No one will have any difficulty in recognizing this counsel to be at once
typically medieval and, from a modern standpoint, bizarre to the point of
intellectual decadence. Yet the scheme proposed is eminently sensible as a
mnemonic both for composing and for delivering a sermon, as well as
providing auditors with a clear structure within which they can grasp and
retain the major points of the sermon.
This example, like many others, combines a variety of mnemonic
structures, which in this case are also reflected in its compositional struc-
tures. 99 The theme is the Passion, and the preacher elects to divide by fives,
a number with evident associations to his subject because there were five
wounds in the crucifixion. A mnemonic rhyme composed of summary
catchphrases organizes the main division in five. Then, in the subdivision,
one remembers the five wounds, imposing upon the structure of those five
another five, the five vowels, or voces, perhaps used because of the word’s
punning association with voce magna, the third division of the text which
deals with the theme humanitatis veritas. (Punning of all sorts is funda-
mental in mnemotechnic; one encounters visualized homophony through-
out the history of written mnemonic advice, as a principle both for forming
images and for association of ideas.) To each of the five vowels is attached a
text beginning with it. Five vowels, five wounds; the basic composition is
complete, combining a scheme of loci and imagines that incorporates a
Elementary memory design
133
numerical grid, on which are ‘‘placed’’ brief texts and other sorts of visual
images. As hermeneutic, an interpretation of their meaning, attaching
these particular texts to the five vowels is grotesque non-sense, but as an
elementary device for retaining and recollecting them – as heuristic – it is
both simple and effective. One can return to the theme of one’s compo-
sition after a detailed exposition of any of its pieces and readily find one’s
place.
Robert of Basevorn’s other examples follow the same principles. For a
division into six of the theme, ‘‘Ego vox clamantis in deserto, parate viam
Domini,’’ he suggests using the six basic syllables of the chant, a pun on
‘‘vox clamantis.’’ To these six notes – ut, re, mi, fa, sol, la – are attached the
texts for the subdivision, as follows:
1) UT filii lucis ambulate; 2) REvertere, revertere, sulamitis, revertere, revertere ut
intueamur te; 3) MIsere animae tuae placens Deo; 4) FAcite dignos fructus
poenitentiae; 5) SOLve vincula colli tui, captiva filia Syon; 6) LAvamini, mundi
estote. 100
This is a most revealing application of the technique called solmization, for
it shows that Robert of Basevorn understood that the device was primarily a
mnemonic, and could thus be utilized in non-musical contexts. It also
suggests how much more broadly the function of a mnemonic was under-
stood then than it is now, when we restrict it to reiterative tasks.
The principle of attaching a syllable to a particular musical value or
‘‘degree’’ was known in antiquity, and indeed is found in virtually every
culture. 101 It is most notably discussed in the Middle Ages in a letter of
Guido d’Arezzo to the monk, Michael. There is evidence that solmization
was used in medieval Europe before Guido, but the Benedictines were
certainly happy to give him credit for having ‘‘invented’’ it, and the system
he describes became the dominant one, indeed the foundation of what
Westerners still use. Guido’s syllables are taken from a hymn to John the
Baptist, ‘‘Ut queant laxis’’; these syllables then became the names of the six
parts of a hexachord, at that time the unit organizing musical tones. As
Guido wrote to Michael:
You will then have an altogether easy and thoroughly tested method of finding an
unknown melody . . . After I began teaching this procedure to boys, some of them
were able before the third day to sing an unknown melody [to them] with ease,
which by other methods would not have been possible in many weeks.102
His system is designedly mnemonic, ‘‘if you wish to commit any note or
neume to memory,’’103 whether you knew the melody before or not, ‘‘quem
scias vel nescias.’’ For the system of places forms a kind of alphabet. Indeed,
134
The Book of Memory
Guido’s analysis of the purpose and function of musical notation seems to
derive from Isidore of Seville’s definition of the purpose of letters, that is, to
hold things in memory, including what we do not already know. Just as the
alphabet enables us to hear again and retain in memory the voices/words of
those who are not actually present and whom we have never heard or seen
in the flesh, so musical notation is, in this way, like an alphabet (and so says
John of Salisbury, in a passage I will examine in the next part of this
chapter).
Here, as well as in any other medieval text, we can understand that craft
rules (artes) were thought to be, as Aristotle says, built up from repeated,
remembered experiences, the principle being to recognize and organize
likeness. This could then be applied, even to things never seen before and
encountered for the first time. This is not mnemonic in the restricted sense
that moderns tend to understand it, but in the larger sense of how all
learning takes place. Learning is regarded as a process of discovering
more effective, efficient, inclusive mnemonics – for memory, as Hugh of
St. Victor says, is the basis of learning. To learn the alphabet (and numbers
were designated by the alphabet, too, we should keep in mind) was to be
possessed of a set of infinitely rich mnemonics – not only a skill that would
enable one to read signs on a piece of vellum, but a key opening the door to
the whole cultural complex of institutionalized practices signified in the
word memoria. This is an analysis of learning that differs from our own in
its emphasis upon memory and memorial cues, and it is very far from
restricted to rote tasks. It places rote in the service of creative thought.
Whatever the number of one’s sermon division, Robert of Basevorn
recommends attaching a set of symbols or markers that incorporate that
number and can be used as a mnemonic for the subdivisions. For example,
besides the two already described, Robert suggests the seven mercies
of God, the eight Beatitudes, the nine orders of angels, the Ten
Commandments, the twelve hours of the day. But eleven, he says, is almost
never used, and never a number greater than twelve, for one will lose track.
This observation is borne out in a catalogue of such numbered groups of
items that was made anonymously and attached, as Book IV, to Hugh of
Fouilloy’s De avibus. Though many threes and sevens, fours, sixes, tens,
and nines are listed, there are only two elevens – and nothing greater than
twelve. 104 It is also useful, Basevorn says, to break up a larger number; for
example, subdividing an initial division of nine into three sets of three,
since there are so many threes th
at finding them presents no problem.
Mary and Richard Rouse comment, concerning a sermon of Alan of Lille
on the Trinity, that it is ‘‘an extravaganza of triplicity’’ – so too they well
Elementary memory design
135
characterize the sermons of Innocent III as amounting to ‘‘a sequence of
lists’’ of threes or fives or fours or sevens. 105 The subject matters so grouped
are the singulars of which Hugh of St. Victor speaks, sorted and stored like
coins of similar size in the properly numbered compartments of the
memorial store-house.
T H E A L P H A B E T A N D K E Y - W O R D S Y S T E M
The number grid which Hugh of St. Victor describes was but one of several
memorial heuristics taught in the Middle Ages. Memory training, we learn
from Quintilian, was an elementary component of learning to read and to
write. The numerical coding seems best designed to remember and to
collate (that is, move discrete pieces of texts about in various combinations)
longer works, like the Psalms, although it clearly can be used for a variety of
other purposes as well. But in addition to such a task, a memory library
needs to have cross-referencing systems, and these also seem to have been
taught among the educational basics.
Quintilian speaks both of memorizing a text in (numerical) order, and
of marking the passages one especially wants or needs to remember with
mental notae. Notae are discussed twice in relation to memory training.
First, describing the system of loci and imagines, Quintilian suggests
imprinting in orderly progression a spacious house with many rooms,
and then marking the items to be remembered by a nota, either an
associative sign (an anchor to remind one of navigation, for instance) or
a key-word, for even people who lose the thread of what they are saying can
put their memory back on track with a single cue. These notae are then
placed in the orderly series of rooms. He prefers, however, a simpler
system, yet one still using notae. After one divides a text into short sections,
and has repeated these to oneself two or three times and perhaps written
them down as well, one then marks passages one especially needs to
recollect with notae which one has invented for oneself.106 Martianus
Capella some 400 years later also advises using notae to mark passages,
and the word is found commonly in memory advice, and ubiquitously
written out in the margins of manuscripts against passages that either the
scribe or a later reader thought to be especially important for remembering.
Notes and places
It is unclear from Quintilian and other writers whether these notae or
notulae are to be inscribed on the physical page or only in memory, as
136
The Book of Memory
Hugh of St. Victor counsels; clearly some were made physically, for
marginal marks do occur in manuscripts, copied by scribes to help users
of the books. But since every reader is advised to make notae, most
mnemonic signs must have been devised mentally, especially when books
were used by many different readers over several generations. In manu-
script margins from the twelfth century on, one finds commonly the word,
nota, addressed to the reader. It is the imperative singular of the verb notare,
‘‘make a note,’’ and it points out an important or difficult passage that the
reader might wish particularly to mark with a nota of some sort of his own
to help remember it.
Mental marking is mnemonically advantageous because each individual
makes up his own system of notulae, his own filing system, and, as we know
from Ad Herennium, one’s own notae and imagines are to be preferred to
memorizing a pre-existent system, because such exercise stimulates the
memory more fully and fixes it more securely. Hugh of St. Victor describes
a completely mental system; Robert Holcot, in a text we will examine again
later, says that as he reads he imagines a memory-image projected onto the
particular text that he wishes to remember, and he gives detailed instruc-
tions in his commentaries for making these. Yet the written manuscript
containing them is a plain one, rubricated but without any drawing or
decoration. Only in the margin, opposite each verbal picture, is written the
single word pictura. Perhaps we might understand the word in this context
not as a noun, but, on the analogy of nota, as a command to the reader that
he is to make a picture from the written description, evidently in his mind. 107
At the end of the English Middle Ages, Stephen Hawes writes of the
orator tucking into his memory ‘‘sundry ymages,’’ not only for the ‘‘mater . . .
lyke to the tale’ but also in a ‘ recapytulacyon’ (‘ re-chaptering’ or ‘ cross-
heading’ ) of each image by its ‘‘moralyzacyon,’’ a general ethical topic or
‘ commonplace’ heading. Much material – how much would have to vary
individually – seems to have been assumed to be filed away not as a
complete text (or not only as a complete text) but as sets of extracts
‘‘noted’’ in the memory. The verb tractare, meaning to ‘‘draw out’’ or
‘‘extract,’ is frequently used of a process of reading as well as of a genre
of composition. To ‘‘tract’’ a text while reading it means to pull out of it the
various matters that one wishes to squirrel away in one’s memorial inven-
tory. Then, when composing a tract (the gerund form, tractandum, is used
to indicate this activity), one begins by collecting on a particular subject all
the material, together with appropriate commentary, that one had previ-
ously drawn out of one’s reading and filed away under a letter or a key-
word or some other nota.
Elementary memory design
137
A late but important description of memory-work by the fifteenth-
century Italian jurist, Peter of Ravenna, who wrote one of the works on
the art of memory most popular in the Renaissance, states that a well-
trained memory is most like a book containing both text and glosses. ‘‘For
I daily read all my lectures of Canon Law without a book; but if I should
have a book before my eyes, I deliver the textual concordances [textum] and
glosses from memory so that I should not seem to omit the least sylla-
ble.’’108 Thus, at the end of the fifteenth century, when printed books had
become almost common, a law professor might still lecture regularly
without a book, and could supply his lecturing apparatus more accurately
and completely, he says, from his memory than by depending on a book.
Peter’s memory-places were arranged alphabetically: ‘‘on the nineteen
letters of the alphabet I have placed twenty thousand extracts of both sorts
of law.’’ I will come back to this passage shortly to discuss exactly how he
says he did this, but for now I want to stress his use of the alphabet as his
ordering device. He uses, he says, the regular alphabet, but also ‘‘human
figures in place of the letters and thus living, vivid images.’’109 For the sake
of vivid images, unusual ones of the sort memory can easily fix on, he can
make use of a sort
of human alphabet to indicate the various letters. He
even suggests using the forms of enticing women for such a purpose: ‘‘illae
enim multum memoriam meam excitant,’’ ‘‘these greatly stimulate my
memory,’’ though, he adds, this is not a technique for those who loathe
women or who cannot control themselves.
Evidence that alphabets were commonly used as a mnemonic ordering
device is scattered but persistent in both ancient and early medieval books.
Aristotle describes (in De memoria) the use of letters of the Greek alphabet
as an ordinary method for ordering memorized material. But if one is using
an alphabet to file lots of different topics, one might well need more than
one set of such places to accommodate one’s material, lest one’s memory be
overwhelmed by trying to crowd too much into one place (a persistent
warning in memory texts). To avoid this, one needs more than one set of
alphabets, containing different forms though having still a rigid, easily
recoverable, order. That is what Peter of Ravenna was doing with his
human alphabets.
It is a curiosity of medieval books that tables of Greek, Hebrew, Coptic,
runic, and even wholly imaginary alphabets are found side by side in a
number of monastic manuscripts. It is also an insufficiently explained
curiosity that books containing the Bestiary, moralized descriptions of
animals both real and imagined, are found commonly in monastery
libraries. Scholars have wondered what function such apparently puerile,
138
The Book of Memory
unscholarly material might serve to justify its preservation. I will return to
this question in Chapters 4 and 7, to make my case more fully, but my
hypothesis is that they functioned not only to delight and intrigue medieval
students but to provide them with mnemonically valuable heuristics,
orderly foundations or sets of mnemonic loci, which can continue to
have value throughout one’s education (a lifetime project). In this context,
such a list of letters does not serve primarily to form words (and indeed in
the case of imaginary alphabets, no words are to be formed) but to supply
sources of notae. They have more in common with notarial abbreviations
and punctuation than they do with the phonemic graphemes of an intelli-
The Book of Memory Page 25
