[Lingtyp] Narrative relative clauses

Daniel Ross djross3 at gmail.com
Wed May 9 22:32:38 EDT 2018

Dear Philippe and Juergen,

I will first quickly respond about the examples in the last email, based on
whether or not there is an adverbial making the relationship explicit. I
see no difference in acceptability at all. But it does seem to flow more
naturally with the adverb there. It's not a problem to leave that implicit,
but certainly makes sense to add it. So I agree with the point of the
argument although those minimal pairs do not demonstrate it as a contrast
in acceptability per se.

More generally, this discussion has made me think of a related phenomenon,
sort of from the opposite perspective.

My research focuses on pseudocoordination, especially of the verb-and-verb
type like English "go and get" or "try and do". In these cases we see a
subordinate-type relationship developing out of an apparently coordinated
structure. Diachronically we can define this as the word AND
grammaticalizing with new functions, beyond just normal coordination.

Although it is not the focus of my research and seems relatively rare
cross-linguistically, there are attested examples where AND grammaticalizes
as a relative clause marker. I briefly mentioned this type based on an
example from Fijian in a conference handout here, in the context of a
typology of pseudocoordination:
(See page 4, column 1 for the Fijian example and references, also mentioned
on page 6 in the list of pseudocoordination types. Unfortunately this
discussion was removed from the published paper due to limited space.)

Old High Fijian (Milner 1956: 35, cf. Dixon 1988: 252)
na waqa ka yali mai a koro
the boat and missing from the village
'the boat which was missing from the village'

In Páez, relative clauses are formed as coordinate clauses, but the
particular conjunction varies based on the switch-reference system (Slocum

In Kenyang, nɛ́ is multifunctional morpheme that is used as a coordinator,
relativizer and other (Tabe & Atindogbé 2017:104-5). (For the same marker
as a clausal coordinator see for example Seguin 1998:58.)

In Degema, the relative linker is nụ́, which differs from the nominal
coordinating conjunction nụ only in tone (Kari 1997:52-3), so I wonder
about a diachronic connection. In Hausa, the comitative dà which is used
for nominal coordination is also used to mark relative clauses (Newman

These are only a few examples I've collected that suggest a (rare)
cross-linguistic grammaticalization path, which in turn may suggest a
functional relationship between coordination and relativization in terms of
information structure for example. Juergen's comments at the beginning of
this discussion (and the various replies) seem to be the relevant other
side of this relationship.

There are also instances where the grammaticalization is not complete and
the reading seems to be pragmatic, based on a coordinate construction. One
particularly clear example is described for !Xun by Heine & König (2013, in
the Khoesan languages volume), where the conjunction *tè 'and' "is by far
the most frequently used conjunction ... and the one expressing the widest
range of funcutions. In addition to marking consecutive (and simultaneous)
events, it may also signal subordinate functions such as introducing
temporal, adversative, reason, manner and other clauses .... In some uses,
*té may even express the notion of a relative clause...":

Mā kē cŋ́ dāʼbā tà yà tci.
1SG PAST see child and N1 come
'I saw a child who is coming.'

I have seen other similar examples elsewhere but don't have them available
at the moment. (I would of course be interested in references to similar
examples if anyone is aware of them, or feel free to email me if you'd like
to discuss pseudocoordination in general and compare notes about the
extensive bibliography I've collected or suggest new references for me.)

Juergen, regarding your original question in general, I'm wondering if you
are asking about a grammatical or pragmatic distinction (or both/either).
Do you expect to find different grammatical encoding for these functions? I
don't really see that for English, and then for another language where this
is such a distinction, I wonder if the new type you're asking about would
be considered a "relative clause" in the traditional sense. But as a
pragmatic reading for relative clauses, I imagine this would be quite
widespread, and then like pseudocoordination there might be a next step
(e.g., 'pseudorelativization' in a parallel sense) where it is no longer
relativization at all. (The term "pseudosubordination" is sometimes used
for, e.g., clause-chaining constructions that have coordinate readings, the
inverse of pseudocoordination where formally coordinated clauses have
subordinate readings. All of these interesting represent intersecting paths
of grammaticalization, potentially even going back and forth if given
enough time, rather than the often proposed unidirectional paths for
certain constructions.)

Thank you for an interesting discussion!

Daniel Ross
PhD Candidate
University of Illinois

On Wed, May 9, 2018 at 5:24 PM, Philippe Bourdin <pbourdin at yorku.ca> wrote:

> Dear Juergen,
> To stick to English and to paraphrase you, it seems to me that when the
> antecedent is a participant of the matrix event, there are a couple of
> routinized lexical devices specialized in making the event relation
> explicit:
> (1a) *Sally gave the cup to Floyd, who proceeded **to smash it to
> pieces. *
> (1b) *Sally gave the cup to Floyd, who went on **to smash it to pieces. *
> Intuitively, what's distinctive about such devices is that they seem to
> have a very special affinity with this type of relative clause — a property
> that run-of-the-mill adverbs such as *immediately* or *then* may not
> have, at least not to the same extent.
> Let me pursue the same thread a little bit. I'm not a native speaker of
> English, but my sense is that there may well be a slight difference in
> acceptability between (2a) and (2b) and between (3a) and (3b):
> (2a)    *Sally gave the cup to Floyd. He then proceeded *
> *to smash it to pieces. *
> (2b)   (?) *Sally gave the cup to Floyd. He proceeded *
> *to smash it to pieces. *
> (3a)    *Sally gave the cup to Floyd. He then went on *
> *to smash it to pieces. *
> (3b)   (?) *Sally gave the cup to Floyd. He went on **to smash it to
> pieces. *
> It's as if, to borrow Anna's term, the relative pronoun in (1a) and (1b)
> exerted all by itself sufficient cohesive force to license *proceeded*
> and *went on*. When you change the hypotactic relation into a paratactic
> one, it might be a bit more natural to insert a cohesive prop, i.e. *then*
> (or *immediately*). But that's a just a hunch and I stand to be corrected
> by native speakers as to the difference in acceptability, which is
> admittedly very slight.
> In any event, if *proceed *and *go on *do have the function I'm
> attributing to them, this would nicely bring symmetry into the system you
> propose, at least for English.
> Best,
> Philippe
> ---
> Philippe Bourdin
> Professeur agrégé / Associate professor
> Département d'études françaises
> et Programme de linguistique
> Bureau YH 264
> Collège Glendon / York University
> 2275 Bayview Avenue
> Toronto, ON, Canada M4N 3M6
> On 2018-05-09 10:38 AM, Giacalone Ramat Anna wrote:
> Dear Juergen,
> in my paper "Persistence and renewal in the relative pronoun paradigm: the
> case of Italian", Folia Linguistica Historica 26, 2005, 115-138, I discuss
> narrative relative clauses and their function in Old Italian. I suggest
> that the emergence and diffusion of relative pronoun *il quale* in Old
> Italian was modeled  on the Latin "connecting relative" (Rosén) or
> relativischer Anschluss (Lehmann) . It was used as a device to enhance text
> cohesion..
> Best
> Anna
> Anna Giacalone Ramat
> Professor Emerita of Linguistics
> The University of Pavia
> Academia Europaea
> Honorary Member of the Societas Linguistica Europaea
> Dipartimento di Studi Umanistici
> Strada Nuova 65
> I-27100-Pavia
> tel. +39 0382 984486
> email: annaram at unipv.it <annaram at univ.it>
> https://www.academia.edu/34500598/CV_CURR
> 2018-05-08 21:10 GMT+02:00 Bohnemeyer, Juergen <jb77 at buffalo.edu>:
>> Dear colleagues -- I’m looking for any leads regarding both in-depth
>> single-language and typological studies on a phenomenon one might refer to
>> under the makeshift labels ‘narrative relative clauses’ or ‘eventive
>> relative clauses’. I will stick here to the former label (NRCs), since the
>> latter is more ambiguous. NRCs are a type of non-restrictive RCs that
>> distinguish themselves from other kinds of non-restrictive RCs by standing
>> in a narrative rhetorical relation to the matrix clause (or put
>> differently, by advancing a narrative story line to which the matrix clause
>> also contributes). Based on European languages, some subtypes could be
>> distinguished based on (i) the “antecedent” of the RC - the matrix clause
>> referent the RC picks up - and (ii) the expression of the semantic relation
>> between the matrix and RC events:
>>         • Antecedent is a participant of the matrix event; event relation
>> implicit:
>>           Sally gave the cup to Floyd, who smashed it to pieces
>>         • Antecedent is the matrix event itself; event relation implicit:
>>           Sally gave the cup to Floyd, which irritated Sam
>>         • Antecedent is the matrix event itself; event relation explicit:
>>           Sally gave the cup to Floyd, whereupon Sam left the room in
>> disgust
>> B and C are presumably structurally distinct from ordinary
>> (non-restrictive) RCs. On the other hand, A-type NRCs are interesting for
>> the form-meaning mismatch or semantic-pragmatic mismatch they involve. A
>> more technical definition of NRCs might be as follows:
>>         • Constructions involving a matrix clause and a dependent clause;
>>         • The dependent clause should share some of the language-specific
>> properties of RCs that set them apart from other types of dependent
>> clauses/predications in the particular languages;
>>         • The matrix clause event and the dependent clause event are
>> causally related and/or spatio-temporally contiguous.
>> I fully expect that the pragmatic functions of NRCs can be partially or
>> wholly fulfilled by other clause combination constructions that do not have
>> the language-specific trappings of RCs. Such functionally related
>> alternative means are very much part of the interest driving this
>> investigation.
>> Thank you in advance for any leads on this topic! -- Best — Juergen
>> --
>> Juergen Bohnemeyer, Associate Professor and Director of Graduate Studies
>> Department of Linguistics and Center for Cognitive Science
>> University at Buffalo
>> Office: 642 Baldy Hall, UB North Campus * Mailing address: 609 Baldy
>> Hall, Buffalo, NY 14260
>> Phone: (716) 645 0127
>> Fax: (716) 645 3825 * Email: jb77 at buffalo.edu * Web:
>> http://www.acsu.buffalo.edu/~jb77/
>> Office hours Tu 2-3:20 /Th 2:30-3:20
>> _______________________________________________
>> Lingtyp mailing list
>> Lingtyp at listserv.linguistlist.org
>> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
> _______________________________________________
> Lingtyp mailing listLingtyp at listserv.linguistlist.orghttp://listserv.linguistlist.org/mailman/listinfo/lingtyp
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20180509/c7cbedad/attachment.html>

More information about the Lingtyp mailing list