Cognate infinitive constructions as complex predicates in colloquial Arabic and in Central Europe

Colloquial varieties of Arabic are known to different types of complex predicates, covering typologically most of the notions proposed within different descriptive traditions (serial verbs, light verbs, auxiliary verbs and others), which themselves represent interesting descriptive and typological challenges (Brustad 2000; Ross 2021 etc.). Another phenomenon that has been observed in Arabic verbal morphosyntax is that of cognate infinitive constructions (CIC), i.e. constructions, in which a verbal form (mostly a finite one, but not necessarily) is accompanied by the infinitive (or the deverbative noun – masdar (MSD)) of the same verb, as in (1-3). The function of this construction has been described in different terms in Arabic linguistics, but recently has found a most thorough account in the seminal work of Iriarte Díez (2021 and 2022), which presents it as a device serving mostly information structure purposes. Interestingly, an isomorphic construction also exists in some languages of the Central European linguistic area (Czech, Slovak, Polish, Hungarian), as described by Lehečková & Januška (2020), illustrated in (4).
The purpose of this paper is threefold – first to approach the CIC in Arabic from the perspective of the study of complex predicates in the descriptive context of other multiverb structures peculiar to the Arabic varieties in question, second comparing it structurally and functionally to the CIC in Central European languages, and third setting it in the general typological context of complex predication. The constructional approach proposed by Lehečková & Januška (2020) serves as a useful framework for that. The fact that CIC has never been approached as an instance of complex predication is probably given by its features which set it obviously apart from other multiverb constructions, namely the fact that it is lexically the same verb that occurs twice in the construction and not a combination of different lexemes or roots. But I argue that when one wishes to explore the realm of complex predication in general, this construction needs to be treated as well, as it can syntactically hardly be accounted for in other terms. In fact, it represents an interesting type of complex predication in both formal and functional terms, as its function combines different aspects of semantics and pragmatics. Furthermore, the similarity of its behavior in Arabic and Central European languages implies its typological relevance.
The present paper presents original data gathered mostly by means of participant observation, mostly from Levantine varieties of Arabic (mostly Syrian and Palestinian), but also from North African varieties. Given the nature of the phenomenon, relying on such natural data is necessary, as it is difficult to obtain by elicitation and very difficult to search for in corpora. Nevertheless, the data is complemented by material obtained in targeted sessions with native consultants. It provides a detailed morphosyntactic characterisation of the construction and develops the account of its actual function and distribution given by semantic and pragmatic factors, arguing namely for its focalising/topicalising function.
The paper shows that however marginal such a structure might be typologically, it represents an important complex predicate type, valid both in descriptive terms and cross-linguistically as a comparative concept (Haspelmath 2010).

Examples
(1) Syrian Arabic
      əd-dāʕəš ḫtaraʕ-ū-hum             əḫtirāʕ.
      DEF-ISIS invent-3PL-3PL.OBJ invent.MSD
      ‘ISIS, they invented it.’
 

(2) Syrian Arabic
      hāda b-yənḥaṭṭ                        ḥaṭṭ.
      this   PRES-3SG.M-put.IMPF put.MSD
      ‘This is to be (simply) put there.’ (put focused)
 

(3) Syrian Arabic (Hauran)
      Hāy lazzag-ū-ha                  talzīg          mū   ḫiyāṭa.
      this stick-3PL-3SG.F.OBJ stick.MSD NEG sew.MSD
      ‘On this one (wound) they just put a plaster, it’s not stitched up.ʼ
 

(4) Czech (Lehečková & Januška, 2020, p. 419)
      Vad-it to asi nikomu ne-vad-í.
      bother-INF it probably noone.DAT NEG-bother-3SG
      ‘It probably doesn´t really bother anyone.’(bother topicalised)