In Spells, Images, and Ma ṇḍalas, Koichi Shinohara constructs a sweeping ritual history of East Asian esoteric Buddhism. Shinohara's central concern is the role of images—painted and sculpted figures of Buddhas, bodhisattvas, and other deities—in Buddhist ritual manuals. Shinohara claims to have identified a pattern of ritual evolution in this literature, as images are introduced into Buddhist rituals only to be later supplanted by practices of visualization. Shinohara makes these claims with modesty—evolution is a mere “conceptual model” (p. xiv)—but his admirable caution and humility cannot conceal the startling ambition of this study.

Part one, “Three Ritual Scenarios,” draws from numerous Buddhist scriptures to illustrate three phases of ritual evolution. In Scenario One, found in early dhāra ṇī scriptures, magical spells are recited for some concrete purpose. The efficacy of the spell might be confirmed through visions of Buddhas. Over time, dhāraṇī scriptures would come to be affiliated with a specific deity or bodhisattva, such as Avalokiteśvara. This development leads to Scenario Two, in which rituals for painting, sculpting, and venerating images of a central deity begin to appear in Buddhist grimoires. This shifts the method of confirmation. Instead of visions of Buddhas appearing before us to affirm the success of our spell, we now rely on image miracles; the image itself quakes, radiates light, or speaks. In Scenario Three, ritual spaces dedicated to individual deities are grouped together into a larger pantheon, forming complex maṇḍalas. Shinohara explores the earliest evidence for the emergence of such an “esoteric pantheon”—the All-Gathering Maṇḍala Initiation Ceremony in Atikūṭa's seventh-century Collected Dhāraṇī Sūtras (T. 901). Shinohara argues that in these large maṇḍalas the image becomes decentered and ambivalent. The overwhelming focus of the rituals is on mantras, mudrās, abhiṣeka, and homa. The most important innovation of this maṇḍala is the initiation ritual, in which the blindfolded practitioner throws a flower into the arena to determine his affiliation with a specific deity. In a bold conclusion to part one, Shinohara traces the key developments in maṇḍalic rites back to the Indian religious world, showing parallels between Atikūṭa's All-Gathering Maṇḍala rite and the rituals of the Gṛhyasūtrapariśiṣṭas.

Part two, “The Evolution of Dhāraṇī Sūtras and the Introduction of Visualization Practice,” delves deeply into dhāraṇī scriptures translated by Bodhiruci (d. 727). Each of these scriptures has multiple translations, allowing Shinohara to analyze changes in ritual programs over time. Shinohara demonstrates an increasing focus on visualization practices from the seventh and eighth centuries onward. In earlier ritual scenarios, we must wait for a vision of Buddhas to appear before us. Later, we conduct rituals around a central image, hoping that the image will confirm our success with a miracle. Now we can directly visualize Buddhas in the ritual space.

In part three, “Toward a New Synthesis,” Shinohara explores the “mature” rituals of visualization found in later East Asian traditions. First, Shinohara shows how Yixing's commentary on the Mahāvairocanasūtra modified the ritual plan of the All-Gathering Maṇḍala through the extensive use of visualization. In this mature system, the practitioner may visualize the maṇḍalic deities and merge them with his own body. The role of images diminishes, as the practitioner becomes the deity. Next, Shinohara explores Amoghavajra's ritual manuals, which reshaped earlier Avalokiteśvara texts into a Vajraśekharasūtra-inspired maṇḍalic system. In these manuals, Avalokiteśvara fades into the background, and the deities of the Vajra group dominate.

Shinohara's work would be impressive enough as a work of “mere” description. Scholars owe Shinohara a debt of gratitude for his thoughtful exploration of a massive and largely unexplored body of literature. But even more important are Shinohara's contributions to the field of esoteric Buddhism. I will conclude with two examples.

First, Shinohara has made a strong argument against Ronald Davidson's account of the origins of esoteric Buddhism. Davidson argues that the Buddhist maṇḍalic rites emerged when Buddhists mirrored and “sacralized” the medieval political world.Footnote ⁵ Shinohara's work argues instead that Buddhists simply appropriated an available repertoire of post-Vedic Indian ritual (p. 88). Shinohara's work will certainly stimulate further studies on the relationship between Buddhist and “non-Buddhist” ritual systems in South Asia.

Second, Shinohara tells a powerful story about the causes of ritual change. Why were images so ambivalent in Buddhist ritual? Once the confirmation of ritual success shifts from a private vision of Buddhas to a publicly available image miracle, the ritual specialist loses “control” over the ritual performance (p. xiii). Shinohara sees the development of complex maṇḍalic rites as an attempt to reassert control (p. 45). New techniques of visualization remove criteria for the ritual's success from the sensible public realm. Shinohara attempts to explain ritual change by thinking of Buddhist ritualists as historical actors, motivated by a desire to control the space of ritual interpretation. This is as much an economic argument as it is a psychological argument. I expect that Shinohara's hypothesis here will be controversial, and will spark valuable debates at the intersection of Buddhist intellectual, ritual, and social history.