The Book of Memory

by Mary Carruthers

  managed to break loose from the lawsuit image to career dangerously around

  on its own in the dark?’’ She suggests perhaps the connection had something

  to do with the Zodiac, or perhaps was just the result of too many lonely

  nights of study. But the explanation is more prosaic. The testicles are in the

  ancient source, of course, but the charging ram bearing them was added in

  the earliest known commentary/gloss on it, Etsi cum Tullius, which is most

  probably the work of William of Champeaux, the teacher of Abelard and

  founder of the school of St. Victor in Paris; thus it dates from the early twelfth

  century.37 This aggressive ram with large testicles – both a pun on testes,

  ‘ witnesses,’’ and a sign for the adversarial nature of the legal proceeding –

  features as well in the other twelfth-century commentaries, and had become

  the standard academic gloss on this passage by the time Albertus wrote.38

  Before proceeding to discuss the circumstances within which the

  Herennian mnemonic was revived, let me summarize the features which

  The arts of memory

  179

  seem to be distinctively medieval in the memory advice we have examined

  in John of Garland, Thomas Bradwardine, and Albertus Magnus. With

  respect to the memory locus, these are: (1) a plain-colored or simply

  patterned background, with just enough suggestion of depth to be able

  to position the images in relation to one another in a scene; (2) the observer

  positioned frontally in respect to this scene, at a distance from which the

  whole composition can be taken in completely at a glance; (3) a clear and

  consistent arrangement of the images, which enables them to be read off in

  sequence – this sequence can be from the center to either side, or, in a

  circular arrangement, clockwise or counterclockwise (we will see examples

  of these in Chapter 7). With respect to the imagines, they are: (1) their

  grouping in a scene in which the order among them is expressed through

  physical action; (2) the use of vivid, unusual, and extreme images; (3) the

  use of images from a variety of sources, including the Bestiary, the

  Calendar, and all other sorts of painted or sculpted forms, such as those

  found commonly in churches and in books.

  The placement of the memory locations is diagrammatic, frequently

  within a grid. One of the most popular of late medieval ethical manuals was

  an allegorical treatment of the game of chess, composed by the Dominican

  friar, Jacopo da Cessola, around 1300. It was one of the works printed by

  William Caxton. In an excellent essay on this work, Raymond DiLorenzo

  commented that it uses mnemonic technique to integrate the ethical

  material which the friar has composed as a set of memorabilia, ‘‘things-

  to-be-remembered.’’ The mnemotechnique used is the chessboard, a grid,

  into which imagines (the chess pieces, described with vivid and unusual

  detail) are fitted. Jacopo prefaces the work, which is basically a florilegium,

  by saying that chess was invented by a philosopher who sought to correct a

  tyrannical king. As they played the game, the philosopher instructed the

  king in the virtues and vices that attached to each piece. Thus the game

  itself became for the king a mnemonic of kingly virtue and responsibility, a

  Rule for Princes presented in a form that embeds its own mnemonic – the

  form of a grid filled with images, familiar to medieval audiences as a basic

  format for the page of memory. 39

  The earliest medieval artes memorativae belong to the thirteenth century –

  or, more precisely, the earliest written medieval artes belong to this period.

  In Alcuin’s dialogue with Charlemagne on rhetoric (ninth century), the

  king asks his teacher, Alcuin, if there are any specific precepts for memoria

  (trained memory), which is the noblest aspect of rhetoric (‘‘nobilissima . . .

  rhetoricae parte’ ). There are none, replies Alcuin, except disciplined

  exercise in memorizing (‘‘ediscendi exercitationem’’), practice in writing,

  180

  The Book of Memory

  and the discipline of cogitation or mental composition. One should

  also avoid drunkenness (which is harmful to any discipline of soul or

  body). Memoria is a store-house, custodian of invention and cogitation,

  of ‘‘things’’ and ‘‘words,’’ and without it ‘‘even the most eminent of the

  speaker’s other talents will come to nothing.’’ Charlemagne’s question

  about precepts is another way of asking whether memoria is an art or not.

  Alcuin takes for granted that memoria is a study, involving discipline,

  training, and practice. But he apparently saw no need (at least for

  Charlemagne) for a complex set of related precepts in addition to the

  elementary principles of memory training. He thought of memory as a

  consciously constructed inventory, a library, a store-house of material in

  the form of both ‘‘words’ and ‘‘things.’’ But his advice to the emperor

  does not go farther, to suggest a systematic art of memory. 40

  This bias, if that is what it is, in favor of teaching rhetoric students a few

  mnemonic principles instead of a systematized art of memory can also be

  found in Quintilian and in the late Roman rhetoricians of the fourth and

  later centuries. In De oratore, Cicero speaks of an architectural mnemonic

  art as so well-known that it needs no elaborate description.41 But, 150 years

  or so later, Quintilian, while not rejecting totally the notion of making an

  art of memory, plays down its utility and that of other such artes. Evidently

  a fashion favoring practice and discipline in the elements of memory

  training over more elaborate, curious (and Greek) artes had begun to

  dominate Roman pedagogy; this bias is consistent with Quintilian’s judg-

  ments against using pre-fabricated mnemonic imagines and all other quick

  prescriptions for memory training. Of course, the Ad Herennium also

  counseled against substituting others’ schemes for one’s own; this is a

  traditional pedagogical caution, the basis of Plato’s warning in Phaedrus

  against substituting textbooks and recipes for teachers and disciplined

  practice. But Quintilian’s bias towards philosophical rhetoric and against

  the sophistical excesses he saw in his contemporaries like Seneca is reflected

  again here in his remarks about the usefulness of artes memorativae. Instead,

  he stresses the elementary aspects of memory training, reserving the various

  arts for it to those of advanced ability – and to teachers of dubious

  reputation (or so he implies).

  The emphasis in the rhetoric schools on the elements rather than on an

  advanced techne of memoria is found still in the fourth century. Julius

  Victor wrote that ‘‘for impressing memory some teach observations of a

  great many places and images, which do not seem effective to me.’’42 Since

  he took a great deal from Quintilian, it is not surprising to find this

  opinion, and it also indicates that during the centuries after Quintilian

  The arts of memory

  181

  wrote, his judgment continued to predominate, despite the fact that

  Cicero himself spoke favorably in De oratore o
f using memory arts based

  on images in places. Julius Victor in turn greatly influenced Isidore of

  Seville and Alcuin. Yet, as we have seen, Julius Victor does give instructions

  in the basic mnemonic preparations: division, memory ‘‘for subjects’’ and

  ‘‘for words,’’ using memorial notae, meditation, and composition. So it

  seems that in later antiquity and the earlier Middle Ages a memorial ars

  such as that taught in the Ad Herennium was regarded in standard peda-

  gogy as marginally helpful at best, and that the focus of memory training

  was on its elements, instilled through practice and discipline, but without

  emphasizing a universal body of principles. To use the traditional cate-

  gories, memoria was taught as a studium, a collection of empirically proven

  guidelines, but it was not a systematic ars, as it apparently had been

  regarded in the time of Cicero and would be again by the end of the

  thirteenth century. The distinction I am making is between what we would

  call a craft, learned by apprenticeship, and a scientific discipline that can be

  taught from a textbook.

  Although arguments from silence are not worth much without positive,

  corroborating evidence, it is interesting that Augustine makes no mention

  of an art of memory, though trained as a teacher of rhetoric and though he

  is one of the great philosophers of memory. Frances Yates attempted to

  make a case for his having been a practitioner of the Herennian art, on the

  basis of his descriptions of memory’s power in Book X of Confessions, of

  which the following is typical:

  See, in the measureless plains and vaults and caves of my memory, immeasurably

  full of countless kinds of things which are there either through their images (as

  with material things), or by being themselves present (as is the knowledge acquired

  through a liberal education), or by registering themselves and making their mark

  in some indefinable way (as with emotional states which the memory retains even

  when the mind is not actually experiencing them, although whatever is in the

  memory must be in the mind too) – in this wide land I am . . . free to run and fly to

  and fro, to penetrate as deeply as I can, to collide with no boundary anywhere. So

  great is the faculty of memory, so great the power of life.43

  The metaphors Augustine uses here are archetypal: one did not need to

  practice the Herennian art to believe that memory was locational in nature,

  like a vast cave with many inner caverns in which all experience was

  inventoried, and that it had immense power which training could enhance.

  Indeed, though copied as one of Cicero’s works, the Rhetorica ad

  Herennium was not much commented on or quoted until late in the

  eleventh century. The earliest ancient references are in scattered sources

  182

  The Book of Memory

  from Jerome (in works from 395 and 402) through Priscian (early sixth

  century). None of these are to the memory section, it should be noted. Its

  influence is not extensive until after the mid twelfth century, and even

  later. 44

  When it was first revived, its memory advice was received with skepti-

  cism. John of Salisbury (who knew Quintilian’s Institutes better than most

  of his contemporaries) alludes to arts of memory: ‘‘Seneca most readily

  promised to teach the art for furnishing memories, of which I certainly

  wish I were a master; but as far as I know, he did not actually teach it.

  Tullius seems to have applied himself to this in [Rhetorica ad Herennium]

  but the latter is not of much help to me.’’45 And Geoffrey of Vinsauf,

  writing at the turn of the thirteenth century, dismisses the Ad Herennium’s

  art completely: ‘‘Tullius relies on unusual images as a technique for training

  the memory, but he is teaching himself; and let the subtle teacher, as it were

  in solitude, address his subtlety to himself alone.’’46

  Instead, Geoffrey counsels the traditional elements of memory training:

  division, placing the pieces in a rigid locational order (one, two, three, etc.),

  marking material with mental notae for secure recollection from the

  storage-bin (cella) of memory, respecting the limits of short-term memory

  (brevitas), and not stuffing too much in at one time, for the store-room of

  memory responds well only while it is having fun, and so must be treated

  well (‘‘Cellula quae meminit est cellula deliciarum’’). The problem with a

  method like that described in the Ad Herennium is that it is gimmicky and

  ready-made, whereas no single group of markers (notae) will work equally

  well for everyone. Everyone should make little markers of his own choosing

  and devising, for those are ‘‘safer,’’ tutior, for recollection. One should note

  the concern Geoffrey expresses for the safety of a recollective scheme –

  Thomas of Waleys has the same concern.47

  One other early medieval writer on memory should be considered here,

  because he is credited by Frances Yates as being the one who ‘‘handed on

  the art [of memory] to the Middle Ages.’’48 This is Martianus Capella, a

  fifth-century teacher whose allegorical encyclopedia on the seven liberal

  arts, The Marriage of Philology and Mercury, became influential after the

  turn of the first millennium. Martianus treats memoria in the traditional

  way, as a part of rhetoric. But he is rarely mentioned as a proponent of an

  art of memory and never, to my knowledge, linked with the precepts of

  Tullius. What Martianus describes under memoria are some of the elemen-

  tary rules of ancient mnemotechnique, but nothing that constitutes an art

  of memory, let alone repeats what is peculiar to the architectural art of the

  Ad Herennium. Indeed, the advice on memory training that is advocated by

  The arts of memory

  183

  Cassiodorus (d. 590), whose Fundamentals of Divine and Secular Letters

  became perhaps the most important academic text for the early Middle

  Ages, is not that of Martianus Capella but of Consultus Fortunatianus (a

  late fourth-century rhetoric master), from whom Martianus Capella cop-

  ied a great deal. As we saw in Chapter 3, Fortunatianus’s advice belongs to

  the elementary pedagogy of how to memorize a text, which one can find as

  well in Quintilian. Like Martianus Capella, Fortunatianus represents the

  mnemotechnical pedagogy which Augustine, Marius Victorinus, and other

  teachers in late antiquity valued.

  Martianus Capella says that while memory is a natural talent, it may

  be assisted by training. The craft of memory has only a few rules and

  requires much practice; but by it words and things can be grasped quickly

  and surely. Memory-for-things is easier and less time-consuming than

  memory-for-words, though the latter may be employed to recall hard

  words or particular words one has trouble remembering. Simonides

  invented the art from his experience when a banqueting hall collapsed,

  and he was able to identify the victims because he had attached their names

  to the places where they were seated. Order is thus the key to the precepts of

  memory training. The order is memorized as a set of distinctive mental

  locations into which
one puts images of things to be recalled – for example,

  if one wants to remember a wedding, one would place the image of a girl in

  a saffron wedding veil; if a murderer, a sword or other weapon. ‘‘For just as

  what is written is contained in wax and letters, so what is committed to

  memory is written into areas as if on wax and in page-form [in cera

  paginaque signatur].’’49 But since much labor is required for remembering,

  we should write down what we wish readily to retain. If the material is long,

  divide it into shorter sections for ease of memorizing. It is useful to mark

  with notae things we want particularly to recall. When memorizing, one

  should not read aloud but meditate upon the material in a low murmur. It

  is best to exercise the memory at night, when silence aids concentration.50

  The ultimate sources for all of this lore are De oratore and Quintilian’s

  Institutes, not the Ad Herennium. Martianus himself provides the particular

  examples. A particular order – this order – is memorized (‘‘is [ordo] . . .

  meditandus est’’) as a set of distinctive, well-known (and well-lighted)

  locations (‘‘in locis illustribus’’). Into these locations the likenesses of

  objects (‘‘species rerum’’) and images of subject matters (‘‘sententiarumque

  imagines’’) are gathered (‘‘collocandae sunt’’). Rather than a procedure

  like that for forming places advised in the architectural mnemonic, what

  Martianus counsels is a prototype of advice like that of Hugh of St. Victor

  (and others, as we shall see) – to make a diagram, linea, in one’s mind that

  184

  The Book of Memory

  acts as a schematic ordering device for material that is then put into it. This

  is the commonest type of medieval description for what one does in

  forming memory places, and it is not the same kind of procedure as that

  described in the Ad Herennium. Bradwardine’s sets of places in fives, one

  on top of another as in a matrix, is more like constructing a diagram than it

  is like walking through a house one knows familiarly. The schematic

  quality is distinctively medieval.

  From the time of Quintilian then, memoria seems to have been com-

  monly taught as an elementary practice, consisting of a few basic rules and

  some common-sense advice, eclectic, empirical, pragmatic, but not a