Ti and ki in Pharasiot Greek

Metin Bağrıaçık¹ & Konstantinos Sampanis²

Boğaziçi University^{1, 2}

It has been noted since Dawkins (1916: 654) that in the Modern Greek dialect of Pharasa (hereafter PhG) a certain particle, ti (< Ancient Greek hóti; Dawkins 1916: 654; Andriotis 1948: 52 and Anastasiadis 1976: 259), may optionally be enclitic on verba dicendi or on verbs of mental communication when such verbs introduce quotes (Nicholas 1998: 290ff; Bağrıaçık 2018: 295–298). In this context another particle, ki (< the so-called Turkish complementizer ki), can also follow the Verb=ti complex, either immediately after it (1) or following another constituent of the reporting clause that comes after the Verb=ti complex (2).

(1) Ipen di ki “Allah, Panajia mu, ade to koridzi

said.3sg ti ki God Holy Virgin my this the girl

dhos ta a psiši”

give.imp.2sg 3obj a soul

‘He said “O Allah, O Holly Virgin, give this girl life!”’ (Dawkins 1916: 466)

(2) Tuz a ipo ti so vasilon ki tin gori su

how prt tell.1sg ti to.the king ki the daughter your

irevi ta a fidhi?

want.3sg 3obj a snake

‘how will I tell to the king “a snake is asking for your daughter?”’ (Theodoridis 1964: 306)

For (1)–(2), Nicholas (1998: 291) claims that “the quotative function of ti is so pervasive in [PhG], that syn[t]actically it no longer behaves as a complementiser, but has grammaticalised into a clitic to the linguistic verb, allowing the Turco-Persian complementiser ki to act as the actual quotative”.

We begin this presentation with two fundamental observations: (i) neither ti nor ki is obligatory with any verba dicendi or verbs of mental communication when there is a quote associated with these verbs, and (ii) ti is not attested with any other verb, e.g., verbs of emotion, even in cases where there is a quote associated with them. Based on (i) we claim that neither ti nor ki can be glosses as an “actual quotative” par excellence, and based on (ii) we claim that ti is not grammaticalized into a clitic just to any linguistic verb; but only to transitive verbs when they are associated with a quote. This leaves us with the question what the functions of ti and ki are in a quotative construction. We begin answering these questions by adopting the cross-linguistic observation that there is a close distributional affinity between quotes and bare nominals (de Vries 2008: 65). Following this, we claim that ti in its earlier attestations was a quasi-obligatory clitic resuming the quote. Data from heritage speakers of PhG reveal that ti has been replaced today by the third person resumptive object clitic ta Janse (1998: 539–540), which, unlike ti, is an obligatory resumptive clitic in this context. Returning to ki, we claim, following Bağrıaçık (2017), that it does not function as a quotative particle but rather as a discourse particle that signals the authority of the speaker on the narrating event. Furthermore, it appears with the same function not only in quotative constructions but in a number of apparently unrelated configurations. These observations urge us to conclude that there are no quotative [markers] in PhG — actual or not — contrary to the claim in Nicholas (1988).

In addition to the synchronic analysis, we examine ti with respect to the complementation system of the rest of Asia Minor Greek varieties and other Greek dialects and we attempt to provide an interpretation for the co-occurrence of the particles ti and ki in certain PhG contexts.

Selected references

de Vries, M. 2008. The representation of language within language: a syntactico-pragmatic typology of direct speech. Studia Linguistica 62: 39–77.

Nicholas, N. 1998. The story of pu: the grammaticalisation in space and time of a Modern Greek complementizer. PhD diss., University of Melbourne.