Exploring discourse and ideology through corpora

by Miguel Fuster Márquez (Volume editor) José Santaemilia (Volume editor) Carmen Gregori-Signes (Volume editor) Paula Rodríguez-Abruñeiras (Volume editor)
Edited Collection 292 Pages

Table Of Content

←6 | 7→

Miguel Fuster-Márquez, José Santaemilia, Carmen Gregori-Signes and Paula Rodríguez-Abruñeiras

Insights from corpus-assisted discourse analysis: Unveiling social attitudes and values

The exploration of discourse through corpus linguistics techniques is a rapidly expanding field. Indeed, Corpus-Assisted Discourse Studies (CADS), as introduced and defined by Alan Partington, has become a very popular methodology or approach for the critical and also non critical analysis of discourses. Publications have proliferated over the last decade, thus problematising their objects of study and increasingly widening their scope and focus. These trends have encompassed a wide range of interests and the proposal of applying specific corpus techniques, or combining different methodologies, or drawing on information from various relevant sources, as an appealing strategy. There is already a well-established tradition of disciplines that have benefitted from the potential application of corpus techniques, including applied linguistics, variationist studies, specialised languages, historical linguistics, pragmatics, forensic linguistics, learner language, translation and stylistic studies, among others. Additionally, young scholars who wish to engage themselves in corpus-related research, today have at their disposal handbooks and companions published by the most prestigious publishing houses, and also a large number of public corpora, web sources and software that will pave their way for their research. While some publications have dealt with corpus linguistics overviews others have, more specifically, focused on statistical methods. This is a helpful reminder that quantification is a crucial and distinctive aspect of most kinds of corpus research, perhaps inextricably bound to it. In contrast with earlier linguistic analysis which focused on single texts, or dealt with small amounts of data, dealing empirically with large amounts of text matter in corpora, not infrequently millions of words, and hundreds or thousands of texts, brings with it the need of solid and ←7 | 8→reliable quantification methods. The fact of exploring large amounts of data has meant a change of research priorities, and older alternative purely qualitative methods are simply not suitable. This also means that corpus linguistics, being more thoroughly empirical, has been moving away from the humanities to embrace methods, approaches which are regularly found in the social sciences.

It is increasingly evident that corpus methodologies have already greatly contributed to discourse analysis and shown their potential to unveil political or ideological attitudes, values, or social inequalities. The implementation of corpus techniques, which are increasingly refined every day, allow us access to document social attitudes, to expose injustice or sexism and –hopefully – to fight against discrimination by obtaining results which are more generalisable. Until quite recently, the study of social issues, identity politics or ideological values has favoured qualitative analyses of small collections of texts. Today, however, more and more researchers are convinced that the use of corpus linguistics for social analysis constitutes a very powerful instrument of social inquiry, especially if used in combination with other (qualitative) approaches which require close reading and the interpretation of particularly salient textual features. Corpus-Assisted Discourse Studies (CADS) enjoys today great popularity and academic appeal, partly because of its ability to move back and forth, in a complementary way, between qualitative and quantitative techniques in order to generate new hypotheses and to test existing ones.

Contributions to this volume deal with socio-ideological issues such as political and media discourses, gender-based studies or hate language, which have increasingly become the object of corpus linguistics research, as have their evaluative, emotive or attitudinal dimensions. Emerging trends in research have come from the application of CADS to the interactive discourses contained in social networks. Some contributions in this volume also bear witness to this trend. The contributions to this volume share a committed view on social reality. To meet their goals, on the one hand, they use corpus tools to guarantee a balanced quantitative and qualitative analysis of (large or small) corpora and, on the other, provide us with a critical approach to the ideological nuances present in texts. This book offers a wide range of analyses and insights into burning and/or conflicting social issues. Alan ←8 | 9→Partington’s chapter, entitled “Post-history, post-democracy, post-truth, post-Trump? Really? A corpus-assisted study of delegitimisation via argument strategies: ‘dirty tricks’, evaluation and hyperbole in modern political discourses,” illustrates the balanced use of qualitative and quantitative approaches in the examination of a series of contemporary post-denominations (post-history, post-democracy, post-truth, post-Trump) in order to unveil the ideological delegitimisation of opposing groups or individuals. In this chapter, Partington quite insightfully addresses a number of crucial problems which have to do with the difficult synergy of Corpus Linguistics and more classical qualitative Discourse Studies. For this particular piece of research, Partington has made use of different corpora. In his view, researching ‘macro’ argument strategies (i.e. de/legitimisation) is a way of addressing one of the ‘dusty corners’ (Taylor/Marchi 2018) in corpus linguistics, which tends to favour a (sometimes decontextualised) scrutiny of large bodies of texts by focusing on ‘micro’ strategies. For Partington, macro level features should be “ ‘assisted’ by the researcher’s intuitions” and world knowledge.

Tony McEnery, Helen Baker and Carmen Dayrell are in charge of the second chapter, “Analysing the impacts of 19th-century drought: A corpus-based study”, where they turn their attention to the past in order to access information gaps through textual data about the weather that could be useful to historians and meteorologists, thus proving the usefulness and convenience of carrying out textual analyses of large amounts of historical media texts (in this case, a large sample of 19th century British newspapers reporting on climate conditions in different parts of the country) by means of corpus methodologies. The rigorous examination of discourses can become an indispensible contribution of corpus linguistics to non-linguistic fields such as the one dealt with here in relation to weather conditions in earlier centuries which cannot be accounted for directly through sophisticated modern standard methods. Nevertheless, the authors deem most appropriate to make use of triangulation, hence they use corpus methods in concert with other techniques (concordance geo-parsing and close reading analyses) with a view to reconstructing, as faithfully as possible, media narratives of the droughts and water shortages in different regions of the UK.

←9 |

Political Discourse, one of the favourite targets of corpus-based analysis, is illustrated in this volume by Salvador Enguix-Oliver and Beatriz Gallardo-Paúls’s study of recent political developments related to the extremist far-right Spanish party Vox during 2018 in their contribution “Coverage of the far-right in the Spanish written press: The case of Vox”. With this aim in mind, the authors have selected a number of widely read Spanish quality written newspapers (El País, El Mundo, La Vanguardia and ABC) and digital dailies (El Español, eldiario.es, Público and Infolibre). According to the authors, these newspaper items belong to their own sub-corpus PRODISNET-201, which is part of the larger PRODISNET (Discursive Processes on the Internet corpus). The hypothesis explored in this chapter is that far-right parties like the Spanish party Vox have been, and are still, receiving largely undeserved media coverage in Western democracies, and this media attitude enhances their public visibility which results in unprecedented electoral success. The authors also resort to the use of quantitative methods such as sentiment analysis (using the software Lingmotif, which reveals a basically negative appraisal) and complementing their study with close pragmatic analyses of all texts, since the authors find, for example, that political evaluation in the media is often expressed implicitly. In their conclusion they highlight that negative political evaluations are most often carried out through presuppositions and anomalous implicatures of manner and quality.

In “Evaluation in Theresa May’s political discourse: A study of the PM’s seminal Brexit speeches”, Ana Belén Cabrejas-Peñuelas and Rosana Dolón focus on a burning issue such as the Brexit process, as exemplified in three seminal speeches by the former British PM Theresa May. The speeches were delivered during key moments of Brexit negotations: the Lancaster House speech (January 2017), the Florence speech (September 2017) and the London Mansion House speech (March 2018). The authors home in on the verbal content of this corpus to show how evaluation of status (see Hunston 2000, 2008, 2011) is expressed in political texts, and reveals how corpus linguistics can meaningfully contribute to the study of evaluation and persuasion. The typological framework applied here has the earlier work by Díez-Prados & Cabrejas-Peñuelas (2018) as an acknowledged antecedent. Use has been made in this contribution of the freeware program called UAM Corpus Tool ←10 | 11→developed by Mick O’Donnell (2019). The corpus has been manually annotated to automatically carry out the inferential statistical analysis. According to the authors, a key persuasive strategy in Theresa May’s speeches would be the use of rhetorical devices of “suggestion” against “proof”, and of maintaining an apparently objective stance (Hunston 2011: 27). The linguistic choices made by the former British PM seem to have the projection of a down-to-earth and objective position as a specific persuasive intention.

Gender issues –and, specifically, violence against women (VAW) – feature prominently in the growing research that makes combined use of corpus linguistics and other methodologies, as shown in Leanne Victoria Bartley’s chapter “ ‘Nobody is guilty in football. That’s the first thing to understand’: A corpus-assisted critical discourse analysis of the UK press coverage of the Ched Evans case.” Bartley focuses on judicial events which took place in 2012 in the UK, when the famous footballer Ched Evans was convicted of raping a young woman in a hotel room in Rhyl, North Wales, after a night out together. In 2016, Evans’ case was reviewed, his conviction overturned, and he also received a compensation of £800,000. Bartley explores the approaches of the British press (The Daily Mail, The Guardian, The Mirror, The Sun) during this period by means of a CADS approach, based on Martin and White’s (2005) appraisal theory slightly modified by Bednarek’s (2008) recommendations concerning the subcategory of affect. She makes use of UAM Corpus Tool (O’Donnell 2019), which allows users to employ this in-built paradigm for quantitative and qualitative analysis. Bartley investigates how both the perpetrator and his victim were represented in the British press at different stages of this controversial case. She finds that the media discourses changed over time – the view of the alleged perpetrator went from harsh criticism to a positive image. The focus on the victim, by contrast, went from understanding her inability to consent to sexual intercourse to criticising her inappropriate sexual activities.

In the following chapter, “News values in construing female victims of VAW discourses in the media: A view from CADS,” Sergio Maruenda-Bataller also addresses a gender issue. The aim of this author is to explore how mainstream Spanish and British quality newspapers, during the decade from 2005 to 2015, have construed female victims of ←11 | 12→VAW discursively. Classifying the news values that are used to depict female victims, he examines and compares the treatment that Spanish and British newspapers give to female victims in VAW news reports. For this purpose, the author makes use of a large purpose-built corpus where he applies the Discursive News Values Analysis paradigm as described by Bednarek and Caple (2017), but also relies on Potts et al.’s (2015) study for the use of corpus methods, where the authors also made use of a large corpus to investigate the representation of the Katrina Hurricane in the news. While more research is needed, it is the author’s view that results appear to substantiate the uneven presence of two complementary discourses which are nevertheless inextricably linked: a discourse of death, violence and suffering (thus, predominance of news values such as negativity and impact) and another of institutional and social support, (which realises the news value of consonance).

Alfonso Sánchez-Moya also explores the gender issue of VAW in the chapter “How does violence-motivated online discourse differ from its non-violent counterpart? Insights from a CADS approach.” He makes use LIWC, a sentiment software tool to analyse the discourse of female survivors of Intimate Partner Violence (IPV) in a corpus of online forum messages by women who are undergoing such violence, containing a total of 120,000 words. Insights into this phenomenon with a discursive approach are, in the author’s words, scarce. Some of his conclusions are that, when compared to another neutral/non-violent corpus containing forum messages by women, the discourse in this IPV-related forum appears to be characterized by a narrative-oriented and personal style, focusing more on the here-and-now. Particularly, the sub-corpus dealing with abuse within the IPV corpus shows a severely pessimistic type of discourse. Also, the claims about IPV-related negative emotions and their discursive characterization are validated by this analysis.

In “ ‘We’ll watch TV and do other stuff’: A corpus-assisted discourse study of vague language use in online child sexual grooming,” Nuria Lorenzo-Dus and Anina Kinzel show the practical application of a CADS methodology for very serious sexuality-related issues – i.e. a recent escalation of cases of online child sexual grooming (OCSG) as “an Internet-enabled communicative process of entrapment” in the ←12 | 13→UK. The authors’ aim is to identify the linguistic and rhetorical tactics used by adults convicted of OCSG in order to provide collaboration to law enforcement agencies and educational institutions. The large corpus containing chat logs were compiled between 2004 and 2016. The corpus was uploaded to CQP Web, and VARD has been used by the authors to standardise the spelling. Among their conclusions they highlight the strategic use of implicitness and vague language for manipulative goals, and their results show that the main category used by groomers is that of quality-approximators, followed by vague identifiers and de-identifiers. Briefly, groomers manipulate communications with children as targets in ways which confuse them greatly.

Stefania M. Maci presents the chapter “The narrative of the anti-vax campaign on Twitter”, on medical knowledge as disseminated by digital media. Maci describes the discourse strategies employed by US anti-vaccine activists in a small Twitter corpus of 16,768 tweets (75,960 running words) gathered in October 2018 after some children died from measles in New York City, their aim being spreading fake news about vaccination. The corpus has been semantically annotated by USAS to identify key semantic domains and WMatrix has been used for the automatic processing of the most relevant meanings. Maci also carried out a sentiment analysis, employing Socialbearing.com as its software. The results obtained reveal an overall strong, articulate anti-vaccine discourse displayed by these activists of the anti-vaccine movement in their tweets. Maci claims that, although apparently grounded on scientific truths, these activists favour a conspiracy theory between the government and the pharmacological industry.

And last but not least, Nouf Alotaibi and Jane Mulderrig are the authors of the chapter that closes this volume, “Debating Saudi womanhood: A corpus-aided critical discourse analysis of the representation of Saudi women in the Twitter campaign against the ‘male guardianship’ system.” Again, gender politics features prominently here, as the chapter focuses on a highly controversial topic (‘male guardianship’) affecting the freedoms, and the whole lives, of millions of women in Saudi Arabia. The authors wish to analyse the Saudi women’s attitudes towards the online campaign #سعوديات_نطالب_اسقاط_الولاية (i.e. ‘#EndMaleGuardianshipSystem’ – #EMGC) that started in 2016, and is generating a polarization of Saudi women’s positions into two ←13 | 14→groups – Pro-#EMGC and Anti-#EMGC. The paper seeks to answer three questions: “1) who were the salient social actors in the discourse of #EMGC?; 2) how did the female campaigners represent Saudi women as social actors?; and 3) what were their most prominent actions?” To that end they use a corpus-aided critical discourse approach, bringing together the potential of corpus tools (word list and concordance tools provided by the concordance software ‘AntConc’ – see Anthony 2017) to reveal key textual patterns and the insights drawn from CDA and the representation of social actors (van Leeuwen 2008). This chapter shows, as do the others in this volume, that a synergy between corpus and discourse analytical tools is crucial in identifying, analysing and even advancing social issues – i.e. in this case, the women’s rights issue in Saudi Arabia.

Readers should notice that authors in this volume have relied on data contained in either small or large corpora. It is convenient to bear in mind that, in the same way as there is no specifically recommended methodology attached to the analysis of (critical) discourses, there is no agreed criterion either about the size of corpora suitable in research. Perhaps ‘the bigger the better’ is a good guiding principle to be seen in many published papers, since large corpora are more likely to yield generalisable findings and insights; however, small targeted corpora, as shown here, are perfectly suitable for many kinds of corpus research into discourse.

Also, although not a few analysts believe that the use of corpus linguistics tools is a thoroughly automatic process, where the “machine”, not the analyst, decides the results obtained through quantification, this in fact constitutes a very poor understanding of the way corpus analysis is actually applied. Researchers need to use their intuition and knowledge to decide many aspects throughout their corpus research, not the computer software. For example, as shown in this volume, not a few authors have chosen specific annotation tools and, at times, annotation is also carried out manually. Other authors decide not to annotate their corpus. In any case, no two annotation systems are equal. Therefore, a variety of annotation tools can be used, in the same way as the “kind” and “size” of corpus is decided by the researchers according to their particular objectives. As to the corpus software, the contributions in this volume also show a variety of software tools, platforms, and the answer as to which ones to apply lie entirely in the researcher. As to ←14 | 15→quantification, different statistics are also displayed. All of these are decisions or choices which have to be made by researchers, and come to show the resourcefulness of corpus linguistics and why it is becoming more and more popular in discourse analysis research.

Finally, corpus linguistics has been enriched by the incorporation of larger social, historial, ideological and ethical considerations, brought about by critical approaches to discourse. Certainly, this volume is not the first one to be published around these issues. However, it attempts to provide readers with updated research from leading international scholars in Discourse Analysis. All CADS researchers found in this volume are quite aware of the strength which come from applying a methodological synergy that combines different kinds of quantitative and/or qualitative research. A critical attitude towards the choice of methods and the author’s stance undoubtedly contribute to more fine-tuned analyses and to a more responsible and reflexive practice.

We believe this edited volume is an excellent example of this growing symbiosis between methods, types of corpora, disciplines, software programmes, discourse perspectives, and social and ethical concerns.


Anthony, Lawrence. 2017. AntConc (version 3.5.0), [Computer Software], <https://www.laurenceanthony.net/software>, accessed 10 February 2017.

Bednarek, Monika. 2008. Emotion Talk across Corpora. Houndmills/New York: Palgrave Macmillan.

Bednarek, Monika/Caple, Helen. 2017. The Discourse of News Values: How Organisations Create Newsworthiness. Oxford: Oxford University Press.

Díez-Prados, Mercedes/Cabrejas-Peñuelas, Ana Belén. 2018. Evaluation of “Status” as a persuasive tool in Spanish and American pre-electoral debates in times of crisis. Atlantis: Journal of the Spanish Association of Anglo-American Studies 40/2, 169–195.

←15 | 16→

Hunston, Susan. 2008. The evaluation of status in multi-modal texts. Functions of Language 15/1, 64–83.

Hunston, Susan. 2011. Corpus Approaches to Evaluation: Phraseology and Evaluative Language. London/New York: Routledge.

Hunston, Susan/Thompson, Geoff. (eds.) 2000. Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press.

Martin, James Robert/White, Peter Robert Rupert. 2005. The Language of Evaluation: Appraisal in English. Basingstoke: Palgrave MacMillan.

O’Donnell, Mick. 2019. UAM Corpus Tool: Annotation for the 21st Century, <http://www.corpustool.com/index.html>, accessed 12 June 2019.

Partington, Alan. 2004. Corpora and discourse: a most congruous beast. In A. Partington, J. Morley & L. Haarman (eds.) Corpora and Discourse. Bern: Peter Lang. 11–20.

Potts, Amanda/Bednarek, Monika/Caple, Helen. 2015. How can computer-based methods help researchers to investigate news values in large datasets? A corpus linguistic study of the construction of newsworthiness in the reporting of Hurricane Katrina. Discourse and Communication 9/2, 149–172.

Taylor, C./Marchi, A. (eds.) 2018. Corpus Approaches to Discourse. London: Routledge.

van Leeuwen, Theo. 2008. Discourse and Practice: New Tools for Critical Discourse Analysis. Oxford: Oxford University Press.

←16 | 17→

Alan Partington

Post-history, post-democracy, post-truth, post-Trump? Really? A corpus-assisted study of delegitimisation via argument strategies: ‘dirty tricks’, evaluation and hyperbole in modern political discourses


This chapter will cover a wide range of history, of themes and of linguistic approaches and notions. As the title implies, we intend to examine how delegitimisation of opposing speakers/authors and their stances is attempted, incorporating the use of techniques from the burgeoning area of Corpus-assisted Discourse Studies (CaDS; Partington et al 2013). We will draw on notions of persuasion first systematised by Aristotle, and also more recent theories which have largely arisen from the area of corpus-assisted linguistics, namely text organisation/structures (Hoey 1983, 1991; Partington/Taylor 2018), evaluation analysis (Hunston/Thompson 2000), lexical priming (Hoey 2005) and its subcategory, highly relevant to political discourse, of deliberately forced lexical priming (Duguid 2009a). Attempts to use corpus techniques to investigate argument structures is a much neglected ‘dusty corner’ (Taylor/Marchi 2018) of CaDS, given a historical tendency of corpus linguistics to focus on copious micro- rather than macro-structures and strategies.

←17 | 18→

2Corpus-assisted Discourse Studies (CaDS): capacities and issues

Corpus-assisted Discourse Studies (CaDS) can be understood as ‘the set of studies into the form and/or function of language as communicative discourse which incorporate the use of computerised corpora in their analyses’ (Partington et al. 2013: 10). The advantages of including corpus linguistics techniques in discourse analysis are several. It provides availability of a wealth of data which enables the analyst, among other things, to perceive an overview of the data under analysis. These software techniques enable the researcher to reorganise the data in hand to analyse it in different ways (see for instance Marchi 2018 on how dividing a dataset in different ways can lead to radically different observations). The use of a variety of corpora facilitates the contrast and comparison of different but related datasets, not only for differences between them but also similarities (Taylor 2018). This contrastive virtue cannot be stressed sufficiently; it is only deontologically justifiable to make statements about the thematic contents and behaviours of participants in one dataset of discourse types if we can compare them with those of other discourse types. Finally, the ability to rearrange the data in various ways vastly augments the potential for serendipity, that is, of stumbling upon unexpected findings, some of which may even, as we shall discover here, be counter-intuitive.1

Naturally, as in all scientific endeavours, some ways of proceeding are more or less rigorous than others, more or less self-aware, not to mention more or less honest. Another of the added values of corpus techniques in discourse analysis can be summarised as the capacity of ‘temporary alienation’ of the observer from the object of observation. As is frequently repeated, complete objectivity of observation in the human sciences is unachievable – observers inevitably bring their personal primings into the experiment – but the striving towards a more accurate, defensible and powerful model of the object of observation is a ←18 | 19→vital component of research. The temporary alienation in question is that of the abstraction or removal of the analyst/observers from the object of evaluation and temporarily placing some distance between the interpreter and the interpretation, given that parts of the analysis are carried out by a different entity from the eventual interpreter, that is, the ‘machine’, which has no vested interest in either disproving or corroborating the analyst’s hypotheses. This process reduces analytical subjectivity:

[O]‌nce the software is set in motion the researcher cannot interfere. There is therefore at least one phase of corpus-assisted research, the software analyses […] which the researcher cannot either consciously or unconsciously manipulate. (Partington et al. 2013: 351)

All this is entirely congruous with scientific practice in general well beyond corpus linguistics. Scientific instruments are adopted precisely in order to ‘see’ physical phenomena in new ways in order to accumulate more precise insights into their nature. This breaking out of the closed hermeneutic circle greatly increases the risk of meeting unexpected serendipitous observations, and unexpected observations require explanation, given the challenge they often pose to the original hypothesis on which the research was predicated. New evidence can lead to new learning.

However, there are a number of potential flies in this research ointment. The first to consider is corroboration bias (also known as confirmation bias), in the worst cases of which a researcher actively seeks out examples to prove their case (sometimes called ‘cherry-picking’). In its subtler form, confirmation instinct, researchers are tempted, indeed primed, to notice and remember data which confirm their original hypothesis and even personal biases (Marchi/Taylor 2009). Even what we count as evidence of a certain position when conducting socio-political CaDS can be affected by confirmation impulse. Perhaps the best way to guard against corroboration bias is to consciously seek out counterexamples to our hypothesis. For example, does a right-wing media source ever evaluate immigration positively and if so on what grounds?2 However, even here we need to exercise care; how many ←19 | 20→counterexamples are needed in order to reject or refine a hypothesis? In the real world, one black swan is not enough, it may simply be a mutation. Indeed, researchers need to put up with a certain degree of cognitive dissonance, to allow for the fact that their hypotheses may not be entirely self-consistent.

There are two further problematic issues which at first sight seem to work against each other. The first is the well-known ‘drama-impulse’, the rush to publish striking findings, that is, ones which appear to either substantially add to existing knowledge or to challenge parts of it. Additional research is often needed to corroborate unexpected results. The second has been called the ‘Semmelweis reflex’, that is, either the hesitation to publish – or the difficulty in getting published – research that goes against the prevailing wisdom in the field or, indeed, its dominant ethical norms (see also note 4 on the related concept of the ‘Overton window’).3 In its infancy, corpus linguistics itself had problems gaining legitimacy; basing language description on examples of authentic use rather than trained grammarians’ introspection proved a paradigm shift too far for many linguists (Nelson Francis 1982).

One final issue has been named the ‘not discovered here’ syndrome, otherwise known as academic tribalism, in which one discipline either shuns or is simply unaware of research and discoveries made in others. This has been an issue for corpus linguistics, generally, especially well-put by Mautner ‘the integration of CL is still not part of the mainstream methodological canon [in areas such as sociology or management studies]; their hero is Foucault, not Sinclair’ (Mautner 2016).

←20 | 21→

The data for the set of studies outlined in this chapter were collected in various ways. One frequent procedure was to search Lexis Nexis Academic for lexical items of interest (e.g. post-truth) and, if necessary, download the articles which contained them and use that set of articles as a corpus, examinable for frequency word listing, concordancing and collocate analysis. The first collection phase was the twelve months before a talk on this topic was delivered in April 2018 at the Heidelberg Centre for American Studies. The second collection phase was the six months prior to writing up the present paper in October of 2019. The interrogation and use of the specially-compiled Brexit news corpus is described in Section 6. Reference is also frequently made to the online, free-to-use Corpus of Contemporary American English (COCA; Davies 2009).

3The age of the ‘post-’

Turning to the title of this chapter, in this section we look at a number of items prefixed by post- which can be used to undermine a set of beliefs or a particular stance, and potentially delegitimise groups, entities and persons who are related in some way to that stance. Delegitimisation is here defined as ‘the attempt to discredit the right or ability of an opponent to make a certain claim or argument or to hold a certain power’ (Partington/Taylor 2018: 76). These items may well be related intertextually; post- in other words, may have become a productive item, one coinage spawning others, generally with negative evaluations.

The first term, ‘post-history’ was, famously, coined by Francis Fukuyama (1992) and has become one of the most misinterpreted notions of our day. Fukuyama was not predicting the end of historical events or even progress, but that the blood-soaked clash of ideologies which had distinguished 20thC history was at an end. His thesis was immediately countered by Samuel Huntington’s far more pessimistic vision of the future, arguing that the temporary conflict between ideologies is being replaced by the ancient ‘clash’ between civilizations (Huntington 1996). COCA records only four uses of the item, none of ←21 | 22→which are related to Fukuyama’s thesis, but two are from the field of geo-politics and express a decidedly negative evaluation, for example:

(1)Most Europeans don’t acknowledge the great paradox: that their passage into post-history has depended on the United States not making the same passage. (COCA)

The term ‘post-democracy’, instead, was coined by Crouch in 2006:

to describe a coming age where democratic institutions in Western countries seemed strong, but where power was subtly being transferred to an oligarchic politico-economic elite. According to Crouch’s pessimistic model, elections gave the veneer of democratic control by changing governments, but did not actually change the fundamental substance of politics, which was shaped in private through exclusive dialogue between governments, and globalised business and corporate elites. For Crouch, political debate in a post democracy is a controlled spectacle, managed by communication experts and professional spin doctors, revolving around a prescribed set of issues debated within a narrow ‘Overton window’ by increasingly similar and ideology-free political parties. (Blaxill 2018: 57)4

Interestingly, although clearly intended as a critique of the debasement of democracy from a left-wing, anti-capitalist perspective, much of Crouch’s analysis would chime with right-wing populist movements who decry the hegemonisation of western democracy by so-called liberal élites, who have taken control of decision-making away from ordinary people. COCA contains just two occurrences of the item, one of which makes mention of the European Union potentially being such a liberal post-democracy (see Section 6 on EU-scepticism):

(2)“Europe”, points out National Review editor and syndicated columnist Rich Lowry, “is on the brink of a’ liberal post-democracy,’ a form of government that (usually) respects basic rights at the same time it lacks the mechanisms for ensuring the popular consent that characterizes traditional democracy? The unelected European Commission not the European Parliament initiates legislation and issues regulations.” (USA Today 2012)

←22 | 23→

The next item, ‘post-truth’, was nominated by the Oxford English Dictionary in 2016 as its ‘word of the year’, a word which was felt to have played a prominent role in Anglophone discourse in that year. The item ‘post-truth’ was defined as ‘relating to or denoting circumstances in which objective facts are less influential in shaping public opinion than appeals to emotion and personal belief’. While acknowledging it was not a new term, the dictionary nevertheless claimed it has gone from being ‘a peripheral term to being a mainstay in political commentary’ and is often used in the expression ‘post-truth politics’. COCA records 25 uses of the term, all from 2015 and later. Many of these occurrences are present-time references:

(3)the post-truth climate of our 2017 (A.V. Club 2017)

(4)The nation’s current post-truth moment (The Atlantic 2017)

including five which refer negatively to the current Trump presidency, for example:

(5)This Sunday, after a post-truth election, do the president-elect’s words matter? (NBC 2016)

(6)The question is, after what many have dubbed the post-truth election, are we getting a look at a post-truth presidency? (NBC 2016)

while another three refer to supposed problems of current media communication, for example:

(7)In a ‘‘post-truth’’ era when journalism is under pressure […] (The Atlantic 2017)

(8)And social media, which is becoming the primary news source for many people, amplifies it all. At its worst, it marinates people in conspiracy theories and helps to foster this ‘‘post-truth’’ era of politics. (New York Times 2016)

We return to ‘post-truth’ in section 3.

‘Post-Trump’, meanwhile, is a semi-sarcastic coinage meant to indicate the sentiment that the rise of Donald Trump to the US Presidency has drastically altered the political landscape, not only in the United States but around the world. For some of his denigrators, the advent of Trump was the Apocalypse (from Lexis Nexis):

(9)An environmentalist told journalist Alan Weisman before the 2016 elections that she was considering voting for Trump. “The way I see it,” she said, “it’s either four more years on life support with Hillary, or letting this maniac ←23 | 24→tear the house down. Maybe then we can pick up the pieces and finally start rebuilding.”

The philosophy of “things have to get worse before they get better” has sometimes worked out in the past. But that’s the past.

Unless we stop him, we’ll be rooting around in the post-Trump ashes in vain for the pieces. The house will be gone. And there will be nothing we can salvage to rebuild it. (Foreign Policy in Focus 2019)

And even one of his Republican supporters describes him as ‘an angel of destruction’ (David Brooks, New York Times 2019). The term has certainly caught on. In the six months before beginning writing this chapter a Lexis Nexis Academic search of English-language news sources this article produced 92 instances of ‘post-Trump era’.

Finally, the term ‘post-trust’ (see Singh/Luthra 2019) has become an object of media attention. A Lexis Nexis search of the term ‘post-trust era’ in English-language news sources in the six months previous to writing this article produced 40 results including the headline ‘This is how to fight fake news at work in a post-trust era’ (Fast Company 2019). The 40 articles were downloaded and concordanced for post-trust. The list of entities which people supposedly no longer trust include:

politicians, governments

banks, traditional economics

the media, press

the police

science, scientific organisations


Most of these will not surprise the reader. As has often been noted, corpus techniques often produce results which, in hindsight at least, seem obvious (Louw 1993). Tongue-in-cheek, Partington (2017) has called this ‘hindsight predictability’ or ‘postdictability’. However, again, the frequency of hindsight predictability should not surprise us. It would be a very strange set of tools and approaches which always came up with counter-intuitive results.

Most of these post-somethings, with perhaps the exception of Fukuyama’s ‘post-history’, have a negative evaluation and can be seen, in the light of the next topic of this chapter, as strategies to delegitimise. ←24 | 25→Calling a political system ‘post-democratic’ – for example, the EU in ­example 1 above – undermines practically everything it does. Labelling the media (including social media) as post-truth may be a way of calling into question its abilities to be honest and fair; the phrase post-truth election of Trump delegitimises both the election and the winner. These post- prefixed items also appear to suggest that there was an earlier period in history in which politics and political discourse were, if not exactly in a Golden Age, in some way more open, more democratic and less debased.

4The Golden Age?

The media and press figure prominently among the entities people no longer trust. However, any belief that there was once an era of political media ‘truthfulness’ and reliability meets immediate problems when one researches historical documents. Even someone as generally democratic and libertarian as Thomas Jefferson remarked that the temptation for even the most well-meaning of politicians to mistrust the press could be acute. Before achieving the Presidency he famously wrote:

[…] were it left to me to decide whether we should have a government without newspapers or newspapers without a government, I should not hesitate a moment to prefer the latter. (Jefferson, letter, 1787)5

but after becoming President, he despaired of what today is often called ‘fake news’:

Nothing can now be believed which is seen in a newspaper. Truth itself becomes suspicious by being put into that polluted vehicle (Jefferson, letter, 1807)

I deplore with you the putrid state into which our newspapers have passed, and the malignity, the vulgarity, and mendacious spirit of those who write for them (Jefferson, letter, 1814)6

←25 | 26→

Nor was this disillusion with the press unknown in the UK, judging from the celebrated sardonic verse of the English poet Humbert Wolfe in 1930:7

You cannot hope

to bribe or twist,

thank God! the

British journalist.

But, seeing what

the man will do

unbribed, there’s

no occasion to.

As for the alleged new aggressive divisiveness of modern democratic politics, there is little in contemporary political combat to match some of that seen in the past. As Partington and Taylor relate:

In the 1800 campaign between John Adams and Thomas Jefferson, the former’s campaign referred to Jefferson as ‘the Anti-Christ’, while the latter’s campaign spoke of Adam’s ‘hideous hermaphroditical character which has neither the force and firmness of a man, nor the gentleness and sensibility of a woman’. In 1828, candidate Andrew Jackson’s wife was repeatedly labelled a prostitute. More divisive still was the duel in 1804 between the Vice-President Aaron Burr and Andrew Hamilton, one of the Founding Fathers. Hamilton was shot dead. And finally the election of Abraham Lincoln in 1860 prompted seven southern states to form the Confederate States of America leading directly to the US Civil War, which surely qualifies it as the most divisive US election ever (2018: 207).

If we look back at the citation of Crouch’s views on a supposed ‘post-democracy age’, where political antagonism is merely a sham and political élites are in collusion, it might appear that outside commentators are never happy. Contemporary politics is either overly aggressive or not nearly aggressive enough; in both cases, negatively evaluated.

To conclude, was there ever a Golden Age of Truth before this ‘post-truth’ era? The notion of ‘post-truth’ is certainly popular in the media. Lexis Nexis uncovered 1,111 occurrences of ‘post-truth’ in the 12 months prior to this start of this research. The articles in question were downloaded and using the WordSmith Tools (Version 5) collocates ←26 | 27→and clusters tools, the following sets of items were found to co-occur with post-truth:

world/era/age, politics, culture

fake news, fiction, disinformation, alternative (facts/reality), bullshit, spin, information warfare, new lying

suppression of freedoms

Brexit, Trump’s, populism

peak, current

The last group suggesting that political discourse has recently become dangerously debased. And yet, once again, past documents strongly suggest that political discourse in the past was far from pure:

Political language… is [often] designed to make lies sound truthful and murder respectable, and to give an appearance of solidity to pure wind. (George Orwell 1946)

Doublethink: To know and not to know, to be conscious of complete truthfulness while telling carefully constructed lies, to hold simultaneously two opinions knowing them to be contradictory and believing in both of them. (George Orwell 1949)

5Delegitimising your opponents as a political strategy

Predating Orwell by a good period but sharing his suspicion of public discourse, or rhetoric, was Plato who in his ‘Gorgias’ (386 BC) viewed persuasion as manipulation and the professional orator as a logodaidalos or ‘speech-rigger’(reminiscent of the modern accusation of ‘dirty tricks’). Public decision-making should be left to philosophers not orators. More than ironically, both Orwell and Plato were, of course, supremely gifted in rhetoric. Persuading by adopting the stance or ethos (see below) of the non-persuader, or even the anti-persuader, is a rhetorical ploy in itself, used to great effect by Frederick Douglass in his ‘This Fourth of July is Yours, Not Mine’ (1852):

←27 |

Fellow citizens, pardon me, and allow me to ask, why am I called upon to speak here today? What have I or those I represent to do with your national independence? Are the great principles of political freedom and of natural justice, embodied in that Declaration of Independence, extended to us?’

by Mark Antony in his funeral oration for Julius Caesar:

I am no orator, as Brutus is;

But, as you know me, a plain blunt man

That love my friend …

and by Nelson Mandela in 1990 after his release from prison:

I stand here before you not as a Prophet, but as a humble servant of you, the people.

Plato’s pupil, Aristotle, in ‘The Art of Rhetoric’ (368–322 B.C.) describes persuasion as opportunity and part of everyday life, and divides the skill of persuasion into three separate parts or structural principles, namely:

logos: persuasion by (apparent) appeal to reason, that what you are saying appears true, to make sense, seems relevant and logically coherent;

pathos: persuasion by appeal to emotion;

ethos: convincing an audience that you, the speaker, are worth listening to and have the right, the authority and the correct personality to speak on the matter in hand. This is clearly a rather different category of notion from the other two parts of rhetoric.

Although Aristotle’s model has been extremely influential, in practice it is often difficult to disentangle persuasion by logos and pathos. In real life many arguments have both logical and emotional appeal. For example, most persuasion – including political argument – appeals to some degree to the self-interest of the audience, persuading them that a certain course of action is in their best interests. On the one hand, tax cuts will mean more money in your purse; on the other hand, higher taxes and increased government spending will improve your public services. Is this a logical or emotional means of persuasion? It clearly combines both inextricably.

←28 |

Aristotle’s model has been adopted and revised in modern linguistic theory. In Systemic Functional Grammar (SFG; Thompson 1996; Halliday/Matthiessen 2004), argument by logos and pathos are both subsumed under the ideational (sometimes also known as the experiential) function of language, which describes how speakers and writers represent the world through language, while ethos is studied as part of the interpersonal function of language, dealing with how speakers and audiences interact.

Aristotle’s notion of ethos has also been adopted and updated in modern sociolinguistics as face and facework in politeness theory (Brown/Levinson 1987). According to Brown and Levinson’s theory, all individuals possess face, that is, a ‘public self-image’ (1987: 61) which we all project of ourselves to the outside world. Facework is the behaviour we employ to project that image during social interactions. According to Brown and Levinson (1987), face consists of two complementary parts:

negative face: the basic claim to territories, personal preserves, rights to non-distraction – i.e. to freedom of action and freedom from imposition

positive face: the positive consistent self-image or ‘personality’ (crucially including the desire that this self-image be appreciated and approved of) claimed by interactants.

(Brown/Levinson 1987: 61)

Negative face, or ‘respect my space’ face, the desire for freedom from imposition and freedom of action is recognizable as more or less the kind of phenomenon we mean when we normally refer to politeness (‘first-order’ perceptions, Watts et al. 1992), for instance, we pay attention to another’s negative face when we knock on their door to request their permission to enter.

Positive face, instead, involves the desire to be respected, approved of, even liked and admired. Positive face, then, is a very similar notion to ethos, Aristotle’s first – and, he says, most important – part of rhetoric (Aristotle 2012, Book 1:2), ‘how the speaker/writer projects their personality and stance towards the audience/readership’ (Taylor 2010: 222).

←29 |

In politics, a persuader’s positive face is divided into two types, namely, ‘competence face’, that is, one’s image as well-informed, an expert, in control, authoritative and ‘affective face’, that is, one’s image as likeable, good-humoured, normal, ‘one of us’, ‘a regular guy’ (Partington 2006). There is an inherent problem however, in that the two kinds of face are not always simultaneously compatible. It is not always possible to appear both an expert and ‘one of us’ at the same moment. It becomes a political skill when to privilege the one over the other.

In terms of interpersonal communication and facework, then, political persuaders need to project their own positive face and to attack their opponents’ face, including, very commonly, via the process of delegitimisation, defined, we might recall, as attempting to discredit the right or ability of an opponent to make a certain claim or argument or to hold a certain power (Partington/Taylor 2018: 76). In other words, alongside attacking their arguments (their appeals to logos, reason, and pathos, emotions), an effective persuader can also undermine the ethos, that is the personal credibility, honesty, competence, even the likability of opponents. A key political delegitimising accusation, then, is that an opponent is unfit to perform a certain function. Lexis Nexis search for the term unfit for + office in English news sources in the first week in September 2019 revealed:

(10)Trump has proved again and again that he is unfit for the office he holds (CNN 03/09/2019)

(11)LABOUR MP: CORBYN UNFIT FOR OFFICE (Daily Mail, headline, 09/09/2019)

(12)Dominic Grieve, a former attorney general […] said the British leader [Boris Johnson] was unfit for office. (Reuters 07/09/2019)

Others also potentially unfit for office, according to opponents, included US Senator Joe Biden, South Africa’s Public Protector and Senator Hillary Clinton, as well as ‘female leaders’ in general (see Section 5). Demata (2018) in his corpus-assisted comparative analysis of tweets issued by the Clinton and Trump campaigns in the month before the Election, concordances the hashtags employed, and notes the instance ‘#CrookedHillary is unfit to serve’.

Bernie Sanders’s campaign slogan during the 2016 US primary elections, ‘Not For Sale’ succeeded in projecting his own positive face/ethos as incorruptible and simultaneously delegitimising his ←30 | 31→principal opponent, Hillary Clinton, by implication of dubious campaign funding.

Senator Clinton herself ran the following campaign advert in 2007:

It’s 3 a.m. and your children are safe and asleep. But there’s a phone in the White House and it’s ringing.

Something’s happening in the world. Your vote will decide who answers the call, whether it’s someone who already knows the world’s leaders, knows the military, someone tried and tested and ready to lead in a dangerous world.

It’s 3 a.m. and your children are safe and asleep. Who do you want answering the phone?

The subtext clearly legitimises her as an experienced leader, while delegitimising her opponent, the neo-Senator Barack Obama, implying an ethos as inexperienced and therefore untrustworthy. At the same time she bolsters her own ethos as being as competent and tough as any man.

Other recent attempts at delegitimisation have been cruder, such as the ‘lock her up’ and ‘crooked Hillary’ slogans of Trump supporters, where the ethos of criminal is imposed on Clinton. And sometimes it is not just the politician who is the object of delegitimisation attempts, but their supporters, as in Clinton’s characterisation of Trump supporters as ‘a basket of deplorables’. Insulting a whole group can be a risky strategy however, not to mention unlikely to win any new friends from the other side, and Clinton later apologised. COCA lists 56 occurrences of deplorables, all related to Clinton’s utterance, including examples where the description is eagerly re-appropriated – re-appropriation, of course involves principally reversing the evaluation – by Trump supporters:

(13)Much of that money donated was given by Trump supporters, the Deplorables that have been demonized as racists by some on the hateful far left. (Fox News 2017)

Indeed, attempts at delegitimisation can be wholly counterproductive, as Trump himself describes:

The cost of a full-page ad in the New York Times can be more than $100,000. But when they write a story about one of my deals, it doesn’t cost me a cent, and I get more important publicity. I have a mutually profitable two-way relationship with the media – we give each other what we need […] These media types sell ←31 | 32→more magazines when my face is on the cover, or when I bring a bigger audience to their television show than they normally attract, and by far. And what’s funny is that it turns out the best way for them to get that attention is to criticize me. (Trump 2015: 11)

In other words, the more they attack me, the stronger my ethos becomes.

6‘All-purpose’ delegitimisers

Aristotle writes that ethos is the most important of the three parts of rhetoric. If you can undermine the ethos of your opponent, then their logos and pathos arguments are consequently weakened too. Many of the examples of delegitimisation discussed so far have been of the ad hominem or ad personam kind, the undermining of an individual. It has been argued that there is also a form of ad personam which is gender-based, in other words, ad feminam. The argument runs that it is more difficult for women in politics to project an ethos of strength or toughness. It was put as follows:

Women running for office […] are subject to these two demands: Be a good leader! Be a good woman! While the qualities expected of a good leader (be forceful, confident and, at times, angry) are similar to those we expect of a good man, they are the opposite of what we expect of a good woman (be gentle, self-deprecating and emotional, but not angry). (Washington Post 2016)

In the terms used here, in other words, female candidates may find it more difficult to combine the projection of competence face as well as affective face. Partington / Taylor (2018: 214) collected the following negative media references to Clinton’s gender during the 2016 campaign:

Trump’s claim that Clinton lacks the ‘physical stamina’ to be president (Washington Post 2016)

Hillary Clinton’s charisma deficit is a common problem for female leaders (Quartz 2016)

Hillary Should Play Up Her Feminine Side (Newsweek 2015)

←32 |

Hillary Clinton talks more like a man than she used to (Washington Post 2016)

Hillary Clinton tries warmth as US voters go cold (Financial Times 2016)

Donald Trump calls Hillary Clinton ‘shrill’ (The Boston Globe 2016)

This is a delegitimising argument to be extremely wary of and should not be used as a justification itself for not choosing women candidates.

The boot is on the other foot when the neologism mansplaining is used; COCA records 18 instances, all from 2014 and after. These examples show how the term can be used to delegitimise men’s contribution to any verbal interaction:

(14)The other day, I was pointing out a new data set to a female student and asked if she wanted me to walk her through it. (It took me a couple of hours to get my head around it.) She told me I was mansplaining and that she knew what she was doing. This was brand new, very complex data that we just received. She (nor anyone else) knew we had the data yet or had ever worked with it. (USA, Slate Magazine 2016)

Five of the COCA entries are accusations against Trump and other Republicans in Democrat-leaning sources, for example:

(15)We’ve got to take a quick break. Just ahead, top Trump surrogate Newt Gingrich tangling with Megyn Kelly over Trump’s female accusers? Was it an example of, some are saying, of mansplaining and misogyny? (CNN (Anderson Cooper) 2016)

(16)While we’re discussing Trump’s heretofore undisclosed health issues and his misogynistic mansplaining to a woman who’s absolutely his superior in every way (Salon Magazine 2016 [the woman referred to is Hillary Clinton])

(17)“We so often have this position of Republican men mansplaining our lady parts and how they work, like we’re too dumb to understand science and make our own decisions,” she says. (Mother Jones 2015)

Napoletano / Aiezza (2018) analyse another recent attempt to delegitimise a whole category of supposed opponents. They compiled a corpus of Trump tweets in the six months after the election. They found a large number of items referring to the supposed unreliability and even downright mendacity of the news media, for example: fake news (48 occurrences), dishonest (13), failing (9), wrong (8), biased (5), false (4), inaccurate (5), made up (3), crooked (3), lies (3), phoney (3), fabricated (2), unfair (2), unwatchable (2).

←33 |

Trump did not invent the term ‘fake news’; COCA reveals instances of it appearing increasingly since 1992. But Trump uses the term and similar items repeatedly in an example of what Duguid (2009a) calls ‘forced lexical priming’ in order to attempt to take ownership of the ‘fake news’ narrative. ‘Forced priming’ describes a process whereby speakers or authors frequently repeat a certain form of words to deliberately ‘flood’ the discourse with messages for a particular strategic purpose. Trump’s strategy in his ‘fake news’ narrative is to taint the ethos of the entire media as untrustworthy and possessing an inbuilt anti-Trump bias. It is an all-round, all-purpose delegitimisation that can be used on every and all occasions to dismiss any negative report or opinion appearing in the media. It also serves the purpose of projecting Trump’s own ethos of ‘authenticity’, as the only font of truth, the only one telling the truth in plain language. Thompson calls this strategy ‘authenticism’ and argues that ‘the authenticist prizes simplicity of language […] because he associates simple expressions with honesty of emotion’ (Thompson 2016: 155). Demata, after his corpus-assisted study of Trump’s tweets (see above), adds ‘authenticism is not a new tendency, but it surfaces every time that trust in politicians reaches a low’ (2018: 75) and reminds us of the existence of #BigLeagueTruth, a self-organising internet community of Trump’s ‘fact-checker’ supporters, also dedicated to delegitimising the media.

7Competitive ‘all-purpose’ delegitimising: A case study. Scaremongering over Brexit

In order to study the reporting of the UK EU Referendum, Zuccato and Partington (2018) built two corpora of articles discussing Brexit from four UK newspapers downloaded via Lexis Nexis, using the search terms ‘Brexit OR EU Referendum’. Two right-leaning and largely pro-Leave newspapers (a broadsheet, The Telegraph and a tabloid, The Mail) and two left-leaning and largely pro-Remain ones (again, a broadsheet The Guardian and a tabloid The Mirror) were chosen in order to encompass a cross-section of UK news reporting.

←34 |

The first corpus contained all the articles in these papers which appeared in these papers the three months before the referendum (a total of 1,402 articles, 1,169,000 words). The second corpus consisted of all the articles published by them in the three months after the referendum (1,888 articles, 1,373,000 words).

Reading some of the articles in the pre-Brexit corpus, it became clear that fear was a common theme, and so a concordance was prepared of fear* using WordSmith V5, which threw up the following items: fear, feared, fearful, fearing, fearless, fears, fearsome as well as fear-mongering and fearmongering (all but fearless express negative evaluations). The WordSmith frequency listing of the corpus also contained 122 occurrences of scaremonger*; and by concordancing *monger*, another 15 occurrences of doom-mongering were uncovered.

The accusation of scaremongering was used amply by both the ‘Remain’ and ‘Leave’ sides:



The Mirror



The Daily Mail



The Guardian



The Telegraph



Tab. 1. Occurrences of scaremonger*/scare-monger*

The relatively large number of occurrences in the pro-Remain Guardian is largely a product of its much greater coverage of the Referendum in general. The vast majority of uses are in the form scaremongering, as either a noun indicating the practice of scaremongering or as an adjective modifying items such as tactics, campaign; scaremonger as a verb occurs just twice, as does scaremonger as a noun referring to an individual, e.g.

(18)He accused Mr Johnson of being a “scaremonger on wages”. (Telegraph 15/06/16)0

The item scaremongering appears in four patterns. The first pattern supplies the fullest information on agency – who is accusing whom and of doing what – the fourth pattern supplies the least information, ←35 | 36→namely, just who is the supposed culprit of the act of ‘scaremongering’. In the latter case, the delegitimising purpose of the accusation is clearest; all the audience needs to know is who is at fault. In the following annotation, (A) is the accuser (B) stands for the accused and (T) stands for the topic.

(i)(A) accuses (B) of scaremongering over/about (T), e.g.

(19)Dave Prentis, the head of public sector union Unison, reported the [Breaking Point, anti-immigration] poster to police, saying: ‘This is scaremongering in its most extreme form.’ (Daily Mirror 20/06/16).

Here (A) is Dave Prentis, (B) is the authors of the poster and (T) is immigration.

(ii)(B) is accused of scaremongering over/about (T). No (A), e.g.

(20)One of the big flaws of the Brexit campaign so far has been the failure of its leaders to give any detailed information about what the UK might look like if they are victorious. It’s a serious weakness for the Out campaign, and explains why the Remain camp’s Project Fear scaremongering over lost jobs and trade drying up appears so convincing. (Daily Mail 31/03/2016

Here (B) is ‘Remain camp’s Project Fear’; (T) is lost jobs and trade, but there is no explicit mention of (A), an accuser.

(iii)(B) is accused of scaremongering by (A). No (T), e.g.

(21)Several members of the studio audience directly accused the Prime Minister of “scaremongering” and failing to make a positive case for Britain to stay in the EU. (Daily Telegraph 03/06/2016)

Here (A) is members of a studio audience and (B) is the PM, but there is no explicit (T). A similar example:

(22)Boris Johnson, the former Mayor of London […] accused Mr Osborne and the Remain campaign of “scaremongering” and “endlessly knocking our country”. (Daily Telegraph 16/05/2016)

Here (A) is Boris Johnson (B) is Mr Osborne and the Remain campaign, but there is no (T).

(iv)(B) is accused of scaremongering. No (A) or (T), e.g.

←36 |

(23)Scaremongering by pro-EU campaigners ahead of next month’s referendum has triggered a surge in support for Brexit among businesses, a survey shows. (Daily Telegraph 10/05/2016)

where (B) is the Remain campaign. There is no (T). No (A) (except the writer).

There is a fifth sort, where only the topic of the scaremongering is mentioned, with no accuser (apart from the writer) and no overt accused:

The referendum is over, the decision has been made, scaremongering about “bloody immigrants” for political purposes is no longer needed. (Guardian 05/09/2016)

However, the writer implies blame for a group who ‘scaremonger’ about immigration for political purposes. We can appreciate from all this Aristotle’s point that what is an offensive delegitimisation strategy for one group is another group’s self-evident truth.

In the majority of cases where Remain accuses Leave of scaremongering, it refers to the allegation of spreading false myths about immigration:

(24)The threats coming from the Brexiteers about Turkey joining the EU and millions of Turks coming here are scaremongering in the extreme (Daily Mirror 7/06/16)

The concordances of scaremonger* in the pro-Leave texts, on the other hand, reveal that Remain’s supposed scaremongering is constructed as continuous (with collocates like constant and incessant), but also as unreasonable, bordering on crazed (baseless, demented, ludicrous). In addition, it is perceived as deliberately shocking (blatant, cynical, naked, outrageous), all with negative evaluation. There is a larger array in the pro-Leave texts of specific indications of who is doing the ‘scaremongering’. They include: the Prime Minister Cameron, pro-Remain Ministers (Osborne, Hammond), ‘Downing St’, the Establishment, the Bank of England, Goldman Sachs. Given the fact that, according to the data examined here, only the Remain side was successfully labelled by their opponents as launching Project Fear (there were 132 occurrences in this data, a further example of deliberate forced lexical priming; see Tab. 2 for a selection) it might be inferred that Leave was more successful than Remain in its attempts at delegitimising their opponents as scaremongers.

←37 |

1 box on June 23. In particular the Leave campaign is very exercised by what they say is the Project Fear campaign, so-called, being run from 10 Downing Street. This nom-de-guerre was

2 nd co-founder of Brexit group Grassroots Out, said the Government’s comments were “part of Project Fear”. He told the PoliticsHome website: “Clearly any EU citizen that is legally

3 dramatic, in a long series of warnings issued by the Government as part of the so-called “Project Fear” agenda. Leave campaigners say that it is impossible to forecast economic ev

4 uperstate, the British voted for independence on June 23. To no one’s very great surprise, Project Fear turned out to be a giant hoax. The markets were calm. The pound did not coll

5 cy, access to the single market is www.critical.No doubt I’ll be accused of being part of “Project Fear” for raising these concerns, of lilylivered cowardice, of “talking the econo

6 od reason why Nicola Sturgeon, a faux supporter of the Remain campaign, has lambasted its “Project Fear” tactics. It’s because Scotland’s First Minister knows from experience that

7 nservative Party over Europe and lead to further allegations of scaremongering as part of “Project Fear” from those campaigning for a Leave vote. Speaking ahead of the report’s pub

8 isolation runs deep. The Ottoman shadow is long. The Justice Secretary is right to dismiss Project Fear as craven and defeatist. A vote to leave the dysfunctional EU half-way house

9 ould not be worse off in the event of a Brexit vote, in a blow to the assembled forces of “Project Fear”. The findings came despite warnings issued by the Treasury, the Bank of Eng

10 ields have fallen following the publication of pro-Brexit polls as evidence that so-called Project Fear is not working. Yesterday morning Douglas Carswell, the UK Independence Part

11 s but does not mention them. Why not? Ian Musgrave Laxey Isle of Man SIR - The latest from Project Fear is a possible £40 billion black hole, with which to threaten pensioners. Wh

12 r cent of companies traded with EU countries, yet they were all affected by EU regulation. Project Fear Mr Johnson said the difference between the Leave and the Remain campaigns wa

13 vice I can give is to relay what the Prime Minister himself said before the whole logic of Project Fear kicked in. He said there was no doubt Britain could thrive mightily and we c

14 ain can sink. It is another own-goal. Andrew Warbrick Market Bosworth, Leicestershire SIR - Project Fear is now targeting the elderly. Why doesn’t Mr Cameron go the whole hog and su

15 the various camps. I had expected to come home to find Remain firmly in the driving seat, Project Fear having done its worst, but far from it. The most recent polls show a marked

Tab. 2.A selection of concordance lines: Project Fear

←38 |

8‘Non-logical’ delegitimising argument strategies

Partington / Taylor (2018) list a number of ‘non-rational’, ‘non-logical’ argument strategies which can be used to delegitimise an opponent, included the ad hominen (also known as ad personam), false parallel and false dichotomy. Non-rational they might be, but their frequent employment suggests they often have some psychological efficacy.

8.1Ad hominem

Some of the examples included in the previous section could be counted as ad personam which consists of attacking the personality, ethos, motivation, and so on, of an opponent rather than dealing with their argument. Partington / Taylor (2018) invite us to consider whether an ad hominem argument, such as alluding to a person’s private life, can be used to detract from their argument. The example they ask is whether Thomas Jefferson personal ownership of slaves undermined his message that ‘all men are created equal’? Does it make the aspiration less real, or the notion less true?

8.2False parallel

As an instance of a false parallel they cite the controversial headline ‘global warming deniers are now on a par with Holocaust deniers’ (Boston Globe 09/02/2006). The false parallel kind of delegitimisation has two elements, a questionable comparison accompanied by exaggeratedly negative evaluation. Somewhat similar false parallels were collected by searching Lexis Nexis for the expression the New Nazis, in the three years preceding the writing of this study. In the view of some argument-maker, all the following were accused of being today’s ‘Nazis’:

progressives, liberals, the Democrats, ACLU (American Civil Liberties Union)

the Republicans

feminists (feminazis)

←39 |

Iran and the ayatollahs



the (UK) government

Here are a few examples, collected from Lexis Nexis (my emphasis):

(25)A BBC News reporter is being investigated over a Facebook rant which branded the Tories the “new Nazis”. (UK, Daily Mirror 2016)

(26)Liberals are the new Nazis,” someone scrawled this month in both of the restaurant’s bathrooms. (US, The New Yorker 2018)

(27)the GOP is a fascist party and that Christian conservatives are the new Nazis (Newstext blogs 2017)

(28)Feminazi: the go-to term for trolls out to silence women; Last week, a young female barrister was called a ‘feminazi’ after complaining about a sexist message sent via LinkedIn. (Headline, UK, Guardian 2019)

Most frequent of all, however, was the following set:

Zionists, Israelis, Jews, ‘Zionazis’

Of the 123 documents containing ‘the new Nazis’ recovered by Lexis Nexis, from newspapers and blogs, in this period, 61 related to this accusation – nearly always attributing it to some other source, rather than averring it, but its frequency suggests another example of forced priming (Duguid 2009a), of one or more parties flooding the public discourse with a particular set of lexical terms to construct a narrative, namely that Israelis and maybe Jews in general are ‘the new Nazis’.

(29)In the new anti-Semitic circles and the BDS movement, the establishment of the State [of Israel] is called the Opposite Holocaust, in which the Jewish Nazi victims became the new Nazis (Israel, Arutz Sheva 22/04/2018)

(30)Mr Goldstein said there is ‘no safe space’ left for Jews inside Labour and wherever they go Jews are told ‘Rothschilds run the world, ISIS is a fake front for Israel and Zionists are the new Nazis’. (UK, Mail Online 26/03/2018)

A UNESCO press release (also found on Lexis Nexis news) explains the pseudo-logical steps in this delegitimising antisemitic false parallel:

←40 |

(31)In the view of all these anti-Jewish groups – recycling classical anti-Semitic themes – Jews control the press, own the banks, manipulate the markets, corrupt politics and politicians, and, via Israel, oppress other groups and peoples. More significantly, it is believed that the Jews conspire, à la the Protocols of Zion, to dominate the world for their own benefit. In particular, this self-interest is now understood to mean primarily the interests of the State of Israel […] The Jews are the “new Nazis,” and all the traditional anti-Jewish canards are recycled. (UNESCO Press Release, 23/01/2017).

Such a narrative combines exaggerated delegitimisation with the sheer racist hatefulness of identifying Jewish people with the ideology which tried to eliminate all Jews from Europe and beyond.

8.3False dichotomy

The delegitimisation of people, whole groups and their arguments, can also be effected by using the argumentative strategy of false dichotomy, also known as the excluded middle, in which complex situations are depicted as a binary evaluative choice, in particular, between good (generally agreeing with the argument maker) and evil (disagreeing with them). Partington / Taylor (2018) note a case such as President George W. Bush’s statement ‘Every nation, in every region, now has a decision to make. Either you are with us, or you are with the terrorists’ (20/09/2001), which was later echoed by President Erdogan of Turkey ‘Pick a side. You are either on the side of the Turkish government, or you’re on the side of the terrorists’ (12/01/2016).

The term ‘false dichotomy’ was searched using Lexis Nexis and we found that the delegitimising function of this argument strategy was occasionally given explicit attention:

(32)We Turks are being asked to choose between “stability” and “democracy”. But this is a false dichotomy. Those who believe in it have learned nothing from history. They have not understood that undemocratic nations are unhappy nations, and unhappy nations cannot possibly be stable. (UK Guardian 2017)

(33)The Tyranny of False Dichotomies

When, in the aftermath of the shooting incident in Mandaluyong City, Bato dela Rosa said he would rather have stupid people with good intentions instead of smart people with bad intentions, he was making use of a rhetorical device widely deployed in our political discourse: presenting false ←41 | 42→dichotomies, or making people pick between two things as if these were the only choices.

A similar argument is presented in defense of Rodrigo Duterte’s foul mouth: What would you rather have, a president who curses but cares for the country, or a president who speaks in flowery language but doesn’t care for the people? Or: Which would you choose – a president who jokes about rape, or a president who is ‘raping’ the nation?

[…] We see this as well when people’s political stances are elicited: Are you ‘pro’ or ‘anti’? There are so many issues facing the nation today, but because people simply assume that being ‘Dilawan’ or ‘DDS’ means opposing or supporting everything the government says or does, no middle ground, no nuance, is to be found. Here, the false dichotomy is between criticism and support.

For politicians themselves, the use of false dichotomies is an effective tool to galvanize their base and defend their tenuous positions. But for the country, the result is further polarization – and the oversimplification of the debates that matter. (Philippines Daily Inquirer 2018)


The first part of this Chapter was dedicated to an overview of the state of the art of Corpus-assisted Discourse Studies, including a discussion of both the added values it brings and also the issues it can pose. The second part contained a discussion of the notion of delegitimisation in argumentation. We then moved on to examine some of the argument strategies and even structures used to attempt to effect delegitimisation, using as evidence data deriving from authentic sources, mostly viewed or downloaded from Lexis Nexis and COCA.

As mentioned above, the topic of the present research is somewhat unusual. CaDS excels in uncovering ways in which various sources represent and/or frame some group or issue e.g. how the media represent immigration (Taylor 2014) or Muslims living in the UK (Baker et al. 2013) or how right-wing groups construct a worldview in which minorities are negatively evaluated (Brindle 2016). Some work has been conducted on speech acts such as apologising (Duguid 2015), responsibility avoidance in answering question (Duguid 2009b) and interactive ←42 | 43→teasing in press briefings (Partington 2006: 144–181). In contrast, little work in CaDS has been dedicated to researching large scale ‘macro’ argument strategies such as delegitimisation, given corpus linguistics’s historical tendency to look at large quantities of relatively ‘micro’-contexts. As was seen here, in order to study language structures at a more macro level, corpus techniques are, of necessity, very much assisted by the researcher’s intuitions, not to mention knowledge of the world.


Aristotle. 2012. The Art of Rhetoric. London: Collins.

Baker, Paul/Gabrielatos, Costas/McEnery, Tony. 2013. Discourse Analysis and Media Attitudes: The Representation of Islam in the British Press. Cambridge: Cambridge University Press.

Blaxill, Luke. 2018. Introduction. From ‘Post-Democracy’ to ‘Post-Truth’ in Political Language. Textus 31(1), 57–66.

Brindle, Andrew. 2016. The Language of Hate: A Corpus Linguistics Analysis of White Supremacist Language. New York: Routledge.

Brown, Penelope/Levinson, Steven. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press.

Cockcroft, Robert/Cockcroft, Susan. 2005. Persuading People: An Introduction to Rhetoric. London: Palgrave Macmillan. 2nd ed.

Davies, Mark. 2009. The 385+ million word Corpus of Contemporary American English (1990–2008+). International Journal of Corpus Linguistics 14(2), 159–190.

Demata, Massimiliano. 2018. ‘I think that maybe I wouldn’t be here if it wasn’t for Twitter’. Donald Trump’s Populist Style on Twitter. Textus 31(1), 67–90.

Duguid, Alison. 2009a. Insistent voices, government messages. In Morley, J./Bayley, P. (eds.) Corpus-Assisted Discourse Studies on the Iraq Conflict: Wording the War. London: Routledge, 234–260.

Duguid, Alison. 2009b. Loud signatures: Comparing evaluative discourse styles-patterns in rants and riffs. In Romer, U./Schulze, R. (eds.) Exploring the Lexis-Grammar Interface. Amsterdam/Philadelphia: John Benjamins, 289–315.

←43 |

Duguid, Alison. 2015. Public apologies and media evaluations. In Bondi, M./Cacchiani, S./Mazzi, D. (eds.) Discourse In and Through the Media. Recontextualizing and Reconceptualizing Expert Discourse. Newcastle: Cambridge Scholars Publishing, 146–169.

Fukuyama, Francis. 1992. The End of History. New York: Free Press.

Halliday, Michael/Matthiessen, Christian. 2004. An Introduction to Functional Grammar. London: Arnold. 3rd ed.

Hoey, Michael. 1983. On the Surface of Discourse. London: George Allen and Unwin.

Hoey, Michael. 1991. Patterns of Lexis in Text. Oxford: Oxford University Press.

Hoey, Michael. 2005. Lexical Priming: A New Theory of Words and Language. London: Routledge.

Hunston, Susan/Thompson, Geoff (eds.) 2000. Evaluation in Text. Oxford: Oxford University Press.

Huntington, Samuel. 1996. The Clash of Civilisations. New York: Simon & Shuster.

Louw, William. 1993. Irony in the text or insincerity in the writer? The diagnostic potential of semantic prosodies. In Baker, M./Francis, G./Tognini-Bonelli, E. (eds.) Text and Technology. In honour of John Sinclair. Amsterdam/Philadelphia: John Benjamins, 157–176.

Marchi, Anna 2018. Dividing up the data: epistemological, methodological and practical impact of diachronic segmentation. In Taylor, C./Marchi, A. (eds.) Corpus Approaches to Discourse. London: Routledge, 174–196.

Marchi, Anna/Taylor, Charlotte. 2009. If on a winter’s night two researchers … a challenge to assumptions of soundness of interpretation. CADAAD 3(1), 1–20.

Mautner, Gerlinde. 2016. Mad about method. Challenges and opportunities at the CDA/CL interface. Plenary talk at the 3rd Corpora and Discourse International Conference. Pontignano (Siena), Italy, 02.07.2016.

Napoletano, Antonella/Aiezza, Maria Cristina. 2018. The press war in the post-truth era: A corpus-assisted CDA of the discourse of US political analysts on Trump’s figure and policy. Textus 31(1), 91–118.

Nelson Francis, William. 1982. Problems of assembling and computerising large corpora. In Johansson, S. (ed.) Computer Corpora in ←44 | 45→English Language Research. Bergen: Norwegian Computing Centre for the Humanities, 7–24.

Orwell, George. 1946. Politics and the English Language. Horizon 13, 252–265.

Orwell, George. 1949. Nineteen Eighty-Four: A Novel. London: Secker & Warburg.

Partington, Alan. 2006. The Linguistics of Laughter. London: Routledge.

Partington, Alan. (ed.) 2010. Modern diachronic corpus-assisted discourse studies on UK newspapers. Corpora 5(2), 83–108.

Partington, Alan. 2017. Varieties of non-obvious meaning in CL and CADS: from ‘hindsight post-dictability’ to sweet serendipity. Corpora 12(3), 339–367.

Partington, Alan/Taylor, Charlotte. 2018. The Language of Persuasion in Politics. London: Routledge.

Partington, Alan/Duguid, Alison/Taylor, Charlotte. 2013. Patterns and Meanings in Discourse: Theory and Practice in Corpus-assisted Discourse Studies. Amsterdam/Philadelphia: John Benjamins.

Singh, Shiv/Lurthra, Rohini. 2019. Savvy: Navigating Fake Companies, Fake Leaders and Fake News in the Post-Trust Era. Kindle Edition.

Taylor, Charlotte. 2010. Science in the news: a diachronic perspective. Corpora 5(2), 221–250.

Taylor, Charlotte. 2014. Investigating the representation of migrants in the UK and Italian press: A cross-linguistic corpus-assisted discourse analysis. International Journal of Corpus Linguistics, 19(3), 368–400.

Taylor, Charlotte. 2018. Similarity. In Taylor, C./Marchi, A. (eds.) Corpus Approaches to Discourse. London: Routledge, 19–37.

Thompson, Geoff. 1996. Introducing Functional Grammar. London: Arnold.

Thompson, Mark. 2016. Enough Said. What’s Gone Wrong with the Language of Politics? Bodley Head: London.

Trump, Donald. 2015. Crippled America: How to Make America Great Again. New York: Threshold.

Zuccato, Matilde/Partington, Alan. 2018. Brexit: before and after. A corpus-assisted study of the Referendum campaigns and the immediate aftermath. Textus 31(1), 119–140.

←45 |

1OED, s.v. serendipity, (n): ‘The faculty of making happy and unexpected discoveries by accident’.

2Partington et al. (2013: 335–9) report their surprise at how, in a case study of the UK’s Daily Telegraph, they found that there were more positive than negative opinions expressed on immigration. However, they also noted that these mainly regarded its positive contribution to the UK economy, rarely any cultural benefits.

3The term derives from the name of the Jewish Hungarian physician, Ignaz Semmelweis who, while working and researching in the Vienna General Hospital in 1847, discovered that childbed fever mortality rates fell ten-fold when doctors washed their hands between patients, especially after conducting an autopsy. Semmelweis’s suggestion of simple hand-washing was frequently met with derision by fellow doctors who, after all, were the medical experts, Austrian gentlemen and Gentiles to boot. Semmelweis was finally vindicated by the gradual acceptance of the pathogen or germ theory of disease transmission in the second half of the 19th century. He is rightly remembered as the Saviour of Mothers.

4The ‘Overton window’ is the range of ideas tolerated in public discourse, also known as the window of discourse. The term is named after Joseph P. Overton (Wikipedia).


6The Washington Post online: https://www.washingtonpost.com/news/the-fix/wp/2017/02/17/trumps-war-with-the-media-isnt-new-thomas-jefferson-railed-about-newspaper-lies-too/

7Humbert Wolfe, 1930, in The Uncelestial City. London: Gollancz.

←46 | 47→

Tony McEnery, Helen Baker and Carmen Dayrell

Analysing the impacts of 19th-century drought: A corpus-based study


In the 19th century, the United Kingdom had no equivalent institution to the Environment Agency and there was no national system which formally recorded occurrences of drought. Consequently, environmental scientists have much to learn about the nature, extent and representation of drought events in 19th-century Britain. In this study, we will explore whether it is possible to use corpus-based discourse analysis to reconstruct climate records and hence fill in some of these holes in our knowledge, specifically through the application of corpus methods to historical newspaper data. Using this approach we want to establish that corpora can help us to i) identify drought events; ii) construct a narrative of those events and iii) understand the impacts that drought had on society at the time.

The climate of the United Kingdom is often described as wet or rainy but, in reality, the country is vulnerable to water shortage.1 Droughts are not necessarily caused by lack of rainfall alone and can be worsened by variables such as the rate of water consumption and the efficiency of the water-supply infrastructure. Expanding our knowledge of historical climate conditions will enable hydrologists to ←47 | 48→uncover patterns in drought events of the past and to better understand public attitudes surrounding water scarcity, thus improving our resilience to drought as a society.

To achieve these goals we use close reading, via the scrutiny of concordances and scans of the original hard copies of the newspapers, both to establish data quality and to enable a qualitative analysis of the material. However, that close reading is guided by monthly and yearly frequency distributions to ascertain when references to drought peaked in the newspaper data. It is also supported by collocation analysis and an innovative technique named concordance geoparsing which enables us to focus on droughts which occurred in Britain.

2The weather in the past

In 19th-century Britain, favourable weather was a crucial factor in securing the wellbeing of the population. In 1892, the Glasgow Herald emphasized the importance of a good growing season: “The difference between a good harvest and a bad one means the difference between comfort and misery for a very large proportion of our population, and it is felt more or less by all classes, although they may fail to understand the causes of the difference they feel.”2 A poor harvest would not cause Britons to starve, but it would impact their lives in a myriad of ways, starting with the small stock farmer and extending to regular householders and business owners. Communities in the past were more vulnerable to the effects of drought because they relied upon local sources of water and a scarcity of water entailed an increased risk of water-borne diseases.3 Waddington (2016: 5) has explained how rural communities in particular relied on archaic water supplies even towards the end of the century, often obtaining water from natural sources such as ←48 | 49→wells, springs, or stored rainwater, rather than from commercial companies or the local authority.4

Despite droughts being a regular feature of the British climate, they have received little scholarly attention.5 However, there have been attempts to supplement our knowledge of historical droughts. For instance, Todd et al. (2013) have reconstructed precipitation and temperature series in order to generate drought data from 1697, 1726 and 1767 to 2011 for three regions in south-east England. Todd et al. (2015) have constructed the rainfall record for the lowlands around Carlisle in north-west England in the period 1757–2012. Attempts to reconstruct drought records by researchers in disciplines outside of climatology are rarer. A notable exception is Stone (2014), who has studied account rolls of manors in 14th-century England to uncover how drought impacted communities in southern and eastern English counties.

3Data used in this study

This study uses data from eight digitized newspapers taken from the British Library Newspapers collection, Part I: 1800–1900, which forms part of the British Library Newspapers series.6 Tab. 1 details the newspapers which are included, the dates of publication for each title, including any missing years, and the number of words and texts in each title.7

Dates of publication in 19th century


The Era

1838–99 (1848, 1849, 1850 missing)

448,894,847 words in 3,096 texts

Glasgow Herald

1820–99 (1823–5, Feb 1827-Dec 1843, 1872–8, May-Dec 1879 missing)

1,780,103,461 words in 13,034 texts

Hampshire Telegraph and Sussex Chroniclea

1800–99 (1879 missing)

531,379,257 words in 6,197 texts

Ipswich Journal

1800–98 (1829, 1831–1832, 1897, 1899 missing)

424,191,798 words in 7,000 texts

Northern Echo

1870–99 (1871, 1872, Jan-Jun 1897, 1898 missing)

386,590,121 words in 8,543 texts

Pall Mall Gazette

1865–99 (1872, Jan-Mar 1873 missing)

468,324,154 words in 10,620 texts

Reynold’s Daily

1850–99 (none missing)

289,617,880 words in 2,635 texts

Western Mail

1869–99 (1872, Jul 1875–Jun 1876, 1896, Jul 1897–Aug 1898 missing)

756,039,060 words in 8,718 texts


5,085,140,578 words in 59,843 texts

Tab. 1:The dates and scale of 19th-century newspapers taken from the British Library newspaper archive

a The Hampshire Telegraph and Sussex Chronicle underwent several name changes. Between 1799 and 1802, it was known as the Portsmouth Telegraph; in 1802 it changed its title to Mottley’s Telegraph and Portsmouth Gazette and then, in the same year, to the Hampshire Telegraph and Portsmouth Gazette. Between 1803 and 1899 it was known as the Hampshire Telegraph and Sussex Chronicle.

←49 | 50→

In order to optimize our access to over 5 billion words, we used the powerful corpus analysis system, CQPweb, to search and manipulate these newspaper corpora.8 In the first instance, this enabled us to investigate a more complex search term: ‘drought OR droughts OR droughty_JJ OR droughted OR drought-* OR drouth OR drouths’.9 The ←50 | 51→British Library Newspapers website generates search results consisting of a list of article titles which must be perused in full in order to glean their content. CQPweb, on the other hand, provides concordances, lists of each occurrence of a search term presented with words to the left and right of it, which can be examined in brief much more quickly.

Fyfe (2016: 8) has argued that digitized collections should be approached as “distinct objects with histories of their own”. The digitized British Library 19th-century newspapers tend to include omissions and errors. This is due to the way in which digital texts are generated from images of text using optical character recognition (OCR). Kettunen (2015: 95–96) has described why older newspapers pose challenges for OCR software due to the appearance of the typeface, the quality of the original source and microfilm, the layout of the page, the scanning resolution and file format and so on. The reproduction of newsprint in CQPweb is thus often littered with extra or incorrect letters and omitted letters or words, all of which can generate considerable noise for an analysis to overcome.

Accordingly, a significant drawback of OCRed material is that the errors it contains have an adverse effect upon the accuracy of searches: in other words, a search for drought will almost certainly fail to highlight every appearance of this term in the newspapers under investigation but will also include a number of articles in which words have been wrongly understood by OCR software as spelling drought.10 The main way in which we overcame problems caused by ←51 | 52→OCR errors was to undertake close reading which enabled us to apply qualitative assessments to the newspaper material at every stage of our analysis.11

4Focusing on British droughts

The ways in which newspapers talk about drought can be divided into three broad categories: i) droughts which are reported as occurring in Britain; ii) droughts which are reported as occurring outside of Britain; and iii) references to drought which are mentioned in a more general context. We wished to focus upon droughts which occurred within Britain, so it was preferable to exclude those newspaper articles which referred to drought outside of Britain. Manually sorting the data to exclude foreign references to drought would have been a mammoth undertaking but the application of a technique named concordance geoparsing enabled us to improve this process substantially, both in terms of accuracy and speed.

Concordance geo-parsing is a methodological approach which combines corpus linguistics with GIS (Geographic Information Systems) methods.12 The technique is achieved by means of a number of stages. Firstly, we used corpus linguistics software to extract each occurrence of a search term and a span of twenty words to the left and right of it. These concordances were then geo-parsed in order to identify instances of place-names which were annotated with geo-spatial co-ordinates. The results were manually analysed in order to reduce errors. By sorting results in terms of longitude and latitude, we were ←52 | 53→then easily able to lift out drought concordances which included references to places outside the UK.

Using the categorized query function of CQPweb and our original search term, we created a query with two separate categories. The first category was composed of the drought concordances which included place-names belonging outside of the UK and the second one comprised of all other concordances generated by our search term, in other words, those which referred specifically to drought events within the UK and more general references to drought which were not connected to a specific place. For this study, we used data drawn from the second category thus effectively excluding references to drought which were mentioned in relation to places outside of the UK.

5Searching for droughts in the past – frequencies of drought in 19th-century newspapers and what this can tell us

Figures 1a and 1b shows how many times drought was mentioned in six British newspapers over the course of the 19th century. The data was normalized in order to compare frequencies per million words. The graph provides an indication of peak periods where newspapers tended to be more interested in discussing drought.

For each newspaper, we collated the ten years which had the highest frequency of drought mentions per million words. We found that seven years – 1868, 1870, 1874, 1887, 1893, 1895 and 1896 – appeared as one of these top ten peak years in at least four newspapers. These years correspond with periods of probable drought which have been identified by researchers in hydrology (Cole/Marsh 2006) based on the incomplete records that exist. Our findings add credence to theirs by means of triangulation. Clearly, then, we have found evidence that newspapers talk of drought more frequently during periods of attested drought; this is a hypothesis which speaks to both the news values of the time (droughts were newsworthy) and the consequent benefit of the newspaper data for exploring such issues as extreme weather ←53 | 54→events. Although this is not an earth-shattering conclusion, it does provide reassurance of the utility of our method. However, we are able to go further than Cole/Marsh (2006) because, by digging into our news texts further, we can look for effect, which increases our understanding of the nature and impact of these droughts by recovering the media narrative of drought from the period.13

Fig. 1a:Mentions of drought in three UK newspapers, 1865–1899, excluding non-UK specific references. The start date shown on Figure 1a is 1865 because these three newspapers did not start publication until the second half of the 19th century

6Recovering drought narratives

The newspapers we considered were published daily, biweekly or weekly and this means that we can access information regarding ←54 | 55→changing weather patterns in considerable detail.14 For example, by studying monthly frequency data, we can trace the progress of one particular drought and identify the months of the year when UK newspapers discussed it the most. This information is particularly helpful for environmental scientists as the characteristics of a drought can partly be explained by what point during the year it was most intense – for instance, in recent times, a drought of 2010–12 peaked in early spring so had an intense impact on water resources.15 Regional newspapers are of great use for this task as they can give us an indication of when the populations of certain areas of the country were particularly interested in drought. It is probably fair to assume, for instance, that the Glasgow Herald would include more references to drought in months when Scotland itself was experiencing a dry spell. A second approach we can take is entirely qualitative, reading articles which have been highlighted as ←55 | 56→being concerned with drought, and constructing a narrative of a year of drought. We will demonstrate both of these approaches below.

Fig. 1b:Mentions of drought in three UK newspapers, 1800–1899, excluding non-UK specific references

Fig. 2:Mentions of drought in 1893 by month

7The Great Drought of 1893 – by month

The drought of 1893 was the subject of a scholarly article by Brodie (1894) and was described in retrospect as the “great drought of 1893” by the Northern Echo.16 Figure 2 shows the numbers of mentions of drought by month for 1893 in six newspapers using data which excludes non-UK references.

Figure 2 shows that the northern-based newspapers, the Glasgow Herald and the Northern Echo, mention drought more often in June, whereas the search terms peak slightly earlier, in May-June, for the Cardiff-based Western Mail and the Ipswich Journal. It appears that the drought of 1893 affected the south of England a few weeks before it hit the north. Interestingly, the newspapers seldom discuss the drought at the end ←56 | 57→of the summer but mentions pick up a little in early autumn, perhaps as a result of increased interest due to harvesting.

By exploring the concordances of 1893 using close reading, we can illuminate which parts of the country exhibited anxieties concerning the dry weather first. The first hint of the approaching difficulties was provided by the Ipswich Journal in an article about a local cycling club on 22 April 1893. In the same issue, readers were informed that the weather had been “altogether delightful” but that the lack of rain was becoming increasingly serious.

The dry weather was being felt in Wales by late April. The Western Mail reported that scarcity of water was affecting several Welsh metal works and that work might be forced to stop if the drought continued.17 Although Cardiff was said to still have a large storage of water in its reservoirs, the supply of water in Merthyr, further north, was under threat to the extent that it was going to be cut off to businesses that used it for machinery and motive power, including railway companies.18 Meanwhile, the residents of Treboeth, to the north of Swansea, were very short of water and were concerned at the unsanitary state of their neighbourhood and the associated risk of cholera.19 By 5 May, the Western Mail was reporting that Llanelli, to the east of Merthyr, was approaching a state of water famine and its residents were being visited by detectives organized by the borough supervisor intent on prosecuting those who were found to have wasted water. Here we have an early indication of who might be deemed responsible for the coming drought.

Lack of rain continued to affect the south of England. At the end of April, the Hampshire Telegraph reported that the drought had continued ←57 | 58→for sixty days, unparalleled in the records of modern meteorology, and was causing uneasiness.20 The Glasgow Herald noted that while rainfall had been adequate in Scotland and Ireland, it continued to be lacking in England and Wales.21 Our hypothesis – that the south and south-west of England were affected by the drought of 1893 before the north of England and Scotland – appears to be accurate. However, the drought in England did have a knock-on effect for Scotland. For instance, cattle breeders in the south of Scotland were reportedly finding it difficult to find buyers for their stock due to English farmers avoiding taking on more animals which would be difficult to maintain during drought.22

On 11 May, the Western Mail reported that tinplate manufacturers in South Wales were sinking wells of their own in order to access water. In the densely populated counties of Glamorgan and Monmouth, the watercourses were described as foul and a threat to public health. Medical experts feared outbreaks of cholera and fever, and it was reported that several people had suffered food poisoning in Cardiff due to eating pork pies. The reporter explained that this was caused by ptomaine poisoning attributable to the weather and urged food inspectors and tradesmen to observe greater vigilance. In the same issue, readers were told that the Bishop of St. David’s had recommended the use in Divine Service of the prayer for rain.23 Public prayers for rain were repeated in Wales on 15 May.24 Readers were exhorted to ration their use of water; take care when choosing foodstuffs; and, as a last resort, appeal to a higher authority for rain. This emphasis on personal responsibility was not unusual: “Not to be as economical as possible of water where supplies seem likely to run short is a violation of social morality.”25

←58 | 59→

By May 1893, the drought was impacting on the lives of people residing in England to a considerable extent. A local intelligence column reported that remote districts of Hampshire were charging twopence half-penny for a single pail of water and that between 5 and 6 millions of gallons were being imported into Portsmouth from the works at Farlington and Brockhampton. The directors of the water company were exhorting members of the public not to waste water. Another article in the same paper added that farmers in North Hampshire were being obliged to travel three or four miles to obtain water for their cattle.26 The Agricultural Correspondent for the Ipswich Journal wrote that he had witnessed indications of drought in Suffolk with farmers unable to sow their fields with barley due to the dryness of the earth.27

References to drought sometimes increased when it had been broken and it appears that this dry spell was interrupted in some parts of the country during May.28 For instance, on 23 May, the Western Mail told the story of a special meeting to pray for rain at a chapel in Barry being interrupted by a heavy shower. The shopkeepers present must not have had much faith in their prayers receiving an immediate response because they had hung perishable goods outside their businesses before attending the service: “…they dallied awhile, and then there was a stampede”.29

These welcome showers did not mark an end of the drought. On 10 June, the Ipswich Journal noted that the drought had continued in Berkshire and Reading.30 The hay crop in East Anglia was reported as being the poorest on record; crops were short in East Essex and there was a complete shortage of feed for cattle.31 On 14 June, the Western Mail ←59 | 60→reported many cases of diphtheria and scarlet fever in North Wales which were attributed to the drought. The Commercial Dry Dock at Cardiff had been utilized as a storage reservoir causing inconvenience at the port. The Thames, meanwhile, had become so low by Glover’s Island, Richmond Bridge, and Brendford Eyot, that it was possible to cross the river by foot by jumping from one exposed strip of bed to another.

In Figure 2, the peak month of drought mentions for the Glasgow Herald is July and, accordingly, by 16 June, drought appeared to have reached Scotland. A report citing George Croucher, the observer at the Ochtertyre Station of the Scottish Meteorological Society, noted that there had not been such a deficiency of rainfall since 1855. The Kilbirnie Water Committee requested that the public did not waste water and that all pipes and fittings were in proper order.32 In mid-July, the Northern Echo reported that the Mugdock Reservoir in Glasgow had fallen 4 foot in one day.33 By June, the north and midlands of England were also taking their share in the suffering. A number of riveters employed in a local shipyard were compelled to stop working because of the intense heat.34 The authorities in Birmingham, “usually one of the last places to feel the effects of drought upon the water supply”, had decided to reduce the street water supply by half.35

As in May, showers in late June misleadingly suggested that the drought might be at an end.36 On 3 July, the Glasgow Herald informed readers that the water supply in Kirnintilloch, north of Glasgow, was being cut off for 17 hours between 1pm and 6am. It insisted that the situation in Scotland was becoming increasingly serious for agriculture.37 On 19 July, the Glasgow Herald included a lengthy report on the harvest prospects of Scotland. The corn crop was not expected to do well. Oats, the most important cereal crop in Scotland, had fared better in places where the crop had been sown earlier and in richly manured lands, but was predicted to have a deficiency of around 15 % throughout ←60 | 61→the region. The reports of barley and wheat were more favourable and the root crops, potato and turnip, were thriving. The article noted that hay prices had risen due to the total destruction of crops by drought in parts of England and the Continent and were likely to be nearly twice the price of the previous year.

By mid-July it was reported that the midlands, southern and western counties of England were suffering a shortage of milk due to reductions in cattle numbers.38 The Western Mail reported the sad story of a yeoman from the Mendip Hills district of Somerset who was forced to shoot his remaining cattle after two died from starvation.39 In early August, the Hampshire Telegraph advised readers that the waterworks company was stopping the supply of water at 8pm every evening.40 In the same month, a column on grouse shooting in the Northern Echo noted that fires had destroyed acres of land during the drought in Snilesworth, Wass and Kilburn.41

The impacts of the drought of 1893 continued into the autumn. On 1 September it was reported that the Hartlepool Gas and Water Company had struggled to provide water for its domestic and business companies and, for the first time, they had been obliged to ask their customers to economize their use of water.42 The Glasgow Herald also noted the difficulties faced by the residents of Hartlepool, reporting that water for manufacturing purposes would be cut off, affecting employment.43 On 24 November, it was reported that in some parts of Pembrokeshire drinking water was difficult to procure.44 However, in general, mentions of drought in the newspapers noticeably reduce after the summer although the effects of the dry weather of 1893 continued to affect the water supply into the following year.

The analysis so far has clearly shown that corpus techniques can identify periods of drought in the past and, by moving between distant ←61 | 62→and close reading of the texts, we can reconstruct narratives of drought associated with a drought event in historical newspaper data. However, one thing that is clearly of importance in that narrative, as is apparent from the analysis so far, is the impact on the natural and social environment that the drought had. While close reading is an effective technique for reconstructing narrative, we explored the possibility that by focusing on collocation alone, we may be able to explore the impact of droughts in the newspaper corpus.

8The impacts of 19th-century drought

Shifting our focus to impacts and collocation allows us to explore whether the issues linked to the drought of 1893 that we uncovered mark a change in the kinds of things the newspapers wrote about when they mentioned British drought. To explore this we had hoped only to use collocation, but the quality of the OCRed data made this difficult. Consequently, we worked in a mixed mode, beginning with collocation but then further investigating discourses surrounding drought in the seven years when the subject was shown to be particularly salient in newspapers. This was a lengthy process which demanded a great deal of close reading and often entailed consulting scans of original articles at the British Library website to overcome the issues with the OCR data. The approach was effective, however. The matrix below summarizes the impacts of each drought as they were reported in the newspapers in the 1800s.

←62 |

To populate the matrix, if a newspaper mentioned a particular impact of drought in a particular year, then that box was ticked. In the public health category, we only checked a box if a particular illness had been specifically attributed to drought. In terms of reports of harvest failures, we only checked a box if a crop was reported as being deficient rather than likely to be deficient. This is important as the newspapers often carried reports of predicted failures of a particular plant which turned out to be nothing more than doom-mongering. Also, a checked box does not necessarily mean a crop was deficient throughout the ←63 | 64→entire country. For instance, the Glasgow Herald reported on 20 September 1893 that the hay crop was plentiful in Ayrshire and Renfrewshire but deficient in Dumfries but, in this case, we still checked hay for that year. In most cases, however, failures of the same crop were reported in several newspapers and in more than one place in the country. Lastly, we only checked a box when a report was specific about the type of crop affected. Vague references to disasters regarding ‘fruit’ or ‘cereal crops’ were disregarded. By studying the totals in the column at the end of Tab. 2, we can surmise that turnips and oats, followed by barley and corn, were probably the most vulnerable crops during dry years. Of course, the checked boxes belonging to these crops may also be indicative of their popularity with farmers. It is doubtful, for instance, that kohlrabi was planted to the same extent as turnips.←64 | 65→

Tab. 2:Impacts of drought in seven major drought years of the 19th century

* Data from the Glasgow Herald is missing for 1874 but evidence for drought in Scotland in this year was not found in the other newspapers.

The matrix not only tells us what kinds of things were affected by drought in 19th-century Britain, but gives us a good idea of what journalists of the time considered to be important to their readerships. For instance, while it is likely that the conditions of lawns were adversely affected by water shortage, there are only two mentions of this.45 ←65 | 66→Although 19th-century Britons seemed to care about the produce from their garden, they appear to have been less concerned about its appearance, or at least found it less newsworthy. While golf and grouse shooting were often mentioned as being negatively impacted by dry spells, otter hunting was only mentioned in one article and football achieved no references at all.46

What stands out is the consistency of results. All the newspapers concentrated on the effects of drought in the same wide areas: the growth of crops; cattle farming; the supply of water; public health; and sporting events. The totals in the bottom row of the table suggest that the drought of 1868, closely followed by that of 1893, impacted the country in the greatest number of ways. In both of these years, public health suffered to the largest extent. Droughts in every year covered by Tab. 2 appear to have affected almost every area of the country. The exception is the drought of 1874 which does not appear to have affected Scotland. However, because we do not have access to the Glasgow Herald for this year, we cannot confidently conclude this was definitely the case.


This study demonstrates how corpus linguistics in combination with concordance geoparsing can be used in order to uncover how often, and in what ways, 19th-century newspapers discussed droughts within Britain. It can also help researchers to reconstruct media narratives surrounding drought events. The digitisation of newspaper text and development of advanced corpus analysis systems make such an endeavour possible. The data was approached in a number of ways: we compiled frequency distributions to reveal peak years when drought was prevalent in public discourse; we matched the monthly frequency distributions of drought in 1893 to the narrative of the drought’s progress as ←66 | 67→told to us by 19th-century newspapers; and we compared the impacts of major 19th-century drought events by means of a qualitative analysis. Moreover, we were able to add a geographic dimension to our study. Not only were we able to exclude drought frequencies which related to countries outside of the UK by concordance geo-parsing, we used information from regional newspapers in order to chart the progress of major drought events as they moved through the country.

We found an overwhelming concordance among the newspapers we analysed in terms of when drought affected Britain and its impacts upon the British population. 19th-century Britons regarded drought with uneasiness and were preoccupied by its affects upon agriculture and cattle farming. The newspaper analysis suggests that the south of England and Wales suffered more at the hands of the 1893 drought than the north and Scotland.47 However, the wellbeing of British and Scottish agriculturalists was linked. If English farmers were unable to provide grass to feed their cattle, they would be unwilling to purchase livestock from Scottish farmers.

Drought affected the livelihoods of working people. Newspapers carried accounts of various manufacturing businesses being forced to close due to lack of water and correspondents often noted how many men were out of employment as a result. Yet, newspapers also reported tensions between local residents and businesses over who had first claim to the water supply in times of scarcity. A discourse of blame is also present in the drought articles of 1893. In its simplest form, this was directed at domestic consumers who were deemed guilty of water wastage due to carelessness and extravagance. Such a narrative was often initiated by water companies who accused their customers of over-use as part of a defensive strategy to deflect blame which may have otherwise been directed at themselves.48 Members of the public did not accept water shortage passively. The newspapers reported local ←67 | 68→communities mobilising in order to exhort water companies to increase their supply and even organising meetings to pray for rain.

This study has demonstrated that newspapers are a reliable means of identifying dry weather events, attested by the fact that many of the droughts mentioned in newspapers also matched existing records. It has confirmed our hypothesis that newspapers discussed drought much more frequently during droughty periods. We have been able to take advantage of the daily renewal of news stories as presented by the newspapers to uncover how a dry spell progressed on a day-to-day basis. By discovering when drought was most discussed in contemporary newspapers, we have a starting-point from which we can delve further into the news texts to uncover the effects and scope of each drought, highlighting common or unusual discourses along the way. In this way, we can offer new information to environmental scientists in terms of the nature of historical drought.


Baker, Paul/McEnery, Tony/Gabrielatos, Costas. 2013. Discourse Analysis and Media Attitudes: The Representation of Islam in the British Press. Cambridge: Cambridge University Press.

Brodie, Frederick. 1894. The great drought of 1893, and its attendant meteorological phenomena. Quarterly Journal of the Royal Meteorological Society 20(89), 1–30.

Cole, Gwyneth/Marsh, Terry. 2006. The impact of climate change on severe droughts: Major droughts in England and Wales from 1800 and evidence of impact. Bristol, UK: The Environment Agency.

Donaldson, Christopher/Gregory, Ian/Taylor, Joanna. 2017. Locating the beautiful, picturesque, sublime and majestic: spatially analysing the application of aesthetic terminology in descriptions of the English Lake District. Journal of Historical Geography 56, 43–60.

Fletcher, T. W. 1961. The great depression of English agriculture, 1873–1896. The Economic History Review 13(3), 417–32.

Fyfe, Paul. 2016. An archaeology of Victorian Newspapers. Victorian Periodicals Review 49(4), 546–77.

←68 |

Gregory, Ian/Ell, Paul. 2007. Historical GIS: Technologies, Methodologies and Scholarship. Cambridge: Cambridge University Press.

Gregory, Ian/Hardie, Andrew. 2011. Visual GISting: bringing together corpus linguistics and Geographical Information Systems. Literary and Linguistic Computing 26(3), 297–314.

Hardie, Andrew. 2012. CQPweb – combining power, flexibility and usability in a corpus analysis tool. International Journal of Corpus Linguistics 17(3), 380–409.

Joulain-Jay, Amelia. 2017. Corpus Linguistics for History: The Methodology of Investigating Place-Name Discourses in Digitised Nineteenth-Century Newspapers. PhD thesis, Lancaster University.

Kendon, Mike/Marsh, Terry/Parry, Simon. 2013. The 2010–2012 drought in England and Wales. Weather 68(4), 88–95.

Kettunen, Kimmo. 2015. Keep, change or delete? Setting up a low resource OCR post-correction framework for a digitized old Finnish newspaper collection. In Calvanese, D./De Nart, D./Tasso, C. (eds.) Digital Libraries on the Move: 11th Italian Research Conference on Digital Libraries, Bolzano, Italy. Bern: Springer, 95–103.

McEnery, Tony/Baker, Helen/Dayrell, Carmen. 2019. Working at the interface of hydrology and corpus linguistics: using corpora to identify unrecorded droughts in nineteenth-century Britain. In Baker, P./Egbert, J. (eds.) Using Corpus Methods to Triangulate Linguistic Analysis. New York: Routledge.

Murrieta-Flores, Patricia/Baron, Alastair/Gregory, Ian/Hardie, Andrew/Rayson, Paul. 2015. Automatically analysing large texts in a GIS environment: The Registrar General’s reports and cholera in the nineteenth century. Transactions in GIS 19, 296–320.


ISBN (Book)
Publication date
2021 (March)
Bern, Berlin, Bruxelles, New York, Oxford, Warszawa, Wien, 2021. 292 pp., 25 fig. col., 15 fig. b/w, 31 tables.

Biographical notes

Miguel Fuster Márquez (Volume editor) José Santaemilia (Volume editor) Carmen Gregori-Signes (Volume editor) Paula Rodríguez-Abruñeiras (Volume editor)

Miguel Fuster-Márquez is an associate professor of English at the Universitat de València . His teaching includes English lexicology, English sociolinguistics and corpus linguistics both at undergraduate and graduate levels. His fields of interest are related to the application of corpus techniques to different disciplines. José Santaemilia is a professor of English at the Universitat de València , where he teaches legal translation (English-Spanish/Catalan) as well as professional deontology and ethics and introduction to translation research. His areas of interest include gender and language studies, gender/sexuality and translation, and legal translation. Carmen Gregori-Signes is an associate professor of English at the Universitat de València . She teaches corpus linguistics, (critical) discourse analysis, and grammar both at undergraduate and graduate levels. Her research interests include corpus-assisted (critical) discourse analysis, and the representation of gender in telecinematic and media discourse. Paula Rodríguez-Abruñeiras is an associate professor of English at the Universitat de València . Both her teaching and her research have always revolved around the history of English. She is also interested in corpus linguistics, the new varieties of English, and gender studies.


Title: Exploring discourse and ideology through corpora