Publications of Tamm, A

Scalar Verb Classes. Scalarity, Thematic Roles, and Arguments in the Estonian Aspectual Lexicon

This monograph discusses scalar verb classes. It tests theories of linguistic form and meaning, arguments and thematic roles, using Estonian data. The analyses help to understand the aspectual structure of Estonian. In Estonian, transitive verbs fall into aspectual classes based on the type of case-marking of objects and adjuncts. The book relates the morphosyntactic frames of verbs to properties typically associated with adjectives and nouns: scalarity and boundedness. Verbs are divided according to how their aspect is composed. Some verbs lexicalize a scale, which can be bounded either lexically or compositionally. Aspectual composition involves the unification of features. Compositionally derived structures differ according to which of the aspectually relevant dimensions are bounded.

Tamm A. Intermoduláris megközelítések: az észt ige, aspektus és eset kölcsönhatásai. In: Lutsar K, Pusztay J, editors. Észt-magyar összevetés. Folia Estonica (1 - 14).: Savaria University Press. Vol VIII. Szombathely: Savaria University Press.; 2011. p. 125-49. ( Folia Estonica; vol VIII).
Tamm A. Uralic case: from space to time and beyond. Vol 6. Csúcs S, Falk N, Tóth V, Zaicz G, editors. Piliscsaba: Reguly Társaság; 2011. (Dissertationes symposiorum ad linguisticam; vol 6).

Cross-categorial spatial case in the Finnic non-finite system: focus on the absentive TAM semantics and pragmatics of the Estonian inessive m-formative non-finites

The article discusses cross-categorial case—case in a wider perspective than an argument-marking phenomenon in the nominal domain. Cross-categorial cases mark bare stems, non-finite verbs, or inflected verbs; they serve as clause-linkers and encode tense, aspect, negation, and mood. The spatiotemporal parallel semantics of cross-categorial spatial case compares well to combinations of adpositions with infinitives, as in I go to swim, je viens de manger, er ist am Arbeiten. The article aims at overcoming the various obstacles that have prevented gaining insight into the meaning of rich systems of cross-categorial case in exotic or dead languages by focusing on the more accessible Uralic languages, many of which have more than 20 cases, and rich and regular cross-categorial case systems. The data are collected and tested by various methods and sources from several sociolinguistic and diachronic varieties of smaller Finnic languages: Karelian, Ingrian, Veps, Votic, Livonian, South-Estonian and Estonian. Especially in the Finnic spatial case systems, the correspondences between the nominal and verbal paradigms are regular. The case paradigms of Finnic non-finites derive from paradigms of deverbal nominalizations. Therefore, Finnic allows a unique opportunity to examine a wide variety of in-between categories combining nominal and verbal syntactic and semantic properties within the context of a rich case system; a detailed example is provided about the inessive (internal locative) case m-formative non-finite. Strengthened by areal contact, the bleaching of the inessive meaning motivates the cross-linguistically attested and cognitively grounded development of locatives into progressive entailment in the verbal domain, while the use of locative expressions in discourse has lead to the development of a cluster of implicatures and presuppositions that correspond to the European absentives. Solving the puzzle about the existence of “the Estonian absentive” provides a synchronic snapshot of a gradually developing nominal-verbal split in the in-between categories with inessive. The regular parallels between the Finnic nominal and verbal case call for a broader view on case than previously assumed, and this study goes beyond the well-known nominal semantics of cases.

Scalarity and Dimensionality across Categories. Estonian Pseudopartitive Constructions

Estonian abstract nouns diverge in their morphosyntactic properties, and this reveals several new facts about the semantic structure of nouns and the nature of derivation. Although temperatuur ’temperature’ and soe ’warmth, heat’ are synonyms in Estonian, ’warmth’ can appear in pseudopartitive constructions (PPC), while ’temperature’ cannot (viis kraadi sooja/#temperatuuri ’five degrees above zero’). The article shows that the morphosyntactic behavior is sensitive to the lexical semantic meaning, namely, scalarity. If the lexical meaning of the noun encodes a degree along a dimension, then it can appear in a PPC (’warmth’). If the lexical meaning does not encode a degree, then the PPC is not possible (’temperature’). The degree structure of ’warmth’ is a feature derived from adjectives, an option unavailable for the noun ’temperature’.

On some modal similarities between the -vat infinitival predicates and the partitive objects

This paper discusses the Estonian vat-form, which has both nominal and verbal properties. Its use as the main predicate of the sentence speaks for an analysis as a verb. However, the restrictions on agreement – the inability to inflect for person and number – indicate the historical infinitival or nominal character of that form. Facts about the historical origins as a partitive-marked personal or impersonal present participle, which still retains some of its partitive semantics as discussed in this paper, lead to a historical nominal analys2is of the vat-form. The paper studies the relation between aspect and evidentiality and wishes to point out that the phenomena related to the vat-form sentences show that partitive marking has pervaded the Estonian Tense-Aspect-Mood system more deeply than previously assumed. The partitive marking on predicates and objects is a unique phenomenon that allows us to understand the relations between the mood, modality, and aspect categories from a new angle.

Tamm A. Prosody and types of negation in some Uralic languages. Congressus XI Internationalis Fenno-Ugristarum: Piliscsaba, Hungary, 2010. (Eds.)Sándor Csúcs, Nóra Falk, Viktória Püspök, Viktória Tóth, Gábor Zaicz. Piliscsaba: Reguly Társaság, 2010. 2010:203-4.
Tamm A. Cross-categorial case in the Finnic non-finite verbs. Congressus XI Internationalis Fenno-Ugristarum: Piliscsaba, Hungary, 2010. (Eds.)Sándor Csúcs, Nóra Falk, Viktória Püspök, Viktória Tóth, Gábor Zaicz. Piliscsaba: Reguly Társaság, 2010. 2010:215-8.
Tamm A. Specificity, aspect and the two partitives of Estonian. Vol 6/3. Budapest: Research Institute for Linguistics, Hungarian Academy of Sciences; 2009. (Working Papers in the Theory of Grammar; vol 6/3; no Research Institute for Linguistics, Hungarian Academy of Sciences ).
Tamm A. Igeosztályok, eseménytípusok, bizonyossági fokozatok: evidencialitás és aspektus közös szemantikájának kialakulása az észt nyelvben. Studia Linguistica. Tanár és tanítvány. Írások Györke József és Hajdú Péter tiszteletére 2002-2007. Budapest-Pécs: PTE BTK. 2009:242-52.
Tamm A. Estonian object and adverbial case with verbs of motion (comparative version). In: Pusztay J, editor. Észt-magyar összevetés V. Szombathely: Savaria University Press; 2007. p. 126-42. (Folia Estonica).
Tamm A. Estonian object and adverbial case with verbs of motion. In: Csepregi M, Masonen V, editors. Grammatika és kontextus. Új szempontok az uráli nyelvek kutatásában. Vol ELTE BTK Finnugor Tanszék, Urálisztikai tanulmányok. Budapest: ELTE BTK Finnugor Tanszék; 2007. p. 319-30. (Urálisztikai Tanulmányok; vol ELTE BTK Finnugor Tanszék, Urálisztikai tanulmányok; no 17).

Számítógépes összehasonlító szövegelemzés ügyfélszolgálati tájékoztatók legfontosabb prozódiai elemeinek a meghatározására

A szövegelemzéssel történő hangsúlykijelölés bonyolult feladat. Jelenleg nincs olyan elemző algoritmus, amelyik gépi úton képes a magyar mondatokban a hangsúlyok kijelölésére. Alapvető célunk, hogy az eddig elért és hozzáférhető elméleti nyelvészeti eredményeket, valamint kész mondatelemző algoritmusokat egyetlen, jól körülhatárolható struktúrált számítógépes rendszerré fejlesszük tovább automatikus hangsúlykijelölési kísérletek végzése céljából. Ebben a tanulmányban munkánk végeredményét, egy futtatható programrendszert (elemző) mutatjuk be. Az elemző bemenetére a mondat szöveges formája kerül, majd a feldolgozás során a mondat minden szavát egy hangsúlycímkével látja el. A feldolgozást két szinten végezzük: (i) tagmondatokra bontás, (ii) a hangsúly kijelölése tagmondatonként. Öt hangsúlykategóriát definiáltunk az elemzéshez: (F)=erős hangsúly, (E)=kiemelt, (W)=normál, (N)=hangsúlytalan és (-)= erősen hangsúlytalan (redukált) címkék. Az elemző 12 modulból épül fel, melyben mindegyik modul azonos koncepcióra épül. Az elemzőt egy szűk témakört leíró szöveges állományra fejlesztettük. A hangsúlykijelölés hatásfoka összességében 85%.

Tamm A. Estonian transitive verbs and object case. Butt M, Holloway TK, editors. Universität Konstanz; 2007. (CSLI Publications; no Stanford: CSLI Publications).

Estonian transitive verbs and object case

This article discusses the nature of Estonian aspect and case, proposing an analysis of Estonian verbal aspect, aspectual case, and clausal aspect. The focus is on the interaction of transitive telic verbs (write, win) and aspectual case at the level of the functional structure. The main discussion concerns the relationships between aspect and the object case alternation. The data set comprises Estonian transitive verbs with variable and invariant aspect and shows that clausal aspect ultimately depends on the object case. The objects of Estonian transitive verbs in active affirmative indicative clauses are marked with the partitive or the total case; the latter is also known as the accusative and the morphological genitive or nominative. The article presents a unification-based approach in LFG: the aspectual features of verbs and case are unified in the functional structure. The lexical entries for transitive verbs are provided with valued or unvalued aspectual features in the lexicon. If the verb fully determines sentential aspect, then the aspectual feature is valued in the functional specifications of the lexical entry of the verb; this is realized in the form of defining equations. If the aspect of the verb is variable, the entry’s functional specifications have the form of existential constraints. As sentential aspect is fully determined by the total case, the functional specifications of the lexical entry of the total case are in the form of defining equations. The general ell-formedness conditions on functional structures secure the sensitivity of aspectual case to verb classification.

Tamm A. Representing achievements from Estonian transitive sentences. Proceedings of the Workshop Semantic Content Acquisition and Representation (SCAR) 2007: (Eds.)Magnus Sahlgren, Ola Knutsson. Stockholm:, 2007, (SICS Technical Report T2007-06, Swedish Institute of Computer Science, Stockholm, Sweden),. 2007:28-35.

Perfectivity, telicity and Estonian verbs

This paper discusses telicity and perfectivity in Estonian. Neither of these categories corresponds exactly to the Estonian object case alternation, which is argued to reflect predicate or clause aspectual properties and not the NP-related properties of objects. The aim of this account is to accommodate the systematic compatibility of verb classes with certain clausal aspectual object case marking patterns. The paper proposes a way of understanding the interaction between verbal and clausal aspect in terms of boundedness and formalizes it in the framework of Lexical Functional Grammar (LFG). The features of verbs and case markers determine the clausal aspect from different constituents at the syntactic level of constituent structure. The information is unified at another syntactic level, functional structure.

Tamm A. Estisch. ABC van de CLVV. In: Libellum lexicographorum Willy Martin . Utrecht: Spectrum; 2006. p. 23-5.
Tamm A. On some modal similarities between the -vat infinitival predicates and the partitive objects. In Oszkó Beatrix, Mária Sipos (eds.). Online Proceedings of Word Categories (SzóOSZTÁLYtalálkozó), BUM 5, 2006. MTA Nyelvtudományi Intézet. . 2006.

Kísérlet automatizált szövegelemzési módszerek kialakítására a szóhangsúlyok meghatározásához

Az automatikus szövegelemzés bonyolult kérdéskörébl egy résztéma vizsgálatát tztük ki célul, nevezetesen a hangsúly-kategóriák szavankénti kijelölését meghatározott mondatokban. Az eredményeket a gépi beszédszintézis prozódiai támogatáshoz tervezzük felhasználni. A hangsúly-kategóriákat úgynevezett címkékkel jelöljük a szó eltt a szövegben. A célkitzést két irányból közelítjük: A klasszikus módszernél a címkéket nyelvészeti mondatelemzés eredményébl nyerjük. A másik eljárás lényege nem nyelvészeti központú, hanem egyfajta egyszer felszíni szövegelemzés, melyben nem használunk nyelvészeti módszereket, csupán szólistákat, táblázatokat, egyszer szabályokat. Mindkét elemzési formánál alapkövetelmény az algoritmizálhatóság. Az elemzésekhez ugyanazokat a hangsúly-kategóriákat használjuk, így mód nyílik arra, hogy közvetlenül összehasonlíthassuk a nyelvészeti elemzés eredményét a nem nyelvészeti központú eljárásból kapott hangsúly jelölésekkel. Rávilágítunk mindkét elemzésnél, hogy mely problémák miatt nem kaphatunk teljes érték eredményt sok esetben.

On the grammaticalization of the Estonian perfective particles

This article presents two grammaticalization types of separable perfective particles in Estonian. The bounding particle (BP) use of the semantically most bleached Estonian separable verbal particle ära is demonstrated to display a distinct status in the grammaticalization of aspect. The special character of the BP is illustrated by contrasting the BP ära, on the one hand, with the well-established use of the particle with the same form (ära) and, on the other hand, several other perfective particles, which are referred to here as instances of completive particles (CP). The study of differences brings out the following novel facts about the BP. The particle occurs only in context-dependent, agentive sentences that typically describe the achievement of planned or foreseen endpoints of events. The perfective sentences with the BP contain primarily activity verbs; however, verbs in such sentences can belong to all aspectual classes, and the combinations are transparent. The occurrence of an argument that would serve as a “measure” for the event is thereby optional. The evidence that the BP and verb combinations do not allow for the derivation of adjectives (participles) also demonstrates the difference between the principles of combining a verb and a BP as opposed to a CP.

Estonian transitive verb classes, object case, and progressive

This article examines the relation between aspect and object case in Estonian and establishes a verb classification that predicts many facets of object case behavior. It is demonstrated that the aspectual opposition between perfectivity and imperfectivity correspond to the morphological opposition between genitive/nominative case marking and partitive object case marking. However, case marking of Estonian objects is shown to be an unreliable indicator for aspectual verb class membership. The verb classification proposed here is established on the basis of tests that involve only the partitive object case. These tests employ the Estonian progressive. The tests distinguish verb classes from each other according to the situations they typically describe and predict several conditions of case assignment of patterns.

Reversing the Dutch-Estonian Dictionary to Estonian-Dutch

This contribution evaluates some aspects ofthe reversing ofthe Dutch-Estonian electronic bilingual dictionary database to Estonian-Dutch. The project has linked two monolingual lexical databases and added new lexical and example units with the editor tool OMBI. The links are provided with information about the status of equivalence. The two sources are the Dutch Reference File and an Estonian database ofpolysemous words. The strategies ofderiving correct polysemy representations ofthe Estonian items in the course ofediting the Dutch- Estonian dictionary are be evaluated. Prior to dictionary editing, an Estonian reference file for polysemous words was created. In the course of editing, many missing entries and senses were added. The Estonian reference file consists of three structurally different parts: first, the left side of another bilingual dictionary, second, a database ofamonolingual dictionary, third, a part created specially for the database. It is argued that the high quality ofthe target language database and a correct specification ofthe equivalence information are crucial for successful reversing. Verbal polysemy and its relation to the Estonian object case have posed a major challenge for the project.

Tamm A. Polysemy and cross-linguistic equivalence. Vol 7/1. Budapest: Research Institute for Linguistics, Hungarian Academy of Sciences; 2000. (Working Papers in the Theory of Grammar; vol 7/1; no Budapest: Research Institute for Linguistics, Hungarian Academy of Sciences).