Adam F. Gibbons (Lingnan University), "Bad Language Makes Good Politics"
Forthcoming, Inquiry: An Interdisciplinary Journal of Philosophy
Bad Language Makes Good Politics
Department of Philosophy and Hong Kong Catastrophic Risk Centre
Lingnan University, Hong Kong SAR
In recent years, an increasing amount of work in philosophy has been devoted to how we communicate in political settings.[1] A curious feature of much of this work is its markedly negative character, with a substantial proportion of it focusing on behaviors that are, generally speaking, evaluated quite harshly by both academics and the general public. Think of lying and bullshitting, creating and disseminating fake news and misinformation, virtue-signalling and grandstanding, empty and trite sloganeering, and more. Politics abounds with such behaviors, and the curiously negative slant of the nascent literature in social and political philosophy of language and political epistemology is, one might think, simply responding to this fact. But why are these sorts of behaviors so pervasive in political settings? Why, in short, is politics full of bad language?[2]
Bad Language is Rational
The answer, I claim, is quite simple: politics is full of bad language because extant social and political institutions are structured in such a way that the production of bad language is frequently rewarded. The use of bad language of various kinds provides a reliable means to our ends. Bad language, in other words, is instrumentally rational.[3]
Consider some basic examples. Politicians lie when failing to do so will result in electoral costs. Producers of misinformation are sometimes motivated by the prospect of financial gain. People on social media grandstand for likes or shares, salient rewards for saying socially desirable things. Prominent political figures who want to communicate risky messages to a select audience might use code-words or dogwhistles in order to maintain plausible deniability. These examples (and more) are discussed at length in Section 3 of my paper.
None of this is to say that such behaviors are always rational. The expected benefits of engaging in these behaviors need to be balanced against the expected costs, and sometimes the costs will be too high. Lies can be rewarded, but they can also be penalized; grandstanding might get you some cheap likes, but it also might see you on the wrong end of a pile-on; code-words might enable you to covertly communicate something, but if your code is cracked then your audience might be larger and more hostile than you’d like; and so on. But the point is that these behaviors are rational frequently enough that they proliferate in the political settings we find ourselves in.
Some Proposals for Tackling Bad Language in Politics…
All of this bad language seemingly has bad effects. When politics is full of liars and bullshitters, the quality of the information we encounter degrades, and when the quality of available information degrades, our ability to make well-informed decisions degrades along with it. Widespread grandstanding contributes to growing levels of political polarization, as well as causing us to view one another ever more cynically.[4] Deliberately euphemistic language—such as referring to extrajudicial kidnappings as extraordinary rendition—might obscure certain events or actions which it is in our interest to see clearly. It is therefore unsurprising that many people want to do something about bad language in politics. But what, if anything, can be done?
Much bad language is a predictable consequence of people responding to incentives, and these incentives are, in turn, shaped by background social and political institutions. So one might think that by altering these background institutions we can modify the incentives agents face, diminishing the rewards they hope to attain by using bad language, or even penalizing them for using it.
For example, consider how one might motivate the adoption of fact-checking institutions. Fact-checkers can flag inaccurate statements in such a way that their inaccuracy is highlighted for all to see. If people are made aware of the inaccuracy of the relevant statements, then, in principle, they may become less likely to accept them, less likely to further share them online, and so on.[5] But if people are less likely to accept or share inaccurate statements, agents cannot expect as many benefits from spreading such statements to begin with. Accordingly, the expected benefits of intentional misinformation or fake news are reduced.
For another, consider calls to further regulate speech in order to combat misinformation. On the one hand, we might grant legislators the power to directly penalize people who spread deliberate misinformation or otherwise inaccurate information. On the other hand, legislators could approach the problem indirectly, imposing costs on platforms that fail to sufficiently monitor their content.[6] In either case, the expected costs of producing or sharing misinformation are increased, ideally to such an extent that it is no longer rational to produce or share it: in the former case, legislators themselves increase the expected costs, while in the latter it is social media platforms who are prompted by legislators to increase the expected costs.
Of course, both proposals could be pursued in tandem, and there are other proposals we could consider too.[7] But the general idea should be clear enough: by amending the background institutions with which we interact, we can modify the payoffs and penalties associated with producing bad language in such a way that it may no longer be rational to produce it.
…and Why We Might be Better Off Without Them
At least, that’s how proponents of such proposals see things. However, there are at least two serious problems that institutional responses to bad language face: (1) there is a significant risk that they will be used abused in harmful ways, and (2) there are possible social benefits to some instances of “bad” language, benefits which proposals of the sort outlined earlier would preclude if they were successful. Let’s take each in turn.
The first worry is quite straightforward. There is simply no guarantee that fact-checkers will discharge their duties in an accurate or unbiased manner. They might focus disproportionately on those they view as ideological or political opponents. They might flag certain statements as inaccurate when they are in fact accurate, they might flag certain statements as accurate when they are in fact inaccurate, and so on. Similarly, legislators entrusted with the power to penalize citizens for spreading misinformation may choose to do so in ways that benefit them and their political allies, while at the same time imposing costs on their opponents.
A peculiar trait of many who worry about the harms of misinformation is that they view the people responsible for producing and consuming it as largely biased and irrational, but then fail to consider the possibility that people given the power to police the language of others will also be biased and irrational. But our presumption here should not be that there is a class of people who, contrary to all available evidence, are somehow unencumbered by the various biases that the rest of us are burdened with (and, moreover, that these are the sort of people who end up as fact-checkers or legislators). Instead, we should be alive to the likelihood of fact-checking and further restrictions on speech being distorted by bias and irrationality.
The second worry is less obvious, but it is nonetheless worth taking seriously. As you might expect, typical treatments of bad language focus on its negative consequences. Indeed, bad language is an object of study very often precisely because of its negative consequences. But focusing only on something’s costs is plainly lopsided; instead, we should look at overall costs and benefits to comprehensively appraise it, especially if we are considering institutional proposals to reduce its prevalence.
What sort of benefits could bad language have? Consider John Stuart Mill’s famous remarks in On Liberty about how the censorship of falsehoods may cause our beliefs to become “dead dogma, not living truth”. Among other things, he thought that by preventing ourselves from coming into contact with falsehoods we also prevent ourselves from attaining a better understanding of the truth, for even falsehoods can aid us in understanding the shortcomings of our own views, correct though they may be. Or consider more recent work in philosophy about the possible social benefits of virtue signalling and grandstanding. For instance, Evan Westra argues that virtue signalling and grandstanding can play a positive role in moral change. And Neil Levy argues that virtue signalling has serves a useful epistemic function, providing higher-order evidence related to moral judgements.[8]
Does this mean that bad language in general has net social benefits? Not necessarily. Different forms of bad language will have different ratios of costs to benefits: some may be net-beneficial, while others may more closely fit our preconceptions regarding the social costs of bad language. Moreover, whether any given form of bad language has more benefits than costs is a complicated empirical question. Still, the possibility of net-beneficial bad language should caution us against prematurely adopting institutional reforms designed to reduce its prevalence.
Conclusion
Taken together, these two worries militate against pursuing institutional reforms designed to tackle bad language. At the very least, they should cause us to think more carefully about the overall costs and benefits of different forms of linguistic behavior, curbing our tendency to accentuate the negative and overlook potential upsides. Additionally, we must stop projecting idealized versions of ourselves, free from bias and irrationality, into arguments for proposals such as fact-checking institutions or intensified speech regulation. Envisaging such institutions as constituted by unrealistically impartial or rational people only serves to provide a deeply misleading picture of how these institutions will actually operate once implemented. And this, of course, is highly ill-advised, for while bad language can be very bad indeed, other things can be worse.
References
Brown, E. (2021) ‘Regulating the Spread of Online Misinformation’, In M. Hannon and J. De Ridder (eds.) The Routledge Handbook of Political Epistemology. New York, NY: Routledge
Cappelen, H. & Dever, J. (2019) Bad Language. Oxford, UK: Oxford University Press
Fritts, M., and Cabrera, F. (2022) ‘Fake News and Epistemic Vice: Combating a Uniquely Noxious Market’, Journal of the American Philosophical Association, 1-22. Doi: https://doi.org/10.1017/apa.2021.11
Gibbons, A.F. (2021) ‘Political Disagreement and Minimal Epistocracy’, Journal of Ethics and Social Philosophy 19 (2): 192-201
Gibbons, A.F. (2022) ‘Is Epistocracy Irrational?’, Journal of Ethics and Social Philosophy 21 (2): 251-81
Hannon, M. and de Ridder, J. (2021) The Routledge Handbook of Political Epistemology. New York, NY: Routledge
Hill, J., and Fanciullo, J. (2023) ‘What’s Wrong with Virtue Signaling?’, Synthese 201, 117. https://doi.org/10.1007/s11229-023-04131-4
Khoo, J. (2021) ‘Code Words’, in J. Khoo and R. Sterken (eds.) The Routledge Handbook of Social and Political Philosophy of Language. New York, NY: Routledge
Levy, N. (2021) ‘Virtue Signalling is Virtuous’, Synthese 198: 9545-62
Millar, B. (2022) ‘Epistemic Obligations and Free Speech’, Analytic Philosophy https://doi.org/10.1111/phib.12279
Rini, R. (2019) ‘Social Media Disinformation and the Security Threat to Democratic Legitimacy’, Disinformation and Digital Democracies in the 21st Century. Publication of the NATO Association of Canada. 10-4
Tosi, J. & Warmke, B. (2020) Grandstanding: The Use and Abuse of Moral Talk. Oxford, UK: Oxford University Press
Westra, E. (2021) ‘Virtue Signaling and Moral Progress’, Philosophy and Public Affairs. 49 (2): 156-78
[1] For some recent examples, see the collection of articles in Hannon and de Ridder (2021) and Khoo and Sterken (2021).
[2] I borrow this expression from Hermann Cappelen and Josh Dever (Cappelen and Dever, 2019).
[3] Note that to characterize behavior as rational in this sense is neither to accord it some positive normative status nor to criticize it. Instead, it is merely to capture the suitability of the means agents use to pursue their ends. The ends that agents pursue may be evaluated, whether positively or negatively, on independent grounds. But even agents with ends widely seen as bizarre or morally suspect can behave rationally.
[4] See Tosi and Warmke (2020) for further discussion.
[5] As I note in the paper, this defense of fact-checking makes some empirically dubious assumptions about how people form and maintain beliefs. But I set these aside in the paper, focusing on other problems the case for fact-checking overlooks.
[6] For discussion of the former proposal, see Brown (2021), Fritts and Cabrera (2022), and Millar (2022). On the latter, see Rini (2019).
[7] In addition to the above, I also discuss epistocratic institutions in the paper. I discuss such institutions elsewhere in Gibbons (2021) and Gibbons (2022).
[8] See Westra (2021) and Levy (2021) respectively. But see also Hill and Fanciullo (2023) for some recent criticism of such views.