Category Archives: Working Papers

How language reveals different faces of loneliness?

Leave a reply

How language reveals different faces of loneliness

by Justyna A. Robinson, Gayathri Sooraj, and Caitlin Hogan

The issue of loneliness has been discussed in public platforms more frequently in recent years. Just within the past week newspapers have talked about student loneliness, male loneliness, loneliness epidemic. A recently published WHO Report on Social isolation shows that 1 in 6 people globally suffers from loneliness.  The data on loneliness in the UK shows that nearly half of adults feel lonely occasionally, sometimes, often, or always (see here). Some groups are especially at risk of loneliness, such as, young, disabled people (see here). However, the available reports rarely dive into exploring  how people themselves  describe loneliness or what language choice people make when they talk about  loneliness (but see here). These questions are explored by Justyna Robinson (Concepts Analytics Lab) and Faith Matcham (Psychology) at the University of Sussex , whose work towards creating a personalised chatbot for loneliness intervention was funded by Sussex Digital Humanities Lab. Ultimately, our headline findings are that  

  • Loneliness is generally experienced with similar intensity across demographic groups.
  • But the risk factors differ greatly, particularly according to age.
    • Social media is discussed as a catalyst for loneliness for younger people.
    • Outright social isolation is framed as a more pressing issue for older people.  

For this project we explored data on loneliness collected by Mass Observation Project (MOP). In 2019 MOP issued a survey, the directive on Loneliness and Belonging (see here). The directive consisted of two parts. Firstly, participants were asked to provide five words that they associate with loneliness, e.g. despair, fear, frustration, quiet, and sad (see Figure 1). Secondly, they provided long-form narrative responses to a series of questions related to the broader topic of loneliness.  

Loneliness in five words

70 Writers shared the five words they associate with loneliness. In total, 417 individual words were submitted (sentence responses excluded) with 206 different words.  The top responses included terms such as isolation, sadness, alone. All 206  words expressed various aspects of affective meaning which we measured by adopting the NRC Valence, Arousal, and Dominance (NRC-VAD) Lexicon classification (Mohammed 2025). Each word is assigned a numerical value, from 0 to 1, for each of the three dimensions depending on how strongly they represent a given dimension.

  • Valence refers to the positive (higher score) or negative (lower score) feelings associated with each word.
  • Arousal relates to the intensity of the emotion, with higher scores reflecting greater intensity.
  • Dominance refers to the degree of control exerted by a stimulus, with higher scores corresponding to greater control. To exemplify this, Figure 1 presents the words provided by a younger female of low socioeconomic status, alongside their scores across the three dimensions.
Figure 1: The valence, arousal, and dominance scores for five words provided by a younger female of lower socioeconomic status

Findings highlight remarkable consistency across demographic groups in their conceptualisation of loneliness across the dimensions of valence, arousal, and dominance. Figure 2 highlights that the average NRC-VAD scores for male and female respondents are almost identical.  

A scatter diagram for the valence, arousal, and dominance scores for male and female respondents
Figure 2: A scatter diagram for the valence, arousal, and dominance scores for male and female respondents
Click to explore how valence, arousal, and dominance vary by gender in this interactive visualization

Indeed, statistical analysis identified very few meaningful differences in the data set. This speaks to a remarkable homogeneity in the conceptualisation of loneliness measure by the NRC-VAD scores. The one exception to this was that on the dominance dimension there was an interaction between age and socioeconomic status. While age exhibited minimal differences among those in the lower status socioeconomic group, there was a much greater difference in the higher status group (see Figure 3). We consider age as a binary category for the purposes of this analysis, distinguishing between those who qualified for the state pension at the time of data collection (the state pension age was 65 in 2019) and those who did not.  

Figure 3: The interaction between age and socioeconomic status for the dominance scores

Figure 3 shows that the effect of age on the experience of loneliness is much greater among higher status respondents, with those who are older in this category conceptualising loneliness as something that exerts a greater degree of control (0.34), relative to their younger counterparts (0.19). In particular, there is a large difference between status among younger respondents, with those of lower status (0.28) conceptualising loneliness as exerting greater dominance than their higher status counterparts (0.19). 

Loneliness in long-form narratives

The longer, narrative responses to the directive discussed experiences of loneliness both on a personal level and at a broader societal level as well as speculation as to the causes of loneliness as well as potential solutions. In order to understand the ‘aboutness’ of the data we look at key terms.  

The outputs of modified ‘keyword’ analysis (see our blog about modified keyword analysis here) are presented in Table 1 and the key multi-word terms in Table 2 with the responses to the directive compared against the Ententen 2020 corpus as a baseline (see here). 

The top 25 keywords in the responses to the Loneliness & Belonging directive
Table 1: The top 25 keywords in the responses to the Loneliness & Belonging directive
The top 25 multi-word terms in the responses to the Loneliness & Belonging directive
Table 2: The top 25 multi-word terms in the responses to the Loneliness & Belonging directive

While a full analysis of these results is far beyond the scope of the blog here we consider one case-study, that of age.  

Many of the key terms in Table 2 relate to the theme of age, e.g. young people, old age, old people, and elderly people. We have seen from the quantitative analysis of the affective meaning of the valence, arousal, and dominance that there was very little difference between age groups in the conceptualisation of loneliness (save for the interaction between socioeconomic status and age for the dominance scores). However, the longer-form responses highlight something else. 

While both younger and older people are mentioned in the contexts of loneliness, the risk factors associated with these groups differ. Indeed, the discourses in which mentions of these groups appear highlight that the challenges relating to loneliness are not uniform. For example, social media was often mentioned in the context of young people: 

  • In the news it is reported that loneliness is on the increase, particularly amongst young people. This surprises me as young people are constantly communicating with others on social media. However, perhaps seeing what looks like a perfect life on someone’s Facebook account serves to exacerbate feelings of dissatisfaction and loneliness in people whose lives are not running smoothly.  
  • I understand that social media can cause young people to feel lonely and that other people are living a more interesting life. Unfortunately, it is full of lies but people don’t really understand that.

In contrast, discussions of older people’s loneliness were often framed in terms of isolation in more absolute terms, e.g.: 

  • I know that there are old people today who feel lonely. Particularly following the death of their spouse.  
  • Inevitably, as people live longer and stay in their own homes, isolation can become a serious problem. I believe the older generation can suffer considerably in this respect, particularly in rural communities where the shop and pub are now closed, the bus service has been withdrawn and children have grown up and live elsewhere. 

Ultimately, using a suite of analytical tools, from sentiment analysis, to keyword analysis, to discourse analysis, enables one to garner insight into the conceptualisation of loneliness that is often missed in survey-type data. The discursive nature of this data enables lived-experiences of loneliness and its perceived risk factors to be unpacked in detail- the sort of detail that computational linguistic methods can cut through, to pick out key patterns and affective meanings. Ultimately, the findings from this report will inform the development of a chatbot that will serve as a tool to combat the loneliness crisis.  

If you are interested in working with Concept Analytics Lab, please do contact us at Justyna.Robinson@sussex.ac.uk

About Us

We identify conceptual patterns and change in human thought through a combination of distant text reading and corpus linguistics techniques.

Blog

Looking back: What the Mass Observation Covid-19 diaries can teach us?

“Show me the meaning of being lonely”

How does concept-led research matter?

Identifying key content from surveys

Collaborate with us

‘We’re saying that we trust them but really we don’t’: The discursive framing of TRUST in international trade deals

1 Reply

‘We’re saying that we trust them but really we don’t’: The discursive framing of TRUST in international trade deals

by Justyna  A. Robinson,  L. Alan Winters, Rhys Sandow, Sandra Young, Caitlin Hogan

Abstract

One key consequence of the UK leaving the EU (Brexit) is that it now has full responsibility for making its own international trade policy. In this context, NatCen and the Centre for Inclusive Trade Policy initiated Citizens’ Juries on the topic of trade policy. From the transcripts of these juries, we created a corpus of 317,974 words. Using corpus-assisted discourse analysis, we focus on the concept of trust in trade policy. We find that trust conferred on actors in trade policy is limited. The greatest degree of trust is conferred to experts, on account of their epistemically-elevated position. The government is broadly not trusted. Jurors wished to be consulted about trade policy decisions and be assured that they are based on sound advice, but few wished to have a role in actually making them. Our findings highlight a deficit of trust that could be remedied by greater honesty and transparency from the government.
Keywords: concept of trust, trade policy, Citizens’ Juries, corpus linguistics

1. Introduction

trust1 is a crucial concept in citizenship, being integral to the social contract between government and citizens (see Locke 1960; Gauthier 1986). However, in a UK context, research has shown a decline in public trust in the government. For example, the Ipsos (2023) veracity index found levels of trust in politicians and government ministers to be at the lowest levels since the index began in 1983. Indeed, a report from Carnegie UK (2022) found trust to be the biggest threat to democracy, with 73% of respondents not trusting the government and 76% not trusting MPs. Winsvold et al (2024) find the basis of (dis)trust of political actors relates to their (un)predictability and intrinsic (lack of) commitment. If the government occupies one pole of (non-)trustworthiness, the literature suggests that experts are located at the opposite pole. For example, research has generally shown experts to be perceived as trustworthy (Dommett & Pearce 2019; Angelou et al. 2023), with Ipsos (2023) findings groups such as doctors, professors, engineers, and nurses to be among the most trusted groups in society. In the present article, we investigate the theme of trust in the context of UK international trade policy as it relates to a variety of actors, particularly MPs, government ministers, experts, and the public through corpus-assisted discourse analysis (see McEnery & Baker, 2015).

International trade is a major issue for the British economy. About a quarter of the UK’s demand for goods and services is met from imports and about a quarter of the demand for UK production of goods and services comes from exports. About 6.5 million jobs depend on exports. While the UK was a member of the European Union or its precursors, policy on international trade was determined in Brussels with a little input from the UK and other member governments. Brexit brought full responsibility for UK trade policy back home to a government ill-equipped to discharge it, not least because it had little idea about attitudes to trade among the public.

In the years shortly after Brexit came into effect (2021) the UK had rolled over 30 trade agreements that it had been party to under the EU, agreed new terms for trade with the EU and signed new agreements with, Australia and New Zealand and with the Comprehensive and Progressive Trans-Pacific Partnership (CPTPP) with 11 Pacific nations. The 30 roll-overs (imperfectly) preserved the trading conditions the UK had under the EU, the agreement with the EU offered vastly inferior trading conditions to those which it had as a member, and official estimates suggested that the economic benefits of the new trade agreements were trivial – see Winters (2024).

This was the context in which the Centre for Inclusive Trade Policy (hereafter, CITP)2 was created (funded by the Economic and Social Research Council) and in which one of the Centre’s early objectives was to find out how the public thought about trade policy – its reasoning, the criteria by which it judged policy and policymaking and how it reacted to specific trade-offs. (Trade policy always entails trade-offs – more of A and less of B and/or more for X and less for Y.)  

The chosen instrument was a series of five Citizens’ Juries in which small groups were informed about, discussed and in some cases made mock decisions about trade policy questions. The recordings of these juries’ final discussions provide the raw material for this analysis, in which we look specifically at the questions of whom the public trusts to inform decisions and whom they trust to make them. A few more details are provided in Section 2 and a good deal more in the documents at CITP (2023).

The concept of trust is explored through corpus-assisted discourse analysis, using both quantitative linguistic data as well as thematic analysis of narratives that illustrate the discourses of trust among the Citizen Jurors (hereafter, CJs). In particular, the current analysis serves to answer the research question –- to what extent is trust conferred on actors in UK trade deals? We consider who is trusted and who is not, and what are the underlying motivations for these assertions. The main characteristics of the concept of trust that arise from the CJs’ conversations are:
  • CJs place very limited trust in political actors
  • CJs recognise that trust is critical to political processes
  • CJs feel that there is a lack of options in who to trust
  • CJs express a relatively high degree of trust in experts
  • CJs’ opinion on trusting the public is divided
  • CJs make a distinction between actors they trust to inform and to decide on trade deals

In Section 2 we discuss corpus-assisted discourse analysis as a method of analysing CJ conversations. In Section 3 we present findings which start with an observation that jurors place very limited trust overall in the process of decision-making with regards to trade deals. We then report on the relatively high levels of trust CJs afford to experts, and the limited trust they confer on the public. Specifically, we note the distinction that CJs made between informing the decision-making process and actually making decisions regarding trade policy.

2. Method

The data interrogated in this report come from the transcripts of the final day of the CITP’s Citizens’ Juries on UK Trade Policy – ‘Trade-Offs in Trade Policy’. These were conducted by the National Centre for Social Research’s Centre for Deliberative Research3 (hereafter, NatCen) in Belfast, Bridgend, Doncaster, Paisley, and Reading, between 11th January and 4th February 2023. The aim of the juries was (i) to uncover how UK residents felt about the trade-offs that any act of international trade policy entails, but only after they had been given some (carefully balanced) information about the issues at stake and (ii) more importantly, to understand how they reasoned about the trade-offs as they came to their views. One focus of the latter was whom they trusted to inform the decision (and the public) and whom they trusted to take the decisions.

Each jury was of around 20 people (with no attrition – perhaps because jurors were remunerated) and representative of its local area. Each met five times: four times virtually for 2.5 hours each and once face-to-face in the final all-day session. At each meeting there were some plenary sessions (of the whole of one jury), but most deliberations took place in groups of about six with a facilitator. Each on-line session dealt with a different trade policy area, i.e.

  • The impact of UK trade policy on the rest of the world
  • Balancing trade between territories and sectors of the economy
  • Privacy and international data transfer
  • Food trade and the environment

CITP researchers made carefully-balanced, impartial presentations on aspects of these but then withdrew for the discussions. The face-to-face meeting had one session on each of the four issues based around a specific question, plus an introduction and a wrap-up session. Face-to-face the format was for NatCen to present on an issue, for deliberations to occur and then a vote to be held on which participants briefly recorded the main reason for their vote and then discussed how they voted. We then modified the scenario a little and the process was repeated. The modification approach was designed to reveal the jurors’ trade-offs more clearly. Details of the process are laid out in detail in the documents at CITP (2023).  

The discussions from the final meeting were audio-recorded and fully transcribed by a professional transcription service. The recordings were not perfect, and the transcription noted inaudible segments wherever they occurred. The data were annotated for the speaker type, i.e. juror or interviewer, male or female (inferred from the audio recordings), the scenario that was being discussed, the location of the Jury, and the name of the facilitator.4

The overall CJs dataset (corpus) consists of 317,974 words and 396,380 tokens (token are words plus punctuation). Of these tokens, 139,527 (35.2%) were produced by male jurors, 157,972 (39.9%) by female jurors and the remainder (24.9%) to facilitators. Thus, the overall data is relatively well-balanced in terms of gender (it was perfectly gender-balanced in terms of membership.). Across the five locations, the distribution of tokens is also relatively even. The highest number of tokens come from Belfast (23.9% of total tokens in the corpus) while the joint fewest come from Bridgend and Paisley (18.5%), see Figure 1.

Figure 1: Distribution of tokens in the CJs corpus for each location
In order to investigate how the jurors conceptualise and construct the concept of TRUST in the contexts of trade deals, we employ methods associated with corpus-assisted discourse studies (CADS, see Partington 2004). This methodology combines the quantitative power of corpus linguistics, in which large electronic collections of text may be analysed and common trends identified, with the qualitative perspective of discourse analysis. Corpus linguistics, including CADS, can be done via a range of software tools, including Sketch Engine (Kilgariff et al. 2004; Kilgariff et al. 2014), which is used in this study. This tool calculates, amongst other outputs, frequencies of words and word clusters (known as keywords when compared to a reference corpus5 to detect statistically significant frequencies within a given text), collocations (cooccurrences of a target word with others) and concordance lines (sections of the text itself which facilitate more qualitative analysis). In particular, Sketch Engine was selected for this study due to the utility of its Word Sketch feature, in which a range of grammatical relations (such as modification, verbs with a search term as their object/subject, prepositional phrases, adjectival predication, etc) can be displayed at once. We term these visualisations behavioural profiles. The behavioural profile allows for streamlined analysis of how terms behave or function in the text, the constructions in which they frequently occur, and how they may operate within circulating discourses. The behavioural profiles that we present are not comprehensive of all possible constructions a word appears in. This is so that the visualisations remain readable, as the complete versions contain hundreds of data points, and that the information that is pertinent to our discussion is retained and foregrounded. CADS employs a mixture of these statistical methods with close discourse analysis to identify quantitative linguistic patterns as well as a more qualitative perspective which adds context and detail to these empirical results. By applying these methods to the analysis of the concept of trust we identify people and entities CJs place trust in and what conditions are imposed on that trust within the discourses of the CJs. While the current analysis of trust primarily centres on the word trust as both a noun and a verb (as well as words with the trust stem, e.g. trustworthy) this is also supplemented by less frequent near-synonymous constructions involving verbs such as to rely on and to believe, or where trust is implicit in the context of discourse. This focus on trust is motivated by the distinctiveness of constructions containing trust in the CJ corpus, which is further outlined in the results section. 
In analysing the discourse of CJs, the primary way we approach the concept of trust is through the notion of stance (Du Bois 2007; Jaffe 2009). We consider two types of stance throughout the current analysis. Affective stance refers to the orientation towards the topic of discourse, that is, the stance object, such as positive or negative, see Eckert (2019), while epistemic stance refers to an individual’s certainty in the content of their proposition (see Kiesling 2009). While there are a variety of ways that stance is indexed through language (for review, see Kiesling 2022), our focus is on discourse. Specifically, we consider how the use of particular words or expressions provide insight into attitudes towards trust in the context of trade deals. These expressions include heavily value-laden words (e.g. love, struggle, bad, good) as well as those that are more functional but speak to permissibility and obligation, such as deontic modal verbs (e.g. should, need), and the use of discourse strategies such as hedging (e.g. ‘I think…’ ‘maybe..’).

The final note in this section outlines transcription and presentation conventions of the data and addresses some of the challenges faced by working with natural spoken conversations. Quotations used in the paper are accompanied by the gender of the speaker and the location of the CJ (it is often not possible to identify the individual speaking in large groups). Where more than one juror is quoted, we refer to this as ‘various (speakers), [location]’. Where speech is produced by a facilitator this is represented in italic font. While the facilitator text is not the subject of our analysis, it often provides important context in which the discourses of the jurors can be interpreted. It is important to note that given that in spoken interactional discourse, such as the CJs, much of the meaning being expressed is extra-linguistic and, therefore, is not captured through transcription. There are also many examples of incomplete utterances as well as anaphoric references for which the anaphor is not clear. We include examples that we feel are relatively high in clarity and are not especially difficult for the reader to parse.

3. Results

Both quantitative linguistic analysis and close reading of the texts highlight the concept of trust as being of widespread yet heterogenous in CJs’ conversations.  For example, the word trust is distinctive within the CJ corpus. This is highlighted by several measures. In comparison with the baseline corpus, the word6trust appears within three of the top-ten three- and four-grams (i.e. strings of three or four words). The word trust itself is analysed at both verb (375 occurrences) and noun (103), as well as words which include trust in the stem (trust *), such as trusting (16) and trustworthy (2). Trust* is a relatively frequent stem in CJs corpus as it accounts for 0.12% of the corpus.7
The exploration of the concept of trust in CJs corpus starts with an overview of the use of the verb to trust in the texts. A behavioural profile for the verb trust can be seen in Figure 2.8 When interpreting behavioural profiles, the different colours correspond to the different grammatical relations with trust, the size of the circle represents the raw frequency of the collocation between the word and trust, and the closer the position to the centre of the circle, the stronger the collocation between that word and trust (calculated by logDice score). The segments represent different grammatical relations, e.g. grammatical objects of trust (e.g. trust people) and modifiers of trust (e.g. really trust).
Figure 2: The behavioural profile for the verb to trust.
The results presented in Section 3 are mainly led by findings in Figure 2. On the basis of that analysis we identify salient discursive frames within which CJs place the concept of trust.  Figure 2, shows that trust+not is typical of CJs’ conversations about trade deals. Thus, we discuss the lack of trust, and who is not trusted in Section 3.1. Presence of collocates expert, scientist in Figure 2. leads to the analysis of the trust afforded to experts in section 3.2, as well as trust (and lack thereof) in the public in Section 3.3. Figure 2 shows that two of the top infinitive objects of the verb to trust are to decide and to inform. Thus in Section 3.4 we explore how different types of actors are trusted to have deciding or informing functions within trade policy. We report that experts are trusted to consult and government is trusted (reluctantly) to make decisions.

3.1 The lack of trust

The most frequent collocation in Figure 2, not + trust, with 54 hits across the CJs corpus, highlights the lack of trust afforded by the jurors, with a small number  of participants admitting they don’t trust anyone in the context of trade deals:
  1. I can’t trust anyone (Female, Bridgend)
  2. I don’t trust anybody, to be honest (Female, Reading)

However, this lack of trust is not uniform, and there are some groups of people and institutions who feature in the context of not + trust more than others. This is evidenced by the grammatical objects of the not + trust construction presented in Figure 3.

Figure 3: Objects of not+trust in the CJs corpus.

This lack of trust is principally directed at political actors seen in the collocations between not + trust and government (n=14); politician (n=2); MP (n=1). For example:

  1. Who do I trust to inform? Well, I don’t trust the government […]They are what they call the public school mentality, isn’t it? They haven’t got a clue really.   (Female, Bridgend)
  2. I don’t trust any of them, to be honest. We can go back to the: I don’t trust politicians full stop – which is sad that I feel that.   Aye. (Female, Paisley)
  3. I don’t trust the MPs to [negotiate trade deals] because they will have a vested interest. (Male, Belfast)

The examples above exemplify some of the CJs’ attitudes pertaining to the government. For example, ‘[the government] haven’t a clue really’ reveals a negative affective stance towards the government and their perceived capabilities. Strong epistemic stances (see Section 2) are also demonstrated through the lack of hedging, e.g. ‘I think’, ‘maybe’, or ‘sort of’, while unhedged comments, such as ‘I don’t trust politicians full stop’, reveal a strongly held belief.

Despite this lack of trust in political actors, a number of jurors do comment that they reluctantly trust the government. That is, many have resigned themselves to trust the government in the perceived absence of an alternative:

  1. I have no choice but to trust the government.  We were going to say, like, politicians, it’s like a bad word these days, but we’ve no option but to trust the government in international trade committees, that they’re speaking to people and know what they’re talking about. (Male, Belfast)
  2. Would I trust local government? They’re the only place you can go to.  They’ve got more connection, the local government.  Yes.  It’s almost like choosing your form of execution, isn’t it …?  Yes.  What do you want to do; electric chair or a bullet?  (Various speakers, Bridgend)

It is often unclear from CJs’ conversations whether it is political actors who are untrustworthy in principle or if it is the specific political actors of the day that are being objected to. While the lack of trust in these actors is evident, many jurors resign themselves to the necessity of politicians in the context of trade deals and that there are few alternatives so their trust towards these political actors is given begrudgingly.

3.2 Experts

The actors who are seen more favourably by CJs in the context of trade deals are experts.  A number of collocations between trust and expert occur (see Figure 2). As expert is a semantically broad term, we unpack what/who the jurors refer to when they speak of experts. This can be identified by the behaviour profile for expert, as shown in Figure 4.

Figure 4: The behavioural profile for expert

Figure 4 displays the verbs used with expert as the grammatical object, nouns modified by expert, the modifiers of expert, and the nouns in conjunction with expert. Collectively these serve to highlight the sorts of domains of expertise that jurors referred to. In particular, academic roles collocate strongly with expert e.g. researcher (as both a modifier of expert and in conjunction with expert), university, and the related term scientist (as both a noun modified by expert and in conjunction with expert). Other specific fields such as business, medical, legal, and economist are also mentioned. Additionally, the more generic subject expert is mentioned as well as appropriate and relevant, that is, an expert in whatever is related to the specifics of a particular deal.

The trust afforded by CJs to these experts is clearly expressed, e.g.:

  1. trust the experts of the trades, so example agriculture would be farmers and so on.    (Female, Bridgend)
  2. We really trust the scientists on this question [of food standards] (Female, Paisley)

While the examples above highlight the industries that jurors associate with experts in the context of trade deals (e.g. medical, legal), other collocates in Figure 4 speak to the types of traits that are associated with experts, particularly the modifiers independent and neutral which are qualities that the jurors valued. For example:

  1. [I trust] neutral experts [to inform trade policy] […] I said the other MPs, so Labour, Conservative, whatever, with the vested interest and baggage and maybe backhanders, whereas experts, we hope are above that (Male, Paisley)

The independence that the jurors associated with experts was also associated with international bodies, for example:

  1. Someone like the World Health Organization has a blanket view of what’s good and what’s not as opposed to us deciding that. You’re saying the WHO because they’re independent. Yes, they should be. Or an equivalent independent body. (Various, Reading)

These examples highlight a positive affective stance towards independent international bodies and their perceived capacity to prioritise the public good, rather than particular interests.

Close reading of sentences containing terms for experts and terms which are typically used in conjunction with terms for experts (see Figure 4) also reveals a similarly positive affective stance in terms of advocating for the role of experts in trade strategy, in particular in the context of the perceived elevated knowledge base of experts, relative to politicians, e.g.:

  1. It’s like the experts, the ones that are not in it for financial gain […] so they’re the knowledge. An MP might not necessarily have the knowledge. (Female, Doncaster)
  2. Let the scientists meet and agree the best decision. (Male, Belfast)

Jurors also mention that the government would be trusted to a greater extent if it was clear that they had consulted experts:

  1. Like we’d have more confidence if we knew that they [the government]’ve spoken to the experts in the field or spoken to whoever we felt should be relevant (Female, Reading)

Thus, for some jurors, the trust in experts can serve to offset their lack of trust in the government.

Trade unions, or their representatives, are sometimes framed as a specific type of expert who should be consulted on trade deals, although not without dissenters. For example:

  1. I’d also trust trade unions to have a part in it, because they obviously act on behalf of their members who are the general public. (Female, Belfast)
  2. I think maybe a trade union, they will think for their own interest so I am not sure about the union. (Female, Bridgend)

Another way in which we identify general sentiment is by analysing the verbs that occur in the context of a given noun (see, e.g. Biber 2006). For example, through the analysis of the verbs that are used with expert, we can provide additional insight into how the jurors perceive experts (see Figure 4). The strongest collocate of verbs with expert as object is need. This highlights the ways in which the consultation of experts is framed as imperative:

  1. We need experts who are qualified to comment on world population, world growth, who is dying of hunger (Male, Paisley)
  2. We need experts to decide whether this is […] worth the risk (Male, Paisley)

Ultimately, of all of the groups discussed in relation to trade deals, jurors frame experts most positively. They expressed trust in experts in the context of informing the government. In particular, jurors favour experts due to their perceived independence and epistemically leading position.

3.3 Public

Another group regularly discussed in the context of trust was the public (see Figure 2). The perception of the public in relation to trust is complex. On the one hand, CJs see the public as incapable of having sufficient knowledge to be trusted around policy decisions, for example:

  1. You don’t want to trust the general public to make the decision because we’re not informed enough, or we don’t know enough about it to be able to make an informed decision. (Female, Belfast)
  2. We are slightly more informed, but the rest of the general public would probably know very little (Female, Belfast)

Such attitudes, which speak to the perception of the limited epistemic position of the public, are also evident in the behavioural profile for the noun public (see Figure 5).

Figure 5: The behavioural profile for public

The collocates of public, such as idiot, child, and stupid (see Figure 5) exemplify the public’s perceived naivety and lack of knowledge:

  1. So we put business owners from the specialist sector, and then we originally had public, and then we agreed that the public are idiots (Female, Paisley)
  2. Don’t ask the public. The public are stupid (Female, Bridgend)

While the collocates educate and inform do occur in Figure 5, these are typically used in phrases in which they are negated, highlighting the way that the public are not educated or informed, with respect to trade strategy. For example:

  1. Are most of the public educated enough to make decisions like this? We’re not, are we, I don’t think (Male, Doncaster)
  2. I don’t think the public are informed enough to be involved in making big decisions like this. I think it needs to be left to the experts (Male, Belfast)

On the other hand, the public are conceptualised as being close(r) (adjective predicates in Figure 5) to relevant issues, and, as such, have a right to have a voice, for example:

  1. I was going to say, I don’t think many people trust the government, do they?  Yes, I think also the people, the public are closer to the issues, have personal experience of issues that have affected them (Female, Belfast)

Many jurors highlight the value of consulting the public. However, as the CJs do not think that the public has sufficient knowledge of the intricacies of trade, they typically do not advocate for public involvement in making decisions that directly affect trade policy. This highlights a broader distinction between which actors were discussed in reference to those who were trusted to decide vs inform on trade policy.

3.4 Informing vs deciding

While attitudes towards different types of actors vary greatly, CJs also have a varied opinions on the sorts of responsibilities each type of actor should be tasked with. The jurors make a distinction between the actors that they feel should be involved in the consulting stage of trade deals and the process of decision making. This distinction is highlighted in the following example:

  1. So, yes, who else should inform?  Who…?  Yes, we kind of said that.  We said the professors, the people who know what they’re talking about, the people who’ve studied this, and so we said they should make the decisions.  We said they should inform.  Inform, sorry, not make decisions.  They’re the ones who should inform the government.  Yes, so the decision-makers should be really listening to […]  Because they’re the ones who inform researchers.  It’s like the experts, the ones that are not in it for financial gain, the ones that gain something, so they’re the knowledge.  An MP might not necessarily have the knowledge.  Yes, they can make the decision, or they can put something forward, but they don’t know the knowledge.  They’ve not studied it, so I think they have to listen to experts and researchers, people that know (Various, Doncaster)

Motivated by findings presented on Figure 2, we focus on the decision-making processes in relation to trade deals which are represented by a set of verbs inform/consult and decide (see Figure 2). We generate a visualisation of the collocates that occur as grammatical subjects of verbs to inform and to decide (Figure 6)9. Terms that appear closer to the left of Figure 6. collocate more strongly with to inform while those that appear on the right have a stronger collocation with to decide, while the size of the text relates to the raw frequency of the collocation (e.g. people has a higher raw frequency than Ireland).

Figure 6: The comparative behavioural profile for to inform and to decide (subject position).

Figure 6 shows that scientist and farmers, which are types of experts, are discussed in the context of informing trade policy.  Relatedly, Figure 7 presents a comparative behavioural profile for decide and consult, which is near-synonymous to inform. It shows that public and people, as well as expert collocate more strongly with consult than decide. Thus, the semantic fields of the public and experts are more strongly collocated with inform/consult.

Figure 7: The comparative behavioural profile for to consult and to decide (subject position).

The theme of government is much more strongly associated with decide, such as MP, government, and minister (see Figure 6). Further analysis of the concordances demonstrates the ways in which the jurors framed (members of) the government as those who should be making decisions:

  1. Also we’ve come out with the UK Government making the decision about any trade agreement.  Are they still the best placed to do this, or are there other groups that you think should decide on an agreement like this?  I think so.  I think that’s what they’re there for, isn’t it; they’re the ones that make the decisions but they need to make informed decisions. (Female, Reading)
  2. Yes, I think that was the point you were making earlier on; that that’s their job and that’s what they should be getting on with.  Yes. …  I would hope it’s – nobody else should be making decisions.   Governments, they are elected and they can be dealing with it.  Yes, because it’s the diplomatic system we have.  I would just hope – my hope would be that their decisions are informed by consulting experts; that’s what we were discussing.  (Male, Belfast)

One comment suggests that the government should simply enact the decisions of the public and experts:

  1. Who do you think you trust most to make these decisions?  I don’t think one; I think it should be a combination of the public and the government.  Yes.  With the government initially with the public on a scale of, I reckon, about 70-30 per cent, so the public should be able to make 70 per cent of the choices, with 30 per cent coming from the experts – not the government – and then the government enforcing the decision. (Various speakers, Bridgend)

Jurors suggest that while experts should play an important role in the process, it is the role of government to ultimately make decisions:

  1. So, who do people feel like should be making these decisions about which jobs we might be creating, who should be winning or losing from trade?  I think it should be the government – like we discussed in the last […]  but with people feeding in, whether it be unions or the agricultural […]  What do they have?  Do they have some sort of union?  I don’t know.  Feeding into them and giving them the information, but we can’t have different sets of people making different decisions.  Yes.  In relation to trade policy.  It’d be all over the place.  It’s got to be Central Government I think, making the decision. (Male, Bridgend)
  2. I think government should always be making the ultimate decision because they are elected to do so.  On the advice of the experts.  And us.  A wee bit.  Select few, yes. (Male, Belfast)

The collocations of decide vs inform highlight the distinct processes in the development of trade deals. CJs provide recommendations and have clear ideas on obligations of various actors through a repeated use of deontic need or should. While the public and experts are more likely to be discussed in the context of informing and consulting on trade deals, it is the government who should be ultimately responsible for making decisions. However, a number of jury members, particularly in Reading, were not content with this arrangement, highlighting a desire for an entirely new and bespoke assemblage of people to be responsible for trade deals.

4. Discussion

The CJ data speaks to the limited trust conferred on politicians and political processes. The government was not trusted in absolute terms but were begrudgingly trusted to be responsible for ultimately signing trade deals. Using Cairney & Wellstead’s (2021: 2) typology of trust, we can think of this as ‘trust as necessary to society’, that is, while the trust is not enthusiastically given to the government, it is perceived to be essential for the running of the state. However, the government are not trusted in terms of their reliability, competence, degree of selflessness, or shared identity or values (Cairney & Wellstead 2021).

Experts, despite frequent lack of specification as to who experts actually are, are evaluated positively according to a variety of Cairney & Wellstead’s (2021) trust categories. In particular, their input in informing trade deals was viewed as necessary to society. They were also evaluated positively in terms of competence, (relative) selflessness, reliability, and performance. There are no categories in relation to which experts were routinely explicitly negatively evaluated.  Thus, while then then Secretary of State for Justice, Lord Chancellor, Michael Gove famously claimed that the public have ‘had enough of experts’, this is diametrically opposed to the evidence from the citizens’ juries as well as other sources such as Ipsos (2023). Overall, the epistemic trust for experts is high, that is, they are perceived to have accurate and relevant information that can serve to benefit the UK’s interests in trade deals.

Applying Cairney & Wellstead’s (2021) trust typology to the public, they are not valorised positively in terms of competence. In particular, they are not thought to have the specific knowledge base required for complex geopolitics and economic analysis. However, they are perceived more positively in terms of having shared identities or values and mutual self-interest. That is, they are perceived to be trustworthy in the sense of a willingness to do the right thing, though not necessarily have the competence to carry this out independently, without the involvement of the government and experts. This nuanced view of public involvement in decision making has also been seen in research looking at another complex decision-making area, health (Litva et al. 2002). Employing Arnstein’s (1969) ladder of citizen participation, the responses typically call for a partnership, without extending this participation to delegated power or citizen control.

While participants did not provide detailed rationales for the lack of trust, some did mention Brexit as a contributory factor for their scepticism of government and perceived vested interests, e.g.:

  1. I can’t trust anyone.  Recently, because of Brexit as well, the promises which were made…  Promises, promises, promises.  (Female, Bridgend)
  2. Brexit were a game-changer wasn’t it? It was, and I also feel very manipulated by Brexit, by the media. (Male, Doncaster)

As the need for UK-specific trade deals is a direct consequence of Brexit, the spectre of Brexit looms large in the CJs. However, the scenarios, presentations and notes for facilitators all sought not to avoid re-igniting Brexit controversies, partly to try to maintain a harmonious atmosphere but also because we wished to reach beyond Brexit to more fundamental building blocks of attitudes to international trade. Thus, the relative scarcity of ‘Brexit’ in these conversations is not an indication of its lack of salience.

Conversely, the levels of trust conferred on experts seems to have been boosted by their role in the Covid-19 pandemic:

  1. Bodies like the IMF, World Health Organization during COVID, those international organisations, when they don’t have a single country it’s just like, ‘I’ve got to get elected in the next three years’. These organisations are less relying on elections coming up to win; it’s just like a general benefit of people. (Male, Bridgend)
  2. Well, in COVID we relied on the experts. Yes. For daily briefings and everything, it was the experts you listened to. Yes, if an expert was like, do you know what? This would work out better for you’ and then the MPs is like ‘no, you’re wrong’, I’d be like […] I don’t believe in the government or MPs. (Various, Bridgend)
Thus, the public are using recent events as frames of reference when deciding with whom to bestow their trust. While the Brexit campaign generally served to diminish levels of trust in politicians, the role of experts, and their visibility, in the Covid-19 pandemic has led to their stock rising.11

5. Conclusion

The over-arching finding from the corpus-assisted discourse analysis of the citizens’ juries is that trust pertaining to actors involved in trade deals is severely limited. In particular, the role of politicians was viewed least favourably, while more positive affective stances were apparent in the context of experts. The perceived lack of transparency, the suspicions about vested interests, and the general lack of awareness of the day-to-day life of the public and their wants and needs meant that the levels of trust afforded to political actors was limited. Trust in politicians has never been high (Clements and King 2023) but it is worth remembering that in early 2023, when the juries were held, the British public had noticed that Brexit was not the overt success that had been promised and that in the previous seven months one Prime Minister had resigned in disgrace and the next had precipitated such an economic crisis that she had to resign after just seven weeks. Thus, one might hope for a little more trust in future. 

It is also important to recall that many jurors did, begrudgingly, want ultimate accountability for trade deals to rest with the government and so believe that the government should make final decisions on trade. Jurors believed widely that the public as well as experts should be informed, and in some cases consulted, about trade-policy issues. However, there was little enthusiasm for them to have a decisive role, not least because they felt insufficiently informed about such a complex issue. Their stated trust in experts of various kinds reinforces the latter view.

The Citizen Jurors’ views on trust suggest a considerable institutional deficit in the UK treatment of international trade policy, which is, after all, a major component of overall economic policy. Future governments should speak more frequently and more honestly about the questions of trade policy; they, along with the public and the media, need to accept that given the nature of the trade-offs involved almost no policy will command universal support. Rather, the trick needs to be that the country’s overall stance on trade delivers benefits to nearly everyone and that where it cannot, provisions are made to ease any burdens created. In concrete terms, partly in response to these Citizens’ Juries, the Centre for Inclusive Trade Policy has advocated:

  • greatly strengthening Parliamentary scrutiny of trade agreements and trade policy (Hestermeyer and Horne, 2024, Winters 2024)
  • better engagement with the UK’s devolved administrations and their populations on trade (Petetin et al, 2023), and
  • related to suggestions from some jurors, reforming the UK Board of Trade to provide independent expert analysis and advice on trade policy to the government, Parliament and the public (Henig & Winters 2024, Winters 2024).

Footnotes

  • Small caps indicate the concept, that is, the idea of trust, and italics are used when referring to the word trust metalinguistically. ↩︎
  • https://citp.ac.uk/ ↩︎
  • https://natcen.ac.uk/centres/centre-for-deliberation ↩︎
  • In total there were 90 transcript documents: 5 juries x 3 discussion groups x 6 discussions (4 scenarios plus intro and wrap-up). ↩︎
  • We use the Spoken British National Corpus (2014, Love et al. 2017) as our reference corpus. This is a 10-million-word corpus represents spoken English in Britain. ↩︎
  • This n-gram search does not distinguish between trust as a noun or a verb. ↩︎
  • This contrasts with 0.0013% of the baseline corpus. ↩︎
  • We focus on the verbal rather than the nominal form here for two reasons. The first reason is quantitative. There are nearly four times more tokens of the verbal form (375 against 103). Secondly, a collocation analysis of verbs enables clearer insight into who trust is being conferred on by the jurors, thus the verbal analysis is more felicitous in order to understand the application of trust in the CJs. ↩︎
  • In Sketch Engine, it is not possible to search for both inform and consult simultaneously in comparison with decide, thus we need to conduct the relevant searches sequentially. ↩︎
  • The decide side of the figure refers to both government and Government: the former is generic while the latter refers to a specific government (almost always the UK’s). ↩︎
  • Winters (2020) contrasts the official attitudes towards expertise in Brexit and Covid-19 policymaking. ↩︎

References

Angelou, Angelos., Stella Ladi, Dimitra Panagiotatou, & Vasiliki Tsagkroni. 2023. Paths to trust: Explaining citizens’ trust to experts and evidence-informed policymaking during the COVID-19 pandemic. Public Administration, Advance Online Publication.

 

Arnstein, Sherry R. 1969. A ladder of citizen participation. Journal of the American Institute of Planners, 35 (4), 216-224.

 

Biber, Douglas. 2006. Stance in spoken and written university registers. Journal of English for Academic Purposes, 5, 97–116.

 

Cairney, Paul. & Adam Wellstead. 2021. Covid-19: Effective policymaking depends on trust in experts, politicians and the public. Policy Design and Practice, 4 (1), 1–14.

 

CITP (2023) Research into public attitudes to trade, CITP, University of Sussex,  https://citp.ac.uk/public-attitudes-to-trade

 

Clements, Michael and Laura King 2023. Trust in politicians reaches its lowest score in 40 years, IPSOS, 14th December 2023, https://www.ipsos.com/en-uk/ipsos-trust-in-professions-veracity-index-2023

 

Dommett, Katharine. & Warren Pearce. 2019. What do we know about public attitudes towards experts? Reviewing survey data in the United Kingdom and European Union. Public Understanding of Science, 28 (6), 669-678.

 

Du Bois, John W. 2007. The stance triangle. In Robert Englebretson (Ed.), Stancetaking in Discourse, 139–182. Amsterdam: Benjamins.

 

Eckert, Penelope. 2019. The limits of social meaning: Social indexicality, variation, and the cline of interiority. Language, 95 (4), 751–776.

 

Gauthier, David. 1987. Morals by Agreement. Oxford: Oxford University Press.

 

Henig, David and L Alan Winters (2024). ‘Restructuring the Board of Trade for the Twenty-first Century’. Centre for Inclusive Trade Policy Working Paper 014, https://citp.ac.uk/publications/restructuring-the-board-of-trade-for-the-twenty-first-century

 

Hestermeyer, Holger and Alex Horne. 2024. Treaty Scrutiny: The Role of Parliament in UK Trade Agreements, Briefing Paper 9, Centre for Inclusive Trade Policy, https://citp.ac.uk/publications/treaty-scrutiny-the-role-of-parliament-in-uk-trade-agreements

 

Jaffe, Alexandra. (Ed.). 2009. Stance: Sociolinguistic Perspectives. Oxford: Oxford University Press

 

Kiesling, Scott F. 2009. Style as stance. In Alexandra Jaffe (Ed.), Stance: Sociolinguistic Perspectives, 171–194. Oxford: Oxford University Press.

 

Kiesling, Scott F. 2022. Stance and stancetaking. Annual Review of Linguistics, 8, 409-426.

 

Kilgarriff, Adam., Pavel Rychlý, Pavel Smrž, David Tugwell. 2004. The Sketch Engine. Proceedings of the 11th EURALEX International Congress, 105-116.

 

Kilgarriff, Adam., Vít Baisa, Jan Bušta, Miloš Jakubíček, Vojtěch Kovář, Jan Michelfeit, Pavel Rychlý, and Vít Suchomel. 2014. The Sketch Engine: Ten years on. Lexicography, 1: 7-36.

 

Litva, Andrea., Joanna Coast, Jenny Donovan, John Eyles, Michael Shepherd, Jo Tacchi, Julia Abelson, Kieran Morgan. 2002. ‘The public is too subjective’: public involvement at different levels of health-care decision making. Soc Sci Med, 54(12), 1825-37.

 

Love, Robbie., Claire Dembry, Andrew Hardie, Vaclav Brezina, & Tony McEnery. 2017. The Spoken BNC2014: Designing and building a spoken corpus of everyday conversations. International Journal of Corpus Linguistics, 22(3), 319-344.

 

Locke, John. 1960. Two Treatises of Government. Ed. Peter Laslett. Cambridge: Cambridge University Press.

 

Partington, Alan. 2004. Corpora and discourse: A most congruous beast. In Alan Partington, John Morley, and Louann Haarman (eds.), Corpora and Discourse, Bern: Peter Lang, 9-18.

 

Petetin, Ludivine., Charles Whitmore, and Aileen Burmeister (2023) ‘Addressing barriers for Welsh institutions and civil society to contribute to UK trade policy’, CITP Briefing Paper 6, https://citp.ac.uk/publications/addressing-barriers-for-welsh-institutions-and-civil-society-to-contribute-to-uk-trade-policy

 

Winsvold, Marte., Atle Haugsgjerd, Jo Sagile, & Signe Bock Segaard. 2024. What makes people trust or distrust politicians? Insights from open-ended survey questions. West European Politics, 47 (4),759–783.

 

Winters, L. Alan (2020) ‘Brexit and Covid: Experts – who needs ‘em?’, The UK in a Changing Europe,   https://ukandeu.ac.uk/brexit-and-covid-experts-who-needs-em/

 

Winters, L Alan (2024) ‘How do we make trade policy in Britain? How should we?’ Centre for Inclusive Trade Policy Working Paper 011, https://citp.ac.uk/publications/how-do-we-make-trade-policy-in-britain-how-should-we

 

About Us

We identify conceptual patterns and change in human thought through a combination of distant text reading and corpus linguistics techniques.

Blog

Looking back: What the Mass Observation Covid-19 diaries can teach us?

“Show me the meaning of being lonely”

How language reveals different faces of loneliness?

Accessibility

Collaborate with us