essential human competencies (like the ability to add) rather than techni-
ques (like digital computation) or technologies (like silicon chips). As
M. T. Clanchy says of literacy, ‘‘It has different effects [on human societies]
according to circumstances and is not a civilizing force in itself ’ (my
emphasis).62 The ancients began from the twin assumptions that the
mind already writes when it stores up its experience in representations,
and, as a corollary, that the graphic expression of such representations is
not an event of particular importance, at least for ‘‘ways of thinking about
things’’ – no more important than the sound of an individual’s voice is to
his or her ability to use language. From this viewpoint, the symbolic
representations that we call writing are no more than cues or triggers for
the memorial representations, also symbolic, upon which human cognition
is based. And to mistake one sort of thing for the other would be a
significant error. Writing something down cannot change in any signifi-
cant way our mental representation of it, for it is the mental representation
that gives birth to the written form, not vice versa.
From this viewpoint also anything that encodes information in order to
stimulate the memory to store or retrieve information is ‘‘writing,’’ whether
it be alphabet, hieroglyph, ideogram, American Indian picture-writing, or
Inca knot-writing. Writing is as fundamental to language as is speaking.
We still habitually use ‘‘he said’’ to mean ‘‘he wrote’’; though this idiom has
been adduced as evidence of deeply buried ‘‘aural residue,’’ it can equally
well be interpreted as an acknowledgment that both writing and speaking
are expressions of a more fundamental human competency.
Clearly various societies have felt variously a need to put systems of
mental representation and organization down on some surface, but the
impulse to do this, and the preserved form it may take, has more to do with
the complexity of their social organization, the other groups with which
they come into contact, the nature of materials used and their accessibility,
than with the way in which a human being is able to form and organize
mental representations for cognition, and to understand that they are
representations (i.e. they ‘‘stand for’’ something). I will later discuss a
case wherein a lag of well over a millennium demonstrably exists between
the common use of a particular scheme of mental organization and its first
Models for the memory
37
appearance in written form. Similarly, neither the prevalence nor the form
of written materials in a culture should, I think, be taken as any sure
indication of those people’s ability to think in rational categories, or of the
structures those categories may take. I am not suggesting that technique
and technology have no effect upon human culture; this study is concerned
to identify and describe a number of distinctive features in medieval literary
culture which are sometimes expressed in particular techniques, such as
page layout. But I try not to reify technique, and in particular I think it very
important to recognize that the form in which information is presented to
the mind does not necessarily constrain the way in which such information
is encoded by the brain nor the ways in which it can be found and sorted.
The three are distinct. 63 Classical and medieval philosophers recognized
this when they said that all information, whatever its source or form,
becomes a phantasm in the brain. And those phantasmata are retrieved
by heuristic schemes that need bear little resemblance to the form in which
the information was originally received.
The idea that language is oral, that writing is not a fundamental part of
it, is a modern one, enunciated by linguists like de Saussure and Leonard
Bloomfield, and then generalized into a theory of culture and even of
epistemology by the French Structuralists, chiefly Claude Le´vi-Strauss. It
has become a canon in this theory that Western culture can be divided into
periods characterized by ‘‘pre-’’ and ‘‘post-Gutenberg man [sic],’’ and the
dividing line is marked by a ‘‘veering toward the visual,’’ to use Walter
Ong’s careening image. 64 My study will make it clear that from earliest
times medieval educators had as visual and spatial an idea of locus as any
Ramist had, which they inherited continuously from antiquity, and indeed
that concern for the lay-out of memory governed much in medieval
education designed to aid the mind in forming and maintaining heuristic
formats that are both spatial and visualizable.
T H E S A U R U S S A P I E N T I A E
The second major metaphor used in ancient and medieval times for the
educated memory was that of thesaurus, ‘‘storage-room,’’ ‘‘treasury,’’ and
‘‘strongbox.’’ Whereas the metaphor of the seal-in-wax or written tablets
was a model for the process of making the memorial phantasm and storing
it in a place in the memory, this second metaphor refers both to the
contents of such a memory and to its internal organization. An important
ancient version of the storage-room metaphor occurs in Plato’s Theaetetus,
when Socrates, explaining how one is able to recall particular pieces of
38
The Book of Memory
information, likens the things stored in memory to domestic pigeons
housed in a pigeon-coop. This occurrence attests to the antiquity of the
store-house; indeed, these metaphors, equally visual, equally spatial, seem
to be equally ancient as well. But we should also take careful note of Plato’s
metaphor for memory’s contents, namely domesticated birds. Pigeons have
two salient characteristics for this model: they are raised for nourishment,
and they naturally fly away. In his Institutiones, Cassiodorus uses a signifi-
cant variant of Plato’s pigeon-coop. He describes the structured memory,
this mirabile genus operis or ‘‘remarkable piece of work,’’ as a kind of
inventoried set of coops or animal-pens. One should think of each stall
or coop as a labeled ‘‘topic’’ or place, which not only keeps things, but
informs the otherwise untrammeled brain, directs it, and makes it useful, in
this case for the task of thinking.
Although the metaphor of animal-coops appears to be the same as the
now-maligned filing-cabinet model for human memory, the reference of
this pre-modern metaphor is considerably different, in a way that pro-
foundly affects how we should understand it historically. The ‘‘filing-
cabinet’’ model in Modernist psychology refers to the mind’s success in
finding unaltered, unculled material in memory, material that, like stored
documents, remains unchangingly complete and accurate. In other words,
the modern metaphor concerns memory’s ability wholly to recapture a
past, complete and unaltered, for its own sake. But the purpose of the
metaphor in antiquity and the Middle Ages is not that. Treasuries and
book-chests are not like twentieth-century filing-cabinets. They contain
‘‘
riches,’’ not documents. And their contents are valued for their richness in
terms of their present usefulness, not their ‘‘accuracy’’ or their certification
of ‘‘what really happened.’’ When a medieval abbot wanted to authenticate
the charters of his foundation, he sought out a written document in his
monastery’s library (or he forged one), he did not search for it in the
scrinium of his well-structured memory.65 This is a crucial difference
between the pre-modern and modern notions of the goals of human
memoria. Medieval abbots also cared deeply about authenticity, especially
when the legal standing of their monasteries was at stake, but their own
memory training served a different purpose, and its contents were imag-
ined as alive (animals and birds) or as materials to be used richly in the
commerce of creative thought (coins, jewels, foods).
In the medieval metaphor, memory’s storage structures allow the mind
its truly ‘‘liberal’’ essence, just as training a horse enables the animal to
function at peak ability, to become what its nature intends. As Cassiodorus
says, the inventory of topica memoriae at once structures and channels the
Models for the memory
39
untrammeled intellect (‘‘conclusit liberum ac voluntarium intellectum’’),
because whatever notions (‘‘cogitationes’’) the mind may have entered
upon (‘‘intraverit’’), the intelligence (‘‘humanum ingenium’’) drops into
(‘‘cadat’’) one or another of these places (‘‘in aliquid eorum’’), and having
done this, then coherent thoughts can be formed out of the mobility and
variety of the mind’s experiences.66
The relationship which Cassiodorus conceives between the structures
and their content is quite clear in this passage. Of first importance is the
comprehensive form, laid out in topica memoriae, places of memory
appropriate to rhetoric, dialectic, poetry, and jurisprudence – this structure
(the trained memory) sorts and gathers the knowledge a person has gained.
One is not born with this structure, nor is it passively gained; one con-
structs it oneself during one’s education. And whatever experiences one has
will be channeled by this previously laid-out inventory, and will find their
appropriate place, each contributing its matter to the general store.
Without the sorting structure, there is no invention, no inventory, no
experience, and therefore no knowledge – there is only a useless heap, what
is sometimes called silva, a pathless forest of chaotic material. Memory
without conscious design is like an uncatalogued library, a useless contra-
diction in terms. For human memory should be most like a library of texts,
made accessible and useful through various consciously applied heuristic
schemes. St. Jerome wrote of a gifted young scholar that ‘‘by careful reading
and daily meditation his heart [i.e. by means of his memory] constructed a
library for Christ.’’67 In order to understand something, we must first have
a place to put it, something to attach it to in the inventory of all our
previous experience.
Modern scholars usually translate cogitatio as ‘‘thought,’’ but this con-
ceals a crucial difference in how pre-moderns conceived of what that is
from how we conceive of it. Cassiodorus says literally that ‘‘the mind enters
into thoughts’’; a modern would much more likely say ‘‘the mind thinks.’’
Cogitatio (con þ agito, ‘‘move, rouse’’) is defined in rhetoric (and in Greco-
Arabic somatic psychology) as a combinative or compositional activity of
the mind. It necessarily uses memory because it combines imagines from
memory’s store. One should therefore think of a single cogitatio or
‘‘thought’’ as a small-scale composition, a bringing-together (con þ pono)
of various pieces (as phantasmata) from one’s inventory. The topica provide
both content and structure for these cogitationes. Aristotle – and all succeed-
ing writers – distinguishes between ‘‘general’’ topica and ‘‘particular’’ ones;
basically, in this context, the general topics are structuring devices solely,
such as comparison, contrast, and the like, while the particular topics are
40
The Book of Memory
the collected expressions of an ethico-political concept, such as mercy,
wisdom, justice, and so on. A commonplace is described as being ‘‘made’’
or ‘‘manufactured’’ – Augustine says that Simplicius, when asked for a locus
communis on some subject, would ‘‘make’’ it (‘‘de quocumque loco volui-
mus . . . fecit’’). One does not simply parrot forth some previously recorded
dictum word for word by rote, but builds a ‘‘topic’’ or ‘‘commonplace’’ out
of materials from one’s memorial inventory.
The topica provide the construction materials for thoughts, for whether
one takes a simple Aristotelian position that knowledge is composed of
experience constructed from many memories, or whether one shares with
the (Neo)Platonists the belief that knowledge is a process of remembering
the imprints of hypostatized Ideas, thoughts cannot be made without the
materials in memory. For whatever memory holds occupies a topos or place,
by the very nature of what it is, and these topica, like bins in a storehouse,
have both contents and structure. Every topic is in this sense a mnemonic, a
structure of memory for recollection.
The image of the memorial store-house is a rich model of pre-modern
mnemonic practice. It takes a number of related forms and gives rise to
several allied metaphors for the activity of an educated mind, but all center
upon the notion of a designed memory as the inventory of all experiential
knowledge, and especially of those truths of ethics, polity, and law, which
are copious and rich in their very nature. Synonyms of thesaurus which
are also used for the memory include cella or cellula, arca, sacculus,
scrinium, and the Middle English word male (carry-sack), used perhaps
most famously in Harry Bailly’s comment to the company in The
Canterbury Tales just after the Knight has spoken: ‘‘This gooth aright;
unbokeled is the male!’’ It occurs again, of course, at the end of the tale-
telling, when Harry Bailly cautions the Parson to ‘‘unbokele and shew us
what is in thy male.’’68
Zeno the Stoic (fourth–third century BC) defines memory as th¯esaurisma
phantasio¯n or ‘‘storehouse of mental images.’’69 Thesaurus is used meta-
phorically in both Romans (2:5) and the gospel of Matthew (6:19–20) in
the sense of storing up intangible things for salvation, the Greek being
translated by Jerome as a verb, thesaurizare, thus: ‘‘Nolite thesaurizare vobis
thesauros in terra . . . Thesaurizate autem vobis thesauros in caelo.’’70 The
Rhetorica ad Herennium calls memory the treasurehouse of invented things,
‘‘thesauru[s] inventorum’’ (III, 16), referring particularly to a memory
trained by the artificial scheme which the author proceeds to recommend.
Quintilian, also recommending a cultivated memory, calls it ‘‘thesaurus
eloquentiae�
��’ (XI, 2, 2).
Models for the memory
41
But these metaphorical uses, though they clearly demonstrate the per-
vasiveness and antiquity of the link between thesaurus and a trained
memory, do not help us know how a medieval student, encountering the
word, might visualize the object denoted. Thesaurus refers both to what is
in the strongbox, the ‘‘treasures,’’ as when Augustine speaks of the treasures
of countless images in his memory, ‘‘ubi sunt thesauri innumerabilium
imaginum’’ (Conf., X, 8), and to the strongbox itself. When the wise men
kneel before Jesus in Matthew’s account (2:11), they bring out their offer-
ings of gold, frankincense, and myrrh from opened thesauri: ‘‘et apertis
thesauris suis obtulerunt ei munera, aurum, thus, et myrrham.’’71 The
thesauri of the wise men are portable strongboxes such as a merchant
might carry.
Cella, the word used by Geoffrey of Vinsauf for the memory (which he
calls a cellula deliciarum) also means ‘‘storeroom,’’ as indeed its derivative
form, cellarium, English ‘‘cellar,’’ still indicates. When Chaucer’s Monk
threatens to tell numerous tragedies ‘‘Of whiche I have an hundred in my
celle’’ (VII. 1972), he is more likely using the word to refer to his memory
‘‘cell’’ or store than to a cell in his monastery. After all, a hundred tragedies
housed in books back home are not going to do him any good when having
to tell a tale or two on the road to Canterbury, but stories in his memory-
cell are a different matter. Therefore, though virtually every editor of
Chaucer has thought the Middle English word needed no gloss, it does:
the Monk is saying that he has a hundred stories tucked away in his
memory. The gloss is underscored by the Monk’s mention two lines later
of the ‘‘memorie’’ of tragedies made to us by ‘‘olde bookes.’’
But the Latin word cella has a number of more specialized applications
that link it complexly to several other common metaphors for both the
stored memory and the study of books, as well as words like arca and
thesaurus. Cellae are stalls or nesting-places for domestic animals and birds,
and, by a transference of meaning, small rooms or huts for people (whence
derives the word’s monastic usage, invoked by Hugh of St. Victor as a
The Book of Memory Page 9