Bad shits

« previous post | next post »

I received the following photograph of a sign taken by Son Ha Dinh in Damak, Nepal:

Son Ha helped us so much with these two posts: "Unknown Language #7," "Unknown Language #7: update." Damak is a town in eastern Nepal where Son Ha works in a refugee camp and where he encountered the woman who is featured in those two posts.

I don't think that "bad shits" requires extensive comment, though I will say that the fine gradations of English vowels seem to pose a lot of problems for nonnative speakers.  For example, my wife always had a devil of a time distinguishing among "pin, pain, pine, and pen".  On the other hand, I had trouble clearly differentiating certain consonants in Nepali, e.g., च /c, t͡ʃ/ and छ /cʰ, t͡ʃʰ/, such that the initials of cār चार ("four") and cha छ ("six") came out sounding very much alike, resulting in amusement on the part of my Nepali friends.  (You can hear recordings of the two words here.)

And I also found it challenging to differentiate the three sibilants:  श śa /ɕ, ʃ/, ष ṣa /ʂ/, and स sa /s/. [Note that the ‘s’ in the devanagari for “furnishing” is not the palatal ‘sh’ but the dental ‘s’; more on that below.]

Here's the transliteration of the main part of the shop sign:

maniṣā pharnisiṅg senṭar

damak 11, pharnicar lāīn phon …

(Manisha Furnishing Center, Damak 11 [I believe that is the "ward" number for Nepali addresses], furniture phone line…)

Manisha is a female name associated with a goddess of wisdom. It may be the name of the store owner or, perhaps more likely, a reference to the famous Nepali actress, Manisha Koirala. The next two words are obviously borrowed from English.

According to Philip Lutgendorf:

The Devanagari transliteration contains two “errors” (of a sort common in such signage).  One is the misspelling of “Furnishing” with the sibilant “sa” rather than “śa” or “ṣa” (as one would expect), though such substitution is common in Eastern Hindi and probably in Nepali as well; the reader will pronounce it “sha” in any case. The other is the use of the dental “na” as a conjunct with retroflex “ṭa” (in Senṭar). This is theoretically impossible according to Sanskrit rules; it should be Seṇṭar (सेंटर or सेण्टर), but again, popular signage doesn’t give a…..sheet….about such niceties!

The conjunct for “ng” uses the optional orthography, derived from Sanskrit, of representing the nasal of the glottal class by a full character normally avoided in modern Hindi (which would generally prefer a superscript dot or anusvār, to yield फ़र्निसिंग…).

I asked Leopold Eisenlohr if the last biconsonantal conjunct of the second word ("furnishing") of the sign is exactly equivalent to ङ्ग.  He replied:

Yes, it is exactly equivalent. The slanted stroke under the first letter is a virama, meaning there is no vowel following the consonant. It is usually preferable to make conjuncts out of consonants that have no vowels between them, but in pronunciation and transliteration both spellings are identical.

One thing to note is the fact that the g has no virama sign, so technically what is written should be transliterated pharnisin̄ga, though the maker of that sign assumed people would know not to pronounce the vowel and would not need the virama.

The letter there is the velar nasal usually translated as n̄ or ṅ. It's the velar n because it comes with the velar consonant g (instead of a labial, dental, retroflex, or palatal nasal). The dot between the two letters is part of the n̄/ṅ: ङ

I think I'll let Fred Smith have the last word:

As someone who has suffered plenty of intestinal problems over the last 40+ years of traveling in India, I really understand the "bad shits" sign. It should have been in front of a GI doctor's office.

[Thanks to Deven Patel, Philip Lutgendorf, Fred Smith, and Leopold Eisenlohr]

Update on 8/11/16 — Comment by  Aparna Garimella on this remark by Philip Lutgendorf in the o.p:

"The other is the use of the dental “na” as a conjunct with retroflex “ṭa” (in Senṭar). This is theoretically impossible according to Sanskrit rules; it should be Seṇṭar (सेंटर or सेण्टर), but again, popular signage doesn’t give a…..sheet….about such niceties!"

I'd like to clarify here that Sanskrit rules usually specify the retroflex nasal based on the preceding phonemes (the famous RUKI rules) not the following ones. So,

षेण्टर [ʂeːɳʈʌr], /ṣeṇṭar/


सेन्टर [seːnʈʌr], /senṭar/

are both okay, but not

सेण्टर [seːɳʈʌr], /seṇṭar/


  1. Piyush said,

    March 16, 2014 @ 12:25 am

    Given the usual traditions of naming things in small-town India (and I presume Nepal too), I would imagine that the name of the shop is much more likely to refer to the owner's daughter rather than to the owner himself/herself or to a movie actress.

    Also, I think the Wikipedia article linked in the post is rather misleading on the meaning of the word məniːʂa: (मनीषा) [मनिषा, as in the sign, is either the Nepali variant or a misspelling]. At least in standard Hindi, the word simply means wisdom—sages, 'those who are wise', are thus referred to as as मनीषी, məniːʂi:— with no references to any goddesses. However, it probably does have a more nuanced technical meaning in ancient Indian Philosophy, as the article claims.

    The Wikipedia article also wrongly claims that

    Another meaning of Manisha is daughter of mankind.

    The word they are thinking of is probably मानुषी, ma:nʊʂi:, which is the feminine form of मनुष्य (mənʊʂjə), descendant of Manu. I also cannot find any occurrences of मनीषा in various lists of the names of Durga, and so I am also doubtful of the article's claim that it is one of the 108 names of Devi.

  2. Ellen K. said,

    March 16, 2014 @ 12:47 am

    What is "bad shits" supposed to mean? Frankly, I haven't a clue.

  3. Piyush said,

    March 16, 2014 @ 12:59 am

    I think they wanted to write "bed-sheets".

  4. Stephan Stiller said,

    March 16, 2014 @ 1:03 am

    @ Ellen K.
    "bed sheets"

    Btw not making a distinction between [æ] and [ɛ] (and merging the former into the latter in speech production) is characteristic of many accents, including the well-known German one.

  5. Anubhav Chattoraj said,

    March 16, 2014 @ 1:58 am

    > And I also found it challenging to differentiate the three sibilants: श śa /ɕ, ʃ/, ष ṣa /ʂ/, and स sa /s/.

    There is no phonemic /ʂ/ in Modern Indo-Aryan languages. This isn't a bold claim; it's been part of linguistic consensus since the nineteenth century.

    Colin Masica's The Indo-Aryan Languages reports that Nepali lacks /ʃ~ɕ/ (henceforth /ʃ/) as well. (However, Standard Hindi certainly has /ʃ/.)

    I'm afraid Lutgendorf is off the mark when he says

    > One is the misspelling of “Furnishing” with the sibilant “sa” rather than “śa” or “ṣa” (as one would expect), though such substitution is common in Eastern Hindi and probably in Nepali as well; the reader will pronounce it “sha” in any case.

    The "misspelling" reflects a sociolectal difference; in the eastern parts of the Hindi-speaking belt, /ʃ/ vanishes as one descends the socioeconomic ladder. The people painting such signs probably lack /ʃ/.

    On the other hand, Lutgendorf probably hangs out with highly educated Hindi speakers, who almost universally have /ʃ/. Such speakers, being aware of the /ʃ/-lessness of other dialects, automatically read <स> as [ʃ] when appropriate.

    If Masica is right about Nepali lacking /ʃ/ entirely, then I'd expect that the only Nepali-speakers to read "फ़र्निसिङ्ग" with a [ʃ] would be those who also speak English.

    > The other is the use of the dental “na” as a conjunct with retroflex “ṭa” (in Senṭar). This is theoretically impossible according to Sanskrit rules; it should be Seṇṭar (सेंटर or सेण्टर),

    The <न्ट> conjunct is quite common in Hindi (and, I'm guessing, in Nepali) for transcribing the English /ntə/ (> Hindi/Nepali [ɳʈə]). This is acceptable because Hindi/Nepali both lack a phonemic /ɳ/, and nasals generally assimilate with the following stop.

    But I suspect that the real reason for writing <न्ट> instead of <ण्ट> is that the former just looks better. This would explain why one rarely sees <न्ड> (<nɖ>): it looks worse than the alternative <ण्ड>.

  6. julie lee said,

    March 16, 2014 @ 11:14 am

    I couldn't distinguish "bed" and "bad" when I was taught English as a foreign language at six. The American nun at a small mission in Guiyang , Guiizhou province, in central China, kept saying "bed", pointing to the spelling, and I kept saying "bad". I was in despair, couldn't hear the difference. Later at nine, that problem somehow disappeared, I don't even remember how.

  7. Sili said,

    March 16, 2014 @ 12:11 pm

    Much like the bad joke about the Italian who "wanta shit in my bed and a fuck on the table".

  8. beslayed said,

    March 16, 2014 @ 12:46 pm

    To follow up on Anubhav Chattoraj's comments:

    In fact, historically all modern Indo-Aryan languages lost the old three sibilant series, all of them collapsing it to one. In the west, to /s/, in the east (Nepali doesn't count as eastern in this respect) to /ś/. In Hindi and other languages, /ś/ is reintroduced through a combination of borrowings from Sanskrit, Perso-Arabic, and English. There are certainly Hindi speakers which lack the /s/-/ś/ contrast — I don't think it's necessarily "Eastern Hindi" speakers particularly (though it may include them), but typically less urbane, more rural speakers (the same speakers who will generally lack a /j/-/z/ contrast).

    Back to the specific Nepali context — of course Sanskrit does have a three way sibilant contrast and Nepali "preserves" some of this in terms of spelling conventions/rules (likewise, Nepali has lost distinctions between long and short /u/ and /i/, but the orthography preserves such contrasts — note that this feature of Nepali makes the "shit" for "sheet" substitution even more likely). Which means that characters for the other sibilants do show up, sometimes correctly (as in the name Manisha) and sometimes incorrectly.

    As for the nasals, I have a sense (for Devanagari in general) that English/foreign words are more likely to be represented with "incorrect"/non-homo-organic clusters like "न्ट" and "न्ड" (I've seen the latter frequently for English transcriptions).

  9. leoboiko said,

    March 16, 2014 @ 2:47 pm

    I can't distinguish either the English subtle vowels or the multiple sibilants/affricates of Mandarin etc. I mean, if I sit down with an IPA table and listen to recordings with attention I can pick them apart, but in fast live conversation it all collapses into the same phoneme.

  10. Rod Johnson said,

    March 17, 2014 @ 8:14 am

    What are "the English subtle vowels"?

  11. Will Thomas said,

    March 17, 2014 @ 9:37 am

    My wife was from New York state, and she pronounced 'pin' and 'pen' differently. I always found that odd. I was born and raised in the Southern US, and I pronounce them exactly the same way.

  12. Stitch said,

    March 17, 2014 @ 9:58 am

    Here in Milwaukee, I'm often amused to hear African Americans and native Southerners refer to a writing instrument an an "ink pen." Apparently just calling it a pen wouldn't narrow things down enough.

  13. MaryKaye said,

    March 17, 2014 @ 12:09 pm

    Rod Johnson asks: 'What are "the English subtle vowels"?'

    I'd assume this is something like the distinctions between Mary, merry, and marry, which I personally (Pacific Northwest US) don't have. I always wonder uneasily whether East Coast speakers feel I mispronounce my own name! Also pin/pen and caught/cot, where I have the first but not the second. I don't even have the vague sense, as I do with Mary and merry, that I'm pronouncing them differently inside my head–for me caught and cot are completely homophones and the spelling is just a nuisance.

  14. Will Thomas said,

    March 18, 2014 @ 4:57 am

    Stitch, in other words, not a ballpoint.

  15. Graeme said,

    March 18, 2014 @ 6:12 am

    Nb they also can supply you with a 'doormet'.

  16. Ellen K. said,

    March 18, 2014 @ 7:40 am

    I had not trouble realizing bad=bed. It's shits=sheets that I didn't get.

  17. Stitch said,

    March 18, 2014 @ 3:44 pm

    @Will Thomas:

    People always seem satisfied when I lend them a Bic.

  18. Steven J Fromm said,

    March 19, 2014 @ 7:37 am

    Sheet man, something got lost in translation. Yikes.

  19. Goofy said,

    March 20, 2014 @ 7:26 am

    But Nepali is like Hindi in that you don't need a virama at the end of words that end in a consonant, right?

  20. Richard Wein said,

    March 23, 2014 @ 10:51 am

    It must be a nightmare for learners of English trying to distinguish all the words that differ only by a vowel sound, especially as many depend on regional pronunciation. For amusement I once tried to find the most different-sounding words I could make with a vowel between the same pair of consonants. I came up with 16 for b-d (not counting homophones):

    bad, bared, bayed, bored, bard, bed, beard, bead, bid, bide, bird, bod, booed, bode, bowed, bud.

    In my pronunciation (basically modern RP) the r's in those words are not pronounced.

    To me, "bad" (not good) and "bad" (the past tense of "bid") seem a little different. The latter seems shorter. But I thought the difference too small to count them as two.

  21. Richard Wein said,

    March 23, 2014 @ 10:55 am

    P.S. I forgot one: bouyed. And I think there was another that escapes me now.

  22. Andrew Bay said,

    April 3, 2014 @ 11:15 am

    Ellen K,
    The "Bad Shits" could also be for the need to clean up unexpected diareah. Otherwise known as "not a fart".

RSS feed for comments on this post