How Did Japan Keep Old Music Playable?

On 8 September 1875, the shrine priest Ōi Sugamaro took a group of learners to the musician Tōgi Otowa in Shizuoka. Their admission payments were wrapped together. Shrine funds covered some of the money; Ōi advanced another participant's payment. When he came away, he borrowed a score for the accompaniment.^[1]

These details survive in his diary. They locate musical transmission among introductions, money, and things that had to be carried home. A score was available because someone knew whom to ask. A group of people could learn because someone had arranged more than the distribution of books.

The music and dance concerned belonged to the wider ceremonial world of gagaku. Its courtly traditions offer an unusually well-documented way to investigate a question larger than any one repertory: how can music remain performable after the people who know it have died?

The answer emerges through changing arrangements, from medieval court households and manuscript collections to the coordinated ensembles of the early modern period and the teachers working between Meiji institutions. This is not the history of every kind of Japanese music. It follows a particular inheritance through the work needed to keep people capable of using it.

By the middle of the tenth century, the imperial palace in Kyoto contained a permanent Court Music Office. Related musical establishments served Nara and the temple Shitennō-ji in Osaka. Over time, musicians organized in hereditary households, known as gakke, specializing in aspects of performance. Music and dance also accompanied religious ceremonies beyond the palace.^[2]

These settings gave musical knowledge a recurring use. A ceremony required people to prepare and perform together. The next ceremony created a reason to retain what they had learned. For a household supplying musicians, instruction could connect the work of one generation with the position expected of the next.

Hereditary organization offered continuity without making competence biological. A child born into a musical household still had to learn. What the household could pass on was access to teachers, instruments, repertory, and occasions on which that repertory mattered. The knowledge survived only through work performed within those advantages.

From the late twelfth century, warrior patrons supported gagaku at places including Itsukushima Shrine and Tsurugaoka Hachimangū. Local traditions developed around shrines and temples, sometimes taking forms different from those of the major centers.^[2] The music's geographical spread enlarged the range of communities with reasons to maintain it.

That could make an inheritance less dependent on one location. It also made variation possible. A piece sustained for a particular festival acquired a relationship with that festival's people and practices. Repetition in several places need not produce identical performances, even when participants understood themselves to be continuing related traditions.

There is consequently a distinction between preserving a repertory and preserving one uniform version of it. The former can occur through dispersed use. The latter requires people in different settings to compare their practices and decide what should agree. As gagaku's institutions changed, both tasks became important—and sometimes pulled in different directions.

Music had a substantial written history within this world. In the late twelfth century, the court aristocrat Fujiwara no Moronaga compiled the Sango yōroku for the biwa, a lute, and the Jinchi yōroku for the koto, a zither. Waseda University's Theatre Museum describes copies of the latter and an abridgment of the former, the Sango yōryaku. Besides musical notation, the collections preserve annotations about rhythm, kinds of pieces, and dance procedures.^[3]

Those annotations make the books more than lists of successive sounds. They retain information about organizing a performance. A player consulting a part might also need to know what sort of occasion or movement surrounded it. Written explanation could preserve a connection that the notes alone would leave unclear.

The distinction between a compilation and its surviving copies matters too. A medieval work could reach later users through copying and abridgment. Its survival did not require the first physical manuscript to remain beside every future player. It required people to regard the material as worth reproducing and to keep enough of its organization intelligible.

How much a performer relied on the book could nevertheless change. Ethnomusicologist Kwok-wai Ng's research on the instrumental tradition known as tōgaku challenges a simple progression from oral transmission to increasing dependence on writing. His account identifies an important earlier role for notation alongside the greater prominence of oral learning in later practice.^[4]

The question is therefore not whether the music had writing. It is what the writing was expected to do for a person who already possessed a certain preparation. A concise part can be highly useful to someone who knows the conventions and frustrating to someone who does not. Adding more marks may clarify one difficulty while leaving bodily technique entirely unresolved.

This explains why a surviving score cannot be treated as a recording waiting for someone to press play. Yet calling it incomplete can be equally misleading. A part designed to work with instruction may accomplish its purpose precisely because it does not attempt to replace the teacher. The historical task is to recover that division of work: what the page retained, what the lesson supplied, and what musicians established by playing together.

Warfare exposed the vulnerability of the people and institutions behind the books. The conflicts beginning with the Ōnin War of 1467–1477 disrupted Kyoto's courtly life. Musicians dispersed. Families in Nara and at Shitennō-ji preserved performance practices as well as written materials. In the later sixteenth century, musicians from those centers helped reestablish court performances in Kyoto.^[5]

Their collaboration developed into the sanbō-gakusho, the musical establishments of three centers. A performance at Toyotomi Hideyoshi's Jurakudai residence in 1588 brought musicians together for Emperor Go-Yōzei. Under the Tokugawa shogunate, a further group served at Momijiyama within Edo Castle.^[5]

The recovery depended on what had survived elsewhere. A centralized institution could draw upon dispersed knowledge because other communities had continued to use and preserve it. The record does not justify declaring that every older piece returned. It shows how the continued activity of more than one center supplied resources for renewed collaboration.

An instrument makes another part of that continuity visible. A four-stringed biwa called Seizan, transmitted through the Kii Tokugawa collection, carries records of repairs in 1629, 1745, and 1815. The object's more ambitious associations with Moronaga and a celebrated instrument brought from China remain uncertain; the catalogue expressly distinguishes those traditions from the repair record.^[6]

The dates belong to interventions, not simply to possession. Someone repeatedly decided that an inherited object warranted more work. Repair entered its history alongside the stories explaining its prestige.

An old instrument and an old composition pose related but different maintenance problems. Wood, fittings, and strings require physical attention. A musical part requires people who can make sense of it and coordinate it with other parts. Possessing either resource gives an institution something to work with; maintaining both keeps open a wider range of possibilities.

It would be unsafe to infer the lute's present condition from those old repair dates. Their historical value is more concrete: the object was not merely handed down untouched. Continuing to care for musical material could involve altering and repairing the material itself. Preservation was an activity carried out at particular moments by people willing to take responsibility for the next stage.

One of the most revealing relationships between written and bodily knowledge lies in shōga: singing an instrumental part with mnemonic syllables. The Japan Arts Council's account of gagaku teaching explains how melody and rhythm are learned through this vocal form. The learner's voice participates before, and alongside, playing the instrument.^[7]

This gives a teacher something different to hear from an attempted instrumental performance. Singing a phrase removes some of the immediate difficulty of producing it on a reed or flute. The learner can attend to its shape and timing, then bring that preparation back to the instrument. The method connects remembering with doing rather than treating them as separate stages completed once and for all.

The syllables could also be written. The musicologist Terauchi Naoko compares Edo-period scores for the ryūteki, a transverse flute, and the hichiriki, a reed pipe. In the later standardized practice, hichiriki syllables include the vowel E while ryūteki syllables do not. The older scores show arrangements that do not fit that division. Terauchi identifies several systems and investigates their relationships with particular musical lineages.^[8]

Here, writing preserves traces of a practice commonly described as oral. It allows a historian to ask whether two teachers may have sung an instrumental line differently, even though neither voice was recorded. The result also warns against using a later convention to classify every earlier source.

A further complication appears in a score associated with the nineteenth-century musician Tōgi Fuminari. Terauchi finds two mnemonic systems within it and relates the problem to the musician's connections with more than one tradition.^[8] A named household was not necessarily a sealed container with one musical version inside.

The individual performer could therefore be a meeting point. Learning involved receiving something recognizable from predecessors, but a musician might also encounter other versions through work with colleagues. The historical evidence is most useful when it preserves those encounters rather than being flattened into one line of unchanging transmission.

Shōga helps explain both stability and variation. It gives a phrase a repeatable vocal shape, making instruction possible without demanding that the learner first solve every instrumental problem. But teachers have to supply that shape, and communities have to agree about it. What sounds like a small difference in a syllable can record a larger history of who learned from whom.

The early Meiji court brought musicians from the older centers into a new department in Tokyo. Performing together required decisions about repertory and practice. After extended discussion, standard part-scores were produced in 1876 and 1888, known as the Meiji sentei-fu, or Selected Scores of Meiji. Material left outside the official selection was called engaku, “remote music.”^[9]

Selection answered a practical ensemble problem. A player needs to know how a part relates to those being played around it. Where inherited versions differ, a rehearsal may become the occasion for settling matters that a single household could previously take for granted. A common reference reduces the number of decisions that must be negotiated anew each time the musicians assemble.

An annotated hichiriki manuscript at Nagoya University brings revision down to the page. It preserves a copying statement dated to the third month of 1857. Another note says that revisions made in the third month of 1874 are distinguished with black dots. The catalogue identifies the surviving volume as a later copy, so the earlier statement does not securely date the physical manuscript itself.^[10]

The note tells a user how to read the layers. An older text and an identified revision could remain together, with marks directing attention to what had changed. A musician did not have to treat the presence of earlier writing as a reason to ignore subsequent decisions.

Nor does this one manuscript establish that every instrument or household revised its books in the same way. Its value is that someone thought the distinction needed an explicit key. Changes had to become usable information at the point where a person consulted the part.

The official selection also created a boundary. A piece could have a history without acquiring a place in the new standard repertory. Exclusion did not prove that nobody retained it, but it affected the common reference around which the central ensemble organized its work.^[9]

These were changes in the distribution of knowledge as well as in notation. Moving experienced musicians to Tokyo left shortages in the earlier centers, where amateur groups helped sustain performance.^[9] Central coordination could therefore coexist with a renewed need for instruction elsewhere. A well-organized ensemble at the palace did not automatically supply a teacher to every shrine that wanted music.

This is the setting of Ōi's diary. In her study of Meiji Shizuoka, Terauchi identifies the importance of teachers able to connect local practitioners with courtly expertise. Tōgi Otowa, formerly a musician of the Edo government, settled in Shizuoka while maintaining connections with musicians in Tokyo. Ōi, his pupil, became an instructor in his own right.^[11]

The diary makes that connecting work visible before it became an established network. On 3 September 1875, Ōi recorded a request to arrange lessons in Yamato-mai, a ceremonial dance, and instrumental music for shrine personnel. He went to Tōgi to discuss it. The next day's answer included a delay before the teacher could come and a request that the dance learners formally enter the teaching relationship associated with another practitioner, Tomita.^[12]

Knowledge was available, but access had to be arranged. The diary identifies particular people, their schedules, and the acknowledgment expected in beginning instruction. A general desire to introduce music into a shrine could not determine any of those details.

On 7 September, Ōi accompanied two learners to Tōgi. One was to study the dance and the shō, a mouth organ; the other, the dance alone. The group expanded the next day, when Ōi recorded the bundled payments and borrowed accompaniment score described at the beginning.^[1][12]

The separate assignments are important. The undertaking needed several kinds of participation. Learning a dance did not supply the instrument accompanying it; learning an instrument did not provide another dancer. Organizing the group meant identifying who would acquire which part of a future performance.

The money likewise came through more than one relationship. The diary distinguishes expenses met from shrine resources from an amount Ōi advanced. Its entries cannot produce a complete training budget, but they expose decisions normally concealed by the statement that a shrine adopted a musical tradition. Someone had to turn that intention into paid access for actual learners.

Women participated in this extension of ritual performance too. The study identifies daughters of shrine personnel among those learning the women's dance. Later in 1875, Ōi discussed instruction for his own daughter, Tsune. His entry for 3 March 1876 records a performance at Shizuoka Sengen Shrine: Tsune and Tōgi's daughter Kino danced, while named singers and instrumentalists supplied the accompanying parts.^[13]

The entry brings the months of preparation to an identifiable occasion. The diary now assigns people to parts that the earlier lessons had made possible, alongside experienced musicians who could support the performance. The ceremony supplied a destination for their preparation.

Terauchi follows Ōi's subsequent teaching at other shrines and among Tenrikyō believers. Her argument is that such intermediate instructors helped produce further instructors, extending the network beyond a single connection to Tokyo.^[11] Musical transmission could spread when a pupil became someone other people could approach. Continuity acquired new branches through teaching, rather than through the circulation of scores alone.

The receipt of a score could still mark a sustained personal relationship. An inscription dated 18 November 1898 records the presentation of music to Nabeshima Naohiro in recognition of his long engagement with gagaku. The surviving set includes parts for hichiriki and biwa. Its inscription associates the transfer with acknowledged study, rather than presenting it as an anonymous purchase.^[14]

Naohiro also used musical knowledge to make something new. His manuscript Shinshiki shōkafu, dated 1906, places traditional gagaku notation alongside Western staff notation. One setting joins verses by Sanjō Sanetomi and Naohiro's father, Nabeshima Naomasa, with music supplied by Naohiro. His father had died in 1871.^[15]

This was not merely an older composition being copied into a different script. A later musician was making a setting from inherited words, using more than one notational system. The same collection's account places singing and playing within his family life as well as his former court duties. Knowledge maintained for ceremony could also become a resource for composition and domestic performance.

The examples resist a choice between perfect repetition and the loss of tradition. Medieval compilations preserved structured knowledge. Dispersed musical communities retained resources that could later be brought together. Vocal teaching made parts learnable, while common scores helped ensembles coordinate. Local instructors assembled the people needed to turn those parts into an occasion of music.

Keeping old music playable meant keeping these activities connected. A score needed interpreters; a learner needed instruction; an ensemble needed an occasion and a workable agreement about what to perform. The relationships changed, and sometimes the versions changed with them. The inheritance remained usable when later people could do more with it than identify its age.