‘How Propaganda Works’ reviewed by Nicolas Olsson Yaouzis

How Propaganda Works

Princeton University Press, Princeton, NJ and Oxford, 2015. 353pp., £19.95 hb
ISBN 9780691164427

About the reviewer

Nicolas Olsson Yaouzis is a visiting postdoctoral research fellow at the Institute of Advanced …


Jason Stanley’s How Propaganda Works (2015) has received a lot of attention by analytically minded philosophers and political theorists. This is partly due to the perception that the book’s subject matter has been neglected by mainstream analytical philosophy, and partly due to Stanley’s earlier contributions to philosophy of language and epistemology. In How Propaganda Works he applies insights from these fields to two major questions in political philosophy. In chapters 1-4 Stanley offers an account and taxonomy of propaganda and then uses tools from formal semantics and pragmatics to explain how it works. In chapters 5-7 Stanley moves on to discuss the relationship between inequality and what he calls “flawed ideology”, and proposes an answer to Etienne de la Boëthie’s (1548) question of why negatively privileged masses tend to accept the flawed ideology of the elites (5).

In chapters 1-3 Stanley argues that, in liberal democracies, the norm of reasonableness governs public discourse. ‘To be reasonable is to take one’s proposals to be accountable to everyone in the community’ (105). The purpose of chapter 4 is to answer how one ‘could appear to be reasonable, yet nevertheless undermine reasonableness’ (125).

Stanley argues that seemingly innocuous words and expressions, such as “welfare”, share many important properties with slurs, such as “faggot” (150-1). On Stanley’s interpretation, using a slur communicates both at-issue content and not-at-issue content. So, for example, if you were to utter

F: Johan is a faggot

you communicate the at-issue content that Johan is a homosexual man, and the not-at-issue content that homosexuals are worthy of our contempt. “Johan is a homosexual man” is the at-issue content, as this is what is up for debate when F is uttered. If I assent to F, the at-issue content would enter what Robert Stalnaker calls the common ground, which is the information we have, or presume to have, in common in a discourse (131). If I deny F, the at-issue content would not enter the common ground.

However, even if I deny F, the derogatory not-at-issue content of F would still enter the common ground. By using or accepting the use of “faggot”, I accept its derogatory not-at-issue content. That the not-at-issue content of slurs persists under negation (“Johan is not a faggot”) and quotation (“Sally said that Johan is a faggot”), is often taken to be a specific property of slurs.

Stanley, however, believes that these properties are shared by other expressions, such as “welfare”. According to Stanley even if “welfare” in the American political context has the at-issue content “cash benefits to the able-bodied, working poor”, it also has the not-at-issue content that “Blacks are lazy” (158-9). In support of this, Stanley provides a half-page summary of a psychology study from the 1990s that indicate that expressions such as “welfare” introduce the thought that Blacks are lazy (123-4). Thus, if an American politician was to say that

W: Welfare discourages poor people from taking jobs

what is up for debate is whether cash benefits discourage poor people of any colour from taking jobs. However, if Stanley is correct, W also communicates the not-at-issue content that Blacks are lazy. Because “welfare” resembles “faggot”, a denial of W would not prevent the not-at-issue content from entering the common ground. So if another politician would answer that not-W, she would according to Stanley only deny the at-issue content and not the not-at-issue content that Blacks are lazy: ‘[i]n short, even evaluating the proposal means that one must accept the social meaning’ (158).

One problem with Stanley’s account is that it is far from obvious that “welfare” operates in the same way as “faggot”. The research he quotes does not seem to provide sufficient evidence for establishing that “welfare” in fact works as a slur.

After all, even if the use of the term “summer” induces in me and many other Swedes the thought that strawberries are sweet, the content of this thought does not enter the common ground whenever “summer” is mentioned. Even if both of us would come to think that strawberries are sweet when we hear the phrase “Cameron said that the demand for electricity will not be as high as last summer”, we would not be justified in believing that this information is now common to us. What is lacking in this case is the mutual belief that summer introduces this thought.

In the case of slurs, all competent language users mutually believe that the term has a pejorative not-at-issue content. This explains why a speaker is justified in believing that his interlocutor accepts that homosexual men are worthy of contempt when the interlocutor responds to F by saying that “Johan is not a faggot”. Thus, it is not sufficient to show that “welfare” introduces the thought “Blacks are lazy” in order to show that it enters the common ground by the mere mention of welfare.

Something that indicates that Stanley agrees that the not-at-issue content must be mutually believed is that he retells how a news story from the 1970s about a woman who fraudulently filed welfare claims under four aliases, was exploited by Ronald Reagan (158-9). Stanley concludes that ‘it is scarcely possible for Americans raised during that time not to find the image of a Cadillac-driving Black urban woman popping into their head when they hear the word “welfare”’.

However, without explicit criteria of what an expression has to satisfy in order to work like a slur, it is hard to evaluate whether the little evidence he offers supports the claim that the mere mention of welfare adds that Blacks are lazy to the common ground. Considering how common the use of “weasel expressions” seems to be, it is unfortunate that Stanley does not do a better job of supporting this claim. As it stands, his account will undoubtedly provoke accusations of censorship and political correctness.

A second problem is that it is unclear how chapters 1-4 are connected to the second part of the book (chapters 5-7) where Stanley discusses inequality and flawed ideology. In the introduction Stanley tells us that the analysis in chapters 1-4 ‘explains how effective propaganda exploits and strengthens flawed ideology’ (5). Although this sounds promising, Stanley does not make explicit the connection between propagandistic utterances, the common ground, and flawed ideology. This leaves one guessing how exactly propaganda is supposed to exploit and strengthen flawed ideology.

A third problem, or perhaps peculiarity, is Stanley’s idiosyncratic use of the traditionally Marxist concept of ideology in the second part of the book. Stanley mentions Etienne de la Boëthie’s question about how the oppressed ‘come to accept the flawed ideology of the elite’ (5). Thus, it seems that he intends to use “flawed ideology” in much the same way as it is used by Marxists. Stanley adds the qualification “flawed” to be able to talk about ideology in a neutral way: “We all have ideologies […] and only some of them are flawed in the relevant sense. The ones that are flawed in the relevant sense are the ideologies that are genuine barriers to the acquisition of knowledge.” (200)

Stanley, in other words, is interested in ideologies that are flawed in an epistemic sense. He acknowledges that this distinguishes him from traditional Marxist accounts of ideology, where “flawed ideology” is usually defined in moral or political terms. He writes that “my interest lies in singling out a subclass of ideological belief that is epistemologically flawed. Its moral and political flaws are a consequence of its epistemological defects.” (198)

In addition to barring the acquisition of knowledge, Stanley adds that ‘flawed ideological belief also characteristically contributes to its own unrevisability’. This is apparent in the kind of flawed ideology Stanley associates with the slave owners of the Antebellum South, who held the belief that Blacks were inherently fit for slavery. This belief was, according to Stanley, ideologically flawed because 1) it prevented the white of the Antebellum South from coming to know that they were treating the slaves unjustly, and 2) it was immune to revision because those who questioned it were punished.

There are, however, reasons to believe that Stanley’s definition of flawed ideology is both too wide and too narrow. First, it seems too wide because it includes false scientific theories. Take for example a believer in Newtonian mechanics. She would be prevented from gaining knowledge about the behaviour of objects travelling close to the speed of light, and arguably lack the means of describing falsifying instances that would cause her to revise her belief. It sounds strange to subsume Newtonian mechanics under the same heading as the ideology of the slave owners. Second, the definition seems to be too narrow because it excludes some sets of politically problematic but knowledge-conducive beliefs from being flawed ideology. Take for example an estate agent’s belief that if Blacks move into a neighbourhood, prices will drop. Even if her belief helps her acquire knowledge about future price fluctuations, it can be argued that it shares an important feature with the ideology of the slave owners by also legitimizing the continued oppression and segregation of Blacks. This is an issue separate from the truth condition of the claim.

Stanley could respond that although the set of beliefs facilitates the acquisition of some knowledge, it prevents the believer, perhaps by obscuring the real causes, from acquiring more important knowledge. This answer will not do, however. Stanley still owes us an answer to what it is that makes one set of knowledge more important than another set of knowledge. A Marxist can answer that it is the moral and political effects that makes the difference. Stanley, however, who wants to explain moral and political flaws in terms of epistemological defects, cannot do so.

Once Stanley has established what he means by flawed ideology, he proceeds in chapter 6 to discuss the relationship between unequal societies and flawed ideology:

E: economic unequal societies give rise to flawed ideologies that contribute to maintaining economic inequality.

Stanley says that he will use mechanisms described by social psychology to show how certain flawed ideologies arise under certain social conditions (223). We could say that Stanley offers us an elaboration (to borrow a phrase from G.A. Cohen (2000, 271)) of E to show how economic inequality explains flawed ideology.

Stanley’s elaboration is given in two steps. First, he shows how certain psychological mechanisms cause the elite in an unequal society to adopt legitimizing myths. Second, he shows that because people, as a psychological fact, tend to accept the claims of authority figures and because the elite tend to control positions of authority, the poor will come to accept the myths of the rich.

It is unclear whether Stanley takes himself to provide a descriptively accurate explanation or whether the elaboration is merely intended as a plausible just-so story. As a descriptively accurate elaboration of E, it is doubtful whether it succeeds. However, I think that as a just-so story Stanley’s elaboration successfully answers a challenge posed by Michael Rosen (1996). Rosen argues that theories such as E fail because there is no plausible mechanism of belief-formation that explains how the poor systematically come to accept the favoured ideology of the rich. Stanley’s elaboration of E could be seen as an answer to Rosen’s objection. Unfortunately, although Stanley refers to Rosen (1996) he neither presents his elaboration as an answer to Rosen’s objection, nor mentions Rosen’s scepticism of theories such as E.

These problems notwithstanding, How Propaganda Works makes an important contribution to analytical political philosophy. Although flawed in some respects, the framework described in the first part of the book can be developed to help us make sense of the political rhetoric in Europe and in the US, concerning, for example, the so-called “migrant crisis”. Similarly, (let’s face it) the Marxist framework presented in the second part can provide a point of departure for analytically minded political philosophers who are interested in how flawed ideologies affect non-ideal societies. Both the merits and flaws of How Propaganda Works should influence and provoke other philosophers into taking these important issues more seriously. 

27 March 2016


  • de la Boëthie, Etienne 2012 Discourse on Voluntary Servitude Translated by James Atkinson and David Sices. Indianapolis: Hackett Publishing.
  • Cohen, Gerald A. 2000 Karl Marx’s Theory of History: A Defence Oxford: Oxford University Press.
  • Rosen, Michael 1996 On Voluntary Servitude Cambridge: Polity Press.

Make a comment

Your email address will not be published. Required fields are marked *