On Stanley on Propaganda and (Tarski style) Truth (II)

It might be thought that the theory of ideology has ideology as its subject matter, and so cannot itself contain its subject matter. But this thought too is incorrect. The logician Alfred Tarski provides a proof of the soundness of the axioms of the calculus of classes in a metatheory. But the reason the soundness proof works is that the meta-theory also has axioms that express those same principles. The meta-theory is not an attempt to provide the justifications of the propositions expressed by axioms of the theory to someone who doubts  them. Its task is different: it is to deliver important knowledge about the object theory. Similarly, the task of the theory of ideology is to yield important knowledge about ideology. Even if the theory of ideology is ideological, it can issue in knowledge. As in the metatheory for logic or set theory, a neutral stance is neither possible nor required, Jason Stanley (2015)  How Propaganda Works, 77.

Yesterday, I noted that there are three distinct features of Stanley's approach to propaganda: first, that propaganda amplifies (or reduces) ideals; second, that it does so by nonrational means; third that propaganda is also possible, even endemic, in liberal democracies (49ff). While I propose an alternative approach, Stanley is right that propaganda can be true and/or sincere.

One notable feature of Stanley's approach is that he grants (see the quote above) that his own theory of ideology cannot escape being ideological itself. In fact, I argued that his analysis presupposes a kind of deliberative understanding of public life with accompanying ideals of "fair deliberation and equal participation." This limits his approach in contexts that have very different or conflicting fundamental commitments about politics (and it is one reason why I argue for a different approach).

More subtly, Stanley's account of propaganda has, in addition to the three distinct features I noted before, a kind of consistency or integrity principle built into its exemplary definitions. Recall his definitions:

Supporting Propaganda: A contribution to public discourse that is presented as an embodiment of certain ideals, yet is of a kind that tends to increase the realization of those very ideals by either emotional or other nonrational means.

Undermining Propaganda: A contribution to public discourse that is presented as an embodiment of certain ideals yet is of a kind that tends to erode those very ideals. Jason Stanley 53 (emphasis in original).

In Stanley's hands, propaganda moves one from a status quo (in terms of an ideal) and, in so doing, moves one toward more coherence (supporting propaganda) or toward incoherence (undermining propaganda). This emphasis on consistency embodies an ideal of rationality (but by no means sole ideal of rationality), as is revealed by the very clever and surprising appeal to Tarski when he describes his approach to ideology (quoted above). Recall that in Tarski's famous paper the "proof of the soundness of the axioms of the calculus of classes" is a means toward explicating truth.

For, at a high level of generality, one can understand Tarski's explication of truth as combining in creative fashion a commitment to a correspondence theory of truth (Tarski calls it the "classical" even "Aristotelian" conception of truth [recall]) and, on the meta-level, instantiating coherence as the ideal of rationality. (Of course, Tarski's explication is compatible with deflationary and more substantive approaches to truth.) What is notable about the uptake of Tarski's explication within philosophical circles (but perhaps not eventually in logical circles) is that it left little room left for alternative explications of truth. I have reported how hard it has become, in practice, to understand and discern genuinely alternative approaches to truth. (For example, recall that in the history of philosophy we can find non-classical approaches to truth as (i) a property/quality of an object or (ii) truth as identity of an object with (well) itself.)

Now, Tarski could presuppose a high epistemic status for formal languages (including arithmetic) such that proofs in the metalanguage about the object language can be accorded (fairly or not) the honorific, 'knowledge.' Nothing analogous is in play in Stanley's theory of ideology (or propaganda). So, in Tarski the apparent unwillingness to offer "justifications of the propositions expressed by axioms of the theory to someone who doubts them" is un-problematic (when it comes to the epistemic status of the theory). But Stanley's theory of propaganda presents the phenomenon of ideology (and propaganda) as well as a theory about it. But both are contestable in ways that Tarski's are not. As Stanley notes "the facts are under dispute." (75)

So, while Stanley is undoubtedly right that in such contested and contestable matters "no neutral stance" (77) is possible, his methodological appeal to Tarski is unpersuasive. Some other time, I return to this methodological issue because the problems with it are also revealed in the theory of (social) concepts that Stanley relies on in later chapters (for relevant musings prompted by another work see here).

In this post, I have emphasized the ways in which Stanley's approach to ideology and propaganda exhibits commitment to a particular ideal form of rationality, one that values integrity or coherence between ideals and practice. But there are other more substantive forms of rationality, some of which emphasize contextual judgment, or acting from reasons, or self-legislation, etc. Some of these ideals fit well with the coherence ideal of rationality, others less so. The point is that Stanley assumes a kind of political consensus over his preferred ideal form of rationality as well as (as I noted above) the noble ideals of "fair deliberation and equal participation." Given the significant work these ideals do in his work, the lack of effort to provide substantive justification strikes me as  a reason to deny that the theory that relies on them produces "knowledge."

As an aside, I also worry that Stanley's methodological stance will reinforce the already regrettable tendency in contemporary analytical philosophy to avoid responding to foundational objections. We are supposed to spot others their foundational commitments and our criticisms are supposed to be immanent (or are otherwise thought question begging). But while this is an efficient practice, it also generates considerable moral problems -- as I learned from Amia Srinivasian and Daniela Dover (and undoubtedly I am not doing justice to their subtlety) -- when some groups can simply stipulate their commitments to an intellectual or political community while others not.

The concern I have is that when our understanding of propaganda and ideology is loaded with these ideals, we are likely to miss the occasions when precisely the firm adherence to these ideals exhibit noxious forms of ideology or propaganda. That's not a refutation of Stanley's approach (it can't be), but it suggests we should use it with caution, and we should keep exploring conceptualizations of propaganda and ideology that fit with alternative or different mix of ideals.

Comments

[Jason Stanley]

Dear Eric,

Thank you so much for these posts, from which I am learning a terrific amount. They are meaty and substantive challenges that came at a crazy time in the semester, and for some reason I have now re-discovered them. I had best reply to some before the semester begins. A general point that I want to make is that you seem to be an internalist about justification and knowledge. I am externalist. I think substantive non-circulation justifications end somewhere, and one place they can end is at the principles of rationality. One can indirectly target various principles of rationality; one can show that with the use of an alternative semantics, e.g. supervaluations, one can solve certain problems that are hard to solve without them (e.g., some have argued, vagueness). Or one can show that the principles of rationality fit into a successful and useful social or political theory. And if my theory of propaganda and ideology is successful and useful, then it provides some element of justification for the principles of coherence between ideal and practice that it presupposes. But that's only indirect justification.

I definitely agree that I haven't provided substantive justification that would produce knowledge by internalist standards for the kind of rational structure presupposed by the theory, which is that there ought to be some sort of coherence between ideal and practice. But my failure to provide a substantive explicit justification doesn't matter as much for me, because knowledge cannot be held hostage to this criterion, or we almost never have knowledge.

Be that as it may, I argue that Tarski is in the exact same predicament. He presupposes excluded middle in the meta-theory in order to derive bivalence. So my theory of propaganda faces a problem no more or no less severe than that. Now, there is a complication here, and I think it's the one you raise, but I'm not sure. Tarski's theory of truth is supposed to be a mathematical definition of truth; so Tarski can be taken just to be engaged in a definitional project. But then it's not a project that can justify principles of rationality stated in a given language, since it is an empirical fact that the language means what it does. Hence Putnam's charge that "it fails as badly as it is possible for a theory to fail" in "A Comparison of Something with Something Else". But we can and should avoid this.

Let's take a Davidsonian absolute truth predicate, and consider the kind of justification a semantic theory that employs such a predicate would yield for a formal theory of logic, or arithmetic (in my old paper "Truth and Metatheory in Frege", I argue that this is a good characterization of Frege's metatheory in Part I of the Grundgesetze; I still think this is correct). Such a theory does not provide a substantive justification of logical laws. It presupposes them. Similarly, I do not provide a substantive justification of the contradiction structure between ideal and practice, but presuppose it. The theory can still produce knowledge, if the presupposition satisfies some externalist condition, perhaps safety. And if knowledge requires more than that, an available and explicit "substantive justification", I'm skeptical we have much of it (Roger White's "You Just Believe it Because" is good on this). That's not to say that one cannot provide another structure, or challenge the logic of the relation between ideal and practice I am presupposing. Of course one can challenge it. I take this book to be somewhere in the middle of inquiry, with lots of people before me and lots of people after me. I expect to be wrong. I see that one can argue that dogmatic adherence to an ideal can lead to injustice just as much as undermining it could. But I deny that I have the responsibility of defeating every challenge to a presupposition of the theory of propaganda before the theory of propaganda can produce knowledge.

I think the challenges have to come in the form of persuasive examples. I will say this tersely because it's late. Some examples are going to be able to be cases where one explicitly mentions one ideal (say objective rationality), using it dogmatically to critique others. But in many such cases, one is also making as if such a stance is *reasonable*, that it's the way to take every perspective into account. It isn't, it undermines reasonableness. That's the kind of explanation I would gesture at for some of the cases I think you are considering in the remark about "firm adherence" to an ideal.

That said, I reject any conception of ideals that demands "firm adherence" (I think). My discussion of what our relation to ideals should be is on pp. 171-7, at the end of chapter 4. I reject any model that says we should have *faith* in ideals (isn't this firm adherence?) because it will lead us to "overlook their violations". So I agree with you here. And since I think that being guided by a norm of objectivity requires "systematic openness to the possibility that one is swayed by bias", I fully agree that we "should keep exploring conceptualizations of propaganda and ideology that fit with alternative or different mix of ideals." This is consistent with my theory of propaganda being knowledge-producing.

[James Dennis]

Just a few speculative and tentative observations about definitional matters (since at the moment I happen to be concerned with the phenomenon called "propaganda").

The term 'ideology', like the term 'theory', refers to a conceptual structure (a logical structure of categories and rules of application to reality) that is used to interpret experience of the world and allow us to understand it; but an ideology, when it comes to critique of this structure, is subject to a lesser standard of evaluation (from the point of view of the "metatheory"), one belonging to the practical sphere of the political, governed by opinion (as you pointed out with relation to Arendt and Madison), effective persuasion and the interplay of the power- seeking of political interests; a lesser standard as compared to the ideal principles of truth- seeking (rationality) that govern (ideally) the critique of scientific theories in the sphere of the construction of scientific knowledge. I take it that we are engaged in truth-seeking, and that we are looking at political ideologies as a problem and an object to be understood out there in the world. We can talk about the differences between the principles that govern the critique of ideology (as a praxis) and the principles that govern the critique of scientific theories.

Obviously I haven't yet read Jason's book, but to me the key property of what is called "propaganda" is that it is put out to the world without the expectation that it will be subjected to critical scrutiny, but is rather to be accepted without critique. The acceptance of messages that are problematic is achieved by coercion, by pressure of power. On the other hand, a claim (e.g., a knowledge claim) in normal discourse (just normal, let alone scientific) is put out there with the expectation that it will be subjected to critique (e.g., wrt its truth). A problematic assertion is only accepted after a dialogic process of rational argument. So it would be possible to have a well-meaning and sincere propaganda; it might be interesting to compare such cases with religious dogma.

I agree with your observation that the uptake of Tarski's explication of "truth" has made it hard in practice "to understand and discern genuinely alternative approaches to truth". Even Putnam, in the article Jason mentions, says he has been "dissatisfied for many years with almost everything written about the notion of truth", in particular with Tarski's work. I would guess that Putnam's problem with Tarski's approach is in part that Tarski's "metalanguage", and the formal mode of language use in general, lacks the ability to refer to intentional objects in the manner of natural languages. It may be that it would be possible to critique the axioms of the formal objectlanguage only if the metalanguage is a natural language that can genuinely refer to those axioms and definitions and describe their effective operation. The principles of rationality implicit in the "U-language", to use Haskell Curry's term, are another matter, and a possible object for exploration. This is a huge area of problems, however, and this is all I have time to say right now.