The Islamic Geometric Tradition

Core Concepts

Aniconism: constraint as creative force

Islamic religious art is structured by a theological prohibition rooted in hadith: the representation of living beings challenges God's uniqueness as creator. This principle — aniconism — is not a blanket rule across all Islamic contexts, and its strictness varies between traditions and periods. But in religious architecture, it concentrated artistic ambition onto three channels: calligraphy, abstract floral arabesque, and geometry.

Geometry did not merely fill the space vacated by the figure. It took on its own theological weight. Geometric patterns became a visual analogy for divine order: the circle, having neither beginning nor end, represented infinity and eternity. Tessellations that can extend without limit served as a visual metaphor for God's boundlessness. The concept of tawhid — divine unity, the oneness of God — found geometric expression in the way disparate polygonal shapes interlock into a single coherent field. Pattern was not decoration applied to architecture. It was architecture as argument.

The mathematics underneath

What Islamic craftspeople achieved empirically, mathematicians would not formally describe until the 19th and 20th centuries. Two-dimensional repeating patterns can be classified by their symmetry operations — translations, rotations, reflections, glide-reflections — into exactly 17 distinct groups, called wallpaper groups. Contemporary analysis shows that Islamic geometric patterns from the 10th to the 13th centuries instantiate all 17 groups, meaning craftspeople achieved comprehensive coverage of every mathematically possible two-dimensional symmetry — without knowing group theory existed.

Islamic artisans mastered symmetry operations now formally classified in group theory. Mathematical completeness was an achievement of practice, not of proof.

This is the central puzzle the module keeps returning to: how does a tradition that works without formal proofs produce results that formal mathematics can only retrospectively classify?

---

Narrative Arc

Before girih: the compass-and-straightedge period

The earliest Islamic geometric patterns were constructed through compass-and-straightedge methods inherited from Greek mathematics — particularly from Euclidean geometry as transmitted through translation projects in Baghdad from the 8th century onward. A craftsman or designer would draw a circle, subdivide its circumference, extend lines from intersection points, and build up a pattern step by deliberate step. The method is precise and teachable, but it treats each pattern as a fresh construction from first principles.

These early patterns were already extraordinary. But they had a ceiling. Compass-and-straightedge construction becomes exponentially laborious as complexity increases, and certain symmetry classes — particularly those involving fivefold and tenfold rotational symmetry — are especially resistant to this approach.

The girih breakthrough (circa 1200 CE)

By the 12th century, something conceptually important shifted. Islamic artisans reconceived girih patterns — intricate geometric surface decorations — not as sequences of compass-and-straightedge steps, but as tessellations of a small set of pre-defined polygons. These are the girih tiles.

The five girih tiles are: a decagon (ten-sided), a pentagon, a diamond (rhombus), a bowtie (non-convex hexagon), and a regular hexagon. Every side of every tile is the same length. Every interior angle is a multiple of 36 degrees. The 36-degree constraint is not arbitrary: it ensures all angles are divisible by 5, which is what makes fivefold and tenfold symmetric compositions possible.

Fig 1

Each tile is decorated internally with strapwork lines — partial lines that run across the tile's face. When tiles are placed edge-to-edge, these internal lines connect across boundaries to form the continuous interlaced pattern visible on the finished surface. The tile logic is hidden; the observer sees only the resulting geometric field.

The critical shift is modularity. Instead of constructing each design from first principles, a craftsman could compose patterns by arranging a small vocabulary of pre-made shapes. Complexity became combinatorial rather than additive. And — crucially — a craftsman could produce sophisticated results without necessarily understanding the mathematical principles encoded in the tile set.

Fivefold symmetry and the impossible tiling

The 36-degree angle constraint creates a tension that turns out to be generative. Fivefold and tenfold rotational symmetry cannot tile a plane periodically. You cannot fill a flat surface with a repeating unit cell that has fivefold symmetry without leaving gaps — unlike squares or hexagons, which tile cleanly. This is a mathematical theorem, not a practical difficulty.

Islamic artisans found a way around it: quasi-periodic tiling. Instead of a repeating unit, the pattern propagates by a subdivision rule — larger configurations can be broken into smaller copies of the same tiles. The pattern never exactly repeats, but it also never fails to fill the plane. This is what Western mathematics would later call a quasicrystal structure, after the 1984 discovery of physical quasicrystals and the 1973 mathematical work of Roger Penrose. Islamic artisans solved it in practice five centuries earlier.

The Darb-i Imam: the quasicrystal hiding in Isfahan

The clearest evidence is the Darb-i Imam shrine in Isfahan, built in 1453. Physicist Peter Lu and mathematician Paul Steinhardt published a 2007 analysis in Science showing that the girih tile pattern on the shrine's facade is quasi-crystalline in the precise mathematical sense: all tile edges align along tenfold symmetry directions, the pattern is not embedded within any larger periodic structure, and the design encodes an explicit subdivision rule sufficient to construct an infinite quasicrystal.

This was not a lucky accident. Contemporary Persian scrolls from the 15th century document girih tile constructions at multiple scales, showing that craftsmen were working with hierarchical design methods — large patterns composed of arrangements that, when zoomed out, replicate the structure of the tiles themselves. The mathematical consequence (quasi-periodicity, golden ratio proportions) was apparently arrived at empirically, through the logic of the tile system and accumulated craft knowledge, without formal proof.

Muqarnas: tiling logic goes three-dimensional

Geometric pattern in Islamic architecture does not remain flat. Muqarnas — the stalactite-like vaulting forms that cascade down domes, niches, portals, and squinches — extend the same underlying logic into three dimensions.

Muqarnas developed independently in the mid-10th century in both northeastern Iran and central North Africa, then spread across the Islamic world. Structurally, muqarnas consists of rows of prismatic niche-like elements, each row projecting outward over the one below in a corbel arrangement, producing the characteristic honeycomb appearance. Individual cells are built from a small set of geometric shapes — the same modularity principle as girih tiles — combined according to precise rules. When viewed from below, the result is a geometric composition that appears to dematerialize the ceiling into an infinite recession of nested forms.

The architectural function is the transition between geometrically incompatible volumes: a square room beneath a circular dome. Muqarnas fills the squinch — the corner zone where these geometries must negotiate — and in doing so becomes an occasion for extreme geometric elaboration.

The Timurid dynasty (late 14th–15th centuries) developed the most advanced muqarnas work, creating "squinch net vaulting" where muqarnas filled entire vault surfaces rather than just transition zones. The Persian mathematician al-Kashi provided the first known systematic mathematical analysis of muqarnas geometry in the same period — though it remains unclear whether he was formalizing existing craft knowledge or contributing new mathematical insight.

Regional variation: not one tradition but many

The Islamic geometric tradition is not monolithic. Regional contexts produced distinct aesthetic emphases.

Persian (Iranian) tradition produced the most mathematically adventurous geometric work, including the girih tile innovations and the Darb-i Imam quasicrystal. Muqarnas reached its greatest elaboration under Timurid and Safavid patronage.

Moroccan zellige is a distinct regional form of Islamic mosaic tilework in which individually hand-chiseled ceramic pieces in different colors are fitted together to form elaborate geometric compositions, typically built around radiating star designs. Unlike girih construction — which generates pattern through the interlocking of pre-designed polygons — zellige is a cutting and assembly process: each piece is chiseled from a fired tile to its precise shape and fit into position by hand. Morocco maintained and continued this craft tradition well after it declined elsewhere in the Islamic world after the 15th century, and it remains an active living practice today.

Ottoman work represents an interesting counterpoint: Ottoman artisans showed a reduced preference for geometric patterns compared to floral and vegetal designs. Geometric decoration appears in Ottoman religious architecture primarily in carved wooden elements — minbars, doors — while the dominant surface decoration in mosques became the Iznik tile with its characteristic flowing floral motifs. This is not an absence of visual sophistication but a different aesthetic commitment within the broader Islamic tradition.

---

Annotated Case Study

The Darb-i Imam shrine facade, Isfahan (1453)

What you are looking at: The entry portal facade of a Timurid-era shrine in Isfahan. The surface is covered with a girih tile composition — a continuous geometric field of interlacing strapwork generated by the five standard girih tile shapes.

Why this is the key case: In 2007, Peter Lu and Paul Steinhardt demonstrated that this particular composition is quasi-crystalline in the formal mathematical sense. Their paper in Science showed three things:

1. All tile edges align along tenfold symmetry directions — consistent with decagonal quasi-crystalline structure.
2. The pattern is not embedded in any larger periodic pattern — it cannot be explained as a section of a repeating unit.
3. The composition encodes an explicit subdivision rule: smaller tiles assemble into larger tiles of the same shapes, enabling infinite extension without repetition.

The craft evidence: Contemporary Persian scrolls from the 15th century include design drawings that show girih tile layouts at two scales simultaneously — the large pattern and the small-tile substructure that generates it. This is direct evidence that artisans were working with hierarchical, self-similar construction methods. The mathematical outcome (quasi-periodicity) appears to be a consequence of applying this method rigorously, not a consciously intended mathematical statement.

What it tells us about craft knowledge: The Darb-i Imam case is the clearest example in the entire course of what the craftspeople-mathematical-intuition claim describes: artisans achieving mathematical structures through empirical practice that formal mathematics would not formalize for five centuries. The craft system — the tile vocabulary, the combination rules, the hierarchical subdivision — encoded mathematical knowledge in procedural form. You did not need to know what a quasicrystal was to build one, if your tiles and rules already contained the necessary constraints.

---

Compare & Contrast

Compass-and-straightedge vs. girih tiles

	Compass-and-straightedge	Girih tiles
Origin	Greek mathematical tradition, transmitted through Islamic translation projects	Islamic innovation, emerging circa 10th–12th century CE
How a pattern is built	Sequential geometric construction from first principles: circles, arcs, intersection points, line extensions	Modular composition: a small set of pre-defined decorated polygons arranged edge-to-edge
Designer's required knowledge	Understanding of geometric construction steps; each new pattern requires a fresh approach	Internalized tile vocabulary and combination conventions; complexity is combinatorial
Practical ceiling	High complexity requires exponentially more construction steps	Combinatorial complexity scales more easily; hierarchy (tiles of tiles) unlocks quasi-periodicity
Mathematical outcome	Periodic tilings with classical symmetry groups	Enables quasi-periodic tilings with fivefold/tenfold symmetry; encodes mathematical structures not formally described until the 20th century
Did methods coexist?	Yes — both methods remained in use; neither simply replaced the other	Yes — both methods remained in use; neither simply replaced the other

---

Active Exercise

Exercise: Tile logic reconstruction

This exercise builds the core intuition of the module: understanding how a small vocabulary of shapes can generate a complex field through combination rules, and what mathematical constraints that vocabulary encodes.

Materials needed: Paper, pencil, and optionally printed or cut-out templates of the five girih tile shapes (decagon, pentagon, diamond, bowtie, hexagon — all sides equal, all angles multiples of 36°). Templates are available through resources like mathgrrl.

Step 1: Establish the angle constraint. Before placing any tiles, verify that all five shapes satisfy the 36° rule by measuring or calculating the interior angles. Note that a decagon has 10 sides (interior angle: 144° = 4 × 36°); a pentagon has interior angles of 108° = 3 × 36°; the diamond has angles of 72° and 108°; the bowtie has angles of 72° and 252° (reflex); the hexagon has angles of 120°... wait — 120° is not a multiple of 36°. This is the first interesting problem: how does the hexagon fit the system? (Answer: the hexagon is a special case that serves as a gap-filler in specific compositions; its role is constrained by context.)

Step 2: Build a local composition. Start with a decagon. Attempt to surround it completely using only the other four tile shapes, edge-to-edge, with no gaps and no overlaps. This is harder than it sounds. Record which arrangements work and which fail.

Step 3: Observe the strapwork. If working with decorated tiles (strapwork lines on each face), check whether the internal lines connect smoothly across tile boundaries in your composition. A correctly assembled region will show continuous interlaced lines. A misaligned edge will break the strapwork. Use this as a secondary diagnostic for correct assembly.

Step 4: Attempt a star. The most common girih composition centers on a ten-pointed star (formed by a decagon surrounded by alternating pentagons and diamonds). Attempt to construct this arrangement and then extend it outward by one ring. What tiles does the second ring require?

Reflection questions:

• Where did you hit dead ends? What does a dead end in tile placement tell you about the mathematical constraints the system is encoding?
• Could you have produced this composition through compass-and-straightedge construction? What would that process look like, and at what point does the girih approach become more tractable?
• What would it mean to extend this pattern infinitely? What pattern of choices would you repeat — or would you?