We applied Speech Act Theory (SAT) (Austin, 1962/1975) to examine social activity around vaccination debate in OSM. A Facebook group that brings together anti-vaccination participants (AVP) and pro-vaccination participants (PVP) served as a rich data context. We found that this forum served as an information resource for groups yet undecided on vaccine adoption. SAT provides a lens to better understand the use of language and social interaction as a source of informing. For forum members who sit between, in a valley of dividedness, discussion serves to inform which side of the debate they ultimately adopt. Examining the potential for language to persuade provides insight on how conversations inform everyday life decisions.

We used an approach that examined what people said, as we felt that message delivery is impacted by how people make or frame their statements. Using SAT theory allowed us to break down the conversation post by post to analyze discussion around a vaccine-preventable disease in order to establish a baseline perspective on vaccine debate. Data were scraped from a conversation in the forum where both AVP and PVP share thoughts, opinions, perspectives, and information resources surrounding vaccines and vaccine-related issues. Lending perspective on how speech acts inform debate in the polarized Facebook group, SAT frames the analysis and divides the group conversations into three types of speech acts: 1) locutionary, 2) illocutionary, and 3) perlocutionary.

Our analysis revealed that the majority of participants demonstrated either AVP or PVP stance. AVP and PVP contribute to the conversation through all three speech acts, whereas undecided vaccine participants (UVP) statements are limited to illocutionary acts. Illocutionary statements indicate seeking additional input from the forum. A brief primer on Austin’s Speech Acts will help in understanding our analysis and conclusions.

1. Locutionary acts are declarative statements carrying no intention to persuade or provoke response. AVP locutionary acts tend to be introspective, conveying their beliefs and experiences about vaccines. PVP locutionary acts, on the other hand, indicate both inwardly and outwardly directed conversational acts. Differing themes emerged for both member perspectives: (a) anti-vaccine: 1) conversational hiatus, 2) declaration of staid perspective, 3) call for research on vaccine injury, (b) pro-vaccine: 4) meta-discussion, and 5) unsubstantiated information.
2. Illocutionary speech acts produce an utterance, suggesting warning, promising, or requesting with intent to elicit a response. The speaker has no intention to persuade others to adopt their beliefs or perspective. Rather, these acts form the core of interactive discussion that seek to deeply understand perspective of participants around vaccine. Five different themes emerged for both member stances: (a) anti-vaccine: 1) clarification/substantiation of information, 2) questioning of authority, 3) meta-discussion (b) pro-vaccine: 4) meta-discussion, and 5) information behavior.
3. Perlocutionary acts, intended or unintended, produce a causal effect on the listener. Persuading, convincing, insulting, frightening, amusing, these speech acts provoke the listener to act (or react). Intent is on listener to argue or change their perspective. Five different themes emerged for both member stances: (a) anti-vaccine: 1) lack of rigor/ethics in vaccine research, 2) victimization, 3) disease preferable to vaccine injury, (b) pro-vaccine: 4) debunking AV argument, and 5) redirecting by criticizing information behavior.

Three findings emerged from the study: 1) self-shielding behavior, 2) information-less behavior, and 3) information behavior as noise. We offer explanation of these findings couched in the three speech acts.

• Self-shielding behavior. Forum members stake their side of the vaccine debate through locutionary acts, rather than to attempt to convince the opposition through perlocutionary acts. Accompanying the position statements were locutionary acts attempting to curb an argumentative response. These statements served as an attempt to shield from an argumentative or accusatory response. None of the three speech acts describes barring or deflecting anticipatory response in discussion activities. We believe that further research examining self-protective speech act behaviors could identify a new speech act as a result of speech metamorphosis based on a world characterized by significant divisiveness.
• Information-less behavior. Although evidence-based information is not completely lacking in this forum, the data demonstrate that the majority of posts lack references to substantiating information resources. Information-less statements relegate participants to inquire (illocution) as to the source of knowledge or proof of support. Requests frequently resulted in comments and accusations about information behavior. UVP come to the forum seeking knowledge on the vaccine debate through performance of illocutionary acts demonstrating their information-seeking efforts. These acts call attention that AVP and PVP speech acts are frequently devoid of verifiable or substantiated information. Questions seek both evidence-based information and experiential accounts.
• Information behavior as noise. Provision of information, either in the form of a direct link or reference is the information. What we found is another form of illocutionary act; criticisms of information behavior. Among AVP and PVP participants, these criticisms serve to discredit the opposition. For UVP, these acts can present as noise, embodying neither information nor substantive communication. In our exploration of a divided community, we found a difference between the structure of illocutionary acts and the structure of conversation. More precisely in illocutionary logic, illocutionary force is subdivided into two components, which demonstrate the informative role and subsequently divisive nature of the discussants around the topic. These two components are 1) information construction through communication and 2) meta-information or information noise (Searle & Vanderveken, 1985, pp. 12–20; Vanderveken, 1985).

In sum, we find that the failure to provide substantiating information by participants on both sides of the debate result in criticisms of information behavior. These meta-discussions neither serve to inform nor provide substantive argument contributing to persuasion. In fact, we find that these types of comments dilute the conversation and degrade civility between opposing parties to the debate. Not to mention, they are distracting to the topic of vaccine in general since they focus on negatively expressed opinions about information behavior. We surmise that an intentional focus to provide substantiating information would play a crucial role in debate. This leads us to consider the role of information among speech acts that inform debate. In essence, the application of SAT theory provided an unexpected outcome of our research as it relates to the use and role of information in conversation. And it is one that warrants consideration and examination, particularly in OSM fora around divided information contexts that inform everyday decisions. Particularly contexts where ambiguity (lack of substantiating information, i.e., science or formal documentation) prevails over the entirety of the topic.

Findings beg further research on classification of intent of self-shielding on conversation. As we noted earlier, the act calls for consideration of a new category of SAT as it exceeds the definition of locutionary act as declarative statements. Such utterances seek neither a response nor to convince or sway perspective. They serve to establish the group member’s perspective on a debatable issue. Yet self-shielding carries an intent of barring response from others. Thus, it fails solid classification within any of the three speech acts.

Not only will the vaccine debate continue into the near future, undoubtedly the vaccine debate forum will remain bifurcated. And it will extend, most likely, into the world beyond social media in a COVID and post-COVID era. Of particular note, self-shielding behavior is an act expressing strong messages of what the speaker desires in response…no response. Strategic application of SAT serves to inform communication as well as interpretation of information more intimately. Conscious application in debate may provide a mechanism to speak strategically, empowering effective engagement in divided discussion. Conscious application in reflection on shared information provides further factors to determine information quality and adoption. Strategic application of SAT shifts focuses from speaking our thoughts to one of speaking so that others hear, reflect, and engage in an effective manner. A focus on strategic communication and information sooner rather than later may promote civility, strategic solutions, and perhaps immunity.

References

Austin, J. L. (1962/1975). How to do things with words. The William James Lectures delivered at Harvard University in 1955 (2^nd ed.). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198245537.001.0001.

Blair, D. C. (1990). Language and representation in information retrieval. Elsevier.

Lane, S., MacDonald, N. E., Marti, M., & Dumolard, L. (2018). Vaccine hesitancy around the globe: Analysis of three years of WHO/UNICEF Joint Reporting form data-2015–2017. Vaccine36(26), 3861–3867. https://doi.org/10.1016/j.vaccine.2018.03.063.

Searle, J. R., & Vanderveken, D. (1985). Foundations of illocutionary logic. Cambridge University Press.

Vanderveken, D.. (1985). What is an illocutionary force? In D. Marcelo (Ed.). Dialogue An Interdisciplinary Approach (pp. 181–204). John Benjamins Publishing. https://doi.org/10.1075/pbcs.1

Wilson, P. (1983). Second-hand knowledge: An inquiry into cognitive authority. Greenwood.

The original article on which this Information Matters translation is based is: Bonnici, L., & Ma, J. (2021). What are they saying? A speech act analysis of a vaccination information debate on Facebook. Canadian Journal of Information and Library Science, 44(1). e1–e19. https://doi.org/10.5206/cjilsrcsib.v44i1.13342