How to make Something out of Nothing:

Grammaticalization with Degrammaticalization in the Greek dialectal imperfect

Brian Joseph¹ & Angela Ralli²

Ohio State University¹ & University of Patras²

A standard interpretation of the grammaticalization “cline” sees the devolution of grammatical material such as affixes into ‘mere’ phonology, as with noun plural Umlaut in Germanic or Celtic grammatically relevant consonant mutations, as showing movement towards the endpoint of the cline (Hopper & Traugott 1993).

Since affixes often themselves result from movement along the cline from originally free words, the development of a phonological effect out of an earlier affix or even a once-free word can be taken as consistent with unidirectional movement in grammaticalization from forms with greater independence and less grammatical value to marking that has less independence and greater grammatical value (Haspelmath 2004). In a sense, such examples show the development along a grammatical cline essentially into nothing — the absence of an overt grammatical form, except for a feature of the phonological make-up of a word-form — out of something, i.e. out of a once-freestanding element.

There are cases, however, of the opposite sort of development, whereby something, i.e. an overt grammatical affix, has arisen out of nothing, i.e. out of a mere phonetic effect that has been reinterpreted as having an independent existence as a grammatical marker (Joseph 2011). A case in point is the emergence of the suffix ‑γ- in some dialectal varieties of Modern Greek — Peloponnesian, and some Cyclades and Sporades varieties (Pantelidis 2003) — as a marker of the imperfective aspect (Ralli 2005).

This suffix developed via the reinterpretation and spread of a phonetic glide, ‑j-, that arose originally in the 3sg imperfect tense between the stem vowel ‑a- and the personal ending ‑e, around the end of the 16^th century (Ηolton et al. 2019); thus fila- ‘kiss’ formed a past imperfective /fíla-e/ ‘s/he was kissing’, which was phonetically realized as [fíla‑j‑e] with the glide ‑j- phonetically inserted in the transition between the back vowel ‑a- and the front vowel ‑e. Since ‑j- before a front vowel in Modern Greek is the phonetic realization also of underlying /γ/, this sequence ‑ajewas reinterpreted as phonemically /‑aγe/ and from that, ‑γ- spread to personal forms where ‑j- was not phonetically justified, e.g. with the 1sg ending ‑a-, thus [fílaγa] ‘I was kissing’, from underlying /fíla‑a/. Thus, this grammatical suffix ‑γ- arose out of ‘mere’ phonology.

This development, therefore, shows hallmarks of “grammaticalization” in that, from a functional standpoint, it involves the emergence of material with grammatical function out of material that was less (or even non-) grammatical originally, but, at the same time, in terms of form, it shows counter-directional, degrammaticalizing, movement from a dependent phonological effect to a more independent affix. Thus, such instances represent cases of “degrammaticalizing grammaticalization”, a somewhat anomalous and unusual combination of traits.

These examples and the interpretation given to them demonstrate important points about Greek dialectology, about grammaticalization, and about the study of grammatical change. As for Modern Greek dialects, the developments with -γ- in the imperfect contrast with what occurred in other dialects: Northern Greek has ‑us- (e.g. aγapusa ‘I was loving’, Papadopoulos 1926), and Cappadocian shows inherited -isk, -an-/-in- or combined -inisk-, varying by area (e.g. Delmeso thoriska/thorina ‘I was seeing’, Fertek rotiniska ‘I was asking’, Dawkins 1916). These developments make the creation of imperfect ‑γ- all the more striking as they show that phonologically independent material was available that could have been used.

As for lessons about grammaticalization, we note that: i) grammatical material can originate in ways other than the downgrading of lexical items; ii) grammatical change does not unidirectionally move “down” the cline from less tightly bound to more tightly bound; movement “up” the cline is possible too; and iii) if we focus just on one type of movement involving grammatical material — from less to more dependent — we can miss interesting types of grammatical change.

References

Dawkins, R. 1916. Modern Greek in Asia Minor. Cambridge: Cambridge University Press.

Haspelmath, M. 2004. Directionality in language change, with special reference to unidirectionality in grammaticalization. In O. Fischer, M. Norde & H. Perridon (eds.), Up and down the Cline, 17–44. Amsterdam: Benjamins.

Holton, D., Toufexis, P., Manolessou, I., Horrocks, G., Lentari, S. & M. Janssen. 2019. The Cambridge Grammar of Medieval and Early Modern Greek. Cambridge: Cambridge University Press.

Hopper, P. & E. Traugott. 1993. Grammaticalization (1^st edn., 2^nd edn.: 2003). Cambridge: Cambridge University Press.

Joseph, B. D. 2011. Grammaticalization: A general critique. In H. Narrog & B. Heine (eds.), Oxford handbook of grammaticalization, 193–205. New York: Oxford University Press.

Pantelidis, N. 2003. The active imperfect of the verbs of the “2^nd conjugation” in the Peloponnesian varieties of Modern Greek. Journal of Greek Linguistics 4: 3-43.

Papadopoulos, Α. 1926. Γραμματική των βορείων ιδιωμάτων της νέας ελληνικής γλώσσης [Grammar of Northern Greek dialects]. Αthens: Glossiki en Athinais Etaireia.

Ralli, A. 2005. Μορφολογία [Morphology]. Athens: Patakis.