Forum on Conceptual Structure

This network is intended for people with a professional or technical interest
in the theory of concepts and related topics -- such as semantic ontology, taxonomy, epistemology,
artificial intelligence and cognitive science. We look forward to your insights and participation!
Note: Pseudonyms and email aliases are available for participants who must maintain confidentiality. Please contact us.

Previous message from Forum on Conceptual Structure Next message from Forum on Conceptual Structure Previous message from Bruce Schuman Next message from Bruce Schuman

ON STIPULATION
Print format

From: Bruce S.
Date: Monday, March 10, 2008, 2:55 PM
Subject: On Stipulation
ID: 260023


I wanted to add a few quick thoughts on the nature of "stipulative definition". The way I see it, by understanding semantics and definition in terms of intention -- "stipulated definition" -- we can clear away much of the confusion associated with attempts to ground definition in social convention.

Most of the concepts we use are defined at a high level of abstraction, with a somewhat blurry mapping from that level of abstraction to anything specific or concrete "in the real world". That blur or potential confusion emerges because social convention is at best a statistical thing, something like a "cluster", with an approximate range of meanings (perhaps a little bit like the graphic we are using for this discussion).

Here is a graphic that decomposes (analyzes) the definition of the concept "credit-worthiness":

This is a "stipulated definition" -- as it might be defined by some lender. This isn't an attempt to define credit-worthiness in all cases or situations -- but in a specific case, as defined by some lender or person with an interest in this particular concept.

We might break down this concept in any number of ways -- but this map breaks it down in a specific way, saying that these are the dimensions and factors that matter -- and that other possibilities, in this context, can be ignored. We can suppose that this decomposition-tree could be continued a few more levels to the right side, until it became very specific -- with the bottom branch of the tree defined in exact numeric ranges (margin must be within a bounded range, income must be in a bounded range, net worth must be in a bounded range, etc.)

We can generalize this model in these terms:

This diagram shows a "cut" at the bottom level of each defined dimension -- much as if that bottom level were a linear dimension of measurement, and the "cut" defined a bounded range (in a digital space, numbers are always defined to some finite number of decimal places -- and "the next decimal place", undefined, is always a kind of mystery -- an uncertainty knowable only within bounds).

To generalize this concept to all instances of stipulative definition, we can simply say that "words" (like "credit-worthiness") are usually defined at some level of abstraction, with some kind of potential decomposition tree extending towards lower levels of analysis (higher degrees of specificity). This decomposition is always "implicit" -- it's not explicitly defined or stated, until for some reason we are motivated to make the decomposition exact and explicit and particular.

Another point worth noting -- is that this dimensional diagram illustrates the basic problem with the so-called "liberal arts" -- in that their language and definitions do not extend to a full and solid grounding in empirical specifics -- but remain "floating above the empirical plane" -- with an inherent degree of uncertainty that makes verification in any reliable way impossible. This is why we have created the "Bridge Across Consciousness" project -- with the idea that deep human intuition -- as essential as it is -- is inherently ambiguous as regards particulars. In today's world, we need to be guided by our highest intuitions in a reliable way. We need a bridge...

Towards this end, we can propose an approximate ordering of the categories and disciplines of human knowledge, ranging along this distribution -- high-level inclusive abstractions on the left side (of these diagrams), and empirical specifics and "measurables" on the right side. And we need not offer this model in some rigid format. It, too, even though offered as a general principle, is "local and ad hoc" and specific to some particular moment. It's the general form that is invariant. The specifics are 100% context-dependent.

And this is not a new idea. The ranking or ordering of variables in this way is a common thing in the social sciences. What may be new is the generalization of this idea to all cases of stipulative definition.

*

It is true that any act of communication depends on these abstractions being "more or less understood" by the listening party. Talking to a banker, he may simply speak about credit-worthiness as a broad generalization, without offering specifics. But when the time comes to take out the loan, the numbers will become specific. In the end, the entire assembly adds up to a binary option: yes or no. And where did the specifics come from? They are assigned to this matrix by the bank or the banker, according to their policy. They intend, they stipulate, they create the definition. And, indeed -- they may make the "empirical measurements" that determine the overall yes or no determination.

Link to this message:
http://lightpages.net/lp/message.cfm?messageid=260023&ex=463335&mo=1

About Forum on Conceptual Structure
About LightPages | Networks | Help | Privacy Policy | Contact Us