sort of memorandum of his composition at a particular stage, which he
might reconstruct or revise almost continually, as he worked to perfect his
res. In this sense, the modern notion of a finished work is quite foreign to
medieval authorship. Authors would issue versions of a work which they
still intended to perfect to scribes for copying onto parchment, perhaps in
an effort to secure them, as Anselm wanted to do. Chaucer’s Troilus and
Criseyde circulated in at least one shorter version than the one we now
possess, as its editor, R. K. Root, demonstrated long ago. A note by the
scribe of the St. John’s College, Cambridge, manuscript, before several
blank lines on one folio page, indicates that ‘‘her faileth thyng þt is nat yt
made,’’ ‘‘here is lacking something that is not yet composed.’’77 And the
two versions, F and G, of the Prologue to The Legend of Good Women
clearly represent different publications of the poem. The visions and
revisions of Piers Plowman were once thought to be so unusual as to
constitute prima facie evidence that multiple authors were responsible,
but in fact the phenomenon is not uncommon in kind, though perhaps
Langland was unusual in the degree of his revisions.78 Petrarch habitually
continued to revise autograph copies of his works after he had sent a version
off for transcription and circulation; this is true of both his poetry and
prose works. 79 And the revising process was not limited to the first author.
Readers, in the course of familiarizing a text, became its authors too. No
modern reader would think of adapting and adding to the work of some-
one else in the way that medieval readers freely did, sometimes indicating
the difference by writing their own work in margins, but often not. The
results concern every editor of a medieval text.
Perhaps because of the familiarity of their language, readers seem
especially to have recorded in vernacular texts the ways in which they had
made them their own, by adding passages, incorporating comments,
respecting the res but not necessarily the verba in the manner in which
memoria expands the words during meditation. The more amateur a
production (one can really only use such a word after the thirteenth
century), the more apt one is to find this, since these were made for use
by families or small coteries. Among English works, the manuscripts of
Piers Plowman are particularly notorious. Unlike the professional scribe,
whose job was to copy not to read, an amateur writing for him- or herself
Memory and authority
265
was writing as a reader, whose task was to make one’s own what one finds in
texts. Rather than condemning them for this, we should understand that
such wholesale commentary is a form of compliment, a readerly contribu-
tion to the text’s continuation, and a judgment that it is worthy to be a
public source for memoria.80
G L O S S I N G : M A K I N G A T E X T A N A U T H O R
I have spoken of the medieval book as a support for the various activities of
memoria – no format shows better its compositional, meditational charac-
ter and catena structure than the lay-out of the glossed books, which
developed during the course of the twelfth century in France, particularly
in Paris. They have been the subject of a penetrating study by Christopher
de Hamel, who has demonstrated how they came into being soon after the
compilation of the line-by-line Biblical commentary known as the glossa
ordinaria.81 This standard commentary was ‘‘simply a practical aid for
students beginning on their study of the Bible,’’ as G. R. Evans calls it,
‘‘distilling out the essence of the work of previous centuries’’ to present ‘‘a
manageable and reliable’’ introduction in a brief, summary, and ad res form
that was keyed specifically to the Bible text itself and did not try to address
large questions or explore difficult matters.82
The glossed book’s lay-out, difficult to set up and copy, was reserved for
works that were among the most fully institutionalized (authorized in the
sense I have been developing in this chapter) – the Decretals and compen-
dious books of canon law, the Bible, especially the Psalms and the Pauline
Epistles, and (by the late fourteenth century) the works of certain classical
authors. The lay-out presents graphically the process of this authorization,
for the compiled comments are written all around the author-text, keyed
into it, catena fashion, via red underlinings, heuristic symbols, and other
punctuation (in one early lay-out the Biblical words being commented on
were written out in red and the commentary in ordinary dark ink). 83 They
are also among the most fully decorated of books – and they are those
which must be fully memorized. Catena is a mnemonic lay-out, elementary
in memory training, in which the source text itself serves as the ordered set
of backgrounds into which material is keyed. The complete format was
developed for the commentary compiled by Peter Lombard for the Psalms
and the Epistles, a revision of the older compilation by Anselm of Laon. It
is possible that Peter Lombard himself devised the format, leaving it to
professional scribes to carry out; this would not be unusual for a master to
do, though the format is extremely complex in execution – one thinks of
Herbert
From
commissioned
who
glosses.
the
Becket,
of
.
one
Thomas
1177
to
and
corrects’
1164
advisor
and
between
‘‘Augustinus’
of
secretary
France
figure
was
The
northern
r.
Bosham
in
147
v–
out
146
fos.
Lombard.
carried
,.45 PeterB.of project,
MS.
gloss
this
the
College
with
Trinity
Psalter,
Cambridge,. Bosham’s
7
of
Memory and authority
267
Hugh of St. Victor’s design for his Chronicle, and of Peter of Poitiers’s
diagrammatic Genealogia, both also twelfth-century French products.84
The source-text is written in the center of the page on alternate lines in
the large, careful hand known as textualis formata, and the commentary,
hooked into it via key-words and phrases, is written in a smaller hand
around it. One immediately sees the textual relationship of source to
commentary; one is also provided with a mnemonically useful image to
help to place the commentary safely in one’s memory, catena-style. Recall
how Hugh of St. Victor advises his students to pay close attention to the
shapes of the letters and the colors on the page in order to fix a memorial
present memorable variations of letters (the different hands) and colors,
for each page is unique. These different hands became conventionally used
for these different kinds of text; the large hand developed into fully formed
Gothic script; the smaller, squatter hand was used for commentary, even in
books that did not reproduce a source-text. Clearly, they were used to form
a visual cue to the sort of text with which one was dealing. 85
One of the best-written of the glossed books of Psalms was made for
Herbert of Bosham, chaplain, confidant, and biographer of Thomas
Becket, probably in the late 1160s. 86 It was made in two large books, so
extensive was the project; it presents the texts of both of Jerome’s trans-
lations of the Psalms, the Gallican (based on the Greek translation of the
Septuagint) and the Hebrew (based directly on the Hebrew text), with the
commentary of Peter Lombard, which consists largely of excerpts from
the major patristic commentators. The first volume (see figure 7) is now in
the library of Trinity College, Cambridge (MS. B. 5. 4) and the second in
Oxford (Bodleian Library, MS. Auct. E inf. 6). 87 One of their more original
features is the use of painted figures to help fix the page as a mnemonically
functional visual image. These figures usually inhabit the outermost margins
of the page. The gloss itself is carefully annotated, with the sources of the
commentary identified (Jerome, Augustine, Cassiodorus) and a system of
patterned dots is established as a unique additional signature for each of the
main sources (such indexing patterns seem to be quite common in glossed
lay-outs; one thinks of the use of similar systematic notae for concording
purposes which we see in Grosseteste’s books, for example). The painted
figures are a part of this apparatus: so, a bearded man labelled Augustinus (or
Cassiodorus or Hieronimus) points a javelin at the commentary text. He
holds a banderole on which is written a warning such as ‘ Ego non probo’’ or
‘ Hic michi caveat’ or, as in figure 7, ‘‘Non ego’’; it is significant that these
figures mark where Peter Lombard misidentified or miscopied extracts, or
268
The Book of Memory
where there is other disagreement. In addition to these figures, several of
the Psalms have emblematic pictures painted next to their opening words;
unlike the citational figures, these can occur in the inner margin where the
gloss itself is written, as well as in the outermost one, suggesting that they
too were considered essential in the gloss, and acted as markers for these
particular Psalms. Unfortunately, both manuscripts were badly mutilated
by someone who cut many such figures out of their pages, so that whatever
system there may have been is now lost; it is apparent, however, that certain
of these figures were deliberately repeated. A figure of ‘‘Ethica’’ holding a
dove and a serpent, for instance, appears at the beginning of Psalms 4 and
11; the figure of a soul in flames, with the legend ‘‘Homo in igne, Deus in
homine’’ is repeated at the beginning of Psalms 2, 8, and 81.
This textual format, serving the memorial lay-out of catena, is thus an
applied mnemonic containing numerous visual helps to memory in its
features, and also laying out graphically the relationship of the auctor and
all its progeny, including their disagreements. The way in which the
commentary is woven together around the auctor illustrates how the
authority is understood as source-word rather than as final-word, the way
we tend to understand it now. The pages shown in figure 7 were intended
for study and meditation – they were not to be used by beginning readers
directly but by their teachers. Moreover, Herbert of Bosham’s books are
not a single work but a whole library of materials, an encyclopedia of
related knowledge gathered into its pages. Though drawn on a flat, two-
dimensional page, each place occupied by a chunk of Psalm-text with its
commentaries and glosses, it is conceived like a cube and used like a box or
little room. The structure as a whole is thus composed of cells, like those in
a beehive. Hugh of St Victor anticipated such a structure in his discussion
of how to memorize the psalms (Appendix A), when he describes how one
should conceive the basic linea, or diagram plan, as a set of compartments
or boxes for associated materials, like the money-boxes used to store coins
for rapid recovery. In a properly designed memory, just as on these pages,
the verses of the source will be like a line with many hooks on it, and as one
pulls in one part of it, all the fish will come along. To pull in one text is to
pull all the commentary, as well as other texts concording with it. Source,
glosses, citations, punctuation, and decoration are all married up together
in a single memorial image which constitutes the text; one cannot mean-
ingfully talk for long about one of these strands in isolation from the
others, for that is not how they were perceived. They all serve memoria
practically (they help to form the heuristics of order, unique address,
division, and composition needed for safe storage and retrieval) and they
Memory and authority
269
also image its institutional nature. A work of literature was not taught
in isolation, as an artifact produced by some person long dead whose
particular intention we must now recover, but as an ever-rolling stream
accumulating and adapting over time as it is collated with its multitude
of readers. Collatio means all these things: bringing together of texts,
conversations about texts, and feeding upon texts as one feeds at a com-
munity meal.
The glossed format remained very popular, especially for university use.
A copy of the Gregorian Decretals with two distinct levels of identified
commentary in it (law commentary was usually signed with the initials of
the master responsible for it) plus additional comments in other hands
shows in a wonderfully visual manner how the source-text has been wound
about with generations of commentary. In this manuscript (Bodleian
Library, MS. Lat. th. b 4, originally completed in July of 1241 at Bologna),
the originating text is written in a large formal hand in two columns down
the center of the page, with the commentary about it on all four sides
(see figure 10).88 Margins have been left between the two parts of the book
and between the two columns of text; these contain ink decorations and
comic drawings, such as a fox chasing a rooster (fo. 37) and a cat with a
mouse (fo. 84v), or, as shown in figure 8, dogs and men hunting a savage
boar. But these margins have themselves been written over with other
commentary (though the drawings are usually left alone); in addition
there are interlinear glosses in the main text. So the page presents a text
that consists of a great many margins, those margins traditionally left for
the memorial activities of readers over time. And these responses form an
integral part of the public presentation of the text; indeed, in this manu-
<
br /> script, the originating text seems almost overwhelmed by its margins, as is
perhaps suitable for a legal book. In earlier glossed books, scribes ruled an
extra margin for current readers when the complex of written text and gloss
threatened to crowd out the blank space of the page. Blank space is needed
both to ensure a clean, crisp image of the text (for turba, ‘‘crowding,’’ is a
great enemy of memory) and to allow one to make one’s own notulae. 89
There are manuscripts in which the commentary has been written out in
the margins, while blank space has been left for the main text, unfinished
by the scribes. One such is Newberry Library MS. 31.1, intended for a text
of Fulgentius on the pagan gods, with commentary by Pierre Bersuire (this
text was the standard beginning of Bersuire’s Ovide moraliseé).90 It is an
odd way for a manuscript to be left, for usually the source-text was written
first with the commentary fitted in. This manuscript was used despite what
to us would be an insurmountable handicap, presumably by those who had
270
The Book of Memory
8. University of Oxford, Bodleian Library, MS. Lat.th.b.4, fo. 23v. Fully glossed page
from a Decretals written in Bologna, 1241. Notice how various readers have expanded
the number of margins with their own glosses. In the center margin are drawings of men
and dogs hunting a boar.
Memory and authority
271
the source so well memorized that they attached the commentary to it
mentally. In this regard, it is significant, I think, that this empty page space
has been respected by readers, whose own notes are written in the expected
place, outside the margins of the commentary, as though the unwritten
space was also required in some way for the making of a proper memory-
image keyed to the mentally supplied text – perhaps as a surface onto which
textual cues could be projected in memory. In doing this they were
following what had been standard before the invention of the glossed
lay-out, when line by line commentary on the Bible text was written out
continuously, without all the keys supplied in a glossed book, and a reader
had to supply the appropriate Bible text from the grid in his memory.
So important is the acquiring of commentary to establish an auctor, that
at least one medieval vernacular writer supplied an exemplar of his work in
The Book of Memory Page 48
