The Language of Chess.

A useful roundup by Edwin Battistella at OUPblog:

Chess comes from the 6th century Sanskrit game chaturanga, which translates to “four arms.” The arms refer to the elephants, horses, chariots, and foot soldiers of the Indian army, which evolved into the modern bishops, knights, rooks, and pawns. The chaturanga pieces also included the king or rajah and the king’s counselor, which would later be reinvented as the queen. In chaturanga, the game ended when the rajah was removed from the board—when the king was killed.

Chaturanga was introduced to Persia around 600 AD and the rajah became the shah. Persian chatrang became Arabic shatranj and made its way to Morocco and Spain as shaterej. The word check, meaning an attack on the king, was adapted from the Persian shah. A player would say shah to announce an attack on the king. The expression checkmate came from the situation in which the king is attacked and has no defense: shāh māt means “the king is dead” and this connotation of regicide persists in the Russian name for chess: shakmati. [Sic: Should be shakhmaty.]

In Latin, the game was not named after the killing of the king, but after the attacks themselves—the checks. It was called ludis scaccorum (game of checks) or, when shortened [Sic: This is the Italian form], scacchi. The Latin word for check later gave us the Middle French eschec, which became échecs in the plural and chess in English.

There is much more on various chess terms from French and German; one thing I wish had been pointed out is that it is not just the chess sense of check that is from shah—the entire complex of English meanings comes from the chess term. See the Usage Note at the end of the AHD entry:

Through a complex development having to do with senses that evolved from the notion of checking the king, check came to mean something used to ensure accuracy or authenticity. One such means was a counterfoil, a part of a check, for example, retained by the issuer as documentation of a transaction. Check first meant “counterfoil” and then came to mean anything, such as a bill or bank draft, with a counterfoil—or eventually even without one.

Thanks, Trevor!


  1. Wiktionary suggests that “mate” was “deposed/conquered” rather than “dead”, and in the Farsi entry, we read “surprised / astonished”

  2. Pretty interesting. Given the limited direct western access to Persian in the Middle Ages, I wonder whether the shah-check relationship might have been mediated by the Arabic sheikh. Note the relationship in Spanish, where sheikh is jeque and check mate is jaque-mate. Online etymologies in Spanish mention an arabic word ‘shah’ meaning king, and French dictionaries online go straight to the Persian for the etymology, but I find this dubious, since both languages added a hard sound that seems less likely to derive from the Persian shah, or even from an Arabic borrowing of shah. Both French and Spanish have modern words for Shah that is pretty much just sha. And words for sheikh that sound very like their words for check.

    Clearly shah is the ultimate source. I just suspect some slightly more complex process to arrive at the western forms.

  3. David Eddyshaw says:


    I’ve always wondered about the supposed etymology of “shāh māt” as “the king is dead”; the word order is Persian, but the second word would be an Arabic perfect. I know modern Persian is full of Arabic loans, but that would be outright code-switching. Interesting to see that “māt” may well not be Arabic after all.

    Presumably the Arabs learnt the game from the Persians rather than vice versa, anyway, in the course of its long journey from India.

  4. David Eddyshaw says:

    Actually, I see that both AHD and Chambers say “shāh māt” is “Arabic.” It certainly isn’t a normal Arabic sentence, but the I suppose there’s no reason why an essentially holophrastic expression confined to just this one use would have to be.

  5. David Eddyshaw says:

    The actual current Arabic for “checkmate” seems to be كش مات ki∫ māt. At this point I feel a need for Lameen Souag …

  6. Calling Lameen!

  7. “One such means was a counterfoil, a part of a check, for example, retained by the issuer as documentation of a transaction. Check first meant “counterfoil” and then came to mean anything, such as a bill or bank draft, with a counterfoil—or eventually even without one.”

    Because we spell it “cheque” over here, I immediately wondered if there was a connection between “cheque” and “exchequer”, and found that there was, and it goes via chess – because the exchequer was the cloth with a pattern of squares on which accounts were settled, and it reminded people of a chess board. (It was, in fact, checked.)

  8. Wonderful!

  9. ə de vivre says:

    FWIW, in Turkish you say “şah mat”, and Nişanyan gives “mat” as being from Arabic, “~ Ar māta مات [#mwt] öldü”.

    From my (admittedly pretty rusty) Arabic, “shah māt” would be perfectly good Arabic for “the king (is) dead”. The VSO “māta shah” would be “the king died”. It is odd though that the [h] in “shah” wound up as a [k] in English. I can’t think of any other examples of that off-hand.

  10. Well, the English -k is from French. (Nice to see you back, ə!)

  11. Most of this sounds familiar, but the etymology of check in its other uses is a cool new fact to me. It’s not too common to see a specialized term like this generalized into more neutral uses, much less in a language half a dozen steps downstream from the original source!

    …So lemme guess, did Australian English mate also originally mean “someone who’s cornered and has nowhere to go (except Australia)”? 😉

  12. ə de vivre says:

    Thanks LH. I missed spouting off my half-baked linguistic opinions with people who are at least interested enough to tell me why I’m wrong 😉

    PS. I realized after the editing period for my post ended that I’d completely forgot about the article in Arabic. sigh. I’ve had my head in Akkadian recently, which is articleless (šarru mītu indeed). Anyways, whoever composed “shah mat”, would’ve had to know enough to either take out the Persian copula or the Arabic definite article. Since Europeans usually weren’t great at separating out the Arabic articles (cf. /al-ʕūd/ > /lut/, or early references to the “alquran”), maybe “shah mat” was used as otherwise ungrammatical technical jargon (per David Eddyshaw) in Persian or Arabic, or maybe it passed through another language with enough exposure to Arabic/Persian to take out the grammatical elements (Greek?).

  13. ə de vivre says:

    Now that I think about it, if [al-ʕūd] became [lyt], the corresponding transformation for [ash-shāh] would be [sh-shah], so maybe “shah mat” is a reasonable Frenchification of an Arabic “ash-shāh māt”. Although “luth” would often be used with an article in French, whereas “shah mat” would only be used as a phrase without one. So “le luth” might be closer to “al-ʕud” than “Ø shah” to “ash-shah”. ¯\_(ツ)_/¯

  14. I always understood that “check” in the financial sense derives from Arabic صك ṣakk, attested in the sense of “cheque” during the first millennium AD; see eg . If it doesn’t, it would certainly be a very odd coincidence.

    In Algeria people tend to know chess via French rather than through Arabic, but as far as I can see from Google, kiš is indeed the normal Arabic term for “check”, and kiš māt or kiš malik for “checkmate”. I would guess at a Persian origin: koš is Persian for “kill”, and keš for “pull”.

  15. In Persian, it seems they use both “šɑh mɑt” and “kiš o mɑt” for “checkmate” – and the mɑt here is argued to be a Persian word, not an Arabic one. “That it has nothing to do with the Ar. mâta is further proved by the evidence in the older Persian manuscripts about chess, where the word used for “‘being checkmated’ (mât shodan or shâh-mât shodan) is given as dar-mândan, mândan being the root of the word mât.4″ Sadly, the author doesn’t discuss “kiš”.

    M. E. Moghadam, “A Note on the Etymology of the Word Checkmate,” Journal of the American Oriental Society 58.4 (Dec. 1938), pp. 662-664

    šâh mât in Wiktionary

  16. David Marjanović says:

    Europeans usually weren’t great at separating out the Arabic articles

    That’s the Spaniards. The Italians did much better: off the top of my head Sp. arroz, It. riso > rice.

  17. Some nice discussion at the Encyclopedia Iranica.
    Hebrew has a couple obsolete terms: Mishnaic nardašir ‘backgammon’ or ‘chess’ < Middle Persian nēwardašīr; and Rashi’s iškuki < Old French eschec, still popular with 19th century Hebrew revivalists. Modern Hebrew uses šax, a borrowing from one of the usual European sources.

  18. IIRC the OED says that the sense of “check” meaning to verify is a sense-development from the sense of stopping, the idea being that you stop something to verify it. Ironically, I can’t be bothered to check if I’ve remembered that right.

  19. Is the sense of “check” as a pattern formed of an array of squares also derived from the chess “chequerboard”, then? I think I have heard that before somewhere.

    Is it not possible that كش is a metathesis of the form that begot “check”?

    I played chess a bit in Cairo many years ago, and it was called “shatarang” (with the hard g of the Cairene dialect – don’t know the Arabic spelling). The rook retains the old name of “elephant” (فيل ), the knight is a “horse” (حصان ) and the bishop a “minister” (وزير

Speak Your Mind