home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!stanford.edu!bcm!convex!darwin.sura.net!paladin.american.edu!auvm!CCB.BBN.COM!BNEVIN
- Message-ID: <CSG-L%92111013000853@VMD.CSO.UIUC.EDU>
- Newsgroups: bit.listserv.csg-l
- Date: Tue, 10 Nov 1992 13:53:16 EST
- Sender: "Control Systems Group Network (CSGnet)" <CSG-L@UIUCVMD.BITNET>
- From: "Bruce E. Nevin" <bnevin@CCB.BBN.COM>
- Subject: re: marking
- Lines: 75
-
- [From: Bruce Nevin (Tue 921110 12:59:20)]
-
- (From Bill Powers (921109.1100) ) --
-
- Re "marking" something as having been said, I think as Martin has
- suggested in his Layered Protocol model of communication that we control
- for a perception of the other party having understood.
-
- The universe of perceptions is not linear, but speech is.
-
- The universe of perceptions is not linear, that is, the ensemble of
- pathways for shifting attention consecutively from one perception to
- another that is in some PCT-internal sense "adjacent" forms some sort of
- network or mesh, not a linear sequence.
-
- Yet when we talk about these perceptions, we must use a linear sequence
- of words (some of them reduced in form so as to appear as parts of
- words). This means that we must linearize some particular traversal of
- the net or mesh of perceptions. Or, commonly, successive traversals
- with returns to common nodes represented as repeated words (often
- reduced to pronouns, zero, etc.).
-
- It seems clear that we control for a given traversal of "adjacent"
- perceptions having been talked about: instead of repeating the traversal
- in words, we use one of the reduced forms (e.g. "this is why thus and
- so," where "this" is a reduced form of words already uttered, and
- corresponds to the same traversal of perceptions that those words
- referred to.)
-
- Given that we also control for perceiving that a given traversal of
- has perceptions been understood by the other, such a perception is often
- grounds for producing reduced forms of words in an utterance, even
- though their unreduced forms had not actually been previously spoken.
-
- In the analysis of language form (hold your nose Bill, I'm not so far
- from perceptual control as you might think), you get a more efficient
- system if you postulate a requirement for word repetition across
- conjunctions. This provides a basis for reconstructing intervening
- sentences conjoined under "and" that (ex hypotheosi) had been zeroed
- because they were common knowledge. Thus:
-
- Don't forget your umbrella, it's Wednesday. <==
-
- Don't forget your umbrella, because you use an umbrella when it rains,
- and the paper said it would rain tomorrow, and you might come back
- tomorrow if you miss the train, and you sometimes miss the train when
- you have a late meeting, and you have a late meeting Wednesdays, and
- today is Wednesday.
-
- It seems to me that we use the socially learned structure of language as
- an framework for organizing how our attention traverses our perceptions
- (the combination of perceptions from the environment with perceptions
- from memory and imagination). It seems also that the hierarchical
- organization of perception strongly informs the operator-argument
- structure of language, as you have suggested, Bill (relationship words
- for relationship perceptions, etc.). However, that correspondence is
- muddled by the existence of phrases that correspond apparently to
- unitary perceptions (frozen expressions, idioms, etc.), diverse
- perceptions for the same word or phrase (polysemy, homophony), and so
- forth, but above all because language is a learned, conventional system
- of social agreements, and the universe of nonverbal perceptions is not.
-
- In our speech, we mark something as "I assume that you already perceive
- this" by producing only reduced forms of the words for it. We use this
- in really quite subtle ways to test and confirm (co-affirm) ongoing
- agreements. This is one not entirely obvious way that the constraints
- and conventions of language become a very important tool for creating
- social agreements.
-
- We control for creating and sustaining social agreements. Failure of
- social agreements is probably occasion for intrinsic error in mammals
- (possibly in other creatures as well).
-
- Bruce
- bn@bbn.com
-