Last week I posted an article that summarized some of the progress in dynamic systems theory and cognitive science over the last 20 years. This article takes up that same topic but examines some of the theoretical debates around the use and validity of dynamic systems theory (a good overview of DNS can be found here).
I am only posting the "Introduction" below, so follow the links to read the whole article (pdf).
Fusella, PV. (2013). Dynamic Systems Theory in Cognitive Science: Major Elements, Applications, and Debates Surrounding a Revolutionary Meta-Theory. Dynamical Psychology. dynapsyc.org
Read the whole article (pdf).
Dynamic Systems Theory in Cognitive Science: Major Elements, Applications, and Debates Surrounding a Revolutionary Meta-Theory
Paul V. Fusella
January 15, 2013
“It is the theory that decides what we can observe.” -Albert Einstein
Dynamic Systems Theory (DST) is a broad theoretical framework imported from the physical sciences and used in psychology and cognitive science in the past several decades that provides an alternative to the computational and information-processing approach that has governed main stream cognitive science since the dawn of the cognitive revolution in the mid-twentieth century (Beer, 2000; van Gelder & Port, 1995; van Gelder, 1998; Spivey, 2007). DST views all psychological processes and capacities as dynamic systems which are best described as complex, non-linear, self-organizing and emergent and whereby cognition develops over the life course and occurs over real-time as a probable description of many possible alternatives instead of linear-assembly-of-symbolic-processes (Spivey, 2007; van Gelder & Port, 1995). Psychological capacities are viewed as emerging as more complex unique forms from prior simpler states, moving from chaotic to more stable trajectories in a theoretical state-space that culminate in the manifestation of a specific thought in real-time or a developmental phenomenon over ontogenesis (Spivey, 2006, Thelen & Smith, 1994; van Geert, 1998). There is a sensitivity to initial conditions and a determination by multiple causality, whereby psychological phenomena, be it a developmental capacity or cognition more generally, are softly-assembled (Thelen & Smith, 2003).
This overarching and revolutionary view for cognitive science has been in the works for quite some time perhaps since the cybernetics movement in the mid-20th century but has become more popular in recent years and has been referred to by a number of different and related labels reflecting related ideas and ranging from chaos theory to complexity theory to nonlinear dynamical systems theory. These titles all refer to similar ideas but have subtle and nuanced differences. I choose to use the term dynamic systems theory because this is the term used by most cognitive scientists who subscribe to this viewpoint and who refer to their movement as the dynamical view in cognitive science and refer to themselves as dynamicists so I will continue with that tradition although the lexicon and conceptual hallmarks used are shared by all these related viewpoints.
Specifically, what I mean by the DST approach in cognitive science (and later to what I refer to as the Complexity Theory (CT) approach) is something also related to work in theoretical computer science and artificial intelligence and empirical and theoretical work done there that applies to some distinct intelligent systems but particularly what it can say about the human mind as an intelligent system. Siegelmann (1997, 2003), Bringsford (2004), Kempis (1991), and Penrose (1990) have described the quintessential hallmark of the DST approach of this particular form of intelligent system as being trans-Turing (or super-Turing) and which possesses hyper-computational capabilities; that is brains and other certain forms of intelligent systems perform processes that go above and beyond the Turing-limit with it’s symbolic-serial processing of the traditional digital computer metaphor that has hallmarked much of the work done in cognitive science and which has been motivated by the information-processing (or computational) perspective. By computationalism I am referring to what has been the dominate theoretical framework in cognitive science since its inception, which has been motivated by the development of the digital computer and principally the work of Alan Turing and the Turing machine, and which uses as a metaphor for the mind, a symbolic-algorithmic-serial-processing digital-computer that computes at or below the Turing-limit.
The argument made by most dynamicists is that DST is a more suitable theoretical framework for situating psychological phenomena because it has achieved success in accounting for other phenomena in the natural world as diverse as meteorological phenomena to kinematics of the human body. The brain and mind are part of the natural world so logically they too can be accounted for by the dynamical view and perhaps more completely and accurately than the traditional computational and information-processing approach with its use of the digital computer as the metaphor for mind. The mind is an abstraction for the neurological underpinnings in the brain and these are not machines they are biological organs made up of cells and organic molecules and they are part of the natural world and could arguably be better accounted for by a meta-theory that has been successful in capturing the diverse natural phenomena that dynamic systems theory has been able to do. I am echoing the argument made by the dynamicists and arguing for a paradigm shift in the sense that Kuhn (1962) described, specifically in cognitive science, as a move away-from the computational, toward a dynamical theoretical framework and paradigm with an recognition that the computations made by the human mind are trans-Turing (or super-Turing) and go above and beyond the Turing-limit of the traditional information-processing approach.
Adopting DST reconciles a lot of the debates in cognitive science surrounding the phenomena that is investigated from monism vs. dualism, nativism vs. empiricism, and subjectivity vs. objectivity, not to mention the various anomalies discovered as a result of adopting a computational viewpoint. DST reorganizes the way that phenomena are studied and conceptualized; where some such as Varela, Thompson, and Rosch (1991) have argued for first person methods in the study of consciousness utilizing methods from traditional Buddhist psychology and which reflects a post-positivist view of how science is conducted. However, others including Spivey (2007), Beer (2000), Thelen and Smith (1994), van Gelder and Port (1995), and Schoner (2008) continue to work empirically with a positivist empirical framework in studying various psychological capacities from a DST perspective and these are the mainstream in the field. Thus, be it the manner in which phenomena are empirically investigated or the manner in which theories are constructed, DST is beginning to be accepted as a viable alternative to the 20th century traditions of computationalism and positivism. DST provides an account of cognitive phenomena that is dynamical, embodied, completely situated and ecologically-grounded and the ways that cognitive scientists go about conducting research and theory building is likely to be influenced by these fundamental aspects to this meta-theory.
In this paper, I will set out to provide an overview of the dynamical approach in cognitive science reviewing the more important work that has been done in recent decades and especially at the turn of the 21rst century. I will focus on a review of two recent debates that were published recently: one in 1998 in the journal Behavioral and Brain Sciences and the other in 2012 in TopiCS in Cognitive Science where contributing scientists debate the legitimacy of the dynamical view. During the review of these debates I will touch on the different facets of DST including work done in embodied cognition and ecologically grounded cognitive phenomena and also work that has been done in applying DST to experiments in psycholinguistics wherein the role of the body, the context, and the environment are united in one framework that is guided by DST.
During this review of the work in the application of DST in cognitive science and reviewing the various formal debates that showed up in peer reviewed journals and discussing them critically, I will echo what others have proclaimed in the past and make the argument that DST is ultimately a more suitable theoretical framework for guiding empirical research and theory building in cognitive science and should at some point in the not so distant future, but especially for moving the field forward in this new century, replace the outdated computational and information-processing approach which appears to have run its course. DST has much promise in providing an overarching and unifying theoretical framework for the cognitive sciences but like anything new it is met with staunch criticisms and rejection. However, the more people that join this movement the more the basic principles embodied in the DST approach will become grounded in empirical evidence. I will begin to conclude the review with a recap of the major controversies that adopting DST provokes from its criticizers and will finish with my modest vision for the future role DST can play in reorganizing the way that science is conducted in the cognitive sciences building off the work that has been done in dynamical cognition from the beginning and reacting against and incorporating within it the good that came from the traditional approaches in cognitive science into the future of what this revolutionary meta-theory means for cognitive science.