Real-Time Systems. Design Principles for Distributed Embedded Applications. Herman Kopetz. Second Edition (811374), страница 14
Текст из файла (страница 14)
The study of modelbuilding, problem solving, and knowledge representation forms an important partof the cognitive sciences.2.1.1Problem SolvingHumans have two quite different mental subsystems for solving problems: theintuitive-experiential subsystem and the analytic-rational subsystem [Eps08].Neuro-imaging studies have shown that these two subsystems are executed in twodifferent regions of the human brain [Ami01]. Table 2.1 compares some of thedistinguishing characteristics of these two subsystems.Example: A typical task for the intuitive-experiential subsystem is face recognition, ademanding task that a baby at the age of 6 months can accomplish. A typical task for theanalytic-rational subsystem is the confirmation of a proof of a mathematical theorem.The experiential subsystem is a preconscious emotionally-based subsystem thatoperates holistically, automatically, and rapidly, and demands minimal cognitiveresources for its execution.
Since it is nearly effortless, it is used most of the time.It is assumed that the experiential subsystem has access to a large coherentTable 2.1 Intuitive experiential versus analytic rational. Problem solving strategy (Adapted from[Eps08, p. 26])Intuitive experientialAnalytic rationalHolisticAnalyticEmotional (what feels good)Logical reason oriented (what is sensible?)Unreflective associative connectionsCause and effect connections, causal chainsOutcome orientedProcess orientedBehavior mediated by vibes from pastBehavior mediated by conscious appraisalexperienceof eventsEncodes reality in concrete images, metaphorsEncodes reality in abstract symbols, words,and narrativesand numbersMore rapid processing, immediate actionSlower processing, delayed actionSlow to change the fundamental structure: changes Changes more rapidly, changes with thewith repetitive or intense experiencespeed of thoughtExperience processed passively and preExperience processed actively andconsciously, seized by our emotionsconsciously, in control of our thoughtsSelf evidently valid: seeing is believingRequires justification via logic andevidence2.1 Cognition31knowledge base that represents an implicit model of the world.
This subjectiveknowledge base, which is one part of what we call the conceptual landscape of anindividual, is mainly built up and maintained by experience and emotional eventsthat are accumulated over the lifetime of an individual. Although this knowledgebase is continually adapted and extended, its core structure is rather rigid andcannot be changed easily. Experiential reasoning is holistic and has the tendencyto use limited information for general and broad classifications of scenarios andsubjects (e.g., this is a good or bad person). The experiential system does assimilatethe data about reality in a coherent stable conceptual framework.
The concepts inthis framework are mostly linked by unconscious associative connections, wherethe source of an association is often unknown.The rational subsystem is a conscious analytic subsystem that operates accordingto the laws of causality and logic. Bunge [Bun08, p. 48] defines a causalityrelationship between a cause C and an event E as follows: If C happens, then(and only then) E is always produced by it. We try to get an understanding of adynamic scenario by isolating a primary cause, suppressing seemingly irrelevantdetail, and establishing a unidirectional causal chain between this primary causeand an observed effect. If cause and effect cannot be cleanly isolated, such as is thecase in a feedback scenario, or if the relationship between cause and effect is nondeterministic (see also Sect. 5.6.1 on the definition of determinism), then it is moredifficult to understand a scenario.Example: Consider the analysis of a car accident that is caused by the skidding of a car.There are a number of conditions that must hold for skidding to occur: the speed of the car,the conditions of the road (e.g., icy road), the conditions of the tires, abrupt manoeuver bythe driver, the non-optimal functioning of the computer based skid-control system, etc.
Inorder to simplify the model of the situation (the reality is not simplified) we often isolate aprimary cause, e.g., the speed, and consider the other conditions as secondary.The rational subsystem is a verbal and symbolic reasoning system, driven by acontrolled and noticeable mental effort to investigate a scenario.
Adult humanshave a conscious explicit model of reality in their rational subsystem, in addition totheir implicit model of reality in the experiential subsystem. These two models ofreality coincide to different degrees and form jointly the conceptual landscape of anindividual. There seem to be a nearly unlimited set of resources in the experientialsubsystem, whereas the cognitive resources that are available to the rational subsystem are limited [Rei10].There are many subtle interrelationships between these two problem-solving subsystems, which form the extremes of a continuum of problem solving strategies whereboth systems cooperate to arrive at a solution. It is not infrequent that, after unsuccessful tries by the rational subsystem, at first a solution to a problem is producedunconsciously by the experiential subsystem.
Afterwards this solution is justified byanalytical and logical arguments that are constructed by the rational subsystem.Similarly, the significance of a new scenario is often recognized at first by theexperiential subsystem. At a later stage it is investigated and analyzed by therational subsystem and rational problem solving strategies are developed. Repeated322 Simplicityencounters of similar problems – the accumulation of experience – effortfullearning and drill move the problem-solving process gradually from the rationalsubsystem to the experiential subsystem, thus freeing the cognitive resources thathave previously been allocated to the problem solving process in the limitedrational subsystem.
There exist many practical examples that demonstrate thisphenomenon: learning a foreign language, learning a new sport, or learning howto drive a car. It is characteristic for a domain expert that she/he has mastered thistransition in her/his domain and mainly operates in the effortless experiential mode,where a fast, holistic and intuitive approach to problem solving dominates.Example: A brain-imaging study of the chess-playing strategy of amateurs versus grandmasters investigated the activity in different sections of the brain immediately after a chessmove by the partner.
The amateurs displayed the highest activity in the medial temporallobe of the brain, which is consistent with the interpretation that their mental activity isfocused on the rational analysis of the new move. The highly skilled grandmasters showedmore activity in the frontal and parietal cortices, indicating that they are retrieving storedinformation about previous games from expert memory in order to develop an understanding of the scenario [Ami01].2.1.2Definition of a ConceptIn a changing world, knowledge about permanent and characteristic properties ofobjects and situations must be identified and maintained since such knowledge isof critical importance for survival. This knowledge is acquired by the process ofabstraction, by which the particular is subordinated to the general, so that what isknown about the general is applicable to many particulars.
Abstraction is a fundamental task of the human cognitive system.Example: Face recognition is an example for the powerful process of abstraction. Out ofmany particular images of the face of a person – varying angles of observation, varyingdistance, changing lighting conditions – characteristic permanent features of the face areidentified and stored in order that they can be used in the future to recognize the face again.This demanding abstraction process is executed unconsciously, seemingly without effort, inthe experiential subsystem. Only its results are delivered to the rational subsystem.Abstraction forms categories, where a category is a set of elements that sharecommon characteristic features.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
