Monthly Archive for March, 2005

H. R. Ekbia and A. G. Maguitman. Modeling and Using Context (Proceedings of the conference CONTEXT 2001, Dundee, UK, July), volume 2116 of Lecture Notes in Artificial Intelligence, chapter Context and Relevance: A Pragmatic Approach, pages 156–169. Springer, Berlin, 2001. [url]

————————

This article propose a pragmatic approach to the issues of context and relevance, where some foundamental assumptions of formal logic are still maintained at the theoretical level. Here the authors, usign some conceptualisations of Dewey, show that context has two components: background, which is spatial and temporal; and selective interest, which conditions the subjective matter of thinking.

Thinking for Dewey is a process of inquiry in which a confused, obscure, or conflicting situation is transformed into a determined one. For the authors, this can be achieved formulating a precise set of inference rules for mapping between actions and plans. Interaction with the environment would involve to detect and recognize the context explicitly and to revise the plans according to some reasoning scheme.

Logic is biased in embracing the contextual dependences outlined by Dewey, becaused the adherence to explicit representations is not fully characteristic of human behaviour:

1. Context, often, is not explicitly identifiable;

2. There are no sharp boundaries among contexts;

3. The logical aspects of thinking cannot be isolated from material considerations;

4. Behavior and context are jointly recognisable.

The pragmatic tradition highlights the action-oriented (actions also equals reasoning behavior) nature of intelligence. At the heart of intelligence lies the ability to find out what is relevant in any given situation, and to act accordingly. Pragmatic relevance, has mainly to do with “selecting the search space” rather than with the process itself.

D. Wilson and D. Sperber. The Handbook of Pragmatics, chapter Relevance Theory, pages 607–632. Blackwell, Oxford, 2004. [url]

————————–

This article present the basic theoretical assumption of the relevance theory, developed by Sperber and Wilson. The relevance theory is based on the inferential model, where a communicator provides evidences of her/his intention to convey a certain meaning, which is inferred by the audience on the basis of the evidences provided. The goal of the inferential pragmatics is to explain how the hearer infers the speaker’s meaning on the basis of the evidences provided.

The central claim of the relevance theory is that the expectations of relevance raised by an utterance are precise enough, and predictable enough, to guide the hearer towards the speacker’s meaning. The aim is to explain in cognitively realistic terms what these expectations of relevance amount to, and how they might contribute to an empirically plausible amount of comprehension.

The authors extend this theory not only on the utterance level, but they port it to all the rest of the cognitive phenomena, like thoughts, memories and conclusions of inferences. The search for relevance is a basic feature of human cognition, which communicators may exploit.

An input is relevant to an individual when it connects with background information he has available to yield conclusions that matter to her/him. An imput is relevant when its processing in a context of available assumptions yields a positive cognitive effect: i.e., a positive conclusion. A contextual implication is a conlcusion deductible from the input and the context together, but from neither input and context alone.

Relevance is a matter of degree. Other things being equal, the more worthwile conclusions achieved by processing an input, the more relevant it will be. Other things being equal, the greater the processing effort required, the less relevant the input will be. Thus relevance may be assessed in terms of cognitive effects and processing effort.

Relevance theory treats the identification of explicit content as equally inferential, and equally guided by the communicative principle of Relevence, as the recovery of implicatures.

D. E. Broadbent. Levels, hierarchies, and the locus of control. Quarterly Journal of Experimental Psychology, (29):181–201, 1977.

———————————

This old paper describes a multi-level adaptive control mechanism of problem-solving situation. The general principle explained is that human processing of information takes place on many levels: some of these levels modify or control the operation of others.

The author describe his setup for controlling problem-solving solution, which is constituted by a simulation of bus transport in a city. The users can alter the time interval between the buses entering the city, and they can alter the amount charged for use of the city car’s park. By altering these quantities, they can affect the load on the buses, which is usually expressed in the number of people carried per hundrerd buses. They can also affect the number of empty spaces remaining in the car parks. Before experience of the situation, and again afterwards, the users are questioned on the effect of their interventions on the system. Despite the satisfactory performance atteined in the task, their ability to answer questions about what they did on the system was poor.

Then, the author describes different control mechanisms, showing that the structure that could explain the obtained results of the task is an adaptive controller structure. The key point of the paper is that the transfer of controlshould include the possibility that one section of the program/structure can alter other sections program/structure.