Kannon (Avalokiteśvara)

The Kannon problem

Kannon (Avalokiteśvara, 観音菩薩) is the most-represented figure in the Japanese Buddhist visual record. The bodhisattva of compassion appears in seventh-century gilt-bronze Asuka pieces, in late-Heian Pure Land paintings, in Kamakura sculpture, in Edo Rinpa monochrome, and in 1888 in the canonical Meiji Nihonga work that proposes a Japanese-modern way of painting the figure. Everything in this cluster sits at some point on that arc.

The cluster is held together not by a single iconographic programme but by a typology — the seven canonical Kannon variants codified in the Mikkyō esoteric tradition: Shō Kannon, Senju Kannon, Jūichimen Kannon, Bato Kannon, Nyoirin Kannon, Fukūkenjaku Kannon, Juntei Kannon. Each variant is fixed by an iconographic combination — face count, arm count, attribute — that an attentive viewer can read at a glance once the variant grammar is known.

Reading Japanese sculpture and painting of Kannon is in large part a matter of reading the variant first, the period second, the workshop third. The disambiguation table below carries the seven-variant grid; the articles in this cluster sit against that grid as worked examples.

What the cluster covers

Twenty articles, ordered loosely by what the reader most often needs to look up:

The variant guides — the iconographic reading entries for each canonical Kannon form. These are the front-door pages for someone who has a Kannon in front of them and is trying to identify which variant it is. The five-and-a-bit canonical reading entries (Shō, Senju, Jūichimen, Bato, Nyoirin, Fukūkenjaku) cover the variants with substantial Japanese sculptural and pictorial witness; Juntei is deferred as a post-launch reading entry on present corpus evidence.

The single-work studies — articles built around one named object. The Met’s 19.140 standing Amida is here as a Kannon-adjacent comparandum; the 1254 Sanjūsangen-dō chief image by Tankei is the cluster’s monumental Senju anchor; the Asuka-period Yumedono Guze Kannon is the cluster’s seventh-century starting point; the Met 2019.419.2 Sakai Hōitsu willow Kannon is the Edo Rinpa monochrome entry; the Hashimoto Gahō / Kanō Hōgai 1888 Hibo Kannon is the Meiji terminus.

The specialised pieces — the Met 2013.114 thousand-armed Kannon embroidery (Senju as an embroidered surface), the Met 2016.517 Chōmei-ji pilgrimage mandala (Kannon as the centre of a pilgrim-route programme), the Tokannon-ji 1271 Bato kakebotoke (Bato in the hanging-plaque format rather than as full sculpture).

What to read first

Three entry points depending on what the reader brought:

If the reader is looking at a Kannon and is trying to name the variant, start with the disambiguation table below. The fastest variant marker is face count, then arm count, then attribute. A figure with eleven small heads stacked vertically is Jūichimen; a figure with a horse head on the crown is Bato; a figure with one head and two arms is almost certainly Shō; a figure with many arms is Senju.

If the reader is coming to the tradition for the first time, the Shō Kannon canonical anchor is the cleanest entry point — it explains why the simplest Kannon variant is the one against which the others are read. The Jūichimen iconographic reading guide is the second-best starting page.

If the reader is interested in a specific century, the cluster’s range stretches from c.650 (Asuka period; the Yumedono Guze Kannon, the Asuka gilt-bronze Kannon) through the late Heian (Mikkyō Six-Kannon codification, the Heike Nōkyō gold-on-indigo programme) and the Kamakura sculptural revolution (1254 Sanjūsangen-dō, Cleveland 1919.913) through Muromachi pilgrimage practice (Chōmei-ji) and the Edo monochrome handscroll tradition (Hōitsu) to the Meiji terminus (Hōgai’s 1888 Hibo Kannon).

The Six-Kannon programme

The seven-variant typology is codified late, not early. The most common organising structure that survives in the Japanese visual record is the Roku Kannon (Six Kannon) programme — six variants placed in correspondence with the six realms of rebirth (rokudō). The seventh variant, Juntei, was sometimes added, sometimes substituted, in the Mikkyō transmission lines.

What the Six-Kannon programme records is that Japanese Buddhism reads Kannon as a system of saving variants — each calibrated to a particular realm of suffering, each iconographically fixed by an attribute-and-attribute-count grid. A standing Kannon image in a Heian or Kamakura sub-temple chapel typically belongs to such a set; the single surviving figure is the trace of a programme that has often lost five-sixths of its components to fire, dispersal, or replacement.

That programme structure matters for how the surviving objects should be read. The standing Bato Kannon at a small Yamato sub-temple is not just a wrathful Kannon variant; it is, in its original context, one of six figures arranged in a specific configuration around a central image. The disambiguation table below makes the variant grid legible; the articles in this cluster sit against the table as worked examples of how the variants survive in the record.

Forms that travel

The Kannon corpus moves across media in a way that the other clusters do not. The same iconographic variant appears as standing wooden sculpture (Kamakura), as embroidered hanging panel (Cleveland 1966.513 Amida raigō with hair-embroidery technique; Met 2013.114 Senju embroidery), as hanging bronze plaque (the kakebotoke tradition, with the Tokannon-ji 1271 Bato kakebotoke as the canonical Bato witness in this format), as gold-on-indigo sutra-frontispiece painting, as ink-monochrome handscroll (Edo Rinpa), as full-colour Meiji-period composition.

Cross-media reading is one of the cluster’s recurring exercises. The same variant — say, Senju Kannon — at Sanjūsangen-dō (1254 monumental sculpture, the 1001-figure hall by Tankei) and at the Met (Met 2013.114, embroidered surface, late Heian to Kamakura) is, on iconographic specification, the same figure. Reading the two together is reading what changes when the underlying iconographic grammar moves between media. The articles in this cluster cover this cross-media tracking; the entity primer for each canonical Kannon variant (Shō, Senju, Jūichimen, Bato, Nyoirin, Fukūkenjaku) carries the variant-level iconographic specification that the articles assume.

What the cluster does not cover

Three things sit at the edge of the Kannon cluster and are deliberately not here:

• Chinese Guanyin precedents. Avalokiteśvara’s Chinese reception (Putuo Mountain pilgrimage, the Song-dynasty Guanyin paintings, the Tang sculptural witness) is the prerequisite tradition that Japanese Kannon iconography presupposes. The bodhi corpus is Japanese-focused; the Chinese Guanyin corpus is a future expansion.
• Tibetan / Newar Avalokiteśvara. The four-armed Chenrezig of the Tibetan tradition, the Newar Karunamaya, and other Avalokiteśvara variants outside the East Asian transmission are not in scope.
• Practice / contemplative literature. Articles on Kannon devotional practice, ritual recitation of the Kannon-gyō, the mantra of Avalokiteśvara, and similar non-art-historical material are out of scope. The cluster reads the visual record.

What stays open

The Kannon cluster is comparatively well-covered for canonical Japanese Buddhist art history; the gaps are specific rather than systemic.

The largest current gap is Juntei Kannon — the seventh and most rarely-represented canonical variant. Juntei has substantial Japanese ritual-text witness in the late-Heian and Kamakura Mikkyō traditions but a relatively thin surviving sculptural record in Japan. A dedicated Juntei iconographic-reading guide is on the post-launch slate.

The second open question is the Edo-period popular Kannon corpus. The Hōitsu willow Kannon (Met 2019.419.2) and the Chōmei-ji pilgrimage mandala (Met 2016.517) cover two distinct Edo-period registers, but the broader Edo-period popular Kannon tradition — woodblock prints, household devotional altars, the pilgrim-route literature of the Saigoku Kannon thirty-three-temple circuit — has substantial scholarship that the bodhi corpus has not yet absorbed.

The third is the cross-cluster reading against Pure Land. Several Kannon works in this cluster (the Cleveland 1966.513 hair-embroidered raigō, the Cleveland 1983.75 substituted-Jizō raigō) are raigō compositions in which Kannon appears alongside Amida and Seishi as one of the welcoming descent attendants. These are catalogued under the Pure Land cluster because the raigō programme is the determining structure; their Kannon content is partial. A cross-cluster sibling guide on Kannon in the raigō tradition is a post-launch candidate.

Reading the table

The disambiguation table below carries the seven-variant grid: face count, arm count, principal attributes, canonical realm assignment in the Six-Kannon programme. Use it as a quick-reference card. The articles in the cluster sit against this grid; the entity primers carry the iconographic specification at variant level.