This article proposes a new theory to describe representation and dynamics in cognitive science. It begins with a very brief overview of the history of cognitive science and the predominant theories of cognitive systems: symbolicism, which relies upon the metaphor of mind as computer; connectionism, which relies on more “brain-style” models of processing, conceptualizing computation and representation as occurring through connected networks and nodes; and dynamicism, which argues that the mind is a physical, dynamic system whose state changes over time and whose functions cannot be viewed as discrete or static. According to the article, early accounts of cognitive systems, driven by behaviorism, were limited to the observable end products of thought (limited to inputs and outputs), whereas in the mid-1950s - with the aid of technological advances that produced computers and facilitated a working metaphor for cognitive processes - researchers began looking at internal states, internal processes, and internal representations to understand cognitive processes. As Eliasmith puts it, people began to peer “inside the black box.” The article goes on to suggest the metaphors provided through symbolic, connectionist, and dynamic accounts of cognitive processing, while all helpful analogies, may nevertheless be constraining how cognitive processes are understood. The author proposes a new theory, representation and dynamics in neural systems (R& D theory), inspired both by modern control theory and recent findings from neuroscience. The author suggests modern control theory, which considers internal system state variables in the processing of inputs and generation of outputs, “offers tools better-suited than computational theory for understanding biological systems as fundamentally physical, dynamic systems operating in changing, uncertain environments” (p.6). R& D theory is governed by three main principles: (1) that neural representations are the result of non-linear encoding and weighted linear decoding; (2) decoding is the result of transformations of neural representations carried out by connected populations of neurons, using linear decoding; and (3) “neural dynamics are characterized by considering neural representations as control theoretic state variables. Thus, the dynamics of neurobiological systems can be analyzed using control theory.” The remaining sections of the article provide support for the theory, and presents arguments for the utility of R & D theory above and beyond existing theories for explaining cognitive systems.
Given my limited knowledge thus far of existing theories of cognitive systems, this article in general was quite difficult for me to follow. While it provided a general account of existing theories, numerous references to key concepts relevant to the theories, about which I am unfamiliar, made it difficult at times to fully grasp some of the examples and explanations. I feel this article will be a good one to return to later down the road. Nevertheless, it was encouraging to read accounts of cognitive processing that line up with what I have learned so far about affective processing; specifically, that affect and cognition is best understood as the result of dynamic interactions between different levels of processing by interconnected brain regions. I look forward to re-reading this article when I have had a chance to digest a bit more of the cognitive literature.
2) Treat, T.A., & Dirks, M.A. (2007). Bridging clinical and cognitive science. In T.A. Treat, R.R. Bootzin, and T.B. Baker (Eds), Psychological clinical science: Papers in honor of Richard McFall.
In this article, Treat & Dirks extol the virtues of taking an interdisciplinary approach towards understanding the role of cognition in psychopathology. They discuss the limits of applying research paradigms from cognitive science directly towards answering clinical questions, and the shortcomings of staying within disciplinary boundaries when trying to extend cognitive science to clinical problems, such as through the use of cognitive therapy. They propose a new, integrated approach called “quantitative clinical-cognitive science.” The article goes on to summarize recent studies using this approach, in which the perceptual organization of two clinical populations is investigated. The studies cited each investigated specific differences in the processing of visual information, exploring whether processing was influenced in a disorder-specific manner. For example, using photographs of normatively heavy and normatively thin women with either sad or happy facial expressions, participants were ask to rate two photographs according to how similar they were. Participants who had endorsed bulimic symptoms consistently rated similarity according to body type, disregarding affective information. The same held true in a memory recognition task, where participants endorsing clinically significant eating disorder symptoms had better memory for body shape than affective information. Similar disorder-specific perceptual processing was found in a sample of college men who perceived unwanted sexual advances to be justified. When shown photographs of women who were or were not wearing revealing clothing and varied in facial expressions of affect, participants judged photographs as similar based upon dress and not affect, and demonstrated greater memory for information about physical appearance than facial expressions. The article suggests that attention should be paid to specific types of cognitive processing biases, and that perhaps by understanding disorder or individual-specific deficits in the processing of information we can develop cognitive therapies that more directly target these deficits. Further, the article suggests that research paradigms pay closer attention to tapping into these individual differences in processing when studying clinical phenomena. This suggestion is not only sound, but opens up a new and important approach to investigating how distorted cognitions might fuel psychopathology. For example, it is increasingly apparent that the way in which information is taken in and interpreted can affect both affective and behavioral responding. By understanding how information is being distorted in psychopathology, his way might enable more finely-tuned cognitive therapies that target core processes, rather than general deficits.
3) Thagard, P. (2005). Being interdisciplinary: Trading zones in cognitive science. In S.J. Derry, C.D. Schunn & M. A. Gernsbacher (Eds.), Interdisciplinary collaboration: An emerging cognitive science (pp. 317-339).
This article in general extols the virtues of cognitive science as an interdisciplinary science, recounting the history of both people and places integral to its formation. Thagard discusses how bringing together philosophers, linguists, psychologists, and computer scientists has contributed towards expanding ideas and conceptualizations about how the mind works. He also discusses how the field of cognitive science could not have evolved without the support of forward thinking institutions that allowed the such things as Carnegie Mellon’s early support of a joint appointment in psychology and computer science, as well as joint degrees. He notes early work of many of cognitive science’s “pioneers,” such as Noam Chomsky’s work on linguistics, George Miller’s “The Magical Number Seven, Plus or Minus Two,” and Marvin Minsky, Allen Newell, and Herbert Simon’s work in artificial intelligence. Thagard summarizes by saying the success of cognitive science lies with the establishment of these “fertile trading zones,” in which various disciplines and institutions have come together to share ideas and inform each other. It is through these collaborations that knowledge has flourished. In essence, it appears cognitive science provides a good example of how sharing knowledge across disciplinary boundaries can serve to advance our knowledge well beyond what is gleaned by sticking to our borders, a lesson clinical psychology would benefit from learning, and has begun to do so with growing collaborations between clinical psychology, neuroscience, and social cognition.