11.2.1 Vaccine Hesitancy on Mumsnet

Founded in 2000, the parenting website Mumsnet reported 104 million unique user visits in 2019 and receives 1.2 billion page views per year (Mumsnet, Reference Mumsnet2021). Its forum section, Mumsnet Talk, currently hosts 243 topics via sub-forums organised around a specific subject, such as ‘Children’s health’, ‘Coronavirus’, and ‘Am I Being Unreasonable?’ (AIBU). Vaccinations are one of the topics that contributors to Mumsnet write about. Indeed, a 2017 study found that 29 per cent of parents in England who use the internet as a source of vaccine-related information specifically access Mumsnet (Campbell et al., Reference Campbell, Edwards, Letley, Bedford, Ramsay and Yarwood2017).

As part of the Quo VaDis project, preliminary analysis of a corpus containing 895 threads of vaccination-related discussion, amounting to 6,269,560 words, attests to the prevalence of conflictual positions within family networks with respect to vaccination (reported in Coltman-Patel et al., Reference Coltman-Patel, Dance, Demjén, Gatherer, Hardaker and Semino2022). Furthermore, the analysis demonstrates the presence of a negative, sometimes antagonistic, attitude towards people who may be perceived as ‘anti-vaxxers’. For example,

Despite the contributor’s dismissive attitude towards the position of being ‘against’ vaccination, there is an appeal to rationality (‘unless there is a known family history’) and some mitigation on the basis that certain circumstances might justify ‘vaccine hesitancy’ (i.e., the risk of harm).

In view of this, we report on a small study of the expression and qualification of ideological positions in relation to vaccinations. In what follows, we will focus on contributions to the forum in which individuals dissociate from the identity of ‘anti-vaxxer’ (i.e., someone who is ideologically opposed to vaccination). However, what we will show is that this tends to indicate a more complex and considered position, whereby participants take ‘anti-vaccination’ as a point of reference but explain that the position is insufficient or undesirable in some way that precludes them from adopting the identity unreservedly. Thus, in establishing an ‘anti-vaccination’ position as the nexus, the contributor creates a point of comparison for their own ideas, which are variously likened or contrasted with the notions associated with ‘anti-vaccination’.

Our analysis is conducted using the Quo VaDis Mumsnet corpus: 31,211,157 words consisting of 12,288 threads from 41 Mumsnet Talk topics with Original posts containing the strings ‘vac*’, ‘vaxx*’, or ‘jab*’ (where the asterisk stands for zero or any character or a sequence of characters), with optional prefixes un(-), re(-), anti(-), and pro(-).

11.2.2 Identifying Ideological Positions about Vaccination

We set out to retrieve from the Mumsnet corpus expressions of self-reference in relation to vaccination – that is, occurrences of the first-person singular pronoun I in close proximity to mentions of vaccines or vaccinations. To that end, we established the query I ++* (*vac*|*vax*); this query syntax allowed for variation in the syntactical relationship and premodification of the reference to the vaccine (indicated in the ++* component), as well as variation in the lexical form of reference to the vaccine/vaccination, as indicated in the (*vac*|*vax*) component. The query returned 11,370 matches in 3,203 different texts and generated a list of 5,306 different formulations, the most frequent of which (30+ occurrences) are shown in Table 11.2.

The research team was not interested in reports of people having a vaccination, nor their intention to do so, but rather contributors’ positioning of themselves in relation to labels suggesting attitudes towards vaccinations, such as ‘pro-vaccine’ and ‘anti-vax’. While participants appeared to claim a ‘pro-vaccination’ stance for themselves (‘I’m very pro vaccine’), references to an ‘anti-vax’ position were more commonly dissociative, as in ‘I’m not an antivaxxer’. This can be expected, given the evidence that in these Mumsnet discussions, pro-vaccination contributors sometimes respond rather aggressively to views and behaviours that may be described as anti-vaccination (Coltman-Patel et al., Reference Coltman-Patel, Dance, Demjén, Gatherer, Hardaker and Semino2022). Subsequently, we refined our query to target self-references in relation to an ‘anti-vaccination’ perspective, as follows: I ++* (anti-va*|antiva*|anti va*) for further investigation. This query generated 1,093 occurrences from 480 different texts, which were expressed according to 457 different formulations. The most common are shown in Table 11.3.

The next step of the analysis was to examine concordance lines and determine how the referent of the ‘anti-va*|antiva*|anti va*’ label indicated an ideological position.

11.2.3 Discursive Strategies for Legitimating Vaccine Hesitancy

The first step in our analysis was to separate those occurrences that referred to a third party, rather than the contributor themselves (e.g., ‘I have friends who are anti-vax’, ‘I have not seen any anti-vaxx posts’). Of the 1,093 occurrences returned from our query, 284 (25.98 per cent) actually refer to someone or something else. What remained were 809 (74.02 per cent) occurrences in which participants describe their own ideological position in relation to an ‘anti-vaccination’ perspective. We observed only 6 (0.55 per cent) instances in which contributors claimed an anti-vaccination perspective for themselves:

For the most part, contributors referred to how their beliefs misalign with what they recognise to be ‘anti-vax’, which involved various degrees of dissociation or qualification. For example, in the following extract the author disassociates themselves unequivocally from the anti-vaxxer label:

In contrast, the author of the next extract qualifies their position in relation to the label:

Several contributors problematise the absolute position of ‘anti-vaxxer’ and state that the label is used to dismiss and distance people who express anything other than complete support for vaccines and vaccination schedules:

Furthermore, while the ‘anti-vaxxer’ label is often posited in binary opposition to the ‘pro-vaccination’ label, some posters offer the alternative stance of being ‘pro-choice’ – an argument that Fasce and co-authors (Reference Fasce, Schmid, Holford, Bates, Gurevych and Lewandowsky2023) capture via the ‘Personal autonomy’ theme:

In the previous example, the statement that the writer’s family ‘are all up to date with vaccinations’ is used to establish their credentials as someone who is not anti-vaccination. In our data, the provision of this kind of detail often occurs in close proximity to the negation of the anti-vaxxer identity.

In several instances, contributors negotiate additional positions of vaccine hesitancy along a cline by presenting anti-vax as a matter of degree, as suggested by the expressions ‘I’m not completely anti-vaccination’, ‘I’m not particularly anti-vaccine’, or ‘I am more anti-vaccine than pro-vaccine’. In most cases, however, rather than alluding to a scale of attitudes towards vaccinations, contributors set out to legitimate their position by making explicit in what specific respect they may adopt a critical position towards vaccinations or not take a particular vaccination.

The most commonly cited concern expressed in the data relates to the (potential) harmful effects of the vaccine. This tends to involve arguments that Fasce and colleagues (Reference Fasce, Schmid, Holford, Bates, Gurevych and Lewandowsky2023) capture via the themes relating to the ‘Fears and phobias’ attitude root, such as ‘Safety concerns’ and ‘Dreadful injuries’. In many cases, such concerns were informed by personal or family circumstances of medical histories, reflecting the individual dimension in SAGE’s (2014) categorisation of determinants of vaccine hesitancy:

The provision of details about serious health-related events arguably involves the legitimation strategy of rationalization, in that each writer has very good reasons for their concerns about vaccinations. At the same time, the disclosure of traumatic personal circumstances also appeals to emotion. The first example also explicitly references the writer’s emotional state following the trauma of losing a baby (‘severe anxiety and depression’). Furthermore, we also find examples of mini narratives that serve as cautionary tales and demonstrate Van Leeuwen’s (Reference Van Leeuwen2007) category of mythopoesis.

Relatedly, the novelty of the vaccines is one of the factors leading to speculation about their potentially harmful effects. The 2014 WHO report noted previously mentions specifically the introduction of new vaccines as one of the causes of this kind of vaccine hesitancy (SAGE, 2014: 12). In our data, the strategy for redirecting efforts towards swift development of the vaccines against COVID-19 – in particular – introduced doubt as to whether they had undergone sufficient testing in the first place, alongside concerns regarding the as-yet unknown long-term effects:

The previous examples are typical of our data from the ‘Coronavirus’ topic on Mumsnet in that they all mention previous vaccinations as part of the writer’s credentials as a non-anti-vaxxer. They include different strategies for authorisation (the hyperlink to a scientific paper in the first example) and rationalisation (through references to what is ‘known’).

A very specific approach to the presentation of oneself as a rational agent capable of making decisions involves referencing one’s own ‘research’ (see also the previous discussion on ‘informed’, and Fasce et al.’s (Reference Fasce, Schmid, Holford, Bates, Gurevych and Lewandowsky2023) ‘Do your own research’ theme):

This implicitly suggests a lack of trust in how scientific findings are used in vaccination programmes and policies, as well as a belief in one’s own ability to acquire sufficient expertise to make independent decisions.

The importance of autonomy and independent decision-making is also reflected in the major theme of resisting mandated vaccination. Some writers express disagreement with vaccination-related policies, the coverage of the vaccination programme, or elements of compulsion (see Fasce et al.’s (Reference Fasce, Schmid, Holford, Bates, Gurevych and Lewandowsky2023) ‘Resisting coercion’ theme):

Others make a distinction between ‘necessary’ and, by implication, unnecessary vaccinations, and claim to take up only the former:

In all these cases, the legitimation of the writer’s position relies primarily on their rational ability to assess policy and the necessity of vaccinations on a case-by-case basis. The rejection of mandatory vaccinations also potentially involves moral evaluation, insofar as the writer is concerned about the implications of compulsion for fellow citizens.

Lack of trust is more explicitly present when references are made to the unreliability of governments and pharmaceutical companies:

This kind of argument is captured by Fasce’s team (Reference Fasce, Schmid, Holford, Bates, Gurevych and Lewandowsky2023) in the themes ‘Government cover-up’ and ‘Big Pharma’ (from the ‘Conspiracy ideation’ attitude root) and is included under ‘contextual influences’ in SAGE’s (2014) report. In our examples, a position of mistrust is presented as different and separate from being an anti-vaxxer. The first example also makes explicit how the writer sees trust as less rational than the position they have adopted (cf. ‘blindly trust’).

Finally, a frequent qualification of the writer’s position in relation to vaccination involves the rejection of specific vaccines, or combinations of vaccines, reflecting SAGE’s ‘vaccine/vaccination-specific issues’ as a major determinant of vaccine hesitancy:

Such contributions challenge the wholesale adoption of an ‘anti-vaxx’ philosophy in arguing for the assessment of each vaccine according to its individual merits. The final example reflects a common concern surrounding the number of vaccines that are administered, particularly when seen in relation to the perceived vulnerability of infants.

In summary, the principled selection of concordance lines involving the first-person negation of an anti-vaxxer identity has led us to identify the range of ways in which Mumsnet contributors reject that identity while at the same time legitimating a nuanced position in relation to vaccinations. By and large, this involves the use of Van Leeuwen’s (Reference Van Leeuwen2007) legitimation strategy of rationalization – that is, the process of legitimating a position by referring to knowledge and the cognitive validity of positions presented as reasonable, acceptable, appropriate, or even superior to those who may disagree. On the one hand, our observations are consistent with previous findings that the identity of ‘anti-vaxxer’ is adopted by a relatively small proportion of people, while a much larger proportion have some specific concerns about vaccinations that coexist with taking up available vaccines most of the time. On the other hand, we have shown how the concerns and qualifications that are presented in our data as consistent with not being an anti-vaxxer are captured by existing typologies of determinants of vaccine hesitancy and well-known arguments against vaccination.

11.3.1 Strategies of Legitimation

One such example arose in the analysis of patients’ comments that mentioned age. Specifically, it was found that (particularly older) patients frequently referred to their older age as part of a broader description of their experience using healthcare services. In such cases, Baker and colleagues (Reference Baker, Brookes and Evans2019) argue, patients can index their experience as healthcare consumers and, accordingly, construct themselves as ‘informed’ or ‘expert’ patients (Fox et al., Reference Fox, Ward and O’Rourke2005), in this case as expert healthcare service users, in particular regarding regular standards of healthcare service provision and, as such, what they might reasonably expect. In this way, the patients positioned themselves as having reasonable expectations, thus rendering them as qualified and reasonable evaluators of the services they accessed. As detailed in Chapter 6, such comments were accessed through concordance searches of queries which were determined to be productive for identifying cases in which patients described their age (as opposed to, say, someone else’s age or simply the number of years that they had been seeing a provider).

Such cases could indicate the legitimation of positive feedback. For example, in the following comment, a patient notes his age (‘I am a 55 year old man’) before then describing how he has ‘been a patient with this surgery all my life’. This autobiographical segment was then followed by a positive appraisal of the practice, where the preceding segment served to legitimate this man’s perspective as an experienced patient at the practice, with his evaluations being based not on a single visit but on consistently positive experiences over a long period of time (‘I have always had a positive relationship with … ’).

However, this kind of self-construction of an experienced, ‘expert’ patient identity could also be used to legitimate negative feedback. This accords with Reyes’s (Reference Reyes2011) observation that legitimation strategies are typically invoked in order to justify contentious propositions. Indeed, negative feedback could be viewed as contentious – face-threatening, even (Austin, Reference Austin1962) – when we bear in mind that such comments are directed at the providers themselves. For example, the following comment represents something of an inverse of that seen above. This patient described himself as ‘over 60 years old with a lot of excellent dental practice on my teeth’ before then providing a negative appraisal of the dental practice in question. This autobiographical segment serves two functions: in addition to (self-)constructing the commenter as an experienced, ‘expert’ patient, the evaluation of their past experiences as ‘excellent’ simultaneously presents them as a reasonable and balanced judge of such services, capable of praising services when they are good and criticising them when they are not. This latter point arguably takes on more pronounced importance in the context of particularly severe criticism; indeed, in the following comment, the patient describes the practice as ‘by far the wors[t]’ they have encountered, notes how they are ‘still in a great deal of pain’, and accuses the dentist of being ‘oblivious to the challenges’ that are particular to older patients’ dental treatment requirements.

Additionally invoking what Van Leeuwen (Reference Van Leeuwen2007) terms ‘authorization’ by describing their experience (and thereby indexing a kind of ‘expertise’) in their comments, some patients were also found to deploy argumentation based on ‘moral evaluation’ (Van Leeuwen, Reference Van Leeuwen2007) as means of legitimating their evaluations. As a reminder, moral evaluation is legitimation by (often very oblique) reference to value systems. As an example, the patient writing the following comment constructs himself as not being burdensome to the NHS, which keys into a moral discourse that public health systems should be used only when necessary (Llanwarne et al., Reference Llanwarne, Newbould, Burt, Campbell and Roland2017). The patient evokes this discourse by first noting his age (‘73 year old male’) and history of ‘heart problems’, which might set up an expectation that this patient has complex healthcare needs and would thus have to use the NHS frequently. However, the patient then describes how he consciously does not use the NHS very often, as he is aware of the strain that services are under (‘tended not to visit my GP because I know they are very busy’). Moreover, when he uses such services, he does so rarely, only for routine appointments, and notes how he always attends these (‘I only have a check up once a year which I always attend’), perhaps indicating an awareness of how much missed appointments cost the NHS (Llanwarne et al., Reference Llanwarne, Newbould, Burt, Campbell and Roland2017). We can also note how, at the end of the comment, the patient describes how ‘considering going to A&E … is totally against [his] principles’, explicitly invoking a moral evaluation to again position himself as a conscientious user of healthcare services who is opposed to using Accident and Emergency services for non-emergencies and is aware of the issues that doing so would cause for the NHS (see, for example, Adamson et al., Reference Adamson, Ben-Shlomo, Chaturvedi and Donovan2009). This background information helps establish this patient’s credentials as a genuine and conscientious patient (as opposed to being a ‘time-waster’), which then legitimates his complaint about a lack of appointment availability.

Another area in which patients appealed to value systems to legitimate their evaluations was with respect to ideas and attitudes relating to gender, specifically in the analysis of comments in which patients disclosed their identities as men or women. For example, in the following excerpt, a patient complains about the pain he experienced when having a filling put in at the dentist. He legitimates his account of the amount of pain he experienced by referring to the fact he is a ‘a big 6ft man’. It is implied here that this patient would ordinarily be able to withstand a large amount of pain, and thus the pain that he experienced was exceptionally severe, prompting his complaint.

There were also cases of women mentioning their identities for similar reasons, for example, describing their pain threshold as high by referencing how many times they had given birth or describing how they had given birth with little or no pain relief, as in the following extract.

Overall, then, although Baker and colleagues (Reference Baker, Brookes and Evans2019) did not explicitly set out to identify instances of legitimation strategies being used in the discourse of the patients’ feedback, their qualitative examination of the comments revealed the recurring use of such strategies to contextualise – and, in the process, legitimate – the appraisals of healthcare services given. Such strategies were particularly visible in comments in which patients disclosed aspects of their identities (namely their age and sex identities). Where comments disclosing age tended to invoke a version of ‘authorization’ based on patients’ implicit self-constructions of themselves as expert users of healthcare services, comments mentioning patients’ sex identities tended to draw on moral values to construe patients as having a high threshold for pain tolerance. While legitimation strategies in age-based comments could be used to justify positive or negative feedback, those found in the gender-based comments tended to be used in support of negative appraisals. Baker and colleagues (Reference Baker, Brookes and Evans2019) argue that the use of such legitimation strategies, particularly in the case of negative comments, could represent a means by which patients try to ensure that their comments and feedback will be taken seriously.