The Linguistic Turn in the Early Frankfurt School: Horkheimer and Adorno

Abstract

Was there a linguistic turn in Frankfurt School Critical Theory before Habermas's communications-theoretic one? Might later Wittgenstein and the early Frankfurt School have adopted similar pictures of language? I propose that both questions should be answered affirmatively, focusing on Horkheimer's Eclipse of Reason. I argue that, thanks to the picture of language that Horkheimer and Adorno share with (later) Wittgenstein, we can reconstruct their theory in a way that renders it more defensible. Insofar as the human life form and language are inseparable, language can be an inextinguishable reservoir of what Horkheimer called "objective reason." Recognizing this allows us to answer Habermas's critique of Horkheimer and Adorno. Moreover, paying attention to this inseparableness can enable us to engage in disclosing social critique (for example, regarding current debates about sustainability).

Introduction

A common complaint against early Frankfurt School Critical Theory—especially Adorno and Horkheimer's version of it—is that it is self-undermining: by engaging in a (purportedly) totalizing critique of modern reason, it makes itself impossible. This complaint is particularly known from the second-generation Frankfurt School theorist Habermas, who speculates that the (purportedly) totalizing nature of the critique of modern reason was the result of the particular historical hour in which Adorno and Horkheimer wrote Dialectic of Enlightenment. Habermas states, "It becomes intelligible how the impression could indeed get established in the darkest years of the Second World War that the last sparks of reason were being extinguished from this reality and had left the ruins of a civilization in collapse without any hope."¹

Yet, however comprehensible it might be as an expression of that dark historical time, it is, nonetheless, unworkable as Critical Theory—or so Habermas claims. Specifically, he claims that only by maintaining intact a rational criterion of critique and thereby a conception of reason separate from instrumental reason, can the dominance of instrumental reason in modern society be criticized and counteracted. Without such an alternative conception of reason, Adorno and Horkheimer—according to Habermas—are stuck in a "performative contradiction."² The content of the speech act of saying that everything is ideological (in Marx's sense of false consciousness) stands in contradiction with Horkheimer and Adorno's (or, indeed, with any ideology critic's) performance of that very speech act. Habermas's own work develops from this criticism into the suggestion of communicative rationality as the alternative conception of reason with which to criticize instrumental rationality and the colonialization of the lifeworld by the system (notably in his Theory of Communicative Action).³ By taking a linguistic turn—or more specifically a "communications-theoretic turn"⁴—Habermas seeks to avoid the aporetic dead end into which he thinks Adorno and Horkheimer maneuvered Frankfurt School Critical Theory.

There are reasons for thinking that, in defending Adorno and Horkheimer against this common complaint, one ought to query the very way that Habermas presents the problem that purportedly haunts their theories. But, in this paper, I mainly take a different approach: I take on Habermas's critique on its own terms, answering how (continuing) the Critical Theory of Horkheimer and Adorno is possible, after all.⁵

Crucially, I argue (a) that there already was a kind of linguistic turn in Horkheimer and Adorno's work prior to and different from Habermas's own (later) one; and (b) that uncovering this hitherto overlooked linguistic turn enables us to see why (noninstrumental) reason is, in a sense, inextinguishable for Horkheimer and Adorno. Part of my overall strategy is to draw parallels to the thought of (later) Wittgenstein.⁶ In this, I oppose the contrast drawn by Wellmer between Habermas, who, Wellmer thinks, has integrated the insights of (later) Wittgenstein into his theory successfully, and Adorno, who, according to Wellmer, has not. I suggest (c) that Horkheimer and Adorno's version of the linguistic turn is actually more in line with (later) Wittgenstein's central insights than the version that we can find in Habermas (and Wellmer).

Before embarking on a defense of these three claims, let me note two preliminaries. First, let me comment on what I mean by 'linguistic turn.' This term is widely used in contemporary philosophy, but there is a danger that different authors mean different things by it. I propose that this turn includes accepting two basic tenets:

1. A. All philosophy has to be (also) philosophy of language. There is no standpoint outside of language to grasp the world, and philosophical reflection is no different. It is mediated by language, and one of its central tasks is to reflect on that mediation and the nature of language from within that medium.

2. B. There is no private language. Intersubjectivity is unavoidably inscribed into language and thereby philosophy (both of language and more generally).

As should be obvious, the linguistic turn, so conceived, owes a lot to the philosophy of (later)⁷ Wittgenstein, although others (like, I suggest, Horkheimer and Adorno) might have come to the same conclusions by other routes and expressed them differently. There can also be various versions of what I just described as the linguistic turn. For example, Habermas has a specific version of it that includes, in addition, transcendentalism (or quasi-transcendentalism) and universal pragmatics, but other philosophical theories that accept a version of the linguistic turn may not include these further commitments, or they may even explicitly reject them (Horkheimer and Adorno's version is, as we see below, one example of the latter view).

As a second preliminary, I note that, in this paper, I ignore any differences between the theories of Adorno and Horkheimer that may exist on this or indeed other matters. I treat their works as a common corpus, presenting what I take to be their shared position, whether expressed in coauthored or individually authored pieces. Relatedly, I should also note that what I am presenting as their position is the result of sometimes quite considerable reconstruction. I do not claim that the position was ever explicitly stated as such by them or even that it was clearly before their minds. My claim is that their theory contains important resources that—once reconstructed—allow us to answer Habermas on his own terms.

In section 1 of this paper, I begin by attending to a neglected or maligned text, Horkheimer's 1947 book Eclipse of Reason. While this book is problematic in certain ways, it also contains overlooked resources. I then (section 2) briefly introduce some key features of (later) Wittgenstein's picture of language. I use them to bring out that a similar picture of language is operative in Horkheimer and Adorno's works. I suggest that recognizing this allows us to answer Habermas's objection on his own terms. I then show (section 3) that the thus reconstructed linguistic turn in Horkheimer and Adorno's works is, however, different from Habermas's communications-theoretic version of this turn and suggest that it is superior in realizing the central insight of (later) Wittgenstein's approach. Answering Habermas on his own terms must operate at a high level of abstraction, but I end (section 4) with a more concrete example of how the reconstructed linguistic turn in Horkheimer and Adorno can be put to work in ideology critique, in relation to the idea of sustainability.

1. eclipse of reason—problematic but promising

Starting with Eclipse of Reason is helpful not only because it may be the source of (what I argue is) a key misunderstanding by Habermas and others, but also because this book contains important leads for reconstructing an answer to the common complaint made against Horkheimer and Adorno, and, specifically, for the linguistic turn operative in their theory. Let me expand on the first claim just made, and then develop the second.

To begin, I want to suggest that Habermas and others misunderstand the metaphor employed in the title of the book. To speak of 'eclipse' is not to claim that (objective) reason has ceased to exist, that "the last sparks of reason were being extinguished." Rather, it is to claim that something is—partially or completely—blocking our view of (noninstrumental) reason. Just as with a lunar or solar eclipse, what is blocked from view is nonetheless present throughout, so that what is required is "only" that it reemerge into view.

To be fair, Horkheimer's Eclipse of Reason evolved from a 1941 essay called (in the published English version) "The End of Reason" (as in "reason's coming to an end," not as in "the telos of reason"). Some of the more pessimistic formulations in that essay might suggest the complete passing away of reason (or the relevant aspect or conception of it).⁸

Still, the German title of the 1941 essay was "Vernunft und Selbsterhaltung" (Reason and self-preservation), which does not suggest anything as final as "The End of Reason."⁹ Moreover, anyone who has looked beyond the titles should realize that the story is more complicated: Horkheimer operates not with a unified conception of modern reason as through and through instrumental, but rather with a dual (one is tempted to say "dialectical") conception of reason (and the same holds for Adorno). Specifically, Horkheimer distinguishes between subjective and objective reason. While subjective reason has its place and should not be abandoned, its dominance—its "eclipsing" of objective reason—is problematic. Critical Theory should, according to Horkheimer, counteract this, which requires putting an emphasis on objective reason, so as to help it to reemerge into view (that is, to become more socially operative again).

There is much to be said about the conception of, and distinction between, subjective and objective reason. One of the problems of Eclipse of Reason as a book is that these very matters are far from clear and worked out in it. Still, it would detract too much from the main line of argument to offer here a reconstruction of them. Instead, for this paper I adopt a simplifying solution, involving two steps.

First, let me offer a rough characterization of the distinction between subjective and objective reason. Horkheimer characterizes subjective reason as a property of individual subjects—specifically, a set of capacities that subjects (can) have: means-end reasoning, following logical laws, classifying and distinguishing (Eclipse of Reason, 1).¹⁰ In other words, it is subjective in the sense of referring to the subject's reason (its reasoning abilities). In contrast, Horkheimer understands objective reason as a property of the world (or, at least, the social world).¹¹ It is not formal and dependent on the ends the subjects happen to have, but substantive and object-dependent. 'Object' here is understood more broadly than simply medium-sized material objects like tables and chairs. Indeed, Horkheimer, importantly, makes use of the language of 'situations' to explicate the specific kind of normativity at stake in objective reason: seeing a drowning child or animal requires of those who pass by (and can swim) that they rescue it;¹² someone's being ill requires that we treat their illness as best as we can; something beautiful requires that we appreciate it; and so on.¹³ Each of these situations "speaks a language of itself"; there is "a silent appeal by the situation itself" (7, 21).

Second, whatever Horkheimer and Adorno mean by 'objective reason' exactly, it is that which Habermas claims they think was extinguished—it is that to which they purportedly cannot appeal on pain of a performative contradiction. Specifically, the thought is that subjective reason—often dubbed simply 'instrumental reason'—has become socially dominant in modern society, and thereby eclipsed objective reason (as Horkheimer puts it) or even extinguished it (as Habermas reads him). Important in this is the Hegelian idea that for something to have normative force for us, it has to be socially operative in our practices. In premodern times, objective reason was socially operative in the sense that society then was shaped by religious or metaphysical conceptions, according to which reason was seen as a property of the world. In their awe for the gods people then might have misconstrued objective reason in various ways, but at least their worldviews left room for the idea that reason has objective reality. The issue now is whether this idea has been eclipsed or even extinguished in our disenchanted modern societies.

With this in mind, we can formulate the crucial questions to which readers of Eclipse of Reason want to know the answers: What enables Horkheimer (and Adorno) to criticize the reigning form of reason as unreasonable? Specifically, what enables them to do this, given that, at the same time, (a) they advocate that the alternative form of reason (objective reason) has been eclipsed in a way that seems to suggest that it is no longer socially operative; and (b) they are Hegelians in thinking that only socially operative forms of reason can be mobilized as critical resources (for otherwise we end up with empty formalism and/or wishful thinking)?

Unfortunately, Horkheimer's text is not sufficiently clear as to how he thinks we can answer these questions. It contains a number of seemingly different strategies, none of which is well worked out. I do not review here the different strategies contained in the text, but instead concentrate on the one that is the most promising.

Specifically, I want to draw attention to two passages that appear incredibly puzzling at first but are rich and promising for seeing how residual elements of objective reason are still available to us (and, in a sense, cannot but be so available). The first passage gives us a hint of what philosophy must do to counteract the eclipse of (objective) reason. Strikingly, Horkheimer directs us to language:

Philosophy must become more sensitive to the muted testimonies of language and plumb the layers of experience preserved in it. Each language carries a meaning embodying the thought forms and belief patterns rooted in the evolution of the people who speak it. It is the repository of the variegated perspectives of prince and pauper, poet and peasant.

(Eclipse of Reason, 117–18; see also 126–27; and "Zum Begriff der Vernunft," 30)

There is no explicit mention of 'objective reason' here, but I want to suggest that, for Horkheimer (and Adorno), language is "the repository" not merely of "variegated perspectives" but of objective reason—and that objective reason thereby remains socially operative even in the age of its eclipse. How this can be so, indeed is so, will only emerge in the next section.

For now, let me continue by exploring the puzzlement that the above passage raises. It appears puzzling as to how turning to language can help us in counteracting the eclipse of objective reason, even if language were "the repository of variegated perspectives." Does Horkheimer mean to suggest that Critical Theory should, so to speak, excavate the idea that there is reason in the world—that it is objective—from the layers of experiences preserved in language of past times where people still believed in this? Other comments by Horkheimer in Eclipse of Reason—such as that language reflects "the longings of the oppressed and the plight of nature" (127)—also seem to suggest a backward orientation toward what has become sedimented, deposited, in language. But perhaps the suggestion is not just about excavating the past, but about attending to something ongoing: insofar as reason has objective reality, such as situations giving rise to demands on us, it is possible that this will register in layers of experience (not necessarily conscious ones), and this will then be reflected in the language we use (not least because the experience we have, and the thought forms and belief patterns we adhere to, will be linguistically mediated).

Still, the puzzlement continues: this does not seem promising as a strategy for Critical Theory insofar as Horkheimer does not believe in a golden past in which people actually had a clear view on objective reason.¹⁴ Moreover, Horkheimer's—and Horkheimer and Adorno's—statements from around the same time suggest a different, much bleaker view of language that seems to fit better with the (purportedly) totalizing critique of modern reason and society that Habermas ascribes to them. Eclipse of Reason begins, after all, with an account of how hopeless it is to ask "the ordinary man" to explain what reason is (1; see also 118)—hopeless for Critical Theory because "the ordinary man" would just reproduce in his answer what is socially dominant, that is, according to Horkheimer, formalized subjective reason (not objective reason). Horkheimer elsewhere claims that language is becoming "a mere tool in the omnipresent production apparatus of modern society" ("Zum Begriff der Vernunft," 30, my translation). And there are similar statements in Dialectic of Enlightenment, coauthored with Adorno, painting a picture of language as an ideological tool of domination, rather than as a resource for social critique (xv, 133–36).

Horkheimer says relatively little about language in Eclipse of Reason, but one thing that he does say might provide a clue for overcoming the puzzlement and clearing the way for the promising approach that I claim to be present in this book (and his and Adorno's theory more generally). What Horkheimer does say about language appears in the second crucial but puzzling passage to which I want to draw attention. He writes, "Philosophy helps man to allay his fears by helping language to fulfil its genuine mimetic function, its mission of mirroring the natural tendencies" (127).

This is a curious statement. What could Horkheimer mean when speaking of language's "mimetic function" and of "its mission of mirroring the natural tendencies?" It is customary to refer to Benjamin's musings about language to decode such statements in Horkheimer (and Adorno).¹⁵ But I propose to take a different path, drawing a comparison to the work of (later) Wittgenstein instead.¹⁶

2. a little help from (later) wittgenstein: the inseparability of language and life form

Before presenting a brief sketch of the picture of language that (later) Wittgenstein endorses, let me make the following preliminary remark. I am not saying that Horkheimer and Adorno's view of language was influenced by (later) Wittgenstein. For one thing, that would get the chronology wrong (some of the key insights by later Wittgenstein were, for example, not yet sufficiently publicly available when Eclipse of Reason was published). For another, there is no evidence (that I know of) that either Horkheimer or Adorno was familiar with the work of later Wittgenstein when it became available. It is unclear how much they engaged even with his earlier work. There is mention of the Tractatus in some of their writings, but the way that it is discussed suggests that Horkheimer and, explicitly, Adorno are nothing but critical of it, seeing it as a prime example of positivism and, thus, of a view that they deeply reject. (This also suggests that whatever influence Gilbert Ryle had on Adorno, it did not lead to an explicit repositioning vis-à-vis Wittgenstein's philosophy.) Similarly, I am not aware of any engagement of Wittgenstein with the writings of the Frankfurt School. In sum, what I am claiming is not that there was any positive influence between (later) Wittgenstein and Adorno and Horkheimer, in whichever direction. Instead, the parallel that I go on to suggest is one of similar pictures of language that, for all we know, they developed independently of each other.

What is the picture of language that (later) Wittgenstein endorses and seeks to disclose to us? There is, I take it, no answer to this question that will be entirely uncontroversial, but I submit that the following features will be widely recognized as key. For (later) Wittgenstein, "to imagine a language means to imagine a form of life" (Philosophical Investigations, §19). As interpreted by Garver, this is to say that meaning cannot be captured simply "in terms of words and syntax"; instead, we need to look to how we use language, how it is "integrated into the activities of some sort of living being."¹⁷ This can, among other things, be seen from the way that we would have to proceed in order to learn an unknown language. To learn such a language, Wittgenstein tells us, one would have to ask questions like the following: "In what circumstances would you say that the people gave orders, understood them, obeyed them, rebelled against them, and so on?" (§206). Crucially, one would have to turn to the "common behaviour of mankind" as one's "system of reference" (§206). Language exists and functions in the particular ways that it does because of this commonality in behavior, where 'behavior' is broadly understood to include capacities. For example, human beings are typically capable of learning basic algebra, but they typically do not have perfect pitch; and this is reflected in the language(s) that they operate with. The common behavior that we can use as a system of reference is not something that we know by way of scientific methodologies, but concerns certainties (as different from scientific knowledge) about our form of life; that is, it concerns that which "lies beyond being justified or unjustified; as it were, something animal" (Wittgenstein, On Certainty, §359). In good part this is—as another commentator puts it—because "there is no such thing as a standpoint outside language from which to characterize the relationship between language and the world."¹⁸ Indeed, there could not be such a standpoint outside of language (hence, the first basic tenet included in accepting the linguistic turn, as per the introduction above). Offering justification is part of what we use language for, and perhaps there are even specific justification language games, but these uses and language games presuppose the human life form and cannot ground it. Moreover, it is also not the case that the human life form—or the "world" more generally—does or could ground language. As Winch puts it, "Reality is not what gives language sense. What is real and what is unreal shows itself in the sense language has."¹⁹

What I want to suggest is that a similar picture of language is at work in Horkheimer's (and Adorno's) theory. Here is the first piece of evidence for this: intriguingly, we can find in Horkheimer the very thought that the human life form is something like a system of reference with which we always already make sense of specific formulations, concepts, or theories (including scientific ones), but which only shows itself in the sense that language has, rather than being something that we can explicitly speak about and delineate. In a neglected text from 1946 ("Vertrauen auf Geschichte" [Trust in history]), Horkheimer writes that what it is to be human "cannot be captured by any concept," but "Unnamed, it is at the basis of language and its concepts" (127, my translation).

I introduce more evidence for my ascription below (including passages from Adorno's sole-authored writings), but let me here already point to the upshot of operating with this picture of language for the pressing question at hand: in a nutshell, the fact that language and the human life form are inseparable makes it possible for language to be a repository of not just past ideas about objective reason, but still now socially operative ones, even despite the fact that language can be, and has been, pressed into being an ideological tool of domination. For in virtue of being inseparable from the human life form, language contains traces of this life form, and thus, of objective reason. Language is not just arbitrary (although it contains arbitrary elements). Indeed, it is not even merely conventional (although, again, it contains elements arising from conventions). As a result of this inseparability, the traces of objective reason inherent in our human life form cannot be eradicated completely from language—even the most ideological conventions cannot, ultimately, change the fact that language harbors opposition to what runs counter to the human life form.²⁰

One might be surprised to hear that this is supposed to be Horkheimer and Adorno's view, but I think that there is further evidence for thinking that it is. One key piece of evidence takes the form of two statements, one from the 1940s (prior to Eclipse of Reason) and one from the 1950s (after Eclipse of Reason). I consider these in achronological order. In 1956, Adorno and Horkheimer discussed the possibility of writing and publishing a new, updated communist manifesto. As part of this discussion, Adorno's wife, Gretel, recorded Horkheimer as saying, "Whatever is right about human society is embedded in the language—the idea that all will be well. When you open your mouth to speak, you always say that too" (Adorno and Horkheimer, Towards a New Manifesto, 5). Earlier, in a 1941 letter to Adorno, Horkheimer writes even more suggestively,

Language intends, quite independently of the psychological intention of the speaker, the universality that has been ascribed to reason alone. Interpreting this universality necessarily leads to the idea of a correct society. When it serves the status quo, language must therefore find that it consistently contradicts itself, and this is evident from individual linguistic structures themselves.²¹

This is a rich passage that deserves detailed analysis. But for our purposes here, three points must suffice. First, it is noteworthy that—like Wittgenstein—Horkheimer adopts the view that (to use a by now common phrase from Putnam) "meaning just ain't in the head" of speakers (notably, it is not reducible to their psychological intentions). This fits well with the second basic tenet of the linguistic turn that I outlined in the introduction above.²²

Second, one might think that the second passage above sounds rather proto-Habermasian. Talk of universality and how it "necessarily leads to the idea of a correct society" seems to have the ring of Habermas's discourse ethics avant la lettre—as does the suggestion (in the first quotation) that we cannot but imply some normative standard whenever we speak ("open our mouth"). But, I think, neither quotation needs to be read in a Habermasian way—and, indeed, they are not best read in that way. In the next section, I bring out the difference between Habermas's communications-theoretic version of the linguistic turn, on the one hand, and the version of it that I ascribe to Horkheimer and Adorno, on the other. For now, it suffices to note a key feature in the second quotation: whatever Horkheimer means by universality, the implication of his statement clearly is that there is a mistake to ascribe it to "reason alone." On my reading, what Horkheimer is hinting at here is that we should not primarily locate the idea of a correct society in something like the subject's rational capacities (as in Kant) or intersubjectivity (specifically communicative rationality, as later in Habermas). Instead, there is a connection, indeed inseparability, between language and objective reason—reason as manifested in our life form and in the world—and it is thanks to this inseparability that (to return to the first quotation) "whatever is right about human society is embedded in the language," such that when one opens one's mouth, one commits oneself to this idea.

Third, the final sentence of the second passage is particularly important in the context of thinking about the objection that I put to myself in the previous section—namely, the objection that language is not the repository of objective reason that can be mobilized by Critical Theory, but instead is the ideological tool of domination. The final sentence of the second passage suggests that for Horkheimer (and, by implication, Adorno) language cannot be purely ideological. Indeed, its ideological use involves a linguistic mistake, contradicting itself in a way that "is evident from individual linguistic structures themselves."

At this point, I should add a qualification. The way that I read this passage—in line with how Horkheimer and Adorno's work as a whole can be read²³—is to understand it as suggesting a research program, and its ultimate vindication would require not just the abstract general considerations about possibility advanced here, but the successful completion of this research program in relation to actual concrete examples of ideological use of language. In particular, the thesis that such use contradicts itself, just as a matter of grammar (in the Wittgensteinian sense), would have to be carefully demonstrated by actual studies of such use. In section 4, I indicate the beginning of such a study in relation to debates about sustainability. A fuller study would have to engage in detail with material like speeches or newspaper columns (as Adorno did in relation to fascist propaganda and popular culture). Also, it would be necessary—perhaps partly by drawing again on Wittgensteinian insights—to investigate the different kinds of linguistic mistake that are involved in the ideological use of language.²⁴ Importantly, in many cases the problem is not that the ideological use of language consists simply of empty statements or gibberish. Instead, the problem can be—to take an example that will be relevant to the discussion of sustainability later—that ideology uses a secondary sense of a word (or secondary use of a concept) as if it were primary or on par with the primary one. It is, for example, not empty or gibberish to speak of Tuesdays as "lean" and Wednesdays as "fat" or to ascribe pain to inanimate objects or to pity dolls; but we go wrong linguistically if we position these uses as primary ones, on which the other uses (say ascribing pain to a living being) are dependent (see Wittgenstein, Philosophical Investigations, part 2, 184, and §282). To position them in this way is one variety of separating meaning and use, and one variety of forgetting how we acquire language—one that is perhaps particularly relevant for ideological contexts (I return to this below).

For now, it is important that, insofar as Habermas and others are even doubting the very possibility of having any resources for Critical Theory, if Horkheimer and Adorno's theory were true, the qualification entered in the previous paragraph is not detrimental to the argumentative aim of this paper. On its own terms, the common Habermasian complaint against their theory would be answered, if we can account for this possibility at a general, abstract level.

Let me be clear about what even a full vindication in relation to concrete examples would not establish. Here, it is helpful to consider something Garver says about his Wittgensteinian perspective: "human activities involve norms" reflected in language, some of which are arbitrary, but some of which are "tightly woven into the fabric of our form of life"; and a perspective that recognizes this "therefore contains the seeds of normativity and of a certain sort of transcendence of the merely empirical and merely factual."²⁵ It is crucial that Garver speaks of "seeds of normativity" (n.b. Garver uses the plural; there is no implied monism here). What he calls such seeds, we might call "traces of objective reason" in the context of Horkheimer and Adorno's work. However called, the key point is that the appeal to language (use) does not yield premises that we can deploy (say in normative ethics or political philosophy) to derive conclusions about what ought to be done (say in relation to suicide or the legalization of same-sex marriage). What is generated is just seeds, not—to stay with the metaphor—full-grown trees with which to build an ethical or political system. (Indeed, even the metaphor of seeds is still too restrictive, as normally seeds can only become one kind of tree, not different kinds of trees or things other than trees. Perhaps 'ingredients' or 'resources' is a better metaphor—these can be used and combined in a greater variety of ways than seeds can.)

Translated back to the point about subjective reason's eclipsing of objective reason, we can say that in using subjective reason on its own, we are forgetful of the inseparability of language and our form of life—in Wittgensteinian terms, we are separating meaning and use. Keeping the inseparability in view provides us with resources for Critical Theory, but we have to be careful not to overstate what kind of resources this yields. We get access to traces of objective reason, not fullblown normative premises. It is not that we get access to specific content, such that we could just deduce from this inseparability—for example—whether or not same-sex marriage is morally permissible and should be legalized.

Furthermore, the claim is not that language provides "normative foundations" in the sense that Habermas and others have demanded of Horkheimer and Adorno, but which the latter rejected as an untenable and problematic justificatory project.²⁶ (As noted above, we are facing bedrock claims here that lie, in Wittgenstein's words quoted earlier, "beyond being justified or unjustified"). The claim is, instead, that mobilizing the inseparability of language and our life form can help uncover ideological distortions and reveal the inhuman social reality for what it is. (I provide a more concrete example of this in section 4.)

In light of this, the puzzling statements that Horkheimer makes in the passages quoted above from Eclipse of Reason become intelligible and compelling. Insofar as he operates with a picture of language as inseparable from the human form of life, we can understand how he would think that its mission is to "mirror the natural tendencies." While it remains unspecified in the text whose natural tendencies they are, we can now make sense of them as those of human beings. In attending to how language is about our agreeing in form of life (Wittgenstein, Philosophical Investigations, §241), we let it fulfill its function. Specifically, as inseparable from the human life form, language contains traces of "the structure of the reality" (Horkheimer, Eclipse of Reason, 6) of this life form, and thus, of (at least part of) objective reason. Insofar as our current social arrangement goes against the grain of the human life form, the traces of this life form in language, and thereby language itself, do reflect what the oppressed long for, but are denied by this social arrangement. This denial might be buried by ideological distortions, but by becoming "more sensitive to the muted testimonies of language and plumb[ing] the layers of experience preserved in it," we can uncover these distortions, and, upon eliminating them, let language fulfill "its mission of mirroring the natural tendencies" (117, 127).

Indeed, it also becomes clear why objective reason can be eclipsed, but never extinguished—can never stop being socially operative—as long as there are human beings.²⁷ Language can include arbitrary elements, but it cannot become completely unmoored from the human life form—for then it would not be language anymore—and, hence, it is—and cannot but be—a repository of objective reason. Just as our natural impulses can be repressed, but not eliminated,²⁸ objective reason embodied in the human life form cannot be eliminated from language, but only covered over; and attentiveness to language itself can help us to uncover it. This means that Habermas's objection fails: Horkheimer and Adorno's critique of modern reason is not totalizing, but leaves intact and draws on objective reason, which, although—according to their diagnosis—it is eclipsed in our modern social world, is not extinguished, indeed not extinguishable, but remains socially operative.

One complexity remains. In arguing that Habermas's objection fails, I am not denying that some of what Horkheimer and Adorno are expressing has an air of paradox, albeit perhaps not quite involving a performative contradiction.²⁹ My point so far has, rather, been that their critique is not totalizing. What I want to add here is that with this assurance in place, we might take a different approach in thinking about performative contradictions, aporia, and the like. For Habermas, a performative contradiction or aporia is the end of the road—upon finding it, we need to back up and adopt a different view. That seems to me too hasty a conclusion. Take an everyday example from the context of personal relationships: one person may well say, as part of yet another argument with her partner, "I can't speak to you!" This looks like a performative contradiction: the propositional content of her utterance conflicts with the act of uttering it. Yet, we can recognize, I submit, that the partner would go wrong if his response was to point this out and to argue that the speaker, hence, cannot mean what she said, but must mean something else. We can also recognize, I submit further, that the original utterance, if it works (where this might include that the partner reacts differently than just imagined and there is a shift in perspective and behavior), works because it involves a performative contradiction. In certain contexts (perhaps particularly desperate ones), speech acts of this sort might have the best chance of bringing about the change in perspective, approach, and behavior that a seemingly intractable situation requires to be resolved. Without equating making utterances in the context of personal relationships with writing books like Dialectic of Enlightenment, I think we should not stop at noticing (what look to us like) performative contradictions but, in addition, consider what role they might play in expressing a predicament and how indispensable they might be for shifting the listeners or readers—us—out of a certain perspective, approach, or behavior. Both this and the parallel to Wittgenstein's writings would require more detailed discussion than I can devote to here. I simply want to flag that my defense of Horkheimer and Adorno leaves open these difficult questions of—for want of a better word—style. A fuller understanding of their writings—and those of Wittgenstein—would have to grapple with how these works are written and why they are written in the way(s) they are. If one believes—as I would suggest Horkheimer and Adorno as well as (later) Wittgenstein believe—that language is both a repository of objective reason and that a tendency has unfolded such that much of our public discourses are ideological or inane (or both), then it takes special expressive means to speak about the latter and to try to implode the bewitchment from the inside.

3. two linguistic turns: communic ative action and objective reason

I have shown that we can legitimately ascribe to Horkheimer and Adorno that they partook in the linguistic turn in philosophy, and that this ascription allows us to respond to the common Habermasian complaint made against their theory: theirs is not a totalizing critique of modern reason. However, this does not mean that they—or Wittgenstein—operate with Habermas's own version of the linguistic turn. Unlike Habermas, Horkheimer and Adorno do not, and need not, appeal to a notion of communicative action, where this action is understood other than (and separable from) strategic action guided by instrumental reason. Similarly, they do not need to demonstrate by way of (quasi-)transcendental argument that such communicative action brings with it normative constraints for all forms of discourse and action. And being able to do without this is important and welcome: it avoids certain problems,³⁰ and it makes sense in the context of rejecting, as briefly mentioned above, the justificatory project that Habermas pursues as untenable and problematic.³¹

The linguistic turn in Horkheimer and Adorno is, thus, fundamentally different from the one envisaged by Habermas in his theory of communicative action. Instead of transcendentalism or quasi-transcendentalism, their move is one about the inseparability of language and the human life form.³² Their Critical Theory is neither grounded in universal pragmatics, nor requires appeal to an ideal speech situation, but explicates our language use(s) from within our historical practices in the light of our natural history.³³

Adorno emphasized the importance of such historical contexts from early on in his career: "Words are not ever merely signs of what is thought under them, but rather history erupts into words, establishing their truth-character. The share of history in the word unfailingly determines the choice of every word because history and truth meet in the word."³⁴ Apart from being markedly different from Habermas's view of truth (and validity more generally), this passage also provides further evidence (for the first time from a text by Adorno) that the picture of language I ascribed to Adorno and Horkheimer is similar to that of the (later) Wittgenstein. In Wittgensteinian terms, what Adorno is saying is that meaning is use, such that decoding meaning requires that we always start in medias res, from within our historical practices of language use.

Another way to see this difference between Habermas's and related positions, is, first, to consider a criticism that Wellmer makes of Adorno by appealing to the insights about language by later Wittgenstein; and, then, to trace how Wellmer misses an important point. Wellmer advances the Habermasian criticism that Adorno remains stuck in the paradigm of a philosophy of the subject. Central to this criticism is the idea that Adorno's view "in terms of linguistic philosophy . . . contains a residue of naivety," such that he remains committed to a position outside of language as a basis for critique, objecting to language use as such, rather than tracing "untruth in" it.³⁵ Adorno, on Wellmer's reading, contrasts a subject with an extralinguistic object to which the use of language fails to do justice. Instead, we should operate "within the horizon of intersubjective linguistic praxis."³⁶ This, according to Wellmer, is the lesson that we need to learn from later Wittgenstein to correct Adorno's "naivety" in matters of "linguistic philosophy."

But it is questionable both (a) whether Adorno is indeed as naive in this matter as Wellmer suggests, and (b) whether Wellmer actually draws the right lesson from the work of later Wittgenstein. Consider these matters in reverse order. Wellmer is sometimes aware that for later Wittgenstein, meaning is irreducible to intersubjective consensus—indeed, Wellmer cites the relevant passages from Philosophical Investigations in which Wittgenstein notes that human beings "agree in the language they use," where "that is not an agreement in opinions but in form of life" (§241, emphasis in original). But, ultimately, Wellmer tends to dissolve the idea that meaning is use into the idea that meaning is all about intersubjectivity (see, notably, his approving quotation of Habermas, who equates the new "paradigm of linguistic philosophy" with "intersubjective understanding"³⁷). Wellmer thereby arguably misses the very point of the philosophy of later Wittgenstein to which he appeals. Yes, meaning is not something that is set by individuals on their own (recall the second basic tenet of the linguistic turn); and yes, we need to be habituated into language use by others and reconstitute meaning in our ongoing intersubjective interactions; but none of this means that validity and truth just come down to what is justifiable to other subjects. Instead, the intersubjective initiation to language discloses a world to us, such that (to put it in McDowell's terms) our eyes are opened to "rational requirements, which are there in any case, whether or not we are responsive to them"³⁸—indeed, we might add, these requirements are there, whether or not we reach a consensus about them. In this way, Adorno (and Horkheimer) can maintain a position whereby objects (and situations) lay claims on us, without this meaning that they are adopting a position outside language (that is, without being guilty of "naivety" in linguistic philosophy).

4. unsustainability as social unreason

There is much more to say about these matters.³⁹ What I do in this section is add further evidence in support of my claim that the kind of position that I ascribed to Horkheimer and Adorno provides an interpretative key for making sense of what they say. Moreover, I provide some more concrete texture to the abstract claims that I have made so far.

In particular, it is striking—and in need of explanation—that there are references to the human life form ("humanity") and its meaning in Horkheimer and Adorno's critique of society as irrational or unreasonable. These can be found already in Horkheimer's seminal 1937 text "Traditional and Critical Theory."⁴⁰ But the passage that is particularly important in the context of this paper comes from a later text by Adorno:

The preservation of humanity is inexorably inscribed within the meaning of rationality: it has its end in a reasonable organization of society, otherwise it would bring its own movement to an authoritarian standstill. Humanity is organized rationally solely to the extent that it preserves its societalized subjects according to their unfettered potentialities.

(Critical Models, 272–73, emphasis added; see also Gesammelte Schriften, 20.1:147–48)

The inseparability of the human life form and language is here expressed in the claim that there is a telos in the very meaning of rationality that links any "reasonable organization of society" to one that preserves human beings, both in the minimal sense of securing their survival and in the more demanding sense of allowing them to live "according to their unfettered potentialities." Societies that do not secure this are pathological, irrational.

Leaving Adorno and Horkheimer aside for a moment, we can specify this idea further. Probably the clearest example of how society itself can (be said to) be ill, irrational, or unreasonable, is the issue of the human-made climate emergency. There are obvious negative implications of our warming planet for the health and well-being of individuals—existing and future ones. But the human-made climate emergency also suggests the thesis that society itself is ill: arguably, the very idea of human society implies the task of sustaining humanity for the future, and, hence, a society that systematically endangers that future is a pathological society.⁴¹ By 'society' we mean something orientated toward perpetuity. (Similarly with 'economy,' where this is clearer from the etymology, which links it, via Latin and ancient Greek, to management of the household [oikonomia] and the idea of stewardship.) If this is right, then a humanity-destroying society (or economy) is in a certain sense a contradiction in terms: it does not fulfill its inbuilt purpose. To claim this is not to deny that a humanity-destroying society (or economy) might well exist as a matter of fact. Indeed, we might well be living in such a society—our capitalist economic system looks like it has us heading for extinction. But this does not take anything away from the fact that a humanity-destroying society (or economy) is problematic—including, arguably, problematic according to the very idea of what a society (or economy) is, which includes, I have suggested, the purpose of sustaining humanity into the future.⁴² And, crucially for the purposes of this paper, it does not take anything away from the fact that this problematic nature shows up in language.

While Adorno could not have known about our climate emergency, it is striking that he does reflect on the idea of society along the lines just suggested. I already quoted one relevant passage earlier in this section, so let us now consider a second passage, in which the meaning of something—this time 'society,' not 'rationality'—plays again a crucial role: "Society . . . 'means': objectively aiming at reproduction of life consonant with the state of its powers. Otherwise, societal arrangement—even societalization itself—in the simplest cognitive sense is absurd."⁴³ The appeal to meaning seems, here, more qualified by the placing of quotation marks around 'means,' but this appeal is still pivotal to the passage (which suggests that the quotation marks are for emphasis—perhaps to remind us of something that we sometimes overlook—not for distancing from what is being said). Insofar as language and the human life form are inseparable for Adorno, it is not surprising that he would display such confidence in appealing to meaning. In virtue of this inseparability, there are traces of objective reason in language, in what we mean.

One might object here that even if, in a sense, sustainability is built into the very meaning of society, it is unclear why this matters. In reply, how debates are framed matters enormously to what can be achieved within these debates and within politics more generally. If we present sustainability as something that we should actively adopt as a new aim, then this puts the burden of proof on its defenders to justify why this is, and indeed, what we should do. In contrast, if we realize that sustainability as an aim is already woven into the fabric of our language—specifically, into the meanings of 'society' and 'economy'—in virtue of its inseparability from the human life form, then the burden of proof shifts. I want to suggest that the former framing—the one that presents matters such that the burden of proof is on sustainability—involves a linguistic mistake, specifically the mistake of presenting a secondary use of the concept of society (when we derivatively speak of it as unsustainable) as if it were on par with the primary one (as if it were nonderivative), thereby generating the linguistic illusion that the issue of sustainability is only externally related to the concept of society. This linguistic mistake is not innocent but ideological (in Marx's sense), as this framing is socially dominant and serves a status quo that is, for all the noise about reducing CO₂ emissions, still dominated by fossil fuel lobbies and a focus on economic growth at almost all costs.

Again, we have to be careful not to overstate what, if anything, follows from exposing such linguistic mistakes. No particular policy will automatically follow as a consequence of approaching the debates about sustainability in a way that reconfigures the burden of proof—to think otherwise would be to (mis)take the "seeds of normativity" that language contains for more than they are. Many of the difficult choices will remain, including about possible trade-offs between the well-being of members of current generations and those of future ones. Similarly, while linguistic mistakes can serve as telltale signs of ideology, exposing them is no substitute for a full ideology critique, which would have to include also a social analysis of the concrete ways that powerful groups use their influence to keep the problematic framing in place and of the complex dynamics involved in why they do so.

Still, such an approach will not simply leave things as they are either.⁴⁴ Instead, it reveals a problematic framing of existing debates about sustainability, which provides a counter to certain ideological distortions, notably those moves in the public discussion of the climate emergency that present sustainability as an option that we could pursue or not, as something that has no default primacy but rather incurs the full burden of justification. Further, it provides a research program: tracing how language in discussions of sustainability "consistently contradicts itself" when it "serves the status quo" and how "this is evident from individual linguistic structures themselves."⁴⁵

In this way, we can begin to see how the idea of language as a repository of objective reason can have critical purchase. Specifically, Critical Theory's role would consist in disclosing the way that our current society runs counter to the life form we agree in, clues of which can be found in the language we use; and, thereby, its role would consist in disclosing how this society is irrational, indeed pathological.⁴⁶