Original article
Painting Department, Fine Arts Faculty, University of Granada, Avda. Andalucia s/n, 18071 Granada, Spain
b Materials Science Institute of Seville (CSIC), Avda. Americo Vespucio, s/n., 41092 Seville, Spain
c Centre de recherche et de restauration des muses de France, C2RMF, Palais du Louvre, Porte-des-Lions, 14, quai Fran
cois-Mitterand, 75001 Paris, France
d Analytical Chemistry Department, Science Faculty, University of Granada, Avda. Fuentenueva s/n, 18071 Granada, Spain
Received 30 October 2008; accepted 27 January 2009
Available online 13 November 2009
Abstract
This work focuses on the study of paper and textiles used in the binding of a series of manuscripts that share some specific characteristics that
lead us to speculate on the possibility of a transitional codicological typology from the Arabic to the Christian book in Al-Andalus during the 15th
century. The books we analyzed belong to the collection of the Historical Archive of Malaga, the Archive of Sacromonte Abbey, in Granada, the
School of Arabic Studies and the Library of P.P. Escolapios, also in Granada. Paper physical study was performed by microscopic and spectroscopic
techniques. A routine and objective method, Fourier Transform Infrared (FTIR) spectroscopy, was employed and proved to be a useful technique
for the characterization of cellulosic fibres, main component of paper from the boards and the envelope flap pasteboards, and the fabric lining
from the cover. The results of our research will help us to date, identify and study the evolution of the techniques, proving that the materials and
innovations of the Italian paper manufacturing processes were perfectly known in the south of modern day Spain, before the Christian Reconquest.
2009 Elsevier Masson SAS. All rights reserved.
Keywords: Al-Andalus; Arabic paper; Fibre identification; 15th century; Microscopic and spectroscopic techniques
1. Research aims
The present work focuses on the study of fabric lining and
papers used in the covering and paper used as support in a
series of manuscripts which share particular characteristics
regarding their binding structures that might suggest a
codicological typology characteristic to the transition between
Arabic and European book in Al-Andalus [1,2]. This study was
carried out keeping in mind the differences between Arabic
and European paper, as well as the different manufacturing
processes. We will also consider the potential interchange of
certain elements of theses different processes during the period
of coexistence of both civilizations in the Iberian Peninsula,
when there was a flourishing trade and information exchange
between both cultures. We applied Fourier Transform Infrared
Corresponding author.
E-mail address: adrian@icmse.csic.es (A. Duran).
1296-2074/$ see front matter 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.culher.2009.01.007
51
Table 1
Composition of the different cellulosic fibres.
Cotton
Flax
Hemp
Jute
% cellulose
% hemicellulose
% pectin
% lignin
% other components
82.7
64.1
67.0
64.4
2.8
16.7
16.1
12.0
2.9
1.8
0.8
0.2
0.1
2.0
3.3
11.8
11.5
15.4
12.8
11.6
whiten with lime [5]. The paste was then placed on a special
sieve or mould in order to shape the sheet. Generally, sheet size
depended on the artisans ability to move the sieve. That was
the reason why paper was almost never bigger than 45 65 cm.
To render it suitable for writing, paper was burnished and sized
with certain materials using a brush, making paper less porous
and brighter. European paper was smoother than oriental paper.
While Europe used animal glues, Oriental paper used rice, wheat
or corn starches which improved paper conservation and gave
its bright appearance.
In Islamic manuscripts, each quire consisted of several bifolia
(sheets of paper folded in two) or, sometimes, single leaves were
added to the quires in the fold. The number of bifolia might vary
considerably: from just one bifolia to quires made up to 12 sheets
or more. All the different quires were then sewn together [6].
Few studies on paper identification have been carried out
and they have determined composition only by means of visual
examination using microscopy and scanning techniques. These
methods rely heavily on experience, as they generally generate
qualitative data and require a subjective comparison of results
[79]. Furthermore, degradation and physical damage in documents are mainly due to humidity, solving agents and glues that
can even destroy the characteristic fibre morphology necessary
to an unambiguous identification. The most common fibres used
in textile supports for paintings and in documentary cultural heritage are cotton and flax, although others such as hemp or jute
might also be found.
Vibrational spectroscopy covers a range of techniques,
including FTIR, Raman and Attenuated Total Reflectance (ATR)
spectroscopies that have already been widely employed for the
study and differentiation of cellulosic, proteinaceous and artificial fibres [1013].
However, there have been few reports on the differentiation
of chemically similar fibres within a group, such as cellulosic
plant fibres (flax, cotton, hemp and jute). Edwards et al. [14] have
employed Raman spectroscopy to discriminate among untreated
plant fibres on the basis of peak ratios derived from the associated CH and glycosidic COC vibrations. Garside and Wyeth
[15,16] have employed ATR Spectroscopy, a refinement of conventional FTIR, to characterize these ones.
In this work, we used ATR and conventional FTIR spectroscopies, and since we obtained similar results with both
techniques we decided to use only FTIR, since this technique
is in general easier to perform. Spectroscopic methods for fibre
characterization are based in the different types of bond vibrations and the different composition of cellulosic fibres (Table 1).
All plant fibres have a cellular structure and are largely composed
Sarh al-Yumal
fi l-nahwi (Fig. 1b);
Manuscript 20 from Sacromonte Abbey Archive of Granada
(14th century), a book of medicine of al-Aynazarbi, Kitab
al-kafi fi-l-tibb min al-qarn il l-qadam (Fig. 1c);
Al-Zubaydi grammar book from the School of Arabic Studies
in Granada (1516th centuries) (Fig. 1d);
Moorish Koran from the Order of P.P. Escolapios [Pious
School] Library of Granada (15th century) (Fig. 1e).
Samples were generally taken from strategic zones of the
books that, without damaging the document, provided the best
information (Table 2). In addition to paper samples, textile samples, taken from the fabrics and threads used to reinforce the
book structure, are also included. The analyses were performed
in the Materials Science Institute of Seville. Also, some analyses of samples from the Moorish Koran were carried out at
the University Alfonso X and the Spanish Historical Heritage
Institute.
The examination of the physical features consisted in the
description of paper dimensions and thickness. Thickness values were the mean obtained by dividing the sum of a set of
measurements performed in different parts of the same sheet
and also in different sheets. If wide differences in these values
were found, the highest and lowest values were determined. For
the detection of wire and chain lines distribution, watermarks,
support characteristics and sheet flaws due to the papermaking
process a fibre optic light sheet was used. The measure of wire
lines are taken between 20 mm in all the cases.
In the samples analyzed at the Materials Science Institute,
fibres were treated with a sodium sulphate solution (1% p/v) in
order to remove glues and other undesired materials. The same
methodology was followed in all the cases (textiles used in bindings threads and fabrics and papers). The fibres were placed
on a microscope slide, after fibre separation in order to get a
longitudinal view. The transverse sections (cross-sections) were
52
Fig. 1. Photographies: a: Book I, Malaga Historic Archive; b: Manuscript 16, Sacromonte Abbey Archive of Granada; c: Manuscript 20, Sacromonte Abbey Archive
of Granada; d: Al-Zubaydi Grammar, School of Arabic Studies of Granada; e: Moorish Koran, P.P. Escolapios Lybrary of Granada.
53
Table 2
Provenance of the samples.
Book I Malaga
Sample 1
Sample 2
Sample 3
Sample 1 Manuscript 16
Sample 2 Manuscript 16
Sample 3 Manuscript 16
Sample 4 Manuscript 16
Sample 5 Manuscript 16
Sample 1 Manuscript 20
Sample 2 Manuscript 20
Sample 3 Manuscript 20
Sample 4 Manuscript 20
Sample 5 Manuscript 20
Sample 6 Manuscript 20
Al-Zubaidy Grammar
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
Sample 6
Sample 7
Sample 8
Sample 9
Sample 10
Textblock paper
Upper cover pasteboard
Lower cover pasteboard
Inner cover covering paper
Blue thread warp of the spine reinforcement
Blue thread woof from the spine reinforcement
Natural thread warp of the inner cover lining
Natural thread woof of the inner cover linning
Headband threat
Natural thread sewing of the cover leather
Moorish Koran
Sample 1
Sample 2
Sample 3
Textblock paper
Textblock paper
54
Table 3
Description of manuscripts.
Book
Binding
Textblock
Material
Writing
Material
No. folia
Dimension
textblock
(mm)
Thickness ranges
(mm)
Chain lines
Wire lines
Watermarked
Book I Malaga
Envelope
Parchment
Paper
114
200 160
0.200.30
Not visible
Not visible
No
Manuscript 16 Sacromonte
Envelope
Rag paper
245
295 215
0.190.27
25
Open hand
with a flower
or a six-point
star
Manuscript 20 Sacromonte
Envelope
Non-vocalized
Maghribi
Rag paper
48
305 235
0.260.30
Not visible
13
Not visible
Al-Zubaidy Grammar
Envelope
Brown leather
Maghribi
Paper
90
200 140
0.170.25
Folio 1 to 80
Folio 81 to 88
Folio 90
16
15
13
V-Shaped
No
Medallion
Moorish Koran
Envelope
Maghribi with
characteristics of
Andalusi style and
in Mesut type.
Carefully
punctuated with
red vocalization.
Headings in Kufic
script
Rag paper
108
235 185
0.240.40
Not visible
10
No
5
6
5
Type
55
Fig. 2. Microphotographies: a: fibres of sample 1 from Malaga Historic Archive (scanning electron microscopy); b: fibres of sample 4 from Manuscript 20 of
Sacromonte Abbey (optical microscopy 50); c: fibres of sample 6 from Manuscript 20 of Sacromonte Abbey (scanning electron microscopy); d: bluish fibres of
sample 5 from manuscript 16 of Sacromonte Abbey (cross-section, optical microscopy 50); e: fibres of sample 7 from Al-Zubaydi Grammar (scanning electron
microscopy); f: fibres of sample 1 from Moorish Koran (optical microscopy 25).
56
Table 4
I1595 , I1105 and I2900 cm1 , and R1 and R2 values, corresponding to the 27 samples studied by spectroscopic methods.
I1595
I1105
I2900
R1
R2
Book I Malaga
Sample 1
Sample 2
Sample 3
0.024
0.052
0.028
0.100
0.204
0.062
0.069
0.135
0.057
0.24
0.25
0.45
0.35
0.38
0.49
Sample 1 Manuscript 16
Sample 2 Manuscript 16
Sample 3 Manuscript 16
Sample 4 Manuscript 16
Sample 5 Manuscript 16
Sample 1 Manuscript 20
Sample 2 Manuscript 20
Sample 3 Manuscript 20
Sample 4 Manuscript 20
Sample 5 Manuscript 20
Sample 6 Manuscript 20
0.040
0.033
0.027
0.022
0.021
0.031
0.021
0.033
0.039
0.046
0.052
0.043
0.142
0.110
0.053
0.050
0.135
0.088
0.039
0.093
0.108
0.316
0.034
0.088
0.070
0.038
0.037
0.076
0.042
0.034
0.056
0.072
0.205
0.93
0.23
0.24
0.41
0.42
0.23
0.24
0.84
0.42
0.42
0.16
1.18
0.38
0.38
0.58
0.57
0.41
0.50
0.98
0.70
0.64
0.25
Al-Zubaidy Grammar
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
Sample 6
Sample 7
Sample 8
Sample 9
Sample 10
0.041
0.059
0.029
0.026
0.025
0.015
0.021
0.015
0.026
0.035
0.070
0.098
0.056
0.143
0.120
0.052
0.041
0.022
0.133
0.155
0.058
0.070
0.043
0.093
0.083
0.046
0.032
0.020
0.084
0.108
0.58
0.60
0.51
0.18
0.21
0.29
0.50
0.68
0.19
0.23
0.71
0.84
0.67
0.28
0.30
0.33
0.64
0.73
0.31
0.32
Moorish Koran
Sample 1
Sample 2
Sample 3
0.029
0.032
0.045
0.065
0.070
0.089
0.053
0.059
0.073
0.45
0.46
0.51
0.56
0.54
0.62
Fig. 3. Infrared spectra from: a: sample 3 from Malaga Historic Archive; b: sample 3 from the Al-Zubaydi Grammar; c: sample 4 from the Al-Zubaydi Grammar;
d: sample 6 from Manuscript 20 from Sacromonte Abbey.
57
Fig. 4. Representation R1 R2 of the 27 samples studied by infrared spectroscopy, defining the different types of cellulosic fibres.
apart from the Grammar book with a 1.4 ratio. The imprints left
by the mould are visible.
In a generic way, we can talk about two different types of
paper clearly differentiated. The first type is the paper used in
the Koran, in Manuscript 20 and in Book I showing a poor refinement degree since rests of fibres, threads or other textile used
as raw materials in the manufacturing process are visible under
transmitted light. Paper is not satin or watermarked. In Book I
the wire lines are not visible. In Manuscript 20, chain lines and
chain-sewing marks are not visible. In the rest of the documents,
generally five chain lines are present, and also 10 to 13 wire lines.
Leaf thickness is not homogeneous and varies considerably from
0.20 to 0.40 mm.
The second type of paper has a clear Western influence. It
was used in Manuscript 16. It is watermarked and chain and
wire lines are clearly visible. The mould was probably metallic
and left five chain lines that run parallel to the axis of the spine
and 25 wire lines separated 20 mm. This type of paper is thinner
than the first type ranging from 0.19 to 0.27 mm, depending on
the folium.
However, the Al-Zubaydi Grammar features from both types
can be found. As we mentioned before, this manuscript dates
from the end of the 15th century to beginning of the 16th century
and different types of paper are present, all poorly refined and
not satin. Paper thickness is similar to the previous one varying
from 0.17 to 0.25 mm. However, the number of chain lines in
20 mm varies from 10 to 16. Chain lines are irregularly arranged,
perpendicular to the spine axis and generally there are five per
leaf, although some leaves show six.
All the samples, so much those of paper as the employees
in the binding, were analyzed morphologically using optical
and scanning electron microscopy, also on cross-sections. Due
to microscopic techniques limitations, a routine and objective method, FTIR spectroscopy, was employed repeatedly to
58
confirm results and proved to be a good technique for differentiating the types of cellulosic fibres in paper and fabric lining.
According to the spectroscopic and microscopic analysis,
lower values of R1 and R2 were calculated in the spectra
corresponding to those samples where cotton presence was
previously suggested by microscopic methods. So, we can
confirm the presence of cotton in paper samples: sample 2 from
Malaga Book, samples 1 and 2 from Manuscript 20, and also,
in samples 2 and 3 from Manuscript 16. Cotton is also present
in fibres from textiles used in other parts of these documents,
such as reinforcement areas of the book structure, especially in
the Al-Zubaydi Grammar.
If we take into account that the raw material used in Arabic papermaking was mainly flax and hemp and that cotton was
mainly used in Christian Europe, the presence of cotton fibres in
paper in Manuscript 20 of the collection of Sacromonte Abbey
from early years of the 14th century (702 AH/1302 AD)
and also in Manuscript 16, in pasteboard of the manuscript from
Malaga and in fabric lining from the cover of these documents
can make us think that this raw material was incorporated to the
manufacturing process in Al-Andalus. The presence of watermarks in both the second and the Al-Zubaydi book ratifies the
fact that the innovations of the Italian papermaking process were
perfectly known and used in the production of Arabic paper in
Al-Andalus before the Christian reconquest.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
Acknowledgments
[14]
[15]
[16]
[17]
[18]
[19]
[20]
References
[1] T. Espejo, A. Beny, Book 1 from the collection of arabic manuscripts from
the Malaga Province Historical Archive: an example of andalusi binding, in:
[21]