Select Page




This paper is aimed at describing the phonology of some historical loanwords into Polish Lovari Romani. Lovari Romani is a Vlax-type dialect spoken by some hundreds of Roma living in southern Poland. The article investigates various loanwords coming mainly from Greek, Romanian and Hungarian. It also provides the reader with the inventory of phonemes belonging to this dialect and with some analyses based upon various phonetic parametres.

The phonological inventory of the Lovari Romani dialect which Tadeusz Pobożniak (1964) codified features 35 phonemes, according to its author. It evinces 5 vowels (/a, e, i, o, u/) with no opposition of length and 30 phonemes. The Lovari Romani consonant phonemes can be classified by taking into account several criteria:

  1. the manner of articulation:

Plosives: ( /p, b, k, g, ʄ, t, d/), out of which some evince aspiration (/pʰ, kʰ, tʰ/);

Vibrants: (/r/);

Fricatives: (/s, z, ʃ, ʒ, ɕ, ʑ, f, v, x/);

Affricates: (/ts, ʨ, ʥ, ʧ/);

Nasals: (/n, ɲ, ŋ, m/);

Approximants: (/j, l, ł, ʎ/).

  1. the place of articulation:

Bilabials: (/p, b, pʰ, m/);

Labiodentals: (/f,v/);

Alveolars: (/t, d, tʰ, n, r, s, z, ts, l, ł/), alveolo-palatal (/ɲ/);

Post-alveolars: (/ʃ, ʒ, ʧ/);

Palatals: (/ʄ, j, ɕ, ʑ, ʥ, ʨ/), palatal-lateral (/ʎ/);

Velars: (/k, g, kʰ, ŋ, x/).

III. the manner in which the mobile parts are involved in the articulatory process:

Labials: (/p, b, pʰ, f, v, m/);

Apicals: (/t, d, tʰ, n, l, ł, r, s, z, ts/);

Pre-dorsals: (/ʄ, ɲ, ʎ, ʃ, ʒ, ʧ/);

Dorsals: (/j, ɕ, ʑ, ʨ, ʥ/);

Post-dorsals: (/k, g, kʰ, x, ŋ/).

  1. their  phonation. Lovari Romani makes use of
  2. voiced consonants: (/b, d, g, r, m, n, ɲ,ŋ, l, ł, ʎ, ʒ, ʑ, ʥ, ʄ, z, v/);
  3. voiceless consonants: (/p, t, tʰ, k, kʰ, ts, ʃ, ɕ, ʨ, ʧ, x, s, f/).

The classification above is my own classification, while Pobożniak (1964) makes some mistakes when he classifies Romani consonants. For instance, under the label ‘semivowels’, he places the palatal approximant (/j/), which is correct, the voiced alveolar trill (/r/), the voiced alveolar lateral approximant /l/ and the voiced alveolar velarized approximant /ł/. Though the latter can evince a non-syllabic vocalic allophone /ṷ/, which emerged under Polish dialectal influence, he is clearly wrong when it comes to the laterals (/r, l/), since they do not evince vowel-like qualities, at least not in this dialect. All the distributions of /r, l/ show they are always accompanied by a vowel, the vowel being the nucleus of the syllable.

Another mistake that Pobożniak (1964) makes is related to offering the status of phoneme to a certain sound. He distinguishes between the voiced alveolar nasal /n/ and the voiced palatal nasal /ɲ/, which is worth prasising, since one can find minimal pairs such as /ˈpʰen/, which is the Imperative form of the verb phenel (‘to say’) and /ˈpʰeɲ/, which means sister. However, he offers the status of phoneme to an allophonic variant of the voiced alveolar lateral approximant /l/, namely the voiced alveolar velarized approximant /ł/, though they are not in complementary distribution and the author does not bring any arguments in favour of this choice. Moreover, he chooses to use the sign anusvara (<ṁ>), where many scholars have used the grapheme <n>, which signals the actual etymology of the voiced alveolar nasal.

The phonological features of the dialect that he tries to standardize are listed below:

reduction of the affricates (/ʧʰ, ʤ, ʧ) to the fricatives (/ʃ, ʒ/) with subsequent palatalization to (/ɕ, ʑ/) under Polish influence. The palatalization also affected the inherited phoneme /ʧ/, which turned into /ʨ/ in some lexemes. However, the native phonemes /ʃ, ʒ/ were preserved in inherited words that did not feature any affricate and in some loanwords, while the native phoneme /ʧ/ is also stable in some native words. Eg.: /briˈʃin/ ‘rain’, but /ɕaˈvo/ ‘Romani lad’, /ˈʨor/ ‘thief’, but /ˈʧik/ ‘mud’;

assimilation in native lexemes featuring the sound sequences /vn, bn/, the fricative becoming a nasal. E.g. khamni < khabni ‘pregnant’, somnakaj < sovnakaj ‘gold’;

metathesis in some Romanian loans. E.g.: dolmut ‘long ago’ < Rom. demult ‘id.’;

dejotation of Romanian loanwords. E.g.: aba ‘already’<Rom. abia ‘hardly’;

prothetic /i/ in Romanian loanwords featuring the initial consonant cluster /sk/. E.g.: iskirin ‘(they) write’ < Rom. scrie ‘id.’;

denasalization of voiced palatalized alveolar nasal /nʲ/ featured by some Romanian loanwords. E.g.: luja ‘Monday’ < Rom. lunea ‘on Monday’;

aphaeresis encountered in Romanian loanwords featuring a vocalic onset. E.g.: glinda ‘mirror’ < Rom. oglinda ‘id.’, lovina ‘beer’ < Rom. olovină ‘id.’;

lateralization of Romanian intervocalic alveolar trill /r/. E.g.: felastra ‘window’ <  Rom. fereastră ‘id.’;

accidental loss of aspiration. E.g.: diklo ‘kerchief’ (SIR dikhlo ‘id.’);

monophthongation of the Romanian diphthong /oa/ evinced by some Romanian loanwords. E.g.: rota ‘wheel’ < Rom. roa‘id.’;

fronting to /e/ of the schwa (/ə/) featured by some Romania borrowings. E.g.: ketana ‘soldier’ < Rom. cătană ‘id.’, feri ‘only’ < Rom. fă‘without’;

palatalization of the Romanian voiceless velar stop /c/ and the slightly palatalized Romanian alveolar nasal /nʲ/, which turn into /ʨ/ and /ɲ/, respectively. E.g.: raćija ‘whisky, spirit’< Rom. chie ‘plum brandy’, ńamco ‘German’<Rom. neamţ ‘id.’;

fronting to /i/ of the Romanian close central unrounded vowel /ɨ/ comprised by some Romanian lexemes. Some examples are inke ‘yet, still’ < Rom. ȋncă ‘id.’, kirpa ‘rag’ < Rom. cârpă ‘id.’, intrego ‘whole’ < Rom. ȋntreg ‘id.’;

paragogic /n/ in Greek loans ending in the vowel /i/. E.g. skamin ‘bench, stool’ < σϰαμνί ‘id.’, papin ‘goose’ < παππί ‘id.’, karfin ‘nail’ <  ϰαρφί ‘id.’;

dejotation in Greek loanwords featuring the sequence /ja/, leading to the disappearance of the glide /j/. E.g.: tranda ‘thirty’ <Gk. τριάντα ‘id.’;

affrication of the voiced velar plosive /ʄ/ in Greek loans, which turns into the voiced alveolo-palatal sibilant affricate /ʥ/ . E.g.: stai ‘cap’ < Gk. σϰιάδι ‘id.’, lulu‘flower’ < Gk. λουλοῦδι ‘id.’, simai ‘pawn’ < Gk. σημάδι ‘id.’;

apocope of the voiceless alveolar fricative /s/ in some Greek loanwords. E.g.: foro ‘town’ < Gk. φόρος ‘id.’, krisi ‘court’ < Gk.  ϰρίσις ‘id.’;

prothetic /j, v/ in Greek loans whose onset evinces the structure /VC/. E.g.: jefta ‘seven’ <Gk. ἐφτά ‘id.’, vorta ‘verily’ < Gk. ὀρϑά ‘straight, upright’;

sporadic fortition of intervocalic /v/, leading to /b/ and sporadic lenition of intervocalic /s/, leading to its voiced counterpart /z/. E.g.: kakabi ‘kettle’<.Gk.ϰαϰϰάβη ‘id.’, mizmeri ‘noon’ < Gk. μεσημέρι ‘id.’;

palatalization of the alveolar nasal /n/ in some Greek loanwords via debuccalization. E.g.: sapuji ‘soap’ < Gk. σαποῦνι ‘id.’;

despirantization of dental fricatives /θ, ð/ with subsequent occlusivation to /t, d/. E.g.: drom ‘road’ < Gk. δρόμος ‘id.’, vorta ‘verily’ < Gk. ὀρϑά ‘straight, upright, right’;

voicing of Greek labiodental fricative /f/. E.g.: sviri ‘hammer’ <  σφυρί ‘id.’;

affrication of voiced and unvoiced velar stops /ɟ, c/ in Hungarian loans. E.g.: emečo ‘fruit’ < Hung. gyümölcs ‘id.’, karća ‘card’ < Hung. kártya ‘id.’;

unrounding of the close front rounded vowel /y/, which turns into its unrounded counterpart /i/. E.g.: bino ‘sin’ < Hung. bün ‘id.’, firisi ‘saw’ < Hung. fürész ‘id.’;

unrounding of the mid front labial vowel /ø/, becoming a mid front unrounded vowel /e/. E.g.: dźemečo ‘fruit’ <Hung. gyümölcs ‘id.’, kexegindas ‘he coughed’ < Hung. köhögni ‘to cough’, erekre ‘forever’ <. Hung. örökre ‘id.’;

velarization of the voiceless glottal fricative /h/ via debuccalization, which turns into a voiceless velar fricative /x/. E.g.: xiro ‘news’ < Hung. r ‘id.’, kexegindas ‘he coughed’ < Hung. högni ‘to cough’;

affrication of the voiceless velar stop <c> in Armenian loanwords. E.g.: morći ‘skin’ < Arm. morth ‘id.’, paćiv ‘honour’ < Arm. pativ ‘id.’;

epenthetic /v/ in some Hungarian loans ending in a vowel. E.g.: xordovo ‘barrel’ < Hung. hordó ‘id.’, xalovo ‘net’<Hung. háló ‘id.’;

closing of the mid back rounded vowel /o/ to a close back rounded vowel /u/ in Hungarian loanwords. E.g.: zubuno ‘coat’< Hung. zubbony ‘id.’, duhano ‘tobacco’ < Hung. dohany ‘id.’, kuldušo ‘beggar’ < Hung. koldus ‘id.’;

depalatalization of liquids /ʎ/ and nasals /ɲ/ in word-final position in Hungarian loans who were assigned masculine gender, the aforementioned sounds turning into their unpalatalized counterparts /l, n/. E.g.: duhano ‘tobacco’ < Hung. dohany ‘id.’, foglo ‘hireling’ < Hung. fogoly ‘captive’, zubuno ‘coat’< Hung. zubbony ‘id.’;

closing of an open back rounded vowel /ɒ/ found in some Hungarian loans to a mid back rounded vowel /o/. E.g.: kolopo ‘hat’ < Hung. kalap ‘id.’;

unrounding of the open back rounded vowel /ɒ/ found in some Hungarianisms, which turns into an open central unrounded vowel /a/. E.g.: ćak ‘only’ < Hung. csak ‘id.’;

closing to /i/ of a mid front unrounded vowel /e/. E.g.: feńkipo ‘photograph’ < Hung. fenykép ‘id.’, firisi ‘saw’< Hung.  fürész ‘id.’;

fronting to the mid front unrounded vowel /e/ of the mid back rounded vowel /o/ found in some Armenian loans. E.g.: koter ‘piece’ < Arm. kotor ‘id.’, xumer ‘dough’ < Arm. xmor  ‘id.’;

fronting to /i/ of the open central unrounded vowel /a/ present in some Armenian loanwords . E.g.: čikat ‘forehead’ < Arm. čakat ‘id.’.


Pobożniak, T.1964. Grammar of the Lovari Dialect. Kraków: Drukarnia Uniwersytetu  Jagiellońskiego.



(sursa foto:

About The Author

Leave a reply

Adresa ta de email nu va fi publicată. Câmpurile obligatorii sunt marcate cu *