1 Onomastics and the Arthurian tradition

For the founders of our discipline who sought solid evidence to assess texts and their mutual relationships, literary onomastics – ‘the study of names in literary texts (with ‘literary’ defined as broadly as possible)’^[7] – provided convincing answers to questions regarding the identification and circulation of the matter of Britain. An onomastic perspective contributes to an appropriate evaluation of texts according to their intertextuality, a critical element, fully at play in Arthurian literature.^[8] Indeed, onomastics often allow for verification of the relations between various traditions.^[9] Together with other evidence, onomastics can be the final piece in the puzzle regarding a character’s identity. If a character does not only act similarly in two romances, but also bears a similar name, he may indeed be one and the same character. Onomastics, when supported by critical and linguistic-etymological investigation, for instance, makes it possible to postulate that Careticus in the Historia Regum Britanniae, Cadwagan Vras of the Welsh Mabinogion, Garadue of Biket’s Lai du Cor, Carados of the Première Continuation du Perceval constitute literary representations of the same character and not cases of homonymy.^[10] Moreover, proper names, which are often linguistically conservative, can resist translation: in the form of calques, they can be transferred in full from one language to another (from Old French Joieuse Garde to Italian Gioiosa Guardia), or covered with a coat of linguistic patina (from Old French Biaus Descouneüs to English Lybeaus Desconus) and such proper names suggest direct contact between two texts in two different languages. Within multilingual textual traditions, proper names such also make a significant contribution to the reconstruction of a text’s linguistic stratification, and to the identification of the complex relationships between direct and indirect witnesses. All these aspects are considered by Francesco Carapezza, who notes that:

Il ‘nome nel testo’ si offre insomma a svariati tipi d’indagine critica ma costituisce al contempo un locus criticus in senso tecnico, essendo il nome proprio soggetto per sua natura a varianti formali, variazioni intenzionali, occultamenti, fraintendimenti ed errori di copia che ci possono informare sulla storia della tradizione e della ricezione del testo in esame.^[11]

Finally, by reversing the perspective (from literature to history), and since historical onomastics often offers precise chronological information, onomastics can help to ascertain details of the diffusion of some subjects with a certain level of accuracy. Numerous cases in point are offered by Pio Rajna’s and Pierre Gallais’ contributions on the reception of matière de Bretagne in Italy and France respectively, as well as by the ‘enromancement du nom’, a phenomenon acutely identified by Michel Pastoureau.^[12]

2 Onomastics and fin de siècle philology

The issues mentioned above were already clear to the very first generations of philologists and medievalists. In addition to purely academic reasons, there are also cultural-historical implications which account for the development of certain fields of study within the scholarly tradition.^[13] This is particularly true for German medievalists, who took the lead in compiling indexes and repertories of proper names, as they would soon also do for dictionaries and inventories of all kinds. Onomastics has in fact been a focus of interest of Romantic intellectuals since Wilhelm Grimm’s Die deutsche heldensage (1867).^[14] Similar was the ambitious project of Ernst Förstemann – supported by his close friend Grimm – of the Altdeutsches Namenbuch, the first volume of which, dealing with anthroponomy, was printed as early as 1856.^[15] One cannot help but notice that the German primacy of literary onomastics in that period is also linked to the complex ideological (re)construction of a national history, a topic evidently beyond the scope of the present investigation.^[16] However, Grimm and his companions’ cultural-political project also passes into the terrain of historical linguistics, to which onomastics is closely connected.

Outside of Germany and its specific context, Europe saw the compilation of medieval literary onomastic indexes at the end of the 19th century. In Great Britain William George Searle produced the Onomasticon anglo-Saxonicum (1897), which shares Förstemann’s methodologies and employs both historical and literary sources.^[17] France, instead, engaged in the collection of indexes relating to literature: Alfred Franklin’s Dictionnaire des noms, surnoms et pseudonymes latins de l’histoire littéraire du Moyen Age (1100–1530) was published in 1875, and is an excellent and purely practical manual, aiming to identify authors referred to by different names.^[18] It does not include fictional characters and focusses instead on Medieval Latin writers. However, less than a decade later, French vernacular fiction would be investigated as well. Pole position in these endeavours, however, does not belong to France, but Germany: in Greifswald in 1882, PhD candidate Fritz Seiffert defended a thesis entitled Ein Namenbuch zu den altfranzösischen Artusepen.^[19] The project, directed by Eduard Koschwitz (1851–1904), aimed to produce an onomastic index from a corpus of Arthurian French romances. Koschwitz, a former student of Gröber and Böhmer, was a renowned medievalist with special expertise in Provençal material, who in 1881 was only 30 years old and had only recently started teaching Romance Philology in Greifswald, where he would later become rector, before moving to Marburg and Königsberg. Interestingly, his œuvre does not seem to reflect any specific interest in Arthurian literature and the topic may have been suggested by some more experienced scholar he had contact with.^[20] A short announcement of Seiffert’s thesis was published in the ‘Chronique’ section of the journal Romania: ‘Ein Namenbuch zu den altfranzœsischen Epen, Teil I … (von) Fritz Seiffert, in-8 °, 45 p. (dissertation, de Greifswald). – Ce n’est ici que l’introduction d’un livre qui doit bientôt paraître; nous en reparlerons. L’idée est bonne, mais l’auteur aurait besoin de plus de préparation’.^[21] Unfortunately, the editors of Romania did not have the opportunity to discuss the thesis again, because the full publication never took place.^[22] From what we can reconstruct from the published material, the thesis considered a corpus of fourteen French verse romances from the 12th and 13th centuries, as well as Geoffrey of Monmouth’s Historia Regum Britanniae.^[23] It is, however, important to acknowledge Koschwitz as having conceived the first onomastic investigation of French Arthurian literature. Although he was a convinced nationalist, especially after events in Alsace and Lorraine, and despite being an honorary member of the ‘Vereins Deutscher Studenten’, Koschwitz never denied the leading position of medieval French literature over other contemporary canons.^[24] In addition, it should be noted that the German professor turned to the then neglected field of romance. Up to that point, medievalist research had concentrated instead on texts with presumed historical relevance such as epics and chronicles, which were important from an ideological and nationalistic perspective, while Arthurian literature had little to offer for these purposes.

In 1904 it was the turn of Ernest Langlois’s Table des noms propres de toute nature compris dans les Chansons de geste,^[25] a publication that did not arise by chance. It resulted from a ‘prix ordinaire’ announced by the Académie des inscriptions et des belles-lettres in 1899 for a work which would ‘relever les noms propres de toute nature qui figurent dans les chansons de geste imprimées antérieures au règne de Charles V’.^[26] Including chansons de geste ‘de toute nature’, the structure of the Table, which was imitated by later scholarship, consists of onomastic entries, each of them with variants considered significant, a concise description of the name’s referent (character, place, country, etc.), and a bibliographical reference to the point in a text where the name occurs. Langlois also included names that would be expunged from later indexes, such as Jesus. The whole project was actively promoted by Paul Meyer and Gaston Paris, the two most influential philologists of the time, who both were members of the Académie.^[27] Langlois’s Table therefore provides a model for the many onomastic works that follow, and it also serves to expand the cultural horizons of late positivism, which played a significant role in structuring Romance philology.

At the same time, other areas of the academy were also taking an interest in onomastics. In September 1902, the Royal Flemish Academy announced a competition for an ‘Een Onomasticon of Lijst van Persoonsnamen der Middelnederlandsche Letterkunde’.^[28] It is not surprising that this project was also set within a nationalist perspective. The individual behind the competition was Willem de Vreese, not only a professor and librarian but also a fervent patriot and leader of the ‘Jong-Vlaamsche Beweging’^[29] under German occupation in the First World War.^[30] The outcome of this competition was somewhat different to what had happened in France. The Reports of the Inspection Commission of the Academy reveal that there was only one candidate: the linguist Charles P. F. Lecoutere, a professor in Leuven. His work left the Commission unsatisfied, who proposed to award him the 600 francs prize only on the condition that he complete and expand his work. There is, however, no trace of what happened to the 600 francs, or indeed to Lecoutere’s work.^[31]

Beyond this, what must be emphasised is that the creation of mediaeval onomastic indexes from the second half of the 19th century ought to be understood from the perspective of the needs of the time. The scant interest shown in Arthuriana is due to the fact that it is fictional literature: and for this very reason Koschwitz’s pupil’s pioneering project takes on even more significance.

3 The Onomasticon Arthurianum: the announcement

In 1891, we finally reach the Onomasticon Arthurianum. In the peculiar fin de siècle context we have just outlined, the first – and indeed the only – enterprise of collecting and cataloguing the entire Arthurian onomastic corpus takes place. Besides its classificatory spirit, the analytic and methodological frameworks of the time also favoured the project of creating a multilingual thematic index, and a number of philologists at the time made use of onomastics (and toponymy, specifically) for the reconstruction of the ‘sources’ of certain textual traditions. The use of names for such purposes was already a long-standing practice of the Romantics with Walter Scott paradigmatic in this respect.^[32] In a more academic context, the fathers of Romance philology – Gaston Paris and Ferdinand Lot, and, on the Italian side, Pio Rajna^[33] – had also dealt with questions of onomastics. In 1891, in the margins of a work on the Grail legend in Middle German, the English folklorist Alfred Nutt optimistically claimed:

Il me semble qu’une des œuvres dont l’étude des romans arthuriens profiterait le plus serait la compilation d’un Onomasticon Arthurianum qui tiendrait compte de l’ensemble des textes tant manuscrits qu’imprimés. Ce serait là une œuvre gigantesque, mais qui pourrait être menée à bonne fin si tous les érudits qui s’occupent de ces études y apportaient un concours actif.^[34]

It is not by chance that the first reference to the Onomasticon Arthurianum appears within the Revue Celtique: Celtic studies, in great expansion at the time, made particular use of onomastics. The echo of Nutt’s desiderata did not take long to spread: a review dedicated by Romania to the issue of the Revue Celtique in question highlights the statement and offers words of encouragement:

C’est une idée à laquelle on ne saurait qu’applaudir. Elle avait reçu en 1882 un commencement d’exécution dans la thèse d’un jeune docteur de Greifswald, M. Fritz Seiffert; mais Ein Namenbuch zu den altfranzösischen Artuspen est malheureusement resté à l‘état d’ébauche.^[35]

The issue was therefore important to Paul Meyer and his companions: Romania gave wider exposure to Nutt’s proposal and at the same time confirmed that Seiffert’s work still remained incomplete. The doors were then wide open for a European Onomasticon. However, it was American academia that was to pick up the baton. From 1898 to 1905, under the supervision of William Henry Schofield from Harvard, Alma Blount, a young anglicist from Radcliff College, which then was the female counterpart of the all-male Harvard College, worked on the compilation of an index, whose name cannot be explained by polygenesis: the Onomasticon Arthurianum. The project is strictly in line with Schofield’s academic interests, as an expert in Norse literature, but also a renowned comparativist and founder of the Harvard Studies in Comparative Literature series.^[36]

Blount’s project involved the creation of an onomastic repertory based on a corpus which included all known Arthurian texts at the time. The work, conducted manually, was far too ambitious for one person: Nutt had already indicated the need for shared efforts and coordination. In 1910, Blount surrendered to the evidence and deposited her notes in the Treasure Room of the Widener Library at Harvard University, for the benefit of other scholars. To her great disappointment, the Onomasticon was never completed, and Blount ended her academic career in 1935 as an English professor at Michigan State Normal College (now Eastern Michigan University), having never taken up the work again.^[37]

In 2022, the University of Harvard digitised the notes deposited by Blount more than a century earlier.^[38] The material is particularly interesting to explore. The scholar left the American library a considerable collection, consisting of more than 23,000 handwritten slips, sorted in alphabetical order. An internal note from 1921 indicates:

Compilation was begun in ‘98-‘99 and has been continued at intervals in the Harvard, Cornell and Chicago libraries and in the British Museum and Bibliothèque Nationale. The lists, both of person and place names were completed for all available printed texts, up to the recent Lancelot publications by Dr. H. O. Sommer, which are yet to be added.^[39]

It is probably not a coincidence that the Onomasticon stopped short of including the French prose tradition, which would have led to an extraordinary expansion of the corpus. An extensive preparatory dossier is attached to the onomastic files: the materials, almost all of which are handwritten, include summaries and descriptions of the texts used, bibliographical references, notes taken on flyers, and clippings from newspapers or academic journals. Blount’s interest in the manuscript tradition also cannot be underestimated: her research was often conducted directly on manuscripts, as evidenced by her visits to North American and European libraries.^[40] Also of great importance is the collection of information to disambiguate cases of homonymy, which are extremely frequent in Arthurian literature: it is in all respects a precise study, carried out with scholarly rigour and methodological attention. The generous digitisation of the dossier makes it possible to observe what the author had on her desk and thus to reconstruct the methodologies and criteria of such portentous manual research which is inconceivable today.

For a long time, Blount’s unpublished work was the only reference for Arthurian onomastics. The notes must have benefited from a certain circulation, even if probably only in North America, given that in 1926 Francis Peabody Magoun Jr., the well-known medievalist and professor at Harvard, published an article in the newly founded journal Speculum in which he gave the list of the abbreviations employed in the Onomasticon.^[41] From 1926 onwards, journals made a few cursory mentions of the project, but it would take a quarter of a century, and another world war, before we were to hear about it again in concrete terms.

4 Onomasticon Arthurianum: what is there

In September 1949, the Modern Language Association of America, then in its sixty-fifth year, held its annual meeting. Among the various discussion groups, we note the group ‘Comparative Literature III: Arthurian Romances’ (Chairman, Albert W. Thompson, Washington State Coll.; Secretary, Robert W. Ackerman, Stanford Univ.)’. The Onomasticon Arthurianum is at the centre of the group’s business: Margaret Schlauch, a renowned medievalist, discussed the ‘Problems of Compiling a Name-Index for the Old Norse Arthuriana’. The formation of the International Arthurian Society is announced, but above all we read about the ‘Report of the Committee on the Onomasticon Arthurianum’.^[42] The content of the Report, which is not further specified, is, however, easy to imagine from later publications. In 1952, with the support of Roger Sherman Loomis and William Nitze, the anglicist Robert William Ackerman, the secretary of the aforementioned group session, published his Index of the Arthurian Names in Middle English.^[43]

Ackerman’s first aim is to fill the gaps in Blount’s work, which, he argues, neglected Middle English texts.^[44] Criteria and methodologies are clearly spelled out: the corpus includes ‘all the Middle English versions of the Arthurian legend except the chronicles’,^[45] for a total of 28 texts, taken from around 40 manuscripts and incunabula.^[46] The names of God, Jesus, religious festivals, days of the week, months and ethnic names are not considered, because ‘none of these could be regarded as of great significance to Arthurian studies’.^[47] Every single entry in the Index includes, in addition to a brief description, a selection of variant readings, accompanied by references to the loci of the corpus.^[48] Furthermore, in keeping with the comparativist perspective from which Ackerman undertook, the work ‘incorporated […] comments about scribal blunders and connections with other names, especially those in Old French texts’.^[49]

Ackerman’s work was also intended as a first step towards the creation of the notorious Onomasticon Arthurianum, for which the author offers a complete description.^[50] The announcement is however formulated in duly cautionary terms: ‘the present index may perhaps be further regarded as one stage, the first, in an effort to compile an Onomasticon Arthurianum’.^[51] Mindful of Blount’s example, Ackerman advocates a coordinated project, but with linguistically distinct units. He also expresses the urgency of dealing with French Arthurian literature, which in his opinion is quantitatively and qualitatively dominant.^[52]

Meanwhile, in France, Louis-Fernand Flutre was working on a new resource that would significantly improve the scholarly landscape: the Table des noms propres avec toutes leurs variantes figurant dans les romans du Moyen Age écrits en français ou en provençal.^[53] Several years earlier, in 1943, the Académie des inscriptions et belles-lettres had again announced a prize, this time for the preparation of a list of ‘noms propres contenus dans les romans français et provençaux du moyen âge’^[54]: the work would mirror what Langlois had done at the very beginning of the 20th century. The Académie entrusted its project to Flutre, an eminent philologist (with special expertise in the textual tradition of the Faits des Romains) but also a scholar with broad horizons^[55]. This is all the more so since he would later make contributions that extend into to onomastics and far beyond the boundaries of Romance philology: in 1957, he published two toponymic monographs, one focussed on the Lozère department, the other on the French colonies in West Africa^.[56] The result of Flutre’s research for the Académie was published almost twenty years after he received the award. The work brings together a vast corpus, comprising more than 200 critical editions of mainly romances (and also 21 Nouvelles, lais, débats, the chantefable Aucassin et Nicolette, and so on) covering a time span from the 13^th to the 15^th century. The work, although in need of updating, has stood the test of time and remains the reference point for onomastic investigations in Gallo-Romance fictional literature to this day: taking into account all the Arthurian romances, in prose and verse, published at the time, the Table is still the most extensive (albeit not the most detailed) tool for the study of Arthuriana in Old French. However, one cannot help but notice that, in the author’s extensive preface to the Table, there is no mention of American works: neither of Blount, nor of the Onomasticon, not even of Ackerman’s near-contemporary Index.

At that time, Arthurian literature remained a primary focus in Anglo-Saxon studies. To French scholars, by contrast, the matière de Bretagne was a matter for Britain. Indeed, the first Arthurian index to encompass French literature would be the work of a British scholar working in Canada. after announcing it as early as 1957,^[57] Gerald D. West, Professor of French at McMaster University, published his Index of Proper Names in French Arthurian Verse Romances (1150–1300) in 1969, followed in 1978 by a second volume dedicated to the prose romances, given that ‘in view of the quantity of texts and the complex and differing nature of the material, a satisfactory result could only be achieved by a separate treatment of verse romances and prose Romances’.^[58] West argues that this division according to form is warranted on practical grounds, but he does not examine the possible drawbacks of such an approach. The Arthurianity of the texts, on the other hand, is discussed here for the first time. West determines that the texts of his corpus must in fact possess ‘some justification for the title Arthurian’.^[59] He notes for the prose romances in particular:

The prose romances from which the entries in this Prose Index have been drawn are those dealing with Arthur and his associates, Merlin, the Knights of the Round Table, Tristan, and the Grail. I have excluded those works in which Arthur’s name occurs, or in which he may even appear in person, if the principal subject matter in non-Arthurian or if the works belong to other cycles.^[60]

Beyond the discussion of theoretical issues, West’s work displays a good deal of systematic and methodological rigour. His period of focus is well delimited, and exceptions to the norm are duly accounted for.^[61] Significantly, the structure of the two Indexes is the same as Ackerman’s, and the names excluded are also the same.^[62] In short, West’s work explicitly places itself in the continuity with Ackerman’s, cited from the outset as a model. Moreover, the author himself declares that ‘it is hoped that the Verse Index and the Prose Index may both be regarded as contributions to the comprehensive Onomasticon Arthurianum’.^[63]

The resource presented by this Canadian professor offers detailed and refined onomastic data. Each unit consists of a lemmatized form, followed by all the varia lectio for the name, as far as available editions allow to account for it. In the case of verse romances, the readings are presented in chronological order, from Wace’s Brut to Girart d’Amiens’ Escanor, but leaving weighty texts such as Froissart’s Meliador to one side.^[64] West notes: ‘for the sake of convenience, […] the romance of Chrétien de Troyes have been grouped together; the same procedure has also been adopted for the Tristan poems, and for the various Continuations de Perceval’.^[65] For the proses, however, the variants are arranged alphabetically because the chronology of these texts, as West rightly notes, is ‘a vast subject requiring perhaps some re-investigation’.^[66] Each variant is accompanied by an indication of the verse or page on which it appears in the critical edition; if the name appears more than ten times, the first ‘ten appearances only are recorded’.^[67] The descriptions accompanying each onomastic unit are detailed and come with a rich and pertinent bibliography and references to the repertories of Ackerman, Langlois and, to a lesser extent, Flutre.^[68] Clever exceptions to the general structure of the Indexes (for example, in the treatment of cases of homonymy) mean that the work is not only fully systematic in its approach, despite the considerable volume of texts covered, but also, to use an anachronism, truly user-friendly.^[69]

The Indexes were greeted enthusiastically even by the most demanding critics. Flutre himself calls West’s work ‘plus qu’une table ou un index, […] une somme géographique et biographique, une encyclopédie onomastique des œuvres étudiées’^[70] and soberly concludes that ‘tout cela fait de l’Index of proper names in French Arthurian verse romances un livre complet, précis, bien ordonné, dont on peut en toute sûreté recommander l’usage à ceux qui étudient notre ancienne littérature’.^[71]

Despite his hopes, West’s work represents a setback since it also underlines the incomplete status of the Onomasticon. In this sense, it is appropriate to shine a light on studies carried out afterwards, such as Christopher W. Bruce’s Arthurian Name Dictionary, published by Garland in 1999. The title seems to promise the completion of the long-awaited Onomasticon, but it is in fact a quite different project. The dictionary, very rich and easy to consult, is intended as a catalogue of Arthurian characters and not as an onomastic tool. Bruce, a policeman from Cambridge, Massachusetts, clearly states his research goals, which Norris J. Lacy, in his foreword, accurately summarises as a ‘comprehensive guide to Arthurian characters’.^[72] It only takes a quick glance to see the Dictionary’s philological limitations: the corpus of texts examined, built entirely iuxta propria principia, extends from the Venerable Bede to Alfred Tennyson; names are often ‘lemmatised’ by loosely translating them into contemporary English; variants, when present, are provided arbitrarily; there is no indication of the textual loci where the indexed names are to be found. In a recent reflection on mediaeval literary onomastics, Christine Ferlampin-Acher reiterates that the Dictionary ‘reste généraliste, laisse de côté de nombreuses figures secondaires, ne prend pas en compte la diversité des formes des noms et mêle onomastique et relevé de motifs, comme l’Épée dans le Perron’.^[73] Bruce’s work is, in sum, suited more as a means of satisfying generic curiosity than as a basis of systematic, scholarly investigation.

To complete our tour d’horizon on the state of the art, let us consider some tools that are not strictly philological, but may nevertheless prove useful for consultation. The Dictionnaire des lieux arthuriens by Goulven Péron and the Arthurian Place Names of Wales by Scott Lloyd are dedicated to toponymy.^[74] Obviously, these useful and up-to-date resources nevertheless cannot address the problems raised by the particular status of a proper name in the medieval text.^[75] A fortiori, the same considerations apply to the numerous thematic dictionaries dedicated to Arthurian literature.^[76] Despite significant contributions and despite the availability of relevant digital resources, the Onomasticon Arthurianum therefore remains a chimera: a project often mentioned but as yet still unrealised. Could it ever be achievable?

5 Onomasticon Arthurianum: what is not there

Having briefly touched upon the main points surrounding the development of the Onomasticon, it may be useful to draw up an inventory not only of what exists, but also of what is missing. As far as Arthurian literature is concerned, numerous linguistic areas remain unexamined. To make up for these shortcomings, it may prove useful to provide a brief account of those resources which, although not strictly Arthurian, nevertheless constitute an important well of information.

Despite the very rich Arthurian tradition in Middle High German – and despite the early onomastic vocation and the presence of numerous indexes – there does not seem to exist an equivalent to West’s or Ackerman’s work for German literature^[77]. There is, however, a very useful repertory for the Arthurian field, although it relates only to anthroponyms: the Catalogue of Names of Persons in the German Court Epics. The Catalogue’s publication date, 1992, should not be misleading. As the editor Martin H. Jones specifies ‘the catalogue […] which is published here for the first time was compiled by the late Frank W. Chandler and submitted as his thesis for the M. A. degree of the University of London in May 1936’.^[78] The broad conception of ‘Court Epics’ allowed Chandler to include many Arthurian romances in his corpus: ‘the term “Court Epic” has been interpreted as widely as possible, embracing all poems which are of sufficient length to be called epics, from the fragments of Graf Rudolf to the fifteenth-century Lorengel, provided that such poems treat in some way or other with knights and their deeds’.^[79] Chandler’s index thus contains copious and useful data, but its inadequacies must also be noted. Firstly, Jones, Senior Research Fellow in the Department of German at King’s College London,^[80] merely republishes the thesis, with some bibliographical updates.^[81] The sources from which the Catalogue is constructed are the late 19th- or early 20th-century critical editions on which Chandler had worked. Moreover, name variants, arbitrarily provided, can only be retrieved through cross-references: for example, the entry ‘Gâwân’ redirects to the main entry (‘see Gawein’)^[82], but from the entry ‘Gawein’ it is not possible to access the varia lectio of the anthroponym. These absences are, however, consistent with the objectives of Chandler’s thesis, the subtitle of which reads: An Examination of Literary Sources and Dissemination, together with Notes on the Etymologies of the More Important Names. Again, one is confronted with an resource focussing more on character than on onomastics: the term ‘sources’ refers to the cultural horizon, outlined above. At the end of the 1980s, Friedhelm Debus, one of the most authoritative voices in onomastics, announced the start of a research project, to be carried out by computer processing, aimed at collecting all proper names from mediaeval Germanic texts.^[83] The initiative does not appear to have been completed. Debus, however, founded the series Documenta Onomastica Litteralia Medii Aevi (DOLMA), in which important monographs have been published: in addition to relevant theoretical acquisitions, each volume also offers rich study material from the Index nominum.^[84]

Mediaeval Dutch literature, which consists of a more manageable volume of text, benefits from a very useful tool: the Repertorium van Eigennamen in Middelnederlandse Literaire teksten or Répertoire des noms propres dans des textes littéraires en Moyen Néerlandais (REMLT).^[85] The REMLT includes all literary texts in Middle Dutch from the earliest attestations (13th century) until 1568. The project, coordinated by Willem Kuiper at the Meertens Instituut KNAW in Amsterdam, began in 1992 and is still ongoing, being continuously updated. The extensive and complete repertory (including anthroponyms, place names, names of objects and animals, names of authors) can be consulted online via a simple and intuitive interface. There are no variant readings, as the authors clearly state,^[86] but the breadth of references to the most comprehensive onomastic tools, the precision of the textual quotations and the multiple cross-matches make REMLT an accurate and reliable resource. The same considerations also apply to the Diccionario antroponímico del ciclo amadisiano (DINAM) edited by Maria Coduras Bruna of the University of Zaragoza.^[87] Although centred on the Iberian Amadis-Corpus, the dictionary, which can be consulted online, proves to be a useful resource for investigating Arthuriana. Here too, the varia lectio of names is not considered, but the rich and up-to-date bibliography given for each onomastic entry provides consistent critical breadth.^[88] Coduras Bruna is also the author of a useful monograph, emblematically entitled Por el nombre se conoce al hombre. Estudios de antroponimia caballeresca, where the Arthurian onomastic tradition receives some degree of investigation.^[89]

A clear map of the scholarly terrain emerges from this bibliographical tour d’horizon. Some paths have been comprehensively mapped out, others are only lightly sketched, and others remain completely uncharted. Arthurian onomastics has been well investigated in Middle English and French literature; for other literatures there are reliable tools, albeit subject to certain restrictions, be that genre, theme, or textual tradition; for other areas, however, our map remains blank. The onomastic study of Italian and Norse literature, for example, still contains large lacunae. And yet even the available resources, however praise-worthy, appear today to be in need of updating for a considerable number of reasons. First of all, the substantial increase in editions and bibliographies makes it necessary to consider new texts and traditions, examining, in particular, the many variants present in manuscripts in a more consistent and decisive manner. Besides, it is important to investigate what remains the most neglected chronological period. The 14th and 15th centuries are still largely under-investigated, thus leaving a wide time span uncovered.^[90]

It is also worth rethinking the practical principles on which the existing repertories were founded. Indeed, new and emerging technologies offer substantial opportunities that can no longer be ignored: the time has come for Arthurian onomastics to rely on computational tools. Digital data processing makes it possible to extend analyses to vast corpora, which cannot be managed by hand. Linguistic and philological-textual competence clearly remains a necessity, and this can only be offered by a scholar trained in the Humanities: no computer programming can imitate correct interpretation of the varia lectio. But the digital humanities make it possible to collect, represent and interpret large quantities of often complex and contradictory data more easily. Projects such as the REMLT and DINAM – which can be consulted remotely and are completely open-access – are virtuous examples.

To conclude, the digital humanities provide extraordinary opportunities for the creation of precise and sophisticated onomastic databases, which are necessary both to represent little-studied linguistic fields and to replace tools that remain fundamental and methodologically exemplary, but are outdated today.

One cannot help but notice that, despite the innovation which the 20th century brought us in many fields of the Humanities, critical tools in the field of literary onomastics have undergone a very moderate development. It has been said that the compilatory and encyclopaedic efforts, steeped in positivist rigour play a role within the cultural context of the time. The same is true of our own approach to the same problems, even if, in many cases, critical categories have evolved: the present neglect of the term ‘source’ is illustrative of the constitution of a new scholarly order. The sheer mention of the term could signal a methodological debt to German positivism and hence prevent any project from receiving funding. Yet, literary onomastics remains firmly rooted in Arthurian medievalism precisely because it bears a huge potential of providing solid data about the circulation of the Arthurian material within a given linguistic and cultural area and at a given time. It is therefore surprising to note the sorry state of the art regarding onomastic resources, especially in view of the vast development of the digital humanities.^[91] Most recently, Ferlampin-Acher hoped for the creation of an Arthurian onomastic dictionary, ‘un travail titanesque, nécessairement collectif et numérique’.^[92] Almost a century and a half later, we seem to be hearing Alfred Nutt’s words resound. The multiplication of theoretical contributions, the interlinguistic and interdisciplinary perspectives, the co-ordination of engineers competent at philological projects – and vice versa, of philologists well-versed in the digital domain –, all contributes to a sense that the time is ripe to complete this undertaking.^[93] The importance of cooperation (not only between philologists and computer scientists, but also between philologists of different languages) must be underlined once again. Working on modular architectures, establishing common operational criteria, defining gradually achievable objectives will make it possible to make ‘titanesque’ works much more accessible. Furthermore, sharing good practices can mean significant improvements. In general, a good critical and digital methodology for Arthurian onomastics works for, for instance, both Norse and Italian texts and traditions. The work of some will inevitably fall by the wayside, since implementing methodologies and structures will mean that someone will have to lead the way. Yet undoubtedly, it will mean that many can find a ready-made framework to work with rather than having to work from scratch: sharing is not only possible, but necessary – provided, of course, that a shared language with which to communicate can be established.^[94]

In the Oxford Advanced Learner’s Dictionary the word ‘chimera’, in its second meaning, is equivalent to ‘an impossible idea or hope’. Under the right conditions, the Onomasticon Arthurianum could be more than mere hope, or an idea appearing now and then in some academic article, but a concrete project, which today is conceptually and technically possible.