Re: SCWC 32: Discussion: IMPLEMENT

On Jun 13, 8:32 pm, Angus Rodgers <twir...@xxxxxxxxxxx> wrote:
On Wed, 13 Jun 2007 16:31:34 -0700, "Peter T.

Daniels"<gramma...@xxxxxxxxxxx> wrote:
On Jun 13, 5:39 pm, Angus Rodgers <twir...@xxxxxxxxxxx> wrote:
On Wed, 13 Jun 2007 17:38:05 -0000, Flying Tortoise

<purple....@xxxxxxxxxxxxxx> wrote:
Tell that to a court transcriber. It is all but impossible to
accurately record spoken language in standard orthography.

I once taped a conversation (between myself and one other
person) lasting about 40 minutes, and later transcribed it,
in its entirety, complete with all hesitations, repetitions
and deviations, without the slightest difficulty (except,
of course, that it was a long and tiring job).

You might be astounded by *The First Five Minutes*, a work produced by
C. F. Hockett et al., ca. 1959, which attempted to provide a fully
explicit transcription of the first five minutes of a psychiatric
session. It is a _very_ large book. (But then, you've said you acquire
books but don't read them.)

Can you give me some idea of what is meant here by "fully
explicit"?

They sought to record _every_ aspect of communicative interaction that
could have any meaning at all.

But that's nothing: Deborah Tannen has made a 30-year career of
analyzing a single Thanksgiving dinner from ca. 1975. Every one of her
books, both technical and popular, draws extensively on that single
database for her groundbreaking studies of conversational interaction.

In a psychiatric context, that could involve
much interpretation of a heavily loaded episode of human
communication (both verbal and non-verbal). For limited
practical purposes it is, I assure you, easy (and boring)
to transcribe a conversation - no doubt missing out vast
amounts of communication which a Sigmund Freud or a Derren
Brown might pick up on, but which are irrelevant to said
practical purposes.

Derren who?

Writing is a substitution for speech, the next best thing when direct
speech is unavailable; it is not speech, nor is it a 'recording' of
speech. It only *appears* to parralel speech.

By a startling coincidence, I have read all six of the Harry
Potter books (to date) aloud to my daughter, and a sequence
of words came out of my mouth that exactly paralleled the
sequence of words printed on the pages! What are the odds
on that happening? It must be magic.

Ms. Rowling's sequence of words bears little resemblance to the
sequence of words you, or she, would have used if you were telling the
story -- with exactly as much detail -- rather than reading it.

For instance, you would probably have used very few nonrestrictive
relative clauses in your narration.

Don't worry, I believe you. (Someday I might even bother to
learn what a nonrestrictive howsyourfatherthingy is.)

Ah, yes: the use of the comma (of punctuation generally) is apparently
not taught at any level of British schooling, as is manifest in every
piece of literature (popular or technical) emanating from a British
publisher. (The lack of teaching has been corroborated by several
British friends.)

By a similar amazing coincidence, electronic computers are
able to do decimal arithmetic without using a pencil; and
they can display graphics without using paint. They must
work by magic, too.

Was that meant to be somehow relevant to your topic?

Of course, but don't sweat it. Usenet threads like this can
go all over the place if everyone follows up every subthread.

Grammar has a lot to do with formal structure, and formal
structure is, by definition, independent of the medium
of representation.

You _really_ need to learn something about language. Written and
spoken language are _vastly_ different.

Again, I believe you. I was reacting to a denial that
written and spoken English parallel each other in any
way. A denial of that denial is not an assertion that
they parallel each other in every way, or even in most
ways. Simple logic, no?

(If you don't believe me, please re-read the post I
was responding to. If I somehow took it too literally,
do please tell me what nuance I missed.)

Mr. Tortoise can quite hold his own (except on the matter of phonics
vs. whole-word).

Or, more accurately, it is dependent
only on certain formal properties of the medium, such as
linearity. A phrase structure grammar, for example, can
generate terminal strings consisting of formal "tokens",
and these strings can then generate several representations
in different media. There would be no point, and a lot of
redundancy, in writing completely separate grammars for
spoken and written English!

How clever of you to have picked up on a graphic device that was
current in linguistics about 40 years ago!

Chomsky type 2* grammars have been the bread and butter of
computer science for longer than that, at least since the
first Algol report in 1958, if my memory serves (but it
often doesn't - head like a sieve, in fact). This is not:
(a) a graphic device; (b) restricted to the linguistics
of natural languages; (c) obsolete; or (d) merely a clever
toy. (I have actually used the things, you know; written
programs to interpret them into parsers, with semantics.)

*(At least, I seem to remember that it was 2. I could look
it up, but why bother?)

I don't know -- those categories are from IIRC 1953, before he had
seriously begun to study English grammar -- and never made it into
linguistics.

At no point have I claimed that such a simple device comes
anywhere near describing the full complexity of language.
(Nor, of course, did Chomsky.)

You keep reading things into my words that I definitely
didn't mean to put there. (I hope I'm not doing the same
to you or Flying Tortoise. Of course you should tell me
if I am.)

If mention of phrase structure grammars seems irrelevant
(or quaint) to you, for some reason, then just think of
whatever kind of grammar you prefer: does it not contain
a level of description at which there are structures that
could be instanced indifferently in speech or writing?
Is there not something like subject-verb-object in there?

Of course. But to say that is to say nothing. _All_ language(s) has/ve
those.

If not, how /does/ it parse something like "the cat sat
on the mat"?

If you consult a corpus, you'll find that such a sentence virtually
never occurs in natural language. No utterance occurs without a
context.

And does it insist that a sentence like
that can only be either spoken or written, but not both?
Can it not account for the frickin' obvious relationship
between the written sentence (that I have just typed and
caused to be displayed as pixels on your screen) and the
spoken sentence (which unfortunately you cannot hear me
say)?

I'm still waiting to discover whatever embarrassingly
simple way it is that I keep misinterpreting the things
that you and FT say (sorry, I mean write, I mean type),
which seem to me to deny what every schoolboy knows
(and surely even every professional linguist knows).

That's the problem. What every schoolboy "knows" tends to be very,
very different from what every professional linguist knows.

The graphic device is quite irrelevant to the question of the
qualitative difference between spoken and written language.

Read my lips (oh I'm sorry, you can't): I AM NOT DENYING
THAT THERE ARE PROFOUND DIFFERENCES BETWEEN SPOKEN AND
WRITTEN ENGLISH. There, can you "hear" me "shouting"?
Does that help at all?

And this isn't just some pedantic technical point, of
interest only to academic linguists; it is sheer common
sense.

Common sense about language tends to be considerably more off the mark
than such common-sense observations as that the sun revolves around
the earth.

I am as aware as anybody of the limitations of common sense.
When you find a complete replacement for it, I'm sure the
world will beat a path to your door.

Language, like many other things, is best described at
several simultaneous levels, and /of course/ spoken and
written English differ enormously at the physical level
of representation, and /of course/ they resemble one
another enormously at the more abstract level at which
syntax functions, and to which semantics is (er ... are?)
attached.

Your "more abstract level" is the level at which _all_ human languages
have a great deal in common.

No, it isn't. So there. Yah, boo, sucks. Yo momma.

Well, what do you expect me to say? You're just being
silly. (Perhaps having a subtle joke at my expense, I
don't know.)

Perhaps you imagine that I'm talking about some (real
or imaginary, I don't know) "universal grammar" such
as everyone started babbling about when Chomsky started
talking (oops, writing) about it? Well, I wasn't.
Think again.

They differ immensely on all the levels in between.

Meanings are not attached directly to graphical or
auditory representations of words (sentences, phrases).

Oh, do you have a theory of semantics all your own, then?

Huh? I wish I had a theory that would explain what you
are on about.

For Heaven's sake, I only meant this: whatever the word
"cat" (for example) means, it is not solely and directly
the meaning attached to the sequence of letters C-A-T,
nor is it solely and directly the meaning attached to the
sound I make when I read the letters C-A-T aloud. It
would be absurd to discuss the meaning of "cat" with
sole reference to the way the word /looks/ or /sounds/.

No, it wouldn't. The vast majority of all human languages have never
had a written form, and the written form is of no interest whatsoever
to the general linguist (although, as a linguist, my specialty happens
to be writing systems).

Like it or not (and you seem bizarrely not to, but I
/must/ be misinterpreting you), there is a word "cat",
which can be written and/or spoken. That is all I
mean about a more "abstract" level, not tied either
to writing or speech (and not divorced from either).

I may have described the level of syntax as "abstract"
in its independence of the physical medium (or rather,
media) of representation, but it is rather these media
that are abstract, and formal syntactic units that are
concrete, because of their connection with semantics.

Sorry, that's simply incoherent and hence unintelligible.

You mean you don't understand it, and you think that that
is all my fault. I'm sure /some/ of it is my fault. My
point, which I will now try to say (sorry, type) in other
words, is simply (very simply! - this, as I keep saying,
is nothing but common sense) that the word "cat", which
I just described as "abstract" (because of its partial
independence from particular physical representations),
is, for practical purposes, more "concrete" than those
completely arbitrary representations (there is nothing
about cats that demands that the noun for them should
be spelt C-A-T, or be pronounced ... well, you know!),
because it is connected somehow with its /meaning/, the
concept of a cat. (And I repeat that I do not want to
oversimplify the vast subject of the philosophy of
language, about which I /admit/ I know almost zilch,
since I read so few books ...)



I don't mean to oversimplify the latter connection -
because philosophers, linguists and other eggheads have
written libraries about the arbitrariness (considered
separately) and incommensurability (considered together,
as in questions of translation) of natural languages -
but it seems indisputable (at least

...

read more »-

I wonder how much more there is ... (I can't click "read more" without
losing my response so far.)

You _really_ would do better to come to sci.lang, where there are
people who can explain things more clearly for you.

.

Relevant Pages

  • Re: THE QUIZ...
    ... >>>You are of course free to use 'language' in this very broad sense. ... > Grammar is what it is, irrespective of the theoretical approach or model. ... by Chomsky, however, not a practical one; ... language in animals - I did not study linguistics and Chomsky but I ...
    (sci.lang)
  • Re: Polynesian and South American place names
    ... You are suggesting that Japanese gen, ... >> linguistics that would be required to make such a claim ... You appeared to be arguing that it was absurd to suppose a language ... We know in that case of the athabascans ...
    (sci.archaeology)
  • Re: Different kinds of Linguistics
    ... Analytic linguistics takes language data and tries to explain it. ... Corpus linguistics is the most obvious example. ... _corpus-based_ grammar, and my ideal has been to account for all or ...
    (sci.lang)
  • Ebooks Forum
    ... Prague Linguistic Circle Papers. ... First Language Attrition, Use and Maintenance. ... Comparative Linguistics and Grammaticalizatio ... Bilingual Education and Bilingualism, ...
    (sci.med.nutrition)
  • Re: Different kinds of Linguistics
    ... Analytic linguistics takes language data and tries to explain it. ... Corpus linguistics is the most obvious example. ... Ref p4 of his "Functional Grammar of Nunggubuyu", ...
    (sci.lang)