Showing posts with label Japan. Show all posts
Showing posts with label Japan. Show all posts

24 November 2017

Japanese Reception of the Chinese Origins Thesis

As I prepare material for my book on the Heart Sutra, I have been collating published responses to Jan Nattier's thesis that the text was composed in Chinese and (back)translated into Sanskrit (Nattier 1992). I suggested in a previous essay that the reception of Nattier's thesis in Japan has been and remains decidedly anti. New evidence of this has emerged in the form of an article by Ishii Kōsei (2015), translated by his English-speaking former student Dr Jeffrey Kotyk

Unfortunately, much of the research done in Japan is only ever published in Japanese and is thus inaccessible to the majority of Buddhist Studies researchers in the West. The linguistic burden is high in our field. I have varying levels of skill in Pāḷi, Sanskrit, and Medieval-Chinese, but adding modern Japanese just to get access to secondary literature is not feasible. A review article of the Japanese reception of Nattier's article by some qualified scholar is a desideratum, but since Prajñāpāramitā is a tiny niche in Buddhist Studies, it is unlikely ever to happen. 

Ishii is apparently writing in a milieu in which there have already been well-received attacks on Nattier's thesis of a kind that we have not seen in English. He cites publications by Fukui Fuminasa and Harada Wasō, but these apparently focus on the conjecture that Xuanzang might have been responsible for making the Sanskrit translation from Chinese. The conflation of the Chinese origins thesis with the Xuanzang as translator thesis is unhelpful. Nattier leaves open the possibility but, in the end, does not commit to Xuanzang being the translator. On the other hand, the evidence for Chinese origins is very strong. Ishii seems to think that it is because we Western scholars of Buddhist Studies are "not specialists in this respect" that we have fallen for Nattier's thesis, rather than the strength of her arguments.

Ishii thus see his article as contributing some details to an existing (Japanese)  consensus in the face of a general credulity and ignorance in the West. Without access to that consensus, we are forced to take his article on face value, which I'm sure does not do it justice. Be that as it may, I will briefly outline the main points of Ishii's article and then review his methods and conclusions. I may say that my own published research has touched on many of the issues that Ishii has raised and I am thus in a relatively unique position to comment. I am very much a specialist in this respect (see my list of publications).


A Precis of Ishii (2015)

Ishii begins by referencing Nattier's 1992 article with a focus on the idea that Xuanzang might have been involved in editing and translating it from Chinese to Sanskrit. The bulk of the article deals with the opening sentence of the Heart Sutra and with Nattier's translation of it, which Ishii suggests follows the Chinese text, largely on the basis that Nattier omits a word-for-word translation of svabhāva  (1992: 155). 

While Nattier is explicitly translating from a modified version of Conze's critical edition, Ishii refers only to the diplomatic edition based on several hand-copies of the Hōryūji manuscript, produced by Müller in 1884 (though he refers to this as a "critical edition", it is clearly not). In order to attempt to refute Nattier, Ishii launches into a lengthy exposition showing that the word svabhāva is present in the Sanskrit text, but absent in the Chinese, and that the passage overall has given translators some difficulty. He tries to establish a case for the word svabhāva being dropped by a Chinese translator (as it is dropped by Nattier). 

Ishii spends a good deal of time speculating on how to translate the Sanskrit text into Chinese, twisting it this way and that according to rules which may be obvious to his Japanese readers, but which are not at all clear to me. His point seems to be that one may, through a series of arbitrary changes, rearrange a Chinese translation of the Sanskrit, to fit the pattern of Chinese one finds in T251 (the standard Heart Sutra in East Asia). However, on face value the Sanskrit and Chinese texts are simply different. I am told that this may reflect the Japanese practice of rearranging Classical Chinese texts into the Japanese word order and only then interpreting them, a procedure known as  kaki-kudashi, 書き下し.

A particular problem is that the Sanskrit has three phrases, marked by the present participle caramāṇo, "practising") and two verbs with meaning "look" (vyava√lok) and "see" (√paś) - both using the pleonastic particle sma indicating the past or the present-in-the-past tense. One of the problems in Chinese is that there are only two verbs in this sentence, i.e., "practising" (行) and "clearly-seeing" (照見). Ishii seems to be saying that the latter is in fact two verbs in two distinct phrases, but rearranged in a series of aesthetic changes so that the two verb characters are together at the beginning of the two phrases, in the order verb1 verb2 phrase1 phrase2

Ishii then discusses the 照見 combination in Chinese literature (two examples) and the vyavalokayati sma/paśyati sma combination in Sanskrit. However, he seems to show that 照見 is used as a binomial verb - the two characters have to be taken together, rather than as two separate verbs, which undermines his case. He argues that, though the phrase 照見五蘊皆空 ("[he] saw the five skandhas were all empty") occurs nowhere else in Chinese, translating it as two phrases does not make sense. 

Next Ishii brings up the commentaries of Kuījī (Ji in the article) and Woncheuk. Ishii notes that Kuījī does not mention a Sanskrit text and that he used a minor variant of T251, which has an extra character  等 (Sanskrit ādi = English "etc") in two places. Woncheuk was also aware of this variant, and finds ādi in his Sanskrit text, though, of course, his commentary is on the text of T251. It is very likely that these two commentaries established T251 as the authoritative text of the Heart Sutra down to the present. Neither man mentions the differences between the versions in the introductory section. As Ishii hints, had a Sanskrit text been available, it would have been incumbent on the commentator to comment on differences, if only because Sanskrit texts were considered authoritative (this was the entire rationale behind Xuanzang's journey to India, after all).

Ishii reveals that his primary goal is still to criticise Nattier's omission of a word for word translation of svabhāva. He has spent 6 of the 8 pages of the article showing this, though we may say that this is an obvious point and one that has little bearing on the larger issue of where and when the Heart Sutra was composed.

Having laboured this point, Ishii briefly discusses the phrase 真實不虛, "true and not false". The Tang dynasty commentators all take this as a standalone phrase; however, Ishii claims that the Sanskrit manuscripts read "satyam amithyātvāt, prajñāpāramitā ukto mantra" which is the way Nattier translates it. Ishii uses the same method to translate the Sanskrit into Chinese, producing something different than the present Chinese text. Ishii seems unaware that Nattier is following Conze's edition, and that Conze's edition gives this passage as:
Tasmāj jñātavyam: prajñāpāramitā mahā-mantro mahā-vidyā-mantro ‘nuttara-mantro’ samasama-mantraḥ, sarva-duḥkha-praśamanaḥ, satyam amithyatvāt. Prajñā-pāramitāyām ukto mantraḥ. 
On this basis, then, Ishii declares that Nattier's thesis is a mistake and untenable. Had I been reviewing this article prior to publication, I would have argued that it need major modifications before being published. As it stands, the argument is difficult to follow and the evidence does not support the conclusion. 



Critique of Ishii (2015)


Core of the Thesis

Nattier's thesis mainly revolves around the core section of the Heart Sutra, which is a quote from Kumārajīva's text of the Large Sutra (T223). The Chinese Heart Sutra, especially T250 is identical with T223. T251 is identical, but missing a line at the beginning and one in the middle; and a few technical terms are "spelled" according to innovations introduced by Xuanzang. The Sanskrit Heart Sutra, by contrast, is a strangely unidiomatic paraphrase of the Sanskrit Large Sutra (compared to either the Gilgit recension or the later Nepalese recension).

The Sanskrit Heart Sutra contains a number of words or phrases that are hapax legomena (one of a kind), whereas the Sanskrit Large Sutra has a string of stock phrases. The Sanskrit Heart Sutra is unidiomatic in almost every place where it is possible to use a nonstandard synonym, that is, outside the settled technical vocabulary of Buddhist jargon.

There is no doubt in my mind, despite some minor slips on Nattier's part, that the thesis is accurate. I think I have the smoking gun for this, but have not yet had time to check all of the details and write it up. So far as I can tell the term sarvabuddhāḥ tryadhvavyavasthitāḥ "all the Buddhas existing in the three times" is a translation of a phrase that only ever occurs in Chinese, i.e., 三世諸佛. This is literally, "three time all buddha", but we would translate it as "all the buddhas of the three times". Sanskrit texts always use the wording atītānāgatapratyutpannāḥ buddhāḥ instead, i.e., "past, future, and present buddhas". There is no way that the Sanskrit Heart Sutra could be anything but a translation from Chinese, produced by someone unfamiliar with Sanskrit Prajñāpāramitā idiom. I need to do a very thorough check on the various texts, but I think this conjecture will stand up to scrutiny and provide definitive proof of the Chinese origins thesis.

Whatever minor flaws we may find in Nattier's analysis, the main conclusion that the Heart Sutra was composed in Chinese is already beyond reasonable doubt. While I would be interested to get more insights into the problems that Japanese scholars see, I cannot imagine how they think they have disproved the thesis. Ishii has certainly not done so in this article, though, strangely, he provides quite a good summary of the evidence presented by Nattier. However, Ishii does not even touch on this central problem or any of the evidence for it, but concentrates instead on peripheral and seemingly trivial issues that have no impact at all on the issues at hand.

Both of the passages that Ishii comments on are outside the core part of the text; i.e., not part of the quoted section, but part of the original composition that accompanies it, one in the introduction and one in the concluding passage.


Flaw in the Introduction

Before addressing Ishii's comments in the introduction I need to point out that I have showed that Conze (and, for that matter, Müller) made a mistake in his edition. In the first (three phrase) sentence, pañcaskandhās is nominative plural and vyavalokayati sma is intransitive, both of which are nonsensical and make the sentence impossible to parse as Sanskrit. In fact, as some manuscripts allow, the noun should be in the accusative plural, pañcaskandhāṃs (simply add anusvāra to dhā). If we do this, pañcaskandhāṃs becomes the object of vyavalokayati sma. The result is a sentence that can be parsed and that does not require any punctuation (Attwood 2015).

Without solving this problem the Sanskrit sentence cannot be parsed or translated without fudging things. Both Nattier and Ishii fail to notice anything amiss, here. But, then, so do all other scholars, apparently.  In this respect, the Heart Sutra is a curiously neglected text, given its popularity. My next published article will identify and solve another simple error in Conze's edition (in Section VI) that has also gone unnoticed (the flaw is already outlined in my essay Red Pine's "Vagaries of Sanskrit grammar" 13 October 2017, but the article will give rigour to the conjecture).

The main problem that Ishii highlights, other than Nattier's failure to provide a word-for-word translation of svabhāva, is that the Chinese has two phrases and the Sanskrit three phrases. If we assume that the Sanskrit is original, then we expect three phrases in the Chinese, as well. In order to make three phrases, Ishii proceeds to rearrange the characters 照見 to make one verb into two verbs, each applying to two different parts of the sentence. 照 can, in fact, mean "inspect, regard" which is what vyavalokayati means, so in that sense this procedure makes a certain amount of sense.

However, Ishii's method seems to require us to believe that Chinese has no syntax rules. We know that Buddhist Chinese does follow syntax rules, albeit that it sometimes follows medieval Chinese and sometimes Indic rules. Ishii's method is a classic case of making the data fit the hypothesis. It is a post hoc rationalisation. His method is not sound, and not consistent with established principles of philology.

In all of this procedure it is never explained why a Chinese translator would omit the word svabhāva from their translation if it occurs in the Sanskrit text, nor why they would condense three phrases down to two. Nothing is explained. 

Assuming that we ignore the overwhelming case of a Chinese origin for the core section, there is no way to establish precedence by comparing the number of phrases in a given passage outside the core. In my work on the epithets of the mantra (Attwood 2015) I showed that the number of epithets varied from 2 to 8 in unpredictable ways. Note also that Conze's English translation of his Sanskrit, has an fourth phrase as he struggled to turn his garbled Sanskrit into comprehensible English.


True and Not False

It is ironic that Ishii should bring up 真實不虛, because the Sanskrit is clearly a mistranslation of the Chinese. Although the combination of 真實 and 不虛 is common in Chinese, the combination of satya and amithyā never occurs in Sanskrit outside the Heart Sutra, where is is one of several hapax legomena. Although Ishii provides several examples of the use of 真實不虛 in Chinese, he never gives the Sanskrit equivalent. Since we know that it is not satyam amithyātvāt, it would be most interesting to see what the equivalent is. 

However, the problem here is deeper: satyam amithyātvāt is nonsensical as it stands. Amithyā does not mean "false"; i.e. ,it is not an antonym for satya, which would be mṛṣa or even asatya. Mithyā, on the other hand, is the antonym of samyañj, and it means "wrong" (as in "going about something the wrong way, against the grain, in the wrong direction"). Worse, in fact 虛 isn't an antonym of 真實, "true", either, but, instead, means, "hollow, empty; vain, pointless". The passage does not mean "true and not false"; it means "true and not in vain". And amithyā cannot be construed as a good translation of this. And the word in Sanskrit that might correspond to this is tucchaka. A better English translation would thus be "true and effective". A better Sanskrit translation would be satyaṃ atucchakaṃ. Again, I hope to publish something on this, but it is another case of something that ought to have been obvious to anyone who reads Buddhist Sanskrit texts. 

Syntactically, in Chinese both qualities are predicates of prajñāpāramitā (there is no suggestion that one is the cause of the other). It makes no sense at all, in Sanskrit, to take satyam amithyātvād with the following passage. Amithyātvād is weird: the wrong word in the wrong form in the wrong case. It is not the weirdest thing about the Sanskrit Heart Sutra, but I find it hard to believe that it has not caused other scholars to scratch their heads.


Miscellaneous Criticisms

It is strange that Ishii would use Müller's diplomatic edition rather than the critical edition by Conze. Despite being flawed in places, it is still the result of comparing many different manuscripts. At one point Ishii refers to "most of the extant Sanskrit manuscripts", but he does not cite any one of them. We have to wonder what sources he consulted, or whether he referred to Conze's notes in his edition? In which case, why not use that edition as his Sanskrit source?

At one point Ishii makes a big deal of the Chinese translations of the extended version of the Heart Sutra T253, T254, T255, and T257. He must surely be aware that there is no dispute that these are translations from Sanskrit. The dates are clearly recorded in Chinese and that they come from a much later period. They have no bearing on the matter of which language the text was composed in. Citing them doesn't help his case at all.

Thinking about Woncheuk's reference to a version with 等 (ādi) in it, Lusthaus (2003) also tries to make something of this. But so what? The version is no longer extant and was not canonised - no one saw it as important enough to preserve. And as before, it doesn't affect the main arguments. Ishii and Lusthaus both fail to see that, although Woncheuk appears to have had a Sanskrit text, he does not treat it as authoritative. Rather, he comments on T251 as the authoritative version of the text. So does Kuījī. Under what circumstances does a Sanskrit "original" (as Lusthaus calls it) not trump a Chinese translation in early medieval China? In fact, both Kuījī and Woncheuk were aware that the Heart Sutra was not a sutra, and Kuījī at least knew it contained a quote from T223  (see Nattier 1992: 206-7, n.33). So this is not news. It is quite likely that it is precisely these two commentaries that establish T251 as the authoritative text in China and its cultural sphere. This is entirely inconsistent with the pair having a Sanskrit "original".


Conclusions

The text of the Sanskrit Heart Sutra is so far from the idiom of Prajñāpāramitā Sanskrit literature, Buddhist Sanskrit literature, or any other kind of Sanskrit literature, that the fact itself is (or ought to be) remarkable. The Heart Sutra stands alone in the entire body of Sanskrit literature and is only related to the other Prajñāpāramitā texts by its use of jargon. This is not consistent with being composed in India. It is consistent with having been composed in China by someone proficient in Sanskrit, but without any great knowledge of idiom. This could not have been Xuanzang - who was more familiar with Sanskrit Prajñāpāramitā idiom than anyone in China at the time. I think the mistakes highlighted by Huifeng (2014) also helped to cement the Chinese origins thesis. The translator has misread the Chinese text at times and has struggled to find the Sanskrit vocabulary to express the Chinese concepts at others. Again, this is inconsistent with a monk in an Indian Sanskrit-using context. The translator was relatively isolated.

I admit, I was hoping for something a bit more challenging from Ishii and I found the article quite disappointing. He concentrates on peripheral issues and provides no refutation of the very strong evidence put forward already (and added to by Huifeng and myself in the last couple of years). The methods are not sound and the conclusions are weak and do not derive from the evidence presented. It looks like a tendentious throwing together of evidence to support a preconceived conclusion. "It is inconceivable that the Heart Sutra was composed in China, therefore it wasn't. QED." But this is hardly the standard of argumentation and reasoning we expect from a senior academic.

Like other scholars before him, Ishii has simply overlooked the grammatical errors in the Sanskrit text, which I am less and less inclined to forgive in professionals. After all, professionals are, on the whole (with a few notable exceptions), very hard on me when I dare to encroach on their territory and do not meet their high standards. So yes, let's have high standards, but that includes not being duped into accepting simple grammatical errors in our texts. 

We should, of course, not judge Japanese scholarship more generally on the basis of this single example, even though Ishii is a senior member of the Japanese Buddhist Studies establishment. We can hope that the article does not reflect the state of the art in Japan. However, it is not a good sign that such a weak and confused article could be published in a peer-reviewed journal at all. 


~~oOo~~



Attwood, Jayarava. (2015). Heart Murmurs: Some Problems with Conze’s Prajñāpāramitāhṛdaya. ​​Journal of the Oxford Centre for Buddhist Studies, 8, 28-48. http://jocbs.org/index.php/jocbs/article/view/104

Attwood, Jayarava. (2017).  ‘Epithets of the Mantra’ in the Heart Sutra. Journal of the Oxford Centre for Buddhist Studies, 12, 26–57. http://jocbs.org/index.php/jocbs/article/view/155

Huifeng. (2014). ‘Apocryphal Treatment for Conze’s Heart Problems: “Non-attainment”, “Apprehension”, and “Mental Hanging” in the Prajñāpāramitā.’ Journal of the Oxford Centre for Buddhist Studies. 6: 72-105.http://www.ocbs.org/ojs/index.php/jocbs/article/view/75

Ishii, Kosei. (2015) 『般若心経』をめぐる諸問題 ―ジャン・ナティエ氏の玄奘創作説を疑う = ‘Issues Surrounding the Heart Sutra: Doubts Concerning Jan Nattier's Theory of a Composition by Xuánzàng.’ Journal of Indian and Buddhist Studies (Indogaku Bukkyogaku Kenkyu), 2015, 64(1), 499-492. (Translated by Jeffrey Kotyk).

Lusthaus, Dan. (2003) 'The Heart Sūtra in Chinese Yogācāra: Some Comparative Comments on the Heart Sūtra Commentaries of Wŏnch’ŭk and K’uei-chi.' International Journal of Buddhist Thought & Culture. September, Vol. 3: 59-103.

10 July 2009

Kūkai's journey to China : Kentōshi Ships and Weather

To see my Google Map click here


One of the marvels of modern technology is that we have easy access to all kinds of information. I've been trying to visualise Kūkai's journey to China and to understand the scale of it. Using the internet I was able to locate a journal article which discusses the detail of the journey, then using Google Maps I have been able to visualise it and get a sense of the scale of it. The route outlined here relies on an article by Robert Borgen in Monumenta Nipponica.*

Kentōshi (遣唐使), which means 'Envoy to the Tang' i.e. mission to Chinese court of the Tang (T'ang 唐) dynasty,** was used to describe both the people and the ships they went on. We don't have much definite information about the vessels, but it is assumed that they were built on the model of the Chinese junk which were developed in China during the Han Dynasty (220 BCE - 200 AD) which were being used for ocean voyages by the 3rd century. Such Chinese ships visited Japan for trade. We know that the Japanese and the Koreans definitely used Chinese junks as models for later ships. It's often stated that because the ships had a flat bottom and no keel that they could only use the sails when the wind was directly behind them. However the boats used a very large rudder which projected well below the bottom of the ship, and did much the same job as a keel, i.e. it stopped the wind pushing the boat sideways when sailing to windward. They could probably have managed to sail close hauled at between 45-60° to the wind. Which in fact means that they could sail in much the same way as an early square rigged European ship such as Magellan had sailed around the world in.

The idea that the Japanese were poor sailors seems to be an assumption related to their decision to sail in the typhoon season, but as I pointed out in an earlier post (Why did Kūkai sail in summer?), the Japanese envoys were concerned to get to the Tang court on New Years day in order to offer their tribute at the appropriate time, and this must have over-ridden the concerns of the sailors. In fact the Japanese were highly attuned from ancient times to the annual changes to their climate wrought by the monsoon, and I find it very unlikely indeed that they did not understand the wind patterns. Note also that by Kūkai's time, in the early 9th century, envoys from the nation of Po-hai (north of Korea) to Japan regularly timed their journeys to take advantage of seasonal winds.

It's very often stated that the winds were against the ships sailing across the sea to China, but the prevailing wind during the summer monsoon in that region is from the south-east. This means that the Kentōshi ships, sailing south and west, were most likely cutting across the wind - a favourable geometry for sailing. With a wind from the south-east (135°) they could probably have sailed in any direction from say 0° - 75° and 195° to 360°. In fact a line joining Tanoura to Ming-chou is at about 252-3° which in sailing terms is a 'close reach' and probably well within the capabilities of the ships.

It is quite unlikely that they could have made the journey at all if they had to row ships that probably weighed over 100 tons all the way, and it does not seem so unreasonable to me that they relied on sails most of the time - even sailing north from Fu-chou to Ming-chou. Note that all four ships of the mission survived a typhoon, some of them two typhoons, and a 500 mile ocean crossing so they must have been reasonably well built. European ships of a similar size and square rigged could make about 5-7 knots, and, allowing for variable wind conditions and given that they would have paused during the night when they could, I initially guessed that they might average about 20 or 30 miles per day.

Previous missions would have made a quick jump across the straights of Korea probably via Tsushima Island, a journey of about 150 miles with a longest stretch of open water of about 35 miles. On a good day the Kentōshi ships could have sailed that distance in a single long day. From there the boats would have hugged the coast all the way to China. However in the 7th century Japan's long term enemy Silla had, with the help of Tang China, unified the whole Korean peninsular under their rule, leaving the Japanese with no bases on the mainland and a more powerful antagonist as neighbour.

The four Kentōshi ships left from Naniwa (modern day Ōsaka) and headed for Hakata (Fukuoka) on Kyūshū Island, a distance of 330 miles most of which is in the usually pacific Inland Sea. Note however that in 803 when the mission first sailed the boats were almost wrecked by a (rare) storm in the Inland Sea. From Fukuoka the ships hugged the coast of Kyūshū down to Tanoura (since merged into Ashikita), the last stop before heading west across the East China Sea. We don't know how good the navigation techniques were at this time, though simply sailing west would mean hitting China at some point, but the ships ideally would make land near the modern city of Shanghai or north of there. They left from Tanoura on the 6th day of the 7th month of Enryaku 23 (ca 14 August 804).

Of the two ships that completed the journey in 804 Ship Two is said to have taken about two months to get to Ming-chou (near modern Ningpo). Now here is a puzzle: Abe, Hakeda, and others give this time frame, but Abe says that the Vice Ambassador who lead Ship Two died in Ming-chou on the 25th day of the 7th month of Enryaku 23. This is a mere 19 days after leaving Tanoura. So, assuming this is not a misprint, either the Vice Ambassador died at sea less than half-way across, or Ship Two made very good time crossing the 540 miles, averaging about 30 miles a day. The latter figure is not unreasonable if they met no more storms, and my other assumptions are correct.

Ship One, the ship that Kūkai was on, took much longer to make the crossing, coming to land on the 10th day of the 8th month (ca 17 September 804) after 34 days at sea. They landed near the city of Fu-chou, in Fukien province (modern day Fuzhou, Fujian). It is sometimes said that this was 1000 miles south of where they intended to be, however the map above makes it clear that the distance from Fu-chou to Ming-chou by sea is about 390 miles, and by land about 360 miles to Hang-chou (using a route something like that suggested by Borgen). In a straight line Ship One covered about 750 miles in the crossing, which means they averaged at least 22 miles per day. In fact we know that they didn't go in a straight line because they were blown off course by the typhoon.

On the return journey (late June early July of 805) which was apparently without major incidents Ship One took nineteen days to make landfall at Tsushima (the island in the Strait of Korea); while Ship Two took twenty eight days to arrive at Hizen on Kyūshū Island. This is about 29 and 19 miles per day respectively - quite comparable to the outward journey suggesting that 20-30 miles per day is a good measure of the average speed of the ships.

typhoon over the East China Sea
Typhoon Tokage near Japan
Image Courtesy NASA Earth Observatory
Typhoons make a rather wavy line as they progress towards Japan from the Pacific Ocean, typically they follow the prevailing winds which spiral out from a massive region of high pressure over the Pacific and into a low over continental Asia. In August the typical typhoon would swing around Kyūshū and head up the Sea of Japan - though a lot of variation has been observed. As the typhoon approached the wind would have swung around initially from the south-west, to the west - the winds swirl in anti-clockwise to the centre, and have become a tight knot by the time they reach Japan. On the western side of the storms the winds are blowing more or less to the south and this explains how Ship One might have been sent far southwards. Ship Two somehow escaped this. The trailing edge of the typhoon seems to have blown Ships Three and Four eastwards back to Japan, though this suggests that there was already a significant distance between them and Ships One and Two by this stage.

Borgen's article is an important source of information about ships 3 and 4 from the Kentōshi flotilla - but that is another story. Hopefully you can see that using Google maps in this way really does makes the scale of the journey clearer, and you find my route plausible where I have supplied details not vouchsafed by history. The historical sources are vague on the construction and design of the ships, but I hope my reinterpretation of the Japanese as intelligent and able boat builders and sailors is both welcome and sound - I hate it when historians assume that people are stupid because they (the historians) don't understand what was going on!


Notes
* Borgen, R. The Japanese Mission to China 801-806. Monumenta Nipponica, Vol 37(1), 1982, p.1-28. In this article I also indirectly cite or use information from: Abé, Ryūichi. The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse. (Columbia University Press, 2000); and Hakeda, Y.S. Kūkai : major works : translated and with an account of his life and a study of his thought. (New York : Columbia University Press, 1972).
** I tend to use the Pinyin version of Chinese transliteration with Wade-Giles equivalents in parentheses at the first occurrence. If there is only one transliteration it is Wade-Giles and I don't have a Pinyin version. Some names have changed substantially since Kūkai's time.

For other materials related to Kūkai and his voyage see my Kūkai bibliography.

Aug 2010 Update.

Since writing this essay I have studied the Diary of Ennin (Ennin, E.O. Reischauer (Translator] Diary: Record of a Pilgrimage to China in Search of the Law) paying particular attention to his records of wind and sailing directions. Although he records about a dozen combinations, the ships he sails on never seem to sail into the wind, and only run before it. It now seems more likely to me that the ships couldn't manage anything more than a broad reach - about 45° either side of the wind direction, i.e. that they could not use a head wind. I've noted that the prevailing wind at the time of year is from the South-East (or perhaps the East) and this may tally with their leaving from quite far south on Kyūshū - they expected to make leeway to the North while travelling West. Although my lines on the map are straight it seems likely the storm blew them far to the south, and that they then sailed North/N' West to make landfall. I have no idea if the could accurately determine latitude.


09 May 2008

Why did Kūkai sail in Summer?

Anyone familiar with the story of Kūkai will know that his journey to China in 804 began by sailing from Nagasaki out into the Sea of Japan. It is usual to comment on the relative lack of seaworthiness of the Japanese ships, and on the lack of nautical knowledge of the sailors since they sailed at a time when the winds were against them, meaning that the rudimentary sails could not be used; and when typhoons regularly swept in wrecking any ships daring to be out of harbour. This is a given in all the biographies in English.

However as long ago as 1995 TŌNO Haruyuki cast doubts on this way of telling the story, at the same time as questioning another long held belief: that the Japanese Emperors presented themselves as equals to the Chinese Emperor, and that the Chinese went along with this. This latter is interesting because it sheds light on the nature of the embassies sent from Japan.

Tōno shows that there is evidence to throw doubt on the supposed equality of the two emperors. It is true that as early as 607 a mission to the Sui dynasty emperor Yang-ti (隋煬帝 ) presented a letter which described the Japanese emperor as Son of Heaven, the title of the Chinese Emperor, however Yang-ti saw this as an affront.

Tōno's article concentrates on the embassies to T'ang China. In 632 a Chinese imperial envoy clashed with the Japanese court over protocol and did not read the letter from the Chinese Emperor. Tōno suspects that this was an attempt to subdue the Japanese. Note that this was a period of massive expansion westwards, with Chinese troops pushing on past the Tarim basin, where they were stopped by an Arab army also intent on expansion. It was the time of the greatest extent of the Chinese Empire.

Until 663 the Japanese were influential in the Korean peninsular. However in that year the Paekche (from whence Buddhism was introduced into Japan in 552) were defeated by a coalition of the T'ang and Silla, despite being shored up with Japanese forces. In 668 the alliance defeated the Koguryo thus unifying Korea. Although the Japanese continued to see Po-Hai (in present day Manchuria) as a tributary state, Tōno points out, from this time onwards it would not have been possible for the Japanese to insist on equal status. Indeed the embassy of 671 can be seen, according to Tōno, as a declaration of surrender!

After a break of 30 years another embassy was sent to the T'ang court in 702. It was at this time that the Japanese concede to paying tribute every twenty years. This was a pragmatic move on the the part of the Japanese in the face of a rampant T'ang state in the process of crushing opposition in other quarters. Evidence of this promise, more or less hushed up at home, is seen in a letter from a monk on Mt T'ien T'ai who is asking for permission to pass on information to the Japanese monk Ensai in 840 where he mentions that "... and they [the Japanese] have promised to pay tribute once in twenty years" (p.45). This would not have been common knowledge in Japan, and though careful records of many other occasions were kept, letters from the Chinese Emperor were mostly lost. In one letter from the Chinese Emperor 735 begins by writing "I order the king of Japan..." (p.52).

It obvious that in the Japanese mind Japan was the centre of civilisation. The ritsuryō code for instance, despite being modelled on a T'ang Chinese legal code, refers to other nations including the Chinese as barbarians. Tōno cites the fact that no one of the royal family ever went to China as this would have admitted to the Japanese people that they were subordinate.

Although Tōno does not mention it, we could also comment on the relative weakness of the Japanese nation until the reforms of Kanmu began to take effect. Japan had been essentially bankrupted by a succession of natural disasters and the flurry of temple building that ensued as a remedy, and by a number of expensive and sometimes disastrous military campaigns against the Ainu. In Kūkai's day there was forced labour and military service. Many people were homeless, and farming so difficult that many left the land to become beggars. In the face of a strong and dynamic T'ang Japan would have looked weak, and perhaps it is only the long sea distance that prevented them from being assimilated along with other neighbours.

Tōno's conclusion is that the embassies to the T'ang court were to offer tribute as agreed in order to keep the Chinese Emperor from casting a military eye eastwards. It is this fact which gives us the clue to why the embassies were sent when they were. As I mentioned it is common knowledge that Summer is a bad time to sail to China; and it is assumed that the Japanese were simply ignorant of the seasonal winds. However Tōno reminds us that emissaries from the Po-hai state regularly visited Japan at the time, and judging by their arrival and departure dates they were adept at using seasonal winds. (p.58) Tōno also argues that the Japanese ships were more sophisticated than has previously been thought, that they used cloth sails in addition to bamboo matting. However they did lack keels which meant they could not use the sails unless the wind was behind them.

The offering up of tribute to the Chinese court was ideally done at the New Year celebrations - the Chinese year beginning on the second full-moon after the winder solstice, usually sometime in February. The average travelling time to China for all of the missions, which can be worked out from a chart in Tōno's article, was six months. This meant leaving in the 6th month, or late summer (July or August) in order to arrive in time for the ceremony in January or February. Far from being ignorant of nautical and seasonal knowledge the Japanese probably knew exactly what to expect, but were forced for political reasons to attempt the crossing at this time. The knowledge of what to expect was probably what accounted for the reluctance of Japanese officials to go on such trips.

After Kūkai's trip in 804-6 only one more Embassy was sent to the T'ang court. Perhaps this was because it was clear, even in 806, that the T'ang dynasty was falling apart. It staggered on until 906 but was racked by civil strife and war. In other words there was no longer any threat to induce a offering of tribute, and Japan had gotten onto a firmer footing as well. Thanks to Kūkai the Heian period was one of a flowering of Japanese culture as distinct from imported Chinese culture.


TŌNO, Haruyuki. "Japanese Embassies to T'ang China and their Ships," Acta Asiatica. 1995 v.69: 39-62.
image: Illustration of a Chinese ship of the type that would have visited Japan during the Edo period (from Tōno article).