SOURCE: Millett, Bella. “The Origins of Ancrene Wisse: New Answers, New Questions.” Medium Aevum 61, no. 2 (1992): 206-28.

[In the following essay, Millett presents evidence that the author of Ancrene Wisse was more likely a Dominican than a Victorine.]

Although it is a book-length work, the early Middle English guide for recluses known as Ancrene Wisse gives us very few clues to its origin. Its author is anonymous. So is its audience: all that we know for certain is that the work was originally addressed to three well-bred (gentile) sisters, who ‘in the flower of your youth renounced all the joys of the world and became recluses’,¹ and that it was later revised (in the version surviving in the Corpus MS) for a group of recluses which had grown to twenty or more.² The only definite indication of its date is the author's assumption in the revised version that the recluses would be visited by both Dominican and Franciscan friars; since the Dominicans arrived in England in 1221 and the Franciscans in 1224, the revised version at least must have been produced sometime after 1224. As for its place of origin, the author writes in a variety of West Midlands English which probably originated in northern Herefordshire or southern Shropshire; but there are no indications of locality in the text itself, apart from a passage in the revised version which praises the community of recluses because, while they have now begun to spread ‘toward Englondes ende’ (‘towards the border of England’), they still maintain as much uniformity of observance ‘as if you were a single religious community of London or Oxford, Shrewsbury or Chester’.³

There have been many attempts to identify the author, audience, date, and place of origin of Ancrene Wisse more precisely, but—as Roger Dahood's 1984 survey of the literature makes clear⁴—none of them has been conclusive enough to gain universal acceptance. However, the solutions offered by Eric Dobson in The Origins of ‘Ancrene Wisse’ (1976),⁵ although they have been questioned, have not been seriously challenged,⁶ and his book remains the fullest and most influential discussion of the problem.

Following up a line of investigation first suggested by Derek Brewer in the 1950s,⁷ Dobson explored in detail both the Augustinian and the West Midlands connections of Ancrene Wisse, using a careful accumulation of circumstantial evidence to identify the work's most likely point of origin. In his first chapter, he argued that Ancrene Wisse must have been written by an Augustinian canon. Some earlier scholars had seen the author's obvious interest in pastoral care, and his comparative lack of interest in mystical experience, as evidence that he was a secular cleric rather than a member of a religious order;⁸ but Dobson identified three passages in the surviving manuscripts, all likely to be authorial, in which the writer refers to the practice of the religious community to which he belongs. The lack of direct influence from the Benedictine Rule, and the enthusiasm for the friars expressed in the revised version, made it improbable that he was a monk; and although Ancrene Wisse includes some striking parallels with Dominican legislation, Dobson believed that on geographical and chronological grounds (to be discussed more fully below) the author could not have been a friar. The most likely remaining possibility, given the frequent parallels in Ancrene Wisse with Augustinian practice, was that he was an Augustinian canon; this might account for the apparent links with Dominican usage, since the earliest Dominican legislation was based on Augustinian customs. In the second chapter, Dobson argued that the author must have belonged to one of the independent congregations of Augustinian canons, which followed a stricter rule than the ordinary Augustinian communities. He noted in particular the close connections between Ancrene Wisse and the statutes of the Premonstratensian canons; but also some apparent divergences, which he took as indications that its author belonged to an order ‘greatly influenced by or interested in the statutes of Prémontré but not itself Premonstratensian’.⁹ In the third chapter, he narrowed down the possibilities still further: if the author belonged to one of the independent Augustinian congregations, the only house at which Ancrene Wisse could have been written, taking the evidence of its dialect into account, was Wigmore Abbey in northern Herefordshire, a house of Victorine canons. This allowed him to concentrate his search for author and audience on the neighbourhood: in the fourth chapter, he argued that the original audience of three well-born sisters could be identified with the ‘sisters formerly living in the Deerfold’ near Wigmore, who are mentioned in a charter of c. 1250; and in the fifth that the author was Brian of Lingen, a member of the local gentry described in a list of founder's kin from the nearby house of Augustinian nuns, Limebrook Priory, as ‘canonicus secularis monesterii de Wigmore’.

The Origins of ‘Ancrene Wisse’ was received, in Dahood's words, with ‘admiration mingled with varying degrees of caution and skepticism’¹⁰—admiration for a cross-disciplinary study which defined the terms of the problem with great precision and broke new ground in its investigations, but also some reluctance to accept a solution which depended on a delicately balanced structure of hypotheses, based on evidence which was often incomplete or ambiguous. Fifteen years later, it is possible to check some of Dobson's conclusions against the findings of more recent research. How well have his arguments stood the test of time?

When the evidence is re-examined, a clear overall pattern emerges. The main thesis of his first two chapters—that Ancrene Wisse shows the influence of a tradition of monastic legislation inherited by the Dominicans from the Augustinian canons—has been considerably reinforced by later research. But the later stages of his argument, where he maintains that this tradition could have reached Ancrene Wisse through the Victorines, and that Ancrene Wisse itself must be of Victorine origin, now seem much less firmly based.

Even when Dobson was writing, the case for the influence of Victorine legislation on Ancrene Wisse was far from self-evident. The closest parallels in Ancrene Wisse to contemporary monastic legislation seemed to be rather with the earliest Dominican constitutions, and while the Dominicans had drawn heavily on Augustinian customs, the model they used was not the Victorine customary, the Liber Ordinis, but the Premonstratensian statutes.¹¹ Dobson acknowledged that the Augustinian parallels in Ancrene Wisse seemed to be predominantly with Premonstratensian rather than Victorine legislation,¹² he believed, however, that both Premonstratensian and Dominican authorship had to be ruled out on geographical and chronological grounds.

The problem lay in the form of Middle English in which Ancrene Wisse was originally written, the West Midlands literary dialect christened by J. R. R. Tolkien ‘Language AB’.¹³ The term ‘AB’ refers to the sigla of the two early thirteenth-century manuscripts which preserve it in its purest form, the Corpus MS of Ancrene Wisse, Cambridge, Corpus Christi College, MS 402 (A), and Oxford, Bodleian Library, MS Bodley 34 (B), which includes the five religious prose works (Seinte Katerine, Seinte Margarete, Seinte Iuliene, Hali Meiðhad and Sawles Warde) known as the ‘Katherine Group’. The development of this literary dialect must have taken place well before Ancrene Wisse was written (Dobson thought it likely that its distinctive spelling-system was evolved no later than the last quarter of the twelfth century),¹⁴ and the area in which it could have been developed is geographically limited.¹⁵ Dobson concluded that the author must have been a member of a well-established centre, either in northern Herefordshire or in southern Shropshire, where ‘language AB’ had been developed and where Ancrene Wisse and the works of the ‘Katherine Group’ were produced; but this would necessarily exclude the possibility that he was a Premonstratensian or a Dominican. There was no Premonstratensian house in the right area at the right time (the nearest was the abbey at Halesowen in north Worcestershire, where the canons did not move in until 1218), and the Dominicans did not even arrive in England until 1221. Dobson looked instead for ‘an Augustinian house of the requisite size and length of life, a member of an independent congregation’ within the right dialect area, and concluded that by far the most likely candidate was the abbey of Victorine canons at Wigmore—‘Only Wigmore Abbey fits.’¹⁶

This premise of Victorine authorship underlies much of the discussion of sources in The Origins of ‘Ancrene Wisse’; one of Dobson's main aims was to demonstrate that the practices the author describes in his work, both those he attributes to his own order and those he recommends for the recluses, could be connected to a specifically Victorine tradition. But if we take his work further in the light of more recent studies, the evidence points to a rather different conclusion.

The source-study of monastic legislation in the twelfth and early thirteenth centuries has to be undertaken with a proper awareness of its hazards. One problem is that we are often working with incomplete information: the survival-rate of early texts is extremely low, partly because of the normal attrition of time, partly because texts of this kind were heavily used, continuously revised, and sometimes deliberately destroyed when they became obsolete.¹⁷ Another is that monastic legislators tended to borrow extensively (often without explicit acknowledgement) from their predecessors. Even close and extended correspondence in content and wording need not indicate the direct dependence of one legislative work on another; it may mean only that both are part of the same general legislative tradition.¹⁸ Only the comparative study of a wide range of monastic customaries makes it possible to form an impression of which features are distinctive to a particular source and which are common property. But where the legislation on a particular point covered in Ancrene Wisse has survived in more than one contemporary or near-contemporary source, it is at least possible to determine which of these sources the text is closest to; and this may in turn help to indicate its likely origin.

Since Ancrene Wisse is a rule for female solitaries rather than for a male community, it cannot be assumed that all its prescriptions reflect the practice of the author's own order. Some were certainly taken over from Aelred of Rievaulx's twelfth-century rule for recluses, De Institutione Inclusarum,¹⁹ and others may have been adapted for the particular needs of the recluses. Our most important clue, therefore, to the order to which the author of Ancrene Wisse belonged is what he tells us, in the three passages identified by Dobson, about the practice of his own community.

The first, which is the only one to occur in more than one manuscript, appears in both the original and the revised version.²⁰ The author is advising the recluses on how frequently they should take communion:

People care less about what they have often. For this reason you should take Communion, as our brothers do, only fifteen times a year: (i) Christmas Day (ii) the feast of the Epiphany (iii) Candlemas Day (iv) a Sunday midway between that and Easter, or Lady Day if it is near the Sunday, because of its importance (v) Easter Day (vi) the third Sunday after Easter (vii) the feast of the Ascension (viii) Whit Sunday (ix) the feast of the Nativity of John the Baptist²¹ (x) the feast of St Mary Magdalen (xi) the Assumption (xii) the Nativity of the Blessed Virgin (xiii) the feast of St Michael (xiv) the feast of All Saints (xv) the feast of St Andrew.

The practice of taking communion fifteen times a year is only recorded for the Dominicans and (rather later) the Franciscans.²² In 1249, the Dominican annual general chapter decreed that in future the brethren should take communion as often as they were tonsured;²³ and the list of fifteen dates for tonsuring, as recorded in the Dominican constitutions from 1216 onwards, is quite closely similar in content and wording to the list in Ancrene Wisse:

Que rasura et tonsura fiant hiis terminis: prima in Nativitate, secunda inter Nativitatem et Purificationem, partito tempore, tertia in Purificatione, quarta inter Purificationem et Pascha, quinta in Cena Domini, sexta inter Pascha et Pentecosten, septima in Pentecoste, octava inter Pentecosten et festum Petri et Pauli, nona in festivitate eorundem, decima in festivitate sancte Marie Magdalene, undecima in Assumptione sancte Marie, duodecima in Nativitate ejusdem, tertiadecima in festo sancti Dyonisii, quartadecima in festo Omnium Sanctorum, quintadecima in festo beati Andree.²⁴

Vincent McNabb had argued in 1934 that the similarity in both the number of communions and the list of dates was evidence for the Dominican authorship of Ancrene Wisse;²⁵ Dobson, pointing out that the Dominicans only formally adopted the practice in 1249, saw it as more likely that ‘the Dominican practice must have been derived from whatever earlier order produced Ancrene Wisse … the English book, in fact … is an important primary document for the origins of the regulation’.²⁶ He noted that the linking of the dates of tonsuring and communion by the Dominicans in 1249 had been anticipated in the manuscript tradition of Ancrene Wisse itself. The version which lies behind the Titus and Pepys MSS (London, British Library, MS Cotton Titus D.xviii; Cambridge, Magdalene College, MS Pepys 2498) had been partially adapted for a male audience; and in this version, which almost certainly antedates the Dominican decision of 1249, the number of haircuts recommended is increased from the four which had been enough for a female audience to fifteen, thus matching the number of communions.²⁷ Dobson may have been right in assuming that the Dominican practice was based on that of an earlier order; but if so, the evidence now indicates a Premonstratensian rather than a Victorine origin. The lists of rasura dates in the second and third codifications of the Premonstratensian statutes (dating respectively from the mid-twelfth century and 1236-8) are considerably less close to Ancrene Wisse than the Dominican list. In 1979, however, A. H. Thomas published an article on a recently identified version of the statutes, in a manuscript of French provenance now in the Mitchell Library, Glasgow, intermediate between the second and third codifications.²⁸ This version includes a chapter, ‘De rasura et tonsura’, with a list of fifteen dates which is clearly closely related both to the list in Ancrene Wisse and to that in the Dominican constitutions:

Que rasura et tonsura fiat hijs terminis: prima in Natali Domini; secunda inter Natale et Purificatione, partito tempore; tercia in Purificacione; quarta [MS adds et] inter Purificationem et Pascha, partito tempore; quinta in Pascha; sexta inter Pascha et Pentecosten, partito tempore; septima in Penthecoste; octaua inter Pentechosten et festum Sancti Iohannis Baptiste, partito tempore; nona in festo Sancti Iohannis; decima in festo [MS feste] Sancte Marie Magdalene; vndecima in Assu[m]ptione Sancte Marie; duodecima in Natiuitate eiusdem; terciadecima in festo Sancti Dionisij; quartadecima in festo Omnium Sanctorum; quintadecima in Aduentu Domini.²⁹

The Victorine Liber Ordinis of c. 1200, on the other hand, has a list of rasura dates which is quite different in both content and wording from any of these three lists:

Tempora tonsurae haec sunt: a festo Omnium Sanctorum usque ad Pentecosten quatuor ebdomadarum emenso spatio; a Pentecosten usque ad festum Omnium Sanctorum post tres septimanas, nisi forte uel in aestate, uel in hieme aliqua imminens festiuitas, uel alicuius occasionis necessitas terminum constitutum tribus aut quatuor diebus praeuenire uel transire deposcat.³⁰

It looks as if Ancrene Wisse is drawing here not on Victorine customs, but on a tradition of legislation shared by the Premonstratensians and the Dominicans.

The second passage which may describe the practice of the author's order is found only in MS Cotton Nero A.xiv:

This is how our lay brothers say their Hours. For Matins on work-days twenty-eight Paternosters, on feast-days forty; for Vespers, fifteen; for every other Hour, seven. Before Matins, Pater noster and Credo, kneeling to the ground on a weekday, and bowing on a feast-day; and then whoever can shall sing Domine, labia mea aperies, Deus in adiutorium meum intende, Gloria Patri, Sicut erat, Alleluia, and in Lent Laus tibi Domine, rex eterne glorie. After the last [Pater noster],³¹Kyrie eleison, Christe eleison, Kyrie eleison, Pater noster, and after the ‘Amen’ Per dominum, Benedicamus Domino, Deo gratias. And begin and end in the same way at all the other Hours; except that at Compline whoever knows how to shall begin Conuerte nos, deus salutaris, and at all the other Hours Deus in adiutorium without Domine, labia mea. If any of you wishes to do it in this way, she will be following here, as in other observances, a great deal of our way of life, and I strongly recommend it.

Although this passage appears in only one manuscript, and overlaps with similar material elsewhere in the text, Dobson may well have been right in arguing that it was the author's own addition, ‘though doubtless it was an ill-considered one of which he may have thought better (since it is not in the definitive Corpus revision)’.³² Both the first-person address to the audience and the tone of the instructions suggest the author's own voice rather than that of an interpolator, and the Nero MS includes other additional material which is probably authorial.³³ Since the history of the passage appears to be complex, and also rather different from Dobson's tentative reconstruction,³⁴ I have discussed it in detail in an Appendix; what follows here is a summary of the conclusions the evidence suggests. The closest surviving parallel is in the early Dominican constitutions. While the Victorine Liber Ordinis does not include prescriptions for lay brothers, the regulations for their hours in the late thirteenth-century customary of Barnwell Priory, which followed Victorine practice, are quite different; and the corresponding regulations in the Premonstratensian statutes of the late twelfth and early thirteenth centuries, while they show some links with the passages in the Nero MS and the Dominican constitutions, are much briefer, omitting some of the details they share and differing on others. Dobson, because he did not believe that Dominican origin was possible, argued that the Premonstratensian, Nero and Dominican prescriptions must all be related to an earlier Augustinian source. The evidence now suggests that he was right; the source, however, is more likely to have been Premonstratensian than Victorine. The ultimate source of all three passages is the description of the lay brothers' hours in the early twelfth-century Cistercian Usus Conversorum; but it seems to have influenced them indirectly, through the legislation of the independent Augustinian congregations (which drew heavily in its early stages on Cistercian customs). Their closest surviving antecedent is the chapter on the lay brothers' hours in the constitutions of the Arrouaisian canons, which goes back to some time before 1147; this, however, alters some details shared by the Usus Conversorum and the three later texts, so it cannot be the direct source of the later versions. These must have been derived from an intermediate version between the Usus conversorum and the Arrouaisian regulations; and since the early legislation of Prémontré influenced that of Arrouaise, the intermediate version is most likely to have been Premonstratensian. It is also possible to read the later Premonstratensian regulations as a supplement to the prescriptions in the intermediate version rather than (as Dobson suggested) a simpler alternative to the common tradition; if this reading is correct, there is a good case for assuming that the Dominican regulations were based directly on early thirteenth-century Premonstratensian practice.

The third passage in which the practice of the author's order is mentioned occurs only in the revised version found in the Corpus MS: ‘In a-mel dei we seggeð ba Placebo ant Dirige efter þe mete graces, i twi-mel dei efter Non; ant ȝe alswa mote don’³⁵ (‘On a one-meal day we say both Placebo and Dirige following the grace after meals, on a two-meal day after Nones; and you may do the same’). As Dobson explained, this refers to the recitation of the Vespers (Placebo) and Matins (Dirige) belonging to the Office of the Dead on fast-days and other days. Dobson did not cite any Dominican parallel for this passage, and it is true that no direct evidence is available for the period we are concerned with, the first half of the thirteenth century, since the practice is not mentioned in the early Dominican constitutions and no early Dominican ordinarium survives; but the ordinarium of Humbert of Romans, who completed his definitive revision of the Dominican liturgy in 1256, prescribes:

Officium defunctorum hoc modo agatur in conventu cum agendum fuerit singulis septimanis:- Post nonam diei dominice siue alterius diei, uel post prandium tempore ieiunij, vel etiam post vesperas preterquam in Quadragesima si priori melius videatur, dicantur uespera pro defunctis cum uigiliis. Poterunt tamen differri interdum uigilie usque post uesperas diej quando vespere defunctorum dicuntur post nonam si uisum fuerit.³⁶

What this indicates is that the normal Dominican practice in the mid-thirteenth century was the same as that recommended in the revised version of Ancrene Wisse: the Vespers and Matins of the dead were said together, after Nones in the non-fasting season and after the single meal in the fasting-season, although the prior had authority to defer either both or the Matins alone until after the ordinary Vespers of the day if he thought fit. It is not clear how far back this practice goes; but what evidence we have indicates that the Dominicans may have begun to develop a liturgy of their own by the death of Dominic in 1221, and that it was certainly in existence by the mid-1230s.³⁷ Dobson's analysis of the regulations given in the Premonstratensian Liber Ordinarius³⁸ shows that they are markedly less close than the Dominican regulations to the practice described in Ancrene Wisse. In the fasting-season, Placebo is said post gratias refectionis, but the remainder of the Office is postponed either until before Vespers (outside Lent) or until the servers have eaten (in Lent); and in the non-fasting season, the Office as a whole is said before Vespers, not after Nones. But the same seems to be true of the Victorine regulations. While the Liber Ordinis of Saint-Victor is less explicit than the Premonstratensian Liber Ordinarius, there is no reason to assume that Victorine practice was closer to that of Ancrene Wisse; Dobson noted that the Office of the Dead was sometimes recited, ‘presumably on fast-days’, immediately after the graces following the meal,³⁹ but also that the chapter on the Vespers and Matins of the Dead ‘immediately precedes those dealing with Vespers and Compline … the order suggests that at St Victor the Office of the Dead preceded the ordinary Vespers’.⁴⁰

All three passages, then, in which the author tells us about the practices of his own order, can be paralleled in Premonstratensian and/or Dominican practice; but there is no positive evidence to support a connection with Victorine practice, and some evidence against it.

The difficulty persists if we extend the search more widely: there is no textual evidence, either in Ancrene Wisse itself or in the texts related to it, which can be used with confidence as evidence for Victorine authorship. Dobson argued that at least a circumstantial case could be made:

there is evidence, none of it perhaps conclusive in itself but nevertheless concurrent, to connect Ancrene Wisse and its group with St Victor—verbal parallels with the Liber Ordinis, the knowledge of Victorine writings (especially the commentary on the Augustinian Rule), and the independent descent of Sawles Warde and the corresponding chapters in the Victorine compilation De Anima from a distinctive redaction of the Anselmian dialogue De Custodia interioris hominis.⁴¹

However, Dobson himself conceded that the parallels he identified with the Liber Ordinis were ‘for the most part not very particular’, and although he identified ‘two striking phrasal agreements’,⁴² both are only three words long, and may simply reflect the tendency of monastic customaries to express similar content in similar wording. This lack of convincing parallels can be contrasted with the very clear links with Premonstratensian and Dominican legislative tradition, not only in content (as in the examples just given) but—as Dobson noted⁴³—in the structure of the work itself. When the Premonstratensian statutes received their first major overhaul, in the middle of the twefth century, the codifiers, probaby infuenced by the structure of Gratian's Decretum (c. 1140),⁴⁴ divided them by subject into four main sections (distinctiones), further subdivided into chapters (capitula), and added a preface emphasizing the need for uniformity of observance and explaining the new layout. This structure was taken over by some later orders who based their legislation on the Premonstratensian statutes, including the Dominicans (who retained much of the preface, but reduced the number of distinctiones to two); and the author of Ancrene Wisse seems to be working within the same tradition. He not only divides his work into eight destinctiuns, sometimes subdivided into chapitres,⁴⁵ but he introduces it by a preface which shows the direct influence of the Premonstratensian/Dominican model; there are no comparable parallels with the Victorine Liber Ordinis (which, like most other twelfth-century customaries, is divided only into chapters). The Premonstratensian statutes were certainly influential beyond the order itself in the twelfth and early thirteenth centuries, but it is hard to see why a Victorine author should draw on them to what looks like the complete exclusion of the regulations of his own order.

As for the second point, any debt owed by the author to Victorine authors need not indicate his own order, as Dobson himself freely acknowledged:

A man may read many books, written by members of a variety of orders or institutions, or of none, and yet not belong to the same society as even the writers by whom he is most influenced; it is risky to judge his very nationality by his intellectual debts … In general I do not think that the intellectual and literary debts of Ancrene Wisse and the other works of its group can be used as evidence of the order to which their authors belonged.⁴⁶

But Dobson did make one exception to this rule, and that is his third main supporting argument for Victorine origin: that the source-text of one of the works of the ‘Katherine Group’, Sawles Warde, had Victorine connections, and this suggested that the ‘AB centre’ itself was a Victorine community:

How did it come about that, in spite of the wide early circulation of MSS of the original text [Anselm's De Custodia] (including a good early copy belonging to Hereford Cathedral), the West Midland author of the free and independent English adaptation used … the same distinctive version of the Latin as formed the basis of the redaction embodied in the composite De Anima? It can hardly be doubted that the latter was compiled in the abbey of St Victor in Paris, since it appears to survive only in Victorine texts and is ascribed, dutifully but falsely, to Hugh of St Victor.⁴⁷

Dobson's answer was that Sawles Warde must have been written in an English daughter-house of Saint-Victor, which ‘might well have in its library manuscripts copied from Victorine exemplars’.⁴⁸ But Wolfgang Becker, in a 1980 article, showed that the readings shared by Sawles Warde and the De Anima text can be found together elsewhere in the manuscript tradition of Anselm's De Custodia, and that Sawles Warde and De Anima must have used different source-texts, so there is no need to assume a connection between them.⁴⁹

This review of the evidence shows that recent work has, if anything, added to the difficulty of the original problem which Dobson was trying to solve: the apparent conflict between the textual evidence of Ancrene Wisse, which suggests that its closest affinities are with the tradition of monastic legislation running from the Premonstratensians to the Dominicans, and the linguistic evidence, which suggests that neither Premonstratensian nor Dominican origin is possible. Dobson took the linguistic evidence as his starting-point, and tried to find ways of accommodating the textual evidence to it; but his research did not produce the positive results which would have confirmed his theory, and later work has eroded the case for Victorine origin still further. Perhaps it is time to approach the problem from the opposite direction: assuming that the author was Premonstratensian or Dominican, how do we account for the linguistic evidence linking him and his work with a scribal centre which cannot have been a Premonstratensian or Dominican house? The simplest way round the difficulty would be to abandon the assumption that he must have belonged to that centre at the time when he wrote Ancrene Wisse. It is reasonable to assume that he both had originally, and maintained, some kind of connection with the ‘AB centre’: his own written English seems to have been a form of ‘language AB’,⁵⁰ and the centre must have had access to his revised fair copy of Ancrene Wisse to produce the excellent Corpus text. But he may only have received his schooling there; or he may have been a member of the centre, but left it to join the order to which he belonged when he wrote Ancrene Wisse.⁵¹ This solution is less elegant than Dobson's, but it is not intrinsically improbable, and it offers us a way of reconciling the textual and linguistic evidence without doing unnecessary violence to either.

If the author was not a Victorine, which order did he belong to? Various kinds of evidence—his obvious interest in confession, what he says about the practice of his own order, and the references to religious orders in the text—suggest that it was the Dominicans rather than the Premonstratensians. The hypothesis of Dominican authorship is not new—it was first suggested by J. B. Dalgairns in 1870,⁵² later maintained in a series of articles by the Dominican Vincent McNabb,⁵³ and revived in a 1954 article by Clare Kirchberger.⁵⁴ But it has never been accepted by mainstream Ancrene Wisse scholars, partly because of a long-standing reluctance, which is only now disappearing, to accept the post-1216 date that Dominican origin would involve,⁵⁵ partly because McNabb, the main advocate of the theory, presented his evidence in a way that did it rather less than justice.⁵⁶ Nevertheless, it is time to look at the case for Dominican origin again.

In the first place, it helps to explain the otherwise surprising emphasis on confession in Ancrene Wisse. As has been noted more than once,⁵⁷ almost half of the work is taken up with an extended discussion of temptation, confession and penance, which includes some material explicitly directed towards a wider audience than the recluses themselves: towards the end of the distinctio on confession, the author says, ‘My dear sisters, this fifth section, on Confession, is relevant to everybody alike; so do not be surprised that I have not addressed you particularly in this section’. This suggests an interest in pastoral care which is much more Dominican than Premonstratensian. The Premonstratensians had originally engaged in preaching and missionary activity, particularly in Germany, but by the early thirteenth century there was little difference between their way of life and that of the older monastic orders.⁵⁸ While they could, and sometimes did, serve the parish churches they owned, there is no firm evidence for the practice in England before the second half of the thirteenth century, and even then the canons' motives seem to have been more financial than evangelical.⁵⁹ The Dominicans, on the other hand, were actively involved in the implementation of the programme of pastoral reform laid down by the Fourth Lateran Council in 1215; one of their main functions was to assist the bishops with their increased pastoral workload by preaching, the hearing of confessions, and the provision of spiritual advice.⁶⁰ Both the emphasis on confession in Ancrene Wisse and the author's tendency to look over the shoulders of his primary audience to a wider readership⁶¹ seem to reflect this pastoral concern.

In the second place, what the author says about the practice of his order, even though it is sometimes closer to Premonstratensian legislation than to the surviving Dominican regulations, seems on balance to point towards Dominican authorship.

The earliest of his three references, the list of communion dates which appears in the original as well as the revised text, is rather closer to the list of rasura-dates given in the Glasgow MS of the Premonstratensian statutes than to that in the Dominican constitutions; it agrees with the Dominicans against Glasgow on one date, the feast of St Andrew (30 November), as against Glasgow's vaguer in Aduentu Domini, but with Glasgow against the Dominicans on two: Easter Day, as against Maundy Thursday (Cena Domini), and the feast of St John the Baptist (24 June), as against the feast of Sts. Peter and Paul (29 June). The most significant of these agreements, since it helps us to date the passage in Ancrene Wisse, is on the feast of John the Baptist. This feast also appears in the list of rasura dates in the 1236-8 codification of the Premonstratensian statutes; but the earlier Premonstratensian statutes, like the Dominican constitutions, have the feast of Sts. Peter and Paul instead. As A. H. Thomas has pointed out,⁶² the substitution of the feast of John the Baptist in the Premonstratensian regulations is likely to have taken place after the Dominicans had borrowed their rasura regulations in 1216, giving us a terminus a quo for the currency of the Glasgow list; and a terminus ad quem is provided by the 1236-8 Premonstratensian statutes, with their revised list of rasura dates. Although the Glasgow list and the one in Ancrene Wisse are not identical, the close correspondences between them suggest that the author's community was directly influenced by Premonstratensian practice during the time when the Glasgow list was current.

The significance of the second reference, to the lay brothers' hours, is harder to determine, as it is not entirely clear what the Premonstratensian practice was. If (as I have suggested above) the Premonstratensian routine was in fact closer to Dominican practice than the surviving Premonstratensian statutes might seem to indicate, the routine described in the Nero MS would fall midway between Premonstratensian practice and the practice recorded in the Dominican constitutions; otherwise, the Nero routine would be considerably closer to Dominican than to Premonstratensian practice.

The third reference, which is also the latest in date, points unambiguously to Dominican origin. The successive versions of the Premonstratensian Liber Ordinarius show that throughout the thirteenth century the Premonstratensian method of reciting the Office of the Dead was different from that described in Ancrene Wisse,⁶³ whereas the Dominican method, at any rate from the evidence of the earliest Ordinarium (1256) to survive to us, was identical.

What is striking about the first two of these three passages is their ‘intermediate’ character: while they clearly belong within the legislative tradition transmitted from the Premonstratensians to the Dominicans, they seem to hover somewhere between the recorded practices of the two orders, not quite matching the surviving legislation of either. The most probable explanation for this is that they are of Dominican origin, but date from a transitional period, the two decades following 1216, when Premonstratensian and Dominican customs were running closely alongside each other and local Dominican practice was still far from uniform.⁶⁴ Their links with Premonstratensian legislation could then be explained by the initial and continuing influence on Dominican practice of the Premonstratensian statutes. The third and last reference, which directly reflects Dominican practice, strongly reinforces the case for Dominican origin. The reverse hypothesis, that the author of Ancrene Wisse belonged to a Premonstratensian house under Dominican influence, is much less probable. The Premonstratensians were an influential and well-established order, with a legislative tradition going back to the early twelfth century and a high degree of central organization,⁶⁵ they would be most unlikely to feel the need to borrow the customs of a newly founded order which had only recently borrowed theirs.

There remain the references to religious orders in the text of Ancrene Wisse. The only two orders specifically mentioned are the Dominicans and the Franciscans, who are referred to twice, each time in additions to the original text. The first reference is found in the Vitellius French version, and in the revised version in the Corpus MS. The original text warns the recluses to be on their guard against clerical visitors, who may be wolves in sheep's clothing—‘Put little trust in seculars, even less in religious’—but the addition makes an exception for the friars: ‘Our Dominican and Franciscan friars follow such a way of life that everybody might be amazed if any of them “turned his eye towards the woodland glade”.⁶⁶ Therefore, whenever any of them comes out of charity to instruct you and give you comfort in God, if he is a priest say before he leaves, “Mea culpa; I confess to almighty God and to you …”’. The second passage appears only in the revised version in the Corpus MS, and again refers to the special treatment of visiting friars: ‘No man should eat in your presence except by your director's leave, general or special: general for Dominican and Franciscan friars, special for all others’. Dobson comments on the first passage, ‘If [the author] had himself been a friar, he would surely have written “any of us” rather than “any of them” … and however friendly he may have been to members of the other Order, he could hardly have written with such complete impartiality of “vre freres prechurs ant vre freres meonurs”.’⁶⁷ But the first objection is not conclusive (particularly if the author himself was not in a position to visit the recluses), and the second is debatable. While the Augustinian canons in general, and the Premonstratensians in particular, might well have felt a close affinity with the Dominicans, who had adopted Premonstratensian customs, they would not have had the same reason to feel close to the Franciscans. On the other hand, there was certainly a special relationship, at any rate in the early stages, between the two mendicant orders: the Dominican constitutions recommend that ‘Fratres minores sicut et nostri caritative et hilariter recipiantur et secundum facultatem domus pie et honeste procurentur.’⁶⁸ And it is probably significant that the author of Ancrene Wisse does not mention any other religious order in these passages: if he was not a friar, surely he would have exempted at least his own order from his generalizations about lecherous religious and his blanket prohibition of visits to the recluses without special permission?

There is a third passage in the text which may also allude to the author's order, but its interpretation is more uncertain. In the revised version, the author praises the recluses for living in ‘one manner of life, as if you were a single religious community of London or Oxford, Shrewsbury or Chester’.⁶⁹ The main point of the comparison is that the recluses, scattered as they are, nevertheless reach the same high standards of uniformity as an organized monastic community; but the additional details may indicate that the author had a specific model in mind. As McNabb long ago pointed out,⁷⁰ all four of the towns mentioned were the sites of early Dominican priories. Oxford (1221) and London (1224) were the first two Dominican foundations in England, and Shrewsbury (by 1232) and Chester (by 1236) the first two in the West Midlands; they would be a natural point of reference for an early English Dominican writing for a West Midlands audience. If so, the image may help us date the revised version of Ancrene Wisse more precisely, with the foundation of the priory in Chester as a terminus a quo.

The assumption of Dominican authorship for Ancrene Wisse would explain a number of its features which are otherwise problematic: its heavy emphasis on confession, its close relationship to the Dominican constitutions and liturgy,⁷¹ and the author's enthusiasm for the friars. It would also help us to give a more precise terminus a quo for both the original and the revised version than has been possible up to now. The original version must be after 1216, the date of the formation of the Dominican order and the drafting of its earliest constitutions, and is likely to post-date—probably by some years—the arrival of the Dominicans in England in 1221; while the revised version, if (as is possible) it refers to the Dominican house at Chester, cannot be dated much before 1236.⁷²

But at the same time this assumption re-opens questions of provenance, audience and authorship which seemed to have been at least provisionally settled. If it is correct, Wigmore Abbey cannot be the community to which the author belonged when he wrote Ancrene Wisse; it could still, in theory, have been the ‘AB centre’, but there are other possibilities—particularly since we no longer need to assume that the centre was a house of regular Augustinian canons, or even that it was a monastic community at all.⁷³ Dobson's identifications of author and audience also need to be reconsidered. Even without the assumption of Dominican authorship, they rest on uncertain foundations: the three sisters to whom Ancrene Wisse was originally addressed cannot have been ‘the sisters formerly living in the Deerfold’ near Wigmore, who are the accidental product of a misread abbreviation,⁷⁴ and Brian of Lingen, ‘canonicus secularis monesterii de Wigmore’, was not necessarily a member of the Victorine community at Wigmore Abbey.⁷⁵ If we assume Dominican authorship, Brian must in any case be ruled out as author, and there is no longer any reason for us to limit the search for the original three sisters to the immediate neighbourhood of Wigmore.

The importance of Dobson's study of the origins of Ancrene Wisse now seems to lie less in the specific answers he offered than in his precise definition of the problems involved, and of the kind of evidence which would be needed to solve them. His investigation of the relationship of Ancrene Wisse to twelfth- and thirteenth-century monastic legislation, much fuller and more systematic than that of previous scholars, opened up an area which had not been adequately explored in the past; and although the research on which this article is based has led me to different conclusions from his, it is essentially a continuation of his work. The author, audience, and place of composition of Ancrene Wisse are still unidentified, and perhaps unidentifiable; but the outlines of its broader institutional and cultural context are beginning to emerge more clearly.

Notes

1. London, British Library, MS Cotton Nero A.xiv, f. 50^r: The English Text of the ‘Ancrene Riwle’, edited from Cotton Nero A. XIV, ed. by Mabel Day, EETS, os, 225 (London, 1952), p. 85/26-7. In all quotations from manuscripts and diplomatic editions, I have expanded abbreviations and modernized punctuation for greater readability.

2. Cambridge, Corpus Christi College, MS 402, f. 69^r/13 28: The English Text of the ‘Ancrene Riwle’: ‘Ancrene Wisse’, edited from MS Corpus Christi College Cambridge 402, ed. by J. R. R. Tolkien, EETS, os, 249 (London, 1962).

3. MS Corpus 402, f. 69^r/27-8, 20-1.

4. Roger Dahood, ‘Ancrene Wisse, the Katherine Group, and the Wohunge group’, in Middle English Prose: a Critical Guide to Major Authors and Genres, ed. by A. S. G. Edwards (New Brunswick, NJ, 1984), pp. 1-33 (pp. 8-11).

5. E. J. Dobson, The Origins of ‘Ancrene Wisse’ (Oxford, 1976).

6. See the reviews by O. Arngart, English Studies, LIX (1978), 154-5; Siegfried Wenzel, Speculum, LIII (1978), 354-6; and Gillis Kristensson, Studia Neophilologica, LIII (1981), 371-6.

7. D. S. Brewer, ‘Two notes on the Augustinian and possibly West Midland origin of the Ancrene Riwle’, Notes and Queries, CCI (1956), 232 5.

8. The point is most fully discussed by Gerard Sitwell in his ‘Introduction’ to The Ancrene Riwle, trans. by M. B. Salu (London, 1955), pp. vii-xxi.

9. Origins, p. 13.

10. Dahood, ‘Ancrene Wisse …’, p. 11.

11. See the thorough study of sources in A. H. Thomas, De oudste Constituties van de Dominicanen: voorgeschiedenis, tekst, bronnen, ontstaan en ontwikkeling (1215-37), Bibliothèque de la Revue d'Historie Ecclésiastique, 42 (Louvain, 1965), esp. ch. iii.

12. Dobson, Origins, pp. 109-11.

13. J. R. R. Tolkien, ‘Ancrene Wisse and Hali Meiðhad’, Essays and Studies, XIV (1929), 104-26.

14. Dobson, Origins, pp. 122-6.

15. Gillis Kristensson, in his review of The Origins of ‘Ancrene Wisse’, is agnostic on Dobson's localization to northern Herefordshire or southern Shropshire (‘Nothing in my material [for A Survey of Middle English Dialects 1290-1350, II: The West Midland Counties] disproves localization of the AB language to Herefordshire-Shropshire, but on the other hand nothing proves it’; he thinks that Staffordshire, which belongs to the same dialect subgroup, remains a possibility (Studia Neophilologica, LIII (1981), 374)); but Jeremy Smith, who has been working recently on the localization of Ancrene Wisse manuscripts, comes to much the same conclusion as Dobson: ‘The balance of the argument would … continue to suggest a localization for AB-language in North Herefs. or the southern tip of Salop; but the form for “when” is some indication that a more northerly location might be possible. At our present state of knowledge, I'd be reluctant to accept a more northerly position than mid-Salop’ (private letter, 14 January 1991).

16. Dobson, Origins, p. 172.

17. See Thomas, De oudste Constituties, p. 79, for a general discussion of the problem; William R. Bonniwell, A History of the Dominican Liturgy 1215-1945, 2nd edn (New York, 1945), p. 18, on the possibility that the Dominicans deliberately destroyed almost all their earliest liturgical texts. Of an estimated 4,000 mediaeval manuscripts of the Carthusian rule, only seventeen survive, of which the three earliest are presentation copies to other monastic houses: see Guigues I^er, prieur de Chartreuse: Coutumes de Chartreuse, ed. par un Chartreux, Sources chrétiennes, 313, Série des textes monastiques d'Occident, 52 (Paris, 1984), pp. 93-105.

18. See Thomas, De oudste Constituties, pp. 125-7, for references, particularly on the links between the legislation of the Cistercians, the Carthusians and the independent Augustinian congregations.

19. Aelredi Rievallensis Opera Omnia, 1: Opera Ascetica, ed. by A. Hoste and C. H. Talbot, Corpus Christianorum, Continuatio Mediaevalis, 1 (Turnhout, 1971), pp. 635-82. Alexandra Barratt, ‘Anchoritic aspects of Ancrene Wisse’, Mæ, XLIX (1980), 32-56, makes a good case for Carthusian influence on Ancrene Wisse, but she is right to be tentative about specifically legislative influence (p. 46): the regulations she cites on bloodletting, clothing, fasting and abstinence can all be paralleled elsewhere, and there are no correspondences striking enough to prove a direct debt.

20. See Dobson, Origins, pp. 9-11, for a discussion of the textual evidence.

21. For this translation of Midsumer Dei, see MED, s.v. ‘mid-somer’ 2.(b); also the earlier French translation of Ancrene Wisse, and the Latin translation (see The French Text of the ‘Ancrene Riwle’, edited from British Museum Cotton Vitellius F vii, ed. by J. A. Herbert, EETS, os, 219 (London, 1944), p. 303; The Latin Text of the ‘Ancrene Riwle’, edited from Merton College MS. 44 and British Museum MS. Cotton Vitellius E vii, ed. by Charlotte d'Evelyn, EETS, os, 216 (London, 1944), p. 165).

22. See Peter Browe, Die häufige Kommunion im Mittelalter (Munster, 1938), pp. 81-2.

23. Monumenta Ordinis Fratrum Praedicatorum Historica, 3: Acta Capitulorum Generalium 1, ed. by B. M. Reichert (Rome, 1898), p. 44, lines 9-12. Browe (see n. 23 above) cites no evidence for his statement that the Dominicans had originally taken communion less frequently.

24. Dist. i, ch. xx (ed. Thomas, De oudste Constituties, p. 331).

25. Vincent McNabb, ‘The authorship of the “Ancren Riwle”’, Archivum Fratrum Praedicatorum, IV (1934), 53-5.

26. Dobson, Origins, p. 64.

27. Ibid., pp. 64-5.

28. A. H. Thomas, ‘Une version des statuts de Prémontré au début du xiii^e siècle’, Analecta Praemonstratensia, LV (1979), 153-70.

29. Dist. iv, ch. xv (Glasgow, Mitchell Library, MS 308892, f. 31^v).

30. Ch. lxii, De ordine tonsurae (Liber Ordinis Sancti Victori Parisiensis, ed. by L. Jocqué and L. Milis, Corpus Christianorum, Continuatio Mediaevalis, 61 (Turnhout, 1984), pp. 242-3).

31. See Dobson, Origins, p. 73 n. 4: this is what the sense must be, but the ordering of points in the Nero passage makes it ambiguous at first sight.

32. Ibid., p. 8.

33. Ibid., pp. 6-9.

34. Ibid., pp. 73-84.

35. MS Corpus 402, f. 6^v/7 9.

36. London, British Library, MS Add. 23935, f. 35^ra, De officio defunctorum. This manuscript was Humbert's own copy: see Bonniwell, A History of the Dominican Liturgy, pp. 94-7. The Ordinarium is edited by M. Guerrini, Ordinarium iuxta Ritum Sacri Ordinis Fratrum Praedicatorum (Rome, 1921).

37. See Archdale A. King, Liturgies of the Religious Orders (London, 1955), pp. 331-8; Raymond Creytens, ‘L'ordinaire des frères prêcheurs au moyen âge’, Archivum Fratrum Praedicatorum, XXIV (1954), 108-88.

38. Dobson, Origins, pp. 59-62; see L'Ordinaire de Prémontré d'après des manuscrits du xii^e et du xiii^e siècle, ed. by Pl. F. Lefèvre, Bibliothèque de la Revue d'Histoire Ecclésiastique, 22 (Louvain, 1941), ch. lxv (p. 113).

39. See Dobson, Origins, pp. 58-9; Liber Ordinis Sancti Victoris, ed. Jocqué and Milis, ch. xxxv, lines 171 4 (p. 174).

40. Dobson, Origins, p. 59; see Liber Ordinis Sancti Victoris, ed. Jocqué and Milis, ch. lix (pp. 239-41).

41. Dobson, Origins, p. 170.

42. See ibid., pp. 110-11.

43. Ibid., p. 109.

44. See Les Statuts de Prémontré au milieu du xii^e siècle, ed. by Pl. L. Lefèvre and W. M. Grauwen, Bibliotheca Analectorum Praemonstratensium, 12 (Averbode, 1978), p. vi.

45. MS Corpus 402, f. 4^r/19-23, 26; and cf. f. 111^r/21-2.

46. Dobson, Origins, pp. 141-2, 146.

47. Ibid., p. 153.

48. Ibid.

49. ‘The source text of Sawles Warde’, Manuscripta, XXIV (1980), 44-8.

50. This is indicated by the ‘Scribe B’ annotations in the Cleopatra MS of Ancrene Wisse, which Dobson convincingly argued were the author's own (The English Text of the Ancrene Riwle, edited from B. M. Cotton MS. Cleopatra C. VI, ed. by E. J. Dobson, EETS, os, 267 (London, 1972), pp. xciii-cxl); Dahood (‘Ancrene Wisse …’, pp. 3-4) has questioned the security of this identification, but it remains the simplest and most plausible interpretation of the evidence.

51. When the Dominicans came to England, they recruited not only students but ordained priests, and some members of the older religious orders: see W. A. Hinnebusch, The Early English Friars Preachers (Rome, 1951), pp. 261-71.

52. See n. 82 below. [The Scale (or Ladder) of Perfection, written by Walter Hilton, with an Essay on the Spiritual Life of Medieval England, ed. by J. B. Dalgairns (1870; repr. London, 1901), p. ix.]

53. ‘The authorship of the “Ancren Riwle”’, Modern Language Review, XI (1916), 1-8; ‘Further light on the “Ancren Riwle”’, Modern Language Review, XV (1920), 406-9; ‘Further research upon the Ancren Riwle’, Review of English Studies, II (1926), 82-5 (see also pp. 85-9, 197-8, 199-201); and ‘The authorship of the Ancren Riwle’, Archivum Fratrum Praedicatorum, IV (1934), 49-74.

54. Clare Kirchberger, ‘Some notes on the Ancrene Riwle’, Dominican Studies, VII (1954), 215-38.

55. See Dahood, ‘Ancrene Wisse …’, pp. 8 ii. Dobson himself dated the original version of Ancrene Wisse at c. 1215 21, between the Fourth Lateran Council and the arrival of the friars; and was prepared to concede, at least in theory, that a date as late as the early 1230s was possible (see Dobson, Origins, pp. 15-16).

56. For an admirably fair summary of the merits and demerits of McNabb's arguments, see Dobson, Origins, chs. i and ii (esp. p. 14 n. 2).

57. See Kirchberger, ‘Some notes on the Ancrene Riwle’, pp. 230-1; Sitwell, in The Ancrene Riwle, trans. Salu, pp. xviii-xii.

58. See H. M. Colvin, The White Canons in England (Oxford, 1951), pp. 1-25.

59. See ibid., pp. 272-88.

60. See Hinnebusch, The Early English Friars Preachers, pp. 279-96.

61. See Dobson, Origins, pp. 250-3.

62. Thomas, ‘Une version des statuts de Prémontré’, pp. 167-8.

63. See n. 40 above; also Michel van Waefelghem, Liturgie de Prémontré: Le Liber Ordinarius d'après un manuscrit du xiii^e/xiv^e siècle (Louvain, 1913), pp. 578-9.

64. Humbert of Romans in the mid-thirteenth century contrasted the variations in Dominican practice which had persisted up to his own time with the relative uniformity of the older monastic orders: ‘Non solum in aedificiis et in habitu, sed etiam in consuetudinibus quibusdam tam in officio divino quam in aliis multis varietas per provincias diversas, immo per domos ejusdem provinciae reperitur quoad multa’ (Expositio magistri Humberti super constitutiones Fratrum Praedicatorum, in B. Humberti de Romanis Opera de Vita Regulari, ed. by J. J. Berthier (1889; repr. Turin, 1956), 1, 6; trans. by Simon Tugwell, Early Dominicans: Selected Writings (London, 1982), p. 142); Humbert goes on to discuss the reasons for this diversity and to recommend greater uniformity in the future.

65. Humbert of Romans says that the Dominicans adopted the customs of the Premonstratensians because they surpassed all other orders of Augustinian canons ‘in vitae austeritate, in observantiarum pulchritudine, in discreto maximae multitudinis regimine per capitula generalia, et visitationes, et hujusmodi’ (Expositio super constitutiones, in B. Humberti de Romanis Opera de Vita Regulari, ed. Berthier, 1, 2-3; see M. H. Vicaire, Saint Dominic and his Times, trans. Kathleen Pond (London, 1964), p. 207).

66. Odo of Cheriton (†c. 1247), telling the fable of the wolf who tried to become a monk but found that his thoughts kept turning to rams and lambs, makes a similar point about religious who cannot give up their animal appetites, and illustrates it with a snatch of alliterative verse which he says is proverbial:

    Thai thu wolf hore          hodi to preste,
    tho thu hym sette          salmes to lere,
    evere beth his geres          to the groue-ward.

    Even if you ordain a grey wolf as a priest, and set him to learn psalms, his natural inclinations are still to the wild.

    See The Fables of Odo of Cheriton, ed. and trans. by John C. Jacobs (Syracuse, NY, 1985), Fable 34 (pp. 92-3); Cecily Clark, ‘A mediaeval proverb’, English Studies, XXXV (1954), 11-15; and the contribution by Andrew Breeze in the present issue (pp. 284-8 below).

67. Dobson, Origins, p. 15.

68. Dist. ii, ch. xxxiv (ed. by Thomas, De oudste Constituties, p. 366); this provision probably dates from between 1221 and 1235. The emphasis on the special friendship between the two orders is repeated in Dominican capitular decrees of 1236 and 1242 (though in terms which suggest that internecine jealousies were already a problem): see Acta Capitulorum Generalium, I, 9 (lines 16-26), 24 (lines 4-5).

69. MS Corpus 402, f. 69^r/19-21.

70. Vincent McNabb, ‘The authorship of the “Ancren Riwle”’ (1916), p. 4.

71. I have not discussed all the liturgical parallels; for further examples, see the articles by McNabb and Kirchberger cited in notes 55 and 56 above.

72. Alan Fletcher argues from other internal evidence that both Ancrene Wisse and Hali Meiðhad are likely to post-date the coming of the Franciscans to England in 1224: see Alan J. Fletcher, ‘Black, white and grey in Hali Meiðhad and Ancrene Wisse’, forthcoming in MÆ, LXII (1993).

73. See Dobson, Origins, pp. 126-35, for a detailed survey of the religious centres within the ‘AB’ dialect area.

74. The charter refers to fratres, not sorores: see Sally Thompson, Women Religious: the Founding of English Nunneries after the Norman Conquest (Oxford, 1991), p. 34 n. 126.

75. It is puzzling that a member of a community of regular canons should be described as secularis. Dobson's explanation was that Brian was described as secular because he was working ‘among men’, seconded from the community to serve a parish church or act as domestic chaplain (see Dobson, Origins, pp. 319-23). E. O. Blake, however, has pointed out another possibility to me: that monasterium here may mean not ‘monastery’, but ‘collegiate church’ (the parish church of Wigmore was a small collegiate church, served by three canons), in which case Brian need not have been a canon of Wigmore Abbey.