The Geometry of Paradise: Symmetry, Mathematics, and the Alhambra

Introduction

In the heart of Granada, Spain, the Nasrid palaces of the Alhambra stand as one of the most stunning achievements of Islamic art and architecture. Built between the 13th and 14th centuries, the intricate geometric patterns that cover nearly every surface of the palace complex represent something far deeper than ornamentation: they are a meditation on infinity, order, and the divine, expressed through the universal language of symmetry.

What makes the Alhambra uniquely significant to mathematicians is not merely the beauty of its patterns, but their structural sophistication. The Nasrid artisans, working without any formal concept of group theory — a mathematical framework that would not emerge in Europe until the 19th century — created tile work that realises an extraordinary range of symmetrical structures. The question of exactly how many of the mathematically possible planar symmetries appear in the Alhambra has generated one of the most fascinating debates in the history of mathematics and art.

This article explores the mathematics of symmetry, the classification of repeating patterns into the 17 wallpaper groups, and the remarkable story of the Alhambra's geometric legacy.

---

The Four Fundamental Symmetries

Every repeating two-dimensional pattern is built from combinations of just four rigid motions — transformations that preserve distance and orientation:

Symmetry	Description	Symbol
Translation	Shifting every point by a fixed vector	$T$
Rotation	Turning around a point by a fixed angle	$C_n$ or $R$
Reflection	Mirroring across a line	$\sigma$ or $m$
Glide Reflection	Reflection followed by translation along the mirror axis	$g$

A pattern is periodic if it has at least two non-parallel translations — that is, if the whole pattern repeats in two independent directions, forming a lattice. The mathematical classification of such patterns asks: given a periodic tiling, which subset of these four symmetries does it obey?

---

The Crystallographic Restriction

Not every rotation is compatible with a periodic lattice. If a pattern repeats by translation in two directions, the set of allowed rotations is severely limited. This is the crystallographic restriction theorem:

A periodic two-dimensional pattern can only have 2-fold, 3-fold, 4-fold, or 6-fold rotational symmetry. 5-fold, 7-fold, or higher rotations are impossible in a periodic tiling.

The proof is elegant. Let $R$ be a rotation by angle $\theta = 2\pi/n$ about some point, and let $\mathbf{v}$ be the shortest non-zero lattice vector. The vector $R\mathbf{v} - \mathbf{v}$ must also be a lattice vector (since the lattice is closed under rotation). Its length is:

$|R\mathbf{v} - \mathbf{v}| = 2|\mathbf{v}|\sin\!\left(\frac{\theta}{2}\right)$

For this to be an integer multiple of $|\mathbf{v}|$ (by the lattice property), we need:

$2\sin\!\left(\frac{\theta}{2}\right) \in \mathbb{Z}_0$

Since $0 \leq \sin(\theta/2) \leq 1$, the only possibilities are:

$n$	$\theta$	$\sin(\theta/2)$	$2\sin(\theta/2)$	Allowed?
1	$360^\circ$	0	0	Trivial
2	$180^\circ$	1	2	Yes
3	$120^\circ$	$\sqrt{3}/2 \approx 0.866$	$\sqrt{3} \approx 1.732$	Yes
4	$90^\circ$	$\sqrt{2}/2 \approx 0.707$	$\sqrt{2} \approx 1.414$	Yes
5	$72^\circ$	$\approx 0.588$	$\approx 1.176$	No
6	$60^\circ$	$1/2 = 0.5$	1	Yes
$\geq 7$	$\leq 51.4^\circ$	$< 0.434$	$< 0.868$	No

This restriction explains why honeycombs have hexagonal cells, why snowflakes have six-fold symmetry, and why the Alhambra's periodic tilings only exhibit rotations of order 2, 3, 4, and 6.

---

The 17 Wallpaper Groups

The full classification of periodic planar patterns was worked out in several stages. Evgraf Fedorov (1891) proved there are 17 possible symmetry groups for periodic tilings, though his work went unnoticed in the West for decades. Independently, George Pólya (1924) and Paul Niggli arrived at the same result.

The 17 wallpaper groups are categorised by the types of symmetry operations present and the relationships between them. In orbifold notation (developed by John Conway), each group has a compact symbol:

Each symbol breaks down as follows (using orbifold notation):

• Digits (3, 4, 6) indicate the highest order of rotation
• * indicates a reflection axis
• x indicates a glide reflection without a corresponding pure reflection
• o indicates only translations (the "primitive" group)

For example, *p4m (442) has:

• 4-fold rotations at the centre of each square
• 2-fold rotations at edge midpoints and square corners
• Mirror lines in four directions, spaced at $45^\circ$

This is the symmetry group of a standard chessboard.

---

The Great Alhambra Debate

The Alhambra's patterns are so mathematically rich that they sparked a controversy spanning over 80 years.

Edith Müller (1944)

In her 1944 doctoral thesis Gruppentheoretische und Strukturanalytische Untersuchungen der Maurischen Ornamente aus der Alhambra in Granada, Edith Müller conducted the first systematic analysis of the Alhambra's tile patterns using the recently developed framework of wallpaper groups. She identified 11 distinct groups among the patterns accessible at the time.

Grünbaum and Coxeter (1980s)

In the 1980s, mathematician Branko Grünbaum re-examined the Alhambra and claimed that Müller had missed two groups. He found 13 groups. The legendary geometer H. S. M. Coxeter confirmed this count, and for a period, the consensus settled at 13.

The Claim of All 17 (1987)

Then came the bombshell. In his 1987 book José María Montesinos: Classical Tessellations and Three-Manifolds, the Spanish mathematician argued that the Alhambra contains patterns realising all 17 wallpaper groups. This was an extraordinary claim — it would mean that medieval Nasrid artisans had intuitively discovered every possible way to repeat a pattern in the plane, five centuries before mathematicians classified them.

Montesinos' argument was controversial because several of the groups he identified are highly subtle. For instance, pgg (22x) has only glide reflections, no pure reflections — a pattern whose symmetry is notoriously difficult to spot by eye.

Recent Findings (2000s–2010s)

Rafael Pérez-Gómez conducted extensive surveys of the Alhambra Museum and newly restored areas, including parts of the palace complex that were previously inaccessible or had been covered by later plasterwork. His research confirmed the presence of pg, p2, pgg, and p3m1 in these newly accessible areas — groups that had been central to the debate.

The current scholarly consensus leans toward the Alhambra containing at least 15–16 of the 17 groups, with the remaining one or two still contested. What is beyond dispute is that the Alhambra represents the most extensive and sophisticated use of planar symmetry groups in any pre-modern building tradition.

---

Nasrid Motifs: Geometry in Practice

The Nasrid artisans developed a distinctive visual vocabulary of geometric motifs, each with its own symmetry properties:

The Pajarita (Bow-Tie)

The pajarita ("little bird") is one of the most common motifs in Alhambra tile work. It consists of two interlocking rhombus-like shapes forming a bow-tie pattern. This motif typically appears in patterns belonging to groups cmm, p4m, or p4g, depending on the arrangement and colouring. M.C. Escher was so captivated by this pattern during his 1922 visit to the Alhambra that it directly inspired his tessellation work.

The Bone and Airplane Patterns

Other characteristic Nasrid motifs include the hueso ("bone") pattern, an elongated hexagonal shape, and the avión ("airplane") pattern, which resembles a stylised four-pointed star. These motifs often appear in borders and as filler elements between larger geometric stars.

Nine-Fold Rosettes

One of the most mathematically intriguing features of Nasrid art is the appearance of nine-fold rosettes. These are star-shaped rosettes with nine points, which are not constructible with classical ruler and compass (since $9 = 3^2$, and constructing a regular nonagon requires solving a cubic equation).

Yet they appear in the Alhambra's stucco work and muqarnas ceilings. The artisans achieved this through an ingenious technique using cartabones and ataperfiles — sets of right-angled triangles with specific angle ratios that, when combined, approximate the $40^\circ$ angle of a regular nonagon with remarkable precision. This reflects a practical, empirical geometry that operated outside the Euclidean tradition but achieved equivalent results.

---

Rosettes: Cyclic and Dihedral Groups

The rosettes found in Alhambra ornamentation beautifully illustrate the distinction between two fundamental families of finite symmetry groups:

• Cyclic groups $C_n$: rotation only, $n$ equally spaced rotations around a centre
• Dihedral groups $D_n$: rotations plus reflections, $n$ rotation axes in addition to the rotations

A rosette with only rotational symmetry (no mirror lines) belongs to $C_n$:

$C_n = \{R^0, R^1, R^2, \ldots, R^{n-1}\}$

where $R^k$ denotes rotation by $k \cdot 360^\circ/n$.

A rosette with reflectional symmetry belongs to $D_n$, which has $2n$ elements:

$D_n = \{R^0, R^1, \ldots, R^{n-1}, \sigma_0, \sigma_1, \ldots, \sigma_{n-1}\}$

where $\sigma_k$ is reflection across the $k$-th axis.

The Alhambra contains rosettes with $n = 5, 6, 8, 9, 10, 12, 16$, and even higher orders. The eight-pointed star (derived from $D_8$) and twelve-pointed star ($D_{12}$) are particularly common, as they emerge naturally from the 4- and 6-fold symmetries of the underlying tile lattices.

---

Conclusion

The Alhambra represents a unique confluence of art, mathematics, and spiritual expression. Its Nasrid builders, working within an artistic tradition that forbade figural representation, channeled their creative energy into geometric abstraction of extraordinary sophistication. Without formal algebra or group theory, they discovered and deployed symmetries that mathematicians would not systematically classify for another 500 years.

The debate over whether the Alhambra contains all 17 wallpaper groups may never be fully resolved — some of the patterns have been lost to restoration, and the surviving fragments are open to interpretation. But what is beyond doubt is that these 13th-century artisans achieved a level of geometric mastery that continues to astonish mathematicians, artists, and visitors alike. In the tiled walls of the Alhambra, abstract group theory is rendered in glazed ceramic — a silent testament to the universal language of symmetry.

---

Further Reading

• E. Müller, Gruppentheoretische und Strukturanalytische Untersuchungen der Maurischen Ornamente aus der Alhambra in Granada (1944)
• B. Grünbaum, "What symmetry groups are present in the Alhambra?" Notices of the AMS (1984)
• J. M. Montesinos, Classical Tessellations and Three-Manifolds (1987)
• D. Schattschneider, "The Plane Symmetry Groups: Their Recognition and Notation" The American Mathematical Monthly (1978)
• R. Pérez-Gómez, "The Fourfold Alhambra: Symmetry and Architecture" (2012)