I recently ran across the word pagoda in a sense unknown to me (in the OED’s words: “A gold or silver coin of higher denomination than the rupee, formerly current in southern India”), so of course I looked it up, and found that it’s a much more complicated word than I had thought. Hobson-Jobson has a long entry beginning “This obscure and remarkable word is used in three different senses,” which you can see at Google Books here and at Digital Dictionaries of South Asia here (scroll to the bottom, and keep hitting “next page”); the OED (entry updated March 2005) has the following etymology:

< Portuguese pagode (1516 in sense 2a [an image or carving of a god; an idol], 1525 in sense 1a [a Hindu or (in later use esp.) Buddhist temple or sacred building, typically having the form of a many-tiered tower with storeys of diminishing size, each with an ornamented projecting roof], 1697 in sense 3 [A gold or silver coin of higher denomination than the rupee, formerly current in southern India]), of uncertain and disputed origin (see note below). Compare French pagode (1553 in Middle French in senses 1a and 2a; 1545 as paxode in sense 1a), Italian †pagode (1554 in sense 1a, 1587 as pagodo in senses 2a and 3; also †pagod (a1652 in sense 3), pagoda (18th cent.)), Spanish pagoda (1585 in sense 1a in the source translated in quot. 1588 at sense 1aα.; 1563 as pagode), Dutch pagode (1596 in sense 3 in the source translated in quot. 1598 at sense 3α., early 18th cent. or earlier in senses 1a and 2a; also as †pagood (1726 or earlier)), all apparently < Portuguese.

The ultimate origin of the Portuguese pagode is uncertain and disputed. It was once thought to be < Persian but-kada idol temple < but idol + kada habitation, but now seems more likely to be either < Tamil pākavata devotee of Vishnu ( < Sanskrit bhāgavata pertaining to the Lord (Vishnu), worshipper of Vishnu or the goddess Bhagavati: see below), or < Tamil pakavati (name of a) goddess ( < Sanskrit bhagavatī goddess, alternative name of the goddess Kali). Sense 3 arose from the fact that the image of the goddess was stamped on the coin (compare quot. 1598 at sense 3α.).

The stressing of the α forms has varied: ˈpagod occurs in Butler’s Hudibras (compare quot. 1664 at sense 2aα.); Pope has paˈgod as well as ˈpagod.

The initial stress in pagod can be seen most delightfully in Butler’s Hudibras (1664: “Their Classique-model prov’d a Maggot/ Their Directory an Indian Pagod”). Of the other citations, I think my favorite is:

1950 O. Sitwell Noble Essences 11, I beheld opposite.. the lean, elongated form of Lytton Strachey, hieratic, a pagod as plainly belonging as did the effigies to a creation of its own.


  1. David Eddyshaw says:

    The word “pagoda” invariably reminds me of “Lucy Church Amiably”, a work which defies description, although I suppose it says something about it that ever since I read it, the word “pagoda” invariably reminds me of “Lucy Church Amiably.”

    Those marijuana brownies evidently had something to do with it.

  2. January First-of-May says:

    I’m a coin collector, and while old gold coins had never been part of my collecting interests (far too expensive), I’m reasonably aware of all sides of coin collecting.

    In particular, the gold coin called “pagoda” is rather well known for its inclusion in the Proclamation of 1799 detailing money to be used in the Australian penal colony.

  3. Bathrobe says:

    Persian but-kada idol temple

    I assume there is no connection between but here and East Asian words for the Buddha…?

  4. George Gibbard says:

    I have read that the Persian word ‘idol’ derives from Sogdian, and the Sogdians were Buddhists. So I think it does derive from Buddha.

    Meanwhile kada derives from the Indo-Iranian root ‘dig’, since buildings were at some point in Central Asia dugouts. Classical Persian has kada, xān and xāna in related meanings (xān ‘inn’; xāna is the normal word for ‘house’).

    Whitney (1885) has “khan, khā, ‘dig'” for Sanskrit. Modern Persian has kan-dan ‘to dig’ in addition to the related ‘building’ words above, and the Arabic word xandaq ‘ditch’ must come from Persian. My proposal (no doubt reproducing earlier proposals I haven’t read) is PIE *kVnH- ~ *knH- > Proto-Indo-Iranian kanə- ~ khā-, with subsequent levelling in either direction as regards aspiration.

    Borrowings from Iranian include Ukrainian/South Russian хата (‘house’ or ‘hut’?) and Hungarian ház ‘house’.

  5. George Gibbard says:

    Meanwhile Persian xanda ‘laughter’.

  6. George Gibbard says:

    Whitney’s ‘dig’ has the supine khánitum, justifying my laryngeal (> Skt -i-).

  7. George Gibbard says:

    Meanwhile Persian kadxudā ‘landlord’.

  8. George Gibbard says:

    Forgot to add Sogdian kty ‘house’ (, and -kant in the names of Central Asian cities.

  9. Trond Engen says:

    Meanwhile Persian kadxudā ‘landlord’.

    I was going to suggest khankhan.

  10. Trond Engen says:

    Borrowings from Iranian include Ukrainian/South Russian хата (‘house’ or ‘hut’?) and Hungarian ház ‘house’.

    It’s borrowed into Proto-Uralic, yielding e.g. Finnish kota and N. Sami goahti. The Finnish word is homonymous to Swedish kota “hut”, which can be either a borrowing from Finnish or a derivation of Gmc. *kot “hut” (> Eng. cot), itself maybe an Iranian loan in Germanic. This gets even messier by introducing hut < IE *(s)kew-. I wonder if the PIE “hut” word (~*kutjan-) could have been borrowed into Uralic from PIE and from there to PII for a special type of house.

  11. Following up on cot, the word more usually used now is cottage, which I had thought was one of those familiar English words with a native root and a French affix. But no: it was apparently created in Normand from Old Norse kot ‘id.’ to mean ‘a cottage with its associated land’ (cf. burgage ‘town house with its plot of land’), borrowed into English, where it formed a doublet with cot, and eventually (in the 19C) back into Standard French.

    Cot ‘small bed’, the more normal usage of the word now, is straight from Hindi khat ‘hammock’ < Skt < something like Tamil kattil ‘bedstead’.

  12. Antonio Silva says:

    From Houaiss and Aurelio Portuguese etymological dictionaries you get:

    From Sanskrit bhagavati, via Dravidian idioms (malay pagôdi, tamil pagōdi)

    Funny that in Portuguese this word “pagode” can also mean fun, amusement with a rowdy tinge.

  13. David Marjanović says:

    Gmc. *kot

    Some kind of *kuta-? PGmc. didn’t have a short *o; that sound arose by “a-Umlaut” from *u followed by *a or on the way to Proto-Northwest-Germanic.

  14. Trond Engen says:

    Yeah, sorry, you must be right. I did that from memory. It’s not in B&L, and being on my way out I didn’t bother to look elsewhere.

  15. FWIW, Orel has PGmc. *kutan “from dial. Iran. *kuta-, cf. Av kata- ‘chamber’ “, referring to an article by a Viggo Brøndal in Acta Philologica Scandinavica III, p. 1-31. I haven’t checked whether there is any evidence for such an Iranian dialectal variant.
    Pokorny puts it under *geu- “bend” (Pok. II 394), together with other words derived from an extension *gu-do- meaning “hole, pit” etc., so either originally also designating a dugout dwelling or metaphorically a substandard dwelling.

  16. J.W. Brewer says:

    Note the Sogdian influence in:

    “Bloomin’ idol made o’ mud —
    Wot they called the Great Gawd Budd —
    Plucky lot she cared for idols when I kissed ‘er where she stud!”

    Note also the unusual three-way STRUT/GOOSE/FOOT merger in the particular variety of BrEng the poet is attempting to represent via eye-dialect . . .

  17. I grew up pronouncing Buddha with the FOOT vowel, presumably a spelling pronunciation because of the doubled consonant. So then all we need is a FOOT/STRUT non-split, which is well-attested.

  18. The FOOT vowel is the only one given in Daniel Jones’ Pronouncing Dictionary; the GOOSE vowel is not even a bracketed alternative. And in fact it’s still apparently the only UK version; the OED has “Brit. /ˈbʊdə/ , U.S. /ˈbudə/.”

  19. J. W. Brewer says:

    Hmm. I must never have heard (or at least noticed) a BrEng speaker actually say “Buddha” out loud. I know about the FOOT/STRUT non-split (I guess a more historically accurate label than merger) but think of it as a North-of-England thing, odd here only because I think of Kipling’s enlisted-ranks characters as generally speaking cockney (or at least a stage-dialect version of it). But maybe I’m wrong about that, or maybe he varies his non-posh dialect from character to character?

    A further complication is that as I understand it the unsplit North-of-England varieties have approximately what I would think of as the FOOT vowel where I have STRUT rather than vice versa, but Kipling’s eye-dialect spellings cue me to think all three words should be pronounced with ʌ rather than ʊ or u — especially since “stud” is only non-homophonous with “stood” if you have the split, and “stud” is on the STRUT side of it.

  20. Well, it wouldn’t surprise me if some minority had only STRUT and not FOOT, just as in the LOT-THOUGHT merger area, some say LOT (= PALM, typically) and some THOUGHT. But then again Kipling may have just been doing the best he could in order to snatch at a rhyme.

Speak Your Mind