The Review of Metaphysics

The overall aim of Rockmore's book, announced on the first page, is to defend "epistemic constructivism." He understands this as a position comprising two claims: first, that "we do not and cannot know the real" and second, that "we know only what we construct." This view is seen as alternative to what Rockmore calls the "standard approach," according to which "to know is to cognize the real, reality, or the world." That approach, as Rockmore asserts and tries to demonstrate, has been and remains "dominant" in Western philosophy since Parmenides, in whose claims he suggests it originates. If Parmenides' original claims—including especially the controversial fragment 3, which appears to assert (on one reading) the identity of thinking and being—could be vindicated, "this would at long last demonstrate the approach to cognition as knowing the real." But, Rockmore argues, they cannot, leaving us with the only alternatives that he sees as remaining: either an overarching skepticism, to be rejected, or the constructivist alternative he favors.

The narrative of the book proceeds largely by means of high-altitude summary of the historical contexts and views of a large number of thinkers from the ancient, early modern, and Kantian and post-Kantian periods, as well as twentieth-century and contemporary Continental and analytic philosophers. While chapters 2, 3, and 4 aim to document the centrality of the "standard" nonconstructivist approach in (respectively) Plato, early modern rationalism, and Lockean empiricism, chapters 5 through 7 document what Rockmore sees as the emergence in the tradition of the "constructivist" alternative, from roots in Bacon, Hobbes, and Vico to the historical watermark represented by Kant's "Copernican Revolution" and (what Rockmore sees as) the further positive development of the constructivist project in Fichte, Hegel, and Marx. Chapter 8 turns to twentieth-century contributors to the project, including especially those describing themselves as pragmatists, and chapter 9 turns to (very) recent positions in Continental and analytic philosophy that appear relevant to the debate, including "new" and "speculative" realisms, Markus Gabriel's ontology of sense, and Irad Kimhi's recent and radical [End Page 827] argument for an anti-Fregean conception of force and content from a position committed (in some sense) to the originally Parmenidean identity of thought and being.

Many of the issues concerning knowledge and ontology that Rockmore treats are, as he acknowledges, more standardly treated in terms of the opposition of "realism" and "idealism." Rockmore rejects, however, both the exclusiveness and indeed the usefulness of that distinction, holding (for example) that Plato is in a relevant sense both an "idealist" and a "realist" and indeed that "all theories of knowledge are realist; none are anti-realist." What is apparently more decisive in producing the constructivist position that Rockmore favors is his rejection of the diverse forms of "representationalism" he finds in Plato, the early moderns, Carnap, and the early Wittgenstein, among others. As Rockmore treats them, these views require "subjective" representational intermediaries such as ideas, sense-data, or protocol sentences as well as the possibility of inferring from these to objects or objective knowledge. Citing Berkeley's historical arguments, echoed by Kant, against the Lockean primary/secondary quality distinction, Rockmore concludes that any argument requiring such an inference from the subjective to the objective is irremediably circular and that, more generally, "[t]here is no way to show … that ideas, representations, or other cognitive intermediaries between subject and object in fact represent." This makes way for the constructivist alternative that Rockmore sees as most centrally developed in the post-Kantian tradition of Fichte and Hegel. Here (for example in Fichte), the term "representation" (or "presentation") refers no longer to the intermediary of a mind-independent object but, rather, to the "object for us." Accordingly, for the subject that is active in the construction of its own concepts on the basis of experience, the only possible knowledge is knowledge of concepts and cognitive objects that are "mind-internal," in this sense.

Despite the vast...