Joseph Rouse's How Scientific Practices Matter constitutes a radical move towards what John McDowell might call the re-enchantment of nature. Rouse refuses any systematic distinction between the natural and the normative, or between the space of reasons and the space of causes. He seeks a sustainable naturalism, based on two principles: rejecting appeals to supernatural forces or objects, and resisting arbitrary philosophical impositions upon science as it is actually practiced. He wants to reclaim naturalism from those who equate it with any form of reductionism that seeks to explain normativity, meaning, intentionality, the mental, or kindred concepts in terms of natural objects or processes assumed to be intrinsically free of such things. But according to Rouse, even those recent philosophers who have actively resisted such reductionism, and have argued instead for the essential place of the normative in nature (prominently including John McDowell, Robert Brandom, and John Haugeland) have retained vestiges of a pernicious dualism between the natural and normative domains.

The first five chapters of the book are predominantly historical and critical. They examine and synthesize several major strands and figures of twentieth century philosophy, including phenomenology, logical positivism, Quinean naturalism, interpretivism (represented by Davidson, Dennett, and Rorty), social science studies, and what he calls ‘feminist science studies’ (represented in particular by Karen Barad and Donna Haraway). With deft erudition, Rouse discloses twentieth century philosophy as a shared history of various manifestations of the dualism of norms and nature, together with ongoing attempts to heal or reduce away this dualism. These manifestations include systematic oppositions between the social and the natural sciences, a priori necessity and empirical contingency, first and second nature, and intentionality and causality, among others. According to Rouse, all of these versions of the dualism are doomed in virtue of foreclosing the possibility of the intelligible reconciliation or co-habitation of the two sides. “Once the normative and the causal dimensions of the world have been conceived incommensurably as self-enclosed domains,” he writes, “they can never be effectively brought back together” (183). But they cannot be left severed from one another without reconciliation either, he argues, at least because the natural world that our knowledge practices disclose must make enough normative contact with those very practices to enable them to be accountable to that world. Thus “causality must be understood as always already normative, and normativity as always already causally efficacious” (183).

The second half of the book—chapters 6 through 9—tries to make good on this programmatic mandate with a wide-ranging constructive analysis that purports to give accounts of nature, inquiry, embodied agency, modality, and normative authority. Though he ultimately rejects the distinction between the normative and the natural, chapters 6 and 7 roughly concern the materiality, causal efficacy, or ‘naturalness’ of norms, while chapters 8 and 9 take up the essential normativity of nature and causal laws. In this half of the book, Rouse's primary interlocutors are practitioners of two strands of contemporary analytic pragmatism. The first could be called “Pittsburgh School Pragmatism,” represented paradigmatically by Wilfrid Sellars, John McDowell, Robert Brandom, and John Haugeland, and rooted in anti-reductionist and anti-representationalist commitments in the philosophy of mind and language. The second could be called “Science Studies Pragmatism,” represented by figures such as Arthur Fine, Nancy Cartwright, and Peter Galison, and characterized by a focus on the local and concrete practices of science in lieu of a search for unified, aprioristic theories of scientific knowledge. Rouse takes the lessons and commitments of both types of pragmatism very seriously. Indeed, to his credit, Rouse's general method is to work through precise disagreements with figures and texts with which he is overwhelmingly sympathetic.

Rouse's positive position is multifaceted, but the main thesis is that both inquiring subjects and the objects of inquiry are ontologically derivative upon phenomena of ‘intra-action’ between the two. The term of art, borrowed from Karen Barad, is intended to avoid the implication that either party is determinate prior to or independently of their process of interacting with one another. Rouse argues that concrete disclosive intra-actions—including traditional concrete epistemic practices such as inquiring, measuring, and observing, but also abstract practices of thinking and reflecting, along with less obviously epistemic practices such as skilled coping—themselves (provisionally and dynamically) constitute the boundaries and nature of both inquiring agents and the situation and objects disclosed. They do so by making a ‘cut’ between inquirer and object, neither of which has a prior fixed character, based on the local shape of their relations of concern, normative accountability, and practical engagement. This cut is not only normative but also material, for practices are material processes involving causal interactions with their objects and situations. As Rouse puts it, “practices constitute objects-in-phenomena by configuring a setting within which the boundaries between one object and another make a difference, and are answerable to the boundaries they make” (289). The idea is that the natural world is inherently normative, in the sense that it is only insofar as it matters how parts of the natural world are and behave that these parts can be said to have a determinate character at all. To be something in particular is for there to be something ‘at stake,’ in an indefinitely large and contextually specific set of ways, in how things turn out with respect to that thing. This is not to say that we bring objects into being by granting them significance through our practices. On the contrary, Rouse argues that the idea that objects are granted their normative status by us relies on an unacceptable voluntarism that undercuts the possibility of real normative accountability to how things are.

This book should be essential reading for those who are dialectically engaged with either Pittsburgh School Pragmatism or Science Studies Pragmatism, as well as for scholars interested in the history of twentieth century philosophy, and it constitutes an important contribution to metaphysics, epistemology, and philosophy of science more generally. Rouse's command of philosophical literature is broad and subtle, and he synthesizes diverse discourses respectfully and convincingly, sensitive to both their continuities and their differences.

Rouse's discussion of Robert Brandom's inferentialist semantics in Chapter 6 is particularly satisfying; there, he argues that despite his own explicit pragmatism, Brandom fails to take sufficiently seriously the way that our discursive and inferential practices are, in the first instance, concrete bodily practices that always engage our perceptual and practical skills. Indeed, many philosophers of yet a third pragmatist stripe have lately turned to embodied practice as a privileged site of epistemic activity, where embodied practice is contrasted with the abstract internal manipulation of representations. But whereas most such philosophers insist on the primacy of embodied practice as opposed to discursive practice, Rouse carefully understands discursive practice as embodied practice. While the recent emphasis on embodiment has provided important philosophical correctives and insights, far too often embodied practice serves as a philosophical black box used to defer rather than to explain crucial epistemological problems. In contrast, Rouse makes good on the popular move to the body. He is much more informed about science and careful with philosophical texts than most proponents of embodied knowledge, and he doesn't simply posit a bunch of embodied practices in the background and leave their claims to epistemic authority unchallenged. Instead he carefully shows how the normative structure of all knowledge practices, including those that are highly discursive and abstract, can only be understood once we take their embodied character seriously.

Rouse's challenging and ambitious book does not lend itself easily to a pithy response, but I will gesture quickly at two dimensions of his account that worry me. First, a recurring theme of the book is Rouse's rejection of ‘epistemology’ in favor of ‘ontology.’ What he actually rejects is epistemology quite narrowly construed, namely the pursuit of conditions of certainty and theories of justification, abstracted from the metaphysics of the knower and the known, and premised upon the presumption that “‘knowledge’ demarcates a coherent, surveyable domain of inquiry” (140). One could easily read Rouse, not as rejecting epistemology, but instead as pursuing a broadened epistemology that includes all thought about how subjects try to disclose truth, and take or abdicate responsibility for their successes and failures at doing so. Clearly Rouse's decision to reserve the term ‘epistemology’ for the narrower project is a strategic one, but I am concerned that by bringing the entire contrast class under the rubric of ‘ontology,’ he risks rhetorically reinforcing the dualistic separation of epistemic and ontological inquiry that facilitated the impoverished version of epistemology to start with, and that his own work should help overcome.

Second, for Rouse, our normative practices play an essential role in the constitution of determinate material phenomena: the very “ebb and flow of tides…depend[s] on the emergence of discursive norms” (291). But this constitutive dependence cannot be a temporal dependence, for it is important that it turn out that historical inquiry can disclose a determinate past consisting of objects and phenomena that have the proper kind of causal links to our current situation. And such inquiry must now disclose a world of objects that, though we now have stakes in their character, managed, prior to any inquiry, to exist independently of such stakes. Rouse's own naturalism requires him to preserve these intuitions, as he recognizes. And indeed, he expressly rejects the hopeless idea that inquiry creates determinacy out of some kind of formless, unschematized content or matter. But his temporal rhetoric makes it difficult to see how he will make good on these commitments. Rouse regularly talks about determinacy ‘emerging’ in the context of practice, and of phenomena ‘becoming’ determinate in intra-action with agents. He rejects the idea that “the world somehow already comes naturally composed of discrete objects” (313). Yet it seems to me that this is just what historical inquiry does and must reveal, even if we have to be careful not to interpret this as a transcendent, foundational truth. He will need to find a subtler temporal language for talking about how our practices now can both constitute and correctly disclose a past with appropriate naturalistic connections to the present.