Skip to content
Topics

Search History Omakase

Start typing to search.

    ← Back

    HO Guides

    Inamuragasaki: The Battle After the Breakthrough

    Why reaching Kamakura’s shore did not mean that the city had been won.

    Opening

    On the eighteenth day, Ōtsuka Kazunari reported breaking through a barricade at Inamuragasaki. On the twenty-first, his brother died in fighting beside a shrine gate near the shore. On the twenty-second, another member of his following was killed in a valley inside Kamakura.

    These dates, in the fifth lunar month of 1333, belong to the account of service he submitted after the fighting. He wanted official recognition and reward. The surviving copy connects three places through the actions and losses of one group of attackers.[1]

    It is a different way into the fall of Kamakura from the famous story of Nitta Yoshisada throwing a sword into the sea and finding a path where the water had been. Kazunari’s account begins with a breach and continues for days.

    Inamuragasaki was an entrance to a battlefield, not a switch that turned the whole city from defended to conquered. Understanding the attack means following what could still happen after men reached the other side of the headland.

    1. Several Fronts, One City

    Nitta Yoshisada’s forces attacked the seat of the Hōjō-led warrior government in 1333. Kamakura’s geography made them confront several approaches: the hill passages to the north and northwest, the western corridor at Gokurakuji, and the coastal margin around Inamuragasaki. The medieval war tale Taiheiki organizes the assault around three principal fronts, at Gokurakuji, Kewaizaka, and Kobukurozaka.[2]

    That arrangement is useful even without accepting the tale’s enormous troop totals. An army outside one approach did not occupy all the others. Defenders positioned at a narrow passage could obstruct movement into the city, while fighting elsewhere might continue beyond their immediate view.

    The western sector contained a particularly important relationship. Gokurakuji’s landward approach and the shore around Inamuragasaki led toward the same urban side of the hills. A force finding a coastal passage could threaten men still facing an attack at the pass. Conversely, reaching the beach left the attackers among positions from which defenders could counterattack.

    The Taiheiki describes this problem in physical terms. On the landward side were steep terrain, barriers, and shields. Along the shore were obstacles made with felled branches, while boats offshore threatened the attackers with arrows from the side. These are features of its battle narrative, not an excavated inventory of the defenses.[3]

    The combination explains more than a simple statement that Kamakura was naturally impregnable. A narrow route is an advantage only while someone can defend it. A force that has passed one obstruction may still be exposed to another. The relationship between positions matters as much as the difficulty of any single path.

    For that reason, the historical place is larger than the familiar headland. It includes the western approach behind it, the other entrances occupied by opposing forces, and the internal valleys to which combat eventually moved. These are connected parts of the attack, not stages of a modern walking itinerary.

    2. The First Entry Did Not Hold

    Kazunari’s service account places him under the command of Ōdate Muneuji on the eighteenth day. He says that they broke a kuruma-sakamogi barrier at Inamuragasaki and fought there. The expression identifies an artificial obstruction; it does not describe an empty beach awaiting its first human crossing.[4]

    Another medieval narrative, the Baishōron, places troops from Yoshisada’s army beyond Inamuragasaki that afternoon, with houses on the shore burning. Its early entry into Kamakura therefore precedes the culminating night crossing familiar from the Taiheiki.[5]

    The accounts cannot be joined into a single timetable merely by assuming that every difference describes a separate, perfectly recorded maneuver. They do, however, preserve an important common difficulty: the coastal sector had to be fought over. Entry and control were different achievements.

    Muneuji did not survive the struggle. The Taiheiki describes a Hōjō counterattack after the western defenses were penetrated, followed by fighting in which he was killed. Near the western approach, Jūichininzuka—the Mound of Eleven—preserves a local tradition of Muneuji and the final men of his force dying after defeat.[6]

    The mound marks remembrance, not a surveyed diagram of that counterattack. Its tradition of the last eleven and the war tale’s account of Muneuji’s death should not be combined into an archaeological identification of eleven bodies. What they bring back into the landscape is a defeat suffered by men on the side that eventually won.

    That reversal is essential to understanding the headland. Getting past it could put an attacking force inside the contested area without giving it secure possession of the routes behind or above it. The shore could be reached, lost, and approached again.

    Kazunari’s next dated claim is for the twenty-first day. He was then serving under another Ōdate commander, identified in the city history as Muneuji’s son. The change is small beside the tale of an entire government’s destruction. Within the history of one attacking contingent, it makes continued fighting after a commander’s death visible.[4]

    3. What the Sword Accomplishes in the Story

    The Taiheiki brings Yoshisada to the western sector on the night of the twenty-first. He confronts the headland and the water, prays, and offers his gold-mounted sword to the sea. The tide withdraws, a broad passage appears, and the boats that threatened his men are carried farther offshore.[7]

    The scene removes two obstacles at once. It opens ground for the army and displaces the force that could shoot into its flank. The ensuing attack comes behind defenders who are already facing another enemy. Even this miraculous narrative contains a recognizable military relationship: a defended front becomes vulnerable when an opponent reaches its rear.

    The prayer supplies another explanation, one that should not be edited out to make the story sound modern. Yoshisada invokes the sun deity Amaterasu, associated in the passage with Dainichi, the cosmic Buddha, and appeals to a dragon deity of the sea. His appeal connects the imperial cause he serves with the opening of a road.[7]

    In the tale, the landscape responds to a claim about rightful service. The sea’s withdrawal does more than demonstrate good timing. It gives the campaign religious meaning and makes an immense political change intelligible through one man’s visible act.

    That is not evidence that the sword caused the water to move. It is evidence for how the war was narrated. Turning the prayer into a secret signal or a deliberate trick to encourage the troops would require a different source. Plausible modern psychology cannot be substituted for the episode the text actually gives us.

    Nor is the comparison with the Baishōron simply one between a miraculous tale and a rational account. The latter also describes unexpected tidal withdrawal and attributes it to divine protection, but its passage does not include the sword offering. Kamakura’s library reference service explicitly distinguishes the two versions.[8]

    The sword is therefore a particular narrative concentration of a wider coastal problem. It gives readers an action to remember and a reason for the result. Kazunari’s barricade, the earlier penetration, and the counterattack restore the repeated military work around that remembered action.

    4. A Tide Calculation Needs a Beach

    In 1915, the astronomer Ogawa Kiyohiko tried to examine the tidal story through calculation. He used tidal constants measured at Aburatsubo on the Miura Peninsula, assuming that they could stand in for those at Inamuragasaki and that their relevant characteristics had not changed radically since 1333.[9]

    He plotted a sequence of predicted water levels around the date of the Taiheiki crossing. The resulting curve replaced a simple contrast between “sea” and “dry land” with a changing series of levels. It also brought him to a difficulty that a tide table alone could not solve: the height of the route itself.

    The lower point of a tide does not necessarily expose a usable path. Whether it does depends on where the ground lies in relation to the water. A ledge might emerge only at an unusually low level; a higher margin might be accessible more frequently. Changing that assumed threshold changes the answer even when the tidal curve stays the same.

    Ogawa accordingly considered alternative relationships between the beach and average water level. He did not possess a measured cross-section of the route in 1333. His exercise was an attempt to test the literary account under stated assumptions, not the recovery of a medieval tide observation.[10]

    The useful result for reading Inamuragasaki is a question rather than a crossing time. Which part of the coast had to become passable, for how long, and under what conditions? Even a correct prediction of water level would leave the number of men, their movement, obstacles, and opposing forces to be explained.

    The present coast supplies no effortless substitute. A recent engineering investigation used instruments and seabed information to study currents around Inamuragasaki and the movement of sand from the neighboring beach. Its observations concern the modern shore, not the army of 1333.[11] They underline the practical importance of measuring a coastal setting instead of treating its familiar outline as a complete description.

    A modern person’s ability to pass somewhere at low tide would answer a modern question. It would not identify the historical line of advance, still less prove that an armed force could use it while opponents defended the adjoining ground.

    5. The Fighting Above the Water

    The hills immediately behind the western shore belong inside the explanation. After surveying the area, the archaeologist Matsuo Nobukata discussed the remains of Buppō-ji, a temple on the slope above Sakanoshita. He proposed that a narrow route between the Inamura and Ryōzen headlands had continued below the temple toward the town.[12]

    The proposal broadens the possible problem from wading around a projecting rock to controlling a coastal corridor beneath elevated positions.

    Matsuo also drew attention to service reports describing fighting at Ryōzenji’s gate and on the ridge during the twenty-first day. His reconstruction connects the assault on those positions with the eventual advance into Kamakura. The proposed synchronization with an exact low-tide hour remains a hypothesis; it is not necessary to accept that hour to recognize the importance of fighting above the beach.[12]

    A gate and a height could affect movement along the shore without occupying the same strip of ground. That relationship explains why taking one obstacle did not settle the sector. Men could be through a barrier and still face arrows, a counterattack, or a defended position farther along.

    Kazunari’s own report returns us to a specifically dated action rather than a reconstructed panoramic battle. On the twenty-first, he says, he fought beside the Hama-torii, the shore-side shrine gate, where his younger brother Narimitsu was killed.[13]

    The report does not describe every movement between Inamuragasaki and that gate. It records another place where his contingent encountered an enemy after the earlier breach. The geographical sequence is important even where the continuous route between its points cannot be recovered.

    The western battle thus involved more than choosing sea or land. Shore, approach, gate, and overlooking slope formed connected problems. An attacker’s progress depended on what remained defended around the route, not just on whether the next few yards were dry.

    6. A Valley at the End of the Advance

    On the twenty-second, Kazunari’s account places his group at Kasaigayatsu, inside Kamakura. There, he reported, his follower Nakano Iemune was killed. The progression from coastal obstruction to shore-side gate to inner valley is the backbone of his claim.[14]

    The same final sector contains Tōshōji, the Hōjō family temple associated with the deaths of Hōjō Takatoki and many of his followers. In the Taiheiki, Takatoki withdraws there as fire and fighting spread through the town. The narrative turns from the defense of the entrances to the defense of a remaining refuge.[15]

    That change of scale matters. Holding a strong position within a city is not the same as controlling the city’s approaches or maintaining its government. Kazunari could still meet lethal resistance in the inner valley after his side had reached the shore days earlier.

    Archaeology at Tōshōji has recovered stonework, a paved way, a gate trace, building remains, and evidence of fire, as well as roof-tile fragments bearing the Hōjō three-scale crest. These findings establish a substantial religious compound with a history of destruction; they do not identify a particular surviving room in which Takatoki died.[16]

    The remains were reburied for protection. The named “Harakiri Yagura” nearby belongs to the site’s commemorative tradition, not to an authenticated death chamber. Its area is closed because of rockfall danger.[16]

    Tōshōji’s place in this account is not simply to supply a dramatic ending. It shows how the location of resistance changed. At first, barriers and passes could keep attackers away from the town. At the end, fighting continued in a refuge within it. The collapse of the Hōjō government did not require every defender to cease resisting at the instant a coastal path was crossed.

    7. Turning a Route into a Claim

    Kazunari did not submit his list of places merely to help later historians reconstruct the attack. He asked for a certificate of official inspection, both as evidence for the future and as a basis for reward. He named two warriors who, he said, had witnessed his service.[17]

    Dates and places made the claim specific. The report tied one action to Inamuragasaki, another to the shore-side gate, and another to Kasaigayatsu. Deaths within his group added losses to be recognized. Witnesses offered a means of checking his account beyond his own assertion.

    The surviving copy also includes an administrative note recording receipt of a certificate and the forwarding of the original to an official concerned with rewards. That tells us the claim entered a process. It does not tell us what land or other compensation was ultimately awarded.[17]

    This perspective gives the battlefield a social geography. The army was composed of groups whose service could be narrated separately. The leader’s victory and a participant’s claim belonged to the same campaign but answered different questions. Yoshisada’s story explained the fall of a government; Kazunari’s explained why particular service should not disappear inside that larger result.

    His account is consequently interested evidence, not a neutral survey. Yet its purpose made distinctions worth recording that a single heroic scene could absorb: different days, different commanders, different positions, and different dead men. Those details make the attack route historically useful without turning it into a continuous line that can be retraced today.

    8. What the Headland Keeps in View

    Inamuragasaki received national historic-site designation in 1934 under a name that explicitly identifies it as the place of the tradition of Yoshisada’s crossing. The protected landscape preserves the association; the designation is not a finding that the sword episode happened literally.[18]

    Read beside Jūichininzuka, the western approaches, and Tōshōji, the headland carries a history larger than its most memorable scene. The mound recalls attackers who died before victory. The passes locate other fronts. The inner valley marks resistance after penetration. Kazunari’s document joins those changing circumstances to a claim that someone could inspect and remember.

    The sea remains essential. It constrained movement, exposed men to danger, and supplied the war tale with its great transformation. But a path through water was useful only in relation to the defended ground beyond it.

    The fall of Kamakura becomes clearer when entry is separated from possession. At Inamuragasaki, an army’s way in was the beginning of further problems: keeping moving, overcoming positions, surviving counterattacks, and carrying the fight into the town. Kazunari’s three dates restore that distance between a way in and a victory.

    Next topic · HO Guides

    Tokiwa: A Thousand Poems in a Warrior’s House

    How a gathering at Hōjō Masamura’s residence brought hospitality, literary skill, and the problem of precedence under one roof.

    Continue reading
    Sources and NotesView sourcesHide sources

    1. [1]

      Ōtsuka Kazunari, service claim, 元弘三年六月日 (1333, sixth lunar month, day unspecified), surviving copy in the Ōtsuka documents, transcribed in 図説北茨城市史, “大塚郷と大塚氏,” pp. 59–60: https://adeac.jp/kitaibaraki-lib/texthtml/d00010/d00010-mp000010/ht000410. See the complete displayed historical text and its administrative continuation. The essay uses the documentary passage, not the city history’s later paraphrase where it accidentally repeats the twenty-first for an event the document dates to the twenty-second. The physical original and photograph were not examined. All medieval month-and-day references retain the contemporary lunisolar calendar.

      Return to the reference ↑
    2. [2]

      Taiheiki (太平記), book 10, “鎌倉合戦事,” original Japanese transcription: https://ja.wikisource.org/wiki/太平記/巻第十. See the relevant assault sections. This is a medieval war narrative consulted through a public-domain-text transcription, not a collation of manuscripts or a field report. The named fronts are used; its troop totals, speeches, and judgments about commanders are not adopted as factual measurements or independent assessments.

      Return to the reference ↑
    3. [3]

      Taiheiki, book 10, “稲村崎成干潟事,” opening description of the western approach, barriers, felled branches, and offshore vessels, same URL as note 2. The article explains the relationship between these positions within the narrative. No surviving tree barrier or excavated fleet has been identified by this research.

      Return to the reference ↑
    4. [4]

      Kazunari’s claim in note 1, clauses beginning “今年元弘三五月十八日” and “同廿一日.” The first names 新田大館殿; the second names 新田大館孫二郎殿. The city history identifies these as Muneuji and his son Yukiuji. The precise construction of the 車逆茂木 barrier is not reconstructed from the compound term. Its existence here is the participant’s reported claim.

      Return to the reference ↑
    5. [5]

      Baishōron (梅松論), passage beginning “五月十八日の未刻計に,” in the older edition identified as 校註日本文学叢書10, Baishōron pp. 30–31; original-language excerpt reproduced at https://trushnote.exblog.jp/24983094/. Only the quoted medieval passage, not the blog’s modern television comparison or historical verdict, is used. The full edition was not accessible. Matsuo’s discussion cited in note 12 also addresses this early entry and the subsequent withdrawal, but the two reproductions are not counted as independent medieval testimony.

      Return to the reference ↑
    6. [6]

      Taiheiki, book 10, “赤橋相摸守自害事付本間自害事,” for the counterattack and Muneuji’s death; Kamakura City Tourist Association, “十一人塚,” historical description: https://www.trip-kamakura.com/facility/detail.php?id=192. The latter expressly presents the last eleven deaths and burial as tradition. Accounts differ in chronology and manner of death; the prose uses the early reversal without combining them into one exact scene or asserting identified remains beneath the mound.

      Return to the reference ↑
    7. [7]

      Taiheiki, book 10, “稲村崎成干潟事,” complete sword-offering passage and the ensuing advance. The invocation connects 天照太神, 大日如来, the sea’s divine power, and the imperial cause. The essay paraphrases this religious logic without presenting the prayer as a verified speech or inventing a morale stratagem. The displacement of boats and attack upon the defenders’ rear are features of the tale’s account.

      Return to the reference ↑
    8. [8]

      Baishōron, passage beginning “爰に不思議なりしは,” old-edition p. 31, reproduced in note 5; Kamakura Central Library, reference response on Yoshisada’s crossing, dated November 12, 2006, later entered into the National Diet Library’s Collaborative Reference Database: https://crd.ndl.go.jp/reference/entry/index.php?id=1000150410&page=ref_view. Read the response for its comparison: tidal withdrawal in both narratives, sword offering in the Taiheiki only. Its bibliography does not constitute this writer’s reading of the listed books.

      Return to the reference ↑
    9. [9]

      Ogawa Kiyohiko (小川清彦), “太平記「稲村ケ崎長干のこと」の話,” 天文月報 8, no. 1 (April 1915), printed pp. 5–8: https://www.asj.or.jp/geppou/archive_open/1915/pdf/191504.pdf. See all article pages, including the curve on printed p. 7. Printed p. 6 states the use of Aburatsubo constants and the assumptions about their applicability across place and time. This historical calculation is examined as an argument with premises, not adopted as a newly validated reconstruction of 1333.

      Return to the reference ↑
    10. [10]

      Ogawa, printed pp. 7–8, discussion of the shore’s elevation relative to mean water level and alternative passability thresholds. No modern numerical recalculation, lunar-to-Gregorian conversion, or precise crossing time is supplied. Ogawa’s closing conjectures about Yoshisada’s intentions and national military character are not adopted. The paper predates the 1923 earthquake and cannot be treated as having corrected for it.

      Return to the reference ↑
    11. [11]

      Uda Takaaki (宇多高明), “稲村ケ崎沖での海浜流の実測と数値モデルによる検証,” technical material published by Kanagawa Prefecture: https://www.pref.kanagawa.jp/documents/120978/03shitirid6shiryou.pdf. See the introductory explanation, instrument-location/bathymetric map, observation-period slides, and conclusions; See relevant maps and figures. The monitoring period shown runs from November 2024 into January 2025. The essay uses only the modern investigation’s scope, not a historical shoreline reconstruction or a precise storm date; a year inconsistency in one concluding slide is not silently reconciled.

      Return to the reference ↑
    12. [12]

      Matsuo Nobukata (松尾宣方), contributions to “[座談会]中世鎌倉の発掘—仏法寺跡と由比ヶ浜南遺跡をめぐって,” Yūrin 430, September 10, 2003, sections “元弘の乱の帰趨を決した稲村ヶ崎一帯” and “新田義貞軍が仏法寺に籠もっていた鎌倉方を打ち破る”: https://www.yurindo.co.jp/yurin/article/430/2. See the relevant survey and battle-reconstruction discussion. Matsuo reports the Ryōzenji fighting from military-service documents and proposes a route near Buppō-ji. Those particular documents were not independently transcribed here; his reconstruction and exact tidal synchronization are attributed hypotheses, not an additional eyewitness account. The essay does not derive measurements from the historical precinct illustration.

      Return to the reference ↑
    13. [13]

      Kazunari’s claim in note 1, “浜鳥居脇” and the death of his younger brother 三郎成光. “Shore-side shrine gate” explains the place-name without identifying a surviving present gate as the original object. No continuous itinerary between this point and Inamuragasaki is reconstructed.

      Return to the reference ↑
    14. [14]

      Same document, “同廿二日、葛西谷に押寄せ合戦致す” and the death of 中野新三郎家宗. This is a claim of service and a death in the valley, not proof that the follower died inside Tōshōji’s main hall.

      Return to the reference ↑
    15. [15]

      Taiheiki, book 10, “鎌倉兵火事付長崎父子武勇事,” for the movement into Kasaigayatsu/Tōshōji as fighting and fire spread; Kamakura City, historical section of the entry in note 16. Exact traditional death totals and dramatic final speeches are omitted.

      Return to the reference ↑
    16. [16]

      Kamakura City, “史跡東勝寺跡,” sections 発掘調査結果 and 「腹切りやぐら」について, updated September 25, 2025: https://www.city.kamakura.kanagawa.jp/treasury/toshojiato.html. The city summarizes investigations in 1975 and 1996–97, reburial, and the access restriction. The latter was checked September 16, 2026. The summary does not establish that every burned deposit belongs to a single day in 1333.

      Return to the reference ↑
    17. [17]

      Kazunari’s claim and its continuation in note 1: request for 御一見状, two named witnesses, “一見状給り了ぬ,” and forwarding of the original in connection with the reward administration. These passages support the documentary sequence, not proof of an eventual award or every reported exploit. The geographical and institutional interpretation is the essay’s inference from the primary text.

      Return to the reference ↑
    18. [18]

      Agency for Cultural Affairs, “稲村ヶ崎(新田義貞徒渉伝説地),” Cultural Heritage Online, title, designation date, and explanation: https://online.bunka.go.jp/heritages/detail/162461. The entry gives March 13, 1934, and explicitly identifies the crossing tradition. The essay does not treat legal protection of a place of tradition as verification of a miracle.

      Return to the reference ↑

    Table of Contents

    1. 00Opening
    2. 011. Several Fronts, One City
    3. 022. The First Entry Did Not Hold
    4. 033. What the Sword Accomplishes in the Story
    5. 044. A Tide Calculation Needs a Beach
    6. 055. The Fighting Above the Water
    7. 066. A Valley at the End of the Advance
    8. 077. Turning a Route into a Claim
    9. 088. What the Headland Keeps in View
    Sources and notes

    Reading preferences

    Text size
    Glossary intensity

    Choose how much explanation appears as you read.

    Saved on this device. All terms remain available in the glossary.

    Browse the glossary →
    History Omakase

    Opening reading…