Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.philol.msu.ru/~otipl/new/main/articles/daniel/Daniel%20ReportedIllocution.pdf Дата изменения: Wed Nov 4 19:00:38 2009 Дата индексирования: Thu Dec 31 23:45:32 2009 Кодировка:

Reported illocution: data from several Daghestanian languages Michael Daniel daniel@qub.com

Moscow State University

1. GENERALITIES

Direct vs. indirect reported speech Direct RS `impersonates' the reported speaker; indirect RS re-tells the speech act from the actual viewpoint. Functional background behind switching from direct to indirect RS might be making it less energy-consuming for the hearer to analyze the information. Reference tracking The most salient feature is the shift of deictic viewpoint to that of the speech act participant (change of pronouns). (1) English/Russian He said, I am sorry. -> He said that he was sorry Subordination Direct RS in English / Russian looks like parataxisis a juxtaposition of two clauses, one introducing the speech act, the other its content: (2) English/Russian He said, I will wait In indirect RS in English/Russian, the speech act is introduced by the main clause to which the content is subordinate: (3) English/Russian He said that he would wait. Categories unavailable in indirect RS Categories impossible (at least in their primary function) in indirect reporting In English/Russian: e.g. imperatives, addresses (vocatives), invectives... (4) English/Russian He said, forgive me. -> He asked (me) to forgive him (but not *He said that forgive me)

1


(5) English/Russian He said, go away. -> He told me to go away. He said that I should go away. (but not *He said that go away) Here, switching to indirect RS requires switching to a different speech verb and/or syntactic construction indicating the manipulative character of the reported speech act. Note, however, that in oral Russian discourse the constraints are not the same as in the literary norm. Another non-reportable category is address. (6) English/Russian He said, John, come and help me out -> He called John by name and asked him to come and help him out. (But not *He said that John, come to help me) (7) English/Russian He said, you bastard, you robbed my house -> He called him names and said he robbed his house. (*He said that bastard! he robbed his house) (Paducheva 1996 more or less identifies this property with insubordinatability)

2. OVERVIEW OF RS STRATEGIES IN SEVERAL DAGESTANIAN L-S.
Bagvalal (Kibrik et al. 2001), Godoberi (Kibrik et al. 1996).

Languages and sources ­ Archi (Kibrik et al. 1977; corpus of glossed texts; elicitations with

Bulbul Musaeva), Agul (Merdanova with Daniel and Ganenkov 2006), Cahur (Kibrik et al. 1999),

Competition of several strategies: (juxtaposition of two clauses) (8) Archi: zero strategy (parataxis) jasa
now you.pl.DAT

zero-strategy, similar to direct speech in European, paratactical

wez

beloved-4

annu-t

a

4.do(IMP)

say.PF-EVID

bo-li

Now do what you want to do, he said. (9) Agul: zero strategy dada pu-ne zun hika-se father(ERG) say.PFV-PF I drive.IPFVFather said he'd drive the car.
FUT

masin car

2


(10) Bagvalal: zero strategy h
then we.excl-ERG say

isi-r

hei o-sa:

b-as-imi-se-e

N-tell-?-PRH-RPRT name name-OBL.M.SUP.LAT

ali alijeic-s

this-OBL.M.SUP.LAT we.excl(NOM) there

"isi

ongiri eka-b

be-PTCP.N

Then we told him: don't tell Ali Alijevich that we were here. Clitic strategy: The other (clitic strategy below) uses a reportative clitic, often transparently derived from the basic speech verb. Some properties of the clitic strategy (in at least some languages): with it) ­ as in Bagvalal ex. above: in some cases, excludes the presence of matrix speech verb, as in Archi or in Agul: (11) Agul: reportative clitic bans lexical speech verb pu-ne *dada HЭni bawa uza-se-aj father(ERG) say.PFV-PF cow mother(ERG) milk.IPF-FUTDad said mom will milk the cow (i.e. don't worry about it)
REPORTED

does not require any matrix speech verb (although may combine

Is a predicative head of the sentence, licensing its own arguments (ergative for the Speaker, a lative form for the Addressee), as in Archi (12) Agul: two ergatives dada HЭni bawa uza-se-aj father(ERG) cow mother(ERG) milk.IPF-FUT-REPORTE Dad said mom will milk the cow (i.e. don't worry about it)
D

jamu abaj k'al-r-si that parents.ERG die-EVID-REPORT Parents said he had died

(13) Archi: an auxiliary depending on the reportative clitic
ED-CVB

or even has its own (partial) paradigm, as in Archi: edi-li be.PF-EVI

D

Note: Archi reportative is a clitic which forms phonological unity with its host (normally, the example above).

verb of the RS), but may have morphosyntactic links outside its phonological host ­ as in the

verb=[REPORTED AUXILIARY] Third strategy is complementizer, but it is rare. Archi and Bagvalal lack speech verb complemetizer; Cahur grammar suggests Cahur has one,

3


but it is not obvious how different it is from the Bagvalal case; probably, Agul has one: (14) Agul: complementizer strategy jaa dada pu-ne uc

mic

qua-se

puna

father (ERG) say.PFV-PF REFL today here.to come.back.IPFV-FUT CMPL

Father said that he would come (back) here today. All strategies use finite predicates in reported speech clauses Finally, two points irrelevant for the purposes of this presentation ­ just for general information processes and purposive clauses (in all languages considered) or hearsay evidential: (15) Agul: outside speech verbs proper gadaji meni q'u-ne boy (ERG) song (ERG) make.PFVThe boy sang that he loved the girl.
PF

The strategies cover a wide range of usages, including mental

uci-s me rus kande-a puna refl-DAT this girl want-PRS CMPL

(16) Agul: hearsay evidential dad bagah quwa-se-aj

father tomorrow come.back.IPFVFather is coming back tomorrow, they say. (17) Archi: `X-called person'

FUT-RPRT

azi nust'apa:-r-si l:aIma-t:u bosor edi man be.PST Hazhi Mustapa-RPRT-CVB rich-ATTR There was a rich man called Hazhi-Nustapa (18) Archi: purposive subordination ita ju-w-mu
then this-1-OBL.1(ERG) inside person(NOM) 8.be-INTRG-VERIF say.PF-CVB whistle SAY.PF-EVID

jasul adam

i-r-kus

bo-li

sit+ bo-li

He wistled in order to know whether there was anybody inside the palace. (lit. to know whether anybody is inside, having said)
English/Russian, and the clitic strategy corresponds to subordinative idirect strategy in English/Russian.] [The first idea that comes to mind is that zero strategy corresponds to zero strategy in

3. REFERENCE ISSUES
Daghestanian languages may use direct ~ indirect ~ logophoric mode of reference in the same morphosyntactic strategy (clitic strategy). Cf. Archi:

4


(19) Archi: first person refers to the reported speaker (direct reference) er-t:-ib olo ik-mul-ce-s hot-ATTR-PL we.excl.GEN heart-PL-OBL.PL-DAT s:ax-du-t w i -t tapanci-li-n gulla b-ezd--to-r. gun-OBL-GEN bullet III-enter.PF-POT.NEG-NEG-REPORTED rusty-ATTR-IV your.SG-IV The bullet of your rusty gun will not enter our hot hearts, they say. (20) Archi clitic strategy: first person refers to the actual speaker (indirect reference) to-w-mu asba-r, zu ez ow-qe-r that-I-ERG caution-RPRT LOG.ERG I.DAT do.IV.PF-POT-REPORTED Just wait, he says, he will show it to me (he will teach me a lesson), he says. (21) Archi clitic strategy: logophoric marking to-w-mu zon zu-l:u l:a cili-si c'eba:-r that-I-ERG I LOG.OBL-COMIT together Azerbaijan-ALL go.1IMP-REPORTED He tells me, let's go to Azerbaijan together (with him). The same RS clause may combine two different modes of reference (cf. also the last Archi example): (22) Archi: direct + logophoric ita saat os-mi-n rigi-li-t
then

time(NOM) one-OBL-GEN time.span-OBL-SUP(ESS) say.IPF-CVB.AUXDEP <1>become.PF

wr-si

iti

zu-n-er wa-s

LOGOPH.1.OBL-GEN-RPRT pocket-OBL-IN(ESS) inside apple(NOM) 4.be.PF-RPRT you.sg.OBL-DAT want-CVB 4.find.PF-COND-RPRT

zip-l-a

jasul ans

ed-er

an-si

o-mcis-er

At 1 AM my husband tells me that he has an apple in his (jacket) pocket, if you (i.e. I) want one. (23) Bagvalal: direct + logophoric in-sa bisdi-b as q'oca-m-o wec'e-ala log-OBL.M.DAT you.PL-GEN.N money want-N-CVB not.be-RPRT <...> I don't want your money ­ he said <...> Possible mode of reference combinations are direct + logophoric or indirect + logophoric< apparently not direct + indirect. Previously, I have argued that these are cases of incosistent direct ~ indirect opposition. However, now I note that there are no examples where direct and indirect modes of reference are combined; either of them may combine with logophoric reference, but not with each other (at least in my examples and in the examples quoted in grammars). Thus, if we admit that logophoric reference has a different nature from direct / indirect RS distinction, the utterances quote above become consistent ­ at least they obviously do not 5


place the hearer in an awkward situation trying to decipher which of the `I' and `you' pronouns has been used in the same clause with direct or indirect mode of reference. zero strategy: using direct reference and no logophorics combines them with logophoric (when applicanle); logophoric reference may be obligatory (Archi, Agul) reference complementizer strategy (Agul) requires logophorics and indirect clitic strategy: using direct and indirect modes of reference and

4. RE-ARRANGED MAPPING

(The explanation that I present below, may be in slightly my own terms, in fact is suggested by Svetlana Toldova in a small paper of 1999 ­ or at least one may come to these conclusions following her lines) It follows from (Toldova 1999) that the two main strategies of reporting speech in Daghestanian should be paralleled typologically not with the typical direct vs. indirect speech means in English/Russian, but with weaker indirect reportative means like the Russian , (while Agul complementizer strategy may well be a closer parallel to English/Russian that/ subordinative strategies). (24) Russian: the case of and (a)
he:NOM



i

say:PRS:3SG

,

i,

he:NOM RPRT ill:M:NOM:SG

, ,



you.sg:NOM

hearer

.
go:IMP

He says that he is sick and that you should go. (b)
he:NOM



i

say:PRS:3SG

,

i,

I:NOM

RPRT ill:M:NOM:SG

, ,



you.sg:NOM

speaker

.
go:IMP

He says that he is sick and that I should go. As I said before, these utterances may be ambiguous (just as Daghestanian utterances discussed above) in the sence that there are no formal signals which consistently indirect) Either direct or indirect, but consistent. The parallel is not only good for reference tracking issues: typologically matches Daghestanian reportative clitics in that it: mode of reference is used ­ however, they are either consistently direct or

6


originates from a speech verb () is not a subordination strategy

allows both direct and indirect reference of personal pronouns has strong evidential (hearsay) connotations that on the left of the figure but that on the right of the figure. The problem is that one first attempts the left side of the table, trying to compare the two central strategies in English/European on the one hand and in Daghestanian, on the other. European
direct

In other words, the correct mapping of reporting strategies is not the one

Table 1 ­ Correspondance between major strategies Daghestanian zero European zero Daghestanian
direct indirect

zero

zero clitic

indirect

subordinative (wrong approach)

clitic

subordinative

subordiinattiive subord na ve (more efficient approach)

5. REPORTING IMPERATIVES

Imperatives are readily reported not only in the zero strategy, but also in the clitic strategy, and even in complementizer strategy in Agul. (25) Godoberi (Haspelmath in Kibrik et al ed., 1996: 185) im-u-di t'alab hi was-u-c'u-ru in-s-o qard--u

kaati

father-OBL-ERG demand do.PST son-OBL-CONT-EL self.OBL-OBL.M-ERG letter(NOM) write-IMP-CIT

Father demanded of his son that he writes a letter. (26) Bagvalal (Kalinina in Kibrik et al ed, 2001: 518) h isi-r hei o-sa: "isi
then we.excl-ERG say

this-OBL.M.SUP.LAT we.excl(NOM) there

ongiri

eka-b

be-PTCP.N N-tell-?-PRH-RPRT name name-OBL.M.SUP.LAT

b-as-imi-se-e

ali alijeic-s

Then we told him: don't tell Ali Alijevich that we were here.

7


(27) Archi (Kibrik 1977, examples) za:r-si marci-maj naI' oq'e-r I.CONT-ALL every-PL.ERG milk give.IMP-REPORTED Everybody tell me to give milk. (28) Agul (Merdanova et al. 2006) naq' dada pu-ne za-s jaa mic

qisaw

puna

yesterday father (ERG) say.PFV-PF I-DAT today here.to come.back(IMP) CMPL

Father told me yesterday that he would come (back) here today. In terms of the Table above, imperative reporting has an Almost complete match between Daghestanian and English/Russian (except for Agul complementizer strategy). But the more important question is: Are imperatives reported in strategies other than zero strategy indicators of direct speech? On one hand, they have very good reasons to be non-reportable indirectly and are not reported in indirect speech in English/Russian. They are as strongly attached to the speech act as deictics are, and even stronger: imperatives require a speaker and a hearer, and if something happens with the speaker ~ hearer reference, why should imperatives be left untouched? imperatives require presence of the hearer-addressee because they include a manipulative component These are arguments in favor of saying that imperatives are intrinsically direct speech category. However: with indirect reference items, as in (the examples like that are not unique): (29) Archi: imperative combines with indirect reference sal-a bo-li parca-li-s mele-tu-t lo bo-li,-- zon cor oci-s zaba-r-si ow-li unlike direct reference of personal pronouns, imperatives are combined

fox-OBL(ERG) say.PF-EVID king-OBL-DAT male-ATR-4 child(NOM) 4.do.PF-EVID say.PF-EVID I.NOM name(NOM) 4.stand-INF come.IMP-RPRT-CVB.AUX 4.AUX say.PF-EVID

i",-- bo-li.

The fox said: a child was born to the king, I am being summoned to give him a name (lit. I to give the name come they say) Probably, morphosyntactic factors there ­ ban on subordinated imperatives? (Paducheva 1996) ­ accounts for the fact that English/Russian do not allow imperatives in indirect speech while Daghestanian allow with clitic strategy ­ the latter is clearly less subordinative (but, again, why Agul with the complementizer strategy then?) (we are ready to say that it all boils down to subordination, as Paducheva who largely identifies the non-quotability and non-subordinatability)

8


another explanation ­ reporting imperatives by imperatives is just a typological option substituting for special lexical and subordination strategies used in English/Russian (cf. ex. 4 and 5 above ­ but then it follows that imperatives are no problem for indirect reporting, contrary to what has just been discussed, and it is unclear why English invents special strategies to report them indirectly in the first place)

6. REPORTING ADDRESSES
Addresses are irreportable indirectly: Addresses, on the contrary, are just as irreportable in indirect contexts as in Russian/English, or almost. If we come back to the previous context, the primary reaction to its version with the pronoun substituted by a vocative form was firmly rejected: (30) Archi: clitic strategy bans vocatives a. *sl, b. sl, cor oci-s zaba-r zaba

fox(NOM).EXCL name(NOM) 4.stand-INF come.IMP

Hey fox, come to give a name (to the child)! ­ they said. cor oci-s Hey fox, come to give a name (to the child)! ­ they said. (31) Archi: vocative requires `external' interpretation buwo, zaba-r
mother.EXCL come.IMP-RPRT

fox(NOM).EXCL name(NOM) 4.stand-INF come.IMP say.PF-EVID

bo-li

Mom, he's calling (me or you or someone else) speaker but of the actual speaker) *naq' dada pune
say.PFV-PF

(lit. `Mom (speakeri's mother), come, heji says' - not the mother of the reported

(32) Agul: vocatives irreportable in complementizer strategy za-s
I-DAT

an

k'irk'

mic
here.to

qisaw

puna

yesterday father (ERG)

dear sonny

come.back(IMP) CMPL

Father told me: sonny, come back here. (33) Agul: vocatives irreportable in the clitic strategy *dada gadaji-s ja an k'irk' mic saw-aj
father (
ERG)

son-

DAT

VOC

dear

son

here.to

come (IMP)-

REPORTED

Dad says to the boy, sonny, come here. Just to be totally frank ­ here is the only example I found of reporting an address in combination with reportative particle (according to Cahur grammar, even in subordinative context)

9


(34) Cahur (Ljutikova, Bonch-Osmolovskaya in Kibrik ed. 1999: 525) ­ possible counterexample ic- iwho-jn: gade, dal hil-e-w

girl-ERG say.PF-A boy

stick 4.give-IMP-RPRT

The girl said, hey boy, give me the stick. But on the whole, vocatives clearly tend to be limited to zero strategies. The possible reason might be the same as one of the reasons for imperatives: manipulative address requires presence of the hearer-addressee because it includes a component (in this case, manipulating the addressee's communicative behavior) This is similar to imperatives; however, the imperatives are reported in all strategies, while vocatives are reported with zero strategy only. possibly, some morphosyntactic factors such as vocatives being morphosyntactically independent clauses, `satellite utterances' (close to Paducheva's approach; what about reporting utterances with fronted topic? but that would not explain why imperatives are not reported in English/Russian) Anyway, here is the final version of the table __ above is as follows: Table 1 - updated European

Daghestanian

reporting

reporting

vocatives

reference

vocatives

reference

direct

direct

zero

zero

imperatives

imperatives

logophoric

reporting

reporting

reference



clitic

10


reference

reference

indirect

indirect

subordinative

subordiinattiive subord na ve

7. CONCLUSIONS

reportative clitics, is not parallel to subordinative indirect speech in English/Russian but to the strategy attested in European only peripherally - strategy in Russian (Toldova 1999 and consequences) By and large, reporting vocatives belongs to the prototypical direct speech domain Imperatives are however easily reported in any strategy in Daghestanian The reason why imperatives and vocatives should behave differently is unclear (it is even unclear whether this happens because the table __ is wrong, or because imperatives behave differently, or the division of labour between direct and indirect is diffrerent in different languages.

One of the main strategy of reporting speech in Daghestanian, the use of

Kibrik 1977: .. . . . ., 1977. Kibrik et al. (eds) 1996: Godoberi. Lincom Europa. Kibrik et al. (eds) 1999: . . Kibrik et al. (eds) 2001: . , , . . Kibrik, in prep.: e-corpus of Archi: A.E. Kibrik, A. Arkhipov, M. Daniel, S. Kodzasov (supported by the NSF grant `Five languages of Eurasia', with software support by A.D. Nakhimovski and Tom Myers) Merdanova, Daniel, Ganenkov 2006: Reported speech in Agul (a paper read at the Paris Caucasian Morphosyntax Workshop in December of 2006) Paducheva 1996: .. . . 11

REFERENCES


Toldova 1999: .. '99 ( ).



12