John's quote isn't just a return to Ed's original query about the circular,
tautological, mutual definition of "noun" and entity, or "verb" and
process. I think John's quote points to two basic postulates of SFL:
a) The ineffability of grammatical categories. This follows from the
general philosophical position that meta-language doesn't stand outside
language, but is only language folded back upon itself. A
linguistic category is going to be defined linguistically, and in no other
way.
b) Form is explained by function, but function in turn has to be explained
(and cannot be explained by form). This follows from quite specific
instances of explaining form by function, such as the observation that if a
language has tenses, these will be applied to verbs and not to nouns, the
observation that if a language has number, that number is readily applied
to entities but less readily so to processes, that it is easier to have
proper nouns than proper verbs (verbs that describe singularities) etc.
Ed says that a) and b) are research programmes and not axioms, and of
course he's right. Wierzibicka's universal semantics is a rather Quixotic
attempt to research a) and any time I sit down to try to come up with
systemic-functional description of Korean, I am trying to do b).
Notice that Ed's initial misgivings were that the arguments sustaining a
natural relationship between meaning and wording were mere axioms. Now he
is giving the opposite objection--that they are NOT axiomatic but rather
research programmes, and therefore empiricist. So now Ed needs to be a
little more specific about what kinds of arguments sustaining the natural
link between meaning and wording WOULD be acceptable.
I have offered three: philosophical (language is not a supernatural
phenomenon but is in fact one of many semiotic systems used by a naturally
evolved species of animal), phylogenetic (language does seem to have
evolved from non-linguistic systems that Peirce would call iconic to those
that he would call inidexical to those he would call symbolic), and
ontogenetic (children begin learning language by
attempting--unsuccessfully--to recapitulate phylogenesis, by going from a
bistratal system to a tristratal one).
Ed would say that the first argument is merely axiomatic, and the last two
arguments are empirical. OK, Ed, but the first argument isn't MERELY
axiomatic because it engenders not one but two research programmes. And
those last two aren't MERELY research programmes; they aren't just waiting
for patterns to emerge from data. They are linked together and they are
linked to the first argument.
The ontogenetic argument is actually one instance of the phylogenetic
argument, and the phylogenetic argument in turn logically derives from the
philosophical one. Form does indeed have to be explained by function, and
function does indeed have to be explained by something else. In his last
ISFC plenary in Vancouver, Halliday did tell us what that something else
was: he argued that it was something called "semiohistory", the evolution
of meaning.
(I notice, Ed, that you don't touch the philosophical argument at all--this
is perhaps why you keep finding empiricism under the rug....)
dk
2022년 3월 6일 (일) 오후 10:23, Edward McDonald <laomaa63@xxxxxxxxx>님이 작성:
Dear David and all:
I'm sure this time I'm definitely way beyond the statute of limitations,
but I'm squeezing in a little more reflection partly in the context of Chris's
posting of the week almost gone on theoretical vs descriptive categories,
and also because of the recent discussion around Wendy's transitivity
question, which seemed to me - even factoring in my well-known lack of
interest in the grammar of my mother tongue - to be largely going round in
circles.
I find your arguments for the posited "naturalness between semantics and
lexicogrammar" - from "nature", phylogeny, and ontogeny - interesting, but
they still seem to me more in the way of axiomatic postulates - "this is
what a natural relationship should look like" - than evidence, let alone,
as I said before, a research program. My scepticism stems from my
impression that as soon as people start getting into the detail of "shared
commonalities", the kinds of things that they identify tend to look
suspiciously like the semantics of a particular language or set of
languages.
Anna Wierzsbicka's "natural semantic metalanguage" - which she defends as
almost non-theoretical, as simply "empirical evidence" based on, as it
were, scouring a great range of data - is for me fundamentally undercut by
its insistence on a componential style of semantics, which *my*
"empirical experience" - as language learner, language teacher, and
translator (not to mention language describer) tells me is a totally
simplistic and distorting model of how linguistic meaning operates.
I recently came across a great quote from the Chinese linguist Yuen Ren
Chao - one of the very greats of modern linguistics in my book - about an
earlier attempt at a "universal semantics". He starts with the arresting
claim - "There is no true co-ordinate conjunction “and” in Chinese" - and
then quietly adds a deadly little footnote:
Morrish Swadesh once compiled a “culture-free vocabulary” of two hundred
words, intended for use in eliciting native equivalents when recording the
languages of all cultures. In alphabetic (and, therefore, semantically
quite arbitrary) order, the first group of five words were “all,” “and,”
“animal,” “ashes,” and “at.” Except for “animal” and “ashes” it would be
very difficult to find word-for-word Chinese equivalents for these words.
That is, three of the first five words were not at all culture-free, but
were features of the English language, or at least of the Indo-European
languages.
(Yuen Ren Chao, Notes on Chinese Grammar and Logic, 1955, p. 34)
I suppose what I'm looking for would be a kind of cross-linguistic
methodology that would be both empirically rich enough to take in a
typologically
representative spread of languages, and theoretically sophisticated enough
to frame its generalisations in robust ways. This may well be the kind of
model Sapir and Whorf were working towards before their tragically pr
emature deaths - and before their supporters rashly trapped them in a
"hypothesis" which was never how they framed the issue and which should n
ever have been framed that way.
But I also think - and this may be rash of me to utter on an SFL list - that
no one theory of language is going to have everything that's needed for the
task. I would love to see a really solid cognitive theory of language -
with the cognitive based squarely on what we know of cognitive processes
in general, and linguistic cognition in particular, and not just a semantics
in disguise - matched up against a really solid social theory of language
- ok I'm happy to let SFL stand in for that one ha ha - and see what
generalisations
actually held from both points of view, and where were the deep points of
disagreement. Where theories disagree is often the most interesting place
to look, in my experience. People like David Butt would no doubt have some
ideas in this area - and I'm sure there's useful and instructive work being
done out there.
But the final question must still be: are the language sciences - and the
social sciences and the cognitive sciences and of course philosophy - ready
for such an attempt - to really take seriously the claim that each language,
in a carefully defined sense, "creates its own world". Are we mature enough
to move past "the almost religious alarm and disgust that greeted the ideas
of Sapir and Whorf" (to quote myself) and convincingly dethrone "Standard
Average European" from its still almost unchallenged position as the "default
model"?
best to all
Ed
On Wed, Feb 16, 2022 at 4:19 PM 데이브드켈로그_교수_영어교육과 <dkellogg60@xxxxxxxxx>
wrote:
Ed:
If I understand you correctly, you worried that the claim that there is a
natural relationship between meaning and wording is circular, because it
rests on evidence such as the fact that nouns (which tend to take number
and gender rather than tense or aspect) realize entities (which tend to be
countable and classifiable) while verbs realize actions (which unfold over
time). Number seems to have a natural relationship with entities and tense
with time. Similarly, there are common nouns to represent universals and
proper nouns to represent individuals, but there do not appear to be proper
verbs to represent unrepeatable acts. This appeared circular to you because
it says that nouns have these properties, and of course anything having
these properties are nouns.
That this is NOT a circular argument should become clear if you apply it
to the relationship between wording and sounding. It is not true that the
word "loud" has be be spoken loudly and only marginally true that words
that begin with nasals have something to do with your nose. So the
relationship between meaning and wording is at the very least MORE natural
than that between wording and sounding (or rather--between wording and
articulation, because the relationship between wording and prosody is
actually natural).
But I also deny that the fit between word classes and semantic categories
(or, for that matter, the "context-metafunction hook-up" that Hasan always
talked about) really is the crucial evidence. First of all, the
philosophical argument for a natural relationship is unanswerable: we're
part of nature, not some supernatural force. Secondly, there is the
phylogenetic argument: I can't think of any element of semantics or
phonology that doesn't have some obvious counterpart in the animal world
(this is why Halliday insists on "consciousness" rather than "mind").
Thirdly, there is the ontogenetic argument: until roughly two or three,
children have a "natural" semantics and a "natural phonology" and their
first attempts at putting the two together (e.g. Nigel's "intonational"
grammar) appear to be natural (i.e. prosodic and not articulatory) as well.
For the very young child, it is actually true that the word "loud" has to
be spoken loudly, and that words that begin with nasals have something to
do with sneezles, sniffles, snot, snogging and snoutishness.
dk
David Kellogg
Sangmyung University
New article in Mind, Culture, and Activity:
Vygotsky’s pedology of the adolescent: the discovery of sex and the
invention of love
50 free online copies available at:
https://www.tandfonline.com/eprint/CTJKPMUKGBANU5SNYG8T/full?target=10.1080/10749039.2022.2030362
New article with Irina Leopoldoff-Martin and Bernard Schneuwly.:
"Developing Concepts of Development in Vygotsky’s pédologie: Introduction
to the Special Issue" in Mind, Culture and Activity.
https://www.tandfonline.com/doi/full/10.1080/10749039.2022.2026972
50 free online copies available at:
https://www.tandfonline.com/eprint/DNZTW3URNHIIMCYBKWX6/full?target=10.1080/10749039.2022.2026972
New book of translations from L.S. Vygotsky's pedological work with
Nikolai Veresov.
L.S. Vygotsky's Pedological Works, Volume 2: The Problem of Age
Downloadable pdfs at:
https://link.springer.com/content/pdf/bfm%3A978-981-16-1907-6%2F1.pdf
and
https://link.springer.com/content/pdf/bbm%3A978-981-16-1907-6%2F1.pdf
