chapter is numbered, in Roman numerals with alternately red and blue
colored elements. These numbers also have distinctively drawn colored pen
lines and circles in them – these, plus the colored initial, for which there are
two or three different forms for the commonest letters, give each folio
opening a unique appearance. The running heads at the top of each
opening are also distinctively colored and drawn with differentiated pen
ornamentation. Since this Bible is a small one, the differences in the design
of the running heads from one opening to the next are quite noticeable.
The Psalms, together with collections of maxims like Cato’s Distichs,
were the elementary reading text throughout the Middle Ages, from late
antiquity onward.66 And Hugh of St. Victor evidently regarded their
mastery as a beginning task, among the puerilia. But in the majority of
medieval Bibles, the Psalms, unlike the divisions in all the other books, are
unnumbered. Moreover, whereas in Carolingian Bibles the Psalms are
written in verses, Bibles of the thirteenth century and later characteristically
wrote them out as units of prose. The reason for this change undoubtedly is
that the Bible at this time began to be issued as a single bound volume,
often quite small, even pocket-sized, whereas in the earlier Middle Ages it
was usually issued in separate volumes of full folio size.67
However, in these later Bibles the verse divisions are indicated by
colored initials, alternately red and blue. Each psalm begins with a large
colored initial, sometimes fully decorated, other times just drawn large, red
or blue alternately. It is interesting that the Huntington Library’s manu-
script of Richard Rolle’s fourteenth-century English Psalter (HM 148)
preserves this format for the Latin text of the Psalms written above
Rolle’s translation of it and of Peter Lombard’s commentary. Each
Psalm’s verse division has its initial letter colored alternately red or blue,
the first initial of each Psalm is a large blue letter with distinctive red pen
decoration surrounding it, and the color scheme is repeated, like a color-
code, for the English translation, and the commentary which follows.
These markers alternate between red and blue. An effort is made in the
writing of the book to distinguish each memory-sized chunk as a unique
122
The Book of Memory
visual image. Since in fixing one’s images for storage (which required time,
and at least two or three separate passes over the written text), one made
sure to use the same codex – as all memory advice stressed – the mnemonic
usefulness of such decorative elements is apparent.
Even what we hear must be attached to a visual image. To help recall
something we have heard rather than seen, we should attach to their words
the appearance, facial expression, and gestures of the person speaking as
well as the appearance of the room. The speaker should therefore create
strong visual images, through expression and gesture, which will fix the
impression of his words. All the rhetorical handbooks contain detailed
advice on declamatory gesture and expression; this underscores the insist-
ence of Aristotle, Avicenna, and the other philosophers, on the primacy
and security for memory of the visual over all other sensory modes,
auditory, tactile, and the rest. Hugh as well insists that acoustically received
material must be translated to visual terms and so fixed in memory. We
recall that for the memory-artist, S., it made no difference whether material
to be retained was presented to him orally or in writing – his visualization
technique was the same.
Though the psychological principles are identical, such image schemes
are far more sophisticated, complex, and learned. The commonness of the
number-grid is indicated by its use to format the Bible, and by its traces in
other scholarly contexts, such as the style used to cite the Decretals and
other non-Scriptural texts.
Orality, literacy, memory, and citation of texts
The numerical grid imposed on Scripture by its division into numbered
chapters and verses was first printed in the Geneva Bible of 1560. But
Scripture had been divided into brief segments before Jerome, and indeed
dividing by chapter goes back to the scholars of the Masoretic text. (We
should remember that ancient texts were commonly copied without any
divisions except for the fixed number of syllables that constituted a line of
writing – students learning to read were taught how to divide up the text on
their own. This was the practice for writing texts through about the third
century AD.) Jerome divided in the classical fashion of cola and commata; so
did Cassiodorus, who used not only these divisions of the text but chapter
divisions and headings as well.68 The late medieval Bible is divided into
numbered chapters, using the scheme of Stephen Langton, together with
the Cassiodorian division of the book of Psalms into cola and commata,
which Alcuin had made standard. The essential utility of such division for
Elementary memory design
123
memorizing the Biblical text was obvious to Langton, who glosses as
follows Jerome’s warning, made in his prologue to the Vulgate text of
Joshua, that both reader and copyist should diligently preserve the divi-
sions of the text: ‘‘by divisions, that is by chapters . . . which truly are most
valuable for discovering what you want and holding it in your memory.’’69
Dividing was not a casual task: Langton took thirty years to work out his
chapter divisional scheme, perfecting it through classroom testing.70
Langton’s chapter divisions superseded various earlier divisional
schemes. This whole matter of the relationship between how scholars
laid out a text in their memories, cited it when they made reference to it,
and how scribes composed it on a manuscript page is well illuminated by a
study of these habits prior to about 1200 AD. The problem is important
because some anthropologists and historians of technology seem to assume
that there is a direct and simple correlation between the form something
takes in writing and the way a person is able to think of it, in the same way
that a washing-machine’s design determines how clothes washed in it will
be washed. The fashion for defining writing as a technological innovation
of the same sort as television and the automobile, or the heavy plow and
moveable type, seems to me fraught with difficulties. Why I think so will be
apparent from the following study of citational habits, page lay-out cus-
toms, and the use of the mental numerical grid for remembering Scripture
during the early Christian and Carolingian periods.
It is apparent from remarks of St. Augustine, who used the pre-Vulgate,
Old Latin text of the Bible, that the Psalms at least were taught to him and
his audience in an order that was numerically designated. In his commen-
taries on the Psalms, which began as sermons, he prefaces his comments on
Psalm 118 (119), with an observation about his compositional habits. He
apologizes for delaying so long, though he has expounded
all the others,
partly in sermons, partly by dictating additional material, but Psalm 118 is
long and extremely difficult.71 He refers to this Psalm by number (‘‘psal-
mum centesimum octauum decimum’’). Of another Psalm, 125, he says,
‘‘Now you remember, according to the order taught to us, this Psalm is one
hundred twenty-five, which is among those Psalms whose title is ‘A song of
degrees’’’ (my emphasis). 72 From the manner of address, it would appear
that this commentary began as a sermon, as many did, and Augustine
clearly assumes that everyone in his audience will know which is Psalm 125,
‘‘according to the order taught to us.’’ Similarly, in commenting on Psalms
100, 104, and 105, he refers to them by number, though his usual habit is
simply to begin by referring to ‘‘this Psalm’’ (iste Psalmus) without further
identification. He does not find mystical significance in these numbers;
124
The Book of Memory
indeed, his attitude towards their numbering seems quite purely practical.
Yet, commenting on Psalm 150, Augustine engages in a memorable tour de
force of associations, expounding several possible numerical divisions of the
Psalms; clearly their numbers were known also to Augustine’s lay audien-
ces. Such compositional dilation upon the secretum of numbers, is a trope,
and it remained so in medieval preaching. No doubt the preachers believed
that a numerological mysterium was divinely concealed in the texts (inter-
preters usually believe that what they find is ‘‘in’’ the text). Yet it is also true
that the technique is a most effective inventory tool, open to use by mortal
authors – witness Dante’s Commedia and, in the English tradition, the
work of the Pearl poet. It is worth remarking that any heuristic instrument,
such as a number, can be treated as a reified sign. A mathematician once
explained to me the difference between numerology and number-theory
this way: the one (numerology) wants to explain everything, the other
explains nothing but itself.
Referring to Psalms by number is not, however, Augustine’s style of
citation when marking a quotation. It was the universal custom in early
medieval scholarly composition not to use numbers at all when citing
sources. Passages from Scripture especially are often quoted without any
attribution; the audience was expected to recognize the source from mem-
ory. New Testament texts may be introduced by a phrase such as ‘‘the Lord
said’’ or ‘‘So the Apostle says,’’ without naming the particular gospel or
epistle being quoted; prophetic books of the Old Testament are cited by
name (‘‘Isaiah says’’) sometimes, but the historical books often are quite
unattributed. And the Psalms, perhaps the most frequently quoted of all, are
almost always completely unattributed; even the phrase ut ait in psalmis, ‘‘as
it says in the Psalms,’ is unusual. Augustine was being quite untypical of
himself when, in the Enarrationes, he cites a text as being ‘‘in Regnorum
secundo libro,’’ ‘‘in II Kings.’’73 Occasionally he will introduce a quotation
from a Psalm by in alio psalmo, ‘‘in another Psalm,’’ but usually he simply
quotes a verse in complete confidence that his audience will know it.
Jerome often refers to the Psalms by number, even in order to introduce
a quotation from one. He also uses the less exact (from our point of view)
citational style of Augustine, but, for example, he introduces a quotation
from Psalm 44 by writing, ‘‘Legimus in quadragesimo quarto psalmo,’ ‘‘We
read in Psalm 44,’’74 and one from Psalm 9 by ‘‘de quo in nono psalmo . . .
dicitur,’’ ‘‘concerning which [the devil] it says in the ninth Psalm.’’75 Books
of the Bible, including the Pauline epistles and historical books of the Old
Testament are often cited by name, unlike Augustine’s practice. Jerome
does this not only in his commentaries, but also in his letters. For
Elementary memory design
125
instance, in his letter (no. 53) to Paulinus of Nola on the study of Scripture,
he introduces a quotation of Psalm 118 by its number.76 Jerome also
numbered and referred by number to the book-divisions of his own
works; thus in the introduction to Book 8 (octavus liber) of his
Commentary on Isaiah, he refers to ‘‘books six and seven above.’’77 Such
variations in citational style continue through the early Middle Ages; Bede,
for example, in his treatise on the figures and tropes of Scripture refers to
the Psalms by number. But Carolingian manuscripts are the first written
books to indicate occasionally in the margin the name of the book from
which a quotation is cited, although such information was frequently not
copied by later scribes.78
It would seem from early references to Psalms by number, so at odds
with the citational conventions then prevalent, that one should not pre-
sume that the form of a citational style limits the abilities of scholars and
their audiences to conceive of and designate texts in other ways. It is
obvious from the works of these fourth-century Christians that the
Psalms not only had a fixed order, but that they were taught and memo-
rized by their number in that order. The earliest monastic rules, including
Benedict’s, discuss the order of the Psalms for divine office by referring to
their numbers (see chapters 9 and 18 of the Benedictine Rule). Indeed, the
method described by Quintilian of dividing a long text into short sections,
memorized seriatim or per ordine, lends itself to just such a method of
memorial storage, and suggests that the numerical grid principle is very
ancient indeed. Quintilian’s advice itself implies numbering the segments
produced by divisio of a long text, so that one can join the second piece to
the first, and so on (Inst. orat., XI. ii. 37).
There is also clear evidence that the verse divisions of each Psalm were
numerically designated as early as the time of Augustine. The word versus is
used by Augustine to refer to the divisions within the Old Latin text
(though it was Cassiodorus who established the divisional scheme used
in the medieval text of the Psalms)79 and these verses have numbers. So, in
his commentary on Psalm 6 (referred to as ‘‘sextus psalmus’’ in his com-
ments on Psalm 11’s titulus), 80 he speaks of the first verse and second verse
of the Psalm. These verses are somewhat different from those similarly
designated in the Vulgate text. Though the divisions themselves differ,
clearly the principle of dividing by verses was known. These divisions were
imposed on the text only in memory. They are not marked in the early
Christian codices we possess – indeed even word divisions are not indi-
cated.81 It thus appears that versions of the numerical grid system described
by Hugh of St. Victor for memorizing and mentally concording the
126
The Book of Memory
Scriptural text were taught commonly at least by the fourth century, and
probably were applied to texts well before then.
Alcuin regards the numerical order of the Psalms as
so ancient and fixed
that he indulges in some exegesis regarding the symbolism of a particular
Psalm’s place in that order. In his commentary on Psalm 118, he says that he
has reckoned the patterns worthy of comment (‘‘eruendas rationes’’) for all
the numbers in the Psalms, ‘‘that is, why are all the penitential Psalms
consecrated in a number containing sevens? and why is Psalm 118 divided
into twenty-two sections, of which each one has eight verses? or what is the
reason for there being fifteen Psalms designated in their titles ‘a song of
degrees’?’’82 Invoking the text from Wisdom (11:21) that God created the
universe (including Scripture) by number, order, and measure, he proceeds
to analyze the significance of the numbers in the format of Psalm 118. For
our purposes, however, what is significant is the simple but profound fact
that Alcuin thought of the text as formatted in terms of a numerical grid,
twenty-two sections of eight verses each, and that he also thought that the
Psalm, which itself had a number in the order of the whole book (118),
incorporated subsets of numbered divisions in sections and verses. This is
the mnemonic format which Hugh of St. Victor taught in 1130, and is in
the tradition of numerically designated divisions which we find in the
fourth century.
But the first Bibles to write out the complete format of chapter and verse
are mid-sixteenth-century, and virtually no medieval scribes wrote in the
numbers of the verses in any Scriptural book; chapter numbers were copied
in routinely only beginning in the thirteenth century.83
There is, in other words, a lag of well over a millennium between the
time the numerical grid was certainly in common use for dividing
Scripture, and its first complete appearance on a physical page. It is clear
in this instance that people had laid out the Bible on a grid in their
memories for over 1,000 years before they bothered to express that grid
in writing, and for at least 400 years before they thought it important even
to suggest it in their scholarly citational habits.
In his treatise on sermon composition, Thomas of Waleys makes a
comment that is illuminating of this whole matter. Advising preachers
how to cite the Bible and patristic authorities during their sermons,
The Book of Memory Page 23
