Back to Main Category

Working Papers in Sociolinguistics No. 38
Southwest Educational Development Laboratory
Austin, Texas 1977
and in
Thomas A. Acton and Donald Kenrick (eds.), Romani Rokkeripen To-divvus: The English
Romani Dialect and its Contemporary Social, Educational and Linguistic Standing.
London: Romanestan Publications, 1984.  Pp. 89-122.

The Social and Linguistic Development of Angloromani

Ian Hancock

The ethnic speech of the Romanies in Britain, and of their descendants who have been sent, or who have voluntarily emigrated to other parts, of the English speaking world2, presents a singular problem in terms of its linguistic classification. This speech, which may be called Angloromani3, has been discussed by various scholars during the past century or so, although the earliest and most significant statements about the language appeared in the mid 16th century.

The hypothesis presented here is that Angloromani developed as a specifically reduced contact language within the first half-century after the arrival of Romanies in Britain4, between English and Gypsy outlaws. When the Romanies entered the British Isles around AD 1500, they found a very considerable population of native “sturdy beggars”, i.e. healthy and often skilled, though unemployed, men who were obliged to support themselves through petty crime because of the breakdown of the guild system. There were also a great many legitimate beggars, forced onto the roads because of the dissolution of the monasteries, which traditionally took care of the poor, and because of the ending of the Hundred Years’ War and the War of the Roses in the 15th century (Chamberlin). Because of the isolated situation of the British Isles, and because Romanies arrived there during the earlier European diaspora5 and have not had contact with any other non-British languages (including other dialects of Romani, the original inflected, non-reduced language) since, the contact variety has gradually supplanted the once more widely spoken, inflected sourceform. This has happened to the extent that today, only a tiny minority of the Romanies in Britain is acquainted with it6.

It may be fairly safely assumed that upon arrival in England, probably from northern France or the Netherlands-north German coast area, the Romanies had no prior acquaintance with English7, and that it was the inflected, non-reduced language that was spoken8. The first verified samples of inflected Romani from Britain, or indeed from anywhere in the world, is dated 15479, and shows little interference from English; Romanies had been in Britain for probably between fifty and a hundred years at this time (cf. footnote 4):

“Good morrow!”

Lach ittur ydyues!

“How far is it to the next town?”

Cater myla barforas?

“You are welcome to the town”

Maysta ves barforas

“Will you drink some wine?”

Mole pis lauena?

“I will go with you”

A vauatosa

“Sit down and drink”

Hyste len pee

“Drink, drink for God’s sake!”

Pe, pe, deue lasse!

“Maid, give me some bread and wine”

Achae, da mai manor la veue

“Give me some meat”

Da mai masse!

“Maid come here; a word with you”

Achae, a wordey susse!

“Give me some apples and pears”

Da mai paba la ambrell!

“Much good may it do you”

Iche misto!

“Good nightl”

Lachira tut!10

The first reference to Angloromani as a newly devised speech, distinct from the above, is found in Harman’s Caveat (1566 or 1567):

As far as I can learne or understand by the examination of a number of them, their language—which they terme peddelars Frenche or canting—began but within these XXX yeeres or lyttle above.

The argument supported in this discussion is based upon an interpretation of Harman’s statement by John Camden Hotten in his Slang Dictionary (1864) at pp. 5-12, which introduces the idea that Angloromani developed as a “compromise” language:

Harman, in 1566, wrote a singular, not to say droll, book . . . wherein the history and various descriptions of rogues and vagabonds are given, together with their canting tongue. This book, the earliest of its kind, gives the singular fact that within a dozen years after the landing of the Gipsies, companies of English vagrants were formed, places of meeting appointed, districts for plunder and begging operations marked out, and rules agreed to for their common management. In some cases, Gipsies joined the English gangs; in others, English vagrants joined the Gipsies. The fellowship was found convenient and profitable, as both parties were aliens to the laws and customs of the country, living in a great measure in the open air, apart from the lawful public, and often meeting each other on the same by-path, or in some retired valley, but seldom intermarrying, or entirely adopting each other’s habits. The common people too, soon began to consider them as of one family, all rogues, and from Egypt. The secret language spoken by the Gipsies, principally Hindoo, and extremely barbarous to English ears, was found incomprehensible and very difficult to learn. The Gipsies, also, found the same difficulty with the English language. A rude, rough and most singular compromise was made, and a mixture of Gipsy, Old English, newly coined words and cribbings from any foreign, and therefore secret, language, mixed and jumbled together, formed what has ever since been known as the canting language, or pedlar’s french; or, during the past century, st. giles’ greek [emphasis added].

In addition to the language-learning difficulties faced by each group, it is probable that, like the Chinese in the Canton Pidgin situation (Hall, 1966:8), the Romanies were content to have the English know the restructured variety but withheld knowledge of their own inflected language from them in order to maintain their separateness within the overall outlawed society.
A further reference from the same period as Harman turns up in a note by J. Harison in the 1578 printing of Raphael Holinshed’s Chronicles, in which he also suggests that the language was consciously created. Harison noted that the Romanies and the British outlaws

. . . have devised a language among themselves, which they name Canting, but others pedler’s French, a speeche compact thirty years since of English, and a great number of od words of their own devising, without all order or reason; and yet such is it as none but themselves are able to understand; the first deviser thereof was hanged by the neck, a just reward no doubt for his desartes . . .

The next important quotation is found in Samuel Rowlande’s The Runnegate’s Race (1610), in which the writer gives an account of the meeting between the leader of the English vagabonds, Cock Lorrel, and the leader of the Gypsies, Giles Hathor a little over a century earlier,

. . . to parle and intreete of matters that might tend to the establishing of this their new found gouernment; and first of all they think it fit to deuise a certaine kinde of Language; to the end that their cousenings, knaueries, and villanies might not be so easily perceiued and knowne in places where they come.

This statement agrees with the findings of Harman who, as already indicated, believed this distinct form of speech to have developed separately from the original inflected variety initially brought into Britain, and not to have existed earlier than, about 1535.  Samuel Rid, writing in 1612 about the Romani population there at that time, noted

. . . many of our English Loyterers joyned with them, and in time learned their crafte and cosening.  The speache which they used was the right Egyptian language, with whom our English men conversing, at least learned their language.

This last statement is significant since it may be argued that so few items actually (or possibly) traceable to Romani occur in the early vocabularies of rogues’ slang—Hotten himself lists 26 in his discussion, and they are almost non-existent in Grose’s standard work on the subject—that the 16th and 17th century writers may have been referring to Cant rather than to Angloromani, since the two are sometimes confused: Ainsworth, in the preface to his Rookwood (1834) speaks of “ . . . the practiced patterer of Romany, or Pedlar’s French”. McPeek (1969:38-89) has most recently supported this view:

Gypsies are credited with assimilating the jargon of their beggar rivals without adding anything of their own to it . . . and pedlar’s French is similarly free, almost untouched by Romani. Furthermore, though the Gypsies themselves employed pedlar’s French in their activities, they maintained their own language essentially free of cant elements.

In his extensive six-volume London, Charles Knight (1841, I:154) made the same point:

It is a common misconception to confound this cant phraseology of our ordinary thieves and beggars, consisting of a few peculiar terms and modes of expression mixed with and engrafted upon the language of the country, to the grammatical forms of which it is entirely accommodated, with the wholly distinct and foreign speech of the Gypsy people. The latter is another language altogether, having as little connexion with the English as the Hindostanee has, to which indeed, or to its fountain-head, the Sanscrit, the Gypsy tongue appears to be nearly allied.

He goes on to attribute the cause of this confusion to such works as Bamfylde Moore Carew, Guy Mannering and others.

Lehmann (1983:214) is guilty of the same; he provides a text in “18th Century gypsy” language in his book on linguistics, but with the exception of the word pal, it contains no actual Romani.  The confusion of identity and speech between ethnic Romanies and other non-Romani traveling populations was already well established by the 1700s:

In a box of the stone jug I was born,
Of a hempen widow, a kid forlorn—fake away!
And my father, as I’ve heard say—fake away!
Was a merchant of fakers gay,
Who cut his last fling with great applause,
Nix, my doll pals, fake away!
To the tune of hearty choke with caper sauce—fake away!
The knucks in quod did my schoolmen play—fake away!
And put me up to the time of day,
Until at last there was none so knowing,
No such sneakesman or buzgloak going.

It should be noted however that the secret nature of Angloromani would have kept knowledge of its lexical and grammatical characteristics, if not its actual existence, from outsiders; cf. the case of Shelta (Hancock, 1974, 1984), which is probably of even greater age than Angloromani, but which was unknown to the outside world until 1876.
The words in English slang which have been treated in the literature (Baumann, Leland, Hotten, Matsell, McPeek, Turner, &c., and see especially Grant, 1998) as possibly being of Angloromani origin are listed here, although the majority of them are more likely to be from a non-Romani source:

bamboozle, bandy, bar, bazaar, berk, bite, bivvy, bloke, blazon, bosh, cadger, can, caroon, cat, cheese, chin, ohiv(e), chivvy, choor, chull, churl, cock, cockal, cocker, conk, cory, cosh, cove, cur, cushy, cutter/couter, dad(dy), dando, dik/dekko, dig, divvy, dook, drum, duffer, gad (about), gadgie, gaff, gibberish, gorger, gybe, hook & snivey, hookey, jalopy, jerry, jib(b), jockey, jomer, keister, ken, lil, loafer, lobs, lolly, lollipop, lour(e), lunan, lush(y), mammy, maund/maung, mindge, mockers, moke, moll, moo, moosh, mo(r)t, mug, mull, muller, mumper, muns, naff, nark, niggle, pal, pandgie, parn(e)y, peck, peels, pen, posh, punch, raclan/racklaw, rig, rocker, Romany, romy/rumy, row, rum, Salomon, sap, saulohomus, shaver, shero, shiv, slang, snack, stir, stur(a)bin, tanner, tiny, toff, trash, vocker, wasted, wongga/vongga and yack.

No written samples of Angloromani, and until 1780 no more of the inflected variety in Britain11 turn up until the 19th century. Borrow, in his book on the Spanish Gitanos (1841) refers briefly to Angloromani, and in his two later books, Lavengro (1851) and its sequel The Romany Rye (1857) he includes some discussion and words and phrases passim. In 1864 Hotten, mentioned above, discussed the formation of Angloromani, and nine years later, in 1873, Leland’s important English Gypsies and their Language appeared, followed a year later by George Borrow’s Romano Lavo-lil, though it should be stressed that these last titles contain a certain amount of fanciful linguistic material.

Smart & Crofton (1875), an enlargement of an earlier treatment by Smart (1860), remains the most comprehensive description of the inflected dialect as far as it was retrievable in mid-19th century England. This devotes pp. 273-278 to texts in what is called the ‘New or Broken Dialect,’ i.e. Angloromani.

The first passing comparison with a pidgin was made by Leland (1876) in his book on China Coast Pidgin English. In a brief statement on the second page, he note that the Chinese Pidgin “ . . . corresponds exactly with the posh and posh or corrupt Rommany dialect spoken by English gipsies, in which Hindi-Persian words follow the English structure.” So taken with the nature of Romani was he, that he even included a poem in the language in his Hans Breitmann Ballads (1889:46), a book of verse in German-English dialect. The first scholar to treat Angloromani as a special case in any depth was James Cresswell Clough, in his On the Existence of Mixed Languages (1876), in which he discusses, among others, Chinook Jargon, Sranan, Jamaican Creole, African American English and the Lingua Franca. He devotes a number of pages to “Gypsy or Romanny,” introducing a further spelling for the word and concluding (rather rashly, since the fully-inflected language survives quite healthily with several million speakers) that

Romanny grammar has been almost entirely destroyed in the contact with Europeans. Thus the English Gypsy makes feminine and masculine words to agree with one another indiscriminately, the first step towards the extinction of grammatical and towards the adoption of natural gender. Though he uses his own plurals, he already forms all his cases by means of English prepositions, instead of Romanny inflexions; and in the conjugation of the verbs he as often uses the English as the correct system, saying I del I give, instead of delo; I del’d, I gave, instead of deliom, and if I had del’d If I had given, instead of deliomis.  Romanny is, therefore, mixed in vocabulary, pronunciation, and grammar.

As a side note, mention may be made of a privately printed booklet by one Hall Desmond, entitled The Language of the Gypsies (ca. 1890). Unless this was a pseudonym for Clough, Desmond’s publication must represent the most blatant plagiarism for, with the exception of the final paragraph, it is identical in every detail with Clough’s earlier statement. Neither of these sources is listed in Black’s Bibliography.

A significant suggestion was made in 1887 by Ribton Turner who, in a discussion .of Rowlande (1610) said that he believed the Romanies to have arrived in Britain speaking a ‘perfect language’, i.e. inflected Romani, but rather than having created a contact language after arrival, simply stimulated the British outlaws to embellish an already-existing cant with words from Romani.  This is the position I take myself. Turner identifies the speakers of this cant as the Gwestwr12 or vagabonds who were considered a public nuisance in England prior to 1330, and who were usually of Welsh or Irish origin. He says (1887:467):

It is probable that after the arrival of the gypsies in England, about the year 1505, the example of their language, which was perfect in itself, stimulated English vagabonds to polish and improve their Cant so as to make it a current medium of speech, but no single individual has ever yet invented a spoken language, and cant words must have existed long before the time of Henry viii.

Once again, in 1889, Leland took up the question of the origins of Angloromani in his preface to Barriere & Leland’s Dictionary of Slang, Jargon and Cant. In a fairly comprehensive discussion of the origin and dispersion of the Romanies (pp. v-xxiij) he mentions some of the 16th century writers, and refers to the possibility that Angloromani was consciously devised.

Some forty years later, the noted linguist Leonard Bloomfield (1933:471-475) discussed the linguistic characteristics of American Angloromani, and while not committing himself to a typological classification of the speech, he includes in the same section references to foreigners’ mixed speech, Tok Pisin, Chinook Jargon, Afrikaans and Sranan, and regarded the use of non-English words in a “normal variety of sub-standard English” as being a conscious practice. Reinecke (1937:77-78), basing his discussion on Bloomfield, included Angloromani in his monumental dissertation on marginal languages, but unlike Bloomfield felt, correctly, that “the Gypsies do not view their speech as a facetious jargon [Bloomfield’s words], but as a true language.” In the same year, Mencken (1937:696-697) included Angloromani as the last entry in the appendix, “Non-English dialects in America” to his The American Language, but adds nothing to the earlier sources.

The first linguist actually to refer to Angloromani as a creolized language was Gray (1939:38). His observations were derived wholly from secondary sources and not at first hand, and as a result he over-estimates both the extent of anglicization and the proportion of Romanichals (Angloromani speakers) to other Romani groups in the United States:

. . . among the English Gypsies . . . we can trace the steady degeneration of a language with its own systems of grammar and vocabulary into a creolised dialect with Gypsy words and an extremely sub-standard English. Finally, among the great majority of English and American Romani, the language has become an English approaching more and more nearly the standard type, with an ever-diminishing number of Gypsy terms, many of those which survive serving chiefly as cant to prevent non-Gypsies from understanding what is being said.

Gray’s position regarding the development of the language was clearly that which was later taken by Kenrick (below), and perhaps by McPeek also, i.e. that the restructured contact variety represents gradual ‘degeneration’, to use Gray’s own word, and that this is a continuing process.

The next reference to the language, and the first specifically to discuss Gray’s position that it developed as a creole, was Hancock (1970), which interpreted Hotten’s discussion, repeated above, from a creolist’s point of view (concluding that it was not a pidginized or creolized speech; see Hancock, 1985:89). This appeared in the Journal of the Gypsy Lore Society that year, and was rebutted a year later by Kenrick (1971), A brief rejoinder to Kenrick by Hancock appeared in the same volume (Acton, 1971). Since that time, Hymes (1971:67) has made reference to Angloromani, allowing for the possibility nevertheless that “the uninflected variety of Anglo-Romani should be defined as pidginized. That is, [it shows] simplification in outer form, reduction in inner form, and restriction in role, together with a confluence of traditions . . .”, and the language has been given a section of its own in Reinecke’s Bibliography of Pidgin and Creole Languages (1975:622-631).

To summarise Kenrick’s arguments, he sees Angloromani as being of comparatively recent development, i.e. as having crystallized within the past 100-150 years, and as having come about because of the gradual loss of command of the earlier inflected language, possibly because of intermarriage with non-Romanies. He further treats it as a register of English, and compares it with Jewish English, that is, as being English with “exotic” lexemes in the syntactic slots. Kenrick’s point of view is based primarily on the documented evidence of the gradual loss of the inflected language among the informants of various investigators over the years up to the present time. He quotes in particular the statement of an old man who in 1863 told Smart and Crofton that when he was a young man, the old people spoke tacho pooro Romani lava, “real old Gypsy words”. Although the old man may have been referring to morphology rather than lexicon, since his own vocabulary here is entirely Romani-derived (but had he adhered to native morphology he would have said tache purane Romane lava) , the argument does not detract from the point being made here. Kenrick also states (1971:7) that it was a popular notion during the 16th century that Romanies weren’t a discrete people, but merely a “made up” group consisting of Englishmen and Englishwomen who dyed their skin, and that accounts of their “made up” language have to be interpreted with this in mind. But while it is true that references to this do occur, probably initiated by Dekker, Romanies were nevertheless real in renaissance England, dyed or not, and there is no reason to suppose that Angloromani was any the less real; cf. Alvarez’ statement in The Spanish Gipsie by Midleton & Rowley (1653), where he speaks of

Gipsies, but no tanned ones; no red-ochre rascals umbered with soot and bacon as the English gipsies are.

In my reply to Kenrick (Hancock, 1971), I disagreed with his comparing Angloromani with Jewish English typologically. This was perhaps precipitous of me in terms of language type, since both are similar in their composition, i.e. each consists of English with lexical inserts from another language, but their respective uses overlap only to an extent, and the non-English words in Angloromani far outnumber those in Jewish English; furthermore the latter only admits intrusions from open-list items; Angloromani retains some non-English prepositions, pronouns and conjunctions13.  In terms of perceptions of each register too, Angloromani is called Romani (Rumnis) or Broken Romani by its speakers; Yiddish English isn’t called Yiddish or even Broken Yiddish.  What complicates the situation is that for most of their speakers, neither speech community now has access to “full” Romani or to Yiddish to draw upon to supplement their non-English components, and those Romani and Yiddish words that are used are gradually being lost.  When their number dwindles to just a dozen or so, even calling either register an ethnolect becomes scarcely justifiable.  Nevertheless when Angloromani loses a non-English item, it can and does compensate, for example calling “sand” lŏn-chĭk (“salt” + “dirt”) because the original word (chishai) has been lost.  Yiddish does not do this.

In summary, there are sufficient references from the 16th and early 17th centuries to a contact language used between Romanies and non-Romanies for it to be possible that Angloromani became distinct from Romani during that period. That there are no texts from that time is not remarkable in view of the secret nature of the language; in fact only one verified text in inflected Romani itself is known to us from that date.   It is clear that English—and hence Angloromani—had already become the dominant language of much of the Romani population in Britain at this early date, since the phonological changes accompanying the shift from Early Modern to Late Modern English) also affected the Romani words it contained.  One such sound shift for example was that of [er] to [ar] (cf. sartain, varsity, sergeant, clerk, &c., now mostly back to [er] again), thus the original [er] in Romani erti- “forgive, excuse” (< Romanian) became ātav- in Welsh Romani, at the same time demonstrating loss of postvocalic /r/; [er] in verdo “waggon” (Common Romani vurdon) becomes Angloromani vadə or vâwdə, and terno “young” becomes Angloromani tâwnə with loss of /r/ and shift of [a] to [aw], cf. Welsh Romani tarno, Common Romani terno.  This latter, which is characteristically British Romani, reflects yet another soundshift which began in South-Eastern (but not South-Western) English in the early modern period (between ca. 1500-1700), cf. [græs], [hæf], [læf] to [gra:s], [ha:f], la:f]; also cf. southern Angloromani gâwjə, yok, yog, yorə (“non-Romani”, “eye”, “fire”, “egg”), &c., with northern (especially Scottish) gadji, yak, yag, yaro, &c., and Common Romani gadjo, yakh, yag, a(n)ro.

It appears that over the centuries Angloromani became more current among Romanies in Britain than did the inflected language. For families today, English is the first language learned in life (hence Angloromani being referred to as an ethnic rather than as a native language here). In some of these families, as Kenrick points out, (1979), children do not achieve fluency in it until their eighth or ninth year. Thus Angloromani has gradually become the ethnolect of the community, while inflected Romani has been retained alongside it only by especially conservative families in dwindling numbers (particularly in Britain) in songs and proverbs, &c., or by isolated groups, at least into the mid 20th century, such as the Lockes or Woods in central Wales.

There are a great many Travellers of non-Romani ancestry in Britain and the United States who are able to an extent to speak Angloromani, and the same situation is found in Scandinavia.  Johansson, in his description of reduced Swedish Romani (1977:5) says of the ethnic composition of its speakers that while a Romani element is discernible, its ro1e has certainly been exaggerated in earlier treatments (“dess roll har tidigare säkert överdrivits”). The same is true in Spain, where Caló has been acquired by Quinquí and other non-Romani groups. The situation is paralleled by the Afro-Seminole case, (Hancock, 1978), where African fugitives in Florida transmitted their English-derived creole, Gullah, to the (mainly) Creek Indians, and in turn acquired Creek identity and culture, yet each population remaining genetically distinct despite their living together for over one and a half centuries.

Angloromani cannot be regarded as a pidginized language in the light of e.g. Taylor’s 15 criteria (1971:295); it retains a fair degree of productive morphology, albeit deriving almost entirely from English, and it lacks many of the linguistic features generally associated with pidgins. It did, however, originate from a contact situation, and does conform to Hymes’ requirements, quoted above. And in particular, it exhibits very extensive internally generated innovative lexicon, perhaps far more extensive than that found in actual pidgins and creoles; and to a lesser degree, morphological devices originating in the source language (Romani) have been readapted, and continue to be productive in ways not paralleled in either English or Romani. Nor may Angloromani fit Todd’s (1975) or Platt’s (1975) definition of a “creoloid”, perhaps coming closer to the Shelta (Hancock, 1974) or Mbugu (Goodman, 1971) situations, for which adequate linguistic descriptions are still lacking.

Similar linguistic situations elsewhere 

The Caló (Hispanoromani) language in Spain and parts of South America, and in Portugal, where it is known as Calão, has already been likened structurally to Angloromani. Many works have appeared dealing with this language, mostly in the form of wordlists (e.g. that edited by Christof Jung, 1972; see also Bakker, 1995). The most recent lexical study of Caló made at first hand is that of McLane (1977), while Wagner (1938) discusses lexical expansion in this dialect, and gives examples which have clear parallels in Angloromani; thus words are created by punning, e.g. loriazo “March (month)”, Spanish “marzo” (<Caló loria “sea” < Persian lor + Spanish “mar” + -azo) , or calquing, e.g. barbalo “elegant” (< Caló barbal “air” < Sanskrit vatavat,  + Spanish “aire”, + -alo), Lillac “Thomas”, Spanish “Tomás” (< Caló lillar “take” < Sanskrit labh- + Spanish “tomar”), or Molancia “Valencia” (< Caló molar “be worth” < Sanskrit mulya + Spanish “valer”), &c. The early social situation was also surprisingly similar to that in sixteenth century Britain; according to Bercovici (1929:170-172),

Gypsies came to Spain at the beginning of the 15th century. Official documents prove that they were already in Castile in 1490 . . . they joined the wandering calabrezos living in Spain, who may or may not have been calabrezos, but were kettle-makers, iron-workers and nomads. These various groups of calabrezos and Gypsies intermingled so that the Gypsies lost their identity, and were also called ‘calabrezos’ by the people, and referred to as such in various documents. The edicts and laws of the time speak of them as “calabrezos and caldareros . . . people who pretend they are Egyptians and speak all languages.”

Intermarriage with the Moors further increased the non-Romani element in the Gitano population: “it is this Moorish blood in the Gitanos which makes them look so different from the other Gypsies, and which has in so many respects so changed the character of the people” (loc. cit.).

In Scandinavia the similarities are even greater, and indeed may share their early historical and linguistic development with the English situation.  The first record of a Romani presence in Denmark indicates that they had been banished to that country by James IV of Scotland in July, 1505. Their arrival in Sweden via Denmark is dated 1512, and they were being abandoned on the Norwegian coast from 1544 onwards, in an attempt to rid Britain of their presence.  Scandoromani is discussed in Hancock (1992).  A typologically similar Hellenoromani variety called Dortic has been reported from Greece (Triandaphyllidis, 1924; Anon., 1930).

A Description of Angloromani

Angloromani is spoken in various forms along a continuum, which is lexical rather than grammatical. While speakers have the notion of “deep”, i.e. basilectal speech, this refers more to the use of esoteric vocabulary than to a knowledge of non-English grammar. In some families, however, where the inflected variety remained in use longer, certain frozen (non-analyzable) inflections may still be encountered. This is generally true for the language in songs, oaths, &c., for all such speakers of Romani ancestry. Non-Romani speakers of Angloromani are not as a rule familiar with deeper forms of the language, or with many of the folkloristic aspects of Romanichal culture. A typical speaker of deep Angloromani might, for example, use /yŭv/ instead of “he”, /lĭs/ instead of “it” in object position, /mŏrə/ instead of “our” and so on, in the pronominal system, items which have generally been replaced by their English equivalents.

Discussions of British Romani by Ackerley, Winstedt and others in The Journal of the Gypsy Lore Society have predictably concerned themselves with describing the speech of those individuals having some knowledge of the traditional inflected language; and while that has also been called “Anglo-Romani”, when spelt that way in the present study it refers only to that now defunct variety (see footnote 1).  Where Angloromani speakers come into contact, but don’t as a rule socialize, with speakers of inflected dialects of Romani, their respective languages are scarcely recognised as being even lexically related14, and Vlax Roma are commonly referred to as “Turks” or sometimes “Rooshians” or “Roos” by American Romanichals.

Since there is no regularized orthography in widespread use for Angloromani, English spelling conventions normally being used (but see Macfie, 1910), and since it is necessary here to distinguish between English- and Romani-derived morphemes, the former are written here in italic script, and the latter in a modified phonemic spelling [NOTE: The original format of this paper used the symbols of the International Phonetic Alphabet.  Since these won’t reproduce on all computers, I have modified that orthography and use only those symbols available in the Latin-extended-A(Multinational-A Roman) insert set in Microsoft Word, as follows:]


as in CALM (kam)


as in COT (kŏt)


as in HAT (hăt)


as in COAT (kōt)


as in AISLE (ail)


as in OIL (oil)


as in SAUERKRAUT (sauəkraut)


as in PUT (put)


as in PAW (pâw)


as in JUNE (jūn)


as in STEP (step)


as in CUP (kŭp)


as in THEY (dhey)


as in THING (thĭng)


as in AMERICA (əmĕrikə)


as in THAT (dhăt)


as in HIT (hĭt)


like the CH in LOCH NESS


as in HEAT (hīt)



In American Angloromani, vowels have their North American English values.

Phonological approximation to English

Angloromani phonology is English phonology; the only non-English phoneme to remain for some speakers is the uvular fricative [x] (Loch Ness = lŏx nes) and even this is common in Scots dialects of English. Generally, however, historic [x] corresponds to [h] or [k] in Angloromani:

hŏbən, hŭkəno (< Romani xaben, xoxano) “food”; “deceitful”

The two /r/ phonemes of Romani, the flap [r] (as in Scottish English) and the voiced uvular fricative (like a French or German ‘r’) had probably coalesced into a single flap phoneme in the northern Continental dialects before the language was taken into Britain; they are not distinguished in German Sinte, for example. In Angloromani this has become a roll [r] or, more rarely, an alveolar stop [d]:

bâwrə,  churī~chudī (< Common Romani baro, čhuri) “big”; “knife”

In Britain, rhotic Angloromani is commoner than non-rhotic Angloromani, even in non-rhotic parts of the country.  Thus ker (“house”) rather than keəAmerican Angloromani is uniformly rhotic, even admitting intrusive (non-historical) R’s in some words: gâwrjər, vâwrdər “non-Gypsy,” “wagon” (gâwjə, vâwdə in British Angloromani).
In Angloromani, the phonetic distinction between aspirated and non-aspirated stop consonants has been lost, or else has become a voiced/voiceless contrast:








ker, kər

“make, do”






“brightness, light”

Grammatical features


Most nouns follow the regular English system of pluralization, and have lost the distinction of gender and case found in inflected Romani:





a juk

a dog

dui džukle

dūi juks

two dogs

i džuklesa

wi’ the juk

with the dog

For a number of items, an historic plural /- (y)âw/ may occur, sometimes in combination with English {-s} ([s]~[z]~[iz]) and sometimes weakened to /-(y)ə/:


a yŏk

an eye

dui yakha

dūi yŏkâw/dūi yŏkəs/dūi yŏksx

two eyes

This is the Common Romani plural -a, which occurs most commonly with consonant-final and feminine subject-case nouns; sap, sapa “snake(s)”, rakli, raklia “girl(s)”. This termination in Angloromani shows the typical retracting in that language of historic /a/ to /âw/, unstressed /ə/.

It may apply to nouns historically belonging to another plural class:













The Romanian plural –uri, common in the Vlax dialects, occurs with a few athematic items in Angloromani, e.g. rushərī “rushes,” carrotərī “carrots”.

The possessive construction usually takes ’s (s’), though in some of the northern counties of England a zero genitive was formerly found, e.g. in the speech of some of the Bosses, Hearnes and Youngs:

dulə’s mandi’s

those are mine

Petey dad ker’s dōstə bâwrə

Petey’s father’s house is very big

Such unmarked possessives are restricted geographically, and are not typical of the language.

There remains in common use a productive nominalizing morpheme -bən, of Indic origin, which may join with verbs, adjectives, &c., to produce nouns:









A less common nominalizing suffix is -məs (< Greek -:`l):

hoinəs, h>wnəs





make, do


doings, action

In the inflected language, the genitival endings for masculine and feminine nouns before both singular and plural objects, are as follows:


“man’s” (m.)


“woman’s” (m.)


“man’s” (f.)


“woman’s” (f.)


“men’s” (m.)


“women’s” (m.)


“men’s” (f.)


“women’s” (f.)


“man’s” (m. & f. pl.)


“woman’s” (m. & f. pl.)


“men’s” (m. & f. pl.)


“women’s” (m. & f. pl.)

In Angloromani, derivatives of these constitute another set of productive suffixes, though they may be applied to items that historically are of either gender or number, and in addition they have no possessive function. In British Angloromani they have the forms  -(m)engrə  -(m)engrī,  -(m)eskrə, -(m)eskrī, - (m)engrīs and -(m)eskrīs.  In American Angloromani these have generally collapsed to -mengər(s):

















Sometimes these different endings serve to distinguish different items sharing the same root morphemes:



(< del “give”)








chef, cook

(< hŏbən “food” < hâw “eat”)









(< lĭvənə “beer”)








alcoholic (person)

(< pī “drink”)




Occasionally there are even triplets formed in this fashion:


boxing glove

(< kur “hit, strike”)


boxer, fighter



This does not seem to be productive in the contemporary language, where kuvər “thing” is its equivalent:

dĭkin’ kuvər

window (“looking thing”)

dūd kuvər

lamp (“light thing”)

papĭn kuvər

gamekeeper, forest ranger (“goose thing”)

tăvin’ kuvər

cigar (“smoking thing”)

Shift of class is evident for a large number of items employing English-derived morphemes also:



(< brishen “rain”)

smenty tud

“creamy milk”

(< smentnə “cream”, + tud, “milk”)


All distinctions of gender and case have been lost in Angloromani. This may be illustrated with the first person singular personal pronoun:




1 me



2 man



3 maya


me (emphatic)

4 mandi


me (prepositional)

5 maŋi

to, for măndī

to, for me (dative)

6 mansa

wi’ măndī

with me (instrumental)

7 mande

from măndī

from me (ablative)

8 miro


my (masculine singular)

9 miri


my (feminine singular)

10 mira


my (feminine oblique)

11 mire


my (plural and masculine oblique)

Inflected Romani has both prepositions and postpositions. The former e.g, əprey “up”, təley “down”, pâwdəl “across”, &c.), have been retained in Angloromani, while the latter have been replaced by English prepositions, with the possible exception of -sa “with”, which may, according to Smart & Crofton (1875:133) and Borrow (1874:58) function prenominally. This has not been heard by the present writer.

As well as losing postpositions, Angloromani has also lost the reflexive possessive third person singular and plural pronouns which in inflected British (i.e. Welsh) Romani are pesk- “his/her own’ and peŋ- “their own”, as opposed to lesk- “his”, lak- “her”, leŋ- “their”.

The pronouns which remain in most common use are măndī, first person singular, and tūtī, second person singular, and for some speakers also the second person plural. Other pronouns may be heard less frequently, and include mīrə, mirī, mi, mī or mŭ “my,” yŭv or lĕstī  “he,” yoi or yūwī “she”, lĭs “it”, mĕndī “us”, “we”, mâwrə “our” and ledi(‘s) or lendi(‘s) “they” (“their”). These derive mostly from inflected oblique (i.e. non-subject) forms of the pronouns in the historical source.

Possession is indicated for the invariable forms with {-s}:

tutī’s ruzlĭbən an’ măndī’s gŭzələs  “your strength and my brains”


Romani verbs belong to four classes, and are inflected for person, tense and number. In Angloromani the morphology is English, the native stem often corresponding to the third person singular present indicative in the inflected language.  There is historically no infinitive in Romani, but in the Central and Northern dialects the third person singular has come to assume this function, cf. Czech Romani kamav te džal “I want to go”, kamav jov te džal “I want him to go.”





măndī dels

“I give”


yŭv, yoi dels

“he, she gives”


tutī lels

“you take”


yŭv, yoi lels

“he, she takes”


we jăls (jels)

“we go”


yŭv, yoi jăls (jels)

“he, she goes”

Sometimes the surviving form is the historic imperative:


 măndī chĭns

“I cut”


 măndī mŏngs

“I beg”


 măndī prăstərs

“I run”




The present indicative takes {-s} throughout:

(tu) rakeres

tutī rŏkərs

“you talk”

(yov) džanel

he jĭns

“he knows”

(me) kamav

măndī kŏms

“I like; I want”

(amen) dikhas

we dĭks

“we see”

The preterite is formed with {-ed}:

(yon) maŋle

ledī mŏnged

“they begged”

(yon) lasterde

they lŭstərd

“they took an oath”

(yoi) garadas les

she gărĭv’d it

“she hid it”

Past participle verbal forms survive in Angloromani, but are not consciously related to their present tense form, functioning only as adjectives:



“buried” (< prăv “bury”; prăstə also means “run”)



“broken” (< pŏgər “break”)



“intoxicated” (< pī “drink;” influence from “pickled”?)

Note too



“wet” (< brishən “rain”?  Common Romani “wet” is kindo)

The BE verb is usually English, although for some speakers the singular Romani forms remain in use: shŏm “(I) am”, shăn “(you) are”, sī “(he) is”. The third person sī often serves for all persons, however, a basilectal speaker might say shŏm a shərengrə “I’m a boss”, or perhaps măndī sī a shərengrə, although in the inflected language, this would mean “I have a boss,” and would be interpreted in that way by a diminishing number of speakers. Most commonly, the English verb is used: măndīs a shərengrə.
Use of the BE verb to express obligation is still heard from some people but seems to be losing ground in favour of must or got to: shŏm tə jel sĭg “I must go soon”, lit. “I am to go soon” (tə here may be traceable either to English to or to Romani te).

For the past tense, the regularized third person singular has been taken from the inflected language:




(me) somas

măndī săs

“I was”

(tu) sanas

tutī săs

“you were”

(yov) sas

he (yŭv) săs

“he was”

(ame) samas

we (mendī) săs

“we were”

(tume) senas

you (tutī) săs

“you were”

(yon) sas

they (ledī) săs

“they were”

Other verbal constructions follow the English model:

šundomas len

măndī’s shuned ’em

“I’ve heard them”

dikhlan les?

did tutī dĭk it?

“did you see it?”

yon marena len

ledī’ll get mâwrd

“they’ll get killed’

yon munjerde i Luther

Luther had been chĭv’d

“Luther had been cheated”


tutī was chĭnin’

“you were cutting”

hagasa i gores

we’re gonnathe gerə

“we’re going to kid the man”




BE is optionally deleted:

akai si tire patrinya

əkai tuti’s păterəns

“here (are) your pages”

Sinclair (1915:737) gives examples of this from American Angloromani (his orthography):

tūte kūšta diken muš

“you (are a) fine looking man”

tūte rinkna râne

“you (are a) pretty lady”

Sinclair also lists hi as a variant pronunciation of sī “is,” which is characteristic of the northern dialects of Romani and from which Angloromani is derived. This no longer seems to occur in Angloromani in Britain, nor has it been encountered by the present writer in the language in the United States15.


The commonest negating morpheme is kek, with the variants kekə(r), kăk, kăkə(r). There is in addition a negative imperative mâw “don’t,” thus

măndī can kek ker lĭs

“I can’t do it”

tutī kek’ll be able to dik the mush from əkoi

“you won’t be able to see the man  from here”

mâw rŏkər, you dĭv, the gâwjər’s shunin’

“don’t talk, you dummy, the non-Romani is listening”

kekər! măndī’ll kər it mĭ kŏkrə

“no! I’ll do it myself”

Also “no,” “not” with nouns:

there’s kekə pani left in kŭvə kurī

“there’s no water left in this bucket”


These conform to the English pattern, comparatives and superlatives taking –er and –est if the adjective is of one or two syllables, and more and most if longer:






"more intelligent"

o hučikaneder


"most intelligent"


wafədī (văsəbəl in the US)



more wafədī


o basaveder

the most wafədī

“the worst”

Adverbs derived from adjectives retain the adjectival form. Some items contain the frozen native adverbial ending -ĭs but this is no longer productive in Angloromani.

må ker lubenes

stall actin’ lubnĭs

“stop acting wantonly”

(yov) rivel pes romanes

he ridəs rŭmnĭs

“he dresses Gypsy-fashion”

The English -ly is commonly found:

(yoi) giyel but rinkenes

she gĭllies right shukəly

"she sings very sweetly”

ja sig

jâw sĭg, jâw sĭg(ə)ly

“go quickly”




Than is expressed by [an] (< than), or more rarely by [nə] (< nor)

kuvə's much more shĭlənə nor Luther’s tan         "this is much colder than Luther's place"

Reduplication is employed for emphasis, as indicated a century ago by Smart & Crofton (p. 48), thus:


"soon", "quick"


"very quick"





dūvrī, dūvərī



"very distant"


These are kŭvə or kuvə "this" and "these", and duvə "that” and "those". Some speakers retain a form dulə for "those". The demonstratives are frequently used in combination with the adverbs əkai/əkoi "here" and ədoi "there": kŭvəkai "this here,” duvədoi “that there”:

kŭvəkai’s lesti’s                     “these are his”

Miscellaneous grammatical items

Although Angloromani is a register of English rather than of Romani, it is characterized by an untypically high rate of retention of grammatical (as opposed to lexical) items from the latter language.  As stated above, they are practically non-existent in blend-registers such as Yiddish English, though a similar distribution is found in such languages as Chinook Wawa, Michif, Media Lengua, etc.










“before” (< French avant que)


“all;” “which;” “what”

kŏn, kŭn





“as;” (relative) “that”

ta, tə, ən(d)







It is in its internally-generated vocabulary that Angloromani is most like the Creoles17. Speakers have shown remarkable inventiveness in keeping the lexicon as free from English (or recognisable English) as possible, the reason being that since the structure already derives from that language, use of English vocabulary too would result in a language which would be English itself. Thus, even where items in the source language did not exist, e.g. “aeroplane”, “television”, &c., equivalents for these have been produced from the existing morpheme stock to expand the lexicon. This is also true for items which, although occurring in the source language, have subsequently become lost in Angloromani, e.g. the words for “gun”, “wheel”, “sand”, “peace”, &c.


Incoining is the process whereby new lexemes are created from morphemes already available in the language, the combination or interpretation of which has no exterior model (in this case, in either English or Romani). Incoining differs from coining, which is the creation of new items with no historical source, for example ‘Kodak’ or ‘blurb’ in English. Incoinings current in Angloromani include:


“barber” (lit. “hair cutter”)


“hook” (lit. “crooked needle”)


“pepper; mustard” (lit. “bite” + -mengrī)

dĭkin’ muktə

“television” (lit, “looking box”)


“telescope” (lit. “far” + “see” + -mengrə)


“bee” (lit. “sweet + flea;” see below)


“turkey” (lit. “red lady”)


“toffee-apple” (lit. “stick” + “red” + “apple”


“copper” (lit. “red tin”)


“carrot” (lit. “red root”)


“sand” (lit. “salt dirt”)

năshin’ chĭnəmengər

“telegram” (lit. “running writing”)

năshin’ pâwnī

“waterfall” (lit. “running water”)


“electric lights” (lit. “new lights”)


“boat” {lit. “water waggon”)


“scarf” (lit. “half” + “neck”

prăstərin’ săstə

“bicycle” (lit. “running iron”)


“plough” (lit. “ground’s waggon”)

rukengrī-ījəs  or ~chŏkəs

“leaves” (lit. “trees’ clothes”)

rŭmnī-rŏkərin’ chərklə

“parrot” (lit. Romani-talking bird”)


“telephone” (lit. “speak” + “hear” + -engri) UK

shūn ’n puch

“telephone” (lit. “listen” + “and” + “ask”)  US


“aeroplane” (lit. “iron bird”)


“locomotive” (lit. “iron horse”)


“orange” (lit. “sour apple”)


“liquor” (lit. “hot water”)


“cigar” (lit. “smoking thing”)


“cocaine” (lit. “bad” + “medicine”)


“reefer” (lit. “bad” + “smoke”)

The same semantic generalization evident in some of these examples is also found in dialects of inflected Romani in the United States, cf. Angloromani pŏbə (also pŏp, pŏbī) “apple”, used in the sense of “fruit”, chĭk “dirt” to mean “gasoline,” or vâwdə “waggon”, used in North America to mean “vehicle” generally.  A good example is gudlə-bĭshəm for “bee,” which contains two semantic shifts: gudlə means “sweet” in Common Romani (gudlo), but means “sugar” in Anglromani, while bĭshəm is from Common Romani pišom “flea,” generalized in Angloromani to mean “insect,” hence “bee.” Similarly in American Vlax mečka, originally “she-bear”, may be applied to any large animal, such as a lion, gorilla, and so on, and pužljano “spider” is used for any insect (discussed with further examples in Hancock, 1976a). It is also apparent that some of these may be of some antiquity in the language, since they retain inflections with their Romani rather than Angloromani function (rukengrī-chŏkəs, for example).


These are puns, though used without humorous intent, on the primary English interpretation of the Angloromani item (given first):


1. “way, road”

2. “way, manner”


1. “two”

2. “to” (see Ackerley, 1929:119)


1. “gorgeous”

2. “in a non-Romani manner” (< gâwjə)


1. “mare”

2 “mayor”


1. “barn”

2. “baron, lord”


1. “hole”

2. hevy “holy, sacred”


1. “piece, portion”

2. “peace, quiet”


1. “(finger)-nail”

2. “nail (for hammering)”


1. “cold, chilly”

2. “cold, catarrh”


1, “four”

2. “for” (see Ackerley, 1929;119)

tūlo, tūlə

1. “fat, obese”

2. “fat, grease”


1. “cap, hat”

2. “captain, chief”

There are, in addition to these, many proper nouns formed in the same way; thus, the Continental Rom in the United States are sometimes referred to as nŏngə-folki that is, “Bear People”, from their earlier practice of exhibiting performing bears, but where nŏngəin fact means “bare, naked”. Similarly, various Romnichal families have Angloromani names punned on their English surnames: the Stanleys are known as the Barengrīs, for instance, the syllable /stan-/ being similar to the English word “stone”, which in Angloromani is ba(r). The Coopers are called the Vâwdərmeskərs, “Waggon people.”  Among town-names formed in the same way, may be mentioned būkə-panī-găv “Liverpool” (lit. “liver water (i.e. pool) town”) in Britain, and lŏn-pani-găv “Seattle”, where the first syllable is taken to be “sea” (lŏn-pani “salt water”, in Angloromani), in the United States.  Other non-punned American town-names have been listed by Gatchet (1897), and include purənə-găv “Boston” (lit. “old town”), măchəneskī-gav “Gloucester, Mass.”, (lit. “fishy town”) and sĭgmengə “Chicago” (lit. “fast” + -mengə).  Fowkes (1977) discusses the same town-naming processes in inflected Welsh Romani.


Calquing, by its nature, requires knowledge of two languages, since they are usually translations of idioms rather than direct lexical correspondences.  Examples in Angloromani include:


“sacked, fired from one’s job” (gŭnə “sack,” “bag”)


“a heavy blow; a “hot one’” (tătə, tăto “hot”)


“testes” (ba “stone;” cf. dialectal English stones).



Semantic extension

Examples of semantic extension have been mentioned above, where the word for “apple” is used for “fruit”, the word for “waggon” is used for “vehicle”, and so on.  Other examples include:


1. “song”

2. “newspaper; bible”


1. “leg”

2. “wheel”


1. “alone”

2. “peaceful; tranquil”

Disguised English items

The commonest enclitic applied to English items is -əs (< Greek -@H)





bluəsə (?)



“bull” (perhaps to avoid conflict with bul “buttocks”)


“by and by; soon”


“tobacco smoke” (< “fog”; also Cant)

Miklosich (1880, Vol. X:6) discusses this in Angloromani and gives examples of the same suffix in other dialects, e.g. Hungarian Lovari baratos “friend” (< Hungarian barato + -os), Scandinavian bankus “bank” (< Norwegian/Swedish bank + -os), Czech Romanes hrmišagos “weather” (< Czech hrmeti + Hungarian -sag- + -os).  See also Barannikov (1931:120-121).  It may also be added to native items, e.g. nŏkəs “nose.”

Among other disguised English items are found











Items from English provincial dialects


“bor,” i.e. “pal, friend”


“swankie,” i.e. “beer”


“tippoty,” i.e. “antipathy, ill-feeling”


“yarbs,” i.e. “herbs”



With the exception of swŏnkī, the above items occur in Gullah with the same forms.

Items from Slang and Cant








“make, mend”




“beer; a drink”


“glass; window” (dukərin gleys, “crystal ball”)








“woman; wife”


“pint; glass; a drink”






“mat, carpet”


“wait; stop; quit”


“horse brush”







Mixed morphemes

English-derived morphemes may combine with those from Romani in ways other than those already described:


“away” (drŏm “road; way”)


“ago” (jâw “go”)


“afraid” (trăsh “fear”)


“together” (ketənī “together”)


“tonight” (ratī “night”)

Collapsed items

A few items with the same surface form have different forms in the source language, and maintain the two meanings in Angloromani:


1. “luck” (< baxt)

2. “hunger” (< Common Romani bokh)


1. “sit” (< beš-)

2. “year” (< berš)


1. “penis” (< kar)

2. “thorn” (< kanro)


1. “white” (< parno)

2. “cloth” (< berand)


1. “nice; secret” (< šukar)

2. “dry” (< šuko)


1. “hear” (< šun-)

2. “moon, month” (< čhon)


1. “fear” (< traš)

2. “thirst” (< truš)

The reverse is evident in the two words răt “blood” and ratī or radī “night”, which have evolved separate phonetic shapes but stem from identical-sounding items in Romani (rat “blood”; “night”).


Arnold (1894) lists vasovul rak as “bad talk” where rak is back-formed from the verb răkər/râwkər “speak.”  I have not heard this; “speech” or “talk” is generally râwkrĭbən or râwkrin’.

Semantic specialization

Semantic specialization has taken place between sŏv and sūtī, derived from the present and preterite forms of the verb “sleep” respectively, in the inflected language. In Angloromani, the former is retained in its original sense, while the latter has come to mean “copulate”. A similar split has taken place in Lomavren (Armenian Gypsy), which has undergone restructuring in the same way as Angloromani), where sutav is the root morpheme meaning “sleep” in that language, and the form sevel deriving from the historic indicative, means “be in prenatal confinement”, or “give birth”. Other restructured or partially restructured Gypsy languages/dialects seem to have had to select one or the other of these forms; thus, the only surviving words for “sleep” in Domari (Nawari Gypsy, spoken in the Middle East) and in Scandoromani are sučar and sutta respectively, from the original preterite root vsut-”, That dialect probably dropped sova because of the Swedish verb sovawhich also means “sleep”. In Spanish Caló, on the other hand, the verb is sobar, and it is the sut- form which has been lost. In inflected Welsh Romani the regularized preterite form sovdom exists alongside sutiom.

Semantic confusion

There is at least one instance of semantic reversal, where Angloromani loko “heavy”, derives from the inflected sourceform meaning “light”. Misapplication is evident in Angloromani kunyəs “knees”, where kunia in inflected Romani means “elbows”. The word chŏng, meaning “knee” in inflected Romani (čåŋ) , is more commonly used to mean “hill” in Angloromani.

Items from Other Romani Dialects

Arnold (1894) lists a number of items in American Angloromani which appear to have been acquired in that variety from other Romani dialects spoken in the United States.  Among them are


“beer” may be < English, or from Vlax birja


“German,” “Dutch” (cf. Chekener gadschkene)


“book” (cf. Vlax kliška)


“priest” (cf. Vlax popo)


“button” (cf. Vlax kočak; the general Angloromani word for this is  krafni, i.e. “nail”)


1. From the British & Foreign Bible Society’s translation of St. Luke XV (1973):

But sor dulla cherus, the puro chavvi sas avri, adre the puv. Kana lesti jelled posho the ker, he shooned gillying and kelling. So he rockered to yek of the butiengries, ta putched lesti; “What’s going on?”, ta the gero pukkered lesti, “Tiro pral’s akai, so your dadrus has mored a tulli tickni gruvni because he’s avved back kushti ta sasteem.

(“But all that time, the old son was out in the field. When he got near the house, he heard singing and dancing. So he spoke to one of the servants, and asked him “What’s going on?”, and the man told him “You brother’s here, so your father has killed a fat little cow because he has come back well and healthy”).

2. From the catalogue of Romanestan Publications, London, England, which produces and distributes materials in Angloromani:

The Travelling Folki’s bokalo for ledee’s kokero lils adre the Romani jib. We want to dick but lils adre the pogardi jib and adre the puri jib . . . adre our chairus we can ker more cushti by chinning lils than by koring gavvers. We want to del the Travelling Folki the lils that will ker them ruzlo adre our tem. Kana sig.

(“The Traveling Folk are eager [lit. “hungry”] for their own books in the Romani language. We want to see many books in the Broken language and in the Old language . . . these days [lit. “in our time”] we can do more good by writing books than by antagonizing policemen. We want to give the Traveling Folk the books that will make them strong in our country. Right away).

3. From a personal letter received from an Angloromani speaker:

Mandi sikkered it to the godgo. He penned that miro tickno lil sas kushti, but I’d kerred some bongo wafedi lavs in it. Then with his kokoro vast he chinned mandi a translation adre Romanes of an angitrakeri gilli, but mandi’s atrashed as it’s kek but kushti.

(“I showed it to the non-Gypsy. He said that my small book was fine, but that I’d made some mistakes [lit. “wrong bad words”] in it. Then with his own hand, he wrote me a translation in Romani of an English song, but I’m afraid  that it’s not very good”).

4. Sample of Angloromani from Smart & Crofton (1875), showing the state of the language a century ago:

. . . some waver Romani folki sas adoi as mendi didn’t  jin. Yon atch’d talé a bitto drom sor by lendi kokeros. They were more copper like adré lendi mooiaw dan mendi . . . Kavakei folki were waver temengri Romanies, don’t you jinéss, and Wester . . . kom’d to rocker wi’ lendi, but bless you rei, he couldn’t jin posh of sor lendi rockeropen.

(“...some other Romani people were there, that we didn’t know . . . They  stopped down a little road all by themselves. They were more copper-like in their faces than us . . . those people were foreign [lit. “other country”] Romanies, don’t you know . . . and Sylvester . . . wanted to speak with them, but bless you, sir, he couldn’t understand half of all their talk”).

5. Sample of Angloromani taken from a Canadian-born Romnichal in Toronto, during Summer, 1968:

Tuti kers mandi te pench oprey o puro chiros vonka mandi shomas a tarno chavi a-jivin’ along of the rig of the sastero vardo’s drom. Amendi sas kekno but lovey andrey adovo divuses. Miro dad sas naflo but of the chiros, ta mandi ta mandi’s prals and pens sas te jal along of the rig of the sastero vardo’s drom te lel o wongar fon the rig of the drom and chiv lis andrey the gono to rigger lis kerey to chiv andrey amendi’s yag te tatter amaro ker.

(“You make me remember [lit. “to think on”] the old time(s) when I was a young boy, living alongside the railway tracks [lit. “iron waggon’s road”]. We didn’t have much money in those days. My father was sick much of the time, and I and my brothers and sisters had to go along the side of the tracks to get the coal from the side of the way, and put it in the sack to carry it home to put on our fire to warm our house”).

6. Sample of American Angloromani, transcribed from a tape-recording made at Galveston, Texas, in. August, 1975:

We kăkə vels drey Texas wi’ the vegrəs you jĭn like we used to kər būdī bərsh aglâw; ŭknə there’s dōstə būdī goddam permanent griffs drey the găvs ’n they kăkə puch us so much no more.

(“We don’t come into Texas with the carnivals, you know, like we used to do many years ago. Now there are too many [lit. “enough many”] goddam permanent griffs [i.e. “shows”] in the towns, and they don’t ask us so much any more”).  [Since this was written, the Stanley family has established a permanent funfair (“carnival” in American English) in San Marcos, Texas].



First attestations of Gwestwr In England


First unofficial reference to presence of Romanies in Britain


First official reference to presence of Romanies in Britain


First unverified sample of inflected European Romani (Britain)


First verified sample of inflected European Romani (Britain)


First reference to a newly-devised speech (in Harman)


Second reference (in Holinshed)


Third reference (in Rowlande)


Fourth reference (in Rid)


Second published sample of inflected British Romani


McElligot makes the first reference in print to Shelta         


Borrow’s first reference to Angloromani


Borrow’s second reference


Borrow’s third reference     


First scholarly description of Romani in England (Smart)


Hotten develops Harman’s account


Leland’s first reference to Angloromani  


Borrow’s fourth reference


Smart & Crofton’s major linguistic study


Clough’s treatment of Angloromani as a “mixed language”


Leland announces his discovery of Shelta


Turner suggests Angloromani is relexified Cant


Leland’s second reference


Bloomfield discusses Angloromani as a “mixed language”


Reinecke modifies Bloomfield’s view


Gray refers to Angloromani as a creolized language     


Hancock develops Hotten’s suggestion          


Kenrick opposes contact situation view; suggests language loss


Hancock argues against this   


Hymes discusses Angloromani as a pidginized language


Angloromani is given a section in the Bibliography. of Pidgin and Creole Languages


Hancock summarizes his arguments 


Kenrick summarizes his arguments


l. An expanded version of a paper first presented at the Conference on New Directions in Creole Studies, held in Guyana, South America, in August, 1976. In preparing the present draft I was helped in particular by my friends Donald Kenrick, Thomas Acton, Ronald Lee, Morey Pinfold, Lanie Pinfold, Moses Harn, and a number of other friends and relatives who did not wish their names to be included here. Thanks go to these, and to Magnus Ljung, who provided translations of the Swedish sources.  [Thanks too to Butch Lee and family for additions to this expanded (2005) version].

I afterwards spent the night at the house of a white planter, who told me that, when he was a boy, he had lived at Alexandria. It was then under the Spanish rule; and, “the people they was all sorts. They was French and Spanish, and Egyptian and Indian, and Mulattoes and Niggers”
— “Egyptians?”
“Yes, there was some of the real old Egyptians there then.” “Where did they come from?”
“From some of the Northern Islands”
“What language did they speak?”
“Well they had a language of their own that some of ’em used among themselves— Egyptian, I supposed it was, but they could talk in French, and Spanish too”.
“What color were they?”
“They was black, but not very black. Oh! they was citizens, as good as any; they passed for white folks.”
“Did they keep close by themselves, or did they intermarry with white folks?”
“They married mulattoes, mostly, I believe. There was heaps of mulattoes in Alexandria then—free niggers—their fathers was French and Spanish men, and their mothers right black niggers. Good many of them had Egyptian blood in ’em too.”
. . . The Egyptians were probably Spanish Gipsies; though I have never heard of any of them being in America in any other way.

There are a great many Romanichals in the USA, especially in the South. Salo (1977, ms. p. 7) suggests that they may constitute “the largest among the Gypsy groups” in North America, though they are clearly outnumbered by the Continental Rom. It is widely believed among southern Romnichals that the late Elvis Presley descended from the Nawgin (Scottish Romanichal) families of Pressley and Smith, and his former manager Tom Parker is also said to be of Romnichal stock.

Ironically, while the earliest Romanies were brought to America as unwilling emigrants, the U.S. Government banned their immigration in 1885, According to Trigg (1973:224) “in the latter half of the nineteenth century, many more Gypsies, mostly from Slavic countries, were to arrive in the United States. By 1885, however, Gypsies were excluded by immigration policy, and many returned to Europe.”  During the early 19th century, several thousand others were transported by the English to the penal colony at Botany Bay in south Australia, once the American colonies were closed off as a result of U.S. independence.

3. In Angloromani, the language is called pŏgədī-jĭb “broken language”, pŏsh-tə-pŏsh “half and half”, or simply rŭmnĭs “Romanes”. Inflected British Romani is usually called puri-jib “old language”.

4. The generally accepted date of arrival of Romanies in Britain is 1505 for Scotland, and 1512 for England. The earlier year is the date of the first known, official reference to their presence, and is found in the accounts of the Scottish Lord Treasurer.  However, MacRitchie (1894:20) cites a source dated ca. 1452, which reported that “ . . . a Company of Saracens or Gipsies from Ireland infested the country of Galloway [in Scotland]”, and another (1894:27) telling of one Marion of Polwarth who, in 1470, “conveyed a letter by the hands of Johnny Faa, captain of a gang of gipsies, to George Home, the young Baron of Wedderburn, her lover”. There are descriptions of people who may well have been Romanies, but who are not specifically identified, dating back to the early decades of the century before that, and Bercovici (1929:263), perhaps rashly, mentions the original [British] Gypsies “who arrived in the isles in the eleventh or twelfth century”.

5. There have been two main waves of migration westward since arrival in southeastern Europe in (probably) the early 14th century; the first after the collapse of the Balkan-Muslim trading empire in the 15th and 16th centuries, and the second after abolition of four centuries of slavery in eastern Europe, in 1864 [and now a third since the fall of communism in 1989].  The Romanies in Britain, and in Spain and Scandinavia, where the same linguistic situation is found, arrived during the time of the first migration.

6. For purposes of comparison with the inflected British dialect of Romani that now barely survives only in Wales, consult Bourgeois (1910), Sampson (1926), Tipler (1957) and Lockwood (1975). For more information on the relationship which existed between the Romani and English outlaws and the settled population, see Salgado (1972), which includes a reprint of Harman (1556), in modernized spelling.

7. Unless it had been with English speakers in England’s territories in France prior to the loss of Calais in 1558. There are several French-derived items in Angloromani: avŏnkə “before” (< avant quo), bĭtə “little” (< petit), dăs~des “cup” (< tasse), pŏvərə “poor” (< pauvre), spĭngə “pin” (< espingle), &c.   On phonological grounds the word pench “think,” is unlikely also to be French (< penser) as has been claimed, and is probably from British Romani pinčer- “know, understand” (prindžer- “recognize” in other dialects).  It is also clear from the nature of Angloromani that its Romani component derives from the northern dialects, now spoken mainly in Germany and France, which include Welsh Romani. Comparisons have been made throughout with the latter dialect.

8. The fact that it is still spoken in parts of Britain is clear evidence of this.  Lockwood (1976:243) notes that the Welsh Romani speakers, whose language he describes, “ . . . remembered that their forebears had come from Somerset in the eighteenth century”.

9. R. Hodgson has recently drawn attention to two short verses in Romani, apparently dated 1517 and some time in the 16th century, and reproduces them in his JGLS article [but cf. Hancock, 2004:90-91 for a discussion of this].

10. Taken from Boorde (1547 [1870:218]). The English has been modernized here, while the Romani remains uncorrected. These sentences have been analyzed by Miklosich (1874:759-765), who rewrites them as follows: Lači tutty dyves!; Cater myla barforas?; Mayst aves barforas; Mole pis lauena; avaua tosa; hys tele n pee; pee, pee, deuel asse; Achae, da mai manor la veue; Da mai masse!; Achae, aw ordey, susse; Da mai paba la ambrell; Iche misto!; Lači rat tut!. 

11. Pinkerton.

12. A Welsh word, literally “(unbidden) guests”, also written Westour in Middle English, In modern Welsh guesteion.

13. The examples Kenrick gives are Jewish English The galach ganav’d the gelt and Angloromani The rashi chored the wongur “The priest stole the money”.

14. This has as much to do with the Slavic and Balkan (i.e. non-English) linguistic and cultural features of the “Rooshians” as it has with unfamiliar native lexicon and phonology.

15. Sinclair seems to have been an innocent object of fun during his visits with his informants; in his glossary he lists the non-Romani brokla (i.e. “broccoli”) for “cabbage”, and for karri, korri he has “domestic cock”, when in fact it means “penis”. The same sort of thing can be found in other vocabulary lists (see Hancock, 2004).

16. See Hancock (1975) for a discussion of processes of lexical expansion in Sierra Leone Krio.

List of works consulted

ACKERLEY, F.G., 1929. “Anglo-Romani gleanings”. Journal of the Gypsy Lore- Society, 3rd. series, VIII (3).112-134.
ACTON, Thomas A., 1971a.  Current Changes amongst British Gypsies and their Place in International Patterns of Development.  Oxford.
ACTON, Thomas A., 1971b, Mo Romano Lil,  Oxford.
AINSWORTH, Wm. H. , 1834, Rookwood, London & Philadelphia, 2 volumes.
ANON., 1930.  [Review article of Triandaphyllidis, q.v.], Journal of the Gypsy Lore Society, 3rd ser., 9(1):38-46.
ARNOLD, Frederick S., 1894.  Romany Vocabulary: Being a Ms. Collection of All the Forms and So Forth of English Gypsy Words Now Current Amongst Gypsies in America.  New Jersey.
BAKKER, Peter, 1995.  “Notes on the genesis of Caló and other Iberian Para-Romani varieties,” in Matras, 1995, pp. 125-150.
BARANNIKOV, A.P., 1931. “Ukrainski vpliv na Ciganski Govirki Ukrain,” Na Movoznavčomu Fronti, 1:116-123.
BARRIERE, A., & Charles G. Leland, 1889. A dictionary of Slang, Jargon and Cant, embracing English, American and Anglo-Indian Slang, Pidgin English , Tinkers’ Jargon, and Other Irregular Phraseology. Oxford.
BAUMANN, H., 1887.  Londonismen: Slang und Cant. Berlin (Pp, ci-cvi on Angloromani).
BERGMAN, Gösta, 1964. Slang och Hemliga Språk. Stockholm.
BERCOVICI, K., 1929. The Story of the Gypsies. London.
BLACK, George F., 1914. A Gypsy Bibliography. London.
BLOOMFIELD, Leonard, 1933, Language, London.
BLOUNT, Ben, & and Mary Sanches, (eds.). Sociocultural Dimensions of Language Change. New York.
BOORDE, Andrew, 1547. The First Boke of the Introduction of Knowledge.  Reprinted in 1870, F.J. Furnival (ed.), Cambridge.
BORROW, George, 1841.  The Zincali, or an Account of the Gypsies in Spain.  Two vols. London.
BORROW, George, 1851. Lavengro: the Scholar ,the Gypsy, the Priest. London.
BORROW, George, 1857. The Romany Rye: a Sequel to ‘Lavengro’. London.
BORROW, George, 1874. Romano Lavo-Lil: Word-Book of the Romany, or English Gypsy Language. London.
BOURGEOIS, Henri, 1910. “Esquisse d’une morphologie du Romani gallois”, Revue de Linguistique et de Philologie Comparée, 43:179-190.
BRITISH & FOREIGN BIBLE SOCIETY, 1973. Shavved then got Latchered: Luke XV in Romany.  London.
BURNETT, Swan M., 1889. “A note on the Melungeons”, American Anthropologist, October, 1889:347-349.
CASSIDY, Frederic, & R.B. Le PAGE, 1967. Dictionary of Jamaican English. Cambridge.
CHAMBERLIN, E.R., 1965. Everyday Life in Renaissance Times. London.
CLOUGH, James C., 1876. On the Existence of Mixed Languages. London.
COELHO, Adolpho, 1892. Os Ciganos de Portugal, com um Estudo sobre o Calão. Lisbon.
COLLYMORE, Frank, 1970. Notes for a Glossary of Words and Phrases of Barbadian Dialect. Bridgetown, Barbados.
DeCAMP, David, & Ian F. Hancock, (eds.), 1974. Pidgins and Creoles: Current Trends and Prospects. Washington.
DESMOND, Hall, ca. 1890. The Language of the Gypsies. Douglas, Isle-of-Man.
DILLARD, Joe, (ed.). Perspectives on American English, The. Hague.
FINCK, Franz N., 1907. Die Sprache der Armenischen Zigeuner. St. Petersbourg.
FOWKES, R., 1977. “Onomastic sophistication of the Gypsies of Wales”. Names, 25(2).78-87.
GATSCHET, A.S., 1897. “Ethnographic notes: Gypsy town names”. The American Antiquarian and Oriental Journal, 19: 334-335.
GJERDMAN, Olof, & Erik LJUNGBERG, 1963. The Language of the Swedish Coppersmith Gypsy Johan Dimitri Taikon. Uppsala and Copenhagen.
GOODMAN, Morris, 1971. “The strange case of Mbugu”, in Hymes, 1971.243-254.
GRANT, Anthony, 1998.  “Romani words in non-standard British English and the development of Angloromani,” in Matras, 1998, pp. 165-192.
GRAY, Louis H., 1939. Foundations of Language. New York.
GROSE, Francis, 1785. Classical Dictionary of the Vulgar Tongue. London.
HALL, Robert A., 1966. Pidgin and Creole Languages. Ithaca.
HANCOCK, Ian F., 1970. “Is Anglo-Romanes a creole?” Journal of the Gypsy Lore Society, 49(1/2).41-44.
HANCOCK, Ian F., 1971. [rejoinder to Kenrick] in Acton (1971a:15-18).
HANCOCK, Ian F., 1974, “Shelta, a problem of classification”, in DeCamp & Hancock, 1974:130-137.
HANCOCK, Ian F., 1975. “Lexical expansion within a closed system”, in Blount & Sanches, 1975:161-171.
HANCOCK, Ian F.,  1976a. “Patterns of English lexical adoption in an American dialect of Romanes”, Orbis, 25(l).83-104.
HANCOCK, Ian F., 1976b. “Romance vs. reality: popular notions of the Gypsy”, Roma, 11(1):7-23.
HANCOCK, Ian F., 1978. “Observations on the Afro-Seminole Creole of Brackettville, Texas”, in Dillard, 1978:
HANCOCK, Ian F., (ed.), 1979a. Romani Sociolinguistics. The Hague.
HANCOCK, Ian F., l979b, “Gypsies”, in Thernstrom, 1979.
HANCOCK, Ian F., 1984.  “Shelta and Polari,” in Trudgill, 1984:384-390.
HANCOCK, Ian F., 1985.  Review article of Judith Okely, The Traveller Gypsies. (Cambridge: The University Press), Journal of Intercultural Studies 1985:89-93.
HANCOCK, Ian F., 1992.  “The Social and linguistic development of Scandoromani,” in Jahr, 1992, pp. 337-52.
HANCOCK, Ian F., 2004.  “The concocters: creating fake Romani culture,” in Saul & Tebbutt, 2004:85-97.
HAYWARD, James, 2003.  Gypsy Jib: A Romany Dictionary.  Wenhaston.
HARISON, J., 1578. Description of England, in  Holinshead.
HARMAN, Thomas, 1566 (or 1567). A Caveat or Warening for Common Cursetors Vulgarly Called Vagabones. London. Reprinted in Salgādo,1972:79-153.
HODGSON, R., 1973. “Early Romani”, Journal of the Gypsy Lore Society, 52(3/4):87-89.
HOLINSHED, Raphael, 1577. Firste Volume of the Chronicles of Englan, Scotlande and Irelande, three volumes, London. Second printing, 1578.
HOTTEN, John C., 1864. The Slang Dictionary. London.
HYMES, Dell, (ed.), 1971. Pidginization and Creolization of Languages. Cambridge.
IVERSEN, Ragnvald, 1944. Secret Languages in Noway. Oslo.  Three volumes.
JAHR, Ernst Håkon, ed., 1992. Language Contact: Theoretical and Empirical Studies.  Berlin.
JOHANSSON, Roger, 1977. Svensk Rommani.  Uppsala.
JONES, Alex., 1834. “American Gypsies”, American Journal of Science and Arts, 26:189-190.
JUNG, Christof, 1972. Wortliste des Dialekts der Spanischen Zigeuner. Mainz.
KENRICK, Donald, 1971. “Anglo-Romani today”, in Acton, 1971:5-14.
KENRICK, Donald, 1978 “Romani English”, in Hancock, 1979a.
KNIGHT, Charles, 1841-1844.  London.  London: Charles Knight Co. In six vols.
KOSTOV, Kiril, 1973. “Zur Bedeutung des Zigeunerischen für die Erforschung grammatischer Interferenzerscheinungen” , Balkansko Ezikoznanje, 16(2) :99-113.
LEHMANN, Winifred, 1983.  Language: An Introduction.  New York: Random House.
LELAND, Charles G., 1873. The English Gipsies and their Language. London.
LELAND, Charles G., 1876. Pidgin English Sing-Song, or Songs and Stories in the China English Dialect. London.
LELAND, Charles G.,1889. The Breitmann Ballads. London.
LOCKWOOD, W.B., 1975.  Languages of the British Isles, Past and Present. London.
MACALlSTER, R.A.S., 1914. The Language of the Nawar or Zutt, the Nomad Smiths of Palestine. London.
MACFIE, F. Scott, 1910.  A System of Anglo-Romani Spelling.  Privately printed monograph, Liverpool.
MACRITCHlE, David, 1894.  Scottish Gypsies under the Stewarts. Edinburgh.
MATRAS, Y., ed., 1995.  Romani in Contact.  Amsterdam & Philadelphia.
MATRAS, Y., ed., 1998.  The Romani Element in Non-Standard Speech.  Wiesbaden.
MATSELL, George W., 1859. Vocabulum; or, the Rogue’s Lexicon. New York.
McLANE, Merrill F., 1977. “The Calo of Guadix: a surviving Romany lexicon”, Anthropological Linguistics, 19(7).303-319.
McPEEK, James A.S., 1969. The Black Book of Knaves and Unthrifts. University of Connecticut Press.
MENCKEN, H.L., 1937. The American Language. 4th edition,  New York.
MIDLETON, Thomas, & William Rowley, 1653. The Spanish Gipsie. London. Act II, scene 1.
MIKLOSICH, Franz X., 1874-1878. “Beiträge zur Kenntniss der Zigeunermundarten”, Kaiserliche Akademie der Wissenschaften, Philosophisch-historische Classe. Parts 1 & 2, Vol. 77.759-792, Part 3, Vol. 83535-562, Part 4, Vol. 90:245-296.
MIKLOSICH, Franz X., 1872-1881. Ueber die Mundarten und die Wanderungen Zigeuner Europa’s, 12 parts, Vienna.
OLMSTED, Frederick L., 1861. Journey in the Seaboard Slave States.  New York.
PINKERTON, W., 1780.  A Collection Taken Down from the Mouths of Gypsies in Somersetshire by a Clergyman Resident There in 1780. London.
PLATT, John T., 1975. “The Singapore English speech continuum and its basilect ‘Singlish’ as a ‘creoloid’“, Anthropological Linguistics, 17(7).363-374.
PRINCE., J.D., 1907. “The English-Rommany jargon of the American roads”. Journal of the American Oriental Society, new series, 11,271-308.
REINECKE, John E., 1937. Marginal Languages: a Sociological Survey of the Creole languages and Trade Jargons. Doctoral dissertation, Yale University.
REINECKE, John E., et al., 1975.  A Bibliography of Pidgin and Creole Languages.  Honolulu.
RID, Samuel, 1612.  The Art of Jugling, &c.  London.
ROMANESTAN PUBLICATIONS, 1969-. Patteran [periodical Angloromani catalogue], London.
ROWLANDE, Samuel, 1610.  The Runnagate’s Race. London.
SALGĀDO, Gamini, (ed.), 1972. Cony-Catchers and Bawdy Baskets. Harmondsworth.
SALO, Matt, 1977. Gypsy Ethnicity: Implications of Native Categories and Interaction for Ethnic Classification. Unpublished manuscript.
SAUL, Nicholas & Susan Tebbutt, eds., 2004.  The Role of the Romanies: Images and Counter-Images of the ‘Gypsies’/Romanies in European Cultures.  Liverpool: The University Press.
SCHUCHARDT, Hugo, 1881. “Die Cantes flamencos”, Zeitschrift für Romanische Philologie,
SINCLAIR, A.T., 1915. “An American Romani vocabulary”. Bulletin of the New York Public Library, 19(10):727-738.
SMART, Bath C., 1860 (1863). “The dialect of the English Gypsies,” an appendix to the Transactions of the Philological Society , 1862-1863.
SMART, Bath C., & H.T. CROFTON, 1875. The Dialect of the English Gypsies. London.
SMITH, Jess & Robert Dawson, 2002.  The Scottish Traveller Dialects.  Blackwell: Samizdat.
TAYLOR, Douglas, 1971. “Grammatical and lexical affinities of Creoles”, in Hymes, 1971:293-296.
THERNSTROM, 1979. (ed.). The Harvard Encyclopedia of American Ethnic Groups. Cambridge, Mass.
TIPLER, Derek, 1957. “Specimens of modern Welsh Romani”. Journal of the Gypsy Lore Society, 36(1/2).9-24.
TODD, Loreto, 1975. “The case for the creoloid”, paper presented at the Hawaii Conference on Pidginization and Creolization, Honolulu.
TRIANDAPHYLLIDIS, M.A., 1923-4.  “Eine zigeunerische Geheimsprach,” Zeitschrift für Vergleichende Sprachforschung, 52:1-42.
TRIGG, Elwood B., 1973. Gypsy Demons and Divinities. Secaucus.
TRUDGILL, Peter, ed., 1984.  Language in the British Isles.  Cambridge: The University Press.
TURNER, C.J. Ribton, 1887. History of Vagrants and Vagrancy, and Beggars and Begging. London.
VIRGINIA MAGAZINE, THE, 1894. [untitled article, anon.], 11(1), page 100.
WAGNER, Max L., 1938. “Stray notes on Spanish Romani”, Journal of the Gypsy Lore Society, 15(1/2).134-138, 16(1/2).27-32.
WINSTEDT, Eric 0., 1950. “Anglo-Romani gleanings”, Journal of the Gypsy Lore Society, 26(1):102-110, 28:50-61.