title harimochi, kazuro 沖縄短大論叢 = okinawa tandai...
TRANSCRIPT
Title Ergativity and so-called 'ergative of'
Author(s) Harimochi, Kazuro
Citation 沖縄短大論叢 = OKINAWA TANDAI RONSO, 2(1): 1-12
Issue Date 1986-03-31
URL http://hdl.handle.net/20.500.12001/10609
Rights 沖縄短期大学
Ergativity and so-called 'ergative of'
Kai:uro Harimochi
'Ergative of' is a term introduced in Bolinger {1977), refering to the
preposition in such a construction as:
[0) It was foolish of Mary to go there.
This paper examines the validity of qualifying the preposition as ergative,
taking into account ergativity in general and in English.
1. Ergativity
What is characteristic of ergativity in general is that, morphologically
and/or syntactically, S (the subject of an intransitive verb) and P (the ob
ject of a transitive verb) are treated in one way, whereas A (the subject of
a transitive verb) is treated in another. In languages that have ergativity
A is overtly marked as ergative but S and P are marked as absolutive, and
S and P are morphologically more siniple than A.
ergative
absolutive { ergative-absolutive
system
To quote some examples:
[1) a. Vas
boy-Abs.
A
) -- nominative
s
p accusative
nominative-accusative
system
v- ekerula.
Sg.Masc.Abs.- run
'The boy runs.'
- 1 -
b. Jas j- ekerula.
girl-Abs. Sg.Fem.Abs.- run
'The girl runs.'
c. Vas-as jas j- ec:ula.
boy-Erg. gilr-Abs. Sg.Fem.Abs.- praise
'The boys praises the girl.'
Sin [La] (Vas) and Sin [Lb] (Jas) and Pin [Lc] (Jas) are all absolutive
but A in [I.e] (Vasas) is ergative. The verbs in [La] and [Lb] agr~e in
gender with their S respectively, but the verb in [I.e] agrees in gender
with its P.
[2] a. Bayi yara bani-nYu.
Masc.-Abs. man-Abs. come-Tense
'The man came here.'
b. Balan dYugumbil baiJgul yara-IJgu balga-n.
Fem.-Abs. woman-Abs. Masc.-Eng. man- Erg. hit -Tense
'The man hit the woman.'
c. Balan dYugumbil bani-nYu.
Fem.-Abs. woman-Abs. come-Tense
'The woman came here.'
d. Bayi yara baiJgun dYugumbi-ru
Masc.-Abs. man-Abs. Fem.-Erg. woman -Erg.
'The woman hit the man.'
In the Dyirbal language, where sentences in [2] come from, there is no
conjunction used in co-ordination, and two sentences are just juxta
posed with optional deletion of the appropriate noun from the second
clause. So the possible co-ordination is
-2-
balgan-n.
hit - Ten:
[2.a+d] Bayi yara baninYu, ba!Jgun dYugumbiru balgan.
'The man came here and the woman hit (him}.
[2.d+a] Bayi yara ba!Jgun dYugumbiru balgan, baninYu.
'The woman hit the man and (he) came here.
[2.c+b] Balan dYugumbil baninYu, balJgul yaralJgu balgan.
'The woman came here and the man hit (her).
[2.b+c] Balan dYugumbil balJgul yaralJgu balgan, baninYu.
'The man his the woman and (she} came here.'
[2.a+d], Pin [2.d] (Bai yara} is deleted as it is the same asS in [2.a].
[2.d+a], S in [2.a] is deleted for the same reason as for [2.a+d]. In
.c+b], Pin [2.b] (Balan dYugumbil} is deleted as it is the same asS in
.c]. In [2.b+c], Sin [2.c] is deleted for the same reason as for [2.c+b].
in [2.b] and [2.d] are not deleted for they are not the same asS or P.
ms the co-ordination in Dyirbal is based on the ergative-absolutive sys
m.
Generally, English is not understood as an ergative language posseesing
ch ergative characteristics as we have seen, but is taken as an accusative
1guage where Sand A are treated alike but Pis treated differently. In the
llowing sentences;
[3] a. He mo'Ves.
b. He moves it/her.
[4] a. They move.
b. They move it/her.
subjects (S for [3.a] and [4.a], and A for [3.b] and [4.b]) are nomina
•e, and both Ps are accusative. Besides, in respect to the agreement of
-3-
number between the noun and the verb, Sand A agree with the verb in the
same way, but P doesn't. Case agreement between S and A and the agree
ment of number between S/A and the verb are morphological character
istics of accusative languages. And also among the following sentences
(cf. (2] );
(5] a. The man came here.
b. The man hit the woman.
c. The woman came here.
d. The woman hit the man.
the possible couplings without any conjunction are;
(S.a+b] The man came here, hit the woman.
(5.b+a] The man hit the woman, came here.
(S.c+d] The woman came here, hit the man.
(S.d+c] The woman hit the man, came here.
because in English co-ordination, the subject of the latter half can be de
leted only if it denotes the same as the subject of the preceding half. If
not, some conjunction must be employed and the subject of the latter half
usually cannot be deleted. Thus English is morphologically and syntactical
ly based on a nominative-accusative system.
But there seems to be some reason for us to be prudent in defining all
English subjects as nominative. Although Indo-European languages are
understood to be based on a nominative-accusative system, there is an as
pect of this language family where the distinction between nominative and
accusative cases is not made; i.e. the nominative form and the accusative
form of a neuter noun are always the same, never being morphologically -4-
distinguished from each othel English has been in the process of losing the
tripartite declension of gender except that of the third person singular pro
noun. And the declension is less based on grammatical gender than on the
distinction of sex and the distinction between animacy and inanimacy. In
Indo-European languages it is comparatively rare for animate nouns to be
neuter; they tend to be either masculine or feminine. And the English pro
noun it represents neutrality in gender less than it represents inami
macy. So we may assume that the concept of neuter gender is app
roximately identifiable with the concept of inanimacy.
In conformity with this assumption the following sentences are sugges
tive in a way that entitles them to a further look.
[6] It moved.
[7] He moved.
[8] She moved it.
[9] She moved him.
To present the features of animacy and agentivity of the noun and pro
noun in these sentences:
Sentences Subject
AN AG Verb
Object
AN AG
[6] It moved. Vi
( ) moved. + ·······················-·-············-·········· ---··········-··· ---·-------······
[7] a. He moved. + Vi
b. He moved. + + [8] She moved it. + + Vt
[9] She moved him. + + Vt +
- 5·-
[6) and [8) are correspondent to each other in the same way as [7) and
[9) are. As for the pair of [6] and [8), S in [6) and Pin [8) have com
mon features of animacy and agentivity, and we may say that S and P are
treated in the same way because the nominative form and the accusative
form of a neuter, or inanimate (on our assumption), noun are never distin
guished morphologically from each other.
As for [7) , it permits two interpretations according to the agentivity of
S. One interpretation is that he moved of his own free will; +agentive. The
other is that he was forced or obliged to move against his will, or that he
moved as the result of being carried on a wheelchair or a stretcher;-agen
tive.
According to the defmition of cases in Case Grammar by Fillmore
(1968); 4
Agentive: the case for the typically animate perceived instigator of the
action identified by the verb.
Instrumental: the case of the inanimate force or object causally in
volved in the action or state identified by the verb.
Dative: the case of the animate being affected by the state or action
identified by the verb.
Factitive: the case of the object or being resulting from the action or
state identified by the verb, or understood as a part of the verb.
Locative: the case which identifies the location or spatial orientation of
the state or action identified by the verb.
Objective: the semantically most neutral case, the case of anything re
presentable by a noun whose role in the action or state identified by
the verb is identified by the semantic interpretation of the verb it
self; conceivably the concept should be limited to things which are
- 6-
affected by the action or state identified by the verb. The term is
not to be confused with the notion of direct object, nor with the
name of the surface case synonymous with accusative.
S. in [7.a] is Dative, Sin [7.b] is Agentive, and Pin [9], as no case ap
pears more than twice in a clause, is Dative. [6] and [8] are correspondent
to each other in the same way as [7.a] and [9] are. But [7.b] and [9] are
not correspondent because P of the causative transitive verb in [9] is not
agentive as Quirket al. (1972) say:
"thus we understand from They stood Joe against the wall that Joe
reached that position without the help of his own volition." 5
What (7.b] can be correspondent to is an analytic serttence with a causa
tive auxiliary;
[10] She made him move.
which seems to be neutral as regards agentivity of P. As can be seen from
[9] and [10], the difference in the agentivity of P dose not take form
morphologically, nor does the difference in the agentivity of S in [7].
There is no language, in Indo-European languages nor in so-called ergative
languages, in which S is morphologically marked to present the feature of
its agentivity. 6
- 7-
But, being based firstly on the inanimate-neuter assumption, secondly
on the fact that the nominative form and the accusative form of a neuter
noun in any Indo-European language are never distinguished, thirdly on
the inclination of S in [7.a] and P in [9] to be Objective rather than
Dative or in between them, we may safely consider Pin [9] to be in the
S position of (7] without any morphological modification.
[11] *Him moved.
[11] is only a potential sentence in English ·and never really exists. In
other words, English S in nominative case is representative ofNominative-S
and Objective-S which stand for agentive S and non-agentive S respectively.
Erg. - A
,- Nominative-S
J Opbjective-S z Abs. l Objective-S
A }Nom.
>Nominativ..S
REPRESENTATION 7
2. So-called 'ergative-or
Bolinger argues that in the sentences like
[12] The departing of the guests was welcome.
[ 13] *The staying of the guests was welcome.
-8-
P - Ace.
the of-phrase names the agent rather than the subject, and that the same of
appears in the infinitive construction with adjectives like [ 0 ] ;
[0] It was foolish of Mary to go there.
(14] It was confusing of John to mix up the explanations.
[15] *It was confusing of the text to mix up the explanations.
Thus revealing the reasoning behind his naming the 'ergatiye-of:
We conclude that of is used for an action originating in or proceed
ing FROM an agent - a kind of ergative case relationship. 8
The of-phrase bears two points of similarity to A (the subject of a transi·
tive verb) in ergative languages, and they seem to reinforce Bolinger's
background. One point is that the of-phrase is in the oblique, i.e. non
absolutive, case. The other is that it presents the agent of the verb con
cerned. The first point of similarity can be admitted readily, because the
eargative form of a noun in ergative languages is often identical with an
other case, most often the genitive or instrumental: But the second point is
subject to some qualifying commentary.
For the reason that the adjective in [16] modifies both the person and
his/her action identified by the verb, while the adjective in [ 17] modifies
the action only, the syntax of [16] must be more complex than that of
[ 17] in the sense that the former is loaded with the latter.
(16] It is unpleasant of you to demand such close attention.
[ 17] It is unpleasant for you to demand such close attention.
- 9-
The syntax of [171 is:
[18]
----~------NP AUX VP .....___
~ tns I Pred
~I A~j You demand such close Pres be unpleasant
attention
The agent-action relation between the of-phrase and the infmitive in
[161 is mapped from the categories dominated by the lower S in [181,
while such a relation in the sentences in [ 11 and [ 21 is observed on the
surface level, Le. ·it is mapped from the caterogies immediately domi
nated by the initial S.
Not to bypass this difference, I would like to refer to the preposition in
question only as 'agentive-_gf, for agentivity is the distinctive feature of
the deep-structure subject of the embedded S introduced with 'agentive..of'.
The deep-structure subject can be either agentive or non-agentive when
the embedded Sis introduceli: with the!complementizer for-to or Poss-ing; I
[191 It is hard for a surgeon to conduct five operations a day.
[201 It is silent for water to evapolate.
[211 I prefer his departing.
[221 I prefer his staying.
-10-
1 Comrie, B. (1978). p. 338.
2 ibid. p. 348.
3 Lyons, J. (1968). p. 293.
Notes
4 Fillmore, C. (1968). pp. 24-5.
5 Quirk, R.et al. (1972). p. 352.
6 Lyons, J. op.cit. p: 357.
7 A. Martinet (1979) refers to this representation as "shunting."
8 Bolinger, D. (1977). p. 141.
9 Trask, R. L. (1979). p; 385.
-11-
References
Back, E. & Harms, R. (eds.) (1968). Universale in Linguistic.
Theory: Holt, Rinehart and Winston.
Bolinger, D. (1977), Meaning and Form:Longman.
Comrie, B. (1978). 'Ergativity'. In Lehmann (ed.) (1978: 329-394).
Curme, G. 0. (1931). Syntax: D. C. Heath & Company.
Fillmore, C. (1968). 'The Case for Case'. In Back & Harms (1968: 1-88).
Hiromi, Nakano. (1970). 'Eigo ni okeru Ergativity Kosatsu'.
In Eigogak:u 5. 34-50.
Lehman, W. P. (ed.) (1978). Syntactic Typology : University ofTexasPress.
Lynos, J. (1968). Introduction to Theoretical Linguistics :
Cambridge University Press.
Martinet, A. (1979). 'Shunting on to Ergative or Accusative'.
In Plank (ed.) (1979: 39-43).
Plank, F. (ed.) (1979). Ergativity: Academic Press.
Plank, F. (1979). 'Ergativity, Syntactic Typology and Universal Grammar'.
In Plank (ed.) (1979: 3-36).
Quirk, R.et al .. (1972). A Grammar of ContemPorary English: Longman.
Silva, G & Thompson, S. (1977). "On the syntax and semantics of adjec
tives with 'it' subjects and infinitival complements in English".
In Studies in Englishl. 109-126.
Trask, R. L. (1979). 'On the Origins ofErgativity'. In Plank
(ed.) (1979: 385-404).
-12-