Portuguese Words in Japanese: A Linguistic Archaeology

I

The Words That Wouldn't Leave

The Portuguese were in Japan for ninety-six years. They arrived by accident in 1543, built a trading empire on silk and silver, converted three hundred thousand people to Christianity, and were expelled under penalty of death in 1639. The Tokugawa shogunate then spent the next two centuries systematically erasing every trace of Catholic Europe from the Japanese archipelago. They demolished churches, executed priests, tortured converts, deported mixed-heritage children, and confined the only remaining Europeans, the Protestant Dutch, to a tiny artificial island where they were stripped of their Bibles, their weapons, and most of their dignity.

They did not manage to erase the words.

Today, a Japanese person wakes up, eats pan for breakfast, fastens the botan on their shirt, smokes a tabako, carries a kappa against the rain, drinks from a koppu, and has tempura for dinner, and at no point in this sequence of entirely Portuguese-derived vocabulary does it occur to them that they are speaking the language of the Southern Barbarians. The words have been in Japanese so long, have been so thoroughly absorbed into the phonological and orthographic systems of the language, that their foreign origins are effectively invisible. They feel native. They feel warm. They feel like Japanese.

This is a story about how that happened, about the mechanisms by which Portuguese vocabulary entered the Japanese language, the techniques by which it was disguised as something indigenous, and the accidents of history that allowed it to survive two and a half centuries of national isolation.

II

Three Doors

Portuguese words entered Japanese through three distinct channels, each contributing a different category of vocabulary that reflected the specific nature of the interaction.

The first door was trade. The Portuguese arrived as merchants, middlemen who exploited the Ming Dynasty's ban on Japanese commerce to position themselves as the sole conduit for Chinese silk entering Japan and Japanese silver flowing out. The goods they carried had no existing Japanese names because they had never existed in Japan. Buttons, European-style capes, drinking glasses, playing cards, flasks, tobacco, each arrived with its Portuguese label attached, and each was adopted because there was no alternative. Botan from botão. Kappa from capa. Koppu from copo. Karuta from carta. Furasuko from frasco. Tabako from tabaco. The language expanded to accommodate the objects, and the objects came named.

The second door was religion. The Jesuits who accompanied the merchants, and who, as the articles on Xavier and the Christian Century on this site describe, became deeply entangled with the commercial operation, needed a specialised vocabulary for evangelisation. They introduced bateren (priest, from padre), Deusu (God, from Deus), kirishitan (Christian, from cristão), kurusu (cross, from cruz), rozario (rosary, from rosário), and iruman (brother, from irmão). This religious vocabulary had a dramatically different fate from the commercial terms: when Christianity was banned, the words were banned with it. They survived only in the mouths of the kakure Kirishitan, the hidden Christians who preserved garbled Portuguese prayers called orasho (from the Latin-Portuguese oratio) in secret for over two hundred years, long after the speakers had forgotten what the individual words meant.

The third door was food. The Portuguese transformed the Japanese diet by introducing deep-frying techniques, refined sugar, eggs as a primary ingredient in confectionery, and a range of New World crops. This culinary exchange, detailed in the article on the Nanban culinary legacy, embedded Portuguese vocabulary into the most intimate and quotidian domain of daily life. Pan from pão. Tempura from têmpora or tempero. Kasutera from Pão de Castela. Konpeitō from confeito. Bōro from bolo. Kabocha from Camboja. These are words that people use every day, in kitchens and bakeries and supermarkets, without the slightest awareness that they are conducting a daily memorial to the Nanban period.

III

The Art of Disguise

The survival of Portuguese words in Japanese across two and a half centuries of enforced isolation was not inevitable. It was engineered, by a writing system that could make foreign words look native.

The technique was ateji: the practice of assigning kanji characters to foreign words based on their phonetic value and, often, their semantic resonance. Kanji are Chinese characters, each carrying both a sound and a meaning. By choosing kanji whose sounds approximated the Portuguese word and whose meanings evoked the object being described, Japanese writers could disguise a loanword as something that looked, on the page, entirely indigenous.

Tabako, tobacco, from the Portuguese tabaco, was written as 煙草. The first character means “smoke.” The second means “grass.” Smoke-grass. A person reading the kanji would see a perfectly logical Japanese compound describing a plant you set fire to and inhale. The Portuguese origin was invisible.

Kappa, raincoat, from the Portuguese capa, was written as 合羽. “Fitted feather.” A poetic description of a garment that fits over you like plumage and protects you from rain. The kanji told a little Japanese story about the object. The Portuguese etymology was buried.

Konpeitō, the star-shaped sugar candy from confeito, was written as 金平糖. “Gold flat sugar.” The characters are phonetically approximate and semantically evocative, but they describe a Japanese confection, not a Portuguese one.

This orthographic camouflage was devastatingly effective. Once a Portuguese word had been dressed in kanji, it ceased to look foreign. It could sit comfortably in a Japanese text alongside native vocabulary and Chinese borrowings without triggering any sense of linguistic intrusion. When the sakoku edicts sealed Japan from Portuguese contact, the words that had already been absorbed into kanji were functionally invisible to the censors. They were not perceived as Portuguese. They were perceived as Japanese words that happened to describe objects the Portuguese had introduced.

IV

The Sound of Adaptation

Japanese is a language of open syllables, nearly every syllable ends in a vowel, and consonant clusters are not permitted. Portuguese, like most Romance languages, is full of consonant clusters, closed syllables, and nasalised vowels. Fitting Portuguese words into the phonological structure of Japanese required systematic modification, and the patterns of that modification are so consistent that they serve, for linguists, as a kind of fingerprint, a reliable method for tracing words back to their Portuguese origins even when every other trace has been erased.

The most important technique was vowel epenthesis: the insertion of a vowel to break up consonant clusters that Japanese phonology cannot tolerate. The default epenthetic vowel was /u/, producing transformations like frasco → furasuko and cruz → kurusu. But there was a catch: inserting /u/ after the consonants /t/ or /d/ would trigger affrication in Japanese, changing the sound in ways that distorted the original word beyond recognition. To preserve the Portuguese /t/ and /d/ sounds, Japanese speakers inserted /o/ instead. This is why copo became koppu (with /o/ rather than /u/ after the final consonant) and why vidro became biidoro.

Nasalisation mapping solved the problem of Portuguese terminal nasal vowels, which Japanese phonology could not reproduce. The Japanese moraic nasal /N/, the only consonant permitted in syllable-final position, served as a universal replacement. Portuguese pão, with its nasalised diphthong, was truncated to pan. Botão became botan. Sabão became shabon. In each case, the nasal quality of the Portuguese ending was preserved in compressed form, mapped onto the single available Japanese consonantal coda.

Fricative substitution handled the sounds that Japanese simply did not possess. The fricative /v/, which appears throughout Portuguese vocabulary, was consistently replaced with the bilabial stop /b/. Veludo (velvet) became birōdo. Vidro (glass) became biidoro. The liquid consonants /l/ and /r/, distinct in Portuguese, merged into the Japanese alveolar lateral flap, a sound that sits somewhere between the two and represents neither precisely.

These adaptations were not random. They followed rules, phonological rules that applied consistently across the entire corpus of Portuguese borrowings. A linguist who understands the rules can work backward from a modern Japanese word, reverse the epenthesis, restore the nasalisation, substitute /v/ for /b/, and arrive at the Portuguese original with a high degree of confidence. The adaptations that disguised the words in speech also, paradoxically, preserved their origins for anyone who knows where to look.

V

The Dictionary That Froze Everything

In 1603, Jesuit scholars in Nagasaki published the Nippo Jisho, the Vocabvlario da Lingoa de Iapam, a Japanese-to-Portuguese dictionary containing over thirty-two thousand entries.

The dictionary was a missionary tool: a reference work designed to help European priests learn Japanese so they could preach, hear confessions, and navigate the complex honorific systems that governed Japanese social interaction. It was compiled by men who had spent years in the country, who spoke the language fluently, and who had an institutional incentive to record it with precision. The entries were transcribed using the Latin alphabet, capturing the phonetic reality of Late Middle Japanese at the exact moment when Portuguese loanwords were being actively absorbed into the spoken language.

The Nippo Jisho is, for linguists, a time capsule. It documents the state of the language before sakoku froze it, recording colloquial trade terms, culinary vocabulary, and daily-life words alongside the formal literary language. It proves that terms like pan, koppu, and botan were already in common use in Nagasaki by the early seventeenth century, before the words had time to be forgotten or attributed to some other source. The Jesuits, in building a tool for their own institutional purposes, inadvertently created the evidentiary foundation for four centuries of historical linguistics.

VI

What Survived and What Didn't

When the Portuguese were expelled in 1639, the fate of their vocabulary in Japanese split along a sharp line.

The religious words died, or rather, they went underground. Bateren, Deusu, kirishitan, kurusu, the entire lexicon of Christian evangelisation was suppressed alongside the faith it served. These words became dangerous to speak. They survived only among the hidden Christians of Kyushu, preserved as phonetic fossils in the orasho prayers that were transmitted orally across generations, their meanings gradually fading as the communities that spoke them lost contact with the global Church. When the hidden Christians emerged after Japan's reopening in the 1850s, their prayers contained Portuguese and Latin words that the speakers could pronounce but no longer understand, linguistic relics of a connection that had been severed two centuries earlier.

The secular words survived. Pan, botan, koppu, tabako, kappa, the vocabulary of everyday objects and actions, persisted because the objects persisted. Japan still had bread. It still had buttons. It still had tobacco and raincoats and drinking glasses. The words that named these things could not be eradicated without eradicating the things themselves, and the Tokugawa had no interest in banning bread. The ateji disguise helped: written in kanji, the words looked Japanese, and within a generation of the Portuguese expulsion, they were Japanese, as far as anyone was concerned.

The Dutch, confined to Dejima, inadvertently reinforced some of the Portuguese borrowings. Dutch and Portuguese share enough phonetic territory, kop and copo, for instance, that the continued presence of Dutch merchants using similar-sounding words for similar objects helped cement vocabulary that might otherwise have drifted or been replaced. The Dutch were a phonetic buttress for a Portuguese foundation.

VII

The Doublets

The Meiji Restoration of 1868 reopened Japan to the world, and with the reopening came a flood of new foreign vocabulary, this time overwhelmingly from English, German, and French. The proportion of Portuguese-origin words in the Japanese loanword inventory shrank dramatically. Today, Portuguese accounts for roughly two percent of Japanese foreign loanwords, against eighty to ninety percent for English.

But the Portuguese words did not disappear. What happened instead was the creation of linguistic doublets: pairs of words, one Portuguese-origin and one English-origin, that describe similar but subtly distinct objects or concepts. The older Portuguese word occupies the traditional, familiar, domestic register. The newer English word occupies the modern, Western, cosmopolitan register.

Koppu (from Portuguese copo) means a glass tumbler without a handle, the kind of vessel you drink water from. Kappu (from English cup) means a mug with a handle, the kind you drink coffee from. The distinction is not arbitrary. It reflects the chronological layering of Japan's foreign contacts: the Portuguese introduced the handleless drinking vessel; the English, arriving centuries later, introduced the handled one.

Kappa (from Portuguese capa) now refers to traditional rain gear, the kind of waterproof garment associated with fishermen, cyclists, and children. Reinkōto (from English raincoat) refers to the modern Western-style garment. The Portuguese word has become the vernacular, the homely, the everyday. The English word carries the connotation of sophistication and modernity.

Biidoro (from Portuguese vidro) survives only in specialised contexts, traditional Nagasaki glass toys, historical references, the occasional literary flourish. Gurasu (from English glass) is the standard modern term. The Portuguese word has been pushed to the margins of the lexicon, preserved as an archaism that carries the faint aroma of Nagasaki's Nanban past.

These doublets are a linguistic stratigraphy, a record, preserved in vocabulary, of the successive waves of foreign contact that shaped modern Japan. The Portuguese layer sits at the bottom, warm and worn and barely visible. The English layer sits on top, bright and dominant and unmistakably foreign. Between them, two centuries of silence.

VIII

The Myth and the New Wave

A note on a persistent falsehood. The Japanese word arigatō, “thank you”, sounds remarkably like the Portuguese obrigado, which also means “thank you.” This coincidence has generated an enduring folk etymology claiming that the Japanese word is a Portuguese borrowing. It is not. Arigatō derives from the native Japanese adjective arigatai, a compound of ari (“to exist”) and katashi (“difficult”), meaning “rare” or “precious”, a word documented in the eighth-century Man'yōshū poetry anthology, seven hundred years before the Portuguese reached Japan. The resemblance is pure coincidence.

The same applies to the Japanese sentence-final particle ne, which functions like the Portuguese tag né (a contraction of não é, “isn't it?”). Both are used at the end of sentences to seek agreement. Both sound alike. Both are entirely unrelated. The Japanese ne has been in documented use since the eighth century. Typological convergence is not etymology.

Meanwhile, the Portuguese linguistic presence in Japan is experiencing a modest twenty-first-century revival. Since the late 1980s, a wave of dekasegi, Japanese-Brazilians migrating to Japan for factory and service-sector work, has introduced modern Brazilian Portuguese terms into Japanese communities with significant Brazilian populations. Shurasuko (from churrasco, barbecue) and football terms like boranchi (from volante, defensive midfielder) have entered casual usage in these communities, creating a new layer of Portuguese vocabulary that sits atop the Nanban-era borrowings like a fresh geological deposit on ancient bedrock.