The View from CiCewa (Bantu): Towards a Cartography of Passive Voice(s) and Developing Terraling

Abstract

In this talk, I will establish a cartography of Passive (and passive-like) Voice(s) for Chichewa (Bantu), use it to explore the observed crosslinguistic variation in passive constructions across languages, and implement the findings in a (pilot) comparative Terraling dataset for passive constructions that can serve as the basis for further current or future theoretical investigations.

Like many Bantu languages, Chichewa has extensive verbal morphology, with both a postverbal applicative morpheme and two passive(-like) morphemes (Alsina & Mchombo (1990) [1], Alsina (1999) [2], Dubinsky & Simango (1996) [3]). The applicative morpheme -ir adds a variety of arguments/adjuncts to the verbal stem, not only goals or benefactives, but also instruments, or locative adjuncts. The stative morpheme (-ik) occurs in middles and potentials. The passive morpheme -idw occurs not only with canonical (agent-theme) and low oblique applicative passives, but also with non-canonical passive applicative constructions (instrumental, locative, ...). Applicative and statives/passive can co-ocur in different orders V-stat-Appl, but V-Appl-Pass, or V Pass Appl the latter restricted to Instrumental and locative applicatives.

Various interactions determine the cartography of Passive-(like) Voice(s) in Chichewa. Passive Voice(s) and Applicative interact with the argument structure ("little" vP) and the hierarchy of "obliques"/ PPs/applicatives (Schweikert (2005) [4]; morpheme orders reflect the syntactic derivation and give insight into the order of operations; stative and passives occur in different regions in Cinque's functional hierarchy (Cinque (1999) [5]) etc. Taken together, this leads to a general cartography of Passive Voice(s), and the conclusion that not only stative, but also passive voice(s) in Chichewa can be merged at different heights in the hierarchical structure. Because Passive can be merged above Locative, Locative can be passivized. Because Passive can be merged below APPL, passive voice can be merged below Instrumental and Locative, and only themes can be passivized in this order. This investigation provides important insights into some fundamental theoretical questions. What is the element that introduces the external argument a (light)verb: little v? ("yes") or (Kratzerian) Voice ("no"), and where is it located (above locative adjuncts or below) ("both are attested")? What are the formal properties of Passive Voice: is (Passive) Voice ever involved in argument structure (assigning or reducing argument structure?). Or is it simply a vehicle, devoid of any semantic properties, that serves to map arguments on their syntactic positions ? We briefly explore how variation in height of Merge further leads to a new typology of passive-like constructions crosslinguistically, and show how to concretely implement the findings in the open crosslinguistic Terraling-dataset for passive-like constructions https://terraling.com/groups/13.

References

[1] Alsina, Alex & Sam Mchombo. 1990. The syntax of applicatives in Chichewa: problems for a theta theoretic asymmetry. Natural Language and Linguistic Theory 8(4). 493-506.

[2] Alsina, Alex. 1999. Where's the Mirror Principle? The Linguistic Review 16(1). 1-42.

[3] Dubinsky, Stanley & Silvester Ron Simango. 1996. Passive and Stative in Chichewa: Evidence for Modular Distinctions in Grammar. Language 72(4). 749-793.

[4] Schweikert, Walter. 2005. The order of prepositional phrases in the structure of the clause, vol. 83. John Benjamins Publishing.

[5] Cinque, Guglielmo. 1999. Adverbs and functional heads: a cross-linguistic perspective. New York: Oxford University Press.