Concerning Hobbits: Welsh Fairies in Oxford * (26.08.13 by Simon J. Cook) - Comments

The very first thing that strikes one, in reading the Mabinogion, is how evidently the medieval story-teller is pillaging an antiquity of which he does not fully possess the secret; he is like a peasant building his hut on the site of Halicarnassus or Ephesus; he builds, but what he builds is full of materials of which he knows not the history, or knows by a glimmering tradition merely; – stones “not of this building,” but of an older architecture, greater, cunninger, more majestical.

Matthew Arnold; quoted by John Rhys in his preface to The Text of the Mabínogían and other Welsh Tales from the Red Book of Hergest, 1887, pp. ix-x. (1)
Alan Lee - Mabinogion - peredur son of efrawg


This essay explores J.R.R. Tolkien’s engagement with John Rhys’s reading of Welsh fairy stories. Rhys was the authority on all things Celtic in the Oxford of Tolkien’s undergraduate days. In his opinion, Welsh tales of fairies were echoes of dim memories of ancient encounters with a non-Indo-European aboriginal population of the British Isles. It is argued that Tolkien discovered hobbit stories (i.e. stories told about and initially by hobbits) behind Rhys’s image of Britain’s pre-Celtic inhabitants.

The essay begins with a discussion of Rhys’s place in the history of scholarship and then outlines his reading of Welsh fairy stories. The main part of the essay consists of an enquiry into the different ways in which Tolkien can be read as responding to Rhys’s reading. It is first shown how Tolkien imagined fairies and fairy tales as belonging to England rather than to Wales. It is then argued that Tolkien believed that, even if he had been wrong in his particular conclusions, Rhys had nevertheless been right to ask what lay behind the Welsh tales. Finally, it is suggested that Tolkien discovered that Rhys had actually produced (an admittedly somewhat confused) account, not of fairies, nor of pre-Aryan aborigines, but of hobbits.

Taken as a whole, the essay both departs from and contributes to recent Tolkien scholarship. In focusing upon Rhys over and above the actual fairy stories that Rhys wrote about, the essay leaves the highway of English literature for the untrodden ways of the history of ideas. Such a path, however, leads to a fresh perspective upon the recently discussed question as to whether Tolkien’s ‘mythology for England’ eventually became a ‘mythology for Britain.’ (2) In the conclusion it is suggested that there is perhaps more English imperialism than British fraternity in this particular instance of Tolkien’s encounter with the study of Welsh folklore.

John Rhys and Welsh Fairies

… in those Keltic tongues which lie so close at our doors, there is a rich mine of future research and discovery, and the Professor who is about to represent them in this University ought to find numerous and eager pupils.

A. H. Sayce, A Lecture on the Study of Comparative Philology, Oxford, 1876, pp. 28-9.
Alan Lee - Mabinogion - the dream of rhonabwy

In 1877 John Rhys (1840-1915) became Oxford University’s first Professor of Celtic. A founding father of modern Celtic studies, Rhys is primarily remembered today for dividing the Celtic language family into P-Celtic (Welsh, Old Cornish, and Breton) and Q-Celtic (Gaelic, Manx). But Rhys took his field of study to encompass “Celtic philology in that wider sense of the term which would embrace not only the study of Celtic speech but also of Celtic archaeology and history, of Celtic religion and folk-lore, of Celtic myth and saga.” (3) He thus sowed his narrowly linguistic studies into a rich interpretive tapestry, other threads of which included both the history and the folklore of the Celtic-speaking peoples. He argued, for example, that the P-Celtic languages had arisen on the Continent when the speakers of the Q languages conquered and mixed with an indigenous “non-Aryan race”, producing a mixed people, the P-Celts, who spoke a form of Celtic the pronunciation of which was heavily influenced by the non-Celtic language of the native aborigines. (4) Rhys then posited two distinct Celtic invasions of the British Isles, the first by the Q-Celts, who he called Goidels, the second by the P-Celts, who he called Brythons.

Rhys was one of a revisionist generation of late Victorian comparative philologists. In 1876 the Oxford Assyriologist Archibald Sayce had called for a shift of philological attention from the “extreme East” to the “extreme West”, declaring the time ripe for “Keltic, the western frontager” of the Indo-European language family, to be brought to the forefront of enquiry. (5) Sayce knew which way the philological winds were blowing. Within a decade he was hailing the “revolution” in comparative philology associated with a recent demonstration of the greater antiquity of the Western over the Eastern branches of the Indo-European languages. (6) Henceforth the original homeland of the speakers of proto-Indo-European, the so-called Aryans, was taken to lie somewhere in Europe. On a purely linguistic level this “revolution” propelled the languages of North-Western Europe firmly into the vanguard of philological research, contributing to that flowering of interest in the Celtic as well as the Teutonic languages that Tolkien encountered as an undergraduate at Oxford. At the same time, it fed into a new historical model of language dissemination. In place of an older picture of the spread of the Indo-European languages by way of mass folk migrations out of Asia and into Europe, younger philologists like Sayce and Rhys envisaged prehistory as a series of warrior invasions both within and out of Europe, one result of which was the imposition of Indo-European languages upon conquered natives populations. (7) Not only language change (such as the derivation of P-Celtic) but also the generation of folklore could now be explained in terms of this invasion model.

In 1901 Rhys published a two-volume study of Celtic Folklore. The previous year he had drawn upon what must already have been an almost completed study when delivering the presidential address to the anthropological section of the British Association for the Advancement of Science. (8) In this address Rhys set out “to substitute for the rabble of divinities and demons, of fairies and phantoms that disport themselves at large in Celtic legend, a possible succession of peoples” that, chronologically, have inhabited the British Isles, and “to each of which should be ascribed its own proper attributes.” (9) While most of these peoples had invaded Britain, Rhys interpreted stories about fairies as referring to an aboriginal, non-Aryan and matriarchal race, who had been driven into the hills of Wales and Scotland in the face of successive incursions of new peoples. Thus, the “key to the fairy idea is that there once was a real race of people to whom all kinds of attributes, possible and impossible, have been given in the course of uncounted centuries of story-telling” by those peoples who came after them. (10)

Rhys used philology in conjunction with archaeology in order to unlock the secrets of the prehistoric past, which he took to be preserved in his day only in folk stories and in the very landscape itself. He contended that the first race to dwell in the British Isles “consisted of a small swarthy population of mound-dwellers, of an unwarlike disposition, much given to magic and wizardry, and living underground: its attributes have been exaggerated or otherwise distorted in the evolution of the Little People of our fairy tales.” (11) Turning to archaeology in order to arrive at some hard evidence of this aboriginal fairy race, Rhys explained how “certain underground — or partially underground — habitations in Scotland are ascribed to the Picts”:

Now one kind of these Picts’ dwellings appear from the outside like hillocks covered with grass, so as presumably not to attract attention…. But one of the most remarkable things about them is the fact that the cells or apartments into which they are divided are frequently so small that their inmates must have been of very short stature, like our Welsh fairies. Thus, though there appears to be no reason for regarding the northern Picts themselves as an undersized race, there must have been a people of that description in their country… (12)

Tolkien and Rhys

When Tolkien came up to Oxford in 1911, Rhys was the grand old authority on all things – and more importantly, all words – Celtic. Rhys would have provided the young undergraduate with his point of entry into the emerging scholarly world of Celtic studies. In the academic year 1914-15 Tolkien probably attended Rhys’s lectures on the Mabínogían, the collection of prose stories that Rhys had edited and published in 1887 as the first volume of The Red Book of Hergest. (13) But the influence of Rhys in these years extended far beyond the lecture hall. In 1914, having been awarded the Skeat Prize for English, Tolkien purchased John Morris-Jones’ Welsh Grammar. In this book Morris-Jones, a one-time research fellow at Oxford, acknowledges his “deep obligation to my teacher Sir John Rhys”. He proceeds to assert the correctness of Rhys’s division of Celtic languages into P and Q groups, explaining that this classification has replaced an older one that itself had its basis in Rhys’s first book, Lectures on Welsh Philology (1877). (14) In so attributing to Rhys both old and new classifications Morris-Jones bears testimony to the authoritative mark that Rhys had stamped upon the Celtic scholarship of pre-war Oxford.

The mark of Rhys’s scholarship can be detected directly at various points in Tolkien’s writings. This is true both with regard to detailed argumentation and general historical vision. In 1932 Tolkien wrote a note on ‘The Name “Nodens”’ – an etymological enquiry into the nature of a hitherto unknown god referred to in some inscriptions discovered in what is now the border-region between Wales and England. Carl Phelpstead has observed how closely Tolkien’s note followed Rhys’s Lectures on the Origin and Growth of Religion (1886). (15) Turning from scholarship to linguistic creation, Mark Hooker, in his recent book Tolkien and Welsh (2012), suggests that Tolkien replicated Rhys’s P and Q distinction within his invented Elvish language family: “Sindarin and Welsh are ‘P’ languages,” he writes, while “Irish and Quenya are ‘Q’ languages.” (16)
Alan Lee - Mabinogion - gereint son of erbin

Tolkien’s enduring adherence to Rhys’s general conception of the complex linguistic and ethnological history of the British Isles is evidenced in both his very early literary creations and his late scholarly meditations. In his editorial commentary to the second volume of The Book of Lost Tales, Christopher Tolkien observes that his father’s early post-war writings contain the idea that Britain had been invaded by “hostile peoples named the Guiðlin and the Brithonin (and in one of these notes also the Rúmhoth, Romans)”. (17) The Guiðlin of these notes are, presumably, Rhys’s Goidels or Q-Celts, while the Brithonin are his Brythons or P-Celts. Nearly four decades later, in a scholarly talk on ‘English and Welsh’, Tolkien reiterated Rhys’s underlying model of British history. Tolkien here described both Celtic and Germanic forms of speech as “invaders” and suggested that the sword had played a role in their vanquishing of the native non-Indo-European languages of Britain. But the actual speakers of these native languages, Tolkien insisted, had never been driven from the land, and they had blended their blood with all the subsequent invaders to produce the ethnically mixed present day population of the British Isles. (18) Thus the linguistic and the ethnological histories of Britain were related but distinct; or, to put it in the terms established by the comparative philologists of late Victorian Oxford, language was no certain guide to race. (19)

To the best of my knowledge, Tolkien never engaged directly with Rhys’s account of Welsh fairies. Nevertheless, drawing as it did upon his wider linguistic and historical research, Rhys’s particular take on fairies was indicative of his approach to Celtic folklore as a whole and it is inconceivable that Tolkien was not aware of it. Simply put, Rhys’s explanation of the aboriginal source of Welsh fairy stories formed part of the academic atmosphere in which Tolkien lived and the scholarly air that he breathed during his early years at Oxford. Furthermore, the young Tolkien was particularly likely to prick up his ears at any scholarly discussion of fairies – he came up to Oxford already interested in fairies and left with his head full of them, as we know from his numerous fairy poems of these years. (20) Indeed, Tolkien later claimed that it was as an undergraduate that he had discovered that the two foundations of his own later fairy stories – mythology and language – were not “opposite poles of science and romance – but integrally related”. (21) Such a discovery was surely the result of an encounter with those “wider” philological investigations taught at Oxford that wove together ideas concerning language, history, folklore, myth and saga. It is thus not only a legitimate but also a potentially fruitful line of enquiry to attempt to discern the various ways in which Tolkien might have responded to Rhys’s reading of the Welsh fairy stories.

English Elves

As John Garth has described and Dimitra Fimi has explored, Tolkien’s Elves began life as Victorian fairies. (22) Nevertheless, the mark of Oxford’s scholarly stamp may be detected even upon some of Tolkien’s pre-war fairy poems. ‘A Song of Aryador’, for example, composed at an army camp in September 1915, identifies fairies as “the shadow-people”, the original inhabitants who still linger in a land now occupied by men. (23) As Garth points out, -dor is the Elvish for land (as in Gondor and Mordor) and Aryador, presumably, was associated by Tolkien with the original European homeland of the Indo-Europeans or Aryans. (24) So where Rhys had argued that prehistoric Britain had been inhabited by a pre-Celtic – and so pre-Aryan – population that the newcomers came to think of as fairies, Tolkien in this early poem suggests that from the very beginning of their history the Aryans had dwelled in lands previously inhabited by fairies. Of course, there is this fundamental difference between Rhys’s scholarship and Tolkien’s early poetry and prose writings: where the one provides a realistic explanation of fairies in terms of a philological history, the other uses that philological history as the backdrop to real fairy stories.

Lying in his hospital bed in 1916 recovering from trench fever, Tolkien took the first steps in the development of his earlier fairy poems into a full-blown English mythology. Christopher Tolkien describes how, in this immediate post-war period, his father was exploring the idea of a fairy land called Tol Eressëa that, long ago, had been moved across the ocean from the West to become the island of Britain (one part being torn off to become Ireland). (25) From this it followed that Britian had been inhabited by fairies long before the first invasions of the British Isles by “hostile peoples” such as the Celtic-speaking Guiðlin and Brithonin. At this point it might seem that Tolkien had merely substituted real fairies for Rhys’s aboriginal little people, but a fundamental transformation of conceptions is now introduced.
Alan Lee - the hobbit - farewell on the edge of mirkwood

At the centre of Tolkien’s early mythological thought is the idea that fairies have become small and transparent but were not always so. (26) Rhys had assumed that successive invasions of the British Isles had pushed the earlier inhabitants ever further west and north. Initially, the aboriginal little people had been driven into the hills of Wales and Scotland. Ultimately, the invasion of the Anglo-Saxons (or English) had pushed many Celtic-speakers into these same retreats (thereby creating Britain’s ‘Celtic fringe’). With the arrival of these new refugees, the little aborigines had nowhere left to flee (beyond was only the Atlantic Ocean), so they now hid themselves away in the hills. Subsequent sightings and brief encounters with this hidden folk had supposedly generated the fairy stories of Celtic folklore. But Tolkien was now inventing a specifically English folklore. He made his fairies English by insisting that, in the presence of men, fairies do not flee but rather diminish physically and fade almost to the point of invisibility. Thus, Tolkien at one point speculated, in the face of the invasions of Goidels and Brythons and Romans the fairies “faded before the noise and evil of war”. (27) They faded, but they did not flee: which is why the fairies today are found, not in the hills of Wales and Scotland, but on the English mainland. The economy of conception is breathtaking: the same move that transforms Rhys’s aboriginal humans into real fairies also ensures that these fairies are denizens of England rather than Wales.

Having established that fairies were in England before the arrival of the first men, and reside there still, Tolkien explored the idea that the English have a special relationship with the fairies. The Ingwaiwar, or English, the seventh people to invade Britain, were said to be better disposed to the fairies than those who had come before them. For this reason the fairies spoke to them in their own tongue and, indeed, since then “all who wish to speak to the Elves, if they know not and have no means of learning Elfin speeches, must converse in the ancient tongue of the English.” (28) With potential communication between fairies and Englishmen established, Tolkien could now posit an Anglo-Saxon transmission of fairy stories, sagas, poems, songs and lore through to our own day. An individual Anglo-Saxon voyager, at first named Eriol but subsequently reconceived (along with other elements of the conceit) as Ælfwine, (29) was said to have received this body of mythology from the fairies and passed it down the generations so that it reached the modern English. The implication of all this was that the English today “have the true traditions of the fairies, of whom the Íras and the Wéalas (the Irish and Welsh) tell garbled things.” As Christopher Tolkien concludes: “Thus a specifically English fairy-lore is born, and one more true than anything to be found in Celtic lands.” (30)

English Scholarship

It might be hazarded that in his later life as a scholar Tolkien would have come to object to the very fact that Rhys had analysed the Welsh fairy tales rather than explored their literary integrity as stories. (31) Certainly, and as is well known, by the 1930s Tolkien had become extremely critical of scholars who approached fairy stories purely as a source of historical information. This was a primary message of his famous lecture ‘On fairy-stories’, delivered at the University of St. Andrews in 1939. Indeed, one might be forgiven for suspecting that Tolkien had Rhys firmly in mind when he lambasted scholars who treat fairy stories “not as they were meant to be used, but as a quarry from which to dig evidence, or information, about matters in which they are interested.” (32) And yet it seems most improbable that Tolkien did here have Rhys in mind, and this because he seems to have shared the attitude of both Rhys and Matthew Arnold as to the literary merit of the Welsh stories.

Consider the analogy that Tolkien had developed three years earlier, in his 1936 lecture ‘Beowulf: the Monsters and the Critics’:

A man inherited a field in which was an accumulation of old stone, part of an older hall. Of the old stone some had already been used in building the house in which he actually lived, not far from the old house of his fathers. Of the rest he took some and built a tower. But his friends coming perceived at once (without troubling to climb the steps) that these stones had formerly belonged to a more ancient building. So they pushed the tower over, with no little labour, in order to look for hidden carvings and inscriptions, or to discover whence the man’s distant forefathers had obtained their building material. Some suspecting a deposit of coal under the soil began to dig for it, and forgot even the stones. They all said: ‘This tower is most interesting.’ But they also said (after pushing it over): ‘What a muddle it is in!’ And even the man’s own descendants, who might have been expected to consider what he had been about, were heard to murmur: ‘He is such an odd fellow! Imagine his using these old stones just to build a nonsensical tower! Why did not he restore the old house? He had no sense of proportion.’ But from the top of that tower the man had been able to look out upon the sea. (33)
Alan Lee - The Tower of Cirith Ungol

In the context of the 1936 lecture in which it was delivered, the meaning of this analogy is fairly straightforward. The man who inherits the ancient stones and builds a tower is the poet who, drawing upon a corpus of now lost Teutonic literature, composed the poem known today as Beowulf. The various critics, who push over the tower to search for hidden inscriptions or even coal, fail to appreciate that the new poem has literary value in its own right (that the tower provides an ocean vista). Such critics might include the likes of Stopford Brooke who, in his English Literature from the Beginning to the Conquest (1898), identified the “original germ” of the idea of Grendel, the first monster encountered in Beowulf, with the “primeval indwellers of the sea-coast” who were “driven back by the first invaders into the wild moors and rocks of the inland”. (34)

Now, at first sight, this analogy may be readily redeployed to embrace Celtic fairy stories, as well as critics of these Celtic tales such as Rhys. The builder of the tower now becomes, perhaps, one of the fourteenth-century Welsh bards who contributed to the Red Book of Hergest. Rhys is one of the critics who knock over the towers of Celtic folklore in their search for the meaning of the ancient stones found within these towers, or the coal buried beneath them (searching, in the particular case before us, for those aboriginal fairies taken to be the Welsh equivalents of Grendel), thereby failing to see the sea, let alone any far green country or sunrise beyond it.

But Tolkien would surely have resisted such a Welsh redeployment of his analogy. In his considered opinion the Beowulf poet had known what he was about – he had understood his material and he had crafted from it something wholesome and grand. The Welsh bards, by contrast, had neither understood their material nor brought to it an intelligible design. As Matthew Arnold had said and Rhys had repeated (see the epigraph to this essay), the Welsh bards had used materials taken from a tradition of which they were ignorant. And, to change the metaphor slightly, as did Tolkien in a letter of 1937, the productions of the Welsh bards were “like a broken stained glass window reassembled without design.” (35) The Beowulf analogy had made the point that modern scholars had failed to see that from ancient materials an Anglo-Saxon poet had made something new and beautiful. But Tolkien did not believe that the same praise could be bestowed upon the Welsh bards; he regarded their literary compositions (as he put it in the same letter) with “a certain distaste: largely for their fundamental unreason”. And the implication of this was that the chief value of the Welsh fairy stories must indeed reside in the individual and ancient elements out of which they were constructed. Rhys, in other words, was not to be faulted for his scholarly method of analytical demolition. This does not mean, of course, that Tolkien necessarily agreed with Rhys’s specific scholarly conclusions – a point on which we must now reflect further.


Rhys’s demolitional approach to the Welsh fairy stories, then, had been correct. But if the most ancient dwellers in Britain were fairies or Elves rather than, as Rhys had claimed, small mound-dwellers, what, from the perspective of Tolkien’s fiction, was to be made of Rhys’s reading of the Welsh fairy stories? Had Rhys simply produced a garbled account of Elves, or was his picture of aboriginal mound-dwellers a hazy glimpse of something else altogether? The subsequent development of Tolkien’s literary imagination suggests the latter. As any Tolkien aficionado will already have spotted, Rhys’s reading of the literary and archaeological evidence of, respectively, Wales and Scotland points neither to Elves nor to primordial inhabitants of the land, but to hobbits. “In a hole in the ground there lived a hobbit”, begins Tolkien’s masterpiece of children’s literature.This particular hole in the ground turns out to be a comfortable hobbit hole, built into the side of a hill and filled with handsome furnishings and numerous pantries. What is a hobbit? In the first pages of The Hobbit Tolkien observes: “I suppose hobbits need some description nowadays, since they have become rare and shy of the Big People, as they call us. They are (or were) a little people, about half our height”. (36) If the English fairy stories provide the true tradition of the first people to walk upon England’s pleasant pastures and mountains green – that is, the Elves, then the source of the little people found in Rhys's interpretation of the Welsh fairy stories is not an aboriginal race of men, but hobbits.
Alan Lee - The Hobbit - In a hole in the ground there lived a hobbit

Hobbits certainly bear some resemblance to the aboriginals that Rhys read into the Welsh fairy tales. As we have just seen, they are a little people, at least one of whom lives in a dwelling rather similar to those that Rhys describes as “hillocks covered with grass”. In ‘Concerning Hobbits’, (37) the prologue to The Lord of the Rings, Tolkien provided a fairly detailed anthropological survey of hobbits, and his discussion may be profitably compared with Rhys’s account of an aboriginal population of small “mound-dwellers, of an unwarlike disposition, much given to magic and wizardry, and living underground”. On some points the two accounts coincide. “At no time had Hobbits of any kind been warlike” states Tolkien (no scholarly qualifications here). Regarding dwellings, however, Tolkien introduces a note of learned pedantry. It is true that all hobbits had once lived in holes in the ground, but over the course of time many had adopted other forms of abode. In fact, by the time in which Tolkien’s narratives are set only the poorest and the richest of hobbits maintained the old-mound-dwelling custom (the pseudo-sociological observation introducing an illusion of realism). What of magic and wizardry? Well, as any reader of The Hobbit and The Lord of the Rings knows, the two most famous hobbits in history, Bilbo and Frodo Baggins, both went on adventures with the great wizard Gandalf; and one may surmise that, as their stories were retold down the ages, the attribute of wizardry came to be transferred to the hobbits themselves. But hobbits “have never, in fact, studied magic of any kind”. Nevertheless, they “may seem magical”, which is to say that hobbits are practiced in the “art of disappearing swiftly and silently”, and as such possess a “power which we sometimes confuse with real magic”. (38) Here Tolkien can be seen as not merely contradicting the idea that hobbits practice magic but also explaining how both Rhys and the tellers of Welsh fairy stories mistakenly came to associate the little people with magic in the first place.

Not all hobbit characteristics can be read from and into the Welsh fairy tales as Rhys interprets them. Hobbits are not native to the Shire, where most are said to dwell at the time of Tolkien’s stories. Nor does the society found in the Shire appear matriarchal. More generally, hobbits are almost completely devoid of those nefarious and sinister qualities attributed to so many of the denizens of the Welsh fairy stories. Yet all these qualities can be associated with at least one hobbit, if only we know where to look. Having left his comfortable hobbit hole far behind, Bilbo Baggins, the hero of The Hobbit, receives a knock on the head and wakes to find himself in a cave deep under the Misty Mountains. Here he encounters Gollum, “a small slimy creature”. While the narrator immediately declares “I don’t know where he came from, nor who or what he was”, it soon becomes clear that Gollum is not so very different from Bilbo (as Gandalf would remark of this meeting in The Lord of the Rings: “There was a great deal in the background of their minds and memories that was very similar. They understood one another remarkably well…”). (39) As Tolkien meditated upon a sequel to The Hobbit he made this implicit kinship explicit. In early drafts of The Lord of the Rings Gollum is already described as “an ancient sort of hobbit” and “of hobbit-kind, or akin to the fathers of the fathers of the hobbits”. (40) Gollum thus becomes a survival from a distant hobbit past, kept alive unnaturally by his possession of a magical but evil ring – the power of which has almost totally corrupted his character.

Gollum bridges much of the gap between Bilbo Baggins and the aboriginal origin of fairies described by Rhys. In general terms, he can appear as the archetypal evil ghoul of folk legend. After losing his ring to Bilbo, Gollum ventures out of his deep dark cave, whereupon the woods are filled with rumours and dreadful tales – “The Woodmen said that there was some new terror abroad, a ghost that drank blood. It climbed trees to find nests; it crept into holes to find the young; it slipped through windows to find cradles.” (41)
Alan Lee - The hobbit - riddles in the dark

But his particular story also includes several of the features that Rhys attributed to the little people as a whole. Gollum began his life by the banks of the River Anduin, where he was known as Sméagol and where he lived among a large family “ruled by a grandmother of the folk, stern and wise in old lore”. (42) Thus Gollum appears to have been born into a matriarchal society. (43) In ‘Concerning Hobbits’ we further learn that the lands around the River Anduin were the original home of the hobbits, or at least the land of which their earliest tales tell. (44) Thus Gollum is an aborigine, at least in relation to Bilbo Baggins. After turning to evil ways, Gollum’s grandmother had banished him from the ancestral home and he had sought refuge deep under the mountains, where he had “vanished out of all knowledge.” Thus, like the ‘little people’ of the Welsh fairy stories, Gollum has been driven into the hills, albeit on the command of his grandmother rather than by the swords of Aryan invaders. As with the (mistaken) association of hobbits and magic, Tolkien here provides a euhemeristic explanation of the fairy stories collected by Rhys: in this case pointing to an original source, the story of which seems to have been generalized (rather than confused) in its telling over countless generations.

Tolkien did not merely provide a source for Rhys’s stories of little aboriginal people; his reading also contains within it the germs of an explanation for Rhys’s mistakes and confusions. In the description of Bilbo’s encounter with Gollum, to take but one example, Tolkien can be seen projecting the relationship between ourselves and the aboriginal race posited by Rhys into a fictional encounter between a bourgeois hobbit and his ancient aboriginal racial self. But at the same time, and true to character, Tolkien can be seen suggesting within his fiction the true story behind the various fairy tales collected and interpreted by Rhys: Bilbo’s encounter with Gollum gives rise, eventually, to a myth of a primitive, cave-dwelling, aboriginal little people. (45) But in so providing both the ‘true stories’ and the explanations of Rhys’s false scholarly turns, Tolkien was not reiterating Rhys’s presuppositions in a different guise. Rhys’s aborigines are primitive in relation to us, perhaps a rung lower on the evolutionary ladder, and perhaps only a smidgeon of their blood flows in our veins, for it has now been much diluted by the infusions of later racial invasions. Gollum is indeed an aboriginal survival from ancient times; but he is essentially of the same folk as Bilbo, and his savageness is the result of degeneration due to the ring, not lack of evolution. Tolkien is telling this fairy story in his own terms, and he connects his own account with that of Rhys not by conceptual links but by the shifts and manglings that he knows to be endemic to the transmission of an originally oral body of stories and legends.

Conclusion: two Red Books

It remains to note one further step in Tolkien’s literary development. As we have seen, the young Tolkien initially posited an Anglo-Saxon transmission of his elvish tales of the elder days. Eventually, however, the scribes who transmitted this material were transformed from Anglo-Saxons into hobbits (from Ælfwine to Frodo Baggins, as it were), with the ‘Red Book of Westmarch’ becoming the imagined medium of transmission.

Tolkien declares the ‘Red Book’ to be our primary “source for the history of the War of the Ring” that ended the Third Age, but it also contains much scattered information concerning the Second, and even the First Ages. This ‘Red Book’ is said to be a manuscript compiled by several hobbit hands – the first of which belonged to Bilbo Baggins; the original has long since been lost, but several copies have supposedly come down to us. In essence, the ‘Red Book’ is the imagined manuscript source for the published books that we hold in our hands as The Hobbit and The Lord of the Rings. (46)

As has often been pointed out, Tolkien derived the title of this hobbit ‘Red Book’ from the fourteenth-century Welsh manuscript known as the ‘Red Book of Hergest’. This historical ‘Red Book’, we may note, was edited by Rhys together with John Gwenoryn Evans and published in two volumes – The Mabínogían (1887) and The Bruts (1890), both of which were owned by Tolkien. (47)
Alan Lee - The hobbit - Bilbo in Rivendell

Now, we have seen that Tolkien shared the mainstream Victorian distaste for these medieval Welsh stories, likening them to “a broken stained glass window reassembled without design.” But we have also seen that examination of some of these broken shards had provided Tolkien (as well as Rhys) with valuable insights concerning those people who dwelled long ago, in the quiet of the world, when there was less noise and more green. And having identified hobbits as depicted in some of these broken shards, it would appear that Tolkien set a few of these hobbits to work reconstructing the rest of this broken window, and, indeed, building much else besides; which is to say, writing the ‘Red Book’ anew, albeit according to a newly coherent design.

This conclusion raises some interesting questions concerning Tolkien’s relationship to Wales and its cultural heritage, a subject upon which it is easy to take a wrong step. Tom Shippey has noted that Tolkien felt a deep affinity with the Anglo-Saxon Beowulf poet. (48) Certainly, it is striking to observe how readily Tolkien’s allegory of the man who built a tower from ancient stones, quoted above, is applicable to Tolkien himself. But the conclusion to which we have been led in this essay is that, at least in those stories of the Third Age that concern hobbits, the materials Tolkien was reworking in his own literary compositions were as much Celtic as Old English or Norse. (49) Indeed, and as Mark Hooker has discerned, there is something intrinsically Welsh about hobbits. But if Tolkien’s hobbits are (or were) found in Wales, their techniques of literary craftsmanship are rather English than Welsh; which is to say that they (or Tolkien) work their material after the sober and steady fashion of the Beowulf poet, not in the supposedly mad and unreasoning manner of the Welsh bards. For all the Welsh flavour of Tolkien’s hobbits, they are – and are obviously – deeply of England; and the same is equally true of the tales written down in the hobbit version of the ‘Red Book’. The literary tower that is the ‘Red Book of Westmarch’ may be built with much Welsh stone and fitted with windows of Welsh glass, and it may even be erected in part upon Welsh soil, but it is for all that by design an English colonial construction. (50)

About the author

Simon J. Cook is a an independent English-language editor and a freelance scholar. Most of his other academic writings can be found at:


* I would like to dedicate this short essay to my friend Vashti Zarach. I wish to thank Mark Hooker for stimulating email correspondence concerning languages and hobbit archers, and Pieter Collier, for information about Tolkien’s personal library, for drawing my attention to the connection of Welsh and hobbit ‘Red Books’, and for providing me with the opportunity to publish this essay in the Tolkien Library. I am also indebted to Andrew Holgate for prodding me to better clarify my first thoughts on the origin of hobbits and to the learned observations of Avner Wallach concerning dwarves and Númenóreans.

(1) For the original see Matthew Arnold, On the Study of Celtic Literature, London, 1867, p. 61. Arnold, whose text was originally delivered as Oxford lectures attended by the undergraduate Rhys, adds (p. 66): “There is evidently mixed here, with the newer legend, a detritus, as the geologists would say, of something far older; and the secret of Wales and its genius is not truly reached until this detritus… is disengaged, and is made to tell its own story.” Both Rhys and Tolkien, in their different ways, could be said to have taken this lesson to heart.

(2) Dimitra Fimi, ‘Tolkien’s “‘Celtic’ type of legends”: Merging Traditions’, Tolkien Studies, 4, 2007, 51-71, p. 66. The phrase ‘mythology for England’, as Fimi and others have tirelessly pointed out, derives from Humphrey Carpenter’s biography, not from Tolkien himself.

(3) Rhys, Lectures on the Origin and Growth of Religion as illustrated by Celtic Heathendom, The Hibbert Lectures, 1886, Williams and Norgate, London, 1892 (second edition), p. vii.

(4) John Rhys, ‘The Celts and the Other Aryans of the P and Q Groups’, (read 1891), Transactions of the Philological Society 1891-2-3, pp. 104-131, p. 129.

(5) Sayce, A Lecture on the Study of Comparative Philology, Oxford, 1876, pp. 28.

(6) A. H. Sayce, Principles of Comparative Philology (3rd edition, London, Trübner & Co, 1885, pp. vii-viii).

(7) For further discussion of this shift in philological and historical thinking see my ‘The Making of the English: English History, British Identity, Aryan Villages, 1870-1900’, forthcoming in the Journal of the History of Ideas, but available online in a draft version via my page:

(8) John Rhys, Celtic Folklore: Welsh and Manx, two volumes, Oxford, Clarendon Press, 1901; Rhys, ‘Presidential Address to Section H. of the BAAS’, Report of the Seventieth Meeting of the British Association for the Advancement of Science, London: John Murray, 1900, pp.884-896. Like much of the primary material discussed in this essay, Rhys’s presidential address is available online (

(9) Rhys ‘Presidential Address’, p. 889.

(10) Rhys ‘Presidential Address’, p. 885.

(11) Rhys, ‘Presidential Address’, p. 896.

(12) Rhys ‘Presidential Address’, p 887-8. As Rhys says, his information on the dwellings of the Picts was drawn from the works of David MacRitchie; see: The Testimony of Tradition, London: Kegan Paul, 1890 and Fians, Fairies, and Picts, London: Kegan Paul, 1893. For useful background on MacRitichie, and indeed on later nineteenth-century discussion of fairies in general, see Carole G. Silver, Strange and Secret Peoples: Fairies and Victorian Consciousness, Oxford: Oxford University Press, 1999.

(13) For details of Rhys’s lectures see Christina Scull and Wayne G. Hammond, The J.R.R. Tolkien Companion and Guide: Chronology, Houghton Mifflin, 2006, pp. 50, 52, 55, 59.

(14) J. Morris-Jones, A Welsh Grammar: Historical and Comparative, Oxford, Clarendon Press, 1913, p. ix and p.3.

(15) See Carl Phelpstead, ‘Tolkien, David Jones, and the God Nodens’, The Lord of the Rings Fanatics Plaza. Phelpstead also discusses Tolkien’s note in his Tolkien and Wales: Language, literature and identity, University of Wales Press, 2011, pp. 53-7, but here mentions Rhys only in a footnote (note 14 on p. 137). Tolkien’s note was originally published as an appendix to an archaeological excavation report, but has now been republished in Tolkien Studies 4, 2007, pp. 177-83.

(16) Mark Hooker, Tolkien and Welsh: Essays on J.R.R. Tolkien’s Use of Welsh in his Legendarium, Llyfrawr, 2012, p. viii.

(17) J.R.R. Tolkien and Christopher Tolkien (ed.), The Book of Lost Tales, Part II[henceforth BLT I or II], ballantine Books, 1992, p. 299; see also p. 310 and p. 321 for the idea that invasion of the English was the seventh invasion of the British Isles.

(18) ‘English and Welsh’ (1955) in J.R.R. Tolkien, The Monsters and the Critics, Harper Collins, 1997, pp. 162-97, pp. 169-71.

(19) See, for example, A. H. Sayce, ‘Language and Race’, The Journal of the Anthropological Institute, 1876; v, pp. 212-20. Fimi cites the 1955 lecture ‘English and Welsh’ as evidence of “a change in Tolkien’s attitude” with regard to Wales and Celtic folklore (‘“Mad” Elves and “Elusive Beauty”: Some Celtic Strands of Tolkien’s Mythology’, Folklore 117, 2006, 156-170, p. 167). Phelpstead (Tolkien and Wales, p. 35) criticizes Fimi for failing to recognize how much Tolkien’s 1955 lecture anticipates the recent scholarly rejection of an older, romantic idea of the Celts. Neither scholar grasps that Tolkien was reiterating the Edwardian orthodoxy on which he had been reared and from which – changes in emphasis notwithstanding – he never substantially departed. This point bears spelling out. The new ideas that Tolkien is supposed to “anticipate” are, according to Phelpstead, grounded in a new model of Anglo-Saxon invasion as involving “a relatively small Anglo-Saxon warrior elite” imposing “its language and culture on what remained a predominantly British population”. This new account replaced an older historical model according to which the invading Anglo-Saxons “destroyed and displaced the native Britons”. This older model is said to have “held the field from c.1849 to the second half of the twentieth century” (Tolkien and Wales, pp. 18-9 and p.128, note 86). But this is faulty intellectual history: the ‘ethnic cleansing’ model had already been rejected back in the 1880s (it made a partial return in the next century). As Sayce put it in 1887 (sounding tellingly like Phelpstead in 2011), although a few years previously it had been “the fashion to assert that the English people were mainly Teutonic in origin, and that the British population had been exterminated”, it had now been established that “the British population, instead of being exterminated, lived under and by the side of their Teutonic invaders” (‘Address by Professor A. H. Sayce, M.A., President of the Section,’ Report of the Fifty-Seventh Meeting of the British Association for the Advancement of Science held at Manchester in August and September 1887, London, John Murray, 1888, pp. 885-895, p. 892 and p. 893). Tolkien was not anticipating late twentieth-century revisionism; he was repeating late-nineteenth-century orthodoxies! And Tolkien indicates as much with his comment: “the dogs that I have been beating may seem to most of those who are listening to me dead” (Monsters and Critics, p. 173). The root problem with Phelpstead’s reading of Tolkien’s 1955 lecture is that it is presented as a reaction against Arnold’s Oxford lectures, while the relevance and significance of Rhys’s intermediary position and revisionist scholarship is entirely passed over. And it is hard not to suspect that behind this lacuna, as also Fimi’s faulty reading, is an inability of the modern academic mind to comprehend the many points of contact between its own post-modern thinking on identity (non-nationalistic and local) and late Victorian and Edwardian scholarship grounded on ideas of race (for further discussion of the latter, see my ‘Making of England’).

(20) Humphrey Carpenter (J.R.R. Tolkien: A Biography, Houghton Mifflin, 2000, p. 55) gives a poem of 1910 about “light fairy things tripping so gay”. Much of the 1915 material is published in BLT, at least one poem of which, ‘The Cottage of Lost Play’, was composed in Oxford in Tolkien’s undergraduate lodgings (BLT  I, p.19); others were composed as Tolkien went through military training. John Garth (Tolkien and the Great War: the Threshold of Middle-earth, Houghton Mifflin, 2005) provides an excellent narrative of the composition of these early fairy poems, while Dimitra Fimi (Tolkien, Race and Cultural History: from Fairies to Hobbits, Palgrave Macmillan, 2010, Part I) relates Tolkien’s early conceptions of fairies to Victorian and Edwardian popular culture and offers a helpful account of how these fairies transformed into the (no-longer little) Elves of Middle-earth.

(21) Humphrey Carpenter and Christopher Tolkien (eds.), The Letters of J.R.R. Tolkien, Harper Collins, 2012, letter 131 and see also letter 257.

(22) Garth, Tolkien and the Great War; Fimi, Tolkien, Race and Cultural History.

(23) BLT I, 151-3

(24) Garth, Tolkien and the Great War, 21% (having accessed Garth’s book via Kindle my reference is reduced to a percentage as opposed to a page number). In Gnomish, Aryador means the land of shadow; but as Garth notes, Tolkien seems to have created “a small but significant proportion of his Qenya words specifically to show kinship with ancient recorded or reconstructed words”.

(25) BLT I, pp. 1-39, and BLT II, pp. 282-340.

(26) BLT II, p. 290.

(27) BLT II, p. 299. These are Christopher Tolkien’s words, and the wars he refers to are indeed the various invasions of the British Isles.

(28) BLT II, p. 310.

(29) For further discussion of this Anglo-Saxon transmission see Verlyn Flieger, ‘The Footsteps of Ælfwine’, in Verlyn Flieger and Carl F. Hostetter (eds.), Tolkien’s Legendarium: Essays on the History of Middle-earth, Wetport, CT: Greenwood, 2000, pp. 183-98, and Maria Artamonova, ‘Writing for an Anglo-Saxon Audience in the Twentieth Century: J.R.R. Tolkien’s Old English Chronicles’, in David Clark and Nicholas Perkins (eds.), Anglo-Saxon culture and the modern imagination, Boydell & Brewer, 2010, pp. 71-88.

(30) BLT II, p. 295.

(31) Note that in the second volume of his Celtic Folklore Rhys was at pains to explain that, “as I have no wish to earn the displeasure of my literary friends, let me hasten to say that I acknowledge the latter, the creatures of the imagination, to be the true fairies… the other folk – the aborigines whom I have been trying to depict – form only a sort of substratum, a kind of background to the fairy picture” (Rhys Celtic Folklore, II: 669-670).

(32) Tolkien, ‘On fairy-stories’, in The Monsters and the Critics and Other Essays, edited by Christopher Tolkien, London: George Allen & Unwin, 1983, p. 129.

(33) J.R.R. Tolkien, ‘Beowulf: The Monsters and the Critics’, Proceedings of the British Academy, 22, 1936, 245-95; quotation on pp. 248-9.

(34) Andreas Haarder and Tom Shippey (eds.), Beowulf: the Critical Heritage, Routledge, 1998, p. 493.

(35) Letters, letter 19.

(36) Tolkien, The Hobbit, chapter 1. Note how, immediately following this introductory description of hobbits, the narrator comments with regard to one particular hobbit family, the Tooks: “It was often said (in other families) that long ago one of the Took ancestors must have taken a fairy wife. That was, of course, absurd…”

(37) The earliest draft of this account of hobbits dates to 1938-9, i.e. soon after the publication of The Hobbit, and is printed in J.R.R. Tolkien, The Return of the Shadow (ed. Christopher Tolkien), Houghton Mifflin, New York, 2000, pp. 310-14. The history of subsequent drafts may be found in J.R.R. Tolkien, The Peoples of Middle Earth (ed. Christopher Tolkien), Houghton Mifflin, New York, 1996.

(38) This last quotation is not taken from the prologue found in the published version of Lord of the Rings but from its early draft, now found in Return of Shadow, p. 311.

(39) Lord of the Rings, ‘Shadow of the Past’. But even in the first edition of The Hobbit Tolkien seems to have had a sense of the encounter of Bilbo and Gollum as that of the modern and the archaic. As Tom Shippey points out, Gollum’s riddles “associate him firmly with the ancient world of epic and saga” while “Bilbo replies to Gollum’s ancient riddles with modern ones” (Shippey, J. R. R. Tolkien: Author of the Century, Harper Collins, London, 2000, pp. 25-6).

(40) Return of the Shadow: Section III: ‘Of Gollum and the Ring’: p. 75 and p. 78. The patriarchal turn-of-phrase seems deliberate – this is how a patriarchal society would describe distant ancestors, even if the society of those ancestors was matriarchal.

(41) Tolkien, Lord of the Rings, ‘Fellowship’, ‘The Council of Elrond’.

(42) Tolkien, Lord of the Rings, Part I, ‘The Fellowship of the Ring’, ‘The Shadow of the Past’.

(43) Tolkien would later insist that Gandalf’s use of the word “matriarch” in reference to Sméagol’s grandmother “was not ‘anthropological’, but meant simply a woman who in fact ruled the clan” having, in all likelihood, outlived her husband (Letters, letter 214). Hooker reads this letter as Tolkien backtracking on his original conceptions and projecting Victorian values onto the hobbits (Tolkien and Welsh, 82-5). Be that as it may, what matters here is that the tale of Sméagol unmistakably plants the seed of the idea of an archaic matriarchal hobbit community and, nourished by way of the retelling of the tale down the generations, such a seed is all that is needed to produce the aboriginals of Rhys’s reading of the Welsh fairy tales.

(44) Lord of the Rings, ‘Concerning Hobbits’.

(45) I cannot refrain from here noting a tale by John Buchan, one of Tolkien’s favourite authors. The Watcher by the Threshold (1918 [1902], New York, George H. Doran) tells the story of an Oxford scholar, learned in “the ancient life of the North, of the Celts and the Northmen” and author of “a monograph on the probable Celtic elements in the Eddic songs” (p. 14) who, on holiday in the remote Highlands of Scotland, is captured and held in a cave by a wild aboriginal people from whom he hears, “preserved in a sort of shapeless poetry… bitter words against the Gaelic oppressor, bitterer words against the Saxon stranger… fragments of old religions, primeval names of god and goddess, half-understood by the Folk, but to me the key to a hundred puzzles” (pp. 50-1). The Oxford scholar’s first glimpse of one of these aboriginals is of a figure “little and squat and dark… in its face and eyes there seemed to lurk an elder world of mystery and barbarism, a troll-like life which was too horrible for words” (p. 44). If Gollum may perhaps be discerned in this figure, Buchan is nevertheless wholly in the conceptual world of Rhys – his aboriginals are unevolved hairy cave-men.

(46) And also, of course, The Adventures of Tom Bombadil: and other verses from the Red Book; not to mention “three large volumes, bound in red leather… of Bilbo’s ‘Translations from the Elvish’… almost entirely concerned with the Elder Days.” See ‘Note on Shire Records’ in the prologue to The Lord of the Rings.

(47) John Rhys and John Gwenoryn Evans (eds.), The Text of the Mabínogían and other Welsh Tales from the Red Book of Hergest (Oxford, J. G. Evans, 1887) and The Text of the Bruts from the Red Book of Hergest (Oxford, J. G. Evans, 1890). I am indebted to Pieter Collier for the information that Tolkien owned both of these volumes. Fimi (‘Tolkien’s “‘Celtic’ type of legends”, p.51) provides details of other versions of The Red Book of Hergest owned by Tolkien, while Phelpstead lists the known contents of Tolkien’s Celtic library in the appendix of Tolkien and Wales.

(48) See the first chapter, ‘Tolkien and the Beowulf-Poet’, in Shippey’s Roots and Branches: Selected Papers on Tolkien, Walking Tree Publishers, 2007.

(49) For discussions of Celtic folklore in Tolkien’s work, see Fimi’s ‘“Mad” Elves and “Elusive Beauty”’ and ‘Tolkien’s “‘Celtic’ type of legends”’, Part 3 of Verlyn Flieger’s Interrupted Music: the Making of Tolkien’s Mythology, Kent State University Press, 2005, and Part II of Phelpstead’s Tolkien and Wales. Marjorie Burns (Perilous Realms: Celtic and Norse in Tolkien’s Middle-earth, University of Toronto Press, Toronto, 2005) compares the Celtic and Norse elements in Tolkien’s fiction with much insight; although statements such as “Tolkien valued both branches of English-kind, both the Celtic and the Norse” (p. 174) suggest some Arnoldian confusion.

(50) Phelpstead concludes a useful overview of the relationship of the two Red Books (Tolkien and Wales, pp. 57-63), by noting that ‘marches’ refers to border regions, and that on the border between Wales and England ‘the west march’ is how the English side of the border would be described by the English (p. 63).

Enjoyed this post? Click to get the RSS feed.

Spread the news about this J.R.R. Tolkien article: