Kimiaki Tanaka's book An Illustrated History of the Maṇḍala: From Its Genesis to the Kālacakratantra represents a milestone in research about the development of the Buddhist mandala, made available with this publication in English eight years after the Japanese version (289). What distinguishes this book from other studies on the topic available so far is its approach, the breadth of the sources used—covering multiple languages from South Asia to Japan—and its methodology. Concerning the latter, the author states that he aims “to elucidate the historical development of esoteric Buddhism and the maṇḍala, paying particular attention to proper names, mudrās and mantras” (11). More precisely, the author uses proper names of deities, their gestures or symbols (mudrā), their speech forms (mantra), and the doctrinal categories associated with them to establish connections between textual sources and to draw conclusions from these on the interrelations between texts. He then uses those interrelations to build a development of the ideas surrounding the mandalas that are mentioned or described in them. Considering the languages involved, the fluidity of the terminology used in esoteric Buddhist texts, and the issues around their dating, the chosen methodology sounds much simpler than it is, and in many ways relies on the validity of previous, predominantly Japanese, scholarship around the different sources used, their dates, and their interpretations. This approach is nevertheless extremely productive and ultimately forms the basis of the book's structure and a range of graphics illustrating it.The chosen methodology builds on the fact that historically the most successful esoteric Buddhist texts have largely been created within the Buddhist tradition and, once spiritual gain became their main aim, built on conceptual and doctrinal frameworks already existing within the tradition. This is also apparent from their cross-referencing and the classification of esoteric Buddhist texts that develop along with them.1 In Tanaka's book this becomes particularly apparent with the Kālacakra system, the base ideas of which are analyzed in some detail in chapter 7, “The Maṇḍala of the Kālacakratantra,” the book’s last historical chapter. Five of the six chapters preceding it largely focus on the respective emergence of other key texts exemplifying the development of esoteric Buddhism, the exception being chapter 3, “Maṇḍalas of the Prajñāpāramitānayasūtra,” between the Garbha- and Vajradhātumaṇḍala (see below). The book concludes with a chapter summarizing “The Development of the Maṇḍala and Its Philosophical Meaning.”While the book's aim is to reconstruct the development of the mandala in India, the early chapters are conceived around the famous Japanese twin mandalas of Vairocana—the Womb World mandala (Genzu mandara) and Diamond World mandala (Kue mandara), respectively—the earliest extant version of which can be dated to 859–880 ce.2 While this is understandable, given that they are among the earliest representations that have come down to us, it has to be kept in mind that they also represent very specific interpretations.3Given the methodology and the topic, the text of the book is dense, full of details and references, and at times difficult to follow, especially where it presumes previous knowledge or summarizes different positions taken in Japanese scholarship. Here it builds on a dialogue that is not reflected as such in English but which is interesting in itself. Basic knowledge of the development of esoteric Buddhism and its key texts, thus, makes the book a more rewarding read, but it can also be used to gain some of that knowledge. Further, the book is an invaluable source for references, illustrations of well-known and obscure mandala representations, as well as references and ideas for interconnections that are worth following up in greater depth.In this review I summarize the content of the respective chapters and comment on them in the light of my own research and teaching on early esoteric Buddhist art from India and Tibet. My main aim is to provide a rough guide to the publication, complement the information provided by the book where appropriate, and to point out the few instances where the book takes a problematic position.Considering the definition of the term mandala, it needs to be noted that, in my view, the continued expansion of the semantic range of the term—culminating in the understanding by C. G. Jung and in Western psychology more broadly (briefly referred to on pp. 280–81)—as a universal symbol has severely distorted the academic perspective on the subject.4 This fact may also be reflected in Tanaka's own definition of the mandala in the introduction: “Here, I define a mandala as an icon that represents the worldview of Buddhism, or Buddhist cosmology and philosophy, by arranging Buddhist deities in accordance with a specific pattern” (2). Clearly this definition is based on the appearance and doctrinal background underpinning the twin mandalas in Japanese esoteric Buddhism as well as the broader usage of the term mandara in Japanese Buddhism. It is remarkable that neither the ritual function of the mandala nor the actual meaning of the term, indicating a circle or an arrangement with center and surrounding, is adequately represented in this definition. In my opinion those two elements are crucial for the understanding of the early mandala, the former explaining its function and the latter the fact that the circular arrangement referred to by the term is the central assembly of deities.5The importance of the central arrangement also becomes clear from Tanaka's first chapter, “The Genesis of the Maṇḍala.” The author builds an arch from early triadic compositions, via preaching scenes described for cloth paintings in the Mañjuśrīmūlakalpa, representations of buddha fields, five Buddha arrangements, and arrangements of the eight great bodhisattvas to wrathful deities and center arrangements in different shapes. The chapter points out that deities in some mandalas can be traced to the preaching scenes at the beginning of Mahāyāna sutras. The example singled out here, the mandala deriving from the Bhaiṣajyagurusūtra (figs. 1.5 and 1.6), may be misleading insofar as no early representations of this mandala are known.6 In my opinion, the origin of the mandala arrangement for the deities of this sutra remains unresolved, and a Tibetan imperial date for the composition (19) is highly unlikely.Another aspect discussed here, the “landscape maṇḍala” (setsue mandara) accounts for a specific, Japanese interpretation of preaching scenes as mandara. In this regard, Tanaka cites the interesting case of the Mahāmaṇivipulavimānakalparāja, which describes an icon of a preaching scene and a mandala composed of the same set of deities (21). This indicates why Tanaka begins his account with generic preaching scenes, and why his definition of the mandala does not emphasize its ritual function, since it would exclude the former as representing a mandala. In Japanese Mandalas, Elizabeth ten Grotenhuis distinguished this specific group and others deriving from it as “mandara,” using the Japanese transcription for mandala, to emphasize their local Japanese origin.7 Tanaka's discussion of the preaching scenes in the context of his publication might imply to the reader that there is an Indian precedent for this understanding of the mandala, but that is not the case.The following account of ideas around directional Buddhas eventually leading up to the five esoteric Buddhas of diverse tantra demonstrates nicely the value of Tanaka's methodology. In addition to the eastern and western Buddhas, deriving from the respective pure land traditions, he also traces the southern and northern Buddhas back to earlier mentions in texts ascribing a particular direction to them. Figure 1.11 converts those relationships into an extremely informative graphic that is perfect for teaching this subject. The short discussion of Buddha representations in the four cardinal directions (27–29) is rudimentary, as potentially earlier examples are found in Gandhara and Kanaganahalli.8 In the latter monument the Seven Buddhas of the past and Ajita ( = Maitreya) are distributed around the stupa at a later phase, providing another base for identifying an exoteric version of directional Buddhas as further exemplified by the reliquary of Śrī Kṣetra.9 In other words, the discussion of this phenomenon in art and architecture cannot be reduced to cases in which the direction of Buddhas is explicitly stated. This is an example for the approach of the publication at large; the purpose is not to be comprehensive but to discuss only those examples that (potentially) represent an earlier development of something ultimately included in a mandala.In my opinion, depictions of the eight bodhisattvas in a nine-fold grid are the earliest Buddhist depictions of a mandala in India. Tanaka's chapter on the genesis of this group cites several seventh-century Chinese textual sources describing them and their importance for the Womb World mandala. Further, Tanaka emphasizes compositions around Buddha Vairocana seated in meditation, but it is important to note that this composition can be presided over by different Buddhas as inherent in the system of the three families that underpins many of the cited textual sources. Counteracting the three poisons that keep us in saṃsāra, the Buddha, lotus, and vajra families that comprise the three-family system can be represented by a variety of deities representing the required qualities.Finally, this chapter considers wrathful deities, uṣṇīṣa-deities,10 and variations of central circle shapes used for a central deity surrounded by eight secondary ones. Covering a wide range of somewhat disparate topics, this chapter does not stand on its own but lays the groundwork for what is discussed in the following ones. As such it does not adequately represent the emergence of esoteric Buddhism and its rituals associated with the mandala, about which much more could be said on the basis of the textual and visual material that has come down to us—in particular, the dhāraṇī literature, which in the last decades has been studied in considerable detail.11Chapter 2 focuses on “The Emergence of the Garbha-maṇḍala,” the Womb World mandala mentioned above, and puts it in dialogue with a considerable range of other mandalas with what I call an “asymmetric distribution of deities” that is characteristic of many of the earlier mandalas. In asymmetric mandalas, deities are distributed unevenly, more on one side than on the other, and/or deities of different qualities or gender are shown on the same level. Often these mandalas expand on the concept of three families mentioned above, which Tanaka's chapter documents. His figures illustrating the principal elements of the described mandalas and the distribution of deities inside the mandala are extremely informative and demonstrate the different applications of the three-family concept. Equally informative is the discussion of the different interpretations in Japan and Tibet, which demonstrates the importance of the commentaries as well as other factors, such as the influence of further developed mandalas, on the respective appearance of the Garbha mandala.The fact that the surviving depictions of the Garbha mandala are much later than the textual sources at their base, and that those textual sources are fully understandable only if their explanatory commentaries are taken into account, complicates the discussion of this chapter. As is to be expected, the chapter pays particular attention to the Japanese version and the evolvement of its axial symmetric composition, which contradicts most of the mandalas presented in the chapter. In the conclusion it is assumed that this was already a concern of earlier masters of the tradition (83), but we have to keep in mind that their interpretation might have been tinted by the existence of more developed Yoga Tantra mandalas. The chapter demonstrates the complexity of the main mandalas developed at this stage—largely the seventh century—and the competition of ideas around all-encompassing mandalas.12 That the Garbha mandala is considered “the culmination of a maṇḍala system that developed in early esoteric Buddhism” (85) is obviously the result of the historical process that ultimately brought this mandala to the fore, probably because of its focus on Vairocana.From the perspective of Indian esoteric Buddhism, chapter 3, “Maṇḍalas of the Prajñāpāramitānayasūtra,” is probably the most surprising, as this text has not received much attention outside Japan. Of course, in the Sanskrit world this text is better known as Adhyardhaśatikā Prajñāpāramitā or Prajñāpāramitā in 150 verses, and different versions of it were translated by Xuanzang (602–664; translation 660–663 ce), Bodhiruci (translation 693 ce) and Amoghavajra (translation 763–771 ce), the latter two closer to each other (87).13 As Tanaka sees it, the main innovation of this text is for the named deities to personify “doctrinal propositions” in groups of four and five deities (88), but he references this text throughout the first half of the book, including with the eight-spoked wheel in chapter 1 and in relation to a distinct group of the eight bodhisattvas deriving from it.Among the important topics covered in this chapter are mandalas centered on Vajrasattva, including a seventeen-deity mandala of Vajrasattva, which is identified with the so-called Rishu-e mandara, the ninth mandala in the Japanese representation of the Diamond World mandala (Kue mandara). Here Tanaka also references mandalas in the Alchi Sumtsek, Ladakh, India (early thirteenth c.; fig. 3.9) and Jampa Lhakhang in Lo Manthang, Mustang, Nepal (second quarter of the fifteenth c., figs. 3.5 and 3.6), but in my opinion both of them can be explained on the basis of the Sarvatathāgatatattvasaṃgrahatantra alone. More important, Tanaka also discusses the five-deity mandala of Vajrasattva with the four inner offering goddesses that is represented in a number of early examples from India to Dunhuang.Somewhat confusingly, the second half of the chapter is dedicated to a detailed discussion of the Paramādyatantra and its mandalas, the root texts of which have been translated from the eleventh century onward. It is discussed here because of its links to the Prajñāpāramitānayasūtra. This complex discussion involving a rich commentarial tradition to all the sources mentioned until this point is interesting for the development of Yoga Tantras more broadly and its relationship to later depictions.Overall, the relevance of the Adhyardhaśatikā Prajñāpāramitā for the development of ideas around the mandala more broadly remains unclear, as the innovations ascribed to this text largely build on Amoghavajra's interpretation of it as it has been canonized in Japan. There is, in fact, no conclusive evidence that this text was of similar relevance in the development of Indian esoteric Buddhism. As Tanaka points out (91), its specific set of eight bodhisattvas has not yet been identified among the Indian depictions of this topic. However, it is clear that the doctrinal content of the Adhyardhaśatikā Prajñāpāramitā aligns well with the Vajradhātu mandala and that the mentioned deities form a base of Paramādyatantra (fig. 3.11).Chapter 4, “The Emergence of the Vajradhātu-maṇḍala,” is largely a systematic account of the different groups of deities assembled in the main mandala of the Sarvatathāgatatattvasaṃgrahatantra, the Vajradhātumahāmaṇḍala. This discussion brings together some of the strands covered in the earlier chapters, including the relation to the main texts covered in the previous chapter. It also assembles some of the evidence for Vajradhātu-mandala deity depictions in India, in particular the representations of the main Buddhas of this system with the secondary deities surrounding them in their respective quarters.In one section Tanaka takes up the five esoteric Buddhas again and presents a rather awkward theory of their emergence (154–57). Trying to rationalize the legend of the iron stupa—better known in the Sanskritic world as the Dhānyakaṭaka stupa—at which the Buddha taught the tantras, and its frequent identification with Amarāvatī, Tanaka explains the five āyaka pillars on each side of the Amarāvatī stupa as potentially standing for one of the secondary Buddhas flanked by the four bodhisattvas that accompany them in the Vajradhātu mandala (155). Even without the contradicting evidence of Kanaganahalli, where only four such pillars have been found on the eastern platform and are depicted in stupa reliefs, this is a stunningly ahistorical outlier for Tanaka's book.Toward the end of the chapter, early representations of the Vajradhātu mandala in Japan and Tibet are taken up again as evidence for its importance but without discussing any of their details. Other Yoga Tantras are covered in the final section in equally summary form, mainly considering if they are to be counted among the early Yoga Tantras or should be regarded as a later development. These include the mandalas deriving from the different version of the Sarvadurgatipariśodhanatantra and the Dharmadhātuvāgiśvaramañjuśrī mandala. In that regard, it is surprising that the five-Buddha system distinct for the Durgatipariśodhana cycle does not receive further attention even though it was mentioned with the Paramādya-Vajrasattva mandala in the previous chapter (111).Further, it is important to note that while the Indic Vajradhātu mandalas are comparable to those in Japan, they are best interpreted within their own historical frameworks as distinct interpretations. In this regard we might consider the ten mandalas of the Alchi Sumtsek (briefly mentioned at 147–48). While their selection is partially comparable to the nine mandalas of the Kue mandara, the interpretation of the mandalas is clearly distinct from the Japanese ones. For example, the Ekamudrā mandala, in Japan represented by a single image of Vairocana, is represented in Alchi by a thirty-seven-deity mandala of Vajrasattva featuring only offering goddesses and bodhisattvas of the fortunate aeon as secondary deities. Tanaka uses parts of this mandala as comparison to the Rishu-e mandara traced back to the Prajñāpāramitānayasūtra (see above).Beginning with the discussion of “The Emergence of the Guhyasamāja-maṇḍala” in chapter 5, the character of the account changes, as the text almost exclusively relies on Indian and Tibetan sources. The chapter thus presents a rationalization of the internal evidence found in the relevant textual sources of the tradition and their later interpretation without building on their relation to Japanese expressions of the mandala. This is most apparent in the discussion of the emergence of deities in sexual embrace (173–78), which curiously is not linked back to the Vajradhātu mandala. There the generation of the secondary mandala deities through sexual intercourse is already implicit in the gesture of Vairocana—the gesture of highest awakening (bodhyagrīmudrā), in which the index finger of the right hand is embraced by the left hand—and his immediate entourage of four goddesses who, in some texts, are designated as the “mothers of the families” of the deities in the respective quarters. However, in the Guhyasamāja corpus, sexual symbolism is both explicit and expressed in a transgressive language.In this chapter, the two main mandalas deriving from this corpus—the Mañjuvajra and Akṣobhyavajra—are discussed under the headings of the Jñānapāda- and Ārya-schools, respectively. It is not surprising that early representations of the deities of these mandalas in India and Tibet did not show the deities in sexual embrace. Other interesting aspects covered in this chapter are the core thirteen deities of the mandala and its relationship to the Forty-two Peaceful Deities mandala, for which there is evidence in Dunhuang, and the symbolism of the mandala deities with matter comprising the universe, its perception, and the body parts of the practitioner (fig. 5.10). The chapter briefly mentions other so-called father tantras,14 such as those of different forms of Yamāntaka and the Māyājāla mandala, for which Tanaka also identifies Indian sculptural evidence for some of the main deities (fig. 5.15).Chapter 6 discusses the “Maṇḍalas of the Mother Tantras” with an emphasis on the Hevajra and Cakrasaṃvara cycles. Their derivation is seen in the Samāyogatantra, the early date of which appears to be confirmed through a description by Amoghavajra that appears to be based on it (200–201). Tanaka further proposes to link this tantra, in the manifestation of the all-encompassing Six-Family Maṇḍala, back to the seventeen-deity mandala of Vajrasattva of the Paramādyatantra (203–7). He also credits it with the introduction of Heruka deities, the (semi-)wrathful main deities of mandalas, an assembly of Heruka forming the sixth, and in some interpretations central, family of the Samāyogatantra (203–9). Of course, the proposal that the obscure term Heruka may in some way have derived from Hercules (207) should be dismissed.More important here is the fact that wrathful deities arise in the Buddhist pantheon as Buddha-forms with the ability to convert or suppress Hindu- and pan-Indian deities to Buddhism, Śiva in particular. Tanaka cites this aspect in relation to the deities converted in the Vajradhātu cycle by focusing specifically on the mother-goddesses (208–9), but does not go into any detail concerning its relevance for the iconography of Cakrasaṃvara himself or for the emergence of the charnel ground imagery. Instead, the sections on the Hevajra (209–13) and Saṃvara (214–22) focus on key mandalas from these cycles and their symbolism (222–26). Further mother-tantra mandalas are mentioned briefly (226–28).In other words, while the links to the Samā-yogatantra made in this chapter are important, many aspects of the emergence of those mandalas are only touched upon, if mentioned at all. Further, presenting the father and mother tantras separately, the derivation of their attribution to the vajra family of Buddha Akṣobhya remains underexplored. In this regard, there is a clear trajectory from the attendant Vajrapāṇi as the force of the Buddha in Gandharan art, to Bodhisattva Vajrapāni's power-countering aggression in the three-family system, to the vajra-family taking on its symbolism in the five-Buddha family system of the Vajradhātu, and the attribution of deities forcefully converting Śiva and his entourage to the vajra-family.Chapter 7, “The Maṇḍala of the Kālacakratantra,” considers the symbolism that has been combined in its conception. Tanaka's methodology of tracing iconographic and doctrinal aspects of the Kālacakra in earlier Buddhist literature comes to full fruition here as it mirrors the conceptual nature of this mandala. In a concise and clear account, he demonstrates that the Kāyavākcittapariniṣpanna Kālacakra mandala draws on a range of ideas already present in earlier tantras, and combines them in a holistic manner with its equivalent on a macro- and microcosmic level, the latter being the body of the practitioner. Thus, after accounting for the historical information available on the Kālacakra and introducing its two main mandalas, the second one being the Kālacakra Mahāsaṃvara mandala that has a round palace with eight doors, Tanaka sets out its concept of non-duality through the inclusion of theories underpinning earlier tantras (239–45).In terms of appearance, the most important aspect of the Kālacakra mandala is its base in its own theory of cosmology, which Tanaka subsequently lays out (245–55). As Tanaka demonstrates, it is the color scheme of the cosmos that takes precedence over that established through earlier tantras, in particular the Vajradhātu and its five-Buddha family system to which subsequently developed mandalas were related as well. In that connection it needs to be emphasized that while some Yoga Tantra mandalas already associated the mandala with the cosmos and included elements of it, such as its distinct color scheme of the five Buddhas (246) or the continents (see also 278–80 for a summary), it is only the Kālacakra system that cross-identifies them and includes the four elements in the buildup of its mandala. The circular discs of the elements that make up the cosmos, which are conceptually embedded in the measurements of the mandala palace (fig. 7.9) and visually represented as rings surrounding the outer palace, also make the outside of this mandala inevitably round. The circular enclosure of the mandala may have its precedent already slightly earlier, see Tanaka's discussion on this aspect on pages 263–64, but it may well have been the success of the Kālacakra that made it an almost universally adopted element of the mandala representation.Having worked through the development of ideas surrounding the mandala in the second half of the first millennium, Tanaka concludes his study with “The Development of the Maṇḍala and Its Philosophical Meaning,” where he uses his previous studies of Tibetan mandala texts and sets of mandala depictions to reflect on the comparative differences between mandalas. On the visual side, Tanaka considers the differences in the geometric appearance of the mandala, the representation of its deities relative to the main deity, and the color schemes used for the quarters of the mandala palace in relation to the five Buddha families. The latter is important, as it provides a lot of information on the conceptual background for the respective mandala depictions. Tanaka then summarizes the doctrinal interpretations of different parts of the mandala palace or its deities, each subchapter providing another perspective on a subject already covered in earlier chapters.Overall, Tanaka's comprehensive study on the genesis of the mandala succeeds in laying out some of the underlying relations between the different esoteric Buddhist sources and the mandalas described in them, but one has to be aware that his presentation is tinged in two ways. On the one hand, much of his presentation around the early tantras is seen through the eyes of Japanese scholarship on the emergence of the twin mandalas. As such, the interpretations of Amoghavajra and Kūkai (774–835) loom large, whereas those of the later Indian interpreters of the same sources receive relatively little attention. On the other hand, the treatment of the later tantras strongly relies on much later Tibetan classifications and interpretations, an example being the classification of the Highest Yoga Tantras in father, mother, and non-dual tantras. Equally, the fact that Tibetan depictions of mandalas are based on the discourse apparent in Tibetan literature and become more and more like each other over time comes too short in Tanaka's account.The main strength of the publication is that it documents concisely the development the mandala underwent over at least half a millennium, from its emergence to its most fully developed form as exemplified by the Kālacakra. Laying out that development, the book not only provides insights into this process but also demonstrates that the mandala as we know it is a product of an accumulation of visual and symbolic properties. More important than the account of this process is the network of interrelations established throughout, the exploration of which is helped by a comprehensive index. As such, the publication succeeds to bridge the obvious discordance between the esoteric base for the mandala and its visual depiction as preserved today in different Buddhist cultures, and it provides a good base for further inquiry into any of the stages of its development.