One of the recurring themes in medieval agricultural manuals, which are mainly based on ancient agronomic knowledge, centres around the correlation between the geography of a region and the effects of this geography on the growth of various plant types and their specific requirements. In Filāḥa al-Nabaṭiyya (Nabatean Agriculture), compiled in 291/903 in Lower Mesopotamia, Ibn Waḥshiyya (d. 318/930) proposed two reasons to explain this relationship. First, he argued that everything in the world possesses a nature (ṭabʿ ) composed of the four elements of fire, air, water and earth. The differences between the nature of each region and each plant explain, for instance, why the dust that covers a vine's leaves is harmful to the plant in ‘the whole clime of Bābil (Babylonia)’, while it might be beneficial in another region. His second explanation contends that the unique properties (khuṣūṣiyyāt) of each region are connected to its position ‘in respect to the rotation of the sun and the other stars which rotate in the sphere (al-falak)’. While the present book does not aim to provide an in-depth analysis of the various cosmological and philosophical aspects of agricultural manuals, such examples attest to the perception of agriculture as a human engagement with nature which is strongly dependent on the geography. A primary explanation for this concern, as mentioned by Ibn Waḥshiyya, is the recognition that the most crucial factors influencing plant growth include water, air, the ‘warmth of the sun’ and notably, soil. However, as several scholars have emphasised, the scholarly literature on gardens of Islamicate regions has often overlooked the relationship between gardens and their surrounding landscapes.This oversight is particularly evident in the conceptualisation of ‘the Islamic garden’, which as Petruccioli pointed out, ‘has been considered a specific, self-contained entity removed from its context – its surroundings, the city, and the environment’.

Problem: The idea of continuity in garden traditions

Studies on ‘Islamic’ garden history have followed two paradigms whose generalising and essentialist character has drawn criticism from various scholars over the years. The first argues that the concept and types of gardens in the Islamic period – drawing on extant examples only from the fifteenth century onwards – were of ancient Persian origin, thus presupposing the idea of a continuous tradition. The second paradigm brings together gardens from different regions with Islamic rulers as ‘Islamic’ gardens, foregrounding a religious attribute for gardens in an extremely large and disparate geographical area including Arab Spain, North Africa, Syria, Iraq, Iran and the adjacent Persian-speaking regions up to India, despite the diversity of their regional and cultural traditions.

The idea of an ancient ‘Persian’ origin of ‘Islamic’ gardens and its continuity into the early modern period has been voiced since the beginning of the twentieth century. It mainly relates to the interpretation of the term chahārbāgh, found in the Persian textual sources from the fifth/eleventh century onwards, as a formal type of garden, indicating a garden divided into four quarters by two cross-axial water courses or paths. It is argued that this type can be traced back to pre- Islamic Iran, continued throughout the medieval Islamic periods and was transferred to other Islamic-ruled regions as well. Some scholars have also related the fourfold layout and chahārbāgh to the mythical Garden of Eden, from which, according to some traditions, such as the book of Genesis, four cosmic rivers, two of which were the Euphrates and Tigris, flowed towards the four corners of the earth.

As an intricate interplay of architecture and nature, a garden comes to life through its plants, living creatures, and the sensory experiences of scents and sounds. In comparison to a garden's solid structures, however, even fewer physical traces of the organisms that once populated it have survived over the centuries. Archaeozoological and archaeobotanical findings, in some cases, can help to identify certain varieties of living creatures and discern the plantings within a garden or the locations of certain plants. Environmental studies, including soil analysis, also provide valuable insights into whether an area was exposed to sunlight or shaded, featured high-growing or low-growing vegetation, and embraced a dense or sparse plantation. However, so far, no such investigations have been carried out in the Abbasid palace gardens of Lower Mesopotamia. Nor has archaeological evidence regarding horticulture, such as soil contours that could reveal the potential remains of flowerbeds, been recorded thus far. Furthermore, unlike ancient Mesopotamia, no representations of the royal gardens, depicting their plants or living creatures, are known to exist from the Abbasid period. Consequently, literary evidence remains the only source of knowledge regarding the flora and fauna of the Abbasid palace gardens in Mesopotamia. Nevertheless, the textual sources also come with their own set of limitations. They do not provide a comprehensive list of all plants or living creatures that existed in the royal gardens, nor do they offer details of planting arrangements or density. These sources, however, offer vivid glimpses into various species in the royal gardens and shed light on the types of plants and animals that were particularly appreciated. They also contribute to our understanding of how two main factors – prestige and climatic conditions – influenced the choice and organisation of plants within these gardens.

The physical remains of the Abbasid palace gardens in Lower Mesopotamia have largely disappeared and archaeologists have mostly ignored these gardens. In a letter to the orientalist Carl Heinrich Becker, dated 25 March 1911, Herzfeld wrote, ‘In Berlin ist man über die ornamentierten Wände sehr erfreut. Und Seine Majestät hat 15,000 Mark gestiftet. Ich hoffe noch weitere Mittel zu bekommen … .’ Herzfeld's publications and personal notes attest that the prime interest of the German expeditions during the two campaigns in Samarra that took place from 1910 to 1913 was merely to clarify the general plans of the buildings and specifically to discover new varieties of stucco wall decorations. As a result, the gardens, as vast open spaces adjacent to the palace buildings, were largely regarded as unimportant to the aims of their investigations. Subsequent excavations at Samarra led by the Iraq Directorate-General of Antiquities since the 1930s have predominantly concentrated on specific structures, and the garden expanses rarely had a place in their investigations.

Archaeological evidence attests to the existence of gardens attached to three palaces at Samarra: Balkuwārā, the Caliphal Palace and Jaʿfarī. The physical traces of the garden of Jaʿfarī Palace are minimal and primarily indicate its proximity to the river and a large rectangular pool, which have already been discussed in Chapters One and Three. Thus, this chapter will focus on the gardens of the Caliphal Palace and Balkuwārā, which both have particular significance to the study of gardens of the Abbasid period but for distinct reasons. Balkuwārā is the only site where a clear outline of the garden is evident, whereas certain key features of the garden of the Caliphal Palace can be discerned which permit reasonable speculations as to its layout and principal design.

Gardens built during the period of Abbasid rule in their Lower Mesopotamian heartland have hitherto been one of the least investigated subjects in the fields of garden studies and Islamic architecture. This lack of inquiry can be primarily ascribed to the inherent nature of these gardens, which have left fewer tangible remnants, coupled with a general disinterest in excavating these expansive areas. Yet, excavations at some palace complexes, particularly those in Samarra during the early twentieth century, complemented by aerial photographs, offer a glimpse into the gardens associated with these palaces and their relationships with broader architectural and environmental settings. Furthermore, textual sources from the period contain a wealth of information about the palace gardens, which offer detailed descriptions, necessary context and historical insights. Combining this meagre material evidence with contemporary textual sources transforms these fragments into a valuable resource, offering a comprehensive and nuanced understanding of the spatial arrangement, functions and cultural significance of these overlooked landscapes. This interdisciplinary approach not only permits the examination of these gardens from diverse perspectives but also forces us to revisit some prevailing notions regarding the spatial arrangement and function of the adjoining covered spaces. These ideas, formulated primarily by the archaeologists who excavated these palaces in the early twentieth century, have been adopted but not fundamentally questioned by later scholars.

In some Abbasid palaces of Samarra, a distinct arrangement of halls, which so far have been identified as the throne hall, audience hall or reception hall block, is recognisable. This configuration comprises four longitudinal halls set at the four sides of a square chamber that is thought to have been covered by a dome. Each hall opens directly or through a portico onto a courtyard or a garden. Based primarily on archaeological findings from two palaces – Balkuwārā and the Caliphal Palace – Herzfeld proposed in the early twentieth century that the cruciform complex functioned as the throne hall, whereby the central chamber was used for private audiences and the semi-open halls (iwans) for public audiences. Herzfeld traced the architectural precedent for this arrangement to the early Abbasid palace of Abū Muslim (d. 137/755), the governor of Khorasan, which stood at Marv and was described by the fourth/tenth-century geographer al-Iṣṭakhrī. According to al-Iṣṭakhrī, the governmental palace (Dār al-ʾImāra) contained a domed chamber (qubba) ‘in which [Abū Muslim] used to sit. … The domed chamber (qubba) has four doors, each leading to an iwan, …, and in front of each iwan is a square courtyard (ṣaḥn).’ Herzfeld, followed by Creswell, further suggested that several palaces described in textual sources as having been built during second/seventh–third/eighth century in the Levant and Iraq and referred to as Qubbat al-Khadhrāʾ (translated as ‘The Green Dome’ or ‘The Dome of Heaven’), probably featured a similar arrangement of audience halls.

Edinburgh Studies in Islamic Art is a venture that offers readers easy access to the most up-to-date research across the whole range of Islamic art. Building on the long and distinguished tradition of Edinburgh University Press in publishing books on the Islamic world, it is a forum for studies that, while closely focused, also open wide horizons. Books in the series, for example, concentrate in an accessible way, and in accessible, clear, plain English, on the art of a single century, dynasty or geographical area; on the meaning of works of art; on a given medium in a restricted time frame; or on analyses of key works in their wider contexts. A balance is maintained as far as possible between successive titles, so that various parts of the Islamic world and various media and approaches are represented.

Books in the series are academic monographs of intellectual distinction that mark a significant advance in the field. While they are naturally aimed at an advanced and graduate academic audience, a complementary target readership is the worldwide community of specialists in Islamic art – professionals who work in universities, research institutes, auction houses and museums – as well as that elusive character, the interested general reader.

Looking at the deserted remnants of the palace gardens at Samarra today, it is perhaps too easy to forget that these grounds once featured vast and elaborate waterworks. Indeed, there are numerous references to large pools, basins, fountains and water channels in contemporary descriptions of gardens and courtyard gardens of the Abbasid palaces in Lower Mesopotamia. However, like the gardens themselves, the water features have rarely been subject to archaeological investigations. The fragmentary evidence is primarily drawn from aerial photographs and archaeological surveys conducted in the early twentieth century, particularly those pertaining to the Caliphal Palace of Samarra. These sources, while sparse, still offer vital clues regarding the abundance of water and its indispensable role in shaping the design of these gardens. By combining this limited material evidence with abundant textual sources, this chapter presents a picture of the diverse water features that once adorned these gardens. It seeks to illuminate their critical role in both garden design and the everyday life at court, offering a comprehensive understanding of how water was intricately interwoven into the fabric of these built environments.

Birkas

Textual sources indicate that a prominent feature of Abbasid palace gardens was a large pool, referred to as a birka. The definitions of birka offered by contemporary philologists help us to understand some of its formal features. The Buyid vizier and lexicographer Ibn ʿAbbād (d. 385/995)1 defined birka as a structure ‘similar to a basin (ḥawḍ) dug into the ground’.2 Kitāb al-ʿAyn (compiled about 184/800 in Lower Mesopotamia) offered a more detailed definition, describing birka as a structure ‘similar to a basin (ḥawḍ), which is dug into the ground and does not have raised sides (aʿḍād: pl. of ʿiḍd) above the ground's surface’.