written by Kitty Liu (University of Cambridge), recipient of the Philological Society’s Anna Morpurgo Davies Masters’ Bursary.

I am grateful to the Philological Society for their generous award of the Anna Morpurgo Davies Masters’ Bursary for the year 2023–24, which allowed me to research morphophonological change in Tibetic languages during my MPhil at the University of Cambridge. This report gives an overview of my research and presents some of my findings.

Tibetic languages are varied, fascinating, and understudied. Diversity within the modern Tibetic family is comparable to that of Romance (Tournadre & Suzuki 2023: 44) and the rich textual traditions and historical status of Old and Classical Tibetan roughly parallel those of Old and Classical Latin. I explored changes affecting the complex syllable structure and verb stem alternations of Old/Classical Tibetan, focusing on the conservative modern Amdo varieties (spoken in and around China’s Qinghai Province). All modern Tibetic varieties show syllable structutre reduction compared to Old/Classical Tibetan. For example, reflexes of bsgrubs ‘complete-PFV’ include /ɣɖʐəp/ (very conservative; Amdo Tibetan) and /˨ʈup/ (very innovative; Central Tibetan) (Bielmeier et al 2018). Old/Classical Tibetan verbs had up to four monosyllabic stem alternants, primarily expressing tense and mood, while modern varieties have up to three, expressing aspect and mood (Zeisler 2004). One stem has thus been lost and the remaining stems’ functions have been restructured during the development of verb paradigms. My thesis explored the intersections between these changes, demonstrating that phonological and morphological developments are often closely interwoven, with the study of one necessitating an understanding of the other.

Investigating these changes led me to incorporate different methodologies, including different approaches to sound change, paradigm-focused approaches in theoretical morphology, computational corpus research and computational corpus linguistics, as well as eliciation sessions over Zoom with a Tibetic speaker from the Amdo region. I really enjoyed getting to work with this wide array of methodologies, and my range of analytical approaches was reflected in the breadth of my findings.

Some aspects of my findings were specific to Tibetic linguistics (e.g. estimating the number of Classical Tibetan verbs in each verb class; developing a hypothesis for contrastive voicing for Old Tibetan nasals), but I also made a number of observations that are of interest for linguistics more generally. I briefly describe two of them here.

1. Amdo verb paradigms, both in pre-existing and my new data, demonstrate syncretism between any two of the three slots (imperfective, perfective, imperative) at the exclusion of the third. This means that Amdo verbs contravene the *ABA constraint in morphology, which postulates that in three-slot paradigms, there is one two-slot syncretism that is disallowed/-preferred (cf. Andersson 2018). All three combinations of two-slot syncretisms are synchronically prevalent in Amdo verbs, but analysis of sound changes and paradigm levelling show that imperfective-imperative syncretisms are derived from a more restricted set of diachronic processes than imperfective-perfective and perfective-imperative syncretisms. Imperfective-imperative syncretisms (where the perfective is not syncretic with them) appear to only arise due to phonological mergers, and never from the extension of the imperfective form to express an imperative meaning or vice versa. I hypothesise that this is due to the extent of semantic differences between imperfective and imperative stems. This finding adds nuance to existing discussions on *ABA in the morphology literature. 

• Old/Classical Tibetan verb paradigms are interesting for discussions of calculating paradigm complexity. The Old/Classical Tibetan verb paradigm is considered impossible under older measures of complexity such as Carstairs-McCarthy’s (1994) ‘No Blur Principle’, which involves counting the number of possible realisations for each paradigm slot, as, in Old/Classical Tibetan, each verb paradigm slot has many possible realisations and the realisations for different slots often overlap. However, under Ackerman and Malouf’s (2013) predictability-based quantification of paradigm complexity (‘conditional entropy’), Old/Classical Tibetan achieves an average score on par with paradigms that are possible under the No Blur Principle. Thus, Old/Classical Tibetan verbs lend support to the utility of predictability-based conceptualisations of paradigm complexity including conditional entropy.

Weaving together the topics and methodologies discussed above, my research process showed me the value of interdisciplinarity between different linguistic subfields in diachronic research. My thesis achieved a ‘Pass without corrections’, which is the highest grade for my course. I plan to prepare parts of it for conference presentations and for publication. I am currently a research assistant for the AHRC-funded ‘Emergence of Egophoricity’ project, assisting the construction of part-of-speech-tagged historical Tibetan corpora. I am also planning to apply for a PhD in the upcoming admissions cycle, to further pursue topics in historical Tibetic linguistics and develop the skills I gained in my MPhil.

References

Ackerman, F. & R. Malouf (2013). Morphological organization: The Low Conditional Entropy Conjecture. Language 89(3): 429–464. doi: https://doi.org/10.1353/lan.2013.0054

Andersson, S. (2018). (*)ABA in Germanic verbs. Glossa: a journal of general linguistics 3(1): 119. doi: https://doi.org/10.5334/gjgl.733

Bielmeier, R., G. L. van Driem & M. Volkart (2018). Comparative Dictionary of Tibetan Dialects (CDTD). Volume 2: Verbs. De Gruyter. https://doi.org/10.1515/9783110554076

Carstairs-McCarthy, A. (1987). Allomorphy in inflexion. Croom Helm.

Tournadre, N. & H. Suzuki (2023). The Tibetic Languages: An introduction to the family of languages derived from Old Tibetan. LACITO-Publications. https://lacito.cnrs.fr/the-tibetic-languages/