Thursday, 15 October 2009

"Thrice Honoured Moon": The Mystery of the Nepalese Inscribed Khukuris

The khukuri is a traditional Nepalese knife, and is part of the equipment of Nepalese Gurkha soldiers:

In 2001 I began my collection of Nepalese khukuris with a British Army Service model from Himalayan Imports/BirGorkha Khukuri:

Since then I've collected a number of khukuris, both modern and antique--though this number has been sadly rather small due to my graduate student salary, I have had the opportunity to view a variety of beautiful examples of antique khukuris of my friends on various forums (e.g. Bladeforums, Sword Forum, IKRHS etc.). Some of the most (linguistically) interesting ones began to appear in 2002/3 when Atlanta Cutlery [AC] started selling old khukuris from a hoard they'd acquired from Nepal. A number of these khukuris have were found to have rather mysterious inscriptions in Nepali stamped on the blade spines.

Before turning to how these inscriptions were deciphered, do let's first consider the etymology of khukuri (because this is a linguistics/philology blog after all).

Note: I have never seen an etymology given for Nepali khukuri. Turner [1] in his Indo-Aryan dictionary surprisingly does not treat khukuri (this is particularly surprising since he served as a Gurkha officer). Thus the following etymology is not a rehearsal of any previous study.

The ultimate Proto-Indo-European root of Nepali khukuri (खुकुरी) must be PIE *kes- "to scratch" (see Pokorny [2]:1.585, Watkins [3]:41). It is from the zero-grade form of this root with -eu- extension (i.e. *kseu-) that derive both Greek ξυρόν (ksurόn) "razor" and Sanskrit kṣurá- "razor" (in the Ṛgveda), "sharp barb of an arrow" (in the Rāmāyaṇa), as well as Sanskrit kṣurī "knife, dagger".

From Sanskrit kṣura and kṣurī come a number of words in Modern Indo-Aryan languages for "knife, dagger" [1], though these are mainly ch- forms (rather than kh- forms, see below), and even late Sanskrit shows ch- forms like churī/chūrī "knife, dagger", also churikā- (in the Kathāsaritsāgara). Modern Indo-Aryan forms include: Hindi churā "dagger, razor", Nepali churā "razor", Punjabi churā "large knife", churī "small knife", Gujarati charo "large knife", Bengali churi "knife" (and Assamese suri "knife"), as well as Armenian Gypsy/Romani čhuri "knife" --- and, interestingly, also Sindhi churī "knife with a hooked blade": apparently denoting an object rather like a khukuri!... [Update [16.10.09]: Berkley at IKRHS provides some images of Sindhi knives, some of which are a bit khukuri-shaped, commenting "...All of which leaves the general impression of a large, curved knife, used for chopping and slicing, with a smaller side knife. Sound familiar?"] In any event, Sanskrit kṣurī looks like the most plausible source for Nepali khukuri.

The change from k to kh (in kṣurī > khukuri) reflects a general sound change in Indo-Aryan (cp. Sanskrit kṣétra "field" > Hindi khet), and indeed in Pali we find khura- "razor" (also Pali churikā "knife") and in Prakrit khura- "razor, knife". Modern Indo-Aryan kh- forms [1] include Nepali khuro "head of a spear; ferrule of a stick; pin at the top or bottom of a door", Sindhi khuryo "grass-scraper; tip of silver at the bottom of a scabbard", Assamese & Bengali khur "razor", Oriya khura "razor", Hindi khurā "iron nail used to fix ploughshares", and Sinhala karaya "razor".

Therefore we can derive a hypothetical pre-Nepali form *khurī. At some point this form must have undergone (partial) reduplication of the first syllable, thus *khukhurī. The loss of aspiration of the second kh appears to reflect some dissimilatory process similar to Graßmann's Law, thus khukurī (and this is indeed the Nepali spelling of the word [खुकुरी], however it is pronounced as khukuri, reflecting the more general loss of long:short distinctions for high vowels in spoken Nepali).

Back to the inscriptions on the Atlanta Cutlery [AC] hoard khukuris:

When I (and my fellows on Bladeforums) first saw these khukuris, we noted that the inscriptions had a recurrent pattern, consisting of three parts: (1) the formula श्री ३ चन्द्र [śrī 3 candra], (2) a three digit number [in the form x/xx], and (3) another word or words in Nepali. [Note: these three parts don't always occur in the same order.]

I initially translated the formula śrī 3 candra literally. Now Sanskrit śrī derives from the verb śri "to burn, to flame, to give off light", but śrī itself (as a noun) usually denotes "good luck, fortune, prosperity, auspiciousness", and Śrī is another name for the Goddess Lakshmi, who is the goddess of prosperity. However, śrī is also used as an honorific title; candra means "moon" in Sanskrit. On this basis, I originally translated the formula as "thrice-honoured moon":
[Note that the spelling candra varies quite a bit: from cadra (with no nasal) to caṁdra to candra to caṁndra (hyperform)]

However, perusing Matthews [4]:239, I found that śrī 3 was the title taken by the Rāṇā Prime Ministers of Nepal (and śrī 5 was the title taken taken by the King of Nepal). It then became clear that śrī 3 candra referred to (Śrī 3) Chandra Shamsher Jang Bahadur Rana, who was Prime Minister of Nepal from 27 June 1901 to 26 November 1929, who had his name inscribed thus (śrī 3 candra) on many weapons produced (and/or used?) during this period.

The next piece of the puzzle fell into place during a conversation I had with Jonathan R.S./Spiral(twista) on IKRHS (see here), when the (3) part of one of the inscriptions matched with name in list of Nepalese battalion names (the Bhawānī Dāl battalion):

While some of the inscriptions have the battalion names in full (like the above Bhawānī Dāl), others use only a part of the battalion name (singh for the Singh Nath battalion), or two letter abbreviations (like na.go. for Naya Gorakh battalion, as shown below)--which is why the identification of this part of the inscription was initially difficult.

Benjamin Judkins/Finnarm unraveled the final bit of the mystery of the Nepalese inscriptions (see here for his original post). He discovered in an appendix to Guns of the Gurkhas by John Walter [5] an analysis of the various types of inscriptions found on Nepalese military rifles. Judkins remarks:
The method of marking military material in Nepal was apparently very similar to the system used in the UK. After 1856 the British army went to a system in which first a unit designation was given, followed by a weapon number. Previously a soldier’s identification number had been used. So for instance, Walter states that "13 L D 153" should be read 13th Light Dragoons, 153rd weapon.

The same approach seems to have been adopted in Nepal and many rifles have been positively identified using these principals. It has been used on the kukris as well. The appendix lists the names of about 150 Nepalese battalions, and it gives two letter abbreviations for some of these names that have been employed on rifles.
So the (2) part of the khukuri inscription denotes the sub-unit and weapon number [i.e x/yz = subunit x / weapon no. yz].

Thus these AC khukuri inscriptions follow the pattern:

(1) śrī 3 candra [= Prime Minister Chandra Shamsher Jang Bahadur Rana (r.1901-1929)];

(2) Battalion Name or Abbreviation;

(3) Sub-Unit No./Weapon No. [x/yz].

(Many thanks to my fellow forumites at Bladeforums, IRKHS, and elsewhere for many conversations on these inscriptions over the years; particularly Jonathan R.S./Spiral(twista), Berkley, Rod Allen, & Benjamin Judkins/Finnarm.)

[Update: I've started a (multi-author) repository for khukuri inscriptions here: Khukuri Lipi ]

[1]Turner, Sir Ralph Lilley. 1966-1985. A comparative dictionary of the Indo-Aryan languages. 4 vols. London: Oxford University Press. [reprinted, Delhi: Motilal Banarsidass, 1999.]
[2]Pokorny, Julius. 1958. Indogermanisches etymologisches Wörterbuch. Bern and München: Francke Verlag.
[3]Watkins, Calvert. 2000. The American Heritage dictionary of Indo-European roots. Boston: Houghton Mifflin, 2nd edn.
[4]Matthews, David. 1998. A course in Nepali. London: Curzon.
[5]Walter, John. 2005. Guns of the Gurkhas. Norwich, Norfolk: Tharston Press.


  1. I don’t know of any hard evidence for it, but my first thought was to wonder whether the Malay keris, kris — which has been borrowed into English in the form, kris, “a wavy, double-edged dagger” — is related to this Indo-European root. Malay itself, of course, is Austronesian, not Indo-European, but might the root have travelled to southeast Asia? What do you think?

  2. Jason - it seems highly doubtful. But I don't know enough about Austronesian to rule it out. If the word was borrowed into Javanese I would think Indo-Aryan would be the only IE language which could be the donor. So I would expect a u vowel rather than e. But who knows. I would be tempted to look for a Austronesian etymology first though (if I knew anything about Austronesian that it).

    Though I always think of kris/keris as having wavy blades, the middle one in this photo from the kris Wikipedia page looks rather khukuri-ish.

    I don't own any krises, though I do have a couple of handsome Javanese goloks.

  3. Note that Platts, not necessarily convincingly, puts it in with words meaning 'bent' or 'curved'.

  4. Platts is confusing his stops here. He cites कुकड़ी kukṛī to refer to the Nepalese knife, but in Nepali it's खुकुरी khukurī, and even in Hindi the form is कुकरी kukrī or the like, with r, NOT the retroflex .

  5. 1. Himalaya: Malaya 2. Black Cap : Songkok 3.Physical appearnce: I went to Kathmandhu- Nepal in 1982(18yrs) everyone greeting me in Nepal languge 4.Rice consume 5. Why the Malay wear black cap in tropical climate(Thai not,Laotian not,Burmese not, Cambodia not): hat or cap, normally wear in cold weather or hot weather.If there any connection with Nepalese? This is only an instinct not academist.