Journal of Historical Linguistics - Volume 13, Issue 2, 2023

Prior to widespread contact with Russian, Khanty (Uralic; Finno-Ugric) did not have overt conjunctions or phrasal coordination. Instead, Khanty texts from the late 19th–early 20th centuries only include examples of conjunction-less clausal juxtaposition, which was used for both clausal and phrasal coordination. By comparing Khanty texts over the 20th century, we trace the emergence of overt conjunctions and coordination of phrasal constituents. We show that overt conjunctions first appeared in the context of clausal coordination, followed by coordination of smaller phrases. Based on novel elicitation data, we demonstrate that, in contemporary Khanty, (i) overt conjunctions are commonplace, (ii) coordinated clausal constituents may be derived via phrasal coordination or clausal coordination with conjunction reduction/ellipsis, but (iii) ellipsis of syntactic heads is banned (nouns & postpositions) or dispreferred (verbs). Based on this diachronic picture, we conclude that the coordination of phrasal constituents only emerged in Khanty once overt conjunctions became available. We derive this correlation from the Maximize On-line Processing principle (Hawkins 2004), and show that this maxim, usually invoked in the context of speech planning and production, can be successfully applied to modelling language change.

Contrastive secondary palatalization is a feature typically associated with Slavic. However, this contrast is present only in some contemporary Slavic languages, such as Ukrainian, Eastern Bulgarian, Russian, and Upper and Lower Sorbian. Thus, a question arises as to whether the secondary palatalization contrast represents a Common Slavic inheritance, and how it should be reconstructed. Providing such a reconstruction is important for the field of Slavic historical phonology, as well as for the general understanding of the development of consonant inventories with palatal consonants and the development of secondary palatalization contrasts in the world’s languages. By considering several historical scenarios, we show that /r/ : /r^j/ is the only secondary palatalization contrast that can be reconstructed to a pre-stage common to all of Slavic. While pursuing the reconstruction, we use supporting evidence from the typology of sound change and the typology of consonantal inventories in the world’s languages, as well as relative chronology.

The present study contributes to the description of the coding of anticausatives and causative-anticausative alternation in Classical Armenian based on conventional typologically-oriented questionnaires and a dataset of verbs attested in the Armenian translation of the Bible. The synchronic evidence is then analyzed from a diachronic perspective with an outlook on modern varieties of Armenian and the Proto-Indo-European reconstruction.

It is argued that discriminating between the derivational and inflectional tiers allows for a better explanation of the morphological links between the patterns used to code causative-anticausative alternation in verbs of different semantic types. According to preliminary quantitative estimations, the overall valency orientation of Classical Armenian is dominated by a nondirected equipollent coding strategy. Altogether, when coded by transitivizing pairs, causative-anticausative alternation shows preference for noncausal verbs of the a-conjugation, the paradigm of which typically combines equipollent perfective forms with labile imperfective ones.

An improved descriptive model enables a better cross-linguistic alignment of coding patterns as illustrated by comparing Classical Armenian to other cognate languages. Classical Armenian provides additional evidence on the typologically common split in the coding of causative-anticausative pairs. Whereas the transitivizing and equipollent (and marginally suppletive) patterns are diachronically stable within the prehistory of Armenian, the labile strategy constitutes a major innovation typical for the imperfective part of the verbal paradigm.

While semantic change is notoriously idiosyncratic, cross-linguistic evidence suggests some general trends in the directions and mechanisms of semantic shifts. Notable among these are trends applying to the target domain of the human body, a domain that has received considerable attention due to its universality. However, broad surveys of many languages risk missing significant details. Data from the Dictionnaire étymologique et cognitif des langues romanes (DECOLAR) on the etymologies of terms for 97 body parts in 14 Romance languages calls some proposed trends into question. In particular, counterexamples are found to a supposed unidirectional shift from visible parts to the wholes that include them. Analysis of individual changes reveals contextual factors that can cause a lexical trend to not apply. The findings contribute to a more complex model of metaphorization and metonymization, the primary processes involved in lexical semantic change.