British heatwave discourse (1985-2023): from ice cream to armageddon?

Abstract

How has public discourse about climate change evolved? This paper answers this question through the paradigmatic case of British heatwaves. These are traditionally considered a fortunate break from dull weather, making them a least likely case for the emergency of a discourse of doom. With a topic-modeling analysis of British national newspaper articles on heatwaves from 1985 until 2023 (N = 35,127), we show that a longstanding Romantic heatwave discourse eventually buckled, and that Apocalypticism finally became the dominant genre in the last decade. A supplementary hermeneutic analysis then indicates and explains complexity within this broad trend. 1980s stories already noted routine heatwave problems, while many recent ones continue depicting positive lifestyle implications. Within the Apocalyptic genre itself climate change is today deemed a factual causal force, whereas in the 1980s and 1990s it was a possible carrier of future dangers. In connecting the genre perspective from literary theory to big data method topic modelling, our approach is parsimonious, novel and replicable in other national contexts.

Full text

Environmental Sociology
ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/rens20
British heatwave discourse (1985–2023): from ice cream
to armageddon?
Thijs van Dooremalen & Philip Smith
To cite this article: Thijs van Dooremalen & Philip Smith (22 Apr 2025): British heatwave
discourse (1985–2023): from ice cream to armageddon?, Environmental Sociology, DOI:
10.1080/23251042.2025.2493332
To link to this article: https://doi.org/10.1080/23251042.2025.2493332
© 2025 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group.
Published online: 22 Apr 2025.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rens20

British heatwave discourse (1985–2023): from ice cream to armageddon?
Thijs van Dooremalen
a
and Philip Smith
b
a
Institute of Security and Global Affairs, Leiden University, The Hague, The Netherlands;
b
Sociology Department, Yale University, New
Haven, CT, USA
ABSTRACT
How has public discourse about climate change evolved? This paper answers this question
through the paradigmatic case of British heatwaves. These are traditionally considered
a fortunate break from dull weather, making them a least likely case for the emergency of
a discourse of doom. With a topic-modeling analysis of British national newspaper articles on
heatwaves from 1985 until 2023 (N = 35,127), we show that a longstanding Romantic heatwave
discourse eventually buckled, and that Apocalypticism nally became the dominant genre in
the last decade. A supplementary hermeneutic analysis then indicates and explains complexity
within this broad trend. 1980s stories already noted routine heatwave problems, while many
recent ones continue depicting positive lifestyle implications. Within the Apocalyptic genre
itself climate change is today deemed a factual causal force, whereas in the 1980s and 1990s it
was a possible carrier of future dangers. In connecting the genre perspective from literary theory
to big data method topic modelling, our approach is parsimonious, novel and replicable in
other national contexts.
ARTICLE HISTORY
Received 3 December 2024
Accepted 10 April 2025
KEYWORDS
Climate change; extreme
weather; heatwaves; genre
analysis; public discourse;
longitudinal; mixed-methods
text analysis; Britain
Introduction
Heatwaves are recurrent features of social life. Analytically
distinct from expected hot weather, we are as likely to
nd one in Iceland as in Algeria. The heatwave is an
anomaly. Temperatures are signicantly hotter than nor-
mal for any given place 3 days in a row (Met Oce 2023a)
and so are noteworthy as ‘unusual’ or as an ‘event’. They
interrupt routine weather and call for comment. When
associated with climate change, heatwaves speak to the
place of humans in the ecosystem, and the fate of the
planet. But they can also be read as opportunities for fun,
for a picnic or a trip to the beach. This paper moves this
duality of meanings center stage as a methodological
resource. Inspired by Raymond Williams’s (1977) work
on dominant, emergent and residual cultural patterns it
looks at developments of these two meanings in recent
decades. It does so with an eye to benchmarking any
shifts in language and interpretation that might have
taken place by deploying the concept of ‘genre’ to orga-
nize scattered knowledge and information.
Looking closely, we see that heatwaves bring
many advantages to the sociological researcher of
the meanings of climate change (see also Batziou
2022; Hopke 2020; O’Neill et al. 2023). They have
something of the quality of a natural experiment.
Unlike government carbon policies, climate confer-
ences and scientic reports that might also generate
deliberation, these are experienced as exogenous to
any proximate human agency. They are not fully
predictable, but at the same time they demand
a societal response. The resulting discourse responds
directly to the events, and is not shaped in normative
ways like surveys, interviews or political environ-
ments. Indeed, we might think of heatwaves as gen-
erating unobtrusive measures in the form of words,
images, editorial choices, and existential moods. They
are also broadly similar to each other – they all
involve ‘hot weather’. Moreover they can be dated.
These qualities make comparisons of event discourse
over time in the same data set legitimate – this is not
a case of comparing apples and oranges. Unlike
other weather events and disasters such as ooding
rivers, blizzards, hurricanes and earthquakes, heat-
waves cover extensive territory (West and Smith
1997) and are probably more easily viewed by ordin-
ary people as intuitive symptoms of climate change.
Because the signs they produce impact ‘everyone’
and involve ‘heat’ they permit stronger inferences
about any purported ‘national’ shifts in climate cul-
ture. If the evidence from both social media and
survey research shows that extreme weather events
are a fruitful location to investigate public orienta-
tions towards climate change, this is particularly true
for heatwaves (Konisky, Hughes, and Kaylor 2016;
Sisco, Bosetti, and Weber 2017).
The meaning of the weather in Britain
In this paper we seek to redeem this promise and
potential. Our particular focus is on the media frames
CONTACT Thijs van Dooremalen t.j.a.van.dooremalen@fgga.leidenuniv.nl
ENVIRONMENTAL SOCIOLOGY
https://doi.org/10.1080/23251042.2025.2493332
© 2025 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting
of the Accepted Manuscript in a repository by the author(s) or with their consent.

through which heatwaves have been understood in
Britain. Like many other places in temperate regions
where the major air masses collide and depressions
march one after another from west-to-east, the British
weather has traditionally been experienced and inter-
preted as suboptimal, unexceptional, drab, windy and
wet. These properties have mapped onto its day-to
-day unpredictability and impersonality to make it the
favored topic for inoensive, solidaristic everyday con-
versation, especially among strangers (Harley 2021).
The British can safely express disappointment, hope,
pleasure and low level suering through routinized
talk. Over the centuries the indierent weather has
seeped into the fabric of social life to the point where
it will form collective memories, motivate backyard
science, shape television viewing routines, and be
deployed to explain an allegedly phlegmatic national
character (Colls 2002; Golinski 2007; Harley 2021).
Much of the time British weather really is temperate
according to the data. Annual rainfall is about
1000 mm and it happens on half the days of any
given year. Daily high temperatures generally range
from 5 to 25 °C depending on the season, and have
done so for centuries according to the renowned
Central England Temperature Dataset (Met Oce
2022). But by the same token it is this cool temperate
predictability and experiential mediocrity that has
made heatwaves notable, exciting and enjoyable for
the Brits. These become truly memorable events that
shape cognition of the past and structure generational
nostalgia even decades after they took place (Harley
2021). Such has been the traditional, positive response.
Yet in recent decades scientists have warned of
extreme weather events. Thus, today the heatwave is
potentially a sign of climate danger and, also in Britain,
a more negative reading is widely available.
Genres of climate discourse
How are we to think about these contending positive
and negative framings of the heatwave in Britain? To
answer this question in a parsimonious way we draw
upon the genre model of risk perception (Smith and
Howe 2015, 51–52), which is loosely derived from
Aristotle’s Poetics via Northop Frye’s (1957) modernist
eorts at formulating a comprehensive theory of nar-
rative forms and the relations between them. At the
price of a radical simplication of discursive complex-
ity, this approach facilitates generalizing comparative
sociology much as it assisted the evolution of com-
parative literature as a eld. We will see in this paper,
for example, how a multiplicity of topics emerging
from our data on the heatwave can be organized –
and hence a confusing, crowded picture of many small
ndings can be made more easily ‘knowable’ if each
topic is treated as an example of a more general type.
Finally we can see the woods, not just the trees. The
genre model of risk has been shown to be widely
applicable over a range of social issues and problems.
For example, recent analyses have looked at the dan-
gers of illicit drugs and organized crime (Barrera 2017),
COVID-19 (Morgan 2020) and articial intelligence
(Binder 2024). It argues that such events, challenges
and threats do not speak for themselves nor dictate
policy without some kind of translation. The various
bits of information are inserted into larger, commonly
found cultural structures known as ‘genres’ that
instruct on ‘what to do’ and ‘what is really happening’.
They speak of how much ‘risk’ there is and how con-
cerned we should be about the challenge at hand.
Crucially for our task we have a proof of concept, as
the issue of climate change has already been demon-
strated to be framed in this way. Smith and Howe
(2015, 50–70) show how its implications and risks
have been ramped up (‘inated’) and down (‘deated’)
over the decades by experts, politicians and commen-
tators through a series of genres that have structural-
poetic properties. Extreme weather events are ltered
through these cultural structures that identify conse-
quences to action and inaction, allocate blame, predict
the future, and provide a sense of risk. But there is no
consensus. Over the years there have been a series of
faltering attempts to mobilize what Smith and Howe
(2015, 58–63) call an Apocalyptic genre. Seemingly
adapted from Biblical stories and eschatologies of glo-
bal doom, this speaks to catastrophic planetary scale
emergencies, world historical shifts, the need for
urgent action to ward o a doomsday, to growing
portents of looming disaster and so forth. Rather than
being fatalistic, in their account Apocalypticism is a call
to arms. Heroic action, solidarity and sacrice are
needed to face down a challenge to civilization itself.
But in the case of climate change – and especially
the British heatwave – the Apocalyptic genre has to
struggle against a Romantic one. Derived from
Aristotle’s ideas about comedy as a pro-social genre
for storytelling, this argues human ingenuity has
always solved problems and that things generally
turn out for the best (Smith and Howe 2015, 39–40,
66). Greater levels of social cohesion, collective happi-
ness and solidarity – not chaos and antagonism – are
viewed as eventual likely outcomes to a confusing
present. Complacency can result from the Romantic
genre’s cultural logic. It can also come from mixed
messages about genre that oer no clear guidance
on how to think about the problem of climate change.
This situation is identied by Smith and Howe (2015),
53) as one of ‘genre confusion, a confusion that has
blocked it [climate change, TvD and PS] from becom-
ing a full-edged, universal social drama that would
compel decisive public action and institutional reform’.
Published in 2015, Smith and Howe’s claims about
stunted apocalypticism and genre confusion might
now be seen as in need of test or challenge.
2T. VAN DOOREMALEN AND P. SMITH

A YouGov survey for the British Met Oce in 2023
showed 76% of Britons saw climate change playing
a role in extreme weather, and 59% said that they
were attempting to make sustainable decisions in
their personal lives (Met Oce 2023b). Indeed, there
appears to have been a signicant uptick in climate
concern just after 2015 (Barasi 2019), making the tim-
ing of Smith and Howe’s book unfortunate. The ‘heat-
wave’ oers a dened location for bench testing the
genre model’s viability. It is also a location to explore
the local constellations of meaning specic to
a particular experience of climate/climate change
rather than to abstractions about climate change ‘in
general’. Lastly we note that we can illustrate how the
genre-centered investigation of the heatwave (or other
dened anomalous events) is an exportable methodo-
logical resource for a generalizing social science of
climate change culture.
But what are we to predict would be the granular
indicators of genre in this case? Specied to the event
of British heatwave discourse, the Romantic genre
might be expected to generate talk of luck and festiv-
ity; indications of sensory embodied pleasures (food,
drinks, swimming, sunbathing, beach) and upward
comparisons to locations associated with glamourous
lifestyles. In contrast, the Apocalyptic genre argues
that we are heading for disaster. Here there would be
mentions of death and destruction, the breakdown of
technological systems and infrastructure, warnings of
doom, and downward comparisons to locations asso-
ciated with poverty, chaos and misery. Associated with
this we might sometimes nd calls for urgent action,
explications of the need for personal and collective
sacrices, demands for accelerated and radical techno-
logical innovation. There is a sense that ‘business as
usual’ is no longer an appropriate response.
Mapping British heatwave discourse: varying
expectations
Informed by poll data (Barasi 2019) our hypothesis,
contra Smith and Howe (2015), is that a growing
awareness of climate change has seen themes asso-
ciated with the Apocalyptic genre start to rmly
replace those of the Romantic genre. This hypothesis
aligns with claims in the wider literature on climate
sentiment. Indeed, scholars have stated that current
climate change framing in many locations across the
globe is one of ‘climate emergency’ (McHugh, Lemos,
and Morrison 2021). Or that ‘climate projects’ (Araos,
Bhardwaj, and Klinenberg 2024) have taken shape on
various societal levels (e.g. Green Deals on the state
level, individual citizens aiming for envirnmontal
change on the community level), indicating
a growing acknowledgment of the problem. Rather
than being seen as a happy blip in a typically wet
and cold year, we expect that the heatwave will
increasingly be framed as a sign of doom, even in
Britain. We write ‘even’ here, because given the glor-
ious status the heatwave has traditionally had in the
country the Apocalyptic genre is facing a much bigger
challenge than in locations where hot weather is con-
sidered part of a normal summer, such as Southern
Europe or California. Indeed, we believe our paper
takes on a least likely case (Gerring 2007) for such
a genre shift.
Scholars coming from a diversity of disciplines have
studied the framing of British heatwaves before. For
instance, Carvalho and Burgess (2005) made an analy-
sis of the framing of climate change in the British press,
including some heatwave cases, between 1985 and
2003. They conclude that actors quoted in the press,
over the years, were able to ‘socially learn’ from science
and policy-making in relation to handling climate
change. Hopke (2020) compared the mentioning of
climate change in heatwave coverage, among others
in the UK, between 2013 and 2018, and show that they
were on the rise. By contrast Batziou (2022) analyzed
the framing of a big heatwave in 2018, and indicates
that links with climate change were hardly present.
Finally, O’Neill et al. (2023) investigated newspaper
visuals in the newspaper coverage of a 2019 case
(which also hit Britain) and claim that, given the sever-
ity of the climate crisis, those were too much focused
on ‘fun’.
These last two publications invite us to formulate an
alternative hypothesis to the one we just outlined,
namely that not much has changed in the framing of
British heatwaves, and that increasing knowledge
about the dangers of climate change has not resulted
in a rise of the Apocalyptic genre. Such an hypothesis
aligns with claims made in the wider social science
debate about climate change. For instance, Brulle and
Norgaard (2019) argue that the current state of aairs
is (still) one of ‘social inertia’, in order to ‘avoid the
cultural trauma’ that humanity is confronted with.
Bowden, Nyberg, and Wright (2021), Bruggemann
and Engesser (2017), and Merkley and Stecula (2021),
among many others, draw attention to the persistence
of voices of denial and forms of ‘avoidance’ in their
empirical studies of routine media reporting. The ten-
dency to challenge, downplay, pivot or ignore the
diasterous implications of climate change whenever
events trigger news items has also been noted by
journalists themselves. For example, Hartsgaard and
Pope (2021) suggest their colleages are pulling their
punches thanks to timid news room norms, and so are
failing to communicate that ‘civilization is literally
under attack’.
We believe that we are better positioned to test
these varying hypotheses than the heatwave studies
we just summarized, because we add: 1) a focus that
considers the representation of heatwaves not only in
relation to climate change but also to broader issues
ENVIRONMENTAL SOCIOLOGY 3

such as lifestyle; 2) a longitudinal perspective, span-
ning a few decades, which enables considering heat-
wave discourse from the moment climate change
became better known as a public problem, in the
1980s (Smith and Howe 2015), until today. This enables
pointing out when exactly a possible genre shift might
have occurred.
Data
As data we use British national newspaper articles from
the last four decades (January 1985 – August 2023)
with references to heatwaves. National newspapers
are central public sphere spaces, where public elites
(journalists, politicians, scientists) express their views
on societal and political matters and that guide the
larger public in their opinions on these matters (Bröer
and Duyvendak 2009; Koopmans and Olzak 2004). This
makes them indicative of dominant political and soci-
etal frames concerning climate change. Of course, the
rise of social media might have reduced their impor-
tance (e.g. Gilardi et al. 2022; Stewart and Hartmann
2020). However, given our aim to compare develop-
ments from the 1980s until today, newspapers are by
far the best pick. We include newspaper articles men-
tioning both domestic and foreign cases of heatwaves,
because as our genre model explains, the situation
abroad is often 'domesticated' (van Dooremalen and
Duyvendak 2025) as either a positive or negative mir-
ror or prediction of the situation at home (‘just like Italy
we are basking in this European heatwave’ vs. ‘in
Florida people are dying because of the heat, accord-
ing to science similar heat will become normal in
Britain in just two decades’). Quarantined discussions
of just the British situation are rare and heatwave
interpretation cannot be easily ringfenced through
any narrow methodological nationalism.
The online database NexisUni covers the content of
many British newspapers since the early 1980s. This
enables mapping if and how talk of climate change has
changed over the course of the last decades, from the
moment it became a topic of public concern (Smith
and Howe 2015).
In our analysis, we include a wide variety of British
national newspapers, both ‘quality’ broadsheet publica-
tions (The Sunday Times, The Times, The Daily Telegraph,
The Sunday Telegraph, The Observer, The Guardian, The
Independent, Financial Times) as well as ‘popular’ tabloids
(The Sun, Daily Mail, Sunday Express, Daily Mirror, Daily
Star, Daily Star Sunday, Morning Star). Together with
a research assistant, we downloaded all articles from
these newspapers with references to heatwaves (search
terms: ‘heatwaves’ and ‘heatwave’) for the years in which
the newspapers are available in NexisUni (see Appendix
A). This presented us with a dataset of 35.127 articles.
As Appendix A indicates, not all newspapers are
available for our full period of analysis. For instance,
tabloid newspapers are only included from the 1990s
onwards. This might create possible bias. However, we
tested for this and our results seem to be generalizable
across newspapers of dierent types and political lean-
ings throughout the decades (see results section).
Appendix A also shows why we chose to let our ana-
lysis advance specically in the year 1985: from then
onwards, more than 2 British newspapers are available
in NexisUni.
Methods
Our data analysis consists of three phases. In the rst
phase, in order to measure the prominence of heat-
waves in British public discourse, we indicate for four
newspapers from our data set how many articles men-
tioning heatwaves they published throughout our per-
iod of analysis (Cf. Ungar 1999).
In the second phase, we use the automated content
analysis method topic modeling, more specically
Latent Dirichlet Allocation (LDA), to map large scale
patterns in the framing of the heatwaves. Topic mod-
eling is par excellence a method to explore the pre-
sence of genres in large amounts of texts: it indicates
‘bags of words’ that often show up together in the
texts, which represent the dominant themes that are
visible in a corpus (DiMaggio, Nag, and Blei 2013;
Maier et al. 2018). Importantly, it can be a control on
wild or idiosyncratic interpretation.
Finally, in the third phase of analysis we engage in
a more traditional hermeneutic analysis of some of the
key topic-modeling ndings. Qualitative researchers and
cultural theorists have indicated how ‘coding’ and ‘count-
ing’ are often misleading and generate a dangerous
veneer of positivist certainty (Biernacki 2014). They
argue we need to immerse ourselves in the data to see
what is going on. Here we do this to provide a more thick
description and to show how a ‘bag of words’ or set of
themes organized in a genre can take on dierent inec-
tions in each period. Our intention with this mixed meth-
ods approach is to present a resource that is both
systematic and deep and that oers a methodological
prototype for future scholarship.
Analysis
Mapping heatwave mentioning
Given that attention for heatwaves might dier accord-
ing to a newspaper’s genre and political leaning, for
the rst phase of analysis, we selected four of them
that dier on both features:
- broadsheet, leftist The Guardian;
- broadsheet, rightist The Times;
- tabloid, leftist Daily Mirror;
- and tabloid, rightist Daily Mail.
4T. VAN DOOREMALEN AND P. SMITH

This allows us to see whether attention trends for
heatwaves are general or specic per genre or political
leaning. For instance, we could imagine that leftist
newspapers for environmental reasons are more
prone to report on them than rightist ones or that
tabloids heavily write about them when the tempera-
ture allows for fun activities (so, when it is not too
warm). Because newspaper sizes can dier consider-
ably both between and within cases (over time), we
constructed a control measure, by calculating the size
of the papers for each year of the analysis. We did this
through looking up how many newspapers articles
were available for them in NexisUni during eight ran-
domly chosen weeks per year. The numbers that we
use further on (in Figure 1) include this control mea-
sure, which for any given year consists of: for the news-
papers that did not have the smallest size, dividing
their number of articles about heatwaves by the num-
ber of times they were larger than the smallest
newspaper.
Mapping genre developments
For the second phase, we want to measure the entire
heatwave discourse, thus including all articles from our
dataset (all 35.127 together). This involved much more
work than the rst phase. All of the steps of this
analysis, up to the coding part, were done in R.
First, we gured that given that there were far more
articles published about heatwaves in the last decade
of our analysis (2014–2023) compared to the rst one
(1985–1989) it could be problematic to run topic mod-
els for our research period as a whole (see Figure 1).
That might result in an overrepresentation of topics
stemming from this last decade (if they indeed dier
between the decades), which would not give us so
much details about what happened in the period of
1985–1989.
Next, given that in some individual years one spe-
cic heatwave caught a lot of newspaper attention (e.g.
the 1995 case, which made for the warmest British
weeks since 1976) and we want to map the general
framing about heatwaves, we also gured that making
individual corpora per year would be problematic. That
is why we settled for making 8 sub corpora, of periods
of 5 years (4 years, for the 2020–2023 period):
1985–1989; 1990–1994; 1995–1999; 2000–2004;
2005–2009; 2010–2014; 2015–2019; 2020–2023.
Others have used topic modeling before for such
a longitudinal research design (van Dooremalen and
Uitermark 2021). Choosing this approach implies that
for each of the 8 periods, we are dealing with dierent
topics, containing dierent words, that we then want
to capture under the same codes. To overcome this
problem, we developed a coding scheme, that we
could apply to all the topics, which we nally checked
for intercoder reliability (see below).
In order to prepare these eight sub corpora for
a proper topic modeling investigation, we took several
steps (all by using the R package Quanteda (Benoit
et al. 2018)): we tokenized them, removed common
English words, and trimmed the remaining words,
mainly to delete words that only showed up a few
times or in a few texts (Maier et al. 2018, 100–102).
Next, we ran topic models for our each of the sub
corpora (by using the R package LDA). Given that the
corpora diered a lot in size, we experimented with
dierent topic numbers for each period, ranging from
5 to 100, in order to search for the most optimal topic
coherence. This method of trial and error is
a commonly accepted way of doing this (Maier et al.
2018). In the end, our topic numbers ranged from 20
for the rst (1985–1989) up to 50 for the last period of
analysis (2020–2023). In total, our dataset consists of
255 topics.
We then went through all the topics and gured that it
would be informative to give two codes to each topic
(see DiMaggio, Nag, and Blei (2013) and van Dooremalen
and Uitermark (2021) for categorizing topics derived from
newspaper data on several dimensions):
(1) Indicating whether it belonged to the
Romantic or the Apocalyptic genre of heatwave
discourse. The Romantic genre included
topics – see section ‘Genres of Climate
Discourse’ above – where heatwaves were pre-
sented as something fun, exciting, liminal,
oering opportunities for the body et cetera.
The Apocalyptic genre consisted of topics in
which heatwaves were framed as dangerous,
costly, and disruptive, or pointing towards
a problem that should be solved (e.g. the cli-
mate crisis). The topics framing heatwaves in
a substantive way, yet one that was not clearly
relatable to one of the two genres, should be
coded Neutral. The ones that could not, should
be coded as Unclear.
(2) Expressing a more specic heatwave subject
that was being raised in the topic, such as ‘heat-
wave as a reason to go to beach’ or ‘heatwave as
a sign of climate change’. We gured that most
of the topics could be included under the fol-
lowing specic subjects:
●Romantic: Travel/Holidays; Going to the beach;
Drinking wine/enjoying nice food; Summer
clothing; Positive economic eects; Going out/
to festivals.
●Apocalyptic: Climate change; Death/Physical
trouble; Natural disasters; Infrastructural pro-
blems; Economic trouble.
●Neutral: Continuation of ordinary life; Sport
events; Political events; Gardening; Weather
reports; Economic situation.
ENVIRONMENTAL SOCIOLOGY 5

The rst author then coded all the 255 topics by
using these two coding schemes. A student assis-
tant independently coded a sub sample, as
a reliability check. The intercoder reliability score
(Krippendor 2013, 277–287) for the genre coding
was 0.91 and for the subject coding it was 0.82.
These are a very good and a good score, respec-
tively, indicating that our measurements of both
types of coding schemes are reliable. Thus, even
though their exact words dier, we are dealing
with topics containing similar contents in the var-
ious sub corpora, that are comparable through time.
Appendix B provides three examples of topics from
our analysis and their codings.
Hermeneutic Analysis
A basic lesson of structuralism and poststructuralism
is that meanings do not arrive in society as jumbled
bundles of nouns, adjectives and verbs. Rather they
have a syntax and are strung together in grammars
and rhetorics that amount to a patterned discourse.
Only with a close reading could we really see how
the elementary ‘bags of words’ collected in our quan-
titative work clustered together in coherent packages
of meanings (Cf. Biernacki 2014) and how there
might be subtle shifts in the ways the genres were
playing out over the decades. To this end, we ana-
lyzed representative articles from our topic-modeling
results. We selected them by making a probability
check in R, which measures the probability that a text
contains a specic topic (Maier et al. 2018, 111). We
went for texts with a probability score of 0.85 or
higher on topics belonging to those specic heat-
wave subjects that came out as dominant ones in
our ndings throughout the years, such as Beach and
Climate change (see Results section 2). We found that
each time at least 20 newspaper articles of our topics
of interest had this probability score. If there were
more, then we went for the 20 articles with the
highest score.
Results
Increasing (and then exploding) mentionings
Figure 1 indicates the development over time in the
references to heatwaves as made in four British
newspapers, with a control for newspaper size. It
presents some clear patterns. First of all, it is very
obvious that for all four newspapers there was a very
steep increase in mentions over the years. Second,
this increase is not per se linear. It comes with small
shocks up until 2016, at which point the references
really start exploding in a consistent way (perhaps
with the exception of the Daily Mirror). Third, in the
early decades of our sample the density of references
does not dier that much between the newspapers.
However, in the period of explosion, the two broad-
sheets (The Guardian and The Times) start to routinely
mention heatwaves more often than their tabloid
counter parts (albeit the Daily Mail catches up again
with The Times in the last year of analysis). This might
imply that the timing of a shift in discourse is led
mostly by the broadsheet newspapers (see also the
end of the next section). Most importantly, Figure 1
simply documents that the heatwave is a salient
theme in public discourse throughout the last dec-
ades, not an obscure residual category or discursive
outlier which newspapers only write about once or
twice a year and thus presents problems for validity
in inquiry.
From genre confusion to apocalyptic dominance
The key hypothesis we aim to test in this article is
whether a genre shift has occurred in British heatwave
discourse. Figure 2 shows the percentages of topics
belonging to the categories Romantic, Apocalyptic and
0
100
200
300
400
500
600
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
The
Guardian
The Times
Daily Mirror
Daily Mail
Figure 1. Development of heatwave references in four British newspapers (1984–2023).
6T. VAN DOOREMALEN AND P. SMITH

Neutral per 5 year timeslot. It indicates that a genre
shift has indeed taken place. The Romantic genre has
decreased over the years, whereas the Apocalyptic one
has increased.
Including the category ‘Neutral’ in the gure does
not help much for our research purposes. Therefore,
in Figure 3, only the Romantic and Apocalyptic gen-
res are included, and this makes for a clearer over-
view. We can see that in the rst three periods of
analysis (1985–1999) a situation of genre confusion,
as described by Smith and Howe (2015), still exists.
Both genres are more or less equally present. This
could be seen as the rst phase of British heatwave
discourse of the last decades.
In a second phase (2000–2014) the Apocalyptic
genre increases, whereas the Romantic one decreases,
albeit the dierences between the two are not that big
yet. There is still genre confusion. However, in the last
phase (2015–2023) the Apocalyptic genre totally starts
to dominate. In the nal period of analysis (2020–2023)
it is more than 4 times bigger than the Romantic one.
Thus, we can conrm our hypothesis: a genre shift has
occurred in British heatwave discrouse – from genre
confusion to the dominance of the Apocalyptic genre.
Figures 4 and 5 give more depth to these developments.
They present the most salient specic subjects that are
coupled to heatwaves within the two genres. From
Figure 4 it becomes clear that the rise of the Apocalyptic
genre is mainly the result of the rise of heatwave dis-
course in relation to climate change. During the phase
in which this genre is very dominant (2014–2023), the
percentages for the topics on natural disasters and death/
physical trouble do not go up that much anymore (or
even slightly go down). The climate change topics, on the
other hand, increase a lot, especially in the last period
(2020–2023, an increase of almost 10% of the entire con-
tent to the period before). This suggests that the rise of
the Apocalyptic genre is not so much due to heatwaves
being more often seen through their direct negative
consequences (ending in people suering and/or the
events being considered natural disasters). It is due to
increasingly seeing them as part of the bigger, often
future related, problem of climate change.
For the specic subjects within the Romantic genre,
we nd that the two most salient ones are heatwaves
in connection to Beach and Travel/Holidays. Both have
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Apocalyptic
Neutral
Romantic
1985-1989
1990-1994
1995-1999
2000-2004
2005-2009
2010-2014
2015-2019
2020-2023
Figure 2. Development of genres in British heatwave talk (1985–2023).
0%
10%
20%
30%
40%
50%
60%
Apocalyptic
Romantic
1985-1989
1990-1994
1995-1999
2000-2004
2005-2009
2010-2014
2015-2019
2020-2023
Figure 3. Developments for Apocalyptic and Romantic genres (1985–2023).
ENVIRONMENTAL SOCIOLOGY 7

gone down in visibility (Figure 5), while Beach has over
the years remained the most dominant specic sub-
ject – albeit only with a score of 3% of the entire
newspaper content for the last period of analysis.
It is important to note that our core nding gener-
ally holds for all the newspapers we analyzed. That is:
we checked whether there were biases in heatwave
discourse, depending on broadsheet vs tabloid or poli-
tical leaning. We did so by comparing the scores for the
left leaning, broadsheet The Guardian with the right
leaning, tabloid Daily Mail, as representing two
extremes on the spectrum, for the periods
1995–1999, 2005–2009, and 2020–2023. And the
trend of a visible shift from genre confusion to apoc-
alyptic dominance holds in both cases. Even though
The Guardian has more Apocalyptic than Romantic
topics in each of these periods (at least 8%), it still
brought up Romantic ones rather frequently in each
of them too, even in the years 2020–2023 (10% of their
total content). In the Daily Mail for the period
2020–2023 more than 40% of the content relates to
the Apocalyptic genre, whereas only 16% concerns the
Romantic one.
Whilst this nding might appear surprising, we sug-
gest there might be an elective anity between
Apocalyptic representations and the hyperbole, sensa-
tionalist critique of authority and emotional enlistment
that are the stock in trade of the British tabloid press
(see Boyko (2008) for very similar results during the
2000s). Ignoring climate change might sell fewer
copies than sounding the alarm bell. The examples
given below capture this phenomenon in more detail.
Genre complexities: 1980s and early 1990s versus
2020s
So far we have painted with a broad brush daubed
in quantitative paint. Our large corpus clearly indi-
cates a general shift in coverage. In terms of
Raymond Williams’ framework (Williams 1977),
Apocalyptic representations of climate change
moved from emergent to dominant status, and
Romantic representations became residual. The bal-
ance of power in representation has been recon-
gured over just four decades. But there is more
going on than this simple, if compelling, story
0%
5%
10%
15%
20%
25%
30%
35%
Deaths/physical
trouble
Climate change
Natural disasters
1985-1989
1990-1994
1995-1999
2000-2004
2005-2009
2010-2014
2015-2019
2020-2023
Figure 4. Developments for Apocalyptic subjects (1985–2023).
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
Beach
Travel/holidays
1985-1989
1990-1994
1995-1999
2000-2004
2005-2009
2010-2014
2015-2019
2020-2023
Figure 5. Developments for Romantic subjects (1985–2023).
8T. VAN DOOREMALEN AND P. SMITH

suggests. As Richard Biernacki (2014) indicates, we
need to be attentive to the contours of the items
making up the sampling frame rather than assume
a homogenous set.
Let’s dive into this complexity with reference to
the extreme ends of our sampling period:
1985–1994, when there was still genre confusion, vs
2020–2023, when Apocalypticism dominates. For
each of these periods, we analyzed newspaper arti-
cles that had a very high probability score (see
Analysis section 3) for the specic subjects that
came out as dominant ones from our topic-
modeling analysis (see Figures 4 and 5): Beach and
Travel/Holidays for the Romantic genre and Climate
Change, Death/Physical Trouble and Natural
Disasters for the Apocalyptic one.
Romanticism back then: pleasures with problems
Even in an era when ‘ice cream’ storytelling was
common we see discussions of immediate, practical
problems connected to heat that hint at possibilities
for greater social disorganization, yet do not t easily
into either a Romantic or Apocalyptic rubric. These
were frequently woven into newspaper articles that
tripped our Romantic genre coding, appearing as
a ‘sting in the tail’. What makes holidays, beach
trips and barbeques newsworthy, after all, is not
only that people will enjoy them but rather that
they generate challenges for individuals and indus-
tries to make the right choices. For example, a travel
report in The Sunday Times noted that a ‘heatwave
ve days’ might make it essential for readers to
‘avoid the swarms of holidaymakers buzzing around
honeypot tourist traps’ (Binns 1989). So genre confu-
sion is not always a case of some media items and
authors using a romantic framing and others choos-
ing to highlight problems. It can take place even
within a single article as it attempts to map out
positive and negative aspects of a newsworthy event.
In the same vein, while hot summers oered
retailers a ‘boost of up to 20% in sales of some
summer related products’ such as ice cream and
cold beverages there were problems for ‘marketing
teams looking for stock guidance’. This was because,
according to the Meteorological Oce’s Gordon
Higgins, ‘it is impossible to make accurate predic-
tions any more than a week in advance’ (Rowland
1987). So positive news about any much-awaited hot
spell was routinely tempered in the same by reports
of trac jams, parking problems, supply shortages,
hose pipe bans and so forth. These are annoying,
manageable, low order, and transient problems com-
pared to global catastrophe, but problems none the
less. They do not sit neatly with our binary genre
analysis.
Apocalypticism back then: possible future risks
Turning to more clearly Apocalyptic items in this earlier
period, we found a surprising interest in climate
science and climate modelling. Heatwaves, it seems,
oered a chance to bring this esoteric matter with its
imagery of doom to the table even four decades ago.
But what is dierent is the existential mood. Compared
to today we see a lack of urgency and a continual
emphasis on uncertainty. Climate science stories of
the time liked to dwell on the signal to noise diculties
of proving climate change was or would take place. ‘As
the nation enjoys a May heatwave it seems that hardly
a day passes without some weather record being sur-
passed’ wrote science correspondent Steve Connor
(1990) in The Independent. True to Apocalypticism
Connor conjures imagery of ‘melting glaciers and
polar ice sheets. That will result in sea levels rising up
to 75 cm, inundating vast tracts of land’. Yet these
horrors he says will take place in the distant future,
and the theme of the article is in fact that it is hard to
know the truth: ‘scientists say it is too early to worry
about a greenhouse eect . . . they need at least 30
years to detect long term changes’. Indeed, Connor
even comments that ‘there are few indications Britain
is undergoing long term climate change’.
Likewise, an item in The Guardian on climate change
from February 1989 starts by noting that ‘the last roses
in summer look like hanging on until March and the
lavender in seaside Sussex hasn’t stopped owering
either’. These sound like positive things associated
with the Romantic genre, but they are just a stylistic
prelude to an extended prognostication on extreme
weather events, possible switches in the Gulf Stream
and rising sea levels. The bulk of the article is given
over to discussing the slow nature of climate change,
the existence of random uctuations in the weather,
and the consequent diculty of causal attribution: ‘ . . .
it isn’t the greenhouse eect at work. Or rather, if it is,
no meteorologist, climate modeler or professional
weather watcher will say so’ (Radford 1989).
In such reportage from three and a half decades ago
the Apocalyptic dangers are adumbrated or known.
But for their authors they exist decades or centuries
away. All the right words or ‘topics’ are in place, ready
to be captured by our topic models. But the
Apocalypticism of this era tips towards a ‘wait and
see’ action pattern and it sits alongside a Romantic
(‘ice cream’) discourse that notes irritations to everyday
life as well as pleasures.
Romanticism Today: enduring pleasures
Just as coverage in the past can look somewhat like
that of today, the converse is also possible. And so it is
that although becoming residual (Williams 1977),
Romantic genres of reportage remain visible in our
ENVIRONMENTAL SOCIOLOGY 9

sample. Heatwaves continue to be a subject for cele-
bration and positive comment regarding upcoming
embodied pleasures, especially in tabloids. The Daily
Star, for example, exuded a sense of jubilation as it
reported England would be hitting the mid-20s in
May 2022 thus becoming, albeit briey, ‘hotter than
Mexico’. ‘Parts of the capital' it continued ‘hit 23.3C
yesterday, a temperature far higher than the average
of 16.2C for the month in England. Supermarkets are
ready to scramble for a burgers-and-beers rush as Brits
ock to beaches and parks to enjoy the hot weather’
(Daily Star 2022).
Translating this modality upscale, The Guardian’s
travel section suggested its readers enjoy the privilege
of wild swimming in upscale and bucolic settings:
This summer’s heatwave turned the usually formal
surrounds of Derbyshire’s grand Chatsworth House
into an almost tropical setting, with people leaping
from trees and rope swings into the River Derwent, or
wading from shallow beaches to admire the splendor
of one of England’s most magnicent houses from the
water. On the deeper side of the river are the sloped
lawns of the estate; on the other, open parkland where
deer and sheep trot down between oak trees to the
water’s edge. (Bowes 2020)
This travel and leisure item in this ecologically aware
broadsheet has no mention of the irony that the
growth in the popularity of wild swimming in Britain
is facilitated by planetary disaster.
A proximate mechanism that explains these conti-
nuities is probably the sticky, deeply institutionalized
professional codes and production routines shaping
the newspaper articles that make up our sample. The
ndings of foundational studies on ‘producing the
news’ (Gans 2004; Hall 1980) showed that pragmatic
conventions determined what was ‘newsworthy’ and
how it was packaged and presented. We know this is
even the case for science journalism (Dunwoody 2008).
And the classical templates continue to remain valid in
the digital age (Lecheler and Kruikemeier 2016).
On top of all this, production codes, writerly con-
ventions and civilizing missions vary over fragments of
the journalistic eld, creating a dierentiated story
feed that further facilitates mixed messaging. It is all
too easy to forget that newspapers don’t only print
‘news’ or undertake ‘serious’ investigations. Business,
travel, food, and gardening features also have their
tropes and editorial understandings of what is ‘news-
worthy’ when there is ‘heat’ around. In the case of
lifestyle pages, particularly, we see a pleasure-driven
mandate to serve up vicarious consumption possibili-
ties, narrate a solidaristic collective identity and pro-
vide optimistic commentary on cultural trends
(Hanusch 2010). This might explain why O’Neill et al.
(2023) found so many ‘fun’ projections of heatwaves
when they made an analysis of newspaper coverage of
a case from 2019: the press probably mainly uses
pictures to present its romantic rather than its apoca-
lyptic sides, and product dierentiation sees lifestyle
items treat the heatwave as an opportunity, not
a threat.
Apocalypticism Today: present dangers
Where we most clearly nd a new sensibility is in
core apocalyptic material involving discussions of
climate science, climate change knowledge and
climate doom that are stimulated by the arrival of
a heatwave. Much has been made in the literature
of the problems of translating scientic knowledge
to publics, of misinformation, of distrust in expert
systems, and of conrmation bias and other psy-
cho-social mechanisms supporting climate skepti-
cism (Dunlap 2013; Kahan 2015; Smith and Howe
2015). In this context it is surprising to see such
buy-in to the reality of climate change and the
credibility of the science – at least at the level of
the text. Grammatically speaking, climate change
by the 2020s has shifted from the subjunctive to
the indicative mode: from future possibility to pre-
sent fact, from ‘might’ and ‘could’ to ‘is’.
As a shared common sense, it now oers
a capsule explanation for assorted extreme weather
events and the misfortunes they bring. Many items
speak to unparalleled global trends or to other kinds
of extreme weather elsewhere in the world. For
example, a feature in The Times from 2021 adum-
brates a US summer of hurricanes, and oods as
well as wildres. It notes that 64% of Americans
‘live in a county that experienced a multi-day heat-
wave this summer’. Quotes from scientists are then
used to oer context, describing these as ‘a result of
climate change’ and warning that ‘it’s just a small
preview of what’s going to happen if we don’t start
stopping emitting greenhouse gases’ (Charter 2021).
Reports of wildres in Greece, Australia and
California often took on similar properties as carriers
of portents and lessons.
But what of those potentially benecent heat-
waves at home in the UK that would push aside
the damp and chilly summer? Here too we see
a shift. The heatwave of July 2022 saw the UK record
its rst 40C temperature and serves as an exemplar.
Now the dominant theme in the buildup was
admonition and warning, with the connection to
climate change no longer questioned. A story in
The Guardian lists possible disruptions to transport
and schools, urges people to stay indoors, to apply
sunscreen and stay hydrated, and warns of possible
death (Murray 2022). There is a strong sense of the
kind of liminality that comes from social chaos rather
than that of the social carnival of the heatwave trip
to the beach. Whereas back in the 1980s it was
impossible to pin the tail on the donkey of climate
change here we nd in the middle of the story the
10 T. VAN DOOREMALEN AND P. SMITH

‘no quibbles’ explanation for the extreme heat that is
bringing these dangers:
Scientists said the link between climate change and
extreme heatwaves was now clear ‘This shows the UK
is already on a warming trend when it comes to heat
extremes,’ said Dr Mark McCarthy from the Met Oce.
“Human-caused climate change has set us on a course
to see temperature extremes in the UK that would be
highly unlikely under a ‘natural’ climate. (Murray 2022)
Likewise, the tabloid coverage has moved. Pleasures
were now risky, climate change science, once the ques-
tionable realm of those they dubbed ‘bons’ and ‘egg-
heads’, had become an unproblematic body of
knowledge which to add context and build fear. An
item in the Daily Mirror warned of health issues even
for the t and healthy and of ambulance services being
overwhelmed. It goes to indicate societal disorganiza-
tion on a substantial scale ‘a danger of disruption to
transport, food supplies and could see everything from
schools to nuclear power plants being closed’. And the
cause: ‘Heatwaves have been made hotter, longer and
more frequent by climate change, and experts have
warned of the need to adapt homes and cities in the
UK for a future of more intense summer heat’ (Weston
2023). Looking back on the heatwaves a few months
later the same paper reported on grim excess mortality
data from the Oce of National Statistics: ‘Thousands
more people died than expected this summer as
a result of Britain’s baking Saharan temperatures’. It
recalled wildres that 'left villages looking like war-
zones’, hosepipe bans, swimmers in diculty and
stranded trains. Then it quotes scientist Isabel Oliver,
who instructs the reader ‘a warming climate shows we
must adapt to living safely with hotter summers in the
future’ (Craig 2022).
Real change, with overlap and continuities
Our one sentence summary: if Apocalyptic messages
are becoming both more prevalent and less dident,
and Romantic ones more residual, some smudging and
mixed messaging remains. When compared to earlier
analyses of British heatwave discourse, our ndings
conrm those from Carvalho and Burgess (2005) and
Hopke (2020), who found a similar trend – albeit they
studied it for far shorter periods and we have a far
more expansive data set. In reference to the analyses
of Batziou (2022) and O’Neill et al. (2023), we would say
that focusing on a specic heatwave case or the con-
vention-driven use of photographs obscures the wider
transformation that has occurred over the last 4 dec-
ades. Informed by Williams (1977) and highlighting
both a big trend and some muddy waters our research
oers both a replication and an explanation of the
mixed messages to be found in this body of existing
research. Change is real, but we cannot expect it to
happen without overlap and continuity. Because
climate discourse in the media remains layered and
somewhat contradictory, research ndings are likely
to reect this reality. Results are going to be extremely
sensitive to methodological choices about research
design, sampling frames and visions of scope
conditions.
Conclusion
We are now in a position to return to the question
posed in the rst sentence of our Abstract. Heatwave
discourse has indeed become more Apocalyptic and
more often linked to climate change – at least in the
case of British newspapers. In the terms established by
Raymond Williams (1977): what was emergent in the
1980s has become dominant today. The nding holds
even for right-of-center and tabloid newspapers (e.g.
the Daily Telegraph, Daily Mail). Within the wider social
science debate on climate change sentiment, we thus
nd support for the thesis of those scholars arguing
that contemporary climate framing is one of ‘emer-
gency’ (McHugh, Lemos, and Morrison 2021) rather
than of ‘avoiding cultural trauma’ (Brulle and
Norgaard 2019). With regard to timing we show that
the decisive shift has taken place only in the last dec-
ade (Figures 1–4). Hence, Smith and Howe (2015) were
just about right speaking of ‘genre confusion’ in their
2015 book, but they were also shooting at a moving
target.
Other than this central empirical conclusion, we
have made various additional contributions. Our
topic models led to a close look at media materials.
This enabled us to grasp why two contrasting gen-
res can persist at the same time, often in the same
newspaper. By looking to thematic dierentiation in
conventional textual product (e.g. aspirational life-
style pages versus social disruption news) we have
found a mechanism explaining why shifts in heat-
wave discourse are not instantaneous and zero sum.
Importantly we indicate that explicit, argumentative
climate denial (Bowden, Nyberg, and Wright 2021;
Bruggemann and Engesser 2017; Merkley and
Stecula 2021) is not the only game in town when
it comes to accounting for the incomplete triumph
of Apocalypticism. We showed it is often set aside
merely by a change of subject: say from British
hospital admissions to the British wine industry.
Furthermore, we have demonstrated that the genre
model of risk perception is a viable way to organize
climate discourse data and explanation. It oers
a generalizable, parisomonious resource for both the-
ory building and comparative analysis. We suspect our
story will hold up more widely, beyond the least likely
case of Britain (Gerring 2007). We certainly expect to
nd it in zones plagued by excessive heat on a regular
basis (Greece, Spain, California etc.). This hypothesis
could easily be tested with the theory, concepts and
ENVIRONMENTAL SOCIOLOGY 11

methods used here – they are entirely exportable. We
invite eorts at falsication, replication, or
improvement.
Will this shift in dominant discourse result in
major changes in political will, electoral decisions or
personal lifestyle choices? In other words: is it
enough to tackle the climate crisis? This is
a question that is both too political and too big to
answer in this paper. But minimally the shift in genre
is already demanding new forms of accountability
and gestures of concern, as is covered by the con-
cept of ‘climate projects’ (Araos, Bhardwaj, and
Klinenberg 2024). Indeed, we know from research
on discursive opportunity structures (Bröer and
Duyvendak 2009; Koopmans and Olzak 2004), that
what happens in public discourse often has serious
consequences on what citizens as well as politicians
can (not) say and do. That said, we acknowledge talk
is cheap. As Norgaard (2011) observed, denial can
take more slippery forms than the outward rejection
of climate science. A double reality can exist, where
the severity of the climate crisis is acknowledged in
discourse, but not in practice.
A nal stone we left unturned relates to agency.
Where does the genre shift come from? In our empiri-
cal analyses, we paid more attention to the fact that
and when this shift has taken place than why it hap-
pened. But we do want to posit that the growing
frequency and intensities of heatwaves (Beckett and
Sanderson 2022; Rousi et al. 2022) per se cannot fully
account for it. For instance, were these trends being
interpreted in contexts where climate change denial is
still rampant, then they would probably be trivialized
or decoupled from the phenomenon of climate
change. As research has shown, extreme weather
does not ‘naturally’ have an impact on climate senti-
ment, but only when audiences are open to see the
link (Hazlett and Mildenberger 2020; Ogunbode et al.
2019). There is elasticity in the stimulus-response sys-
tem. Our ndings (in particular Figure 4) also suggest
that the immediate negative eects of heatwaves are
not directly responsible for the genre shift. For exam-
ple, in the period that the Apocalyptic genre became
dominant the scores for those topics most closely
related to death and physical troubles did not go up.
However, it would be surprising if recent heat-
wave trends were not in some way a contributing
factor. The heatwave cannot speak for itself, but it
does oer a ‘newsworthy’ opportunity (or encour-
agement?) for humans to talk about what it means.
The time period in which the major genre shift
occurred, the last decade, intersects with an expan-
sion of climate change consciousness: IPCC reports
getting more worrying; the rise of Greta Thunberg as
a public advocate of alarmism; and the spread of the
sometimes playful but often severely disruptive
(hence newsworthy) protests of Extinction
Rebellion across the globe (Gardner, Carvalho, and
Valenstain 2022). And, indeed, Figure 4 clearly shows
that it is a rise of heatwaves being seen as a sign of
climate change that goes hand in hand with the
genre shift.
Indicating how the causality for this shift ows
between nature, cultural tropes, governments and citi-
zens is up for future research. That said, we have at
least shown that climate science and activism are no
longer howling in the wind. ‘Social learning’, as
Carvalho and Burgess (2005) would have it, has really
taken place. Today, fears about climate change dom-
inate British newspaper discourse on what was fre-
quently considered a desirable apotheosis of the
country’s unreliable summer – the heatwave.
Acknowledgements
Earlier versions of this paper were presented at the Annual
Meeting of the American Sociological Association 2023
(Philadelphia), the World Congress of Sociology 2023
(Melbourne), the Dutch/Flemish Sociology Conference 2023
(Ghent), the Cultural Sociology East & West Conference (Yale,
September 2023), the James Coleman Symposium 2024
(Utrecht) and the Northwestern Culture & Society Workshop
(Chicago, October 2024). We would like to thank the organi-
zers of the respective conference panels and seminars for
giving us the opportunity to share our work and the partici-
pants for their feedback.
Disclosure statement
No potential conict of interest was reported by the
author(s).
Funding
The work was supported by the HORIZON EUROPE Marie
Sklodowska-Curie Actions [101066910].
Notes on contributors
Thijs van Dooremalen works as an Assistant Professor at
Leiden University’s Institute of Security and Global Aairs.
He is interested in how events and crises shape social and
political life. The research for this paper was conducted when
he was a Marie Curie Postdoctoral Fellow at KU Leuven’s
Centre for Sociological Research.
Philip Smith is Professor of Sociology and Director of the Yale
Center for Cultural Sociology. He has written widely on social
and cultural theory. Claims in his book with Nicolas Howe
Climate Change as Social Drama are put to the test and
updated in this article.
ORCID
Thijs van Dooremalen http://orcid.org/0000-0001-7184-
775X
Philip Smith http://orcid.org/0000-0002-2164-7773
12 T. VAN DOOREMALEN AND P. SMITH

References
Araos, M., A. Bhardwaj, and E. Klinenberg. 2024. “The Social
Life of Climate Projects.” Sociological Forum 39 (2):
149–152. https://doi.org/10.1111/socf.12995 .
Barasi, L. 2019. “Guest Post: Polls Reveal Surge in Concern in
UK About Climate Change”. Accessed August 15, 2023.
https://www.carbonbrief.org/guest-post-rolls-reveal-
surge-in-concern-in-uk-about-climate-change/ .
Barrera, D. J. 2017. “Drug War Stories and the Philippine
President.” Asian Journal of Criminology 12 (4): 341–359.
https://doi.org/10.1007/s11417-017-9253-x .
Batziou, A. 2022. “Climate Change and the Heatwave:
Searching for the Link in the British Press.” Journalism
Practice 16 (4): 681–701. https://doi.org/10.1080/
17512786.2020.1808515 .
Beckett, A. D., and M. G. Sanderson. 2022. “Analysis of
Historical Heatwaves in the United Kingdom Using
Gridded Temperature Data.” International Journal of
Climatology 42 (1): 453–464. https://doi.org/10.1002/joc.
7253 .
Benoit, K., K. Watanabe, H. Wang, P. Nulty, A. Obeng,
S. Müller, and A. Matsuo. 2018. “Quanteda: An R Package
for the Quantitative Analysis of Textual Data.” Journal of
Open Source Software 3 (30): 774. https://doi.org/10.21105/
joss.00774 .
Biernacki, R. 2014. “Humanist Interpretation versus Coding
Text Samples.” Qualitative Sociology 37 (2): 173–188.
https://doi.org/10.1007/s11133-014-9277-9 .
Binder, W. 2024. “Technology as (Dis)enchantment: AlphaGo
and the Meaning-Making of Articicial Intelligence.”
Cultural Sociology 18 (1): 24–47. https://doi.org/10.1177/
17499755221138720 .
Binns, R. 1989. “Extended Leave.” The Sunday Times, April 2.
Bowden, V., D. Nyberg, and C. Wright. 2021. ““We’re Going
Under”: The Role of Local News Media in Dislocating
Climate Change Adaptation.” Environmental
Communication 15 (5): 625–640. https://doi.org/10.1080/
17524032.2021.1877762 .
Bowes, G. 2020. “Post Splosh.” The Guardian, September 14.
Boyko, M. T. 2008. “The Cultural Politics of Climate Change
Discourse in UK Tabloids.” Political Geography 27 (5):
549–569. https://doi.org/10.1016/j.polgeo.2008.05.002 .
Bröer, C., and J. W. Duyvendak. 2009. “Discursive
Opportunities, Feeling Rules, and the Rise of Protests
Against Aircraft Noise.” Mobilization: An International
Quarterly 14 (3): 337–356. https://doi.org/10.17813/maiq.
14.3.kx87723664558jm4 .
Bruggemann, M., and S. Engesser. 2017. “Beyond False
Balance: How Interpretive Journalism Shapes Media
Coverage of Climate Change.” Global Environmental
Change 42:58–67. https://doi.org/10.1016/j.gloenvcha.
2016.11.004 .
Brulle, R. J., and K. M. Norgaard. 2019. “Avoiding Cultural
Trauma: Climate Change and Social Inertia.”
Environmental Politics 28 (5): 886–908. https://doi.org/10.
1080/09644016.2018.1562138 .
Carvalho, A., and J. Burgess. 2005. “Cultural Circuits of Climate
Change in U.K. Broadsheet Newspapers, 1985–2003.” Risk
Analysis: An International Journal 25 (6): 1457–1469. https://
doi.org/10.1111/j.1539-6924.2005.00692.x .
Charter, D. 2021. “Fires, Drought or Hurricanes Plague a Third
of America.” The Times, September 6.
Colls, R. 2002. Identity of England. Oxford: Oxford Univeristy
Press.
Connor, S. 1990. “Turtle Tourists May Mark Climate Shift.” The
Independent, May 6.
Craig, E. 2022. “Summer 2022 Saw Britain’s Deadliest Ever
Heatwave.” Daily Mail, October 7.
Daily Star. 2022. “Now We’re Cooking.” May 7.
DiMaggio, P., M. Nag, and D. Blei. 2013. “Exploiting Anities
Between Topic Modeling and the Sociological Perspective
on Culture: Application to Newspaper Coverage of U.S.
Government Arts Funding.” Poetics 41 (6): 570–606.
https://doi.org/10.1016/j.poetic.2013.08.004 .
Dunlap, R. E. 2013. “Climate Change Skepticism and Denial:
An Introduction.” The American Behavioral Scientist 57 (6):
691–698. https://doi.org/10.1177/0002764213477097 .
Dunwoody, S. 2008. “Science Journalism.” In Handbook of
Public Communication of Science and Technology, edited
by M. Bucchi and B. Tench, pp. 15–26. London: Routledge.
Frye, N. 1957. Anatomy of Criticism. Princeton, NJ: Princeton
University Press.
Gans, H. J. 2004. Deciding What’s News. Evanston:
Northwestern University Press.
Gardner, P., T. Carvalho, and M. Valenstain. 2022. “Spreading
Rebellion?: The Rise of Extinction Rebellion Chapters
Across the World.” Environmental Sociology 8 (4):
424–435. https://doi.org/10.1080/23251042.2022.2094995 .
Gerring, J. 2007. “Is There a (Viable) Crucial-Case Method?”
Comparative Political Studies 40 (3): 231–253. https://doi.
org/10.1177/0010414006290784 .
Gilardi, F., T. Gessler, M. Kubli, and S. Müller. 2022. “Social
Media and Political Agenda Setting.” Political
Communication 39 (1): 39–60. https://doi.org/10.1080/
10584609.2021.1910390 .
Golinski, J. 2007. British Weather and the Climate of
Enlightenment. Chicago: University of Chicago Press.
Hall, S. 1980. Encoding-Decoding. London: Routledge.
Hanusch, F. 2010. “The Dimensions of Travel Journalism.”
Journalism Studies 11 (1): 68–82. https://doi.org/10.1080/
14616700903290569 .
Harley, T. A. 2021. “Nice Weather for the Time of Year: The
British Obsession with the Weather.” In Weather, Climate,
Culture, edited by S. Strauss and B. Orlove, pp. 103–118.
Abingdon: Routledge.
Hartsgaard, M., and K. Pope. 2021. “The Media is Still Mostly
Failing to Convey the Urgency of the Climate Crisis.” The
Guardian, June 3. Accessed November 20, 2024. https://
www.theguardian.com/commentisfree/2021/jun/03/
media-climate-change-crisis-emergency .
Hazlett, C., and M. Mildenberger. 2020. “Wildre Exposure
Increases Pro-Environment Voting within Democratic but
Not Republican Areas.” The American Political Science
Review 114 (4): 1359–1365. https://doi.org/10.1017/
S0003055420000441 .
Hopke, J. E. 2020. “Connecting Extreme Heat Events to
Climate Change: Media Coverage of Heat Waves and
Wildres.” Environmental Communication 14 (4): 492–508.
https://doi.org/10.1080/17524032.2019.1687537 .
Kahan, D. M. 2015. “Climate-Science Communication and the
Measurement Problem.” Advances in Political Psychology
36 (S1): 1–43. https://doi.org/10.1111/pops.12244 .
Konisky, D. M., L. Hughes, and C. H. Kaylor. 2016. “Extreme
Weather Events and Climate Change Concern.” Climatic
Change 134 (4): 533–547. https://doi.org/10.1007/s10584-
015-1555-3 .
Koopmans, R., and S. Olzak. 2004. “Discursive Opportunities
and the Evolution of Right-Wing Violence in Germany.”
The American Journal of Sociology 110 (1): 198–230.
https://doi.org/10.1086/386271 .
Krippendor, K. 2013. Content Analysis: An Introduction to Its
Methodology. Los Angeles: Sage Publications.
ENVIRONMENTAL SOCIOLOGY 13

Lecheler, S., and S. Kruikemeier. 2016. “Re-Evaluating Journalistic
Routines in a Digital Age.” New Media and Society 18 (1):
156–171. https://doi.org/10.1177/1461444815600412 .
Maier, D. A., P. Waldherr, G. Miltner, A. Wiedemann, A. Niekler,
B. Keinert, B. Pfetsch, et al. 2018. “Applying LDA Topic
Modeling in Communication Research: Toward a Valid
and Reliable Methodology.” Communication Methods and
Measures 12 (2–3): 93–118. https://doi.org/10.1080/
19312458.2018.1430754 .
McHugh, L. H., M. C. Lemos, and T. H. Morrison. 2021. “Risk?
Crisis? Emergency? Implications of the New Climate
Emergency Framing for Governance and Policy.” Wiley
Interdisciplinary Reviews: Climate Change 12 (6): e736.
https://doi.org/10.1002/wcc.736 .
Merkley, E., and D. A. Stecula. 2021. “Party Cues in the News:
Democratic Elites, Republican Backlash, and the Dynamics of
Climate Skepticism.” British Journal of Political Science 51 (4):
1439–1456. https://doi.org/10.1017/S0007123420000113 .
Met Oce. 2022. “Hadley Center Central England
Temperature Dataset (HADSET)”. Accessed February 15,
2023. https://www.metoce.gov.uk/hadobs/hadcet/ .
Met Oce. 2023a. “What is a Heatwave?”. Accessed May 25,
2023. https://www.metoce.gov.uk/weather/learn-about
/weather/types-ofweather/temperature/heatwave .
Met Oce. 2023b. “Over Half of Public Making Low Carbon
Decisions”. Accessed August 15, 2023. https://www.metof
ce.gov.uk/about-us/press-oce/news/weather-and-
climate/2023/over-half-of-british-public-say-they-are-
making-conscious-decisions-to-live-a-low-carbon-lifestyle
-and-they-want-to-do-more .
Morgan, M. 2020. “Why Meaning-Making Matters: The Case of
the UK Government’s COVID-19 Response.” American
Journal of Cultural Sociology 8 (3): 270–323. https://doi.
org/10.1057/s41290-020-00121-y .
Murray, J. 2022. “UK Braces for Record Temperature as First Ever
Red Heat Warning Comes into Eect.” The Guardian, July 18.
Norgaard, K. M. 2011. Living in Denial: Climate Change,
Emotions, and Everyday Life. Cambridge, MA: MIT Press.
Ogunbode, C. A., C. Demski, S. B. Capstick, and R. G. Sposato.
2019. “Attribution Matters: Revisiting the Link Between
Extreme Weather Experience and Climate Change
Mitigation Responses.” Global Environmental Change
54:31–39. https://doi.org/10.1016/j.gloenvcha.2018.11.005 .
O’Neill, S., S. Hayes, N. Strauβ, M. N. Doutreix, K. Steentjes,
J. Ettinger, N. Westwood, and J. Painter. 2023. “Visual
Portrayals of Fun in the Sun in European News Outlets
Misrepresent Heatwave Risks.” The Geographical Journal
189 (1): 90–103. https://doi.org/10.1111/geoj.12487 .
Radford, T. 1989. “Don’t Bank of Bananas in Bognor.” The
Guardian, February 9.
Rousi, E., K. Kornhuber, G. Beobide-Arsuaga, F. Luo, and
D. Coumou. 2022. “Accelerated Western European
Heatwave Trends Linked to More-Persistent Double Jets
Over Eurasia.” Nature Communications 13 (1): 3851. https://
doi.org/10.1038/s41467-022-31432-y .
Rowland, S. 1987. “Retailers Hope for Long Hot Summer.” The
Financial Times, May 7.
Sisco, M. R., V. Bosetti, and E. U. Weber. 2017. “When Do
Extreme Weather Events Generate Attention to Climate
Change?” Climatic Change 143 (1–2): 227–241. https://
doi.org/10.1007/s10584-017-1984-2 .
Smith, P., and N. Howe. 2015. Climate Change as Social
Drama: Global Warming in the Public Sphere. Cambridge:
Cambridge University Press.
Stewart, E., and D. Hartmann. 2020. “The New Structural
Transformation of the Public Sphere.” Sociological
Theory 38 (2): 170–191. https://doi.org/10.1177/
0735275120926205 .
Ungar, S. 1999. “Is Strange Weather in the Air? A Study of US
National Network News Coverage of Extreme Weather
Events.” Climatic Change 41 (2): 133–150. https://doi.org/
10.1023/A:1005417410867 .
van Dooremalen T., and Duyvendak. 2025. “To Foreignize or
to Domesticate? How Media Vary Cross-Nationally in their
Degrees of Incorporating Foreign Events.” Journalism 26
(1): 108–127. https://doi.org/10.1177/14648849231223576 .
van Dooremalen, T., and J. Uitermark. 2021. “The Framing of
9/11 in American, French, and Dutch National Newspapers
(2001–2015): An Inductive Approach to Studying Events.”
International Sociology 36 (3): 464–488. https://doi.org/10.
1177/0268580920966829 .
West, B., and P. Smith. 1997. “Natural Disasters and National
Identity: Time, Space and Mythology.” The Australian and
New Zealand Journal of Sociology 33 (2): 205–215. https://
doi.org/10.1177/144078339703300205 .
Weston, K. 2023. “Exact Data UK Heatwave Could Peak at
a Blistering 37C.” Daily Mirror, June 13.
Williams, R. 1977. Marxism and Literature. Oxford: Oxford
University Press.
14 T. VAN DOOREMALEN AND P. SMITH

Appendix A: Newspaper Overview
Broadsheet Publications
Name Days of Publication
Years Available in Nexis
Uni
The Guardian Daily 1984–2023
Financial Times Daily 1982–2023
The Sunday Times Sundays 1985–2023
The Times Daily 1985–2023
The Observer Sundays 1990–2023
The Independent Daily 1990–2023
The Daily Telegraph Daily 2000–2023
The Sunday Telegraph Sundays 2001–2023
Tabloid Publications
Name Days of Publication
Years Available in Nexis
Uni
Daily Mail and Mail on Sunday Daily, Sundays 1992–2023
Daily Mirror Daily 1996–2023
The Sun Daily 1996–2023
Sunday Express Sundays 1999–2023
Daily Star Daily 2001–2023
Morning Star Daily 2001–2013
Daily Star Sunday Sundays 2005–2023
Appendix B: Topics and their Codes
Example 1 Example 2 Example 3
Period 1985-1989 Period 2005-2009 Period 2020-2023
House Wimbledon Climate
Beach Match Change
Town Sport Crisis
Hotel Run World
Best Tennis Report
Restaurant Heat Extreme
Good Court Global
Place Play Action
Pool Murray Future
Room Open Impacts
Sun Final Risk
Even Won Environmental
Travel Set Government
Great Ground Countries
Go Well Need
Genre Genre Genre
Romantic Neutral Apocalyptic
Specific Subject Specific subject Specific Subject
Travel/Holidays Sport events Climate change
ENVIRONMENTAL SOCIOLOGY 15