Emergent Constructivism Jedediah W. P. Allen and Mark H. Bickhard Lehigh University ABSTRACT—Passive versus active ontologies for modeling the nature of representation impose powerful constraints on the conceptual possibilities for the different versions of constructivism. The neoconstructivism outlined by N. S. Newcombe (2011) is convergent with an active, action- based approach to representation; however, it does not directly address the issue of representational emergence. If cognition is fundamentally emergent from (inter)action, then an emergent constructivist approach to development is necessary to fully transcend the limitations of the passive ontologies inherent to nativist and empiricist perspectives. KEYWORDS—emergence; constructivism; representation; nativism; empiricism; action In her article ‘‘What Is Neoconstructivism?’’ Nora Newcombe (2011) presents a neoconstructivist approach to developmental research. The stated tenets of neoconstructivism have both a descriptive aspect and a prescriptive aspect—describing con- temporary themes and shifts in developmental research, as well as advocating those themes and shifts. We endorse all of them, both descriptively and prescriptively, but would like to suggest that they do not go far enough in certain directions to be able to give a full prescriptive orientation. Our discussion can be developed from consideration of the core term constructivism. In a broad sense, constructivism can be understood as positing internal processes that create new inter- nal organization, not previously available, for future system func- tioning. This notion comes in several flavors, and the differences among them, so we suggest, make a difference—a very important difference. In particular, constructivism is an umbrella term that has fundamentally different meanings depending on underlying assumptions about the nature of representation. A basic distinc- tion can be drawn between passive and active ontologies for modeling the nature of representation. In their empiricist versions, passive models of representation derive from a general framework in which the world is assumed to ‘‘impress’’ itself into the mind. In-the-moment representation is constituted by encoding the world via transduction (Bickhard & Richie, 1983; Fodor, 1981; Fodor & Pylyshyn, 1981), whereas, longer term learning involves encodings that are acquired via induction, association, or statistical input processing (Bickhard, 2009b; Popper, 1968). Thus, for these models, there is some sense in which learning takes place, and so the idea that ‘‘something’’ is constructed begins to get some traction. However, passive encoding models of representation impose serious conceptual constraints such that the strongest sense of construc- tivism possible is construction out of already available (represen- tational) atoms—a combinatoric constructivism. Such a constructivism may, in addition, posit the ability to make use of prior constructions (i.e., combinatoric constructivism can be recursive), but these can never go beyond the combinatoric space generated by the foundational base of representational atoms. Importantly, nativist models of representation posit equally pas- sive ontologies. Whereas nativist models differ from empiricist models in that they posit a base set of representations available innately, instead of being ‘‘impressed’’ by the environment, they share the assumption of a fixed generative base set of representa- tional atoms, and are equally limited to the combinatoric space that they generate. In both cases, the mind is passive, either rela- tive to the environment or relative to the presumed innate repre- sentational base. Historically, neither approach has ever succeeded in accounting for representation (Bickhard, 2009b). 1 Correspondence concerning this article should be addressed to Mark H. Bickhard, Psychology Department, Lehigh University, 17 Memorial Drive E., Bethlehem, PA 18015; e-mail: mhb0@ lehigh.edu. ª 2011 The Authors Child Development Perspectives ª 2011 The Society for Research in Child Development DOI: 10.1111/j.1750-8606.2011.00178.x 1 It is easy to posit various empirical factual correspondences that might be claimed to constitute representation—causal, informational, lawful, struc- tural, so forth—but a model of how any of these could possibly account for the normativity of representation—of how representation could be true or false, how representation of falsehoods and nonexistents could occur, how representational content could be modeled—has always eluded such efforts (Bickhard, 2006, 2009a, 2009b, 2009c). In current practice, accounting for content for such empirical correspondences that are claimed to be represen- tational tends to be relegated to innatist assumptions. Volume 5, Number 3, 2011, Pages 164–165 CHILD DEVELOPMENT PERSPECTIVES