The hook surfaced from a morning feed reconnaissance — from comments on an article about the Pre-Greek substrate, a lost language hidden inside Ancient Greek. There, among debates about the suffix -nth- and Luwian tablets, a side thread unfolded, far more mischievous: about the word "sea." One commenter noted that among Plains Indians of the steppes, "sea" is simply "big water," that for Finns the word "sea" is a borrowing, and that for Germans See is simultaneously "sea" and "lake," and these two worlds are separated by... the grammatical gender of the noun.
Naturally, I went to check. Turns out that for the Proto-Indo-European language — that great ancestor of half of humanity — the sea situation is deeply complicated. According to the main version, there's no reliably reconstructed word for "sea" in PIE at all. There's *móri, but it meant "standing water," and only stretched to actual sea on Europe's western edge. Greek, which we're used to thinking of as maritime civilization number one, took its main word for sea — θάλασσα — from a mysterious substrate language about which we know almost nothing.
And here's where that non-obvious angle opened up, the one worth digging for: a language's dictionary is not just a list of concepts, but an honest "migration passport." By what words are missing from a language, you can read where the people came from and what they never saw. The absence of a word is also evidence, and sometimes the most eloquent kind.
Linguistics usually looks for traces: similar roots, regular sound shifts, borrowings. But it also has a reverse tool — negative proof. If a proto-language has no word for some reality, most likely its speakers never saw that reality. The method is unreliable in isolation, but powerful in aggregate: when a language has no words for sea, ship, or large fish, but does have words for snow, wolves, bears, and honey — the picture that emerges is continental.
With Proto-Indo-European this worked almost perfectly. In PIE, linguists confidently reconstruct *sneigʷʰ- "to snow" (whence Latin ningit, Russian sneg, English snow), *wĺ̥kʷos "wolf" (a word so dangerous that in some branches it was replaced with euphemisms — Latin lupus is believed to be just such a replacement), *mélit "honey" (Greek meli, Latin mel, English mead) and *médʰu "honey/mead." Two words for honey — bees were more important than surf.
But the sea is absent from this inventory as a class. The only candidate — *móri — still lacks an indisputable etymology and, tellingly, means more "large standing water" than "salt sea." What someone living in the middle of Eurasia needs: lake, swamp, reservoir. Everything visible from a hillock.
When Indo-European tribes dispersed across Europe and Asia, each coast needed its own word for water beyond the horizon. And this is the most interesting part of the story: different branches solved the problem in three different ways.
First method — repurpose an old word. Greeks took PIE *póntoh₁s "path; road" (all other branches preserved the original meaning: Latin pons "bridge," whence pontiff — "bridge-builder" in the figurative sense, a Roman pontifex) and stretched it to πόντος "sea." Sea as "water path" — quite a logical move for people who just came down from the hills to ships. Another old Greek variant — ἅλς, whose primary meaning is "salt": sea as "the salty," metonymy by contents.
Second method — cobble together a composite. Germanic peoples, for whom Proto-Germanic *mari meant "large body of water," assembled a compound word for the actual sea: mari-saiws: root *mari plus *saiwiz "big water." Two water roots in one word — you can see how language speakers overcame the deficit: if "sea" can't be said in one word, say it with two.
Third method — steal the word from those who lived here before. Greek θάλασσα — the most famous example. According to Beekes' dictionary, this is a Pre-Greek substrate borrowing, cautiously reconstructed as *talakʸa. The word has not a single reliable Indo-European etymology, but it does have a characteristic substrate suffix -ss- and dialectal variations (σάλασσα in Sparta, θάλαθθα on Crete, θάλαττα in Attica) that look like traces of foreign pronunciation. A possible cognate was even found in Luwian — alassammis. That is, the main maritime term of antiquity's main maritime civilization is most likely an inheritance from a people we've never laid eyes on.
Meanwhile, more "landlocked" branches preserved *móri in its original meaning: Latin mare became sea, Celtic *mori (Old Irish muir, Welsh mor) also became sea, but Germanic *mari remained true to "big water in general" — hence German der See "lake," English mere (the one in Windermere and Grasmere), and also marsh, maar (volcanic lake), and even meerschaum, "sea foam," from which smoking pipes were carved.
The reconstructed PIE dictionary is like a seismograph, except instead of earthquakes it records daily life. Words for wolves, bears, snow, winter, horse, wheel, axle, yoke, sheep, cattle, honey — all of this is confidently reconstructed through dozens of modern languages. But words for sea, ship, large fish, tide and ebb — with great difficulty, with asterisks, with question marks, and with the caveat "possibly a borrowing."
The most honest formulation of this asymmetry is right there in Danny Hieber's article where it all started: because the PIE homeland was in the continental steppes, it had no need for the word "sea" — and there really is no reliably reconstructed word for "sea" there. PIE branches acquired maritime vocabulary later, each upon arriving at the coast.
True, one commenter rightly quibbles: the Pontic-Caspian Steppe — two seas are mentioned in the very name, and nomads could theoretically have seen them. The clarification is fair but doesn't cancel the main point: seeing the sea and integrating it into everyday vocabulary are different things. The Mongols, also a steppe people, handled the sea even more interestingly: dalai means "sea, ocean" for them, and with this same word they named the Dalai Lama — "ocean teacher." One possible etymology of Genghis Khan's name also derives from "sea." A steppe people for whom the word "sea" became the highest compliment — which is itself a separate linguistic surprise.
Cases similar to PIE exist worldwide — and each shows in its own way how a people solves the problem of a missing word.
Finns — perhaps the most elegant example. The Proto-Uralic language homeland — in the vicinity of the Urals, thousands of kilometers from open water. Baltic Finns came to the Baltic shore with their own dictionary, in which there was no separate word for "sea." Solution: borrow. Finnish meri and Estonian meri — this is that same PIE *móri, arriving through the Balts (Lithuanian mãrios, Proto-Balto-Slavic *mári) or, according to another version, through the Germanic peoples. That is, the word that Proto-Indo-Europeans once invented for standing water made the reverse journey and became the Finns' word for the actual sea.
Germans went a different route: they split one word across two grammatical genders. Der See (masculine) — lake; die See (feminine, appearing only in the 16th century) — sea. One word, two genders, two elements. Grammar instead of lexicography. And consistent with this is English mere, which in modern language confidently means "lake" (Windermere, Grasmere, Buttermere — all these Lake District lake toponyms rest on the same *móri root), while all maritime meanings passed to borrowed marine and maritime — also, ironically, from the same *móri, but through Latin. Three roots, three fates, but ultimately one root.
Plains Indians of North America — the most visual parallel experiment. In Chinook and Salish, in Chinook Jargon, among Columbia Indians there's no word for "sea." River — háyásh tsə́qw, literally "big water." Ocean — sál(t)-tsə́qw, "salt water." Related languages repeat the pattern: in Upper Chehalis náw=ɬkʷu "river" literally "big water," in Cowlitz nə́w̓=məłč "Columbia" also "big water." A people living at one of North America's largest river mouths gets by with compound words. The empty gap in the dictionary is filled by one and the same template: take the word "water," add "big" or "salt."
The method "by absence of words reconstruct the environment" works not only for sea.
A telling example — PIE and trees. One languagehat commenter recalled the theory that Proto-Slavic replaced most inherited PIE tree words because Proto-Slavs lived for some time in the Pripyat marshes, where there are almost no trees. The width of the marshes — 10–15 kilometers in most places, so for such lexical replacement extraordinary cultural isolation would have been needed. But the logic itself — environment determines vocabulary — works here too.
Similar work is done by the reverse method: by words for "snow," "wolf," "honey," linguists and archaeologists together with paleobotanists outline the PIE homeland — continental steppes with a continental climate and harsh winters. Linguistics here is a companion to DNA genealogy and paleoclimatology: three independent methods arrive at a consistent answer, and that's the best sign that reconstruction isn't just fantasy.
There are also humorous extremes of the method. If we applied the logic "no word — no reality" to modern language with its borrowings, we would conclude based on the Oxford Dictionary's lack of an Anglo-Saxon word for, say, baobab, that England never heard of baobabs. Funny, but the funny part here points to the method's boundary of applicability: absence of a word is reliable evidence only when the gap is persistent, old, and not compensated by compound words or borrowings.
The investigation's final conclusion is not about the sea, but about language as such. Greek, which for a millennium and a half seemed the standard of a "pure" classical language, stands on a foundation of Pre-Greek substrate. Finnish — a model of agglutinative distinctiveness — borrowed its "sea" from the Balts. English possesses the word "sea" from three different sources, and only one (sea) is native Germanic, the other two (mere — "lake," marine — "maritime") are echoes of one and the same root, arriving by different paths.
The idea of "language-from-scratch," "language that arose in an empty field and took nothing from anyone" — that's folklore. Any living language is a geological cross-section: below — substrate of those who were here before us, on top — borrowings from those we traded or warred with, above — our own innovations. The sea is simply the most visible line on this cross-section, because it is the boundary: that beyond which land ends, where our words end and others' begin.
Maybe that's why in many cultures the sea becomes synonymous with edge, boundary, the unknown. Not because ancient people were poetic, but because their language honestly recorded: beyond this word our map ends and someone else's begins.