In my ongoing Kurosawa retrospective, I’ve reached 1954 and Seven Samurai, which I’ve seen more often than any other of his (it’s one of my favorite movies, and of course whenever I read DeWitt’s great novel I have to rewatch it). I’ve been following along with the excellent commentaries (two different tracks, one with five film scholars!) on the Criterion set, and at one point there’s a reference to the “shamisen player.” I’ve run across that form before, but I think of the instrument as a samisen, and I finally got irritated enough to investigate. My memory that it used to be known as samisen turns out to be correct; the OED (entry from 1909) has only that form even in Japanese (“Etymon: Japanese samisen”), and Webster’s Third (1961) uses the headword samisen, adding “also samsien […] or shamisen.” (I was unaware of the form samsien, and Google Books finds only 19th-century uses.) But here in the 21st century, Wikipedia has it under Shamisen, starting the article “The shamisen or samisen (三味線), also known as sangen (三絃) (all meaning ‘three strings’), is a three-stringed traditional Japanese musical instrument derived from the Chinese instrument sanxian,” and AHD has “sham·i·sen (shăm′ĭ-sĕn′) also sam·i·sen (săm′-).” So why the difference?
Wiktionary has an interesting and complex etymology:
The instrument derives from the Okinawan 三線 (sanshin). Originally called 蛇皮線 (jabisen, literally “snakeskin strings”) in Japanese, so named for the way the Okinawan instrument’s soundbox is traditionally covered in snakeskin. The traditional jabisen instrument was imported into the Sakai area of Osaka during the Eiroku era (1558–1570), then later modified by biwa luthiers to have the square-shaped shamisen soundbox of today.
The reading jabisen shifted over time to 蛇味線 (jamisen), replacing the 皮 (bi, “skin, leather”) character with 味 (mi) for phonetic reasons, i.e. as ateji (当て字). Then jamisen changed to shamisen, replacing the 蛇 (ja, “snake”) character with 三 (sha, usually read san, “three”) for semantic reasons. The sha reading for the 三 character is irregular.
The shamisen reading is first cited to a text from 1580.
AHD derives it straightforwardly from san ‘three’ + mi ‘taste, touch’ + sen ‘string’; the references for the Okinawan etymology are in Japanese, so I can’t evaluate them. Thoughts?
Incidentally, the etymology of samurai itself is more complex than I would have expected; Wiktionary:
⟨sa morapi1⟩ → /saburapi/ → /saburafi/ → /samurawi/ → /samurai/
Sound shift from saburai (see below), itself the 連用形 (ren’yōkei, stem or continuative form) of classical verb 侍ふ (saburafu, “to serve”, modern saburau).
The shift from saburaf- → samuraf- is due to the development of the 女性語 (josei-go, “women’s language”) in Middle Japanese. A different development of this b is seen in 候 (sōrō, “service”, classical saurafu).
The usage note is also worth a read:
In modern popular usage, the terms bushi and samurai are often used in both English and Japanese somewhat loosely to refer to any soldier or warrior during Japan’s feudal age (prior to the Meiji period). In historical contexts, these terms may be used with more specific senses, wherein bushi refers to any professional warrior, and samurai refers more strictly to a hereditary social class. See also Samurai on Wikipedia.
All I know is that when I lived in Japan from Showa 63 to Heisei 1, we all said „shamisen”. I can’t recall having ever heard „samisen”.
I’m not able to say much on the etymology, but I will note that the version of the instrument from the Tsugaru Peninsula is called theTsugaru-jamisen (津軽三味線, つがるじゃみせん). Presumably, the rendaku (sequential voicing) form corresponding to shamisen (三味線, しゃみせん) is being used here (with sh at the start of the second element in a compound regularly changing to j) rather than this being a survival of the older pronunciation.
Rendaku usually (though not exclusively) applies to words using the native kun’yomi readings, so it is interesting that it seems to be occurring here when shamisen would seem to be an on’yomi reading. Maybe the irregular sha for 三 has something to do with it.
Well, we left Japan in Showa 40. I wonder when general usage changed?
I lived in Japan more recently (though not due to youth, see below), and I only heard shamisen. I strongly suspect, given the explanation on Wikipedia, that jabisen/shamisen is the only pronunciation that has ever been common among native speakers. It is pretty common for words to be assigned ahistorical kanji years after they enter the language, though admittedly forcing an irregular pronunciation is less common.
To me, it seems pretty likely that the English term samisen is a mistake by people educated enough to know how the characters ought to be pronounced, but not knowledgable enough to know how it is actually pronounced in Japanese.
I have made this same mistake. My wife and I were on vacation in the Cotswolds. I got lost on the way to Cirencester, and had to ask a local for directions (this gives you some idea of how long ago this was). Being a woman who prides herself on being more worldly than the average American, I asked the way to “Seer-un-stur.” The man looked at me like I was the biggest idiot he had ever met, and asked if I was talking about “Sye-ren-ses-ter.” Turns out, the Brits have one town that they pronounce exactly the way you would pronounce it if you were an American who had never stepped outside of Bethel, Arkansas, and I was blissfully unaware because my knowledge all came from books.
Maybe shamisen is a similar linguistic man trap for tourists. I like to think so.
I play both the shamisen and the sanshin, and I have had plenty of exposure to people talking about both.
To the best of my recollection, I’ve never encountered the pronunciation “samisen” in the wild, from professional players, instrument makers, specialist wagakki (Japanese instrument) sellers, or random Japanese people. I don’t think it’s current, if it ever was, and I’ve been travelling to Japan since 1999 (or Heisei 11 if you prefer!).
I have heard the word “jamisen” (ja 蛇 for snake, I assume) on a few occasions from mainland Japanese referring to the Okinawan sanshin.
I’ve always assumed it was just rendaku in “Tsugaru jamisen”. Both “Tsugaru shamisen” and “~ jamisen” are valid, but the dialects of north-eastern Japan (where Tsugaru is) do tend to voice consonants, so there might be a regional influence.
In Worcester, Leicester, etc., the second syllable is unstressed, the vowel sound is dropped, and the two occurrences of /s/ merge into one, giving the familiar simplified pronunciations in two syllables. In Cirencester, though, it is not the second but the third, and has stress (albeit secondary). This blocks the above simplification.
My own childhood residence in Japan was during Shōwa 48 through 51, and the google books ngram viewer says it was indeed during that fairly brief period while I was there that the trendlines crossed, with the “shamisen” spelling becoming the majority variant and the “samisen” one the minority. By early in the Heisei era the “sh” variant was more than twice as common.
FWIW I think of the “sh” as normative and have no specific memories of exposure to the “s” version, although I accept that I have no sufficiently clear memory of seeing the word written out in romaji in 1974 to be certain that early exposure might not have been overwhelmed by subsequent.
To me, it seems pretty likely that the English term samisen is a mistake by people educated enough to know how the characters ought to be pronounced, but not knowledgable enough to know how it is actually pronounced in Japanese.
I’m sure you’re right, but what’s interesting to me is that at some point (apparently in the late ’60s) English-speakers gave up the s- version and settled on the sh- one. It’s certainly not automatic for anglicized forms to be replaced by more authentic ones, though I guess it’s been getting more common in these more cosmopolitan times.
Wikipedia (as distinct from wiktionary) says that “In Western Japanese dialects and several Edo period sources, it is both written and pronounced as samisen.” And the wiktionary piece hat linked to gives (under “Etymology 2,” which I find a bit confusing) “samisen” (i.e. さみせん not しゃみせん) as a “less common” variant pronunciation in Japanese. So I don’t know why we would assume that the earlier dominant spelling in English was a pure misunderstanding as opposed to an adaptation of a variant pronunciation extant within Japanese.
I know very little about historical shifts in Japanese pronunciation, but are there any other lexemes where there has been a shift from sa/さ to sha/しゃ or indeed a shift in the other direction, whether word-initially or otherwise?
The 1954 edition of the Kenkyusha’s New Japanese-English Dictionary (ed. Senkichiro Katsumata) has both variants, shamisen and samisen, with the sh- variant as the main entry. The translation is “a Japanese balalaika”.
Wikipedia (as distinct from wiktionary) says that “In Western Japanese dialects and several Edo period sources, it is both written and pronounced as samisen.” And the wiktionary piece hat linked to gives (under “Etymology 2,” which I find a bit confusing) “samisen” (i.e. さみせん not しゃみせん) as a “less common” variant pronunciation in Japanese. So I don’t know why we would assume that the earlier dominant spelling in English was a pure misunderstanding as opposed to an adaptation of a variant pronunciation extant within Japanese.
Good point, and I vote for an adaptation of a variant pronunciation.
Cirencester is a special case among special cases: “even the locals at some point realised they had gone too far, so they settled for a spelling-pronunciation” – after having devised no less than five shorter variants all the way down to “sister”.
Presumably you’ve just seen Ikiru 生きる again, then, Hat?
If forced to choose at gunpoint between Ikiru and Seven Samurai, I’d say that Ikiru is the greater movie. (I am sure that Watanabe is the reincarnation of Kambei.)
My daughter won’t let me play Takashi Shimura singing Gondola no Uta ゴンドラの唄 because it makes her cry. Actually, it makes me cry too …
I know very little about historical shifts in Japanese pronunciation, but are there any other lexemes where there has been a shift from sa/さ to sha/しゃ or indeed a shift in the other direction, whether word-initially or otherwise?
I don’t think this is any kind of regular historical sound change. Rather, the samisen pronunciation is an eggcorn: a perfectly real pronunciation, but remodelled by analogy/folk etymology. Three strings, so sa(n)-misen.
It’s the same process whereby Kusaal has borrowed Hausa lambu “orchard, garden” (ultimately from Songhay) as lɔmbɔn’ɔg, by analogy with bɔn’ɔg “ricefield, swamp.” A lampaddy.
(There are English examples of eggcorns made good too, but I can’t think of one off the top of my head.)
Presumably you’ve just seen Ikiru 生きる again, then, Hat?
Quite so.
If forced to choose at gunpoint between Ikiru and Seven Samurai, I’d say that Ikiru is the greater movie.
I remembered your love for Ikiru and thought about that very question; they’re so different that it’s hard to make a direct comparison — it depends to some extent on whether you prefer a concentrated story with a single focus or a complex one with a lot going on. I think I have to some extent made a transition from the former to the latter, and if forced to choose at gunpoint (bang! and a hero falls) I would opt for the samurai flick. But Shimura is ineffably great in both, playing very different characters. What an actor!
What an actor!
Youbetcha. I think my single most favourite scene in Ikiru is when Watanabe sees off some gangsters who have come to threaten him off his project by just looking at them (no words.) Shimura manages to make this completely convincing.
Shimura was of course also Dr. Kyohei Yamane in the original (and far the best) Godzilla movie, from the very same year as Seven Samurai. I think it’s a measure of his acting skill, rather than my own denseness, that I didn’t actually realise this until my younger son pointed it out.
(There are English examples of eggcorns made good too, but I can’t think of one off the top of my head.)
Crayfish, aerie/eyrie. There are quite a few, but if you didn’t look up “folk etymology”, I won’t either.
Godown?
I think it’s a measure of his acting skill, rather than my own denseness, that I didn’t actually realise this until my younger son pointed it out.
One of the commentary tracks mentioned an occasion when Mifune was being honored and someone praised his performance in I Live in Fear (生きものの記録, ‘Record of a Living Being’), which is next on my list (I’ve never seen it); another panelist blurted that he loved that movie but hadn’t realized that Mifune was in it, to which Mifune wryly responded “I am sorry my performance was so inadequate.” He realized, of course, that it was actually high praise. I remember my own astonishment when I saw him in his pre-samurai roles and realized that hey, he can do things other than gibber and jump around!
@David E.: This seems to differ from the classic eggcorn situation, because of interference between the actual “three” word and the not-too-dissimilar-sounding actual “snake” word for the same referent. Changing the orthography of the “snake” word to swap in the kanji you’d naturally use for a calque of the Okinawan “three” word was just asking for something wacky to result. If you ask me being somehow snake-related seems like a cooler distinctive attribute of a musical instrument than being three-stringed. But they didn’t ask me.
…And having seen I Live in Fear, I too wouldn’t have realized that the lead character was played by Mifune if I hadn’t already known. It’s not a good movie, but what an acting job!
I find Ikiru feels very different from most of Kurosawa’s films from that period. Some of it is the tone, but a big part of it is obviously that it’s the only film he directed from (I think) Drunken Angel to Red Beard that doesn’t have Mifune in it. Shimura, of course, does a tremendous job carrying the film, but it’s striking (as I remember from the first time we attended a months-long Kurosawa film festival in Cambridge) when suddenly Mifune, who we had gotten accustomed to always seeing in a major role, just isn’t there.
I don’t really have an opinion about what Kurosawa’s best film is. He covered so much material and different genres. However (as I think I have stated before), I think Kurosawa’s most characteristic film is Yojimbo, because he really makes use of his stable of favored actors; Toshiro Mifune, Tatsuya Nakadai, Isuzu Yamada, Daisuke Katō, and Takashi Shimura are all in it!
It’s a wonderful film, and underrated (like Leone’s Fistful of Dollars) because it’s “just” a fun action movie.
Yeah, you can’t really go wrong with ダシール・ハメット as source material.
The 時代別国語辞典 室町時代編 (Muromachi Period Dictionary of the Japanese Language) has both samisen and shamisen. In the entry for samisen there is a quote given which is dated 1588. This entry redirects the reader to the entry for shamisen. The entry for shamisen (orthographically here shiyamisen) is more detailed but the quote given is from the Portuguese Nippo Jisho (1603-4), a later date than the quote for samisen. The Portuguese transcribed the word as xamixen, where the orthographic x’s represent the [sh] sound, i.e. [shamishen]. For those still paying attention the second [sh] sound is there because the Nippo Jisho was based on the contemporary Kyushu variety of Japanese which palatalised /s/ before /e/ as well as before /i/. (Still does in fact.)
The Nara Period dictionary in this series of dictionaries does not have an entry for either shamisen or samisen, and unfortunately the Heian Period volumes have never been published.
Thanks very much! It’s odd that the Heian volumes have never been published; Japan is a wealthy country, and you’d think there would be enough interest that someone would take the plunge.
I don’t know Japanese, but I have listened to the Yoshida Brothers. Their Wikipedia page says they play “shamisen”. I wondered if the page had ever said “samisen”, and been changed, but no, the oldest revision (from 2006) says “shamisen”.
FWIW.
Yeah, in this century it’s consistently “shamisen”; I was wondering how far back the change happened.