the litterae of writing are representations both of a voice and of words
236
The Book of Memory
spoken. But the words (voces) in turn are signs of things – res, the concept
one keeps coming back to in all pre-modern discussions of rhetoric,
language, and the role of memoria.
The word auctor was thought to be derived from the verbs agere, ‘‘to act,’’
and augere, ‘‘to grow.’’ A second word, autor, was related through etymol-
ogy to Greek autenth¯es, ‘‘authentic,’’ but medieval dictionary-makers dis-
tinguished the two words, one spelled with a c and one without, quite
carefully. Auctoritas derived from auctor, and was defined by Hugutio of
Pisa in about 1200 as ‘‘sententia digne imitatione,’’ ‘‘a saying worthy of
imitation.’’ Thus, both authority and author were conceived of entirely in
textual terms, for an author is simply one whose writings are full of
authorities. And an author acquires authority only by virtue of having his
works retained sententialiter in the memories of subsequent generations.3
Author-texts are retained and imitated ad res because it is there, not
in their actual words, that their authority lies; this is the assumption in
Hugutio’s definition. It is related directly, I think, to the memorial dis-
tinction between iteration and imitation, recitare and retinere, memoria
verborum and memoria rerum. Both the word auctor and the later synonym,
originalis, are related closely to the traditional metaphor of literature as a
great river flowing over time from a fons or source. So, when Jerome wrote
of the Bible as originales libri, he meant something like ‘‘originating texts,’’
progenitors of a whole family of textual descendents, especially commen-
taries and other adaptations, which are the indication, or authorization, of
a work’s institutional standing in the public, communal memory. 4
Let us look again at Petrarch’s comment, through the mouth of Augustine,
concerning the meaning he finds in the Cave of the Winds passage in the
Aeneid, for Petrarch a major auctor:
I cannot but applaud that meaning which I understand you find hidden in the
poet’s story, familiar as it is to you; for, whether Virgil had this in mind when
writing, or whether without any such idea he only meant to depict a storm at sea
and nothing else, what you have said about the rush of anger and the authority of
reason seems to me expressed with equal wit and truth. 5
Petrarch has responded to this text’s res, amplifying it through first famil-
iarizing it in his memory (by divisio) and then writing it anew in words that
do not reproduce the actual language of the Aeneid, but rather adapt its
sense to Petrarch’s own situation (compositio). Were we to think of this
exegetically, we could say that Petrarch gives these lines a tropological
interpretation. But the point is that his interpretation is not attributed to
any intention of the man, Virgil, but rather to something understood to
Memory and authority
237
reside in the text itself. Authorial intention in itself is given no more weight
than that of any subsequent reader who uses the work in his own meditative
composition; the important intention is within the work itself, as its res, a
cluster of meanings which are only partially revealed in its original state-
ment. Petrarch supports his reading by appealing to some of the words in
this passage that suggest anger and turmoil, but he does not suggest that
these meanings were ever Virgil’s intent – they may or may not have been,
it is unimportant. What keeps such a view of interpretation from being mere
readerly solipsism is precisely the notion of res – the text has sense within it
which is independent of the reader, and which must be amplified, dilated,
and broken-out from its words, as they are processed in one’s memory and
re-presented in recollection. Amplifying is an emotional, image-making
activity as we have seen, and it is just this quality that makes it ethically
profitable. More importantly than growth in knowledge, reading produces
growth in character, through provisioning – in memoria – the virtue of
prudence.
C O N S I D E R T H E B E E S
Composition starts in memorized reading. The commonest way for a
medieval author to depict himself is as a reader of an old book or a listener
to an old story, which he is recalling by retelling. In monastic meditation,
the old book was something in the Bible; later in the Middle Ages, old
books were also secular and vernacular. It is well to consider the common-
place of the reader/author as a bee. This trope came most familiarly to the
Middle Ages in the form given to it by Seneca, whose letters were a
standard item in the medieval curriculum of rhetoric, or ars dictaminis as
it came to be more commonly called during the twelfth century. 6 Seneca
wrote: ‘‘We ought to imitate bees, as they say, which fly about and gather
[from] flowers suitable for making honey, and then arrange and sort into
their cells whatever nectars they have collected.’’7 Composition begins in
reading, culled, gathered, and laid away distinctively in separate places, ‘‘for
such things are better retained if they are kept separate’’; then, using our
own talent and faculties, we blend their variety into one savour which, even
if it is still apparent whence it was taken, will yet be something different
from its source (‘‘ut etiam si apparuerit unde sumptum sit, aliud tamen esse
quam unde sumptum est appareat’’).
I mentioned this variation of the store-house metaphor for memory in
Chapter 1; like the rest of those metaphors, this one too should be under-
stood not as a mere decoration but as a complex model of the process of
238
The Book of Memory
composition and authorship. As a model it is a variant also on that of
digestion, as indeed Seneca realized: ‘‘the food we have eaten, so long as it
retains its original character and floats in our stomachs as a mass, is a
burden; it passes into tissue and blood only when it has been changed from
its original form. So it is with the food which nourishes our mind . . . We
must digest it; otherwise it will only come into our acquired memory-store
[memoria] and not pass on to become part of our own abilities [ingenium].’’
Merely to store memory by reading is an incomplete process without
composition, for composing is the ruminative, digesting process, the
means by which reading is domesticated to ourselves. Indeed the two
tasks require one another – Jerome echoes this same principle when he
says that there is no point to reading if one does not also compose and
write.8
This familiar trope of the bee has been extensively analyzed as a model
for classical, medieval, and Renaissance assumptions concerning the nature
of literary imitation, the relationship of authors to their antecedents, and
the changing way in which these assumptions have been understood.9
I have no desire here to rehearse work which has been so ably and
thoroughly done. But the scholars who hav
e analyzed it most comprehen-
sively have started from the preoccupations of Renaissance writers defining
themselves against a medieval world which they sought to reject (while
often unaware, inevitably, of the extent to which they were themselves still
its products), and in the course of such analyses a particular medieval
understanding of the trope has, I think, been rather slighted and misrep-
resented. To get at this particular medievalness, I must emphasize that, in
this trope, composition, like reading, is assumed to depend on a memory
properly stored with discrete, immediately recoverable loci. For Seneca,
memoria is not an alternative to creativity (which is how I would under-
stand his use of ingenium in this passage) but the route to it. He does not
disparage memoria, but only its undigested, parrot-like use.
Memory is the matrix of all human temporal perception. This too is a
medieval commonplace, nowhere so eloquently explored as in the final
books of Augustine’s Confessions. Memoria makes present that which is no
longer so in actuality; indeed, as we have seen, this temporal understanding
of memorial representation is more emphasized, at least in medieval
analyses, than its mimetic one. Prudence, the ability to make judgments
in a present context about both present and future matters, is founded upon
memoria, and traditionally was represented with three eyes, looking to past,
present, and future. But memory remains, by its nature, of the past – a thing
cannot be in memory until it is past. This insistence is basic in medieval
Memory and authority
239
Aristotelian (and Augustinian) psychology, as I emphasized in Chapter 2.
Therefore, to say that memory is the matrix within which humans perceive
present and future is also to say that both present and future, in human time,
are mediated by the past. But the past, in this analysis, is not something itself,
but rather a memory, a representing of what no longer exists as itself but only
in its memorial traces.
It seems to me that this is quite different from insisting, as Renaissance
and modern scholars have done, that the past is mediated by the present.
The change in emphasis, in the direction of mediation, if you will, is
critical. It seems to be typical of modern (Renaissance) consciousness to
give the past, like other scientific subjects, objective status apart from
present human memories. As a result, perhaps, a Renaissance scholar
worried that the past had been distorted through the mediation of the
present, and sought to recover or resuscitate the dead past itself. In his book
on Renaissance literary imitation, Thomas M. Greene has emphasized how
the language of necrology and revival is woven into those scholars’ anti-
quarian and philological projects.10 By contrast, Augustine journeyed
through his memory not to find his past but to find God, his present
and future. And it is clear that Augustine assumed that the way to God lay
only through the re-presenting of his past in memory; he has no interest in
his past except as it provides him with a way and ground for understanding
his present.11
Few features of medieval scholarship are so distinctive as an utter
indifference to the pastness of the past, to its uniqueness and its integrity
on its own terms, as we now would say. Ordinarily, medieval scholars show
no apparent interest in archaeology or historical philology, and the repre-
sentation of classical or Biblical figures in medieval dress is continually
amusing to modern audiences of their art and literature. Yet it is evident as
well that medieval scholars realized that language and societies had changed
significantly over time; Chaucer and Dante commented famously on the
phenomenon, but the acknowledgment is found often much earlier as well.
Indeed, the division of modern from ancient was first formulated at the
beginning of the Middle Ages. 12 It simply doesn’t seem to have been
thought to be of paramount importance.
This omnitemporality in medieval thought, as Erich Auerbach called it, is
usually attributed to a prevailing belief in the eternity of God and consequent
emphasis upon divine continuity in human history. The dogmatic cause is
surely important, but no more so, I think, than the particular character of the
medieval institution of memoria by means of which texts of past authors are
constantly related in and through present minds (the dual meaning of the
240
The Book of Memory
English verb relate is important here, for it captures both the positioning and
re-speaking of these texts). The sole relic of the ancients with which medieval
scholars vigorously concerned themselves was written texts; this choice itself
is interesting, for there were other artifacts of antiquity still readily visible.
But only the letters which compose texts can speak (for letters are signs of
voices/ words no longer present); only they can be related to the present and
future, and thus become the material of prudence.
The most comprehensive model of the medieval view of what consti-
tuted memoria is the medieval book itself, especially those fully marked up
codices, punctuated and ornamented to the last, precise hair (distincti ad
unguem, to adopt what Ambrose said of his textual labors). As codicologists
speak of paper or parchment or stone as a support for writing, so the book
itself is the chief external support of memoria throughout the Middle Ages.
In its lay-out and ordering, it serves the requirements of readers who
expected to engage it in their own memories. It also often records the
memorial gatherings of a whole community of readers over time, present-
ing in its multiple margins the graphic display of a catena or chain of
comments upon the source-text. The distinctive format of the glossed
book, used especially for law compendia and certain Biblical texts, but
later also for secular authors, is the most satisfying model of authorship and
textual authority which the Middle Ages produced (see figures 10 and 11), as
many scholars have recognized.13 Let me emphasize once more, that a book
does not mimic a memory; its relationship to memoria is not that of a
mirror or copy, any more than letters on parchment mime their contents.
Their relationship is functional; the book supports memoria because it
serves its requirements, some of which are biological, but many of which,
in the memorial cultures of the Middle Ages, were institutional and thus
conventional, social, and ethical.
C O M P O S I N G A W O R K : I N V E N T I O N
But before examining the finished product, the book itself that both results
from and furnishes memoria, it will be helpful in the rest of this chapter to
consider in detail how the act of making a text was thought to proceed, in
order to stress its origins in the activities of memory. I think it will become
clear during my discussion that the terms ‘ oral’ and ‘ written’’ are inadequate
categories for describing what actually went on in traditional composition.
I would propose the term ‘ memorative composition’ instead, and stress its
previous chapter.
Memory and authority
241
It is clear, both from descriptions of pedagogy and from the practices of
individual writers, that much of the process of literary composition was
expected to occur mentally, in mature authors, according to a well-defined
method that had postures, settings, equipment, and products all its own.
The drafts that resulted were designated by different names, which do vary
a bit according to the particular writer, but each of which denotes a fairly
well defined stage of composition. These are, first, invention, taught as a
wholly mental process of searching one’s inventory. It involves recollection
primarily, and occurs with postures and in settings that are also signals of
meditatio; indeed, it is best to think of invention as a meditational activity,
and indeed Quintilian so designates it. This meditation involves both the
discovery and disposition of the subject matters, and it results in a product
called the res, a term familiar also from the pedagogy of memory training.
More complete than what modern students think of as their outline, the res
should, according to Quintilian, be formed fully enough to require no
more than finishing touches of ornamentation and rhythm. In other
words, the res is like the rougher drafts of a composition, with much
room left still for shaping, rearranging, and adjustment.
The post-invention stage involves elaboration, of both form (disposition
of parts) and style. Its products are called dictamen; it might, but need not,
involve writing instruments. As will become clear, the dictamen is most like
what we now call a formal draft; a number of versions, each unfinished,
could be involved in its making. Compositio covers three closely related
activities: formalization, or taking one’s res and giving it final form as a
composed piece; correcting, by both adding and emending, but also by
comparing and adjusting the revisions to make sure the words fit one’s res
in intention and adequacy as much as possible (changing one’s res drasti-
cally at this stage would indicate a lack of proper invention); polishing,
artfully adjusting one’s expression to make it stylistically striking, persua-
The Book of Memory Page 43
