Converbs

This paper is a study of Korean converbs in relation to the neighboring languages of Northeast Asia. It is based on the descriptive and theoretical advances made in modern linguistics and provides an extensive analysis of the key issues regarding converbs in Korean as well as ten other neighboring languages. The thesis consists of five sections including an introduction and conclusion.
The introduction consists of a background of the converb and its use in typological settings. Chapter 1 is subdivided into three parts, the first of which describes exactly what forms can be considered converbs in Korean and relates them to their traditional description in Korean grammar. This is followed by a description of Korean converbs categorized into those containing or lacking temporal values. Chapter 1 concludes with a short section regarding the origins and evolutionary paths of converbs. Chapter 2 is a comparison of the Korea data above with converbs in neighboring languages. These languages can be roughly grouped into two sets, the Transeurasian languages and the Protoasiatic languages, neither a language family in the traditional sense. Chapter 3 consists of a discussion on several trends within the converbal systems on Northeast Asian languages. The paper ends with a short conclusion.
The Korean data is not atypical for the region. While it is true that Korean has a larger than average number of converbs, this is not unique to Korean for Nivkh contains nearly the same number. Korean’s lack of different-subject converbs and the paucity of posterior converbs is the overall trend for the region. It is also common for a language to have one highly contextual converb with a wide variety of possible interpretations.
Keywords: Korean, Converbs, Northeast Asia, Transeurasian, Paleoasiatic, Linguistic Typology

This work is a descriptive grammar of Denjongke, or Sikkimese Bhutia (also known as Lhoke or Sikkimese) (ISO 639-3 sip), an endangered Tibeto-Burman, Tibetic language spoken in the Indian state of Sikkim. The study is based on original fieldwork conducted over more than six years. The theoretical framework is functionalist-typological and may further be characterized as an application of Basic Linguistic Theory. The discussion is data-oriented and aims to describe Denjongke in its own terms.

The grammar begins with an introduction to the language and the people who speak it and continues with a description of phonology and an introduction to word classes, suffixes and clitics. The remainder of the grammar presents a mainly functionally-oriented description, starting with phrase-level syntax and proceeding through clausal syntax to discourse. The last chapter on vocabulary and the texts in the appendix provide a cultural window into Denjongke speakers' life.

Denjongke is an incipient tone language with 43 consonants and eight vowels. Nasalization and length are contrastive in vowels. The analysis establishes four major word classes (nouns, verbs, adjectives and adverbs) and eleven minor word classes. Many nouns and verbs have ordinary and honorific equivalents. Controlled vs. non-controlled verbs occur in phonologically related pairs. Unlike many Tibetic languages, Denjongke does not make a clusivity distinction in first person plural pronouns. There are five case-marking enclitics some of which may be stacked.

In syntax, the marking of A argument shows signs of both syntactic and pragmatic control. The marking of P argument is sensitive to animacy, specificity and affectedness. Denjongke has a rich array of copula forms, which mark three evidential values: personal, sensorial and neutral. The semantically oriented category “personal” differs from the more syntactically-oriented Lhasa Tibetan category “egophoric”.

Denjongke is a clause-chaining language and has a wide variety of adverbial clauses, which are expressed through various constructions, including ten converbs. Relative clauses are a subclass of genitive-marked constituent-modifying clauses. The last chapter addresses some interesting features of Denjongke lexicon, including ideophones, i.e. vividly descriptive words which are semantically, phonologically and morphologically distinct from other words.

Abstract
First, it is argued that, in the Second Scheme of the Emphatic Construction, the use of the Uninflected Relative Forms (traditionally “Nominal Verb Forms”) as initial circumstantial clauses is similar to the use of adverbial participles and adverbial relative clauses in other languages. Accordingly, the construction is identified as a detached adjectival verb form construction (here “Detached Relative Form Construction”, DRF-Cx), in which the Uninflected Relative Verb Form serves as a “converb”, i.e., a less inflected, adjectival verb form that is used adverbially. In a second line of thought, it is hypothesized that the Emphatic Construction proper (“First Scheme”) was born from a Verbal Balanced Sentence Construction with two identical Uninflected Relative Forms plus an additional adverbial phrase, in which the second ‘twin’ Uninflected Relative Forms was omitted due to its semantic redundancy. In contrast to earlier accounts, this scenario explains simultaneously a) the construction’s semantic layout, i.e., the sequence ground––focused foreground and the effectively absolute tense interpretation of the Uninflected Relative Form; b) its morphosyntactic layout, i.e., Uninflected Relative Form––adverbial phrase); and c) its paradigmatic fingerprint which is similar to that of Nominal Sentences. Altogether, this analysis takes notably 1) the Adverbial Sentence Construction, 2) the Detached Relative Form Construction (“Second Scheme of the Emphatic Construction”), and 3) the Emphatic Construction proper, alias “Circumstance Focusing Construction” (CF-Cx) as three different, unrelated constructions, –– the latter, i.e. the CF-Cx, however, being related to the Verbal Balanced Sentence, i.e., a Nominal Sentence.
Keywords: Emphatic Construction, Balanced Sentence, Detached Participle Construction, Absolute Participle Construction, Adverbial Participle Construction, Relative clause, Converb, Pragmatic focus,
Adverbial Sentence, Great Hymn to the Aten 12.
Table of Contents
§1 Assumptions
§2 Grammatical framework
§3 The First and Second Scheme of the Emphatic Construction
§4 The “Second Scheme” as a Detached, uninflected adjectival verb form construction
§5 The “First Scheme” as “Emphatic Construction” proper
§6 The Emphatic Construction proper as a truncated Balanced Sentence Construction
§7 Summary

Sri Lankan contact Malay (SLM) and Portuguese (SLP) share sprachbund-discordant features, including pre-verbal functional markers for TMA and negation. Yet their negation strategies also differ. In SLM, negation morphology is a diagnostic for the finiteness status of verbs. SLP verbs are contrastively negated, based on aspectual (not tense/finiteness) contrasts, and participles in adjunct clauses have distinctive non-finite negation. SLM marks finiteness status on matrix auxiliaries in a biclausal periphrastic construction. In the SLP construction, auxiliary and participle cannot be independently negated and the auxiliary cannot be separate from the verbal complex, arguing against biclausal status. SLM marks negative polarity in quantified nominal constituents, but has no negative concord, whereas SLP has negative concord, but relatively little negative polarity marking.

The study describes the morphology, syntax and semantics of non-finite verb forms in contemporary literary North Saami. According to the received view, there are about a dozen non-finites in North Saami, but the morphosyntax and semantics of these have not yet been investigated thoroughly. In the present study, these verb forms are re-evaluated from the functional-typological perspectives of general linguistics: they are further classified into the subcategories of infinitives, participles, converbs and action nominals.

A major portion of this study discusses adverbial non-finites or converbs and comparable action nominal constructions, the types of constructions in which many of the independent converbs originate. On the basis of abundant data on actual language use, it is concluded that contemporary North Saami possesses both old and more recently developed non-finites that have not received due attention in earlier descriptions of the Saami languages, whereas certain alleged non-finites are better analyzed as belonging to the lexicon rather than productive inflection. An important result of the study is that there are as many as six converbs and five participles in the language. The study provides a comprehensive description of the use and mutual relations of these verb forms.

In addition to the literary language and traditional dialects of North Saami, the study also addresses corresponding non-finites in the immediately neighboring Saami languages. Attention is also paid to the influence of the surrounding national languages (Norwegian, Swedish and Finnish) on the morphosyntax of the Saami languages, especially in translations.

The North Saami system of non-finites is undergoing a process of restructuring, and the rise of new non-finites through agglutination and analogy indicates that the Saami languages have developed not only in the direction of the fusional western European type of language, but recent developments also point back to an agglutinative type characteristic of the Uralic languages. Contemporary North Saami exhibits considerable morphological variation that may diminish in the future either as a result of unconscious change in the language or through conscious language planning.

(An unpublished paper from 2013 containing some opinions on the historical phonology of Turkic which I no longer entertain)
The ambiguous verbal suffix -zU occurs only twice in the whole Old Turkic corpus. In the introduction, it is concluded that it is a virtually obsolete causal converb.
Having analysed the sound change *ǰ > z in Turkic by means of various cases, *-r2U and *-ǰU are presented as two equally possible Proto-Turkic reconstructions of the converbial suffix -zU. Since internal evidence is indecisive in reconstructing -zU as *-r2U or *-ǰU, external evidence is taken into account.
The Mongolic converbs in -ǰi ~ -ǰU and -rUn are suggested as two possible cognates of -zU. Both are discussed in detail, and the Mongolic converb in -ǰi ~ -ǰU is ruled out due to its semantic discrepancy. It is concluded that Mongolic -rUn might be the cognate of Turkic -zU whereas Mongolic -ǰi ~ -ǰU is a borrowing of the Turkic converb in *-dI.

The Japanese -te form is often cited as a typical example of a converb. This master's thesis provides a summary of the most important definitions of the term 'converb' and a short investigation of the -te form and its classification as such.

German paper.

The Turkic language Kazakh has a remarkably large set of eighteen auxiliary verbs, which express various tense-aspect-modality (TAM) values in combination with a lexical verb. This thesis presents the first multivariate analysis of auxiliary verb constructions (AVCs) to precisely define the function and distribution of every auxiliary in Modern Spoken Kazakh.
Based on corpus data and first-hand fieldwork, I demonstrate that contemporary Kazakh has 28 AVCs, each with a distinctive function. I argue that some morphologically or syntactically ambiguous constructions, hitherto analyzed as AVCs, should be treated as lexical verbs in separate clauses. Previ-ously undescribed AVCs are described, including an unexpected case of stem alternation. The distribu-tional analyses demonstrate that some AVCs are sensitive to syntactic parameters, such as the independ-ence of the clause they head. Others are expressions of a purely semantic feature and thus are insensitive to syntax, such as the modal abilitative. Thus, AVCs are grouped into six classes in order to contrast distinctive distributional behavior with characteristics shared across constructions. The analysis assumes that auxiliaries are periphrastic and thus are part of the lexical verb’s paradigm. Therefore, alongside AVCs, synthetic TAM expressions are investigated and the results include precise descriptions and a novel contrastive analysis of three past tense expressions.
The description is complemented by an HPSG style analysis in order to present the system using a rigorous, feature-based approach. This is the first attempt to formalize a large auxiliary system with implemented solutions that lay the grounds for future work on the diachrony of auxiliaries. I propose novel semantic features that account for distinctions including boundedness, phase specification and focus. The main contribution is a systematic, synchronic, fine-grain examination of every Kazakh aux-iliary verb, which makes this complex system available for the general linguist, as well as specialists of TAM and periphrasis.

The Estonian des-construction is used as both an implicit-subject and an explicit-subject converb. This article concentrates on the subjects of both and also compares them. In the case of implicit-subject converbs, it is argued that not only the (semantic) subject of the superordinate clause can control the implicit subject of the des-converb, but also the most salient participant (which can sometimes even be the undergoer) of the superordinate clause. The article also discusses under which conditions the undergoer of the superordinate clause can control the implicit subject of the converb. In the case of explicit-subject converbs it is demonstrated which subjects tend to be explicitly present in the des-converb and which are the main properties of the structure and usage of explicit-subject des-converbs.

This paper discusses the syntactic properties of the general perfective converb in Rutul. It can head either an independent predication or a subordinate clause. Clausal relations tests applied to the converb suggest that, when combined with another clause, it may be either coordinate or subordinate. Furthermore, the general covnerb may either fall or not to fall within the scope of main clause illocutionary force and tense operators. I suggest that the converb may have different syntactic size in different constructions: it is a CP when used independently, but, when combined with another clause, it is either a CP or an AspP.

Essays in the History of Languages and Linguistics. Dedicated to Marek Stachowski on the occasion of his 60th birthday. Edited by M. Németh, B. Podolak, M. Urban. Kraków 2017. Pages 57–80.

This contribution focuses on the synchronic description and etymology of the Written Manchu word nofi ‘person (used after numbers higher than one)’. We conclude that nofi continues an analytic construction expressing collective numbers plus a converb and therefore historically should be better seen as a derivational suffix rather than as an autonomous lexical word, in spite of its synchronic status as measure word (numeral classifier) which is a result of Chinese influence.

The author takes up the “V1 conjunction wa V2” complex verb construction in Ainu, an endangered language spoken in north Japan that is genetically and typologically unrelated to Japanese. Assuming that the “V1 conjunction wa V2” construction corresponds to the “V -te V” complex predicate construction in Japanese, the author exploits diagnostic tests comparable to those applied for Japanese V -te V complexes to see whether the Ainu construction makes up one predicate or two, leading to nine  subtypes of the construction. By comparing Hokkaido Ainu with their Japanese V1 -te V2 counterparts, she suggests that six subtypes with V2s meaning EXIST, COME, GO, SEE, GIVE, PUT, but not DISAPPEAR, FINISH, or HEAR, are attested in both languages and speculates that some of the shared subtypes may have developed in Hokkaido Ainu under the influence of Japanese.

Статья излагает результаты исследования, посвященного параметрам выбора между финитными формами глагола и краткими причастиями в Новгородской первой летописи младшего извода. В статье сформулирован круг основных функций конструкции с кратким причастием каждого типа (настоящего и прошедшего времени в именительном и дательном падежах). Демонстрируется, что использование в летописи конструкций с причастиями во многих случаях обусловлено дискурсивными критериями второстепенности предикативной единицы (ее аспектуальной фоновостью, прагматической презумптивностью, стереотипностью сочетания ситуаций). Но о полном соответствии критериев выбора причастного способа оформления и критериев второстепенности предикации говорить нельзя. Для определенных глаголов причастные конструкции могли иметь особенные функции, однако появление некоторого количества причастных конструкций оказывается не мотивированным никакими прагматическими и вообще содержательными факторами. / The article expounds the results of an investigation dedicated to the parameters of the choice between finite verb forms and short participles in the Novgorod I chronicle (the younger recension). The article formulates a list of main functions of every type of short participle constructions (present and past tense participles in nominative and dative case constructions). The use of short participles in the chronicle is shown to be conditioned in most cases by discourse features typical of backgrounded predicative units (aspectual backgroundedness, pragmatic presupposition, stereotypical combination of situations). But there is no full correspondence between the criteria for choosing a participle construction and the criteria for the backgroundedness of the predicative unit. For some verbs participle constructions could have special functions, but the appearance of some participle constructions is found to be not motivated by any pragmatic and substantial factors at all.

In this talk I deal with converb clauses in Udmurt (Finno-Ugric, Permic). Converbs are non-finite verb forms whose main function is to express adverbial subordination (Haspelmath 1995). I will argue that converbs in Udmurt do not constitute a unified class syntactically and morphologically.
The discussion will touch upon a recent theoretical proposal on converb clauses (Weisser 2014) and several case studies of individual languages, such as Hungarian (Bartos 2009), Tsakhur (Kazenin & Testelets 2004) and Kazakh (Ótott-Kovács 2015). The language data come partly from own fieldwork as well as from descriptive grammars.
Weisser (2014) argues that all converb clauses are adjoined to the matrix Spec,vP. Given the variety of functions the Udmurt converbs display, I propose that they can have various attachment sites in the matrix clause (in line with Bartos 2009). The height of attachment correlates with several syntactic properties, such as the possibility of an overt subject, the choice of controller, and the (surface) position of the converb clause with respect to the matrix verb.

A morphological asymmetry is shared by certain Dravidian (and Finnic) languages. The phonological shape of a negation element is dependent on the finiteness of the verb it negates. Pragmatic factors are identified that could motivate the development of this shared asymmetry, using evidence from the grammar of a Dravidian-influenced contact language. I will show that contrastive finiteness marking (finite and non-finite morphology) can facilitate the development of pragmatically-motivated linear reordering of affirmative clauses and negated clauses in order to accommodate new information structure conventions, extending the contrast to negated verbs by expanding the functional range of a negative imperative marker. Radical contact languages resulting from collective adult second language acquisition in naturalistic social contexts are typically presumed to feature reduced functional morphology, in which only highly salient contrasts, such as temporal contrasts, are formally instantiated. If a formal finiteness contrast and other relatively marked properties ("complexity") could develop in a highly analytic contact language that did not previously have them, this suggests that such a sequence of changes is in fact as plausible among genetically-unrelated languages in a sprachbund as it is over longer periods of time in genetically-related languages. We can observe this by examining grammatical change in a language that previously lacked both a finiteness contrast and a corresponding negation asymmetry, but which developed both the contrast and the asymmetry as a result of contact with a genetically-unrelated language that has analogous properties. Sri Lankan Malay (SLM) has undergone grammatical change due to contact with Dravidian (primarily Sri Lankan Muslim Tamil). Several of these changes involve verb morphology and syntax, and are plausibly motivated by discourse-pragmatic triggers. Consideration of tense and (non-)finiteness phenomena, as well as their reflexes in SLM negation, suggests a discourse-pragmatic motivation for these changes. Two discourse processes could conspire to motivate the development of the new morphology in SLM. The first is a clausal asymmetry, in which the predicate representing the most recent event is ordinarily in focus, indicated by tense morphology and position of the clause relative to clauses referring to subsequent events. The second is the communicative need to reassign focus in certain contexts to a temporally non-primary clause, one referring to an event that did not take place first. In spite of a constraint in Dravidian languages blocking the marking of functional contrasts under negation (so that only a negation morpheme can be prefixed to the verb), negation morphology encodes an obligatory finiteness contrast, optimally supporting these information-structuring processes. The clause describing the most recent event in a sentence remains visibly finite under negation, when a temporally secondary clause is focused.
Linguistics and Language Behavior Abstracts (LLBA)

Las lenguas quechua (Sudamérica) despliegan un sistema de conmutación de la referencia (switch-reference) que cuenta generalmente con dos elementos: -pti (discontinuidad de la referencia) y -spa (continuidad) (van Gijn 2016; Dreidemie 2012). La literatura para
el quichua santiagueño también propone una forma -s, descripta como apócope de -spa (Albarracín 2016; Albarracín & Alderetes 2015, entre otros). El recorrido de este trabajo explora las propiedades semánticas y sintácticas de cada una de estas construcciones (-pti, -spa y -s). El objetivo de este estudio es argumentar que las formas en -s no forman
parte del sistema de CR (-spa/-pti), ilustrando la especificación funcional y características sintácticas diferenciadas de los sistemas de CR y los converbos: los últimos se especializan en relaciones estrechas entre eventos y son sintácticamente más reducidos que las construcciones de CR.

The aim of this paper is to analyse participles in Ancient Greek taking into account recent studies on the so-called converbs. In particular, the paper will focus on conjunct participle constructions. The first section shows the importance of new criteria for classifying these constructions (§ 1) and also offers an overview of the notion of converbs and of their
syntactic, semantic and pragmatic features (§ 2). The second section (§ 3) applies converb classification criteria to conjunct participles in Ancient Greek, with particular reference to present and aorist stems. This analysis leads to the identification of three different types of
conjunct participles, i.e. coordinative, clause modifier and predicate modifier.

This talk deals with the morphosyntactic evolution of the gerund in Late Latin. The morphosyntactic categorization of the gerund is determined on the basis of the external syntax of the clauses it heads. Distinguishing between nominal gerunds and adverbial gerunds, I show that towards the Romance development the gerund specializes as an adverbial gerund, an evolution which can best be described as a converbalization process. This evolution, however, is more nuanced than has been traditionally assumed: a closer look at the evolution concerned reveals that the converbalization of the gerund is to a large extend due to the functional competition between the nominal gerund and the emerging prepositional infinitive construction.

This paper concerns the converb forms expressing Simultaneity in Izhma Komi, Northern Khanty and Moksha belonging to the Finno-Ugric group of languages. The existing typological classifications of temporal relations and simultaneity relations in particular either are not detailed enough or lack rigorous criteria and thus appear not to be sufficient for understanding the usage and the distribution of temporal converbs. The study attempts at building a more detailed typological classification of Simultaneity functions which accounts for the data of the languages under consideration. The analyzed parameters of variation include the viewpoint aspect of the events, clause modification type, givenness of the conveyed information and the pragmatic type of the predicate. Special attention is paid to the discourse-pragmatic properties of the forms which bring new insights into the discussion.

В докладе предлагается и иллюстрируется понятие синтаксического дублирования. Оно применяется к случаям, когда глагол в подчинённой клаузе копирует форму глагола из  главной клаузы. При этом сама копируемая форма может быть разной, что отличает дублирование от стандартных случаев управления. Дублирование усматривается в конструкциях с фазовыми глаголами, трёхчастных конструкциях со временным придаточным и сравнительных конструкциях.

In Beja, a Cushitic language, adverbial subordination may be expressed by the use of converbs. In the first part, this work analyses the syntactic and semantic functions of the causal and sequential converbs, the second one being diachronically derived from the first one by the suffixation of a coordinating morpheme. Within the framework of enunciative theory, the second part studies the order change of converb clauses and uses an instrumental analysis of intonation patterns to demonstrate the articulation between syntax and prosody and the role of the latter in the interpretation of the enunciative, pragmatic and semantic values of the utterances.

The conditional forms of a verb marking action which caused other action expressed by the main predicate of the sentence are presented in many languages. Researchers state the contradictory points of view concerning semantical-grammatical features of these forms in the Caucasian languages. In article we tried to define the place of the studied forms in the system of the Chechen verb and also their role in hypotaxical constructions.
Principal and auxiliary means of expression of conditional semantics in the Chechen language are distinguished in the article. Thus, the most characteristic and frequent in use are the forms formed by the affixes -ħ and-ħāră. In addition to them the following markers of conditional meanings are presented in the Chechen: the interrogative form of the indicative mood, the present-tense form with the -q particle, and the converb formed by adding the -čă affix to one of the stem of the past.
Forms of verb ending with -ħ and -ħāră, which are the main condition markers in polyclausal conditional sentences, are presented in different points in the works of scientists. Thus, some authors specify them as moods, others consider them as adverbial forms of a verb. Besides, there are contradictory views concerning the very type of sentences in which the forms function. Some linguists consider clauses formed around such forms as subordinate ones in a complex sentences, another point of view is advanced by the opponents of the theory of the developed hypotaxis in the Chechen language, distinguishing the subordinate conditional clauses as conditional non-finite construction.
The analyzed forms have some specific characteristics distinguishing them among the other converbs in Chechen language. Thus, they have category of absolute time and the own subject different from the subject of the main predicate. Forms of verb ending with -ħ and -ħāră differ from each other mainly on the basis of reality or unreality of the marked condition. Forms of verb with a suffix -ħ function as a predicate of a subordinate clause in the conditional sentences where the main predicate is a verb in the form of a future tense or the imperative. Not the alternative reality, but real-conditional one is expressed in such sentences. Forms on -ħāră function as a predicate of a subordinate clause as a part of the subjunctive sentence, marking alternative reality in all tense structure, often the time for the sentence is defined by a predicate of a subordinate clause.
Analysis of the grammatical and semantic features of the conditional forms in Chechen language allowed us to draw some conclusions on this question.
In our opinion, the developed taxis and subject-object relationships within the clause, predicated by conditional form, and also the nature of the relationship of this clause with the main clause in the structure of the polyclausal construction give us a reason to define these constructions as complex sentences.
We believe that it is unreasonable to define the conditional forms as forms of mood, by means of that the special meaning of these forms can’t be considered as modal. A special function of the conditional form is the realization of the syntax connection between clauses as part of a complete conditional sentence. We haven’t detected any other semantic, grammatical and modal functions or meanings at the forms. That is, the special formants -ħ and -ħāră, forming conditional forms, are only the means of connection for parts of the sentence and on the grammatical meaning they are equivalent to the conjunctions.

Remarks on possessive-suffixed passive simultaneous converb forms in Finnish

The article provides a description of heretofore neglected Finnish converb clauses, in which the passive simultaneous converb in -taessa/-täessä ‘when V-ed; while being V-ed’ is followed by a possessive suffix, most often the third-person possessive suffix -Vn. Based on data drawn from a body of modern written Finnish, and accompanied by the results obtained from a questionnaire survey with native speakers, the article claims that person-marked passive converb forms like lausuttaessaan ‘when uttered’ and perustettaessaan ‘when established’ need not be considered downright erroneous contaminations by L2 learners or careless L1 writers. Instead, they are widely accepted in sentences where the possessive suffix refers to the patient of a transitive verb in the passive converb form, identical with the non-agentive theme argument of the main clause, as in Varasto oli perustettaessaan Euroopan suurin [storage be.PST.3SG establish.PASS.CVB.3 Europe.GEN large.SUP] ‘When established, the storage was the largest one in Europe’.

En retraçant l'évolution du gérondif et du participe présent du latin au français médiéval, je propose une nouvelle hypothèse sur leur rapport catégoriel en français. En particulier, j'argumente que dans le passage du latin au français les deux catégories ont fusionné dans une nouvelle catégorie, qui ne correspond ni au gérondif ni au participe présent latins (cf. De Smet 2014). Adoptant une perspective typologique et partant de la distinction entre catégorie descriptives et concepts comparatifs (Haspelmath 2010), je défend l'hypothèse que la catégorie émergente en français correspond à celle du converbe.