Re: RT list: On the conceptual-procedural distinction (again)

Fraser (2006) was not the first person to argue against the idea that a linguistic element must encode either conceptual information or procedural inforation but not both. I argued the same in my 1996 DPhil thesis and also in a 1998 article:

Nicolle, Steve (1998). A relevance theory perspective on grammaticalisation. Cognitive Linguistics 9: 1-35

The thrust of the article is that since grammaticalisation involves the change of typical conceptual expressions (e.g. verbs) into typical procedural expressions (e.g. tense markers) there must be a point at which a linguistic expression undergoing grammaticalisation encodes both conceptual and procedural information. An analysis of 'be going to' shows that it expresses both procedural and conceptual information, as would be expected given that it is a recently grammaticalised construction.

However, I do not think that pronouns necessarilyencode both conceptual and procedural information. Pronouns in English encode information concerning person, number, gender and case. However only person is common to all the personal pronouns in English: you specifies that an instance of second person reference is at stake, but fails to distinguish number, gender or case (this latter being specified structurally) except for the genitive form. Given that the procedural information encoded by I takes the form of an instruction to identify its referent by first identifying the speaker (i.e. the first person), information concerning person must be procedural. Since you encodes no more information than that concerning person, it must be an exponent of procedural encoding alone, and so this pronoun at least encodes no conceptual information.

If procedures are computations there must be something to be computed; this, I think, is behind the informal definition of procedural information as information which constrains the manipulation of conceptual representations. If this is correct, my suggestion is that the conceptual representations which are manipulated in line with the procedural information encoded by pronouns are ones which are projected from the logical entries of predicators (verbs), which specify what the predicator needs to combine with to yield a well-formed logical form (see Sperber & Wilson 1986:86). This is in line with Denny (1986), who proposes that predicators give rise to 'sortal expectations'. My suggestion is that these take the form of semantically underdetermined sub-propositional conceptual representations of their arguments. In a procedural account, both pronouns and classifiers/agreement markers constrain the manipulation of these conceptual representations thereby aiding the addressee's search for the intended referent.

In addition to procedural information concerning person, case, etc. which differs between pronouns, all members of the set of pronouns (in a given language) encode procedural information relating to the relative accessibility of mental representations of intended discourse referents (pace Ariel). This takes the form of an instruction to find a highly accessible discourse entity within the utterance interpretation context, which is compatible with both the specific procedural information encoded by a particular pronoun and the conceptual representation projected by the logical entry of the relevant predicator. Pronouns reflect and exploit the principle of relevance by constraining the addressee's search for intended referents to the set of highly accessible discourse entities. Individual pronouns in different languages also narrow the search by encoding information relating to one or more of number, person, gender, case, noun class, etc. This information is subsumed by the particular procedural information encoded by first and second person pronouns (an instruction to identify the intended referent by first identifying the speaker or addressee) since speakers and addressees are always highly accessible within an utterance interpretation context. In the case of third person pronouns, however, the general procedural information relating to accessibility distinguishes the referents of pronouns from those of descriptions, names, demonstratives, emphatic pronouns, etc.

I believe it is the combination of these two kinds of procedural information (number, person, gender, case, noun class, etc. on the one hand, and accessibility on the other) which characterises pronouns as procedural expressions.

One publication dealing with pronouns from an RT perspective is:
Cram, David & Paul Hedley (2005) Pronouns and procedural meaning: The relevance of spaghetti code and paranoid delusion. Oxford University Working Papers in Linguistics, Philology and Phonetics 10:187-210.

For an argument against procedural semantics, see:
Bezuidenhout, Anne (2004) Procedural Meaning and the semantics/pragmatics interface. In: Claudia Bianchi (ed) The Semantics/Pragmatics Distinction. CSLI. Pp.101-131.

Hope this is of interest.

Steve Nicolle

Reference
Denny, Peter J. (1986). The semantic role of noun classifiers. In: Colette Craig (ed.) Noun Classes and Categorization. Amsterdam: John Benjamins. Pp.297-308.

  ----- Original Message -----
  From: Minh Dang
  To: relevance@linguistics.ucl.ac.uk
  Sent: Tuesday, November 27, 2007 1:59 PM
  Subject: RT list: On the conceptual-procedural distinction (again)

    
  Dear all,

  I am looking into the conceptual-procedural distinction, and I would be grateful if you could share with me your opinions on the following.

  1) As far as I know, the standard RT position on personal subjective pronouns such as I, he is that they are procedural though I have known of no one specifying exactly what procedural information is encoded in each. I would imagine the procedure encoded in he is something like this: look for a male person, but I have problem with I. Is it: look for the speaker/utterer or else? I also have problem with the it in It's great to hear from you.

  In addition, what is the RT position on personal objective pronouns such as me, him? And what is the RT position on other pronouns such as one, someone, somebody, anyone, nobody, nothing?

  Also, what is the RT position on proper nouns such as Blair, Scotland, Amazon?

  2) I read in Blakemore (2006) that because is not procedural but conceptual. Now, if because encodes a concept, it should be amenable to semantic compositionality. The best examples I can think of are simply because, just because, partly because, but honestly I am not sure. Could someone clarify this for me please? I am also wondering what logical and encyclopaedic properties of because are, for these are defining properties of concepts according to Wilson (unpublished lecture 2002-3).

  In addition, I assume that the RT position on prepositions is that they are conceptual. But again, it seems to me they do not look susceptible to semantic complexity as suggested by Fraser (2006)

  3) It has been argued by Dor (2003) that 'newspaper headlines are designed designed to optimize the relevance of their stories for their readers: Headlines provide the readers with the optimal ratio between contextual effect and processing effort, and direct readers to construct the optimal context for interpretation.' My understanding of the above quote from the abstract of the article is that newspaper headlines are in effect procedural. Now, if Dor is right, I would like to say two things. First, a quick look at newspaper headlines reveals that they are often loaded with conceptual expressions. For example, BECKS FACES CROATIA AXE; SICK NOTICE; NIGHT FOR PRIDE, TERRY READY TO ROAR... all taken from The Sun of 19 Nov 2007. Thus, I find it difficult to say that these headlines are not conceptual. Second, if it is true that newspaper headlines are procedural or relevance optimisers (I think it is true), it should be possible to argue along similar lines that the topic sentence in a passage, titles of books, subtitles, abstracts, summaries, tables of content, preface, illustration on the cover, paragraphing, chaptering, choice of conceptual words in an utterance, intonation, stress, formating (bold, italic, ..), pause, punctuation, silence, and so on, all serve the same function, i.e. optimising relevance or procedural. It looks like everything that is used is relevance optimisers/procedural! What do you think? For me, I think concepts and procedures are like two sides of one and the same coin: no coin has only one side, be it conceptual or procedural. When we talk about procedural meaning of an expression, it seems to me that we are not talking about its 'per se' meaning. Rather, it looks like we are talking about function(s) of the expression. Of course, I am assuming that function and meaning are two different things of one and the same thing - by the latter 'thing' I mean an expression or a linguistic form, or even more broadly anything that is used for commutative purpose.

  Thank you very much for reading. And thank you very much more for any comments you are going to have.

  Minh

  Quick Reference

  Blakemore, D. (2006). Meaning, Procedural and Conceptual. In Encyclopaedia of Language and Linguistics, Elsevier.

  Dor, C. (2003). On newspaper headlines as relevance optimizers. Journal of Pragmatics, Vol 5, Issue 5, pp 695-721.

  Bruce Fraser (2006) "On the conceptual-procedural distinction". Style. Spring-Summer 2006. FindArticles.com. 27 Nov. 2007. http://findarticles.com/p/articles/mi_m2342/is_1-2_40/ai_n17113874

------------------------------------------------------------------------------
  Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now.
Received on Tue Nov 27 13:36:32 2007