ï»ż
Back to Main Category

Copyright
Pater Bakker (ed.)
The Typology and Dialectology of Romani
Amsterdam & Philadelphia: John Benjamins (1998), pp. 65-89

GEORGE BORROW’S ROMANI

Ian Hancock

0.  Introduction

George Borrow (1803-1881) has stood as the acknowledged source of inspiration for countless Romanophiles (as well as Romanophobes) ever since his literary heyday in the 19th century; in fact Brian Vesey-Fitzgerald saw himself as quite “unfashionable” (1944:x) because he was one of the few who didn’t make his “first acquaintance with [Gypsies] in the pages of George Borrow.”  From Prosper MerimĂ©e, who “drew from Borrow his inspiration for Carmen” (Ridler, 1996:55), to criminologist Detective Samuel Haines whose monograph on American Gypsy crime relied largely on the “Romano La Volil” (sic; 1989:2), George Borrow’s writings have stimulated the creative muse for innumerable writers about Gypsies for more than a century and a half. 

Borrow spent the greater part of his life studying languages, a love which was kindled while he was a boy learning Latin and Greek at Norwich Grammar School, and he was already able to translate some twenty as diverse as Armenian and Middle Welsh “with facility and elegance” by the end of his teen years (Ridler, 1996:451-453).  By the time of his death at the age of 78, he had dealt in greater or lesser detail with eighty more (op. cit., 427).

As one of two sons of a military family, he moved about England with his father’s regiment and developed a love for the British countryside and its inhabitants later reflected so elegantly throughout his writings.  So strong indeed was the lure of rural England, that instead of pursuing the legal education his father had intended for him, he left his parents to take up with Traveller families on the roads, acquiring a taste for the Romani language which he subsequently took with him to Spain, Hungary, Romania, Russia and elsewhere.

Few figures in Romani Studies have been so roundly praised nor yet so heartily criticized as George Borrow, or have prompted such extremes of reaction.  His all-consuming interest in Gypsies condemned him to serve as a prime example of “feebly inhibited genetic development” in a report by the Director of the U.S. Department of Experimental Evolution (Davenport, 1915:23) for example, while on the other hand John Sampson, the greatest scholar of Romani ever to have lived, was moved to dedicate his monumental grammar of Welsh Romani to Borrow.  His inscription, in Devanagari script, reads ki Borrow, kai but berơendi dudyerdas m’o drom akai, ta akana asala ’pre mande peske briơindeskeriate (“to Borrow, who for many years lit my way here, and who now smiles at me from his rainbow”) (1926:iv).  Yet his contemporary Laura Smith wrote (1889:134) that his book was “most incomprehensible” and “could content no one [because] it hovers between romance and reality, and can have done but little towards establishing a more friendly feeling between Gorgios and Romanies.”  A century later Audrey Shields, in her study of the Gypsy stereotype in Victorian literature believed that Borrow’s idealization “did as much harm as writers who denigrated Gypsies” (1993:167), a sentiment echoing that of British Member of Parliament John Wells, one of the few government representatives sympathetic to the Romani situation, who claimed that “George Borrow has done more harm to the cause of those of us who wish the Gypsy community well than almost anyone else” (quoted in Reid, 1962:37),1 John Geipel more recently lauded him as the “savant . . . who provided posterity with its largest single source of information on the Indian roots of the ‘secret’ language of the gitanos” (1995:112).  That he has prompted such widely differing responses probably accounts for the fact that the study of the man and his life remains so intriguing to this day; and it was his involvement with the Romani people and language in particular which has left the most indelible mark.

1.  George Borrow and the Romani Language

Three features in particular characterize George Borrow’s relationship with Romani: firstly, that his knowledge of its structure was surprisingly poor, given his acquaintance with so many other languages; secondly, that he freely mixed dialects, and thirdly, that he was not above creating lexical entries and grammatical forms of his own.  Behlmer (1985: 241) criticized both Borrow’s knowledge of English Romani and Spanish Romani, saying that his Romano Lavo-Lil was “shot through with absolutely ludicrous errors in etymology.”

True to his style, Borrow spared no pains in cloaking the origins of Gypsies and the Romani language in mystery, a mystification which in fact was forced, at least in one respect, for he was aware of the Indian identity of the Roma by the time that he published his work The Zincali in 1841, being already familiar with the works of Grellmann, Whiter, Marsden and Richardson, which last work also contains the vocabulary collected by Bryant.  He had even seen Andrew Boorde’s book, which he mentions in The Romany Rye -- though he seems to have missed the Romani it contained.  So on page 104 of the second volume of The Zincali, (1841) he stated quite plainly that

Gypsies. . . are the descendants of a tribe of Hindus, who, for some particular reason, had abandoned their native country.

Despite already being aware of this, he asked in Lavengro, published ten years later in 1851 (in Chapter 17),

Rommany Chals! . . . whence did they come originally?  ah! there is the difficulty.

and the concluding statement of The Romany Rye, which appeared six years later in 1857, clearly gives the impression that the Indian connection had only at that moment occurred to him:

“I shouldn’t wonder,” said I as I proceeded rapidly along a broad causeway, in the direction of the east, “if Mr. Petulengro and Tawno Chickno came originally from India.  I think I’ll go there.”

Nowhere in those two books is the Sanskrit origin of Romani discussed, although there and elsewhere Borrow does compare other languages with Sanskrit, for example Welsh, in Wild Wales (1862:599-600).  Even in the Lavo-Lil (1874:4), his final work, he did not elaborate upon his observation that

the Gypsy language. . . [is] decidedly of Indian origin, being connected with the Sanscrit or some other Indian dialect,

although he did wonder in Lavengro (p. 316) whether it was perhaps “the mother of all languages in the world.”  Angus Fraser suggested to me that this notion may have had its origin in Whiter (1800:xxvij, see also Grosvenor, 1908, Fraser, 1995 and Ridler, 1996), who himself believed that Romani “as it is now spoken, may probably be considered as the most ancient form of Speech, which is at present extant in the world,” and whom he toasts in the same book in Chapter 24.2

2.  Borrow’s Knowledge of Romani

Because of his penchant for embellishment, and because so many of his samples are transparently of his own creation (though he was by no means such a sprucer as was his contemporary Leland), it would be useful to ascertain the extent of his knowledge of Romani, as well as the degree of its accuracy.

Perhaps his first exposure to it took place according to his own account about 1810, when he was seven years old, and when he met Jasper and his family for the first time.  The words he heard during that initial encounter were bengui “devil” tawny “young,” sap “snake,” sap-engro “snake-charmer,” Romans (sic) “Gypsies,” gorgeous “non-Gypsy” and the non-Romani Cant word mumper.  Although he didn’t recognise the language being spoken around him, he knew enough to say “it wasn’t French.”  He was not to meet Jasper again after this for nearly ten years.

Although this episode was described in Lavengro, which was published in 1851, it was not the first time that Borrow had introduced Jasper, whom he had already mentioned ten years before in an appendix to the second volume of the second edition of The Zincali. In those pages he provided vocabularies and texts in British Romani, and compared it with the dialects spoken in both Spain and Hungary.  Jasper Petulengro and his brother Tawno Tickno were, in real life, Ambrose and Faden Smith; but changing the names of people and places, or sometimes hiding them behind initials, was part of Borrow’s mystifying style.  Furthermore, while Jasper is presented in the stories as being older than George Borrow, he was in fact one year younger3.

Borrow visited Russia in 1833 when he was thirty, and remained there for two years.  After a period of time spent back in England, he left again for Spain, and between 1836 and 1840 made three visits to that country.  In 1844, he took an extended trip through Hungary and Romania to Turkey, in each place seeking out Romanies and collecting linguistic material from them.  It was clear that he recognized the linguistic unity of Romani, thinking of it as one language consisting of many dialects differing by more or less retention of the original vocabulary and original grammar.  Perhaps this is why he felt at liberty to mix them so freely, though it is more likely that he in fact believed that his readers would be unable to distinguish one dialect from another.  Thus in Wild Wales (Chapter 98) he had Romanichals speaking Hungarian Romani, and elsewhere he put Spanish Romani in the mouths of Transylvanian Romanies (e.g. the entries busno and errai in Winstedt, 1952), and British Romanichals; the very first Romani word in Lavengro, “bengui,” in fact, has the Spanish Caló spelling and pronunciation rather than the British Romani beng.  This means, of course, that he was not reporting accurately, which seriously diminishes the value of his material—referred to by Thomas Acton (in p.c.) as “deeply unauthentic”—for Romanologists, a number of whom in the 19th century based their own work at least partially upon it.  Smart & Crofton remarked upon this in their own book, referring to “. . . the intrinsic evidence in his writings that many of his words have been procured from various and wide-spread sources” (1875:xij).  This is evidenced by the fact that well over half of the ca. 120 words they give as unattested in English Romani (op. cit., 157-163) are listed as having originated in the works of Borrow. Knapp, his biographer, called it a “a kind of philologico-literary gazpacho” (1899: 23).

3.  The Creation of Romanies and Romani

In a paper I wrote some years ago (1981:14-15), I made reference to Douglas’ notion of Bongo-Bongoism, the practice of some scholars of faking or misrepresenting data with the assumption that their audience knew nothing about the topic, and was therefore not in a position to challenge their claims.  Borrow was especially guilty of this, although he should not be judged too harshly for that.  Approaching his work on Romani as linguists may be frustrating and disappointing, or even amusing sometimes, but coming to it from the point of view of the literary critic is something else entirely.  Borrow’s writings are widely enjoyed precisely for their ponderous and self-involved style; Croley (1996: 186) calls it “precious.” He  had a particular fondness for this, and modelled his prose upon the numerous 18th century works his library contained.  He was, after all, writing for a particular audience.  Helyear sums this up very well when she says (1972:82) that Borrow

complied, consciously or not, with the requirements of Victorian society, to the detriment of realism.  He obviously sought literary success, and consequently had to satisfy his public’s tastes and suit its intellectual esthetics.

Victorian readers relished being titillated by tales of the savage and the exotic in the Sunday afternoon safety of their drawing rooms, and Borrow’s writings typified very well the popular middle class literature of the period.  In fact he tended to overdo it, and was criticised in the press for being too prone to philosophizing and flights of moralistic fancy.  In retrospect, we might reproach Borrow for presenting the Romani population in too romanticized and idealized a light—Helyear points out that nowhere does he “drop a single hint about the hardship of their life,” instead sustaining the more attractive image of “picturesque outsider.”  Like Frederick Ackerly, whose only observation in his review of Potra’s volume on the enslavement of Gypsies in Romania (1942:69-71) was that it was a pleasure and a delight to read and that it gave him a chance to practice his Romanian, Borrow too had nothing to say about that ultimate violation of humanity.  Although he visited Romania while slavery was still everywhere in effect, for example, he was not moved sufficiently to comment upon it:

I visited Wallachia with the express purpose of discoursing with the Gypsies, many of whom I found wandering about, the men supporting themselves by smithery, and the women by telling fortunes, but the generality [were] employed in the brick-fields making bricks like the Ishmaelites of old in Egypt (in Knapp, 1899, ii:44).

Although Lavengro and The Romany Rye were not especially well received at the time of their publication, and Shields (1993:113) believed that in particular it was the “long, virulent diatribe against Scott and Catholics . . . that readers found excessive,” and that the “picaresque style, lacking moral earnestness, was less fashionable than it had been in the previous century,” they nevertheless had a singular effect upon both contemporary and later writers.  Borrow’s style was very cleverly parodied in an article in the Pall Mall Magazine by Sir Arthur Conan Doyle of Sherlock Holmes fame, in an article entitled “Borrow-ed scenes” (Conan Doyle, 1913).

Lavengro appeared in 1851, the year of the Great Exhibition in London, when the British Empire was enjoying the height of its power on every continent.  The Romany Rye was published in 1857, only two years after the appearance of Gobineau’s Essay On the Inequality of the Human Races and two years before Darwin’s equally influential Origin of the Species.  Colonialism and new ideas about evolution and racial superiority stimulated a particular interest in foreign, and especially non-western, peoples and cultures.  Readers’ imaginations did not need to be transported to Borneo or Zululand or Nepal, when this dark and mysterious eastern population occupied their very doorstep.  The Illustrated London News (1873:503) drew attention to this.  Describing a Gypsy community “within an hour’s walk of the royal palace,” it urged

. . . a few serious reflections upon social contrasts at the centre and capital of the mighty British nation, which takes upon itself the correction of every savage tribe in South and West Africa and Central Asia.

Ideas about these “savage tribes” and the superiority of the colonizing powers gave rise in time to the kind of racist dogma expounded by Houston Chamberlain (1899) and specifically elaborated upon against Romanies by Alfred Dillman (1905), in a document which laid the foundation for the National Socialists’ concept of Gypsies as subhumans vis-à-vis the German master race, and which later paved the way for the racial policy of ethnically cleansing the Third Reich by attempting to exterminate the entire Romani population—adopted in modified form since the Holocaust by CeauƟescu in Romania into the 1980s (Hancock, 1993:27).

The Industrial Revolution (ca. 1730-1850) had also brought tremendous changes to urban society by the time of Victoria’s reign, having created a foul and unhealthy mechanized environment vividly described in the works of Dickens and others.  But Borrow’s idealized Gypsies lived apart from all this and above it, noble savages untouched by civilization, representatives of a vanishing rural era who had refused to relinquish it for the sake of progress.  This has been well discussed by Wilson (1986:33-39).  Shields (1993:113) calls Borrow’s attitude towards his Romani acquaintances “preachy and condescending,” and says that he was “worried that their aspirations to gentility [would] corrupt them.”  That the real experience of British Romanies didn’t mirror the literary descriptions of it presented no dilemma, since those who obviously didn’t match the idealized image were simply dismissed as being “not real Gypsies,” but mumpers, diddicais, pikeys, people with little or no Romani ancestry who got the “True Romanies” a bad name (see in particular Mayall, 1988).  Here again, the particular devaluation of individuals regarded as being genetically mixed between Gypsy and white was to re-emerge in a frightening way in Hitler’s Germany half a century later.

Although many authors before Borrow’s time incorporated Romani characters into their works, it was not until the appearance of George Borrow’s books, as Angus Fraser states (1993:15),

. . . that the literary and dramatic clichés received a clear and credible challenge.  Borrow conveyed something of the real nature of Gypsies and their life-style, language and attitudes, and provided a picture that offered an exciting alternative to the previous hackneyed views.

But Fraser was choosing his words carefully when he referred to Borrow’s exposition of “something of the real nature of Gypsies,” for his writings are an exasperating mixture of fact and fancy, and separating the two is not easy.  Borrow clearly imagined that he had some special status, reflecting what Shields has appropriately called “the rye phenomenon” (1993:164).  For Borrow this provided something of a conflict because, as Helyear says, “he mingled with ruffians and people of the road, but insisted upon his being [regarded nevertheless as] a scholar” (1972:82).  It might say something of his aloofness that he missed entirely learning of the existence of inflected Romani in Wales—a discovery that would have been of tremendous importance to him—despite his exploration of that country and the book resulting from it.  Ken Lee brought my attention to the information in Jarman & Jarman’s book The Welsh Gypsies (1991:151), that

George Borrow missed the opportunity of hearing this language when he was in Bala, for he drank in the front room of the inn, while the local Wood family of Gypsies drank in the rear room with the rest of the ordinary Welsh people of the town.

To be fair to Borrow, he spent only three days in Bala altogether, and there is no evidence that the Woods were in the inn while he was there, or that they were even in the district.  Nevertheless during his entire stay in Wales, he encountered Gypsies only once, and those were Boswells, from England.

Despite his rather distant behaviour, he regarded himself as a “brother,” at least in his own terms, almost certainly a self-designation just as his status as a rye was.  We shall never know whether his Romanichal acquaintances saw him as he saw himself, but Mrs. Hearne’s attempt to murder him with a poisoned cake (Lavengro, Chapter 71)4 is a good indication of their general attitude.  Philip Allingham (1934:5) came to realise what seemed never to have occurred to Borrow, despite his being told repeatedly and in many ways to mind his own business:

I decided to steal out of London as quietly and as quickly as possible and join the Gypsies.  That the Gypsies would not be too keen to have me, did not then occur to me.

Paspati (1870:30) too admitted to receiving a less than warm reception on occasion:

Some have chased me from their tents with nasty words and gestures.  Others haven’t responded to my questions, continuing with their work without acknowledging my presence . . .

This perception of Gypsies simply as sources of data, as objects of study, rather than as people with sensibilities of their own, was one characteristic of the early ryes.  Describing his visit to the Baltic lands in 1908, for example, Bernard Gilliat-Smith (1909:154) wrote

I could see clearly five or six black tents pitched on the left side of the road under some birches.  I would fain have stopped, waked the Baro, and there and then collected material of interest for the Gypsy Lore Society.

 Like some modern ryes, Borrow was not above giving others the impression that he was himself a Gypsy; in Lavengro (Chapter 110), when asked by the man in black Reverend Fraser “are you then, a Gypsy?,” he replied “what else should I be?”5. He has also been taken to task for putting far too cultured a brand of English in the mouths of the Romanichals (Helyear even calls it “pompous”).  Pakomovna tells him, when she visits him in the dingle, “do you allow me to officiate upon your hair,”—a hortative construction obsolete in English even by Borrow’s day—while in the Book of Wisdom in the Lavo-Lil, the heavily class-marked pronoun “one” turns up in his list of homilies in Romani, justifying the charge that “his Gypsies all spoke like bishops,” thus when yeck’s tardrad yeck’s beti tan oprey, kair’d yeck’s beti yag ta nashed yeck’s kekauvi. . . (“when one’s put one’s little tent up, made one’s little fire and hung up one’s kettle. . .”).  Equally transparent are the English origins of his translations, thus the mush savo kek si les the juckni-wast oprey his jib “the man who has not got the whip-hand of his tongue.”  While calquing upon English idioms certainly characterizes Angloromani, constructions of this kind lack the ring of authenticity (see Tilford (1953) for a discussion of the English spoken by the Romanies of Borrow’s acquaintance).

4.  Borrow As Romani Scholar

As a linguist, he was prone to exaggeration and sometimes quite impressive creativity; Ridler (1981:329), referring to Caló, relates how in a letter to the British and Foreign Bible Society Borrow wrote that within a day of having first arrived in Spain, he located some Gypsies and “we began conversing in the Spanish dialect of [Romani], with which I was tolerably well acquainted.”  It is significant that in his defence of Lavengro, which appeared as an appendix in The Romany Rye, he chided his critics for not having spotted his “deliberate” mistakes in the Armenian and Welsh samples in his books, but he did not mention Romani.  It is possible that some of his errors in the latter language were likewise deliberate—certainly they have allowed us to spot plagiarism in the writings of others. Collie (1982: 231) gives examples of Borrow’s becoming flustered and actually running away when confronted with questions concerning his knowledge of foreign languages.

Ridler also remarked upon his ability to undertake a complete re-translation of his earlier version of St. Luke thirty years after having left Spain and any further contact with CalĂł speakers.  Her implication is that much of his new version was in what we might call Borromani, a concocted dialect not actually spoken by anyone.  Ridler goes on to provide examples of such created forms, e.g. his word for “incense” for which he used the Spanish word incienso in the original version, and which became usur-gudlo in the new version, compounded from usur “smoke” and gudlo “sweet.”  The first item, usur occurs only in Pott, who got it from Borrow’s Zincali in the first place; it isn’t in any other CalĂł wordlist.  And the second item, gudlo, has the English and not the Spanish Romani form, which is gulo. Another example of lexical creation Ridler cites is Brono Aljeñicato for “Pontius Pilate,” or PĂČncio Pilato in Spanish.  To Borrow’s ear, the name (in English moreso than in Spanish) sounded like puente and pila, which mean “bridge” and “fountain” in Spanish.  According to Borrow, the CalĂł words for these are brono and aljeñicato respectively, and both have been listed in his Zincali lexicon as legitimate CalĂł, and were furthermore entered as such, like usur, without comment by Pott in his dictionary (1844-II:433); but in his correspondence with the British and Foreign Bible Society he indicated that he made up both words himself, from Sanskrit and Arabic. 

The same creativity is evident in his English Romani versions.  Thus in the appendix on this dialect in the second volume of The Zincali, he includes his translation of the Apostle’s Creed; to the line “I believe in the Holy Ghost,” he appends the following note:

The English Gypsies, having, in their dialect, no other term for ghost than mulo, which simply means a dead person, I have been obliged to substitute a compound word.  Bavalengro signifies literally a wind thing, or form of air.

This word, respelt as bavol-engro, then turns up as the only entry for “ghost” in the Romano Lavo-Lil, published in 1874, while at the entry mullo in the same vocabulary, which is the actual English Romani word for both “ghost” and “dead,” only the meanings “dead man” and “dead” are given.

While it would be possible to reconstruct the lexical and grammatical characteristics of the kind of British Romani Borrow was hearing, such an undertaking might be more usefully achieved by looking at the work of others, e.g. by Whiter or Vallancey, because of Borrow’s tendency to mix dialects.  But even from his own material it is clear to see that English Romani is a member of the Northern European dialect group, sharing much in common with Sinti (e.g. such items as jin-, tikno, miro, grai, haier-, ma, chomoni, juvel, stif(o)-p(r)al, mukh (“know,” “small,” “my,” “horse,” “understand,” “don’t,” “something,” “woman,” “brother in law,” “let,”—compare Southern Romani jan-, tsino, mu(n)rro, gras, hatyer-, na, vareso, juvli, salo, mekh).6  Like Sinti, the infinitive seems to be modelled upon the third person singular indicative (Sinti kamav te jal “I want to go” rather than kamav te jav--cf. Central Romani kamav te jan), although Welsh Romani, like Vlax, does not do this.  Like Vlax but unlike the Central or Northern dialects, Welsh Romani negates “is/are” as nai, recorded by both Borrow and Smart & Crofton for English Romani.  Borrow correctly analyses this as na + hi, although unlike Northern and Central Romani, British (Welsh, Scottish and English) Romani, like Vlax, belongs to the /s/ group, thus si, san, lesa, sĂ„r rather than the /h/ group (hi, hal, leha, har, “is/are,” “am,” “with it,” “like/as”).  British Romani preserves the original negative construction with preverbal na, now lost in most dialects of Sinti, which has postverbal  ga(r) (< German).

The most authentic Angloromani Borrow reproduces, i.e. recorded verbatim from Romanichals, is probably found in his “beti rockrapens” and to a lesser extent in his collection of English Gypsy songs in the Lavo-Lil, but even here, material of his own creation is quite evident.  Thus in the section entitled “The English Gypsies,” he includes a song called “Tugney Beshor” which contains the line cauna volĂ©lan “when they fly,” using the Spanish root for “fly,” an item he does not enter in his dictionary.  In the same song is the word artavĂ vam, glossed as “we’ll forget.”  His knowledge of Romani grammar seemed to have deserted him by 1874, since this is an aorist construction, not a future one, and he also appears to have forgotten its real meaning, which is “forgive” rather than “forget” (in English Romani bister), but even as early as 1844 he was confusing tense and number endings, listing camenna “they (will) love” as “I love” and chorava “I (will) steal” as “I have stolen” (Winstedt, 1951).

In the dingle episode in Lavengro (Chapter 83) he is describing the manufacture of a horseshoe, and refers to the fire’s “tongues of flame” as vagescoe chipes, using the CalĂł word for “tongues” (vagescoe is probably a misprint for yagescoe here, though the form is continental rather than British, which would be yogeskro), and to an Aanvil” as a covantza, a Slavic-derived item listed in his Hungarian Romani vocabulary (Winstedt, 1951) and by Paspati (1870), and recorded by Smart & Crofton (1875, along with volĂ©lan and artavĂ vam) as only having been encountered by them in print in Angloromani, and in the works of Borrow. 

Borrow was evidently influenced by the Romani he heard in Eastern Europe, for even Jasper’s wife Sanspirella Heron was renamed with the Balkan-sounding Pakomovna in his stories.  The same Continental influence is evident in three unusual words which occur in Chapter 26 of Lavengro, where Jasper Petulengro is muttering to himself about the weather: dearginni “it thundereth,” villaminni “it flameth” (referring to the lightning) and grondinni “it haileth.”  With the exception of vĂ­lamo “lightning” in Bohemian Romani (JeĆĄina, 1886), the items themselves turn up in no other recorded dialect,7 though it is clear where Borrow found them, for their origins are, respectively, in Hungarian dörga and villĂĄm, and Romanian grindini.  It might now be possible to temper Winstedt’s caution that “it would be rash to attempt to define the precise dialect of Borrow’s ‘CzigĂĄny’” (1951:49); the morpheme {-in-} is characteristic of Hungarian Lovari loanverbs (like {-sar-} in other Vlax dialects), and the athematic  termination  {-i}  (more properly   {-ij}) for third person singular {-il} is likewise indicative of Hungarian phonological interference in the Lovari Vlax spoken in that country and in Transylvania.  Clearly these items were recorded during his visit to Hungary and Romania in 1844, seven years before he wrote Lavengro, although only rondĂ­ni “hail” turns up in his Hungarian Romani vocabulary (Winstedt, 1952:56).  The spellings both with and without initial /g/ may reflect trouble Borrow had in attempting to represent the Vlax voiced uvular fricative in this word. In Welsh Romani, “thunder” is devlĂ©ski gĂłdli “God’s noise” and “hail” is briĆĄindĂ©ske bĂĄra “rain stones;” “lightning” is molĂłna (from Slavic)8.  Smart & Crofton include all three of Borrow’s words, but mark each of them as unattested; the fact that two of the items don’t occur in Winstedt’s list lends support to his suggestion (1952:47) that Borrow may have collected more material in Cluj than has so far been located.

Borrow errs in his explanation of the plural morpheme, which he spells <-or> and which he sometimes applied to words indiscriminately and incorrectly (for example bauor “mates,” from English dialect bor “chum”)9.  Because he spoke a non-rhotic dialect of English, /r/ was not articulated in his speech except before vowels; after vowels, it merely indicated length.  This confusion is evident in such conflicting orthographic representations as both <saulo> and <sorlo> “early,” <sap> and <sarp> “snake,” <villaminni> and <villarminni> “lightning-flashes” and so on.  Thus what he wrote as <-or> in fact represents the sound “-aw”, which corresponds to the Common Romani plural postconsonantal morpheme {-a} and reflects the regular phonetic shift of [a] to “aw” (Sampson’s <Ă„>) in British Romani (e.g. his shockor <cabbages” for Common Romani ĆĄaxa &c.).  But on the basis of this non-existent /r/, Borrow assumes that the English Romani plural may be derived from the Romanian neuter plural {-uri}.9  It is possible, however, that by Borrow’s day, an intrusive /r/ could have found its way into a number of words, since in Angloromani as spoken in North America, it is clearly present in such words as gorjer, vonger, kekker and so on (where in British Angloromani these are pronounced gawja, vonga, kekka, &c.).  This behaviour of /r/ suggests that Borrow was dealing with south-eastern varieties of British Romani, since Welsh Romani is in the main rhotic (cf. feder, určos, sĂ„r, &c.).  Rhotacism as a phonological feature of English began to disappear from the south-eastern dialects in the early modern period (1500-1700), but survives today in Welsh English and in the South-Western dialect area, in which the KĂ„le lived before moving into Wales.  That phonological changes in English should affect the phonology of Romani spoken in that country suggests not only a high degree of bilingualism, but that the population probably became English-dominant early on.  A similar sound shift in English which has affected Romani is that of [er] to [ar] (cf. sartain, varsity, sergeant, clerk, &c.), thus the original [er] in erti- “forgive, excuse” (< Romanian) becomes ātav- in Welsh Romani, also demonstrating loss of postvocalic /r/, which is Borrow’s artav, [er] in verdo “waggon” (Common Romani vurdon) becomes Angloromani va:da or vawda, and terno “young” becomes Angloromani tawno with loss of /r/ and shift of [a] to [aw], cf. Welsh Romani tarno, Common Romani terno.  This latter, which is characteristically British Romani, also reflects a soundshift which began in South-Eastern (but not South-Western) English in the early modern period, cf. [grĂŠs], [hĂŠf], [lĂŠf] to [gra:s], [ha:f], la:f]; also cf. southern Angloromani gawja, yok, yog, yora (“non-Romani”, “eye”, “fire”, “egg”), &c., with northern (especially Scottish) gadji, yak, yag, yaro, &c.

Other misinterpretations which bear comment are Borrow’s assumption that the genitive postposition, legitimately a nominalizer in Central and Northern Romani, in its various forms (-engro, -eskro) is a free morpheme meaning “thing” or “fellow,” thus boro drom engroes “highwaymen,” (Lavengro, 358), also listed with the same interpretation in his Hungarian Romani vocabulary (Winstedt, 1952:110), and that the sociative postposition {-sa}, meaning “with” or “by,” could be detatched and made into a preposition: come sar mande, “come with me”.  It is entered as a separate word not only in his Lavo-Lil but also for Caló in the glossary in The Zincali, and turns up in Leland’s English Gypsy songs and in Smart & Crofton.  Sampson (1907:95) doubts that this ever existed as a separate preposition, suggesting that these various compilers, like Borrow himself, misinterpreted the postposition as they heard it.11

His inventiveness is betrayed by his poor knowledge of Romani grammar in his discussion of a verse he calls “the oldest specimen of English Gypsy at present extant, and  perhaps  the purest . . . at least as old as the time of Elizabeth” (1923:11).   The lines in question, with his own translation, are

  Coin si deya, coin si dado         “who’s your mother, who’s your father”
  Pukker mande drey Romanes,  “do thou answer me in Romany”
  Ta mande pukkeravava tute.     “and I will answer thee”

Coin is the form of Common Romani kon which he imported into the Romanichal dialect from his vocabulary of Spanish Romani, though it is not attested in any other dialect; deya and dado are both vocatives and cannot follow the verb si, the former taken from his Lovari wordlist, and not found in Britain (though he does have the British form daiya elsewhere in his book). In the second line, Romanes is an adverb, though he doesn’t seem to have realized this; he has it glossed as a noun in his dictionary, and believing it to be a noun has it following the preposition drey.  In the third line, he has the inflected forms mande and tute (“to me,” “to you”) functioning as personal pronouns as they do only in Angloromani, while pukkeravava is a causative verb form, meaning “I’m being made to tell.”  The verses, in British Romani, ought to have read kon si tiri daj, kon si tiro dad, phuker mange romanes tha (me) phukerava tuke.

Borrow also treated the bound affix {-i-sar} (or {-a-sar}), a marker of loanverbs, as though it were a free morpheme, and entered it into his Lavo-Lil as the independent item asarlas “at all.”   However, his own example originates in the misanalysis of we can’t help asarlus, which he glosses as “we can’t help at all,” when the construction he heard was we can’t helpasar les, “we can’t help it.”  It is possible, however, that this development was in part a legitimate one in English Romani, since Smart & Crofton (1875:52-54) list many examples of asár meaning “also” which clearly demonstrate a reinterpretation of earlier {-i-sar}.  But it is not paralleled in Welsh Romani nor has the form and use seemed to have survived in modern Angloromani.  This athematic morpheme is particularly characteristic of the Vlax group of dialects, and is used with loanverbs in Welsh Romani too, though with one exception only in imperative constructions (Sampson, 1926:117-118).  Another indication of Vlax influence is the non-final affix {-n-}, a marker of athematic adjective-derived adverbs, thus drago, dragones, mundro, mundrones, compared with thematic baro, bares, tikno, tiknes, without the {-n-}.  Thus we have such Borromani forms as weshenjugalogonés, bolli-menggreskoenés, dinnileskoenés, dovodoiskoenés and so on.  This last is compounded from dovo “that,” + odoi “there,” and is given to mean  “in that way,” while  that  item  and  the  two  before it  include -(e)sko, which consists of the singular masculine oblique nominal morpheme {-es} plus the Vlax genitive {-ko} (rather than the Northern Romani {-k(e)ro}), and followed by the adverbial {-es} linked with non-athematic {-n-}.  In actual Angloromani, dovodoi is a legitimate determiner meaning “that there,” but not even in inflected British Romani could it combine with grammatical particles meant for nouns and adjectives.  Yet such forms are presented as “genuine Gypsy. . . clear-sounding and melodious, and well-adapted to the purposes of poetry” (Lavo-Lil, page 11; see also Hancock, 2003). 

George Borrow probably never dreamt that his work would come under scrutiny a century and a half after it appeared; but at the time it was written, interest in Gypsies and the Romani language was marginal, and remained so until almost the end of the 20th century.  His writings were evaluated by his contemporaries not at all for the Romani they contained, but for his observations and descriptions and opinions in other areas.  Those few individuals, such as Pott (1844), Ascoli (1865), Paspati (1870) Miklosich (1872) and Colocci (1889) who had an academic interest in Romani, on the other hand, used his material uncritically, and were scarcely interested in his writings otherwise.

5.  Conclusion

Fifteen years ago, Angus Fraser summed up a talk he gave on George Borrow at The University of East Anglia12 with the words “I wish I could feel sure that a new generation of Borrovians is growing up.”  His wish has clearly been fulfilled; there is now a society and a journal dedicated wholly to Borrow’s life and works, and an international conference on the man has become an annual event, though whether the impetus will be maintained since his death in 2001 remains to be seen.  As specialists in Romani, we must be careful to separate our judgment of Borrow’s linguistic expertise in that language from our judgment of him as a prose writer and raconteur.  In the latter role, he holds a special place; and even as a linguist dealing with other languages, as Ridler (1996) has so magnificently documented, he demonstrated a remarkable genius.  It is all the more puzzling, therefore, that he dealt so inadequately with the one language for which he is remembered best of all.13 As a student of Gypsies, he must only be regarded as prime example of the term he himself created, a Romany Rye—for today, the word has been taken back by Gypsies, and is a term of disdain.  It is no coincidence that Kalderash Romanies in Europe and America use the word rai for a policeman or an authority figure, meanings also shared in England by the present-day descendants of George Borrow’s Romani brothers and sisters.

NOTES

 1. The original is in Hansard (see Wells, 1961).

 2. Borrow’s observations on the Iranian element in Romani (1841(ii):111-112) are worth reproducing (as an adjunct, for example, to Hancock, 1995a):

“Still more abundant . . . than the mixture of Greek, still more abundant than the mixture of Sclavonian, is the alloy in the Gypsy language, wherever spoken, of modern Persian words, which circumstance will compel us to offer a few remarks on the share which the Persian has had in the formation of the dialects of India, as at present spoken.
“The modern Persian, as has been already observed, is a daughter of the ancient Zend, and, as such, is entitled to claim affinity with the Sanscrit, and its dialects.  With this language none in the world would be able to vie in simplicity and beauty, had not the Persians, in adopting the religion of Mahomet, unfortunately introduced into their speech an infinity of words of the rude coarse language used by the barbaric Arab tribes.”

 3. The age and identity of Jasper is the topic of an article by Fraser (1996:7-12).

 4. Borrow was probably poisoned not by “Mrs. Hearne” (Martha Boss in real life) but by her daughter Joni (Borrow’s “Leonora”).  See Fraser, 1996:7-12.

5. It is interesting that Borrow should have given this impression, considering his disdain for others who have done the same thing.  Describing himself in The Romany Rye as one who “affects to be neither Frenchman, nor German,” he scorns those who “affect the airs of Spaniards [and who . . . ] make Tom-fools of themselves by sticking cigars into their mouths, dressing themselves in zamarras, and saying carajo!” (1857, II: 273).

His  adversary Charles Godfrey Leland actually became angry at being referred to as a non-Romani (1882:121-2):

“As we came up the street, I saw the man talking with a well-dressed, sporting-looking man, not quite a gentleman, who sat cheekily in his own jaunty little wagon.  As I passed, the one of the wagon said to the other, speaking of me, and in pure Romany, evidently thinking I did not understand—“dikk’adovo Gorgio, adoi! (Look at that Gorgio, there!).  Being a Romany Rye, and not accustomed to be spoken of as a Gorgio, I looked up at him, angrily, when he, seeing that I understood him, smiled, and bowed politely in apology.  I laughed and passed on.”

Like the use of salo (“brother in law”) among Balkan Romanies as a disparaging term for a cuckolded husband, the designation rye, for some ryes at least, seems to have had a more specific in-group meaning: managing to bed a Romani woman.  Thus in a letter dated November 6th, 1908 Augustus John wrote to fellow gypsilorist Scott Macfie

“I have recently taken it upon myself to confer the title of Rai upon a friend of mine—one Percy Wyndham Lewis, whose qualifications, the having coupled and lived in a state of copulation with a wandering Spanish romi in Brittany, seemed to me upon reflection to merit the honourable and distinctive title of our confraternity.”

 6. Scottish Romani on the other hand appears to have entered Britain via Scandinavia rather than from the northern European coast; see Hancock, 1977.

 7. In Kalderash Vlax these words are trĂ©zneto or rrĂłndjeto for “thunder” and bĂĄbica for “hail,” both from Romanian, and strĂ©lica for “lightning,” from Slavic; no Lovari Vlax wordlist contains Borrow’s three items.

 8. The same metaphorical compounding, especially typical of Northern Romani, is found in Sinti, thus “thunder” is devleskero čiro (“God’s noise”), and “lightning” is devleskeri jakh (“God’s eye”).

 9. Bor (from Old English bār, and retained in the word “neighbour”) is listed by Wright (1898-I:345) as meaning “a term of familiar address” and localized to Cumberland, East Anglia and Essex.  Borrow’s created plural should have been *boror rather than “bau-or.”

10. The Rumanian-derived plural morpheme {-uri} does, however, seem to have made its way into British Romani otherwise; see Hancock, 1984:103 and 1995b:30.

11. Sampson (1907) demonstrated convincingly that sar in British Romani was a literary creation, though Peter Bakker (in p.c.) points out that this may not be the case for Caló, referring to texts collected by de Luna (1951) in which sar is used prepositionally.  Norbert Boretzky has also documented its prepositional use in some Balkan Romani dialects.

12. “George Borrow: The romance and the reality,” lecture given at the University of East Anglia, June 12th, 1981.

13. This would suggest strongly that he worked best from written materials, and perhaps had a photographic memory; Romani was the only language in his working repertoire for which he had no access to a comprehensive grammatical description.

Works Mentioned in the Text

Ackerley, Frederick G., 1942.  [Review of Potra, 1939], Journal of the Gypsy Lore Society, 21(1):68-71.
Allingham, P., 1934.  Cheapjack.  London.
Anon., 1879.  “Gypsy life round London.”  The Illustrated London News, April 29th, p. 503.
Ascoli, G.J., 1865.  Zigeunerisches.  Halle: Julius Fricke.
Behlmer, George, 1985. “The Gypsy problem in Victorian England,” Victorian Studies, 48(2):231-253.
Boorde, Andrew, 1547.  The Fyrst Booke of the Introduction of Knowledge.  London: Copland.
Borrow, George, 1841.  The Zincali, or an Account of the Gypsies of Spain.  London: John Murray.  2 volumes.
Borrow, George, 1843.  The Bible in Spain. [Dent 1947 edition].
Borrow, George, 1851. Lavengro: Scholar, Gipsy, Priest. [Collins 1910 edition].
Borrow, George, 1857.  The Romany Rye: A Sequel to “Lavengro.” [Murray, 1924 edition].
Borrow, George, 1862.  Wild Wales. [Collins undated edition, with an introduction by Brian Rhys].
Borrow, George, 1874.  Romano Lavo-Lil, Word Book of the Romany or, English Gypsy Language. [Murray 1910 edition].
Borrow, George, 1874.  Romano Lavo-Lil, Word Book of the Romany or, English Gypsy Language. [Alan Sutton 1982 edition].
Bryant, Jacob, 1785.  “Collections on the Zingara or Gypsey language,” Archéologia, 7:387-394.
Burke, Peter, & Roy Porter (eds.), 1995.  Languages and Jargons: Contributions to a Social History of Language.  Polity Press.
Chamberlain, Houston Stewart, 1899.  The Foundations of the Nineteenth Century.  London.
Collie, Micheal, 1982.  George Borrow, Eccentric.  Cambridge: The University Press.
Colocci, Adrian, 1889.  Gli Zingari: Storia d’un Popolo Errante.  Turin: Loescher.
Conan Doyle, Arthur, 1913.  “Borrow-ed scenes,” The Pall Mall Magazine, 52:348-357.
Croley, Laura Sagolla, 1996.  “Wanderers and Settlers:” Vagrancy, Begging, and the English Middle Class, 1780-1867.  Unpublished doctoral dissertation, The University of Pennsylvania.
de Luna, José Carlos, 1951.  Gitanos de la Bética.  Madrid: Ediciones y Publicaciones Españolas S.A.
Dillmann, Alfred, 1905.  Zigeuner-Buch.  Berlin: Widsche.
Douglas, Mary, 1970.  Natural Symbols.  New York: Vantage Press.
Fraser, Angus, 1993.  “Authors’ Gypsies.”  Antiquarian Book Monthly, 20(2):10-17.
Fraser, Angus, 1995.  “Borrow, the Palgraves, and the Worships,” The George Borrow Bulletin, 9:00-00.
Fraser, Angus, 1996.  “The parallel universe of Borrow’s Gypsies,” The George Borrow Bulletin, 11:7-12.
Geipel, John, 1995.  “The ‘secret’ language of the Gypsies of Spain,” in Burke & Porter,1995:102-132.
Gilliat-Smith, Bernard, 1909.  “Russian Gypsies in Lithuania, July 1908,” Journal of the Gypsy Lore Society, n.s., 3(2):154-155.
Gobineau, Arthur Compte de, 1855.  Essai sur l’InĂ©galitĂ© des Races Humaines.  Paris.
Grellmann, Heinrich M.G., 1787. Disseration on the Gipsies.  (Matthew Raper translation), London: Elmsley.
Grosvenor, Lady Arthur, 1908.  “Whiter’s ‘Lingua Cingariana’,” Journal of the Gypsy Lore Society, n.s., 2(2):161-179.
Haines, Samuel M., 1989. The Travelers.  Privately-circulated report on Gypsy Crime, Swindle Section, Dallas Police Department.
Hancock, Ian, 1977.  The Social and Linguistic Development of Angloromani.  Working Papers in Sociolinguistics No. 38.  Austin.  Reprinted in Kenrick & Acton (1984), pp. 89-122.
Hancock, Ian, 1981.  “Talking back.”  Roma 6(1):13-20.
Hancock, Ian, 1984.  “Romani and Angloromani,” in Trudgill (ed.), 1984:367-383.
Hancock, Ian, 1993.  “Antigypsyism in the new Europe.”  Roma, 39:17-35.
Hancock, Ian, 1995a. “On the migration and affiliation of the Domba: Iranian words in Rom, Dom and Lom Gypsy,” Papers of the First International Conference on Romani Linguistics, Y. Matras, (ed.), Amsterdam: Benjamins, pp. 29-59.
Hancock, Ian, 1995b.  A Handbook of Vlax Romani.  Columbus: Slavica.
Hancock, Ian, 2003.  “The Concocters: Creating fake Romani culture”, in Saul & Tebbutt (in press).
Helyear, Annick, 1972.  The British Traveller: The Romani Civilization, its Mythical Presentation in the English Literature, and the Social Problems Today.  DiplĂŽme d’Etudes SupĂ©rieures d’Anglais, UniversitĂ© d’Aix en Provence.
Jarman, A., & Eldra Jarman, 1991.  The Welsh Gypsies: Children of Abraham Wood. Cardiff: University of Wales Press.
Jeơina, P. Josef, 1886.  Romáƈi Čib, oder die Zigeunersprache.  Leipzig.
Kenrick, Donald, & Thomas Acton, eds., 1984.  Romani rokkeripen To-divvus.  London:  Romanestan Publications.
Knapp, William I., 1899.  Life, Writings and Correspondence of George Borrow, Derived from Official and Other Authentic Sources.  London: Murray.  Two vols.
Leland, Charles G., J. Palmer & E. Tuckey, 1875.  English-Gipsy Songs.  London: TrĂŒbner.
Leland, Charles G., 1882.  The Gypsies.  Boston: Houghton, Mifflin & Co.
Marsden, William, 1785.  “Observations on the language of the people commonly called Gypsies,” Archéologia, 7:382-386.
Mayall, David, 1988.  Gypsy-Travellers in Nineteenth Century Society.  Cambridge: The University Press.
Miklosich, Franz, 1872-1881.  Ueber die Mundarten und die Wanderungen der Zigeuner Europa’s.  Vienna: Karl Gerold.
Miklosich, Franz, 1874-1878.  BeitrÀge zur Kenntniss der Zigeunermundarten.  Vienna: Karl Gerold.
Paspati, Alexandre G., 1870. Etudes sur les TchinghianĂ©s ou BohĂ©miens de l’Empire Ottoman.  Constantinople: KoromĂ©la.
Potra, Gheorghe, 1939.  Contribuƣiuni la Istoricul ƹiganilor dßn Romùnia.  Bucharest.
Pott, August F., 1844.  Die Zigeuner in Europa und Asien.  Halle: Heynemann Verlag.
Richardson, David, 1801.  “An account of the Bazeegurs, a sect commonly denominated Nuts,” Asiatick Researches, 7:457-485.  Bryant’s vocabulary, pp. 481-485.
Reid, Alastair, 1962.  “The Travellers,” The New Yorker, August 18th, pp. 37-40.
Ridler, Ann M., 1981.  “Sidelights on George Borrow’s Gypsy Luke.” The Bible Translator, 32(3):329-337.
Ridler, Ann M., 1996 [1983].  George Borrow as linguist: Images and contexts.  Doctoral thesis submitted to the Council for National Academic Awards, Oxford, 1983.  Published by the author for private circulation as George Borrow as a linguist: Images and Contexts. 1996.  All in-text references are to this later version.
Sampson, John, 1907.  “sar, ‘with’,” Journal of the Gypsy Lore Society, n.s., 1:95-96.
Sampson, John, 1926.  The Dialect of the Gypsies of Wales.  Oxford: The Clarendon Press.
Saul, Nicholas, & Susan Tebbutt (eds.), 2003.  The Role of the Romanies: Images and Self-Images of “Gypsies”/Romanies in European Cultures.  Liverpool: Liverpool University Press.
Shields, Audrey C., 1993.  Gypsy Stereotypes in Victorian Literature. Unpublished doctoral dissertation, New York University.
Smart, Bath C., & Henry T. Crofton, 1875.  The Dialect of the English Gypsies. London: Asher & Co.
Smith, L.A., 1889. Through Romani Songland, London: Stott.
Strong, T.B., & S.A. Stondemire, (eds.), South Atlantic Studies for Sturgis E. Leavitt, Washington: The Scarecrow Press.
Tilford, J.E., 1953.  “A note on Borrow’s bookish dialogue,” in Strong & Stondemire, 1953:199-206.
Trudgill, Peter, (ed.), 1984. Languages in the British Isles.  Cambridge: The University Press.
Vallancey, C., 1804.  “On the origin and language of the Gypsies,” Collecteana de Rebus Hibernicis, Volume 6, Ch. 2, pp. 61-120.
Vesey-Fitzgerald, Brian, 1944.  Gypsies of Britain: An introduction to their history.  London: Chapman & Hall, Ltd.
Wells, John, 1961. [Comments regarding Borrow].  Hansard (Commons), Vol. 650, cols. 829-830, December 1st.
Whiter, Walter, 1800.  Etymologicon Magnum.  Cambridge: Hodson.
Whiter, Walter, 1822-25.  Etymologicon Universale.  Cambridge University Press. In three volumes.
Wilson, Nerissa, 1986.  Gypsies and Gentlemen.  London: Columbus Books.
Winstedt, Eric Otto, 1952. “Borrow’s Hungarian-Romani vocabulary.”  Journal of the Gypsy Lore Society, 29(1/2):46-54, 103-115, 30(1):50-61.
Wright, Joseph, 1898.  The English Dialect Dictionary.  London: Henry Frowde.  In six volumes.