The Self-Organization of Intentional Action
Type de matériel :
1
Earlier I proposed that top-down interlevel causality should be reconceptualized as the operation of second-order contextual constraints. Following that line of reasoning, I now conclude that intentional action consists in a trajectory whereby emergent, second-order contextual constraints of conscious, dynamical brain organization selectively constrain lower-dimensional motor and speech processes. And these constraints do so in virtue of the meaningful and conative properties of the intentions that they embody. Constrained by such top-down commands, individual motor and speech processes are regulated such that the resulting behavior preserves the dynamic organization that embodies the intention's content. As is the case with all structured structuring structures, intentions are thus causally effective by constraining (entraining) their components in a manner homologous to the way in which the virtual governor of a beam of coherent laser light synchronizes or slaves the random waves of the individual atoms that constitute it: by altering their probability distribution - their rate of firing, for example. Proximate intentions are the changes in the motor neurons' natural firing frequency. Ontologically, behavior that is the top-down projection of self-organized semantic constraints onto lower-level motor or speech processes constitutes an act-token. As such, intentions can function as the action's formal and final cause. From the perspective of complex dynamics, proximate intentions (do x now) are therefore not separate blueprints or templates to which the behavior refers; nor are they represented as (by?) a symbol. Much less are intentions external causes operating like Newtonian forces pushing the (rest of the) body into action. Functioning as second-order contextual constraints, proximate intentions are dynamical attractors within a prior intention's self-organized semantic space. They drive behavior in a particular direction because they are changes in the frequency distribution of entrained motor processes. It is therefore more accurate to think of proximate intentions as a dynamical system's operator (EDELMAN, 1987), an endogenous control knob that takes the network from one location in state space to another. Both prior and proximate intentions embody top-down, contextsensitive constraints which restrict behavior to within certain regions of semantic space. As such they are ongoing formal causes. Because the contextual constraints of attractors within semantic space embody act-types, individual behavioral trajectories executing proximate intentions constitute act-tokens. One must keep in mind, however, that a particular proximate intention is never the only existing attractor, as we know from studies on akrasia. Nor is it always the one that wins out -although, by definition, it usually does. From a dynamical systems perspective, however, individual act-tokens are unbroken trajectories within an intention's semantic space as its dynamics project and cascade - without equivocation - to lower-level motor and speech levels, and terminates in behavior that, ex hypothesi, semantically fits the intention. And, once again, that anticipated terminus is the action's goal or purpose. Its final cause. This approach offers a plausible account of the way in which Danto's claim that "action is a matter of making the world fit one's representations" might be physically realized. The context-dependent constraints of any dynamical organization make its behavioral output conditional on the origin. In the case of actions, their trajectories thus embody the cognitive content of the intention from which they issue and which, with the help of feedback circuits, flows into the behavior to its completion. Using conditional probability, the concepts of equivocation and noise (as described by DRETSKE, 1981) can be used to calculate whether that trajectory is compromised or not. This reconceptualization of cause in terms of the operations of contextual constraints allows both for the possibility of self-cause as well as for structuring or formal causes that guide and direct as well as initiate behavior- all within a scientifically acceptable framework. It also allows for a rethinking of teleology. The trajectory that implements the intention aims at a behavioral output that reproduces the invariant relations of the intention. Causes of action and their behavioral effects have traditionally been conceptualized as two independent and disconnectable items. Once the volition (intention, etc.) fired, its job was done; the motor units got to do their thing next (separately next). Requiring that action be understood as an unequivocal trajectory within and constrained by the dynamics of a self-organized semantic space eliminates the atomicity of intentions and volitions. An action's trajectory crosses dynamical coordinates, from the conscious, semantic and affective levels of neural organization to those governing raw motor control. Since individual neurons implicated in the motor system for blinks might the same ones implicated in winks, the difference between a wink and a blink, is not to be located, reductionistically, in the individual neurons and muscles involved. Instead, we must look to the entire trajectory - origin, and constrained pathway and output. For behavioral output to constitute an act-token, the pathway and output must have been entrained into and constrained by the dynamics of its origin in a cognitively organized semantic space, whose properties and laws monitored and directed the trajectory.
Réseaux sociaux