Against Irrationalism in the Theory of Propaganda

Abstract

According to many accounts, propaganda is a variety of politically significant signal with a distinctive connection to irrationality. This irrationality may be theoretical, or practical; it may be supposed that propaganda characteristically elicits this irrationality anew, or else that it exploits its prior existence. The view that encompasses such accounts we will call irrationalism. This essay presents two classes of propaganda that do not bear the sort of connection to irrationality posited by the irrationalist: hard propaganda and propaganda by the deed. Faced with these counterexamples, some irrationalists will offer their account of propaganda as a refinement of the folk concept rather than as an attempt to capture all of its applications. The author argues that any refinement of the concept of propaganda must allow the concept to remain essentially political, and that the irrationalist refinement fails to meet this condition.

Introduction

According to many accounts (see Stanley 2015; Ross 2002; Marlin 2002; Ellul 1973), propaganda is a variety of politically significant signal with a distinctive connection to irrationality. Depending on the account, the irrationality may be supposed to be theoretical, or practical; it may be supposed that propaganda characteristically elicits this irrationality anew, or else that it exploits its prior existence. The view that encompasses such accounts I call irrationalism. The purpose of this essay is to argue that irrationalism is misguided.

I want to be clear up front about the success conditions of the project that I take irrationalism to be an attempt to complete. I do not think an inquiry into the nature of propaganda can consist of analysis of a single pretheoretic concept of propaganda or determination of the extension of the term propaganda according to a single stable pattern of use. I suspect that the exact content of the concept and extension of the term differs across communities. For example, it has always been common for leftists in the United States and elsewhere to refer to their own promotional materials as propaganda, yet this clearly does not entail a negative judgment; whereas political discourse emanating from communities closer to the center of the political spectrum use the term propaganda almost exclusively as a term of condemnation. While it is possible that these different patterns of usage reflect an agreement about the extension of the term but a disagreement about the normative status of things in that extension, what is more likely is that they reflect genuinely different, though related, term extensions and genuinely different, though related, concepts. Against such a backdrop, the selection of any one pretheoretic concept for analysis, coupled with claims about the authoritativeness of that analysis as an account of propaganda tout court, could only be arbitrary.

The projects remaining open are then (1) a genealogy or a comparison between different concepts, without any claims to the primacy of one over the other and (2) an explicative, analytical, or ameliorative definition of propaganda motivated by specific theoretical purposes. The latter is my focus in this essay, and I take irrationalism to be an attempt at it. When engaged in explicative definition, as W. V. O. Quine put it, ‘the purpose is not merely to paraphrase the definiendum into an outright synonym, but actually to improve upon the definiendum by refining or supplementing its meaning’ (Quine 1951: 24–25). The refinement in this case is guided by a desire to arrive at a concept that serves the purposes of political theorizing per se.

1. Preliminary Remarks on Propaganda

Instances of propaganda are signals, where a signal is anything that transmits some information or content. In what follows, I use the term information without any factive commitment: information, for my purpose, can be true or false. The thing to observe from the literature naturalizing signalhood is that something can carry information while functioning for other purposes. A footprint in the dirt is a signal bearing information of interest to a tracker or conservation biologist, though for the creature that created it may be a mere incidental consequence of getting where she was going. In H. P. Grice's (1957) famous example, grey storm clouds are a signal (carrying the information that rain is on its way) though they are also nonsemantic meteorological phenomena. I think we should accept that propaganda includes all sorts of phenomena that are signals in addition to being other things. Where I discuss hard propaganda and propaganda by the deed below, I introduce some cases of propaganda that are also tokens of other sociopolitically and materially significant types.

As the footprint and raincloud examples also make clear, signals need not be intentionally produced, as such or at all. While I am, in this essay, neutral on the question of whether propagandistic signals must be intentionally produced qua signals, what is evident, I think, is that propaganda does not have to be intentionally produced qua propaganda. This is because propaganda can be produced by structures or systems in which no one individual intends for the signal to perform aims characteristic of propaganda. Here one may think of structure much as Iris Marion Young does, where ‘structure denotes a confluence of institutional rules and interactive routines, mobilization of resources, and physical structures . . . The term structure also refers to wider social outcomes that result from the confluence of many individual actions within given institutional relations, whose collective consequences often do not bear the mark of any person or group's intention’ (2003: 4).

That propaganda can be produced by structures is an idea familiar from such classics of the literature as Edward Herman and Noam Chomsky's Manufacturing Consent (2011), which advanced a ‘propaganda model’ of the US media ecosystem. Herman and Chomsky present data (from roughly the mid-1960s to the 1980s) suggesting, for instance, that mainstream media coverage of atrocities perpetrated by the United States’ adversaries was significantly greater and more sympathetic to victims than was the coverage of atrocities perpetrated by the United States itself or by its client states (2011: 37–86). But they do not take this to be the result of, say, deliberate orchestration by a propaganda minister at the helm of all US media. Rather, they characterize it as the net result of a variety of different pressures on newsrooms, such as the need to attract advertisers and to stay cozy with state sources. The cumulative effect of major media outlets’ navigation of these practical concerns is, on this presentation, a propagandistic media environment. Requiring that a signal (or, to perhaps stay closer to Herman and Chomsky's intent, set of signals) be intended to bring about the ends characteristic of propaganda would exclude such cases, and I take that to be undesirable.

I do of course feel the force of the concern that signals that just happen, flukishly, to bring about whatever effect (E) a given account says is distinctive of propaganda, perhaps ought not themselves be called propaganda—clearly the intention requirement, while flawed for the above reasons, is an attempt to positively articulate what it is required for a signal's effects not to be flukish. I also feel the force of a worry about the inverse: that is, probably cases that fail to achieve E but were in some sense designed to do so should be considered propaganda. These complications are not special to the consideration of propaganda; rather, they are endemic in theoretical work that focuses on the meanings of signal-type terms (such as tell, show, warn), nontechnical usage of which seems to shift from one side of the action/non-action and success/non-success distinctions to the other. Here I will rest content with just a negative and disjunctive articulation of a stance on these issues: propaganda is a signal that non-flukishly achieves, or was disposed to achieve, some effect, E. In what follows, I argue against a particular view about what E is.

2. On Irrationalism

The irrationalist says that propaganda is a kind of signal the characteristic effect of which (E) has something to do with audience-side irrationality. The ‘audience-side’ qualifier makes clear that the irrationalist view does not envision propaganda as encompassing all signals that were, say, produced in a fit of irrationality by the signaler. In what follows I talk about irrationality per se as the focal feature of irrationalism, but in all cases, it is audience-side irrationality that I mean.

I argue against irrationalism as such, rather than against any one particular irrationalist account. And I construe irrationalism as a focalist view: that is, I take it to be a view about what the core or ‘focal’ feature of propaganda is.¹ The irrationalist account says that the core feature of propaganda is audience-side irrationality, but allows that tokens of the type may bear slightly different relations to this core feature. This reflects the fact that different views that I would classify as irrationalist seem to connect propaganda to different varieties of irrationality and render the nature of this connection differently.

Particular accounts that I categorize as irrationalist describe propaganda as the ‘manipulation of the rational will to close off debate’.² as ‘speech that irrationally closes off certain options that should be considered’ (Stanley 2015: 48, 49 on the ‘classical’ sense of propaganda); as ‘the organized attempt through communication to affect belief or action or inculcate attitudes in a large audience in ways that circumvent or suppress an individual's adequately formed, rational, reflective judgment’ (Marlin 2002: 22, my emphasis); and as ‘an epistemically defective message used with the intention to persuade a socially significant group of people’ where epistemically defective persuasion stretches to cover not only falsehoods and instances of misleading but also the use of ‘spurious’ means, like emotional arousal, to persuade (Ross 2002: 24, my emphasis).

Worth singling out is the influential discussion in Jason Stanley (2015), which I take to offer an irrationalist account. He regards it as possible that his conception of propaganda is a version of the ‘classical’ view articulated above (2015: 48) and, accordingly, inclines toward the view that ‘propaganda runs counter to rational principles’, focusing on a variety of propaganda (‘undermining demagoguery’) that conspires to use liberal democratic ideals against themselves, with this contradiction (irrationally) unnoticed by the audience because of their preexisting flawed ideology (2015: 57–58). The case is even clearer in Anne Quaranto and Stanley (2021), who declare right out of the gate that they define propaganda as ‘an argument that bypasses reason’ (125).

Stanley's (2015) account can seem to connect propaganda most tightly to theoretical irrationality. But as Quaranto and Stanley (2021: 135) acknowledge, irrationalism per se also encompasses views that connect propaganda to other varieties of irrationality. For instance, while Jacques Ellul's view seems to entail a rejection of the idea that (‘modern’) propaganda aims to bring about theoretical irrationality, he explicitly commits to a connection between propaganda and practical irrationality: ‘Propaganda is very frequently described as a manipulation for the purpose of changing ideas or opinions, of making individuals “believe” some idea or fact, and finally of making them adhere to some doctrine—all matters of mind . . . This line of reasoning is completely wrong . . . the aim of modern propaganda is no longer to modify ideas but to provoke action. It is no longer to change adherence to a doctrine, but to make the individual cling irrationally to a process of action’ (1973: 25, my emphasis)

Likewise, a view that articulates propaganda's characteristic effect as the creation of an alienation from one's own perspective, or a false consciousness (in the tradition of Marx and Engels [1932] 1972; Marcuse 1964; Horkheimer and Adorno 2002) can be read as a variety of irrationalism, whether this alienation is cashed out as theoretical rationality (one comes to form unsupported beliefs about oneself) or in volitional or practical terms (one comes to form desires that are in some sense false, and which conflict with one's true underlying desires).

In emphasizing—and, in what follows, criticizing—what this broad range of views has in common, my approach bears some connection to that of Cory Wimberly (2017, 2020). Wimberly emphasizes that a number of superficially different worries aired over the last century about the manipulation of public opinion are alike in worrying that public opinion has become somehow ‘false’—though this falsity comes in epistemological, ethical, and ontological varieties, depending on the thinker. Wimberly's objection is that propaganda as such need not falsify anything at all: often propagandists seek to ‘transform subjects and their conduct’ (2017: 108), creating genuinely new publics rather than repressing or covering over a persisting true one. This point about falsity is distinct from my point concerning rationality, but they are similar in spirit. Both my view and Wimberly's emphasize that, in connecting propaganda too closely to one particular flavor of normative breach, we have missed propaganda's true scope, and the scope of the threat it can present.

Different varieties of irrationality gesture at an incoherence somewhere in the system comprising the agent and her actions. To regard irrationality as the focal feature of propaganda is then to connect propaganda to a state of incoherence in the individual—some accounts might say that this connection lies in the fact that propaganda renders the individual irrational, whereas others might say that propaganda exploits irrationality that is in some sense already there.

To oppose irrationalism, one need not argue that instances of propaganda are never connected, causally or otherwise, to irrationality of any kind—just that this is not always the case, and so should not be characterized as the focal feature of propaganda per se. Below I discuss two varieties of propaganda that fail to bear the sort of connection to irrationality posited by the irrationalist.

3. Against Irrationalism

Some cases that turn out to help make the case against irrationalism are those in which an instance of propaganda works—and is designed to work—by conveying information beyond the nonnatural or speaker meaning of the signal. In particular, I have in mind cases that vary from the paradigm of speaker meaning familiar from Grice (1957) in that they resemble showing more than telling. The difference between telling and showing is the difference between a person saying (in English), ‘I speak fluent Urdu’, and their in fact proceeding to hold forth in Urdu; it is the difference between my saying ‘Nikita Khrushchev once took off his shoe and banged it on a desk of the UN General Assembly’ and my offering you a photo of him doing it. Whereas Grice says that speaker meaning, which is characteristic of ‘telling’ cases, occurs when a signaler produces a signal ‘with the intention of inducing a belief [in the audience] by means of the recognition of this intention’ (Grice 1957: 384), showing cases fall short of this standard in that, while the signaler may intend to bring about a belief (or some similar attitude), and may intend the audience to see that they, the speaker, had this intention, they do not take it that recognition of their intention will be necessary in order for the audience to come to have the relevant belief (for further discussion of showing and blended showing and telling cases see Neale 1992; Sperber and Wilson 2015). When I present you with the photo of Khrushchev and the shoe, your recognition that I want you to believe this occurred is causally otiose when it comes to bringing about that belief in you—it is this further sort of evidence, the photo, that does the trick. Below I demonstrate how instance of so-called hard propaganda and propaganda-by-the-deed both function more on the model of showing than telling. I also make clear how, for this reason, they function as counterexamples to irrationalism.

3.1 Hard Propaganda

The first class of propaganda that fits this bill is brought out below, in Lisa Wedeen's ethnography of Syrian political life in the 1970s–1990s: ‘In official Syrian political discourse, President Hafiz al-Asad is regularly depicted as omnipresent and omniscient. In newspaper photographs he appears as the “father”, the “combatant”, the “first teacher”, the “savior of Lebanon”, the “leader forever”, or the “gallant knight” . . . [But] no one in modern Syria, neither those who orchestrate official praise nor those who are forced to consume it, believes that Asad is the country's “premier pharmacist”, that he “knows all things about all issues,” or that he actually garners 99.2 percent of the vote in elections’ (1999: 1).

Haifeng Huang, whose own work primarily concerns contemporary China, dubs the sort of propaganda that Wedeen draws our attention to ‘hard propaganda’ (2015, 2018). Hard propaganda involves messages that are ‘crude, heavy-handed, or preposterous’, which can be ‘seen through by citizens’ and which therefore ‘do not induce persuasion’ (Huang 2018: 1034).

One conclusion about such cases of propaganda might be that they are just ineffective, but that they do still aim to generate belief, and perhaps to do so via epistemically spurious means. But Wedeen and Huang both argue that such propaganda functions strategically to bring about some end other than belief in the speaker-meant content of the speech act. In broadcasting articles of absurd jingoism or leadership-cult doctrine, the state aims to demonstrate that it can dominate the media environment and command attention even for absurdities. That is, it aims to demonstrate its power, and so intimidate dissidents into silence. Huang for instance argues, ‘Chinese citizens frequently dislike and ridicule the state's flagship TV news program Xinwen Lianbo, [but] the fact that the regime easily bombards the nation with the program daily at 7:00 p.m. manifests its power’ (2018: 1034–35). One function of hard propaganda is then to convey the belief that just as the propagandist (in the above cases, the state) is powerful enough to dominate a media environment, they are powerful enough to crush resistance.

If the hard-propaganda analysis of such cases is correct, and they display the state's might specifically because they are ‘preposterous’, a question may arise as to why the speaker-meant content of such propaganda would have a political flavor at all. If flagrant deviation from conversational norms is part of the mechanism by which such propaganda works, why would we not find more state news programs just saying things like ‘giraffes are flamingoes!’ ‘snow is purple!’ and ‘squares are circular!’, or even descending into actual gibberish? Why bother crafting absurdities germane to the greatness of the leader or infallibility of the state, if there is no expectation that the propaganda signal will actually induce these beliefs and the only thing that matters is the ability to broadcast the patently outrageous? A partial answer is that some propaganda is only ‘hard’ for certain audiences—while one audience is not susceptible to being persuaded, and can only be intimidated, another may be persuadable. For this second audience, the topic of the message then is significant; for the signal to perform its requisite function vis-à-vis both audiences it must be both audacious and a statement of support for the leader or state. Huang (2015) thinks that this is the right analysis of the Chinese propaganda environment he analyzes. But if Wedeen is right that virtually no one believed the leadership cult messaging of Hafiz al-Assad's Syria, this cannot explain the persistent political themes of his regime's propaganda. For this explanation, we must look to another function that these signals play, according to Wedeen: to act as scripts or templates for the audience's own conduct. Wedeen attests that the strategy of compelling individuals to participate actively in pro-regime spectacles of varying scales was common, with official state festivities often including audiences of university faculty and students corralled on their campuses and transported in on buses (1999: 2, 68), for instance. And this compelled participation often included adopting, even augmenting, the official propaganda rhetoric, as in the following anecdote about a Syrian military regiment: ‘One day a high-ranking officer visiting the regiment ordered the soldiers to recount their dreams of the night before. A soldier stepped forward and announced: “I saw the image of the leader in the sky, and we mounted ladders of fire to kiss it”. A second soldier followed suit: “I saw the leader holding the sun in his hands, and he squeezed it, crushing it until it crumbled. Darkness blanketed the face of the earth. And then his face illuminated the sky, spreading light and warmth in all directions”. Soldier followed soldier, each extolling the leader's greatness’ (Wedeen 1999: 67).

The dream-fabricating exercise recounted here asked the soldiers to generate images that fit with the rhetorical and symbolic formulae disseminated via state propaganda, and so in effect to produce propaganda of their own. It is evident that the high-ranking officer is not concerned so much with whether the soldiers actually had the dreams that they recount, as he is with their performance of compliance with the formula, which involves acting as if they were true believers in the leadership cult, with their mental lives as saturated with the prescribed iconography as such dreams would suggest.

This scripting function of hard propaganda makes evident why the specific political content is important even where there is little expectation that this content will be believed. We must conjecture that forcing an individual to say something deeply contrary to their beliefs on a subject that is of great importance to them is degrading in a way that forcing them to utter gibberish is not. Forcing me to effusively praise a leader I regard with contempt (or even ambivalence) is a distinctive way of compromising my political agency. A core function of this ‘politics of as-if’, on Wedeen's account, is to bring it about that the subject ‘comes to know about himself, and about the others, that each can be made to subordinate to state authority not only his body, but also his imagination’ (1999: 81). On Wedeen's account, this knowledge functions to undermine the sense that one has a coherent set of political values, which brings about a ‘depoliticization’ of the individual (1999: 81). She quotes a former employee of the Syrian Ministry of Culture as saying that the function of the official discourse was to: ‘monopolize public discourse absolutely, to kill politics, to eliminate politics as a means of defending oneself or expressing oneself’ (1999: 45). It is notable that not only does this goal not require belief in the speaker meaning of the state propaganda, but it actually requires that the individual be forced to say and do things they do not believe in.

Hard propaganda can bring about the belief that the regime is powerful, and, via its prescriptions for individual conduct, the belief that one has oneself become complicit in the regime. The beliefs that I have just described are not obviously formed in an irrational way. Where I come to believe that the regime is powerful, it is because I have seen evidence of this—the expensive and logistically demanding domination of a media environment, for instance. And where I come to believe that I can be made to comply with the regime's prescriptions, as anathema as these may be to the values I had heretofore taken myself to hold, it is because I have seen evidence that this is so—my own humiliating recitation of state rhetoric or participation in state spectacle.

So hard propaganda does not function by inciting its targets to theoretical irrationality—that is, to an inconsistency among beliefs. Nor, I think it is clear, does the conduct that it elicits seem clearly practically irrational; I think we should judge that the individual who chooses to back down from a confrontation with a powerful state, or even to act as if they were a true believer in a leadership cult, may be acting in precisely the way that makes sense given a desire for survival.

There is however one remaining form of irrationality that the irrationalist might discern in (some) instances of hard propaganda. One conception of rationality owed for instance to Nietzsche and to Freud, concerns in the first place rationalization: a capacity to create, rather than to discover, a coherence among characteristics or events. A specifically modern sense of irrationality, then, is instantiated when, as Richard Rorty has put it, the agent ‘becomes irrational not in the sense that she has lost contact with reality but in the sense that she can no longer rationalize—no longer justify herself to herself’ (1989: 177). And one might suspect that this is precisely what is going on with the depoliticization effect that Wedeen notes. Indeed, Wedeen gestures at Rorty's comment that, ‘people can, their torturers hope, experience the ultimate humiliation of saying to themselves, in retrospect, “Now that I have believed or desired this, I can never be what I hoped to be, what I thought I was. The story I have been telling about myself—my picture of myself as honest, or loyal, or devout—no longer makes sense. I no longer have a self to make sense of’’’ (1989: 178).

The inability to rationalize does not point to an incoherence in the individual at any one time, but posits that a particular sort of alteration over time counts as incoherence—specifically, the kind of alteration that stands in the way of attributing to oneself any absolute values or beliefs in the story that one tells about oneself. Whereas the individual may have taken their political commitments to be core to who they were, being compelled by ambient threat of force to praise the regime compels them to contend with evidence that preserving their own life, status, or safety, was more important to them than absolute adherence to their principles. So while there is a perfectly coherent decision-theoretic story to be told about the agent who consents to reproduce state rhetoric, this story will cast those desires that the agent took to be fundamental to themselves as contingent and conditional after all. Perhaps the resulting inability to rationalize is evidence that even hard propaganda is to be characterized in terms of a connection to something deserving to be called irrationality.

My method in this paper has been to be expansive about which sorts of rationality are encompassed by the irrationalist account, and so I will not now deviate by trying to rule this variety of irrationality out of court. What I will say in response to the above concern though is that not all hard propaganda interferes with the agent's ability to rationalize, since not all instances of hard propaganda are accompanied by the expectation that the audience compromise themselves by repeating back a variation on the message. Those that do not will still count as counterexamples to irrationalism.

3.2 Propaganda by the Deed

The second variety of propaganda that I think motivates rejection of irrationalism is propaganda by the deed, a kind of propaganda originally associated with late nineteenth-century European anarchist thought. Propaganda by the deed, unlike mere theoretical propaganda or propaganda of the idea, generally took the form not of posters, pamphlets and speeches, but of protests, assassinations, acts of industrial sabotage, and insurrection. In general, propaganda by the deed consisted of actions designed to demonstrate, and not merely assert, the power of the working class and the fallibility of the state. To head off a potential misunderstanding, to say that an assassination is an act of propaganda is not of course to deny that it is also an act of murder, any more than attributing signalhood to a rain cloud (i.e. ‘that cloud means a storm's coming’) amounts to denying that the cloud is also a meteorological phenomenon. Acts of propaganda by the deed make it particularly evident that propagandicity may be a dimension of events in our political environment without exhausting those items’ political and material significance. While the exact origins of the idea and of the expression propaganda by the deed are uncertain (for a discussion of these origins, see Cahm 1989: 76–91), one early occurrence is an 1877 article by Paul Brousse (which is sometimes, though perhaps erroneously, also attributed to Peter Kropotkin). On Brousse's articulation, propaganda by the deed is: ‘a way of grabbing [the] people’s attention, of showing them what they cannot read, of teaching them socialism by means of actions and making them see, feel, touch’ ([1877] 2004: 150). On Brousse's account, such propaganda is critical in coming to have any influence over the masses of peasants and workers who spend ‘most of the time toiling eleven and twelve hours per day’, and therefore ‘make their way home worn out from fatigue and have little inclination to read socialist pamphlets or newspapers’ ([1877] 2004: 150). An intriguing additional rationale for propaganda by the deed was given by Italian anarchists Errico Malatesta and Carlo Cafiero who had previously written in an 1876 letter to the Jura Federation, ‘The Italian Federation [of anarcho-socialists] believes that the insurrectional act which is intended to affirm socialist principles by deeds, is the most effective means of propaganda and the only one which, without deceiving and corrupting the masses, can penetrate down to the deepest levels of society’ (as quoted in Guillaume 1910: 116, my emphasis).

Here is a concern not just with getting through to their intended audience, but with doing so without in some way harming that very audience. This speaks of a general concern about the compatibility of propaganda with anarchist political ideals. Many discern a general mistrust of representation as a theme in anarchist thought (May 1994; Cohn 2005; Colson 2017), where this mistrust goes beyond the rejection of representative democracy and trickles down to all sorts of other ethical, epistemological, and metasemantic questions. The anti-representational perspective that is relevant to us here is one that is suspicious of the representations of political persuasion, on the grounds that ‘in giving people images of who they are and what they desire, one wrests from them the ability to decide those matters for themselves’ (May 1994: 48). Propaganda by the deed, like hard propaganda, is propaganda that shows rather than tells—and in comments like the above we see that opting to show rather than tell was a very deliberate strategic response to a variety of pragmatic and ethical concerns.

To be explicit about why propaganda by the deed functions as another class of counterexample to irrationalism, it will be helpful to talk about a specific example. Although the term has come to be associated mainly with acts of violent insurrection, and indeed with terrorism, the perpetration of violence was no necessary condition on propaganda by the deed on at least early conceptions. Indeed, an example that Brousse gives is of an 1877 protest in Berne, Switzerland, which was able to function as anti-state propaganda because the police attacked the protestors for displaying their revolutionary red flag. Whereas ‘the Swiss bourgeoisie nurtures in the mind of the Swiss working man a prejudice that he enjoys every possible freedom,’ this spectacle ‘was a practical demonstration laid on for Swiss working folk in the public square, that they do not, as they thought they did, enjoy freedom’ (Brousse 2004: 151). What we should observe about this case is that, like the hard propaganda discussed before, it may function to instill a belief in its audience (in this case, the working class of Switzerland), but it does so by providing evidence—and perceiving that people are not free to engage in certain kinds of political speech without being assaulted by the apparatus of state functions as perfectly good evidence for the conclusion that they do not enjoy ‘every possible freedom’. There is no clear violation of rationality here.

It is no accident that propaganda that works by showing rather than telling should systematically stand as a counterexample to irrationalism. Showing is a hybrid between natural and nonnatural meaning precisely because it involves an intentional deployment of a natural sign token—that is, it involves an agent deliberately directing attention to an object or event that carries some information, p, in the sense that the presence of the object or occurrence of the event is to some degree positively correlated with the truth of p. Instances of (successful) showing then, by definition, involve directing the audience's attention to evidence (of some strength), with the intention of getting them to update their attitudes on the basis of this evidence. And updating one's attitudes on the basis of evidence is the paradigmatic rational activity. Far from hoping to exploit some irrationality in their audiences, both hard propaganda and propaganda by the deed make essential use of their audiences’ rationality.

4. What Role Should a Concept of Propaganda Play?

I have presented two classes of propaganda that are not connected to audience-side irrationality in the way posited by the irrationalist. One kind of response to these counterexamples to irrationalism will proceed by discerning some way in which the above cases do bear a connection to some sort of irrationality after all (an approach that I gave some expression to at the end of my discussion, above, of hard propaganda). Given that I have characterized irrationalism as a focalist view rather than one that posits a specific connection to a specific form of irrationality as a necessary condition, my rebuttals can probably never be entirely definitive: I can perhaps never ensure that, for any one of my putative counterexamples, there is no exotic variety of irrationality I had not thought of with which the case bears some sort of connection I had not thought of. I take this to be in the nature of a dispute over the aptness of a focalist account, rather than a defect of my argument in particular.

In any case, I think that the more common response to the above cases is not to affirm that they all in fact involve audience-side irrationality but to make explicit that the project here is not after all to offer an account that captures all uses of the term propaganda; this sort of irrationalist will happily admit that neither hard propaganda nor propaganda by the deed count as propaganda at all on their account, because they are looking to offer a refined concept of propaganda, rather than to analyze some prior concept that stood behind all folk uses of the term. There is no doubt more to be said about why the irrationalist would come to favor a different concept of propaganda than did Brousse, for example, and why different communities and ideologies might give rise to these differing concepts—inquiry in this direction would amount to the comparative or genealogical project at which I gestured in the introduction. The project that interests me, however, is the explicative one—my claim is that, given a certain set of compelling theoretical interests, the irrationalist's preferred concept is not the one we should put in the crosshairs of a theory of propaganda. This is because the central function we should want the concept of propaganda to play is to serve social and political theorizing. Among all the signals in the world and indeed all the signal-types that appear in politics, the propaganda theorist is tasked with demarcating a type that is significant to political theorizing per se. The trouble with the irrationalist's concept is that it is just not distinctively political at all.

The psychodrama of individual internal incoherence is a story that can be told without reference to a political community or project beyond the individual. While a political project may advance by imposing incoherence on the individuals who are grist to its mill, it is important to observe that this makes only for an occasional causal connection between propaganda and the political. Regardless of how precisely one characterizes the political, there are clearly nonpolitical causes of individual irrationality as well as nonpolitical effects of individual irrationality, so even a causal connection between the irrationalist's conception of propaganda and the political will be inconstant and seemingly non-fundamental. Propaganda, on the irrationalist's conception, is a phenomenon squarely in the domain of action theory, or perhaps epistemology; but it will have only occasional, incidental bearing on social and political theorizing.

I do not offer a positive account of propaganda in this essay. But it does make sense to ask here whether the objection that I am laying at the feet of irrationalism will not apply in equal measure to any plausible alternative view. And I do not think it will. The sort of view that would evade this objection is one that posited as propaganda's characteristic effect something constitutively political. Politics is intersubjective: it has to do with relations and activities between individuals. A view that has the capacity to forge the kind of tight connection between propaganda and politics that I think we should want will be one that locates propaganda's distinctive effect not at the level of the individual agent but at the level of more complex social entities. Call this a holist account of propaganda.

Hannah Arendt endorses a holist view of propaganda in The Origins of Totalitarianism, where she argued that the ‘true goal of totalitarian propaganda is not persuasion, but organization of the polity’ (1994: 361). For Arendt, the level at which propaganda's characteristic effect is registered seems to be a movement, polity, or other organized structure (for a discussion, see Hyska 2018, 2021). I myself have elsewhere articulated a view related to Arendt's, on which propaganda's characteristic effect is to create or destroy group agency (Hyska 2021). Wimberly (2017, 2020) reads as a holist to the extent that he emphasizes propaganda's modification of macro-level traits of publics. And the ‘practice first’ account of propaganda offered by Olúfẹ́mi Táíwò (2018) is another sort of holist account, as Táíwò treats ‘the public mental representations that make up the content of the common ground’ rather than any individual-level representation as ‘the paradigm target of political intervention’ by propaganda (2018: 307–8).

Laying forth an argument for holism in general, let alone for a particular holist account, is beyond the scope of this essay. But the presence of these views in the literature should reassure us that not all views will fall afoul of the objection I raise against irrationalism.

Finally, I think that the reluctance to let go of irrationalism in some quarters stems from a feeling that, without tying the concept of propaganda closely to that of irrationality, the concept can no longer perform the critical function that we are used to it playing. It is true that, if one gives up irrationalism, to say that a signal is propaganda is no longer ipso facto to say that the signal is responsible for inciting or exploiting the nonnormative state that is irrationality. But this does not mean that the term becomes useless for the purposes of critique. For all that I have said here, the right positive account of propaganda might define propaganda in terms of some other normatively problematic condition, in which case to call a signal propagandistic will still be, ipso facto, a criticism. But I note that something like existing critical practices involving the term propaganda can be retained and made sense of even if the term is not rendered inherently evaluative in this way. On any account that is consistent with the imperative to cast propaganda as something politically significant per se, propaganda will, by its nature, have designs on the way that people coordinate to exercise power (or do not). A typical use of the term will then be to point out the presence of a political agenda behind the signal; that is, to point out that an individual agent has intended the signal to produce, or a system has adapted to give rise to the signal precisely because it will have some tendency to produce, a set of (political) states or actions beyond mere comprehension by the audience. Of course, to associate a signal with an agenda is not itself to allege that the signal is problematic in any way—the familiar Austinian notion of perlocution picks out the speaker's act of trying to induce further actions beyond content and force recognition in the audience, and describes cases as mundane as the utterance ‘Can you pass the salt?’ (which is of course uttered with the hope that the audience not only will understand it but will indeed pass the salt—an ‘agenda’, but hardly a troubling one). Nor, on most accounts of the political, will it turn out to be inherently problematic to have a political agenda. But where we bother to point out the presence of a political agenda behind a signal, we do so because there is something notable about it; uses of the term propaganda will therefore often serve to focus attention this agenda. Once pulled into focus, the political agenda behind a signal can of course be critiqued in lots of different ways—the basic irrationalist impulse may find expression in criticizing its advancement of irrationality for instance, while other forms of critique may target other troubling formal features or else the substantive political vision that it serves. On a non-irrationalist account, the term propaganda can still serve these different varieties of critique even without their specific critical accusations being built into the term's definition.

5. Conclusion

My argument has been about a specific, somewhat marginal political phenomenon, propaganda. But the argument I have given is not without resonance elsewhere in social and political philosophy. One upshot of my view is that the terms of critique we use to diagnose the woes of our politics or our political discourse, cannot (or at least should not) all be cashed out in terms of irrationality. This harmonizes with the trend toward questioning whether the phenomenon of political polarization is reliably the result of irrationality either (Kelly 2008; Singer et al. 2019; Dorst 2020; Nielsen and Stewart 2021; Pallavicini, Hallsson, and Kappel 2021). The claim in that literature has been that, while the tendency for individuals’ views to converge with those within their political group and diverge from those in different or opposing groups may look like it must rest on some nonnormative treatment of evidence, this result is in fact consistent with a system full of good Bayesian conditionalizers. Polarized, like propaganda is a term frequently used to express a critique of political discourse; a theme running through this work and mine is that faults in political discourse are not essentially connected to individual irrationality.

Existing uses of the term propaganda cover cases that do not involve audience-side irrationality. Beyond forcing us to exclude such cases, the irrationalist account makes propaganda over as a phenomenon in the domains of action theory or epistemology, rather than one within political philosophy. Clearly, I hope that my argument may have discouraged the reader from embracing irrationalism. But if I have merely clarified that debates concerning the nature of propaganda are fundamentally about whether to focus discussion on politics itself or upon the individual constituents of a polity, I think this, too, has been worthwhile.