In article
<13887824.1145.1334013591976.JavaMail.geo-discussion-forums@pbbnw6>,
DKleinecke <
dklei...@gmail.com> wrote:
> On Sunday, April 8, 2012 11:27:20 PM UTC-7, Nathan Sanders wrote:
> > In article
> > <12080790.1383.1333931410504.JavaMail.geo-discussion-forums@pbgg10>,
> > DKleinecke <
dklei...@gmail.com> wrote:
> >
> > > There are languages in which utterances are mostly verbs. There are
> > > nouns,
> > > of course, but they are few and far between.
> > >
> > > I am curious to what extent there are languages in which utterances are
> > > mostly nouns. There are verbs, of course, but they are few and far
> > > between.
> >
> > Are you referring to type frequency (i.e., frequency of what's listed
> > in the dictionary) or token frequency (i.e., frequency of what's used
> > in conversation)?
>
> I intended what you call token frequency. See my reply to Peter.
Which then raises the question: why? Or rather, why not type
frequency instead? Or why not both? You could very well find a
language with a disproportionately large number of verbs in the
lexicon, but a disproportionally large number of nouns in utterances,
and then find another language with exactly the opposite pattern.
> > If the latter, which kinds of conversations? Spoken, written,
> > technical, casual, speeches, between friends, between family, between
> > employer-employee, ...?
>
> I hate to do this - yes. All and/or any. The more focused the better.
I would easily believe that different styles of conversation could
have different frequencies. And if they do, how would these different
styles be weighted with respect to each other? What if casual speech
is verb-heavy, but formal speech is noun-heavy?
> > And for "noun" and "verb", do you just mean full words, or would you
> > count, say, an incorporated noun stem as a noun? For example, in
> > Nahuatl, you can say:
> >
> > (1) niccua tlaxcalli
> > ni-c-cua tlaxcal-li
> > 1S-3S-eat tortilla-ABS
> > 'I am eating tortillas'
> >
> > But you can also say:
> >
> > (2) nitlaxcalcua
> > ni-tlaxcal-cua
> > 'I am eating tortillas' (lit. 'I am tortilla-eating')
> >
> > Does (2) contain a noun and a verb, or just a verb?
>
> I assumed one could separate out phonological words and assign them to verb,
> noun or other.
That's a bit of a dangerous assumption. It's not always obvious where
word boundaries are in any given language, because some languages
don't have any word boundary phonology that is distinct from morpheme
or syllable boundary phonology, and if every word boundary coincides
with a morpheme or syllable boundary, you can't tell where words end
and begin.
> Your second example is, in my opinion, a single word. I know
> next to nothing about Nahuatl but I would have to assure myself somehow that
> the tlaxcal in (2) was an actual incorporated noun and not a noun subjected
> to derivation into a verb.
Morphosyntactically, "tlaxcalcua" behaves as a verb (for example, it
shows subject-verb agreement with the ni- prefix) and "tlaxcal" when
part of "tlaxcalcua" does not behave morphosyntactically as a noun (it
doesn't allow any nominal affixes, like the usually obligatory
absolutive -li, or plural, or possession, or anything nominal
morphology).
But in terms of communicative content, and number and type of lexical
stems, (1) and (2) are equivalent. They only differ in the order of
the lexical stems (V-N in (1) versus N-V in (2)), and the required
inflections (nominal and verbal in (1) versus verbal only in (2)).
> > Are you counting different meanings of the same noun/verb as different
> > nouns/verbs, and if so, how distinct must the meanings be to count as
> > different words? For example, a "set" could be a sub-unit of a tennis
> > match, or a formal mathematical object, or a group of repeated
> > exercises (a set of push-ups), or a kit (a chemistry set), or a single
> > electronic apparatus (a television set), or a group of songs in a
> > single musical session, or a constructed location for a film, with
> > many of those being clearly semantically related, but still
> > identifiably distinct uses.
>
> This was strictly a formal syntactic question.
Finding a universal way to separate out the syntax from the lexicon,
the morphology, or the semantics is quite probably an utterly
impossible task.
> All nouns, whatever their
> meanings count exactly the same and all verbs likewise.
But the question is, what counts as a noun? Are you talking about
noun stems (a lexical definition)? Are you talking about words
(whatever those may be...) that exhibit nominal morphology? How are
you defining what is and is not a noun?
For example, what about gerunds, which behave partly like nouns (they
can be possessed, be in subject position, and take adjectival
modifiers: "my continuous smoking bothered him") and partly like verbs
(they can take nominal direct objects, verb phrase complements, and
adverbial modifiers: "defiantly smoking a cigarette bothers him", "my
having smoked a cigarette yesterday bothered him")?
Compare "they watched my dancing" to "they watched me dancing" to
"they watched as I danced". As you move from one sentence to the
next, the instance of "dance" seems more verb-like and less noun-like.
And English also has something like noun incorporation, as in "I went
grocery shopping" (cf. "I went shopping for groceries", which clearly
has one noun)? And what about compounds, like "he's a rat catcher",
or "there's a mousetrap"?
It's just not clear to me that nouns and verbs can be so easily
catalogued in English, let alone in a wide variety of languages with
vastly different morphosyntax.
> If there is any
> semantics to be done it would be restoring implicit copulas in languages with
> non-verbal sentences.
Do those restored copulas count towards the number of verbs? If so,
what about other types of missing words, like the missing noun in "I
have two dogs, and John has one" or the missing verb in "Mary ate
fish, and John, ham"?
> > What about different verb+particle constructions that aren't fully
> > compositionally transparent? For example "look forward" means
> > 'anticipate', "look out" means 'be cautious', "look after" means 'take
> > care of', "look over" means 'inspect', and "look up" means 'search for
> > in a reference', "look into" means 'invesitage', and you have to learn
> > the meanings of each verb+particle separately from the others.
>
> These questions matter, of course, in communication, but not in my question.
> All the particles belong to class "other" and are not counted either way.
> The analysis of these verbal adjuncts is an extremely interesting - but it
> was not what I inquired about.
It would matter if you were interested in type frequency (I didn't
know when I asked these questions which frequencies you were
interested in).
> > > The most noun-heavy texts I have ever examined are English (American to
> > > be
> > > precise) engineering magazines. Not only are they filled with nouns but
> > > what
> > > verbs do appear are generic and bland. Almost nothing is communicated via
> > > verbs.
> >
> > Communicative content is another matter altogether from noun/verb
> > frequency!
>
> I agree - but I think there is a correlation. My text analysis techniques put
> in front of me sentences where all the non-verbal material has been replaced
> by symbols. In the material in question it was impossible to form any notion
> of what was going on from the verbs alone.
I don't see how that's any less true for nouns. If you know that
"dogs" and "cats" are the only nouns in a sentence, you have no idea
which did what to which, if anything even happened at all! The full
sentence could be "the dogs chased the cats", or "the dogs were chased
by the cats", or "the dogs saw the cats", or "the dogs wanted to chase
the cats", or "the dogs didn't chase the cats", or "the dogs and cats
did nothing", or "dogs and cats are natural enemies", or even "there
were no dogs and cats"!
> > And I'm not sure I agree with you that verbs don't communicate much.
> > There are very few meaningless or redundant verbs in English. There's
> > the "do" that shows up in yes-no questions (though it usually conveys
> > the tense of the sentence, so it isn't completely meaningless). I
> > guess one could argue that the existential "be" in there-existentials
> > doesn't add any information that the "there" doesn't already add,
> > except again the tense, of course. Likewise for the "be" used in
> > passives, and maybe the aspectual helping verbs "be" and "have".
> > There may be a few more, but nearly all verbs add something to the
> > meaning of an utterance, even if it's just the tense.
>
> I agree - which is why finding a text where verbs communicate very little is
> so striking. Ordinary English is, in my opinion lacking any statistics, not
> especially noun-heavy.
>
> > > Are there any useful statistics in the literature? Most corpus studies
> > > are
> > > not focused enough. How about other languages than English?
> >
> > My hunch is that, no, the literature isn't useful, in part, because
> > the question is difficult to frame, and even once framed, it may not
> > even ask anything important (and thus, no one cared enough to answer
> > it).
> >
> > But I don't know the corpus linguistics literature very well, so my
> > hunch could likely be wrong.
>
> I could put on my Whorfian hat and claim it must make a difference in the way
> you think.
Why you put on such a weird hat?