A Sellarsian Approach to the Normativism-Antinormativism Controversy

1. Introduction

In his recent book—Explaining the Normative—Stephen Turner (2010), one of the leading figures of contemporary philosophy of the social sciences, argues that normative facts are not themselves necessary for the explanation of human social practices. The latter can instead be completely accounted for in non-normative, causal, or dispositional terms (e.g., adaptive habits of action). He calls the contrary position—that is, the view that normative facts cannot be completely accounted for in non-normative terms—“normativism” and places Sellars, among others, squarely within the normativist camp. In this article, it shall be argued that not only Turner misdescribes Sellars’ view of normativity but also that Sellars’ own view provides the key to a proper resolution of the debate between the normativist and the anti-normativist. Drawing on an early Sellarsian article (“A Semantical Solution to the Mind-Body Problem”), we identify a common underlying and unquestioned assumption shared by both normativists and anti-normativists (including Turner himself) and attempt to overcome it by drawing a distinction between causal or explanatory reducibility and logical or conceptual reducibility of the normative to the non-normative. Pace anti-normativism, to explain the normative in non-normative terms, does not entail that normative discourse is logically or conceptually reduced to non-normative discourse, while pace normativism, the fact that normative discourse cannot be derived from non-normative discourse (or defined, without loss, in its terms), does not entail that the former is also explanatorily irreducible to the latter. Moreover, this Sellarsian approach can provide a viable alternative to views according to which the only way to resolve the controversy between normativism and non-normativist naturalism is through a non-explanatory understanding of normativity—exemplified, for example, by hermeneutic phenomenology.

2. Turner’s Anti-normativism and the Normativist Response

As Turner lucidly shows, normativists—that is, those who hold that normativity is explanatorily irreducible to non-normative concepts or processes—use a standard argumentative strategy against non-normativists. They argue that normative facts cannot—that is, logically cannot—be explained in non-normative terms. This is because a necessary condition for an adequate explanation of any phenomenon—including, of course, normative facts—is the availability of an adequate or correct description of this phenomenon: a description that will include those (essential) features of the phenomenon that differentiate it from other, non-normative phenomena. However, normative facts cannot be adequately described without reference to their essential features, namely, the fact that they are facts that can be valid or invalid, that is, the description of which essentially includes their evaluation as “good” or “bad,” or, more generally, as correct or incorrect. Therefore, if one attempts to explain those normative facts of “correctness” or “incorrectness” (“rightness” or “wrongness”) with the exclusive use of non-normative vocabulary, what one in fact does is change the subject; one explains something else (e.g., the non-normative “correlates” of normative phenomena such as certain neurophysiological events or certain adaptive habits of action ingrained by processes of social conditioning) rather than the normative phenomena under consideration. Natural objects, properties, and relations, as conceived at the non-normative level, are not sufficient for explaining the essential, defining characteristics of normatively laden reality. It seems, therefore, that there is a domain of special, sui generis facts or “phenomena,” which, although it cannot be non-normatively explained, is nonetheless perfectly real, since the facts of this sui generis “normative reality” are presupposed for the existence and function of any rational belief forming processes, including science, that is, the very practice that attempts to explanatorily reduce normative facts. Basic normative phenomena such as thought, understanding, rationality, the language of beliefs, intentions and action, and so forth are always already integral parts of our experience of reality, which is why any attempt to account for them in terms of non-normative concepts and processes is doomed to failure since it misdescribes the explanandum—the very phenomenon to be explained, to the point of absolutely losing it from view (Turner 2010, 186-92). According to this line of thought, there exists a domain of genuinely worldly facts that imbue our practices with normative force, that is, with the capacity to motivate and effect changes in physical (non-normatively structured) processes, by exerting a sui generis kind of “influence” in them (“normative force”). Absent those essential normative elements, our very being-in-the-world would be incomprehensible; yet, this normativity falls outside the scope of any actual and possible empirical-scientific redescription and explanation.

Turner argues that the above normativist conclusions are metaphysically inflationary and end up creating an a priori explanatory gap between the natural world, as described by the natural and social sciences, and “normative reality,” as revealed by the above philosophical (transcendental) arguments. In fact, in a previous work (The Social Theory of Practices), Turner (1994) had already pointed out that normative phenomena such as rationality, understanding, the language of belief, intention, action, and so forth are themselves understood as essentially collective abstract entities (e.g., shared rules, premises, presuppositions, tacit knowledge, traditions, shared skills), which, by being “shared” by individual members of a group, can adequately explain or ultimately justify their actions performed in a context of a certain human practice. According to Turner, the attribution of a direct causal force to what is essentially an abstract and collective entity cannot explain how the actions of individual members of a practice can be causally efficacious in the physical world except by “partaking” in this mysterious, semi-autonomously existing, collective entity, from which the members of the practice somehow “download” the information needed to motivate and produce their individual actions (Turner 1994, 45, 62-63, 100; 2002, 16-20). Moreover, Turner (2010) in his work Explaining the Normative places his argument against the causal efficacy or autonomy of practices, construed as abstract collective entities, within a larger context in which this time, the “enemy” is normativism, that is, the view that precisely attempts to justify this sui generis kind of “worldly” efficacy of human practices by holding that this sui generis efficacy is due to their being essentially normative, where the notion of “normative force” in this context is understood not only as something constitutively irreducible to scientifically construed causal forces but also as something that has the power to change physical reality in its own, sui generis—that is, “normative” —manner. However, if, as Turner suggests, “every time something goes on normatively, something goes on also causally” (Turner 2007, 63-64; 2010, 147) and if a necessary condition of mastering normative discourse is a certain learning process that cannot be normative all the way down but is, in essence, a causal process of behavioral conditioning based on human natural reactions of an irreducibly non-normative character (Turner 2010, 145-48), ¹ then it seems that the way is open for a resolutely naturalist view to the effect that the causal efficacy of human practices, whatever else the latter involve, can be adequately understood in terms of the formation, sustenance, and transmission (e.g., through ritualized actions) of behavioral habits of action, which, although uniform “externally,” actually do have internal and personal consequences that vary from individual to individual (Turner 2002, 12; 2010, 173-80). On the basis of arguments of the above kind, Turner concludes that the mysterious addition of an irreducible and, at the same time, causally efficacious normative layer of reality in an otherwise non-normatively described world can be resisted if we naturalize the normative by holding that

normative inferential relations, as they are actually operative in the social world, are redescriptions of, or idealizations of, causal processes. At no point is a special new “normative” fact inserted into the relevant processes. The same facts are redescribed in normative terms. These terms have no special privilege. They do not correspond to an essence. There is a non-normativist alternative description at every step. (Turner 2010, 147)

Of course, proponents of normativism would not accept the above analysis. For example, they have often argued (see, for example, McDowell 1994) that they do not at all posit a metaphysically inflated “normative reality” alongside the ordinary natural world and its causal processes; instead, the concept of “normative reality” functions as a reminder of precisely our ordinary concept of “nature” as the latter is disclosed from the inside of our multifarious human practices, a fact that is obscured, or even lost from view if we take it that the explanatory methods of the scientific image exhaust the domain of the “natural” (i.e., if we accept the scientia mensura principle, as the latter is lucidly formulated by Sellars: “In the dimension of describing and explaining the world, science is the measure of all things, of what it is that it is, and of what it is not that it is not” ² ; Sellars 1956, §41). With this theoretical move, normativists attempt to metaphysically deflate normative discourse, by locating it in “the rough ground” of our ordinary human practices, in the context of which all human interaction with others and the natural world is conducted. If there is any “world” or “reality” to which the normativist is committed “ontologically” by accepting normative facts as real, it is the “world of practices” or “practical reality” (see, for example, Rouse 2002), which is as robustly naturalistic as it gets, provided that we overcome what is probably the most deep-seated philosophical prejudice of our times (reinforced and sustained by a certain metaphysical interpretation of scientific practices): that the domain of the “natural” is exhausted by what can be non-normatively described and explained by the natural and social sciences and does not include “second nature” as the latter (its distinctive “space of reasons” intelligibility) is revealed by a normal human upbringing within society (McDowell 1994, xix-xx, 66-86). This normativist philosophical viewpoint, at this level of generality, can be equally represented by both “normative naturalism” (or “naturalist normativism”), wherein nature is seen as intrinsically normative in form (Rouse 2002) and “naturalised Platonism,” in which the normative demands of reason are considered as real, causally efficacious, independent of our opinions, and, at the same time, belonging within the “space of nature”—since they are thought of as a “product” of the development or “growth” of our “second nature” within a context of essentially situated and “embodied” human practices (McDowell 1994, 77-86).

It may seem, of course, that this debate sooner or later reaches a stand-off in which the naturalist accuses the normativist of postulating an unnecessary extra level of causally efficacious abstract entities or properties (“practices,” “normative facts”), which, moreover, stand by definition outside the order of ordinary (natural and social) scientific explanation, whereas the normativist retorts that those sui generis properties of normative practices are mundane, ordinary facts, essential to our self-understanding as agents who can perform actions on the basis of reasons within the natural world (but, at the same time, irreducible to non-normative explanations due to the fact that certain normative facts within our practices are always presupposed in every non-normative explanation of a practice as a whole). Predictably, the “practice-normativist” could continue by pointing out that the only reason for finding these platitudes intellectually unsatisfying is the unjustified, intellectually non-obligatory conviction that only science can be the final arbiter what is “natural” or “causally efficacious,” while the anti-normativist would point out that there is no need to rethink our empirical-scientific concept of nature, since all those normative facts that the practice-normativist considers as “mundane” and “ordinary” are, in fact, heavily laden with theoretical presuppositions (e.g., about ways of causal influence), which are not themselves irreducible to empirical-scientific explanation (see, for example, Turner 2002, 138-41, 160-62). ³

In what follows, we shall propose that a potentially fruitful way out of this controversy between the normativist and the non-normativist can be outlined on the basis of Sellars’ view on normativity. It must be noted that Sellars is one of the major philosophical figures that appear in Turner’s book Explaining the Normative. There, Sellars is placed squarely within the normativist camp, as one of his major representatives, but, as we shall see in the next section, his view is far more nuanced and sophisticated, having affinities with both normativism and anti-normativism, when the latter are properly understood. Moreover, Turner devotes a whole chapter of his book (chapter 5, titled “Communities, Collective Intentions and Group Reactions”) to an analysis and critique of Sellars’ theory of collective intentionality (as the ultimate source of normativity). A detailed description and critique of what we think is actually a mischaracterization of Sellars’ position in this matter by Turner is beyond the scope of this article. However, it is our contention that the basic Sellarsian argumentative strategy for the overcoming of the normativist-antinormativist controversy, which shall be developed in the next section, can be also used for showing the problematic character of Turner’s interpretation of Sellars’ views on collective intentionality.

3. The Sellarsian View of Normativity as the Key to the (dis)Solution of the Normativist-Antinormativist Controversy

We propose that there is a properly Sellarsian response to the debate between the normativist and the non-normativist. It is based on the idea that the whole controversy is permeated by a systematic ambiguity in the meaning of the terms “understanding,” “irreducibility,” or “non-derivability” as the latter are used in the following expressions of the normativist thesis (all of which occur time and again in Turner’s 2010 book): “the normative is irreducible to the non-normative,” “the normative cannot be understood (accounted for) in non-normative terms,” “normative terms cannot be defined in (derived from, translated in) non-normative terms.” And the same systematic ambiguity is also an essential feature of the non-normativist negations of the above normativist theses (including Turner’s attack to normativism). According to Sellars, a crucial distinction should be drawn in this case between logical or conceptual irreducibility—that is, the impossibility of defining or translating the content of a concept in terms of another or deriving the former from the latter—and causal or explanatory irreducibility, where a concept can be causally reduced to another if the latter can causally explain the former, without making any essential use of descriptions or judgments with normative content (Sellars 1953, §6). ⁴ (Sellars draws this distinction to solve a parallel problem with regard to the possibility of reducing the ethical “ought”—for example, the concept of “obligation”—to “is,” that is, to physical, psychological, and social [non-normative] facts. As is obvious, the problem of the reduction of ethical “ought” to a purely descriptive “is” is actually a special case of the wider normativist-antinormativist problematic.) Pace non-normativism, normative facts can indeed be irreducible to non-normative facts if this irreducibility is understood as being logical or conceptual in character (i.e., if it concerns relations of logical or conceptual entailment between normative and non-normative facts), ⁵ but at the same time, pace normativism, this does not exclude the possibility that normative facts can be, in another (causal-explanatory) sense, reducible to non-normative facts. Again, pace non-normativism, the possibility of explaining the normative in non-normative terms, does not entail that normative discourse can be logically/conceptually reduced to non-normative discourse, while, pace normativism, the fact that normative discourse cannot be defined, translated without loss in terms of non-normative discourse or derived from the latter does not entail that the former is also explanatorily irreducible to the latter. The impression to the contrary stems from an underlying assumption that is tacitly shared by normativists and anti-normativists alike: that an explanation of the normative in non-normative terms necessarily goes hand in hand with a logical or conceptual reduction (derivation) of the former to the latter (Sellars 1953, §9). This shared background assumption—which essentially amounts to a deductivist understanding of explanation—explains why although non-normativists, for example, rightly (by Sellarsian lights) assume that normativity can, indeed, be explanatorily reduced to non-normative phenomena, they also (wrongly) think that this necessarily implies that the former can also be conceptually reduced to non-normative (e.g., biological, neurophysiological, psychological, social) categories. ⁶ The above tacit assumption also explains why although normativists, on their part, rightly hold that normative phenomena cannot be conceptually reduced to non-normative phenomena, they also (wrongly) think that this necessarily implies the impossibility of explaining the former in terms of the latter. ⁷

Now, Sellars’ above-mentioned conception about the relations of the normative to the non-normative is based on the further and more fundamental Sellarsian thesis to the effect that a proper philosophical investigation of the “logical grammar” of normative discourse reveals that the latter, despite first appearances, is not in the business of describing and explaining states of affairs in the world; rather, what it actually does is specify certain functions or roles that various empirical (linguistic or non-linguistic) items play, which serve as standards for assessing their actual “behaviour.” Thus, the job of normative discourse is to prescribe and evaluate (not describe and explain) the “behaviour” of certain “empirical items” (including human behavior) according to standards or “ideals” about their proper “empirical realization” (see, for example, Sellars 1957, §79). A system of normative prescriptions is not an explanatory theory for the description of reality (or the mind), but rather a system that aims at motivating actions for changing empirical reality, in case it fails to accord with certain “standards” or “goals” (Sellars 1949, §17-18). ⁸ This profound difference in function among our descriptive-explanatory practices and our practice of normative assessment (the former places empirical items in what Sellars 1967 calls “the logical space of causes and effects” [p. 137], while the latter places them in “the logical space of reasons”), combined with the fact that, for Sellars, the very essence of personhood—that is, its identity and individuation conditions—is normative, rather than descriptive or explanatory, ⁹ is the ultimate reason why normative discourse cannot be conceptually reduced to non-normative discourse. If, per impossibile, such a conceptual reduction of the normative to the non-normative could actually be carried out, not only would everything normative be extinguished from the world but our very sense of personhood—which is what makes us human in the first place—and, with it, the very point of our practices (including our scientific practices) would also be completely lost from view ¹⁰ (Sellars 1963, 18).

With the above clarifications in mind about the role of normative discourse, we are also now in a position to make full sense of the (in)famous Sellarsian scientia mensura principle “In the dimension of describing and explaining the world, science is the measure of all things, of what it is that it is, and of what it is not that it is not” (Sellars 1956, §41, emphasis added), without committing ourselves to the absurd idea of a complete elimination of the normative from all areas of discourse or all dimensions of human existence. ¹¹ It may well be that, according to Sellars, and in line with anti-normativists, normativity is completely eliminated from an inventory of the basic ontological categories (categorical concepts) with which the spatiotemporal and the “sensible” world is adequately described and explained. ¹² But, pace anti-normativism, it does not follow that normative phenomena are, for example, just “useful fictions,” causally impotent “epiphenomena,” which are subjectively “projected” to objective reality (Mackie 1977, 38-42, 48-49; Friedman 1998, 244-45) or, to mention just another well-known view in the vicinity, something that “can only be real if it is really something else” (i.e., non-normative phenomena) (Fodor 1987, 97). In Sellars’ view, an eliminativist account of normativity would indeed be correct only if the job of normative discourse was to describe and explain states of affairs in the world (or, in the mind), ¹³ while projectivism about normativity, as described above, would stand a chance only if the notion of objective knowledge was itself a descriptive and explanatory (i.e., non-normative) concept. However, as was mentioned above, according to Sellars, the function of normative phenomena in our practices—including, crucially, the “knowledge language-game”—is different from that of describing and explaining states of affairs in the world. ¹⁴ The job of normative facts is to prescribe and evaluate human behavior according to certain standards of correct action. ¹⁵ Thus, in the above precise sense, normative “phenomena” can indeed be considered as real, objective, and not reducible to the scientific image, although it can equally be said, without the slightest contradiction, that those very normative “phenomena” can be adequately explained only with the use of non-normative concepts ¹⁶ (see also O’Shea 2010, 459-70).

4. A Sellarsian Response to Turner’s Arguments against Normativism

We are now in a position to see, in more detail, how our Sellarsian view of normativity would escape Turner’s criticism of “normativism,” while acknowledging an important truth in it, which is lost from view if it is considered as being simply false.

Turner takes it that normativism makes use of highly controversial arguments. Some of the most problematic features of this “family” of arguments according to Turner include (1) queerness, that is, “admitting the existence of an unusual class of facts [ . . . ] involving entities with a spooky relationship to the actual causal world,” (2) the problem of description, that is, “the need to insist on the unique validity of particular descriptions of events that require explanation in terms of these unusual facts,” (3) the problem of the construction of normative facts themselves

which were made by analogies -rules which are not ordinary explicit rules, but tacit; spaces, for example, of reason, which were not spatial; presuppositions that are not, as in Euclid, actually presupposed, but were rather conditions for the possibility of the normative facts that are retrospectively constructed with the purpose of explaining them,

and (4) a constant reliance on circular notions

such as appeals to communities, for example the community of rational beings, which were not communities in any sense other than that they are composed of people who adhered to the rules that the fact of community is supposed to explain. ¹⁷ (Turner 2010, 180)

I think that the Sellarsian response to these charges against normativism will by now be obvious: Sellars would wholeheartedly accept Turner’s critique of normativism with regard to, for example, the “queerness” of normative facts (their putative causal efficacy), the fact that normative descriptions are immune from redescription and reconceptualization, and the patent circularity of normative explanations. The significance of the Sellarsian distinction between descriptive-explanatory and prescriptive-justificatory domains of discourse and his insistence that normative discourse has nothing to do with describing and explaining can be now clearly discerned. With this theoretical move, he is in a position to side with the anti-normativist with regard to the issue of the explanatory circularity of normativist arguments (including the explanatory account of normativity in terms of “we-intentions” that Turner attributes to Sellars) while acknowledging an important truth in the normativist “circular” transcendental ¹⁸ arguments: the latter (including, pace Turner, Sellars’ actual account of normativity in terms of “we-intentions”) are correct insofar as they are construed as a conceptual analysis (à la Strawson) of a domain of discourse or practice, provided that this philosophical task of conceptual analysis is strictly distinguished from the task of describing and explaining states of affairs in the world (something that is not always appreciated by adherents of “conceptual analysis-style” transcendental arguments). ¹⁹

5. Anti-normativism Proper, Pace Turner, Incorporates Central Elements of Normativism

Moreover, this Sellarsian line of thought about normativity holds the key to a proper (i.e., non-eliminativist and non-projectivist) answer to a classic objection, this time leveled against the anti-normativist “explanatory reductivist” view, which has been already mentioned by Turner, to the effect that an explanatory reduction of the normative to the non-normative is conceptually impossible since it simply “changes the subject.” According to this normativist objection, what is actually explained in terms of non-normative phenomena are not normative phenomena as such, but rather, at most, their non-normative correlates. And this kind of explanandum is essentially such (e.g., non-normatively described neurophysiological state, biological adaptations, habits of action) that it does not even make room for distinctions essential to the very being of the phenomenon to be explained. For example, in case the explanandum is that of intentional action, the very distinction between intentional action performed on the basis of reasons (potentially in conflict with one’s subjective desires and impulses), and unintentional, impulsive, reflex, or otherwise purposeless bodily movements is construed, at best, as a distinction that is “subjectively imposed on reality,” and is, therefore, essentially lost from view as a potentially objective distinction, if the content of the explanans is non-normative. However, this critique against anti-normativism—which, in our view, is, indeed, compelling against eliminativism and projectivism—overlooks the fact that, in Sellars’ more nuanced view, when one attempts to understand the relations between the normative and the non-normative level, the conditions of identity and individuation of the former do not disappear from view, even if it is considered as capable of being explanatorily reduced to the latter. The conditions of identity and individuation of the normative remain absolutely intact and objective, provided that this special, sui gereris, “normative” mode of understanding ourselves and the world is conceived not in descriptive or explanatory terms but rather in prescriptive and evaluative terms, which aim at regulating (not describing or explaining) our actions. Only in this way—that is, only if the normative point of view remains firmly in place—is it possible for a non-normative explanation to be about normativity, rather than something entirely non-normative, situated at a sub-personal, for example, neurophysiological,or “pre-personal,” for example, habitual level. ²⁰ This is the kernel of truth contained in the normativist thesis.

However, as was briefly pointed out in note 14 above, this profound truth, that is, the affirmation of the objective purport of the normative point of view, is systematically distorted by the (according to Sellars, misguided) normativist’s insistence that the distinctive irreducibility of the normative level lies in a sui generis dimension of description and explanation of reality. According to this latter normativist view, the only way to preserve normativity in our worldview is to admit the reality and sui generis character of normative descriptions and explanations that are immune from explanatorily driven redescription and reconceptualization in the context of empirical-scientific inquiry (on pain of eliminating normativity altogether from our self-conception and losing thereby our grip on absolutely central elements of the “phenomenology” of our own personhood, which is what distinguishes us from being mere “complex physical systems”). On the contrary, in Sellars’ view, this latter possibility of redescription and reconceptualization of our categories, concepts, and conceptions pertaining to normativity on the basis of explanatory scientific concerns of our ongoing empirical inquiry is, indeed, genuine. However, pace normativism, this possibility, far from eliminating normativity and “personhood” from our worldview, would, if actualized, effect a change in the descriptive and explanatory conceptual recourses with which we refer to the world that is the causal antecedent of our normative responses. We gradually come to have an alternative understanding of this world with regard to its material constitution, evolution, and causal efficacy, and, in this way, our conception of normativity itself changes significantly. In the last analysis, the latter is revealed to be not a sui generis reality in which some ethereal or otherwise inexplicable “demands of reason” inhere, but rather a reflection of our most embracing logical, modal, and ethical commitments, as the latter are expressed in what Sellars calls “we-intentions” (see note 9). And a proper understanding of the material (evolutionary, historical, social-psychological) antecedents of our normative responses not only does not eliminate normativity (or personhood) from our practices but also actually paves the way for its potential material realization, by liberating persons from unforeseen, uncontrollable, and hitherto completely unknown practical restrictions and impediments. ²¹

Now, this does not, of course, mean that the Sellarsian position sketched out above is in agreement with anti-normativism, for example, of Turner’s stripe. The basic problematic element of Turner’s anti-normativist position, from a Sellarsian point of view, is that, although Turner rightly insists on the possibility of a successful explanatory reduction of normative concepts to non-normative concepts, he does not see that neither does that fact undermine the radical conceptual/logical irreducibility of normative to non-normative discourse nor does it compromise the objective purport of the normative point of view. Turner seems convinced (and in this, he is in complete—albeit, tacit—agreement with the normativists) that the knowledge obtained by conceptual systems that provide adequate causal explanations of certain domains of “phenomena” is the paradigm of knowledge and exhausts objective knowledge in general. ²² This impression is even stronger in the case of the scientific image, especially to the extent to which the latter purports to provide adequate explanations of normative discourse itself in non-normative terms. However, according to Sellars, this view overlooks the fact that the knowledge “language-game,” although a necessary condition of describing and explaining the world, is not itself in the business of description and explanation. Thus, the concept of “objective knowledge” cannot be the exclusive possession of a conceptual system that operates in the dimension of description and explanation, even if, as in the case of the Sellarsian scientific image, this system has absolute cognitive authority in those areas of discourse. Successful empirical-scientific causal explanations only exhaust our knowledge about “matters of fact,” “what is the case (and why),” that is, about “descriptive” knowledge; yet, knowledge is more than knowledge about what is the case (and why). It is also knowledge about what ought to be the case, that is, “normative” knowledge, and this latter kind of knowledge can certainly make (instrumental) use of empirical-descriptive-matter-of-factual knowledge, but its content is not constituted by the latter. As Sellars himself says, “Not all knowing is knowing how to describe something. We know what we ought to do as well as what the circumstances are” (Sellars 1957, §107).

At this point, we can use another crucial Sellarsian distinction, that between what a proposition says or asserts and what information one “conveys” by using this proposition to say something, to distance Sellars’ view from Turner’s anti-normativism. Sellars makes his point as follows: “Whatever users of normative discourse may be conveying about themselves and their community when they use normative discourse, what they are saying cannot be said without using normative discourse” (Sellars 1953, §66). Thus, although the information conveyed by normative discourse (e.g., physical, historical, psychological, social facts culminating in certain habits of action) can, in principle, be described and explained in non-normative terms, what is asserted by the former, that is, its (prescriptive) semantic content, cannot be conceptually reduced to what is asserted in non-normative (descriptive-explanatory) terms. ²³

Now, an interesting consequence of the above analysis is that one could, for instance, even accept as meaningful Turner’s hypothesis to the effect that certain non-normative cognitive mechanisms, operating at the non-intentional, pre-linguistic level (e.g., “mirror-neurons”), can be cited in an explanation of the binding force of normative discourse as the latter is expressed in actual practice, ²⁴ yet, at the same time, one can deny, on purely conceptual grounds, that those putative non-normative descriptions and explanations are part of the semantic content that is asserted in propositions expressing normative language and insist that the truth or falsity of what is thereby asserted can be objectively assessed with the use of standards internal to normative discourse (e.g., intentions that are warrantedly assertible, that is, intersubjectively reasonable and universally applicable) (deVries 2005, 271-72). This essentially normative content is not descriptive and explanatory, and the raison d’être of normative language is not to describe or explain what is the case in some “realm” of “facts” (in the world, the mind, the “transcendental subject,” behavior, practice), but rather to provide standards of assessment—and possible change—of worldly, behavioral, or practical “facts” in accordance with ideals of (proper) function and performance as the latter are expressed in intersubjectively reasonable and universally applicable “we-intentions.”

At this point, Turner might object that his anti-normativism does not eliminate normativity altogether from our self-conception, but rather just shows that it does not have any explanatory import in our practices, while still allowing for the possibility that normative discourse is just different in its aims and methods from descriptive-explanatory discourse. Indeed, following Friedman (1998), he treats normativity as something we impose on a causal world, as a matter of our “fashioning a normative prescriptive lens” to the world through which to view it (Turner 2010, 7-8). Is that not enough for saving whatever can be saved from normativity within a resolutely “disenchanted” naturalist, anti-normativist framework?

Unfortunately not, from a Sellarsian perspective. Although the latter is by far the most radical naturalist perspective in twentieth century analytic philosophy (combining naturalism, nominalism, and scientific realism), from its standpoint, Turner’s “rehabilitation” of normativity, due to his not being crystal clear about the radically different kinds of irreducibility—or reducibility—relations among normative and non-normative concepts, amounts to a certain “projectivism” about normativity. And this latter thesis, as we saw, fails to account for a central feature of normativity, that is, its objective purport, without which the values that underlie and “animate” our everyday epistemic practices—including, crucially, the epistemic values of scientific practices ²⁵ (on the epistemic authority of which the projectivist thesis itself relies)—would be mere arbitrary expressions of our subjectivity, undermining thereby the epistemic rationale of the very practice that is responsible for providing objective standards of correct description and successful (adequate) explanation. Normative standards are an essential part of the “generative mechanism” of our epistemic practices and, as such, they cannot but have objective purport, on pain of undermining the claims to objectivity or adequacy of the very descriptive and explanatory resources of our practices. Turner himself admits that although the standards of correct explanation are themselves part of the stream of ordinary scientific explanation, the question whether those standards are the correct ones is a normative matter (Turner 2010, 194). Yet, by being reluctant to ascribe objectivity and truth (potential intersubjective validity and universal applicability) to normative discourse, he is deprived of the necessary conceptual resources for making full sense of the claims to adequacy of (his favored social-scientific) descriptive and explanatory discourse in the first place.

Turner could respond, here, by strictly distinguishing explanations themselves from the evaluation of explanations and insisting that they proceed in parallel domains of thought that never interact. Consider, for example, the following suggestive passage:

One way of handling claims about validity would be to treat them as questions that have nothing to do with explanation. Judging the validity of a law, for example, can be treated purely as a normative problem to be answered solely by reference to normative considerations. Whether we can get agreement on these considerations, or even whether they make sense, has nothing to do with the problem of explaining what people actually do. The problem of explanation and the problem of evaluation proceed, so to speak, in different containers of thought. Explanation never needs to take anything from the container marked “validity, correctness, etc.” except when we evaluate explanations themselves—but explanation is something distinct from the evaluation of explanation. (Turner 2010, 187)

Indeed, “explanation is something distinct from the evaluation of explanation” (this is something with which Sellars would wholeheartedly agree), but notice that Turner himself, at the same time, admits that when we evaluate explanations themselves we cannot do this unless we place ourselves firmly within the “space of reasons” (and its distinctive mode of “intelligibility,” as opposed to “space of causes” intelligibility). Now, as we saw above, by adopting Friedman’s “projectivist” account of normativity, Turner compromises the objective purport of normative discourse, that is, that on the basis of which alone can one make full sense of talk about the “evaluation of explanation.” The crucial point here is that explanation is, of course, distinct from the evaluation of explanation (as Turner rightly suggests), but only on condition that the latter is genuinely possible. If, as was claimed above, this possibility is undermined, this also affects ordinary scientific (causal) explanation, since the very intelligibility of the latter depends on the existence of criteria that distinguish (in a reasoned, principled, non-arbitrary way) between correct and incorrect, adequate, and inadequate explanations. Ordinary scientific explanation cannot be legitimized unless it comes with correctness or “adequacy” conditions, ²⁶ and this claim to objectivity—that is, to the possibility of expressing prescriptive categorical truths—is the ineliminable core of the framework of normative discourse; the latter cannot be treated as a mere projection of normative properties in an otherwise non-normative world, on pain of losing our grip on the world (and on what would count as a “projection” upon it) altogether. ²⁷

6. The Sellarsian View of the Normative as an Objection to the Alleged Necessity of Non-explanatory Accounts of Normativity

We have argued that Sellars’ conception of normativity can successfully resolve the normativist-antinormativist controversy, as the latter is construed by Turner in his book Explaining the Normative. But, of course, this way of resolving the above debate does not seem to be “the only game in town.” What does our Sellarsian view on normativity have to say with regard to the possibility of providing alternative (non-Turnerian as well as non-Sellarsian) solutions to the debate on the status of normativity?

In the Sellarsian view developed in this article, the normativist-antinormativist controversy was resolved by insisting (1) on the indispensability of normative discourse with regard to the very intelligibility (our very conceptual grasp, in what Sellars calls “the order of signification”) of the non-normatively structured world and our place in it, and (2) on the provision of an explanation of the existence and function of the normative level itself in non-normative terms. This form of solution presupposes, among other things, the legitimacy of drawing a strict methodological distinction between the normative and non-normative level of discourse. However, in a recent article (Ginev 2013), it is argued that the proper (dis)solution to the debate between the normativist and the anti-normativist lies in providing (pace Turner and Sellars) a resolutely non-explanatory account of normativity, by erasing (pace Sellars) both the ontological and the methodological distinction between the a-normative causal level and the normative level; only in this way, it is claimed, is it possible to successfully respond to a potentially devastating skeptical epistemological argument, that is, Fries’ trilemma, a version of which can be readily applied in the debate on the status of normativity, against both normativism and anti-normativism.

After briefly describing Fries’ trilemma, we shall argue that, pace Ginev, one does not automatically fall prey to Fries’ trilemma just by drawing a methodological distinction between the normative and the non-normative and insisting on the need to explain normativity. The impression to the contrary stems from a confusion about explanation—a failure to distinguish between explanation understood as ontological reduction and explanation construed as conceptual derivation—which also emerged in our previous Sellarsian critique of Turner’s position but, as we shall see, is equally present in positions that not only reject Turner’s anti-normativism but, more radically, put into question the legitimacy of the whole normativist-antinormativist controversy.

Let us now proceed to a brief sketch of Ginev’s argument against normativism and anti-normativism, beginning with his explication of Fries’ trilemma and its relevance to the normativist-antinormativist debate. Fries’ trilemma, adapted to the case of the normativist-antinormativist controversy, states that

in facing the need of making intelligible normative orders of social activities one cannot avoid the choice between three evils: either (1) to spell out the roots of normativity in terms of a foundational metaphysical conception that draws and legitimizes an ontological demarcation line between the causal-natural and the social-normative, (2) to commit to various kinds of “naturalistic fallacy” and “normativist fallacy” that derive the normative from the causal-natural (or vice versa) in a manner that cannot avoid circular arguments, or (3) to admit that there is a potentially infinite cascade of levels of (normative) justification of the norms creating social orders. (Ginev 2013, 276)

Now, according to Ginev, insofar as both normativism and anti-normativism aim at explaining normativity and insist on upholding a strict methodological distinction between the causal and the normative, they end up in the following unsatisfactory position (which, ultimately, cannot avoid the second, and eventually the third horn of Fries’ trilemma):

Explanations derivate or reduce the causal or the normative to an “ultimate reality” of normative regulations or causal relations. Depending on the kind of explanantia in these scenarios, the normative events or the non-normative causal processes have only contingent validity, and must be revealed as “epiphenomena” to what is essential and constitutive in the (causal or normative) world as the “underlying world.” (Ginev 2013, 279)

Ginev argues that even such sophisticated opposing views in the normativist-antinormativist debate as Rouse’s “naturalist normativism” and Turner’s “anti-normativist naturalism” fall prey to the above objection, although both are careful not to assume an ontological hiatus between rationally organized human action and causally ordered natural events (i.e., a thesis that directly leads to the first horn of Fries’ trilemma).

According to naturalist normativism, the normativity of our practices (including, crucially, our scientific practices) cannot be derived by elements that supposedly inhere in some kind of “non-normative nature” underlying it. It is only in the performances of our ongoing practices that genuine, non-reductive normativity can be found and seen for what it actually is. In the last analysis, this view postulates a “primordial” “practical reality” (i.e., what is shared by agents when their actions are answerable to criteria of correct or incorrect performance) from which both the “social world” and the “natural world” are reified abstractions. ²⁸ And it is precisely the mutual accountability of performances with respect to the possibilities the practice makes available that provides the explanans that explains how practices “generate” normativity (simultaneously being accountable to it). However, it is a consequence of this view that the non-normative causal level—itself normatively structured ²⁹ —has only contingent validity, being in the last analysis an “epiphenomenon” to what is essential and constitutive in the “underlying normative world.” In this sense, according to this analysis, naturalist normativism cannot escape the second horn or the Fries’ trilemma: it commits the “normativist fallacy” in its attempt to derive the causal-natural from the normative, in a manner that cannot escape circular arguments.

On the contrary, as we have already seen, the anti-normativist (typically represented by Turner) does not see any reason why any putatively “ultimate” normative explanans cannot be itself explained in non-normative terms. For example, the naturalist normativist’s basic explanatory concept, that is, that of the “mutual accountability of performances” can be readily explained in terms of adaptive habitual behavior, which is a purely a-normative causal notion. The normativity of our practices is not a sui generis causal power of theirs, irreducible to empirical-scientific explanation, but rather an imposition on our part of a “normative prescriptive lens” on the world, through which we view it, albeit without in the least causally affecting it (Turner 2010, 7-9). Yet, this view, in Ginev’s analysis, proves to suffer from exactly the same structural deficiencies that plague its exact opposite, that is, naturalist normativism: it is a consequence of anti-normativist naturalism that the normative level has only contingent validity and is revealed to be nothing more than an “epiphenomenon” (normative “lens” without causal powers) to what is essential—that is, causally efficacious—in the non-normative world. Therefore, anti-normativist, exactly like normativism, cannot successfully escape the second horn of Fries’ trilemma: it commits the “naturalistic fallacy” in its attempt to derive the normative from the causal-natural by explanatorily reducing the former to the latter.

Ginev concludes that both naturalist normativism and anti-normativist naturalism fall prey to Fries’ trilemma, the main reason being that both try to figure out explanatory strategies of how the normative operates. As Given says, “Both positions are predicated on the postulation of an ultimate reality as an actual presence and the commitment to the search for explanatory foundations.” He proposes that “getting rid of Fries’ trilemma requires a non-explanatory approach to the meaningful orders constituted within human practices (including the orders of natural-scientific research)” (Ginev 2013, 283), which questions the legitimacy of the problem of the explanatory priority of one domain over another (normative over causal or vice versa), by uncovering a problematic background assumption of the latter: normativity’s essential isolability from the practices to which it is actually applied. And he contends that a Heidegger-inspired hermeneutic phenomenological construal on normativity—in which the internal (non-explanatory, non-extrinsic) relation between a rule and its applications (celebrated by Wittgensteinians) is constituted by the “hermeneutic circularity of interpretation”—is the best candidate for providing an adequate (non-explanatory) solution to the problem of normativity (Ginev 2013, 283-93).

Now, is the above argument compelling? Is it really the case, for example, that to understand normative phenomena as a proper and legitimate domain to be explained by an independently—that is, non-normatively—described explanans necessarily leads to Fries’ trilemma? From our Sellarsian perspective, the answer should be obvious: this is not so, and the impression to the contrary stems from failing to distinguish between explanation understood as ontological reduction and explanation construed as conceptual derivation (of the [normative] “ought” from the [descriptive] “is”). Ginev uses the terms “reduction” and “derivation” interchangeably (with regard to explanation) (Ginev 2013, 279), but this, of course, is not at all obligatory. The same goes for “naturalist normativism” and its opponent, “anti-normativist naturalism,” which both conflate those very different uses of “explanation,” ³⁰ and should be for this reason, rather than those actually offered by Ginev, criticized for committing the “normativist fallacy” and the “naturalistic fallacy,” respectively. What is more, according to our Sellarsian viewpoint, at a deeper level, the identification of explanation as ontological reduction with explanation as conceptual derivation is reinforced by a certain picture about explanation coupled with the conviction that the business of all non-logical concepts (be they normative or non-normative) is to describe the world: this latter assumption, taken together with this picture about explanation, that is, the view to the effect that the relation between the explanandum and the explanans is that of deriving (deductively recovering) a description of the explanandum from the conceptual resources of the explanans (e.g., laws and principles of a theory), results in making it all the more natural to view the possibility of an explanatory reduction of the normative to the non-normative as identical with that of a conceptual derivation of the former from the latter. ³¹

Notice that if the above distinction regarding the notion of explanation is correctly drawn, pace Ginev, it does not follow that in the very attempt to explain the normative in non-normative terms one automatically commits the “naturalistic fallacy” or the “normativist fallacy.” That is, only if explanation is understood as an amalgam of ontological reduction and conceptual derivation of the normative from the non-normative or vice versa does it follow that the sheer attempt to explain normativity necessarily results in “naturalistic” or “normativist” fallacies; therefore, it does not follow that the mere invocation of the possibility of an explanation of normativity is sufficient for a philosophical position’s falling prey to the (second, and eventually, the third horn of) Fries’ trillemma. Neither, of course, does it follow that, to repeat Ginev’s diagnosis:

Explanations derivate or reduce the causal or the normative to an “ultimate reality” of normative regulations or causal relations. Depending on the kind of explanantia in these scenarios, the normative events or the non-normative causal processes have only contingent validity, and must be revealed as “epiphenomena” to what is essential and constitutive in the (causal or normative) world as the “underlying world.” (Ginev 2013, 279)

Ginev’s above basic diagnosis of why the very attempt providing an explanation of normativity is philosophically problematic holds water only if one overlooks the distinction between explanation as ontological reduction and as conceptual derivation (although it will be remembered that both naturalist normativism and anti-normativist naturalism commit the normativist and naturalistic fallacy, respectively, albeit not because of their very attempt to explain normativity, but rather precisely because they are not sensitive to our Sellarsian distinction with regard to explanation). It seems, therefore, that the recourse to decidedly non-explanatory philosophical viewpoints, such as a Wittgensteinian ethnomethodology or a Heidegger-inspired hermeneutic phenomenology, is not necessary for a proper understanding of normativity.

7. Concluding Remarks

To recapitulate, it should be noted that, in essence, the debate between normativism and anti-normativism is not a narrowly philosophical one, confined within the contours of academic philosophy, but actually represents two fundamentally different kinds of intellectual “reaction” to what can be seen as the most fundamental change in worldview in the movement from the traditional to the modern societies: the disenchantment of the world, that is, the gradual overcoming of “magical” or, in general, “anthropomorphic” ways of understanding the world and our place in it, through the relentless depersonalization of explanation. On one side, normativists (including, for that matter, “non-explanatory” normativist views) have sought to defend the distinctiveness of the human (e.g., its rationality and freedom) against the expansion of naturalistic explanation. On the other, anti-normativists have tried to causally explain human experience, against the mystifications of various non-empirical ways of understanding the world (including more sophisticated forms of the latter, such as transcendental philosophy), although they have too often explained it away. Sellars, on his part, attempts to extract what he takes to be the profound insights of both normativism and anti-normativism while unearthing their (shared) confusions so as to delineate a viable alternative conception of normativity, which, in the last analysis, is nothing less than an expression of his own bold intellectual vision of man-in-the-world in the face of the continuing cultural process of the disenchantment of nature.