Thomas suggests that although some cite the Psalms by number, others
omit such numerical citation, referring instead generally to ‘‘the Psalter,’’
since the Psalms are so commonly known:
Next, whenever a particular Psalm is cited, it is also referred to as the fortieth or
thirtieth, or in another similar manner. Others however omit such citation, and
Elementary memory design
127
only say: ‘‘This is found in the Psalter,’’ and the reason is that the Psalms are generally
well known. 84 (my emphasis)
This comment by Thomas would seem to explain why, even in late
medieval written Bibles, the Psalms are commonly unnumbered while in
all the rest of the books the chapter numbers are included. The Psalms are
unnumbered because they are so well known, having been memorized as a
first book – people have their own schemes for recalling them (some recite
them with their numbers, some do not), and a cue is enough to bring forth
the whole text in recollection. In written texts such as moral treatises and
treatises on preaching, well-known texts such as the Psalms and Epistles are
commonly referenced by the first few words of a phrase only, for example,
‘‘Beatus vir &c.’’ Or in some treatises, one finds a verse referenced solely by
the first letters of the words, on the model, adduced earlier, of ‘ b. u. q. n. a.’
for Psalm 1:1. These references could only be useful to readers who had
memorized verbatim the texts being cited.85
Thomas of Waleys says, though, that thematic sentences taken from
every other part of the Bible must be cited by chapter and book. This
comment indicates how widely used some sort of numerical grid was as a
mental filing system. The style for citing the Decretum, used in glosses and
other contexts throughout the Middle Ages, was according to the number
of the large division (distinctio, quaestio, or causa) followed by the intro-
ductory word or phrase of the particular canon referenced, ‘‘xxxij c.
multorum’’ or ‘‘xxiv c. cum itaque.’’ If two or more canons in a given
quaestio or distinctio began with the same words, they might be distin-
guished as, for example, ‘‘c. Si quis i, Si quis ii,’’ etc. Similar forms were
used for citation of the various volumes of decretal collections compiled
after Gratian’s catalogue. 86 Obviously a canon lawyer was expected to have
pretty well memorized the entire set of Decretals in order to be able to use
these references without a great deal of wasteful ‘‘turning the pages.’’
William of Ockham, in his Dialogus, speaks of his relief at having stocked
his memory with so much material, including all the Decretals, when he
found himself exiled in Munich; Peter of Ravenna, a fifteenth-century
Italian lawyer, claimed (as we will see) that adopting his version of an
alphabetical mnemonic enabled him to learn all the decretal collections.
Yet there were clearly risks in citing too many numbers while one spoke.
Thomas of Waleys advises that supplementary authorities, Biblical and
patristic, should not be fully cited by book and chapter number, but by
book alone. If one insists on citing every chapter number it often happens,
when bringing together several authorities, that a chapter number falls out
of the preacher’s memory, and one then could bring forth another section
128
The Book of Memory
than the one intended. Wherefore, he says, he praises the practice of former
times when preachers did not cite chapter numbers in their delivery, but
only when they wrote their sermons down. He concludes, ‘‘I think their
method was better because it was safer [tutior] than the modern method by
which chapter numbers and books are cited together.’’87 Before dismissing
this as the advice of an irredeemable pedant terrified of being caught out by
a colleague, let us consider it in terms of the practical difficulties attendant
on the numerical grid as a method for recalling text.
The grid serves essentially as a coding and filing system for the orderly
and ready recollection of material. In using a mnemonic, however, one
does not normally call out its coordinates, whether one’s grid be the
architectural places or this numerical one. The grid is essentially arbitrary,
in the sense that it is not part of the material placed within it, any more
than a file is part of what it contains. However, wrong coordinates will send
one to the wrong part of the file, and thus to the wrong text, causing a
devastating loss of place. Thomas advises articulating only the name of the
work cited, a safer thing to do since it isn’t a part of the grid. It is not the file
itself, only the name of the file. Indeed it is striking that he advises avoiding
articulation of the code whenever possible, even of the Psalms, if one can
get away with it. Only in the more secure situation of a corrected, written
version should the numerical grid be articulated, when any miscitation can
be worked out. And his advice is given for safety’s sake.88
Not every preacher followed Thomas’s advice in this matter; he himself
suggests that it is a bit old-fashioned, and of course the sermons we have
preserved are all in written form, with all citations given. But it is a
significant point that medieval citations are given before the text more
frequently than after it. The anterior position serves to cue the mental
grid. Our mode of citing after a quotation, in parentheses or a footnote, is
designed solely to send a reader to a printed source that he or she must find
elsewhere. Our citations are referrals, not access cues to a mental grid of texts.
Indexing systems devised in the monasteries and, later, the universities
seem all to have served the dual function of being both tools for finding
texts initially and also for noting them in the memory, as mnemonic hooks.
Not only numbers could serve such a purpose: the Dominican system,
devised for the Bible concordance, of citing by book, chapter number, and
letter (A–G) to indicate the position of a text within the chapter was widely
disseminated. The Rouses describe a bookmark of the thirteenth century,
which explains the key to the pagination scheme of the volume in which it
is found, a Cistercian manuscript in which pages are marked with a
combination of letters and dots. 89 One side of the bookmarker says: ‘‘Ut
Elementary memory design
129
memoriter teneatur alphabetum taliter ordinatur: a .a :a a:,’’ etc. Evidently
this reference scheme was designed in part to provide a memory grid for the
readers. On the reverse of this bookmark, the strip of parchment is divided
into seven equal sections, marked A through G, the location scheme also
used by the Dominicans within chapter divisions. One held this side
against the edge of the page to provide the finely tuned location for the
text. A number of other such bookmarks are known.90
As Richard and Mary Rouse say of all these various coding and filing
schemes, ‘‘they emerge with striking suddenness in the West, to the point
that one may say that, probably before 1220, certainly before 1190
, no such
tools existed; and that, after the 1280s, the dissemination and new creation
of such aids to study were commonplace.’’91 Of the sudden proliferation of
written tools there is no doubt; as there is also no doubt that the same order
(the Dominicans) chiefly responsible for creating them is also chiefly
responsible for the dissemination of mnemonic devices based on places,
including the Herennian one. It is also true that during the thirteenth
century schemes for memory training were disseminated into the general
culture in France and especially in Italy, via vernacular translations and
adaptations of the architectural and other memory-training techniques.
We therefore should not assume that these multitudinous study aids
replace memory as a fundamental tool; instead, they often were thought
of as memory systems first and manuscript aids second. This suggests that
reading and memorizing were taught as they were in antiquity, as one
single activity, and, further, that the monastic understanding of what one
does in reading, so well described by Dom Leclercq, not only persisted but
became part of general culture in the thirteenth, fourteenth, and later
centuries, for reasons that had as much to do with the moral value of
memoria in meditation and prayer as with its utility. But this is a subject for
later chapters.
Furthermore, the proliferation of written guides to these various heu-
ristic schemes in the twelfth and thirteenth centuries does not mean that
there were no indexing mnemonics in use earlier. Hugh’s Preface describes
an indexing scheme in great detail, long before 1190, and, for reasons I have
indicated, it seems most likely he was describing a well-known, persistent
technique. Indeed it is likely that all these schemes – numbering of
chapters, alphabetizing (the memory scheme which Aristotle described in
De memoria was an alphabetical one), page indexing by number and/or
letters – are medieval adaptations of schemes known since antiquity as
memorial schemes. With the Rouses and other scholars, I suppose that a
chief reason for the proliferation of written indexing schemes was that
130
The Book of Memory
more people had need of organized textual material and that there were
more texts to be consulted after the twelfth-century infusions from Spain
and Sicily. Once the scholarly community was enlarged beyond the mon-
astery, the need for written transmission was greater, and priests and friars
who need to preach have need of preachers’ tools. But this did not occur at
the expense of memory; indeed the written schemes themselves are to be
used as memory grids, ut memoriter teneantur.
One other intriguing practice which indicates the tenacity and perva-
siveness of the methods described by Hugh of St. Victor is that of quota-
tion itself, and the related practice of what Hugh calls ‘‘gathering,’’
colligendum. The medieval verb quotare first makes its appearance, accord-
ing to the Revised Medieval Latin Word List, early in the thirteenth century
(which does not mean it was unknown earlier). Derived from the adjective
quot, ‘‘how many,’’ it meant ‘‘to number’’ a book, dividing a longer text into
numbered subdivisions, such as chapters. Robert of Basevorn discusses
quoting in chapter 31 of his Forma praedicandi, written in 1322. After the
prayer, the theme is restated and quoted, ‘‘quotandum quantum ad librum
et capitulum.’’ For authorities other than the theme, it is not necessary to
give the section number ‘‘secundum modum modernum’’ (apparently, like
Thomas of Waleys, Robert did not like the new custom of citing every
authority by chapter number), though one may number books in a series
(like I and II Kings), or one may also give a chapter number in a book
which explains another book (Gregory’s Moralia in Job, for instance, which
comments direcly on Job). The correct form is, ‘‘Gregorius primo, vel
secundo, Moralium super illud Job.’’ But masters and professors of theol-
ogy may ‘‘quote’’ (numerically divide) anything:
It is permitted, indeed it is their prerogative, for masters and professors of sacred
theology to quote everything in a carefully worked-out manner [exquisite], since in
regard to this matter much deference is given them by others, because of the
testimony of their excellent teaching and display of humility regarding their
responsibilities, and because of the dignity and honor of their rank.92
So the practice of quoting, marking, and numbering a text for citation
seems to have been the special prerogative of the most learned members of
the university, who alone are able to quote exquisite.
The quotation of patristic texts is a feature of thirteenth-century scholar-
ship, and made possible the compendia of authorities designed for sermon
composition, such as distinctiones collections, textual concordantiae, and
other libri rerum memorandarum. Often in these collections, each quote is
cited by book and chapter number (and in some by the further refinement
of A–G division), the two basic coordinates of the grid that serve not only
Elementary memory design
131
to help find the quotes in a complete text of their parent work but to place
them in a memory file, so that, considering Avaricia, for example, would
bring forth a number of texts, each labelled in a mental file called
‘‘Avaricia.’’ Indeed, for most preaching friars, working far from large
libraries, the citations could only have served as memory coordinates.93
It is likely that the passion of many late medieval preachers for number-
ing sermon divisions – what has been called the scholastic or university
method of the thirteenth century – is related to the nature of this numerical
grid. Thomas Waleys says that numerical division of the theme in the
modern style is not just a gimmick, but most useful for both the preacher
and his auditors:
Indeed if only one division of the theme be made, still that division will be
beneficial as to those subjects, as much for the preacher as for the hearer. For
the moderns began not just because of a vogue, as others believe, to divide the
theme, which the ancients did not customarily do. Especially it is useful for the
preacher, because division of the theme into separate parts affords an opportunity
for dilation in the later development of his sermon. For the hearer truly it is most
useful, because when the preacher divides the theme and afterwards develops the
parts of the division in order and clearly, both the matter of the sermon and the
form and manner of the preaching is more easily understood and retained; that
will not be the case if the preacher proceeds unclearly or without order and by a
confused structure.94
Of special interest in this advice is the assumption on Thomas Waleys’s
part that both preacher and auditor relied on the numerical system for
retaining discourse, for the key both to successful dilation and to retention
is orderly division by number. This manual was written within a university
setting, in which preacher and heare
rs alike were scholars, but the same
method was used in preaching to the laity.95 The fourteenth-century
Dominican preacher, Giordano of Pisa, whose sermons were taken down
by dictation while he spoke, used the method of division constantly. Since
he preached commonly two or three times a day, Friar Giordano often only
got through the first few divisions of his sermons, though he always
announced all his themes in order in his introduction. When exhausted,
he would promise to continue where he left off at a later time – and he did,
invariably picking up exactly where he left off. These sermons were taken
down as they were preached; they were not later corrected. 96
Robert of Basevorn’s quoting schemes
Several excellent examples of how thematic division works mnemonically
can be found in Robert of Basevorn’s treatise on preaching. For a sermon
132
The Book of Memory
on the Passion, Robert suggests the following ‘‘division’’ of the text: ‘‘Jesus
iterum clamans voce magna emisit spiritum,’’ Mt. 27:50 (I will quote the
Latin text here because the word-play, essential to its mnemonic effective-
ness, cannot be reproduced in an English translation):
Hic notantur: Passionis utilitas, Jesus; patientis potestas, clamans; humanitatis
veritas, voce magna; patiendi libertas, emisit; separationis acerbitas, spiritum.97
This is the basic division into five, a portion of the theme plus a rhyming
catchphrase which in some way (not always clear to me, I confess) abstracts
the major idea to be developed from each of the five divisions.
Then comes the subdivision, also in five parts:
Subdivitur tunc sic: Quinque sunt vocales, scilicet A E I O U, quae omnem vocem
faciunt. Sic quinque vulnera Christi omnem sonum, sive doloris sive gaudii faciunt.
Vide in manibus A et E: [Jer. 31:3] ‘‘Attraxi te miserans’ etc. [Is. 49:16] ‘‘Ecce in
manibus meis descripsi te’’; I in latere: talem enim figuram imprimit vulnus
lanceae, hoc est ‘‘ostium arcae’’ quod ‘‘in latere,’ etc. in Gen. [6:16] et in Joan.
[20:27]: ‘‘Infer digitum tuum huc et mitte in latus meum, et noli esse incredulus,’’
etc.; O et U in pedibus: [Ps. 8:6] ‘‘Omnia subjecit Deus sub pedibus ejus.’’ Ideo ut
consequaris dicas facto: [Job 23:11] ‘‘Uestigia ejus secutus est pes meus.’’98
The Book of Memory Page 24
