The Book of Memory
Page 21
pausing over what they have to say.23 Both are disciples of Quintilian and
Cicero (the Cicero of De oratore and De inventione, not the ‘‘Auctor ad
Herennium’’).
Julius Victor, more or less a contemporary of Augustine’s, alludes briefly
and disparagingly to the art of memorizing by loci and imagines. For him,
memory as a subject in rhetoric is chiefly important for composition:
‘‘Memory is the firm grasp of things and words for the purpose of
invention.’’24 The goal is plainly stated: memory is not retention for its
own sake but rather the raw material of further creation. Julius Victor cites
Cicero’s De oratore, I, 18 (a passage that became famous in the later Middle
Ages), to the effect that memory is a treasure-house (thesaurus) of every-
thing an orator needs, safe custodian of the verba and res required in
thought and invention. But Quintilian is his major source. Paraphrasing
Institutio, II. vii. 2–3 and XI. ii. 37, he advises learning by heart both your
own and others’ written compositions, but recommends particularly, as
Quintilian does, that one learn the best compositions of the orators,
historians, and other worthies by heart, for then one will carry within
oneself models for imitation and sources for substance and style.
For both written and purely memorial activities, divisio and compositio
are most useful. He who correctly puts together the basic structure of an
oration can never err (errare).25 This will hold true not only when one
arranges the questions to be addressed, but in expounding them as well – if
the first and second and so on are bound together correctly, then one will
securely remember the whole following content. In its root meaning, Latin
errare does not invoke the need for accurate representations of external
objects, but rather the dangers of chaotic, unplanned mental movement. It
papal
of
principles
the
.
summarizing) 1241
July
(distinctio
topic
Bologna,
in
legala scribe
the
marks
by
eagle
An
signed
v.
21
fo., Decretals,
4
the
of
Lat.th.b.
copy
MS.
a
in
Library
power,
Bodleian
Oxford,
of
University.4
Elementary memory design
109
invokes the concept of staying on a path and not straying aimlessly. Having
clearly marked routes – a map or ‘‘disposition’’ – is basic to a task conceived
in such terms. Similar advice concerning the advantages of properly divid-
ing a theme for the sermon is a feature of late medieval artes praedicandi, as
we shall see. Divisio means dividing a text into short segments for mem-
orizing, and compositio, putting the segments together in an order, arrang-
ing them properly from one to two to three, and so on.26 The advice is a
clicheóf classical rhetoric, from the pen of a very ordinary professor of the
subject, but it is a most valuable cliche´.
One needs, as Aristotle said, a starting-place, a beginning clearly marked
‘‘One.’’ By dividing and composing, a person is constructing a series of
numbered sequences for each text one memorizes, whether it be one’s own
work or a set piece from one of the great stylists of the past. Quintilian’s
otherwise cryptic advice (Inst. orat., X. vii. 7) to orators always to have
before them the modus and finis of their speech, ‘‘and for this division is
absolutely necessary,’’ makes some sense when one understands divisio in
its mnemonic context. The order of numbers cues both the modus or ‘‘way’’
and the finis or ‘‘goal’’ for a speech; it allows for digression and all sorts of
extempore speaking, while keeping one from losing one’s way, or forget-
ting how much one has left to cover or one’s chief points.
The mnemonic requirement for a firm starting-point also gives a
practical context for the critical importance given in medieval commen-
tary to the title of a work, an emphasis that often seems to us bizarre.
Titulus is derived from ‘‘titan,’’ says Remigius of Auxerre, because it is the
illuminating ‘‘sun’’ of the entire text. The titulus was one of the basic
analytic categories or circumstantiae that every student had to know about
a work. Mnemonically speaking, the starting-point of a text is its title;
everything else in both the text itself and its accompanying commentary
will be linked in an order from this point. 27 The Benedictine Rule
required that each brother be given a codex at the start of each Lent,
which he was to read ‘‘in order from the beginning.’’28 Reading in order is
a mnemonic requirement, as well as an injunction to read the whole book
and not just the bits we like (a habit less likely in Benedict’s time than it is
now, when we are far more accustomed – and able – to skim our regularly
printed books). As he makes the book his own in his memory, each monk
must give particular care to its starting-points for those are the key to its
order, and thus to his ability to recall it. ‘‘Begin at the beginning’’ was
made a moral duty in the Rule in part because of its fundamental
mnemonic importance, for memoria, as mindful attentiveness, was the
necessary attitude of meditation.29
110
The Book of Memory
Fortunatianus, also of the fourth century, wrote an art of rhetoric in
three books in the form of a dialogue. 30 For what he says about memory,
Fortunatianus relies almost entirely on Quintilian’s advice in Institutes, XI, 2.
Memory is a mixture of natural ability and artifice, or training; nature
is served by art and art aided by nature. Simonides, Charmadas, and
Metrodorus Scepsis taught the art of memory; what Simonides learned
from his experience at the banquet was that memory is best aided by orderly
marked-out places (sedes) in the soul; when we have such places, we can
bring things together, whether we compose in writing or mentally (ad scripta
vel cogitata).
Fortunatianus then says that the best procedure for memorizing is first
to divide a long piece into sections. Next we memorize by constant and
intense concentration (continua et crebra meditatio), and then we join one
piece to the next in numerical order, until we have learned the whole.
‘‘What assists memory the most? Division and composition; for order
serves memory powerfully.’’31 Those passages we find hard to memorize
should be additionally marked with notae. We should repeat often what we
have learned, and write passages down on wax tablets. To exercise our
memories, we should begin by memorizing poems, then orations, and then
harder material such as legal writings. Reciting in a low voice or murmur is
also a very useful technique (voce modica et magis murmure). We also retain
> better and recall more clearly what we have learned at night, when dis-
tractions are few. (Much of this advice is also in Martianus Capella, who
copied many passages from Fortunatianus; its antecedents are in
Quintilian.)
One of the commonest and oldest distinctions made in memory advice
is between memory for things and memory for words; we find it already
in Dialexeis. The meaning of ‘‘memory for words’’ is clear, even to us, for
it denotes the exact repetition of a word that we identify with memoriza-
tion. What exactly constitutes ‘‘memory for things’’ is somewhat less self-
evident. The distinction drawn in Ad Herennium is probably the clearest
of the ancient accounts. There, memory for things means organizing
memorial cues by means of a composite scene of mental images associated
with various key-words and subjects (illness, poison, heir, will, witnesses).
Memory for words also involved constructing images, but seriatim, follow-
ing the exact syllables of the original words, and is to be used to set words in
foreign languages or proper names (the sort of exercise ‘‘S’’ performed to
recollect Italian words he did not understand).
Fortunatianus considers when one ought to select one method over the
other. Should we always learn word-for-word (ad verbum)? Only if time
Elementary memory design
111
permits; but if it doesn’t we should retain only the main matters (res), and
suit our own words to them later, according to the occasion. 32 It is a very
bad practice to have to excuse ourselves and refresh our memories by a
prompt, or by reference to a book. If your memory is poor or time is short,
do not tie yourself down by trying to speak word-for-word from memory,
for if you should forget even one word in a series it will lead to an awkward
pause or to silence. So it is best to remember res rather than verba, for one
can suit words to the res as occasion demands (de tempore) and not run the
risk of needing prompting or forgetting altogether.
Word-for-word rote memorizing of a number of outstanding literary
and Scriptural texts was also always considered to be the essential base of
education. Fortunatianus and Julius Victor both were addressing students
who learned a set of canonical poetic texts, works in meter by Virgil, Ovid,
and Horace. Quintilian advises acquiring such a memorial foundation in
earliest education, and we have John of Salisbury’s admiring account of
how his master, Bernard of Chartres, set daily memorizing exercises of this
kind for his pupils. 33 But many writers gave paraphrases of texts, even when
manuscripts containing the complete text were available to them. The
reason is not far to seek. They are quoting from memory sententialiter,
according to the matter or res, rather than word-for-word. The amount of
material learned by each method probably varied according to an individ-
ual’s talent and time, as Fortunatianus here acknowledges. One finds
even the poetry of the Aeneid sometimes quoted approximately. Pierre
Richećites two Merovingian examples: Aeneid, I. 90, ‘‘Et crebris micat
ignibus aether,’’ becomes ‘‘Crebris micantibus ignibus ex aethere,’’ and
Aeneid, II. 794, ‘‘Par levibus ventis volucrique simillima somno,’’ is rendered
as ‘‘Par levibus ventis similisque somno volucri.’ 34 In each instance of
adaptation the chief words are remembered but the syntax is completely
altered in the first, moderately changed in the second. Fundamentalists will
object that the author’s sense has also been altered, and they will be correct.
But by late classical and medieval standards, the res in both instances is
preserved, even though the verba are changed.
The important thing to realize is that such alteration can result from a
choice made consciously by the memorizer and writer – it does not
automatically reflect poor training or a faulty manuscript. On its face,
alteration like this is completely neutral. Modern scholars tend to assume
that accuracy of reproduction is a function of continual access to written
texts, and thus that the extent of an author’s reliance upon his memory can
be gauged in inverse proportion to the fidelity of his quotations. I think this
is a naive assumption. It is clear from what Fortunatianus says that he urged
112
The Book of Memory
his students always to memorize the canonical texts in the first instance.
Only if one is pressed for time should one fall back on memorizing the res
(it is also clear that res could mean anything from a summary aphorism to
all the main words of a text). From advice in various early monastic rules, it
is clear that students then were obliged to memorize important texts
accurately and in full.35 Monastic reading, of course, meant meditatio, as
Dom Leclercq has so well described it: ‘‘The meditatio consists in applying
oneself with attention to this exercise in total memorization; it is therefore,
inseparable from the lectio. It is what inscribes, so to speak, the sacred text
in the body and in the soul.’’36
The Rule of Ferreolus observed, ‘‘anyone who wishes to be worthy of the
name of monk is forbidden to be ignorant of letters; he must also hold all of
the Psalms in his memory’’37 – psalmos totos, ‘‘in their entirety’’ – a task that
commonly took two to three years, though gifted individuals could manage
it in six months. The mnemonic practice which informs this command also
informs the admonition of the sixth-century (?) Regula magistri that monks
on a journey who had not yet mastered the Psalms should provide them-
selves with tablets on which the text was written, so that when they stopped
they could sharpen their memory with the help of their companion.38 This
requirement is also derived from ancient mnemonic practice: recall
Quintilian’s comment that it is of great value when one is practicing in
order to fix one’s memory, to have someone else read aloud the material, so
that one can check the accuracy of one’s recollection against the reading.
The book which Christians, both clergy and educated laity, were sure to
know by heart was the Psalms. It has been observed that Augustine wrote
not only in Latin but ‘‘in Psalms,’’ so imbued is his language with their
phrasing and vocabulary; the same could be said of any number of
Christian writers. Nor was it only the Psalms which were so intimately
known; Pierre Courcelle has shown the extent to which Gregory the Great
and other writers might be said to write ‘‘in Augustine.’’39 Such intimacy
can be achieved only by long and thorough familiarity with a text as a
whole, not just a few aphorisms. As Richeóbserves, ‘‘to know how to read
was to know one’s Psalter’’; even Merovingian lay aristocrats, especially the
women, knew at least a few Psalms verbaliter, word-for-word. 40
Accuracy of recollection was a helpful skill to nurture in an age of few
manuscripts, many of uncertain quality. The Regula magistri counsels that
the scriptures be retained in memory partly so that if a codex has lacunae or
lacks the commen
tary (textum lectionis), the missing parts could be sup-
plied or expanded from memory. 41 There is no point in debating whether
or not their faith in the accuracy of memory was misplaced, for some
Elementary memory design
113
individuals have highly accurate recollection and others do not. The point
to understand is rather that one’s memory was expected to be not only
copious but accurate (the reverse, one might observe, of expectations now).
Nor is this characteristic only of the period from the sixth through ninth
centuries, when the availability of books was at a minimum. Writing his
Dialogus while in exile in Munich and away from the libraries that nurtured
him earlier, William of Ockham apologizes not for the inaccuracy of his
memory but for the fact that circumstances prevent him from having access
to all the latest documents in the ongoing controversy, and so his treatise is
incomplete – it was imperfect in that sense. Thomas Aquinas, living in a
century far richer in books than the sixth, still stored his reading from
various libraries in his memory, to be pulled forth as a seamless golden
chain, Catena aurea. The ability perfectly to replicate the contents of one’s
memory again and again, forwards, backwards, and in all sorts of combi-
nations, remained a revered skill at least until the end of the Renaissance.
‘‘Memoria rerum’’
The ‘‘inaccuracy’’ we find so frequently in medieval citation can often be
the result of a deliberate choice on the authors’ part, either at the stage of
initial memorizing or (and I think more frequently) at that of composing.
Medieval scholars’ respect for accuracy in copying texts has been repeatedly
demonstrated (despite complaints about particular errors), and it is justly
observed that without the labors in scriptoria throughout the Middle Ages
virtually no ancient literature would be known to us today. Yet the same
people who honored the exact copying of even non-Christian texts quote
these same works erratically, at times precisely, at times so paraphrased and
adapted as to alter them almost beyond recognition. The one sort of
activity is verbatim memorization, the other memoria rerum, memory of
subjects. Both can employ mnemonic techniques, such as those just
described. In the traditional ancient and medieval education in liberal
arts, word-for-word rote memorizing is associated particularly with gram-