[Lexicog] part of speech for phrases

[List Facilitator]

----- Original Message -----
From: <rrhodes at cogsci.berkeley.edu>
To: <lexicographylist at yahoogroups.com>
Sent: Friday, January 16, 2004 7:20 PM
Subject: RE: [Lexicog] part of speech for phrases


> John,
> I'm not sure I agree with the end of your argument. The fact
> that the English word saw is both noun and verb without any
> derivational marking doesn't require that you point that out in
> either of the entries.
> This question touches on the relationship between morphology
> and syntax. The usual principle is that dictionary entries list the
> part of speech that is reflective of the EXTERNAL syntax of the cited
> form. That a particular form happens to have the INTERNAL morphology
> of different part of speech is irrelevant. That is the assumption
> behind the answer John Roberts gave Ron Moe in the exchange about
> part of speech for phrases (quoted below).
> The choice to assign part of speech on the basis of external
> syntax is not arbitrary. It is dictated by languages with little
> internal morphology like English and Chinese. There is only the
> external syntax to tell us what the part of speech is.
> Where the confusion arises is that the typological norm for
> morphologically complex languages is that particular morphologies are
> aligned with the external syntactic word class. As such, in normal
> case you can infer the external syntactic class on the basis of the
> morphology. But it ain't necessarily so. The fact that a language has
> verb forms that can be used as syntactic nouns without being overtly
> derived means that they are nouns. (There's a parallel argument about
> languages that treat nouns predicatively, but I'll pass on that for
> now.)
>
> As far as the dictionary goes, one way to do it is to list
> the most "neutral" nominalization and make a note that sends you to
> the grammar to see how such deverbal nominals are treated. So
> I'd argue your version
>
> >i'base vt cut with
> >we'base n saw
>
> is a principled way to do it, and not an uncomfortable compromise you
> were "driven to".
>
> Rich Rhodes
>
>
> >On Fri, 16 Jan 2004, Ron Moe wrote:
> >>  I agree that we need to look at lexical units using each of the three
> >>  criteria. But my understanding is that 'part of speech' means
'inflectional
> >>  and/or syntactic class'.
> >
> >In dealing with Siouan languages I have been driven to conclude that
these
> >are actually not quite the same thing.  A lot of things that are
> >morphologically verbs (or halfway in from being verbs) are used regularly
> >in syntactic and lexical terms as nouns.  For example, 'house' might be
> >'he dwells' or 'he dwells there'.  In some languages this might be
> >possessed with nominal schemes, while in others 'my house' is 'I dwell
> >there'.  Similarly, 'saw' (the tool) might be 'he cuts things with it'
and
> >'my saw' might be 'my one cuts things with it' or '(the thing) one cuts
> >things with that I have' or 'I cut things with it'.
> >
> >There are pure noun forms, and nouns can be used as verbs to a certain
> >extent, but the only derivational mechanisms are verb derivation and
> >clausal syntax, so any derived form has to be a verb, or a clause, the
> >latter often reduced to a compound.  There no specific mark of
> >nominalization.
> >
> >Since many of these verbish nouns are quite lexicalized, and may have
> >mandatory "indefinite object" prefixes added that would only be added to
> >the verb when the object was specifically "indefinite," it's not quite
> >possible to do something like
> >
> >i'base vt cut with, n saw
> >
> >You would actually have to say something like
> >
> >i'base vt cut with
> >we'base vt-indef cut things with, n saw
> >
> >But the vt-indef is productive and any vt can have that prefix, so you're
> >driven to
> >
> >i'base vt cut with
> >we'base n saw
> >
> >But we'base is verbal in derivational morphology and may well be
> >inflectable as a verb, so you're faced with something like
> >
> >i'base vt cut with
> >we'base [vt-indef]n saw
> >
> >I'm inclined to conclude that noun and verb in morphological terms is
> >something different from noun and verb in syntactic and/or lexical terms.
> >
> >(Exx. are Omaha-Ponca)
> >
> >JEK
>
>
> On 16 Jan 2004 John Roberts said:
>
> >
> >On 15 Jan 2004 Ron Moe said:
> >>
> >>So here's my question: Does anyone know of something written on the
subject
> >>of labeling the part of speech for multi-word lexical items? Can anyone
> >>clarify the issue or give examples from your language? For instance the
MDF
> >>manual is good on principles for determining the parts of speech of a
> >>language, but says nothing about phrases.
> >
> >The basic problem is that there are three sets of criteria that you can
> >appeal to in defining what a word is and these criteria are independent
of
> >defining a unit's syntactic function. You can use phonological
(phonological
> >unit), morpho-syntactic (morphological unit) or lexico-semantic (lexemic
> >unit) to define a word unit. These do not always converge so that all the
> >criteria form a unit.
> >
> >For example, each of the English examples below are phrasal constructions
of
> >some type but they are all unitary lexemes. 'off duty' and 'by-product'
are
> >phonological words but the others are not.
> >
> >an off duty policeman    'off duty' is a PP functioning as an adjective
> >'by-product' is a PP functioning as a noun
> >He worked round-the-clock. 'round-the-clock' is a PP functioning as an
> >adverb
> >a hit-and-run accident     'hit-and-run' is a V and V phrase functioning
as
> >an adjective
> >'pass up' is a V + P phrase functioning as a verb
> >'kick the bucket' is a V + NP phrase functioning as a verb
> >
> >On the other hand, they each function as a unitary part of speech, such
as
> >adjective, adverb, noun or verb. They should also all be entered in an
> >English dictionary because they are unit lexemes.
> >
> >John Roberts
> >Linguistics Consultant
> >SIL WEG
>
>
>
> --
> ******************************************************************
>
>   Richard A. Rhodes
>   Department of Linguistics
>   University of California
>   Berkeley, CA 94720-2650
>   Voice (510) 643-7325
>   FAX (510) 643-5688
>
> ******************************************************************