Clift Irony in Conversation

Irony in Conversation
Author(s): Rebecca Clift

Source: Language in Society, Vol. 28, No. 4 (Dec., 1999), pp. 523-553
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/4168962
Accessed: 09/02/2010 01:33
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=cup.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Language
in Society.
http://www.jstor.org
Language in Society 28, 523-553. Printedin the United States of America
Irony in conversation
REBECCA
CLIFT
Departmentof Language and Linguistics

Universityof Essex, WivenhoePark
Colchester, Essex C04 3SQ, Great Britain
[email protected]. uk
ABSTRACT
This article proposes the adoption of Goffman's concept of "framing"to

characterizeironyacrossits forms;the suggestionthatthis framingis achieved
by a shift of footing reveals links between verbal irony and other forms of
talk. Examinationof irony in conversation shows how the shift of footing
allows for detachment,enabling the ironist to make evaluations in response
to perceived transgressions with reference to common assumptions. It is
both the constructionof an ironic turn and its placement in a sequence that
make for the discernible shift of footing, and thus the visibility of the frame;
with irony, conversationalexpectations of what constitutes a next turn are
fulfilled on the level of form, but underminedon the level of content. This
analysis shows the extent to which irony is affiliative, and reveals its hitherto unacknowledgedsubtlety of effect and range of attitude. (Irony, pragmatics, conversation, framing, footing)*
Recent years have seen a variety of characterizationsof linguistic irony. Given
the range of perspectives - pragmatic,psychological, and literary- the essential
focus of such studieshas remainedconstant.The overwhelmingconcernhas been
with verbal irony, and the object of investigation has been the ironic sentence,
either in isolation (cf. Wilson & Sperber 1992) or in the context of a constructed
text (Bolloba's 1981, Jorgensenet al., 1984, Gibbs & O'Brien 1991, Dews et al.,
1995, Giora 1995).' I arguehere that this focus has producedtheoretical models
which, put crudely, are at once too narrowto reveal what irony is, and too broad
to illuminatewhat it does. I suggest thatan adequatecharacterizationof linguistic
irony may be best attainedby a considerationof irony across its forms (dramatic/
visual/situational as well as linguistic) and that, similarly, an exploration of its
functions is most clearly revealedby investigationof its basic site: conversation.2
The following is thus a preliminaryattemptto expand the traditionaldomain of
inquiry- first, by providinga top-down characterization,going beyond the usual
focus on specifically linguistic irony; and second, by a bottom-upexaminationof
the process by which EXPLICIT irony emerges, and of the interactionaluses to
which it is put.3 I hope thereby to suggest that a model which reveals the basic
C 1999 CambridgeUniversity Press 0047-4045/99 $9.50
523
REBECCA
CLIFT
characteristicsof irony in its diverse forms will also enable us to examine its most
common - and fleeting - realization.
PRELIMINARY
CHARACTERIZATIONS
The traditionalview of verbal irony, originatingin classical rhetoricand emerging by way of the philosophy of language,4holds that that the ironic utterance
means the opposite of its literal form:
Statedvery crudely,the mechanismby which irony works is thatthe utterance,
if taken literally,is obviously grossly inappropriateto the situation.Since it is
grossly inappropriate,the heareris compelled to reinterpretit in such a way as
to renderit appropriate,and the most naturalway to interpretit is as MEANING
THE OPPOSITE OF ITS LITERAL FORM [emphasis added]. (Searle 1991:536; see
also Bolloba's 1981:327; Brown & Levinson, 1987:226)
This is evidently an attemptto formulatethe sort of divergencebetween a speaker's words, vs. what he might mean by his words,5thatwas perceived to lie at the
heart of irony - a divergence seen most starklyby revealed misunderstandings:6
(1) Video
1 Pete
2
3 Jenny
the boys want a video because apparentlywe're the only

people in Penn without a video,
no? . we haven't got one either.
(-)
5 Pete
6 Jenny
no . definitely we are the O:Nly people in Penn without a video,

hehho:hI s:ee::(h)
(2) Dante (S & A are talking aboutA's abscessed tooth and his imminentvisit to the dentist)
1 Sue
I really don't think you should uh (.) stint on descri(h)bing the pain you
know
2 Andy -+ yes. (1) have you see:n (2) the illustratedD(h)a::nte:
3
(2)
4 Sue
no I haven't.
5 Andy
n(h)o:: I mean6 Sue
O::H::I ISEE:: ri:(h)ght (.) to the d(h)entist
Such examples show thatthe speakersare, in some sense, not sincere in the turns
subsequentlyrevealed as misunderstood;they also make clear thatthe traditional
formulation,with its neat algorithmof negation, is hopelessly inadequatein capturing the precise characterof this insincerity. Several analysts (among them
Kaufer 1981, Sperber & Wilson 1981, Clark & Gerrig 1984, Williams 1984,
Haverkate 1990, Martin 1992, Barbe 1995) have acknowledged such shortcomings. Alternativeshave included Grice's proposal (1975:53) that irony flouts the
conversational Maxim of Quality (cf. Levinson 1983), or that it is a mode of
indirect negation (Giora 1995); or they have attemptedto accommodatethe traditional model within speech act theory (Haverkate 1990) or politeness theory
(Barbe 1995). Yet ultimately such proposalsprovide refinementsof, not alternatives to, the traditionaloppositionalmodel: Samuel Johnson'sdefinition of irony
524
Language in Society 28:4 (1999)
IRONY IN CONVERSATION
as "a mode of speech in which the meaning is contraryto the words" (1755) is
essentially upheld. Three recent proposals which highlight different aspects of
verbal irony offer radical departures.7
IRONY:
ECHOIC
INTERPRETATION,
PRETENSE
OR THEATER?
Sperber & Wilson 1981 provide the first radical alternativeto the oppositional
model in drawingon the traditionallinguistic distinctionbetween the USE andthe
self-referentialMENTION of a word or utterance,8characterizingirony as a form
of echoic mention- a view subsequentlyreconsideredby Wilson & Sperber1992
as a form of echoic interpretation.9By proposing an account rooted in echo,
Sperber& Wilson short-circuitthe traditionalmodel at a stroke; a range of phenomena unaccountedfor under the oppositional model or one of its derivations
can be reanalyzed as echoic. Indeed, reference to the ironies in exx. 1-2 would
seem to bear them out; echoic interpretationis immediately more plausible a
characterizationthan any otherhithertoproposed.The speakerin both echoes an
interpretationof a thoughtor opinion while at the same time dissenting from what
is echoed. Ordinarytalk furnishes some startlinglyprototypicalexamples; in the
following, the echo is particularlyclear.10
(3) Change (B is A's elderly father)
1 Anne
does it thelp if you pu:t your feet fla::t, (.) bend your feet towa- bend
2
your legs towards you a bit.
3
(2)
4 Anne
no bend them towards you- bring your legs up. (.) Is that
5
better?or not,
6
(-)
7 Bill
(Yes thankyou --)
8 Anne
makes a cha:nge=
9 Bill
=Ye::s. (.) It's a cha:nge, (smiles at Anne)
10
(3)
11 Bill
hehe[heh
12 Anne
[hhhehehuh.hh uh (1) huh huh come on Dad take these tablets
On the face of it, Bill's it's a change is a directecho of Anne's assertion;however,

the subsequentlaughter of both parties suggests that it is not a straightforward
agreement.Customarymarkersof agreement- a repetition on a straightfalling
tone, or an upgraded evaluation (Pomerantz 1984:65) - would have served to
endorse the positive implication of Anne's assertion that the change is good. But
instead, Bill echoes on a fall-rise intonation,the conventional sign of non-finality
(Cruttenden1986:102), which suggests doubtregardingthatwhich it asserts and
thereby underminesit. Sperber& Wilson's claim that the ironist simultaneously
echoes and dissociates from that echo certainly appears to find support here.
Even when the origin of the echo is not obviously present, as is overwhelmingly
the case, it may not be difficult to identify:
525
REBECCA
CLIFT
(4) Yugoslavia(S has asked G if he has been to Turkey;this was recordedwhen the civil warin the
formerYugoslavia was just beginning)
1 Gus
I was nea:ronce but I went to Yugoslavia instead. (1) Uh:m,
2
(1)
3 Sarah
Oghmm?.(.) I'd steer clear of that, (.) (as well [now),
4 Gus
[we::ll this was when it was
5 Gus
-+
[(reasonably)peaceful (1) TPEA:ceful:
6 Jo
[yeahhe he he
7 Gus
(1) uh:m socialist people,
hehehehe
8 Sarah
9
10
Gus
11
12
13
14
15
16
17
18
19
Jo
Gus
Sarah
huhuhas they a:re, (1) [hehehe?

[with deep- deep fraternalbonds,
(1)
yeah that's [right
[betwee:n the um (1) [separatebut equal republics,
[no: trouble(----)
indee:d.
(2)
or so we thou:ght
Gus
Sarah
ye:s. (0.2) Well Tthat'swhat they used to say,

ye::s.
-4
-X
Sarah
-4
Gus
Sarah -X
Sarah -*
19
20
21
right (.) uh,

(1)
(-)
The description of peaceful socialist people with deepfraternal bonds between

the separate but equal republics,in sounding like a piece of propaganda,has the
quality of a quote. The laughterof the two women, as well as their agreements
(4:11 and 4:14) in the same ironic vein, display their understandingof Gus's
statement as such - an understandingratified by Gus in 4:20, that's what they
(presumably the people themselves, or at least their self-appointedrepresentatives) used to say. On the second peaceful, Gus is not recognizably speaking "in
his own voice," but in thatof anotheror others;afterthe one-second pause, there
is a clear articulatoryshift on PEA:ceful. Gus's reiterationof the word, but at a
higher pitch and with emphasis on the first syllable, marksa transitionto a more
deliberatearticulation;the shift to the highly emotive abstractidealizationsserves
to distinguish it as somehow quoted, ratherthan originatingfrom him. The parallels with dramaticperformanceare striking. Bauman, in his study of performance in oral literature,talks of occasions in which
the act of speaking is itself framed as a display, objectified, lifted out to a
degree from its contextual surroundings,and opened up to scrutinyby an audience ... Performancethus calls forth special attention to, and heightened
awarenessof, both the act of expression and the performer.(1993:182-83)
WithSperber& Wilson's accountof echoic interpretation,the dramaturgicalcharacteristics of irony remainimplicit. By placing such characteristicsat the center
of their own theories of irony, Clark & Gerrig 1984 and Haiman 1990, 1998
provide the other radical alternativesto the oppositional view. Clark& Gerrig's
proposal- that irony is in fact a form of pretense, with a speaker"pretendingto
526
IRONY
IN CONVERSATION
be an injudicious person speaking to an uninitiated audience" (1984:121) attemptsto addresswhat they regardas the deficiencies of the initial Sperber&
Wilson model of irony as echoic mention. Their principalobjection - that mention is too weak a notion to characterizeirony - appearsto have been addressed
in Sperber& Wilson's reanalysisof their own model as echoic interpretation;but
Clark & Gerrig's notion of pretense is notable for its attemptto shift the focus
from the utteranceto the participants:
Suppose S is speaking to A, the primary addressee, and to A', who may be
present or absent, real or imaginary.In speaking ironically, S is pretendingto
be S' speaking to A'. What S' is saying is, in one way or another,patently
uninformed or injudicious ... A', in ignorance, is intended to miss this pretense, to take S as speaking sincerely. But A, as part of the "innercircle" ... is
intended to see everything - the pretense, S's injudiciousness, A's ignorance,
and hence S's attitudetowards S', A', and what S' said. (122)
Haiman'sproposal regardingirony and sarcasm(made explicit in the title of his
1
1990 article) is equally rooted in dramaturgy:
I wish to propose very seriously that the best metaphorin terms of which to
understandsarcasmandirony is thatof the stage and screen, with its frequently
exploitedcontrastbetween(phony,pretend)"reel"playactingand(God's) "real"
truth. One sarcastic perspective is essentially that of the actor on stage who
steps out of characterand sharesasides with a privileged omniscient audience,
inviting them to deride the other membersof the play, who, unlike the sarcast,
are seen to be playing a role in the limited world of the stage ... (1998:26)
On the face of it, these formulationsappearequally adequateto characterizethe
conversationalironies so far considered, with ex. 4 providing a particularlyrobust example, as one based on echo. Yet what is especially compelling about the
models of both Clark& Gerrigand Haimanis their potential applicationto other
forms of irony in additionto that which is purely linguistic. This is a new departure for studies of irony - and, as will become apparent,a significant one. Possible links between verbal and other forms of irony have been otherwise largely
neglected by linguists;12 witness Sperber'sconfident assertion that
there may exist interesting relations among (different forms), but there is no
reason to expect them to fall undera single unified theory of irony. (1984:130)
So while Sperber& Wilson make no claims for their model of irony beyond the
verbal, it remainsthe case thatinstances of dramaticirony,for instance, are in the
main achieved verbally. Duncan's words in Macbeth on arrivingat Glamis Castle, where his murdererslie in wait,'3 are a classic case of dramaticirony; yet the
echoic model would give a misleading accountof why they areironic, suggesting
that they are somehow an echoic interpretationof "an attributedthought or utterance"(Wilson & Sperber1992:65). In this case it is not the speakerwho is the
527
REBECCA
CLIFT
ironist, but the playwright,who is using the audience's knowledge to renderthe

character'swords ironic; the echoic model fails to make such a distinction."
The pretenseandtheatermodels, with theirinsistence on the assumptionof an
audience as an integralpartof irony,thus appearto capturea characteristicthatis
missing from the echoic account;in so doing, they promise to illuminate ironies
beyond the purely verbal. Clark& Gerrigcite JonathanSwift's Modestproposal
as one instance of a work whose irony is not adequatelyexplained by recourseto
echoic interpretation."5
Their observationsmight also be applied to a ratherdifferentbutequallycelebratedinstanceof literaryirony:thatin JaneAusten'sEmma,
where the irony,unlike Swift's, is not recognizableby referenceto what might be
deemed reasonable,but only by knowing the story's ending. The irony thus becomes evident only on the second reading(the pretensemodel would claim that,
on the first reading,we are the "uninitiatedaudience").Yet Clark& Gerrigclaim
even wider application for their model, suggesting that the notion of pretense
illuminates what they call "theirony of fate," and what might more generally be
termed "ironyof situation."16 Following Fowler (1965:305-6), Clark & Gerrig
suggest thatwhat links linguistic and dramaticirony with that of situationis "the
presence of two audiences- one in on the secret, the other not" (1984:124); they
give as an example, "Ironically,George bought a brandnew Studebakerthe day
before the automobilecompany announcedit was going out of business" (ibid.).
Beyond the intrinsicawkwardnessof theirformula,with its parallelspeakersand
audiences, lies a more substantialobjection on groundsof adequacy.17 In the first
place, any notion of pretense vanishes with Clark& Gerrig'sexamples of situational irony:in the Studebakerexample, the irony lies in the fact that speakerand
audience (in Clark & Gerrig's terminology) are conversing after the announcement by the automobilecompany;they constitute,we must assume, the audience
that is "in on the secret,"the secret in this case presumablybeing hindsight.Yet
hindsightis hardlya "secret"- we eitherhave it or we don't - andpretense,as we
have seen, is a redundantnotion in such a case anyway. Nowhere is this clearer
than in the savage irony of the following newspaperreport (note that it is presumably only the irony that makes it news, a fact suggested by the headline):
Killer cured, then executed.Joliet, Illinois - George DelVecchio, a 47-year-old
child-killer who underwentsurgeryearlier this month to relieve a heart condition, was executed by lethal injectionearly yesterday,prisonauthoritiessaid.
He underwentan angioplastyto clear a blocked arteryafter a heart attacklate
last month. (The Independent,23rd November 1995)
On the evidence of this and the other examples given by Clark & Gerrig, the
notion of pretense as such appearsnot to fit all manifestationsof irony, although
it is easy to see how such a notion finds a place in their formulation.Pretense is
the predictable outcome of their apparentlyreasonable proposal that irony involves a double audience. If a double audience is involved, then it is the next
logical step to assume thatone of these is knowing, andthe otheris not. However,
528
IRONY
IN CONVERSATION
the doubleaudiencecharacterizationis itself undermined,andthe formulastretched

beyond the breakingpoint, when it meets the novel that begins:
You are aboutto begin readingItalo Calvino's new novel, If on a winter'snight
a traveler.Relax. Concentrate.Dispel every otherthought.Let the world around
you fade. Best to close the door; the TV is always on in the next room ....
(Calvino 1982:9)
It ends:
And you say, "Just a moment, I've almost finished If on a winter's night a
traveler by Italo Calvino." (205)
Unlike the previous cited examples, the assumptionof an unknowing audience is
unsustainablehere. To the contrary:the authoris conspiring to make the reader
only too awareof the act in which she is engaged; the irony lies in revelation, not
pretense.Indeed, it is the very revelation of the normalpretense- the suspension
of disbelief - that is ironic.
Looking beyond the literary to other artistic manifestations of irony proves
equally problematicfor the pretensetheory:if, as Clark& Gerrigclaim, pretense
can really explain other forms of irony beyond the verbal, then it should be possible to explain how architecture,such as the Centre Pompidou, or music, as in
the ironies of Bach, can be ironic. Yet the pretense theory is at as much of a loss
as the echoic theorywhen confrontedby such variedinstantiations.To this extent,
Haiman'salternativetake on the dramaturgicalmodel - with his referenceto "the
actor on stage who steps out of characterand shares asides with a privileged
omniscient audience, inviting them to deride the other members of the play"
(1998:26) - seems far nearerto capturingthe knowingness of irony. However,
Haiman's metaphoritself leaves certain issues unresolved, such as the relationship between the actor who steps out of role and the authorof the play in which
the otheractorsarestill engaged, andthatbetween theirrespective statesof knowledge. In addition,its insistence on a tone of derision is surely heavy-handedif it
aims at applicationto irony in general. Surely the tone of JaneAusten is anything
but derisive, and the irony of Calvino's writing is, if anything, celebratory.18
These concerns aside, perhapsthe most urgentproblem relates to the sequential
natureof the metaphor,with the actor stepping out of character.This assumes,
alongside the change in role for the actor,a concomitanttransitionin the audience
from credulity to knowingness, and thus a transition in time. While Haiman's
metaphorof theateris thus the most powerful in its applicationto various forms
of irony, the duality and double perspective of irony remainelusive. It is not that
the actoris steppingout of role, but thathe is somehow playing a role and SIMULTANEOUSLY stepping out of character.
Thus all three theories of echo, pretense,and theaterprovide valuable insights
into the nature of irony - its echoic quality and its apparentorientation to an
audience;but these insights are not ultimatelyadequateto reveal the featuresthat
529
REBECCA
CLIFT
are common to irony across its forms. Any alternativeproposal must profit from
their insights while exploring how an analysis of irony might be illuminatedby
work in other domains.
FRAMING
AND
FOOTING
IN
INTERACTION
In earlierwork (Clift 1995, 1996)1 have suggested thatboth literarycriticism and

sociology have furnishedobservationsstrikinglyreminiscentof those regarding
echo and pretense. Bakhtin'sconcept of "doublevoicing" in literature,with the
associated notions of polyphony and heteroglossia, provides the clearest parallel
with Sperber& Wilson's concept of echo. In his observationson heteroglossia in
the Russian novel, Bakhtin notes:
The speech of the narrators[in Dostoevsky, Gogol, Pushkin and others] is
always ANOTHER'S SPEECH(as regardsthe real or potentialdirect discourse of
the author)and in ANOTHER'S LANGUAGE (i.e. insofar as it is a particularvariant of the literarylanguage that clashes with the language of the narrator).
Thus we have in this case "nondirect speaking" - not IN language but
THROUGH language, through the linguistic medium of another- and consequently througha refractionof authorialintentions. (Bakhtin 1981b:313)
Bakhtin's literary perspective, which predates Sperber & Wilson's account of
echo by half a century,'9finds resonance in Goffman's observationfrom a sociological perspective:
In daily life the individualordinarilyspeaks for himself, speaks, as it were, in
his "own"character.However, when one examines speech, especially the informal variety, this traditionalview proves inadequate ... When a speaker
employs conventionalbracketsto warnus thatwhat he is saying is meantto be
takenin jest, or as mererepeatingof words by someone else, thenit is clear that
he means to stand in a relation of reducedpersonal responsibilityfor what he
is saying. He splits himself off from the contentof the wordsby expressingthat
their speakeris not he himself or not he himself in a serious way. (1974:512)
While neitherBakhtinnorGoffmanarespecificallyconcernedwith irony,Bakhtin's reference (1981a:44) to the adoption of "another'sstyle ... in intonational
quotationmarks"in termsof irony andparody,20as well as Goffman'sdistinction
between the speaker speaking "for himself" and speaking "in a relation of reduced personal responsibility for what he is saying," are implicit in the echoic
account. Irony may, on this view, be seen as one form of what is clearly a more
general strategicuse of language, dependingon the recognitionof a "voice"that
is the speaker'sown - a sincereexpression of his belief - as distinctfrom one that
threatensto underminethat belief. That distinction is made by what Goffman
elsewhere (1974:53) calls "framing,"with certain strategies or "conventional
brackets"used to signal the existence of a distancingframearoundwhat is said.2'
Bakhtinin fact points to the clearest manifestationof framingwith his reference
530
to quotationmarks;the surrenderof authorialresponsibility that they mark is a

boon to the newspaperheadline writer,striving for economy of message:
Inquiryinto "police sex" at Cromwell St
Howard opposed racism law "to protect Rushdie"
Budget "leak"dampens tax-cut hopes (The Independent,25 November 1995, pp.1-2)
The framingof the particularphraseswith quotationmarksallows the writersto

"split [themselves] off from the content of the words," in Goffman's phrase making clear to their readersthat they are not to be held personally responsible
for such opinions. In English SPEECH, of course, no such systematic cues exist.22
For example, the radio reporter,unlike the newspaperjournalist,has to find verbal means of ensuring that he is seen as neutral,as purely the bearerof the message he is carrying,not its originator- particularlywhen that message is deemed
to be sensitive:
(5) Quote, unrecorded(Introductionto storyon UN Secretary-Generalon "Today,"BBC Radio 4,
6.8.92; JH is presenterJohn Humphrys)
JH
The UN Secretary-GeneralBoutros Boutros-Ghalihas in an interview with
the New YorkTimes said that criticism of him in the British press was and I'm here quoting him - "maybebecause I'm a wog"
In his pursuitof the theatermetaphorto characterizeirony andsarcasm,Haiman

identifies quotation as one of the "stage separators"(1998:27) that indicate an
overlaid metamessage.23The quotation above clearly counts as one such instance. However, if we apply Haiman'smetaphorof the stage actor stepping out
of role to the above example, it only provokes the question identified earlier
concerning the relationshipof the playwright and the actor in the play. Haiman
does not make clear what any analysis of the above quotationshould, namely that
there are evidently three perspectives to take into account: the UN Secretary
General's,the radiopresenter's,and thatof (or, to be more precise, that attributed
to) the British press. On Haiman'smodel, the radio presentermight be the actor,
and the words he quotes might be the play from which he distances himself. Yet
the othercrucialelement, the perspectiveof the SecretaryGeneral- which might,
in Haiman's schema, correspond to the playwright - is left out of Haiman's
equation.
However, an alternativeis to invoke Goffman'sdistinction(1979:17) between
(a) the "animator"of an utterance(the person articulatingthe words - here the
radio presenter);(b) its author(the composer of those words - here the Secretary
General);and (c) its principal(who is committedto the propositionthey express
- here, purportedlythe British press). Then it is possible to capturethese distinct
perspectives.The possibility of confusion among animator,author,and principal
here demands that the portion of the utterancecontaining the offensive item is
selected and explicitly framed as a quotation,so that the presenteris not seen to
be subscribingto the view he is reporting.The speaker's adoptionof a particular
531
REBECCA
CLIFT
perspective - of, say, animator,or author - is what Goffman 1979 terms his
"footing"vis-a-vis what he is saying.24The presenter'sstrategyin emphasizing
his footing as one of neutrality,to maintain an impartial stance in the extract
above is recognizableas one of those describedby Clayman 1988, 1992 in studying the achievement of neutralityby television news interviewersin interaction
with their guests. The presenter,denying a possible attributionof principal and
deflecting authorship,disaffiliates himself from the use of the offensive and incendiarywog. Claymandemonstratesthat,althoughneutralitymay be claimed by
speakersby the use of such strategies,it is only maintainedcollaboratively- and
that "correspondingly,the footing throughwhich it is achieved is also a collaborative production" (1992:194). When that collaboration is refused or withdrawn,it disruptsthe trajectoryof the sequence, threateningthe statusof the talk
as an interview:
(6) Interview (On "The Worldat One," BBC Radio 4, 17.1.93; debate on the Calcutt report on
privacy and the media. KC is the Home Secretary,KennethClarke;B is the interviewer,Nick
Clarke, who has just asked him a question)
1 KC
I think that Calcuttand others who are foa:ming on in the way that you:'re
f:oaming on- [
2 NC [I'M only trying to representthe foaming,=
3 KC -*
=W(h)ell in that case you're doing it very adequately.
4
NC
KC
->
Thank you.
I think they're ...
The emotive assertionthatthe interviewerisfoaming on in the same vein as those

on whom he should be reporting objectively, and thus by implication aligning
himself with them, promptsthe interviewer'sinterruptionin 6:2 and a three-turn
insertion sequence. This explicitly addressesthe issue of footing, and therebythe
prioritywith which the interviewertreatsthe maintenanceof the animator/author/
principaldistinction;indeed, it emerges as a preconditionof the continuationof
the interview itself. Such an exchange shows the difficulty of maintaining"the
complex journalisticrequirement... of being interactionally'adversarial'while
remainingofficially neutral"(Clayman 1992:196). On a more general level, the
efforts to which journalistscan go, in orderto deny affiliation with what they are
saying, serves to reveal the default assumptionin talk: that animatorsare indeed
standardlyidentified with principals,unless efforts are made to inform us to the
contrary.
The parallel between the strategies used by journalist and ironist is striking.
Both signal a lack of commitmentto whatthey are saying in the very act of saying
it. By its nature,thejournalist'sfooting as journalistis sustained,while thatof the
speaker as ironist is fleeting. In conversation, it is standardlyassumed that animator, author,and principal are one and the same, except when sanctioned by
activity (as is thejournalistor actor);but the ironist, as we shall see, effects a shift
of footing from committed participantto detached observer.25What makes the
ironist different from the journalist or actor,however, is that the distinction between animator,authoror principalis one thatthe ironistonly lays claim to, while
532
IRONY
IN CONVERSATION
in reality he is all three. While the journalist's neutral footing is designed to

achieve the verbalequivalentof quotationmarks,a speaker'sshift of footing into
irony achieves thatwhich, it will be arguedhere, characterizesall irony:framing.
IRONY
AS
FRAMING
Adopting Goffman'smetaphorof framingto the characterizationof irony makes

it possible to capturethe simultaneouspresence of two dimensions of meaning:
what, for want of more elegant terminology,I shall refer to as "inside"and "outside" meanings, the one framing the other, which lies within it. Traditionalaccounts - adheringto a one-dimensionalmodel that sees irony as the result of the
straightforwardreversal of an utterance - ignore the fact that two aspects of
meaning must be perceived at the same time to make sense as irony. So it is not
thatone dimension cancels the other,but thatit is necessary to make sense of the
other.Ex. 4 makes this particularlyclear.We can see that what the emphasis and
deliberationon the secondpeaceful achieves is a shift of footing, and this is what
renders the statement so reminiscent of a quote; the ironist thereby ostensibly
denies authorshipof what he is saying (attributingit, presumably,to Communist
dogma) and lays claim only to "animator"status. Given that he makes no overt
attemptto disaffiliatefrom whathe is saying (as do thejournalistsin exx. 5- 6, for
example), he implicitly lays claim to the default assumption that he is sincere.
Thus two things are subtly achieved by the footing shift - detectable, as we shall
see, often by a combinationof WHAT iS said and HOW, as well as by expectations
regardingthe next conversationalturn,ratherthan the speaker'sexplicitly drawing attentionto it. By virtue of the shift, the utterancelays claim to an animator/
author/principaldistinction; by virtue of its subtlety, it implicitly suggests that
the animatoris principal.But in the very process of shifting footing, the ironist
frames what is said, thus becoming principalof an outside - framing- meaning.
ECHO,
PRETENSE,
THEATER,
AND
FRAMING
Elements of the echoic, pretense and theatermodels clearly persist in a framing

model of irony: all are characterizedat some level by a double perspective. Yet
the problems raised by these are potentially resolved by appeal to the framing
model. Such a model, for example, avoids the theoreticalfix in which the echoic
model finds itself - namely, that an echo presupposesan origin which, as Wilson
& Sperberseem to concede, may not always be easily identifiable:
The thoughtbeing echoed may not have been expressed in an utterance;it may
not be attributableto any specific person, but merely to a type of person, or
people in general; it may be merely a culturalaspirationor norm. (1992:60)
Of course, as we have seen, the echoic model appears to fit some cases very
adequately;but cases that Wilson & Sperberclaim are echoes are equally well
explained in terms of framing. So, in ex. 3, where Bill's utteranceit's a change is
533
REBECCA
CLIFT
an echo of what Anne has just said, it is not just this that makes it ironic, but the
shift of footing effected by the intonation- which frames the utteranceand thus
altersthe assumptionthatchange is good. The footing shift thus provides us with
an explanationfor the apparentsincerity of Anne and the apparentinsincerity of
Bill.
As with the echoic model, characteristicsof the pretense and theatermodels
are preservedby a framingaccount.The dramaticqualities of irony,so evident in
ex. 4, are presenteven when the ironist clearly implicates himself:
(7) Vital moment
1 Mike
2 Julia
3 Mike
-
4
5
6
7
8
9
Steve
Mike
Steve
Julia
Sarah
Mike
you mentionedgoing (1) to somewherelistening to people talkingaboutfood

(at one time), I rememberwhen I hadn't been in France for very long, (.)
going from Paristo: (.) Strasbourg(1) andhearing,(1) hh- s- listening to two
people b'side me in the train. A:ll the wa:y it must have been about four
hou:rs or mo:re (.) talking about foo::d and (smiling; faster) THIs was a
culture shock to ME::! (.) and they- (.) every- (.) where we went through,I
remembervery much (.) the haricotverts de Soissons this was very important
[Soissons was=
[heh
=one of th- (.) I rememberthat- (I mean) I often thou:ghtI must go back
to ta:ste them but I don't know what it was [that was- THATwas a=
[hah
T
=[l'VAItal
moment of the Ttrip =
[hm::
[[hhuhhuhuh
[[huhhehhh=
=then we got on to something else and it was something el(hh)se ...
Mike's narrativereconstructionof his experience constitutes that experience as

an event. As narrator,he is both a participantin the ongoing talk and a participant
in his own narrative;he switches from presenting his perspective - this was a
culture shock to me - to apparentlyanimatingthe perspective of his travel companions, first by means of the quote, the haricots vertsde Soissons, and second by
his commentthatclearly conveys his interpretationof theirinterestin the subject:
this was very important.He thus presentsus with a double perspective, which is
neatly collapsed in his verdict thatthat was a vital momentof the trip. The frame
that detaches Mike from commitment as principal of the inside meaning is an
evaluation of that meaning. The irony provides the narrativeclimax, both characterizing the foregoing narrativeas an episode and confirming the speaker's
dual perspective on it: as participantin the realms of both inside and outside
meaning, with the principalas the subjective"I"who is the ironist-narratorof the
event ratherthan the objective "me"within the frame who is participantin it.26
The framing model thus goes some way towardendorsing the dramaturgical
characterof the models of Clark & Gerrig and of Haiman. To this extent, the
ironist - in being author,animator,and principalis playwright,performer,and
character.Bauman's comment, quoted earlier, that "the act of speaking is itself
FRAMED as a display" (1993:182; emphasis added) now appearseven more pertinent;the framingof an utteranceachieves the double perspectivethat the pres534
IRONY
IN CONVERSATION
ence of an audienceprovidesin the theater.Dramais essentially a framedmedium,

and in situationsof dramaticirony the playwrightexploits the animator/author/
principaldistinctionto producean evaluationof thatwhich is being animated.As
we have seen, it is ourknowledge thatmakes for dramaticirony;in Macbeth,that
knowledge frames for us Duncan's words, which for him have only inside meaning; the double perspective is that created by the playwright, with the dramatic
situation itself constituting the frame. Indeed, the same is true of any irony in
dramaticform. Browning'sdramaticmonologs,27for example, areilluminatedby
appeal to the author/animator/principaldistinction, where accounts in terms of
echo, pretense, and theatersimply confuse. The double perspective broughtinto
existence by framingis ultimatelywhat lies at the heartof all irony, and provides
the one characteristiccommon to all of its forms.28
FRAMING
BEYOND
THE
VERBAL
Whateverthe shortcomingsof the dramaturgicalmodels of irony,theirallusion to

dramaticform does capturea crucialfeature:the sense of an audience, implicit in
the notion of irony as performance.This, above all, is what characterizesnonverbal irony, and which, through the notion of framing (unlike the notions of
echoic interpretation,pretense, or theater)links it to its verbal realization. In the
process of framing, the ironist draws the observer's attention to that which is
inside the frame, and to its relation to the frame. Thus the situationalirony (and
therefore,in this case, newsworthiness) of the "Killer cured, then executed" report lies in the way the events are framed,even down to the title itself: The events
are brought together so that one is seen to reflect on the other, with the frame
existing in the relationbetween them. To this extent, hindsightconstitutes a powerful frame for situationalirony, because many such ironies are evident only in
retrospect.The irony is broughtinto being only by the presence of an observer,
who in this case is also the ironist; in creatingthe frame by arrangingobjects or
events, he creates a metalevel of communicationthat transformsus all into observers.29This characteristicis what makes artistic ironies, where the frame may
be present in a very real sense, so particularlypotent - the most emblematic,
perhaps, being Magritte's Ceci n'est pas une pipe over the painting of a pipe.
Architecturalironies, so prevalenta feature of Postmodernism,lack the explicit
and visible frame;but the irony exists, like Magritte's,in the juxtapositionof two
contradictoryelements, one serving to evaluate the other. In Postmodernarchitecture the inside meaning has taken the form of our established expectations of
what buildings "should"look like, andthen has underminedthese. Thus the Centre Pompidouinverts a traditionalarchitecturalconceit by placing on the outside
that which is normally hidden, explicitly drawing attention to it; the apparent
fragility of the Louvre's glass pyramids makes ironic play on our prototypical
knowledge of pyramidsas solid. The irony of such architecturelies in such selfconscious juxtapositions,30and the frame exists in the relation between the two
535
REBECCA
CLIFT
elements. The fact thatbuildings, situations,and works of art may be considered

ironic is yet more evidence of the need to re-assess existing models, on the one
hand, and of the case for framing,on the other.
IDENTIFYING
IRONY
Whereas shifts of footing that serve to frame what is being said by the journalist
are relatively explicit, those made by the ironist are less so. In identifying such
shifts, a strikingmusical example may serve as a startingpoint. Auer, discussing
a passage from Bach's St. MatthewPassion, shows how a switch of key into the
"almostprimitivelytransparentC major"(Bach's regularkey for signaling irony)
attributesto the High Priests[thesingers]candidsincerityandchildlike straightforwardness... there is a clash between the expectationsbuilt up so far in the
story, accordingto which the High Priests are sly and malicious, and the particular harmonies underlying their words now, which suggests the opposite.
The conclusion of this inferencingcan only be thatthe High Priests'words are
to be understoodas ironic, i.e. thatthey mean somethingdifferentfrom what is
said. (1992:3)
The shift of key serves, in effect, to shift the footing. The crucial reference here
is to the "expectationsbuilt up so far";in the realm of visual ironies, of course,
those expectations take the form of one element (a pictureof a pipe, the shape of
a pyramid)which is then underminedby another(the text, the building material).
Self-contained verbal ironies - "one-liners"that can stand alone, independentof
interactionalcontext - work in much the same way, evoking well-known phrases
while simultaneously up-ending our expectations of them. Thus Dorothy Parker's reputedcomment on KatherineHepburn,She ran the whole gamut of emotionsfrom A to B, works as irony because it elicits the response associated with
another utterance,She ran the whole gamut of emotionsfrom A to Z, which it
parallelsin structureand(in AmericanEnglish pronunciation)sound- andwhich,
given the expansive associations of the words run and gamut, we might have
expected. In the mismatchbetween this expectationand what is actually said lies
the irony.The polarityestablishedis anothercharacteristicof the ironic utterance:
and its presence, sometimes manifestedas inversion, may be one reason why the
traditionalview of irony as the product of inversion (and the basis for Grice's
assertion that irony is a flout of the Quality maxim) has persisted.
With such self-contained ironies, the shift of footing takes place within the
domain of the utterance.As we have noted, studies of verbal irony have hitherto
largely confined themselves to such examples. But when conversationalirony is
considered, the inappropriatenessof analysis in terms of utteranceslifted from
any conversational context becomes clear. We are then dealing not with constructedor evoked contexts, but with irony emerging out of "expectationsbuilt
up" across across a sequence.3' The form that these expectations can take is
highlighted by a misunderstanding:
536
(8) Selfridges
1 Sarah
2 Jane
3
4 Sarah
5
6
Bob
Sarah
7 Bob
8
9 Jane
10 Bob
... some of the Italian restaurants.(.) like the one near the dentist.
yeah.
(1)
you can have a good big tasty plateful there (.) at lunchtime, (.) I'm not
sure they're stilloperating, (.) when I passed the other day.
m.
there Tis something there but I don't know whether it's the same place,
[(or if it's-)
[that'sthe one you took me to (1) near Selfridges 'isn't it,'
(3)
m[m[well we had actually to walk a:ll the way down the street and back
because you got it wrong (---) (.) [you-
11 Jane
[oh that's right.
Many commentatorshave in fact attemptedto single out markersof irony like

the vocal equivalent of Bach's key shift: conventionalized, invariantand generalizable to all cases. Thus Cutler 1974, Muecke 1978, Roy 1978, and Brown &
Levinson (1987:222) all suggest nasality;Haiman(1998:22) lists "'spitting it out
with a sneer,'nasalization,exaggeratedduration,deadpanmonotone, sing-song,
caricaturedcourtesy, formality, and sympathy ..." However, if self-contained
and constructedironies arethe object of analysis, it would seem naturalto look at
how irony is signaled in the articulationof the utteranceitself (there being nowhere else, analytically,to look). Certainlythe data collected suggest that irony
may be - but is not necessarily- accompaniedby visual cues, such as smiling; but
what the conversational data can show is that the suggested paralinguistic and
articulatorycues arenot always present.The exchange in ex. 8 shows none of the
paralinguisticor articulatoryfeaturestraditionallysuggested as ironic cues. The
irony relies on the mutualknowledge thatthe restaurantwas in fact nowhere near
Selfridges. Ratherthan in any exaggerationof delivery, the irony is discerned in
the structureof the turn:The one-second pause preceding the ironic description
near Selfridges (thus locating it firmly in relationto somewhere else, ratherthan,
say, "lowerWigmore Street")suggests a certain deliberationand precision, and
the confirmation-seeking isn't it is designed to secure a response, effectively
framing and markingoff the irony. The turn itself also interruptsthe prior turn
(8:6), which is concerned with establishing whether it is the same restaurantas
the one visited, ratherthan specifying where it is (Janehas alreadyagreed, in 8:2,
that it is near the dentist). The interruptivereversion to identifying the restaurant's location marks a shift in the focus of the talk, with the lack of response to
the query implicit in the priorturn.
What this misunderstandingmakes particularly clear is how the irony marked itself by a shift of footing - serves to shift the activity at hand: an
attempt to identify a place, in this case, switches into what is effectively a
tease. Similar shifts are visible in the other examples, which serve to raise the
question, inextricably linked to sequencing considerations, of what it is that
irony is doing.
537
REBECCA
IRONY
CLIFT
AS EVALUATION
Perhaps the simplest observation relating to what irony does in the cases here
concerns its outcomes: The response of the addresseeto recognized irony is routinely laughter(as in ex. 3) and/or a continuationof the irony (as in ex. 4); both
serve to accept the footing shift. But it is also evident thatboth types of response
constitute not only an acceptance of the changed footing, but simultaneously
perform an agreement with that which is asserted from the new footing. Such
agreementsare responses to an implicit evaluation that the irony delivers.
Evaluation,as can be seen, is implicit in the framingthat characterizesirony.
The framing serves metaphoricallyto invite the observer/audience to share the
ironist's perspective. Recognizing this implicit invitation depends both on the
design of the turn and on its sequential placement. To take a particularlyclear
example of evaluation, we turnagain to Mike's summaryconclusion to his narrativein ex. 7, thatwas a vital momentof the trip.The ironyhere, in common with
so many, is recognizable because it relies on common understandingsand assumptions and on accepted standardsof behavior to which the speaker makes
appeal. Thus, it is clearly absurd,by what we recognize as normal standards,to
characterizearrivingin a place famed for its haricots verts as a vital momentof
the trip. Evaluationsthus make referenceto such normsand standards,which the
ironic utterancethrows into focus by invoking them - and, often, by apparently
contraveningthem (i.e. on the level of the inside meaning). So it is only by reference to the generallyheld norm- say thatrainis bad and sunshineis good - that
it's a beautifulday is ironic in a context where it is evident that it is pouringwith
rain. Such ironies are markedby theirextremity,and indeed they often make use
of extreme case formulations(Pomerantz 1986) to emphasize the impossibility
of what is being asserted. Thus the irony of the only family in Penn without a
video - and, from a longer sequence, the following:
(9) Meringues (S is talking about trying to buy meringues at the local supermarket)
1 Sarah
... when I went in last ti:me I said to the: lad on the- the lad (.) who said no:,
and I said could you ring the manager plea:se. (.) and say have you got any
meri:ngues and he Tra:ng the manager and said a customer wants to know
if we've got any meri:ngues. (.) or SHE: wants to know if we've got any
meri:ngues and the answer was (.) INO::. (.) mhuh
2
(-)
3 Gus
oh they do::. Hm.
4 Sarah
NOT like, you know could you lsay to the customer .hhh I'm sorry we
haven't got any but we're g(h)etting them s(hh)oo(hh)[::n
5 Jane
[hahaha
6
(2)
-*
7 Gus
(we have) th(haha)at's right th(haha)at's right. and DON't come round
-*
A:Sking for them [(.)EVer again.
8 Jane
[mhuhuhuh
9 Sarah
hehe it's a bit li:ke that actually,
Extremecase formulationssuch as don 'tcome roundaskingfor themEVeragain

are often components of what Torode, in his discussion of humor, has called
538
"impossible descriptions,"whereby an extraordinaryreality is momentarilyacknowledged and shared(1996: 1). These extraordinaryrealities may be mediated
through other personas such as the boys in ex. 1 lobbying for a video, or the
manager in ex. 8; but they may equally be a fantasticalversion of the speaker,as
in ex. 2, where the speaker,the "I,"projects a different "me,"removed in space
and time, with different interactants- or a different version of the same interactants.The following exchange shows the speakerpresentinga differentversion
of himself, but the distinction between the extraordinaryrealities of irony and
humoris evident:
(10) Turningprofessional (B, aged 93, has been talkingaboutplaying tennis with his granddaughter S and son-in-law M)
1 Bill
... keeps you fit though I suppose.
2
(2)
that's true I suppose, mhm.
3 Mike
4
(9)
5 Bill
however I sha:n't (.) turnprofessional now, I don't think I shall,
6
7
8
9
10
11
Susan
-4
Bill
Susan
Bill
Susan
no?
no not rea:lly. (.) though I've got the forms to sign, but
I don't think I shall go?
you'll just advi:se people,
just advi:se- in the- in the wrong direction. Hehe[he
[huhuhuh
Both irony and humor present us with a double perspective that invokes two
incongruousworlds: the possibility of could or should be, glimpsed in the face of
what is. Both set up expectationsthatthey subsequentlyoverturn.Bill's comment
thathe will advise people in the wrongdirectionmarksa shift from the preceding
irony into straightforwardnon-serious talk. The humor here emerges from the
inversion of the standardassumptionsthatadvice is beneficial; the state of affairs
invoked - misleading advice - is self-contradictory,literally an impossible description.To be ironic, Bill would have had to have said something like advise in
the right direction, which is internallyconsistent in belonging to the wider realm
of the impossible world evoked. As it is, advise in the wrong direction stands
outside the boundariesof the imaginaryworld andis in fact nearerthe truthof the
situation; it is not an assertion from which the speaker can in actuality claim
detachment.In contrast, the claimed distinction between animator,author,and
principal is always present in irony, because the impossibility of the world invoked calls the sincerity of the speakerinto question.
Sometimes the extraordinaryrealities are such purely by dint of the fact that
they question what is only too obvious. It is patently the case that a student of
Frenchwill speak French in France, and that one cannot wear twenty-four pullovers at once:
(1 1) French (S is a studentof French who is about to visit France)
1 John you'll be speaking French in Franceprobably,
2
Sue
3 John
IOh No:::
(smiling) I figured that ou:t on my o:wn.
539
REBECCA
CLIFT
(12) Pullovers (from Svartvik & Quirk 1980:312; (J and K are married)
1 Jim
there's twenty-fourwarm pullovers to be knittedbefore Januaryfor a
2
starter
3
(2)
4 Ben you can't wear them all at once
5 Kath
oh no no for differentpeople
In seeking reassuranceregardingsomething ludicrously obvious, the ironist

invokes an alternativeworld where these certainties are not so certain. The alternativeworldproposedby ironyis thuscharacterizedby the expressionof doubt
where thereshouldbe certainty(exx. 11-12), or the expressionof certaintywhere
there should be doubt (ex. 1).32 These framed, impossible worlds, within the
realm of the inside meaning, illustratethe dramaticqualities of irony;they reveal
what Goffman (1979:25) calls "ourgeneralcapacity to embed the fleeting enactment of one role in the more extended performanceof another."These juxtapositions of the extraordinaryandthe actualarewhatinvests ironic evaluationswith
their particularpower: Dorothy Parker'sA to B quip is so devastatingbecause it
invokes expectationsbased on the A to Z ideal, thus makingthe gulf between the
ideal and the actual apparentwithout being explicitly stated.
The footing shifts that make ironic evaluations possible are thus identifiable,
certainly in part from the content of the evaluations themselves, which usually
tends towardthe extreme.Yet it is only in its sequentialcontext thatthis extremity
can be seen as ironic: thus the apparentvehemence of and don't come round
askingfor themever again marksan abruptdisjunctionof the particularsequence
in which it is embedded,on articulatory,grammatical,and sequentialgroundsfirst by emphaticstresses and comparativelyloudervolume, second by the use of
come, and finally by the deflection from the polite proposalof Sarah'spriorturn
(on which, with and, it is built) into the exaggerated rudeness and improbable
hostility of Gus's turn.
Sequentialdisjunction,in which what is projectedby a priorturnis apparently
subverted,is thus a majorfeatureof irony.That this subversionis only apparent
is due to the distinction between an interactional"slot" and the activity accomplished by it; as Sacks observes:
Certainactivities not only have regularplaces in some sequence where they do
get done but may, if their means of being done is not found there,be said ... to
not have occurred,to be absent ... For example, the absence of a greeting may
be noticed ... Observationssuch as these lead to a distinctionbetween a "slot"
and the "items"which fill it, and to proposing that certain activities are accomplished by a combinationof some item and some slot. (1972:341)
In the examples collected, we can see that the ironic turns occur routinely in
positions where evaluation is expectable. One such prime site is story completion, after which there is a defined slot for the recipient's appreciation:
540
One reason for the story recipient's slot upon story completion being a structural place is ... that in it recipients must display appreciationof story completion. Anotheris that, not being affordedover the course of a story occasion
for displaying their understandingof the story, there is an issue, upon story
completion, of story recipientsdisplaying their understandingof the story,and
there is a range of ways of doing so. (Schegloff 1984:44)
Thus both story-teller and recipient must ensure a display of understandingby
recipient.We can see in ex. 9 how this display is sought by the storyteller;Sarah's
emphasis in 9:4 of the point of her story - the lad's rudeness - suggests that she
takes Gus's initial response in 9:3 as insufficient (delayed, apparentlynot addressed to what she has said, and lacking the laughterto respond to Sarah's).Her
second attemptelicits Jane's laughter,along with Gus's laughter,agreement,and
response. Effectively, Gus's and don't come round asking for them ever again
performs the agreementlacking in his previous turn;this itself elicits an affiliative response from both Jane and Sarah.
The irony in ex. 9 provides an evaluative summaryof the preceding talk; it
tells us nothingnew aboutthe egregious behaviorof the lad and the manager, but
serves simply to affirmthe stance of the story-teller.In ex. 7, similarly,the storyteller himself uses irony to summarizewhat he has just said abouthis trip, by his
animatingof his travelingcompanions'view that this [the haricots verts de Soissons] was very important.In both these exchanges, the potentially extraordinary
natureof the incident recountedhas been established with evaluation implicit in
the telling. In the following, the trivializing little Chinese hats has told us all we
need to know aboutthe speaker'sopinion of attemptsto make GerrardStreetlook
Chinese:
(13) Chinese hats
1 Julia
2 Mike
3
4
5
6
7
8
9
Julia
Mike
Julia
Mike
Steve
Mike
Julia
-*
well isn't- aren't the- (.) Sjgns in Ge- in GerrardStreet- GerrardStreettwo s- two or three streets [(.) they're in Chine:se aren't they
[well (.) there may- (.) I- (.) I haven't s- I
haven't seen that but I'm not surprisedcos they've (.) found- yes (.) but
then they've got telephone (.) [telephone booths with little (.) Chinese=
[t(hehe)elephone b(h)oo:ths
=h::at[s on I mean=
[hats on heheh
=they're huhuh=
huhuhuh
i(hehe)t's just so so: Chine:se- heheheh
heheheh
The storyteller'sirony here, as in ex. 7, constitutesthe summaryconclusion to the

discussion, which has alreadymadethe participants'attitudesclear.In both cases,
the ironic content may seem sequentially disjunctive, yet the activity accomplished by the slot it occupies is fully expectable as one of the "regularplaces," in
Sacks's words, for such an activity to get done. What is distinctive about ironic
evaluations is thus that the item which occupies the evaluative slot would seem
541
REBECCA
CLIFT
on the face of it to disruptthe expected sequence, while the slot itself continuesit;
item and slot, in other words, are in apparentconflict. Sacks emphasizes how
such sequencing rules are invoked: "The rules of conversationalsequences are
the first rules to be used" (1992:418).33 It is the starkdisjunctionthatcontributes
to the visibility of the frame and allows for identificationof the irony; in those
cases where irony is misunderstood,the mismatchbetween slot and item may not
be so apparent,particularlyif ironic evaluationis positioned where it is not structurallyprovided for. In exx. 1-2, furthermore,the ironist's attitudehas not been
explicitly set out in advance of the irony itself - as it was in exx. 7, 9, and 13,
where the extraordinarycharacterof what was being discussed was already established before the ironic contribution.Thus Pete's switch of activity - from
reportingwhat the boys want, to ironicizing it - as well as Andrew's shift into an
impossible world in ex. 2, are both missed by their interlocutors.
The visibility of the frame,andthus the recognitionof irony in such cases may
be seen to be not so much dependenton conventionalizedparalinguisticcues, but
ratheron expectations as to what constitutes an appropriatenext turn in a conversational sequence: expectations that appearto be subverted by an apparent
mismatch between the next slot in a sequence and that which fills it. Thus the
common assumptionthat irony is characterizedby linguistic and paralinguistic
markersof exaggerationper se is misleading; such markersare only used when
an ironic turn is explicitly overbuilt. It is, then, this apparentmismatch of item
and slot that serves consistently to performone activity: evaluation.
IRONY
IN
INTERACTION
Determiningexactly what is being evaluated throughirony on each occasion is

less straightforwardthanmight at first be apparent;existing accountshave tended
to assume that the targetof any irony will be straightforwardlyidentifiable.34In
some cases this is so: Those that neatly fit the theory of irony as echoic interpretation also provide us with a source that becomes the target. Thus Bill's it's a
change (an explicit echo of Anne) and Gus's and don't come round asking for
themever again (a fictional attitudeimputedto the shop manager)both find their

targets in the source. Yet other cases suggest a less straightforwardrelation between origin and target. So the target of she ran the whole gamut of emotionsfrom
A to B is KatherineHepburn;but contra Sperber& Wilson, it has no source in

anything the target has said - unless we contrive an explanation by somehow
imputingto Hepburnthe opinion that she ranthe gamutof emotions fromA to Z.
An utterance such as peaceful socialist people with deep fraternal bonds between
the separate but equal republics in ex. 4 has origins, as we have seen, in com-
munist propaganda;but the target seems to encompass both communist ideals

and the people of the formerYugoslavia.The targetof the following exchange,
for example, is less clear than may at first be apparent:
542
(14) Hello (E and J have just been discussing whetheror not, as sisters, they have similar tastes.
Hello is a magazine consisting of tittle-tattleand photographsof the rich and famous)
1 Gus
that's right. (.) and you no doubt share the same passionate fondness for
a great many things.
2
3
4
5
6
7
8
9
10
11
12
(-)
Emma
Gus
Emma
Gus
Jane
hehe:::
u:hm (.) [u::h (.) Erghtu::h .u::h
[(smiling) oh we have so: much in co:mmo::n, hehehe
[uh:m- OH I know ONE thing you're terriblyfond of both of you.
[huhuhuh
(2)
what's that,
Jane
Gus
13 Jane
14 Emma
15 Gus
16
17
18
19
20
Jane
Gus
Jane
Gus
Jane
21
Emma
(1)
u:hm (.) Hello:.
()
huhuhuhu?ye::(huhuhu)s::. tha(haha)t'sright. huhuhuhu?

ye::(heheh)s::
I(2) (smiling) there's- there's a- there's- DO I detect
[a- DO I detect a-=
[(it's a family trait).
=do I detect a COmmonELement=
ye(hehe)s
=the hi::ghbro::wvery refi::n:ed[taste (.) that's right.
[huhuhuh
yep.
The sympathybetween the speakersis evident from Jane and Emma's appreciation (in lines 13-14, 18, 20, 21) of Gus's teasing. Although they seem initially to
be the target- despite Emma'sheavily ironic we have so much in common,which
reveals her complicity in the teasing - Gus's reference to highbrowvery refined
taste itself seems slightly sardonic,and so seems to shift the target.Compare,for
example, "sophisticated,"which would have adequatelypoked fun at such lowbrow taste by playing on the common equationof sophisticationwith something
positive to which most would aspire. In contrast,highbrow and very refined are
not qualities to which one might necessarily aspire,with their overtones of snobbery and elitism. Thus the spin that Gus's utteranceputs on the ironic exchange
seems to undermineits function hitherto, as the range of the target broadens to
encompass those who might aspire to such qualities. In doing so, of course, the
addresseesare let off the ironic hook; if theirtaste is by implication lowbrow and
crude, neitherare the opposites, highbrowand very refined,desirable.The target
here is thus neither stable or easily recognizable; and, contrary to the echoic
account, it is not invariably identifiable with a source - if only because, as we
have seen, a source itself is not as easily determinedas the echoic account suggests.35This is underlinedby ironies thataredirectedagainstoneself, as in ex. 10,
mentionedearlier.The irony there, with its invoking of an impossible world, is a
kind of linguistic distancingmechanismfrom the all too obvious reality;it allows
the speakerto be gently self-deprecating,and therebyto show himself in control
of his currentcircumstances,not a victim of them. Justas the footing shift in news
interviews is promptedby the professionaldemandsof journalisticneutrality,the
543
REBECCA
CLIFT
footing shift to irony allows for a similar detachment,enabling the speaker to

allude to a sensitive andpotentiallysombersubject.This is irony as facilitator,the
curious paradoxbeing thatit invokes the realityin the very distortingof it. Justas
an actor or doctor is sanctionedby role to say (or do) what would not ordinarily
be sayable (or doable), the shift of footing into irony allows the ironisteffectively
to take on the persona of another to say the unsayable. In a recent television
documentary,the friend of an AIDS patient who is seriously ill in hospital is
shown walking up to him and saying:
(15) Apple sauce (unrecorded)
you didn't finish your apple sauce (.) after all that trouble (.) I took about four hours to
make it
Just as in ex. 10, where Sue's continuationof the irony initiatedby Bill in lines 5,
7, and 8 accepts the "impossibleworld"thatBill proposes, the intimacybetween
the participantsis itself revealed by the irony.The apparenthostility deliveredon
the level of "inside meaning"is never actual. Such ironies are touching because
we know that, if the speakerwere less intimatewith the addressee,what was said
might seem cruel and perverse; it thus draws attention to the intimacy of the
relationship.As Irvine, in anothercontext,36 points out, "Insome relationshipsa
speaker needs no lines of retreatat all, for the relationshipitself provides one"
(1993:129).37 This intimacyis what makes it possible for Bill to speak as if he has
real choices, which he obviously has not; the patient is being treatedas a person
who is robustand resilient to complaint,when in reality he is weak and dying. In
its play on participants'perceived identities, irony bears some resemblance to
teasing - which, on Drew's account,
demonstratesthat recipient identities or categories ARE OCCASIONED either in
recipients' own talk prior to being teased, or in the teases themselves. From
among the indefinite numberof identities someone may possess, in the sense
of categories to which they may belong, one or some of those identities are
being occasioned in and throughthe teasing sequences. (1987:249)
This account shows that teases can be sufficiently close to reality to be close to
the bone (Drew 1987:246), despite being playful and humorous;but the polarizing characteristicof irony means thatwhen irony is sympathetic,as in exx. 10 and
15, it is evident that it is untrue.Through enabling us temporarilyto become
someone else, irony thus gives us access to subjects that otherwise might be
deemed too sensitive; throughbecoming another,the ironist paradoxicallysides
with his addressee. Irony is simultaneouslyassertion and denial: a way of mentioning the unmentionable.
In revealing the sympathy that can underlie irony, examples like the above
underminetraditionalaccountswhich assume a uniformlynegative tone;38Grice,
for one, is vehement: "I cannot say something ironically unless what I say is
intended to reflect a hostile or derogatoryjudgement or a feeling such as indig544
nation or contempt" (1978:124). Grice's assertion is supported by Haiman

(1998: 19);39 by Sperber& Wilson, who furthercontend that the ironical attitude
is invariablyof the rejectingor disapprovingkind (1986:239); by Thomas 1986,
who claims thatironyis characterizedby an impoliteforce conveyed by utterancelevel politeness,40and by Barbe, who claims that irony indicates criticism "of an
action or utteranceor general stance of anotherparticipantor participants... or
of circumstancesbeyond control" (1995:10). Yet as we have seen, such assessments are misleading - particularly,as in the previous two examples, when the
subject matteris sensitive, and there is sympathybetween the participants.Even
where an obvious intimacy is missing, it is not necessarily the case that hostility
is in evidence. Indeed, Mike's that was a vital momentof the trip in ex. 7 shows
his wry amusementat the enthusiasmof his fellow passengers;he is not indignant
or contemptuous,pace Grice. The whole episode is told with obvious relish;
Mike delights in his fellow passengers' enjoyment, implicating himself in the
experience, while at the same time he steps back from his own partin it to see the
absurdityof their enthusiasm- and, potentially,his own. Thus it does not necessarilyfollow thatthe frame'sdissociationof speakerandassertionentails a wholly
negative response to what lies within it. Thus the evidence of the examples so far
suggests that Grice's references to hostility and contempt are overstatements.
Levinson seems more moderatein his observationthat "ironiesseem TYPICALLY
used to makecriticisms"(1983:161;emphasisadded).However,since ironymakes
a degree of detachmentpossible, it is unsurprisingthat it should be used overwhelmingly to make negative evaluations;there is, after all, little need to dissociate oneself from positive judgments.
Thus, far from being uniformly hostile and directed to one clear target, it
seems that irony is considerablymore flexible in its range and more subtle in its
outcomes thanhithertorecognized. It is the affiliative qualities of irony thatmake
this possible. Without such affiliation, the hostility of irony becomes evident,
especially when the evaluation offered is likely to be one with which the addressee will not agree. This most commonly occurs, of course, when the addressee is the target;in such cases, even understandingthe irony - for which one
must share enough of the ironist's assumptions in the first place - effectively
makes the addresseecomplicit in the attack.The following exchange shows how
the addresseewithholds the affiliation routinelydisplayed by the addressee after
sympatheticironies:
(16) Yes Mommy (Sacks 1992:421)
Roger:
Ken, face it. You're a poor little rich kid.
Ken:
Yes Mommy. Thank you.
Face the music.
Roger:
Al:
Okay. Now you've got that er outta yer system. Now you're a poor
little rich kid we've told you that.
It is this difference of tone that distinguishes sarcasm as a particularform of

irony:" an irony that does not seek to affiliate, and whose negative evaluations
545
REBECCA
CLIFT
(often directed against the addressee) may thus be perceived as hostile. Indeed,
Grice's observations on irony, although failing to capturethe subtleties of tone
and sympathy of which irony is capable, seem in this light to be considerably
more appropriateto the unidimensionalnegativity of sarcasm.
CONCLUSION
To characterizeirony as framedevaluation is to point to how it is achieved, and

to what it achieves. The main alternativesto the traditionaloppositional model
captureimportantcharacteristicsof irony - the distinctive sense of anothervoice
and its dramaturgicalflavor;but none of these accountsis situatedwithin a framework that explains how irony emerges, or what it can be used to do. As we have
seen, a characterizationof what it is links inextricablyto what it does.
TakingGoffman'smetaphorof framingas the basis of a new characterization
enables us to see how verbalirony is linked to otherforms, andto see this framing
as achieved by a shift of footing also reveals the links between verbal irony and
otherforms of talk. It shows how the shift of footing achieves detachment,allowing us to makeevaluationswhich rangein tone fromhostile to sympathetic.These
evaluationsareoften responsesto perceivedbehavioraltransgressions.Such evaluationsstronglyimplicatea certaincategoryof response,namelyagreementordisagreement;overwhelminglyin thesedata,theresponsehasbeenagreementthrough
laughteror continuedirony. In such cases, irony is a facilitator:The affiliation it
makes possible allows the ironistto enterpotentiallysensitive interactionalterritory.Analysis of conversationalirony reveals it to be considerablymore subtle in
its effects, andgreaterin its rangeof attitudes,thanhas been assumedby previous
studies groundedin self-containedironies or (re)constructedcontexts. The unidimensional hostility once thoughtto be characteristicof irony is, on this account,
seen to be moreappropriatefor a definitionof sarcasm- which, althoughachieved
by the same means as irony,actively works to disaffiliate itself from the object.
As well as revealing hithertounacknowledgedsubtletyin the use of irony,the
analysis of conversationalexchanges also shows contraryto previous assumptions, that it is not necessarily systematic paralinguisticcues that markout conversationalirony;rather,irony emerges from the placementof the turnitself. The
use of conversationaldatain suchcases shows whatexaminationof self-contained
ironies cannot. The irony does not necessarily lodge in the articulationof the
utteranceitself.
As we have seen, irony routinely occurs in positions where evaluations are
expectable;butthe apparentevaluationsthey deliver, on the level of inside meaning, are anythingbut ironic. In comparisonwith what precedes them, these evaluations- in being eithermanifestly all too true,or else absurdly,ludicrouslyfalse
- invoke extraordinary,impossible worlds. Thus conversationalexpectations of
what constitutes a "next"action are subvertedby irony, which emerges from the
546
IRONY
IN CONVERSATION
peculiartension between a conversationalslot and the items that fill it. In effect,
slot and item are at odds.
It is, then, a combinationof the constructionof an ironic turnand its placement
that makes for a discernible shift of footing, and thus the visibility of the frame.
In this respect, of course, how we identify irony is but one aspect of the global
issue of how we come to identify anythingas an instanceof anythingat all. To this
end conversationaldataprovide a useful entrance,giving us a means of exploring
what irony is throughan account of what it can be used to do.
NOTES
* I am most gratefulto David Britain,Anita Pomerantz,Ad Putter,andAndrew Spencerfor helpful
discussions and useful comments on an earlier version of this article. Paul Drew, David Good, and
RachaelHarrisalso readversions of my work, when it formedpartof a Ph.D. thesis on conversational
misunderstandings,and offered invaluable guidance. I am, in addition, deeply indebted to an anonymous referee from Language in Society whose observations on an earlier draft prompted me to
reconsider my presentationof certain issues, particularlywith regardto the work of John Haiman.
Many ensuing improvementsin clarity are owed to this referee. The study would have been impossible withoutthe conversationaldata;the extent of my debt to those who suppliedthem will be evident
to anyone reading this. My thanksto all.
i As this article was going to press, I found one exception: Kotthoff 1998, who considers four
fragmentsof irony in a Germanconversationalcontext. Kotthoff's focus is less on the conversational
uses of irony than on its links with other forms of what she calls "stagedintertextuality"(p. 1; see fn.
23). Barbe 1995 examines examples of irony in conversation,but her examples are recalled, with the
result that articulatorydetails and the wider interactionalcontext are missing. Roy 1978 elicits ironic
utterancesunderexperimentalconditions, so thatthe precedingstretchof talk is available, but the data
collected cannot be considerednaturallyoccurring.Haiman(1998:193) uses a questionnaireto elicit
sarcasmin a range of play-acted scenarios, but once again the context has been predeterminedby the
researcher.
2 I use "conversation"here - ratherthan the generally adopted conversation-analyticterm "talkin-interaction"(which is generally taken to refer to talk in general) - because my data come from
naturallyoccurringconversationaltalk.
3 Hutchby & Drew 1995 examine an IMPLICIT irony in a stretch of talk. Theirs is a subtle and
sophisticatedanalysis of how irony emerges across a conversationalsequence out of thejuxtaposition
of two turns;their study remainsthe only one, to my knowledge, that examines irony in a sequential
context. Theiranalysis is not concernedwith ironyperse, but is an illustrationof "how 'next position'
can be treatedas a systematic locus in which participantsin talk-in-interactionuse essentially local
interpretiveresources to establish and maintaina sharedorientationon salient aspects of social reality" (1995:187).
4 Quintilian(De institutioneoratoria, VIII, vi, trans.H. E. Butler) claims that the ironist intends
to convey "otherthan what he actually says."
5 "Speaker"in all cases designates the ironist, "addressee"the recipient. For the sake of argument
(and incidentally,in keeping with the majorityof the occurrencesin the data),I assume a male speaker
and female addressee.
6 All exchanges cited have been tape-recorded,unless stated, and come from my own data. Each
has a numberand title for ease of reference. Names of participantshave been changed. Significant
turnsare markedby an arrowthus: ->. Data from elsewhere are transcribedas at source; my own data
are transcribedaccording to the following conventions:
indicates point at which current
speaker'stalk is overlappedby another's talk
Jo
Gus
yeah that's [right

[betwee:n
547
REBECCA
CLIFT
distinguishes pairs of overlapped

utterances
Mike =[IV::Ital moment of the trip =

Steve
[hm::
Julia [[hhuhhuhuh
Sarah [[huhhehhh=
indicates continuous contribution Mike [thatwas- THATwas a=
by one speakerover two lines of
Steve [hah
transcription
Mike = [TV::Ital moment
or no gap between one speaker's
contributionand another's
NC
KC
I'M only trying to

representthe foaming,=
=W(h)ell in that case you're
doing it very adequately.
indicates a cut-off; i.e., a speaker Sarah well isn't- aren'tthe- (.) signs in Gebegins to say something and then
in GerrardStreet- GerrardStreetrestarts
underlining
indicates stress
Mike
CAPITALLETTERS indicate segments louder thanrest Mike

of talk
?isn't it,?
(2)
.)
that wasTHATwas
indicates quieter or softer arti- Bob

culation relative to surrounding
talk
near Selfridges 'isn't it,'
indicate prolongationof immediately prior sound
KC
others who are foa:ming on
indicates lapsed time in seconds
Jane
indicates micropause
Gus
what's that,
(1)
u:hm (.) Hello:.
indicates rising intonation
Anne
Is that better?
indicates a continuing intonation Gus

(most commonly a fall-rise)
with deep- deep fraternalbonds,
indicates a falling intonation
Gus
I was near once but I went to Yugoslavia instead.
indicates animatedtone
Mike
THIs was a culture shock to ME::!
indicates segment starting on a

relatively high pitch comparedto
surroundingtalk
Mike
TV::Italmoment of the Itrip
.hhh
indicates an inbreath; without a

dot indicates an outbreath. The
length of the row of 'hh's indicates the length of the breath
Sarah .hhh I'm sorry we haven't got any ...
-- __)
recording unclear. Dashes indi- Bill

cate number of syllables heard,
and parenthesizedwords indicate
possible hearings.
(Yes thank you --)
7 Two of these proposals have engaged directly with their competitors; see the exchange in Journal
of Experimental Psychology: General 13:1 (1984), initiated by Jorgensen, Miller & Sperber. Clark &
Gerrig's criticisms of their model are followed by Sperber's response.
8 The distinction between the USE and MENTION of a word may be seen in the distinction between
the following occurrences of Hannah:
a) There's Hannah.
b) "Hannah" is a palindrome.
In (a), Hannah refers to a person; in (b), it refers to a word. In (a), Hannah is used; in (b), Hannah is
mentioned.
548
9 The later work acknowledges that the earlier definition is too restrictive: Mentions of an attributedthought or utteranceare literal interpretation,while echoic interpretationsmay be literal or
non-literal.
10 A neat literaryexample of an exact ironic echo (as opposed to an interpretationof one) occurs
in Owen's poem Dulce et decorumest, with the epigramat the beginning - dulce et decorum est pro
patria mori- echoed at its end. The resultis thatwhat at the beginning is a bold assertionof patriotism
is transformedby the end, following the catalog of futility, into a statementutterly hollow.
11The title of Haiman'sessay - and indeed his subsequentbook (1998), subtitledSarcasm,alienation and the evolution of language - identifies sarcasmas his primaryconcern.As will become clear,
particularlyin my section on irony in interaction,I identify sarcasmas a specific, non-affiliative form
of irony, fully subject to the usual (in my view, mistaken) assertions regardingirony: that it is necessarily hostile and denigratory.Haiman,in contrast,sees the distinction between irony and sarcasm
as lying in intention: "To be ironic, a speaker need not be aware that his words are 'false' - it is
sufficient thathis interlocutorsor his audience be aware of this ... To be sarcastic,on the other hand,
is to be aware that your words are false" (1990:188). Again, "Irony,unlike sarcasm, may be both
unintentionaland unconscious"(1998:20). It is indisputablethat irony may be unintentional;but my
ultimateproposalof irony as framingmakes clear thatit is only in the seeing (in the case of situational
ironies, with the aid of 20/20 hindsight) that such utterances,situations, etc. are regardedas ironic.
(Thus the fact that my lottery numbers come up the week I forget to buy my ticket is ironic only
because I make the connection between my favored numbers,the fact that I always otherwise buy a
ticket, and the fact thatthis week is the only lapse.) If irony is by definition groundedin hindsight, an
attemptto establish a distinction on the basis of consciousness/intention seems beside the point.
12 A notable exception are Littman & Mey (1991:131), who focus on situational irony RATHER
thanverbalirony on the basis of theirclaim that"anironic statementis an utteranceof a speakerwhich
refers to certain aspects of an ironic situationto make a point."Accordingly, they develop a computational model of irony based on three types of ironic situations, which they call intentional, serendipitous, and competence irony. In privileging situationalover verbal irony, their account does not
addressthe natureof the links between the two.
13 "This castle hath a pleasant seat; the air / Nimbly and sweetly recommends itself / Unto our
gentle senses" (Macbeth, I.vi. 1).
14 The fact that Sperber& Wilson's account of echoic interpretationis a reworkingof their original proposal of irony as echoic mention, which they concede was over-restrictive,in itself seems to
indicate a distancingfrom the problems involved with the notion of echo. As it is, "INTERPRETATION
of an ATTRIBUTED thoughtor utterance"(emphasis added) seems to be edging away from any appeal
to what we normallybelieve to be echo. In this model, there is potentially no limit to what one might
attributeto a speaker.
15 "I have been assuredby a very knowing Americanof my acquaintancein London, that a young
healthy child well nursedis at a year old a most delicious, nourishingand wholesome food, whether
stewed, roasted, baked, or boiled, and I make no doubt that it will equally serve in a fricassee, or a
ragout"(Swift, A modestproposal, 1729).
16 Haiman acknowledges that "situations may be ironic" (1998:20), but he does not elaborate
furtheron how his stage metaphormay be applied to such situations.
17 Sperberpoints out thatthis rendersirony indistinguishablefrom parody.Since my concern here
is conversationalirony,it has not been my aim to focus in depthon the distinctionsof irony vs. parody,
or of irony vs. sarcasm. However, given that the difference between irony and sarcasm will become
clear in the later stages of my analysis, I offer some brief comments on irony and parodyhere. In the
same way that sarcasmwill emerge as a particularform of irony, albeit restrictedby tone, parody is
also a form of irony, but restrictedin this case by form. It is irony groundedin stylistic imitation of
another/others;mimicked exaggeration (often, but not exclusively humorous)is largely its point (a
particularlyclear example of this is ex. 4). But whereas sarcasmis characterizedby hostility, parody
may be celebratory;as Dwight Macdonald writes in the introductionto his masterly anthology of
(written)parodies,"Mostparodiesarewrittenout of admirationratherthancontempt"([19601 1985:13).
The pretense theory does not distinguish between irony and parodybecause it does not stipulate the
basis of the pretense; Sperberis therefore reasonable in his observation that pretense conflates the
two.
18 Time magazine, quoted on the cover of the 1982 Picador edition of If on a winter's night a
traveler..., referredto the novel as "a love letter on the wry but irresistablepleasure of reading."
549
REBECCA
CLIFT
19
Bakhtin'saccount of Pushkin'scharacterizationof Lensky's style in Eugene Onegin is a striking adumbrationof Sperber& Wilson's account. Lodge (1988:125) judges that Bakhtinwas writing
in 1940:
This novelistic image of another's style ... must be taken in INTONATIONAL QUOTATION MARKS
within the system of directauthorialspeech ... thatis, takenas if the image were parodicand ironic
... Lensky's representedpoetic speech is very distantfrom the directword of the authorhimself as
we have postulatedit: Lensky's languagefunctions merely as an OBJECT of representation(almost
as a materialthing); the authorhimself is almost completely outside Lensky's language (it is only
his parodic and ironic accents that penetratethis "languageof another"(Bakhtin 1981a:44).
Note thatBakhtincollapses the distinctionbetween ironyandparodyhere,althoughhis commentsare
consistent with my proposeddistinctionbetween irony and parodyin fn. 17.
20
"Double-voiced discourse is always internallydialogized. Examples of this would be comic,
ironic, or parodicdiscourse, the refractingdiscourse of a narrator,refractingdiscourse in the language
of a characterandfinally the discourseof a whole incorporatedsense. All these discoursesare doublevoiced and internallydialogized" (Bakhtin 1981b:324).
21 Haiman (1998:8) refers to Goffman's notion of frame in its capacity as "'a code or set of
principles for the interpretationof any ongoing activity' (Goffman 1974:10-1I)." Haiman also discusses shifters(indexicals) as an example of linguistic framesin the broadercontext of his mainthesis
on the autonomyof language, and furthermore"the insincerity and inconsequentialityof language"
(op. cit., 7). What he does not do, however, is put framingat the center of a theory of irony,which is
the concern of my proposal.
22 Lyons (1982:110-l l) notes that such distancing may be achieved grammaticallyin Frenchby
the quotative conditional ("conditionnelde citation"):Le premier ministre est malade 'The prime
minister is ill'; Le premier ministreserait malade 'We understandthe prime minister to be ill'.
23 The distinctionbetween Haiman'sproposalof the stage metaphorto characterizeirony and my
eventual one of framingis mirroredin Haiman'sselection of terminology:"I propose to call devices
which demarcate art from life STAGE SEPARATORS (what Goffman 1974 calls FRAMING CUES)"
(1998:27). Kotthoff's reference to irony (1998:1) as one form of "stagedintertextuality,"the prototype of which she regardsas quotation,mightbe seen as an implicitendorsementof Haiman(although
without an explicit mention).
24 Bakhtin, talking of heteroglossia in the novel, once again adumbratesa contemporaryformulation of the same phenomenon:"Acomic playing with languages, a story 'not from the author'(but
from a narrator,posited author or character),characterspeech, characterzones and lastly various
introductoryor framinggenres are the basic forms for incorporatingand organizingheteroglossia in
the novel. All these forms permit languages to be used in ways that are indirect, conditional, distanced"(1981b:323).
25 The move to what Goffman calls a situation of "reducedpersonal responsibility,"markedin
news interviews by shifts of footing, has been been investigatedin a varietyof speech situations.Thus
Isaacs & Clark 1990 discuss what they call "ostensibleinvitations"by referenceto a range of felicity
conditions to be fulfilled; these are identified as one of a class of ostensible speech acts, and are
related to other types of non-serious language use. Labov 1972 and Kochman 1983 focus on the
distinction between real and ritual insults in Black verbal dueling; Labov sees the refuge from responsibility as lying in ritual:"Ritualsare sanctuaries;in ritual we are freed from personal responsibility for the acts we are engaged in" (1972:168). Rituals, in otherwords, constitute"frames"which
separatethe speakerfrom commitmentto his utterance.
26
The "me"that Mike refers to within the frame is what Urban (1989:27) calls "dequotativeI."
27
The duke in "My last duchess," for example, cannot be said to be "echoing"anyone, unless
Wilson and Sperberwould claim that it is the authorwho is doing the echoing.
28
Kierkegaardcapturesthis sense of simultaneityin his assertionthat"theironic figure of speech
is like a riddle and its solution possessed simultaneously"([1841] (1965:265).
29 This of
course is the basis of my argumentagainst Haiman's distinction between irony and
sarcasm on the groundsof intentionality;see fn. 11.
30
Indeed, CharlesJencks's comments on Postmodernismin general capturethe double perspective thatrendersso much of it ironic: "in several importantinstances [Postmodernarchitecture]is ...
doubly coded in the sense that it seeks to speak on two levels at once: to a concerned minority of
550
architects,an elite who recognize the subtle distinctions of a fast-changinglanguage, and the inhabitants, users, or passers-by,who want only to understandand enjoy it" (quoted in Watkin 1986:573).
This of course bears a strikingresemblanceto Clark& Gerrig'scomments on irony, and yet the irony
of such architecturelies - like Calvino's address to the reader- in its knowingness ratherthan its
pretense.
31 Indeed, Hutchby & Drew's analysis of an implicit irony (1995) reveals how the turn-taking
system itself provides a resource for irony (see fn. 3).
32 Grice's account of irony as a flout of the Quality maxim can, in this light, be seen as inadequate
to cover such cases.
33 With self-contained ironies, of course, the same applies, although expectations are set up and
underminedwithin the domain of a single utterance.
34 In their early work on irony, when they claim it is a form of echoic mention, Sperber& Wilson
propose that "an ironical remarkwill have as naturaltargetthe originators,real or imagined, of the
utterancesand opinions being echoed" (1981:3 14). They furtherpredictthatwhen thereis no specific
originatorfor the utteranceor opinion echoed, therewill be no victim. Their subsequentreassessment
of irony as echoic interpretation(Wilson & Sperber1992) does not explicitly addressthe issue of the
target,suggesting thatthe link between irony and targetis not as straightforwardas thatimplied by the
mention theory.
35 Both Hymes 1987 and Brown 1995 generally endorse Sperber & Wilson's model, though it
should be stressed that they refer only to the account of echoic mention, ratherthan that of echoic
interpretation;however, they show that the echoic account is problematic in this respect. Hymes
(1987:317), applying the echoic model to ClackamasChinook, shows how a particularroutine fails
to fit the claim that the originatorof something quoted/mentioned is its target;Brown (1995:161),
applying the same model to Tzeltal, shows that the target is not necessarily clear.
36 Irvine's observations are made with reference to verbal abuse among the Wolof.
37 Brown & Levinson's model of politeness classifies irony as an off-record strategy (1987:69);
this allows the speakerto assert sincerely that the off-record interpretationis the one intended, if the
literal one causes offense. It thus detractsfrom the seriousness of the face threatwithout detracting
from the seriousness of the subject.
38 This again is possibly a function of the type of data used for many analyses, namely selfcontained ironies. It is not implausible that the self-contained ironies examined are chosen for being
memorableand witty, and thereforeare more likely to be savage and wounding.
39 Haimanstates that"thehumorin sarcasm(as in irony) lies in the contrastbetween the speaker's
flatteringor sympatheticwords ... andhis or her hostile intentions"(1998:21). Yet ex. 15 shows irony
working in exactly the opposite way to what Haiman claims; the inside meaning is an expression of
hostility, and the outside is one of obvious sympathy.
40 The misleading contentionthat irony is invariablyhostile and disapprovingseems widespread
and may stem from an identificationof distance with hostility. The fact that the speakeroften implicates himself in the irony is an observationonly made possible by looking at the ironic utterancein its
conversationalcontext; without hearing the stretchof talk preceding that was a vital momentof the
trip, and without knowing the speaker/addresseerelationshipwithin which after all that trouble - I
took aboutfour hours to make it is said, we have no way of knowing the degree of hostility involved.
41 Sarcasmand irony are often used interchangeably(cf. Muecke 1970, Sperber& Wilson 1981,
Clark& Gerrig 1984). Haimandistinguishes irony from sarcasmon the basis of intentionality(see fn.
11).
REFERENCES
Atkinson, J. Maxwell, & Heritage, John (1984), eds. Structuresof social action: Studies in conversation analysis. Cambridge& New York:CambridgeUniversity Press.
Auer,Peter (1992). Introduction:JohnGumperz'approachto contextualization.In PeterAuer &Aldo
Di Luzio (eds.), The contextualizationof language, 1-37. Amsterdam:Benjamins.
Bakhtin,Mikhail M. (1981a). From the prehistoryof novelistic discourse. The dialogic imagination,
41-83. Austin: University of Texas Press.
(1981b). Discourse in the novel. The dialogic imagination, 259-422. Austin: University of
Texas Press.
551
REBECCA
CLIFT
Barbe, Katharina(1995). Irony in context. Amsterdam:Benjamins.

Bauman,Richard(1993). Disclaimersof performance.In JaneHill & JudithIrvine(eds.), 1993:182-96.
Bollobds, Eniko (1981). Who's afraid of irony? An analysis of uncooperativebehavior in Edward
Albee's Who'safraid of VirginiaWoolf?Journal of Pragmatics 5:323-34.
Brown, Penelope (1995). Politeness strategies and the attributionof intentions:The case of Tzeltal
irony.In EstherGoody (ed.), Social intelligence and interaction, 153-74. Cambridge& New York:
CambridgeUniversity Press.
_____
& Levinson, Stephen C. (1987). Politeness: Some universals in language usage. Cambridge
& New York:CambridgeUniversity Press.
Calvino, Italo (1982). If on a winter's night a traveller. London:Picador.
Clark, HerbertH., & Gerrig, RichardJ. (1984). On the pretense theory of irony. Journal of Experimental Psychology: General 113:121-26.
Clayman, Steven E. (1988). Displaying neutrality in television news interviews. Social Problems
35:472-92.
(1992). Footing in the achievement of neutrality:The case of news-interview discourse. In
Paul Drew & John Heritage (eds.), Talk at work: Interaction in institutional settings, 163-98.
Cambridge& New York:CambridgeUniversity Press.
Clift, Rebecca (1995). Misunderstandingsin conversation. Ph.D. dissertation, University of Cambridge, U.K.
(1996). Irony in conversation. (Essex working papers in linguistics, 8.) Colchester, UK:
University of Essex.
Cruttenden,Alan (1986). Intonation.Cambridge& New York:CambridgeUniversity Press.
Cutler,Anne (1974). On saying what you mean without meaning what you say. Chicago Linguistic
Society 10:117-27.
Dews, Shelly; Kaplan,Joan;& Winner,Ellen (1995). Why not say it directly?The social functions of
irony. Discourse Processes 19:347-67.
Drew, Paul (1987). Po-faced recipients of teases. Linguistics 25:219-53.
Gibbs, RaymondW., Jr., & O'Brien, Jennifer(1991). Psychological aspects of irony understanding.
Journal of Pragmatics 16:523-30.
Giora, Rachel (1995). On irony and negation. Discourse Processes 19:239-64.
Goffman, Erving (1974). Frame analysis. New York:Harper& Row.
(1979). Footing. Semiotica 25:1-29.
Grice, H. P. (1975). Logic and conversation. In Peter Cole & Jerry L. Morgan (eds.), Syntax and
semantics 3: Speech Acts, 41-58. New York:Academic Press.
(1978). Furthernotes on logic and conversation.In PeterCole (ed.), Syntaxand semantics 9:
Pragmatics, 113-28. New York:Academic Press.
Haiman, John (1990). Sarcasmas theater.CognitiveLinguistics 1:181-205.
(1998). Talkis cheap: Sarcasm, alienation, and the evolution of language. Oxford & New
York:Oxford University Press.
Haverkate,Henk (1990). A speech act analysis of irony.Journal of Pragmatics 14:77-109.
Hill, Jane H., & Irvine, JudithT. (1993), eds. Responsibilityand evidence in oral discourse. Cambridge & New York:CambridgeUniversity Press.
Hymes, Dell (1987). A theory of verbal irony and a Chinookan patternof verbal exchange. In Jef
Verschueren & Marcella Bertucelli-Papi, The pragmatic perspective, 293-37. Amsterdam:
Benjamins.
Hutchby,Ian, & Drew, Paul (1995). Conversationanalysis. In Jef Verschuerenet al. (eds.), Handbook
of pragmatics, 182-89. Amsterdam:Benjamins.
Irvine, JudithT. (1993). Insult and responsibility:Verbalabuse in a Wolof village. In Hill & Irvine
(eds.) 1993:105-34.
Isaacs, Ellen A., & Clark,HerbertH. (1990). Ostensibleinvitations.Languagein Society 19:493-509.
Jorgensen, John; Miller, George A.; & Sperber,Dan (1984). Test of the mention theory of irony.
Journal of ExperimentalPsychology: General 113:112-20.
Kaufer,David S. (1981). Understandingironic communication.Journal of Pragmatics 5:495-5 10.
Kierkegaard,S0ren (1966). Theconcept of irony. translatedby Lee M. Capel. Princeton,NJ: Princeton University Press. [Originallypublished in 1841.]
Kochman,Thomas (1983). The boundarybetween play and non-play in Black verbal dueling. Language in Society 12:329-37.
552
Kotthoff, Helga (1998). Irony, quotation, and other forms of staged intertextuality.(InLiSt: Interaction and linguistic structures,5.) Konstanz, Germany:University of Konstanz.
Labov, William (1972). Rules for ritual insults. In Sudnow (ed.), 120-69.
Levinson, Stephen C. (1983). Pragmatics. Cambridge.
Littman,David C., & Mey, Jacob L. (1991). The natureof irony: Towarda computationalmodel of
irony. Journal of Pragmatics 15:131-51.
Lodge, David (1981), ed. Modern criticism and theory. London: Longman.
Lyons, John (1977). Semantics, vol. 1. Cambridge& New York:CambridgeUniversity Press.
Macdonald,Dwight (1960), ed. Parodies: An anthologyfrom Chaucerto Beerbohm- and after. New
York:Random House. [Reprinted,New York:Da Capo, 1985.]
Martin, Robert (1992). Irony and universe of belief. Lingua 87:77-90.
Muecke, Douglas C. (1978). Irony markers.Poetics 7:363-75.
Pomerantz,Anita (1984). Agreeing and disagreeing with assessments: Some features of preferred/
dispreferredturn shapes. In Atkinson & Heritage (eds.), 1984:57-101.
(1986). Extremecase formulations:A way of legitimizing claims. Human Studies 9:219-30.
Roy, Alice M. (1978). The function of irony in discourse. Text 1:407-23.
Sacks, Harvey (1972). An initial investigation of the usability of conversationaldata for doing sociology. In Sudnow (ed.), 1972:31-74.
(1992). Lectureson conversation, volume 1. Ed. by Gail Jefferson. Oxford: Blackwell.
Schegloff, EmanuelA. (1984). On questions and ambiguitiesin conversation.In Atkinson & Heritage
(eds.), 1984:28-52.
Searle, John R. (1991). Metaphor.In Steven Davis (ed.), Pragmatics: A reader, 519-39. Oxford &
New York:Oxford University Press.
Sperber,Dan (1984). Verbalirony:Pretenseor echoic mention?Journal of ExperimentalPsychology:
General 113:130-36.
______
& Wilson, Deirdre (1981). Irony and the use-mention distinction. In Peter. Cole (ed.), Radical pragmatics, 295-318. New York:Academic Press.
(1986). Relevance: Communicationand cognition. Oxford: Blackwell.
Sudnow, David (1972), ed. Studies in social interaction. New York:Academic Press.
Svartvik,Jan, & Quirk,Randolph(1980). A corpus of English conversation. Lund, Sweden: Gleerup.
Thomas, Jenny (1986). The dynamics of discourse: A pragmatic analysis of confrontational interaction. Ph.D. dissertation,University of Lancaster,England.
Torode,Brian (1996). Humouras impossible description:Humourand horrorin calls to a consumer
helpline. Paperdelivered at 5th InternationalPragmaticsConference, Mexico City, 4-9 July.
Urban,Greg (1989). The "I"of discourse. In BenjaminLee & Greg Urban(eds.), Semiotics, self, and
society, 27-52. Berlin: Mouton.
Watkin,David (1996). A history of western architecture.2nd edn. London: Laurence King.
Williams, Joanna P. (1984). Does mention (or pretense) exhaust the concept of irony? Journal of
ExperimentalPsychology: General 113:127-29.
Wilson, Deirdre, & Sperber,Dan (1992). On verbal irony. Lingua 87:53-76.
553

Clift Irony in Conversation

Uploaded by

Copyright:

Available Formats

Clift Irony in Conversation

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Clift Irony in Conversation

Uploaded by

Copyright:

Available Formats

Irony in Conversation

Author(s): Rebecca Clift

Language in Society 28, 523-553. Printedin the United States of America

Departmentof Language and Linguistics

This article proposes the adoption of Goffman's concept of "framing"to

the boys want a video because apparentlywe're the only

no . definitely we are the O:Nly people in Penn without a video,

Language in Society 28:4 (1999)

On the face of it, Bill's it's a change is a directecho of Anne's assertion;however,

huhuhas they a:re, (1) [hehehe?

ye:s. (0.2) Well Tthat'swhat they used to say,

right (.) uh,

The description of peaceful socialist people with deepfraternal bonds between

Language in Society 28:4 (1999)

ironist, but the playwright,who is using the audience's knowledge to renderthe

Language in Society 28:4 (1999)

the doubleaudiencecharacterizationis itself undermined,andthe formulastretched

In earlierwork (Clift 1995, 1996)1 have suggested thatboth literarycriticism and

Language in Society 28:4 (1999)

to quotationmarks;the surrenderof authorialresponsibility that they mark is a

The framingof the particularphraseswith quotationmarksallows the writersto

In his pursuitof the theatermetaphorto characterizeirony andsarcasm,Haiman

I think they're ...

The emotive assertionthatthe interviewerisfoaming on in the same vein as those

Language in Society 28:4 (1999)

in reality he is all three. While the journalist's neutral footing is designed to

Adopting Goffman'smetaphorof framingto the characterizationof irony makes

Elements of the echoic, pretense and theatermodels clearly persist in a framing

you mentionedgoing (1) to somewherelistening to people talkingaboutfood

Mike's narrativereconstructionof his experience constitutes that experience as

Language in Society 28:4 (1999)

ence of an audienceprovidesin the theater.Dramais essentially a framedmedium,

Whateverthe shortcomingsof the dramaturgicalmodels of irony,theirallusion to

elements. The fact thatbuildings, situations,and works of art may be considered

Language in Society 28:4 (1999)

[oh that's right.

Many commentatorshave in fact attemptedto single out markersof irony like

Extremecase formulationssuch as don 'tcome roundaskingfor themEVeragain

Language in Society 28:4 (1999)

(smiling) I figured that ou:t on my o:wn.

Language in Society 28:4 (1999)

In seeking reassuranceregardingsomething ludicrously obvious, the ironist

Language in Society 28:4 (1999)

The storyteller'sirony here, as in ex. 7, constitutesthe summaryconclusion to the

Determiningexactly what is being evaluated throughirony on each occasion is

themever again (a fictional attitudeimputedto the shop manager)both find their

A to B is KatherineHepburn;but contra Sperber& Wilson, it has no source in

munist propaganda;but the target seems to encompass both communist ideals

Language in Society 28:4 (1999)

huhuhuhu?ye::(huhuhu)s::. tha(haha)t'sright. huhuhuhu?

footing shift to irony allows for a similar detachment,enabling the speaker to

Language in Society 28:4 (1999)

nation or contempt" (1978:124). Grice's assertion is supported by Haiman

It is this difference of tone that distinguishes sarcasm as a particularform of

To characterizeirony as framedevaluation is to point to how it is achieved, and

Language in Society 28:4 (1999)

Language in Society 28:4 (1999)

yeah that's [right

distinguishes pairs of overlapped

Mike =[IV::Ital moment of the trip =