Wilfrid Sellars was not only one of the last systematic philosophers of the twentieth century but continues to be relevant in light of his impact on the development of analytic philosophy in America and abroad. Sellars developed influential arguments for the rejection of the given in epistemology, a unique account of non-conceptual content, a commitment to scientific realism, and a functionalist approach to meaning—all of which have received substantial attention in the past 30 years. Moreover, Sellars’ conception of normativity—a conception that posits normative accounts of language, action, and agency as entailing a sui generis dimension of language, one seemingly ‘over and above’ naturalistic descriptions of agency—has been especially influential, motivating currently popular inferentialist and conceptualist accounts of everything from language and material inference to meaning and consciousness (Brandom 1994; McDowell 1994).

While Sellars’ work from the mid-1950s forward has been the subject of critical collections (Castañeda 1975; Delaney 1977; deVries 2009; O’Shea 2015) and interpretive books (deVries 2005; O’Shea 2007); his earliest publications (ranging from 1947 to the mid-1950s) have received substantially less attention. The practical exclusion of Sellars’ earliest papers has two important consequences: (1) Sellars’ philosophy is depicted as an internally consistent and singular position and (2) the historical context surrounding Sellars’ early work is largely ignored. Thus, Sellars’ later philosophy is not depicted as developing out of his earlier thought, a depiction that would be responsive to the contextual shaping of his peers and intellectual inheritance, but as a holistic position not marked by frequent substantive changes.

The point of this book is to offer a historical account of Sellars’ early thought that both situates his attempt to formalize pragmatics among his peers and historically grounds his conception of language and linguistic rules as it developed from his earliest, under-analyzed publications to his most cited works in the 1950s and 1960s. By starting from a historical perspective, I argue that three main conclusions challenge the currently favored, ahistorical vision of Sellars’ philosophy:Standing as distinct historical and philosophical statements, all three claims are interwoven once Sellars is understood from a historical perspective. Although pure accounts of syntax and semantics were common in the 1940s, Sellars’ pure conception of pragmatics appears as a novel project—one attempting to formalize models of linguistic behavior without using behavioral or psychological facts (defining features of pragmatics in the 1930s and 1940s). As even Sellars himself recognized, a formalist definition of philosophy would need to be abandoned if a distinctly philosophical conception of language, one that explicitly incorporates psychological and sociological facts concerning linguistic behavior, was to include normative considerations. Sellars’ reliance on such facts, though not exhausting his conception of normativity, is largely responsible for the tension between normative and naturalistic commitments in Sellars’ philosophy.

The tension between normativity and naturalism can be framed in numerous ways (e.g., as a conflict between rationalism and empiricism, intuitionism and emotivism, anti-realism and realism), but one of the main issues concerns what is required of explanations between differing, sometimes conflicting, conceptual frameworks. Throughout Sellars’ work, one finds a consistent attempt to “mesh” the normatively laden concepts and terms of the ‘manifest image’ framework¹ with the descriptive and postulated categories and entities found in the sciences. In Sellars’ earliest publications, this concern manifests itself as the project of properly demarcating factual from non-factual discourse in order to characterize how a formal analysis of language could ‘properly’ situate the formal, and therefore philosophical, dimension of concepts alongside empirical concepts. This project is framed as offering requirements for characterizing what Sellars calls ‘empirically meaningful languages’, although such requirements are construed as non-factual concepts that do not interact with descriptive or factual characterizations of historical languages. Thus, Sellars’ early project is beset by one form of this tension: giving an account of language that is stridently non-factual, but providing necessary concepts for any explanation of language.

This tension is also seen once psychological and sociological facts are available for philosophical explanation of agency and linguistic practices—explanations, by definition, unavailable from a formalist point of view. What makes Sellars’ later philosophy possible is the abandonment of his early formalist meta-philosophy. These considerations, I shall argue, are readily apparent when one traces the historical development of Sellars’ conception of normativity. Without these changes to his early meta-philosophy, Sellars’ later meshing of manifest and scientific conceptual frameworks could not have developed into the interlocking system of theoretical and practical reasoning for which he is presently known.

In addition to offering a developmental and historically grounded account of Sellars’ early publications, I also explore the philosophical value of pure pragmatics. While some of Sellars’ early ideas, have been touched on recently (e.g., Brandom 2015), most of Sellars’ early publications are simply absent from the literature. Even though Sellars himself frequently cited some of his early publications, the papers that fall under ‘pure pragmatics’ make infrequent contributions² to his overall philosophy. Nonetheless, the views and formulations of ideas contained within pure pragmatics cannot simply be dismissed without examination; just because Sellars, and even most of Sellars’ contemporaries, did not comment on how all of his arguments ‘fit together’ does not mean our default position should be to see them as inherently problematic. Despite their ignored status, the arguments and ideas in pure pragmatics should be explored and historically situated before it is assumed they are irrelevant for the rest of Sellars’ philosophy.

Chapter 2 sets the historical context within which Sellars’ initial publications developed. I characterize the then-dominant understandings of semantics and pragmatics as addressed by Rudolf Carnap, Charles Morris, and others. Once these notions are established, I set the stage for Chap. 3 by tracing the reception history of Carnap’s shift into semantics as it is found in the works of philosophers at the University of Iowa (where Sellars began his philosophical career). Specifically, I argue it is not primarily Carnap’s conception of semantics and pragmatics should be read as the impetus for Sellars’ formal treatment of pragmatics but, instead, a combination of the reception history of Carnap’s semantics and then-dominant conceptions of semantics and pragmatics. Bergmann’s and Hall’s interpretation of Carnap, and the juxtaposition their differing interpretations represent, helps explain why Sellars saw a need for the formal articulation of pragmatics—an anomalous project for the time period—and why such a project relies on a misreading of Carnap’s philosophy.

Carnap’s philosophy, especially the meta-philosophy found in his Logical Syntax of Language, plays a monumental role in shaping Sellars’ initial attempts to formulate concise philosophical and meta-philosophical positions. Formalism plays a demarcational and legitimizing role for definitions of philosophy; Carnap’s syntactical phase saw the adoption of a restrictive understanding of what kind of claims or subject matter ‘legitimately’ counts as philosophical. It is the idea that non-factual philosophical concepts are legitimate insofar as they count as logical or syntactical characterizations or explanations of language that anchors formalist definitions of philosophy (Carnap 1937, pp. 279–80). As I discuss in Chap. 2, Sellars mimics (though does not identically reproduce) this kind of meta-philosophy, arguing the difference between philosophical and non-philosophical concepts (and, therefore, what counts as the ‘legitimate’ subject matter of philosophy) turns on their specifically formal nature. Problematically, Sellars fails to clarify exactly how all philosophical concepts are formal and why philosophical concepts (once formally characterized) should be seen as necessary for adequate accounts of language.

Sellars’ attention to meta-philosophy, as well as his privileging of an ostensibly formal conception of philosophy, is explicit throughout his earliest publications. Although I discuss this in Chaps 2 and 3, the role of meta-philosophy in Sellars’ work should be quickly clarified. Counting as a preoccupation for most of his career, Sellars’ very first publication starts by both loosely defining the nature of philosophical concepts and bemoaning the then-current state of confusion over the ‘proper’ subject matter of philosophy.³ Developing a meta-philosophy that clearly defines the philosophical dimension of concepts generates two specific benefits: it allows philosophy to carve its own non-factual dimension of conceptual analysis (one that cannot be overtaken by the increasing reach of the sciences) and, subsequently, it determines the ‘proper’ subject matter of philosophical claims. The latter claim is particularly important, as what counts as justifying or supporting a philosophical concept (as both specifically philosophical and required to characterize ‘first-order’ problems) presupposes some stance on the relevant kinds of justifications. In other words, in order to demarcate the ‘properly’ philosophical from the non-philosophical dimension of concepts (and thereby secure a theoretical role for philosophy), Sellars must determine what kinds of claims count as specifically philosophical. By defining philosophy as a wholly formal pursuit, Sellars is—by definition—ruling out factual considerations as relevant for philosophical concepts.

This understanding of ‘legitimate philosophical concepts’ clashes with the then-accepted understanding of semantics and pragmatics. While the early Carnap had no interest in addressing meaning or meaningfulness as a legitimate philosophical category (or, more so, accounted for philosophical conceptions of meaning in terms of syntax) during the syntax phase of his philosophy, Sellars gives a formalist rendering of traditionally semantic and pragmatic concepts in order to mark them as specifically philosophical notions.⁴ The intellectual progression in the 1930s and 1940s saw the role of philosophy shrink in determining semantic and pragmatic concepts. In particular, pragmatics was conceived of as a branch of study generally limited to descriptions of linguistic usage or practices—that is, factual accounts of actual occurrences of language (e.g., anthropological studies of communities and their linguistic habits and practices). Although pragmatics was the original source for all investigations of language, semantics and syntax could (arguably) be abstracted away from empirical investigation in order to reconstruct these subjects in purely logical or structural terms. Because of its connection to empirical investigations of actual languages and practices, pragmatics was generally, though not exclusively, seen as resistant to formalization. Thus, the employment of a formalist meta-philosophy in developing a non-factual conception of pragmatics is seemingly a conceptual confusion. Sellars’ early conception of pragmatics, while obscure, is similar to then-accepted conceptions of pragmatics, yet attempts to offer a specifically formal account of prag...