12,080
edits
Changes
Directory:Jon Awbrey/Papers/Inquiry Driven Systems : Part 5 (view source)
Revision as of 19:12, 2 August 2011
, 19:12, 2 August 2011→5.1.2.8. The Formative Tension: + raw text
=====5.1.2.8. The Formative Tension=====
=====5.1.2.8. The Formative Tension=====
<pre>
A lassie all alone, was making her moan
Lamenting our lads beyond the sea:
"In the bluidy wars they fa', and our honor's gane an a',
And broken hearted we maun die."
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
An incidental purpose of this reconnaissance is to point out the more problematic features of interpretation, and especially the properties of sign comprehension that are observed to be salient in natural settings. It is these more problematic aspects of interpretation that ultimately place the greatest demands on "artificial intelligence research" (AIR), at any rate, if it is going to support an understanding of the ways that human beings use real languages in real practice, or if it is ever going to supply a medium for modeling the ways that human beings develop an ability to produce and to understand real texts.
I am sensitive to the possibility that there are many features in the situation of the present inquiry, perhaps due solely to my description, that make it appear like an artificial, an implausible, or a specialized set of circumstances. I begin by discussing a number of these features, after which I argue that they are really quite typical of any situation where one has to interpret a problematic text in a problematic language. A text or a language is "problematic", of course, only in relation to a prospective interpreter. To clarify this sense of the word "problematic" it helps to introduce the following definitions.
1. A "fully interpretive language" (FIL) is one whose attributes are fully filled out in all three directions of natural language use, to wit, along syntactic, semantic, and pragmatic dimensions.
2. A "fully interpretive grammar" (FIG), more commonly referred to as the "full grammar" of a language, is a body of knowledge that generates or specifies a FIL, incorporating the full details of its syntax, semantics, and pragmatics.
3. A "problematic text" is one that seems to make some sort of sense, but whose meaning, if any, is not entirely clear on first reading.
4. A "problematic language" is one whose "full grammar" is not yet available for articulation by the user in question, no matter how well the user is able to employ the language itself. It ought to be obvious that all "natural languages" are problematic for their customary users.
As I stood by yon roofless tower,
Where the wa'flow'r scents the dewy air,
Where the houlet mourns in her ivy bower,
And tells the midnight moon her care:
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
One of the functions of the epitext that I am weaving through the main text of this discussion is to keep constantly before the mind the more problematic features of natural languages, to illustrate the diversity of their utilities, and to notice a few of the elaborative, figurative, and imaginative devices that are not so easily formalized.
It is true that poems and programs share a conscious definition, coming to their collective senses under the head of "effective description" (ED). At least, it ought to be clear to the impartial observer that there is a close kinship between "the words that do" to engender action, that start it toward the hope of eventual success, and "the words that fit" a finite ambition, that point it toward the realms of eternal truth, and that this relation amounts to an accord that unites their intentions or a resonance that serves to bind their performance into a concerted whole, so I leave the reader to rede which is which and to judge on a case by case basis what the individual occasion demands.
It is also true that abstract category theory, in the ways that it affords a view of formal analogies, is largely a study of mathematical metaphors. But aside from all that, there is much else that mathematicians, poets, and programmers are bound to see alike and ought to share in common.
Under many natural circumstances, the only way to unravel the meaning of a problematic text is to place it in the field of influence of a FIL, typically as embodied in a variety of different interpreters, and to see how it is led to develop along the prevailing lines of interpretive force. In corresponding circumstances, approached in a complementary fashion, the only way to uncover the structure of a problematic language is to scatter a sample of signs and texts throughout its field of influence, and then to observe how these literal test particles are led to develop along interpretive lines, and if there is a coherent sensibility in force.
The winds were laid, the air was still,
The stars they shot along the sky,
The tod was howling on the hill,
And the distant echoing glens reply.
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
One of the abiding tasks of "artificial intelligence research" (AIR) is to figure out how natural languages do what they do for the human mind. This task amounts to articulating the FIL's that humans actually use, and thus to arrive at the "fully interpretive grammars" (FIG's) that generate these FIL's, shaping their syntax, semantics, and pragmatics. The main tool that one has in this task is "formalization", the process that devises formal models for the ongoing processes of interpretation.
Until one develops a battery of formal methods for exploring fields of interpretive influence and for tracing lines of interpretive force, one tends to be impelled by signs without understanding how or why one is moved by them, to wallow around in interpretive phenomena with little control over what develops, and to wander aimlessly through domains of apparent significance and evident meaning with no insight into their underlying structures and generative forms. Consequently, an attempt to avoid all formalization, though it appears at first sufficient to the gaining of experience, is not sufficient to the gaining of understanding, and therefore ultimately leads to the impoverishment of experience itself.
Unfortunately, there are equally pernicious tendencies that arise in the attempt to formalize experience and thus to arrive at formalized models. There is the tendency, in pursuing formalizations of a difficult subject, to settle on a premature formalization, that is, a narrowly circumscribed set of models, or an overly simplistic typology for addressing the topic, and then, in a vain attempt to avoid further difficulties by dictating to the subject how it ought to behave, to think that a partially successful formalization gives one the right to bar the subject from leaving the charmed circle swept out by its survey, or else to think that one can afford to ignore all aspects of the subject that do not fit within it. This temptation seems to arise on a recurring basis in the history of every formal science, being so well known from the dawn of awareness that its pattern is emblazoned in myth under the name of "Procrustes".
The burn, adown its hazelly path,
Was rushing by the ruin'd wa',
Hasting to join the sweeping Nith,
Whase roaring seemed to rise and fa'.
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
Generally speaking, formalizations constitute an indispensable utility in adapting the intellect to its environment and in helping it to build up relatively stable systems of organized knowledge. But it takes a "force" to make things "fit", since the formal picture is partial to its own bias, and the intellectual image distorts the world as much as it adapts to it.
The "violence" that is involved in a formalization, that is, the degree of its departure from nature, is observed to become especially acute each time that a specialized community of inquirers succeeds in developing a fresh array of formal methods and formal models, and thereby acquires a standardized resource that appears to equip them with an expanded range of formal powers. For those inquirers, these are competencies whose winning is not easy and whose salve they do not wish to give up, as they think they might by thinking too much about the lineaments of its ligaments and the limitations of its liniments. Consequently, one finds that the following sort of situation typifies the state of formal inquiry: Against every conscious caution that the formalization is only partial and in spite of every conscientious concession that the object domain is vastly more complex that any representation can contain, the tendency persists, once a formalization is fixed, to treat it just as if it were already and always would be perfectly adequate to the object.
The cauld blae North was streaming forth
Her lights, wi hissing, eerie din:
Athort the lift they start and shift,
Like Fortune's favors, tint as win.
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
Now, looking over firth and fauld,
Her horn the pale faced Cynthia rear'd,
When lo! in form of minstrel auld
A stern and stalwart ghaist appear'd.
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
And frae his harp sic strains did flow,
Might rous'd the slumbering Dead to hear,
But O, it was a tale of woe
As ever met a Briton's ear!
Robert Burns, As I Stood by Yon Roofless Tower, [CPW, 570]
He sang wi joy his former day,
He, weeping, wail'd his latter times:
But what he said it was nae play!
I winna ventur't in my rhymes.
Robert Burns, As I Stood by Yon Roofless Tower,
[CPW, 570]
</pre>
=====5.1.2.9. The Vehicle of Communication : Reflection on the Scene, Reflection on the Self=====
=====5.1.2.9. The Vehicle of Communication : Reflection on the Scene, Reflection on the Self=====