Definitions of Conjunction and Disjunction
Ben & Mandy Pehrson
ben-mandy_pehrson at SIL.ORG
Thu Oct 12 05:59:57 UTC 2006
Hi all,
I'm sorry I have been silent in this discussion after I brought up the
exclusive/inclusive distinction. I'm not going to be able to give much time
to this, but I thought I'd at least chip in with my agreement concerning the
distinction in RST between relation definitions and formal realizations. But
the same thing could be said to argue that RST should have two different
relation definitions for two different relations that happen to be expressed
with the same form in some languages. To me, the difference between
exclusive disjunction and inclusive disjunction seems like a fairly
significant distinction. However, if we do want to group similar relations
under one definition, I think the proposed definition with the part about
"not necessarily exclusive" in parentheses does a good job of allowing for
inclusive and exclusive disjunction.
I appreciate the example given that reflects exclusive disjunction at one
level and inclusive disjunction at another:
> (1) A: What d'ya wanna do tonight?
> B: We could go to the movies or have a quiet evening home.
I'm sorry I am not able to answer how any languages that differentiate
between inclusive and exclusive disjunction would mark this. I suspect that
if a language doesn't have a form that can be used ambiguously for inclusive
or exclusive 'or' it would not be able to encode such a double meaning. The
reply would either use the inclusive 'or' to express the multiple desires
that the person has, or the reply would use the exclusive 'or' to express
the mutually exclusive possibilities.
By the way, the only source I was thinking of concerning a language with
different forms for exclusive and inclusive 'or' was taken from an article
by Ettien Koffi concerning the translation of 'or' in African languages. He
writes, "Ditammari a language of northern Benin, has three or-type
coordinators; ke and yaa are used only for exclusive coordination, while yoo
is used to link inclusively."
Ben
> -----Original Message-----
> From: RST Discussion List [mailto:rstlist at LISTSERV.LINGUISTLIST.ORG] On
> Behalf Of Gisela Redeker
> Sent: Friday, September 15, 2006 5:32 AM
> To: RSTLIST at LISTSERV.LINGUISTLIST.ORG
> Subject: Re: [RST-LIST] Definitions of Conjunction and Disjunction
>
> Hi Ben,
>
> I see your point, but I think it gets tricky with presentational
> disjunctions ("speech act" relations in Sweetser's (1990) terms) as in B's
> answer in (1):
>
> (1) A: What d'ya wanna do tonight?
> B: We could go to the movies or have a quiet evening home.
>
> Here the semantic alternatives ('we could do X' OR 'we could do Y'), and
> thus the choices offered to A, are mutually exclusive, while B is at the
> same time saying that both options are things she might want to do tonight
> ('I'd like doing this AND I'd like doing that'). -- How would cases like
> this be marked in the languages you are thinking of?
>
> Another problem I see with your argument is that RST has (wisely, I think)
> stayed away from linking the relation definitions directly to linguistic
> realization ('cues', cohesive signals). Like all coordinating conjunctions
> (see, e.g. Schiffrin 1987, Redeker 1991), 'or' (in English) has a wide
> range
> to uses and thus does not differentiate between relations that are very
> clearly different on other grounds. An example where 'or' means neither
> inclusive nor exclusive disjunction are corrective uses like the
> afterthought in (2) and the speech act conditional in (3):
>
> (2) We could go to the movies... Or let's just have a quiet evening home.
>
> (3) We could go to the movies tonight - or do you have to work late again?
>
> The relation signalled with 'or' in (2) should probably be analyzed as
> ANTITHESIS, and in (3) as UNLESS -- both relations that no one would
> suggest
> lumping together with disjunction.
> I don't think the relation DISJUNCTION is or should be defined by the
> occurrence or 'or' (or its equivalents in other languages). On the other
> hand (as e.g. Knott and Sanders (1998) have argued for linguistic markers
> of
> coherence relations in general) it is presumably no coincidence that many
> languages use the same means (conjunction, cue) for marking inclusive and
> exclusion disjunctions. As Ben sys in his post, the two are very similar
> and
> speakers may sometimes want to remain vague and not mark the distinction.
>
> Gisela
>
>
> -----Original Message-----
> From: RST Discussion List [mailto:rstlist at LISTSERV.LINGUISTLIST.ORG] On
> Behalf Of Ben & Mandy Pehrson
> Sent: vrijdag 15 september 2006 4:04
> To: RSTLIST at LISTSERV.LINGUISTLIST.ORG
> Subject: Re: [RST-LIST] Definitions of Conjunction and Disjunction
>
> Maite,
>
> As far as disjunction is concerned, there is a difference between
> exclusive
> disjunction and inclusive disjunction. Some languages use different
> 'conjunctions' to mark each type of disjunction. If the RST definition
> that
> is used for disjunction states "An item presents a (not necessarily
> exclusive) alternative for the other(s)," then these two different types
> of
> disjuction are lumped into one relation. Perhaps that's fine for English
> since we use 'or' to mark both kinds of disjunction and sometimes it may
> be
> ambiguous. But even in English there are ways to differentiate the
> relations
> 'and/or' make the intended function more clear. While I think the
> definition
> as it stands is meant to be inclusive of more than one kind of relation,
> it
> could be interpreted to more narrowly define what we mean by 'and/or' but
> not the exclusive use of 'or'.
>
> Ben
>
> ________________________________________
> From: RST Discussion List [mailto:rstlist at LISTSERV.LINGUISTLIST.ORG] On
> Behalf Of Andy Potter
> Sent: Thursday, September 14, 2006 8:20 PM
> To: RSTLIST at LISTSERV.LINGUISTLIST.ORG
> Subject: Re: [RST-LIST] Definitions of Conjunction and Disjunction
>
> Maite,
>
> I don't know how you'll know when a consensus has been reached, but it
> seems
> to me that the definitions recovered by Gisela capture disjunction and
> conjunction in a way that is both useful and consistent with the spirit of
> RST.
>
> Andy
>
> ----- Original Message -----
> From: Maite Taboada
> To: RSTLIST at LISTSERV.LINGUISTLIST.ORG
> Sent: Wednesday, September 13, 2006 10:39 AM
> Subject: Re: [RST-LIST] Definitions of Conjunction and Disjunction
>
> Hi all,
>
> I had a student this summer doing annotations, and she started using the
> disjunction and conjunction relations, because they were listed listed in
> the RST coder. Below are a few examples of the text that she coded with
> those relations. I'm not sure they all apply, since I haven't gone over
> them, and she was working without precise definitions (she sort of came up
> with definitions based on the examples she found). She found many more
> disjunctions than conjunctions, and most, if not all, of them had only 2
> nuclei.
>
> As for why the relations are not listed on the web site any more, I simply
> don't know. I transferred whatever was on the original site around mid-
> 2004,
> and it looks like Bill had removed them by then. I don't know why.
>
> If there's consensus in the list about definitions, I'd be happy to post
> them again on the site with examples.
>
> - Maite
>
> ---------------------------------------------------------
> Examples:
>
> Disjunction
> He either had Langdon flashback to a lecture he gave in a class somewhere
> (yawn)
> or he had two or more characters discuss the issue to death.
>
> Will Peter ever get out,
> or will he die in the tower?
>
> She can't get a good night's sleep,
> [disjunction, nucleus1] because either Grandma is snoring [disjunction,
> nucleus2] or somebody is breaking into her house [sat., result] and waking
> her up.
>
> either that,
> or he was specifically looking for a movie contract for this story,
>
> Apparently he just liked the name Betsey or kept forgetting he'd already
> used it.
>
>
> Conjunction
> This didn't make me like the story any less nor did I find it hard to
> follow-
>
> Disney provides great access to transportation and every cast member is
> ready to provide detailed directions and tips for getting to your desired
> destination quickly.
>
> ---------------------------------------------------------
>
>
> At 19:20 12/09/2006 +0200, Gisela Redeker wrote:
>
> I have been using the following definitions, which I am pretty sure I got
> from Bill Mann's page in 2003:
>
> Relation
>
> Constraints on each pair of N
>
> Intention of W
>
> Conjunction
>
> The items are conjoined to form a unit in which each item plays a
> comparable
> role
>
> R recognizes that the linked items are conjoined
>
> Disjunction
>
> An item presents a (not necessarily exclusive) alternative for the
> other(s)
>
> R recognizes that the linked items are alternatives
>
> What I like about these definitions is their flexibility: I distinguish
> subject matter and presentational uses (for these and for the multi-
> nuclear
> relations of LIST and SEQUENCE).
>
> I've only now discovered that the current list of definitions on the RST
> site no longer includes these -- does anyone know why?
>
> Best regards,
> Gisela
>
> Gisela Redeker, Professor
> Department of Communication and Information Sciences Faculty of Arts,
> University of Groningen P.O.Box 716, NL-9700 AS Groningen g.redeker at rug.nl
> tel:
> +31-50-3635973 fax: +31-50-3636855
> http://www.let.rug.nl/~redeker
>
>
> Mick O'Donnell wrote:
>
> Hi Chris, Jelisaveta,
>
> The original document describing was RST:
>
> William C. Mann and Sandra A. Thompson 1987 "Rhetorical Structure Theory:
> A
> Theory of Text Organization". ISI Technical Report ISI/RS-87-190.
> (available from: http://www.sfu.ca/rst/05bibliographies/report.html)
>
> It mentions disjunction under multinuclear relations (p73). However, from
> a
> quick look, I don't think any details are given of this relation.
>
> Conjunction is not mentioned.
> However, there is a Joint relation (p76), which I think is too weak for
> conjunction: Joint asserts no relation between nuclei, while Conjunction
> should assert some relation amongst the nuclei.
>
> Note however that Bill Mann did add Conjunction to the relation sets for
> both classical and extended RST.
>
> Mick
>
>
>
>
>
>
>
> Jelisaveta Safranj wrote:
>
> Dear Chris,
>
> I have found something in Discourse Tagging Reference Manual written by
> Lynn
> Carlson and Daniel Marcu.
>
> Disjunction is a multinuclear relation whose elements can be listed as
> alternatives, either positive or negative.
>
> Examples:
> [Call it a fad.] [Or call it the wave of the future.]
>
> In the aerobic phase, for instance, lactic acid and lactate are still
> produced, [but they are consumed by less active muscles] [or metabolized
> in
> the liver] and so do not accumulate.
>
> Conjunction is not mentioned at all.
>
> Hope it helps
> Jelisaveta
>
>
> ___________________________________________________________
> $0 Web Hosting with up to 200MB web space, 1000 MB Transfer
> 10 Personalized POP and Web E-mail Accounts, and much more.
> Signup at www.doteasy.com
>
> _____
> Maite Taboada
> Assistant Professor
> Department of Linguistics
> Simon Fraser University
> 8888 University Dr.
> Burnaby, B.C. V5A 1S6
> Canada
>
> Tel: 604-291-5585 Fax: 604-291-5659
> mtaboada at sfu.ca - http://www.sfu.ca/~mtaboada
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/rstlist/attachments/20061012/97caee3d/attachment.html>
More information about the Rstlist
mailing list