Introduction I
INTRODUCTION
1 he canon of St. Anselm's genuine finished works can be very clearly defined. I his was the first point that needed to be established in modern
Anselmian studies. In the Middle Ages the body of his writings was encumbered with a large baggage of spurious works, and the Maurist edition (1675; 2nd edit., 1721) of Dom Gabriel Gerberon perpetuated this state of confusion until recent times. It was only in 1923 that Dom
Andre Wilmart began the systematic work of distinguishing the genuine works from the host of spurious additions, and the succession of his articles from 1923 to 1932 laid the foundation on which all later workers have built. Without undue rashness, it may be said that the task of dis engaging the hard core of Anselm's works has now been completed.
Anselm's own works stand out stripped of the fictitious additions of several centuries.
The main factor in making possible this new clarity was undoubtedly the excellence of the early manuscript tradition. To this may be added
Anselm's own fastidious literary taste, which distinguishes his work from most imitations. Finally, there is the testimony of his biographer, who took care to record the circumstances in which most of his works were written. The result is that we now have a picture of St. Anselm as an author which is not likely to be much altered by future discoveries.
Nevertheless the line between the genuine and the spurious is not at all points equally clear. There are degrees of inauthenticity, and not all the works which must be excluded from the inner company of finished works merit total expulsion. To adapt one of Anselm's own images, there is an inner citadel where all is safe and assured: these are the works which stand beyond all doubt. But outside the citadel there is a bailey, less safe, but still part of this main structure. It is with this that we are here concerned. To put the matter in less picturesque language, we are con cerned with the work of a small group of Anselm's friends and disciples, who recorded his words, transcribed his writings before they were com pleted, and preserved his unfinished drafts and notes. The evidence of their activity has been preserved in comparative abundance. The very excellence of the early manuscript tradition, which makes it possible to draw a clear line between genuine and spurious works, has also ensured the preservation of much that cannot be assigned to either of these clearly defined categories. Roughly speaking, the works in this inter mediate zone fall into two classes. They are either reports of Anselm's
C 3501 B
INTRODUCTION
words written by one of his close companions, or they are remains of unfinished works or notes rescued by zealous disciples from the oblivion into which they would otherwise have fallen.
The reports of Anselm's words consist mainly of records made by Anselm's secretaries, Eadmer and Alexander, between 1093 and 1109.
These two monks of Canterbury had unique opportunities for hearing and reporting Anselm's words during these years. There is not the slightest reason to doubt their good faith, and—whatever we may think of their skill or standards of accuracy—the substance of what they report can confidently be accepted as a record of Anselm's thoughts. Eadmer was active as a reporter mainly in the years before 1100, and Alexander mainly after this date, so their records give a continuous picture of Anselm's mind as it appeared to two close observers throughout his years as archbishop. Eadmer's reports survive partly in his Vita Anselmi and partly in the Scriptum de beatitudine perennis vitae, which is printed below; all that survives of Alexander's writing is in the Dicta
Anselmi and the Miracula, which are here printed as a whole for the first time.
Of Anselm's literary remains which form the remainder of this volume, it is impossible to speak in general terms. Each item must be considered on its merits. But it can be said that drafts and fragments of Anselm's work, whether written at his dictation or copied from his own notes, were sought with great eagerness by some of his disciples. Anselm him self refers to this snatching of unfinished works from his pen in his preface in the Cur Deus Homo. The process can be illustrated in the defective texts of the De Incarnatione Verbi and in the logical notes which are preserved in Lambeth MS. 59.1 This interest in even the smallest fragments of his work survived his death and was responsible for the preservation of several of the texts which we print. Naturally, however, that which is preserved without the sanction or supervision of the author cannot claim the same status as works completed in his lifetime and circulated on his own initiative. Works which are preserved in an incomplete form by disciples may be added to, and words which were remembered may be the basis of developments which can scarcely claim to represent the thoughts of the master. In this way we begin the steep descent from genuine records to more or less appropriate em broidery, to imitation, and finally to the completely spurious. Nothing that we print here is pure and perfect Anselm: it is Anselm incomplete, or as others heard or thought they heard him. A number of doubts
1 F. S. Schmitt, Ein neues, unvollendetes Werk des hi. Anselm von Canterbury, BGPTM xxxiii. 3, 1936, and below, p. 333.
INTRODUCTION 3
therefore hang over these texts which no amount of discussion can wholly dissipate. It is the purpose of the detailed discussions which follow to define the nature of these doubts, and to indicate the extent to which these texts preserve an accurate record of Anselm's words and thoughts, and extend our knowledge of his teaching and personality.
LIBER ANSELMI ARCHIEPISCOPI
DE HUMANIS MORIBUS
PER SIMILITUDINES
THE WORK This treatise can best be described as a systematic treatment of virtues and vices within the context of the religious life. The first half (cc. 1-71) consists of two long sections mainly analytic in character. Chapters 1 to 46 deal with the Will and give an elaborate account of the three forms of self-will—delectatio, exaltatio, and curiositas—which are analysed in intricate detail (1-36) and illustrated in a short series of allegories (3746). This leads to an account of the fourteen rewards and punishments of the obedient and disobedient Will (48-71), very similar in doctrine to the sermon De xiv beatitudinibus reported by both Eadmer and Alexander. But it is noticeable that whereas the sermon was essentially about the joys of heaven, there is here much greater emphasis on the punishments of hell, and on the intermediate condition of men in this life. These systematic expositions are followed by a series of allegories occupying the remainder of the work (72-146) which enforce and expand the doctrines of the early chapters. These allegories may broadly be divided into two parts: cc. 72-97 are chiefly concerned with the virtues and vices within the monastic life; cc. 98-146 discuss and illustrate religious virtues and vices and the gifts of the Holy Spirit in a broader context. Naturally the allegories of the second half of the work do not develop a logical argument. They illustrate some of the main aspects of a few chosen themes. But the unity of thought and the originality of illustration are very striking. Although we cannot compare this treatise with the Secunda Pars of the Summa Theologica of Thomas Aquinas in completeness and systematic rigour, in its own way it is remarkably complete. Without much exaggeration it may be called the first attempt at a systematic study of the psychology of the religious life.
TITLE The wide variation between the early manuscripts indicates that there was at first no fixed title. But in the last chapter there is a phrase —'Haec
INTRODUCTION
de humanis moribus per similitudines suo interiectas loco diximus'— which seems to be the foundation of nearly all the titles later attached to the work. 1 his phrase gives the best possible definition of the scope of the work, and we have adopted it in a slightly modified form. Among the varieties of title in the early manuscripts the superscription of F deserves special note: 'Liber Anselmi archiepiscopi Cantuariensis de morum qualitate per exemplorum coaptationem.' But this and similar phrases proved to be too clumsy for general acceptance and gradually the simple, though rather misleading, title De similitudinibus became almost uni versal. By the time that this title became general, however, the work had been greatly expanded, and the original text had been lost to view.
Hence the title De similitudinibus has been generally accepted as belong ing to the expanded version of the text. We have preserved this con vention and have reserved the title De moribus for the primitive text which is printed here for the first time, while keeping the title De similitudinibus for the larger and more popular version of the text.
THE DIVISIONS OF THE ORIGINAL TEXT
The manuscripts of the work in its original form show a close agree ment in dividing the work by means of a large capital letter at forty-five places. Nearly all these divisions correspond to the beginning of a new similitude, and apart from these obvious breaks there was little attempt to divide the work into sections. These breaks in the text seem to belong to the original state of the material, but the same cannot be said of any of the rubrics. All the manuscripts have some rubrics in the text, but there is almost no measure of agreement between them—certainly none that cannot be explained by later interpolation. The rubrics of the best of our manuscripts (F) and its copy (F2) will serve to show the lack of system in the earliest rubrics and the absence of agreement between the manuscripts. In F there are only nine rubrics within the text, at cc. 21,
26, 37, 38, 83, 84, 90a, 96, no; but they agree only twice with the rubrics of the nearest allied manuscript (D) and only once with any other manuscript: c. 21 De Exaltatione FD c. 26 De Curiositate FD c. 37 De Propria Voluntate F Item tractatus Anselmi cantuariae archi episcopi de propria voluntate P There seems therefore to have been very little attempt at systematic editing of the work in its earlier form. This omission, as we shall see, was soon remedied, but not until the work was undergoing substantial alteration.
INTRODUCTION
THE AUTHOR
The only author mentioned in any manuscript, and in virtually all the manuscripts from the beginning, is Anselm himself. For reasons which are obvious, this attribution cannot be accepted sans phrase: the work is not mentioned by his biographer, it is not included in the manu scripts of Anselm's works made at Canterbury and elsewhere during his lifetime or immediately after his death; in many places the style is rough, and, at least in comparison with Anselm's early works, unadorned and sometimes faulty. One possible way of meeting these difficulties is to ascribe the work to a reporter of Anselm's words, like Eadmer or Alexander. Alexander's own claims have been canvassed more than once, partly because of his own assertion that some of the material which he had collected had been copied and stolen by others, and partly because the revised version of the work, which alone has been printed until now, contains considerable portions of Alexander's Dicta. But now that the original text has been restored, his claim can no longer stand: the differences in style and matter between Alexander's Dicta and the De moribus are sufficient to eliminate the possibility that a single author was responsible for both works.
If the De moribus was put together after Anselm's death by some reporter of the archbishop's discourses, in much the same way as the Dicta, then the name of the reporter is unknown to us. This would be surprising, for it is highly unlikely that anyone who w'as not in close touch with Anselm for a considerable time could have gathered enough material from his talk to put together this work, and there is good reason to think that we know the names of all Anselm's close monastic com panions during the last sixteen years of his life. Yet it seems that none of them could have been responsible for the present work: Eadmer and Alexander are excluded by the style and character of their known works and the absence of the slightest connexion between these works and the De moribus; Baldwin was clearly uninterested in speculation and probably quite unliterary; Eustace, who was a monk of Bee and not a constant companion, scarcely had the necessary opportunities.
The difficulty of naming a possible reporter of Anselm's sayings is not of course a conclusive obstacle to the view that such a man may have existed. But there is another difficulty which must be mentioned. Any one who could put together this work from notes of Anselm's talk in the monastic chapter, at table, in sermons, and the like, must have had not only very unusual opportunities for listening and recording, but also unusual powers of organizing his material. Both Eadmer and Alexander
INTRODUCTION
report discourses, sometimes of considerable length, but neither of them attempted to put them together as a single coherent whole. Eadmer gave his reports a biographical setting. Alexander gave his material a logical continuity so far as this was possible, but the results show how little could be done in this way (see below, pp. 23-24). When all allowances have been made for the gaps in the present work and the necessary lack of system in its allegories, we may still doubt whether such a feat of organization would have been possible without guiding lines provided by Anselm himself. The present work after all is a treatise and not a series of Dicta.
This brings us to the possibility that the work was not formed by a reporter putting together his own notes after Anselm's death, but from drafts of a work of Anselm left unfinished when he died. As we have observed, there is other evidence for the existence of such drafts, and though they are not on the scale of the present work, they have some similar characteristics. These fragmentary drafts generally appear with out any title or clear subdivisions; and though in many ways clearly unfinished, they develop in great detail some of the lines of thought initiated in Anselm's completed works. These are the characteristics of the present work. In several places there is a more developed exposi tion than we find elsewhere of arguments which Anselm is known to have used. Yet there are many signs of incompleteness. Originally there was no title and little or no division into chapters; there are many infelicities of style combined with vigour and precision of language.
There are many signs of hasty composition, and a few errors in the manuscripts which can only be explained as the legacy of uncorrected dictation (e.g. pp. 56. 30; 83. 7; 86. 19-28). There is no trace, at least in the original text, of the smoothing hand of a reviser; and in one place
(c. 31) the argument badly needs a few sentences of elucidation which it would not have been difficult to supply. These characteristics do not suggest that a reporter was at work in the way that Alexander worked on his Dicta or Eadmer on his De beatitudine. Both the style of the work and the peculiar features of its manuscript tradition seem more consistent with an unfinished draft than with a series of reports of Anselm's words put together by another hand.
We are left with a choice between two broadly divergent views of the origin and immediate authorship of the work. Either it is the work of a reporter—and there are difficulties in this view; or it is basically the work of Anselm himself—and here too there are difficulties. On the latter view it seems puzzling that Anselm's known disciples at Canter bury did not make any use of, or apparently attach any importance to,
INTRODUCTION
a work of such considerable bulk and originality. No completely satis fying answer can be given to this puzzle. All we can say is that frag mentary and unfinished works seem easily to have been lost sight of, and the manuscript tradition suggests that other unfinished works of Anselm became dispersed about the time of his death. The significance of this fact will become clearer when we discuss the curious manuscript tradition of the De moribus.
In indicating these two different views and the difficulties attached to each of them, it should be added that there is a middle way between the two extremes. Many of the difficulties we have noted would be explained if the treatise came not from a reporter of Anselm's casual talk and sermons, but from the hand of an amanuensis, working under Anselm's direction and dictation, though without his final supervision. Several of the stylistic and verbal errors in the manuscripts seem to arise from just this mode of composition. Moreover, one of the companions of Anselm's later years was in a position to produce a work of this kind.
This was Boso, the monk of Bee who came to England specially to act as an assistant in Anselm's philosophical work in 1094-5 while the Cur
Deus Homo was being composed, and again in his last years from 1106 to 1109. He returned to Bee when Anselm died, and there is no evi dence that he took any part in preparing or disseminating Anselm's work after this date. If we knew what notes he left behind or what happened to them we might have more light on the origin of this w'ork. In the absence of this light, it can only be said that here we have the fullest, and in some cases the most mature, results of Anselm's thought on sub jects which were already occupying his mind when he was prior of Bee.
Eadmer's description of his interests at this early stage of his career may serve as a preface to this work:
Origines insuper et ipsa, ut ita dicam, semina atque radices necne processus omnium virtutum ac vitiorum detegebat, et quemadmodum vel haec adipisci vel haec devitari aut devinci possent luce clarius edocebat.1
This passage precisely defines the scope of the De moribus.
COMPARISON WITH OTHER EVIDENCE FOR ANSELM'S THOUGHT
At a number of points the De moribus deals with subjects discussed in various works of Anselm, in reports of his words by Alexan der in the Dicta Anselmi, and by Eadmer in his Vita Anselmi and De
1 Vita Anselmi, i„ 8.
INTRODUCTION 9
beatitudine perennis vitae. The main passages for comparison are the following:
Chapter i. See De concordia praescientiae et praedestinationis et gratiae dei) iiiII_I3 (e<^- Schmitt, ii. 278-87), written in 1107-8; and cf. De libertate arbitrii, vii (Schmitt, i. 218-19) and De conceptu virginali
(Schmitt, ii. 143).
This comparison shows that the early sections of the De moribus are an elaboration of an analysis of the will already proposed in Anselm's earlier w7orks, and especially in the last chapters of the last work which he completed.
Chapters 21-25. See Epistola 285 [iii. 75], first part, written in 1103; and cf. Dicta, c. 2.
The letter of 1103 distinguishes three simple modes of pride (in aestimatione, in voluntate, in opere) which in all their combinations pro duce a total of seven possible modes. The De moribus adds to these simple modes a fourth, in locutione, so that the number of possible com binations has now risen to fifteen. The addition of the fourth simple mode is a necessary refinement which fills a gap in the analysis of 1103, and it must therefore be considered an advance on the scheme proposed in Epistola 285. In Dicta, c. 2, Alexander reports an analysis on the same lines as the letter.
Chapters 48-71. See Proslogion, xxv (Schmitt, i. 118-20), written in 1078; and cf. De beatitudine (below, pp. 273-91) and Dicta, c. 5.
These chapters describe the fourteen joys of heaven and pains of hell, and there is a wealth of supporting evidence about Anselm's thoughts on this subject in the passages listed above. The Proslogion is at least twenty years earlier than any of the other sources and it is sig nificantly less logical in its arrangement: by contrast the other sources agree among themselves with some minor differences. The fourteen joys are divided into joys of the spirit and joys of the body. All the sources agree on the first group, but the De moribus and the first recension of the Dicta have a slightly different order from the Proslogion and the De beatitudine. It is a small point, but it is probably not acci dental.1 In the second group, which comprises the joys of the body, the differences between the Proslogion and the other sources are more
1 The joys of the soul in Proslogion and De beatitudine are sapientia, amicitia, con cordia, potestas, honor, securitas, gaudium: in the De moribus and the first recension of the Dicta the order of potestas and honor is reversed: later Alexander corrected the Dicta on the basis of the De beatitudine and adopted the other order (see below,
PP137-9)-
INTRODUCTION
evident and the superiority of the later arrangement could easily be demonstrated:
De beatitudine, Dicta, and De moribus Proslogion 1. pulchritudo 2. velocitas (agilitas De moribus) 3. fortitudo 4. libertas 7. diurnitas (longaevitas De moribus) 5. sanitas i. pulchritudo velocitas fortitudo libertas corporis 3. longa salubris 4. satietas 5. ebrietas 6. melodia 7. voluptas I vita 6. ( voluptas To complete the comparison it may be added that the Proslogion does not mention the miseries of the damned which are opposed to these joys. In Eadmer's De beatitudine and Alexander s Dicta these opposites are briefly described at the end of the sermon; in the De monbus the) are described step bv step as a contrast to each joy, and their discussion forms an integral part of the whole account. A comparison with the two accounts of Eadmer and Alexander will show that De moribus is here the more developed work. The De moribus alone, in conformity with its general purpose, adds to these extremes an account of the inter mediate states which are experienced in this life.
Chapters S2-S3. See Epistola 121 [ii. 29] (Schmitt, iii. 261,11. 34-42), probably written about 1086.
The thought expressed in this letter is the same as that which is illustrated in the similitudes in these chapters.
Chapters 100-8. See Epistola 2S5 [iii. 75], second part, written 1103; and cf. Dicta, c. 1.
The letter of 1103 describes three membra humilitatis. In the De moribus these are described much more elaborately as seven steps of humility, and these correspond to the seven steps of humility in Dicta, c. 1.
Since the account in Epistola 2S5 is compressed into a single sentence no clear comparison can be made, but in a general way there is a cor relation between the first membrum humilitatis and step 1, the second membrum and steps 5, 6, and 7, and the third membrum and steps 2,3, and 4.
Chapter 122. See De conceptu virginali et de originali peccato, c. iv.
(Schmitt, ii. 143-5), written 1099-1100.
In this chapter of the De conceptu there is not only a brief mention of
INTRODUCTION
t he affectiones and usus of the instrumentum volendi which is the startingpoint of the De moribus, hut also a passage (Schmitt, ii. 144, 11. 4-21) which is closely related to the doctrine developed in De moribus, c. 122.
Although these parallels cover only a small part of the whole bulk of De moribus they are sufficient to give the work a distinct place in the development of Anselm's thought, especially in his last years.
I o these parallels should he added the following passages in Eadmer's Vita Anselmi and be beatitudine and in Alexander's Dicta, in these parallels there is no trace of direct borrowing, but the thoughts are the same, and they provide further guarantees of the Anselmian origin of various parts of the be moribus: cc. 39 80: see Vita Anselmi, ii. 21; Dicta, c. 10 c. 42: see Vita Anselmi, ii. 11; Dicta, c. 11 cc. 4345: sec Dicta, c. 4; De beat. p. 287 c. 78: see De beat. p. 280 c. 89: see Vila Anselmi, ii. 11 cc. no-11, 118: see Dicta, c. 16 c. 123: see Vita Anselmi, i. 31 c. 124: see Dicta, c. 13
DATE AND PLACE OF ORIGIN
The material contained in the be moribus must have been in existence, probably at Canterbury, at the time of Anselm's death in 1109. The earliest manuscripts F and D are not later than about 1130. It is certainly strange to find that they come not from Canterbury, but from the Augustinian priory of Llanthony on the Welsh border. Anselm had been closely associated with the foundation of this house, but there is no evi dence that any early canon of Llanthony was among his pupils. Never theless, the strong evidence that the dissemination of the work took place from Llanthony or some neighbouring foundation can scarcely be dis regarded. There are only six manuscripts of the work in its original state: the two earliest (FD) come from Llanthony, the third (F2) is a copy of one of these, the fourth is now in Hereford Cathedral Library and was probably always there, the fifth belonged to Gloucester Abbey and is a copy of the fourth, and the sixth, which belonged to Saint-Germain-desPres but was perhaps written in England, is closely related to F and D.
Both Hereford and Gloucester were closely connected with Llanthony.
The second prior of Llanthony, Robert of Bethune, was bishop of Here ford from 1131 to 1148, and the canons lived with him for a year in 1136 while their new priory was being built; and this new house, Llanthonia secunda, was just outside Gloucester. So the links with this part of
INTRODUCTION
England are impressively consistent. By contrast there is no evidence that Christ Church, Canterbury, ever had a manuscript of the De moribus d certainly no existing manuscript comes from this most obvious home of Anselm's works. In addition to this there is the curious fact that, although Eadmer and Alexander knew and used each other's works, they show no sign of knowing the De moribus, nor does the De moribus in its original form show any knowledge of their works.
It would be a mistake to attach too much significance to these facts, but they seem to suggest that at some date very close to the time of Anselm's death the materials which made up the De moribus were no longer at Canterbury. It may be significant that the earliest copy of the De custodia interioris hominis ascribed to Anselm, and an otherwise unknown early draft of the De Incarnatione Verbi, are both found in the Hereford manuscript which contains the De moribus.1 2
THE REVISED AND ENLARGED TEXT: DE SIMILITUDINIBUS^
Shortly after the completion of the De moribus, a second and much enlarged recension was made by incorporating suitable material from Alexander's Dicta and from Eadmer's Vita Anselmi and De beatitudine perennis vitae. The revision also introduced a large number of small stylistic alterations and divided the work into 192 short chapters. It was in this form that the work became widely known in the Middle
Ages and was finally printed in the editions of Anselm's works. The effect of the revision was to lengthen the work by about a quarter, partly by adding new chapters and partly by enlarging some of the existing sections. The new chapters were nos. 112-18 and 147-92. Of these,
112-18 and 147-74 came from the Dicta, and 176-80 and 186-92 from the Vita Anselmi. The source of chapters 175 and 181-5 is unknowm, but 181-2 are closely related to the Anselmian miscellany in MS. Bodley 561, and their inclusion indicates that the compiler had access to some fragmentary records of Anselm's teaching. The chapters which were extensively revised and enlarged were 42-45, 47-71, 89, 99-108, 123, and 124. Of these, 42-45, 99-108, and 124 were revised on the basis of Alexander's Dicta; 89 and 123 on the basis of Eadmer's Vita
Anselmi; and 47-71 on the basis of his De beatitudine. Although the reviser omitted some of the material in his sources which might suitably
1 The only item in the medieval catalogue which may be a copy of the De moribus is a work with the title Ex dictis et moribus Anselmi (M. R. James, Ancient Libraries of Canterbury and Dover, no. 67, p. 24). But this is more likely to be the De similitudinibus, which in British Museum MS. Royal 5 E xiv has the title Libellus de moribus factus ex dictis Anselmi archiepiscopi.
2 Schmitt, ii. 2-28.
3 P.L. 159, 605-708.