Search

Predmet disertacije je sintaksa i semantika adnominalnog pridevskog vida u srpskom jeziku. Primenjena je kartografska generativnosintaksička i formalnosemantička metodologija u modelovanju empirijskih činjenica, prikupljenih ekscerpiranjem iz književnoumetničkih tekstova iz 19. i 20. veka, dnevne štampe, sa Interneta, te pomoću sudova gramatičnosti na osnovu jezičke intuicije izvornih govornika. Iako se često u lingvističkoj literaturi proglašava gotovo neutralisanom, utvrdili smo da se kategorijom pridevskog vida u srpskom jeziku markiraju (ne)određenost (starost u diskursu), (ne)jedinstvenost, epistemička (ne)specifičnost, vremenski lokalna ili globalna interpretacija pridevske modifikacije, kvantifikativna i generička (ne)restriktivnost, te supsektivnost ili intersektivnost. Određeni vid prideva je dvosmislen u pogledu pobrojanih kategorija, dok se neodređenim vidom prideva obeležavaju neodređenost, nejedinstvenost, epistemička nespecifičnost, vremenski lokalno čitanje, kvantifikativna i generička restriktivnost, te supsektivnost pridevske modifikacije. Iako morfološki markiran, određeni vid je semantički markiran, dok se morfološki neobeleženi neodređeni vid ponaša kao semantički markirana kategorija. Postuliran je odgovarajući sintaksički model, koji sadrži dva domena podeljene determinatorske fraze (DP) – domen klase / vrste, lociran ispod funkcionalne projekcije partivnosti, i domen individue / objekta, koji se nalazi iznad projekcije za kardinalni broj. Predložena je analiza u kojoj se određeni vid prideva dobija specifičnim slaganjem prideva sa referencijalnim obeležjima određenosti ili specifičnosti u nekom od dvaju DP domena, čime je objašnjena dvosmislenost određenog vida. Prikrivenim pomeranjem prideva u tzv. logičkoj formi (LF) i vezivanjem za operator neke više funkcionalne projekcije, poput projekcije subjektivne evaluacije govornika, kvantifikatorske ili projekcije aspektualne komponente glagola, moguće je da pridev izađe iz domena DP-ja, čime izbegava slaganje i ostaje u neodređenom vidu. Ovako je interpretirana pojava neodređenog vida prideva u određenom (diskurs-starom), jedinstvenom ili epistemički specifičnom okruženju, u prisustvu kvantifikatora i kod vremenski lokalnog čitanja modifikacije. The subject of the dissertation is syntax and semantics of adnominal adjective aspect in Serbian language. A cartographic generative syntactical and formal semantics’ methodology have been applied in modeling of empirical facts, collected by excerpting from 19th and 20th century literature texts, daily press, the Internet and by means of grammaticality judgments based on language intuition of native speakers. Although in linguistic literature it is often claimed to be almost neutralized, we determined that adjective aspect can mark (in)definiteness (discourse-giveness), (non-)uniqueness, epistemic (non-)specificity, stage- / individual-level interpretation, quantifier and generic (non-)restrictiveness, as well as subsectivity / intersectivity. The definite adjective aspect is ambiguous regarding all of these categories, while indefinite adjective aspect can mark indefiniteness, non-uniqueness, epistemic non-specificity, stage-level interpretation, quantifier and generic restrictiveness, and subsectivity. Although morphologically marked, definite adjective aspect is semantically non-marked, while the morphologically non-marked indefinite adjective aspect behaves as the semantically marked form. An appropriate syntactic model has been postulated, which contains two split-DP domains – kind DP, located beneath the partitive projection, and individual DP, above the cardinality projection. In the proposed analysis, definite adjective aspect is a result of a specific type of agreement between the adjective and the referential features of definiteness or specificity in either of the two DP domains, which explains the ambiguity of the definite adjective aspect. By a covert movement of the adjective in LF and binding to an operator of a higher functional projection, such as speaker’s subjective evaluation projection, quantifier or verb aspect projection, it is possible for the adjective to exit the DP domain, escape the agreement and stay in the form of indefinite adjective aspect. This is how we interpreted indefinite adjective aspect in definite, unique and specific contexts, in the presence of a quantifier and in stage-level interpretation.

We conducted a comparison of preverbal and postverbal subjectverb agreement in an elicited production study carried out with 60 speakers of different language varieties in South Slavic, spanning Bosnian, Croatian, Serbian, and Slovenian at six local universities. These languages have three genders, and we measured the response type and total reaction time for production for the nine possible combinations of two plural noun phrases. Elicited production experiments for the preverbal and postverbal versions were conducted separately, with 54 items and 54 fillers. A sample target item involved a model sentence, immediately followed by a replacement phrase, and participants had to produce the entire sentence that resulted from combining the new replacement subject noun phrase (a coordination, such as the books and the magazines), with the remaining predicate, making morphosyntactic adjustments where necessary (e.g. in agreement). Our analysis of the gender agreement reveals that preverbal versus postverbal positioning make a large difference in the availability with the linearly furthest conjunct: while highest agreement is possible preverbally (e.g. the FN and NF columns), lowest-conjunct agreement is essentially unattested postverbally. In terms of reaction times, we find longest latencies for the conditions where speakers have the most grammatical options to choose among, suggesting that all three strategies are in principle available, although constrained by syntactic (hierarchical) and morphological factors (default masculine and relative markedness). The consequences of these results will be discussed with respect to three recent theoretical models of closest-conjunct agreement in South Slavic (Boskovic 2009, Puskar & Murphy 2014, Marusic, Nevins & Badecker 2015).

Psych verbs or experiencer verbs are a class of verbs that entail emotional states on the part of one of the participants in the situation they denote (Dowty 1991). This class of verbs raises important challenges for some of the core theories and assumptions in the Generative Framework such as the Uniformity of Theta Assignment Hypothesis (Baker 1998) or Binding Theory (Chomsky 1986). Previous research suggests that assuming that these challenges are only apparent can lead to improvements of the theories in question as well as yield important insights into the category of experiencer verbs itself (Belletti and Rizzi 1988; Pesetsky 1994; Landau 2010).This dissertation tackles the data from Serbian psych verbs from the standpoint of the classification of the same category of verbs in English informed by cross linguistic research (Levin 1994). The empirical contribution of the dissertation lies in the fact that the data are described and classified following the crosslinguistically-established formal criteria (cf. Levin 1994) while drawing on the previous descriptions, which were primarily informed by functionalist approaches (Štrbac 2006; Milenković 2017).The descriptive investigation gave rise to a series of theoretical questions pertaining to the argument structure of these verbs, the thematic roles that they assign and the syntactic status of different types of NPs that are found with these verbs, case assignment mechanisms, and the role of the morpheme SE. These questions are tackled separately, in dedicated chapters, within the framework of Distributed Morphology (Halle and Marantz 1993).Regarding the role of SE, the data from Serbian psych verbs (based primarily on event modifier licensing) motivate the existence of a semantic class between anticausatives and reflexives which I label ‘semi-reflexives’. The proposed class of ‘semi-reflexives’ is then used as a basis to bridge the gap between the two opposing views on the nature of the anticausative SE (Chierchia’s 2004 ‘Reflexive view’ and the ‘Standard Analysis’ a la Parsons 1990 or Schäfer and Vivanco 2016). This solution opens the door to understanding the lack of psych verb anticausatives in English.Oblique case-marked bare NPs expressing experiencers or stimuli with some Serbian psych verbs are argued to be arguments by showing that they cannot be analyzed as adjuncts or complements, and the origin of different oblique case forms (specifically, genitive and dative) is attributed to two different applicative heads (Source Applicative and Goal Applicative) building on Pylkkänen (2008).The possibilities of deriving n-participles and -(n)je nominalizations from Serbian psych verbs were explored under the assumptions that passive participle formation is governed by argument structure properties (Embick 2004) and -(n)je nominalizations are derived from passive participles (Marvin 2002; Simonović and Arsenijević 2014). The ambition behind these investigations was to reveal important facts in both directions (i.e. about the formation and internal structure of passive participles and -(n)je nominals and about the argument structure of psych verbs that underlie them). In a nutshell, it is argued that n-participles can be derived only from those verbs that contain an agentive component (VoiceP), and a somewhat modified version of the hypothesis that -(n)je nominals can be derived only from those verbs that also derive n-participles can be maintained.

U disertaciji se eksperimentalnom metodom ispituje uticaj gramatičkih karakteristika koordiniranih konjunkata, kao što su vrednost gramatičkog roda, vrednost gramatičkog broja, linearna i hijerarhijska bliskost, kao i semantičkih osobina poput osobina referencije, deljenje modifikatora, na slaganje glagola sa celim koordiniranim subjektom (tzv. razrešeno slaganje), sa prvim konjunktom ili sa poslednjim konjunktom. Cilj je da se dâ model procesiranja slaganja u srpskom jeziku, ali i da se odgovori na značajna teorijska pitanja kao što su da li se slaganje dešava i na interfejsu sa fonologijom, te kakva je međusobna uloga roda i broja u procesu slaganja u srpskom jeziku. Cilj je i da se utvrdi koji faktori utiču na slaganje glagola u rodu sa prvim članom koordiniranog subjekta, kao i da li srpski jezik podržava hipotezu prema kojoj su rod i broj odvojeni i odvojeno učestvuju u procesu slaganja. Kako rezultati sprovedenih produkcijskih eksperimenata i eksperimenata sa sudovima gramatičnosti kojima se testira slaganje glagola sa koordiniranim modifikovanim subjektom pokazuju, slaganje glagola u rodu sa poslednjim članom konjunkcije najčešće je upotrebljen obrazac slaganja. Ovaj obrazac je značajno manje produkovan za uslov u kome je prvi član koordiniranog subjekta modifikovan nego kada je poslednji član modifikovan ili su oba člana modifikovana. Za prvi uslov slaganje glagola u rodu sa prvim članom konjunkcije je značajno više produkovano nego za preostala dva uslova. Kada je linearna distanca između prvog člana konjunkcije i glagola minimizovana, umanjena je šansa da se glagol složi u rodu sa poslednjim članom konjunkcije koji mu je linearno najbliži. Naši rezultati potvrđuju da je linearna distanca faktor koji utiče ne samo na slaganje glagola sa poslednjim članom konjunkcije, nego i na slaganje glagola u rodu sa prvim članom konjunkcije, kao i da hijerarhijska i linearna distanca zajedno određuju izbor kontrolora slaganja u okviru jedne gramatike.10 Rezultati produkcijskih eksperimenata i eksperimenata sa sudovima gramatičnosti kojima se testira efekat sinkretizma na slaganje glagola u rodu potvrđuju da je razrešeno slaganje značajnije umanjeno kada se glagol slaže sa koordiniranim sinkretičnim subjektom mešovitog roda i broja nego kada se slaže sa koordiniranim nesinkretičnim subjektom. Naši rezultati potvrđuju da sinkretizam teži množinskoj interpretaciji, kao i da postoji manja šansa da se glagol složi u rodu sa prvim članom koordiniranog subjekta koji je u množini, nego sa poslednjim članom. Kada je množinski konjunkt na mestu poslednjeg člana konjunkcije, vrednost za broj biva udružena sa efektima koji dolaze iz fonologije (efekat sinkretizma i efekat linearne bliskosti), što jača linearno slaganje, pa je razrešeno slaganje značajnije umanjeno. Rezultati idu u prilog hipotezi da se slaganje dešava i na interfejsu sa fonologijom. U sprovedenom eksperimentalnom istraživanju potvrđena je pojava kombinovanog slaganja na materijalu srpskog jezika, što je suprotno ranijim tvrdnjama Maretića (1899), Korbeta (1983a), Boškovića (2009) za srpski jezik, te Marušiča i dr. (2015) za slovenački jezik, a u skladu sa novijim istraživanjima Arsenijevića i Mitić (2016a) za srpski jezik i Fuhsa i dr. (2015) za španski jezik. Obrazac kombinovanog slaganja, kao obrazac u kome je je kontrolor za slaganje u rodu član ženskog roda u jednini, dok je kontrolor za slaganje u broju u množini, potvrđuje da u srpskom jeziku postoje situacije u kojima vrednost za rod može da prevagne u konkurenciji sa vrednošću za broj. Rezultati idu u prilog hipotezi da su rod i broj u srpskom jeziku odvojeni, i da mogu odvojeno učestvovati u procesu slaganja. This dissertation will use the experimental method to examine the influence of grammatical characteristics of coordinated conjuncts, such as the value of grammatical gender, the value of grammatical number, linear and hierarchical distance, as well as semantic properties such as referential properties, division of modifiers, on the verb agreement with the whole coordinated subject (i.e. resolved agreement), on the first conjunct agreement, or last conjunct agreement. The aim is to give a model for processing agreement in the Serbian language, but also to respond to significant theoretical questions such as whether agreement also occurs in the interface with phonology, and what is the role of gender and number in the process of agreement in the Serbian language. The aim is to also determine which factors influence the verb gender agreement with the first member of the coordinated subject, as well as to determine if the Serbian language supports the hypothesis according to which gender and number are separate and they separately participate in the agreement process. The results of the production experiments and acceptability judgments experiments which test the verb agreement with the coordinated modified subject show that verb agreement with the last conjunct is the most frequently used pattern of agreement. This pattern is significantly less productive for the condition where the first member of the coordinated subject is modified than when the last member is modified, or when both members are modified. For that same condition, the first conjunct agreement is significantly more productive than for the other two conditions. When the linear distance between the first conjunct and the verb is minimized, that also reduces the chances of gender agreement of the verb with the last conjunct which is linearly closest to it. Our results confirm that linear distance is a factor that affects not only last conjunct agreement, but also the verb gender agreement with the first conjunct. The results also confirm that the patterns of agreement compete in determining the agreement controller in the same grammar.12 The results of elicited production experiments and acceptability judgments experiments which test the effect of syncretism on the verb gender agreement show that resolved agreement is more significantly reduced when the last member of the coordinated subject is plural, than when the first member of the coordinated subject is plural. These results confirm that linear distance more significantly reduces the chances of unsuccessful agreement than hierarchical distance. In the conducted experimental research the occurrence of mixed agreement in the Serbian language has been confirmed, contrary to the claims of Maretić (1899), Korbet (1983a), Bošković (2009) for Serbian, Marušić et al. (2015) for Slovenian, and in agreement with recent research conducted by Arsenijević and Mitić (2016a) for Serbian, and Fuchs et al. (2015) for the Spanish language. The pattern of mixed agreement, as a pattern where the controller of the gender agreement is female singular, while the controller for number agreement is plural, confirms that there are situations in the Serbian language where gender value can prevail in competition with the number value. The results support the hypothesis that gender and number are separate in the Serbian language, and that they can separately participate in the agreement process.

The aim of the dissertation entitled Topicalization andLeft Dislocation in English and Serbian is to describeand offer a generative account of the syntactic andinformation-structural properties of topicalization andleft dislocation in the languages in question, twosuperficially similar preposing structures whichexpress the same propositions, but are not felicitous inthe same context. The analysis is not contrastive in thesense that we are looking for English-Serbiancounterparts or vice versa, but the tertiumcomparationis are the phenomena of topicalizationand left dislocation. The fundamental diagnostics ofdifferentiating between the two variants of leftdislocation identified in Serbian is laid out, viz.Hanging Topic Left Dislocation and Contrastive LeftDislocation. The dissertation also explores howinformational structuring of an utterance determinescontextual choices. The lexicon provides the input tothe computational system which by means of featuresbuilds structure via phases and also gives rise todisplacement. The data related to information structurecome from the numeration. The results of the researchindicate that the notion of a topic should bedeconstructed in a combination of the features [+/-a(nchored), +/-c(ontrastive)], similarly to López(2009), anchored in the sense of Birner & Ward(1998), and contrastive in the pragmatic sense of Titov(2013), whereby [+/-c] is parasitic on [+a].Topicalization in both English and Serbian, andContrastive Left Dislocation in Serbian mark topicsvii[+a, +c], whereas Hanging Topic Left DislocationHTLD in both English and Serbian marks topic [+a, -c]. It is argued that adding pragmatic features inaddition to formal ones relevant to the derivation bythe operation which forms the numeration does notoffend the Inclusiveness Condition either. Theinterpretation of an element marked as a topic is theresult of its featural content and its syntactic position,reflecting the interaction of syntax, prosody andpragmatics. Pragmatic features are valued butuninterpretable in the numeration rendering thesyntactic object containing it active for syntacticoperations. Probe (pragmatic) features triggerAgree(ment), but not displacement. It is an edgefeature that drives movement or Internal Merge. It isargued that topicalization in both English and Serbianis generated by movement. The edge feature on C(Force) licenses the TopP if it is required for theinterpretation and if it is structurally possible, asobserved by Jiménez-Fernández & Miyagawa (2014).Unlike in English, the non-phase head T in Serbiancan inherit an edge feature from C and license theTopP in the Spec,TP in those structures which are saidto be incompatible with topicalization in English. Thehead licensing the left-dislocated element is realizedas comma intonation, as argued by Emonds (2004). Itis argued that Hanging Topic Left Dislocation in bothEnglish and Serbian is derived by base-generation ofthe left-dislocated constituent in its surface position(adjoined to a CP), whereas Contrastive LeftDislocation in Serbian by movement (also to aposition attached to a CP). What moves is theresumptive pronoun, and then co-reference isestablished upon adjoining of the left-dislocatedelement via the operation Match or Match+Agree ofBoeckx (2003), which is the only way not to violatethe Inclusiveness Condition. Although bothTopicalization and Contrastive Left Dislocation markcontrastive topics, they have different discoursedistributions, as confirmed by our corpus, thus thisadjoining of a left-dislocated element is justified. Inthe case of Hanging Topic Left Dislocation, coreferencebetween the left-dislocated element and theresumptive pronoun is established via the operationMatch of Boeckx (2003). If the resumptive pronoun isa clitic, it moves to the second position in its intonational phrase triggered by the phonological requirement. Hanging Topic Left Dislocation in both English and Serbian marks referential topics and it isalso a topic-promoting device in Serbian, as argued onrelying on our corpus., Cilj doktorske disertacije pod nazivom Tematizacija ileva dislokacija u engleskom i srpskom jeziku je daopiše i ponudi generativni prikaz sintaksiĉkih iinformacijsko strukturnih osobina tematizacije i levedislokacije u pomenutim jezicima, dveju naizgledsliĉnih struktura za pomeranje reĉeniĉnog elementa uprednje polje koje izraţavaju iste propozicije, ali neodgovaraju istom kontekstu. Analiza nije kontrastivnau smislu da traţimo englesko srpske ekvivalente iliobratno, već je tertium comparationis pojavatematizacije i leve dislokacije. PonuĊeni su i osnovnidijagnostiĉki testovi za razlikovanje dva oblika levedislokacije identifikovana u srpskom, naime levedislokacije odvojene teme i kontrastne levedislokacije. Disertacija takoĊe istraţuje kakoinformacijsko strukturiranje iskaza odreĊujekontekstualne izbore. Leksikon obezbeĊuje ulaznuinformaciju sistemu sintaksiĉkih operacija koji putemobeleţja gradi strukturu u fazama i takoĊe dovodi dopomeranja. Podaci vezani za informacijsku strukturusu dati u numeraciji. Rezultati istraţivanja ukazuju napotrebu da se pojam teme rašĉlani na kombinacijuobeleţja [+/-a, +/-c], sliĉno Lópezu (2009), anaforiĉnou smislu Birnera & Warda (1998), i kontrastno upragmatiĉkom smislu Titove (2013) pri ĉemu je [+/-c]zavisno od [+a]. Tematizacija i u engleskom i usrpskom jeziku i kontrastna leva dislokacija u srpskomobeleţavaju teme kao [+a, +c], dok leva dislokacijaodvojene teme i u engleskom i u srpskom jezikuobeleţava teme kao [+a, -c]. U radu se dokazuje da nidodeljivanje pragmatiĉkih obeleţja pored formalnih obeleţja relevantnih za derivaciju putem operacije koja formira numeraciju ne narušava uslov ukljuĉenosti. Interpretacija elementa obeleţenog kao tema je rezultat njegove kombinacije obeleţja i njegove sintaksiĉke pozicije, što odraţava interakciju sintakse, prozodije i pragmatike. Pragmatiĉka obeleţja su vrednovana, ali netumaĉiva u numeraciji ĉineći sintaksiĉki objekat koji ih sadrţi aktivnim za sintaksiĉke operacije. Upravna (pragmatiĉka) obeleţja uzrokuju slaganje, ali ne i pomeranje. Obeleţje ivice je ono što pokreće pomeranje ili interno spajanje. Argumentujemo da je tematizacija i u engleskom i u srpskom jeziku generisana pomeranjem. Obeleţje ivice na upravnom elementu C (Force) dozvoljava TopP ako je to neophodno za interpretaciju i ako je strukturno moguće, kao što su formulisali Jiménez-Fernández & Miyagawa (2014). Za razliku od engleskog, nefazni upravni element T (upravni element obeleţja vremena) u srpskom jeziku moţe da preuzme obeleţje ivice od C i dozvoli TopP u Spec,TP u onim strukturama za koje se smatra da nisu kompatibilne sa tematizacijom u engleskom jeziku. Upravni element koji dozvoljava levu dislokaciju se realizuje kao intonacijska pauza, kao što predlaţe Emonds (2004). Argumetujemo da je leva dislokacija odvojene teme i u engleskom i u srpskom jeziku nastala generisanjem levo dislociranog konstituenta u mestu realizacije (pridruţenom CP projekciji), dok je kontrastna leva dislokacija u srpskom nastala pomeranjem (takoĊe u poziciju pridruţenu CP projekciji). Ono što se zapravo pomera je rezumptivna zamenica i onda se po pridruţivanju levo dislociranog elementa uspostavlja koreferentnost putem operacije uskladi ili uskladi+sloţi Boeckxa (2003), što predstavlja jedini naĉin da se ne naruši uslov ukljuĉenosti. Iako i tematizacija i kontrastna leva dislokacija obeleţavaju kontrastne teme, one imaju razliĉite diskursne distribucije, što je potvrdio naš korpus, stoga je ovo pridruţivanje levo dislociranog elementa opravdano. U sluĉaju leve dislokacije odvojene teme, koreferentnost izmeĊu levo dislociranog elementa i rezumptivne zamenice se uspostavlja putem operacije uskladi (Boeckx 2003). Ako je rezumptivna zamenica klitika, ona se pomera u drugu poziciju u svojoj intonacijskoj frazi, što jeuzrokovano fonološkim zahtevom. Leva dislokacija odvojene teme i u englesko i u srpskom jeziku obeleţava referencijske teme i takoĊe je sredstvo unapreĊivanja teme u srpskom, kao što se argumentuje na osnovu našeg korpusa.