The Book of Memory

by Mary Carruthers

  Diagrams

  A prominent feature of medieval book art is the use of diagrams. The

  Canon Tables, written intercolumnia – between arched columns that are

  decorated with birds and floral motifs – are one very early example, which

  we find in books from the earliest medieval centuries (figure 6). The typical

  lay-out of the Calendar, marked by images of the Zodiac and (later) the

  labors of the months, is another. In fact, one of the more intriguing aspects

  Memory and the book

  325

  of Hugh of St. Victor’s Ark-picture is its gathering of many common

  medieval diagrams into one place, and, if one adds the columnar format he

  employed for his Chronicle, it is clear that Hugh was accustomed, both in

  teaching (lectio) and in his own meditational composition, to using vir-

  tually every major genre of diagram common in the twelfth century –

  ladders, trees, circles, columns, maps, and genealogical charts – all enclosed

  within the rectangular shape of the memorial page. Even more interesting

  is Hugh’s assumption that basic pedagogy consists in the ability to use such

  structures readily to perform tasks of mnemonic calculation (collating

  texts, for example). His own practice indicates that such recollective skill

  is, in large measure, what he is referring to in his aphorism, ‘‘the whole

  usefulness of education consists only in the memory of it’’ (‘‘In sola enim

  memoria omnis utilitas doctrinae consistit’’).

  In many ways, Hugh of St. Victor is a twelfth-century version of

  Quintilian, not an innovator of technique so much as an admirably clear,

  practical guide to the best pedagogy of his time. This pedagogy employed

  visual aids and diagrams to an extent that many modern historians have

  noted. For example, in his descriptive catalogue of English Romanesque

  illuminated manuscripts, C. M. Kauffmann writes that ‘‘the pictorial dia-

  gram . . . became one of the typical features of Romanesque art.’’79 But

  most of these diagrams had been in common use for centuries before the

  twelfth, as mnemonic compositional aids – one thinks, for example, of

  the ladder with which Benedict organized the steps (gradus) of humility in

  the seventh chapter of his Rule.

  Indeed, it is a much-remarked-on medieval characteristic to treat the

  space in a full-page drawing diagramatically, that is, with images placed in

  specific locations, often grouped about a large central figure, often in an

  architectural setting, often with related images enclosed in roundels or

  other geometric forms, usually with a border, and commonly with inscrip-

  tions, like tituli or rubrics, to be associated with the figure and to help

  associate the figures with one another. The justification for this practice is

  mnemonic necessity. The framework of the page provides a set of orderly

  loci; furthermore, this frame remains constant while the images in it change

  from page to page – that is the manner of a diagram, and it is also the

  manner of the page of memory, imagines rerum imposed upon a set of

  geometrically defined places in an orderly framework or grid.

  Examples of this format span the whole period that we call the Middle

  Ages. The Gospels of St. Augustine (Cambridge, Corpus Christi College

  MS. 286), made in Italy or Gaul in the sixth century, was at Canterbury

  during the Middle Ages, and is thought to have been brought to England

  326

  The Book of Memory

  29. Cambridge, Corpus Christi College MS. 286, fo. 125r (‘‘The Gospels of St. Augustine’’;

  Italy or Gaul, sixth century). A picture-page, showing a grid format and frontal perspective

  towards figures set against a plain background. Compare the instructions for forming

  memory locations in Thomas Bradwardine’s De memoria artificiali.

  from Rome by St. Augustine himself. A full-page picture (figure 29) groups

  figural scenes of the Passion week, many containing textual tags, into a grid

  of twelve cells, each set in a border. The whole rectangular construction is

  enclosed in another border painted around the vellum page. The outer

  Memory and the book

  327

  border demarcates the area or flat surface of the page, just as the memory-

  page in Hugh of St. Victor’s Ark pictura has a border enclosing it. The loci

  are small quadratures within this, three across in each of four tiers, clearly

  separated from one another by a red-colored border. Some of the individ-

  ual imagines are given a rudimentary background; others are set in what has

  been called ‘‘hierarchical’’ space, wherein the figures are grouped in such a

  way that the main figure, the first in the order, appears in the center and

  larger than the subsidiary figures grouped to its right and left.

  The background is simple, a single color, providing no more than the

  barest suggestion of a setting for the figures. And the perspective of

  the viewer is from the front and central, the qualities that define distancia

  in the mnemonic advice that has survived from later centuries. In fact, this

  page exemplifies so many of the rules for making mnemonic locations

  described by Bradwardine that one might, at first, wonder if he must have

  seen it; the manuscript was, after all, at Canterbury. Yet the details of

  Bradwardine’s examples are not those of the drawings in the Gospels of

  St. Augustine. Their similarity in principle, however, suggests strongly that

  the same general rules that governed the making of the sixth-century

  Gospels’ page were still a part of medieval mnemonic pedagogy in the

  fourteenth century.

  The picture-page in this Gospels provides a complete set, in order, of

  imagines rerum for the Passion story, drawn not from a particular gospel

  but synthesizing all four accounts. The picture-page comes between the

  gospels of Mark and Luke; it is thought that similar pages, with images

  from the earlier life of Christ, were placed before Matthew and between

  Matthew and Mark, and that a page showing events from the Passion to the

  Ascension preceded John. Such a placement is usual in early medieval

  Gospels, which were large books designed for study within a monastery

  community.80

  Picture-pages in medieval books had cycles of popularity. A much later

  example is a Flemish Psalter of the second half of the thirteenth century (The

  Morgan Library MS. M.183), which contains four such pages. M. R. James’s

  catalogue description of this manuscript demonstrates clearly how the

  individual subjects of the images are inserted into a framework that is

  stable in all four of these pages – stable not only in shape and configuration,

  but in the type of content that occupies the different parts of the frame.81

  Again, the places within the frame are separated clearly from one another

  by borders; the imagines rerum refer to sixteen main events in the life of

  Christ and to various saints. This framework is quite elaborate, though its

  relationship to the common grid-format is evident.

  328

  The Book of Memory

  Such imagines rerum, either painted only mentally or also on parchment,

  share many features with the teaching diagrams
which the later Middle

  Ages produced in abundance. Diagrams were a common feature of medi-

  eval classrooms. One of the better-known, and most widely used, was in a

  tree form, composed by Peter of Poitiers towards the end of the twelfth

  century. It is the genealogy of Christ, called Compendium historiae in

  genealogia Christi or, simply, the Genealogia, and it is quite like the linea

  genealogia we encounter in Hugh of St. Victor’s De formatione arche,

  composed a half-century earlier.

  Beginning with Adam, the names of Christ’s predecessors are listed in

  order, with a brief biographical sketch. The names and their relationships

  are indicated by lines and, usually, circles, with pictures of the chief events,

  such as the Temptation of Adam and Eve and the Flood – often also

  the diagram known as the Tree of Jesse is included. A version of the

  diagram seems to have had an independent existence by the early thirteenth

  century. An early thirteenth-century chronicler reports that Peter of

  Poitiers had ‘‘designed for the benefit of poor clerics trees painted on

  skins that displayed the history of the Old Testament.’’82 Much copied,

  often incorporated into universal chronicles from the twelfth through

  the fifteenth centuries; the Genealogia was also translated into several

  vernaculars, including English, and had an exceptionally long run in the

  classroom.

  Frequently, the Genealogia was written on a continuous roll of parch-

  ment rather than in a codex. Three examples in the Bodleian Library are of

  some interest. An English codex of the second quarter of the thirteenth

  century (MS. Laud Misc. 151), which contains both this text, in Latin, and

  Peter Comestor’s Historia scholastica, belonged to Cardinal Morton

  (d. 1500) at one point in its history, and the first leaf pictures both his

  coat of arms and a rebus for his name, a drawing of a cask, or ‘‘tun,’’ with

  the letters ‘‘mor’’ written on it. A roll (MS. Barlow 53), dating from the

  early fifteenth century, is an English translation of the Genealogia, with its

  preface. And finally, another roll, MS. Lat. th. b. 1, also English and from

  the first half of the thirteenth century, is especially interesting because its

  companion roll (MS. Lat. th. c. 2), which may have been attached to it at

  one point, contains diagrams of the Seven Deadly Sins, the Seven Gifts of

  the Spirit, and so on. The genealogy roll begins with a drawing and

  exposition of the Menorah or seven-branched candlestick, proceeding to

  enumerate a number of items in sevens and also in threes before it gets to

  Peter’s diagram itself. Such a digest of preaching aids would be a useful tool

  for classroom lecture as well as for sermon making. 83

  Memory and the book

  329

  Peter of Poitier’s preface makes it clear that his format is a memory aid.

  Considering the length of holy writ, he says, he has digested the genea-

  logical information and presented it in this form so that students may

  retain it in memory, as in a money-bag (quasi in sacculo). He uses the same

  image for memory used by Hugh of St. Victor in his Chronicle Preface, that

  of a sacculus or compartmentalized money-pouch, in which coins of various

  sizes can be carried, each in its place without loss or confusion. As I showed

  in my first chapter, this image is a common variant of the archetype of

  memory as thesaurus or arca. And Peter goes on to say that his scheme,

  though useful for all, is particularly so for those students who feel over-

  whelmed by the bulk of historical material; with the aid of his diagram, that

  daunting prolixity, having been made subject to the eyes, can be commit-

  ted to memory and its usefulness realized.84

  In his section on rhetorical memory, Geoffrey of Vinsauf (who was

  Peter’s contemporary) emphasizes the mnemonic value of formae: ‘ loca,

  tempora, et formae and other similar notulae are sure ways to lead me to

  [what I have stored away].’’85 He does not define what these formae are. The

  word can mean ‘‘shape’’ or ‘‘outward appearance.’’ But it can also mean

  ‘‘geometric shape’’ or ‘‘outline’’ – the sort of thing that a diagram is,86 and

  given the fondness of the twelfth-century classroom for diagrams, this was

  what Geoffrey probably meant. Diagrams formed of concentric circles,

  rose diagrams (which are not confined to windows), the primitive chart

  that forms the Genealogia, Hugh of St. Victor’s columns of persons, places,

  and dates for his Chronicle, ladder diagrams, and the veritable forest of trees

  used to illustrate the relations of parts of a subject whether it be preaching

  or vices and virtues – all such formae are common features in manuscripts

  too, especially in the twelfth and thirteenth centuries.87

  There is one more famous diagram worth considering here, described by

  John of Garland. I discussed in Chapter 3 John’s instructions for compos-

  ing a three-columned memory tablet that was to include all the voces

  animantium (chiefly animal names), words in various languages, and so

  on, in alphabetical order; some manuscripts of John’s text include a partial

  schematic drawing of this, but it was clearly designed as a mental structure,

  infinitely expandable as such schemes are. John also mentions a diagram

  ‘‘quam pre manibus habimus,’’ ‘‘which we have in our hands,’’ the so-called

  Rota Virgili or Virgil’s Wheel. 88 This circular diagram is also drawn in

  some manuscripts of the Parisiana Poetria, and it shows a series of con-

  centric circles divided into three wedges, within which comparisons, like-

  nesses, occupations, animals, plants, and implements are written that

  pertain to John’s subject, the three different kinds of audience to be

  330

  The Book of Memory

  considered in determining stylistic decorum. Virgil’s Wheel was clearly a

  mnemonic diagram that his students held; it is likely that it could be

  physically manipulated, as its concentric circles suggest.

  This figure of the rhetorical Rota Virgili may provide the connection

  between the Latin word rota, ‘‘wheel,’’ and the English phrase ‘ by rote.’’ Rote

  is ‘ of obscure origin,’’ according to OED, although it appears as though it

  should be derived either from Latin rota or from French rote, ‘ way, route.’ As

  OED states, however, there has been ‘ no evidence to confirm the sugges-

  tions.’ 89 My review of the evidence does not confirm a derivation either, but

  it may help to strengthen the suggestion in favor of the Latin over the French.

  ‘ By rote’’ appears in English in the fourteenth century meaning both

  ‘ reciting (prayers, speeches, and the like) from memory’’ (an essentially neutral

  usage) and, in a pejorative way, ‘ reciting unintelligently or by formula.’’ For

  Chaucer, the word on its face is neutral, meaning ‘‘from memory exactly,’’ a

  synonym of verbatim memory. The first clearly pejorative uses of the word

  (which are from the fifteenth and sixteenth centuries) modify ‘‘rote’’ with

  ‘‘mere’’ or ‘‘pure.’’ ‘‘Rote’’ had also a generalized meaning of habit or
<
br />   custom, that which memory produces. The earliest OED citation, in a

  poem by William of Shoreham written probably in the 1330s, refers to a bad

  habit as a ‘‘wikked rote’’; habitual knowledge is what the classroom rotae

  were designed to instill. John Gower speaks of the ancient pagans’ ‘‘rote’’

  (custom) in his Confessio amantis (VI, 1311), interestingly in the context of

  knowing magical arts and sorcery (things that scholars, whose memories

  were sure to be trained, would dabble in).

  The clearest link between English ‘‘by rote’’ and the rhetorical tradition

  which included using the Rota Virgili diagram is in Chaucer’s Pardoner’s

  Tale, when the Pardoner describes his rhetorical tactics and how he adopts

  his speech to his audience:

  Lordynges, quod he, in chirches whan I preche,

  I peyne me to han an hauteyn speche

  And rynge it out as round as gooth a belle,

  For I kan al by rote that I telle.

  My theme is alwey oon, and evere was:

  Radix malorum est cupiditas.

  (VI, 329–334)90

  The Rota Virgili showed schematically how to adopt one’s speech to partic-

  ular kinds of audiences, something at which the Pardoner is especially adept.

  Notice how he distinguishes between his theme, which is always the same,

  and his ability to vary it to suit the particular occasion and congregation, ex

  tempore dicendi, a boast he repeats in lines 412–420 when he describes how

  Memory and the book

  331

  he has insulted particular enemies in his sermons. I have hardly settled the

  etymological puzzle here, but the possibility of a link with memory devices

  like the Rota Virgili is suggestive. It is also easy to see how, if it referred to a

  mnemonic schoolroom diagram, ‘‘speaking by rote’’ could come to mean

  ‘‘speaking by mere formula.’’

  Wheel-diagrams were not limited to the teaching of rhetoric. Wheels

  that move and are intended to be manipulated are common enough

  artifacts in manuscripts. For example, two moving wheels are found in

  Bodleian Library MS. Bodley 177, a collection of scientific pieces made up

  of four separate manuscripts written in the last part of the fourteenth

  century. On fo. 63v there is a simplified astrolabe, made of two freely

  moving concentric parchment disks, and on fo. 62r a wheel that reveals the