The primary aim of the fellowship is a conceptual analysis of recent research that traverses the evolution-cognition-culture boundary. Specifically, I will relate an evolutionary model of human group configurations and their related social cognitive processes (Caporael, 1997, 2004; Brewer & Caporael, in press) to recent work in multiple inheritance systems (Griesemer, 2005) and generative entrenchment (Wimsatt, 2003; Wimsatt and Griesemer, in press). I propose that such work illustrates the emergence of a New Kind of Theory (NKT) in the human sciences, one that contrasts with the familiar single factor linear causal model typical of cognitive theories.

Like many other scholars in the mid-1980s and later, Davydd Greenwood (1984) argued that, when it came to humans, sociobiology was hardly the groundbreaking achievement that it claimed to be. Darwinism was “tamed” by assimilating it to earlier, pre-Darwinian views about human nature, souls, progress, and the practice of science as a reductionist enterprise (Caporael, 1994; Latour & Strum, 1986; Midgely, 1987; Nelkin, 2001. See also Freyd, 1983 for a discussion of mechanism). A significant consequence is that theorizing evolution, mind, and society occurs in historically distinct, sometimes competitive, disciplines united by a taken-for-granted conceptual framework that is more a product of shared cultural background than of scientific logic (Caporael, Dawes, Orbell & Van de Kragt, 1989; Markus, Kitayama & Heiman, 1996). It cannot deliver the revolutionary promise to shed more light on psychology because the gene’s eye view already presupposes knowledge of the phenotype. Its main features include the notion that humans are more or less rational, self-interested individualists; mind is analyzable as a disembodied abstraction, and society is a uniform aggregate of individuals, or at most, dyadic decision makers. The gene’s eye view of neo-Darwinism was easily invoked to “explain” these features; what it could not do is overturn or challenge them. For that, we need a new kind of theory, one that facilitates transversals across levels, models, and perspectives (Griesemer, in press).
The issues of using the gene’s eye view for understanding human cognition and sociality are the same as those described by Buss (1987) for preferring a hierarchical evolutionary perspective over genic selectionism for the evolution of multicellularity. The phenomena studied from a hierarchical perspective can be framed as gene-level selfish innovations, but that leaves evolutionists in the role of bookkeeper offering post hoc analyses about why one or another feature was better than alternatives. For example, few one would deny today that humans have evolved in groups. However, the gene’s eye view is at a level of analysis that cannot distinguish between social and asocial organisms. Grouping itself and downward causal effects on individuals are simply set aside, usually as misguided errors of group selection (Gaulin & McBurney, 2001). Consequently, “social intelligence” is equated with Machiavellianism and strategic competition, deception, and exploitation (Byrne & Whitten, 1998). Only the hierarchical view can lead to novel predictions and to the organization of a broader range of previously unrelated observations. Indeed, the gene’s eye view typically gives an evolutionary gloss to common folk psychology (not surprising given that the gene’s eye view itself is an anthropomorphic heuristic), while ignoring a wealth of counter-intuitive, yet-to-be organized, body of experimental work on social cognitive phenomena such as labile ingroup favoritism, errors attributing causality, and automaticity of behavior. 
In contrast, a multilevel selection perspective can organize a broad range of existing phenomena and suggest novel predictions. Caporael (1997, 2001, 2003) suggested a model for the topography of the selective environment for humans based on a consideration of tasks that are necessary for survival and reproduction and on research about group size (see Appendix A). The model consists of four face-to-face subgroup configurations—dyad, task group, deme (or band), and macrodeme (or macroband) organized as a nested hierarchy, or demic structure (Hull, 1988). A core configuration is the joint function of group size and activity. Configurations provide a context for tasks or activities that are specific to that level of organization. That is, each group configuration affords functional possibilities and coordination problems that do not exist at other levels. From an evolutionary perspective, the relevance of tasks is not the activity per se, but rather the set of social and cognitive processes that enable the activity. 
Humans can be selfish and exploitive, as the gene’s eye view suggests, but phenotypic self-interest is constrained because humans are also obligately interdependent, unable to survive and reproduce without a group. While humans themselves may daily face sociobiology’s “central problem” of altruism versus self-interest, the central scientific problem is understanding the evolution of coordination at any level of analysis whatsoever (Boinski & Garber, 2000; Brewer & Caporael, in press). The synergisms and conflicts of multiple levels of selection are literally written into the human body as bipedalism, omnivory, a long developmental period and dependency on collective knowledge. The recurrence of these and other features, from generation to generation, in ontogeny, in daily life, and arguably in human evolutionary history reveals the real missing links between theories of evolution and mind. Such recurrence, over different scales of time, is the essence of evo-devo. 
Proposed Research: Iterations, patterns, temporality, levels of organization, templates, scaffolds, constraints, scales and cycles, shifting contexts and changing perspectives are counterpoint terms for parsing observations in the world without demanding that they be reduced to fundamental elements or particles. Lovelace, Darwin, Baldwin, Turing, and more recently Simon, Mandelbrot, Buss, Wolfram—and readers can add others to this list—have led the way to playing with different perspectives on complexity and simplicity: the sciences of the artificial, fractals, evo-devo, and “New Kind of Science.” My fellowship objective is to continue this kind of play, and without making any intellectual commitments, take off from Wolfram’s euphonious phrase, and refer to a New Kind of Theory (NKT) to describe efforts to recast the human sciences in a multilevel mould. Griesemer (in press) correctly assumes that NKT will not be just a merger of older classical theories, which assumed that fundamental entities, such as cognition and culture, were separable. 
I intend to undertake a conceptual analysis of recent work that traverses (Griesemer, in press) evolution-mind-culture boundaries, with a focus on work that incorporates co-evolutionary and/or evo-devo notions (e.g., Jablonka & Lamb, 2005; Laland, Odling-Smee & Feldman, 2000; Richerson & Boyd, 2005; Geary, 200x). The aim is to identify what comparative dimensions are used in such work, and what dimensions might be worth adding. The main vehicle of the project will traverse levels of analysis between my research on core group configurations and two other perspectives on humans and evolution. Griesemer’s (2005) “reproducer perspective” concerns multiple inheritance systems, of which genes may form a possible part. Wimsatt’s “generative entrenchment” perspective concerns processes of stasis and change, including constraints on change that are an incidental—but hugely significant—to later events, be they human developmental, the design of artifacts, or institutional or cultural change. These three perspectives share a resonance in their use of counterpoint terms. Each perspective on its own is novel, and each would be enhanced by drawing on the others.
The reproducer perspective is a general evolutionary account of units (of which genes are one example) that recur in a genealogical relationship generation-to-generation. The reproducer perspective both draws upon and is a contribution to multi-level selection theory (Szathmáry & Maynard Smith, 1995).) The relationship between parent-offspring units is not merely resemblance, but also one of material overlap. That is, offspring units (whether genes, cells, organisms, or social groups) are made from physical parts of the parents. Offspring are organized so as to have the capacity to develop—which is the capacity to acquire the capacity to reproduce. The core configuration model also draws on multi-level selection, but as a source of a vocabulary for the “repeated assembly” of entities including those at a psychological and social level of analysis, and not as a general account (Caporael, 2003). Repeated assembly enabled psychologists to talk about recurrence, which is the attraction of an evolutionary perspective, without having to make commitments to specifically causal genes or other dichotomies in nature nurture dualism. Individual behavior is the material overlap that connections genes and culture, and core configurations are part of the capacity for development. Generative entrenchment (Wimsatt, 2001, in press) is the recursive developmental addition to the Darwinian principles of variation-selection-retention. It refers to biological, cognitive or cultural structures that are generated (reproduced or repeatedly assembled) over time such that they are part of a developmental sequence. Such parts of structures have downstream effects; the greater the downstream effects, the more deeply entrenched is the part. At a cultural level, deeply entrenched elements will be harder to shift than more shallowly entrenched parts. 
The intellectual significance of the fellowship work lies in describing what comparative dimensions might be useful across somewhat different NKT projects. It also contributes a more complete theoretical framework of embodied social cognitive processes that can fill the gap between genes and culture in coevolutionary theories. The expected product will be a paper on the results of the research and a proposal for a special issue to Biological Theory.