Stemmler’s doctoral dissertation : “Just do it! Guilt as a moral intuition to cooperate–A parallel constraint satisfaction approach,” (post Simultaneous Feeling and Deciding) also made me aware of the concept of grounded cognition. Stemmler indicates that his findings are mainly in line with current approaches of grounded cognition. Models of grounded cognition assume that cognitive processing and conceptual knowledge is grounded in the perceptual and action systems. According to Stemmler in these approaches, the experience of an emotion is based on the way a situation is temporarily conceptualized or categorized. Conceptualizations of emotions represent abstract conceptual constructs that aggregate information from different perceptual and action systems. Since these approaches assume that conceptual knowledge is stored with relation to other information which was co-activated within a current situation, the situations itself can activate knowledge, similar to the assumption of network-models of emotion. For instance, situations may be accompanied by memory-retrieval of similar situations which then have to be adapted to the current situation. Hence, present and past situational information is integrated in a coherent fashion. Situational information, memories from the past and current affective feelings can be integrated to form a meaningful gestalt.
Stemmler states that grounded cognition models imply that information processing within a situation is directed by mechanisms of pattern completion, hence how people “decide” may be due to action-tendencies of situational information able to render an incoherent pattern coherent. Since not all information may be able to “fill the blank” within a temporal representation, conceptualizations from the past may be retrieved.
Lawrence W. Barsalou is one of the most referenced researchers on this idea (“Grounded Cognition”, Annual Review of Psychology, 2008, 59: 617-45.). He notes that for the past 50 years, theories have assumed that the brain is an information processing device somewhat similar to a computer, and that the information processing properties of the brain (cognition) are separate from the brain’s systems for perception, action, and introspection (self-thought, emotion, metacognition, etc.). Barsalou suggests that these new theories propose instead that the cognition is deeply grounded in these systems, relying heavily on simulations of perception and action, being grounded in the body, and being situated in the environment. The significant change in the past 10 years is the explosion of new empirical findings that support this approach, plus more sophisticated theories that explain them.
Barsalou provides this example. As an experience occurs (e.g., easing into a chair), the brain captures states across the modalities and integrates them with a multimodal representation stored in memory (e.g., how a chair looks and feels, the action of sitting, introspections of comfort and relaxation). Later, when knowledge is needed to represent a category (e.g., chair), multimodal representations captured during experiences with its instances are reactivated to simulate how the brain represented perception, action, and introspection associated with it. According to this account, a diverse collection of simulation mechanisms, sharing a common representational system, supports the spectrum of cognitive activities. The presence of simulation mechanisms across diverse cognitive processes suggests that simulation provides a core form of computation in the brain.
For me the idea shows cognition not occurring by juggling symbols in the brain like a computer, but cognition being simulations that actually reach out to the information gathering senses. This idea of simulation is in line with the concept of parallel constraint satisfaction. We keep running iterations until coherence is maximized. To do this, we may have to decide that the visual information needs to be tweaked a bit or that the higher level brain predictions need to be changed.