[Lexicog] how to handle phrases

[Ron Moe]

?????Ken Keyes raised several questions about phrases that I would like to
see discussed. The Qazaq phrase  "az qaldy" "little left" may simply be a
straightforward way of saying that most of something has been used: "There
is little left of the flour." = "The flour is almost gone." In this case it
is not a lexical phrase. However if it is used in other contexts: "We are
little left there." = "We are almost there," then I would say it is an
idiom. The English phrase 'not quite' is like this. You can say, "We are not
quite there," but you can't say, "We are quite there." The phrase 'not
quite' is an indivisible whole and is filling the same slot as 'almost' or
'nearly': "We are almost/nearly/not quite there." It should be noted that
you can say, "The bottle is quite full," and "The bottle is not quite full."
But the two expressions are almost synonymous.

Here are the issues I see in dealing with phrases: First, phrases constitute
a fairly high percentage of lexical items in English. The percentage may be
less in the polysynthetic discussed lately on this list, but I have had
several people report to me that phrases constitute about 25% of the lexical
entries in their dictionaries. I believe I already raised this issue on this
list.

Second, the part of speech label for a phrase should not be the same as for
a single word. A phrase like 'on the other hand' may function semantically
like a conjunction, and may fill the same syntactic slot. So it may be
tempting to call it a conjunction. However I would prefer calling all
phrases 'phrase' and then subcategorizing them. But at this point we run
into serious difficulties. Can we call 'on the other hand' a 'conjunction
phrase'? What about idioms that combine a verb and noun, like 'kick the
bucket'? Semantically this may be the equivalent of the verb 'die', but
syntactically it is not a verb. It is a syntactic verb followed by a noun
phrase. So we might want to simply label it as an 'idiom'. There is one more
difficulty here. A noun phrase has a noun as its head. But 'on the other
hand' does not have a conjunction as its head. It looks like a preposition
phrase. When we classify words, we are talking about three different
factors:

(1) Inflectional category--'boy' is inflected differently than 'man'. In
Bantu languages each noun belongs to an inflectional category that
determines the singular and plural prefixes and also affects concordial
agreement.

(2) Syntactic category--'the man' 'John' and 'I' can all fill the subject
slot in a clause. Most every language has a syntactic category 'noun' that
generalizes what slots the word can fill--head of a noun phrase, subject,
object, complement of a preposition phrase, etc.

(3) Composition--'backpack' is a compound noun composed of two nouns, 'on
the other hand' is compositionally a preposition phrase.

It has been my observation that these three issues get confused in practice.
We also confuse syntax and semantics when we talk about 'count nouns' and
'mass nouns', or 'transitive verbs' and 'intransitive verbs'. Although it
can be argued that these labels reflect phenomena that operate on both the
semantic and syntactic levels. If a language has inflectional categories, it
is normal practice to label them in the dictionary. In Bantu dictionaries we
place the noun class immediately after the part of speech (muntu, bantu
n.1/2 person, someone, people...). What we call 'part of speech' is usually
syntactic class. We don't normally label the composition of a lexical item,
although we sometimes note it (foregone [fore + go] adj. ...).

So when we label the part of speech of a phrase, what do we label? There is
no inflectional category. Some phrases, like 'on the other hand', fill an
obvious syntactic slot. So do we call it a 'conjunction phrase'? A
syntactician may claim that 'the day after tomorrow' is a noun phrase. But
it functions as an adverb. So do we call it a noun phrase or an adverbial
phrase? One possibility would be to establish the practice of
differentiating between 'noun phrase' and 'nominal phrase', 'verb phrase'
and 'verbal phrase', 'adverb phrase' and 'adverbial phrase'. The first in
each pair would refer to a phrase with the respective element in the head
slot. The second in each pair would refer to an idiomatic phrase that
functions syntactically (and semantically?) like the respective word class.
I just looked at David Crystal's 'A dictionary of linguistics and phonetics'
and he distinguishes 'adverb' from 'adverbial' in a similar way. I don't
particularly like 'conjunctional phrase' but I can live with it. Since part
of speech is normally abbreviated in a dictionary, these could be
abbreviated as v.phr, n.phr, adj.phr, adv.phr, conj.phr, etc. Longman labels
some phrases this way, but leaves other phrases unlabeled.

The third issue with phrases is whether to make them main entries or
subentries. If they are made subentries, should they be subentries under
each of their constituent words or just the first word? If just the first
word, then do we create minor entries so that the user can find them? For
instance how would we enter 'on the other hand'? One option would be to put
it as a subentry under 'on':

on prep. ...
   on the other hand conj.phr. (used when introducing another opinion that
is as important to consider as the one you have just given) [Longman]

Another option would be to make it a subentry under 'hand':

hand n. ...
   on the other hand conj.phr. ...

We could then create a minor entry which would alphabetize under 'O':

on the other hand conj.phr. see under 'hand'.

The third option would be to make 'on the other hand' a major entry and
possibly create a minor entry:

on the other hand conj.phr. ...

hand, see also 'on the other hand'

A fourth issue is how to determine what is a lexical phrase and what is
merely collocation. There is grading between the two and no clear cut off
point. There are several tests. The primary one is semantic--does the phrase
mean something different than the sum of the parts? Another is
syntactic--does the phrase function in an unusual way? For instance 'on the
other hand' is always sentence initial or set off by pause: "The zebra, on
the other hand, has fewer ribs than the horse." So syntactically and
phonologically it functions like a conjunction. There are phonological and
other tests but I'll quit. I'd like to see a nice list of tests.

Ron Moe
  -----Original Message-----
  From: Kenneth Keyes [mailto:ken_keyes at sil.org]
  Sent: Friday, August 06, 2004 1:11 AM
  To: lexicographylist at yahoogroups.com
  Subject: RE: [Lexicog] Almost


  Ron, et al,



  In Qazaq (one of the languages respresented on this list), there is a
rarely used word for "almost", "derlik." However, it more common to use an
adverbial phrase consisting of a reduplication of the verb stem with the
gnomic future -Vr (often called "aorist") and its negation, -MVs e.g.,
zheter-zhetpes, tiijer-tiijmes, bolar-bolmas.etc. E.g. Bazardan alynghan
pajdasyn kuen koeryge zheter-zhetpes. = "The profits taken at the Bazar are
barely sufficient for daily living." Bajdyng ajaqtary uzengige
tiijer-tiijmesten koek bajtal shaba zhoeneldi. = "Just before the rich man's
feet had almost reached the stirrups, the mare started to gallop."



  Bolar-bolmas has the the unfortunate connotation of insufficient,
inadequate, irrealis.



  Another way to say this is there is "little left"  i.e., "az qaldy."



  If you search for the Russian equivalent for almost "pochti" in
Russian-Qazaq dictionaries, you won't find an entry. That is because the
language handles it with an adverbial phrase. Am I correct in calling this
an adverbial? At least it is periphrastic. How should this be added as a
dictionary entry?



  Ken



  -----Original Message-----
  From: Benjamin J Barrett [mailto:gogaku at ix.netcom.com]
  Sent: Friday, August 06, 2004 7:25 AM
  To: lexicographylist at yahoogroups.com
  Subject: RE: [Lexicog] Almost



  In Japanese, there are three concepts for this. One is adverbial such as
almost dry (hotondo), another for the adjective such as almost $1000 (ni
chikai, gurai), and then one for "almost happened but didn't" (tokoro, sou,
etc.) I think you could make an argument to collapse the adverb and
adjective classifications; you might even catch a native Japanese speaker
using the wrong one in a performance error. These are in the near category
(chikai means near). I can't think of any word that covers both of them at
the moment,though.



  The almost but didn't seems clearly different, though there are some cases
that might cross such as "almost full" and "almost done".



  Benjamin Barrett

  Baking the World a Better Place

  www.hiroki.us

  -----Original Message-----
  From: Ron Moe [mailto:ron_moe at sil.org]
  Sent: Thursday, August 05, 2004 5:01 PM
  To: lexicographylist at yahoogroups.com
  Subject: [Lexicog] Almost

  I've been analyzing the domain 'Almost'. Longman's Language Activator
gives
  the primary meaning of the English word 'almost' as 'almost a particular
  number, amount, time, age etc.':

  'almost $1000' 'almost full' 'almost an hour' 'almost 50 years old'

  If this is true, then the other meanings would be secondary:

  someone has almost reached some place: 'almost there'
  something has almost reached some state: 'almost dry'
  something is almost finished: 'almost done'
  someone almost did something: 'almost fell'
  something almost happened: 'almost overflowed'
  something is almost the best/worst (the most extreme example of some
  quality): almost the best
  something is so similar that it is almost the same: almost human
  something is almost some quality: almost perfectly round

  This is a nice example of semantic chaining, and one can readily recognize
  the semantic links. However does this concept occur in other, especially
  non-Indo-European, languages? Does the concept include the range that the
  English word has? Would speakers recognize a semantic link between the
  translation equivalents of 'almost' in the examples above?

  English has a nice set of words in this domain:

  nearly, close to, close on, approaching, nearing, getting on for, not
quite,
  be pushing, just about, practically, virtually, all but, as good as, to
all
  intents and purposes, verging on, bordering on, more or less, pretty well,
  pretty much

  Many of these are secondary meanings of movement verbs or words in the
  domain 'Near'. It would appear that in English we conceptualize 'almost'
via
  the metaphor 'move near' or 'be near'. Is this also true in other
languages?
  In this case is the primary sense the one in 'to almost reach a place'? At
  least it seems that this is conceptually the more salient sense.

  Ron Moe
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lexicography/attachments/20040806/c2e24ad9/attachment.htm>