THE DUTCH WOMEN LEFT BEHIND IN TAIWAN AFTER 1662

The surrender of Fort Zeelandia on 1 February 1662 ended the Dutch East India Company’s colonial government in Taiwan. It did not give every family the same ending. Those inside the fortress could evacuate under negotiated terms; people captured elsewhere had already faced imprisonment, separation and, according to surviving Dutch accounts, enslavement. The women in those records cannot be reduced to a single story of abandonment.

The distinction is central: the surrender agreement promised releases and prisoner exchanges, while published records from the siege and its aftermath describe people already killed or held captive. A promise in a treaty is evidence of an obligation, not a complete record of its fulfillment.

Illustration note: the hero is an AI-generated artistic reconstruction of a seventeenth-century harbor. Its people, clothing and scene are imagined. It does not identify an actual captive or reproduce a documented farewell.

A colony built around a harbor

Koxinga’s war reached Taiwan

Zheng Chenggong, known in European sources as Koxinga, was a Ming loyalist fighting the Qing, not simply a pirate seeking loot. Taiwan offered a defensible base and resources for his struggle. His army arrived in spring 1661, captured positions and took people prisoner. Zeelandia resisted an assault, and the confrontation became a prolonged siege. Andrade’s peer-reviewed study places the attack within that wider conflict.

Fort Provintia, at Sakam across the harbor area, fell early in the campaign. It was a different fort from Zeelandia. That distinction matters because people held outside Zeelandia did not automatically share the protection, provisions or eventual evacuation of the defenders inside it. The surrender treaty concerned the remaining fortress, not a population safely assembled in one place.

Families inside Zeelandia endured crowding and disease while the VOC attempted relief from Batavia, today Jakarta. Reinforcements did reach Taiwan, so the siege should not be described as an absolute absence of help. Yet relief failed to reverse the military situation. Andrade’s account also follows an officer who wanted to take his wife and children away but was ordered to leave them in the fort while he sailed on a dangerous expedition. Family separation was part of the siege itself.

What the surrender treaty actually promised

Governor Frederik Coyett and his council agreed an 18-article settlement dated 1 February 1662. Company fortifications, weapons and property were to pass to Koxinga. Other clauses allowed provisions for the voyage to Batavia, private possessions subject to inspection, government archives and an armed departure with military honors. Evacuation was therefore negotiated, not forbidden. Read the translated treaty, articles 2–8.

Captivity in the surviving records

A particularly revealing source is the day-journal entry of Commander Caeuw for 21 October 1661, published in William Campbell’s 1903 collection. Two boys reached Zeelandia from the occupied town. The entry describes one as enslaved to a Dutch official and the other as free-born but serving in an enslaved position in a local official’s household. Their identities and vulnerability matter: they were not neutral observers conducting an investigation. Campbell, record 116, pages 326–327.

The journal records their report that Dutch prisoners and ministers had been executed, that some people had been sent to China, and that Dutch women remained under guard at Provintia. It also records a claim that a daughter of the minister Anthonius Hambroek had been made Koxinga’s concubine. These are documented reports within a contemporary journal; they are not testimony preserved in the daughter’s own words.

The same entry supplies a major warning. The boys were subsequently tortured and accused of being sent to spy on the fortress, although the journal says they maintained their account. This does not erase the record, but it limits the certainty with which individual claims can be repeated. The responsible formulation is that the journal reports the daughter’s captivity and coerced sexual status, not that her entire life has been independently reconstructed.

Another source is a letter written by the minister Joannes Kruyf from Negapatam on 13 October 1662, after leaving Taiwan. He described the deaths of fellow ministers and reported that their wives, with others, had been carried into slavery. His letter also describes his arrival with his own family. The same document thus records both survival through departure and the captivity of others. Campbell, record 117, pages 327–328.

Kruyf was a bereaved participant, not a detached chronicler. He framed the catastrophe in religious and anti-enemy language. His account is important evidence that enslavement was reported immediately after the defeat, but it cannot supply a verified total, an individual biography for every captive, or proof of every detail transmitted through wartime rumor.

Forced unions and the missing question of consent

Captivity, slavery, marriage and concubinage are not interchangeable terms. A captive could be transferred into a household or subjected to a sexual relationship without freely choosing it. Where a record describes a marriage but preserves no account from the woman, the ceremony or label alone cannot establish consent. Nor should a modern retelling romanticize such an arrangement as a love story.

Caeuw’s journal also reports the marriage of a sixteen-year-old girl, described as having a father from Quinam and a Japanese mother, to an official at Sakam. Her consent is not recorded. The passage shows that the households caught in the conflict were not exclusively European. It supports examining constrained unions across origins, but not inventing her later fate or declaring the precise circumstances of her marriage fully known.

Indigenous women were not background figures

For Indigenous women, the relevant history did not begin with Dutch surrender. The National Museum of Taiwan History’s discussion of John Robert Shepherd’s research shows why Siraya marriage and reproductive practices require cultural context rather than a missionary’s moral judgment. Colonial accounts filtered women’s lives through European concerns about conversion and household organization. Indigenous societies had their own institutions, obligations and forms of authority.

The evidence consulted here does not permit a reliable island-wide count of Indigenous women enslaved during the 1661–1662 campaign. It would be misleading to insert them into an unverified European captivity narrative. Their exclusion from many named prisoner lists is a limitation of the record, not evidence that the war and change of government left them untouched.

After the ships left

Dutch evacuation ended a colonial government, not every captive’s ordeal. The October 1661 journal distinguishes people held in Taiwan from people reportedly transported to China. Kruyf’s later letter recalls enslavement but does not resolve each person’s subsequent life. The treaty’s timetable cannot safely be converted into a statement that all prisoners returned, any more than the captivity accounts justify claiming that every Dutch woman remained behind.

For the women whose later lives disappear from these documents, silence is not a license to supply descendants, rescue dates or reunions. Establishing an individual fate would require matching names across letters, prisoner negotiations, church registers and company correspondence. Even then, spelling changes, unnamed relatives and unequal recordkeeping can prevent a secure identification.

How to read a colonial archive

The Zeelandia diaries are administrative records, not a transparent account of everyone’s experience. Academia Sinica explains that regular entries survive from 1629 and that the originals are held in the Netherlands’ Nationaal Archief. Modern editions and translations make them accessible. Campbell’s compilation is useful, but its translated extracts, later commentary and the 1675 defense of Coyett must not be treated as a single eyewitness voice.

The Nationaal Archief’s VOC guidance reminds readers that a trading company could wage war, administer justice and operate through coercion. Historian Leonard Blussé’s work on women in VOC Batavia also challenges the assumption that colonial society’s female population can simply be described as Dutch. Reading these sources critically means asking whose names were recorded, who supplied an allegation and whose own words are missing.

A defeat remembered differently

The conquest replaced VOC rule at its Taiwanese center with Zheng authority and redirected the island’s political development. For Dutch survivors, the story could become one of lost livelihoods, bereavement and the defense of officials blamed for defeat. A military victory narrative, however, does not explain the experience of a captive woman, just as a colonial victim narrative cannot make Dutch rule innocent.

These women matter because the military ending and the human ending were different events. Some families left; some people had already died; others appear in reports of captivity and slavery, then fall beyond what the consulted documents establish. Remembering them responsibly means retaining that uneven evidence, without turning uncertainty into spectacle or allowing the surrender ceremony to stand in for every life.

For the wider maritime setting, explore PRESDA’s history of the Bruneian maritime empire and Majapahit’s trade and political networks. These provide regional context, not evidence about individual Taiwan captives.