ArticlePDF Available

Case marking in infinitive (ad- form) clauses in Old Georgian

Authors:

Abstract and Figures

A specific feature of both Modem and Old Georgian is that case marking of subjects and objects is sensitive to the choice of tense/aspect. This paper focuses on a construction that was found in complementation in Old Georgian (5th-l 1th centuries) where alongside with finite forms, an infinitive began to develop. Generally, this was a verb-noun in the adverbial case (-(a)d): tesva 'sowing' -> tesva-d 'sow'. As the infinitive lacks expression of tense/aspect it is not able to assign case to its arguments in the same way as a finite verb does in Georgian. In this paper we will show that case marking of the direct object (and sometimes of the subject) of the infinitive is determined by the tense/aspect of the matrix verb. Non-finite forms in Modern Georgian include participles and masdars (verb-nouns). Participles are declined as nouns, are formed from the finite forms of the verb, and usually they have the same functions as adjectives. Masdars are also case marked like nouns, but are formed from the fmite forms of the verb and usually have the same functions as nouns. In Old Georgian, a third non-finite form is found. Formally, it is a masdar in the adverbial case. This form has been called infinitive even though it has been observed by several authors (Martirosovi 1955; Dzidziguri 1989; Chkhubianishvili 1972) that it differs from what is called infinitive in, for instance, Indo-European languages. In order to avoid confusion, we will call this form the ad-foim, where ad refers to the adverbial case marker -(a)d. Before turning to masdars and ad-foims in complementation, it is necessary to have a look at case marking in simple sentences in order to understand how case is assigned to masdars as heads of complement clauses and to the objects of ad-fotms.
Content may be subject to copyright.
Lund University, Dept. of Linguistics 1
Working Papers 46 (1997), 167–183
Case marking in infinitive (ad-
form) clauses in Old Georgian1
Manana Kobaidze and Karina Vamling
A specific feature of both Modern and Old Georgian is that case marking of subjects and
objects is sensitive to the choice of tense/aspect. This paper focuses on a construction that
was found in complementation in Old Georgian (5th-11th centuries) where alongside with
finite forms, an infinitive began to develop. Generally, this was a verb-noun in the adverbial
case (-(a)d): tesva ‘sowing’ –> tesva-d ‘sow’.
As the infinitive lacks expression of tense/aspect it is not able to assign case to its
arguments in the same way as a finite verb does in Georgian. In this paper we will show that
case marking of the direct object (and sometimes of the subject) of the infinitive is
determined by the tense/aspect of the matrix verb.
Non-finite forms in Modern Georgian include participles and masdars (verb-
nouns). Participles are declined as nouns, are formed from the finite forms of
the verb, and usually they have the same functions as adjectives. Masdars are
also case marked like nouns, but are formed from the finite forms of the verb
and usually have the same functions as nouns.
In Old Georgian, a third non-finite form is found. Formally, it is a masdar in
the adverbial case. This form has been called infinitive even though it has been
observed by several authors (Martirosovi 1955; Dzidziguri 1989;
Chkhubianishvili 1972) that it differs from what is called infinitive in, for
instance, Indo-European languages. In order to avoid confusion, we will call
this form the ad-form, where ad refers to the adverbial case marker -(a)d.
Before turning to masdars and ad-forms in complementation, it is necessary
to have a look at case marking in simple sentences in order to understand how
case is assigned to masdars as heads of complement clauses and to the objects
of ad-forms.
1We would like to thank the Swedish Institute for support during the research period while
working on this article. Also, we would like to express our thanks to Darejan
Chkhubianishvili of the Institute of Linguistics at the Georgian Academy of Sciences for
consultations on various aspects of the infinitive in Old Georgian.
2MANANA KOBAIDZE AND KARINA VAMLING
1. Case marking in the simple sentence
The nominative, ergative and dative cases are those which occur in the
marking of subjects and objects of finite verbs. (The dative has some other
functions too, for instance to express locative and benefactive relations.) The
genitive case shows up in the marking of the dependents of the masdars,
although the dative also is allowed in this position. Apart from the syntactic
cases, the set-up of cases includes the semantic cases: the instrumental, the
adverbial and the vocative.
The Old Georgian case markers listed in table 1 (Schanidse 1982:36). Note
that the nominative marker is realized as /-i/ following a stem ending in a
consonant, and as /-j/ when the stem ends in a vowel.
Table 1. Case markers. Singular Plural
‘Bare’ stem2---eb
Nominative -i / -j (n)-i -eb-i
Ergative -man t(a) eb-man
Dative -s(a) " eb-s(a)
Genitive -is(a) " eb-is(a)
Additive -isa " eb-isa
Instrumental -it(a) " eb-it
Adverbial -ad / -d " eb-ad
Vocative -o (n)-o eb-o
A common feature of both Old and Modern Georgian is that the case
marking of the subject and objects is sensitive to the choice of the tense form
in the sentence. As shown in (1a-c), which appear in different tenses, the
subject appears in the nominative, ergative and dative cases.3
(1) a. mgel-i ßç’ams cxovar-sa
wolf.NOM S3SG.O3SG.eat.PRSlamb.DAT
‘The wolf eats the lamb’
b. mgel-man ßeç’ama cxovar-i
wolf.ERGS3SG.O3SG.eat.AOR lamb.NOM
‘The wolf ate the lamb’
c. mgel-sa ßeuç’amies cxovar-i
wolf.DATINV.S3SG.O3SG.eat.PF lamb.NOM
‘(Apparently) the wolf has eaten the lamb’ (Schanidse 1982:172-3)
2Glossed as ABS.
3The person (and number) of the subject, direct object, indirect object (in this order) is
indicated initially in the verb forms, as in: vubr
ane ‘I ordered him (it)’:
S1SG.O3SG.O3SG.order.AOR. In certain verb forms so-called inversion takes place: the logical
subject is marked as an indirect object, both in the verb form and case marking. This is
indicated in the glosses by INV. The ad-form is marked by AD in the glosses.
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 3
Finite forms are grouped into three series, where each series is related to
one case marking pattern. The verb forms included in the three series in Old
Georgian are listed in table 2 (Schanidse 1982:79-80).
Table 2. Series of finite verb forms in Old Georgian.
IPresent c’ers (S3SG.O3.write.PRS etc.)
Present Iterative c’ern
Imperfect c’erda
Imperfect Iterative c’erdis
Imperative I c’erdin
Conjunctive I (/Future I) c’erdes
II Aorist c’era
Iterative II c’eris
Imperative II c’eren
Conjunctive II (/Future II) c’eros
Mixed Future c’erodis
III Perfect uc’eries
Pluperfect ec’era
Iterative ec’eris
Conjunctive III ec’eros
Not all verbs show such a difference in case marking related to the choice
of the finite verb form. It is however, relevant for transitives and active
intransitives.
Another group of verbs is characterized by having experiencer subjects.
The case marking differs from the first class in assigning dative case to the
experiencer subject, and the nominative case to the object if any, independent
of the choice of tense.
The remaining verbs, including one-, two- and three-argument verbs,
constitute a more mixed group. The subject is usually low in activity and the
increased valency is often due to the incorporation of relations such as the
benefactive and locative case into the verb. These verbs have a stable case-
marking pattern, not depending on the choice of tense. Nominative case is
assigned to the subject and dative case to the object(s).
The case marking patterns of the three verb classes are summarized below.
As shown in table 3, there is no accusative case for direct objects. Case
marking of direct objects depends on the series of the finite verb.
2. Matrix predicates with ad-forms as complement predicates
A rather wide range of matrix predicates in Old Georgian take ad-forms as
their complement predicate. Characterized from a semantic point of view (cf.
4MANANA KOBAIDZE AND KARINA VAMLING
Noonan 1985), these matrix predicates include manipulatives, desideratives,
aspectuals, modals. A sample is given below:
brΩana ordered
aiΩula forced
arc’muna persuaded
isc’rapa strove for
egulebis intends
hnebavs wants
µer-içina found necessary, wanted
ßesΩina continued
qel-q’o started
umµobes ars is better
µer-ars is necessary
In the following sections the masdar and the ad-form will be examined with
respect to some verbal and nominal features such as case marking and
adverbial modification.
3. Comparing masdars and ad-forms
The masdar is case marked as an ordinary noun. Therefore, it is to be
expected that a masdar in the object position of a matrix verb in the series I
and II will differ in case marking. This is illustrated by the following examples
where the matrix verb in (2a) is in the present, motivating the dative case for
its direct object. The matrix verb in (b) is in the future II, assigning the
nominative case to its direct object (Chkhubianishvili 1972:139). Example (c) is
in the perfect, i.e. in the third series. Here, the nominative case is assigned to
the object.
(2) a. çuen visc’rapit monagebta ßek’reba-sa
we S1PL.O3SG.strive.PRS property.PL.GEN collection.DAT
‘We strive for the collection of property’
Table 3. Case marking patterns for the three verb classes.
Verb class Subject Object(s)
Series
(1) I. Nominative Dative (DO), Dative (IO)
II. Ergative Nominative (DO), Dative (IO)
III. Dative Nominative (DO), Genitive+P[d(a)]
(2) Dative Nominative
(3) Nominative Dative (IO)
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 5
b. uk’eutu çuen ara visc’rapot akave
if we not S1PL.O3SG.strive.FUT.II here
aqoca-j mati
destruction.NOM their
‘If we will not strive for their destruction …’
c. ßeni adgili dagit’evebies.
your.SG.NOM place.NOM INV.S2SG.O3SG.leave.PF
‘You have left your place’ (shush: kart. krest.I.1.:135)
The syntactic relations in the masdar phrase are usually similar to the
relations within the NP. The masdar marks its ‘object’ by the genitive case, as
shown below.
(3) çuen visc’rapit monagebta ßek’reba-sa
we S1PL.O3SG.strive.PRS property.PL.GEN collection.DAT
‘We strive for the collection of property’
However, it was not a rare exception in Old Georgian that an object of a
masdar was assigned the dative case (see below, section 9).
When examining the ad-forms in different positions, it becomes clear that
they do not change for case. The ad-forms end in -(a)d (the frozen adverbial
case), in the positions corresponding to the ones in (2a-c). The matrix verb in
(4a) is in series I, (b) in series II, and (c) in series III, corresponding to direct
objects in the dative (I) and nominative (II, III) cases.
(4) a. rajsa maiuleb çuen gandgomad
why S2SG.O1PL.force.PRS I.PL.(funct. DAT) go away.AD
mrtisagan.
God.GEN.from
‘Why do you force us away from God’ (sin. mr. 118.2)
b. xolo mevic’q’e sit’q’uad
and I S1SG.O3SG.begin.AOR speak.AD
‘And I began to talk.’ (shush.: kart.krest.I.1:139)
c. tavadsa iesos ec’q’o
himself.DAT Jesus.DAT INV.S3SG.O3SG.begin.PLUP
q’opad meocdaatesa c’elsa
do.AD thirtieth.DAT year.DAT
‘Jesus himself had started to fulfil his mission at 30 years of age’
(luka. 3.23)4
Both the masdar and ad-forms may take objects. However, the case
assigned to such objects may differ, as is shown in the following sections.
4Examples from the Bible are literal translations based on the Georgian text.
6MANANA KOBAIDZE AND KARINA VAMLING
4. Case assigned to direct objects of ad-forms
4.1 Objects of ad-forms in direct object position
Before considering the case assigned to objects of ad-forms, one has to
differentiate two functions of the ad-forms. Such forms can occur as either
complement predicates, or as predicates of purpose clauses. We will first turn
our attention to ad-forms as complement predicates.
As has been pointed out above, the case assigned to subjects and objects in
finite clauses depends on the series of the finite verb. As both masdars and ad-
forms lack tense, they also lack a way of differentiating series. Consequently,
they cannot govern their objects in the same way, as do finite verbs. Despite
this fact, objects of ad-forms (as complement predicates) appear in the dative
or nominative case in the same way as the objects of finite verbs5.
Alternations between case marking patterns due to the choice of the matrix
verb forms from different series appear here, although the ad-form does not
itself indicate tense. The argument of the ad-form is in the dative in (5a) (series
I), and in the nominative in (5b) (series II).
(5) a. titoeuli matgani isc’rapda
everyone.SG.NOM of.them.NOM S3SG.O3SG.strive.IMP
tesvad k’actmoq’uareba-sa
sow.AD love.of.mankind.DAT
‘Everyone of them strove to sow the love of mankind’
(Chkhubianishvili 1972:149)
b. … isc’rapa … adginebad ek’lesiasa ßina
S3SG.O3SG.hasten.AOR revive.AD church.DAT in
sactur-i borot’-i
temptation.NOMevil.NOM
‘… hastened to revive the evil temptation in the church’
(Chkhubianishvili 1972:149)
It appears as if the tense of the matrix verb has the effect of determining
the case marking not only within the finite VP, but also in the ad-form phrase
as suggested by Chkhubianishvili 1972. As expected from this hypothesis, the
direct object of a transitive verb in series I (imperfect) takes the dative, and the
direct object in series II takes the nominative in the examples above, as does
the direct logical object of an ad-form in these positions.
The same rule also applies in series III (6), a fact that has not been discussed
in the literature before6.
5In addition, objects of ad-forms in the genitive or genitive-adverbial case are also found
(Martirosovi 1955:54, Chkhubianishvili 1972:77).
6We are grateful to Guram Kartozia for giving us access to unpublished material from the
Rustaveli fond in Tbilisi.
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 7
(6) a. anu kmnad raj gwisc’avies
or do.AD what.NOM INV.S1PL.O3SG.learn.PF
gank’urnebisatwis k’actajsa.
cure.GEN.for man.PL.GEN.SG.GEN
‘or what we have learned to cure men’ (A 1115. 15v. 24-25b)
b. da vidre d∞eindlad d∞emde vervis
and until of today.ADV.C day.ADV.C.till nobody.DAT
uk’adrebies aebad igi maßinßißisatvis
INV.S3SG.O3SG.dare.PF take.AD it.NOM then fear.GEN.for
sp’arstasa (resp. sp’arstajsa)
Persian.PL.GEN.SG.GEN.
‘and until these days nobody has dared to take it because of the fear
of the times under the Persians’ (luars. mart.I:416.3)
c. p’at’ivi xat’isaj p’irmßojsa
veneration.NOM image.SG.GEN.SG.NOM the first born baby.GEN
mis saxisa mimart aslvad gwisc’avies
that.GEN face.GEN towards ascend.AD INV.S1PL.O3SG.learn.PF
‘We have learned to ascend the veneration of image towards the face
of that child’ (S-384, 422.5a)
Consequently, the ad-form direct object is a semantic argument of the ad-
form whereas case marking is determined by the matrix predicate.
4.2. Direct objects of ad-forms in other positions
The observation that the series of the matrix verb determines the case marking
of the logical direct object of the ad-form holds in other positions as well. A
verb like hnebavs ‘he wants it’, marks its logical subject (experiencer) by the
dative case and the direct object (source) by the nominative (cf. table 3). The
logical direct object of an ad-form in the object position of mnebavs and
gegulebis below is, as expected, marked by the nominative.
(7) a. mnebavs xivlad adgomajca misi
INV.S1SG.O3SG.want.PRS see.AD ascension.NOM his
‘I want to see his ascension’ (Chkhubianishvili 1972:87)
8MANANA KOBAIDZE AND KARINA VAMLING
b. romeli gegulebis sakmed,
what.NOM INV.S2SG.O3SG.intend.PRS do.AD
q’av adre
S2SG.O3SG.do.IMPERATIVE.II quickly
‘… what you intend to do, do quickly’ (i.DE.13.27)
The ad-form clause may also appear as a syntactic subject of a matrix verb.
When looking at monovalent verbs like
µ
er-ars ‘have to, need to’, the logical
direct object of the ad-form is marked by the nominative case in the same way
as a syntactic subject of the matrix predicate whose syntactic subject should be
in the nominative case (cf. table 3)7.
(8) µer-arsa micemad xark’i k’eisarsa anu ara?
S3SG.necessary.PRS give.AD tax.NOMemperor.DAT or not
‘Is it necessary to give tax to the emperor or not?’ (mark.12.14)
The masdar micemaj corresponding to the form micemad also has the
ability to govern its object in dative case (see section 10). Concerning the
logical indirect object, compare this data with section 6.
4.3. 1st and 2nd person pronouns as objects of ad-forms
So far we have only considered case marking of third person NPs. First and
second person objects are not case marked in finite clauses. Compare the
examples below, where the object is represented by a personal pronoun in
(9a), and a full NP in (9b).
(9) a. ßemip’q’robs is me
S3SG.O1SG.catch.PRShe.NOM I
ßemip’q’ro man me
S3SG.O1SG.catch.AOR he.ERG I
ßevup’q’rie mas me
INV.S3SG.O1SG.catch.PF he.DAT I
‘He catches/caught/has caught me’
b. ßeip’q’robs is k’acsa
S3SG.O3SG.catch.PRShe.NOM man.DAT
ßeip’q’ro man k’aci
S3SG.O3SG.catch.AOR he.ERG man.NOM
ßeup’q’ries mas k’aci
INV.S3SG.O3SG.catch.PF he.DAT man.NOM.
‘He catches/caught/has caught the man’
7As masdars could be used with either active or passive meaning, it is also possible to
suggest the passive form as an original form for the ad-form (miecema xark’i). In such a
case, xark’i is a logical subject of the ad-form and a syntactic subject of the matrix verb.
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 9
The behaviour of the first and the second person pronouns as arguments of
the ad-form is very significant.
Considering examples like (10b), Chkhubianishvili 1972:78 notes that there
are no cases where the object of the infinitive (either direct or indirect) is
expressed by personal pronouns of the first or the second persons without
postposition. As the direct object of a finite verb this pronoun has the form me
(10a), whereas it appears in the genitive case with the postposition da (
ç
em-da)
with the ad-form (b).
(10) a. ßemip’q’robs me
S3SG.O1SG.catch.PRSI (by function DAT)
‘He is catching me’
b. gamoxuedit ßep’q’robad çem-da
S2PL.go out.AOR catch.AD I.GEN-to
‘You went out to catch me’ (m.DE.26.55)
However, a close examination of the material reveals that bare first and
second pronouns without any postposition do appear with ad-forms. Two
conditions have to be fulfilled for this to occur: the first/second person
pronoun has to appear in a complement clause (not a purpose clause); and it
has to be represented by an agreement prefix in the matrix verb.
(11) mun ver ßemilon
ç
uen
there not S3PL.O1PL.can.CONJ.II we (funct. DAT)
ßep’q’robad mtavarta mat bnelisata
catch.AD sovereign.PL.ERG those.PL.ERG darkness.SG.GEN.PL.ERG
‘The sovereigns of darkness will not be able to catch us there’
(mamata sts. 230.10)
Compare the forms below. The ad-form clause in (12a),
is a complement of
the matrix verb, and at the same time, its object
ç
uen is reflected in the matrix
verb
ße-m-iΩlon
(
m
- is a formant of the first object person), so
ç
uen appears
without any postposition. Therefore, it occurs in the same form as an object of
the ad-form (12a) that it does after a finite verb (12b).
(12) a. ße-m-ilos man ßep’q’robad
ç
uen
‘He can catch us’
b. ße-m-ip’q’robs
ç
uen
‘He catches us’
c. gamoxuedit tkuen ßep’q’robad
ç
uen-da
‘You went out to catch us’
10 MANANA KOBAIDZE AND KARINA VAMLING
In (12c), the ad-form is not a complement of the matrix verb. And, since
the ad-form cannot govern its object in the same way as the finite verb does,
the object of the ad-form is marked by a postposition.
(13) is another example where the first person of the logical direct object of
the ad-form is marked as an object of the matrix verb.
(13) arca sxuamandabadebulman ßemilos
not.too other.ERG born.ERG S3SG.O1PL.can.CONJ.II
ç
uen ganq’enebad siq’uarulsa mrtisasa
we (by function DAT) part out.ADlove.DAT God.SG.GEN.SG.DAT
‘… and no one of creatures will be able to part out us from love of
God’ (romaelta mimart, 8. 39)
This behavior of first/second pronouns is another confirmation of the
integration of the ad-form clause into the matrix clause.
5. Case assigned to logical subjects of the ad-form
It is common for the ad-form clause to appear in the indirect object position8
of the matrix verb. In this position, the logical subject of the ad-form is
marked as the syntactic indirect object of the matrix verb as in (14).
(14) ubrΩana mas dadebad µaç’wi kedsa missa
S3SG.O3SG.O3.order.AOR he.DAT put.AD chain.NOM neck.DAT her.DAT
‘He ordered him to place a chain on her neck’
(shush. kart. krest.I.1: 138)
The ad-form clause may also appear as the subject of a monovalent matrix
predicate. In such a case (15a), the logical subject of the ad-form is marked by
the nominative case. In (15b), the matrix verb is transitive, and the ad-form
monovalent. The subject case assigned by a transitive verb in the second series
is the ergative case, and it is this case which appears here.
(15) a. huadvilejs ars mankanisa sabeli qurelsa
easier.ABS S3SG.be.PRSmachine.GEN rope.NOM hole.DAT
nemsisasa ganslvad, vidre mdidari ßeslvad
needle.SG.GEN.SG.DAT go through.AD than rich.NOM go in.AD
sasupevelsa mrtisasa
Paradise.DATGod.SG.GEN.SG.DAT
‘It’s easier to lead a rope through the eye of a needle, than it is for a
rich man to enter into the kingdom of heaven’ (Haem. mark. 10.25)
8Indirect object is to be understood as objects that appear in the dative case in the first and
second series, cf. table 3.
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 11
b. ic’q’o kroladkar-man
S3SG.O3.begin.AOR blow.AD wind-ERG
‘The wind began to blow’ (Chkhubianishvili 1972:73)
6. Case assigned to indirect objects of the ad-form
In examples with indirect objects of the ad-form, the case assigned is always
the dative. (16) shows an example in series II, and (17) shows an example in
series III.
(16) araravin ik’adra miaxlebad µuarsa
nobody S3SG.O3SG.dare.AOR come closer.AD cross.DAT
‘Nobody dared to come closer to the cross’
(Chkhubianishvili 1972: 75)
(17) numca vis tkuengansa
not.Particle (’let’) who.DAT (here: anybody) you.from.DAT
uk’adrebies amieritgan uc’esosa
INV.S3SG.O3SG.dare.PF this time.INSTR.fromimproper.DAT
sakmesa qelq’opad monast’ersa amas ßina çemsa
action.DAT begin.AD friary.DAT this.DAT in my.DAT
samk’widrebelsa
proprietary.DAT
‘Hence none of you should have dared to begin any improper action
in the friary belonging to me.’ (vahanis kvab. gang. 62.27)
We have not found any examples where the matrix verb is in the III series,
and the ad-form occurred with both direct and indirect logical objects.
7. The ad-form in purpose clauses
Apart from complement clauses, the ad-form occurs in purpose clauses as
well. When purpose and complement clauses are compared, a difference is
noted in the factors governing case marking. The verb series of the matrix
verb is of importance for case marking only in complement clauses, no such
regularity is observable in purpose clauses (Chkhubianishvili 1972:92-93). The
most common situation in purpose clauses is that the direct object of the ad-
form is marked by the dative case.
(18) movida … smenad sibrne-sa solomon-is-sa
S3SG.come.AOR … listen.AD wisdom-DAT Solomon-GEN-DAT
‘He came to listen to Solomon’s wisdom’
Also, there were other non-finite forms in Old Georgian that were able to
govern their object in dative case (cf. sections 10 and 11).
8. Nominative case instead of the expected dative case
8.1. Complement clauses
12 MANANA KOBAIDZE AND KARINA VAMLING
In some examples, where the object of the ad-form is expected to be assigned
the dative case, it appears in the nominative case.
(19) q’ovelnive ßen-gan elian
all.PL.NOM you.SG.(funct. GEN)-from S3PL.O3SG.wait for.PRS
mocemad sazrdeli mati ÷amsa
give.AD food.NOM their.SG.NOM time.DAT
‘Everybody is waiting to get their food from you in time’
(ps.103.27)
In some cases, the nominative case of an object that was expected to be in
dative case may be explained by adjacency factors. For example:
(20) xasc’avebdit mat damarxvad
S2PL.O3PL.O3SG.teach.IMPERATIVE.I they.PL (funct. DAT) keep.AD
q’oveli romeli gamcen tkuen
all.NOM which.NOM S1SG.O2PL.O3SG.report.AOR you.PL (funct. DAT)
‘Teach them to keep everything what I have reported to you’
(xanm. lekc.28.20)
The nominative case of the ad-form object qoveli seems to be conditioned
by the nominative case of the adjacent member of the following clause romeli,
that is assigned nominative case by finite verb gamcen.
8.2. Purpose clauses
The nominative case may also appear on objects of ad-forms in purpose
clauses. However, in this position, the nominative case is found more rarely
than the dative case according to Chkhubianishvili 1972:93.
(21) da c’arvida moq’vanebad tavisa
and S3SG.go.away.AOR bring.AD head.here:REFL.PRON.SG.GEN
twisisa coli
his.ADDIT wife.NOM
‘And he went away to bring himself a wife’
Such rare examples as (19-21) show the ability of ad-forms to take objects
in the nominative case (that differs markedly from other non-finite forms in
Old Georgian). Therefore, the ad-form has a certain degree of verbal features,
but influence from foreign languages in translated literature cannot be
excluded. Furthermore, as has been mentioned above, the dative case is found
not only with ad-forms, but also with other non-finite forms in Old Georgian.
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 13
9. Dative case assigned by masdar
Like indirect objects of the ad-form, an object of a masdar may also be
assigned the dative case.
(22) giirs tkuen uprojsad damorçilebaj
INV.S2PL.O3SG.be worth.PRS you (funct. DAT) more obey.NOM
br
anebasa.
order.DAT
‘The order is worth obeying for you’ (kim. I. 110.8)
Example (8) – repeated here for convenience as (23a) – includes both a
direct and an indirect object of an ad-form. The same dative case occurs with
the masdar of the verb micemaj (23b).
(23) a. µerarsa micemad xark’i k’eisarsa anu ara?
S3SG.necessary.PRS give.AD tax.NOM emperor.DAT or not
‘Is it necessary to give tax to the emperor or not?’ (mark.12.14)
b. µerarsa xark’isa micemajk’eisarsa?
S3SG.necessary.PRS tax.GEN give.NOM emperor.DAT
‘Is it necessary to give tax to the emperor?’ (m.DE. 22.17).
In these examples, micemaj occupies the syntactic subject position that was
held by xark’i in (23a), and xark’isa has become the argument of the masdar
micemaj. As a result, it has changed its case to the genitive. Compare the
behaviour of the indirect object k’eisarsa, which retains its dative case in both
examples (23a) and (23b).
A masdar may assign dative case to its logical direct object as well. In (24a),
the logical object of the matrix verb is a masdar marked by nominative case.
The noun k’ari ‘door’ appears in the dative case as the object of the masdar.
Compare this with (24b), where the object position of the matrix verb is
occupied by an ad-form, and k’ari ‘door.NOM’ is marked by the nominative
case.
(24) a. ubrano dak’ralvaj k’arsa ek’lesiisasa
S1SG.O3SG.O3SG.order.CONJII keep.NOM door.DAT of church.DAT
‘I shall order him to defend the door of the church.’
b. ubrano dak’ralvad k’ariek’lesiisaj
S1SG.O3SG.O3SG.order.CONJ II keep.ADdoor.NOMof church.NOM
‘I shall order him to defend the door of the church.’
(sin. mr.70.16)
14 MANANA KOBAIDZE AND KARINA VAMLING
10. Dative case assigned by other non-finite forms
As noted above, the ad-form along with masdars and participles may assign
the dative case to their objects. Dative case assigned by non-finite forms has
been studied by Shanidze 1980:143-144, 1976:58; Imnaishvili 1957:441-52,
690, 720; Kiziria 1963:187, and others. In Old Georgian there was another
non-finite form that governed its object in dative case too: the comparative
degree of adjectives (Shanidze 1976:158).
The dative case marking is considered to be connected with the verbal
origin of these forms. The comparative degree of adjectives is also a finite
form from its origin in Georgian (Shanidze 1980:143-144). Even a noun with
the semantics of a participle could assign the dative case to its object
(Imnaishvili 1957:692).
(25) ara var vaç’ar sit’q’uasa
not S1SG.be.PRS salesman.ABSword.DAT
‘I am not a salesman of words.’
Other examples are given below: comparative degree of the adjective with
dative case in (26a) and participle with dative case in (26b).
(26) a. q’ovelive siborot’e umcires arn siborot’esa
all.ABS evil.ABS less.ABS S3SG.be.PRS.ITER evil.DAT
mas dedak’acisasa
that.DAT woman.SG.GEN.SG.DAT
‘Any evil is less than the evil of woman’
b. daq’udebaj ars momp’ovnebel gwirgwinsa
loneliness.NOM S3SG.be.PRSobtainer.ABS crown.DAT
brc’q’invalesa
briliant.DAT
‘Loneliness is obtainer of a brilliant crown.’
11. Modifiers of masdars and ad-forms
Looking for further support for the verbal character of ad-forms and masdars,
one finds, in the case of adverbial modifiers, that they behave similarly.
Actually, according to Chkhubianishvili 1972:43-44, they pattern in the same
way as do finite verbs. In the examples below, both the ad-form in (27a), and
the masdar in (b), are modified by the adverbial picxlad ‘fiercely’.
(27) a. muntkuesveubrana cemadmata
then S3SG.O3SG.O3SG.order.AOR beat.AD them.ADDIT
k’uertxita picxlad
stick.INSTR fiercely
‘… then he ordered him to beat them fiercely with a stick’
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 15
b. ubrana cema-j mata
S3SG.O3SG.O3SG.order.AOR. beat-NOM them.PL.ADIT
k’uertxita picxlad.
stick.INSTR fiercely
‘… he ordered him to beat them fiercely with a stick’
The masdar can be modified by a declinable noun as well (Jorbenadze
1995:14).
12. The loss of ad-forms in Old Georgian
The resemblance in case marking of finite and non-finite forms is observed in
the earliest texts. Later, the dative case marking of objects of masdars and
participles has been gradually lost, and the genitive has become the only
possible form in this position (Shanidze 1976; Imnaishvili 1957; Chikobava
1953). Still the governing of objects in dative case by participles (28) is also
rarely retained in Modern Georgian (Kiziria 1963:187; Imnaishvili 1957:690).
This government of objects appears more frequently in some dialects of
Georgian than in others.
(28) naxvas moc’q’urebuli
see.DAT thirsting.NOM
‘Thirsting for seeing’
The ad-form gradually disappeared from the literary language, beginning
around the 10th century AD. The part of the system that changed first was
the dative/nominative case marking with ad-forms being replaced by the
genitive marking. Chkhubianishvili notes that the infinitive construction
remained for a longer time in religious texts and other texts written in a
higher, archaic style. The ad-form was also found in purpose clauses for
several centuries longer.
Although the ad-form has by now completely disappeared from the
modern literary language9, the form is still found in some peripheral dialects
spoken in the mountains of northern Georgia such as khevsur, tushetian, mtiul,
upper racha and others (Chkhubianishvili 1972:155-156). It has remained
mainly in the purpose clause use, as in the tushetian example below:
9One type of purpose clause in Modern Georgian is formed by a future participle in the
adverbial case as the subordinate predicate, having its object in the genitive case. This form is
derived from a participle, not a verbal noun as in Old Georgian.
ßevedi samzareuloßi sadilis gasak’eteblad
1.enter.AOR kitchen.(DAT).in lunch.GEN do/PART.FUT-ADVERBIAL
‘I entered the kitchen in order to prepare lunch’
16 MANANA KOBAIDZE AND KARINA VAMLING
(29) … rusis µaric ßemosuliq’w ßamilis daç’erad
… Russian army.NOM.too S3SG.came Shamil.GEN catch.AD
‘The Russian army came to get hold of Shamil’
13. Conclusions
As we have seen, Old Georgian exhibits a number of cases where a notional
argument of the ad-form, the subject or direct object, syntactically appear as a
dependent of the matrix predicate. This phenomenon may thus be described in
terms of raising (cf. Boeder 1989:73-74).
To summarise, we find examples of:
(1) direct object to direct object raising (5a-b), (6a-b)
(2) direct object to subject raising (8)
(3) subject to indirect object raising (14)
(4) subject to subject raising (15).
The marking of the raised object includes not only case marking, but also
object agreement in the matrix verb (11), (13).
References
Boeder, Winfried. 1989. ‘Zur Typologie der Satzverknüpfung in den Kauka-
sischen Sprachen’. Annual of Ibero-Caucasian Linguistics 16, 67-87.
Chikobava (çikobava), Arnold. 1953. ‘masdarisa da mimeobis ist’oriuli
urtiertobisatvis kartulßi’ [The relation between the masdar and participle in
Georgian in a historical perspective]. iberiul-k’avk’asiuri enatmecniereba
5, 33-49. Tbilisi: Mecniereba.
Chkhubianishvili (çxubianißvili), Darejan. 1972. inpinit’ivis sak’itxisatvis
vel
kartul
ßi
[On the problem of the infinitive in Old Georgian]. Tbilisi:
Mecniereba.
Dzidziguri (iiguri), Shota. 1989. rtuli c’inadadebis p’roblema kartul ena
ß
i
[The problem of complex sentences in Georgian]. Tbilisi: Mecniereba.
Imnaishvili (imnaißvili), Ivane. 1957. saxelta bruneba da brunvata punkciebi
vel kartul
ßi
[Declination of nouns and functions of cases in Old Georgian].
Tbilisi: Mecniereba.
Jorbenadze (jorbenae), Besarion. 1995. kartuli enis morpologia [The
morphology of Georgian]. Tbilisi: Mecniereba.
Kiziria (k’iziria), Anton. 1963. mart’ivi c’inadadebis
ß
edgeniloba
vel kartul
ß
i
[The simple sentence in Old Georgian]. Tbilisi: Sakartvelos mecnierebata
ak’ademiis gamomcemloba.
Martirosovi (mart’irosovi), Aram. 1955. ‘masdaruli k’onst’rukciis
genezisisatvis vel kartulßi’ [On the genesis of masdar constructions in Old
CASE MARKING IN INFINITIVE CLAUSES IN OLD GEORGIAN 17
Georgian]. iberiul-k’avk’asiuri enatmecniereba 7, 43-60. Tbilisi:
Mecniereba.
Noonan, Michael. 1985. ‘Complementation’. In T. Shopen (ed.), Language
typology and syntactic description 2, 42-140. Cambridge: Cambridge
University Press.
Schanidse, Akaki. 1982. Grammatik der Altgeorgischen Sprache. Tbilisi:
Tbilisis universit’et’is gamomcemloba.
Shanidze (ßanie), Akaki. 1976.
veli kartuli enis gramat’ik’a [Grammar of
Old Georgian]. Tbilisi: Tbilisis universit’et’is gamomcemloba.
Shanidze (ßanie), Akaki. 1980. kartuli enis gramat’ik’is sapu
vlebi
[Foundations of Georgian grammar]. Tbilisi: Tbilisis universit’et’is
gamomcemloba.
Sources
Haem. mark.: haemet’i t’ekst’ebi. Tbilisis universit’et’is moambe III. 1923.
I.DE: kartuli otxtavis ori
veli redakcia sami
ß
at’berduli xelnac’eris mixedvit. Tbilisi 1945.
Kim.I: kartuli hagiograpiuli
eglebi I. Tbilisi 1918.
Luars. mart.: luarsabis mart’viloba:
veli kartuli lit’erat’uris krest’omatia I. Tbilisi 1946.
Luka: see I.DE.
M.DE: see I.DE.
Mamata sts.: mamata sc’avlani X da XI saukuneta xelnac’erebis mixedvit. Tbilisi 1955.
Mark.: see I.DE.
Ps.103.27: psalmis
veli kartuli radakciebi X-XIII s-is xelnac’erta mixedvit I. 1960.
Romaelta mimart: axali a
tkumaj uplisa
ç
uenisa ieso krist’esi. Tbilisi 1963.
Shush: kart. krest.I.1: c’amebaj c’midisa ßußanik’isi : kartuli enis ist’oriuli krest’omatia, I,
1. Tbilisi 1982.
Sin. mr.: sinuri mravaltavi 864 c’lisa. Tbilisi 1959.
Vahanis kvab. gang.: vahanis kvabta gangeba. Tbilisi 1939.
Xanm. lekc.: xanmet’i lekcionari. Tbilisi 1944.
Manuscripts: A 1115, S-384.
Article
One Pontic Greek variety, Romeyka of Of, Turkey, today preserves a robust infinitive usage. Comparing the current infinitival distribution in Romeyka with previous stages of Greek, I argue that: (a) the Romeyka infinitive has roots in Ancient Greek due to preservation of the construction prin “before” with infinitive, which remains extremely productive, but which did not survive in other varieties into early medieval times and is only found as a learned construction in ‘high’ registers of the Medieval Greek record; (b) neither the survival of the plain and personal infinitive, nor the emergence of the inflected one can be due to contact with Turkish; (c) the Romeyka infinitive, part of a conservative medieval variety with Hellenistic features, once cut off from other medieval varieties (as early as the 11th c. ce and as late as the 16th c. ce), was reanalyzed as a negative polarity item. Such reanalysis feeds into the discussion that NPIs belong to various syntactic categories, such as nominal NPIs, NPI adverbs, NPI verbs, NPI focus particles, minimizers and now an infinitive, too.
Article
Dummy head noun; Georgian; Head-final word-order; Inclusive/exclusive; Opposition; Initial consonant cluster; Kartvelian; Laz; Medioactive; Mediopassive; Middle voice; Mingrelian; Number suppletion; Old Georgian; Polypersonal verb; Prefix sequence constraint; Reduplication; Relative pronouns; Reported speech; Slot filling; Sound symbolism; Suffix sequence constraint; Suffixaufnahme; Svan; Vigesimal Lingua 115 (2005) 5‐89
Zur Typologie der Satzverkniipfung in den Kaukasischen Sprachen'. Annual of Ibero-Caucasian Linguistics
  • Winfried Boeder
Boeder, Winfried. 1989. 'Zur Typologie der Satzverkniipfung in den Kaukasischen Sprachen'. Annual of Ibero-Caucasian Linguistics 16, 67-87.
Arnold. 1953. 'masdarisa da mim∞eobis ist'oriuli urtiertobisatvis kartulßi' [The relation between the masdar and participle in Georgian in a historical perspective]. iberiul-k'avk'asiuri enatmecniereba 5
  • Chikobava
Chikobava (çikobava), Arnold. 1953. 'masdarisa da mim∞eobis ist'oriuli urtiertobisatvis kartulßi' [The relation between the masdar and participle in Georgian in a historical perspective]. iberiul-k'avk'asiuri enatmecniereba 5, 33-49. Tbilisi: Mecniereba.
Darejan. 1972. inpinit'ivis sak'itxisatvis Ωvel kartulßi
  • Chkhubianishvili
Chkhubianishvili (çxubianißvili), Darejan. 1972. inpinit'ivis sak'itxisatvis Ωvel kartulßi [On the problem of the infinitive in Old Georgian]. Tbilisi: Mecniereba.
Shota. 1989. rtuli c'inadadebis p'roblema kartul enaßi [The problem of complex sentences in Georgian
  • Dzidziguri
Dzidziguri (ΩiΩiguri), Shota. 1989. rtuli c'inadadebis p'roblema kartul enaßi [The problem of complex sentences in Georgian]. Tbilisi: Mecniereba.
Ivane. 1957. saxelta bruneba da brunvata punkciebi jve/ kartulsi [Declination of nouns and functions of cases in Old Georgian
  • Imnaishvili
Imnaishvili (imnaisvili), Ivane. 1957. saxelta bruneba da brunvata punkciebi jve/ kartulsi [Declination of nouns and functions of cases in Old Georgian]. Tbilisi: Mecniereba.
Aram. 1955. 'masdaruli k'onst'rukciis genezisisatvis Ωvel kartulßi
  • Martirosovi
Martirosovi (mart'irosovi), Aram. 1955. 'masdaruli k'onst'rukciis genezisisatvis Ωvel kartulßi' [On the genesis of masdar constructions in Old Georgian]. iberiul-k'avk'asiuri enatmecniereba 7, 43-60. Tbilisi: Mecniereba.
Ωveli kartuli enis gramat'ik'a
  • Shanidze
Shanidze (ßaniΩe), Akaki. 1976. Ωveli kartuli enis gramat'ik'a [Grammar of Old Georgian]. Tbilisi: Tbilisis universit'et'is gamomcemloba.
Akaki. 1980. kartuli enis gramat'ik'is sapuΩvlebi
  • Shanidze
Shanidze (ßaniΩe), Akaki. 1980. kartuli enis gramat'ik'is sapuΩvlebi [Foundations of Georgian grammar]. Tbilisi: Tbilisis universit'et'is gamomcemloba.
kartuli otxtavis ori Ωveli redakcia sami ßat'berduli xelnac'eris mixedvit. Tbilisi 1945. Kim.I: kartuli hagiograpiuli Ωeglebi I
  • I De
I.DE: kartuli otxtavis ori Ωveli redakcia sami ßat'berduli xelnac'eris mixedvit. Tbilisi 1945. Kim.I: kartuli hagiograpiuli Ωeglebi I. Tbilisi 1918.
luarsabis mart'viloba: sveli kartuli lit'erat'uris krest'omatia I. Tbilisi 1946. Luka: see I.DE. M.DE: see I.DE. Mamata sts.: mamata sc'avlani X da XI saukuneta xelnac'erebis mixedvit
  • Mart Luars
Luars. mart.: luarsabis mart'viloba: sveli kartuli lit'erat'uris krest'omatia I. Tbilisi 1946. Luka: see I.DE. M.DE: see I.DE. Mamata sts.: mamata sc'avlani X da XI saukuneta xelnac'erebis mixedvit. Tbilisi 1955. Mark.: see I.DE.