The New
Rug Society
Presents An
Online Exhibition:

Prayer Rugs & Related Textiles


Tech Notes




Welcome to what we believe is the first on-line exhibition of antique oriental rugs and textiles to be posted by a North American rug society. Each of these pieces is "on loan" from a member of the New England Rug Society. And the introductory essay from NERS member Jeff Spurr adds some valuable new insights into a subject about which much has been written over the past century. Our thanks go to NERS members Robert Alimi, Yonathan Bard, and Jim Adelson as well as Jeff Spurr for the hard work and energy that went into making this exhibition possible. We hope you find it both fulfilling and enjoyable.

Mark Hopkins, NERS President
November, 2002


Some Thoughts on
Islamic Prayer Rugs and Related Textiles
by Jeff Spurr

Islamic prayer rugs have captivated collectors since the beginnings of modern collecting at the end of the nineteenth century. They were also highly appealing to the rising middle classes when they started furnishing their homes with imported rugs in the latter decades of the same century. They are often discussed in rather abstract, decontextualized ways among collectors. Consequently, this essay commences with the two defining terms which link these rugs to their place in Islamic culture and proceeds to a discussion which attempts to sustain this interest in context to achieve a better sense of the location of these interesting objects within the cultures and places whence they originated. It also briefly addresses a general class of Islamic textiles that adopt the form and imagery of prayer rugs and attempts to establish a meaningful relation between them.

Salat:  devotional prayer, comprises one of the five pillars of Islam.  Originally it was to be performed three times a day (dawn, dusk and night), to which were later added the prayers at midday and at mid-late afternoon to make five.  According to G.  Monnot in his article, “Salat,” The Encyclopaedia of Islam, VIII, p. 925-34), the Koran "places the origin of ritual prayer, under divine guidance, at the outset of humanity," (p. 925) it being practiced by all of the prophets back to Abraham.  The most fully realized prayer is communal, hence the Friday prayer, which ideally takes place at a congregational mosque.  Prayer needs to be performed in a state of ritual purity, achieved either by performing ablutions or, in the absence of sufficient water, with fine sand.  This prayer must be performed in a prescribed direction, toward al-Masjid al Haram, the Ka`ba at Mecca.  Detailed information is available in the E.I. article cited above.

Sajjada (or sajjadah):  "sjd:" verbal root for to become lowly or submissive; to prostrate oneself, hence "sujud," a ritual prayer act, and "masjid" (mosque), with the locative "ma" prefix, for a place of prayer, thence sajjada. According to A. Knysh, in his article, "Sadjdjada," in The Encyclopaedia of Islam, VIII, pp. 740-41, the word is not employed in the early compilation of Hadith (the Prophet's sayings) by Bukhari (d. 256 AH, 870 AD), where Muhammad is variously described as performing prayer on his own clothes, his quilt or a mat.  The term does not appear until the early 10th century, where it is initially equated with khumra, the mats described in the Hadith.  In so far as many of the first textiles used for prayer in the Islamic world were mats, the Sumatran rattan prayer mat (#26) in this show might be considered to come closer to the original objects used than those things we usually conceive of as "prayer rugs." That the range of possibilities for an individual prayer panel extends beyond the realm of textiles is made clear by the low wooden platforms, raised off the ground by two six-inch high transverse supports and decorated at their rounded top end with elaborate carving, favored in the Swat region of Pakistan and, perhaps, neighboring areas of Afghanistan.

It is perhaps a truism to state that any material object is the product of a complex web of forces and factors that combine to insure its existence and use in the world.  Concerning prayer rugs and related textiles, the following play obvious roles:  (1) material conditions of production, including the nature of the environment and available materials; (2) the technical capacities of the producers and the possibilities and limitations of available techniques; (3) the cultural determinants, including (a) the religious/ideological context and, specifically, ritually-prescribed behaviors relevant to the use of the objects,  (b) the stylistic tradition within which they are made and the available models from that tradition employed in their design, and (c) both the available and appropriate iconography (whether fully understood by the creator or not) (4) the imagination of the individual weaver (if not fully governed by commercial rules); (5) the status of the producer and user (who may be one and the same or not); and, finally, (6) the more general economic circumstances of production.  All of these will be in play at the moment that any piece is produced.

Most attention in the scholarly literature is devoted to the formal and symbolic aspects of prayer rugs. I do not wish to simply reprise the information provided by the authors I mention here since that is readily available in their articles; however, I may comment on certain points they make and will provide a quick gloss of more extended discussions.  I intend particularly to address some practical issues specifically related to manner of use and context of production and how they apply to the character of the prayer rugs and related pieces.  The two scholars who have seriously addressed the general topic to the greatest effect are Richard Ettinghausen in "The Early History, Use and Iconography of the Prayer Rug, in the Textile Museum catalogue, Prayer Rugs (1974), and Walter Denny in "Saff and Sejjadeh: Origins and Meaning of the Prayer Rug," Oriental Carpet and Textile Studies 3(2).  While in his article, Denny is promoting a specific thesis, the Ettinghausen piece is particularly dense with information, although he does get sidetracked a bit by deciding to address some Anatolian rugs that do not reflect the characteristic prayer rug format.

Images of Paradise in Islamic Art, edited by Sheila Blair and Jonathan Bloom, to which Denny was a contributor, also provides a wealth of information regarding the religious, intellectual and visual environment for the imagery on prayer rugs, although it has a hard time setting limits, the text sometimes leaving the impression that everything floral within Islamic art may be perceived as evocative of paradise, which vitiates rather than underlines the meaningfulness of the primary forms and symbols under discussion.  Furthermore, Denny's new catalogue to the recently opened exhibit at the Textile Museum, The Classical Tradition in Anatolian Carpets, contains particularly useful discussions of filiation of designs within this rich tradition. 

As most people understand, Islamic prayer does not demand the use of a prayer rug.  It does require ritual purity and, after the necessary ablutions, the placement of a textile of some sort between the person praying and the ground will suffice.   Ettinghausen does address use in citing early and late examples in the literature.  He makes clear that ownership or circumstances of use or quality of a sajjada were likely to reflect status, whether hierarchical or elective, as in the instance he cites from the 14th century traveler, Ibn Battuta, who describes the circumstances of ownership and use of sajjadas by members of a Cairene mystic brotherhood (pp. 11-12). Furthermore, Sufi shaykhs throughout the Islamic world owned a special sajjada (or, alternatively, a sheepskin) that served as a badge of office, for which reason they were termed the sajjadanishin, Persian for 'the one who sits on the sajjada,' the sitting shaykh in other words.

Most mosques provide an appropriate prayer surface obviating the necessity of individuals who have come to pray to bring their own.  Traditionally, these floor covers were only exceptionally saffs, that is multiple-niche prayer carpets designed specifically to be placed in a mosque to assist in the organization of mass prayer. It seems that only in the modern era when mass communications have broken down barriers of distance and local traditions and convey normative notions of what Islamic institutions should look like that the saff has truly spread far and wide, often in machine-made form.  In Iran the zilu, a type of large flat-woven rug, was the most common mosque floor covering.  Interestingly, although later ones virtually never adopt the saff format, the oldest known example, a fragment datable to the 14th century, was clearly woven with multiple niches (see Earthly Beauty, Heavenly Art: Art of Islam, M. Piotrovsky & J. Vrieze, p.  66)  North African mosques typically feature large reed mats. Several old Ottoman saffs have survived and there is a reference in the archives to purpose-woven carpets for the Süleymaniye in Istanbul in the 16th century.  Nevertheless, even large Ottoman mosques were often covered by rugs of every description, given the age-old tradition in Turkey of making pious donations of rugs and kilims to them. With warm weather use or at times when large congregations might spill out into the courtyard, a prayer rug or cloth would be necessary for those not assembling inside the prayer hall.  Still, evidence exists here and there revealing that a person might decide to conspicuously mark his high status it by employing a prayer rug of his own in the mosque.

It is impossible to know when the use of pile prayer rugs commenced just as there is a dispiriting absence of physical evidence for early rugs in general despite the stray textual evidence that makes their existence certain.  Denny would have it that the individual prayer rug derived from the saff.  He illustrates early (12th c.) depictions on paper of mosques, which he shows are represented by arcades with mosque lamps hanging in them.  This is symbolic in the sense that a three-dimensional space within which lamps would hang is condensed into two dimensions, with arches depicted as hung with lamps.  Simply put, he argues that the arcade as a motif antedates the mihrab or prayer niche and the manar or minaret as a distinctive architectural characteristic of mosques, and saffs emulated these arcades and thus the early Arab mosque type of the hypostyle hall.  This seems to ignore the clear presence of prayer mats from the early period and some of the evidence for early usage of prayer rugs adduced by Ettinghausen.  There is no doubt that the arcade is a feature of certain mosques, particularly the early hypostyle type, but who is to say that the arch or niche did not first provide a model for the design of a prayer rug only for that design to be repeated with the impulse to weave purpose-made rugs for mosque floors?  Without any concrete evidence, any such hypothesis will remain entirely conjectural.

One factor has been left out of past discussions focusing on the formal and symbolic features of prayer rugs, which has an impact on the issue of the relative paucity of Persian prayer rugs if we exclude the commercial sort that were manufactured in response to the export market.  This is that critical feature of Twelver Shi`ite devotional practice, the "prayer stone," the  mohr used by each individual while engaged in prayer.  The prayer stone is preferably made of unbaked clay derived from the vicinity of one of the shrines to the imams and martyrs of Shi`ism, amongst which the shrine to Husayn at Karbala, Iraq, provides the ideal source and that to Imam Riza in Mashhad the second best source for Iranians.  On the one hand, it is formed of the lowliest stuff––earth––and thus represents the humility of the owner, but it also participates in the holy essence of the martyred imam and confers blessings on the owner, who rests his forehead, or forehead and nose (there may be a larger and a smaller or simply one mohr) on it every time he prays. The stone may be plain or have devotional phrases impressed in its surface.  These plaques are periodically washed by their owners and, due to their propensity to dissolve, may be reconstituted by them.

The point to all this is that the accompanying prayer panel is more likely to be a prayer cloth than a prayer rug.   This textile will also serve as a bokhcha within which to wrap the mohr, more readily accomplished with a cloth than a rug.  It is typically rather small and a prayer rug, if used, should be relatively small and light.  These prayer panels are often undecorated.  The upper classes are liable to privilege fine material and a plain design, whereas the lower classes are inclined to favor more highly decorated cloths.  Although highly decorated, the Persian prayer panel of imported Kashmiri shawl material in the Sackler exhibit would certainly be the property of a wealthy owner.  However, no clergy would use so ostentatious  a prayer panel.  It is important to realize that it would have been used as the container for storing and transporting a prayer stone as well as a surface to pray on.  Of the two types of Isfahani qalamkari (printed and painted cotton textiles), that conform to the arch or mihrab format, it is more likely that the smaller, explicitly architectural type would be used as a prayer panel whereas the larger, with the cypress/tree of life as its central motif, would more likely have been used as a hanging in the home.  (My gratitude to Ahmad Mahalatti Shirazi for the bulk of this information on Shi`ite practice).

These facts help explain the relative dearth of Persian prayer rugs compared to the situation in Anatolia.  It is no accident that the splendid group of early Persian prayer rugs found in the Topkapi palace and only recently convincingly determined to be early were for so long considered nineteenth century products of Turkish looms.  The poverty of comparanda made it difficult for scholars to believe that they could really be sixteenth century.  Since it is almost certain that they comprised a portion of a formal gift sent by the Safavid Shah to the Ottoman Sultan, it is possible that they were produced as the result of an explicit imperial order.

Particularly when we are addressing rugs made largely in the latter decades of the 19th and the early 20th century, it is essential to emphasize that many "prayer rugs," were made for entirely commercial purposes, indeed for export to places where they would never be used in prayer but instead would be subjected to the most inappropriate sort of service underfoot in European and American homes.  They were obviously a popular design for members of the bourgeoisie, who bought them by the thousand.  Thus it behooves us not to get too carried away in our descriptions of these things.  To be sure, prayer rugs were an extremely common accompaniment to the religious life of the pious in Turkey.  This is made clear by the fact that they are so abundantly available in kilim form and in village pile production across the whole of Anatolia (whether emulating more sophisticated commercial examples or not).  Nevertheless, the kustar system introduced by the Russian imperial authorities to promote rug weaving for commercial purposes in the Caucasus and the similar developments in Iran sponsored by private German, British and Tabrizi merchants, produced countless prayer rugs that went directly to market.  Many of these Persian prayer rugs are relatively large and would have ill served the requirements for Shi`ite prayer described above.  While most Caucasian prayer rugs are much smaller, there is little if any ethnographic evidence of their local use.  Two Caucasian saffs are known to this author from the literature.  Whether they were ever used in a mosque is not known.

As already stated, much attention has been devoted to the symbolism of the prayer rug.  This has clarified its formal and symbolic associations with the mihrab, the gateway to paradise and paradisiacal imagery generally, and the lamp as evocative of the the Ayat al-nur, the Light Verse, Qur'an 24:35, which commences "God is the Light of the heavens and the earth; the likeness of His Light is as a niche wherein is a lamp...kindled from a Blessed Tree..." (see Reflections of Paradise..., p. 37).  This quote references a blessed tree, perhaps associated with the "Lote Tree of the Boundary,"one leaf of which could shade the whole community of the faithful, located in Janna, Islam's paradise (or Firdaws, the seventh heaven, both literally meaning "garden") (see the article by L. Gardet, "Djanna," in E.I, II, pp. 447-51). The tree of life most commonly takes the form of the cypress in the Persianate visual culture of recent centuries (the cypress being associated with eternity in Persian literature).  Other symbols, such as the ibrik (ewer for ablutions), and protective talismans could be woven into prayer rugs, such as the Marasali in the Sackler show.

In her article, "The Mihrab Image: commemorative Themes in Medieval Islamic Architecture," Muqarnas IX, pp. 11-28, Nuha Khoury carefully argues that in the medieval period flat mihrabs, which actually were placed in mausolea or provided the form for some tombstones, carried much of the imagery that we associate with mosque mihrabs proper and that they should really be considered as "mihrab images," and "signs of commemoration" (p. 12).  These were associated with sites of particular sanctity, such as the tombs of saintly figures, and often evoked the equation of garden: paradise: life eternal in death.  She also quotes the hadith, "Souls of martyrs take shelter in gold lamps that hang in the shadow of (God's) throne," (p. 18) in elaborating on the symbolic reference of the lamp on these objects.  It is clear that, with time, such imagery found its way willy-nilly into the realm of the prayer rug and that distinctions that might have once held true in the realm of architecture were conflated in this new medium and environment.  Islam is a religion freighted with few primary formal symbols.  I would like to suggest that the arch-mihrab-gate form acted as a multivalent master symbol in Islamic culture that could be represented with a variable specific iconography and could be inflected in particular ways depending upon immediate context.

A large class of textiles exists for which the form and, in some cases, the specific iconography of the prayer rug was adopted sometime in the past.  Most of those extant do not date much earlier than the 17th century and the vast majority are (as with most textiles) from the nineteenth and early twentieth centuries.  This form was particularly common in the qanat (kanat), or portable textile screens surrounding tent complexes, characteristically formed of individual panels, most often decorated with flowering bushes or trees under an arch.  Elite tents themselves, whether Mughal, Safavid or Ottoman, could feature multiple panels of a similar sort on their interiors.  This class includes a great number of domestic hangings that embellished Middle Eastern households, which were often relatively devoid of furniture.  Three examples can be found in this show (#s 23, 24 & 25).  They reflect, to one degree or another, imagery associated with prayer rugs: the arch, the cypress/tree of life and paradisiacal imagery generally.  This is particularly true of the Persian embroidered hanging (#23), with its exceptionally rich iconography.  Islamic popular tradition has it that the peacock, two of which grace this textile, was expelled from the Garden of Eden with Adam and Eve.  The multiplicity of animals evokes the notion of a "peaceable kingdom" and thus an ideal place.  In Central Asia and elsewhere, many important events occur out of doors.  Textiles such as these, and the already-mentioned qanats, are hung to define the space set aside for these events.  Consequently, these textiles can be seen to confer prestige and cast a protective aura over the spaces marked in this way, whether temporary or permanent.

It is not the purpose of this introduction to discuss the prayer rugs and related textiles on display here in any depth.  The contributors have done a fine job of describing their pieces.  It is notable that the Anatolian prayer rugs are all village products, in contrast to the mainly (though not exclusively) commercial rugs on exhibit at the Sackler.  Together, they make for a interesting display of the variety available from that region although only one in the Sackler recalls the coupled column prayer rugs and their multifarious successors that are presently highlighted at the Textile Museum.  It is interesting that, while #4 and #6 show features derived from kilim technique, the one kilim (#7) is the only prayer rug here to feature the upper field panel with ascending plant forms that descended from the original high Ottoman coupled column prayer rug through the Ladik tradition (among others). 

The splendid Marasali prayer rug (#8) makes a nice contrast to the much later but jewel-like example in the Sackler (please refer to my short article on this type of rug, "Marasali – the Kashmir Connection" in Hali 105, July-August 1999, pp. 105-107).  The six Baluch rugs implicitly present a nice case for the delights of variation within a relatively restrictive form and palette and the Baluchi propensity for borrowing motifs, including the possible relationship between the "shrubs" in #21 and the stylized flowering plants in the Beshir example (#22).  The most curious example is the Qarajeh rug (#15), which expresses its identity as a prayer rug (if that it be) in the most oblique of terms, easily overlooked.  We invite comments on this and the other pieces and hope that those visiting this exhibit derive some of the pleasure from viewing these rugs and textiles that NERS members did in collecting and studying them.


Jeff Spurr works at Harvard University serving as Cataloguer for Islamic Art in the Aga Khan Program for Islamic Architecture at the Fine Arts Library, where he documents Islamic visual culture and manages several archival collections of historical photographs of the Middle East, most notably the Harvard Semitic Museum Photographic Archives. A longtime student of rugs and textiles, he is also a collector of textiles and basketry from Borneo, Sumatra, Africa and Central Asia.


NOTE: Please see Hali #127, March-April 2003, pp. 105-111 for a more developed version of this essay entitled Ends and Means: Islamic Prayer Rugs in Context.