Welsh Romani language

The Welsh Romani language (Romnimus, or Romimus, Romanes and Romanī čib; Welsh: Romani Cymreig) is a variety of the Romani language spoken by the Kale subgroup of the Romani people, who have been present in Wales since the 18th century.^[1]

Quick facts Native to, Language family ...

Welsh Romani
Romnimus
Native to	Wales (United Kingdom)
Language family	Indo-European Indo-Iranian Indo-Aryan Romani Northern Romani British Romani Welsh Romani
Language codes
ISO 639-3	`rmw`
Glottolog	`wels1246`

The majority of the vocabulary is of Romani origin but there are a number of loanwords from other languages. Welsh loanwords include melanō ("yellow", from melyn), grīga ("heather", from grug) and kraŋka ("crab", from cranc). There are also English loanwords such as vlija ("village"), spīdra ("spider") and bråmla ("bramble").^[2]

Historically the variants of Welsh Romani and Angloromani (spoken by the Romanichal) constituted the same variant of Romani, known as British Romani before the 18th century.^[3] Welsh Romani is one of the many Northern Romani dialects.^[1] Welsh Romani is closely related to Angloromani, Scandoromani (as spoken by the Romanisael in Sweden and Norway), Sinte Romani (as spoken by the Sinti), and Finnish Romani (as spoken by the Finnish Roma in Finland and Sweden), are closely related to British Romani. For instance, Welsh Romani grai ("horse"), Angloromani grai ("horse"), Sinte Romani grai ("horse") and Finnish Romani graj ("horse") versus the Vlax dialect's Kalderash and Lovara grast ("horse"). The Kale, Romanichal, Romanisael, Kaale and Scottish Lowland Romani are closely related Romani subgroups which stem from the wave of Romani immigrants who arrived in Northern Europe in the 16th century.^[4]

Remove ads

Phonology

More information Labial, Dental ...

	Labial	Dental	Alveolar	Post- alveolar	Palatal	Dorsal	Glottal
Nasal		m			n				ŋ
Stop	p pʰ	b		t tʰ	d	tʃ	dʒ		k kʰ	ɡ
Fricative	f	v	θ	ð	s	z	ʃ	ʒ		x	ɣ	h
Trill			r̥	r
Approximant							j	ʍ	w
Lateral			ɬ	l

More information Front, Central ...

	Front	Central	Back
Close	i		u
Mid	e	ə	o
Open		a

History

Summarize

Perspective

This section does not cite any sources. (September 2025)

By the nineteenth century, scholars widely believed that the fully grammatically inflected Romani language in Britain had declined to the point of near extinction. Anglo-Romanes (Anglo-Romanes was the term for both the "deep" and broken dialect in early material) had undergone significant erosion, losing most of its inflectional system and much of its original vocabulary, and surviving mainly as a para-Romani variety used alongside English. The remnants of “deep Romanes" in Britain were said to be preserved only among a handful of elderly Romanichal speakers and were regarded as little more than fragments of a once fully structured language. Despite this general view, there were reports of more complete Romani usage in Wales during the late nineteenth century around the same time. In 1874, H. T. Crofton reported meeting Oliver Lee, the son of an English Romani father and a Welsh Romani mother, who provided him with examples of what was described as “deep Romani”:

In September 1874 I met with a Welsh Gypsy, Oliver Lee, at Bettws-y- Coed, North Wales. His father was an English Gypsy from the Midland Counties; his mother was one of the Woods, patricians amongst Welsh Gypsies. He was born, and had always lived, in Wales; was about twenty-two years old, but, unlike most of the rising generation in England, he could converse in both deep and broken Romanes, as well as Welsh and English. He and his wife had just been joined by some of her relatives, natives of Worcestershire, but Welsh by adoption; whose children spoke English with a Welsh accent, and some of whom had married amongst the Welsh. I gathered from Oliver that his two aunts, Mary Wood, nicknamed Taw (W., silent), and Caroline Wood, both aged about forty, spoke Romanes habitually, and only used English or Welsh when talking to gaujos. After satisfying myself of Oliver's knowledge of the old forms, I read to him "The Widow's Son," "The Licence,' "Zuba B," and "The Fairies," all of which he interpreted correctly to his companions, the eldest of whom seemed to have a hazy recollection of several of the verbal inflections, and kept exclaiming, "It's just as I used to hear the old folk talking when I were a lad." A reference to the stories themselves will indicate how far the deep Anglo-Romanes corresponds with the current Welsh-Romanes. We did not, however, think we were warranted in concluding that the dialects were so far distinct that we must exclude my notes from the vocabularies...

Sampson, the author of the monumental Dialect of the Gypsies of Wales, was of the opinion that "the examples obtained from Oliver, which are in the hybrid vernacular peculiar to this Welsh branch of the Lees, cannot be considered as representative of the classic tongue of the Woods."

In 1876, Francis Hindes Groome made further contact with John Roberts of Newtown, known as the Queen’s Romani harpist. Roberts was unusual in that he could both speak and write his own language. Groome obtained texts from him, some of which he later published in the Encyclopaedia Britannica (9th ed., 1879) and in In Gypsy Tents (1880). After Roberts’s death, a folk tale in his hand was published in the Journal of the Gypsy Lore Society (1892). Groome’s interest, however, lay primarily in folklore rather than linguistics, and the texts he published, “interspersed as they were with much English” and written in Roberts’s unsystematic orthography, did not "draw adequate attention to the wonderful preservation and structural perfection of the dialect".

The most significant rediscovery occurred in 1894, when John Sampson, then a young philologist, encountered Edward Wood, a Romani harpist, in Bala, North Wales. Sampson was struck by the quality of Wood’s speech, describing it as “hardly less pure than that of Paspati’s Tchinghianés.” He later concluded that Welsh Romani represented “a survival of the oldest and purest form of British Romanī." The Woods were direct descendants of Abram Wood, a prominent ancestral figure in Wales born in the turn of the seventeenth century, who was remembered as the ancestor of the Woods. Sampson recounts:

The harper Edward, son of the harper Adam, son of the harper Valentine, eldest son of Abram Wood, remained for some time my chief instructor, and from this gentle and very intelligent friend, himself trilingual like all the members of his family, I learned to master the new sounds, words and inflections. From Edward, too, I ascertained the habitat and favoured camping-places of the scattered scions of the 'teulu Abram Wd' , with their congeners the Griffithses, Robertses and Prices, and as the result of many pilgrimages throughout North and Central Wales became acquainted with almost every Welsh Gypsy of the original stock—among them many who with Edward himself are now gathered into their ancestor's bosom. By the banks of Tal-y-llyn in 1896, led thither from Machynlleth by one of the harper sons of John Roberts, I found Matthew Wood and his four boys, all of whom were faithful conservators of the language, as well as unequalled raconteurs of the ancient folk-tales with which it is the Gypsy custom to enliven the long winter evenings.

While Sampson’s earliest materials came from Edward Wood, his later collections were obtained from Edward’s cousin Matthew Wood and his family. Sampson first met Matthew Wood in the mid-1890s and maintained contact with him for over two decades. Initial encounters took place near Abergynolwyn, where Sampson collected stories and songs from Matthew, described as a fiddler and fisherman.

In 1909, Sampson established a base near Bettws Gwerfil Goch, Denbighshire, at a site overlooking the river Alwen and the surrounding hill country. He described the location, a spur of Bron Banog, as a “pleasant eyrie” which became his summer headquarters for several years. The place was a regular meeting ground for Romani families, and during this period Matthew Wood and his eldest son Henry (“Turpin”) Wood were settled nearby.

Several members of the extended family played key roles in Sampson’s research. Among the most prominent was Matthew’s mother, Saiforella (Mary) Wood, known as “Tau.,” mentioned decades prior by Oliver Lee to H. T. Crofton. Despite her age, she remained active and possessed of "wonderful spirit", and "inspired by a strong racial devotion", travelled alone across North Wales to visit relatives and family burial places. She was particularly devoted to the churchyard of Llangelynin, on the coast near Tywyn, which she regarded as a shrine because it contained the grave of Abram Wood, her great-grandfather and the ancestral figure of the Welsh Kale. Other relatives contributed as informants. Matthew’s uncle and aunt, Benjamin and Caroline Wood, travelled widely in Anglesey and Caernarfonshire with their sons Cornelius and Adolphus, both fiddlers. In Newtown, Montgomeryshire, one of the earliest Welsh Romani settlements, Sampson recorded from members of the Wood-Roberts branch of the family. At Aberystwyth he collected material from Matthew’s sister, Eldorai Wood, while in Pwllheli he worked with Llewellyn Wood, son of Edmund Wood, a harpist who wintered there. All of these individuals were reported to be competent speakers of Romani and provided Sampson with both linguistic and cultural material. Sampson said:

My collections have been gathered in every part of Wales where members of the clan were to be found, following the Gypsy avocations of harpers, fiddlers, fishermen, horse-dealers, knife-grinders, basket-makers, woodcutters, fortune-tellers, and hawkers. From ancient men and women, their faces a complex of wrinkles, to tiny children out of whose mouths Romanī falls with a peculiar charm, all have been laid under contribution—nay, even those dead and gone long before my own day. For the older Gypsies delight to dwell upon the scenes and figures of their childhood, and in doing so recall the actual words of their forbears and kinsfolk—of the famous harpers Archelaus, Jeremiah, and Theophilus; of William, who had the mishap to be bičadō pārdal or transported; of Alabaina, the yellow witch, who marrying a gigantic gā̊jō of Lleyn founded the sub-clan of Jones; of the Romanī Shahrazad, 'Black Ellen', accredited with knowing 300 folktales, some long enough to occupy a whole night in the recital; and of many another Kā̊lō and Kā̊lī whose sayings and doings seem so familiar that I could fancy I had lived among them myself.

Sampson’s fieldwork was immersive and took over 20 years. He spent long periods travelling with Romani families, recording material at fairs, in barns, at campfires, and in domestic settings. Over the course of his research he filled more than one hundred notebooks with lexical items, grammatical paradigms, narratives, and folk tales. He emphasized that his data consisted entirely of spontaneous usage: “Every Romani sentence given in the Grammar or Vocabulary is the spontaneous utterance of some Welsh Gypsy, reflecting the life and lore, customs, beliefs, thought and feeling of the race.”

The collection process was not straightforward. Sampson noted that speakers often denied knowledge of expressions they had previously used, only to recall them later in different contexts. He also observed that emotions or discussions could provoke sudden recollections, bringing to the surface “a mass of half-forgotten words and expressions recovered from the depths of the past.”

Although he presented "Welsh Romanī" as unusually well-preserved, Sampson also observed signs of decline:

But while I have been fortunate in happening upon Welsh Romanī in its Augustan, or at least its Silver Age, I cannot disguise from myself that decay has already begun to set in, and that another generation or two may see the end of this ancient speech. The Woods, probably as the result of repeated consanguineous unions, are not a prolific race, while inter-marriages with Welsh Gā̊jē, or of late years with the English Gypsy clans of Lee and Locke, have resulted in the deterioration of the language. A surprising number of Welsh Gypsy adults remain single; the Ingrams have long ceased to exist as a co-clan with the Woods; the younger generation of the Roberts have almost lost the tongue of which the veteran harper was so proud; while not a few pure-bred Gypsies bearing the old patrician name (among them Edward himself), marrying aliens, have neglected to hand down Romnimus to their children.

In 1957, Derek Tipler recorded Welsh Romani from the Lee, Roberts, and Jones families and published his findings in the Journal of the Gypsy Lore Society. His work provided one of the few substantial accounts of the dialect following Sampson’s Dialect of the Gypsies of Wales (1926) and confirmed that the language continued to be spoken among these families, despite ongoing phonetic and grammatical changes over the 31 years since Sampson's publication. Scholars within the field of Romani Studies agree that Welsh Romani, as a fully grammatical language, is now extinct. Yaron Matras, a prominent Romani linguist, in discussion upon Romani dialect classification, notes:

British Romani, an independent branch, is now considered extinct. The most thorough and extensive description is Sampson’s (1926) monumental grammar of Welsh Romani or the Kåålē dialect, which was still spoken by a number of families until the second half of the twentieth century (cf. Tipler 1957).

Many scholars do not note that the Welsh Romani recorded by Tipler was already in decline and exhibited notable phonetic and grammatical changes that are absent from Sampson’s earlier recordings.

Remove ads

General Terminology

Summarize

Perspective

This section does not cite any sources. (September 2025)

Like other Romani dialects, the adverb romanes and the noun romanī čib appear. However, in Welsh Romani they are not the predominant terms used to designate the language by its own speakers. Rather, Romnimus—commonly pronounced Romimus—served as the primary term employed by speakers themselves. The latter two do not exist in continental Romani dialects or its closest relative, Angloromani. All three terms refer to Romani in general, with Kā̊lē distinguishing between ‘mā̊rō romimus ("our Romani") and those such as Aŋitrákō rom[n]imus ("English Romani") the term for Angloromani. P’agerdō rom[n]imus ("broken Romani") is likely a synonym for Angloromani.

More information Page, Entry ...

Page	Entry	Ad.	Meaning	Sampson's Examples (1926)
318-319	romanes	adv. [romanō + -es]	Gypsy-like, in Gypsy fashion, to Gypsy taste.-	Dik’áva les mē romanés. It has a Gypsy look to me. Dita, bā̊, romanés bešéla . See, friend, he sits like a Gypsy (i.e. cross-legged). Mištō, čai, kedán ‘dovā romanés. Well done, girl, thou didst that like a Gypsy. Romanés dik’ā mē ī bā̊rī båšimáŋerī: andilē šamas lasa. The harp seems gypsy-like to me: we were used to it from childhood (lit.’ reared with it’).
a. (with vbs. of speaking) To talk Gypsy.-	‘Dolā te rakeréna romanés, romanē si ‘dolā: kamáva len mē. Those who speak Gypsy are the same as Gypsies: I love them.
romnimus, oft. proned. romimus m. 6 c.	[romanō + -mus]	Gypsy language, Romanī.	‘Mā̊rō romimus, our (i.e. deep Welsh) Romanī. Aŋitrákō (vavērt’eméskō ) r., English (foreign) Romanī. P’agerdō r., broken Romanī. Kek lav romnimus, not a single word of Gypsy. Rakerdásas romnimus kitanés ‘lan ī penyā. We used to speak Gypsy together when we were in the convent (lit. ‘before the sisters’). ‘J̌ā̊-sār Aŋitrákerō ši-lō ‘kerát, ‘šiš andéla romimus avrī kek. He is like an English Gypsy to-night, he cannot get a Romani word out.
61	čib f. 2 a.		1. Tongue 2. Talk, chat, impudence 3. Language	Romanī čib, the Gypsy tongue.

Erroneous Terminology

The designation Kalá for the language is found only on Omniglot, where it appears alongside the terms Romnimus and Kåålē. The latter, however, is a historical ethnonym rather than a linguistic label. Importantly, Kalá is not attested as a name for the language in earlier sources or in modern academic literature. Academics, including those by Sampson (1926), Tipler (1957), Wood (1973), Jarman and Jarman (1991), Kenrick (1998), Clarke (1999, 2006), Acton (1997), and Matras (2002), make no use of the term.

A related form, Kalo (without the accent), appears occasionally online as a name for the language. Its earliest documented use occurs in Jarman and Jarman’s The Welsh Gypsies: Children of Abram Wood (1991), where it is employed as the plural form of the ethnonym rather than a linguistic term. It is worth noting that another Romani subgroup, the Calé, have a cognate ethnonym and speak Caló, a distinct Para-Romani dialect.

Remove ads

Grammar

Summarize

Perspective

This section does not cite any sources. (September 2025)

Declension 1

Masculines in -ō (dropped in the inflected cases), nom. pl. -ē, obl. sg. -es-, obl. pl. -en-. Included in this class are a number of adjectives used substantivally, e.g.: bā̊rō 'giant', kā̊lō 'Gypsy', p'urō 'old man', melanō 'sovereign '.