Kaliarda XVIII: Sechidou

Irini Sechidou’s recent paper¹ compares Kaliarda with three para-Romani languages of Greek—that is, mixed languages with some core Romani vocabulary, but which use the grammar of the gadjo language. (She does not use the term para-Romani in this paper, but she does elsewhere). The three languages are Dortika, the builders’ cant of Eurytania; the less well documented Romika of Epirus; and the still extant Romika of Finikas, Salonica, which she has worked on. (I’ve hesitated to call Dortika para-Romani, because its speakers no longer appear to have identified as Romani; but linguistically, it falls neatly into that category.)

The recurring conclusion in her paper (which Montoliu had anticipated going on just Kaliarda) is that Kaliarda is not like the others: it is not a para-Romani, but an artificial language which exploits Romani vocabulary for secrecy and language play.

Sechidou, truth be told, finds the para-Romani variants more interesting; I’m going to restrict myself to what she says about Kaliarda.

p. 561. Kaliarda is not a mixed language like Romika and Dortika, because part of its secrecy vocabulary comes from Greek or other non-Romani languages. Kaliarda is a Greek secrecy languages, which utilises vocabulary from Romani and other languages.

p. 563 Kaliarda, as a typical cant, uses a lot of word creativity and word games, including Greek neologisms.

p. 564. Dortika has over 100 content words of Romani; Finikas Romika has over 200.

Greek Para-Romani has some unassimilated Romani words (the imperative naš “go away” used with a light verb: na kanume nas “we should do naš“), and it can use Romani plurals: balame “bosses”.

p. 565. Finikas Romika also uses some phonetic transformations, appending the Romani demonstrative to conceal Greek words: ɣat-akalu “cat”, ðermat-al “leather”.

Tzitzilis (2006)² has identified 20 core Romani words in Kaliarda; those loans are used as the basis for deriving dozens more lexemes, so that the the actual count of Romani-based vocabulary is more than 300. (Recall that Montoliu counted these more systematically, and came up with higher numbers.)

p. 566. So the basic strategy for Kaliarda vocabulary is not Romani core vocabulary, as it is for para-Romani, but neologisms through derivation and compounding, using both Greek and non-Greek vocabulary. Phonetic transformations are also used in Kaliarda (ksalimari < maksilari “pillow”, vuelo < avelo.)

Romani has a privileged role among source languages for Kaliarda: it is the source of its basic and light verbs (avelo, ðikelo, xalo, benavo, dzao “do, see, eat, speak, leave/expel”), which are highly productive. In particular, dzao has grammaticalised into the prefix dzas- “without”, e.g.

• dzas-kanis “lame” < Kaliarda kania < Greek kani “cane”
• dzas-futis “lame” < English foot
• dzas-bratelos “armless” < bratelo “arm” < Italian bracetto?
• dza-testos “mad” < Italian testa “head”
• dzas-bar-iarikos “curative” < baro “heavy > “disease”
• dzas-provia “hair removal” < Greek provia “fleece”
• dzas-tekno “abortion” < Kaliarda tekno “child” < Romani tikno “small”

(And the dzas-tirax- “kick the bucket = death” compounds I had already posted about.)

Only Romani has contributed verbs directly to Kaliarda; verbs from other languages are mediated via nouns; e.g. English birth < berθa “birth” < berθ-aro “to give birth”.

p. 567. This indicates greater contact with Romani, and that Kaliarda speakers learned Romani through direct contact with Romani speakers.

In para-Romani, there is some Romani morphology, notably Romani plural suffixes for nouns, indeclinable singulars for inanimate nouns, and uninflected verb loans. (As we already saw in Triantafyllidis’ study, there is also some Romani phonology in Dortika.) In Kaliarda, the only Romani grammatical contribution is some free function words, such as apokate “over here” and axatos “this one” < akate “here”. [Once again: Kaliarda is not a Romani language, the way para-Romani languages are.]

p. 569. Para-Romani verbs, when inflected, are mostly based on the frozen 3sg form: Romani dzanela “he knows” > Romika dzanel-izo.

Sechidou does not discuss this, but Romani verbs in Kaliarda are based on either the 3rd sg form (kurav-al-o, x-al-o, av-el-o, ðik-el-o) or the 1sg form (dzin-av-o, ben-av-o). The 1sg forms are not what you would use if you were a Romani speaker, and wanted to map your native language verbs to Greek; this is yet another inconsistency that suggests that Kaliarda was elaborated by people who didn’t speak Romani well. (In fact, it suggests that Kaliarda did not go through a para-Romani phase at all—that Greek speakers jammed Romani “I know, I speak” into Greek verbs without caring about a consistent approach to Romani morphology, the way a Romani-speaker would do instinctively.

p. 570. In his writing on mixed languages as identity vehicles, William Croft³ considers mixed languages positive acts of identity, because they assert a distinct identity via the minority language whose elements they preserve. Polyglot idioms like Kaliarda, Croft identifies as negative acts of identity: because there isn’t a primary language that the idiom aligns with, those idioms are not asserting a distinct (ethnic) identity, but are only used to exclude outsiders. He treats secrecy languages using transformations of an ethnic language in the same way.