A note on on verb/object order and head/relative clause order



Download 485.1 Kb.
Page1/5
Date conversion25.05.2016
Size485.1 Kb.
  1   2   3   4   5
A NOTE ON ON VERB/OBJECT ORDER AND HEAD/RELATIVE CLAUSE ORDER*

Guglielmo Cinque


That some typological relation exists between the order of the object with respect to the verb and the order of the relative clause (RC) with respect to its Head is known since Greenberg (1963). While VO languages (SVO, VSO and VOS) have postnominal RCs, prenominal RCs are found almost exclusively in OV languages.1 In other words:
(1)a VO NRel

b RelN OV


These implications cannot be strengthened by adding NRel VO and OV RelN, because OV languages seem to show no clear preference for either a pre- or postnominal positioning of their RCs. This appears most clearly from Dryer’s (1992a) 543-language sample:
(2) Order of Relative clause and Head and the VO/OV distinction (source: Dryer 1992a,86)2

NRel RelN

OV 37 26

VO 60 1
Dryer’s conclusion that Verb/Object order and Head/RC order do not form a correlation pair in the same sense as Verb/Object and Adposition/Object do is very widely shared. See, among others, Hawkins (1994, 265,273);3 Croft and Deligianni (2001,3); Diessel (2001,446); Song (2001,244); Rijkhoff (2002,307).4

The mere numbers, however, may conceal the existence of a significant generalization relating the order of the verb and its complements to the order of the Head and the RC.

In their chapter 5 (“Relative Clauses”, pp.261-371), Mallinson and Blake (1981) list the 150 languages of their sample according to subject/verb/object order, and according to whether they display RC-Head order, Head-RC order, or both.5 The numerical results largely confirm (ante litteram) Dryer’s results in showing no clear tendency for OV languages (especially if languages with exclusive NRel and those with both NRel and RelN as alternative options are added together):6

NRel RelN both NRel and RelN

OV 5 17 12

VO 109 1 6
However, more telling than the actual numbers is to observe from their table which OV languages allow only the RelN order and which allow the NRel order as the exclusive or as an alternative order. The former group (Ainu, Amharic7, Basque, Burmese, Burushaski, Chibcha, Fore, Japanese, Kannada, Korean, Mongolian, Piro, Sherpa, Sinhala, Sri Lanka Malay, Sri Lanka Portuguese, Tamil, Telugu) appears to contain languages corresponding to Greenberg’s (1963,79) “rigid” type; the latter group (Adyghe, Fur, Galla (Oromo), Hindi, Hittite, Hottentot (Nama), Kanuri, Khamti, Marathi, Nubian, Quechua, Rashad, Sandawe, (Classical) Tibetan, Tigre, Turkish8) appears to contain languages corresponding to his “non rigid type”.9

Assuming this generalization to be essentially right, one could propose the following correlations:


(3)a. If VO then NRel

b. If “rigid” OV, then RelN

c. If “non-rigid” OV, then NRel or both NRel and RelN
Even if possibly correct, such a statement would, however, fail to expose what is at the basis of these correlations. We submit that the correlation between V/O order, and the order of RCs and their Heads is intimately related to the order of complement and adjunct subordinate clauses w.r.t. the verb. In VO languages subordinate clauses follow the V, as they can, typically, in “non-rigid” OV languages (cf. Dryer 1980,130,172). In the same languages, RCs follow the Head. Subordinate clauses, however, do not ordinarily follow the V in “rigid” OV languages, which are more strictly V-final.10 In the same languages, RCs do not follow their Head, either.

The generalization could be phrased more perspicuously as follows:


(4)a In the general case, OV languages that do not allow postverbal subordinate clauses (“rigid” OV languages) do not allow postnominal RCs.

b In the general case, OV languages that allow postverbal subordinate clauses (“non rigid” OV languages) also allow postnominal RCs


If this generalization survives further scrutiny, then there may be a genuine correlation between V/(clausal) O order in the sentence and N/RC order in the DP.11

From the languages in the two Appendices below, which includes the OV languages of Mallinson and Blake’s own sample and a number of other OV languages, it appears that the generalization is basically correct.

Generalization (4) says that in those OV languages in which there can be a post-Head clause in the sentence ([..V Clause..]) there can be a post-Head RC in the nominal phrase ([..N RC..]).

In turn, the possibility for a clause to follow the V or the N seems to some extent related to the presence of initial complementizers. While preverbal and prenominal (finite) clauses have final rather than initial complementizers ([[Clause…..COMP] V/N], postverbal and postnominal (finite) clauses have initial rather than final complementizers ([V/N [ClauseCOMP.….]]).

Hawkins (1990,256) notes that VO languages are exclusively Comp initial, while OV languages are either Comp initial or Comp final (see also Dryer 1992a, sections 4.3 and 4.5,1992b; Diessel 2001):
(5) VO languages S’[Comp S] only

OV languages S’[Comp S] or S’[S Comp]


In the light of what we just observed about V/O order and RC/Head order, the double possibility in complementizer positioning of OV languages, vs. the single possibility of VO languages, leads us to expect that S’[S Comp] will be found preverbally in “rigid” OV languages and S’[Comp S] will be found postverbally in both VO and “non-rigid” OV languages. This appears confirmed by the following passage from Hawkins (1994): “[..] grammars that would potentially generate D [i.e., Comp S V] seem to have an extraposition rule converting D into A [i.e., V Comp S] [..]. This is true for Persian and for German. It is also true for the finite S’ structures of Yaqui and Turkish (cf. Dryer 1980). Moreover, in all the languages mentioned, Extraposition is obligatory in this environment, with the result that these languages exhibit a “left-right asymmetry” [..]: a rightward skewing for sentential direct objects, even in languages that are SOV for non-sentential objects [..].”(pp.263-64).12

Bayer (2001), noting that “Indo-Aryan languages with Dravidian contact often show a dual system of sentential complementation with clause-initial complementizers for clauses in post-verbal position and clause-final complementizers for clauses mainly in pre-verbal position” (p.11), makes the important observation that the initial and final complementizers are lexically different, and cannot be used interchangeably (i.e. “the lexical choice of the complementizer goes hand in hand with word order”, p.15). The so-called ‘quotative’ complementizers, which derive from verbs of saying, are necessarily final. The necessarily initial complementizers, instead, appear to have originated in noun-modifying clauses as relative pronouns (p.18ff).13 More important than their origin, though, is the fact, pointed out by Bayer, that they are differently specialized w.r.t the types of clauses they select, and seem to enter different structures. Observing that with postverbal clauses introduced by an initial complementizer there can be a nominal correlate “in the expected position to the left of the verb”14 (p.21) (cf.(6) from Bengali (Bangla) = his ex. (10)), Bayer suggests that perhaps they always do, and that when nothing appears one should posit an unpronounced nominal correlate:15

(6) chele-Ta e kOtha jane na *(je) baba aS–be

boy-CL this story knows not (that) father come-will

‘The boy does not know it that his father will come

This conjecture appears to be supported by the fact that postverbal finite clauses with initial complementizers (as opposed to preverbal ones with final complementizers) behave the same way as “extraposed” relative clauses and “extraposed” clausal complements of N(P)s. They are “frozen” in place; e.g. they cannot be topicalized (cf. Bayer 2001,18ff).

What all of this suggests is that to be clause initial is possibly a property of those complementizers that are nominal in character; i.e., that appear with RCs, with complements of Ns, and nominalized clausal complements of verbs.16

What is crucial from the present perspective is that such “initial” complementizers/subordinators turn out to be a feature of VO and “non rigid” OV languages.

To judge from Diessel (2001), a similar pattern is displayed by adverbial clauses: “While adverbial clause constructions that tend to precede the main clause/predicate only occur in OV languages in my sample, adverbial clauses that are commonly pre- and postposed occur in both VO languages and a significant minority of OV languages. If we look at the latter more closely, we find that (almost) all of them are marked by an initial conjunction or adverb, while adverbial clauses that usually precede the main clause/predicate always include a final subordinator (i.e., a final conjunction, adverb, or suffix). There is thus a strong correlation between the ordering of main clause/predicate and adverbial clause and the position of the subordinator in the subordinate clause: adverbial clauses including a final subordinator tend to precede the main clause/predicate, whereas adverbial clauses that are marked by an initial subordinator are commonly found in both initial and final position regardless of the order of verb and object.” (p.434). Also see Dryer (1992a, §4.5). Once again, the postverbal positioning of the adverbial clause in VO and, we take, “non rigid” OV languages appears to be a function of the initial subordinator/complementizer.

To summarize, we have suggested that, in OV languages, 1) the presence of prenominal RCs correlates with the presence of preverbal complement and adverbial clauses; 2) conversely, the presence of postnominal RCs correlates with the presence of postverbal complement and adverbial clauses17; and 3) the two correlations are related to the presence, in the three types of clauses, of final and initial complementizers, respectively. The latter claim is supported by the languages in Appendix II only partially, though. Of the 46 OV languages with postnominal RCs and postverbal complement and adverbial clauses considered there, only 13 (Brahui, Galla (Oromo), Georgian, Hindi, Hittite, Marathi, Pashto, Persian, Pima Bajo, Svan, Tùnεn, Turkish, Zazaki) have an initial complementizer in the three types of clauses; 8 (Bagri, Bangla, Gapapaiwa, Latin, Santali, Somali, Xakas, Yaqui) have an initial complementizer in two of the three types of clauses; 2 (Hopi and Teribe) have a final complementizer in two of the three types of clauses; 9 (Ala’ala, Coahuilteco, Evenki, Nama, Sandawe, Sentani, Shipibo-Konibo, Tol, West Greenlandic) show a (mainly final) complementizer for only one of the three types of clauses (the adverbial clause); 2 (Godoberi, Santali) show a final complementizer only for complement clauses; 9 (Desano, Eudeve, Kabardian, Kairiru, Manam, Northern Paiute, Quechua, Skou, Wichita) do not show any complementizer for any of the three types of clauses; and 3 (Canela-Crahô, Kuku Yalanji, Pech) have a final complementizer for all three types of clauses.

Despite this less than perfect correlation between the postverbal/postnominal positioning of the clause and the presence of a clause initial complementizer (which may in part depend on the limited character of the sample), we take the correlation to be real; and to follow from a property, recently discussed by Kayne (2000a, 2001, 2003, 2005), of the (initial) complementizer of finite subordinate clauses (whether complement, adverbial, or relative).
Initial complementizers. On the basis of various considerations, Kayne suggests that clauses are generated in their argument or adjunct position without a complementizer. They then move to their licensing position,18 to the left of which an overt complementizer is subsequently inserted. Exemplifying with German:19

(7)a [nicht [VP[DP[IP Fritz Maria kennt] [NP(es)]] glaubt]]

b [[DP[IP Fritz Maria kennt] [NP(es)]] [nicht [VP t glaubt]]]

c [[IP Fritz Maria kennt] C [[DP t [NP(es)]] [nicht [VP t glaubt]]]]

d (Weil Hans) [daß [[IP Fritz Maria kennt] C [[DP t [NP(es)]] [nicht [VP t glaubt]]]]]

(Because H.) that F. knows M. it does not believe


The property, here relevant, that complementizers have (in VO languages, and in many OV languages; i.e. those of the “non-rigid” type) is that of attracting to their left everything that follows their clausal complement, hence turning (7)d into (8)
(8) (Weil Hans) (es) nicht glaubt daß [er Maria kennt] t .. ’As he doesn’t think that he knows M.’
with the consequence that both the complementizer and the clause “end up” after the matrix verb.20

This movement could be thought of as a kind of ‘intraposition’, the “antisymmetric” analogue of the ‘extraposition’ operation that in earlier stages of the theory was assumed to derive (the string of words in) (8) from (the string of words in) (7)d (Kayne 1994).

If we take the overt (and abstract) complementizers of RCs to have essentially the same attraction property (as in fact Kayne 2000, 318f himself suggests), then the similarity in post-“Head” positioning of the clause in the sentence and in the nominal phrase (i.e., the generalization noted above) can be captured:21
(9)a [[we bought [which expensive book]] expensive book]

b C [[we bought [which expensive book]] expensive book]

c [we bought [which expensive book] C [t] expensive book]

d [that [[we bought [which expensive book]] C [t] expensive book]

e [which expensive book [that [[we bought t ] C [t] expensive book]]

f X [which expensive book [that [[we bought t ] C [t] expensive book]]

g [expensive book X [which expensive book [that [[we bought t ]] t ]] C [t]

e (I lost) the [expensive book X [which expensive book [that [[we bought t ]] t ]] C [t]


As noted, such “initial” complementizers turn out to be a feature of VO and “non-rigid” OV languages.

The case of initial complementizers in pre-Head position, as in (7d) above, is apparently rather marked. We interpret it as arising from the attraction of IP by a non-pronounced lower complementizer (the C of (7)d and (9)d), and from the (marked) property that the higher overt complementizer has of attracting nothing.22

Though again quite rare, the case of initial complementizers in finite RCs also seems to exist. It is generally stated, or assumed, that there are no languages with prenominal RCs that have an initial finite complementizer (e.g., Andrews 1975,44; Downing 1978,394). Yet, Galla (Oromo) (Cushitic), Sílli Greek (which is spoken in Asia Minor, and on which Turkish may have played a role), and Tigre (Ethio-Semitic), appear to be three such languages. See (10)-(12):23

(10) [kan [ kalēsa gale]] namtičča an arge (Galla (Oromo) - Mallinson and Blake 1981,289)

Rel yesterday arrived(finite) man-def I saw

‘I saw the man that arrived yesterday’


(11) [kiát [íra ]] perí (Sílli Greek - Song 2001,256)

Comp saw-I boy

‘The boy that I saw’
(12) [ la [zet fäggr mnna]] ’kl (Tigre - Palmer 1961,27f)

Rel marker oil it-comes from-it crop

‘the crop from which oil comes’
Although they are quite common in preverbal position (as well as in postverbal position) in VO languages, subordinator-initial adjunct clauses normally occur postverbally in OV languages, though some exceptions exist. See, e.g., (13).
(13) [kawu [nji yakin-də-ro ]] bəri bukin (Kanuri - Hutchison 1976,141)24

before water drink.1sg.impf-det-dat meal eat.1sg.impf

‘I will eat before I drink water
In (“rigid”) OV languages instead one typically finds in preverbal position subordinator-final adjunct clauses. See the quote from Diessel (2001) above and Dryer (1992a, §4.5).
Final complementizers. It is tempting to take the “final” complementizers typical of “rigid” OV languages to be the spell out of the lower C of (7) and (9); the one which attracts the “complement” IP or the relative IP (and which is ordinarily not spelled out in VO and “non-rigid” OV languages). See the case of complement clauses in (14), the case of relative clauses in (15) and the case of adjunct clauses in (16):25
(14)a Taroo wa [[Ziroo ga baka da] to] omotte ita (Japanese – Josephs 1976,367)

T. topic Z. nom. mad is Comp thought

“Taroo thought that Ziroo is mad”
b mene [[Madhu se bethane] ke liye] kaha (Hindi – Singh 1977,204)

I M. to sit for said

‘I asked Madhu to sit down’
c Chele-Ta [[or baba aS- be] bole] Sune-che (Bangla – Bayer 1996,255)26

boy-CLF his father come-FUT Comp hear-PST3

‘The boy has heard that his father will come’
(15)a [[Vok rool ?a pee] mii] lawthlawpaa ka mu? (Lai – Kathol and Vanbik 1999,434)

pig food 3sg give Comp farmer 1sg see

‘I saw the farmer who gave food to the pig’
b [[ngbā dzĭ ] ] tss̀tss kā ndzá (Lendu – Kutsch Lojenga 1987/2003,5)

child buy.PrC Comp banana ripen.P Neg

‘The bananas which the child is buying are not ripe’
(16)a [[ọ duma tún timi] seribi] arị́ waịị́ bó-mi (Ijo – Williamson 1965,78)27

3Pl song sing Cont.Past while I turn come-Past

‘While they were singing, I returned’
b [[ enu-nege-pi ] tawa] tarep war-an (Daga – Thompson and Longacre 1985,188)

spear-me-3sg medial lest dance get-1sg past

‘Lest he spear me, I dance about’
In this respect, “rigid” OV languages would lack the higher complementizers of VO languages (the one that attracts VP in the case of complement and adjunct clauses and the ones that attract the relative pronoun and the Head NP in the case of relative clauses). Alternatively, they would have unpronounced ones which fail to attract any material. The existence of languages with postverbal or postnominal (“extraposed”) complement or relative clauses with final complementizers appears to support the second alternative. In these languages, we may assume that the higher unpronounced complementizers retain the property of attracting the VP, or the relative pronoun and the RC Head. See (18),(19), and (20):28
(18)a cu-te i-mã amji jarẽn C [[cu-mã akĩn] na] (Canela-Crahô - Popjes and Popjes 1986,165)

3-Past 1-Tempry self told 3-Temp 2-like subord

‘He told me that he likes you’
b TohᲠslolyáya he C [[wakpála ektá ohįhpaye] ] (Lakota – Dryer 1980,132)

when you.know Q creek to fall Comp

‘When did you find out that he fell in the creek?’
(19)a domer C [[bor ĩ-ga ] li] (Teribe - Quesada 2000,129)

man 1sg see-ABIL REL

‘The man who saw me’
b tthik’íhí C [neyaa yet’ah golọ thehk’é síi] (Slave – Dryer 2003,31)

gun 2SG.son it.with moose 3.shot COMP

N Rel

‘the gun that your son shot the moose with’


(20)a ami ekhane eSe-chi C [[tomar SONge kOtha bol-bo] bole] (Bangla – Bayer 1996,255)

I here come-Pst1 you with speech say-Fut1 Comp

‘I have come here in order to talk with you’

b ?amá k’a way C [[ma híβe] mpes] (Tol - Holt 1999,51)

land dry Cop Neg Pres.rain.3 because

‘The land is dry because it doesn’t rain’


Circumpositioned complementizers/subordinators. The existence of two complementizers/ subordinators sandwiching the complement/adjunct or the relative clause also seems to constitute evidence for the hypothesized unpronounced higher complementizer, as it seems plausible to take the simultaneous appearance of an initial and a final complementizer to be the spelling out of both positions. See (21) for examples of complement clauses, (22) for an example of an adjunct clause, and (23) for examples of relative clauses:29
(21)a tuisi tu?i ke hu hamut bwika-kai (Yaqui – Dryer 1980,fn.7)

very good comp this woman sing-subord

‘It is very good that this woman sings’
b rəpšuû-qi -na ná ya šá tsáwa nεέma-qɔɔ-s Lə̀p-pə-reê (Tibetan – Bayer 1999,fn25)

goat-tail-erg comp-I top meat at all neg-want-comp say-perf/inference

‘The goat-tail said “I don’t want any meat”’
(22) [se mi-wi΄é a] mí-kò fi΄e (Fanti – Welmers 1946,72)

when 1sg-finish when 1sg-go home

‘When I’m finished, I go home’
(23)a mo e jó sáŋ á’á (Banjoun (Ghomala) -Watters 2003,255)

man Rel 3ps see.Past bird Rel

‘…the man who saw the bird’

b ŋa? [naŋ ka-keŋ εŋ ge-ya pola? naŋ] ge-mu ge-meŋ (Jabêm - Ross 2002d,281)

man DEM 1sg-give 3sg 3sg-go:3 Polac DEM 3sg-go.back 3sg-go:1

‘Has the man I sent to Polac come back or not?’


c kpàrà-wā:yi [ālī ei tà nā:b ] (Buli – Hiraiwa 2003,46)

farmer-REL C have cow(indef.) Subord.Particle

‘the farmer who has the cow’
d dispela man [ya em I stap long bus ya] em i redi.. (Tok Pisin - Lyovin 1997,424)

this man C he [3p] live in bush C he 3p ready..

‘This man who lived in the bush was ready..’

Internal complementizers. The case of Bangla “Comp-internal clauses” discussed in Bayer (1996,1999,247;2001,fn.12), Bhattacharya (2001), and references cited there, may possibly be another instantiation of the property that the (higher) finite complementizers have of attracting material to their left in “non-rigid” OV languages. Bayer and Bhattacharya point out that finite complement clauses can have an initial COMP if they follow the matrix verb ((24)a), but can no longer have an initial COMP if the complement clause precedes the matrix V ((24)b). In that case, the COMP is rather internal to the complement clause itself ((24)c). I would like to interpret both cases as consequences of the attraction property of the complementizer. Either the remnant following the complement clause – i.e., the matrix V(P) - is attracted (with the consequence that [COMP S] will be postverbal – see (25)), or part of the complement clause itself will (see (26) (with the consequence that the remnant – the matrix V(P) - can no longer be attracted, but has to stay in situ, to the right of its complement):w
(24)a chele-Ta Sune-che [je [or baba aS–be]] (Bayer 1996,255)

boy-CL hear-Pst3 that his father come-will

‘The boy heard that his father will come’

b *chele-Ta [je [or baba aS–be]] Sune-che (Bayer 1996,255)

boy-CL that his father come-will hear-Pst3

‘The boy heard that his father will come’

c chele-Ta [or baba je [aS–be]] Sone-ni (Bayer 1996,263)

boy-CL his father that come-will hear-neg/Pst3

‘The boy hasn’t heard that his father will come’
(25)a C chele-Ta [or baba aS–be] Sune-che

b [or baba aS–be] C chele-Ta Sune-che

c je [or baba aS–be] C chele-Ta Sune-che

d chele-Ta Sune-che [je [or baba aS–be] C t] (=(24)a)


(26)a C chele-Ta [or baba aS–be] Sune-che

b [or baba aS–be] C chele-Ta Sune-che

c je [or baba aS–be] C chele-Ta Sune-che

d chele-Ta [or baba je [ t aS–be]] Sone-ni (=(24)c)


That je corresponds to the higher complementizer daß (that) of (7), rather than to the lower complementizer C, is suggested by the fact that C may be spelled out as well (together with the ‘final’ complementizer bole seen in (14)c). See (27) (also from Bayer 1996,263f):
(27) [[ chele je poR-be] bole] ami mon-e kor-I ni

boy JE study-Fut3 BOLE I mind-loc do-1 neg-pst

‘I haven’t thought that the boy will study’
Like in complement and adjunct clauses in the sentence, in some OV languages constituents of the RC may also end up to the left of the relative complementizer. This is more obvious (pace Kayne 1994,93) in those cases where the relative and declarative (or interrogative) complementizers have the same form, as is the case, apparently, in Amharic (Demeke 2001,196ff), and Basque (De Rijk 1972,116; Lehmann 1984,59). See (28):30
(28) [lä-saba [-šäT-ku-t] C mäšhaf ]]]] (cf. Demeke 2001,203)

to-Saba comp-sellperf-1s-3ms book

‘a book that I sold to Saba’
As (following Kayne 2000a,2001,2003,2005) we take the post-“Head” positioning of a clause to depend on the presence of an overt (or abstract) complementizer (of the right kind), it could be that a non-rigid OV language with postverbal complement clauses still has only prenominal RCs if the language has no (relative pronoun or) relative complementizer of the same right kind.31 Conversely, it could be that a certain OV language with postnominal RCs introduced by relative pronouns or relative complementizers (of the right kind) has no postverbal clause as it lacks declarative complementizers (of the same right kind). Slave appears to be such a case. It has preverbal subordinate clauses (Rice 1989, chapt.42), but postnominal RCs (with final complementizers) (Rice 1989,chapt.47; Dryer 2003,31).32 In any case, we submit that both such situations are marked, the more general case being that if a language allows postverbal subordinate clauses (i.e., is VO or “non-rigid” OV) then it also allows postnominal RCs. This was seen to be a consequence of a property of higher complementizers.

APPENDIX I

(M&B = Mallinson and Blake 1981)



OV languages of the “rigid” type (no postverbal subordinate clauses; no postnominal RCs):

Ainu (isolate - M&B,276; Tamura 2000), Amharic (Ethio-Semiticsee fn.7 above), Betta Kurumba (South Dravidian – Coelho 2003,78ff,214,223), Burmese (Tibeto-Burman - M&B, 277; Lehmann 1984,183; Soe 1999), Burushaski (isolate - M&B,277), C(h)amling (Tibeto-Burman - Ebert 1997)33, Chibcha (Chibchan - M&B,277), Divehi (Indo-Aryan – Cain and Gair 2000, , Dulong (Tibeto-Burman – LaPolla, 2003), Enets (Samoyed – Künnap 1999,31ff.), Fore (East New Guinea Highlands Languages - M&B,278; Scott 1978), Gadaba (Central Dravidian - Bhaskararao 1998, 347, 350ff), Hakha Lai (Tibeto-Burman – Peterson 2003), Ijo (Niger-Congo – Lehmann, 1984,72; Carstens 2002), Inor (Ethio-Semitic – Chamora and Hetzron 2000,64), Japanese (Altaic - M&B,279), Kannada (South Dravidian – Steever 1987,109;1998,146ff; Sridhar 1990), Koiari (Non-Austronesian Papuan – Dutton 1996); KoDava (South Dravidian – Ebert 1996, chapter 5), KoNDa (South Central Dravidian – Steever 1987,110; Krishnamurti and Benham 1998,266ff), Korean (Altaic - M&B,280), Lahu (Tibeto-Burman – Lehmann 1984,61-63), Lalo (Tibeto-Burman - Björverud 1998), Lushai (Tibeto-Burman - Hillard 1977,339ff, 343); Malayalam (South Dravidian - Mohanan 1982,510; Asher and Kumari1997,§1.1.2)34, Mao Naga (Tibeto-Burman – Giridhar 1994)35, Meithei (Tibeto-Burman – Chelliah 1997), Mongolian (Altaic - M&B,281; Binnick 1979,chapt.III)36, Parengi-Gorum (Munda – Aze 1973,263,300f), Piro (Arawakan - M&B,281), Qiang (Qiangic (Tibeto-Burman) – LaPolla (with Huang), 2003,19f,221), Sherpa (Tibeto-Burman - M&B,282; Givón 1975, 78, 95-96, 99-100)37, Sinhala (or Sinhalese) (Indo-Aryan - M&B,282; Gair 1970,62ff;157ff; Gair 1992,443ff; Gair and Paolillo 1997, chapter 3; Keenan and Comrie 1979,345), Sri Lanka Malay (Creole - M&B,283), Sri Lanka Portuguese (Creole - M&B,283; Smith 2001), Tamil (South Dravidian - M&B,283; Steever 1992,134-136; Annamalai and Steever 1998,117ff), Tauya (Non-Austronesian Papuan – MacDonald 1990,4,289ff), Telugu (South Central Dravidian – Krishnamurti 1998,233f)38, Thulung Rai (Tibeto-Burman - Lahaussois n.d.), Tyvan (Turkic – Anderson and Harrison 1999).39

APPENDIX II
OV languages of the “non rigid” type (postverbal subordinate clauses; postnominal RCs, either as the exclusive, or as the alternative, option):40
Ala’ala (Non-Austronesian Papuan – Ross 2002c)

postverbal complement clauses:



Ia e-‘ou [Koloka ‘ani e-ba] (Ross 2002c,361)

he 3sg-tell Koloka EMPH 3sg-die ‘He told me Koloka had died’

postverbal adverbial clauses:

kau e-da’a luma [‘ola-na melo e-da’a loba] (Ross 2002c,360)

man 3sg-go house because boy 3sg-go garden

‘The man went to the house because the boy went to the garden
postnominal RCs:41

Ate‘ate [a-ika-ia] bosea e-vua-ia (Ross 2002c,352)

Woman 1sg-see-3sg basket 3sg-carry-3sg ‘The woman I saw was carrying a basket’


Bagri (Indo-Aryan – Gusain 2000)

postverbal complement clauses:



˜ socu˜ hu˜ [kε bó jawεgo] (Gusain 2000,66)

I think.Prst am that he go.Fut.3ms ‘I think that he will go’


postverbal adverbial clauses:42
postnominal RCs:

bó admi [jiko kál ayo ho] (Gusain 2000,62)

that man rel. yesterday come-Perf aux-pst.3ms ‘the man who came yesterday’


Bengali (or Bangla) (Indo-Aryan – Bayer 1996,1999,2001)43

postverbal complement clauses:



chela-Ta Sune-che [je or baba aS-be] (Bayer 1996,255)

boy-CF hear-Past.3 that his father come-Fut.3 ‘The boy has heard that his father will come’


postverbal adverbial clauses:44

?tomar ma khuSi hO-be [tumi kolkata-Y ge-le] (Bayer 1996,282)

your mother happy become-FUT2 you Calcutta-LOC go-CondPrtc

‘Your mother will be happy if you go to Calcutta’


postnominal RCs:

ami Sey lok-Ta-ke [je eSe-che] cin-i na (Bayer 1996,256)

I the man-CF-OBJ that come-Past.3 know-1 not ‘I don’t know the man who came’



Brahui (North Dravidian - Elfenbein 1998) In addition to the Dravidian prenominal pattern, Brahui (possibly due to the influence of the neighbouring Indo-Aryan languages - Elfenbein 1998, 409, 411f) has also postverbal finite complement and adverbial clauses and postnominal finite RCs introduced by the same complementizer ki (borrowed from Balochi - Elfenbein 1998,411):

postverbal complement clauses:



ō tēnā ust-atī pārē [ki ī duzziw=ta] (Elfenbein 1998,412)

he-nom his heart-locI say-past-3 that I steal-fut1sg=3sobl

“he said in his own heart that he would steal it”
postverbal adverbial clauses:

ī nā xal-ōī uţ, [ki nī dawn apāsa]? (Elfenbein 1998,404)

I you hit-prt.necess be, that you thus speak-impfc-prs-2s

‘Am I to be struck by you because you are speaking in this way?’
postnominal RCs:

kunē=nē hamē kučak-as [ki drust kē-nē] (Elfenbein 1998,412)

bite-prs3s=2obl same dog-def that knowledge do-prs3s=2obl

“The dog that bites you is the same dog that knows you”
Canela-Crahô (Jê (Amazonian) – Popjes and Popjes 1986)

postverbal complement clauses:



cu-te i-mã amji jarẽn, [cu-mã akĩn na] (Popjes and Popjes 1986,165)

3-Past 1-Tempry self told 3-Temp 2-like subord

‘He told me that he likes you’
postverbal adverbial clauses:

jaco me capi te pĩ hêre jakep [ame to ajpẽn cahhyr prãm te] (Popjes and Popjes 1986,165)

Jaco and Capi Past wood twig cut 3pl Inst Recip beat want because

‘Jaco and Capi cut twigs because they wanted to beat each other with them’
postnominal RCs:

wa i-te rop pupun, [capi te ih-curan ata] (Popjes and Popjes 1986,171)

I I-Past dog see Capi Past 3-kill Dem/Rel

‘I saw the dog Capi killed’

Coahuilteco (Language isolate formerly spoked in Texas – Troike 1981,2004)

postverbal complement clauses:



na-ka·m [ta-x-pa-ta·wex san pa-n] (Troike 1981,664)

1S-hope 1O-2S-Sub-help Fut Rel-1Con ‘I hope that you will help me’


postverbal adverbial clauses:

na-k-ax in, .. [uxwwa·l’ tukwe·-m mak-pa-ču· santupa·yokwe·-n] (Troike 1981,671)

1S-2°--give also sky Dem-2Con 2S:3pO-Sub-carry in order that-1Con

‘I give you also (the indulgences) in order that you carry them to Heaven’

postnominal RCs:



saxpame pinapsa [xami·n xa-p-xo·] tupa·-n (Troike 2004,4)

sins you 2-Sub-know Dem-1C ‘the sins that you know’


Desano (Tucanoan – Miller 1999)

postverbal complement clauses:



yi?i pepi-a [sĩrĩ-a wa-gokũbõ] (Miller 1999,71)

1s think-Non3^Pres die-Perf go-Prob^3fs ‘I think she will die’


postverbal adverbial clauses:

bĩã igo-re karta goha-bãsĩ-a [igo bãsĩ-bo-ro dopa-ta] (Miller 1999,152)

2pl 3fs-Spec letter write-Abil-Nom3^Pres 3fs know-pot-deverb like-Lim

‘You can write her a letter so that she will know’
postnominal RCs:

yi-re su?ri [ãsũ-basa-ra-ye] sãyã-bi (Miller 1999,144)

1s-Spec clothes buy-Ben-Deverb-Clf put^on-Non3^Pst ‘I put on the dress that was bought for me’


Eudeve (Uto-Aztecan – Shaul 1991)

postverbal complement clauses:



nee aguátera-n [dominco-tze amo missa ca vitzá-cauh] (Shaul 1991,102)

1sg know-Pre Sunday-Loc thy mass Neg see-Past ‘I know that you didn’t see mass on Sunday’


postverbal adverbial clauses:

eme-ne suba-m [eme deni-hipsi-cade] (Shaul 1991,90)

thee-1sg like-Pre thee good-heart-Nom ‘I like you because you have a good heart’


postnominal RCs:

hipsem-ta [no vvas-vva mavva-tzem-ta] ovvic (Shaul 1991,106)

people-Obj my field-Ali weed-Nom-Obj call ‘Call the Indians who are to weed my field’


Evenki (Tungusic - Nedjalkov 1997; Bulatova and Grenoble 1999)

postverbal complement clauses:



nungan sa:-re-n [eme-d’enge-vi] (Nedjalkov 1997,25)

he know-nfut-3sg come-part-prefl ‘He knows that he will (be able to) come’


postverbal adverbial clauses:45

nungan ala:t-cheche-n [o:kin girki-n eme-d’e-n] (Nedjalkov 1997,44)

he wait-impv-3sg when friend-3sg.pss come-FUT-3sg

‘He was waiting when his friend would come’
postnominal RCs:46

bi beje-ve [tatkit-tu haval-d’a-cha-ve] archa-0-m (Nedjalkov 1997,36)

I man-accd school-dat work-impv-part-accd meet-nfut-1sg ‘I met the man who worked at school’


Galla (Oromo) (Cushitic - M&B,278,289, Gragg 1972,162-165; Dryer 1992fn.5; Stroomer 1995)47

postverbal complement clauses:



atini hin-beek-tu, [akka bisaani nyaap’a-ii irra ta-u] ? (Stroomer 1995,127)

You Neg-know-2Neg.Pres, that water enemy-Subj on be.present-3Pres.Subord

‘Don’t you know that your enemies are staying by the water?’

postverbal adverbial clauses:

Nuu dandeenee guyyaa sadiillee hinoolluu, [atoo bisani hind’ugini] (Stroomer 1995,126)

… we be.able.1pl.Past day three.also neg.pass.day.1pl.neg.Pres, if water neg.drink.neg.Past

‘…we cannot live even three days, unless we drink water’

postnominal RCs:



nam-tičča [kan kalēsa gale] (sana) an arge (Gragg 1972,162; M&B,289)

man-def Rel yesterday arrived (Dem) I saw ‘I saw the man that arrived yesterday’



Gapapaiwa (Oceanic (Austronesian) - McGuckin 2002)

postverbal complement clauses:



I-vona [da yaghiyaghina ko-na-vovira] (McGuckin 2002,319)

3:NON.PRES-say Comp quickly 2PL-FUT-return ‘They say that you must return quickly’


  1   2   3   4   5


The database is protected by copyright ©essaydocs.org 2016
send message

    Main page