Diathesis

La raison principale des confusions notionnelles, évoquées dans le titre, se cache, selon l’avis de l’auteur, derrière les critères définitoires non homogènes à la base desquels on définit au départ toute catégorie grammaticale. L’analyse pertinente de l’aspect du verbe pronominal est entravée par une (con)fusion entre l’aspect syntaxique et l’aspect sémantique (Aktionsart) du même verbe, ainsi qu’à la «mise en circulation» de certaines paires de notions figurant chez plusieurs
auteurs comme synonymes et comme notions éloignées chez d’autres.

Questo secondo volume della collana CLUB Working Papers in Linguistics raccoglie alcuni dei contributi presentati durante il secondo anno di attività del CLUB – Circolo Linguistico dell’Università di Bologna (a.a. 2016/2017). Il volume contiene dieci saggi a firma di Fabio Ardolino (vincitore del premio CLUB ‘Una tesi in linguistica’ per l’anno 2017), Ilaria Fiorentini, Giuliana Fiorentino, Chiara Gianollo, Eugenio Goria, Elisabetta Jezek, Alberto Manco, Caterina Mauri, Francesco Olivucci, Andrea Sansò.

The topic of this PhD thesis is a classification and description of one-argument intransitive verbs in Georgian in a typological perspective. The analysis consists of two steps: firstly, the classification of the various types of verbs and secondly, a critical investigation of the complexity of the system under study and on the concept of intransitivity in general. The proposed classification begins by separating verbs not derived from other verbs (‘primary intransitives’) from those whose morphology indicates a process of de-transitive derivation (‘derived intransitives’). Primary one-argument verbs, in turn, are divided into two classes, each consisting of verbs with the same morphological, as well as semantic properties: telic verbs (such as “to die”) and atelic verbs (e.g. “to live”). Two-argument verbs are briefly described according to a transitivity scale. Discussed next are derived one-argument verbs which may have up to four readings: passive, decausative, autocausative and antipassive, and finally, verb pairs whose members, one transitive and one intransitive, are both morphologically marked. One-argument intransitivity in Georgian has not yet been investigated thoroughly; this particularly applies to some of its aspects such as the antipassive, of which a detailed study is offered here. The present work is intended to contribute to Kartvelian studies and at the same time, to recent discussion in general linguistics on the concepts of transitivity, valency and voice. Indeed, owing to its large number of verb classes and the close connection between morphosyntax and semantics, Georgian offers excellent data for the study of these concepts.

The causative diathesis –its semantic nature and its processes of formal expression– has been one of the most explored areas in the field of linguistics in recent decades from various disciplines and theoretical frameworks, as well as in numerous languages. In this sense, Latin has been an exception. Proof of the little attention that Latin causativity has claimed to date is the fact that most Latin syntax manuals, both classic and current ones, do not devote a chapter –and, in some cases, not even a page– to the treatment of this question.
The reason for this is that, unlike other languages, Latin does not have a regular morphological procedure for the expression of causativity. This diathesis is expressed in Latin preferably by means of noun‐verbal collocations of the type aliquem morte afficere (literally «affect with death», «kill») or aliquem alicui honorem dare («give someone an honor», «honor»).
The study of the collocations and, more particularly, of the support verbs constractions has also demanded a wide attention both in modern languages and, lately, also in Latin. The object of study of this Thesis is not the most prototypical support verbs, that is, those that develop the argument and event structure of the name without modifying them, but the extensions, and in a very special way the causative extensions.

The se-constructions are used in various functions in Macedonian, similar to those in the other Slavic languages as well as in a number of other Indo-European and other types of languages. The particle se, which derives from the direct object clitic of the reflexive pronoun exhibits a gradual increase in grammaticalization along a continuum of mutually related meanings (Mitkovska 1995). Grammaticalization reaches the highest degree in those constructions which express agent defocusing situations. The first part of this paper discusses the formal and functional characteristic of the se-passive construction in Macedonian.
In the second part reflexive passive constructions in Serbian and Bulgarian are examined. In both languages there are also personal and impersonal forms with similar functional properties. All four types of identification of the demoted agent are also present. This is not surprising, since the three languages are genetically related and in close contact. The purpose of my comparison was to see if some language specific features have produced diversities and in what direction.

Seit Beginn der Erforschung der indogermanischen Sprachen war die eigentümliche Verteilung der Morpheme des Nominativs und Akkusativs auf die einzelnen Stammklassen und Genera Gegenstand vielzähliger Untersuchungen, Spekulationen und Theorien. Bereits 1901 rekonstruierte C. C. Uhlenbeck ein System mit "Agens"- und "Patiens"-Funktion, worauf die spätere (und mittlerweile überholte) "Ergativ-Hypothese" aufbaute. Mit Erforschung der sogenannten "Aktiv"-Sprachen Kaukasiens und Nordamerikas in der zweiten Hälfte des 20. Jahrhunderts wurden naturgemäß Versuche angestellt, die indogermanische Kasusgrammatik auf ein solches System zurückzuführen. Hier machte sich besonders W. P. Lehmann einen Namen, dessen Theorien auf den Ideen von G. A. Klimov und andere namhaften Russischen Linguisten aufbauten.
In meiner Magisterarbeit habe ich untersucht, welche typologischen Strukturen für Aktiv-Sprachen kennzeichnend sind, ob diese sich im Urindogermanischen und in den indogermanischen Einzelsprachen wiederfinden lassen, und wie sich bestimmte typologische Phänomene des Indogermanischen aus Sicht einer Akkusativ-Typologie darstellen.
Es stellte sich dabei heraus, dass es für Akkusativ-Sprachen keineswegs üblich ist, so streng zwischen Subjekts- und Objektsfunktionen zu unterscheiden, wie man im frühen zwanzigsten Jahrhundert glaubte, und dass das indogermanische morphosyntaktische System sehr gut zu der Evidenz passt, die sich in anderen Akkusativ-Sprachen finden lässt.

This thesis examines the passive and the middle voice in Imperial Aramaic, a variety of Aramaic, a Northwest Semitic language, that has been attested for around 3000 years. Imperial Aramaic is in the process of developing from a voice system that distinguishes between active, passive and middle/reflexive/reciprocal voice to one that has different morphological categories for active, medio-passive, reflexive and reciprocal. On the one hand, this transformation of the voice system is apparent in the temporary diversity of morphosyntactic marking of the passive voice in Imperial Aramaic: While the internal passives are gradually disappearing, the t-stems and the impersonal construction increasingly denote passive voice. Combined with morphological restrictions of the stems and verbal forms, this results in complex morphosyntactic patterns used to denote the passive. On the other hand, the profound changes within the voice system result in the extremely high ambiguity of expression of the t-stems, which basically came to denote non-active voice in general. Thus, in Imperial Aramaic a situation presents where one function, the passive voice, is expressed by a number of different morphosyntactic forms, whereas one form, the t-stems, denotes a plurality of functions.

El presente estudio aborda el uso de verbos transitivos en activa intransitiva con un significado próximo a los de la voz media, construcción a la que se ha prestado poca atención en la gramática griega. Los datos han sido tomados de los nueve libros de Historias de Heródoto.

В докладе представлены результаты продолжающегося корпусного исследования пар переходных и возвратных глаголов эмоций (восхищать-восхищаться, обрадовать-обрадоваться). Рассматриваются три параметра, связанных со свойствами участников глаголов эмоций: онтологический класс Стимула, коммуникативный статус Стимула, лицо Экспериенцера. На основе различий между переходными и возвратными глаголами эмоций выводятся типичные дискурсивные функции глаголов эмоций в письменных (прежде всего художественных) и устных текстах (на материале НКРЯ). Типичные для переходных и возвратных глаголов эмоций контексты употребления позволяют предположить, что переходные глаголы эмоций используются скорее для передачи интеллектуальной оценки ситуации, а возвратные глаголы эмоций скорее используются для описания наблюдаемых со стороны объективных проявлений эмоций.

This article reviews the literature on the question of the periphrastic passive in Modern Standard Arabic, tests the hypotheses proposed by Larcher and Girod (1990) on a corpus of Egyptian press and examines this passive in terms of the relationships between categories of diathesis and aspect.

It is common for languages crosslinguistically to code a range of diathetic constructions with the same verb form (Kazenin 2001, Kemmer 1993, Geniušienė 1987). Macedonian, as well as most Slavic languages, employs the reflexive se-construction for that purpose. Since one and the same verb can be used in several types of diathesis, such reflexive verb constructions are polysemous, creating ambiguity and vagueness along the diathesis continuum. In this presentation we look at reflexive constructions which code a number of non-factive situations in which the agent of the predication coded in the main verb is suppressed to some degree. They display a gradience of modal interpretations: potential (1), normative (2), generalising (3).
(1) Svilata teško se    održuva.  Lit. ‘Silk maintains with difficulty.’
        silk-DEF difficult-ADV REFL maintain
(2) Ne se igra so ogan.  ‘You shouldn’t play with fire.’
      not  REFL  play with fire 
(3) Imeto ne se preveduva. ‘The name is not translated.’
      name-DEF not REFL translate
This type of constructions is known in the literature under various names, such as gnomic, quasi-passive, mediopassive or middle, among others. Here they are referred to as “pseudo-passive reflexive constructions’. We claim that the difference between the three types of meaning depends on the degree of implication of the agent in the construction: it is only indirectly evoked in the potential, its presence can be felt in the normative, and a non-referring agent is present in the generalising constructions. Consequently, overlapping between the subclasses is common. For example (3) can be also interpreted as potential (it is difficult to translate the name) or as normative (it is against the rules to translate names).
An electronic survey, testing native speakers’ interpretation, was conducted to check the role of the proposed factors in distinguishing the modal senses of the construction in particular contexts. The responses of 776 speakers confirm the basic assumptions about the conditions for the three basic types of modality in the Macedonian pseudo-passive constructions.

Languages expressing diathesis by morphological voice often extend the use of voice to modal meanings. Typical examples are modal passives and facilitatives. This paper deals with the modal uses of the "detransitive" voice in Georgian, using data from the author's own fieldwork. Georgian has a morphological detransitive in i-R-eb-a (R = root) which derives intransitive constructions from transitive ones. Its main values are passive. Its morphosyntactic features are well-studied,
especially its use as a passive, but its aspecto-modal uses lack thorough analysis. The paper argues that the passive use of the detransitive voice typically includes deontic ('shall') and potential meanings. Deontic meanings are typical for normative speech; the agent is dereferentialized and the construction expresses a general statement (e.g. 'Purchased items are not refundable'). In its potential meanings, the verb expresses whether the process can be realized or not (e.g. 'Does the window open?'). In addition, the passive has generic meanings and is used to express permanent virtual semantic properties of the patient (e.g. 'to be edible'). This third type of modality differs from the potential in that the agent is entirely removed semantically whereas in the former, the
ability of the agent to carry out the process is crucially implied. In conclusion, it will be suggested that the fact that both voice and modal categories are expressed by the same morphology has a decrease of semantic transitivity as its centre: the non-referentiality of one of the core arguments enables the sentence to be interpreted as atelic and/or irrealis.

(Cet article, soumis initialement en 2010, est à la base de la recherche ayant abouti à notre thèse doctorale "Résultatif, diathèse et possession en basque, vieux perse et élamite".
Certaines des vues exprimées ici sont donc dans un état moins avancé que dans la thèse.)
Le parfait périphrastique des verbes transitifs en vieux perse a fait l’objet de nombreuses études. Face à ce tour construit sur un participe passé passif avec un agent au génitif-datif, et qui a évolué, dans certaines langues iraniennes, en un prétérit actif de construction ergative, plusieurs auteurs ont exprimé des vues divergentes quant à la voix que cette forme était censée rendre. Le présent article commence par rappeler les limi-tations de nos connaissances sur la langue des souverains achéménides et inventorie les différents types d’occurrences dudit tour périphrastique. Il passe ensuite en revue une dizaine d’approches susceptibles de livrer des arguments en faveur du caractère actif ou passif du tour, tous arguments utilisés à divers degrés et dans divers buts par des auteurs reconnus. Cette étude montre que la plupart des arguments ne sont pas convaincants en soi, mais le vaste tour d’horizon effectué en les analysant permet de dégager plusieurs certitudes et directions de recherches futures.

Whether morphologically coded or not, the seemingly secondary passivity of many intransitive constructs is not always more “retarded” in view of agentivity or eventuality than the activity of their corresponding active-transitive ones. When a cake bakes well, for instance, it is possible that a given line of cakes is endowed with the capability of being baked with little or no difficulty on the part of the Agent. Without the previously licensed attribute of the food, few
cooks will find it easy to bake the cake adroitly. Likewise, we will have difficulty peeling an onion whose skin is stubborn. The onion which peels well allows you to peel it well. Very
often, the Theme precedes the Agent.

Key words: computational semantics, homogeneous function, medio-passive verb, middle verb.

Among the European languages with a participial passive, there are some that tend to distinguish dynamic passive and object resultative through the use of different auxiliaries, e.g. English, Italian or German. In German, the difference is obligatory, and the German difference between the werden-passive and the sein-resultative had an influence on several neighboring languages (Polish, Sorbian, Swiss Retoromansh, Hungarian). Only partly compatible with this area, there is another subareal in which a recipient passive is built by using verbs with the meaning ‘get’ as auxiliaries. Again, the origin has to be sought in German, but more or less grammaticalized recipient passives can be found in Sorbian (with the auxiliary borrowed in dialects and calqued in the standard), Czech, and Slovak (the auxiliary is a calque here). In the paper, corpus material from Upper Sorbian is compared to former descriptions of the recipient passive in Czech and Slovak, all by taking into account the situation in different varieties of German.
As a result of the comparison, it seems that there is stronger grammaticalization of the recipient passive in Sorbian dialects than in Czech, which is probably due to the fact that the Sorbian construction came into being earlier than the Czech one (evidence from 19th-century Upper Sorbian and Czech is given in the paper). While in Sorbian the construction perfectly resembles its German counterpart (including differences in grammaticalization and style between dialectal use and use in the standard language), in Czech there are some peculiarities in the choice of the full verbs used in the recipient passive that do not agree with either Standard German or East Middle German dialects. For these reasons, the process leading to the formation of the recipient passive can be analyzed as polysemy copying in Sorbian, but as replica grammaticalization in Czech. The recipient passive in Slovak seems to be dependent on Czech rather than on German.

In: Wiemer, B., Wälchli, B., Hansen B. (eds.): Grammatical replication and borrowability in language contact. Berlin – Boston (2012), 559-588. (Trends in linguistics. Studies and monographs 242)

L'articolo analizza l'uso del riflessivo all'interno di un manuale di economia in lingua tedesca edito nel 2015 in sedicesima edizione e adottato in molti corsi universitari germanofoni. Dopo alcune considerazioni generali su come la più autorevole linguistica contemporanea, sia tedesca che italiana, affronta la forma riflessiva, si delinea il campo di indagine: all'interno del corpus prescelto sono considerate tutte le forme verbali collegate al pronome riflessivo. Si passa poi alla disamina dei verbi utilizzati raggruppandoli a seconda della reggenza e delle caratteristiche semantiche. Queste ultime in particolare svelano l'opportunità di rimeditare la diatesi media del greco antico. La forma media infatti si rivela tuttora un'utile categoria per una migliore comprensione del riflessivo, che altrimenti appare piuttosto farraginoso e difficilmente riconducibile a un pensiero coerente. A partire da questi aspetti semantici, sono poi messi in luce alcuni nessi con il passivo e sono proposte alcune possibilità traduttive di forme le quali, a causa delle sottili sfumature rivelate dal riflessivo, sfuggono a criteri unitari e utilizzabili in modo costante. Parole chiave: diatesi riflessiva in tedesco, verbo riflessivo nei testi di economia, problemi traduttivi della forma riflessiva dal tedesco vs. italiano Abstract The article analyzes the use of reflexive in a German handbook for Economics published in 2015 (16. ed.) and adopted in many university courses. After some general considerations on how the most authoritative contemporary scholars, both German and Italian, consider the reflexive form, the field of investigation is clarified: all the verbal forms connected to the reflexive pronoun are considered. We then move on to the examination of the verbs actually used in the corpus, grouping them according to the regency and the semantic characteristics. The latter in particular reveal the opportunity to refer to the middle voice of the ancient Greek. In fact, the middle voice is still a useful category for a better understanding of the reflexive, which otherwise appears rather cumbersome and unable to be subsumed under a coherent thought. Starting from these semantic aspects, some connections with the passive voice are highlighted and some translation possibilities are proposed which, due to the subtle nuances revealed by the reflective, escape uniform criteria of translation (especially automatic one).

In Czech and Sorbian, there is a special construction for the syntactic promotion of the recipient, called indirect or recipient passiv and marked by an auxiliar verb meaning „get“. It is supposed to be a calque from German. In Slovak no such construction has been noticed so far, but there is an evidence from the Slovak National Corpus that such construction exists as well in Slovak.

In the surface expression of agency within passive constructions, languages behave in different ways with regard to the humanness of the agent. Indo-European languages can mark the surface agent through a) 'agentive' morphemes (i. e. morphemes whose primary function is to indicate a high degree of control over the action expressed by the verb), and b) 'instrumental' morphemes (i. e. morphemes whose funetion in active contexts is to indicate an instrument). After an examination of the data offered by various lndo-European languages, a typology of agent noun phrases is proposed. Languages are divided into three groups: Type A languages mark human agents with a type
a) morpheme and non-hurnan agents with a type b) morpheme; Type B languages mark both human and non-human agents with a type a) morpheme; and Type C languages mark both human and non-human agents with a type b) morpheme. Languages of type A bring to the surface a semantic bipartition in the deep structure. Languages of type B give a major relevance to agentivity, while languages of type C stress the demotion of the deep structure agent from the syntactic role of subject.

In this paper our propose is to analyze the Latin collocation poena afficere, ‘to inflict punishment’, a type of collocation having both syntactic and semantic specificities that distinguish it from the prototypical ones: the predicative noun functions as third argument of the support verb, a syntactic scheme that is ideal, as we shall attempt to demonstrate, for the expression of causative predicates. From the examples of poena afficere documented in a large text corpus, we will intend to highlight the main features of this type of collocation. For its description and formalization, we will use the theoretical framework proposed by the Meaning-Text Theory.

ABSTRACT. One of the problems from which the grammatical term "ergativity" suffers derives from the relativity found in the grammaticality of "absolutives." In the sense that the object NP of a transitive VP is not omitted subsequent to the movement into the subject NP of an intransitive VP, the nominal item can be termed so. However, the grammatical argument of the subject NP "Jessie" as in "Jessie milks well," a semantically passivized construction from "Jesse milks Jessie well," is not an “absolute” one due to the change of the thematic role: from an agent to a patient. Thus, the ergative transition can be better understood when the term absolute is replaced by "relatives."

CONCRETE: Ergativity in the literature of linguistics and grammar denote a system of marking grammatical process by which intransitive subjects comes from transitive objects (“absolutive”) whose grammatical subject NP is an agent (“ergatives”). The problem is that the "absolute" NP is quite "relative" in the sense that it moves itself from an object NP to a subject NP. Given the instability of the so-called "absolute" NP in the context of syntax, the label needs to be transferred to the VP: the "absolute" VP. However, the valency of the "absolute" VP also changes. In other words, the morphological absoluteness is counter-balanced by the gradual change in the verbal energy in terms of transitivity. Whether it is called "absolute" or "relative," neither permits us to accept the binary polarity found in the grammaticalness. In terms of morphology, the NP and the VP almost never change. In view of semantics, however, they change significantly. Thus, the "reciprocity" might be a better alternative to the grammatically symmetric inter-changeability.

The Polish-no/-to construction is an arb, i.e. a human impersonal with a similar meaning as the impersonal pronouns man in German or on in French or the 3 impersonal in Russian. The common view that it can be formed from virtually all verbs as long as the referent is human is here contested by an acceptability judgement test. It shows that verbs assigning more agentivity features to the subject are significantly better than verbs assigning fewer agentivity features, and completely un-agentive verbs are just as bad as inanimate referents. This reveals a prominence relation. The details of the findings indicate that the-no/-to construction behaves in a similar but nonetheless different way from other arbs and that the list of agentivity features has to be revised in order to model the prominence relation exactly.

It is well-known that voice as a grammatical abstraction is one of the oldest verb categories that can be found in languages around the globe. Even in Egyptian (from 3200 BC) and Akkadian (from 3000 BC) one can find grammatical expressions of this abstract notion which signifies its coming before the separation of tense from aspect. One could also observe a reflection of classical models over time on Russian and Bulgarian grammatical systems in the 19th century. Closely related to the category of voice in Russian and Bulgarian is the division of verbs into transitive and intransitive.
The presented opposition is inherent to many language families but it can function and be expressed differently in various languages. In this study the opposition is not understood as a phenomenon of categorical nature – they are not categories but
oppositions based on different signs or properties. Additionally, a mention is made of the opinion held by some linguists that the opposition transitivity ~ intransitivity is justified by certain lexico-semantic word-forming and grammatical peculiarities
of verbs.
Voice as a grammatical category in Russian, just like case, occupies an intermediate position between morphology and syntax which makes it necessary for a wider investigation.

This study treats the Macedonian constructions with the reflexive marker se as various types of departure from the prototype transitive situation, i.e., as various types of diathesis. Separate functions exhibit different distribution of the participants and/or their constituents along the syntactic positions, depending on their communicative purpose. They are classified according to semantic criteria, depending on the role of the participant in subject position. From a synchronic point of view, all these constructions cover a broad semantic domain between transitive and intransitive situations on the vertical axis and between active and passive ones on the horizontal axis. The connections between the functions are established according to the properties they share and it is shown how they merge into one another, building a grammaticalisation chain.
The second part is devoted to analysis of the strategies that are employed in English for coding the same type of meanings. Even though the equivalents are represented by a variety of constructions, there are still some parallels, which confirm the hypothesis of conceptual affinity among the functions of the se-constructions.