Yūzū Nenbutsu engi at Cleveland: reading the second scroll

Cleveland 1956.87 is the second of a two-scroll Yūzū Nenbutsu engi from the early 14th century; the first scroll, AIC 1956.1256, holds Ryōnin’s biography. The pair is one of roughly two dozen surviving recensions of an emaki the sect kept reproducing, by kanjin subscription, into the 16th century.

The work, and the missing half

Cleveland 1956.87 is twelve metres of paper. The painted surface runs 1,232 cm at a height of just under 30 cm.

It is one of two scrolls; the other lives in Chicago, Art Institute accession 1956.1256, and the matching accession years are not coincidental. Both halves were sold to the two American museums in 1956 by the Cleveland dealer Howard Hollis & Co., who appears to have separated a Japanese pair only to disperse it across the Great Lakes the same year.¹

The Cleveland catalog is direct about the relationship. The page for 1956.87 records the work as “the second scroll of a two-scroll set” and notes that “the other scroll from this two-scroll set belongs to the Art Institute of Chicago.”²

The first scroll, AIC’s, holds the biography of the founding monk Ryōnin (1073–1132).³ The second, Cleveland’s, opens with Ryōnin’s death at Ōhara and runs through the miracles attributed to the continuing nembutsu practice of his school.

This matters for how a reader treats the Cleveland scroll alone. It is not the founding story; it is what the founding story made possible. The biographical scenes (the 1117 vision of Amida, the Bishamonten apparition at Kurama-ji, the missionary tour through the provinces) are in Chicago.

To read this engi as the original makers intended, the viewer would have to sit with high-resolution scans of both, in the historically prescribed order: AIC first, Cleveland second.

Ryōnin and the 1117 vision

The doctrinal claim the engi exists to propagate is named in the title itself. Yūzū (融通) is two characters: 融, “to dissolve, to blend,” and 通, “to flow, to pass through unobstructed.” Together they name a doctrine of mutual benefit.

One practitioner’s recitation of Amida’s name flows through and benefits all other beings; every other practitioner’s recitation flows back through and benefits the first.⁴ The grammar is reciprocal in a strict sense: the practice belongs to no one in particular and to everyone at once.

Ryōnin (良忍, 1073–1132) was a Tendai monk who left Mt. Hiei in 1095 for the rural retreat at Ōhara, north of Kyoto.⁵ On the fifteenth day of the fifth month of Eikyū 5 (1117), during a session of nembutsu meditation, Amida Buddha is recorded as having appeared to him and pronounced the verse that the school takes as its founding charter:

One person is all persons; all persons are one person. One practice is all practices; all practices are one practice. Ten thousand practices are perfected through the Yūzū Nenbutsu; the Pure Land of the Ten Directions is realized at present.

The verse is preserved in the engi’s own kotobagaki (詞書, the textual passages set between painted scenes); modern scholarship treats it as the post-hoc theological justification of a practice already in motion.⁶

A subsequent vision of Bishamonten at Kurama-ji enjoined Ryōnin to propagate the teaching, and a missionary tour through the central provinces followed. Dainenbutsu-ji (大念仏寺), the school’s head temple at Hirano in Osaka, was founded in 1127 at Emperor Toba’s request.

Ryōnin died at Ōhara’s Raigō-in in 1132 at age 60. He was named Shōō Daishi (聖応大師) by imperial decree in 1773, an indication of how marginal Yūzū Nenbutsu remained in the official Pure Land settlement compared with Hōnen’s Jōdo-shū and Shinran’s Jōdo Shinshū.⁷

The doctrinal positioning is the small detail that organizes everything else. Jōdo-shū (1175) and Jōdo Shinshū (1224) are individualist in their soteriology: each practitioner’s nembutsu reaches Amida on the practitioner’s own behalf. Ryōnin’s earlier and rival framework is collectivist: nembutsu pools.

A Pure Land school whose central claim is mutual benefit needs a register of all the practitioners whose nembutsu it pools, and the myōchō (名帳, donor-and-practitioner register) tradition is unique to Yūzū Nenbutsu among the Kamakura Pure Land schools. It is the doctrinal twin of the engi-reproduction practice this article is about.

What the second scroll shows

A second scroll in an engi pair is structurally an “after.” Cleveland 1956.87 picks up at Ryōnin’s death and runs through episodes in which his teachings, transmitted by his successors, demonstrate their efficacy. The Cleveland catalog’s own description is one sentence and worth quoting in full: “This scroll begins with scenes of Ryōnin’s death and relates episodes illustrating the efficacy of following his teachings.”²

The scroll alternates kotobagaki passages, densely brushed running script in dark ink on the pale ochre paper, with painted scenes separated by gold cloud-bands. Reading right-to-left, the visible pictorial sequence opens with a death-bed and funerary scene: monks gathered around a low platform, a bier wrapped in white textiles, a procession of clerics in formal robes.

Subsequent scenes show interactions between the school’s later representatives and lay believers: pilgrim figures arriving at temple gates, ailing or dying figures attended by clerics chanting nembutsu, the school’s circulating myōchō being added to and signed by patrons.⁸

The pictorial register stays companionable throughout, in the literal sense — figures in middling colored robes at near-domestic scale, no aureoles, no descending Amida triads, no celestial-musician retinues. The salvific work in this scroll is mostly done by people in rooms.

What an unhurried look at the scroll surface yields is a narrative cadence quite different from the Heian-Kamakura raigōzu tradition. A raigōzu (来迎図) compresses Amida’s descent into a single hierarchic moment: the Buddha and twenty-five attendants drop from the upper-right toward the deathbed in the lower-left, and the painting’s whole iconographic argument is the trajectory of that descent.⁹

An engi handscroll like the Yūzū Nenbutsu does the opposite work. It spreads the salvific moment across a sequence of dozens of small, low-key encounters, each rendered at near-domestic scale, each with the figures’ faces drawn in the hikime kagibana convention (hooks for noses, slits for eyes) that places the iconographic emphasis on situation and gesture rather than on individual portraiture.

The engi’s argument is cumulative; the raigōzu’s is instantaneous. The Yūzū Nenbutsu doctrine, where practice pools and flows back, is the engi format’s natural medium.

The painter signature is unattributed, as is conventional for engi-mono of this period. The kotobagaki calligraphy in surviving early-14th-century recensions is sometimes attributed to specific brushes, but the Cleveland-AIC pair carries no such firm attribution in the published catalogs.¹⁰

The painting style places the work in the early Kamakura yamato-e narrative tradition, the same broader idiom that produced the Kasuga Gongen genki-e (1309) and the Ippen hijiri-e (1299).¹¹

The engi as devotional and recruitment apparatus

The Yūzū Nenbutsu engi is one of the most-reproduced narrative scrolls in medieval Japanese Buddhist art, and the reproduction is the iconographically interesting fact, not an incidental one. Roughly two dozen recensions survive across Japanese and Western collections, dating from the early 14th century to the late 16th.¹² No other Buddhist engi-mono was reproduced at this rate over this duration.

Karen L. Brock’s work on the engi-mono genre supplies the framework: an engi (縁起) is at root a “founding-story” narrative, originally Sanskrit nidāna, and the medieval Japanese illustrated engi-emaki functions simultaneously as devotional object, institutional history, and recruitment instrument.¹³

A school commissions an engi to be unrolled at festivals, carried on circuit by traveling clerics, displayed at fundraising events, and read aloud to lay audiences. The unrolling is the act of telling.

The form supposes a body in a room. An engi handscroll is not, like a kakemono or a mandala, a thing one stands before; it is a thing two people unroll together, shoulder to shoulder, advancing the painted surface across a low table at the pace of the reader’s voice.

A twelve-metre scroll like Cleveland 1956.87 takes perhaps thirty minutes to advance from end to end at a recitation pace; an audience around the table watches each scene briefly and the next one immediately. The kotobagaki are read aloud while the painted scenes pass, and the audience watches the founder’s life and the school’s miracles in real-time horizontal procession.

In the Yūzū Nenbutsu case, the act of patronage that funded a recension was itself a doctrinal act, since the patron’s name entered the accumulating myōchō and joined the pooled nembutsu.

The Yūzū Nenbutsu sect appears to have systematized this in a way none of its rivals matched. Takagishi Akira’s 2015 study in Japanese Journal of Religious Studies 42/1 (translated for the Nanzan Institute for Religion and Culture by Aoki Atsuko) argues that the cyclical reproduction of the engi was institutionally bound to memorial-offering programmes for successive generations of Ashikaga shoguns, and Takagishi anchors the corpus to a specific 1314 (Shōwa 3) original — the lost prototype around which every later recension is calibrated.¹⁴

Takagishi sorts the surviving emaki into three groups: the Shōwa group (early 14th c., closest to the 1314 original); the Ryōchin holograph kanjin set produced between 1381 and 1387 under the monk Ryōchin’s subscription drive; and the Meitoku print recension of 1391 and after, which Yoshimitsu patronised and which converted the engi from manuscript-only into a mass-distribution print form.¹⁵

Ryōchin, the named monk who conducted the kanjin (勧進, “subscription drive”) for both the holograph 1381–87 set and the Seiryō-ji manuscript versions, is the documented organizer of one such cycle; the pattern continues into the Bun’an era (1444–49) and beyond.

The economic logic is plain. A new recension required a fundraising drive; the drive required a circuit of public displays of an existing recension; the displays converted lay audiences into patrons whose names entered the myōchō and whose money funded the next recension. The cycle was self-sustaining and doctrinally consistent: every step was an act of yūzū, of mutual practice flowing through individuals.

This reframes how one reads the Cleveland scroll itself. It is not a singular masterpiece in the way an Unkei sculpture is singular; it is one node in a centuries-long replication network.

The early-1300s dating Cleveland gives the work places it inside Takagishi’s first group — the Shōwa cluster, the cohort that travels closest to the 1314 original — and well earlier than the Meitoku print of 1391 or the 15th-century Bun’an-era and Seiryō-ji manuscript versions held by Smithsonian (F1958.11, F1959.13), Harvard (1936.78, 1936.79), MFA Boston (11.4022), and the Japanese institutional collections at Dainenbutsu-ji, Daigo-ji, Chion-in, Eikan-dō Zenrin-ji, Tokyo National Museum, Nezu Museum, and Okura Museum of Art.¹⁶

Whether the Cleveland-AIC pair is the Shōwa-group senior witness or one of several parallel early variants in that group is the question Takagishi leaves open. He treats the Japanese institutional holdings (Dainenbutsu-ji as head temple; Seiryō-ji and Zenrin-ji as ritual-active sites) as his primary documentary base, and the Cleveland-AIC pair is not named in his catalogue — a measurable gap in the Western literature on this scroll.

The recension problem

Twenty-eight or so surviving sets, made over three centuries, all telling the same founding story with the same scene sequence: this is the recension problem in its Yūzū Nenbutsu form.

Different sets vary in scene count (most carry 11–15 painted scenes per scroll across the two-scroll structure), in the calligraphy hand, in the precise iconography of the 1117 vision (some recensions show Amida full-figure descending; others show only a cloud-band and a beam of light), and in the pictorial style.

The early-14th-century recensions sit in the early Kamakura narrative-painting idiom; the 15th-century ones absorb Muromachi suiboku influences; the late ones approach Edo workshop convention.

Cleveland’s early-1300s dating places it firmly in the first recension wave. The pictorial idiom, with its hikime kagibana faces, gold cloud-bands, and relatively restrained gold use compared with the later luxury manuscript versions, supports the museum’s date. The dimensions (29.7 cm × 1232.4 cm for the painting) are within the standard range; the AIC partner is comparable (30.5 cm × 1176.9 cm).

The two surviving Smithsonian Freer scrolls are instructive comparanda. F1958.11 is a kotobagaki-only fragment, text without illustration, suggesting either a deliberate text-only recension for ritual reading or, more likely, the surviving half of a once-complete illustrated set. F1959.13 is illustrated and dates to a later recension. Together they demonstrate the range of the genre: the engi could be reproduced as full-color illustrated set, as text-only ritual document, or anywhere between.

Where scholars disagree is not over the recension landscape, which is well documented, but over which surviving early scroll is closest to a hypothetical lost prototype.

Komatsu Shigemi (1983) implicitly treats the Daiunzan version at Daiunzan Senshū-ji as the textual base for comparison; Tashiro Shōkō (1976) gives weight to the Dainenbutsu-ji holdings as the head temple’s own institutional copies; Takagishi 2015 routes the question back through the 1314 Shōwa original and treats the 1391 Meitoku print as the watershed mass-distribution moment after which the recension landscape becomes dense.

None of these readings has settled the prototype question. Bodhi’s view is that the question may not be answerable: the engi appears to have been reproduced from the start in parallel rather than from a single Ur-text. The Cleveland-AIC pair is one early branch; whether it is the senior branch remains open.

The 1956 split, and what the scroll asks of a visitor

The Cleveland scroll’s documented provenance is short. It was sold to the museum in 1956 by Howard Hollis & Co., the Cleveland-based dealer who supplied much of the institution’s mid-20th-century Japanese material. The Marlatt, Severance, and Whittemore Funds paid for it; the accession number 1956.87 sits within the museum’s normal early-1956 sequence. Nothing about the institutional record signals an exceptional purchase.

What is documented but understated is the simultaneous Chicago purchase. AIC accession 1956.1256, the first scroll of the same pair, was acquired through the Kate S. Buckingham Endowment in the same year, almost certainly from the same dealer.

Sewell’s 1959 publication in the AIC Quarterly treats the two scrolls together; since then, the museums’ presentation has kept them separate. There is no published evidence of a coordinated curatorial reunion exhibition in the seventy years since.

For the museum visitor, the practical implication is that the Cleveland scroll cannot be read as the makers intended without also going to Chicago.

For the digital reader, the implication is more tractable: both museums have released both scrolls under CC0, both are available at high resolution through Cleveland Open Access and the AIC’s open-collection programme, and both are mirrored on the Internet Archive. A reader who unrolls both scans in sequence, AIC first and Cleveland second, sees the engi in its intended order for the first time since the 1956 separation.

The fact that this is now possible, freely, without any visit to either museum, is itself an instance of yūzū. A practice once tied to circuit-traveling priests and physical scroll-unrolling at provincial temple festivals now flows through institutional digitization programmes and open-license commitments.

The mechanism is utterly different. The function (collective access to a story whose central claim is collective access to the salvific name) is recognizably the same.

Footnotes

1. Cleveland’s published provenance lists Howard Hollis & Co., Cleveland, as the immediate prior owner; the AIC’s accession number 1956.1256 dates the Chicago acquisition to the same year. John F. (“Jack”) Sewell published the joint pair in The Art Institute of Chicago Quarterly (April 1959), the standard English-language reference for the Hollis-period split. ↩

2. Cleveland Museum of Art, accession 1956.87, “The Illustrated Miraculous Origins of the Yūzū Nenbutsu School (融通念仏縁起絵巻),” catalog page accessed 2026-04-25; CC0 (Cleveland Open Access). ↩ ↩²

3. Art Institute of Chicago, accession 1956.1256, “Legends of the Yūzū Nembutsu Sect,” catalog page accessed 2026-04-25. ↩

4. For the standard etymological reading see the Yūzū Nembutsu entry on Wikipedia (Q578719), which gives 融 as “dissolve / blend” and 通 as “flow through unobstructed”; the reading is conventional in modern Japanese reference works on the school. ↩

5. Standard biographical sketch from the Yūzū nenbutsu engi’s own opening kotobagaki and from the modern Wikipedia summary (Q11614413). Ryōnin’s documented life dates are 1073–1132. ↩

6. The verse appears, with minor variants, in the kotobagaki of the AIC scroll (1956.1256) at the scene of the 1117 vision. English rendering after Tashiro Shōkō, Yūzū Nenbutsu engi no kenkyū (Tankōsha, 1976), with adjustments for accessibility; the original is Japanese, transliteration and translation ours. ↩

7. Tashiro 1976 is the principal modern Japanese-language treatment; Komatsu Shigemi’s edition in Zoku Nihon emaki taisei vol. 11 (Chūōkōronsha, 1983) gives the standard reproduction with commentary. ↩

8. Per the Cleveland catalog’s summary description and reading the scroll surface against the published reproductions in Komatsu (ed.), Yūzū nenbutsu engi, Zoku Nihon emaki taisei vol. 11, Chūōkōronsha 1983. A scene-by-scene comparison against the Komatsu plates is the operator’s reading; cross-checking against the AIC scroll’s published documentation would refine the count. ↩

9. For the raigōzu compression-into-instant convention, see the bodhi article on the Heian raigōzu descending procession and the Met holdings 45249 and 44602. ↩

10. Sewell 1959 (AIC Quarterly) does not assign the calligraphy to a named hand; subsequent treatments by Komatsu (1983) and Tashiro (1976) similarly leave the attribution open. This is the residual uncertainty: the painters and calligraphers of the Cleveland-AIC pair remain anonymous, in the conventional sense for early-14th-century engi-mono workshop production. ↩

11. Mōri Hisashi’s Sculpture of the Kamakura Period (Heibonsha / Weatherhill, 1974) treats the workshop-production model in passing in its discussion of the contemporary Kei-school sculptural workshops; the analogous reading for narrative painting is conventional in Japanese-language engi scholarship from the 1970s onward. ↩

12. The figure of “roughly two dozen” survivors comes from the Wikidata holding list (Q24899892, accessed 2026-04-25), cross-checked against the institutional records cited individually below. Takagishi 2015 refers to ten-plus extant recensions in his core catalogue and discusses additional fragmentary and 19th-century copies; the precise count depends on whether partial sets are folded in. ↩

13. Karen L. Brock’s published work on engi-mono includes essays on the Kegon engi emaki and her broader treatment of the genre as devotional/recruitment apparatus. Brock’s reading is the standard methodological frame for English-language engi scholarship; the application to the Yūzū Nenbutsu corpus specifically is implicit in the Takagishi/Nanzan synthesis. ↩

14. Takagishi Akira, “The Reproduction of Engi and Memorial Offerings: Multiple Generations of the Ashikaga Shoguns and the Yūzū nenbutsu engi emaki,” Japanese Journal of Religious Studies 42/1 (2015): 157–182. Translated from the Japanese by Aoki Atsuko for the Nanzan Institute for Religion and Culture (PDF accessed 2026-05-13). The 1314 (Shōwa 3) original-creation date is Takagishi’s anchor; his three-group recension typology and the Ashikaga memorial-offering function are the central documented argument. ↩

15. Takagishi 2015, pp. 161–169 on the Shōwa group, the Ryōchin holograph kanjin sequence (1381–87), and the 1391 Meitoku print recension; pp. 169–177 on the Ashikaga memorial-offering programme tied to subsequent reproduction cycles. Takagishi’s grouping is the working consensus in current Japanese scholarship; the Cleveland-AIC pair is not directly catalogued in his article, which focuses on Japanese institutional holdings. ↩

16. Institutional holdings per the Wikidata entry for the engi (Q24899892) and the individual museum catalog pages cited above. The Smithsonian Freer holds two related Yūzū Nenbutsu engi scrolls (F1958.11, the kotobagaki-only fragment; F1959.13, an illustrated later recension); the MFA Boston holds a Seiryō-ji-version copy (11.4022); Harvard holds a two-volume late recension (1936.78, 1936.79). ↩