HO Guides
Ishibashiyama: When the Followers Had to Leave
After the defeat of 1180, protecting Yoritomo required dispersal, local help, and messages that could never tell everyone the whole story at once.
Quick · About 2 min
The story in five points
A short route through the argument. Follow a point into Classic for its context and sources.
Defeat came before support converged
Yoritomo’s intended allies had not all assembled when fighting began at Ishibashiyama. In the chronicle’s account, opponents exploited that separation. A shared cause could threaten an enemy without yet delivering usable help to the men on the hillside.
Read in contextLeaving could preserve a connection
Sanehira argued that followers who found Yoritomo should disperse because a larger party was harder to conceal. Local knowledge and manageable numbers mattered more than the wish to remain together. Support changed form when battlefield defence became hiding.
Read in contextAssistance crossed opposing sides
The chronicle credits Ieyoshi with helping Yoritomo after serving among his opponents, and Shigekuni with sheltering Sasaki fugitives despite fighting against them. Such assistance depended on particular relationships and risks, not automatic refuge with everyone on the same side.
Read in contextEscape required separate arrangements
Survivors left by different routes and boats. Sanehira obtained a vessel through a local resident, while a messenger carried news limited to events before embarkation. Reaching safety and making that survival known were distinct tasks performed by different people.
Read in contextRecovery included those left behind
Awa still required local protection and carefully routed recruitment. Yoritomo also ordered protection for a dead follower’s children before dispersed men reunited. Survival preserved relationships that needed further work; it did not guarantee the later government at Kamakura.
Read in context
Opening
In the Azuma kagami's account of the retreat from Ishibashiyama, Doi Sanehira confronts an unexpected difficulty. Several followers have found Minamoto no Yoritomo alive in the hills. They want to remain with him. Sanehira wants them to leave.[1]
The problem is their presence, not their lack of commitment. A leader with a group around him will be harder to conceal than one accompanied by a local man who knows where to go. The chronicle makes Sanehira argue that separation now could make a future recovery possible.
That is an uncomfortable beginning for a battlefield remembered through men who died beside their lord. Yet the hills above the Ishibashi coast contain both histories. Some followers were killed in the fighting. Others had to get away, take different routes, and discover whether their connections still worked after the army had broken apart.
1. A Battle Before the Supporters Had Assembled
Ishibashiyama lies on the constricted coastal side of the mountains south of present-day Odawara. The protected memorial sites occupy Ishibashi and neighboring Komekami. The wider landscape links wooded slopes with the shore, a setting in which movement could quickly change from marching along an approach to searching for a way uphill.[2]
Yoritomo reached this area shortly after beginning his uprising in Izu. He was challenging the Taira establishment, not yet commanding the government he would later build at Kamakura. The Azuma kagami places his departure from Izu in the eighth lunar month of 1180 and the battle on that month's twenty-third day.[3]
The chronicle describes opposing forces separated by a valley. Ōba Kagechika led men supporting the Taira, while Itō Sukechika occupied ground behind Yoritomo's position. Other intended supporters, the Miura, had not joined him. They were still away from the main battlefield, near a river.
This mattered to the sequence of action. The narrative makes Kagechika's side discuss attacking before the Miura could arrive. A prospective alliance could threaten an opponent without yet providing help to the men waiting on the hillside. Its members needed time and a usable route to become one force.
The chronicle portrays Yoritomo as heavily outnumbered and records his defeat. Among the dead were Sanada Yoichi Yoshitada and his retainer Bunzō, also named Ieyasu. Their commemoration would eventually give the battlefield two of its most enduring focal points.[4]
Our detailed narrative is not an observer's report sent from the valley. The Azuma kagami was compiled long afterward within the world of the Kamakura government. Its accounts preserve names, places, and relationships, but also speeches and episodes shaped into a story of the founder's survival.[5] Following that story closely is useful as long as we do not turn its narrative precision into proof that every exchange was recorded at the time.
2. The Group That Was Too Large to Hide
The retreat did not immediately produce a safe party moving in one direction. The entry for the following day describes repeated fighting as Yoritomo's men moved farther into the hills. Horses were struck, rear guards halted, and relatives turned back toward one another. Tokimasa and his sons became separated from Yoritomo amid the exertion of the fighting and climbing.[6]
At this point the body of a particular person mattered more than an army's position on a map. A man unable to continue uphill could delay those who remained with him. A father turning toward a son could be acting within the same attachment that made the group harder to move. Retreat produced decisions at distances small enough for people to see one another and still be unable to bring everyone along.
When several men reached Yoritomo and Sanehira, their reunion appeared to solve one problem. They had found their leader. Sanehira's objection exposed another: finding him was not the same as keeping him hidden.
The chronicle presents concealment as a task for someone with particular knowledge, rather than a simple benefit of entering a forest. Sanehira believed he could shelter Yoritomo with a much smaller party. More companions meant more people whose movements and needs would have to be managed. The text does not give us their food consumption or an exact hiding place; its argument concerns the difficulty of maintaining secrecy together.[1]
The followers' wish to stay was therefore not enough to determine what should happen next. In the account, even Yoritomo initially appeared willing to accept their company. Sanehira repeated his objection, and the men dispersed.
This was a change in the form of support. On the battlefield, gathering around a leader could contribute directly to his defense. In hiding, the same concentration could expose him. Leaving did not promise that the men would meet again. It left open a possibility that remaining together might close.
3. Help Did Not Follow the Battle Lines
One of the men who found Yoritomo was Iida Ieyoshi. The chronicle had earlier placed him among Kagechika's forces, unable to reach Yoritomo before the fighting. During the withdrawal, it credits him with breaking away to assist the defeated side. Now he arrived carrying Yoritomo's prayer beads, which had been lost during the retreat.[7]
The beads were potentially recognizable. The account says that men in Sagami had seen Yoritomo carrying them during hunting excursions. An ordinary possession could therefore reveal a connection that the fugitives needed to conceal. In a landscape where people knew one another from earlier activities, secrecy involved more than putting trees between two armies.
Sanehira still refused Ieyoshi's request to stay. Useful assistance did not automatically give the helper a place in the hiding party. The same decision had to be made again, despite the value of what he had brought.
Other scattered followers depended on different relationships. On the twenty-sixth day, the chronicle records Kagechika asking Shibuya Shigekuni to take the Sasaki brothers' wives and children into custody. Shigekuni had fought on Kagechika's side, yet he had longstanding connections with the Sasaki. He objected to the demand. When several of the brothers later reached his residence, he concealed them in a storehouse and provided food.[8]
A residence belonging to someone on the opposing side could thus supply assistance, while the fugitives' own households remained vulnerable. The account does not make all local connections available to everyone. Each refuge depended on a particular person accepting a particular risk.
Yoritomo's party likewise required a suitable room, not merely a friendly religious institution. The chronicle says that the quarters of Hakone's head cleric were crowded with pilgrims, making concealment impractical. His brother Eijitsu's lodging was used instead.[9] Support from the clerical household opened a possibility; the distribution of people within its buildings determined which part could actually serve it.
4. Not Everyone Left on the Same Boat
The coast provided another way out, but departure was not one coordinated evacuation. The chronicle places Tokimasa and a party of survivors on a boat from Iwa on the twenty-seventh day of the eighth month. At sea they encountered members of the Miura group, who were themselves withdrawing after fighting elsewhere.[10]
Yoritomo's departure comes the following day, from Manazuru. Here the account identifies a small but essential chain of assistance: Sanehira instructed a local resident named Sadatsune to prepare a small boat. Yoritomo crossed with Sanehira toward Awa, on the eastern side of the water, and was received by men who had arrived before him.[11]
The named resident is an important part of the escape. The existence of a shore does not supply a vessel. Someone must be able to obtain one, prepare it, and connect its use to the people trying to leave. Sanehira's local relationships extended the retreat beyond the places in which he could personally guide Yoritomo on foot.
Meanwhile, others remained inland. The chronicle describes the Katō brothers taking their exhausted father to Izusan before setting out in another direction. The brothers themselves became separated while fleeing pursuit. The movement away from Ishibashiyama was a collection of departures, not a single line that can be drawn from the battlefield to a waiting fleet.[10]
Messages followed routes of their own. Before sailing, Yoritomo sent Sanehira's son Tōhira to Masako with news. When the messenger reached her early in the next lunar month, he could report what had happened before embarkation. He did not know what had happened afterward.[12]
This is one of the chronicle's most revealing limits. News could be accurate and still fail to answer the recipient's most urgent question. Masako's lack of information was not resolved merely because Yoritomo had reached another shore. His whereabouts had to become known through another person's movement, after the event that person was able to witness.
5. Awa Was Another Set of Relationships
Crossing the water removed Yoritomo from the immediate search in the Ishibashi hills. It did not make every household in Awa a refuge. The chronicle reports a local opponent preparing to attack a lodging used by the party. Miura Yoshizumi, acting with knowledge of the district, learned of the threat and moved against him.[13]
The difficulty encountered on the western shore had changed location rather than disappeared. A place to stop still had to be assessed through people who knew the neighborhood. The sea could interrupt pursuit, but it could not supply political agreement on the other side.
A local supporter, Anzai Kagemasu, then advised against simply traveling to the residence of the prospective ally Kazusa Hirotsune. The account has him recommend sending messengers first and arranging for supporters to come forward. Yoritomo's party turned to Anzai's residence while emissaries went out.[14]
The difference is practical. A traveler approaching an uncertain destination exposes himself along the way and depends on what he will find there. A messenger can open negotiations without bringing the principal person into the same danger. Staying in one known household while sending others outward separates communication from the movement of the leader.
Nor could all supporters use the same approach. A message to Kasai Kiyoshige explicitly advised arrival by sea because of the position of his home between the territories of other warrior houses. Recruitment required instructions adapted to where a man lived, not merely an appeal to a shared family name or cause.[13]
The geographical problem had therefore expanded. At Ishibashiyama, separated supporters had failed to become one force in time. During the recovery, routes, messengers, and local hosts were again needed to make distant willingness become physical assistance. The later growth of Yoritomo's following depended on those connections being made; it was not something already contained in the fact that he had escaped.
6. The Children Still at Home
The battlefield's consequences also reached people who had never been part of the retreating group. Near the end of the ninth lunar month, the chronicle records a message to the mother of Sanada Yoichi, one of the men killed at Ishibashiyama. His young children remained at the family's holding. Yoritomo ordered that they be sent to his present location in Shimōsa for safety.[15]
The recorded concern was that local opponents, including Kagechika's associates, might harm them. Their father's death had not ended their exposure to the conflict. The household that sustained a warrior could remain within reach of his enemies after he was gone.
The addressee makes the continuing relationship specific. The instruction went to Yoichi's mother, not to an abstract category of bereaved dependents. Someone still connected to the family's home had to receive it and arrange what happened next. The chronicle records the order; it does not, in this entry, describe the children's subsequent journey.
This gives Yoichi's memorial at Ishibashiyama a history beyond the moment of his death. Commemoration can fix attention on the sacrifice. The message shows a further responsibility arising from it: what could be done for those still alive and still in a dangerous place?
Two days later in the chronicle's sequence, on the first day of the tenth month, many men dispersed at Ishibashiyama rejoined Yoritomo at his temporary quarters in Saginuma.[16] The account does not identify every route by which they reached him. It does explicitly bring the scattering and the reunion into the same history.
Their return was different from the survival of the leader alone. A man hidden successfully in the hills could not constitute an army by himself. The possibilities preserved through separate departures had to become people meeting again. Other relationships, including those with the dead man's household, required attention even while that regrouping continued.
7. A Route Changed by the Dead
The protected battlefield memorials are two distinct places: Yoichi-zuka in Ishibashi and Bunzō-dō in Komekami. Their pairing preserves the association between Yoichi and his retainer. It does not give us a measured boundary of the whole battle or establish that the present structures survive unchanged from 1180.[2]
The Azuma kagami already makes their graves part of Yoritomo's later experience of this coast. An entry for the twentieth day of the first lunar month of 1190, recording his return from a pilgrimage, describes his distress when seeing the graves along the approach to Izusan. His pilgrimage guide considered such mourning inappropriate on the approach to the sacred destination. The account says that the order of future visits was therefore changed.[17]
The passage is about a route as well as an emotion. The graves stood where travel could bring Yoritomo into renewed contact with the defeat. A guide's religious concern then affected how subsequent journeys should be arranged. The coast that had once constrained his escape became a landscape through which later obligations also had to be negotiated.
The chronicler's report of tears should not be mistaken for access to every private memory Yoritomo carried. Its significance is that the narrative of the established leader still gave these two deaths a capacity to interrupt an ordinary sequence of movement. His survival had not made their absence irrelevant.
Modern commemoration concentrates that relationship into named sites. It is easier to point to Yoichi's mound or Bunzō's hall than to all the temporary lodgings, meeting places, and uncertain routes used by the survivors. The memorials make the dead legible. Reconstructing the aftermath from the record restores something less easily marked: the work of keeping scattered people connected.
8. What Dispersal Preserved
Ishibashiyama was a defeat. Calling it an early step toward Yoritomo's later success can conceal how much remained unresolved when the fighting ended. A leader needed a hiding place; followers needed routes out; a boat required local arrangements; families needed news and protection.
The chronicle answers those problems through different people rather than one continuous act of command. Sanehira argued for separation. A man from the opposing force returned a recognizable possession. Another sheltered fugitives whose family he knew. A local resident prepared a vessel, and a messenger carried information that stopped at the moment of departure.
Their actions did not guarantee a later government at Kamakura. They preserved particular lives and relationships that could still be used, renewed, or acknowledged. That is the history between the defeat on the hillside and the reunion recorded weeks later—and the history that gives the battlefield's two memorials more than a story of men who died.
Next topic · HO Guides
Ganjōju-in and the Hōjō Residence: What Prayer Required
The hidden contents of Buddhist images, an unfinished vow, and a convent on former household ground reveal different kinds of work behind a warrior family's religious life.
Continue readingSources and NotesView sourcesHide sources
- [1]
Azuma kagami (吾妻鏡), book 1, 1180, eighth month, twenty-fourth day, paragraph beginning 北条殿父子三人 and Sanehira's argument from 実平云. Original kanbun in the transcription based on Kokushi taikei, vol. 5 (Keizai Zasshisha, 1903): https://ja.wikisource.org/wiki/吾妻鏡/第一. The opening and section 2 paraphrase the chronicle's reported exchange, not an independently authenticated conversation. No particular surviving tree, cave, or modern path is identified as its setting.
Return to the reference ↑ - [2]
Odawara City, “石橋山古戦場のうち与一塚及び文三堂,” heritage description and locations: https://www.city.odawara.kanagawa.jp/field/lifelong/property/cultural/historic/s-ishibashiyama.html. The protected pair is at Ishibashi 470 and Komekami 136; these are memorial locations, not coordinates for the entire battle. No archaeological identification of individual human remains was established in the consulted material. Iwasaki Sōjun, “石橋山の合戦,” in the city's 浮世絵が語る小田原 (originally published in 1992), also supplies the geographical setting: https://www.city.odawara.kanagawa.jp/encycl/ukiyoe/01/. No visitor route or access promise is made.
Return to the reference ↑ - [3]
Azuma kagami, 1180, eighth month, twentieth–twenty-third days, note 1. See the departure, Miura movement, opposing dispositions, and battle entries. Dates are lunisolar, not Gregorian August dates. The text's exact troop totals and dramatic claims of marksmanship are not treated as verified measurements.
Return to the reference ↑ - [4]
Azuma kagami, eighth month, twenty-third day, from 両陣之際隔一谷 through the withdrawal into the hills, note 1. Used for the valley, Sukechika behind the position, the absent Miura, the discussion of attacking before their arrival, and the named deaths. The entry places the Miura near 丸子河 but does not itself expressly make a swollen river the cause of their halt; that familiar explanatory addition is not supplied here. The deceased retainer appears as 豊三家康 in this transcription and 文三 in the modern memorial name.
Return to the reference ↑ - [5]
National Archives of Japan, “吾妻鏡,” in 歴史と物語: https://www.archives.go.jp/exhibition/digital/rekishitomonogatari/contents/08.html. See the description of compilation, scope, and manuscript transmission. The narrative's dependence on a later chronicle remains a substantive limitation throughout this essay, not just a bibliographical qualification.
Return to the reference ↑ - [6]
Azuma kagami, eighth month, twenty-fourth day, opening paragraphs through Tokimasa's separation, note 1. The text reports horses struck, fighting rear guards, kin turning toward one another, and fatigue preventing ascent. The explanation of small-scale movement constraints is a limited inference; no slope gradient, exact route, pace, or casualty calculation is asserted.
Return to the reference ↑ - [7]
Azuma kagami, eighth month, twenty-third day, passages on 飯田五郎家義; twenty-fourth day, paragraph beginning 其後家義奉尋御跡参上, note 1. These give Ieyoshi's initial position, assistance, return of the beads, their familiarity from hunting, and his departure after Sanehira's objection. The danger of recognition is an inference from the text's emphasis on familiarity, not a claim that an enemy actually followed the beads. The better-known Kajiwara rescue episode in the same entry is not needed for this essay's explanation.
Return to the reference ↑ - [8]
Azuma kagami, eighth month, twenty-sixth day, paragraphs beginning 又景親行向渋谷庄司重国許 and 入夜定綱, note 1. See the detention request, objection, and later storehouse hospitality in the original-language transcription. The chronicle praises Shigekuni; the prose reports his described acts without adopting a general moral verdict or inferring that all opponents behaved similarly.
Return to the reference ↑ - [9]
Azuma kagami, eighth month, twenty-fourth day, final paragraph beginning 其後以永実為仕承, note 1. Eijitsu is written 永実 and the head cleric Gyōjitsu 行実. The distinction between the crowded principal lodging and Eijitsu's quarters is explicit. No surviving building is authenticated as that room.
Return to the reference ↑ - [10]
Azuma kagami, eighth month, twenty-seventh day, both the Iwa/Miura passage and the Katō family's movements, note 1. These are different groups and routes. The narrative's precise ages, weather descriptions, and speeches about abandoning an elderly parent are not reconstructed as witnessed scenes.
Return to the reference ↑ - [11]
Azuma kagami, eighth month, twenty-eighth–twenty-ninth days, note 1: 武衛自土肥真名鶴崎乗船 and 実平仰土肥住人貞恒粧小舟, followed by arrival in Awa. Sanehira, not Yoritomo, instructs Sadatsune. The account does not supply the craft's dimensions or an exact charted sea track. Manazuru is necessary retreat context, not a substituted principal Place.
Return to the reference ↑ - [12]
Azuma kagami, eighth month, twenty-eighth day, dispatch of 土肥弥太郎遠平; ninth month, second day, his arrival and lack of knowledge after embarkation, note 1. The episode is used for the limit of the messenger's information. The chronicle's emotional descriptions are not converted into invented internal monologue, and no letter manuscript is claimed to survive here.
Return to the reference ↑ - [13]
Azuma kagami, ninth month, third day, instructions to 葛西三郎清重 and the Awa lodging episode involving 三浦次郎義澄, note 1. Used for the requested sea approach and Yoshizumi's local knowledge. The chronicle's hostile labels and assertions of secret motives are not adopted as the narrator's judgments.
Return to the reference ↑ - [14]
Azuma kagami, ninth month, fourth day, Anzai's arrival, advice, change of lodging, and dispatch of emissaries, note 1. The contrast between sending a messenger and exposing the leader on a journey is the essay's practical explanation of the reported action, not an independently documented general military doctrine.
Return to the reference ↑ - [15]
Azuma kagami, ninth month, twenty-ninth day, paragraph beginning 又被遣専使於佐那田余一義忠母之許, note 1. The text records the children at the family's holding, a concern about possible harm, and the instruction to send them to Yoritomo's current location in Shimōsa. It is not the original dispatched letter and does not verify compliance. No attack on these children is asserted.
Return to the reference ↑ - [16]
Azuma kagami, tenth month, first day, phrase 於石橋合戦之時令分散之輩 through 武衛鷺沼御旅館, note 1. It explicitly records the return of many dispersed men. No complete survivor list or reconstruction of their intervening routes is supplied.
Return to the reference ↑ - [17]
Azuma kagami, 1190, first month, twentieth day, kundoku transcription: https://www5a.biglobe.ne.jp/~micro-8/toshio/azuma/119001.html. See the entire entry, which records Yoritomo's return and explains a future reordering of shrine visits because of mourning at the graves on the approach. This separates the entry date from the earlier encounter it reports. Odawara's heritage page compresses the episode as a return-journey visit, whereas its essay “佐奈田与一と俣野五郎” connects it with the outward pilgrimage: https://www.city.odawara.kanagawa.jp/encycl/ukiyoe/02/. The prose follows the chronicle's sequence rather than selecting either modern summary's compressed wording. It does not authenticate the present memorial fabric as the graves seen in 1190.
Return to the reference ↑