Clark, A. (2013). Expecting the world: Perception, prediction, and the origins of human knowledge.
Journal of Philosophy, CX(9), 469–496.
Last edited by: sirfragalot 26 Jul 2018 10:36:16 Europe/Copenhagen
"Perception [...] is a matter of the brain using stored knowledge to predict, in a progressively more refined manner, the patterns of multi-layer neuronal response elicited by the current sensory stimulation."
perception "is a process of explaining away the sensory signal by finding the most likely set of interacting distal causes"
Arguing that the model he presents does not support the view of the reality of the world being created within us (i.e.
indirect perception), Clark states that: "The internal representations at issue function within us, and are not encountered by us. Instead, they make it possible for us to encounter the world. Moreover, they enable us to do so under the ecologically common conditions of noise, uncertainty, and ambiguity."
Perception is "an active process involving the (sub-personal) prediction of our own evolving neural states."
"Perceptual content, as delivered by such a process of active self-prediction, is inherently organized and outward-looking [...] it reveals ‒ and cannot belp but reveal ‒ a structured [...] external world ... the world thus revealed is inherently meaningful ... It is an external arena populated not by proximal stimulations but by distal, causally interacting items and forces whose joint action best explains the current suite of sensory stimulation."
Clark presents two alternate models of perception:
"What happens when, after a brief chat with a colleague, I re-enter my office and visually perceive the hot, steaming, red cup of coffee that I left waiting on my desk? One possibility is that my brain receives a swathe of visual signals (imagine, for simplicity, an array of activated pixels) that specify a number of elementary features such as lines, edges, and color patches. Those elementary features are then progressively accumulated and (where appropriate) bound together, yielding shapes and specifying relations. At some point, these complex shapes and relations activate bodies of stored knowledge, turning the flow of sensation into world-revealing perception: the seeing of coffee, steam, and cup, with the steaming bound to the coffee, the color red to the cup, and so on.
As I re-enter my office my brain already commands a complex set of coffee-involving expectations. Glancing at my desk sets off a chain of visual processing in which current bottom-up signals are met by a stream of downwards predictions concerning the anticipated states of various neuronal groups along the appropriate visual pathway. In essence, a multi-layer downwards cascade is attempting to "guess" the present states of all the key neuronal populations responding to the present state of the visual world. There ensues a rapid exchange (a dance between multiple top-down and bottom-up signals) in which incorrect guesses yield error signals which propagate forward, and are used to extract better guesses. When top-down guessing adequately accounts for the incoming signal, the visual scene is perceived. As this process unfolds, top-down processing is trying to generate the incoming sensory signal for itself. When and only when this succeeds, and a match is established, do we get to experience (veridically or otherwise) a meaningful visual scene."
"the brain's job is to account for the sensory signal by finding a way to generate, in a kind of rolling present, that incoming signal for itself. To do this, the brain must find the structure in the signal. But the structure in the sensory signal is mostly determined by the structure in the world (making sure that's the case is pretty much the job description if you are a sensory transducer). So the best way to anticipate/match the incoming signal is to discover and deploy internal resources that amount to a kind of 'virtual reality generator' that models the distal elements and their typical modes of interaction (simplistically, if it generates 'car' and 'sudden braking' it might also generate 'smoke from tires'). An agent perceives when the virtual reality generator can use its resources to capture (match, cancel out) the structure of the incoming signal."
Prediction-based models that Clark espouses "learn to construct the sensory signal by combining probabilistic representations of hidden causes operating at many different spatial and temporal scales [...] they must match the incoming sensory signal by constructing the signal from combinations of hidden causes (latent variables). The so-called 'transparency' of perception emerges as a natural consequence of such a process when it is conditioned by an embodied agent's lifestyle-specific capacities to act and to choose. We seem to see dogs, cats, chasings, pursuits, captures [...] because these feature among the interesting, nested, structures of distal causes that matter for human choice and action."
"Perception [...] is the successful prediction of the current sensory signal using stored knowledge about the world. [The model] explains why perception, although carried out by the brain, cannot help but reach out to a distal world; it shows why that 'reaching out' reveals a world that is already structured"
Clark presents another example of the two competing models of perception this time concerning speech recognition (quoting p.936 of Poeppel & Monahan):
"Representations constructed at earlier stages of processing feed immediately higher levels in a feedforward manner...this process proceeds incrementally until access to a ''lexical conceptual'' representation has been achieved. In speech recognition...this involves a conversion from acoustic features onto phonetic representations, phonetic representations onto phonological representations, and finally access of the lexical item based on its phonological structure.
[...] (1) the extraction of (necessarily brief and coarse) cues in the input signal to elicit hypotheses, that while coarse, are sufficient to generate plausible guesses about classes of sounds (for example, plosives, fricatives, nasals, and approximants), and that permit subsequent refinement; (2) the actual synthesis of potential sequences consistent with the cues; and (3) a comparison operation between synthesized targets and the input signal delivered from the auditory analysis of the speech." David Poeppel and Philip J. Monahan “Feedforward and feedback in speech perception: Revisiting analysis by synthesis”, Language and Cognitive Processes 26:7, (2011): 935-95.
Evans, D. (2001).
Emotion: A very short introduction. Oxford: Oxford University Press.
Added by: sirfragalot 28 May 2011 07:52:39 Europe/Copenhagen
Re John Searle's thought experiment, the
Chinese Room. A man in a room has a set of rules for responding to Chinese inscriptions that are given to him. Based on these responses, people outside the room believe the man knows the Chinese language. He doesn't though. Searle argues a computer is like this -- it only interprets, it never knows therefore could never become conscious.
Szabó Gendler, T. (2010).
Intuition, imagination, & philosophical methodology. Oxford: Oxford University Press.
Last edited by: sirfragalot 18 Apr 2013 17:40:14 Europe/Copenhagen
Reasoning (according to Steven Sloman's Two Systems) involves two systems: Associative and Rule-based. The first (System 1) operates on similarity, contiguity, is automatic, uses generalization, soft constraints and is exemplified by intuition, imagination, fantasy, creativity etc.
Rule-based reasoning (System 2) uses symbol manipulation, derives knowledge from language, culture and formal systems, uses hard constraints, can operate on concrete, general and abstract concepts and is exmplified by explanation, deliberation, verification, formal analysis, strategic memory. Division into two systems may be simplisitic but there is certainly not just one system used for reasoning.