Sino-Sanskritic "devil"

« previous post | next post »

One of the most curious and fascinating words I learned during the first or second year of Mandarin study was móguǐ 魔鬼 ("devil; demon; fiend").  Somehow it just sounded right as the designation for what it signified:

Tā shìgè móguǐ 他是個魔鬼 ("He's a devil")

Even the characters, which I have always deemphasized since I began learning Mandarin, seemed appropriate. Guǐ 鬼 ("ghost; spirit; apparition; deuce"), the representation of a bogeyman that goes all the way back to the oracle bone inscriptions more than three millennia ago, was the thing itself.  Although I didn't know the exact meaning of mó 魔, it too had the guǐ 鬼 radical, so I thought of móguǐ 魔鬼 as a "mó 魔 type guǐ 鬼", and I just took it on faith that it meant "devil".

It was only after I started studying the transmission of Buddhism to China that I became acquainted with the Sanskrit word मार māra (“killing; death; Destroyer; demon; Evil One”) and its transcription in Sinographs as móluó 魔羅 (Middle Sinitic muɑlɑ).

So that's where whoever dreamed up the word móguǐ 魔鬼 ("devil; demon; fiend") got the first syllable, from Sanskrit.  They had to invent a new character to represent that syllable, which they handily did by taking the long established graph má 麻 ("hemp; cannabis; flax"), which goes back to the bronze inscriptions nearly three thousand years ago for the sound and joining to it guǐ 鬼 ("devil; demon; fiend") for the meaning.

In the course of preparing this post, I came to know of the ryakuji 略字 (extreme, unofficial simplified kanji [Sinograph]) form of 麻 in Japanese, which consists of the radical 广 + the katakana マ as phonetic “ma” for the on [Sinitic] reading).

The mó 魔 syllable was adopted into Korean as well where we find such words as "maryeok / malyeog 마력 / 魔力" ("witchcraft; witchery; spell; charm") and "mabeob 마법 / 魔法" ("[black] magic; enchantment; sorcery; wizardry").

Modern Sinitic (and by extension other East Asian languages) are full of Sanskrit loanwords, such as the following:

chànà 剎那 ("instant" < kṣaṇa)

chán(nà) 禪(那) ("meditation; Zen" < dhyāna)"

púsà 菩薩 ("Bodhisattva" < pútísàduǒ 菩提薩埵 [shortened by taking only the first and third syllables])

fāngbiàn 方便 ("convenience" < upaya), with which regular readers of Language Log are thoroughly familiar

Most borrowings from Sanskrit into Chinese are sound transcriptions (the first three above), but some are translations (the fourth one above).  Móguǐ 魔鬼 ("devil; demon; fiend") is interesting in that it is half transcription and half translation, though it started out as full transcription, viz., móluó 魔羅 (“killing; death; Destroyer; demon; Evil One”).  It's similar to the far more recent borrowing of English "ice cream" as bīngqílín 冰淇淋.

[h.t. Kendra Dale]


  1. Victor Mair said,

    December 11, 2018 @ 3:42 pm

    Dan Boucher just reminded me of another good example of a Sanskrit –> Sinitic borrowing that is half transcription and half translation, chándìng 禪定 ("meditation"), which means the same thing as chán(nà) 禪(那) ("meditation; Zen" < dhyāna)" mentioned in the o.p.

  2. Ross Presser said,

    December 11, 2018 @ 3:57 pm

    I suppose this is essentially the same word as the Mogwai from the Gremlins movies?

  3. Lars said,

    December 11, 2018 @ 4:04 pm

    @Ross Presser, Victor will be quick to correct me, but mogwai is probably Cantonese. Mogui is Mandarin.

  4. cameron said,

    December 11, 2018 @ 4:22 pm

    Mogwai is also the name of a band from Scotland, and yes, it's Cantonese.

  5. JB said,

    December 11, 2018 @ 8:13 pm

    I've seen some different explanations for the origin of the 魔 character, for example:


    The idea of friction producing heat (afflictions) is a tempting one.

  6. Victor Mair said,

    December 11, 2018 @ 9:06 pm

    There's a whole galaxy of derived terms and usages between Sanskrit māra (“killing; death; Destroyer; demon; Evil One”) and its Sinitic transcription móluó 魔羅 (Middle Sinitic muɑlɑ). There's nothing comparable to that involving 磨 ("friction; grind"). Nor are 魔 and 磨 commonly used interchangeably. Furthermore, Buddhist 魔 is about demons and devils, not afflictions.

  7. Andreas Johansson said,

    December 12, 2018 @ 1:19 am

    Would this be the word I've seen translated as "demon devil" in accounts of the Taiping?

  8. John Swindle said,

    December 12, 2018 @ 5:10 am

    @Andreas Johansson: I think "demon devil" or "demon-devil" translates "閻羅妖" (simplified: 阎罗妖), Yánluó yāo, where Yánluó is Yama, the king of the underworld, and yāo is demon. That's what Hong Xiuquan, the leader of the Taiping Rebellion, called the the earthly minions of the devil who were ranged against him. Still a kind of móguǐ, I suppose.

  9. Chris Button said,

    December 12, 2018 @ 7:28 am

    As an aside, the oracle-bone character for 鬼 is an interesting one in terms of "folk-etymology" (if "etymology" is an applicable word for graphic analysis). The current analysis cited by Wikipedia is almost certainly incorrect (as also is its linked analysis of 異), but it does provide a cute mnemonic device in terms of a devilishly attired figure.

  10. Doc Rock said,

    December 12, 2018 @ 8:19 am

    For those who wish to explore further I recommend Soothill & Hodous, A Dictionary of Chinese Buddhist Terms

  11. Victor Mair said,

    December 12, 2018 @ 8:49 am

    @Chris Button

    It does look like a kneeling (or bent over) man with a huge head, but I don't see a tail on the oracle bone form. Maybe that's a hunchback?

    What do you think it represents?

  12. Victor Mair said,

    December 12, 2018 @ 8:56 am

    From Zeyao Wu:

    Liang Xiaohong (梁曉紅) also analyzed this term in her book Fojiao ciyu de gouzao yu hanyu cihui de fazhan (佛教詞語的構造和漢語詞匯的發展 The Construction of Buddhist Words and the Development of Chinese Vocabulary). She said that at the very beginning of the translation activities, Mara is translated as "磨". Emperor Wu of Liang later changed the 石 to 鬼, and made it become 魔. Considering its two syllables, it became 魔羅. It is also the name of Papiyas. According to Huilin 慧琳 in Tang Dynasty, the 儸 is usually omitted. Since this term is similar to the Chinese term 鬼神, the combination of 魔鬼 appears later. Liang names the category of this kind of phrases as "transliteration+Chinese characters".

    She also exemplifies many other new terms such as "塔 (stupa)+Chinese characters" and "懺 (Ksama)+Chinese characters".

    [VHM: This shows that JB's suggestion was right for the very earliest stage in the transcription of the Sanskrit word māra, but apparently Emperor Wu of Liang (464-549) himself (an ardent Buddhist) was dissatisfied with 磨 ("abrade; friction") as a transcription for the mā- part of māra. Emperor Wu of Liang later changed the stone signific 石 to the demon signific 鬼, and made it become 魔 ("ma- type demon").]

  13. Victor Mair said,

    December 12, 2018 @ 11:06 am

    From Jan Nattier:

    Many thanks for these notes. Very interesting, as always!

    Just a couple of small remarks: first, 魔 is not an abbreviation of 魔羅 (nor is 禪 an abbreviation of 禪那, or 菩薩 an abbreviation of 菩提薩埵). In all of these cases, the shorter form came first (all of these transcriptions are attested already in the 2nd century CE). The longer forms came several centuries later, as Chinese Buddhists started receiving texts written (or recited) in Sanskrit rather than in various Middle Indic dialects, and starting becoming more meticulous (or more persnickety, depending on one’s point of view) at recording the writing, rather than the pronunciation, of these Sanskrit (no longer Middle Indic) terms.

    Second, I don’t see why 方便 should be considered a “Sanskrit borrowing” in Chinese. It’s a Buddhist term, of course (it doesn’t seem to be attested in any pre-Buddhist Chinese source). But when it was first introduced it was not based on Sanskrit, and it was not a translation of upāya.

    The first occurrence is in the works of An Shigao, where it serves as a translation of one of the members of the eightfold path, viz., (samkak)vyāyāma “(right) effort.” Only later did it come to be adopted as a translation of upāya “tactics” (e.g., in the compound upāyakauśalya “tactical skill”).

    So I would think this should be called a “Buddhist translation term” that subsequently gained currency in non-Buddhist Chinese, rather than a “Sanskrit borrowing.”

    But there may be arguments in favor of other interpretations.

  14. Victor Mair said,

    December 12, 2018 @ 12:21 pm

    From Zeyao Wu:

    The original text recording this change is Fanyi mingyi ji 翻譯名義集. In the second volume "Simo pian 四魔篇 The section of Four Devils", it says

    In the ancient translated [Buddhist] scriptures and expositions, the character 魔 mo follows [the character] 石 shi. Since the Emperor Wu of Liang, it is said that mo can annoy people. Thus, the character is preferable to follow gui.

  15. Chris Button said,

    December 12, 2018 @ 3:12 pm

    @ Victor Mair

    The one thing about which I think the Wikipedia article is correct is that the top components of 鬼 and 異 are related. However, I prefer the long-standing interpretation of 異 as a person carrying a basket on their head (as later represented by 戴) which Vernon Fowler (in his 1989 dissertation at UBC under Takashima) extends to 鬼. I need to give it a little more thought but in terms of semantics one might compare an ultimate relationship between "porter" and "fear" in English while noting the clear association between 鬼 and 畏 in terms of the latter sense.

  16. Andreas Johansson said,

    December 12, 2018 @ 3:19 pm

    @John Swindle


  17. Chris Button said,

    December 13, 2018 @ 7:10 am

    I should probably add that while 異 clearly shows a person physically carrying a basket on their head with raised hands, the case of 鬼 is using the "basket" component (viz. 簣) for its iconic value graphically placed above a human form rather than as a pictograph showing any actual carrying.

  18. Adrian Morgan said,

    December 22, 2018 @ 7:18 pm

    Doctor Who fans will recognise "mara".

RSS feed for comments on this post