home *** CD-ROM | disk | FTP | other *** search
- Path: keats.ugrad.cs.ubc.ca!not-for-mail
- From: c2a192@ugrad.cs.ubc.ca (Kazimir Kylheku)
- Newsgroups: comp.lang.smalltalk,comp.lang.perl.misc,comp.lang.pascal.borland,comp.lang.eiffel,comp.lang.cobol,comp.lang.c++.leda,comp.lang.c,comp.lang.basic.visual.3rdparty,alt.computer.workshop.live
- Subject: Re: PROGRAMERS OF ANY LANGUAGE
- Date: 17 Apr 1996 12:43:19 -0700
- Organization: Computer Science, University of B.C., Vancouver, B.C., Canada
- Message-ID: <4l3hknINNgpg@keats.ugrad.cs.ubc.ca>
- References: <Pine.SOL.3.91.960329010021.13209A-100000@harvey> <4kqe6q$1tm@innocence.interface-business.de> <DpxA0t.3vy@bcstec.ca.boeing.com> <4l3a2l$753@gaia.ns.utk.edu>
- NNTP-Posting-Host: keats.ugrad.cs.ubc.ca
-
- In article <4l3a2l$753@gaia.ns.utk.edu>,
- Matt Kennel <kennel@msr.epm.ornl.gov> wrote:
- >Charles DeRykus (ced@bcstec.ca.boeing.com) wrote:
- >
- >: You must admit though that it's lovely to be able to say either
- >
- >: "If you really love me, feed me artichokes".
- >
- >: or,
- >
- >: "Feed me artichokes if you really love me".
- >
- >: Sometimes there's expressive power in saying one rather than
- >: the other. Besides, love needs a lot of latitude :)
- >
- >Yes, to humans there are subtle distinctions between the two.
- >
- >To me the second is more of an immediate request: they want to be fed
- >NOW.
-
- A linguist specializing in syntax would say that the two sentences originate
- with the same grammar structure. According to the most entrenched theories of
- syntax, both sentences conceptually start out as the same phrase structure
- ``tree'', but the former is ``topicalized''. Topicalization is a process
- whereby a sentence is transformed by the movement of constintuents toward the
- front. It is a feature of nearly all natural languages.
-
- It is not possible to account for transformations like topicalization with
- context free grammars, but context free grammars in combination with
- transformation rules form a convenient and workable model that is more
- empirically adequate and revealing.
-
- The above sentence might have the D-structure (the underlying one) that looks
- like:
-
- S
- +-----------------------+-------+---------------+
- | | |
- NP VP S'
- | +-------+-------+ +-------+
- | | | | | |
- N V NP NP COMP S
- -----------
- You feed me arti. if you love me
-
- S is sentence, NP is noun phrase, VP verb phrase S' is a marker for an optional
- sentence, COMP is a complementizer. The grammar in this case would look
- something like:
-
- The ``You'' is deleted under the transformation (traditional grade-school
- grammar teachers might refer to it as an ``understood subject''). This would
- probably be an obligatory transformation, though I'd have to look this up to be
- sure! The S' can travel and adjoin to the topmost S to the left of NP.
-
- Transformations can change the meaning. A declarative can turn into a question,
- for instance, with certain kinds of ``subject-AUX'' inversion or ``Wh-
- movement''.
-
- In this case, the semantic difference between the two sentences is not that
- clear.
-
-
- I find transformational syntax fascinating because it's not something that
- comes up in the field of computer languages, which normally stop at
- context-free grammars. When I was first introduced to this aspect of
- linguistics, I was surprised how far these guys are able to stretch a CF
- grammar by adding these clever transformation rules! At the same time, there is
- a computational flavour to it: a hacker can't help but imagine n-ary tree
- _data structures_ being traversed and manipulated, when faced with this.
-
- >The first is more of an abstract 'predicate', implying the object ought to
- >be the type of person who is willing to feed artichokes, but the
- >artichokes might not have to be procured this very instant.
-
- They may not be procured in the first. The subtleties are difficult in this
- case (as opposed in a case where a sentence undergoes a transformation to
- become a question). Exactly in what way a topicalization movement affects the
- meaning is difficult to estimate (at least from my own narrow understanding).
-
- >'perl' is too dumb to know the difference, so why have the distinction?
-
- Probably to do a weak imitation of the transformational nature of natural
- language syntax, and give the language the semblance of flexibility. It's
- ``syntactic sugar''.
- --
- I'm not really a jerk, but I play one on Usenet.
-