Owing to the large number of unusual typographic characters I prefer for the time being to offer my analysis of the Zaza phoneme system in pdf format. Click here to download the text in pdf format.
Linguistically Correct
Friday 26 July 2024
Thursday 25 July 2024
The Passive Construction in Zaza
Contents |
Abbreviations |
Introduction |
1.
Morphological and Syntactical Preliminaries |
2.
In Search of an Agent |
3.
Embedded Sentences |
4.
Manner Adverbials |
5.
Quantifiers |
6.
Stress |
7.
Locative Phrase and Time Phrase |
8.
Statal Passive |
9.
Nominalisations |
10.
Questions |
11.
Non-Human Agents and Inanimate Agents |
12.
The Adverbs Again |
Conclusion |
Bibliography |
Abbreviations
Abs. |
absolutive |
Obl. |
oblique |
|
Comp. |
comparative |
pl. |
plural |
|
Det. |
determinative |
Phr |
phrase |
|
DO |
direct
object |
Po. |
postposition |
|
fem. |
feminine |
Rel. |
relative |
|
Iz. |
izafet |
sg. |
singular |
|
Loc
Phr |
locative
phrase |
SU |
subject |
|
masc. |
masculine |
V |
verb |
Introduction
Zaza
is an Iranian language closely related to Kurdish and is now spoken in Turkey
where it has been brought by migration. Its grammar has not received a thorough
investigation yet (as of 1978). The present paper, after a short
morphological and syntactical sketch, will describe different aspects of
passive formations in Zaza in view of a general constraint which will have to
be formulated right from the beginning. In the conclusion we will be concerned
with some general theoretical considerations brought about by the Zaza data.
1.
Morphological and Syntactical Preliminaries
Since
most of the examples used in this paper will be in the past, it is necessary to
outline a brief sketch of Zaza syntax in the past, because past tenses present
certain peculiarities. Zaza is ergative in the past; in other words, whilst in
the present the verb agrees with its subject (in the absolutive case), in the
past verb agreement is with direct object when there is one, the subject is in
the oblique case and the direct object in the absolutive:
Present :
SU (Absolutive) DO (Oblique) V (agrees with SU)
Past:
SU (Oblique) DO (Absolutive) V (agrees with DO)
Examples
of verb agreement:
(1) Rεmzɑn |
mεrdɨmi |
kɨʃεnno |
Abs. |
man-Obl. |
kills |
‘Remzan
kills the man’ |
(2) Rεmzɑni |
mεrdɨm |
kɨʃt |
Obl. |
Abs. |
killed |
‘Remzan killed the man’ |
(3) Rεmzɑni |
mεrdɨmi |
kɨʃti |
Obl. |
Abs.pl. |
killed |
‘Remzan killed the men’ |
As a consequence, the verb is invariable (in the past) when there is no direct object (objectless transitive verbs); in the case of intransitive verbs, the subject is in the absolutive case and there is verb agreement. Subjects of passive verbs behave like subjects of intransitive verbs and take the absolutive case.
The
morpheme characteristic of passive is ‑j‑, compare the following examples: εz
vinεnnɑ ‘I see’, εz vinejεnnɑ ‘I am seen’, εz vinejjɑjɑ ‘I was seen’ (mɨ di ‘I
saw’ is irregular).
2.
In Search of an Agent
The
starting point of this study will be the following active sentence:
(4) dʒɨnεk-ɨ |
dɑ |
mεrdɨmi-ro |
woman-Obl. |
woman-Obl. |
man-Obl. |
‘The
woman hit the man’ |
(n.b.:
the verb ‘to hit’ is always accompanied by the postposition ‑ro; as a rule,
postpositions are preceded by the oblique case of the noun).
A
problem arises when this sentence is put into passive: apparently then, there
is no way to express the agent:
(5) mεrdɨm |
dejjɑ
bɨ |
pɨ-ro |
man-Abs. |
hit
was |
|
‘The
man was hit’ |
(n.b.: pɨ replaces the object in the oblique case in the citation form of the verb).
A
further investigation is necessary in order to see if this represents a general
constraint in Zaza or if there is any case where an agent can be expressed. It
should also be noted that the rare examples of passive given by Hadank
1932: 93, 94, 192, 194 are always agentless.
3.
Embedded Sentences
Our
first move will be to check if a passive with a by-phrase is possible in an
embedded sentence; given the following sentence:
sɨlɑsneno |
mεrdɨm-o |
kɨ |
dʒɨnekεr |
dɑ pɨ-ro |
||
Abs. |
knows |
man-Dem. |
Rel. |
woman |
hit |
|
‘Remzan knows
the man whom the woman hit’ |
||||||
passivisation
applies as in simplex sentences, with deletion of the by-phrase:
(7) Rεmzɑn |
sɨlɑsneno |
mεrdɨm-o |
kɨ |
dejjɑ pɨ-ro |
|
Abs. |
knows |
man-Dem. |
Rel. |
was hit |
|
‘Remzan knows the man who was hit’ |
4.
Manner Adverbials
In
Aspects Chomsky observes that the verbs which do not take manner
adverbials freely are those verbs with following NP’s which do not undergo the
passive transformation (Chomsky 1965:103). He adds: “a Verb will […] undergo the
passive transformation only if it is positively specified, in the lexicon, for
the strict subcategorization feature [-NP⁀Manner], in which case it will also
take manner adverbials freely” (Chomsky 1965:104). But one can question if this
statement gives a true image of the phenomena involved in deep structure or is
not merely an observation of phenomena occurring only in surface structure. In any
case, such a cooccurrence is absent from Zaza, since a manner adverbial can
appear whereas a by-phrase is not permitted:
Active
(8) Meriεm |
tʃɨnɑ |
kefɑ |
ʃɨt |
Mary |
cloth |
with joy |
washed |
‘Mary
washed the clothes with joy’ |
Passive
(9) tʃɨnɑ |
kefɑ |
ʃojɑ |
cloth |
with joy |
was washed |
‘The
clothes were washed with joy’ |
The
evidence collected seems to point to Zaza taking manner adverbials freely in
the passive even if there is no by-phrase.
5.
Quantifiers
As
was pointed out by Chomsky in Syntactic Structures, quantifiers present
certain peculiarities as far as a semantic relation between active and passive
is concerned : "we can describe circumstances in which a
'quantificational' sentence such as "everyone in the room knows at least
two languages" may be true, while the corresponding passive "at least
two languages are known by everyone in the room" is false, under the
normal interpretation of these sentences — e.g., if one person in the room
knows only French and German, and another only Spanish and Italian"
(Chomsky 1957:100-101).
Given
the fact that passive sentences with quantifiers seem to have special
properties in English, one is led to ask if there will be, in such a case,
expression of an agent in Zaza; but this is not the case, as can be seen from
the following two examples:
Active
(10)
hyrgy |
mɛrdɨmi |
dɨ |
kɛrgi |
kɨʃti |
every |
man-Obl. |
two |
chickens-Abs.pl. |
killed |
'Every
man killed two chickens' |
Passive
(11)
dɨ |
kɛrgi |
kɨʃjɑj |
two |
chickens-Abs. |
were
killed |
'Two
chickens were killed' |
The
following sentences have also been tried but the result is always that there is
no possibility for a by-phrase to appear:
Active
(12)
qɑtʃkɑne |
bolɨn |
kɨtɑbi |
wɛndi |
children |
many |
books |
read-Past |
'Many
children read books' |
Passive
(13)
kɨtɑbi |
wɑnejɑj |
books-Abs. |
were
read |
'The
books were read' (*by many children) |
Active
(14)
ʒɛw |
mɛrdɨmi |
kɨtɑb |
wɛnd |
one |
man-Obl. |
book |
read |
'One
man read the book' |
Passive
(15)
kɨtɑb |
wɑnejɑ |
book-Abs. |
was
read |
'The
book was read' (*by one man) |
6.
Stress
It
has already been observed that certain passive sentences in English are considered
odd by native speakers when they are uttered with normal intonation. Singh
1976:10 notes that a sentence like 'a muscle was pulled by Max' (corresponding
to the active 'Max pulled a muscle') is doubtful whereas if special structural
or intonation features are added, such a sentence as the following is readily
accepted by native speakers: 'Yesterday, a muscle was pulled by Max,
today a tendon, what will it be tomorrow?'
The
question to be asked here is whether contrastive stress can trigger the
appearance of a by-phrase. The sentence to be translated was: 'The book was
given to him by me (not by you)' and the result was:
(16)
kɨtɑb |
dejjɑ |
e. |
mɨ |
dɑ |
e |
book-Abs |
was
given |
to
him |
I |
gave |
to
him |
'The
book was given to him. I gave him [=dative]' |
Once
more, the conclusion is that no by-phrase is admissible.
7.
Locative Phrase and Time Phrase
The
next move was to consider if passivised verbs can take a locative phrase and a
time phrase and, if so, if the agent would be expressed. As for the agent, the
answer is once more negative. But, clearly, locative phrases and time phrases
can occur in passive sentences:
Loc
Phr (17)
nɑn |
mɨtbɑxdɨ |
verjɑ |
food-Abs. |
in
the kitchen |
was
eaten |
'The
food was eaten in the kitchen' |
(18)
tʃɨnɑ |
rodɨ |
ʃojɑ. |
Meriɛmɨ |
ʃɨt |
cloth-Abs. |
in
the river |
was
washed |
Mary-Obl. |
washed |
'The
clothes were washed in the river. Mary washed (them) = The clothes were
washed by Mary in the river' |
Time
Phr (19)
tʃɨnɑ |
vɨzer |
ʃojɑ. |
Meriɛmɨ |
ʃɨt |
cloth-Abs. |
yesterday |
was
washed |
Mary-Obl. |
washed |
'The
clothes were washed yesterday. Mary washed = The clothes were washed by Mary
yesterday' |
Loc
Phr and Time Phr (20)
tʃɨnɑ |
vɨzer |
rodɨ |
ʃojɑ |
cloth-Abs. |
yesterday |
in
the river |
was
washed |
'The
clothes were washed in the river yesterday' |
8.
Statal Passive
Traditionally,
grammars distinguish between kinetic and statal passives, according to meaning:
(21)
The door was shut (statal)
(22)
The door was shut by John (kinetic)
For
Hasegawa, statal passives are not passive at all and he tentatively assumes
that they are a variety of be + Pred construction; the
last two examples will be analysed thus (Hasegawa 1966:236):
(23)
the door Past be En [Shut] S (Passive)
(24)
the door Past be [En shut] Pred (Statal 'passive')
Since
in English there is a similarity in surface structure between the two passives,
Hasegawa assumes that this relationship must be captured in the underlying
structures but this assumption is questionable for Zaza, as can be shown from
the following examples:
(25)
kɑpɨ |
gɨrote |
bɨ |
door |
shut |
was |
'The
door was shut' (statal) |
(26)
kɑpɨ |
gɨrjɑ |
bɨ |
door |
shut |
was |
'The
door was shut' (kinetic) |
In
Zaza, the surface structure of both constructions is clearly differentiated, as
opposed to English. Moreover, this distinction implies, in the case of the
kinetic passive, that an agent must be 'understood': one is then led to posit,
in deep structure, the presence of a by-phrase or, better, a by+unspecified NP
which is then deleted by a transformation of the following type:
(27)
SD |
X |
en⁀V |
by⁀unspecified
NP |
|
|
1 |
2
3 |
4
5 |
➱ |
|
1 |
2
3 |
0
0 |
|
However,
this poses a problem of recoverability, since the same surface structure
sentence "the door was shut" can have as underlying structures both
"the door was shut by Remzan" and "the door was shut by +
unspecified NP".
9.
Nominalisations
In
Remarks on Nominalisations, Chomsky observes that passivisation can also
occur within complex NP's. For example, in
(28)
John's killing of the chicken
John
is the agent and that can be paraphrased by:
(29)
The chicken was killed by John.
But
even within an NP, expression of the agent is impossible in Zaza; the sentence
(30)
I don't know which chicken John's killing (it) surprised Mary
can
only be translated into Zaza if the agent is omitted:
(31)
nezɑnɑ |
kɨʃjɑjɨ-e |
kɑnʤin |
kɛrgɨ |
Meriɛmɨ-re |
eʤɑjɨb
ɑme |
not-(I)know |
kill-Pass+Iz. |
which |
chicken-Obl. |
Mary-Obl.-for |
surprised |
'I
don't know the killing of which chicken surprised Mary' |
That
is probably the strongest evidence for the complete impossibility of expressing
agents in Zaza. However, certain other remaining possibilities must also be
checked.
10.
Questions
Given
a question of the following type:
(32)
kɑnʤin |
mɛrdɨmi-ro |
dɑ |
ʤɨnekɛr? |
which |
man-Obl.
Postpo. |
hit |
woman-Obl. |
'Which
man did the woman hit?' |
we
would like to know if a question on the agent is possible, that is:
(33)
By which woman was the man hit?
The
only alternative, if we want the question to be on the agent, is to have the
sentence in the active. Otherwise, the only passive interrogative sentence
is :
(34)
dejjɑ |
kɑnʤin |
mɛrdɨmi-ro? |
hit-Pass |
which |
man-Obl. |
'Which
man was hit?' |
Starting
from the translation 'von wem?' given by Hadank 1932 of the Zaza phrase 'kɑmi-rɑ?'
I asked my informant if the following would be a normal sentence of Zaza:
(35) kapɨ
gerjɑbɨ; kɑmi-rɑ ?
'the
door was shut; "von wem"?'
The
answer came at once that it was not possible and that the translation of kɑmi-rɑ
should be 'from whom?' and not 'by whom?'.
11.
Non-Human Agents and Inanimate Agents
So
far, we have only been concerned with trying to find a way for a human agent to
be expressed in Zaza. Now we must explore the possibility of a non-human agent,
say an animal, being expressed. The sentence chosen for that purpose was:
(36)
The caravan was preceded by a dog
It
should be noted that, strictly speaking, it is questionable whether 'by a dog'
should be considered as an agent; in other words, since the agentivity of such
a by-phrase seems rather thin, it should be easier for a by-phrase not to be
deleted. However, even in that case, the agent must be deleted or the sentence
turned into active:
(37)
kɛrvɑni-verrɑ |
qutɨke |
ʃijɛ |
caravan-Obl.-in
front of |
dog |
was
going |
'A
dog preceded the caravan' |
The
next move was to investigate if the same constraints hold with inanimate
agents, starting with the following English sentence:
(38)
The door was shut by the wind
This
time, a literal Zaza translation was possible:
(39)
kɑpɨ |
vɑj-rɑ |
gerjɑ |
bɨ |
door |
'by'
wind-Po |
shut |
was |
'The
door was shut by the wind' |
(n.b.:
-rɑ is described by Hadank 1932:58 as an 'ablative' postposition).
Compare
also :
(40)
bɑxʤɛ |
vɛw-rɑ |
gerjɑ |
bɨ |
garden |
snow-with |
'closed' |
was |
'The
garden was covered with snow' |
There
is no restriction for the corresponding embedded sentences:
(41)
Rɛmzɑni |
vɑ |
(kɨ) |
kɑpɨ |
vɑj-rɑ |
gerjɑ |
bɨ |
Obl. |
said |
that |
door |
wind-Po. |
'closed' |
was |
'Remzan
said that the door was closed by the wind' |
(42)
Rɛmzɑni, |
bɑxʤɛ-u |
kɨ |
vɛw-rɑ |
gerjɑ |
bɨ, |
di |
Obl. |
garden-Dem. |
Rel. |
snow-Po. |
'closed' |
was |
saw |
'Remzan
saw the garden which was covered with snow' |
But
when further questions are asked, one soon realises that it is not sure that an
agent is expressed in the preceding sentences; instead, what looked like
by-phrases was actually described by my informant as complements of cause.
However, this intuition does hot correspond to other uses of the postposition
-rɑ:
Ablative
(43)
Rɛmzɑn |
New
York-rɑ |
ʃɨ |
Abs. |
-from |
went |
'Remzan
left New York' |
Subject
Matter (44)
Rɛmzɑni |
jɑɣɛri-rɑ |
qɨse-kɛrdi |
|
rain-Po. |
talked |
'Remzan
talked about the rain' |
Comparative
(45)
Rɛmzɑn |
Meriɛm-rɑ |
gɨrd-er
- i - o |
|
|
big-Comp.-Det.-is |
'Remzan
is bigger than Mary' |
12.
The Adverbs Again
The
last trial was to apply a remark made by Singh 1976:26 that in a sentence like
(46)
The decision was internationally approved
the
agent is already expressed in the adverb, so that no other noun-phrase need be
expressed. A Zaza equivalent seemed plausible since, as already noted, manner
adverbials can occur with passives. However, an adverb functioning as a
by-phrase cannot appear in surface structure. A sentence like the
following :
(47)
This fact is universally known
was
translated:
(48)
ne |
ɑlɛm |
zɑno |
this |
everybody |
knows |
'Everybody
knows this' |
Conclusion
Clearly
enough, there can be no full passive in Zaza (even what seemed to be an
inanimate agent, pp. 13-14, was felt by my informant as not being equivalent to
a corresponding English by-phrase). This raises the question of agentless
passives in a transformational grammar. Chomsky seems to have taken for
granted, in the early days, that the normal form of passives (at least in
English) was with a specified by-phrase, deleted by a later transformation, as
in Syntactic Structures (p. 81, note 7). In Aspects
he modified somewhat his position and introduced the notion of unspecified
agent to explain sentences of the type "this job is being worked at quite
seriously" (Chomsky 1965:105). However, statistical evidence shows that,
even in English, full passives are much less frequent than agentless passives
(see Bates 1974:16 for references), a fact which has not been taken into
account by Chomsky.
It
is well known that passives allow the hearer's attention to be focused on the
patient. In such a case, one can ask if the same goal can be obtained in Zaza
merely by the word order. The unmarked word order is SOV:
(49) Rεmzɑni Mɑhmud kɨʃt
Obl. Abs. killed
There
is only one other possible word order :
(50) Mɑhmud Rεmzɑni kɨʃt
Patient Agent
Verb
but,
in this case, it is the agent which is emphasized.
Passive
in Zaza shows a strong affinity with intransitive; this is best illustrated by
the fact that, in the past, both passive and intransitive verbs agree with
their subjects in the absolutive case (whereas object1ess transitive verbs are
invariable and transitive verbs with a direct object agree with the direct
object in the absolutive case); in other words, past passive and intransitive
verbs share three features: 1) their subjects are in the absolutive; 2) they
agree with their subjects; 3) they have no direct object (on the relation of
passive and intransitive see Keenan 1975:340-1).
Bibliography
Bates,
J. 1974. An Appraisal of Some Aspects of the Passive Construction in
Transformational Grammar. Cambridge, unpublished dissertation.
Chomsky,
N. 1975 (=1957). Syntactic Structures. The Hague, Mouton.
------------,
1965. Aspects of the Theory of Syntax. Cambridge, Mass., The M.l.T.
Press.
Hadank,
K. 1932. Mundarten der Zâzâ, hauptsächlich aus Siwerek und Kor.
Berlin, Walter de Gruyter.
Hasegawa,
K.1968. 'The Passive Construction in English' in Language 44, 230-243.
Keenan,
E. 1975. 'Some Universals of Passive in Re1ational Grammar'. Papers from the
Eleventh Meeting, Chicago Linguistic Society, 340-52.
Singh,
C.1976. The Limitations of Transformational Generative Treatments of
the Passive as a Syntactic Operation on Underlying Structure. Cambridge,
unpublished dissertation.