Antedating tsunami

« previous post | next post »

The OED's first citations for tsunami:

1897 L. Hearn Gleanings in Buddha-fields i. 24 ‘Tsunami!’ shrieked the people; and then all shrieks and all sounds and all power to hear sounds were annihilated by a nameless shock‥as the colossal swell smote the shore with a weight that sent a shudder through the hills.
1904 Publ. Earthquake Investigation Comm. Foreign Lang. (Japan) xix. 6 Records and reports of earthquakes and ‘tsunamis’.

On ADS-L a bit more than a year ago, there was a discussion of possible antedatings.

In particular, Victor Steinbok on 2/28/2010:

Interesting that the OED missed such a trivial publication as National Geographic for Sept. 1896 for antedating tsunami. In fact, the entire cover story is devoted to it (specifically to one of June 15, 1896, pp. 285-9). The Atlantic was not to be outdone, placing a story, A Living God, in the Dec. 1896 issue of the Monthly (pp. 833-41).

There is also a Tsunami entry in the Helpburn's Japanese-English/English-Japanese Dictionary, 2nd/4th ed., 1872/1888, p. 567/695 (see also entry on p. /166)

> TSUNAMI, [], n., A large waves that rolls over and inundates the land.

Note, however, that 1) this is a dictionary and the words are in Japanese, not adopted in English, and 2) the 2nd edition was published in Shanghai by the American Presbyterian Mission, before it was picked up by international publishers; the 4th and subsequent editions (not sure about 3rd) were published by Z.P. Maruya, Tokyo, Kelly & Walsh, Yokohama/Shanghai/Singapore, and Trubner & Co in London.

The same definition is in Ichikawa's New Pocket Dictionary, also from 1888 (Yokohama).

So the actual antedating is only by one year, but there might be more a bit earlier–just not on GB (majority of hits on {tsunami pre-1900} were false–about half were false dates and half bad OCR, especially the ones to the first half of the XIXth century and those in other European languages).

The National Geographic Magazine for 1896 is available from Google Books, and (as Victor observes) it contains an article by Eliza Ruhamah Scidmore in the September 1896 issue, "The Recent Earthquake Wave on the Coast of Japan", which starts this way:

On the evening of June 15, 1896, the northeast coast of Hondo, the main island of Japan, was struck by a great earthquake wave (tsunami), which was more destructive of life and property than any earthquake convulsion of this century in that region. The whole coastline of the San-Riku, the three provinces of Rikuzen, Rikuchu, and Rikuoku, from the Island of Kinkwazan, 38° 20' north, northward for 175 miles, was laid waste by a great wave moving from the ast adn south, that varied in recorded height from 10 to 50 feet. A few survivors who saw it advancing in the darkness, report its height as 80 to 100 feet.

There is a second instance in the article which is more use than mention, though still italicized:

There were old traditions of such earthquake waves on this coast, one of two centuries ago doing some damage, and a tsunami of forty years ago and a lesser one of 1892 flooding the streets of Kamaishi and driving people to upper floors and the roofs of their houses.

Ms. Scidmore's article includes an interesting side note:

With telegraph telegraph offices, instruments, and operators carried away, word came slowly to Tokyo, and with 50 to 100 miles of mounts roads between the nearest railway station and the seacoast aid was long in reaching the wretched survivors. When adequate idea of the calamity reached the capital and the cities, men-of-war, soldiers, sappers, surgeons, and nurses were quickly dispatched, and public sympathy found expression in contributions through the different newspapers, amounting to more than 250,000 yen, for the relief of the injured. The Japanese journalist and photographer soon fed the public full of horrors, yet the first to reach the scene of the disaster was an American missionary, the Rev. Rothesay Miller, who made the usual three day's trip over the mountains in less than a day and a half on his American bicycle.

The article is illustrated by this photo:

Wikipedia tells up that Eliza Scidmore, among other accomplishments, was the originator of the idea of planting cherry trees in Washington DC.

I'm sorry to say that E. Rothesay Miller's 1901 Sketch of the North Japan Mission contains no account of his encounter with the tsunami's aftermath in 1896.

The Atlantic article that Victor mentions is not really an antedating of the OED's citation — it's the original source of its Lafcadio Hearn quotation. The article is Lafcadio Hearn, "A Living God", and it includes two uses (or perhaps mentions) of tsunami. The first one:

From immemorial time the shores of Japan have been swept, at irregular intervals of centuries, by enormous tidal waves, — tidal waves caused by earthquakes or by submarine volcanic action. The awful sudden risings of the sea are called by the Japanese tsunami. The last one occurred on the evening of June 17, 1896, when a wave nearly two hundred miles long struck the northeastern provinces of Miyagi, Iwate, and Aomori: wrecking scores of towns and villages, ruining whole districts, and destroying nearly thirty thousand human lives. […]

And the second one:

Through the twilight eastward all looked, and saw at the edge of the dusky horizon a long, lean, dim line like the shadowing of a coast where no coast ever was, — a line that thickened as they gazed, that broadened as a coast-line broadens to the eyes of one approaching it, yet incomparably more quickly. For that long darkness was the returning sea, towering like a cliff, and coursing more swiftly than the kite flies.

"Tsunami!" shrieked the people; and then all shrieks and all sounds and all power to hear sounds wre annihilated by a nameless shock heavier than any thunder, as the colossal swell smote the shore with a weight that sent a shudder through all the hills, and a foam-burst like a blaze of sheet-lightning.

This article was apparently reprinted a year later as the first chapter of the work that the OED cites, Gleanings in Buddha-fields.

That same ADS-L discussion also unearthed an article that seems to have been widely reprinted in American newspapers during the last of week of July 1896, which includes the sentence

The dreaded tsunami (sea wave) was not altogether new to some of those who were soon to become its victims.

Google's News Archive has a version from the Luddington (Michigan) Record of July 30, 1896.

In any case, whenever tsunami first found its way into English, it didn't start taking over from tidal wave until after WWII:

And Eliza Scidmore's "earthquake wave", though the most accurate English term, has never had much mindshare.

Interestingly, most recent uses of "tidal wave" seems to be metaphorical.


  1. [links] Link salad wakes up slow and cold | said,

    March 14, 2011 @ 7:36 am

    […] Language Log on the history of the word 'tsunami' in English — When I was a kid, I swear we still called them 'tidal waves'. […]

  2. Bill Walderman said,

    March 14, 2011 @ 7:52 am

    Out of curiosity (and with apology for an off-topic question), is the initial [ts] in the Japanese word a phoneme that is independent of [t] or a word-initial allophone of [t]–sort of like the Japanese equivalent of the High German Second Sound Shift? If I'm not mistaken, something similar happens to initial [t] in Danish. Basboll, The Phonology of Danish, p. 60.

    [(myl) It's an allophone of /t/ before /u/. The same or a similar process turns /ti/ into [tɕi], etc. (Or maybe not.)]

  3. Mary Sweeten said,

    March 14, 2011 @ 8:38 am

    Another not-quite-off-topic subject I've been wondering about: How did "prefecture" become the preferred English word for the Japanese administrative division? Like Japanese "feudalism," it has connotations that I imagine aren't really applicable.

  4. jaymc said,

    March 14, 2011 @ 8:52 am

    Any reason why the Google Ngram graph ends at 2000? If you extend it to 2008, you'll see that "tsunami" surges ahead of "tidal wave"; I'm assuming the 2004 Indian Ocean disaster played some role in this.

    [(myl) A cut-off at 2000 is the default, I believe because the sample in the last few years is both smaller and also significantly different from that of earlier periods. But in this case, I think that your point is probably valid.]

  5. Valentine said,

    March 14, 2011 @ 9:31 am


    IIRC, prefecture was adopted as the official English version of the suffix "ken" during the Meiji Era because the administrator of a ken was an appointed official and the system had been adopted from the French prefectural system. Similarly, the Japanese Kokkai 国会 (lit. "national assembly") was given the English equivalent "Diet" because it had been modeled on the German Reichstag.

  6. blahedo said,

    March 14, 2011 @ 10:03 am

    I had been wondering just a few days ago: what did we say before tsunami, which "feels recent"? I mentally noted that it probably wasn't as recent as I thought but then forgot to follow up; I never was able to come up with "tidal wave" until the LL discussion. And it surprised me, because I never quite realised that they were the same thing!

    In particular, I think of a "tidal wave", naively, as a huge wall of water that crashes down, a sort of epic surf wave, along with a bundle of metaphorical meanings; a tsunami is much more like a storm surge, driving debris ahead of it, possibly for hundreds of yards or even a few miles, and then sucking it all back out to sea. Looking at the above chart, I'm guessing that the increase in "tsunami" coincided with my learning about the quake-driven nature of the thing, and because "tidal wave" was not frequently used as a synonym, I never linked them back up.

    Incidentally, if we're proposing better terminology I'd go with "quake surge" over "quake wave"; the analogy with "storm surge" actually is a pretty good one.

  7. Mary Beckman said,

    March 14, 2011 @ 10:46 am

    Regarding the phonemic status of the affricates, the idea that alveolopalatal [tɕ] is an allophone of /t/ was always predicated on a very abstract, orthographic phonemicization for Sino-Japanese words such as [tɕa] 'tea'. And while that analysis may have been viable for old fogies such as me, it is no longer so, as young children learn words such as [remonti:] (black tea with lemon in it) and [tiɕ:u:] 'tissue' (the latter having totally replaced [tɕirigami]). There are also many young folk that say [tu] instead of [tsu] for words such as the "to" in "Back to the future", and [tsaitogaisuto] < German Zeitgeist was, I think, already in use back in the 1950s by people who talked about such things.

  8. Alan Gunn said,

    March 14, 2011 @ 11:55 am

    "I had been wondering just a few days ago: what did we say before tsunami, which "feels recent"? I mentally noted that it probably wasn't as recent as I thought but then forgot to follow up; I never was able to come up with "tidal wave" until the LL discussion. And it surprised me, because I never quite realised that they were the same thing!"

    I was taught as an undergraduate geology major in the 60's to call them "seismic sea waves," a term I don't think I've heard since. We also learned "tsunami," but didn't use it in ordinary conversation. "Tidal wave" was discouraged because they have nothing to do with tides, though it's hard to see why that should matter much.

    Googling "seismic sea wave" turns up a lot of questions in the form of "what are these?" Usual answer: Tsunamis.

  9. army1987 said,

    March 14, 2011 @ 12:35 pm

    Without context, I would have never thought that “earthquake waves” would refer to water waves.

    @Mary Sweeten:
    Well, I have seen Italian comune (which, as a masculine noun, means nothing more than “what a mayor has jurisdiction upon”) translated in English as commune, where township or municipality would be far more appropriate.

  10. army1987 said,

    March 14, 2011 @ 12:49 pm

    @Mary Beckman:
    Would you consider all of /c ɕ t͡ɕ n̠ʲ ç mʲ ɽʲ ɟ d͡ʑ bʲ pʲ/ to be distinct phonemes in pre-WWII Japanese?

  11. ShadowFox said,

    March 14, 2011 @ 3:17 pm

    @blahedo: If you follow most dictionaries and glossaries, they consider "tidal wave" a misnomer, as tsunami has no relation to tides. In fact, even the wave's behavior is drastically different from tidal, as there is a significant water recession prior to the onslaught of the giant wave.

    A number of other terms, including "earthquake wave" were associated with the one tsunami that was recognized globally prior to the 1896 Japanese disaster (that killed upwards of 20000 people). It was the surge caused by the Krakatoa explosion in 1883. However, one word that is not associated with that coverage–even via mention–is tsunami. If you search Google Books and News Archive for "Krakatoa" and follow reports immediately following the disaster (somewhere between 3 weeks and 6 months, roughly, as news distribution was still fairly slow), you can find all sorts of terminology. Note that even in 1896 the news of the Japanese disaster took 5-6 week to distribute, as the event occurred on 15 June 1896 and the first news reports in the US did not appear until July 25.

  12. ShadowFox said,

    March 14, 2011 @ 3:25 pm

    @army1987–At the time, in the last quarter of the 19th century, no one thought of earthquakes as waves. So the association with water would have been perfectly appropriate. In fact, it may very well have been the only plausible association at the time.

  13. Christy said,

    March 14, 2011 @ 3:35 pm

    I like blahedo's new term of "quake surge." While "tidal" is definitely wrong, so is "wave." A wave breaks on the shore. A tsunami is basically the ocean being displaced. It doesn't break – it just keeps coming. Once I learned the word tsunami, I never used tidal wave again.

  14. Joseph said,

    March 14, 2011 @ 3:38 pm

    Is there any indication of whether mentions of words count as uses of words for the purposes of ante-dating? For example, if "tsunami" had appeared in an English-Japanese dictionary as a headword, that would not count. How is "'Tsunami!' shrieked the people" any different?

  15. Mary Beckman said,

    March 14, 2011 @ 4:01 pm

    @ army1987
    Sorry, I'm old enough to be an old fogey with respect to the *[ti]; it's only in the last year or so that I've managed to say [supageti:] instead of [supagetei] or [supagete:], and I have a clear auditory memory of older speakers saying [supagetɕi:]. But I'm not old enough to have any auditory memory of how speakers said things pre-WWII.

    A mark of how old-fogey this [supagete:] is, btw, was in an on-street poll that the Japanese GQ TV show did, in which they asked young women about the markers that a man is ossanpoi, and one response was, "He says [dezunirando] instead of [dizunirando]."

  16. Edward Carney said,

    March 14, 2011 @ 6:21 pm

    I share Victor Steinbok's curiosity about the OED and its antedating. Several years ago I informed them of an antedating of "loony" (in the sense of their number 1—"Lunatic, crazed, daft, dazed, demented, foolish, silly.") I found a usage in "Bartleby the Scrivener: A Story of Wall Street," which antedates the earliest usage the OED had by some 19 years. their quotation is from Bret Harte's story "An Heiress of Red Dog, collected in a volume published in 1879 and which they date 1872— perhaps from its original publication date.

    'I think, sir, he's a little luny,' replied Ginger Nut with a grin."
    which appeared in Putnam's Monthly Magazine of American Literature, Science and ArtVolume 2, Issue 11, (November 1853).

    Their entry remains unmodified. They're not good at updating antedating.

  17. dirk alan said,

    March 14, 2011 @ 10:01 pm

    word was invented by susan nami. case slammed shut.

  18. Xmun said,

    March 14, 2011 @ 10:30 pm

    Could somebody please explain what the phonetic symbol ɕ represents? Some verbal explanation must be possible. Is it a sibilant, a fricative, or what? Alveolar or labial or something else? I presume it's unvoiced.

  19. Xmun said,

    March 14, 2011 @ 10:32 pm

    Sorry to bother you. Should have Googled. It's a voiceless alveolopalatal fricative.

  20. Ray Girvan said,

    March 14, 2011 @ 11:42 pm

    @Edward Carney: They're not good at updating antedating

    Yeah, same experience here. I felt briefly inspired by the thought of helping, but two years ago I antedated their entry for "diddums" by nearly a decade – see Diddums, then – and got an acknowledgement. But it's not in the Nov 2010 revision, so stuff 'em.

    I just realised that I know the Gleanings in Buddha-fields account in a different guise: a 1950s (?) children's retelling by Elizabeth Clark called Hamaguchi's Harvest. I didn't know until recently that it was based on a real person.

  21. Matt McIrvin said,

    March 15, 2011 @ 1:00 am

    @Christy: Physically, it is a wave; it's just an enormously long one in the direction of travel.

  22. Stuart said,

    March 15, 2011 @ 4:47 am

    The Civil Defence information in NZ telephone used "tidal waves" into the 70s at least, I remember that from my childhood, but "tsunamis" had conquered that realm by the mid-80s.

  23. Jerry Friedman said,

    March 15, 2011 @ 10:29 am

    @Edward Carney and Ray Girvan: I recently sent the OED a few antedatings too. Here's part of the response I got to one (guesstimate, from Google Books):

    "The existing OED entry has not been touched since it was written for the Supplement in 1972, and therefore represents a pre-database view of the evidence. I shall add your helpful references to the revision file, as an encouragement to the revisers to look for much earlier evidence."

    I take this to mean:

    Antedaters aren't going to get instant satisfaction.

    If you found the antedating in an obvious place such as GB, they'll find it too. The most you're doing is saving them a little work (which you might still consider worthwhile).

    What they can really use is antedatings that aren't in databases, or ones that have really weird spellings they might not think to look for, or ones for senses that are hard to search for, or words or senses of words they don't have at all. (I helped someone document the sense of acre as "the length of the side of a one-acre square".)

  24. Mark said,

    March 15, 2011 @ 10:35 am

    I'm going to fail to add anything helpful to this discussion and instead remark on how cool it is that the last graph showing the trends for "tidal wave" and "tsunami" looks like any of a thousand diagrams I've seen in my life that try to explain how a tsunami works. :-)

    Bill's initial gripe about the spelling still seems out of place on LL but it has led to some of the best discussions I read on here in some time. Rather than calling for your usual guarantee I think instead I'll double my subscription fee this week!

  25. Jerry Friedman said,

    March 15, 2011 @ 1:10 pm

    I should also quote the other e-mail I got:

    "Thank you very much for your message. We are always glad to hear from readers who have come across material that may be helpful to us.

    "As I am sure you are aware, the OED is very much a work in progress at the moment. We have revised everything from M to R, and also carried out some non-alphabetical revision; this has covered just over a quarter of the text. Your contribution will be added to the revision file to ensure that it is seen by the revisers when we come to work on the relevant range."

    So if you can antedate something that's been revised, they probably especially like that too.

  26. Mark F. said,

    March 15, 2011 @ 3:34 pm

    It would be interesting to try to find the first source deprecating the term "tidal wave." This seems like an example of expert-driven linguistic change.

    One reason that oceanographers don't like "tidal wave" as a synonym for "tsunami" is that the tide as it comes in is itself a (very slow) wave, and sometimes the term "tidal wave" gets used for that, if I understand this correctly.

  27. Xmun said,

    March 15, 2011 @ 5:17 pm

    Thanks to Ray Girvan for the link to "Diddums, then" and the link therein given to "Babies' trains". Reminds me of the bound volumes of Punch dating from the late nineteenth century that we had in the library of my preparatory school (= AmE private elementary school) in Berkshire in the late 1940s. The school — St Hugh's School, Carswell Manor, Oxfordshire, — still exists, but I don't know if those volumes are still in the school library. Somehow I doubt it.

  28. ShadowFox said,

    March 16, 2011 @ 2:45 am

    @Joseph–Generally, AFAIK, dictionary entries are only OK if it's an English-English dictionary. Foreign language dictionaries are not useful for antedating words on the "foreign" side. If a word is used in explaining another word, that's a different matter.
    Note that in most cases mention in the sense you identify is not included among the examples, although, obviously, there may be some exceptions. Consider, for example, the word "sabotage". If all we have is a dictionary definition of a French word or a French phrase/word quoted in an English-language text, neither identifies the use of the word in English. But, suppose the text says, "This is what the French call sabotage. Sabotage is …" Is this "mention"? I certainly think not, but your mileage may vary. But there may be some important information contained in a French-English dictionary that may help to identify the etymology of the word. Should this information be included? I can easily antedate OED citation for "sabotage" by 2-3 years, yet dictionary entries exist at least 30 years earlier–on the French side, of course. I would argue that, for historical accuracy and cleaner etymology, some of them may well be useful if included, but policy-wise, it may be better to exclude them all.
    @Ray Girvan and @Edward Carney: These complains seem somewhat petty. I've submitted a number of dates and corrections and I don't expect to see most of them for many years. The ratio of the number of editors to the number of words in the OED is very low and it does not even include the neologisms that arrive every day and need to be checked out and integrated. Actually, trying to antedate some of the lemmas gave me a new appreciation of the editorial job in a volume of this, well, volume. But I also grew increasingly frustrated with past editors who, to put it simply, did not always do their homework. Still, those of us who are upset over mistreatment of our own pet words and phrases should never forget that those comprise a small portion of even the original dictionary. And if everyone tried to "help" the editors all at once, the speed of corrections would slow to snail's pace. So the more accurate the work becomes, the more difficult is the editors' job. Please have some appreciation for their work.

  29. ShadowFox said,

    March 16, 2011 @ 2:54 am

    @MarkF–you can see the current OED entry and the 1899 citation clearly explains the preference against the use of "tidal wave" as translation. But it's not merely pointy-headed elitism. There is a tradition of distinguishing tidal waves from waves not caused by the normal action of the tides and, as I mentioned earlier, in references to the destruction caused by Krakatoa, this distinction is particularly clear, even though the word "tsunami" was not in use yet (in English). Although pervasive conflation of two unrelated ideas may well lead to users adopting a term that's initially described as "incorrect", only to become acceptable later, this is not a reason not to point out the distinction when the equivalence is still not fully integrated into the language. And references to tsunami as "tidal wave" are not particularly pervasive. To make a simple point, you can have tidal waves in Hawaii daily–weather you are going to differentiate between some special kind of "tidal waves" (i.e., those useful for surfing) and the generic waves cause by tides or not. But tsunami is a rare occurrence and is fully associated with seismic activity (which a tidal wave is not). If a small number of people conflate the two, it does not make them right, not should their ignorance (willful or not) be imposed on other speakers by omitting the explanation of the distinction.

  30. Richard Hershberger said,

    March 16, 2011 @ 4:35 pm

    Antedating tsunami a bit further, the Philadelphia Inquirer of July 26, 1896 has a report on what I suspect is the same tsunami as National Geographic and The Atlantic recounted, but a bit earlier by virtue of being published in a daily.

    It includes "The dreaded tsunami (sea wave) was not altogether new to some of those who were soon to become its victims." The word is not italicized. I don't know if the Inquirer of that era ever used italics.

  31. Julie said,

    March 16, 2011 @ 11:11 pm

    @Shadowfox: Does anyone really use the phrase "tidal wave" to mean a wave caused by tides? I grew up along the ocean and have never known the word to mean anything other than a surge caused by an earthquake.

    The link posted above by Mark F uses the phrases "tidal wave characteristics" and "tidal wave propagation" in the first few lines, but I wonder if those should be understood as tidal + wave propagation and tidal + wave characteristics.

  32. blahedo said,

    March 18, 2011 @ 9:57 am

    I was driving in to work today and heard Ben Zimmer (among others) on NPR talking about the history of the word tsunami in English! He read the first sentence of the National Geographic quote above; others talked about its etymology in Japanese (harbor + wave) and some of the semantic issues raised here.

    (I should also add that Ben's reading gave it a very distinct /ts/, and several other speakers also had a /ts/, but some had either an indistinct /t/ or no /t/ at all. :)

  33. Glennis said,

    March 18, 2011 @ 6:57 pm

    According to the Ngram viewer (if I'm using it correctly — it is late on a Friday night), even the word 'tsunamigenic' was coined before 1900. It then disappeared for a few decades and resurfaced in the scientific literature in the 1960s. However, my current favourite word, 'teletsunami', doesn't have quite the same pedigree.

  34. Ray Girvan said,

    March 19, 2011 @ 1:28 pm

    I don't know what's going on there. From here it's showing a blip just after 1900 (graph) – but that doesn't correspond to any occurrence in the books.

    [(myl) My guess is that the 1900-ish blip is a work that was mis-dated in the Google Books archive at the time that the n-gram corpus was created, but has since had its dating corrected.]

RSS feed for comments on this post