I was disappointed to hear that the students of Sanskrit in the Eastern Classics program at St. John’s aren’t working much with Devanagari this year, at the discretion of their tutor (we used it a lot last year, with a different one). Apart from the practical consideration that a large number of Sanskrit texts and resources assume knowledge of the script, I found myself explaining to a current student why I think Devanagari is totally awesome. To wit:
In our (the Roman) alphabet, the order of the letters roughly follows that of the Greek alphabet, which itself follows that of the Semitic alphabets. Thus, a-b-c from alpha-beta-gamma from aleph-bet-gimel. But in all of them, this order is to all appearances arbitrary. Not so in India.
All of the consonants in Devanagari, as in most (or is it all?) of the other Indian scripts, are divided into five groups called sthānas—which literally means “standing,” or “position”—according to the place of articulation in the mouth, and the sthānas are themselves arranged according to distance from the throat.
Thus, the first group is articulated nearest the throat: the ka-sthāna.
The second is forward slightly: the ca-sthāna (“ca” being pronounced as what in English we would write “cha”).
The third is not necessarily further from the throat, but uses a further-forward part of the tongue: the ṭa-sthāna. (This is the “retroflex” sthāna, pronounced with the tip of the tongue straight up in the palate. It’s the stereotypical feature of Indian speech mocked in fake Indian accents.)
The fourth is at the teeth: the ta-sthāna.
The fifth is at the lips: the pa-sthāna.
Now, within each sthāna the letters are arranged according to prayatna (literally, “effort”). These begin with the unvoiced, unaspirated consonant, so the first letter in the pa-sthāna is “pa.” Next is the unvoiced but aspirated form, so the second letter in the pa-sthāna is “pha” (pronounced the same as “pa,” but with more breath). Next is the voiced and unaspirated: “ba.” Then the voiced and aspirated: “bha.” Then the nasal: “ma.” This pattern is repeated for all of the sthānas.
Some of these distinctions can be hard for English-speakers to hear and produce, since we pay less deliberate attention to them. For instance, we do have aspirated and unaspirated consonants, but we don’t distinguish them in writing, and most of us are usually unaware of the difference though we hear people who get them wrong as somehow vaguely foreign-sounding. The “p” in “pot” is aspirated; the “p” in “spot” is unaspirated. If you’re a native speaker, you’ll put more breath into the former, though you might not realize it, and you’ll think people sound non-native when they get it wrong.
It also looks like we don’t have so many nasals in English; but we have more than you might think, and just use “n” as a generic nasal-marker. For instance, people sometimes refer to “dropping” the letter “g” from words ending in “ing”; but nothing is being dropped, only replaced. “Talking” ends in a nasal of the ka-sthāna (a “velar nasal”); “talkin” ends in a nasal of of the ta-sthāna (a “dental nasal”). The place of articulation of the sound has changed, but neither form has more sounds than the other. You might say we’re using “ng” to indicate a single sound, which would be indicated with a single letter in Devanagari.
So the consonants in Devanagari are arranged like so:
k — kh — g — gh — ṅ
c — ch — j — jh — ñ
ṭ — ṭh — ḍ — ḍh — ṇ
t — th — d — dh — n
p — ph — b — bh — m
Each row is a different place in your mouth; and then within each row, all of the letters are pronounced with your mouth in the same position, but changing the manner of enunciation. So with your mouth in the position for the “k,” add breath to get “kh,” add voice to get “g,” add breath and voice to get “gh,” and make it nasal to get “ṅ” (which is the “ng” in our “-ing” words).
(The vowels and other letters come before the consonants, and are similarly—though perhaps less obviously—arranged according to manner of enunciation.)
One consequence of this ordering is that, whereas when looking up a word in an English dictionary I find myself singing the alphabet song in my head, when looking up words in Sanskrit I find myself moving my tongue through the different positions in my mouth.
When I first learned this, it seemed like something Tolkien would have used for Elvish. And it’s the main reason I think Devanagari is awesome.
But this is all a fact of the language, not the script. The order of the letters in the alphabet remains the same whether it is written in Devanagari, Tamil Grantha, Newari, etc., etc., etc. Roman transliteration is no different.
Does ‘क’ somehow mean unvoiced, unaspirated velar stop when ‘ka’ does not?
Sanskrit is wonderfully script-independent. To me it’s one of its charms.
The order of the characters is a fact of the script— a fact that is shared, as I suggested in the post, with many other related scripts. (I assume it is common to all systems descended from the Brahmi script.) It’s an incidental historical fact that of this family of scripts, Devanagari became most associated with Sanskrit in the West, and therefore the most widely known outside of India. Therefore this particular thing that I think is awesome about Devanagari is also awesome about a bunch of other scripts that I don’t know.
If you mean that this ordering is a consequence of Sanskrit itself, or of the people who used it, I think there might be a case to make that it’s less fundamentally a fact of the script than a consequence of the culture, and especially with regard to the position of the Vedas. But it’s not demanded by or necessary for the language. And it’s shared by many other languages besides Sanskrit that use the same or related scripts.
All human languages are script-independent. That’s why so many scripts are used across multiple, sometimes even unrelated languages; and why so many languages have been written in multiple scripts. It is an interesting fact that, for a literary language, Sanskrit has seemed unusually unattached to one particular script, and maybe that’s what you’re suggesting. It’s a fact that might reflect the unusual regard for the sounds of the language, over and above the look of it on the page, which regard also resulted in the careful structure of the native scripts.
Devanagari only exists in the order in which it exists because of the underlying structure of the Sanskrit language, the order of which is pre-Paninian (though Panini uses the ordering of the Siva Sutras, for economy of formulation) and therefore pre-writing in South Asia.
Once writing happens the orality of Sanskrit is, in my opinion, a ideologically-driven fiction, as evidenced by the sections of bandha forms of citra in the earliest kavya, as well as, less concretely, the obsessive nature of the protestations to the “fact” of orality.
What is this “underlying structure” of the language which effects the ordering of the characters in the script? It’s not clear to me what you could mean.
My understanding—which may very well be mistaken—is that something like this ordering is used for most or all of the family of brahmi scripts, most or all of which are used for writing languages that are not Sanskrit, and many of which are used for languages that are not even Indo-European. Unless these scripts all inherited their ordering from Devanagari or a precursor to Devanagari which itself was formulated specifically to write Sanskrit (which I do not believe to be the case), it would seem that this ordering is not due to any structural demands of Sanskrit. Perhaps it is due to a cultural feature of the people who inherited the Vedas, but I don’t see how their language (or any one of their languages) is by its nature more or less given than any other to be written in such a phonologically ordered script.
Also, though it is an aside, I am given to understand that there is considerable uncertainty about dating Pāṇini, and controversy over whether he is “pre-writing in South Asia.”
If your point is just that this particular phonological (or phonemic) arrangement predates Devanagari, then I misunderstood you, and I have no objection to that. I did suggest in my post that this ordering was not unique to the one script.
But it is not clear to me how a structure of the Sanskrit language itself is responsible for this, or any other, ordering of sounds.
Yes, the phonetic structure. Which would be part of the structure of the language. I think we mainly have a problem of terminology. For instance, when you say:
“So the consonants in Devanagari are arranged like so:
k — kh — g — gh — ṅ
c — ch — j — jh — ñ
ṭ — ṭh — ḍ — ḍh — ṇ
t — th — d — dh — n
p — ph — b — bh — m”
This is wrong. The consonants of Sanskrit written in an extended version of the Roman script is as presented. The consonants of the Sanskrit language in the Devanagari script are like so:
क – ख – ग – घ – ङ
च – छ – ज – झ – ञ
ट – ठ – ड – ढ – ण
त – थ – द – ध – न
प – फ – ब – भ – म
But the ordering of the phonemes is the same and the ordering predates any script, Devanagari or otherwise. The Devanagari script itself evolved out of Brahmi script, which was developed for a series of Prakrit languages for the Emporer Ashok. The Prakrit languages, in turn, had evolved out of early Sanskrit, most likely in conversation with Dravidian precursors.
I think you may be right that we mainly have a problem of terminology. I think, for one thing, that you are meaning “the Sanskrit language” in a more expansive way than I am.
I take it that the language itself presents a variety of phonemes, but makes no structural demands on a traditional system of ordering— which is why so many (presumably the vast majority) of historical writing systems have arranged their characters more or less haphazardly. In fact a single standardized ordering is not theoretically necessary even for literacy: we could both have learned every letter of the Roman alphabet and become equally literate without having used the same ordering of letters to do so. Certainly no “underlying structure” of English (or of Latin, had we been ancient Romans) would lead us both to arrive at the same ordering independently.
You are right, of course, that in this post I only presented the characters of an extended Roman set, and not the Devanagari equivalents, since most of my audience would not recognize the latter. But this transliteration scheme was not developed as a natural consequence of the structure of Sanskrit, nor would it be likely to have arrived at this arrangement from mere consideration of the language independently of any other script: it was developed specifically to have one-to-one correspondence with Devanagari and related Indic scripts, and inherited its order from them, not from a deeper structure in the languages they are used to write. And now that the ordering is learned in tandem with the script—wherever the ordering originally came from—it is shared with speakers and writers of other languages that presumably have quite different “underlying structures,” including, for instance, Nepal Bhasa, whose “structures” are not even Indo-European, and which would nevertheless appear not to interfere with the ordering, or even to suggest a different one more natural for its own structures.
My main point here was that the ordering of characters in a writing system—though it may certainly be a feature of the interests and preoccupations of the people who formulate it—is most likely not a feature intrinsic to the language itself. In a different history, the languages of India might have been written in arbitrarily ordered scripts, and the languages of the Mediterranean in ones with such orderly phonemic arrangement; I am suggesting that India deserves credit for its totally awesome invention, but that it most likely is not a simple consequence of the structure of the language, which would just as easily have been accommodated by any arbitrary ordering.
I believe it remains contentious (and it might finally be unanswerable) whether the Prakrits evolved from early Sanskrit, or if early Sanskrit was merely one of many related dialects that became a prestige variety.
I have no contention with the (it seems to me quite probable) claim that the phonemic ordering I praise here predates Devanagari. I did make explicit in my original post that this ordering is shared by other, related scripts (no one of which I should consider less awesome than Devanagari, though I don’t know them), and I presume they inherited it from a common source. However, I have no way of evaluating the claim that this ordering predates any script. It seems rather improbable to me, but many claims about ancient India seem rather improbable to me, and some of them are true, so this one might be as well. I’d be interested to see evidence one way or the other. But I would continue to think that it is not properly a feature of the language—since it works just as well for other languages, and other orderings work just as well for this one—even if it is a very ancient and even pre-literate feature of the culture that grew up around the language.