Thursday, August 07, 2008

Sky Marsen - The Role of Meaning in Human Thinking

"The Role of Meaning in Human Thinking" is a cool article from The Journal of Evolution and Technology. This article could be called: A Semiotic Model of Why AI Has Failed So Far. Interesting stuff, if you can wade through the philosophical language.

The Role of Meaning in Human Thinking

Sky Marsen, Victoria University, New Zealand

Journal of Evolution and Technology - Vol. 17 Issue 1 – March 2008 - pgs 45-58

Abstract

The creation of meaning to interpret and communicate perceived phenomena is a fundamental trait of human intelligence. This article explains some major ways in which this is achieved, focusing on language and the perception of embodiment. It examines the representational aspects of natural language, which account for the ambiguity of linguistic signs, and describes how these manifest in metaphor, connotation and emotive expression. The article argues that the human propensity to create meaning lies largely in this representational ambiguity, which underlies all forms of symbolism. However, the ambiguity of natural language has a paradoxical side, since it is also at fault in many shortcomings of human communication, such as misunderstanding and prejudicial stereotyping. This article argues that any attempt to emulate human ways of thinking, for example in Artificial Intelligence research, should take this paradoxical factor into account.

* * *
He, who through vast immensity can pierce,
See worlds on worlds compose one universe,
Observe how system into system runs,
What other planets circle other suns
What varied being peoples every star
May tell why Heaven has made us as we are.

(Alexander Pope: Essay on Man.)
Humans create meaning. In fact, it is a fundamental trait of humans to attach meaning to the objects they perceive in the world, to their relationships with others, to their own physical form, and to the various manifestations of agency encompassed by the category “self” – a trait that is as universal as that of language. The complex operations that characterize human cognition carry this meaning-generating function on many levels. Classifying an object according to selected criteria, attaching value to it, and judging its aesthetic appeal, are all mental operations that, in one way or another, give meaning to the phenomenal world.

This article explores some ways in which meaning is produced, especially with the use of language. Using a semio-linguistic approach, it explains some of the basic principles of human language that affect thinking and underpin communication. Its aim is to discuss some aspects of Human Intelligence that distinguish it from Artificial Intelligence (AI) in its current state, and to suggest some areas that would require improvement if humans are to reach a post- or trans-human stage. I begin with an overview of theoretical approaches to meaning, continue with a description of pertinent linguistic features of communication, and end with an overview of areas where communication is problematic, if not defective.
Go read the whole article.

Here is a key passage from later in the article:
Many professions, for example, favor skill and innovation as opposed to experience, while the linear, climactic, continuity of modernism is rapidly being replaced with the serial, random access of the digital era. Also, many post-modern individuals share a value system with peer groups, often scattered throughout the world, rather than with families or immediate communities, as they did in the past.

Based on modernist ideology, existing conceptual structures and metaphors may restrict recognition of these changes, and may prevent contemporary, pre-posthumans from engaging with them creatively. In fact, a shortage of symbolic forms that would help to legitimize these developments may well account for what is sometimes described as the hypocrisy of contemporary life. It seems that, although human conceptual and representational systems, such as language, are dynamic and adaptive, they do not adapt quickly enough to cater for the socio-emotional upheavals of transitional periods of human evolution, such as the one we are arguably undergoing now.

This is a dense but interesting article that examines the problems of communication in regard to meaning (which isn't to be confused with truth); and it raises issues that must be addressed if we are ever to develop AI that works.


No comments: