Volume 55, Issue 1

The paper provides evidence for the existence of endoclitics in Andi, a Nakh-Daghestanian language of the Avar-Andic branch spoken in the Republic of Daghestan, Russia. In Andi, the additive marker (‘also’) and the intensifying marker (‘even, at all’) behave as enclitics on various types of hosts and as endoclitics when they occur on negative verb forms. In the latter case, the additive and intensifying markers break up the word form and appear before the negation marker. I argue that both the additive and the intensifier are clitics, especially in view of their highly promiscuous attachment. I also show that negative verb forms are morphologically synthetic, so the additive and the intensifier are genuine endoclitics, i.e. clitics that occur inside morphological words. In addition I provide a few parallels for the unusual morphosyntactic behaviour of additive and intensifying clitics in some other Nakh-Daghestanian languages as well as in some languages of Northern Eurasia. Although in these cases the corresponding markers do not qualify as endoclitics proper, the available data hint at a cross-linguistic tendency towards word-internal placement of morphemes with meanings like ‘also’, ‘even’ or ‘only’.

Typological studies have tended to take for granted the default interpretation for English and imposed ‘simultaneity’ as the basic meaning of when -clauses for all languages. This in opposition to the approach taken in reference grammars, which generally report temporal linkage markers like when as encoding one or several meanings. Data from languages other than English show that comparative studies should also be open to the possibility that when -clauses do not always, or only, denote simultaneity. To support this claim and argue against the default interpretation of simultaneity, in this study I explore the range of temporal meanings of when -clauses across languages and provide evidence from Spanish and Yaqui corpora. Unlike English, corpus-based studies show that Spanish equivalent cuando -clauses equally introduce simultaneous and sequential readings, while Yaqui o-/kai -clauses predominantly express sequential meanings. Furthermore, a convenience sample of 28 unrelated languages reveals that, if there is a when -clause in a language, it can locate the event of the adverbial clause earlier, later, or around the same time as the main clause. The analysis of the semantic side of when -clauses demonstrates that there are language-specific tendencies regarding their temporal meanings. On these grounds, I propose that a better understanding of when -clauses can be arrived at by classifying them as ‘unspecific’ temporal clauses. This categorization would motivate a richer analysis of new data and a systematic comparison between unspecific, simultaneous and sequential clauses. Finally, I advance a four-way classification regarding general versus specific markers, and the temporal relations they encode, two of which account for most languages analyzed.

This paper describes the ablauting patterns in Siyuewu Khroskyabs, an understudied Gyalrongic language. Ablaut is only found in verbs containing closed syllables, and ablaut patterns in Siyuewu preserve Proto-Khroskyabs patterns relatively well. After providing a synchronic description of verb-stem functions and ablauting patterns, implicative entropy is used to model Siyuewu’s ablauting status. Entropy measurements reveal Siyuewu to have relatively low ablaut predictability, which may be suggestive of a historically conservative ablauting system. On the basis of this quantitative analysis, the current paper proposes an internal reconstruction of ablauting patterns, and postulates a series of velarised vowels for Proto-Siyuewu. Analogical changes are identified through comparison with other Gyalrongic languages. The reconstructed verb forms and patterns are then compared with neighbouring Gyalrongic languages and the directionality of analogy is discussed.

The paper provides a detailed examination of reflexive strategies in the Kuznetsovo dialect of Hill Mari (Mari, Uralic) filling in an existing gap in the description of anaphoric elements in Uralic languages. Firstly, I focus on simple lexical reflexives derived from the stem ( ə̈ ) škə̈- . Having examined their morphosyntactic and binding properties, I adopt several typological classifications and approach the Hill Mari data from a cross-linguistic perspective comparing them to anaphors in other Uralic languages. Secondly, I consider other reflexive strategies employed in Uralic languages, such as complex (reduplicative) reflexive pronouns and reflexive detransitivization of a predicate, and I demonstrate that these scenarios are unavailable in the variety of Hill Mari under discussion.

This paper investigates the phenomenon of ‘classificatory verbs’, i.e. a set of motion and positional verbs that show stem alternations depending on the semantic features of one of their arguments in Tłı̨chǫ Yatıì (Dogrib), based on field notes and documentary sources of the language. The paper shows that Tłı̨chǫ classificatory verbal categories belong to four semantic subclasses which have inconsistent stem inventories caused by the presence or absence of some semantic features. Stem inventories of locative verb systems vary depending on the scalar [ effort ] feature, and those of motion verbs correlate with the scalar [ agentive ] feature. The paper explains why other semantically related verbs do not show stem alternations and proposes contrastive hierarchies to represent variations in stem inventories intra- and cross-linguistically assuming that the selection of a stem for a particular semantic category follows a series of binary choices that characterize the opposition’s active in the language.

The goal of this paper is to revisit the so-called aspectual se , frequently cited over the last three decades as a new function of reflexive pronouns in Spanish and other languages, which refers to facultative uses of the reflexive pronoun where it has no effect on the valency or diathesis of the verb. I will focus on four empirical problems that such accounts face when dealing with corpus data: the requirement of a delimited object for transitive verbs, the semantic implications of the aspectual function of the reflexive pronoun, the unacceptability of the reflexive pronoun with some predicates, and the fact that these accounts have ignored a number of predicates that also take this facultative reflexive pronoun. I argue that a larger sample of both contexts and verbs, obtained by exhaustively analyzing corpus data, is necessary to improve our understanding of these uses.

The so-called ‘Big Mess Construction’ (BMC) frustrates standard assumptions about the structure of nominal groups. The normal position of an attributive adjective is after the determiner and before the noun, but in the BMC, the adjective precedes the determiner, e.g. that strange a sound , so big a scandal , too lame an excuse . Previous accounts of the BMC are couched in ‘Phrase Structure Grammar’ (PSG) and view the noun or the determiner (or the preposition of ) as the root/head of the BMC phrase. In contrast, the current approach, which is couched in a ‘Dependency Grammar’ (DG) model, argues that the adjective is in fact the root/head of the phrase. A number of insights point to the adjective as the root/head, the most important of which is the optional appearance of the preposition of , e.g. that strange of a sound , so big of a scandal , too lame of an excuse .