 and 6 need not have the damaging implications it would have in an unambiguous advice sequence.
The ability of Cs to maintain a multi-turn sequence about condom use without any response from Ps is also shown in Ex 12:
Ex 12 (US2) [28:1]
C: thee o::nly::(hhh) (0.5) protection (0.5) that's available at this point in histree:: (0.7) ah: is the condom. (0.5) a::nd our recommendation is: (0.5) la:tex only^ (0.9) ah:: preferably american made (0.5) condoms which have some qualidy contro:l (1.0) ah:: which: (1.1) the foreign imports: for the most part do not have. there's a few (0.6) imports that are very high qualidy, (.hhh)but it's:- (0.9) they are an exception rather than the rule. (1.4) C: .hhh the second recommendation is that (1.5) the lu-condoms (.) should be lubricaded
Once again, P does not take up the available slots to acknowledge C's 'recommendation - the 0.9 pause after 'only' (line 4); the 1.0 slot after 'contro:l' (line 5) and the 1.4 second pause after 'the rule.' (line 10). Nonetheless, given the ambiguity of the format, C proceeds to offer a 'second recommendation' (line 11).
This is because P can hear C's 'our recommendation is' (line 3) as indicating
simply information about the kind of advice usually given in this clinic,
i.e. as information-delivery about advice rather than as pure advice-giving.
Although the Information-Delivery Format is sustained by response-tokens,
these can be delayed for long periods without endangering the stability
of the format, e.g. where such delays are hearable as implying only that
the information-recipient is waiting for the completion of the information.
Alternatively, Cs sometimes go in search of a response token following an AIS, as below:
Ex 13 (US2) [B27] (continued)
C: .hhh ah::we recommend that ah:: any kind of
condoms that you use for vaginal sex or
anal sex Iunno f- if you have anal sex on her.
.hhh uh::m (0.7) then use something with
nonoxynol nine.
(0.9)
C: okay?=
P: =mmhmm=
C: =NNoxynal nine is a spermacide.
(0.7)
P: righ.=
In Ex 13, the 0.9 pause on line 5 can be heard as marking the completion of what C has to say about what 'we recommend' about condom use. When P does not take up that slot, C seeks some acknowledgment by her 'okay? (line 6). This obtains an immediate 'mmhmm' from P. Now when C makes available another slot after a slight elaboration, P, having been coached, provides a further response token without being prompted ('righ.', line 10).
Note, however, that C's prompt ('okay?') is hearable as only seeking to establish that P is attending to what C is saying. Moreover, P's two response tokens can mark merely such attendance to information about what 'we recommend'. Contra Heritage and Sefi (1992), they need not be heard as implicit resistance to advice precisely because it is not clear that 'advice' is what is being communicated.
In the three 'condom' extracts above, we have seen an AIS set up using the following formats:
we recommend tha'chu use co:ndoms (25)
our recommendation is: (0.5) la:tex only^ (26)
we recommend that ah:: any kind of condoms that you use (27).
All these sequences depend upon an appeal to an impersonal voice ('we' or our'). As Watson (1987) has shown, when such a voice is used by a single individual, that individual can be heard as 'speaking in an organizational capacity' (271). Our suggestion is that such a form of speech can construct any 'recommendations' that follow as simply information about the official line in this organization.
However, the data reveal other, more subtle ways, through which advice can be constructed as advice-as-information. In particular, even advice couched in a personal voice ('I') can be made ambiguous as Ex 14 will show:
Ex 14 (US2) [B35 A3]
C: um condoms come in all varieties and um in this day and age, you know, I tell women that come through here you know you are the one that is being put at risk, and if your partner refuses to wear a condom ah ah you know you're subjecting yourself to to the risk and if necessary, you keep some for the forgetful males, if it- say well I forgot to get any or you know this that or something else, but basically I you know it is for your protection and you have to decide you know if you're going to keep a supply of them. ((More detailed transcription)) (0.4) C: if you're going to keep any^, (1.0) ah: number one:,
Ex 14 apparently reveals unmistakably personal advice delivered by a personalised C ('I') to a P who is characterised (implicitly) as female and as 'you'. However, closer inspection shows the elegance of C's 'I tell women that come through here' (line 2). This creates a clear ambiguity. C can be heard now not to be advising P but to be informing her about the advice he gives to 'women' in general. Once more, the proof that C monitors the talk this way is demonstrated by the lack of turbulence after P's failure to use the first available slot (0.4 pause, line 12) to acknowledge C's 'telling'.
In Ex 14, C's introduction works as an elegant way of dressing up advice-giving as information-delivery through
embedding the advice in an account of a 'telling' (to 'women that come through here'). The implicit appeal to the form of a 'story' is made explicit in Ex 15:
Ex 15 (US2) [B43B2]
C: MY:: (1.8) story on condoms: that I give to people is that .hhh there are:: all ki:nds of them ou:t there (.) a:nd when uh people well I don't like 'em, they don't feel right. I say experiment with them til you fi:nd some that (0.2) you know (0.5) work for you because .hh there's enough vari:edy in thickness sh:- (1.0) ((closes drawer)) C: SHA:pe(hh) si:ze (0.5) what have you that- (0.4) uh there's going to be something that works for you. (0.4) P: okay.
Note how C's 'advice' ('experiment' with condoms, line 4) is embedded in a narrative ('my story', line 1) about what he tells 'people' about condoms, the objections of such 'people'('well I don't like 'em, they don't feel right', lines 3-4) and what 'I say' (line 4) in response.
In Ex 15, the AIS works by constructing P as the recipient of a 'story'. Hence his 'okay' (line 11) serves to mark his receipt of the story rather than implying positive uptake of unambiguous personal advice.
Of course, as in earlier examples of the AIS, there is nothing to prevent Ps offering stronger acknowledgments. For instance, in Ex 16 below, a 'recommendation' about condoms evokes a 'newsworthiness' token ('ah ha'):
Ex 16 (US2) [32B]
C: (.hhh)AH: we do: recommend that you use co:ndoms [in:- P: [ah ha C: ah:: ah:: (1.6) ah::m: (1.0)
The point, however, is that AIS can tolerate a very broad range of responses (from marked acknowledgments, as above, to response tokens, to several turns with no client response). Conversely, unambiguous sequences of personal advice break down very speedily without marked acknowledgments
By constructing advice-sequences that can be heard as information-delivery, counsellors manage to stabilise advice-giving. A function of maintaining an ambiguous communication format is that the counsellor does not have to cope with the difficult interactional problems of the failure of the patient to mark that what she is hearing is personalised advice and hence to offer more than a mere response-token in reply. For, information-delivery can be co-operatively maintained simply by the client offering occasional response-tokens, like 'mm hmm' (see Silverman:1996,Chapter Three).
A second function of offering advice in this way is that it neatly handles many of the issues of delicacy that can arise in discussing sexual behaviour. First, the counsellor can be heard as making reference to what she tells 'anyone' so that this particular client need not feel singled out for attention about her private life. Second, because there is no step-by-step method of questioning, clients are not required to expand on their sexual practices with the kinds of hesitations we saw above. Third, setting up advice sequences that can be heard as information-delivery shields the counsellor from some of the interactional difficulties of appearing to tell strangers what they should be doing in the most intimate aspects of their behaviour.
The U.S. transcripts considered in this chapter reveal multiple forms through which counsellors avoid explicitly personalising their advice. These forms include:
'We would strongly encourage'
>we would also< encourage
We would suggest
what we're- what we recommend
we recommend
(.hhh)we re:commen:d (1.0) use of a cond(h)om (0.2)
our recommendation is:
THE RECMMENdation is:
THE FIRST PREFerence: (.) is
the official recommendation is
I tell women that come through here
MY:: (1.8) story on condoms: that I give to people
Each form has subtle nuances, varying from a mere 'report' on policy ('THE RECMMENdation is:'; 'THE FIRST PREFerence: (.) is'; 'the official recommendation is'), to an account of what the professional usually does ('I tell women that come through here'; 'MY:: (1.8) story on condoms: that I give to people') to an impersonal recommendation in the voice of 'we'.
These forms play with devices such as an 'if' preface (the 'Proposal of the Situation' documented by Kinnell and Maynard:forthcoming), the 'Oblique Reference' (e.g. 'some people' as noted by Peyrot:1987) and upon an appeal to an impersonal voice ('we' or our'), which, as Watson (1987) has shown, can allow professionals to be heard as 'speaking in an organizational capacity' (271).
These devices set up at least three dimensions of ambiguity, as set out in Table 1 below:
____________________________________________________________
TABLE 1:THREE KINDS OF AMBIGUITY
____________________________________________________________
| Nature of ambiguity | Device | Example |
| Source of advice | Impersonal voice | 'We', 'our' |
| Recipient of advice | Oblique reference Proposal of Situation |
'Some people' |
| Activity | Advice-as-information | 'The recommendation is' |
It is beyond the scope of this chapter to address the different implications of each device, although it is worth observing an interesting feature of the 'impersonal voice' seen in the following extract:
Ex 17 (US2) [B27]
C: o::m:: not (???) the (???) I've heard tha that the heavy duty plastic wrap ye know that stuff on the shelves that pink and blue they have like the microwave stuff not the saran wrap but the real heavy plastic. P: Ah huh C: ^o::m:: the official recommendation is a dental dam. .hhh now where you can get a dental dam I don't know. (1.6) C: do you know what a dental dam is?
As C embellishes his account of safer oral sex, he constructs an AIS using an 'impersonal voice' ('^o::m:: the official recommendation is a dental dam.' line 7). Now C continues by appearing to undercut the practicality of the 'advice' he has just given ('.hhh now where you can get a dental dam I don't know.').
The fact that we do not hear a contradiction seems to arise by the way in which the utterance 'the official recommendation is ...' serves to distance a speaker from the very advice which he is giving. This 'report' form thus serves as a neat device which allows professionals to do what is expected of them (i.e. report the 'official' view), while enabling them, if they wish, to depart from it. Like the AIS itself, the 'impersonal voice' functions by allowing an ambiguity in the interpretation of what is being said and hence a potential ambiguity about the meaning to be attached to P's response..
It must be stressed that that we have not been concerned here with logical ambiguities of the kind that normative communication textbooks may criticise. Such ambiguities are usually identified without reference to sequential organization with the aim of 'getting language into shape'. Instead, we have followed Sacks' concern with what he calls 'sequential ambiguity'. As he puts it:
'Now when I talk of 'ambiguity', there's some special attention needed to the way I want to use it here. One tends to think about 'ambiguity' that, e.g. a word could mean this or that, or that it could mean this, that, or God only knows what else. The sort of ambiguity that I'm interested in specifically is 'sequentialized ambiguity', where the issue is what sort of thing should go next, turning on what this thing might have been' (Sacks, 1992, 671).
Following Sacks, we have been examining how clients can inspect what an utterance by a professional might mean in order to establish 'what sort of thing should go next'. In addition, we have shown how professionals might monitor that 'next' in producing a further turn. Unlike logicians or philosophers, these practical actors do not usually treat ambiguity as a 'problem' in need of a 'solution' (see Wittgenstein:1969).
This method suggests that both students of communication and practitioners should avoid treating 'ambiguities' as problems
in need of solutions. Instead, by playing with ambiguities, both practitioners and clients manage difficult interpersonal relations while displaying considerable interactional skills.
It may well be that pre-test counselling is ideally suited to the use of the AIS since 'the advice cannot unambiguously apply to the client until the test result is known' (Heritage:personal communication). Nevertheless, in many professional-client interactions, advice is expected to be delivered, often with little opportunity to elicit a client's perspective. Apt examples may be general practitioners expected to deliver health promotion messages or community lawyers giving legal 'advice'. In such cases, the AIS works as a neat device to shield both professional and client from some of the delicate implications of what they are doing.
Finally, however, it should be noted that the AIS is not a straightforward solution to communication problems. Like any such 'solution', it contains potential dangers or dysfunctions. For just as the ambiguity it serves to create may manage 'delicate' matters, it may create a further, less helpful ambiguity: are clients' minimal uptakes to be heard as acceptance or resistance to the advice they are given?
It seems that, without specific questions, counsellors will not know how their client is responding. But, of course, as Perakyla (1995) shows, in the context of questions about future 'dreaded' situations, obtaining responses to certain questions is not always an easy matter!
Heritage, J and Sefi, S (1992) 'Dilemmas of Advice: Aspects of the Delivery and Reception of Advice in Interactions between Health Visitors and First Time Mothers'. In P.Drew and J.Heritage (eds.), Talk at Work, Cambridge:Cambridge University Press
Jefferson, G & Lee, J (1981) 'The rejection of advice:managing the problematic convergence of a "troubles-telling" and a "service encounter"', Journal of Pragmatics, 5,5, 399-422
Kinnell, A M and Maynard, D W (1996) 'The Delivery and Receipt of Safer Sex Advice in Pre-Test Counseling Sessions for HIV and AIDS', Journal of Contemporary Ethnography,24,4,405-437
Kofmehl, A M (1992) 'Doing a "Proposal of the Situation" in the Delivery and Receipt of Safer Sex Advice', unpublished M.Sc. thesis, Department of Sociology, University of Wisconsin-Madison
Maynard, D W (1991) 'Interaction and asymmetry in clinical discourse', American Journal of Sociology, 97,2,448-495
Maynard, D W (1992) 'On clinicians co-implicating recipients' perspective in the delivery of diagnostic news'. In P.Drew and J.Heritage (eds.), op.cit. pp.331-358
Perakyla, A (1995) AIDS Counselling: institutional interaction and clinical practice, Cambridge, Cambridge University Press
Peyrot, M (1987) 'Circumspection in Psychotherapy:structures and strategies of counselor-client interaction', Semiotica, 65, 249-268
Sacks, H (1992) Lectures on Conversation, edited by Gail Jefferson with an Introduction by Emmanuel Schegloff, Oxford: Blackwell, 2 volumes
Sacks, H, Schegloff, E and Jefferson, G (1974) 'A simplest systematics for the organization of turn-taking in conversation', Language, 50 (4), 696-735
Silverman, D (1997) Discourses of Counselling:HIV counselling as social interaction, London:Sage
Silverman, D, Bor, R, Miller, R and Goldman, E (1992) 'Advice-Giving and Advice-Reception in AIDS Counselling. In P Aggleton, P Davies and G Hart (Eds), AIDS:Rights, Risk and Reason, London:Falmer Press
Watson, R (1987) 'Interdisciplinary considerations in the analysis of pro-terms'. In G Button & J R E Lee (eds)., Talk and Social Organization, Clevedon, Avon, UK:Multilingual Matters Publishers.
Wittgenstein, L (1969) On Certainty, New York:Harper & Row.
[ Home Page ][ Site Map ][ Links ][ Feedback ][ Submit ][
HIV Prevention Services ]
[ Berkeley Free Clinic ][
Confession Box ][ Counselor
Resources ][ Relationships ][ Theorizing Desire ]