1.Introduction
This chapter takes as its principal impetus a widespread and
somewhat controversial belief in certain parts of world that speakers from
North America “don’t do” irony. Whether this belief is simply an “urban
myth” or a specious linguistic stereotype is one of the questions that
inform the later sections of this chapter. There is however compelling
evidence that the issue has some contemporary everyday relevance insofar as
it manifests frequently in discussions and debates in the public sphere.
Even a rudimentary search of the internet, keying in the quotation in the
title of this chapter, will yield thousands of relevant results. This
internet evidence intimates a number of things. First, it suggests that
interest in this topic is extensively and almost exclusively non-academic in
origin. Second, the sources supporting the evidence are skewed
proportionally towards Britain and Ireland, suggesting that this is where
the debate is principally situated. For instance, the top ten placed
results, in an internet search conducted in English by this author on
November 9th 2015, all derive from sources from the eastern side of the
Atlantic, results which are materialised in “chat”, posts and discussion
pieces dating from 2004 up to January of 2015. Moreover, this irony “debate”
is recorded not just in informal pieces but in more eminent outlets such as
the UK’s Guardian and Independent
newspapers, and alongside these “quality” print media articles was, in fifth
position in the search, a posting on the topic from the BBC’s online news
magazine. These sources suggest that the issue prompts a level of conscious
discussion in Britain and Ireland which is not mirrored in Canada and the
USA. Whatever its terms and wherever its source, though, this is a debate
about perceptions of language practices in general, and, more specifically,
about how different cultural and geographical contexts intersect with
certain kinds of linguistic-pragmatic inferencing strategies. Indeed, there
is good precedent for initiating a linguistic analysis on the basis of “lay”
perceptions or everyday beliefs about language practices (Bauer and Trudgill 1998). Therefore, and without
attempting play to, or fuel, a populist agenda, this article seeks to build
a systematic, empirically-grounded investigation of the cross-cultural
production and reception of irony.
It is of course a commonplace in much linguistic-pragmatic
research to position irony at the centre of many forms of social
interaction, including humour, sarcasm, banter, teasing, politeness, satire
and parody. For instance, Nash offers an overview of the stylistic
intersections between humour, sarcasm and irony, noting the special
“counter-coding” strategies used in the generation the last of these forms
of discourse (Nash 1985: 152).
Cognisant of this broader research backdrop, this chapter narrows the focus
by exploring the way in which ironic situations are processed and
interpreted in different cultural contexts. It is also argued that more
emphasis should be given in pragmatic accounts of irony to the role of
individual language users in both the generation and the classification of
ironic utterances and situations. In pursuit of these related aims, an
online experiment is generated which comprises different narrative
scenarios, to which reactions are gathered from over 300 anonymous
informants world-wide. The scenarios used in the experiment are relayed
through six short 44-word stories, some of which lay good claim to being
considered “ironic”, and others less so. The data gathered from the
informants’ responses offers fascinating insights into structured linguistic
and cultural practices. It also helps challenge and re-cast the central
assumption behind the analysis; that is, the perception of many in the UK
and Ireland that people from North America “don’t do” irony. The study
therefore places ordinary users of language centre stage, allowing them, not
academic researchers, to decide on the ironic status of different kinds of
discourse. The results from the on-line questionnaire, presented later,
reveal not only broad patterns of response across a genuinely international
group of language users, but the data itself throws up many surprises and at
times raises many more questions than it can realistically answer.
The bulk of this chapter is devoted to elucidating a method
for linguistic-pragmatic analysis, to gathering empirical data, and to
offering an extensive discussion and interpretation of the results obtained.
This is therefore not the place to offer an exhaustive account of the
substantial and expanding body of research in linguistic pragmatics on the
discourse of irony; nor is it feasible to present a detailed critical
overview of such work. Suffice it to say, many scholars have advanced their
own models for ironic communication drawing on a host of theoretical
perspectives. As noted, Nash accounts for irony in terms of “counter-coding”
where an utterance is juxtaposed against acknowledged facts, accepted
attitudes and what he terms “broad truth conditions” (1985: 153). In their well-known study of the
trope, Sperber and Wilson postulate an echoic model of
irony which is built on the logical distinction between “use” and “mention”
(1981; and see Wilson and Sperber 1992). From the
extensive range of other current frameworks of analysis, Attardo conceives
irony as “relevant inappropriateness” (2000), while Partington uses a corpus-based approach to talk of
irony in terms of “reversal of evaluation” (2007). By contrast, Clark and Gerrig (1984) view irony as a form of
“pretence”, Utsumi (2000) as
“implicit display”, and Giora
(1995) as “indirect negation”. Additional frameworks include
Barbe’s discussion (1993) of the
implicit and explicit aspects of irony, noting how the former type has to be
discovered by “an initiated audience”, while Kapogianni (2011) explores the connection between
ironic effects and “surreal” elements in message composition, where speakers
generate irony by producing strikingly unrealistic or unexpected utterances.
Separately, Kapogianni (2016)
illustrates the ways in which irony functions at the semantics/pragmatics
interface by postulating three components of the ironic operation: the
vehicle, the input, and the output.
Where each of the studies check-listed above present
individually nuanced models of analysis, there are as many other linguistic
studies that investigate irony more broadly across different genres of
discourse. While Gibbs and Colston
(2007) import insights from cognitive linguistics throughout
their collection of essays, Dynel
(2014) offers a neo-Gricean perspective on occurrences of irony
in contemporary television shows. Benwell
(2004) examines the use of irony as a defence against sexism in
men’s lifestyle magazines whereas Shen
(2009) offers a stylistic investigation of “context-determined
irony” in prose fiction. Other publications have explored irony in forensic
contexts, such as Simpson et al.’s discussion of legal judgements and
rulings which reference, or are prompted by, assessments of irony (2019: 27–30; see also Kelsey and
Bennett 2014). While this overview is no more than a gloss of the rich
variety of linguistic research on irony, there is one particular study that
has helped inform directly the design of the present analysis. This is
Shelley’s conceptualisation of irony as “bicoherence” (2001) which builds an elegant model of
situational irony out of a corpus of media reports about episodes and events
that participants have described as “ironic”. Although research on nonverbal
irony is often separated off from the dominant focus on verbal irony,
Shelley makes a good case for seeing the same organising principles at work
in both ironic situations and ironic utterances. Recourse will be made to
this and other relevant aspects of Shelley’s study in the section that
follows.
The present analysis does not seek to dismiss or reject per se
any of the existing approaches to irony touched on above, although by the
same token, it does not align itself exclusively with any one of the
prevailing schools of thought in linguistic pragmatics. Instead, this study
attempts eclectically to collate different models of analysis, importing
those aspects of a particular model that offer useful theoretical or
explanatory currency. In this context, Simpson (2011) has proposed an umbrella definition for irony,
and although the theoretical justification for these constructs is offered
elsewhere (2011: 39–42), the
definition yields a reasonably serviceable framework that will inform the
experimental design over the remainder of this chapter. The umbrella
definition of irony, along with two associated sub-definitions, runs
thus:
Irony – umbrella definition:
The perception of a conceptual paradox, planned or
un-planned, between two dimensions of the same discursive event.
- A perceived conceptual space between what is asserted and what is meant.
- A perceived mismatch between aspects of encyclopaedic knowledge and situational context (with respect to a particular discursive event).
The idea of a paradox is preferred in the
definitions, rather than the more familiar expressions like “oppositeness”
or “incongruity”, in some measure to help accommodate many of the approaches
to irony touched on earlier. The significance of the
perception of irony is central to the definitions:
irony cannot work without some perception of it, and while much irony
undoubtedly passes us by in everyday interaction, we also perceive irony
that was not planned or intended (see for example Gibbs et al. 1995). Where appropriate, other
aspects of the umbrella definition will be fleshed out over the remainder of
the chapter, but the focus will now shift to the design and methods of the
online experiment itself.
2.Methodology: “Six Little Stories in Need of a Response”
The experiment was designed to elicit reactions to various
scenarios portrayed in six very short narratives. The sources and
foundations for the six stories will be amplified shortly, but the general
structure of the experiment was to capture anonymous informant responses
through an online questionnaire. The only personal details sought in the
questionnaire concerned informants’ age, gender, and, of course,
nationality. Questions soliciting information about educational attainment
were deemed inappropriate on the grounds that informants might feel that the
experiment was designed to test either linguistic proficiency in English or
general intellectual and cognitive skills, or both. Finally, the
questionnaire offered an option for participants, if they wished, to record
their email addresses in order to receive the results of the experiment. On
the understanding that a copy of the final paper in portable data format
would be passed on electronically, 223 of the respondents documented their
contact details in this manner. Interestingly, less than one third of the
emails recorded an academic institutional affiliation – testimony, perhaps,
that the experiment had had the desired reach beyond the academe.
Informants were required to react to the six scenarios
presented sequentially in the questionnaire in such a way that progress to
subsequent stories was only made possible through the completion of the
previous stage. For each story, informants had two tasks. The first, an
obligatory phase, required them to offer in the box provided a
single word that best described their feeling about, or
reaction to, the story they had just read. The second, optional phase
allowed informants to record, if they wished, any other reactions or
feelings they had about the story, and a larger box was provided to capture
this more discursive type of commentary. The questionnaire was created by
and hosted on the online survey tool Questback. A link and invitation to
complete the questionnaire was circulated using social media platforms such
as Twitter and Facebook, and the experiment was disseminated through the
library at Queen’s University Belfast. A Word version of the first page of
the survey is reproduced as Appendix 1.
As argued earlier, the core of the experiment hinges on
stories whose outcomes have not been deemed “ironic” by the academic
researcher, but rather, have had the status of irony conferred upon them by
participants in the discursive event itself. In contrast to the analysis of
invented or contrived data, the (potential) presence of irony in the
relevant stories used in this study is attested by commentary in the public
domain through which these narratives were played out. In this respect,
three of the six stories are “ironic” in the sense that they have been
reported as such by participants in the original context of interaction. Of
the remaining stories, two “non-ironic” stories are taken from research
literature in experimental psychology (more on which below) while the sixth
story has a rather more contentious status insofar as differing commentaries
exist about whether or not it could claim to have some ironic status (again,
more on which below). Of course, at no point in the online instruction brief
was the term “irony” used; informants could offer any single word in the
obligatory phase and importantly, it was made clear to them that the same
word could be used to describe different stories.
As noted in Section 1
above, experimental consistency inhered in the rendering down of each source
story into a short text of forty-four words. In addition to this
triangulation for length, all of the stories were levelled stylistically by
being stripped of any potentially “irony inducing” expressions that might
signal some kind of conceptual paradox between the conjuncts in the story.
Expressions that were avoided included Mood, Comment and Conjunctive
Adjuncts, as in, respectively, “obviously”, “unfortunately” or “however”,
and related expressions of Extension and Enhancement, as in “but” or
“although” (see Halliday
1994: 81–84, 219–220; Halliday and
Matthiesson, 2004: 126–132, 395–413). The first story in the
experiment was drawn from a genuine news report that ran in the British
print media in December 2011. The story concerns a car driver who,
immediately after his anger management class, became involved in a road rage
incident. Under the header “The Filth and the Fury”, the Daily
Mail describes how Philip Croft launched a “four-letter tirade
at a policeman … on his way home from an anger management course”
(Daily Mail 2011). Given that the anger management
course had been made compulsory after an earlier offence, the District
Judge’s summation leaves no doubt about his pragmatic interpretation of the
episode: “There’s a certain irony that on your [i.e. Croft’s] way back from
an anger management session you behaved like this with a police officer.”
Rendered down into the stylistic template set out above, this is how the
story appears on the questionnaire:
Story 1:Mr. Jones had been ordered by a court to take part in anger management classes. On his way home from one of his classes, he was pulled over for speeding. Mr. Jones verbally abused the police officer and was later convicted for the offence.
For ease of reference when assessing the quantified results
later in this chapter, this story will be referred to hereafter as “Story 1:
Anger Management”, although the added descriptor was
not of course present in the online survey. Similar mnemonic descriptors
will be added to the remaining five stories.
The second of the “ironic” stories is derived from data
presented in Shelley’s corpus-based study of irony in electronic news
sources (Shelley 2001). As
outlined in Section 1, Shelley’s
data is developed from press reportage in which the lemma IRONY and its
derivatives appear. One particular episode in his corpus lends itself well
to the present study in that it features both an appropriate narrative event
and a suitable commentary on the event from one its
participants. This is the story of the fire-fighters of Station 20, Clark
County, Las Vegas, who on one hapless afternoon were called out to put out a
fire. Only later were they to discover that the lunch that they had been
cooking, of what seemed in the report to be chicken goujons, had been left
on the stove and had set ablaze the fire station. A spokesperson for the
service captures the episode thus: “It just shows that if it can happen to
us, it can happen to anyone … The irony’s not lost on it.” (adapted
from Shelley 2001: 775–6).
Modified following the stylistic template, and appearing in third position
in the questionnaire, this how the “fire brigade” story reads:
Story 3:Members of the fire brigade were preparing a lunch of chicken goujons. They were called suddenly to attend to a fire in a nearby residence. When they returned, they discovered the chicken goujons had caught fire, and had set the whole fire station ablaze.
The genus for the third of the three “ironic” stories is from
media coverage of an episode involving a power outage at a soccer ground.
The online UK news platform Metro describes how, in
December 2013, fans of English football club Leicester City used their
mobile phones in an attempt to light up their stadium after the floodlights
had failed. Given that the stadium was sponsored by the King Power conglomerate (though strictly, in spite of public perception, not itself an energy provider), the irony of the situation was teasingly captured in comments like
the following: “The floodlights went out midway through the clash at the
aptly named ‘King Power Stadium’ – with power cuts
affecting the whole area” (Metro 2013; my emphasis). This scenario,
the “power cut”, formed the basis for the sixth story on the
questionnaire:
Story 6:A soccer team, sponsored by a national electricity company, was playing its local rivals in an important match. The floodlights went out midway through the contest with power cuts affecting the whole area. The referee stopped the match until the problem was finally solved.
Turning to the development of the “non-ironic” stories, some
useful source material was taken from a particular body of empirical
research on reading and text processing (Egidi and Gerrig 2006; Rapp
and Gerrig 2006). This research explores principally the
psychology of readers’ preferences for narrative outcomes and examines the
way readers react to certain characters’ goals and actions in stories.
Bluntly put, none of the material used in such experimentation is remotely
connected to irony; it is instead focused on how readers tend to identify
with stories with that are consistent with a priori
expectations and how reading time is affected when stories present unusual
or “dispreffered” outcomes. One scenario developed by Egidi and Gerrig (2006: 1322; following earlier
work by Huitema et al. 1993 and
Poyner and Morris 2003)
involves the sun-loving character Dick who likes to swim and sunbathe.
Planning his vacation, Dick visits his travel agent and “ask[s] for a plane
ticket to Florida”. Egidi and Gerrig observe that readers are consistently
less quick to read an inconsistent action in the same
context, as in their alternative scenario when Dick asks his travel agent
for a plane ticket to Alaska. Another interesting story, developed in the
same paper (2006: 1325–6), is when
a man robs a well-known coffee franchise and makes speedily for the Mexican
border. In one variant of the story, however, the robber interrupts his
getaway by stopping at the side of the road for a sleep. To re-iterate,
these scenarios are not connected to the study of irony but are instead
concerned with how readers deal with inconsistent situations where
characters’ actions contradict their purported goal. In this respect, they
make for excellent contrastive material and in the questionnaire they have
been adapted into stories 4 and 5:
Story 4:Mr. Smyth had just robbed a city bank and was making good his escape in his getaway car. He drove at speed towards the border. After fifteen minutes on the road, he pulled over and had a snooze at the side of the road.
Story 5:Ms. Smyth worked in an office in the city. She enjoyed her holidays in the sun and was always looking forward to her summer break in warmer climes. As soon as her last day at work was over, she booked a flight to Iceland.
The final story in the experiment has an unusual provenance
insofar as it draws on a debate about irony that was played out in the
sphere of popular entertainment. The detail of this debate is reported
elsewhere (Simpson 2011: 42–44)
but the substance of it concerns the lyrics of the song “Ironic” by Canadian
singer-songwriter Alanis Morrissette and the subsequent unpacking of these
lyrics by Irish stand-up comedian Ed Byrne. In essence, Byrne interrogates
the semantic and pragmatic foundations of the song, concluding amongst other
things that the only thing that is ironic about “Ironic” is that it was
written by someone “who doesn’t know what irony is”. Byrne’s overall
position is that the singer “keeps naming all these things in the song that
were supposed to be ironic, and none of them are. They are all just …
unfortunate”. Although not the only comedian to probe these lyrics (see
Shelley 2001: 784), Byrne
takes particular issue with the sequence: “And isn’t it ironic, dontcha
think? … It’s like rain on your wedding day” (Simpson 2011: 43).
On the potential of this “wedding day” scenario to embody
irony, Byrne quips with the following proviso: “Only if you’re getting
married to a weather man and he set the date”. Clearly, the debate enacted
here not only offers potential for empirical investigation but the state of
affairs it depicts can easily be integrated into the test materials. It is
the ‘wedding day” scenario therefore that informs the remaining story in the
questionnaire:
Story 2:Ms Smyth had been looking forward to getting married for some time. She and her friend had chosen a fine white dress for the ceremony, which was held on the 3rd of June at a local country club. It rained on her wedding day.
To round off this section, some observations will be offered
on the process of textual composition of the six stories and on some of the
issues and challenges the process raises. Thereafter, and working up from
the data, a model for situational irony (and its comprehension) will be
suggested. This model offers a more fine-tuned account within the broader
terms of the umbrella definition set out in Section 1 above.
As noted, it was important in the composition of the stories
to avoid any kind of evaluative modality that might direct readers to a
specific interpretation. Particularly important was the circumvention of
aspects of text that might signal the presence of the kind of conceptual
paradox that is inherent in irony. Of course, the textual structure that
this circumvention engendered tended to make the six stories seem rather
bland, perhaps disengaged or even somewhat “Hemingway-esque” in feel.
Although necessary, this pattern of levelling did elicit some reactions in
the more open responses on the questionnaire, more on which later. A related
issue arose in other aspects of composition that ran the risk of attracting
prescriptive meta-commentary. For instance, the opening to Story 3 as
originally drafted was “The fire brigade were making their lunch” but
informal feedback from university colleagues at the pre-experimental stage
tended to draw attention to the “lack of agreement” between the singular
subject and plural verb form (cf. “The fire brigade was making
its lunch”). If only to avoid prescriptive reactions to
grammar that are not relevant to the aims of the study, the text of 3 was
altered accordingly (and a similar strategy invoked for Story 6, “A soccer
team was playing … its local
rivals”). Additionally, in the transition from source narrative to test
narrative, some stories were fine-tuned to avoid possibly context- or
culture-dependant associations. For instance, “soccer” was preferred to
“football” because of the wider sphere of reference the former term enjoys
outside the UK and Ireland. A reverse of this transposition is in Story 5:
Vacation, where the USA-orientated destination of
“Alaska” was replaced by “Iceland”, simply on the grounds that the latter
encapsulates a perhaps more universal embodiment of a location with a cold
climate. Similarly, the explicit allusion to the Mexican
border was removed in Story 4: Robbery, again because of
geographical specificity. With these late revisions in place, the survey
opened on the 8th of June 2015 and closed on 31st December 2015. The results
presented in Section 3 below are
based on the 321 responses collected during this period.
One corollary of the bottom-up development of ironic
situations for experimental purposes is that it allows theoretical
constructs to be developed and refined post hoc from the body of data
itself. The observations that follow are sustained by three of the stories:
Story 1: Anger Management, Story 3: Fire
Brigade and Story 6: Power Cut. The situations
portrayed in this group of narratives enable a more nuanced
linguistic-pragmatic account of the umbrella definition to be attained.
These refinements concern especially the second sub-definition of irony;
that is, where irony inheres in a mismatch between aspects of encyclopaedic
knowledge and situational context with respect to a particular discursive
event. This broad characterisation implies the presence of two distinct sets
of knowledge for the understanding of an ironic situation. The first of
these is a more stable conceptual repository that mirrors and parallels
various comparable conceptual units that have been developed in both
discourse analysis and cognitive linguistics. Far from forming an exhaustive
list, analogous terms which suggest themselves are “mental space” (e.g.
Fauconnier and Turner
2002: 40), “frame” (e.g. Goffman
1986: passim), “conceptual frame” (e.g. Minsky 1975: 211), “memory organisation packet”
(e.g. Schank 1982: 95) and
“encyclopaedic entry” (e.g. Sperber and
Wilson 1986: 86–93). Figure 1 illustrates how this more constant conceptual space,
situated in the box on the left, is characterised by a number of attributes:
it is encyclopaedic, universalistic,
stative, atemporal and
enabling.
Figure 1.Modelling situational irony
The first two boxed attributes on the left encapsulate the
conceptual space along the broad lines of the cognate terms in linguistics
just mentioned; that is, as beliefs, memories and assumptions about things
and states of affairs. In contrast, both of these more abstract attributes
collide with the palpable features of the ironic situation, positioned in
the circle on the right of the figure, where generality gives way to the
particular and the local aspects of
the discursive event. The staggered arrow in the centre of the Figure
signals the paradox and captures the mismatch between the two sets of spaces
through semantic relationships of negation, contradiction, and relational or
complementary antonymy (see Jeffries 2010). The next two terms in the left
hand box, stative and atemporal, highlight
more narrowly the framing aspects of this dimension of situational irony.
These aspects can be teased out and brought to the fore through the
grammatical forms used to describe them. For instance, the relevant aspects
of Story 3: Fire Brigade might be cast propositionally as
“Firefighters put out fires”. The expression of this propositional form,
through the so-called “timeless present”, is therefore as a
universal truth. This generic structure, which has no
specific action or time reference, cannot be inflected for aspect or for
other tenses. The same level of generality applies to the core propositions
in the other two stories: Energy companies provide power
and Anger management classes prevent people from doing angry
things. In contrast, the parallel aspects of situational
context portrayed on the right of the Figure are anything but generic. On
the one hand, they capture temporal episodic information
through tenses other than the timeless present; on the other, they employ
grammatical resources that are punctual in the specific
sense of describing action that is grounded in a particular time. Additional
particularised reference is achieved through deictic words, demonstrative
determiners and definite and indefinite reference – all of which serve to
pick out “real” individual actors, participants and places. The fifth
category on the Figure is cued by Shelley’s observation that ironic
situations often have a negative association, or are prompted by a goal that
is unachieved (2001: 784–5).
Shelley describes this aspect of situational irony in terms of “manner”,
which is the main measure of the distance between an ironic situation and
“the goals, concerns, and preferences that a cognizer applies to it” (2001: 785; and see also Partington’s
discussion of irony [2007: passim]
in terms of “reversal” of expectation). Certainly, the three stories here
are all grounded in discourse contexts that are potentially facilitative,
positive or biased towards a good outcome for their participants; in
contexts that are in other words enabling. By contrast, the
actual outcomes in all three stories are dis-enabling in
the sense that the positive expectations are thwarted. In the Figure, each
of the core propositions below the box on the left are opposed through
complementary opposition or antonymy by negation in the corresponding
grammatical realisations on the right. Indeed, this sense of being “let
down” by the contextual outcome may be part of the humour-inducing aspect of
some kinds of irony, but it may also account for qualitative assessments of
certain forms of irony. Although an area of investigation that is beyond the
scope of this chapter, corpus analysis reveals for instance how the lemma
IRONY attracts particularly common collocates, notably in judgements of
irony as “heavy”, “bitter” or “grim” (Simpson 2011: 41–42). For the moment, attention will focus on
the results of the Questback survey and to this effect the next section will
offer an overview and assessment of the data generated by the
questionnaire.
3.Results and discussion
Over its course, this section will assess the broader patterns
of response collected through the survey, before drilling down into some of
the more nuanced qualitative features of subjects’ individual contributions.
The section concludes, by way of denouement, with the final set of
quantified results, aided by tabulated histograms that summarise the
outcomes of the survey. The data set reveals much about consistency of
linguistic-pragmatic processing across often diverse cultures, with an, at
times, remarkable level of convergence in patterns of response to some of
the stories. This convergence is striking when we recall that informants
were permitted only a single word response in the first stage of the online
protocol. Nevertheless, the results also throw up many surprises and
inconsistencies, and in some areas, more questions are raised than can be
answered in a study of this scale. Attention will be given later to where
the results point towards further research questions, or to possible
modifications to the existing methodology and experimental design.
In terms of very raw numbers, the 321 responses which form the
data set break down as follows: North America (NA) produced 169 completions
(52.6% of the total), while the UK & Ireland (UKIre) realised 63
completions (19.6% of the total). Within the latter category, four
informants declared their nationality as “Northern Irish” and two as
“Scottish”, although the broad-based assimilation of these national
affiliations into the umbrella grouping “UK and Ireland” is, it is hoped,
uncontentious. There remain 89 informants from what can be classified,
admittedly less than ideally, as the “Rest of the World” (RoW) group. This
group made for 27.8% of the survey, and of these 89 informants, the largest
national affiliations in size of grouping were Chinese, Spanish, Brazilian
and French, moving respectively, from 20 down to 12 informants in each
group. At the other end of the scale, there were 11 single-nationality
responses, from individual informants as far apart as Cyprus, South Korea
and Sierra Leone. Although the aims of the present study preclude detailed
exploration of this very heterogeneous group, it is still worth factoring in
RoW responses where they intersect in interesting ways with the comparative
analysis of the other two groups.
To generate a visual display of the convergence in response
touched upon above, Figure 2 is a
word cloud that captures all 321 one-word responses to the third story in
the survey.
Figure 2.Single word reactions to Story 3: Fire
Brigade.
In this and the other word clouds that follow, the variants
“irony” and “ironical” have been subsumed into the adjectival form that
dominates the cloud visually, on the basis that, of the three possible
forms, “ironic” was the item most commonly used across the one-word
responses. Based on preponderance, the cloud demonstrates a striking level
of agreement across all nationalities and this includes many informants for
whom English is not a first language. In other words, with only a single
word available to grasp the mishap that befell the luckless fire brigade in
Story 3, the overwhelming bulk of informants instinctively and consistently
draw from the semantic-pragmatic space that is inhabited by the concept of
irony.
While specific numerical data will follow later, it is worth
observing that the orientation to “ironic” renders the other choices barely
visible in Figure 2. The nearest
contenders include “coincidental”, “funny”, “careless” and “stupid”. In
contrast, the word “fire” appears, which tends to recapitulate the
narrative’s subject matter rather than offer an interpretative response.
Although reasons of space preclude the presentation of word clouds for all
six stories, and indeed, of word clouds broken down further by national
grouping, the item “ironic” is dominant, and only markedly less prominent,
in the reactions to the other “ironic” stories; that is, to Story 1:
Anger Management and Story 6: Power
Cut. However, within this preponderance, there are interesting
and nuanced variations around the north American and the UK and Ireland
groups, as the quantified data adduced later will confirm.
Consider how this marked clustering compares to a word cloud
derived from reactions to one of the control stories, such the scenario
involving the bank robber who pauses to sleep during the getaway. Clearly,
the cloud in Figure 3 embodies a
much more heterogeneous set of responses. Visually, whereas the word
“stupid” clearly dominates, many more lexical items make a significant
showing, indicating a much more open-ended set of reactions. Again, some
informants offered a summative recapitulation (“robbery” and “robber”), but
most interpretations converged on the “stupidity” of the protagonist (e.g.
“dumb”, “foolish” or “silly”) or on the seeming pointlessness of the story
proper (“anticlimactic”, “odd” or “nonsensical”). Throughout the data set,
individual free-text comments made for very interesting reading, and to
represent these comments economically, a coding system was developed where
the informant receives a number, based on the chronological sequence in
which each of the 321 completed surveys was received, alongside a
designation of their national grouping. Thus, informant number one hundred
and thirty one from the RoW group is therefore coded as “#113/RoW”. Of Story
4, this informant asks “What happens then?”, whereas #75/UKIre has more to
say: “Mr Smyth is an idiot. Either floor it to the border, or hide”.
Orientating towards the “dispreffered outcome” aspect of the story discussed
earlier, #171/NA says the “[s]tory lacks finality but reader might supply an
ending” [sic, here and passim], while the
acerbic #200/NA, using irony of their own, remarks: “Cool story–you
should tell it at parties. … (sarcasm)”. It is worth noting that this
overall pattern of response is replicated for Story 5:
Vacation, which is the other story where the
character’s actions contradict their purported goal. Here, the item
“contradictory” dominates while other popular choices include “unexpected”,
“strange”, “nonsense” and “confusing”. Again, there is some thematic
replication (“vacation” and “holiday”), while seven informants actually
offered “ironic” as their single word response. On this last choice, it is
noteworthy that the item “ironic” appeared at least once in all six sets of
responses, and this is an issue that will be discussed later.
Figure 3.Single word reactions to Story 4:
Robbery.
There remains one story that has yet to receive attention.
This the story of Ms Smyth’s wedding plans and the subsequent rainy wedding
day. Figure 4 displays the
relevant word cloud. As it turns out, the predominant single-word response
is precisely that supplied in Ed Byrne’s skit on Alanis Morissette’s song
(above); that is, that rain on your wedding day is simply
unfortunate. Other entries in the cloud supplement this
interpretation, through near-synonyms like “unlucky” and “disappointing”.
Most informants therefore see no paradox in Story 2, nor
any of the potential irony-generating conceptual oppositions embodied in
Stories 1, 3 and 6. Again, some thematic replication occurs, as in “wedding”
and “marriage”, but overall, these responses do not tell the whole story.
For a start, some informants offered “lucky”, interestingly in opposition to
its slightly more dominant negated variant. The free-text comments offer a
clue: informants point out that in some cultures, like Italy and Brazil, it
is good luck to have rain on a wedding day. #3/RoW says that “In Brazil
raining wedding day bri s good luck.” while #152/NA offers a personal
account of her own: “It rained before my wedding. And the people told me an
Italian folk saying ‘a wet bride is a lucky bride’”.
Figure 4.Single word reactions to Story 2: Wedding
Day
Aside from these cultural factors, some informants, as the
centre of the cloud shows, do offer the word “ironic” as
their single-word response. The unpacking of this through associated
free-text comments is interesting. Ten respondents (out of the 98 who
offered comments) mentioned the Morissette song by name in their follow-up
comments, as with #4/RoW who remarks: “It makes me remind a musica – Isn’t
it ironic from Alanis Morissette”. Other free-text comments hint perhaps
playfully at the song’s title (“ironic, isn’t it?”). Emerging as something
of a wit, informant #200/NA actually attempts to capture the musical
phrasing of the song thus: “IT’S LIKE RAIIIINNNNNN ON YOUR WEDDING DAY”.
Curiously, allusion to the song did not depend on whether “ironic” had been
offered in the single word space. Having inserted “misfortune” in this
space, #13/UKIre dwells at length in the free-text space on the perceived
non-ironic aspects of the song:
This reminds me of the Alanis Morrissette song, ‘Ironic’ where she sings that irony can be explained as “rain on your wedding day.” It always irritated me because that isn’t irony – just bad luck!.
Amusing as the free-text commentaries often are, they raise
more serious questions about the interface between popular culture and
everyday interpretative strategies. Are the words of a popular, high-profile
singer-songwriter responsible for shaping some of the one-word responses?
Can song lyrics act as a higher-order determinant on the
linguistic-pragmatic inferencing strategies employed by ordinary users of
language? Or more bluntly, have informants used “ironic” for the wedding
scenario simply because Ms Morissette’s song describes it thus? All of this
is probably true to some extent, although this interpretation does not
explain the relatively high incidence of “ironic” in reactions to Story 5:
Vacation, where there is no such informing framework
from popular culture. We will revisit some of these issues in the concluding
section, but to bring this section to a close, it is appropriate now to
“unveil” the statistical breakdown of ironic across all
stories and across all national groupings. Figure 5 comprises a histogram which charts in
percentages all responses that produced the single word “ironic” in the
relevant part of the online survey.
Figure 5.Overall percentage of single-word “ironic”
The histogram’s simple headline message is that, globally,
people “do” irony in some way or another. It also highlights the non-binary
and continuum-like nature of irony processing, where certain situations are
felt to be clearly “more ironic” than others. As predicted, the three
“ironic” stories are displayed as precisely that, although the data shows
that Story 3: Fire Brigade is perceived as the “most”
ironic scenario while Story 1: Anger Management is arguably
the most marginal of the three. To reiterate, when charted against this pool
of over 300 respondents, the perception of irony is variable and is not,
even for Story 3, necessarily the “go to” inference for all
informants. The boundaries around conceptualisations of irony are, moreover,
porous: a small percentage of informants, as Figure 5 shows, claimed irony for the stories that
had been intended as the “control” narratives. Notably, Story 5:
Vacation drew even more ironic inferences than those
for the contested rainy wedding day scenario.
What then does the data say about the central themes of this
chapter? Clearly, North Americans do do irony, but is there
enough in the histogram to sustain a meaningful cross-cultural distinction
between this group and the UK and Ireland group? Reactions to Stories 1 and
6 do suggest a difference, where the incidence of “ironic” in the UKIre
columns is, respectively, 4% and 13% higher. However, the striking column,
which bucks this trend, is Story 3: Fire Brigade, where the
North American group was 8% higher in its rate of ironic interpretation of
the narrative situation. Thus, where Stories 1 and 6 collectively display a
8.5% higher incidence for the UKIre group, this is offset by the 8% variance
in the other direction, as displayed by Story 3’s columns. Moreover, the
possibility that there may be something in Story 3 that promotes a
particular ironic reading for the North Americans (a cultural resonance or
geographical clue, perhaps) is thoroughly nullified because this high score
is matched by a proportional rise in the UKIre group also. That said, the
survey’s results, alongside the free-text comments, did draw attention to
aspects of the composition of the stories and to features of its design that
may require re-modelling. The final section of this chapter deals with such
issues and suggests ways in which the large-scale exploration of (reactions
to) ironic situations may be developed further.
4.Conclusions
It has been the principal aim of this chapter to address and
challenge a popular notion that speakers from North America “don’t do”
irony. The evidence drawn from the survey, with over 300 participants
worldwide, and 169 of them from North America, points to the contrary; that
is, that the findings presented here do not support this folk perception of
a connection between nationality and the capacity to produce and understand
irony. While the results show that there are admittedly some differences in
the degree to which ironic inferences are made, these are too marginal to
make a strong and statistically-significant case at this stage. The one-word
data used as the main part of the survey was also supplemented with
free-text commentaries, and the latter alone would make for a study in its
own right, not least because of the light it sheds on the diversity of the
textual clues that underpin the informants’ responses. Moreover, although
the six stories were rendered down into what was thought to be accessible,
universalistic and context-independent formats, the free-text comments at
times cast some doubt on this kind of design. For example, #18/NA is not the
only informant to comment on the firefighters’ meal when he/she asks of
Story 3: “I wish I knew what chicken goujons were!”. Other informants probed
what they perceived as implicit ideological assumptions in the stories.
Having entered “lesbian” in the single word space for the wedding day
narrative, #121/UKIre entertainingly takes issue with the bourgeois
affectation of the country club scenario presented: “Well, you know, ms then
her friend … but a country club? None of my gay friends would marry there”.
But one of the most significant pointers towards a possible re-think in the
textual composition of the stories concerns respondents’ linking of
characters across the stories in such a way as to place them all in the same
narrative universe. During the experimental design, characters’ names like
“Smyth” and “Jones” were simply intended as generic monikers, acting as
neutral place holders in the stories. It came as a surprise then that many
informants linked these characters across the stories. For instance, several
respondents raised concerns that the bride of Story 2 had made a poor choice
in choosing the somnolent bank robber of Story 4. #88/UKIre writes:
With a fianceé as manically unpredictable as Mr Smyth, I would say rain on her wedding day is the least of Ms Smyth’s worries.
while #183/NA comments:
I hope this isn’t the man Ms Smyth from Story 2 married. If so, poor choices. Also, not the best criminal, is he?.
Clearly, many informants had sought to establish a network of
interrelated stories in the survey and while an interesting interpretative
strategy in and of itself, this search for interconnectedness was not
anticipated at the experimental design stage.
Returning to the data in Figure 5, one of the most important theoretical questions is
what makes one story in the collection more ironic than
another? Part of the answer may, again, lie in composition. The model of
situational irony postulated in Section 2 sets out the generic structures for each of the three
relevant stories in the survey. The core element of Story 3: Fire
Brigade was cast propositionally as “Firefighters put out
fires” which collides with the particular and localised features of the
ironic situation (that is, “These firefighters started a fire”). The most
pithy of the three, this formula also seems the most universally accessible
of the three sets of paradoxes on Figure 1. Alternatively, the abstract components for the
arguably less-accessible anger management scenario, by contrast, take longer
to tease out (“Anger management classes prevent people from doing angry
things”) and although they still satisfy the requisite criteria for an
ironic situation, the results confirmed that this story was less readily
classed as ironic across the sample. Moreover, in the text of Story 3, the
lemma FIRE occurs four times and in three different grammatical
environments, and is further supplemented by the cognate term “ablaze”. This
oversight in textual design may constitute an example of what Nash calls
“the hazard of reiteration” in composition, where a lexical form is
repeated, often clumsily, within close syntactic limits (Nash 1980: 48–49). The inherent paradox of the
situation in Story 3 is thus stylistically more foregrounded than in the
other two stories, making the antonymic contrast easier to access.
Finally, as suggested across Section 3, the data collected in the survey is richer and has
more to offer than can be interrogated fully here. It raises questions about
other variables that might play a part in shaping the results. Whereas the
study asked for self-descriptors of nationality, a future study might bring
ethnicity into play also, although asking informants for a perhaps more
sensitive self-assessment of ethic grouping could prove a challenge. The
survey did, however, collect information on age. Looking at responses from
those born during or after the year 1996 – it seemed somehow appropriate to
use the year the Morissette song was released – yields some interesting
results. This group of younger informants produced, on five of the stories,
irony ratings that were on average 5.1% higher than the ratings from those
born before 1996. This was especially marked on Story 5:
Vacation. In fact, the only story on which this trend
was reversed was (ironically?) the wedding day scenario. It might be that we
are using the expression “ironic” more and more, and with perhaps looser or
more permeable semantic-pragmatic boundaries. The longitudinal study that
this observation warrants must, however, wait for another day.
Acknowledgements
The author would like to thank Diarmuid Kennedy, Senior
Librarian at Queen’s University Belfast, for his invaluable support in
developing this experiment and for his guidance on the use of online
systems. I am grateful also to John Garry, for his advice on the
quantitative aspects of this chapter. I would also like to thank for their
feedback the colleagues who attended various of the conferences, meetings
and symposia at which parts of this research chapter were disseminated, and
especially participants at these universities: Lyon, Sheffield, Nottingham,
Portsmouth, Aix-en-Provence, Liverpool Hope and Manchester Metropolitan.
Finally, I am of course grateful to the 321 respondents
world-wide who completed the online questionnaire and without whose
participation, and often hilarious interpolations, the experiment would not
have been possible.
References
Appendix 1Questionnaire (First Page)
Questionnaire
I am interested in the ways in which people respond to
stories and am currently conducting some research into how we react
to simple, short narratives. I am particularly interested to see if
people in different parts of the world, or people of different
nationalities, react in different ways to stories. I would be
grateful therefore if you would complete the short questionnaire
below.
The questionnaire includes six little stories. After you
read each story, you will be asked for a quick one-word
response. If you feel it is appropriate, you can use the same word
to describe other stories as well.
Later in the questionnaire, you will be asked for more
comments on the stories (if you want to provide them) and for some
information about yourself which will be treated anonymously.
Story 1:Mr. Jones had been ordered by a court to take part in anger management classes. On his way home from one of his classes, he was pulled over for speeding. Mr. Jones verbally abused the police officer and was later convicted for the offence.
If you had to describe story 1 in just one
word, what word would it be? Write it in the box:
Story 2:Ms Smyth had been looking forward to getting married for some time. She and her friend had chosen a fine white dress for the ceremony, which was held on the 3rd of June at a local country club. It rained on her wedding day.
If you had to describe story 2 in just one
word, what word would it be? Write it in the box:
