medieval mnemotechnical principles, not derived directly from the classi-
cal text of Tullius. Albertus wrote about one generation after John of
Garland did, and Bradwardine some ninety years after Albertus.
It is often the confused student who can teach us the most about
unstated general assumptions. One revealing account of memory training
from the early thirteenth century is the garbled adaptation of the
Ad Herennium on memory, found in John of Garland’s Parisiana poetria.
John acquired his education in Paris, and taught there as well, except for
a brief sojourn at Toulouse. He wrote in the 1230s, by which time the
Aristotelian winds of the university were starting to blow upon the
Ad Herennium but had not yet shaped it into respectability. In his treatise,
John recommends as most necessary for poets organizing their material in
invention (composition) the ‘‘ars memorandi secundum Tullium,’’ ‘‘the art
of remembering according to Tullius.’’ But it is evident from what he says
that John of Garland had little of a historian’s concern for what the first-
century BCE Tullius was about. His confusion is instructive, however, for it
is that of a man educated in a different memory keying system, trying to
relate the strange terms of Ad Herennium to those with which he is familiar.
What he was familiar with was the system of keying mnemonically
on locus, tempus, et numerus, as we found those terms used by Hugh of
St. Victor – page design, memory occasion, and an ordering system that
places items to be remembered in a grid design. 6 To this basic knowledge,
acquired as part of his own early grammatical training, John of Garland
tries to add the learned rules of the Herennian mnemonic. The discussion
occurs in Book II, devoted to what John calls selection of readings –
literally, their binding-together (alligare lecta), John’s psychologically
suggestive term for invention. Having discussed what sorts of material to
select and the various tropes and figures (mostly drawn from Geoffrey of
Vinsauf’s Documentum), all in terms of a vaguely defined notion of stylistic
levels, John proceeds to the art of remembering. The discussion is brief
and wholly geared to comprehending what John thinks is being said in
The arts of memory
157
Ad Herennium. After his discussion of memoria, John of Garland proceeds
immediately to give examples of well-composed letters, the subject matter
of ars dictaminis, the ‘‘art of the dictamen’’ (whose intimate basis in the arts
of memoria I will discuss in Chapter 6). Ars dictaminis, and the related ars
notataria or notaria, the ‘‘art of notation’’ or short-hand, in which the
system of Latin abbreviations was taught to students of law and theology,
are both closely associated with discussions of memoria, as these begin
to reappear in handbooks of rhetoric in the later Middle Ages.7 John of
Garland’s placement of his discussion in the Parisiana poetria underscores
one of his basic assumptions, namely that selecting and gathering material
one has read is the heart of successful composition, and that this can occur
only because one has a trained memory.
John of Garland says that the art of remembering is essential for selecting
(‘‘electio’’) and organizing material. So, as Tullius says, we should place
(‘‘disponere’’) in our minds an open space (‘‘area’’), in a place neither too
bright nor too gloomy, because these qualities are harmful to retaining and
selection. The area should be thought of as divided into three principal
parts and columns (‘‘per tres partes principales et columpnas’’). In the first
column, subdivided in three, we place ‘‘courtiers, city dwellers, and pea-
sants,’’ together with their particular concerns, duties, and the things
pertaining to them. If one’s teacher (of rhetoric, in this instance) should
say something in class having to do with these three levels of audience, we
place it in the appropriate section of the first column. In the second column
we mentally mark off (‘‘intelligi distingui’’) our source texts, ‘‘exempla et
dicta et facta [deeds] autentica,’’8 including the source of each, which we
have read in a book or hear in class. To aid in recalling a particular text, we
should also mentally note the circumstances under which we first heard
it, ‘‘the place in which, the teacher from whom, his dress, his gestures,’’ or
the page upon which we read it, whether it was light or dark (referring to
the hair- or skin-side of the parchment), ‘‘the position on the page and the
color of the letters.’’ This advice is noticeably similar to that given 100 years
earlier in Hugh of St. Victor’s Chronicle Preface (see Appendix C).
Finally, in the third column we should imagine written out the sounds
of all sorts of languages and the characteristics of animals (‘‘omnia genera
linguarum sonorum et vocem diuersorum animancium’’) and etymologies,
interpretations, and summary definitions (‘‘ethimologias, interpretationes,
differentias’’).9 These should be in alphabetical order (‘‘secundum ordinem
alphabeti’’). So when the teacher makes an etymological explanation or we
hear a word in a language we do not know, we put it in this column,
together with something that marks it. Word and sign together we gather
158
The Book of Memory
(‘‘collocemus’’) in the third column. Whenever we hear an unfamiliar
word, we relate it by sound to one we know and so, associating one with
the other, store the unfamiliar safely in our memory. The eagle and
summary distinctio shown in figure 4 would seem to derive from such
memory advice as this, though John wrote in Paris and the eagle summary–
picture was drawn in Bologna a decade later, suggesting that John’s
mnemonics were not unique.10
While it is clear that John did not fully understand the Ad Herennium,
he did understand the system he learned himself. And so the Ad
Herennium’s rule of using intercolumnia for backgrounds is understood
as the columnar design of a manuscript page, rather like the design Hugh
devised for his pupils to help them memorize those seventy folio pages of
the most important data for their study of Scripture. There are more
examples of medieval ‘‘misreadings’’ in this text. The Ad Herennium advises
using architectural locations with few people in them, meaning ‘‘open’’ in
that sense, but John freely adapts this rule to his students’ circumstances.
The open area is like a vacant page to be written upon in columnar format,
with subdivisions upon it for the various categories of subject. And on this
vacant page we write in three columns the material we learn: first, what is
relevant to decorum and style; second, all the authoritative texts we read or
hear, which we write down as though they actually were on a page, marked
and colored, each with its source and each also with whatever associative
details of locus or tempus (in Hugh of St. Victor’s sense) help us to recall
them. The making of John’s second column itself requires all the m
ne-
monic utilities that Hugh of St. Victor counsels by way of dividing,
marking, and gathering authorities.
The third column is of particular interest. It is ordered alphabetically in
the memory, as distinctio collections and concordances were on the page.
To help us remember hard words and words in languages we do not know
(such as Hebrew and Greek) and the etymologies, interpretations, and
distinctiones requiring such material, we should associate words we do not
know with those we do. Although there is irreduceable imprecision in
John of Garland’s language here, I think informed guesswork will help us
through some of the murk. The two chief memory aids (notulae of a sort)
which John mentions are ‘‘omnia genera linguarum sonorum et vocem
diuersorum animancium,’’ or ‘‘all kinds of sounds of languages and the
voces animatium of diverse animals.’’11 Using the ‘‘sounds of languages,’’
coincidental homophonies between sounds in a known tongue and the
sound of unknown words, is a simple and effective mnemonic (by omnia
genera John means to include vernaculars as well as Latin). Robert of
The arts of memory
159
Basevorn’s advice to use the syllables of the chant or the five vowels to recall
texts in a sermon division is a slight variation of what John counsels here.
And we recall that Luria’s subject, Shereshevski, recalled the words of
Dante’s unknown Italian by associating their sound with words familiar
to him in Russian.
The voces animantium is a list of the habits and physical features thought
to characterize the animals and birds; it was disseminated both as a simple
listing (the form it takes in such sources as Isidore) and, in an expanded
semi-narrative form, as the Bestiary, versions of which were a standard
element in Western education from at least Alexandrine Greece.12 Scholars
have not entertained the possibility of a mnemonic function for the
Bestiary, but what John of Garland says concerning the usefulness of
voces animantium to help mark material is very suggestive, especially
taken in conjunction with other evidence. As we will shortly see, Thomas
Bradwardine also uses mental pictures drawn from the Bestiary and from
the traditional depictions of the Zodiac signs for mnemotechnical
purposes.
The Bestiary was thought of as a beginner’s book, an entertaining way of
retaining moral precepts. This does not exclude it from being of use to learned
readers too, but its primary audience was student novices and the monks who
taught them. Despite its pictorial and pleasurable content, it was found
commonly in monastery libraries, even those of the Cistercian reform. This
fact is particularly significant, for the Cistercians frowned on idle image-
making, and are responsible for an unadorned style both of architecture and
of manuscript painting. In such an apparently hostile environment, only
some over-riding perception of its utility could account for the Bestiary’s
continuing presence and dissemination among these monks.13
A book from the English Cistercian library of St. Mary’s, Holmecultram,
makes a particularly interesting study from the standpoint of mnemonic
technique. It is an early copy of an Anglo-Norman version of the Bestiary,
composed about 1130 by Philippe de Thaon for Adeliza of Louvain, Henry I’s
second queen. The manuscript is in the British Library now, MS. Cotton
Nero A. v, and the first eighty-two folios of it are the original, twelfth-
century book from Holmecultram. The Bestiary is preceeded by a mne-
monic rhyming poem on the dating of Easter, and by several doggerel
stanzas, also in Anglo-Norman, describing the traditional images of the
Zodiac and the Calendar. The fact that this book is in Anglo-Norman
indicates that it was for students who were beginning their studies and not
for the adult scholars, who would have known Latin (it is the only Anglo-
Norman book in Holmecultram’s library).
160
The Book of Memory
Philippe de Thaon’s Bestiary is presented as a memory-book. In each of its
pictures of the animals, the verses admonish the reader to remember parti-
cular pieces of the description as well as the whole: ‘ Aiez en remembrance .
ceo est signefiance,’’ ‘‘Hold in memory . this is important.’ The Bestiary is
described by its author as a ‘ gramaire,’’ or elementary book, derived from
‘ Physiologus’ (supposed to have been the original author of the Bestiary)
and Isidore of Seville, whose Etymologiae contained a listing of the voces
animantium. Thus the contents of this book were understood to be among
the essentials of a medieval education, for, together with the grid-layout
described by Hugh of St. Victor, the Bestiary and Calendar/Zodiac images
were also elementary mnemotechnical tools, the students acquiring a reper-
tory of images along with their store of basic texts and chants. 14
But how were they used? Here John of Garland supplies an important
clue. The Bestiary descriptions are laid away so that one can use them later
to mark material for recollection. They do not themselves supply an
orderly memory grid, nor (probably) were the Zodiac signs used for a
grid. Instead, one uses the rigid order of numbers and/or alphabet to lay
out one’s basic grid, but then one might use such vivid images to further
mark the material for immediate, secure recollection. What the Bestiary
taught most usefully in the long term of a medieval education was not
natural history or moralized animal fables but mental imaging, the system-
atic forming of ‘‘pictures’’ that would stick in the memory and could be
used, like rebuses, homophonies, imagines rerum, and other sorts of notae,
to mark information within the grid.
It is important that the Holmecultram Bestiary is not illustrated. Its
pictures are entirely verbal, they are not drawn on the page. This forced the
students to make the pictures carefully in their minds, to paint mentally,
thus learning one of the most critical of inventive techniques. (As we will
see in Chapter 7, decoration of one’s mental book remained basic in
mnemotechnique, even after pictorially graphic images became a much
commoner feature of medieval books, including Bestiaries.)
John of Garland counsels that the voces animantium are to be laid
alphabetically on one’s memory-page. We may recall, from Chapter 3,
how Peter of Ravenna says that he had alphabetized lists of all sorts of
material stored in his memory places, which he used to mark the thousands
of textual segments he had collected. Visual alphabets, in which the letters
are given the shapes of animals, birds, or tools, are very common, as
Frances Yates observed, in fifteenth-century treatises on memory, but
derive ‘‘almost certainly . . . of an old tradition.’’15 Often, the name of the
animal or bird also begins with the letter it is made to represent.
The arts of memory
161
Some elaborate drawings are found in Host von Romberch’s
Congestorium, 16 the product of a late fifteenth-century German Dominican,
/>
whose ‘‘congestion’ of lore (and it certainly is that) concerning memorial
heuristics seems to conserve the traditions of his order. Romberch’s schemes
incorporate what is in the Ad Herennium, but add to it an enormous amount
of advice on how to fashion complex grid systems based both on alphabetical
and on numerical orders. Several of the alphabetical ones involve animal
images; in one case, the letters are associated with animals whose names
begin with those letters, so that for A one might think of an eagle (aquila), for
B an owl (bubo), for N a bat (noctycorax), etc. Peter of Ravenna, we recall,
could pull forth a string of beasts beginning with A or B or whatever letter of
the alphabet. In another case, Romberch suggests using animals or various
implements whose figures are bent into the shapes of the various letters. The
printed text (1533) offers several pages of these, a number of which look very
like the twisted shapes of birds and other beasts used to decorate initials in
manuscripts from the early Middle Ages on. Romberch evidently sought to
preserve the mnemonic efficacy of such manuscript decoration, attested to by
Hugh of St. Victor. But instead of seeing these drawn on a page, one used
them mnemonically to mark a text or distinctio (which is basically an ethical
commonplace, a set of texts remembered ad verbum or ad res, which together
define a moral topic).17
A remarkably fierce animal suddenly looms into a Biblical interpretatio
by the Dominican scholar, Hugh of St. Cher (d. 1263), which may help to
shed light on how the voces animantium could be used for organizing
material in the memory. It occurs in his comment on the phrase in medio
umbrae mortis, Ps. 22:4. 18 ‘‘Et nota,’’ he says in a phrase which, I have
already suggested, is both an invitation to remember (in reading) and a
trigger for recollection (in composing), ‘‘quod inter omnes peccatores,
detractores proprie dicuntur umbrae mortis,’’ ‘‘among all sinners detractors
are most fittingly called of the shadow of death.’’ For death indeed spares
no-one but carries off everyone equally; likewise detractors detract from
everything. Wherefore, a detractor is signified by a bear (ursus). A bear has a
great big voracious mouth, just like a backbiter or detractor. And it has
The Book of Memory Page 29
