The pragmatist order of explanation and examples in Brandom, Peirce, Sellars, Rawls, Anderson, and Korsgaard
In Brandom's discussion of historical pragmatisms, and his identification of himself as a pragmatist, he tends to focus on what he calls a "pragmatist order of explanation." The pragmatist order of explanation is one in which metaphysics is explained in terms of practice, instead of vice versa. In the below, I discuss this order of explanation in Brandom's work and also in other examples, from Peirce as representative of historical pragmatism, from Sellars as a neo-pragmatist, and from Rawls and some students as exemplifying pragmatism in ethical theorizing. Certain arguments of Rawls, Anderson, and Korsgaard fit very neatly into Brandom's schema of the pragmatist order of explanation. A current project of mine involves conceiving certain strands of ethical theorizing as pragmatist so this is useful background work for me.
In general, Brandom's semantic project pursues an order of explanation from inference to representation instead of vice versa (see, e.g., Brandom, 1994, Making It Explicit; xvi, Brandom, 2000, Articulating Reasons, 4). Representationalists start with metaphysics and semantics and then validate inferential practices from that base. Brandom instead starts with the base of inferential practice to develop a semantics. This relates to his strategy of "top-down semantic explanation" in which the meanings of sub-sentential elements are given by their roles in the formation of sentences, which get their significance from their roles in inferential practice (AR, 11-12).
This same pattern of priority is evidenced in many specific places in Brandom's project. Consider for example what he calls his "phenomenalism" about truth: setting out to explain taking-true as a practice, and holding that, once you have done so, you have "understood all there is to understand about truth" (MIE, 287). Similarly, in Brandom's account of formal validity, the goodness of material inferences is basic. Logical language comes along to make explicit material-inferential commitments. Formal validity is then defined using the Fregean substitutional strategy, starting with (and depending on) material validities. In Brandom's discussion of action, similarly, preferences are explained in terms of material practical inferences instead of vice versa (MIE, 233-4, 244-53; AR, 31). To have a preference is to endorse (or be disposed to endorse) a pattern of practical inferences, with the inferences themselves being (in some sense) explanatorily basic.
Brandom frequently labels his explanations as using a pragmatist order of explanation, and he connects this order of explanation with the style of reasoning of the historical pragmatists. For example, while Brandom's treatment of truth is not the "stereotypical pragmatism" that "truth is what works," he does agree with historical pragmatist treatments in treating truth as a practical rather than a descriptive notion (MIE, 285-91). Peirce's treatment of truth is pragmatist in exactly the sense of according priority to practice over metaphysics. Truth is a feature of "the experiential method" of inquiry (Peirce, 1878, "How To Make Our Ideas Clear," 13-5; elsewhere called "the method of science," Peirce, 1877, "The Fixation of Belief"). The truth is the belief that is "fated" to be agreed upon by all inquirers at the ideal end of experiential inquiry. Similarly, reality is the subject matter of true belief. The metaphysical and semantic notions of truth and reality are thus explanatorily downstream of the practice of inquiry.
Sellars's discussion of perception in Empiricism and the Philosophy of Mind (1956) also exemplifies the pragmatist order of explanation. Sellars criticizes the approach of the sense-data theorists, who are pursuing a form of phenomenalism about the external world (see also MIE, 292-6). That is, they take it that our knowledge of the external world can be reduced to our acquaintance with sensory qualities. Sellars shows that this project can't succeed -- even if our knowledge of the external world could be epistemically justified in terms of sensory experiences, sense data are not semantically autonomous (EPM, s. 12, 35-6; and see Brandom's study guide in the edition with the Rorty preface). They are in fact meaningless except by reference to the external world properties they represent. The alternative order of explanation offered by Sellars is that sense data discourse is akin to a scientific-theoretical vocabulary that is parasitic on observational practice and discourse in terms of an external world (see Brandom study guide, 162-5).
Ethical theory examples
Elizabeth Anderson's pragmatism about objective value (Anderson, Value in Ethics and Economics, ch. 5) also partakes in the pragmatist order of explanation. For Anderson, objectivity about value is constituted by objective practices of value inquiry, which subject value claims to potential disputation in a "space of reasons" (93-5). It is this practice that makes value claims susceptible to right and wrong answers; a value claim is objectively right if it is taken to be right in a sufficiently objective value inquiry practice.
Rawls's constructivism also privileges a pragmatist order of explanation. The principles he proposes, notably the principles of justice in A Theory of Justice, get their validity by being the product of a procedure that plays a suitable (ideal) role in social-political life. In later writings, Rawls explicitly argues against the reverse order of explanation. In Rawls, 1974, "The Independence of Moral Theory," he argues against pursuing moral theory as grounded in independent metaphysical or epistemological commitments. Instead, moral theory arguments should ultimately be based on distinctively ethical reasons. Even where formal theory informs our conceptions of possible ethical principles and well-ordered societies, the formal is not hierarchically prior to the ethical nor does it displace the methods characteristic of ethical theorizing (e.g. presumably the method of reflective equilibrium, 13).
In "Kantian Constructivism in Moral Theory" (1980), Rawls characterizes the enterprise of moral theory as asking not, "What freestanding moral truths must we govern our lives by?", but rather, "Is there an acceptable set of principles by which we could govern our lives, given what we take ourselves and our collective lives to be?" That is, we do not seek to discover an "independent order" of moral truths (519). Instead, we try to "discover and formulate ... deeper bases for agreement" within the ongoing practices of justification in the public culture of a liberal society (518). "Apart from the procedure of constructing the principles of justice, there are no moral facts" -- at least, not within the scope of the inquiry of justice as fairness (519).
As Korsgaard also later argues, Rawls's form of constructivism is autonomous in the sense that moral truths are truths by virtue of their roles in our practices of deliberation -- they get their authority from us (555-60, 569). Moral realisms like the rational intuitionism of Sidgwick and Moore are instead heteronomous, because they take us to be governed by a moral order that is "fixed" prior to our conceptions, as it were by something outside of us (compare Korsgaard on "substantive" vs. "procedural" realisms, Korsgaard, 1996, The Sources of Normativity, 35-7, and generally 33-47).