Buddhism and languages

« previous post | next post »

Whether you are familiar with Chinese characters or not, try to guess the meaning of the calligraphy on the front of this forthcoming book (the answer is at the very end of this post):

If you are a serious Buddhologist, you end up having to study a lot of difficult languages.  When I enrolled in a program of Buddhist Studies at the University of Washington in the fall of 1967, I went to my advisor, Edward Conze, to ask him which courses I should take.  He looked at me quizzically and said, "Are you serious?"

"Yes, sir," I replied, "I'm serious."

Looking me straight in the eye, he intoned:  "Sanskrit, Tibetan, Chinese, and Japanese".  That first semester almost killed me.  To put it mildly, Buddhist Studies is very language intensive.

Buddhism was born in India and its earliest scriptures were written in Pali.  Soon there were texts in Prakrit, Buddhist Hybrid Sanskrit, Chinese, Tibetan, Japanese, Korean, Sogdian, Khotanese, Uyghur, Khitan, Thai, Cambodian, Mongolian, Manchu, and numerous other languages.

The book whose cover is pictured above is to honor the memory of Antonino Forte (1940–2006), an outstanding Italian Buddhist scholar who was Director of the Italian School of Oriental (later East Asian) Studies in Kyoto for many years. At the institute, Professor Forte welcomed scholars from around the world to use the magnificent research library he had built up, and he generously shared his vast erudition on countless subjects related to East Asian medieval Buddhist history.  Those of us who did research at Nino's institute would hear people speaking languages from around the world — Italian, French, German, English, Japanese, Chinese, Dutch, Bengali, and many others — on a daily basis.  Working in that multilingual environment was a constant reminder of the linguistic richness inherent in the global community of Buddhologists.

Among the more curious linguistic facts about the transmission of Buddhism from India to China is that the earliest translations of Buddhist texts into Chinese were done neither by Indians nor by Chinese, but by Iranian speakers.  One of Forte's specialties was precisely on the Iranian individuals who were involved in the cultural transfer of Buddhism from India to China.

For me personally, the most important linguistic impact of Buddhism was its legitimization of the written vernacular in China, for which see:

Victor H. Mair, "Buddhism and the Rise of the Written Vernacular in East Asia:  The Making of National Languages", Journal of Asian Studies, 53.3 (August, 1994), 707-751.

In terms of vocabulary, upwards of thirty thousand Indic words entered the Sinitic lexicon with Buddhism beginning in the first century A.D., many of which are still spoken in the common language of today, including those for "convenience", "instant", and "meditation".

Buddhism was also deeply involved in the development of linguistic science, especially phonology, in China.

Buddhist Transformations and Interactions will be published from Cambria Press on March 16, 2017 and will be introduced to the scholarly community at the Association for Asian Studies annual meeting in Toronto on March 18.

For those who are curious, the exquisite calligraphy at the top of this post is "Spring" (春 — Japanese haru; MSM chūn) by Lilla Maria Moramarco, wife of the honoree of the volume.


  1. shubert said,

    March 1, 2017 @ 9:46 am

    Congrats to Prof. V.M.

  2. Neil Dolinger said,

    March 1, 2017 @ 11:13 am

    Do calligraphers typically expect others to recognize the standard form within the calligraphic form? If I try I guess I could see 春 within this image, though I don't see the fourth stroke (I may be wrong on the count, it's the stroke down and to the left, after the three horizontal strokes).

    My first guess was that this was something related to 戈 (ge / dagger axe), though spring is certainly a more appropriate image for this book cover!

  3. william holmes said,

    March 1, 2017 @ 12:18 pm

    Prof. M. notes the role of Persian translators in those early centuries. If –per a 6th c. source cited by Hu Shih — Bodidharma (Daruma) was Persian, then Persians would have been pivotal in both the original textual transmission to East Asia and in subsequent de-texting (as it were). (In his 菩提達摩考, Hu Shih also quotes a 7th c. source identifying Bodidharma as an Indian Brahman).
    I second the congrats to Prof. M. on the new book.

  4. liuyao said,

    March 1, 2017 @ 2:46 pm

    More recognizable would be the seal above VHM's name. But what does Li Na 李娜, a very common feminine name (one tennis player, one singer-turned-Buddhist nun, and others) has to do with this book?

  5. David said,

    March 1, 2017 @ 3:32 pm

    Interesting that Conze didn't list Pali as a language that you should study. Did he think that if you master Sanskrit, you sort of 'get Pali for free'? Or did he just not think Theravada was worth the time?

  6. Neil Dolinger said,

    March 1, 2017 @ 4:14 pm

    "But what does Li Na 李娜, a very common feminine name (one tennis player, one singer-turned-Buddhist nun, and others) has to do with this book?"

    Perhaps it is the calligrapher's (Lilla Maria Moramarco's) nom de plume (笔名?)

  7. Chris C. said,

    March 1, 2017 @ 4:41 pm

    "Li Na" seems a reasonable sinitization of "Lilla".

    Is all hanzi/kanji calligraphy so opaque? As a non-reader of either, but as a sumo fan, I often struggle to recognize the characters in a sumo wrestler's tegata.

  8. Victor Mair said,

    March 1, 2017 @ 6:23 pm


    We could only take 4 courses per semester, and anyway I think that Pali had a Sanskrit prerequisite. Conze being Conze, he never would have overlooked Pali.

  9. Jonathan Silk said,

    March 2, 2017 @ 2:50 am

    "Buddhism was born in India and its earliest scriptures were written in Pali." Tut tut, Victor, you know the second part of this is not correct, or at the very least it is highly misleading. Pali is demonstrably an 'artificial language' which, however old the material it may preserve, is not itself the oldest language of the Buddhist traditions. Moreover, the earliest scriptures were of course not written at all, but anyway, they were not composed in Pali.
    The publication of anything more of Nino's oeuvre is a cause for great celebration, and for this, thank you! I look forward very much to seeing the book.

  10. Victor Mair said,

    March 2, 2017 @ 8:35 am

    From Ronald Davidson:

    I don’t know what Jonathan has in mind, but the oldest surviving documents are in Gandhari.

    Even then, many would argue (and all Pali specialists would agree) that the oldest surviving written materials are the earliest Pali documents, especially the Suttanipāta and related verses in the Dhammapada and the Udānavarga, perhaps some of the Jātaka verses. I suspect Jonathan was indicating that Pali is not the “language of the Buddha” and that oral transmission of Buddhist literature happened prior to the advent of Pali per se.

  11. Victor Mair said,

    March 2, 2017 @ 8:38 am

    From Michael Witzel:

    Pali is a western North Indian LITERARY language, not the orginal language of the Buddha who lived in eastern North India (in N. Bihar/S. Nepal) that was more like the Magadhi of Patna, of Asoka’s inscriptions

    We can see that in the occasional verses transmitted in Pali that have eastern forms. Nothing new here: see Lüders, Die Sprache des Buddhistischen Urkanons, about 100 yeras ago. All of this has been substantiated, especially in the work of O.von Hinüber.

    We do not have the Buddha’s original speeches.

    But Pali clearly is the oldest version of his teachings (some before, some after Asoka) . Some fragments exist in Pali-Like Middle Indic that are younger than Pali, and of course now in Gandhari, plus the usual suspects: Sanskrit fragments (Buddhist Hybrid Skt. from Xinjiang) and much more Skt texts from Nepal, Chinese and Tibetan translations….

  12. Victor Mair said,

    March 2, 2017 @ 8:41 am

    From Jan Nattier:

    The consensus among specialists, as far as I am aware, is that the Buddha probably spoke something called "Old Māgadhī" (one of many Prakrit/Middle Indic vernaculars spoken on the subcontinent at his time). Nothing was recorded in writing (or even passed down in memorized form, so far as we know) in this language, though some scholars have pointed to individual lexical items in a few Buddhist texts as likely to be "Magadhisms" (like the bits of Aramaic in the Greek New Testament I suppose).

    Pāli was one of a number of languages in which Buddhist sacred texts were memorized and transmitted, the oddity here being that no one ever spoke Pāli as his or her native language; Jonathan is right, it is an "artificial Prakrit" (there is a classic article by von Hinüber on this, I'm sure I can dig up the reference if you don't find it easily), deliberately put together from more than one Prakrit (= vernacular) to serve as the language of sacred texts in one particular ordination lineage (the one that developed into the so-called Theravāda lineages of today).

    Many other "Buddhist languages" (that is, languages used by various ordination lineages to transmit sacred texts) certainly existed, though we have little or no evidence of them. And as far as written textual remains go, Gāndhārī stands out as having by far the earliest surviving Buddhist manuscripts (as old as the 1st c. BCE).

    So there is actually a quite clearcut answer to the question you asked: the winner is Gāndhārī, hands down. (and this is interesting also as an increasing number of early Chinese Buddhist translations are being shown to contain evidence of having been translated from Gāndhārī originals) Surviving Pāli manuscripts are many centuries later, and most other surviving Buddhist manuscripts already date from a period where the spoken vernaculars were being supplanted by BHS (and eventually, more classical Sanskrit). So the Gāndhārī texts provide a precious window into the "vernacular age" that pre-dated the age in which Buddhists moved away from using their own spoken tongues and began to use Pāli and Sanskrit as "church languages" (I have an ancient article on this in case that topic is of interest).

  13. Jonathan Silk said,

    March 2, 2017 @ 9:15 am

    At the risk of diving further into waters that probably only attract specialists, I should note that there is evidence of some Gandhari texts having been translated from some other, well, language? dialect? Let's not go there… I apologize if, as is suggested by some of the comments like Ron's, I did not state my idea quite clearly enough, but others have, so, nothing further to add except to repeat that I'm delighted that Victor introduced this new book of Nino's work!

  14. David said,

    March 2, 2017 @ 11:32 am

    I thought I remember reading somewhere that based on its dialectal features, as compared to other (pre-)Prakrits and Ashokan inscriptions, Pali probably originated on the central west coastal region of India, perhaps around where Gujarat is now. I think it had to have originally based on a real vernacular at SOME point, tho it is true that no way is it what the Buddha spoke, and that there's virtually nothing from the early period in Pali that isn't related to Buddhism.

  15. DWalker07 said,

    March 2, 2017 @ 11:40 am

    I don't know Chinese characters at all, but that calligraphy on the cover looks very much like … a male member.

  16. Victor Mair said,

    March 2, 2017 @ 3:48 pm

    From Justin McDaniel:

    I assume he means Magadhi. This is a controversial subject. Also, is there any such thing as a non-artificial language?

  17. Victor Mair said,

    March 2, 2017 @ 4:21 pm

    From Vesna Wallace via David White:

    Orally, texts were transmitted in various local languages, but I have never heard of scriptures being preserved in any earlier language than Pali.

  18. Victor Mair said,

    March 2, 2017 @ 4:27 pm

    From Asko Parpola:

    The oldest language of the Buddhist tradition, the language used by the Buddha himself, is an eastern Prakrit, Old Magadhi, of which traces are found in Pali, which came into being when Buddhism moved westwards. Asoka'sinscriptions from around 250 BCE are the oldest preserved texts recording Buddhist traditions. The Buddhist scriptures remained oral until the Pali canon was written down around 80 BCE.

  19. Victor Mair said,

    March 2, 2017 @ 7:09 pm

    From Dan Boucher:

    Jonathan is indeed right that Pali is an artificial language (von Hinüber has a nice article on this), showing as it does influence from both Western and Eastern dialects, as well as Sanskrit forms that became frozen in the language in contrast to typical Middle Indic morphology.

    Given this artificiality, Pali cannot be the oldest language for Buddhism even if the material it preserves represents the oldest redaction of our texts (which may or may not be true). The Buddha certainly spoke an eastern dialect (something akin to Māgadhī), but we don’t have any of those texts (though some Māgadhī forms show up in Pali and in the Aśokan edicts). The oldest written texts we have are certainly the new discoveries of Gāndhārī manuscripts, and their language, while not consistent between texts, is distinctive from other MIAs. But regardless of the earliest written texts, ALL materials as we have them are translations from earlier recensions we no longer have. So no collection can make special claims to being the earliest without some demonstration of why one recension must have preceded another. And such a demonstration is not likely to hold for all texts within any given language. Thus it is quite possible that someone could show (and probably has) that some versions of āgama texts in Chinese could be recensionally earlier than some versions of Pali texts. Lots of new studies of the āgamas coming out of Dharma Drum by Analayo and others that look at exactly these kinds of questions.

  20. Victor Mair said,

    March 2, 2017 @ 7:12 pm

    From Robert Goldman:

    I would say that, regardless of its “artificiality” Pali would be the language of the oldest written Buddhist texts.

  21. Victor Mair said,

    March 2, 2017 @ 10:03 pm

    From Alastair Gornall:

    The oldest Buddhist texts we possess are in Gāndhārī, though these are largely fragmentary. Gāndhārī sūtras often have parallels with equivalent passages in Pali suttas. We can say then that while Pali texts do contain ancient material the language of these texts underwent certain phonetic changes at the hands of later redactors. The earliest Pali manuscript we have dates from the eighth or ninth century and was discovered in Kathmandu (see v. Hinuber, The Oldest Pali Manuscript, 1991). I don’t really buy into the idea of Pali as an ‘artificial’ language though, as what literary language isn’t to some extent?

  22. Victor Mair said,

    March 2, 2017 @ 11:12 pm

    From Whitney Cox:

    I am not a specialist in this, and the sources I can recommend are just the 'classics': there might have been some revisionist thinking about this recently, but I haven't heard of it. The two big waypoints would be the edited volume by Heinz Bechert, Die Spräche der altesten buddhischen Überlieferung/ The Language of the Earliest Buddhist Traditions (Göttingen, 1980; U of C library catalogue entry here), and on this and more generally, Oskar von Hinüber, Das ältere Mittelindisch im Überblick (Vienna, 1986, catalog). My colleague Steve Collins also has an article on this theme, "On the very idea of the Pāli Canon" (recently reprinted in this volume, available on Google Books).

  23. Victor Mair said,

    March 2, 2017 @ 11:17 pm

    From a graduate student in Buddhist Studies:

    I've heard similar statements about Pāli's being an artificial language, and I've also heard people question the dating of the Pāli canon. Nevertheless, I'm pretty sure the consensus opinion is that the Pāli texts represent the oldest strata of written Buddhist material extant. (Though I think – and I may be wrong – the Pāli material should predate writing in India given we don't have much evidence for written material pre-dating Aśoka's columns, etc. So I would guess, like Sanskrit, the language was both "artificial" and spoken before its being written down.)

    Perhaps more accurately: Pāli is the language of the earliest codified and transmitted Buddhist texts I know of – though I'm not sure whether they were the first ones actually written down in ms form.

  24. Victor Mair said,

    March 3, 2017 @ 12:47 am

    From Frederick Smith:

    Pali is the oldest language of the Buddhist texts, but as Jonathan Silk said here, it’s not the oldest language of the Buddhist traditions. It appears from all manner of evidence that there were other Northeastern Indian dialects that preceded Pali, one of which the Buddha likely spoke. Various scholars have written about this. It’s likely that the Pali suttas were composed at least a century and a half after the passing of the Buddha, but that their transmission at some point switched into Pali. I think part of the problem is that there’s not a whole lot of evidence that Pali was a spoken language…. I believe that Oscar von Hinuber, Thomas Oberlies, Richard Salomon, and a few others have written on this.

  25. Victor Mair said,

    March 3, 2017 @ 10:04 am

    From George Cardona:

    A major study, published long ago by Heinrich Lueders, argued that the original canon was in an eastern dialect, which accords with the tradition that Māgadhī was the language of the original teachings and with the fact that Buddhism originated in the Bihar area. When people speak of Pali as artificial, one can equally well say it is a mixed dialect used in the transmission of the texts, although I now forget the details of the arguments put forth by von Hinueber, the major proponent of this view.

  26. Victor Mair said,

    March 3, 2017 @ 7:35 pm

    From Eviatar Shulman:

    Following up on Alastair's response, there is a question of what you mean by 'written'. If you are speaking about manuscripts, then Gandhari (and probably other languages too) give us earlier written materials, which are in many senses in line with the Pali.

    But then, 'written' can also refer to the date of writing, and then what we usually go by is the faint account in the Mahavamsa about the writing down of the texts at Alu-vihara toward the end of the first century BC. If texts were written down in other languages earlier, or if this account is reliable, we simply don't know. But that's the earliest we know of, simply because that's what is under the textual spotlight.

    If 'written', refers to 'composed', then that's a more interesting question. I cannot recall there being any convincing scholarly discussion regarding the language in which the earliest texts were composed. Although I too do not see Pali as 'artificial', I would also think that many of the materials we find in "Pali" were adaptations from languages that were close to Pali but more localized ("Magadhi", for example).

    But then, too, still in the 'composed' zone, I would also not find difficulty taking the earliest surviving texts in Pali (mainly the Atthakavagga and Parayanavagga, which are probably predate Asoka) as relatively original compositions in that language. Much of this is indeed "artificial", in the sense that it comes from different places and stages of the development of the language, but is nonetheless "Pali."

    I don't know if this helps (and do pardon the lack of diacritics), but it's what I've got to share.

  27. Victor Mair said,

    March 3, 2017 @ 7:39 pm

    From Andrew Ollett:

    There was a book on this subject edited by Heinz Bechert:

    Die Sprache der ältesten buddhistischen Überlieferung / The Language of the Earliest Buddhist Tradition

  28. Victor Mair said,

    March 3, 2017 @ 7:55 pm

    From Stephanie Jamison:

    The short answer — assuming you mean "earliest written … preserved in manuscripts [vel sim.] of that same period" — are the scrolls in Gāndhārī Prakrit from the northwest, dated to 1st c. AD. The scrolls were discovered fairly recently and are being worked on by Rich Salomon (UWash) in a joint project of UW and the British Library. Here's the website of the Early Buddhist Manuscripts Project.



    Early Buddhist Manuscripts Project

    The Early Buddhist Manuscripts Project. The Early Buddhist Manuscripts Project was founded at the University of Washington in September 1996 to promote the …


    Of course, as we both know, that's only one possible answer to the "earliest texts" question, and I'm not sure whether Jonathan, who doesn't use "written" is his formulation, is thinking of Gāndhārī or something else. He's not a linguist by training and so "oldest language" is a phrase he can use, though it doesn't mean anything to a linguist without a qualifier like "attested."

    By the way Pali is an "artificial language" — this is something of a buzz phrase in the sec.lit. about Pali and Middle Indic these days — but "artificial" in the same sense that many literary languages are, incorporating elements from different dialects and chronological layers that may not have occurred together in nature (or the pristine nature that linguists like to imagine). It is not an artificial language in the more linguistic sense, of having been consciously created and designed ex nihilo like Esperanto.

  29. Victor Mair said,

    March 6, 2017 @ 8:43 pm

    From Luther Obrock:

    The issue of the oldest language of Buddhism is difficult, and further complicated by the fact that everyone tacitly assumes that Pāḷī is the language that the Buddha spoke. It seems that the Pāḷī canon does indeed preserve early Eastern dialect of Middle Indic, however, this language has been redacted and standardized over time. The oldest *written* evidence that we have is found in the Gandharan mss, which are written in a northwestern dialect of middle Indic, and which also predate the standardization of the Pāḷī canon.

  30. Victor Mair said,

    March 7, 2017 @ 3:05 pm

    From Deven Patel:

    My understanding is that Pali is a literary Prakrit based probably on an ancient Prakrit known as Magadhi, a language of the Bihar region. It is very much possible that the Buddha spoke some form of this language. As the Buddhist "scriptures" don't emerge until several centuries after the Buddha, perhaps we cannot say that Pali is the earliest language of Buddhism. I think others have addressed the issue of earliest "written" language.

    From Oskar von Hinüber:

    Here is my wise answer: Nobody knows. If Dipavamsa / Mahavamsa are correct, Pali was written in Ceylon during the 1st century BC. If C14 is correct, Gandhari may even be older. But why should they be the earliest?

    Sorry for my (and the perhaps general) ignorance

RSS feed for comments on this post