Abstract
Stephen Turner’s discussion of the cognitive and the social can be summed up with the following formula: the cognitive = the computational and the social = the Verstehen bubble. Do the cognitive and social intersect? That is a first approximation. Closer to the truth of what Turner argues, would have been to use the above formula with the following qualified rephrasing. On the standard views in cognitive science and the social sciences, respectively: the cognitive = the computational, and the social = the Verstehen bubble. Turner, though, actually prefers an alternative to both standard views. Before going there, I discuss Turner’s discussion of the standard views, and hence, why the sub-title, A Primer, is only partially appropriate. The sub-title, A Prolegomena, would have been more appropriate if Turner had focussed exclusively on the alternative viewpoint that he prefers and left the standard viewpoints out of the discussion. Indeed, one might wonder: why does Turner devote so much space to discussion of the standard views? My guess is that the purpose of the discussion of the standard views is to show that the basic question of whether the cognitive and social intersect must lead to a No answer for the standard views. Thus, on the basis of the standard views of the cognitive and the social, no unity is possible for the human sciences. But then, why care if there is disunity in the human sciences? Why care, if the cognitive, and for that matter the psychological, require models that are different from the model required for the social? For instance, the models required for Quantum Mechanics are different from the models required for gravity, and the models required for physics are different from the models required for evolutionary biology, and those evolutionary biological models are different from the models required for molecular biology. I think. So, what is wrong with a pluralism of models in both the natural and human sciences? I personally do not know.
A deeper and more remote background to Turner’s discussion of the present state of the cognitive and social sciences might be useful to get a grasp on what the tensions are in the human sciences. I think this exploration of a deeper more remote background will also provide a sharper context for appreciating the nascent and struggling alternative that Turner seeks to launch. This review will function as a Turner viewpoint launch party in the form of a discussion of the deep background of Turner’s viewpoint. How does the launching of Turner’s viewpoint prove beneficial to the further development of both the cognitive sciences and the social sciences? I answer at the end of this review.
We can go back to Aristotle for the very deep background to the connection between the individual and the social: for Aristotle, the human is both a rational and a social animal. We can jump forward to Spinoza and to Marx, where the individual is embedded in the social and then to Lev Vygotsky. For Vygotsky, institutions, in particular schools, provide scaffolding for the cognitive development of children. “. . . scaffolding can be best understood as a process made up of steps in which affordances are taken and lead to changes of capacity, which allow other affordances, which in turn change capacities, and so forth.” Vygotsky’s original example of schooling is a case in point—“each step enables further steps, but each step also requires the next affordance” (140). It is interesting that those in cognitive science/psychology have turned to Vygotsky as a precursor for their developmental theories of how children develop a theory of mind. Turner calls the theory of children developing a theory of mind, as is common I think, the “theory-theory.” Here, in the theory-theory, we have a natural link-up with the cognitive and the social, because the cognitive is built into the social. In the theory of children’s development of a theory of mind, if we trace its background to Vygotsky’s social view of the child, we have a deep background to the unity of the human sciences. Developmental psychological and cognitive studies of individuals are embedded in studies of society. We can even go back to the ancient Greeks where rationality is social rather than a zero-sum game among individuals, for this deep background to current cognitive and social studies. The question now becomes, when did the unity of the human sciences, the social and the cognitive or psychological, disintegrate? When did it become a puzzle: how to get from the individual mind to other minds; from the cognitive to the social? It was no puzzle for Aristotle, and later for Vygotsky, and the current descendants of Vygotsky. How did the tradition of the human as a social animal get interrupted? Ironically, Max Weber, the founder of the individualistic approach to sociology, and the developer of what Turner calls the Verstehen bubble, has an answer: the rise and diffusion of the Protestant Ethic. Can it be the case that the development of individuals tied to hard labor on machines and mass production created the disunity in the human sciences, as well as the fragmentation of the social and the individual? I do not think Turner is going that far, with regard to how the problematic nature of the cognitive (in the individual) versus the social, arose.
Rather, given the problematic of the individual (cognitive) versus the social, can there be a solution within its own terms of reference? The cognitive = the computational, and the social = the Verstehen bubble. Although the Verstehen bubble is a form of what Turner and others refer to as folk-psychology. The Verstehen bubble is individualistic in the cycle of beliefs, reasons, desires, goals, actions, and feedback. The problematic is due more to the computational side of the story regarding the individual. The computational goes on beneath consciousness and depends on algorithms. Although Turner, rather than using the term “algorithm” which is the core of the computational, talks about rule-governing that actually goes into the make-up of algorithmic functioning: “. . . all models of cognitive processes are computational models. The main rival to the standard model, connectionism is just as much a computational model as the more familiar rule-governed, program-like computational models of the standard approach” (56). Moreover, Turner could point out that the connectionist model also uses algorithms though the algorithms are bottom-up and make new algorithms, rather than top-down and determinate in the traditional serial computational model.
The new puzzle of the individual and social is not about how and where do the individual and social intersect. Rather the new puzzle of the individual and the social is how does the individual as assumed by the Weberian in the model of the Verstehen bubble and the individual as assumed in the model of computational cognitive psychology/science, connect? Weber’s estrangement of the individual from the social is actually a sociological question, part of Weberian social theory: the individual in the social world of the Protestant work-ethic is bound to machines, including social machines such as bureaucracies, and so “bowls alone” to use the metaphor of Robert Putnam (2000). (Notice that Putnam emphasizes “social capital”: the hard working individual only connects with the social in terms of using other individuals to develop social networks for climbing social ladders, which are composed of rational individuals playing zero-sum games with each other.) In other words, the descendants of Weber such as Robert Putnam (not discussed by Turner) turn the intersection of the individual and the social into the sociological problem of alienation and anomie. (Turner discusses Émile Durkheim who takes “social facts” as basic not derivative of “facts” stemming from the individual.) Perhaps, if instead of the Verstehen bubble, we started with the social facts of Durkheim, the social problem of the individual has not so much to do with individual psychology in folk-psychological terms, but with the dissolution of the social glue in modern urban society. But again, we do not have a methodological problem: of how the computational applies to the social. We have a sociological problem of how the social fails the individual in the modern world. This sociological problem is very different from the methodological problem. Now, we come back to where I started: how did the methodological problem come about sometime between Aristotle and Vygotsky? Both took the social as primary to understanding the individual. Even back to Plato’s Republic: the methodological principle is that social structure is individual psychology writ large. How did the methodological principle of explaining the individual in terms of the social break down?
I have been drawing a holistic perspective for viewing Turner’s book as dedicated to the solution, or at least, outline of a meta-methodological problem. There is a methodology for the computational model of the individual that has been built up, piece by piece. There is a standard family of methodologies for the social, that at least are cousins to each other, but the family of social methodologies has little connection with the computational methodologies. Though some cognitive scientists are working on mirror neurons where the brain causes individuals to imitate each other, not for flattery, but as an automatic cognitive response that enables learning, and the unconscious or tacit passing on of traditions. Indeed, the program of mirror neuron research pokes into the Verstehen bubble (without bursting it): “Joint attention makes the action ‘social’ in Weber’s sense because joint attention is inherently other-regarding. Moreover it directly links to neuronal processes—joint attention in primates involves mirror neurons” (187). In any case, the meta-methodological problem is Can we make connections among the methodologies for the computational (or cognitive) and the social (other-regarding)? Turner is nothing but complete and exhaustive in discussing the literature around this issue: raising questions, reporting answers, and proposing tie-ins within the standard methodologies to show lines of connection. In specific, the mirror neuron studies he reports look hopeful for the standard model.
However, Turner proposes an alternative methodology to the cognitive and the social that he thinks by-passes the somewhat shaky ground of the standard methodologies. Here, it is in his own nutshell description: We can go back to the notion of affordances, and the recognition that cognition occurs in a world of physical objects which individuals use or respond to . . . The fact that this is also a shared world . . . And there of course other shared objects—texts, documents, works of art, and so forth. (188)
Do we include institutions and social facts as part of the shared objects? If so, we have a unitary methodology for the human sciences: one that looks at the physical and the social as providing affordances for human activity. But still, it is not clear how the cognitive works. Turner does give us a rough outline. I briefly discuss Turner’s approach to the cognitive in the following.
Turner rejects the two main computational models of cognition—top-down (resembling innateness where no learning occurs), bottom-up (resembling induction, where learning is a logically invalid process). However, Turner turns to an apparently noncomputational model that he avowedly borrows from Andy Clark, that Clark calls “predictive coding.” He quotes Clark favorably from an online article called “Predictive Coding,” https://www.edge.org/response-detail/10404 (accessed on May 29, 2019). In Turner’s own words, it is trial and error learning that uses “. . . expectations, not a panoply of pre-existing concepts . . . learning . . . takes the form of reducing signal noise, or irrelevance, in the buzzing, blooming world” (63). What Turner says about both innateness and induction versus predictive coding or how we learn by using our expectations to filter out noise, is reminiscent of Karl Popper (1965): The theory of inborn ideas is absurd, I think; but every organism has inborn reactions or responses; and among them, responses adapted to impending events. These responses we may describe as “expectations” without implying that these “expectations” are conscious . . . an inborn expectation, no matter how strong and specific, may be mistaken. (47)
Popper, according to John Wettersten and William Berkson (1984), first learned these ideas about the psychology of learning from the pioneer of cognitive psychology, Otto Selz (whose life and continued research became a victim of the Holocaust).
This neo-Selzian alternative approach in cognitive science applied to social science seeks how individuals develop expectations about the social: about all the varieties of institutions, rituals, rites, and religions. The neo-Selzian alternative approach also seeks out expectations that develop in our learning within all varieties of institutions, and in families, and in other social structures. Here, we have an intersection of the cognitive and the social. One might say that the cognitive mostly occurs within the social, the infant within the family; the child within the school. So, do we return to the pre-computationalist, such as Vygotsky? Do we return, even to Plato, where individual psychology recapitulates the social? Let me end this review by turning the question with which I began on its head: do we have to explain how cognition develops by applying social studies to how individuals learn in social settings? The meta-methodological question, by going in the reverse direction from the social to the cognitive, becomes the following: Can we develop methodologies strictly for the social where humans are seen as developing their individuality within and often in contrast, if not opposition to the social? Turner should have the last word on the social, for a pluralistic and manifold approach to the social, and therein to how individuals and the social shape each other: “We can think of actual societies as made up of multiple focal points which are the subject of joint attention by different overlapping groups, as the distributed rather than centralized source of multiple modes of coordination” (209).