Arabesque

About Arabesque

Pour ink onto a reed pen in a 16th-century Tabriz workshop. The pen draws one spiral stem — a curving line that loops back on itself, splits in two at a leaf-fork, and returns. A second stem mirrors the first across a vertical axis. A third turns the pair into a half-rosette. Within an hour the page carries a vegetal scroll that will pass to the weavers, the tilemakers, the bookbinders, and the architectural stuccoists working for Shah Tahmasp's court. The pattern is *islimi* in Persian, *rumi* in Ottoman Turkish, and in the wider European vocabulary, *arabesque*. Arabesque is the curvilinear vegetal-and-biomorphic ornamental tradition of Islamic art — distinct from the rectilinear-geometric *girih* family but routinely combined with it on the same surface. The tradition has formal rules: stems split symmetrically, leaves rotate at the split, half-palmettes and split-palmettes are the canonical leaf forms, the scroll fills a frame to a controlled density. Its formative phase runs from the Umayyad palace at Mshatta (8th c.) through Abbasid Samarra (Styles A, B, and C identified by Ernst Herzfeld's 1923 publication on the Samarra stucco) into the high Persian, Ottoman, and Mughal traditions. The Ardabil Carpet (1539-40, Maqsud Kashani, V&A Museum) carries one of the most ambitious worked-out arabesque compositions in any medium. The Iznik tilework of Topkapı Palace (16th c.) and the Mughal pietra dura panels of the Taj Mahal (1632-53) extend the tradition into ceramic and stone inlay.

Mathematical Properties

Arabesque is curvilinear, not polygonal — its formal rigor is rule-based rather than ruler-and-compass. The base rules govern stem behavior and leaf attachment. A primary stem may split into two secondary stems at any point along its run; at the split, the two stems mirror across the bisecting tangent of the parent stem, generating local bilateral symmetry. Leaves attach at points of stem split and at points of stem termination. The canonical leaf forms are the *palmette* (a fan of leaflets radiating from a central stem), the *split palmette* (a palmette divided into two halves rotated outward), the *half palmette* (one side of the divided palmette), and the *saz leaf* (the elongated curving leaf of Ottoman tradition). The composition fills a frame under a density rule: the negative space in any region must be governed by the surrounding positive forms, never left to wander.

Larger arabesque compositions are organized around symmetry axes — bilateral, four-fold, eight-fold, or twelve-fold — that determine the overall structure. The Ardabil Carpet centers a sixteen-lobed medallion with eight cartouches arrayed around it; the field is broken into bilateral halves through the long axis. Iznik tile compositions often use bilateral symmetry through a vertical axis with horizontal repetition under the bilateral rule. The formal rigor is closer to a generative grammar than to a fixed pattern: the rules define what counts as a well-formed arabesque, and the artisan within those rules has interpretive freedom comparable to a calligrapher writing the same text in a specific hand. The Topkapı Scroll documents this rule-set as transmissible craft knowledge — geometric construction guides accompanied by canonical example compositions.

Architectural Use

Arabesque appears across virtually every surface available in Islamic architecture — carved stucco on wall fields, carved stone on facades and portals, glazed ceramic tile on dadoes and domes, painted plaster on vault soffits, inlaid wood on doors and minbars, inlaid stone (pietra dura) on Mughal marble surfaces, and woven textile in carpets and wall hangings. The standard Maghrebi-Andalusi surface program at the Alhambra and the Bou Inania madrasa places arabesque stucco panels in the wall fields above the zellige geometric dado and below the calligraphic frieze — a vertical sandwich where each band carries a different ornamental tradition.

In high Safavid Iran, the Shaykh Lutfollah Mosque (1603-1619) and the Shah Mosque (1612-1638) in Isfahan carry arabesque tile compositions on their domes and qibla walls; the Shaykh Lutfollah dome's interior is a worked-out arabesque-and-medallion composition on a tan-gold ground. In Ottoman Istanbul, the Iznik tile fields of the Rüstem Pasha Mosque (1561-63, Sinan) and the Süleymaniye (1550-57, Sinan) carry saz-leaf arabesque on mihrab walls and dome pendentives. In Mughal India, the Taj Mahal (1632-53) uses arabesque pietra dura inlay on the cenotaphs, the dado panels, and the throne wall; the Itimad-ud-Daulah tomb in Agra (1622-28) is the earlier Mughal pietra dura prototype.

The portable surfaces — carpets, manuscript bindings, lacquered book covers, metalwork — carry arabesque at smaller scale. The Safavid royal kitabkhana workshops under Shah Tahmasp (r. 1524-76) produced illuminated manuscripts where arabesque framing borders are the standard ornamental scaffolding around calligraphy. The Ardabil Carpet is the architectural-scale carpet counterpart — designed not as a floor covering but as a worked-out ornamental program for a shrine floor at Ardabil.

Construction Method

Arabesque begins on paper or parchment with a master draftsman in a court kitabkhana (book-and-design workshop) or an artisan workshop. The draftsman draws the primary stems with a reed pen and ink, applies the splitting rules at each fork, attaches the canonical leaf forms at the splits, and fills the frame to controlled density. The drawing is then traced or pricked through to a working surface — onto tile blanks, onto a carpet cartoon, onto a stucco wall, onto a stone block for carving, onto a marble panel for inlay.

Material-specific construction follows. For Iznik ceramic tile: the design is transferred by pouncing (pinpricks dusted with charcoal) onto a glazed white tile blank, then painted with cobalt blue, copper turquoise, manganese black, and the signature Iznik tomato-red (a fritted slip applied as a thick relief) under a transparent overglaze, then kiln-fired. For Safavid carpet: a master designer produces a full cartoon at scale, the cartoon is divided into bands, and a team of weavers works each band to the cartoon under a head weaver's supervision — the Ardabil Carpet (1539-40, Maqsud Kashani) records its head designer's name in a small inscription panel. For Mughal pietra dura: the design is traced onto a white marble base, the design lines are chiseled out to shallow channels, and shaped pieces of semi-precious stone (carnelian, lapis, jasper, malachite, mother-of-pearl) are cut to fit and set into the channels with adhesive. For carved stucco at the Alhambra or Bou Inania madrasa: the design is transferred to wet plaster, the plaster is carved while still workable, and the finished panel is painted and sometimes gilded.

In every medium, the underlying drawing logic is the same — the same formal grammar of stem-and-leaf rules. The Topkapı Scroll and surviving artisan workshop manuals document this transmission. Apprentices in Safavid, Ottoman, and Mughal court workshops would have spent years working through the canonical compositions before being trusted to invent within the rule-set. The craft transmits through workshop lineage, master-to-apprentice, with the scrolls and manuals as supplements rather than primary sources.

Spiritual Meaning

Arabesque's theological coordinates sit inside the same aniconic framework that shapes Islamic geometric tradition broadly: divine unity (tawhid) cannot be reduced to figurative depiction, so the ornamental tradition develops formal languages that carry meaning without picturing the sacred. Where geometric *girih* tiling carries that meaning through n-fold symmetry and self-similar tiling, arabesque carries it through controlled vegetal abstraction — recognizably alive (plant-derived stems and leaves) but not identifiable as any specific plant. The half-palmette is not a real palm leaf; the split palmette is not a real flower. The forms are abstracted enough that no viewer mistakes them for natural specimens, yet alive enough that the surface reads as cultivated rather than mechanical.

Oleg Grabar's framework in *The Mediation of Ornament* reads this controlled-abstraction as the carrier of meaning between viewer and architecture. The arabesque is not a sign of something else (not a coded message, not a hidden esoteric script) — it is the mediating layer through which the architecture addresses the viewer. Sheila Blair and Jonathan Bloom's *The Art and Architecture of Islam 1250-1800* (Yale, 1994) elaborates this for the post-Mongol period: arabesque under the Timurids, Safavids, Ottomans, and Mughals carries dynastic identity, theological orientation, and craft virtuosity simultaneously, and the traditional viewer would have read these layers without separating them.

The move to avoid is the modern New-Age reading that treats arabesque as 'sacred geometry secret-Sufi-code' for a hidden mystical content. The patterns are publicly transmitted through artisan scrolls (Topkapı, the Tashkent fragments, parallel Maghrebi manuals), discussed openly in surviving treatises, and reproduced in court workshop manuals for centuries. There is no esoteric layer hidden behind the formal surface. The traditional theological framework is real and discussable; the modern romantic reading that adds a layer of secret mysticism on top of the public craft tradition is a 19th- and 20th-century Western projection. The honest spiritual reading is the one the tradition gives itself: aniconic ornament as the formal language through which a sacred or sovereignty context can be marked without depicting what it marks.

Significance

The error to avoid in writing about arabesque is treating it as decoration. It is decoration in the same sense that a sonnet is text on a page — formally constrained, ruleset-governed, and capable of carrying serious content within its constraints. The *islimi* tradition of Persia and the *rumi* tradition of Ottoman Turkey are sister-traditions with shared formal vocabulary: split palmettes, half-palmettes, lotus-and-peony hybrid leaves, double-spiral stems, and symmetry rules that govern how a stem may divide, what leaves may sit at the division, and how the entire composition must fill a frame to a controlled density without leaving negative space ungoverned.

The historical arc begins at Mshatta — the unfinished Umayyad palace facade (8th c., now mostly in Berlin's Pergamon Museum) where vegetal-and-geometric interlace appears in early form on a continuous stone facade. The Abbasid capital at Samarra (836-892) carries the tradition's first articulated stylistic phases. The German archaeologist Ernst Herzfeld, working at Samarra in 1911-1913 and publishing the stucco-ornament volume in 1923 (*Die Ausgrabungen von Samarra, Band 1: Der Wandschmuck der Bauten von Samarra und seine Ornamentik*, Berlin: Reimer), identified three distinct stucco styles in the palaces. Style A is the earliest and closest to late-antique vine-scroll; Style B is more abstracted; Style C — sometimes called the 'bevelled style' — is the most stylized, with deep slanted-bevel cuts that produce a rhythm of light and shadow rather than naturalistic plant forms. Style C is the technical innovation that disengages the tradition from late-antique naturalism and opens the form to its later abstraction.

The high Persian period under the Safavids (1501-1736) is where arabesque reaches its most ambitious worked-out compositions. The Ardabil Carpet (1539-40), made for Shah Tahmasp's shrine to his ancestor Shaykh Safi al-Din, is signed by Maqsud Kashani — the carpet's central medallion radiates an *islimi* composition of split-palmettes and double-stems against a deep indigo ground, with surrounding cartouches each carrying their own self-contained variations. The carpet originally formed a pair; the larger inscribed-and-dated example is at the Victoria and Albert Museum in London, while the smaller un-inscribed companion (reconstructed in the 19th c. using border fragments from the V&A example) is at the Los Angeles County Museum of Art. In Ottoman Iznik, the same vocabulary translates into ceramic tile — the tilework of the Süleymaniye and Rüstem Pasha mosques in Istanbul, and the Harem courtyards of Topkapı Palace, deploys saz-style vegetal scrolls in white-blue-red-green on a white ground, often with one large signature leaf (the saz leaf, a long curved fern-like form) anchoring the composition. In Mughal India, the same tradition translates into stone — the pietra dura inlay of the Taj Mahal (1632-53, under Shah Jahan) sets carved semi-precious stones into white marble in vegetal compositions that read as three-dimensional *islimi*.

Oleg Grabar's *The Mediation of Ornament* (Princeton, 1992) is the central scholarly framework for taking arabesque seriously rather than as either exotic decoration or hidden mystical code. Grabar's argument is that ornament in Islamic art is the mediating layer between viewer and architecture — the carrier of meaning, not the absence of it. The aniconic theological frame (no figurative depiction in sacred or sovereignty contexts, grounded in tawhid) shapes what content the ornament can carry; the formal sophistication of arabesque is what content-carrying looks like once figuration is filtered out. This is not 'secret esoteric knowledge'; it is publicly transmitted craft tradition with theological coordinates that any traditional viewer would have read fluently. The modern Western reception of arabesque — through 18th-19th-c. Orientalist publications like Owen Jones's *The Grammar of Ornament* (1856), Prisse d'Avennes's *L'Art arabe* (1869-77), and the chinoiserie-and-arabesque vogue of European decorative arts — distorted the tradition into ornament-as-exotica. The 20th-century corrective, beginning with Herzfeld and continuing through Grabar, Sheila Blair, Jonathan Bloom, and Eva Baer, has been to restore the tradition's formal rigor and its theological coordinates without re-mystifying it.

Connections

Arabesque belongs in conversation with three sister-traditions and one cross-tradition cousin. Inside Islamic art it routinely combines with the rectilinear-geometric *girih* family on the same surface — the Topkapı Scroll documents both, and the Alhambra carries arabesque stucco panels in fields between zellige geometric dados and calligraphic friezes. The two traditions are formally distinct (curvilinear vs. rectilinear, vegetal vs. polygonal) but they share a doctrinal frame and frequently share a wall. Calligraphy is the third partner: *Kufic* and *Naskhī* scripts often run as banded borders or central inscriptions inside arabesque fields, and the line-quality of the calligraphic stroke and the *islimi* stem are kin. Outside Islamic tradition, the closest formal cousin is the Indian temple lattice and creeper-vine tradition (*kalpalata*, the wishing-vine) seen in Mughal-era jali screens and in the architectural carving of Hindu temples — the Mughal arabesque-and-jali tradition is a real cross-pollination point between Islamic *islimi* and Indic vegetal ornament. Further afield, late-antique vine-scroll (Roman, Byzantine, Coptic) is the direct ancestor of the earliest Umayyad and Abbasid examples — Style A at Samarra is recognizably continuous with Mediterranean classical vine-scroll, and the abstraction into Style B and Style C is the historical pivot that disengages the tradition from its late-antique base. The Western Christian medieval 'arabesque' (Romanesque manuscript marginalia, Anglo-Saxon interlace) is a distant cousin that adapted vocabulary through Mediterranean contact but developed independently.

Further Reading

• Grabar, Oleg. *The Mediation of Ornament*. Princeton University Press, 1992.
• Herzfeld, Ernst. *Die Ausgrabungen von Samarra, Band 1: Der Wandschmuck der Bauten von Samarra und seine Ornamentik*. Berlin: Reimer, 1923. (The original Style A/B/C classification of Samarra stucco.)
• Baer, Eva. *Islamic Ornament*. New York University Press, 1998.
• Necipoğlu, Gülru. *The Topkapı Scroll: Geometry and Ornament in Islamic Architecture*. Getty Center, 1995.
• Canby, Sheila R. *Shah ʿAbbas: The Remaking of Iran*. British Museum Press, 2009. (Safavid art and arabesque tradition.)
• Atasoy, Nurhan, and Julian Raby. *Iznik: The Pottery of Ottoman Turkey*. Alexandria Press, 1989.
• Koch, Ebba. *The Complete Taj Mahal*. Thames & Hudson, 2006. (Mughal pietra dura.)
• Jones, Owen. *The Grammar of Ornament*. London, 1856. (Primary historical source — read with Grabar's critique in mind.)