DAIS - Digital Archive of the Serbian Academy of Sciences and Arts
    • English
    • Српски
    • Српски (Serbia)
  • English 
    • English
    • Serbian (Cyrillic)
    • Serbian (Latin)
  • Login
View Item 
  •   DAIS
  • Институт за српски језик САНУ / Institute for the Serbian Language of SASA
  • Јужнословенски филолог / Južnoslovenski filolog
  • View Item
  •   DAIS
  • Институт за српски језик САНУ / Institute for the Serbian Language of SASA
  • Јужнословенски филолог / Južnoslovenski filolog
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

О неким глаголским конструкцијама у српскохрватском

Some Complement Constructions in Serbo-Croatian

Thumbnail
1981
kojen.glag.konstrukcije.1981.pdf (2.817Mb)
Authors
Којен, Леон
Article (Published version)
,
Институт за српски језик САНУ
Metadata
Show full item record
Abstract
The central features of Serbo-Croatian complementation may be conveniently brought out by analysing different constructions with да, а syntactic operator whose use in complement structures overlaps that of both that and for in English. Within the framework of an extended categorial grammar (Kojen 1979), да in its relevant uses is naturally treated as a sentential nominalizer and assigned to the category n/S: it takes a sentence as an argument and yields a nominal which then occurs either preverbally, as subject, or postverbally, in a variety of positions, as a complement to different types of verb. When such a nominal occurs in subject position, the sentence optionally undergoes Extraposition which permutes the subject and the predicate; for SVO languages, as Kuno has shown (Kuno 1974), extraposed structures of this kind represent a natural solution to the perceptual difficulties created by center-embedding and the consequent juxtaposition of particles (including nominalizers)... which have the function of marking clause boundaries. Rather similarly, when nominalized sentences are embedded under a small but important class of „impersonal” verbs (треба, мора, изгледа, etc.), they optionally undergo Extraction, a transformation which moves the initial constituent of the embedded sentence and places it to the left of the group consisting of the „impersonal” verb and the nominalizer. When the embedded sentence is subject-predicate inform, Extraction is of course just Subject-to-Subject Raising under another name. But, at least for a Slavic language like Serbo-Croatian, the usual name and its implications are clearly misleading since Extraction is just as applicable when the embedded sentence itself canonically lacks a subject: what is fronted by Extraction is the initial constituent of the embedded sentence, whether subject (Изгледа да Ана није тужна -- Ана изгледа да није тужна ), adverbial (Изгледа да овде нема никога -- Овде изгледа да нема никога) or dative nominal (Изгледа да Петру није хладно- Петру изгледа да није хладно). Both Extraposition and Extraction apply optionally to all structures, even those with multiple center-embedding: the latter are excluded by a special filter which marks as ill-formed any string whose associated sequence of category indices contains a sub-sequence (...in/S, n/S, . . .). Араrt from the small class of „impersonal” verbs, which are categorially one-place verbals, S/n, nominalized sentences normally occur as complements to two-place and three-place verbals, (S/n)/n and ((S/n)/n)/n respectively. The former group includes „intentional” verbs, both of the epistemic, believe type and of the non-epistemic, want type (мислити, веровати, знати, желети, хтети, etc.), and the latter various,,control” verbs analogous to promise and persuade (обећати, наговарати, уверавати, дозволити, рећи, захтевати, крити, etc.). It is shown that verbs of both groups undergo Equi NP Deletion, those in the latter group falling into three distinct sub-classes with respect to the problem of „control”: the nominal which „controls” the deletion of the embedded subject is always the main clause subject (very few cases), the main clause object or structurally analogous dative or genitive nominal (наговарати, дозволити, захтевати, etc.), or under differently specifiable conditions either the one or the other (уверавати, рећи, крити, etc.). Да аlso appears with what one might call „operator” verbs, i.e. verbs which from a logical point of view represent either sentential operators (e.g., моћи, смети аnd морати which are roughly equivalent to English can, may and must respectively)or predicate operators (e.g., почети = begin, престати = stop, успети = succeed, etc.). Such verbs should be carefully distinguished, on both syntactic and semantic grounds, from „intentional” verbs subject to Еqui NP Deletion; sentences in which they occur have complex predicates consisting of the ,,operator” verbal („operator” verb plus the nominalizer) and a simple predicate, but there is no reason to construe these as containing a truncated complement clause whose subject has been deleted by Equi. Finally, to mark formally the differences between „ordinary” verbs and verbs which appear in various structures with да, the latter are associated with specific restrictions on permissible derivations: for instance, we shall stipulate in certain cases that at the appropriate step of any derivation for a string containing a form of И the relevant verbal (which may or may not consist of just the given form of V) must take an argument obtained from n/S and S.

Source:
Јужнословенски филолог, 1981, 37, 55-89
Publisher:
  • Београд : Институт за српскохрватски језик

ISSN: 0350–185x

[ Google Scholar ]
Handle
https://hdl.handle.net/21.15107/rcub_dais_5820
URI
https://dais.sanu.ac.rs/123456789/5820
Collections
  • Јужнословенски филолог / Južnoslovenski filolog
Institution/Community
Институт за српски језик САНУ / Institute for the Serbian Language of SASA
TY  - JOUR
AU  - Којен, Леон
PY  - 1981
UR  - https://dais.sanu.ac.rs/123456789/5820
AB  - The central features of Serbo-Croatian complementation may be conveniently brought out by analysing different constructions with да, а syntactic operator whose use in complement structures overlaps that of both that
and for in English. Within the framework of an extended categorial grammar
(Kojen 1979), да in its relevant uses is naturally treated as a sentential nominalizer and assigned to the category n/S: it takes a sentence as an argument and
yields a nominal which then occurs either preverbally, as subject, or postverbally, in a variety of positions, as a complement to different types of verb.
When such a nominal occurs in subject position, the sentence optionally
undergoes Extraposition which permutes the subject and the predicate; for
SVO languages, as Kuno has shown (Kuno 1974), extraposed structures of
this kind represent a natural solution to the perceptual difficulties created
by center-embedding and the consequent juxtaposition of particles (including
nominalizers) which have the function of marking clause boundaries. Rather
similarly, when nominalized sentences are embedded under a small but
important class of „impersonal” verbs (треба, мора, изгледа, etc.), they
optionally undergo Extraction, a transformation which moves the initial
constituent of the embedded sentence and places it to the left of the group
consisting of the „impersonal” verb and the nominalizer. When the embedded sentence is subject-predicate inform, Extraction is of course just Subject-to-Subject Raising under another name. But, at least for a Slavic language
like Serbo-Croatian, the usual name and its implications are clearly misleading
since Extraction is just as applicable when the embedded sentence itself canonically lacks a subject: what is fronted by Extraction is the initial constituent
of the embedded sentence, whether subject (Изгледа да Ана није тужна --
Ана изгледа да није тужна ), adverbial (Изгледа да овде нема никога --
Овде изгледа да нема никога) or dative nominal (Изгледа да Петру није
хладно- Петру изгледа да није хладно). Both Extraposition and Extraction
apply optionally to all structures, even those with multiple center-embedding:
the latter are excluded by a special filter which marks as ill-formed any string
whose associated sequence of category indices contains a sub-sequence (...in/S,
n/S, . . .). Араrt from the small class of „impersonal” verbs, which are categorially one-place verbals, S/n, nominalized sentences normally occur as
complements to two-place and three-place verbals, (S/n)/n and ((S/n)/n)/n
respectively. The former group includes „intentional” verbs, both of the
epistemic, believe type and of the non-epistemic, want type (мислити, веровати,
знати, желети, хтети, etc.), and the latter various,,control” verbs analogous
to promise and persuade (обећати, наговарати, уверавати, дозволити, рећи,
захтевати, крити, etc.). It is shown that verbs of both groups undergo Equi NP Deletion, those in the latter group falling into three distinct sub-classes
with respect to the problem of „control”: the nominal which „controls” the
deletion of the embedded subject is always the main clause subject (very few
cases), the main clause object or structurally analogous dative or genitive
nominal (наговарати, дозволити, захтевати, etc.), or under differently specifiable conditions either the one or the other (уверавати, рећи, крити, etc.).
Да аlso appears with what one might call „operator” verbs, i.e. verbs which
from a logical point of view represent either sentential operators (e.g., моћи,
смети аnd морати which are roughly equivalent to English can, may and
must respectively)or predicate operators (e.g., почети = begin, престати =
stop, успети = succeed, etc.). Such verbs should be carefully distinguished, on
both syntactic and semantic grounds, from „intentional” verbs subject to
Еqui NP Deletion; sentences in which they occur have complex predicates
consisting of the ,,operator” verbal („operator” verb plus the nominalizer)
and a simple predicate, but there is no reason to construe these as containing
a truncated complement clause whose subject has been deleted by Equi.
Finally, to mark formally the differences between „ordinary” verbs and
verbs which appear in various structures with да, the latter are associated
with specific restrictions on permissible derivations: for instance, we shall
stipulate in certain cases that at the appropriate step of any derivation for
a string containing a form of И the relevant verbal (which may or may
not consist of just the given form of V) must take an argument obtained
from n/S and S.
PB  - Београд : Институт за српскохрватски језик
T2  - Јужнословенски филолог
T1  - О неким глаголским конструкцијама у српскохрватском
T1  - Some Complement Constructions in Serbo-Croatian
SP  - 55
EP  - 89
VL  - 37
UR  - https://hdl.handle.net/21.15107/rcub_dais_5820
ER  - 
@article{
author = "Којен, Леон",
year = "1981",
abstract = "The central features of Serbo-Croatian complementation may be conveniently brought out by analysing different constructions with да, а syntactic operator whose use in complement structures overlaps that of both that
and for in English. Within the framework of an extended categorial grammar
(Kojen 1979), да in its relevant uses is naturally treated as a sentential nominalizer and assigned to the category n/S: it takes a sentence as an argument and
yields a nominal which then occurs either preverbally, as subject, or postverbally, in a variety of positions, as a complement to different types of verb.
When such a nominal occurs in subject position, the sentence optionally
undergoes Extraposition which permutes the subject and the predicate; for
SVO languages, as Kuno has shown (Kuno 1974), extraposed structures of
this kind represent a natural solution to the perceptual difficulties created
by center-embedding and the consequent juxtaposition of particles (including
nominalizers) which have the function of marking clause boundaries. Rather
similarly, when nominalized sentences are embedded under a small but
important class of „impersonal” verbs (треба, мора, изгледа, etc.), they
optionally undergo Extraction, a transformation which moves the initial
constituent of the embedded sentence and places it to the left of the group
consisting of the „impersonal” verb and the nominalizer. When the embedded sentence is subject-predicate inform, Extraction is of course just Subject-to-Subject Raising under another name. But, at least for a Slavic language
like Serbo-Croatian, the usual name and its implications are clearly misleading
since Extraction is just as applicable when the embedded sentence itself canonically lacks a subject: what is fronted by Extraction is the initial constituent
of the embedded sentence, whether subject (Изгледа да Ана није тужна --
Ана изгледа да није тужна ), adverbial (Изгледа да овде нема никога --
Овде изгледа да нема никога) or dative nominal (Изгледа да Петру није
хладно- Петру изгледа да није хладно). Both Extraposition and Extraction
apply optionally to all structures, even those with multiple center-embedding:
the latter are excluded by a special filter which marks as ill-formed any string
whose associated sequence of category indices contains a sub-sequence (...in/S,
n/S, . . .). Араrt from the small class of „impersonal” verbs, which are categorially one-place verbals, S/n, nominalized sentences normally occur as
complements to two-place and three-place verbals, (S/n)/n and ((S/n)/n)/n
respectively. The former group includes „intentional” verbs, both of the
epistemic, believe type and of the non-epistemic, want type (мислити, веровати,
знати, желети, хтети, etc.), and the latter various,,control” verbs analogous
to promise and persuade (обећати, наговарати, уверавати, дозволити, рећи,
захтевати, крити, etc.). It is shown that verbs of both groups undergo Equi NP Deletion, those in the latter group falling into three distinct sub-classes
with respect to the problem of „control”: the nominal which „controls” the
deletion of the embedded subject is always the main clause subject (very few
cases), the main clause object or structurally analogous dative or genitive
nominal (наговарати, дозволити, захтевати, etc.), or under differently specifiable conditions either the one or the other (уверавати, рећи, крити, etc.).
Да аlso appears with what one might call „operator” verbs, i.e. verbs which
from a logical point of view represent either sentential operators (e.g., моћи,
смети аnd морати which are roughly equivalent to English can, may and
must respectively)or predicate operators (e.g., почети = begin, престати =
stop, успети = succeed, etc.). Such verbs should be carefully distinguished, on
both syntactic and semantic grounds, from „intentional” verbs subject to
Еqui NP Deletion; sentences in which they occur have complex predicates
consisting of the ,,operator” verbal („operator” verb plus the nominalizer)
and a simple predicate, but there is no reason to construe these as containing
a truncated complement clause whose subject has been deleted by Equi.
Finally, to mark formally the differences between „ordinary” verbs and
verbs which appear in various structures with да, the latter are associated
with specific restrictions on permissible derivations: for instance, we shall
stipulate in certain cases that at the appropriate step of any derivation for
a string containing a form of И the relevant verbal (which may or may
not consist of just the given form of V) must take an argument obtained
from n/S and S.",
publisher = "Београд : Институт за српскохрватски језик",
journal = "Јужнословенски филолог",
title = "О неким глаголским конструкцијама у српскохрватском, Some Complement Constructions in Serbo-Croatian",
pages = "55-89",
volume = "37",
url = "https://hdl.handle.net/21.15107/rcub_dais_5820"
}
Којен, Л.. (1981). О неким глаголским конструкцијама у српскохрватском. in Јужнословенски филолог
Београд : Институт за српскохрватски језик., 37, 55-89.
https://hdl.handle.net/21.15107/rcub_dais_5820
Којен Л. О неким глаголским конструкцијама у српскохрватском. in Јужнословенски филолог. 1981;37:55-89.
https://hdl.handle.net/21.15107/rcub_dais_5820 .
Којен, Леон, "О неким глаголским конструкцијама у српскохрватском" in Јужнословенски филолог, 37 (1981):55-89,
https://hdl.handle.net/21.15107/rcub_dais_5820 .

DSpace software copyright © 2002-2015  DuraSpace
About DAIS - Digital Archive of the Serbian Academy of Sciences and Arts | Send Feedback

CoreTrustSealre3dataOpenAIRERCUB
 

 

All of DSpaceInstitutions/communitiesAuthorsTitlesSubjectsThis institutionAuthorsTitlesSubjects

Statistics

View Usage Statistics

DSpace software copyright © 2002-2015  DuraSpace
About DAIS - Digital Archive of the Serbian Academy of Sciences and Arts | Send Feedback

CoreTrustSealre3dataOpenAIRERCUB