Linguistic evidence suggests that proto-Ainu originated in southern Sakhalin and northeastern Hokkaido before spreading across the archipelago -- into the Kurils, across Hokkaido, and partially into northern Honshu. The Ainu were not newcomers to Sakhalin; they were among its oldest continuous inhabitants. A study by Lee and Hasegawa from Waseda University, drawing on linguistic, archaeological, and genetic evidence, linked the Ainu significantly to the Okhotsk culture that once dominated northern Hokkaido. Oral tradition records that the Ainu displaced an earlier people in central Sakhalin whom they called the Tonchi -- a group that toponymic evidence identifies as the Nivkh. The landscape of Sakhalin, in other words, has been shaped by waves of language and people for far longer than written records suggest.

The earliest written record of Sakhalin Ainu dates to 1643, when the Dutch explorer Maarten Gerritszoon Vries transcribed several sentences during his expedition through the waters around Sakhalin and the Kurils. Over a century later, in 1787, the French navigator Laperouse recorded 161 Sakhalin Ainu words -- a vocabulary snapshot of a language that had never needed writing to survive. The Polish ethnologist Bronislaw Pilsudski conducted more extensive fieldwork in the early 1900s, collecting narratives and conversations from speakers along Sakhalin's southeastern coast. In 1903, a Sakhalin Ainu man named Sentoku Taroji wrote Pilsudski a letter in Cyrillic script -- one of the few instances of the language adapted to a foreign alphabet, a small act of cultural bridge-building between worlds that would soon be torn apart.

After World War II, when Sakhalin passed from Japanese to Soviet control, the Ainu population faced a forced choice that was really no choice at all. All but roughly 100 Ainu living on Sakhalin were deported to Japan, severed from the land that had shaped their language and identity. The last Ainu household on the island died out in the 1960s. In Japan, the displaced speakers maintained Sakhalin Ainu for another generation, but the language was cut off from its landscape, its place names, its seasonal rhythms. Linguists worked to document what remained, eliciting narratives from native speakers who had been born on Sakhalin but lived out their years in Hokkaido. When Tahkonanna died in 1994, she took with her the last living connection to a language that had named Sakhalin's rivers, mountains, and weather for millennia.

Sakhalin Ainu was not simply a dialect of Hokkaido Ainu transplanted northward. It had evolved its own distinct features, including long vowels absent from the Hokkaido variety. Where Hokkaido Ainu preserved certain syllable-final consonants, Sakhalin Ainu softened and merged them through a process linguists call lenition. The language was written in multiple scripts over its documented life: in katakana by Japanese researchers, in Cyrillic by Sakhalin Ainu speakers in contact with Russians, and in Latin-based transcriptions by European ethnologists. Each writing system captured the sounds differently, like photographs taken from varying angles of the same subject -- no single image complete, but together suggesting the shape of something irretrievably lost.

Flying over central Sakhalin today, the forests and river valleys below give no outward sign that a distinct language once mapped every feature of this terrain. The Ainu place names survive on old Japanese maps and in ethnographic archives, fragments of a vocabulary built from intimate knowledge of the land. Sakhalin Ainu is classified as extinct, but what was recorded -- by Pilsudski, by Japanese linguists, by the speakers themselves -- preserves a window into a culture that understood this island on terms entirely its own. The recordings exist. The grammars have been analyzed. The language cannot be revived, but it can still be heard, if you know where to listen.