Documente Academic
Documente Profesional
Documente Cultură
brill.com/jim
Mayte Penelas
Escuela de Estudios rabes, csic
mpenelas@eea.csic.es
Abstract
The two extant fragments of the second volume of Ibn ayyns Muqtabiscurrently
preserved in the Real Academia de la Historia of Madrid and in the Qaraouiyine Library
of Fezbelong to the same codex, which was copied by several scribesup to 14
according to our reckoningin cooperation. As a rule, each copyist undertook the
transcription of one of the seventeen gatherings the codex is made up of; in other
words, the changing of hand normally coincides with the changing of quire. There
are, however, exceptions to this rule. This paper posits that the irregularities of the
codex, as well as its characteristic features, are the result of teamwork, including the
accommodation of the share of text allotted to each scribe into the available space.
Keywords
General Remarks
tents of the manuscript, but rather with the process of it being written. In this
regard, the first noticeable fact is that it was made by several scribes in coop-
eration. As a rule, each of them undertook the transcription of a gathering, but
there are a number of irregularities that will be examined later. Before delving
further into the specific subject of this paper, a more detailed description of the
codex is necessary.
The codexon watermarked paperis currently made up of 196 or 197
leaves,6 numbered with ghubr numerals from 88 to 284. In addition, there are
remains of another foliation, also in ghubr numerals, which was crossed out.
The fragment of Madrid contains 101 leaves (ff. 88188), and that of Fez, the
remaining 95 or 96 leaves (ff. 189284). The latter are loose and laminated.
As regards the foliation, a couple of further remarks must be made: on the
one hand, figure 203 appears in two different leaves, which is the reason why
Makk left out f. 203bis/mukarrar7 (bis) from his edition of M2c;8 on the other
hand, no leaf is numbered with figure 266 (either the leaf is lost or the person
who made the foliation skipped the figure by mistake).
The fragment of Fez has a third foliationin pencil and Arabic numerals
from 1 to 96that was made after both fragments had been separated. Assum-
ing that it did once exist, the fact that there is no jump in the numeration would
suggest that f. 266 was already missing when that modern foliation was made.
By contrast, f. 252 had not yet disappeared. This leaf was lost (or maybe not)
at an indeterminate moment between 2008 and 2009. In 2008, we saw it in Fez
and, unlike all the other leaves, it was not laminated; this led Hassan Harnane
curator of manuscripts of the Qaraouiyine Library at the timeto believe that
it did not belong to the same codex. We informed him that this was a misjudge-
ment, but by the time we returned to Fez the following year, the leaf had already
disappeared. It has not reappeared to our knowledge, but we cannot be certain
of its status. Therefore, we cannot tell whether 196 or 197 leaves of the second
volume of Ibn ayyns Muqtabis survive, of which 95 or 96 leaves are in Fez.
For convenience, in this paper we assume that f. 252 is not lost, even though
the truth may prove otherwise.
account of Hishm is rule (172/788180/796). As for the rest of Muammad is, it would be
included in the third volume.
6 The reason for this indefiniteness will be explained below.
7 The word mukarrar was written below the figure with pencil by a hand much later than that
which made the foliation.
8 F. 203bis and two more pages unedited by Makk (owing to reasons that are not relevant here)
have been edited recently by Luis Molina and Mayte Penelas, Dos fragmentos inditos del
volumen ii del Muqtabis de Ibn ayyn, Al-Qanara 32 (2011), pp. 229241.
With regard to the page setting, each page has a consistent 29 lines, the
written area is 222mm high by 145mm wide, and the interlinear space measures
8 mm,9 with an exception that will be mentioned later. These figures obviously
apply to pages with no titles and to those whose written area is completely
preserved; that is to say, up to f. 196. From f. 197 onwards, the lower part of the
leaves is mutilated, so that an unequal number (ranging from three to six) of
final lines per page is missing.
The justification and the ruling are the same for all but 39 of the leaves:
folio 206, for reasons that will be analyzed later, and folios 237 to 275 (not
counting f. 266), which do not belong to the same codicological unit. The
latter were apparently included later to fill a gap, thereby producing a factitious
codex. Even if none of the added leaves is completely preserved, we can tell
that their written area is smaller (140mm wide) and the interlinear space is
larger (approx. 9mm) than those of the base codex. Accordingly, each page
must have had considerably fewer lines than the pages belonging to the base
codex. In fact, although the number of completely or almost completely legible
lines per page varies between 17 and 19, on the basis of textual evidence we have
been able to determine with almost complete certainty that the added leaves
originally had 23 lines per page,10 in contrast to the 29 lines of the base codex.
[Fig. 1]
Henceforth, we shall refer only to the 159 leaves that are part of the base
codex, not to those factitiously added. Those 159 leaves were arranged in 17
quires, 16 of which are quinions (i.e. ten folios) and one is a binion (four folios),
and were distributed as follows:
9 All the pages are identical in terms of the written area and the ruling; this leads us to think
that a misara was used.
10 See Molina and Penelas, Dos fragmentos inditos, p. 232, n. 4.
figure 1 Ibn ayyn, Muqtabis, ms. Fez, Qaraouyine Library, f. 237r (added fragment) /
f. 236v (quire 16, copyist g)
The 17 gatherings that compose the extant partial copy of the second volume of
Ibn ayyns Muqtabis were produced by more than ten scribes. The changing
of hand is at times observable with the naked eye, while at other times a
rather more thorough examination of the scripts is necessary. Furthermore,
the changing of hand usually coincides with the changing of quire. In other
words, each scribe undertook the transcription of one of the gatherings that
make up the volume. This practice was frequent in the medieval monastic
scriptoria: Before being bound, the book was divided into gatherings; these
were then distributed among several copyists, each of whom transcribed his
stint.11 A distinction must be made between this method and the pecia system
that was used in several European universities between the thirteenth and
fifteenth centuries.12 The difference between the successive pecia system and
the distributive method of the medieval scriptoria is very clear:13 The successive
system allowed the simultaneous production of several copies of a book each
made by one scribe, whereas with the distributive method only one copy was
produced by several scribes in a short period. Obviously, the latter would be the
case for our codex.
According to our examination of the diverse handwritings, up to 14 people
collaborated on the transcription of M2b and M2c (copyists a to m, plus s). We
are not going to describe in detail the peculiarities of the respective scripts as
that would be excessively lengthy; suffice it to say, some of the handwritings
are neat, others careless, some are smaller than others, some are clearer, etc.
Furthermore, each copyist employed his own scribal techniques: some of them
figure 2 Ibn ayyn, Muqtabis, ms. Madrid, rah 11/9370, f. 145r (quire 7, copyist g) / f. 144v
(quire 6, copyist f)
vocalized abundantly, others less so, others did nothing; three of them used
red ink for titles; punctuation marks or catchwords were employed at will, etc.
[Fig. 2]
If 14 people participated in the transcription of the text, and taking into
account that the codex consists of sixteen quinions and one binion, some of
the copyists clearly had to work twice or three times as much as the rest did.
In fact, one of them worked four times as much, as he (or she) took charge of
four quinions and the binion (copyist g). Furthermore, another one was a sort
of supply copyist to turn to when, whatever the reason, the copyist on duty had
to absent himself from his work place. Sometimes, the copyist was only away
for a short time and the substitute had to transcribe just a few lines. In other
cases, the reason for the copyist on duty to leave was more pressing and the
substitute had to transcribe a greater number of lines or even several folios. In
all, the supply copyist contributed to the work four times (copyist s).14
But the matter does not rest there and, as we will see, it is a bit more
complicated.
14 This copyist is easily identifiable by a peculiar feature of his writing: above the words with
ending kf he usually writes a small kf. See Fig. 3.
The fragment of Madrid comprises ten quinions and one binion (plus one leaf
of the following gathering, which is counted in the next section), the work being
distributed as follows:
As the table above shows, in the fragment of Madrid each copyist undertook
the transcription of a gathering except for two cases: a sole person, copyist b,
took charge of quires 2 and 9; and another one, copyist g, made quires 7 and 8
(the binion). In sum, nine scribes carried out eleven gatherings. In addition
to these nine, the supply copyist collaborated by transcribing fragments of
unequal extension in quires 2, 4 and 8:
The main irregularities in the fragment of Madrid concern the gatherings rather
than the scribal labour: quire 8 is not a quinion but a binion, and quire 11 is
an imperfect quinion as it contains only nine folios (one leaf must have been
left over, so the copyist tore it off). Assuming that quire 1, which currently
consists of seven leaves, was originally a perfect quinion, there are no further
observations to be made on the rest of the gatherings of this fragment.
figure 3 Ibn ayyn, Muqtabis, ms. Madrid, rah 11/9370, f. 118v (quire 4, copyists d and s)
The above described method for distribution of work relating to the fragment
of Madrida succession of copyists every ten leavescan be seen as an almost
ideal situation. By contrast, the method used for the fragment of Fez involved
up to three copyists participating in the transcription of a sole quire whereas
a copyist alone transcribed up to three quinions. The fragment of Fez does
not present any irregularity in terms of the gatherings: not includingwe do
insistthe leaves belonging to another codicological unit, this fragment is
made up of six quires, all of which are perfect quinions. The six were tran-
scribed by five copyists plus the substitutewhose contribution to this frag-
ment amounted to no more than four linesin an irregular manner, which
the following table tries to summarize:
Quires 12 and 14 (1st and 3rd of Fez) have no irregularities at all: each was
copied by a hand with distinct and recognizable features (copyists j and m,
respectively). In contrast, the gathering in between is the most irregular in
terms of the distribution of labour. As shown next, quire 13 was the work of
three people, who shared out the task in a curious way:
F. 198r Copyist j
Ff. 198v, 199r Copyist k
F. 199v Copyists k and l
Ff. 200r, 200v Copyist k
F. 201r Copyist j
Ff. 201v, 202r, 202v, 203r, 203v, 203bisr Copyist l
Ff. 203bisv, 204r, 204v, 205r, 205v Copyist k
Ff. 206r, 206v Copyist l
15 As said in the previous section, the first leaf of this gathering is the last one of the fragment
of Madrid.
16 Makk (ed.), Al-Muqtabas min anb ahl al-Andalus, pp. 153154 (introduction); see also
p. 354, n. 4 (edition).
The last leaf, f. 206, of the gathering described in the previous section as being
the work of three copyists (quire 13) visibly differs from the rest. On the one
hand, the watermark it exhibits is not the same as that of the others; on the
other hand, and what is more relevant to this paper, the page setting is different:
the written area is larger and the interlinear space is smaller. Consequently,
there are more than 29 lines to the page (as many as 34 can be counted, and
the number of those missing may be 3 or 4). Apparently, the copyist who took
charge of this gathering (copyist l) realized that the text still to be finished was
not going to fit into the last leaf, so he tore off this leaf and added another one
whose justification and ruling allowed him to introduce a larger amount of text
into it. In addition, he was forced to make his handwriting smaller and more
compressed in order that the text fit into the available space. [Fig. 5; cf. fig. 4]
This feature is typical of collective copying: the scribe must accommodate
the share of text allotted to him into the space at his disposal, either by making
his writing smaller or larger, or by reducing or enlarging the space between
wordsespecially in the last and/or penultimate leaf of the gatheringor by
modifying the page setting or, even, the size of the quire. The example just
described is a clear illustration of this, but there are others throughout the
manuscript.17
As previously said, the tenth leaf of quire 11 was left over, and the copyist
of this gathering (copyist i) solved the problem by directly removing it. This
copyist also had the three last lines of the previous leaf to spare, and he simply
left that space blank (f. 187v). A similar problem was encountered and resolved
by other copyists, in instances where the extra space was of small significance,
i.e. the copyists of quire 1, who had four lines left over in f. 94v [Fig. 6], and quire
3, who had just one in f. 114v.
The script of the person who transcribed quires 2 and 9 (copyist b) is quite
large, to such a degree that near the end of both gatherings he found himself
compelled to make it smaller in order to fit the whole text on the leaf [Fig. 7].18
17 The ability of each copyist to adjust to the available space varies a lot. Thus, some of them
kept a constant pitch throughout their stint, whereas others had to perceptibly modify it
in the final part. Copyist b provides an illustrative example in the two gatherings he made.
See the next footnote.
18 For instance, the first sixteen pages of quire 9 have 57.7 characters per line on average,
whereas in the next two pages the average adds up to 78.1 and 78.4 characters per line,
respectively. By doing so the copyist was endeavouring to make the text fit into the last four
pages of the gathering. He succeeded, in such a way that in the last page of the gathering
(f. 168v) he could return to his normal script, which resulted in the 57.2 characters per line
on average of this page.
figure 4 Ibn ayyn, Muqtabis, ms. Fez, Qaraouyine Library, f. 199v (quire 13, copyists k
and l)
figure 5 Ibn ayyn, Muqtabis, ms. Fez, Qaraouyine Library, f. 206r (quire 13, copyist l)
figure 6 Ibn ayyn, Muqtabis, ms. Madrid, rah 11/9370, f. 95r (quire 2, copyist b) / f. 94v
(quire 1, copyist a)
The above instances are all more or less disguised, only noticeable as the
result of careful examination. However, the example provided by quire 14 is so
blatant that it would not be hidden even from the least informed person. In
order to fill the last page of the gathering (f. 216v), the scribe (copyist m) had to
stretch out the words and to separate them to an exaggerated extent. [Fig. 8]
Evidently, the material to be copied was distributed among several people,
and it seems logical to think that each copyist was given one of the gatherings
that made up the exemplar. In other words, the quires of this copy would corre-
spond to those of the manuscript from which it was copied.19 This hypothesis
seems to be not only the most logical but also the most likely. However, if this
was the case, it raises other questions: Why did one of the copyists have to add
a binion? Did the exemplar also contain a binion? Why did another one have
to tear off a whole leaf, and why did another have to separate exaggeratedly the
19 The amount of text inserted into each page of the exemplar was the equivalent to approx-
imately 25 lines of this copy. We know this thanks to the fact that some leaves of the
exemplar were incorrectly arranged and the copyist of our manuscript transcribed them
without realizing the disorder (ff. 235v236v). On this, see Luis Molina, The Preservation
of Textual Continuity in the Formation of a Composite Codex (forthcoming).
figure 7 Ibn ayyn, Muqtabis, ms. Madrid, rah 11/9370, f. 104r (quire 2, copyist b) / f. 102v
(quire 2, copyist b)
words in the last page of the gathering because of the extra space when neither
of them had an especially small handwriting?
Collective copying was not a usual practice in the Arab world, but it was
not as exceptional as Franz Rosenthal suggested. He wrote that only under
special circumstances [] it was considered advisable to divide the task of
copying the whole among various scribes.20 As an exceptional example, he
mentioned the ten scribes who were commissioned by Amn al-Dawla, vizier
to the Ayybids, to copy the eighty volumes of Ibn Askirs Trkh Dimshq
in the thirteenth century.21 Each made eight volumes and, thanks to this work
method, the transcription of this considerable work was completed in two
years. The exceptionally huge task that Rosenthal takes as an example could
not very well have been completed without collective copying.
20 Franz Rosenthal, The Technique and Approach of Muslim Scholarship. Rome (Pontificium
Institutum Biblicum), 1947, p. 2.
21 Ibn Ab Uaybia, Uyn al-anb f abaqt al-aibb, ed. Nizr Ri. Beirut (Manshrt
Dr Maktabat al-ayt), s.d., p. 725.
figure 8 Ibn ayyn, Muqtabis, ms. Fez, Qaraouyine Library, f. 216v (quire 14, copyist m)
22 Some examples and further references in Franois Droche, Manuel de codicologie des
manuscrits en criture arabe. Paris (Bibliothque nationale de France), 2000, pp. 211212
(Islamic Codicology. An Introduction to the Study of Manuscripts in Arabic Script, trans.
by Deke Dusimberre and David Radzinowicz, ed. by Muhammad Isa Waley. London
(Al-Furqn Islamic Heritage Foundation), 2006, pp. 198199).
23 Franois Droche, La transmission crite du Coran dans les dbuts de lislam. Le codex
Parisino-petropolitanus. Leiden (Brill), 2009.
24 The remaining two leaves are at the Vatican Library (Vat. Ar. 1605/1) and the Khalili Islamic
Collection of London (kfq 60).
25 Rosemarie Quiring-Zoche, A Manuscript Copied in Teamwork? Manuscripta Orientalia
9/4 (December 2003), pp. 6772.
26 Quiring-Zoche, A Manuscript Copied in Teamwork? p. 67a.
rect, then it would be a case similar to ours. Like ours, the codex is divided
into quinions; but, unlike ours, the changing of hand does not coincide with
the changing of quire. Five hands copied the 333 leaves of the shiya of
Shaykhzde codex: the first transcribed 86 and a half leaves; the second one
made 50 leaves; the third, 49 and a half; the fourth, 130 and a half; and the
fifth, the last 15 and a half leaves. Quiring-Zoche observed that in some parts
the copyists made an effort to accommodate a certain amount of text on a
certain number of pages, whereas in other parts this fact is less evident.27
She was inclined to think that it was teamwork (although she was not entirely
certain). In any case, it did not necessarily mean that all scribes worked simul-
taneously.28
Conclusion
The production of the codex unicus of the second volume of Ibn ayyns
Muqtabis was the result of teamwork, perhaps under special circumstances.
What these special circumstances may have been we do not even dare to
speculate. Yet, it seems evident that there was a need to finish the work as
soon as possible, to such an extent that, together with the 13 copyists (not
counting those who made *m2a, obviously) working in the same place at
the same time, there was a person keeping a watchful eye on them, ready to
substitute any scribe who had to leave temporarily, even to transcribe as little
as four lines. Furthermore, there was apparently no intention of carrying out a
homogeneous work as, apart from using the same ruling pattern, each copyist
put into practice his/her own scribal usages. In short, speed of accomplishment
was given priority over uniformity of result. The colophon in f. 284r does not
give any clue to the circumstances under which this copy was made either,
since no information is provided on the date of the copy or on the place where
it was made, let alone on the names of all the copyists who participated in it.29
We can only tell that the codex is most probably datable to the second half of
the fifteenth century with the help of the watermarks in the paper.30
In sum, all evidence suggests that the only extant copy of the second volume
of Ibn ayyns Muqtabis was made in the late fifteenth century in an Andalusi
or Maghribi atelier that was created ad hoc, either for meeting a deadline or in
order to cope with a pressing situation. Anything else belongs to the realm of
speculation.
flower, are largely attested in documents from the late fifteenth century and the early six-
teenth. Several documents preserved at the Archivo Municipal of Crdoba, for example,
exhibit almost identical watermarks; for instance, documents 36, 46 and 50 of Seccin 01,
serie 02 (in boxes 001 and 002). See also Ricardo Crdoba de la Llave and Manuel Cerezo
Villegas, Filigranas del Archivo Municipal de Crdoba (14501550), Anales de la Univer-
sidad de Alicante. Historia Medieval 6 (1987), pp. 407434, esp. p. 410. Another very similar
example is in Charles Mose Briquet, Les Filigranes. Paris, 1907, No. 10790: (www.ksbm
.oeaw.ac.at/_scripts/php/BR.php?IDtypes=102&lang=fr) (last accessed on May 10, 2015).