NOTES ON THE GENESIS OF CALÓ AND OTHER IBERIAN PARA-ROMANI VARIETIES PETER BARKER University of Amsterdam 0. Introduction This paper deals with the different varieties of Romani spoken in or originating in the Iberian peninsula, especially Caló (Spanish Romani). Iberian Romani is taken as the collective name for the varieties of Romani spoken on the peninsula. These are now all reported (perhaps unjustly) to be extinct or close to extinct. The argument presented here is therefore based on publications rather than fieldwork. Virtually all the material published shows the complete loss of the inherited grammatical system and an adoption of the grammatical system of the host region. Below, I present data which may help uncover the genesis of Caló. First, I discuss the mixed Romani dialects in general (section 1) and some historical factors related to the Gypsies in the Iberian peninsula (section 2). Then, I briefly discuss the place of Iberian Romani among Romani dialects (section 3). Next, I discuss some structural features of Caló, in order to assess the nature of the mixture in Caló (section 4), and provide some data on the other Para-Romani dialects of the Iberian peninsula (section 5). After that I briefly mention the social functions of Caló (section 6). Furthermore, I discuss the oldest sources of Iberian Romani in order to assess the possible origin (section 7). With these historical, historical-linguistic, structural and clas-sificatory facts in mind, a hypothesis is formulated concerning the genesis of the Iberian Para-Romani dialects, in particular Caló (section 8). As yet, there has been no attempt to explain the genesis of Caló itself, save in some more general overviews. These were based only on a limited number of sources. 1-26 PETER BAKKER 1 . Para-Romani languages Until today, about ten varieties of Romani have been identified which have a Romani lexicon but which have lost the native grammatical system. Instead, these dialects use the morphosyntax of the surrounding languages (see Bakker & Van Der Voort 1991 and especially Boretzky & Igla 1994). These are called Para-Romani languages, a term originally coined by Marcel Cortiade. They roughly display the following characteristics: the vocabulary is Romani (hence basically Indie), but nearly all the phonology, morphology and syntax are non-Romani. Although all Romani dialects borrow heavily from the languages of the host country, no cases are reported of languages with a Romani grammatical system and a non-Romani lexicon. Until now, Para-Romani languages have been described, documented or mentioned in the literature in connection with the grammatical systems of Swedish, Norwegian, German, Catalan, Portuguese, English, Basque, Spanish, Greek, Persian, Turkish, and Armenian. All these varieties must be seen as languages in their own right, and not as dialects of Swedish etc. or Romani. Several theories have been put forward to account for the genesis of these dialects, ranging from saving a dying language by preserving the lexicon, through gradual massive grammatical borrowing, the conscious creation of a mixed language, relexification to language intertwining - the combination of one lexicon with the morphosyntax of another language. Not all of these are mutually exclusive, but the debate on their genesis is still going on. The Para-Romani language of the Spanish part of the Iberian peninsula is relatively well documented. Both speakers and outsiders have written down and published vocabularies or grammatical studies of this language. Almost all of this, however, is in languages other than English. This study will introduce this language to an English speaking audience. The language is generally called Caló [kalo] in the literature. Undoubtedly the Romani word kalo "black" is the source. It is used both as an ethnic self-designation and as a name for the language. Speakers also call it Romano (Quindalé 1867: 49), a nominalized adjective derived from the noun Rom 'Gypsy, man' with the Romani suffix -aniAano. Caló as a Para-Romani language should not be confused with other languages and slang varieties which are also called Caló. All these languages have in common that they are cryptolectal or in-group languages embedded in Spanish structures. This name for non-Romani languages is reported from places as far as California (Polkinhorn et al. 1983). Although such cryptolectal languages may have borrowed some words from Romani or Gypsy GENESIS OF CALÓ 127 Caló, they will not be discussed here. When we mention Caló in this paper, it refers exclusively to the Gypsy language of that name, as spoken by people of Gypsy descent in Spain, Portugal and South America. As a Para-Romani language, Caló is particularly interesting for three main reasons. First, there is a lot of material, covering two or perhaps even three centuries. There is no other Para-Romani dialect documented over such a long period of time. Second, in contrast to most other Para-Romani languages, Caló also makes abundant use of cryptic devices, apparently meant to conceal the meaning of the communication from outsiders. Third, apart from Caló, which has an (Andalusian) Spanish grammatical system, there appear to be several other varieties of Para-Romani languages which came about under contact with languages of the Iberian peninsula, viz. one with a Catalan grammatical system, one with a Portuguese grammatical system (in Brazil), and one with the non-Romance language Basque. These varieties all seem to be derived from one branch of Romani as they all share some lexical particularities apparently not shared with other varieties of Romani (see section 4). 2. Gypsies in the Iberian peninsula If the description of "Egyptian" acrobats and jugglers by the Byzantine writer Nicephorus Gregoras indeed refers to Gypsies, the first Gypsies may have reached the Iberian peninsula in the first decades of the fourteenth century (Fräser 1992: 48). The earliest undeniable reference to Gypsies dates from 1425, when a group was granted safe-conducts in Aragon and Astúria (Fräser 1992: 76). They are first mentioned in Andalusia in 1462, in Portugal in the first decades of the sixteenth century, and in Navarra in the 16th century. Since then there has been a continuous presence of Gypsies on the peninsula. The number of Gypsies today in the Iberian peninsula is estimated at between 300.000 and 700.000. They live in both urban and rural areas, with major centres in Madrid and Andalusia. For all of them (with the exception of some newly arrived groups), the dominant and first language today is Spanish, particularly the Andalusian dialect. Many Gypsies know at least some Caló words, but even the most knowledgeable speakers may not know more than fifty or a hundred words (McLane 1977, 1985). There seems to have been an ungoing decline of the languages for a long time, resulting in a large-scale shift to Spanish, with an occasional Caló word. 128 PETER BAKKER 3. Lexical classification of Iberian Romani among Romani dialects As any other language, Romani consists of dialects. These are grouped into clusters of dialects, but there is no single accepted classification. In general, these classifications are based on three factors: internal sound developments in words inherited from Indie and Persian languages, shared or different grammatical features and the main source languages of borrowed items (hereby excluding the language of the host country). The first two are of course common in any dialect classification, but in the Spanish case the grammatical system is lost and hence cannot be used. The use of borrowed items is rather typical of Romani dialectology alone. On grounds of the lexicon (phonological peculiarities of inherited words and source language of the borrowings), Caló and the other Iberian dialects have been classified differently (Kaufman 1979, cf. Hancock 1988). There are a few words used in the Iberian dialects of Romani which deviate in certain ways from all other varieties. This means its classification as a different branch seems justified. The clearest cases are the following: (1) The word for "boy" and "girl" are raklo and rakli in Romani, but lakri in the Basque Country, lacrollacrin in Brazil, lacro and lacri in Caló. (2) The words for "father" and "mother"' are dad and dai in common Romani, but bato and bata in Brazil, bato and bati in the Basque Country, bato and bata in Caló, bato (and dai) in Catalonia. Further research may reveal other such Iberian similarities. On the grounds of the lexicon, different classifications have been proposed. Most researchers classify Iberian Romani as relatively isolated among Romani dialects: Kaufman (1979) makes it a separate branch of European Romani, on a par with the Northern, Balkan, Vlach, Greek and Zargari dialects. Kenrick (see Hancock 1988) classifies Iberian Romani as a Balkan dialect on a par with Vlach and southern Balkan Romani. Boretzky (1992) shows that Caló has a significant number of words in common with the Northern dialects, in particular Sinti and Angloromani. Iberian Romani, therefore, seems to be historically a separate branch, most closely related to the Northern dialects and non-Vlach Balkan dialects. 4. Some Romani and Spanish derived features Caló is not a spontaneous ad-hoc mixture of Romani and Spanish. The Spanish part (at least in the 19th century) differs from the varieties of Spanish spoken locally, and in some cases Spanish is not spoken locally (e.g. in GENESIS OF CALÓ 129 Portugal, see below). In some cases, Gypsies speak archaic varieties of a language, e.g. the variety of Catalan spoken by the Gypsies near Perpignan in North Catalonia. In this section, I will present some of the differences. 4.1. Andalusian features The Spanish component of Caló is not ordinary standard Spanish but rather the Andalusian dialect. This is true to a certain extent for all sources of Caló, not only in Andalusia but also in Madrid. Apparently, Caló found its origin in Andalusia, from where it spread to other areas. In Portugal, for instance, Caläo has an Andalusian Spanish rather than Portuguese base. Also, Caló retains some archaic features lost in modern Spanish. For instance, old Spanish ende "since", modern Spanish desde, Caló ende (Keller 1892). A Catalan influence is also suggested in words such as matejo "self", Catalan mateix, Castilian mismo. Andalusian Spanish differs from standard Spanish in a number of points. a) Spanish /Kl (orthographic '11') is pronounced [j]. b) final consonants tend to be dropped, even those of grammatical endings (for- instance plural -s). c) The phonemes /s/ and IQI are to an extent in free variation. d) lál and /g/ between vowels are often dropped, e.g. Spanish asadura, And. saura, Spanish pasado, And. pasao. This is also found in other Spanish dialects. e) 'confusion' between (some?) hi and III. Spanish el, And. /er/ or Id. f) alternation (some words) Cast, /h/, And. lil. g) prothetic morphemes a-, des-, en-, es-. The fact that we find these Andalusian features in all Caló varieties, suggests that they came into being in Andalusia and later spread to other parts of the Iberian peninsula. (For details, see Boretzky 1992: 32). Catalonian Romani shows many conservative phonological features independent from Castilian (Boretzky 1992: 29-34). 4.2. Phonology The phonology of Caló is Spanish; all Romani words are adapted to the phonology of Andalusian Spanish. This means that Romani phonemes not existing in Spanish are lost in Caló. For instance, in Caló there are no aspirated stops. Romani aspirated stops become unaspirated stops in Caló (except /th/ which becomes /č/). It is clear, however, that the aspirated stops were still used when the Gypsies arrived on the Iberian peninsula. The 130 PETER BARKER aspirated /th/ of Romani became /Č7 rather than /t/, for instance chem /čem/ 'land' < Romani them, chute /čute/ Romani thud 'milk' (Boretzky 1992). Further, the voiced affricate lážl became /č/ in Caló and NI became /b/, etc., following Spanish phonology. One of the features of Caló inherited from Romani is its stress pattern, or at least some aspects of it. Whereas Spanish words rarely have stress on the final syllable, this is commonly so in Romani and Caló (inherited from Romani). See for instances the stress markers in example (1) below. This may give Caló a distinct flavour. Phonotactic constraints seem, at first sight, to be the same in Caló and Spanish, in that the syllable structure is identical. However, there are exceptions as well: in Caló there are words ending in -m, but these do not seem to exist in Spanish. A recent study suggests that the phonotactic constraints of Spanish and Caló are not completely identical. This thought-provoking study by Dietz and Mulcahy (1988) compares the combination of letters in a Bible chapter in Castilian and Caló. They conclude on the basis of statistical evidence "that Caló and Castilian differ greatly in the way they form and distribute basic linguistic particles" (i.e. letters' or perhaps, by implication, phonemes). For example, in the samples of the same text in Spanish and Caló versions, the phoneme III (the digraph "ch") was counted 2.056 times in the Caló text and only 242 times in the Spanish text, a difference of nearly 850 % (Mulcahy & Volland 1986: 144 n. 4). How can we explain this? There are several possibilities. With very few exceptions, Caló phonemic structures are possible in Castilian. This is to be expected, since the Romani words are adapted to Spanish phonology. It may be, however, that the frequency of certain combinations of phonemes differ in both languages because the lexicon of Caló remains basically Romani and follows some non-Castilian features of Romani. For instance, Romani /c/, /č/, /čh/ and /th/ all became lil in Caló, which could explain the high proportion of HI. In fact, this is what the results of Dietz & Mulcahy's study suggest. Whatever the explanation is, we agree that Caló cannot simply be a 'Gypsified' version of Castilian (Dietz & Mulcahy 1988: 15). 4.3. Morphology Caló uses Spanish derivational affixes, such as the diminutives -ico, -ho, the superlative -isitno, the adverbial ending -mente, etc.. Apart from that, it GENESIS OF CALO 131 uses a number of prefixes (apparently without any meaning) to verbs, such as a-, en-, de-, des-, some of these derived from Andalusian Spanish. Inflection is also Spanish. For instance, it has the nominal plural inflection -s and whole verb paradigms from Spanish. In Caló, all verbs are conjugated according to the class of Spanish verbs ending in -ar which is the most regular and unmarked class. This class is also used for borrowed verbs in Spanish. From this point of view, the Romani verb in Caló behaves like a borrowed verb in Spanish. Although Caló morphology is so close to Spanish, it does preserve some Romani morphemes, productively or not. A small number of these are used in Caló. First, there is the suffix -pen, forming abstract nouns from verbs and adjectives. Second, there are the gender suffixes -/ (F) and -o (M) used with adjectives in Caló and nominalized adjectives. Whereas the nouns with these endings in Romani are probably lexicalized forms, if they were used with adjectives they would probably still be productive. Third, there are also instances of the noun plural suffixes' -ia(s) for feminine nouns and -e(s) for nominalized adjectives, followed by the Spanish plural suffix -s. Some authors also use Romani case endings, probably only non-productively. In any case, these inflected pronouns are not used like one would expect considering their function in Romani: Caló uses datives or genitives with subject function, and this is ungrammalical in Romani. There are several instances of the case marker -ha for the comitative/ instrumental case, as in romi-ha, dal-ha, sila-ha. One finds the -ha endings for instance in Sinti and Balkan and central dialects. It is undoubtedly related to the Caló preposition sar 'with' and the Romani instrumental case -sa(r)*. There are more 'frozen' Romani morphemes in Caló: many Caló verbs have the Romani personal conjugation markers -av and -el between the verb stem and the Spanish conjugation. They are derived from the Romani first person singular ending and the third person singular ending respectively, but they do not function as such. To these, Spanish inflection is added. Quindalé (1867) mentions a semantic difference between verbs with the -el- element and those without, based on the same root. Those with -el- denote a more intensive action, e.g. chinar "to cut", chinelar "to harvest"; querar "to do", querelar "to execute". This is an innovation in Caló; it does not exist in either Spanish or Romani. Caló apparently makes use of Romani inflection to denote 'aspect' or 'Aktionsart'. Gender assignment in Caló deserves a special study. Here, I will just present some observations. Both Romani and Spanish distinguish masculine ->2 PETER BAKKER and feminine nouns. Assignment of gender of nouns in Spanish is partially based on the semantics (especially with animates), partly on the phonological structure of the stem, especially the final vowel or consonant. In Caló, this is also the case. Quindalé (1867: 51) says that words ending in a consonant or in the vowels -ó, -e, -o, or -u are masculine, whereas words ending in -í, -a, or -i are feminine. 4.4. Lexicon McLane (1977,1985) recorded a few hundred words considered Caló by the Gypsies of Guadix and other parts of Andalusia and Spain. He calculated (1985: 188) that 68 % of the lexical items in his corpus are of Indie (hence Romani) origin and 22 % of Spanish origin (many of those distorted). We üso find the usual sources of non-Indie words in Caló from the pre-European layer of loan elements present in all Romani dialects, such as Armenian, Persian, Greek and Slavic languages. Boretzky (1992), using a wider vocabulary, detected words from Slavic languages, in some cases specifiable is South Slavic, Czech or Polish (Caló dosta "enough"< South Slavic dosta; ľaló kornes "laced boot"< Czech škornje; Caló stajnia "horse-stable" < Polish estanja). There are also a few Hungarian words in Catalonian Romani, such as arany "silver" < Hungarian aranj and ein "colour" < Hungarian szin. Many of these words, however, are not specific to Caló; in fact they can be Found in other Romani dialects as well. Boretzky further identified a number 3f words which Caló hw> in common with the Sinti and Angloromani dialects ^Boretzky 1992: 16), so called 'northern dialects'. Further there are words aken from Germania (Spanish Cant) in Caló (see below 6.3). It should be mentioned that Iberian Romani also seems to have Greek and perhaps Persian) words not attested in other Romani dialects. A few )thers have been identified as Arabic (e.g. (j)azari "ten" < Arabic ?ašara; víoroccan variety). Overall, the Caló lexicon is highly aberrant among lomani dialects because of the singular phonological development, the ;ryptic devices used (see below) and the relatively large number of words of inknown origin. 15. Word Order Word order is the same as in Spanish. GENESIS OF CALÔ 133 4.6. Innovations Although there is a clear Romani component and a clear Andalusian Spanish component, there are also elements which are neither. I focus on two of these: the use of innovative place names and the use of cryptolalic formations. 4.6.1. Place Names Local place names are among the first elements to be used by immigrant groups when using their native language in the land of immigration. Nevertheless, the Gypsies in Spain did not (always) do so. They formed new place names, many with unclear etymons (see below), instead of adopting Spanish place names. This list of place names is extracted from Sales Mayo (1870). Some Caló place-names: Barcelona: bajari (Sales Mayo 1870:8) (Barcelonés: bajanó) (Sales Mayo 1870:8) Extremadura: chim ye manró (Sales Mayo 1870:24) Granada: meligrana (Sales Mayo 1870:47) Guadalquivir: len baró (Sales Mayo 1870:43) Habána: bobaní (Sales Mayo 1870:13) Jerez: borbóreo (Sales Mayo 1870:13) Judea: bordajía (Sales Mayo 1870:13) (Judio: bordajú) (Sales Mayo 1870:13) Londres: llundun (Sales Mayo 1870:45) Madrid: madrilati (Sales Mayo 1870:13,45) Sevilla: safacoro (Sales Mayo 1870:13) Some of these are clearly circumscriptions, such as len baró "big river" for Guadalquivir, and chim ye manró '"and of the bread" for Extremadura. Others are distortions of the existing place names, such as madrilati for Madrid and llundun for London. Some of them are cryptolalic formations, such as Bobaniiox Habána. It is a pun on Spanish haba "bean" and Caló bobi "bean": one part of the word is replaced by the Romani word with the same meaning (see below for other distortions of this type). This indicates that speakers intended to make their speech unintelligible to outsiders by avoiding borrowing, and instead making up new words. We find the same phenomenon with place names in Angloromani (Hancock 134 PETER BAKKER 1984a). However, it is not typical only of Para-Romani dialects: the inflected dialect of Finland and the Sinti dialect in Germany also have cryptic forms for place names. 4.6.2. Cryptolalic formations Cryptolalic formations are not limited to place names. Wagner (1937-1938) discusses a score of these cryptolalic forms for a variety of words. Usually, when a Spanish word contains a syllable which is also an existing Spanish word, then the small word is translated into Romani. For instance, the Spanish name for "March" is Marzo. This word contains the sequence mar, identical to the Spanish word mar "sea". In Caló, the sea is loria (Romani dorjav), hence Marzo becomes Loriazo. Another example is the Caló word for "namesake". In Spanish it is tocayo. This resembles very much the Spanish verb tocar "to touch". The Caló verb bajamb-ar means "to touch", so the Caló word for "namesake" becomes bajambayo. Another example, mentioned by Keller (1892: 171) is Caló sardenar "to condemn", from Spanish condenar, both first syllables meaning "with". A final example is the word for "die, dice" in Caló. The Spanish word is dado, which happens to be homophonous with the Spanish word for "given" (past participle). In Caló, "given" is diňao, the past participle form of the verb dinar. The Caló noun for "die" is therefore also diňao. Cryptolalic formations like these may be devices to keep the language unintelligible to outsiders, such as those who learned some of it (see above). Such processes are common in secret languages. 5. The four Iberian Para-Romani languages Caló is not the only Para-Romani dialect of the Iberian peninsula. In Bakker (1991) I showed that the Romani dialect of the Basque Country was (or is) a mixed dialect like Caló, a Para-Romani dialect or an 'intertwined' language. It has a Basque grammatical system and a Romani lexicon. It is therefore structurally different from Caló, especially since Basque is not an Indo-European language. It is less known that Catalonia also had such a mixed language, with (again) a Romani lexicon and a Catalan grammatical system. This differs from the Catalonian Romani dialects described in Ackerley (1915), which show a reasonably intact Romani grammar (as the only dialect of the peninsula). There are a few texts in a novel in Catalan concerning Catalonian Gypsies that are written in Catalonian Para-Romani. The following text is GENESIS OF CALÓ 135 illustrative. The text and its Catalan translation are both from Vallmiljana 1908 (as cited in Leblon 1982: 63). The analysis and interpretation are mine2. Catalonian Para-Romani De la mutzi d'un aranui van nyisquerbá un dicaló un busnó en diquelava panant: Quin sambanbanó Catalan in source De la pell d'un gat ne varen treure un mocador un home s'ho mirava dient: ! Que es bonic! "They go to take of a neckerchief made of a cat's skin. A non-Gypsy saw it, saying: How beautiful!" Interpretation: De la mutzi d'un aranui of the skin of-a cat un busnó endiquel-ava a non-Gypsy see-3.PAST Van nyisquerbá un dicaló they-go take-off a neckkerchief pan-ant: Quin sambanbanó say-ing: what beautiful The Romani-derived words are mutzi (< R. morthi "skin") dicaló (^ >-. & o, 3 B CUTS >> ^ o (A O CS S3 ° Si á ,S5 C «J 3 W5 M U •u S, >> t— e. 3 Ž O, O §í u C4 fl u o et) O >«J oo i o ej CJ 03 a, a, G >, i ej 3 a g >-. o "> yes S. ä o U "O u 3 « S3 Sß 8 ° c3Ä 8-S c« ?3 O .3 &. u o er o ,53 oo o c C 3 ej •O í«l! o O! S. S (S •«• S -g § -o > m O ö ej ej 5 ej a o c P 3 o ca x § s j?£ .3 Š o s«, ej ej C O O O i-'sall'a u. JS ^ es u ej ej ej " GENESIS OF CALO 141 All the functional elements are of Romani or Spanish origin except for the italicized forms somia for "for" and nu for first person singular, both of which are of unknown origin. One result is clear: the sources show no evolution from a more Romani-oriented variety towards a more Spanish-oriented variety. The first text (Bright 1818) has almost exclusively Spanish function words, whereas later texts have more function words from Romani. In fact, the number of Ro-mani-derived function words used in sources after 1850 is greater than before 1850,* contrary to what one would expect, if Caló is seen as a gradual evolution away from Romani towards Spanish. This could result from regional differences, however, or increasing purism. For the final analysis, all sources should be studied, and in a more thorough way than was possible within the scope of this paper. 7. The oldest sources of Iberian Romani If we want to explain the genesis of Iberian Romani, we have to look for the earliest sources. As yet, no exhaustive historical research has been done concerning mentionings of the Gypsies' languages. However, there are a few early remarks about the language of the Gypsies in Spain. Already in 1608 the Spanish writer P. Martin Delrio, in his book Disquisitionum Magicarum said that the language of the Gypsies "was a vernacular invented by them to replace their native tongue, which they had forgotten" (cited in Spanish in Pabanó 1915: 179)3. One early source claims that the Gypsies around 1600 were fluent in Greek. This may point to a previous presence in a Greek speaking area, presumably not long before. It also seems that the Gypsies lost their language quite early and, then, may have completely replaced it with Caló within 150 years after arrival in Spain. More historical research is needed here. If we look at the two earliest sources of Caló (or Iberian Romani in general), we see that these two earliest sources already show a complete hispanization of the language. I will discuss both of them in some detail. 7.1. Anonymous word list 17 th century The oldest source is a word list found in the Biblioteca Nacionál in Madrid by John M. Hill and published early this century (Hill 1921). It is called Lengua egipciaca; y mas propio: Guirigay de Gitanos [The Egyptian language, more properly, jargon of the Gypsies]. Hill seems to suggest that it 142 PETER BAKKER dates from the 17th century, but he does not indicate how he arrived at this conclusion. The list contains 61 words, almost all of them clearly Romani. Although there are no phrases in the list, it is also clear that the Romani inflection is not inherited. Indeed the words have inflection: verbs end in -ar, and plural nouns end in -s, both in the Spanish and in the Caló columns. There are also some shared derivational endings in both sections, such as -ador (or rather its Andalusian form -aor) for a person who performs an action. Although there are some Romani grammatical morphemes, these are most probably lexi-calized forms, such as gachó and gachi for "servant" and "maid", with the Romani masculine and feminine endings. The -o and -i endings are productive in Romani proper for adjectives, but not for nouns and we only find them used on nouns in this list. The same iox puxó and/nay for "old man" and "old woman" respectively. The x is probably a misreading for r, since the Romani source word is the adjective puro (M), puri (F). If this list is really from the 17th century, it means that without any doubt there were already completely hispanicized varieties of Romani by then. 7.2. Bright The first dated publication concerning Caló, and the first one which contains sentences, is Bright's European travel account. In an appendix, he compares the Gypsy languages of Hungary, Britain and Spain (Bright 1818: lxxviii-xcii). Bright gives roughly 150 words and some 25 sentences. In view of the fact that Bright's book is so hard to obtain, his text material is presented in an appendix to this paper. Bright did not collect the material himself, but he received it from one of his friends (Bright 1818: ix). In the word list many plural nouns end in -s and verbs in -ar. More important, he gives a number of sentences which clearly show that it is Caló we are dealing with and not Romani with inherited inflection. The proportion of Spanish and Romani lexical elements used here differs considerably from one sentence to another, as we see comparing (1) and (2), in which Spanish elements are italicized. (1) ochanaba mangue loque chile (Bright, Caló) know-3 1SG that-which tell-2SG "I know not what you tell me" (Bright, English) (Romani džanav "I know", man-ge "to me", chile ??; perhaps džilav I sing'?) GENESIS OF CALÔ 143 (2) gillaie de mi que no te pueda indicar (Bright, Caló) go-2PL.IMPER from me that NEG you can-3 see-INF "get out of my sight" (Bright, English) (literally: "go from me (so) that he cannot see you") (Romani gel-em "I went", dikh-av "I see", with Andalusian in-; cf. 4.1.) In short, there is no doubt that in 1818 Caló was already the mixed language known from later sources. It even has the Romani dative form used in all grammatical cases, as in the later sources (see also section 5). 7.3. Inflected Romani in the Iberian peninsula There are few sources of inflected Romani from the Iberian peninsula. Apart from the inflected dialect of Catalonia, there are only two brief texts recorded in the Basque country in the 19th century, in a period when all the other sources from the Basque Country already show that the mixed variety was spoken (cf. Bakker 1991). As these were songs and prayers, they can be expected to have been preserved longer. This indicates that the inflected language as a means of communication was lost early outside Catalonia. The texts from the Basque Country will be given in appendix 2. 8. The genesis of Caló No hypotheses have been formulated as to the genesis of Caló. We would like to know, however, when, why and how Caló came into existence. Was it a gradual or a sudden process? Was it conscious or unconscious? Under what circumstances did it happen? The lack of hypotheses concerning the genesis of Caló contrasts with the study of Angloromani and Scandoromani. For the first, there has been a debate, notably between Donald Kenrick and Ian Hancock, as to the genesis of this language. The differences between the two positions can be summarized as follows. Hancock believes that Angloromani was consciously created by Gypsies and British outlaws in the 16th century and that Angloromani co-existed with inflected Romani for a long time (Hancock 1984a, 1984b). Kenrick, however, believes it was a gradual development taking place mostly in the 19th century, whereby more and more English grammatical elements were integrated into Romani (Kenrick 1979). Hancock's ideas were followed up in 144 PETER BAKKER his paper on Scandoromani (Hancock 1992). A third hypothesis was formulated by Boretzky & Igla (1994), in a comparative study on Para-Romani dialects. They suggested a reversal of language shift, at a point when the youth had shifted to the language of the host country, but could still rely on the older generation to conserve the Romani lexicon. These hypotheses were based on parallel cases, not in Caló itself. Caló is sometimes mentioned in these studies. Hancock (1984a) also mentions Caló in his papers on the genesis of Angloromani, showing many parallels in functional and structural features. On the one hand this is no proof in itself, since each case must be studied individually. On the other hand, the parallels are strong and the facts unusual, so that a comparison is justified. In the foregoing we presented (among others) the following facts concerning Caló: - The Gypsy population in the peninsula has been relatively settled for many centuries. - The Spanish elements show archaic features. - The Spanish elements show Andalusian features. - Already in 1609 they were reported to have lost their own language. - Around 1600 they were reported to speak Greek. - In the earliest source of Iberian Romani (17th century?) the language was already thoroughly hispanicized. - The earliest source of Romani contained already Germania lexical items. - Germania in the early 17th century hardly contains Caló words. - The function words show no evolution from more Romani oriented towards more Spanish oriented varieties. - Caló contains distorted Spanish items. - Caló was also known by quite a few non-Gypsies. This is compatible with an early genesis of Para-Romani, i.e. relatively soon after the arrival on the peninsula. If Hill's date for the word list is correct, Iberian Romani was already a Para-Romani language in the 18th century and, if Delrio can be trusted, already around 1600. A few facts can be used as arguments against a gradual development, especially the early sources and the non-evolution of function words. Nothing in itself gives convincing evidence for either of the hypotheses. Nevertheless, an early genesis (16th century) seems most likely, considering the conservatism of some Spanish elements, Delrio's remark and the nature GENESIS OF CALÓ 145 of the early sources. It may have been a conscious creation, perhaps related to an attempt at the reversal of language shift. 9. Conclusions On the Iberian peninsula and Latin America, at least four Para-Romani languages have come into being, most likely independently of one another. Caló combines Iberian Romani vocabulary with Andalusian Spanish grammar, Errumantxela in the Basque Country combines Iberian Romani vocabulary with Basque grammar, around Barcelona an unnamed variety came into being combining Catalan grammar and Iberian Romani, and in Brazil a Portuguese grammar variety came into being. The Para-Romani variety of Portugal is derived from Spanish Caló and is called Caläo. An early genesis for Caló is suggested by some archaic Spanish remnants, pointing to conservatism on the part of its speakers. Perhaps the language already existed in the 16th century, several generations after the arrival in Spain. Regional differences remained, some being'closer to Spanish, others to Romani, as far as the use of function words is concerned. Only in Catalonia the inflected language survived into the 19th century, apparently beside a Para-Romani variety. Further research on Caló is needed on a number of subjects. An etymological dictionary is badly needed, as well as a critical assessment of the sources, including the question who took over what from whom. Moreover, a thorough grammatical study is also needed, whereby both the Spanish and Romani source dialects should be taken into account. NOTES 1 Boretzky (p.c.1993) has observed that in some Balkan dialects (notably Arli) both -kar and -ha are used beside one another, and the Gurbel dialect of Romani has the suffix -sar, and he thinks that the preposition sa(r) and the case ending -sa(r) existed side by side for a longer time (as with most of the other prepositions/ case endings). Caló shows traces of this. 2 I thank M. López Abellán for his help with Catalan. 3 Pabanó wrote: "decia que era un habla inventada por ellos para suplir su idioma nativo, que se les habfa olvidado" (Pabanó 1915: 179). In a superficial search in Delrio's voluminous work, I was not able to locate the exact page for this quotation. 146 PETER BAKKER Appendix 1 THE CALÓ SENTENCES FROM BR Spanish-derived elements are italicized. Las ducais me marel-an Vastel-a-ie cate ochanaba mangue loque chile prastarelo no orobeles mi dai por la estipen de la mangue Apande umd el bundal Abele umd acoté Nagueíe umd Endineme un prajo Abele umda]w\ar Voy a sobeliir Se va a romandifi-ar Naguemos a jonjobar/e Amcabado umd Ábela la pani Esta chai es lili Haber el boqui de un dever terero Que engispo? Se ha endiňŕKfo el pamé a la chai Gillaie de mi que no te pueda indicor No se gille umd porque tereto ir de esta cocorri En el chen de los chindoquenck« el que avela un sacai es un clai Romandiflare con este chavo Malos menguis re jamelan Mai fen tengas tu cuerpo Mai fen tengas Gitano Song: Del estaribel me sacan Montadito en un june ľme van acurruba/ufo (1818). Trouble kills me Sit by me I know not what you tell me run! run Do not weep mother for my health Shut the door Come hither Go! begone Give mé a cigar Come to eat I am going to sleep .He is going to marry Let us go and deceive him You are a thief It rains This girl is very wild To be very hungry What do I see? The money was given to the girl Get out of my sight Do not leave me I fear to go hence alone In the country of the blind he who has one eye is a chief Marry this fellow May the devils eat you (curse) (curse) They take me from the prison mounted on an ass and flog me GENESIS OF CALÓ 147 Por las calles catorre through the streets Utterance: Chavo gillate. que vienen los Doraií á cogerte Date con los carcoj en el Buerengi Be off boy. The officers are coming to take you Give your shoes against your breech Appendix 2 The two texts below were taken from Cénac Moncaut (1855: 345), whose informant was somebody named Sansberro. Sansberro is one of the Cascarots in Ziburu (Ciboure), a fisherman's village on the northern Basque coast, close to Donibane Lohitzun (Saint Jean de Luz). The Cascarots are said to be descendants of marriages of Gypsies with Basques (see Webster 1889). The first text is a Catholic prayer. (Source text) LebaTusquet Echa Bisquet Le Apelinguet Taberamente (French in source) Au nom du Pere Et du Fils Et du Saint-Esprit Ainsi soit-il (Romani) Le Batoske e chaves-ke le apenicě-ke t'avel amende (my translation) For the Father For the Son For the Holy Spirit That he comes to us This short text contains a few words typical of Iberian Romani: bato for "father" (Romani dad. The word apelinguet for "Holy Spirit" is unclear, but it may have to do with peniče (written peniche), the Caló form for 'Holy Spirit', presumably derived Írom Greek pneuma. The interpretation of the second text is more complex, since the translation does not seem to fit the text and it contains some words which are hard not clear. (source text) Usti, usti, chajori mindre foucar moyorré Samendé caracolenge (source translation) Regardez, regardez, cette fille, Avecsajolie figure, Qui va chercher des escargots (Romani interpretation) usti, usti, chajori minre, sukar muiore-sa, amende caracol-enge (corrected translation) stand up, stand up, little girl of me, with a pretty little mouth to us for snails 148 PETER BAKKER REFERENCES Ackerley, Frederick G (1915) The Romani speech of Catalonia. Journal of the Gypsy Lore Society, New Series 8, 99-140. Acton, Thomas & Donald Kenrick (1984) Romani Rokkeripen ToDiwus. The English Romani Dialect and Its Contemporary Social, Educational and Linguistic Standing. London: Romanestan Publications. Bakker, Peter (1991) Basque Romani - a preliminary grammatical sketch of a mixed language. In: Bakker & Cortiade (eds.) 1991, 56-90. Bakker, Peter & Marcel Cortiade (eds.) (1991) In the Margin of Romani: Gypsy Languages as Contact Languages. Studies in language Contact I. Amsterdam: Publikaties van het Instituut voor Algemene Taalweten-schap 58. Bakker, Peter & Hein van der Voort (1991) Para-Romani languages: an overview and some speculations on their genesis. In: Bakker & Cortiade (eds.) 1991, 16-44. Boretzky, Norbert (1992) Romanisch-Zigeunerische Interferenzen (zum Caló). In: Prinzipien des Sprachwandels. I. Vorbereitung. Beiträge zum Leipziger Symposium des Projektes "Prinzipien des Sprachwandels" (PROPRINS) vom 24.26.10.1991 and der Universität Leipzig, Jürgen Erfurt, Benedikt Jessing, Matthias Perl (HGG). Bochum: Studienverlag Dr. N. Brockmeyer. 11-37. Boretzky, Norbert & Birgit Igla (1991) Morphologische Entlehnung in den Romani-Dialekten. Arbeitspapiere des Projektes "Prinzipien des Sprachwandels'" Arbeitspaper Nr. 4. Essen: Fachbereich Sprach-und Literaturwissenschaften an der Universität GH Essen. Boretzky, Norbert & Birgit Igla (1994) Romani Mixed Dialects. In: Peter Bakker & Maarten Mous (eds.) Mixed Languages. 15 Case Studies in Language Intertwining. Amsterdam: IFOTT. 35-68. Borrow, George (1841) The Zincali... With an original collection of their songs and poetry, and a copious dictionary of their language. London: John Murray. 2 Vols. [Reprint 1923: London: Constable. Borrow, George (1843 [1899]) The Bible in Spain. London: John Murray. Bright, Richard (1818) Travels from Vienna through Lower Hungary. Edinburgh: Archibald Constable & Co. Brown, Irving (1922) The knowledge of Gypsy by the Gentiles of Spain. Journal of the Gypsy Lore Society, 3rd series 3, 143-144. Campuzano, R. (1848) Orijen, usos y costumbres de losjitanqs y diccionario de su dialecto. Madrid. [2nd edition, 1851. Facsimile reprint, Madrid 1980: Heliodoro]. Cénac Moncaut, M. (1855) Fragment de vocabulaire gitanos. In: Cénac Mon-caut, M. Histoire des Pyrenees et des Rapports Internationaux de la France avec I'Espagne depuis les temps les plus réculés jusqu'ä nos jours. Paris: Amyot. GENESIS OF CALO 149 Coelho, Francisco Adolpho (1892) Os Ciganos de Portugal com un Estudio sobre el Caló. Lisboa: Imprensa Nacionál. Colocci, Adriano (1888) The Gitanos of today. Journal of the Gypsy Lore Society, First Series 1, 286-289. Delrio, Martin A. (1633 [1608]) Disquisitionum Magicarum. Cologne: P. Henning, [originally published in 1608]. Dietz, Henry G. & F. David Mulcahy (1988) Romani of a third place: a statistical analysis of nineteenth-century Caló and Castillian. In: Carmen DeSilva, David J. Nemeth & Joan Grumet (eds.) Papers from the 8th and 9th Annual Meeting of the Gypsy Lore Society, North American Chapter. New York: Gypsy Lore Society, North American Chapter. 1-17. Fräser, Angus (1992) The Gypsies. London: Routledge. Hancock, Ian F. (ed.) (1979) Romani Sociolinguistics. (International Journal of the Sociology of Language 19). The Hague: Mouton. Hancock, Ian F. (1984a) The social and linguistic development of Angloromani. In: Acton & Kenrick 1984. 89-134. Hancock, Ian F. (1984b) Romani and Angloromani. In: P. Trudgill (ed.) Language in the British Isles. Cambridge: CUP. 367-383. Hancock, Ian F. (1988) The development of Romani linguistics. In: Jazayery, Mohammed Ali & Werner Winter (eds.) Languages and Cultures. Studies in Honor of Edgar C. Polome. Berlin etc.: Mouton De Gruyter. 182-223. Hill, John M (1921) A Gypsy-Spanish word-list. Revue Hispanique 53, 614-615. Hill, John M. (1945) Poésias Germanescas. Bloomington, Indiana: Indiana University Publications 15. Jimenez, Augusto (1853) Vocabulario del dialecto gitano... con varios rezos, cuentos. 2nd Sevilla: Imprenta del Conciliador. [originally: Sevilla, 1846: Gutierrez de Alba] Kaufman, Terrence (1979) Review of Rajendra Weer Rishi 'Multilingual Romani Dictionary.' In: Hancock (ed.) 1979, 131-144. Keller, A (1892) Einfluss des Spanischen auf der Sprache der in Spanien lebenden Zigeuner. Zeitschrift für Romanische Philologie 16, 165-173. Kenrick, Donald (1979) Romani English. In: Acton & Kenrick 1984, 79-88. Also in: Hancock (ed.), 111-120. Lebion, Bernard (1982) Les Gitans dans la littérature espagnole. Toulouse: France-Ibétie Recherche. McLane, Merrill F. (1977) The Calo of Guadix: a surviving Romani lexicon. Anthropological Linguistics 19, 303-319. McLane, Merrill F. (1985) Romani speech domains in Spain and Portugal. In: J. Grumet (ed.) Papers from the 4th and 5th Annual Meeting of the Gypsy Lore Society, North American Chapter. New York: Gypsy Lore Society, North American Chapter. 188-198. 150 PETER BAKKER Moraes, Alexandre Jose de [Mello] (1885) Canconeiro dos Ciganos. Poesia Popular dos Ciganos de Cidade Nova. Rio de Janeiro: B.L. Gamier. Moraes, Alexandre Jose de [Mello] (1886) Os Ciganos no Brasil. Con-tribucáo Ethnographica. Rio de Janeiro: B.L. Garnier. Mulcahy, F.D. & Anita Volland (1986) The Gospel according to St. Luke: a preliminary analysis of the same text in three Romany dialects. In: Joanne Grumet (ed.) Papers from the 6th and 7th Annual Meeting of the Gypsy Lore Society, North American Chapter. New York: Gypsy Lore Society, North American Chapter. 135-145. Pabanó, R.M. (1915) História y Costumbres de los Gitanos. Barcelona: Muntaner y Simón. [Facsimile Reprint, 1980, Madrid: Ediciones Giner] Polkinhorn, Harry, Alfredo Velasco & Mal Lambert (1983) El Libro de Caló. San Diego: Atticus Press. Quindalé, Francisco (1867) [pseud, of F. Sales Mayo]. Diccionario Gitano. Madrid: Oficina Tipográfica del Hospicio. Sales Mayo, Francisco de (1870) El Gitanismo. História, Costumbres y Dialecto de los Gitanos. Con un epitome de gramática gitana, primer estudio filológico publicado hasta el dia, y un diccionario Caló-Castellano, que contiene, ademas de los significados, muchas frases ilustrativas de la acepcion propia de las palabras dudosas. Madrid: Libreria de Victoriano Suárez. Sowa, Rudolf von (1889) The dialect of the Gypsies of Brazil. Journal of the Gypsy Lore Society, First Series 1, 57-70. Tineo Rebolledo, J. (1900) A Chipicalli (la Lengua Gitana). Granada: Im-prenta de F. Gomez de la Cruz. Vallmitjana, Juli (1908) Sota Montjuic. Barcelona: ĽAvená. Wagner, Max L. (1937-1938) Stray notes on Spanish Romani. Journal of the Gypsy Lore Society, Third Series 15, 134-138; 16, 27-52. Webster, Wentworth (1889-1889) The Cascarrots of Ciboure. Journal of the Gypsy Lore Society, First Series 1, 76-84. ROMANI LEXICAL ITEMS IN COLLOQUIAL ROMANIAN CORINNA LESCHBER Free University, Berlin 0. Introduction The subject of Romani borrowings in Romanian generally touches upon different areas of past and present Romanian colloquial speech. It concerns the language of the youth (school children and students) as well as that of soldiers, historical argot and contemporary slang, so-called vulgar language, and the language of newspapers in Romania from the end of the 1800's and into the 1930's. In this contribution I deal with the adaptation of Romani words into Romanian and their semantic developments and stylistic changes. I focus on the question which Romanian words originating from Romani are still in use today, by whom, and with which semantic content, and attempt to find out whether the users of these borrowings are conscious of their origin, and whether the words they use are applied only in certain circumstances. 1. The status of Romani in related literature Sfirlea (1989) points out that there is a regrettable gap in Romanian sociolinguistics: there is no analysis of the effect of linguistic contact between Romanians and the national minorities according to age groups. Steinke (1989) notes in his article on Romanian Sondersprachen (i.e. marginal or special languages) that there is no dictionary which reflects the current state of the Romanian Sondersprache. The lack of such a book is all the more noticeable in that the major dictionaries hardly even touch the Sondersprachen. In a ninety-two page essay, Graur (1934) deals with Romanian words which had their origins in Romani. He illustrates these words with examples taken from the literature and the press. One of the main sources drawn upon is the humorous weekly Veselia (i.e. "Happiness"), published since 1891. Graur points out that the authors of these articles and books tried to create a