The Earliest Statements of Christian Faith in China? A Critique of the Conventional Chronology of The Messiah Sutra and On One God**

ABSTRACT

This paper is the first systematic critique of the conventional chronology of the so-called earliest statements of Christian faith in China, The Messiah Sutra 序聽迷詩所經 and On One God 一神論, which are believed to have been translated into Chinese by the first Christian missionary to the country Aluoben 阿羅本 around the 640s. To achieve this, it has surveyed the early publications and re-read the alleged textual and historical clues. The paper reveals that the common arguments accepted since the 1920s are plagued with defects and that therefore the conventional chronology does not withstand scrutiny. The significance of this paper lies in proposing to reconsider not only the traditional dates of these two documents but the conclusions derived therefrom. Based on some clues, this paper also speculates that The Messiah Sutra and On One God might be two of the latest ancient Chinese Christian sources.

Introduction

The Messiah Sutra 序聽迷詩所經 and On One God 一神論 are two Chinese Christian religious manuscripts discovered in Dunhuang Cave 17 in 1900. Since their first publication, they have been accepted as the earliest statements of Christian faith in China, the first fruits of Sino-Christian relations, and the inauguration of translating the Bible into Chinese. These two documents, according to the common understanding, were translated by Aluoben 阿羅本, the first known Christian missionary to China, who arrived at Xi'an in 635, and were used in his discussions with Emperor Tang Taizong. The former, it is said, was created between 635 and 638; the latter in 641. To date, as shown below, this chronology still prevails. The two sources are universally used to attest to the arrival of the first presence of Christianity in China, known as Jingjiao 景教, which was established by the Church of the East in 635 (hereafter the Tang church) and was allegedly extinguished after 845 during the Tang dynasty.¹

Despite this apparent consensus, their dating to the 640s and authorship is far from certain. First and foremost, neither The Messiah Sutra nor On One God carries a date. Their dates were assigned when they were first published in the early twentieth century, but have never been substantiated. In fact, not a single source ever records that either of the manuscripts was composed in Xi'an by Aluoben and eventually sent to the remote oasis of Dunhuang around the 640s. Moreover, current scholarship seems to have realized that the conventional chronology is problematic. In his article presented to the first Salzburg symposium on the Church of the East, "Past and Current Research on Tang Jingjiao Documents: A Survey," for instance, Nicolini-Zani (2006, 38) points out: "In summary, we can say that the traditional position [of all seven extant Chinese Christian manuscripts—see below] was based on an almost uncritical acceptance of a small amount of uncertain data and on the passive transmission of many unfounded assumptions." More specifically, Riboud (2001, 5) states that the dates of The Messiah Sutra and On One God "are rather uncertain." Sadly, space does not allow these two scholars to offer explications and pursue the issue of uncertainty any further. Their voices, although insightful, are brief and remain marginal.²

This paper does not attempt to offer an alternative dating, as doing so is a daunting task and involves a deep understanding of the ancient Christian presence in China and many other complex issues, such as manuscript production. Instead, this article continues the unfinished work of the above two scholars by thoroughly exposing the uncertainty of the accepted timeline. It will survey the early publications and re-examine the alleged clues, showing that the arguments maintained by mainstream scholars do not hold water. To establish a foundation for the critique, the paper first introduces The Messiah Sutra and On One God before scrutinizing the common arguments for their conventional chronology and drawing wider implications.

Knowing The Messiah Sutra and On One God

In the discussion on the history of Christianity in medieval China, The Messiah Sutra and On One God feature more prominently than five other known Chinese Christian religious manuscripts, which also were removed from Dunhuang Cave 17 and are seen as the texts used by the Tang Christians.³ In terms of length, all other five sources combined—to be more precise, the other six named texts listed below, of which the last two are contained in one manuscript (P.3847)—are just a little bit longer than The Messiah Sutra alone, and yet much shorter than On One God. In terms of chronology, according to the traditional view, The Messiah Sutra and On One God are the oldest manuscripts and the only surviving seventh-century religious texts. Moreover, The Messiah Sutra and On One God seem to be in better condition than other manuscripts. As shown by recent color photos released by the archives in which they are stored, the Kyōu shooku 杏雨書屋, Osaka library of the Takeda Science Foundation 武田科学振興財団, the paper is thick and the ink unfaded.⁴ Despite some water stains, none of the glued sheets has deteriorated, crumbled, or become riddled with wormholes.

In order to more directly show the special position of The Messiah Sutra and On One God, I provide here the common timeline used by scholars of all seven manuscripts together with additional information about their length.⁵

1. 1. Two sources were created in the 640s by Aluoben, and are commonly known as the Aluoben documents.

   The Messiah Sutra has 2,845 characters;

   On One God contains 6,949 characters.

2. 2. Two early-eighth-century manuscripts. They have been exposed as modern forgeries by Chinese scholars (Chen 1997; Lin and Rong 1992).

   Praise of a Great Saint, Pervading Truth and Conversion to the Daqin Jingjiao 大秦景教大聖通真歸法讚 has been dated to the year 720 and is eighteen columns long.

   The Sutra of Origin of Origins of the Daqin Jingjiao 大秦景教宣元至本經 has been assigned to the year 717. Only its final thirty columns have been preserved.

3. 3. Four texts were made in the late eighth century (around the year 781) when the Christian priest Jingjing 景淨 composed the famous Chinese-Syriac bilingual Xi'an Stele 景教碑.⁶ They are known as the Jingjing Documents:

   The Mysterious Bliss Sutra 志玄安樂經 (2,596 characters);

   The Sutra of Origins of the Daqin Jingjiao 大秦景教宣元本經 (465 characters);

   The Praise of the Three Majestics 景教三威蒙度讚 (327 charactrs);

   The Sutra of Reverence 尊經 (277 characters).

In order to facilitate the following critique, four more issues need to be discussed here.

Firstly, The Messiah Sutra and On One God were only perfunctorily documented. According to Haneda, who first described, edited, and identified them as Christian sources, On One God was purchased from a Chinese book-dealer by a Japanese lecturer, Tomeoka Kenzō 富岡謙藏. The latter kept it for a while and then showed it to Haneda, who made the first two earliest, but very brief studies—one in 1918 and the other in 1923. The Messiah Sutra was obtained from an unknown Chinese person by Takakusu Junjirō 高楠順次郎. In 1925, he transferred it to Haneda, who publicized the source a year later. With the exception of what has just been noted, very little is known about their history before or since they were sold to Japanese collectors. No extra information was provided when these two sources were introduced to the West in the early 1930s (Saeki 1951, 1934, 1933, 1932; Moule 1930; Drake 1935). This poor documentation, combined with the inaccessibility of the original manuscripts (another issue that will be discussed shortly), has been a major impediment to research, because they have made the sources a mystery, thereby adding fuel to the theory that both manuscripts are modern forgeries.⁷

Thirdly, there is little doubt that the two sources are Christian religious texts. On One God seems to have come down to us in its entirety. Its three entitled essays, The Parable, II (喻第二), On One Heaven, I (一天論第一) and Shizun on Almsgiving, III (世尊布施論第三), are devoted to clarifying one specific theological theme that is indicated by the general title, On One God (一神論), although Part II comes before Part I. In particular, the third essay matches the Gospel stories about Jesus Christ (the ministry, betrayal, trial, and death of the Messiah and so on). On the other hand, The Messiah Sutra recounts the Gospel stories in a general way. Sadly, its seven badly damaged end columns were cut off by the Chinese seller and replaced by the empty heading we see today.⁸

The fourth outstanding issue is the possible interconnection between the two sources, namely the hierarchy (which text was based on the other). As observed by scholars (Lin 2003, 210–28; Drake 1935; Haneda 1926, 125–27), the two sources display some textual similarities. Their wording and style are close. Their contents are equally difficult. Many sentences are broken and do not lend themselves to straightforward readings. A number of characters, even some key theological phrases, were left out or incorrectly written. Furthermore, judging by the calligraphy, they were written within a reasonably short time of each other by the same hand. Unfortunately, the information is too meagre and not much can be said about the relationship between the two sources.

The Popularity of the Conventional Dating, the 640s

As stated previously, the dating of the texts to the 640s was assigned by modern scholars. All specific dates were first systemically explicated by Saeki (1951, 113–24) in his above-mentioned tome, The Nestorian Documents and Relics in China. Although it amounts to little more than fifteen pages in total, Saeki's investigation of the dates was and still is the most extensive discussion available, and has so far never been seriously challenged.

In the early stages, for instance, it was quickly accepted by Saeki's contemporaries. Moule (1940, 13–14) wrote that the two texts "have been with some reason dated about 640, or not more than five years after the arrival of the mission in 635." Even those who have criticized the exact dates have in principle agreed with this dating, and have merely proposed minor revisions. For example, Foster noted that the date of The Messiah Sutra was based on the accustomed calculation of the Christian era. "The earlier Patristic tradition," he argued (1939, 47), "placed the birth of [. . . Jesus] from two or three years earlier. The year 641 thus becomes 638."

The chronology was also unanimously accepted by second-generation researchers. Among them, to name a few, are Chiu (1987, 192–211), Moffett (1998, 291–314), and Gillman and Klimkeit (1999, 275–78) in the West, and Gong (1960, 24–29), Jiang (1982, 58–59), and Weng (1996, 111–55) in the East. In the current surge of interest stimulated by the growing number of Chinese Christians, The Messiah Sutra and On One God are still overwhelmingly believed by present-day scholars to "constitute the first statement of the Christian faith" in China (Charbonnier 2007, 44). In almost all works published in this new century, such as monographs written by doctoral researchers of the Tang church (Wang 2016; Nie 2010; Zeng 2005; Tang 2004), chapters by historians of Christianity (Foley 2009, 6–15; Ji 2007, 37–80; Kim 2004, 120–27), and sections by all other sorts of experts (Bays 2012, 7–11), the two sources are widely referred to as the Aluoben documents. With the exception of Nicolini-Zani and Riboud, I am not aware that anyone else has openly cast doubts on the conventional dating.

Looking at the universal acceptance and persistence of the traditional chronology, it must be conceded that there are good reasons for it. The persistence, as lately pointed out by today's scholars, was at least partly on account of the fact that the manuscripts remained unassessed for a long time. The two sources have been locked away since they were moved from China to Japan in the 1920s. Excluding the collectors and Haneda, only a handful of scholars have ever had the opportunity to study the original rolls. The only non-Japanese scholar who has had access to the originals was and remains Moule, and he (1930, 58) examined only The Messiah Sutra in the late 1920s. As far as can be inferred from the information available, Saeki never had the chance to examine any original sheets. To date, the original manuscripts are still inaccessible, although the Takeda Science Foundation opened the doors of its Osaka library in 2010 and exhibited them, as well as other Dunhuang manuscripts, to the public for the first time in 100 years. The inaccessibility, as described above, has been a major impediment to research, forcing almost all scholars, especially those who cannot read Chinese, to quote Saeki's and Haneda's transcriptions, translations, and publications and thereby reproduce their errors.⁹

In addition, it must be also observed that there are some motives for sticking to the conventional dating. From the traditional point of view, The Messiah Sutra and On One God were the only known seventh-century religious sources. They are also the core evidence that attests to the first clear presence of Christianity in China. For many scholars, in particular those who are Christians imbued with strong missionary inclinations and those who aim to demonstrate Christianity's long history and influences in East Asia, the conventional chronology presents a secure starting point for the presence of Christianity in China that is just too perfect and too spectacular to give up.

Three Inconclusive Arguments about On One God's Date, 641

According to its proponents, the date of On One God is suggested by three strands of textual evidence. Two frequently cited pieces are neatly summarized by Saeki (1951, 114, my underlining) who, meanwhile, acknowledges Haneda's contribution:

Regarding the date and the authorship, however, no specialist has as yet committed himself to declare his judgment, except Prof. Haneda who pointed out the fact that [… On One God] contains the following passages: "Though it is only 641 years since the time of the birth of the Messiah, 'consisting of the five attributes', yet (His name) is known in all parts of the world." [. . . The other] passage [is], "A point of space in the world of tangibility, for instance, is like the point between Persia and Fu-lin (i.e., Ephraim)." [. . .] And judging from these passages it is right that Prof. Haneda should declare that [. . . On One God] must have been originally written in 641 A.D., at a certain place situated somewhere on this side of the Persian Territory.

The third piece of evidence is that the Liao people (遼人) are disparaged as swine (豬). Saeki claims that this derogative usage was related to a specific historical event: Tang Taizong began his campaign to conquer the Liao area in 631 but encountered massive resistance. The Liao people, Saeki explains (1951, 236),

had become the object of hatred and contempt to the Chinese people with their capital in Ch'angan. It would be no surprise at all to find "the people of Liao" used as the synonym of the word "swine" as we do here. On the other hand, the Nestorian Mission met with a cordial reception of the Emperor T'ai-tsung in 635 A.D., and the Scripture began to be translated into Chinese under the auspices of the Emperor who sent the expedition army to Liao-tung in 641 A.D. Under these circumstances it might not have been impossible that the Chinese officials and courtiers with whom our author had to do a great deal might have called the enemy "people of Liao" by the name of "swine" as is often the case with the Chinese who call the enemy by very nasty names such as "vermin" or "swine."

Basically, Saeki is claiming that, in the context of imperial patronage, Aluoben was simply following the then 'common' practice of Chinese officials and associated the Liao people with pigs. Consequently, in Saeki's opinion, this derogatory usage suggests that On One God was created in this period.

However, none of these three strands of evidence is conclusive. The original Chinese text is very difficult to punctuate. The immediate meaning of its context is unclear. Moreover, Saeki's reading of the Chinese is not accurate. For example, the third piece of evidence was very probably not related to the 630 conquest at all. The usage of 'the Liao people' and 'swine,' as pointed out by Moule (1940, 17), was proverbial. It had been in use since the fifth century as shown by the historical account in the Houhanshu 後漢書.

Turning to the second argument, parsing the original Chinese is exhausting. Usually, three or four complete columns can yield enough information to punctuate ancient texts; however, these four columns are virtually unsolvable: "無接界一 處兩處第一第二時節可接界/處喻如從此至波斯亦如從波斯至拂林無/接界時節 如聖主風化見今從此無接界亦/不起作第一第二亦復不得此一神因此既無."¹⁰ It took me quite a while to match Saeki's translation with the underlined Chinese. Importantly, as indicated by the geographic names, Persia and Fulin (Rome?), this passage seems to deal with something related to space or location. It has nothing to do with the time of the writing of On One God. In point of fact, Saeki appears to misread Haneda's argument. Haneda (1923, 157) takes it as an indication of where On One God was made: a place east of Persia. He does not relate this evidence to the time the source was produced at all.

Finally, let us deal with the first argument, which seems to be the weightiest. Admittedly, the original Chinese, liu bai sishiyi nian 六百四十一年, does contain a time. However, the nub of the problem is that it is hard to ascertain whether it is a date, the year 641, or a time span of 641 years. Should it be a time span, it is not clear which specific year is the starting point for the calculation of this time span. Again, the surrounding Chinese text is obscure. The texts are so ambiguous that I am not sufficiently confident to offer any punctuation here. Even if one accepts Saeki's punctuation (1951, 66, "彌師訶向天下/見也。向五蔭身六百四十一年不過。已於一切處。"), Saeki's version of the Chinese, unlike his above underlined English translation, is far from clear. The last two parts of Saeki's Chinese punctuation are ungrammatical and unintelligible. Some elements have been dropped. The second part, for instance, is meaningful only if (1) the verb xian 見 ('to give birth' in Saeki's English) is inserted after wuyinshen 五蔭身 ('the five attributes') and (2) the final phrase buguo 不過 ('only') is moved from the end of the sentence to a position before the Chinese numeral 641: 向五蔭身[見不過]六百四十一年. Even after such a complex reconstruction, however, the third part is still not readable. Probably Saeki has realized this non-readability. To tackle the obscurity of this third part as well as to make his English translation intelligible, he silently added "His name," a phrase that does not appear in the original Chinese at all.

Moreover, Saeki attempts to convey that Haneda was confident about his interpretation of the Chinese time, 641. Reading the literature, however, it becomes clear that Haneda was by no means convinced of his own interpretation. As a matter of fact, Haneda dealt with the chronology of On One God only very briefly. Curiously, the fairly conspicuous Chinese numeral escaped Haneda's initial attention, and he (1918, 142) mentioned in passing that the source is "a Tang manuscript" (唐代の寫本). In 1923, for the first time, Haneda quotes the Chinese numeral and assigns the year 641 to On One God. However, this dating only appears in a page-long summary of a report of the Tang church presented by Haneda (1923, 156) at a meeting of The Society of Historical Research 史學研究會. In 1931, Haneda again mentions the part containing the Chinese numeral 641. But again he was sparing, and did not offer any arguments or justification for his punctuation of the original Chinese, which differs slightly from Saeki's version.¹¹

Perhaps the best illustration of Haneda's lack of confidence about the date 641 is that he actually refused to assign it to The Messiah Sutra. He (1926, 126) first suggests that The Messiah Sutra was written in "pre-mid-Tang China [mid-eighth century]" (中唐代時代以前), and then vaguely rephrases this claim (1931, 1): the source was created by "early" (初期) Tang Christians. Given that Haneda, as mentioned above, clearly knew that the two Christian sources were written in the same hand within a reasonably short time, it is rather odd that he should have been comfortable with the date '641' of On One God, but not claim that The Messiah Sutra was made around 641.

Three Dubious Observations on The Messiah Sutra's Dating, 635–638

First of all, as demonstrated above, Haneda never assigned The Messiah Sutra a specific date. The date of The Messiah Sutra as 635–638 was first systematically proposed by Saeki. Saeki (1951, 114–15) observed that "Prof. Haneda somehow refrained from expressing his own view definitely regarding the date and the authorship of" The Messiah Sutra, and therefore he claimed that, "it is our first duty to decide the most important question, viz. which is the older of all these documents, whether [. . . The Messiah Sutra] is the oldest of all the Nestorian documents yet found in China, or whether [. . . On One God is] older than [. . . The Messiah Sutra]."

Moreover, it must also be noted that The Messiah Sutra had already been claimed to be the oldest text before Saeki had even written a word about it. Moule (1930, 58) proclaimed briefly that The Messiah Sutra and On One God "are apparently earlier" than the other eighth-century Christian texts, even though he had not studied the latter source because of the inaccessibility of the original manuscript. "A slight indication of early date" of The Messiah Sutra, Moule (1930, 59) writes, is that the cross in The Messiah Sutra is called "a 'tree'" (木) rather than "the 'figure of ten'" (十字), whereas the latter Chinese phrase was not "adopted" until the end of the eighth century.

In addition, the date 635–638 seems to have been circulating in the literature before 1937. As far as I can tell, it first appeared in the introduction to Saeki's translation of The Messiah Sutra (1932, 31): "The date may be given as between A.D. 635–638."¹² As there are no further explanations and this introduction was added by the journal editor Evan Morgan, it is hard to determine who made this claim, or on what grounds. Most likely, the date was influenced by Moule's short assertion above, because his book, Christians in China before the Year 1550, is mentioned in that introduction. However Moule made no such claim in 1930.

Nevertheless, it is Saeki who first systematically examined the date of The Messiah Sutra in his abovementioned book, The Nestorian Documents and Relics in China. In essence, he carried out a comparative study accompanied by a historical survey to determine which of the texts predated the other. As shown below, he basically compared The Messiah Sutra with On One God and associated the making of the sources with a historical event, Aluoben's translation activity. In total, his book (1951, 114, 117, 121) gave three arguments, based on "historical," "philological or ideographical," and "doctrinal" observations. Each argument, according to Saeki, independently proved that The Messiah Sutra was made before On One God. However, none of those arguments can withstand scrutiny.

The Historical Observations

Saeki placed great stress on the historical observations because the fact that Aluoben translated some doctrines into Chinese between 635 and 638 seems to be documented by not one but two sources. One is the 781 Xi'an Stele; the other is the short colophon of manuscript P.3847, which will be discussed momentarily. According to the 781 Xi'an Stele,

In the country Daqin, the Great Virtue named Aluoben, [. . .] reached Chang'an [Xi'an] in the ninth year of the Zhenguan reign [635. . .]. [He] translated books in the library and was questioned [by Emperor Tang Taizong] about the Way in the imperial apartments. Understanding the profound truth, the Emperor thereafter permitted the propagation. In the seventh month (autumn) of the twelfth year of the Zhenguan reign [638], the edict was issued: "[. . .] Aluoben, the Great Virtue from the country of Daqin, has brought scriptures to the Upper Capital from afar. [. . .] As they are beneficial to things and people, they should be promulgated under Heaven. Have the requisite offices construct a Daqin Monastery in the Yining quarter of the capital and ordain twenty-one monks."

大秦國有上德曰阿/羅本[. . . . . .]貞觀九祀至於長安[阿羅本]翻經書殿問道禁闈深知正真特令傳授貞觀十有二/年秋七月詔曰[. . . . . .]大秦國大德阿羅本遠將經像來獻上京[. . . . . .]濟物利人宜 行天下所司即於京義寧坊造大秦寺一所度僧二十一人

This particular record, Saeki (1951, 116) claims,

will show to any one, that between 635 A.D. and 638 A.D., a certain kind of Christian literatures [sic] must have been translated or composed in Chinese either by Bishop Alopen himself or by his men with the assistance of native Chinese scholars, whilst the Emperor [Tang Taizong] himself investigated "the Way" or the outline of the Nestorian doctrine in his own Forbidden Apartments by reading these Nestorian literatures [sic] that might have been written by Bishop Alopen or his men.

Saeki (1951, 116) goes on to associate this record with the making of The Messiah Sutra. He asserts that The Messiah Sutra "bears characteristics to show that the document was written with special intention to give the Emperor [Tang Taizong] the outline or general sketch of the whole Christian Teaching." One of his examples is the Chinese version of the Ten Vows, which includes the worship of the emperors and ancestors. According to Saeki (1951, 117), the Chinese version reveals an integration of both Chinese culture and Christian teachings, and it is presented "to solicit the Imperial favour for the Nestorian Mission just arrived at the Capital."

In this light, Saeki compares the contents of The Messiah Sutra with those of On One God. He (1951, 117) writes:

any one who went through [The Messiah Sutra], as we have it, cannot but be convinced that [The Messiah Sutra] contains a surprisingly complete outline of the fundamental doctrine of Christianity and that it must have been presented to some one in order to teach the outline of Christianity, and that, therefore, it must have preceded [On One God] in the order of reading, since the latter contain[s] more advanced Nestorian doctrines than the former.

Following these observations, Saeki (1951:117) alleges that "we may rightly presume that [. . . The Messiah Sutra] must really be the very first Nestorian Sutra that was ever composed in the Imperial Library between 635 and 638 A.D."

However, these historical observations do not prove that The Messiah Sutra was made by Aluoben sometime between 635 and 638. The part on the emperor and ancestor worship does not necessarily reveal that The Messiah Sutra was written with the special intention of impressing Tang Taizong. Since the emperor and ancestor worship was a Chinese tradition, this part could have been written to impress any Chinese emperor. These observations therefore seem to exaggerate the translation event.

In fact, this reasoning simply toys with possibility. No evidence specifies which texts were translated by Aluoben. While one can argue that it is possible that Aluoben translated these two texts, one can equally claim that Aluoben did not translate them. As a matter of fact, one seems to be in a much better position to conclude that Aluoben did not translate The Messiah Sutra and On One God. The use of the Tang Christian texts is contained in P.3847. This manuscript, as introduced below, documents thirty-five sources used by the Tang Christians. Mysteriously, however, it does not record these two 'earliest' manuscripts.

More importantly, Saeki's argument does not demonstrate that The Messiah Sutra was composed earlier than On One God. The presumed existence of a "surprisingly complete outline of the fundamental doctrine of Christianity" contained in The Messiah Sutra is not adequate enough to demonstrate that The Messiah Sutra was made to show "the outline of Christianity" to someone (Tang Taizong) who knew nothing of it. It could have been created for and read by newly converted people or simply by younger members of the community whose faith had to be nurtured and strengthened. As the purpose of the writing of this source is not known, there are too many possibilities, none of which can count as a reliable historical argument. In addition, Saeki did not present any textual evidence that demonstrates that the doctrines outlined in The Messiah Sutra are less "advanced" than the doctrines set out in On One God. To determine that one doctrine is more advanced than the other is actually subjective. In essence, it is a value judgment, depending on the standards on which one relies. Since Saeki does not explicitly mention his criteria, one cannot evaluate whether they are objective or not.

Lastly, we must mention a common association between the translation event and the difficult texts in the discussion of the chronology of The Messiah Sutra and On One God. It is reasoned that the two documents are so obscure that they must have been composed by a foreigner like Aluoben who had then only just arrived in China and was not really familiar with Chinese.

At the outset, it must be said that Saeki did not explicitly make this association. Neither did he openly critique this reasoning. As far as I know, this rationale seems to have been initiated by Haneda in 1923 when he examined On One God for the second time, and was reiterated in 1926 when he first published The Messiah Sutra and observed the textual similarities of the two sources. Haneda (1926, 125–26) wrote that the (Buddhist) texts with the "strange phrasing" (奇怪な行文), wrong characters, and other textual features similar to The Messiah Sutra and On One God were often made by "foreigners who had not been versed in Chinese" (漢語漢文に通達しない外國人). But Haneda did not claim that it was Aluoben who created the texts.

In the 1930s, however, researchers such as Foster and Drake had already begun to openly associate the translation event with the chronology of The Messiah Sutra and On One God. They explicitly asserted that the two texts must have been made by Aluoben who had just arrived in China in 635. For example, Drake (1935, 679) specifically stated that The Messiah Sutra "may be one of those 'scriptures' that was produced in the Imperial Library, at the Emperor's command, between the years 635 and 638 A.D., by Alopen himself, as recorded on the Nestorian monument [Xi'an Stele]." Foster, on the other hand, attempted to reconstruct a dynamic translation process, in which the translators cooperated with the scribes and yet still could not produce satisfactory texts. To better understand the association between Aluoben's translation with the chronology of the two Christian manuscripts, I shall quote a core paragraph of Foster's reconstruction (1939, 44–45):

We can imagine them [the Christian translators] sitting in the great Library with their scribes. The translator would find chapter and verse in his Scriptures, and explain its meaning in halting Chinese to the clerk. The clerk would ask for further light upon this point and upon that. Then, often only half comprehending, he would write down what he thought was meant. After each sentence or paragraph he would read aloud his Chinese version. The translator would try to check it. But he could not be sure of correcting every blunder. Characters were written wrongly, but, right or wrong, they were strangers to him. Phrases bore a meaning different from that which he intended, but they were unfamiliar, and he had to accept them on trust. It was wearying work—especially for the scribe, whose heart was not in it. Once attention began to flag, all manner of mistakes crept into the text. But the Emperor [Tang Taizong] had commanded them to prepare samples of their gospel, so however ill-equipped for the task, they must proceed.

Admittedly, one cannot deny that there must have been scribes who rendered some help. Neither can one discard Aluoben's translation event or the difficulties of interpreting the texts. However, the association between Aluoben's translation activity and the making of the two sources is awkward. It leads us to believe that, even though Aluoben and his team did not have the necessary command of Chinese, they would still have been skilled enough to have translated The Messiah Sutra and On One God into that language. However, if these two texts are as difficult as they are commonly believed to be, it is hard to understand why Christians would have submitted such incomprehensible texts to Tang Taizong. Even if Aluoben did compose these difficult texts, one has to ponder how Tang Taizong would have managed to understand them; judged that the teaching of Christianity could not do his empire any harm; and finally permitted the diffusion of Christianity in Tang China in 638 by issuing an edict. Perhaps Saeki himself has already recognized the awkwardness of this association. On the one hand, as shown above, he relied on Aluoben's translation activity to establish the possibilities of creating the two questioning manuscripts. On the other hand, he avoided mentioning the difficulty of the texts. In fact, I am not aware that he ever complained about the difficulties of interpreting the sources.

Philological Observations

Philological observations, Saeki (1951, 117) goes on to argue, "will also prove beyond any doubt that [The Messiah Sutra] was written before" On One God. In this part, he analyzes the translations of some words, including the phonetic transcriptions of key theological terms. In particular, he focuses on how "'God,' 'Messiah,' 'Holy Spirit,' 'Jesus-Messiah' and 'Holy-Jesus'" are transcribed (1951, 118). Saeki writes, "Judging from these few decisive evidences taken out of many other proofs, it is natural that we should conclude that [The Messiah Sutra] is older than [On One God] which [was] written in 641 A.D." (1951, 121)

However, these observations are shaky in essence. They rest completely upon the premise that translations go through a process from 'bad' to 'good' and that such a progressive process could shed light on the order of precedence of the sources. Take for instance Saeki's reliance on the translations of the word 'God' to discern which source is earlier than the other. In the two sources, 'God' is translated differently. In On One God, 'God' is yishen 一神, which literally means 'one god.' In The Messiah Sutra, 'God' is fo 佛, a Chinese translation of 'Buddha.' Dealing with these different translations, Saeki (1951, 119) asks:

What does this mean? We believe that this fact clearly shows us that the Nestorian missionaries in 641 A.D. came to know for the first time that the Chinese word "I-shen" (一神) was a better word for the Nestorian document than the word "Buddha" [fo 佛] to designate God, whilst, on the other hand, it shows that the Nestorian missionaries at Ch'angan did not even known of the proper word for God when they composed the Takakusu Document [The Messiah Sutra].

In this quotation, Saeki shows that he honestly believed that yishen was a "better" translation of God than fo. He is assuming that Tang Christian translators went through a progressive process. His claim is that these translators initially used a 'bad' term fo in The Messiah Sutra, and only later became aware that yishen was "proper." Therefore, Saeki argues that The Messiah Sutra was made earlier than On One God.

However, this premise of progressive translation is problematic. It is a rigid mechanical understanding of translation. Recent work in Translation Studies has shown that translation is not merely a linguistic operation, the results of which can be evaluated on a linguistic basis. "Translations are not made in a vacuum. Translators function in a given culture at a given time" (Lefevere 1992a, 14). Translation is a complex dynamic process involving many factors (client, patronage, purpose, and the like), all of which influence translators in one way or another. For instance, the translation project can easily be derailed by the patrons, and certain expressions could be modified in order to suit the dominant ideology in a particular culture. A translated work, as Hermans (2014), Lefevere (1992b), and other scholars confirm, is a manipulated or rewritten product, and there is no ultimate standard by which to evaluate that one translation is necessarily better than any other.¹³

Even if one sticks to the traditional view of translation and believes that a translation can be evaluated in terms of linguistic accuracy, to judge that one ancient translated term is more appropriate than the other is still very tricky. Not only should one be careful about the standard used but one also has to be very familiar with the whole translation process. So far, however, little is known about how Tang Christians translated their religious texts. Processes such as quality control, editing, and selecting original texts are completely shrouded in mystery. Except for the above two brief records, the Christian missionaries' translation activity has never been documented by any other sources.¹⁴ In a word, there is virtually nothing that would allow us to suggest that Tang Christian translators "came to know for the first time" the term yishen is "better than" the term fo.

Doctrinal Observations

In the part on doctrinal observations, Saeki revisits the concept of the emperor and ancestor worship. He also takes into account Christian precepts for daily life and social morals, Christology, and several other points. He notes that what is stressed in one source is not stressed in the other:

We find that no such social morals or Christian precepts for daily life [contained in The Messiah Sutra] were taught emphatically in [On One God]. But in [On One God] the doctrine of Monotheism was emphasized, together with the story of the Fall of Adam, besides the necessity of the vicarious death of the Messiah as well as His death on the Cross, and His Resurrection and Ascension, and even the story of the Pentecost were stated. No such statements were found in [The Messiah Sutra].

(1951, 123)

Therefore, Saeki (1951, 123–24) alleges:

We, therefore, can not imagine that the comparatively difficult theological documents owned by Mr. Tomeoka [On One God] were written before the Takakusu Document [The Messiah Sutra] which gives the outline and general view of Christian precepts for daily life.

However these doctrinal observations are ill-considered. Firstly, the decision about which doctrine is more difficult than another doctrine is a value judgment. As stated above, the results depend on the judgment criteria on which the evaluators rest. Secondly, these observations are not really connected to the question of which source is older than the other. It is simply a given fact that what is found in one source is not necessarily found in another. As long as any two texts deal with different topics, their contents should not be the same. Similarly, the substance of The Messiah Sutra cannot be found in On One God or vice-versa. They are two distinct texts, not two versions of one source. Therefore, what Saeki's observations do is list the given textual variations. They belong to a compare-and-contrast study that examines the differences between The Messiah Sutra and On One God. In essence, they have nothing to do with the time of the making of the two sources.

Conclusion: Implications and Speculations

Although The Messiah Sutra and On One God have been recognized as Aluoben texts since the early twentieth century, their common dating must be abandoned. The traditional argumentation in general is plagued with defects that range from the cursory and subjective treatment of the textual and historical evidence to faulty reasoning and unfounded presuppositions. In addition to the aforementioned motives for maintaining the traditional dating, the persistence of the conventional chronology is largely the result of the fact that the inaccessibility of the original manuscripts forces more recent commentators to reproduce the assertions of the first generations of experts.

Given this argument, it seems that we have to rethink the general conclusions about the seventh-century presence of the Tang church derived from the traditional dating of these manuscripts. Chronology, we must remember, is the very pillar of historical research. It anchors manuscripts in the flow of time, offering the proper time frame into which to fit studies. Dating is one of the most fundamental areas that have to be clarified before any general point is pressed. If a manuscript is wrongly dated, it is wrongly used. Cause is confounded with effect or vice-versa. Conclusions will be founded on false premises. Only with a sound timeline is it possible for inferences to be drawn from the two Chinese Christian manuscripts about the so-called Tang church and for an accurate overall historical sketch of Christianity in medieval China to be produced. In fact, until the chronological problem is settled, no satisfactory research, either historical or theological, textual or cultural, can be written.

The last question concerns the time these two manuscripts were produced. As previously stated, to propose an alternative dating is a task that cannot be completed within the limits of this paper. Based on some clues, however, we may speculate that The Messiah Sutra and On One God might be two of the latest ancient Chinese Christian sources.

The first argument that gives support to this speculation is related to the lengths of the different manuscripts. As demonstrated above, the two documents are much longer than any other known Chinese Christian text. Given this fact, we may wonder why the early Christians made such long texts, whereas their successors preferred to create very short ones. Admittedly, the early pioneers often faced serious challenges. In a new environment, they had to overcome changes in lifestyles, deal with unfamiliar situations, and respond to the pressure to conform. They might have concerned themselves with pragmatic matters more than with nourishing their faith. Intellectually, the creation (translation) of two such long sources, we must admit, would have required a considerable amount of theological training and, more pragmatically, financial support and linguistic capability. Moreover, there would have needed to be a community both stable and sizable enough to require Chinese texts to nourish their faith. That is, it is more likely that the first missionaries created short sources and the later Christians took advantage of their predecessors' experience and produced longer documents.

Secondly, extra information that endorses the above speculation can be gleaned from manuscript P.3847, which was made no earlier than the tenth century.¹⁵ This manuscript has three conspicuous parts. The first part is a verse, The Praise of Jingjiao Three Majestics. The second part is the other religious text, The Sutra of Reverence, which opens with a brief summary of the Trinity and then lists thirty-one saints, among them Luke and Mark, and thirty-five titles of documents allegedly used (translated) by the Tang Christians.¹⁶ The third part is this colophon in eighty-five-characters:

The above is the catalogue of the sutras. The headquarters in Daqin house 530 bu (volumes) of the sutras. All are written in Sanskrit on patra leaves. In the ninth year of the Tang Taizong Emperor, Great Virtue from the West Region, Aluoben, came to China and submitted a petition in his native tongue. Fang Xuanling and Wei Zheng [two ministers] announced permission for the translation. Later, Great Virtue of Our Teaching, Jingjing, was summoned [to translate the sutras]. [In total,] the above 30 bujian (rolls) have been made. The great majority all remain in their leather cases and are still not translated.

謹案諸經目錄大秦本教經都五百卅部並是貝葉梵音/唐太宗皇帝貞觀九年西域太德僧阿羅本屆 于中夏並奏/上本音房玄齡魏徵宣譯奏言後召本教大德僧景淨譯/得已上卅部卷餘大數具在貝皮 夾猶未翻譯

Given the last two parts of manuscript page 3847, it goes without saying that a post-Tang scribe knew about the Tang church and was familiar with the Tang legacy. At least, he clearly knew that Aluoben and Jingjing translated thirty-five texts, whose titles are listed in The Sutra of Reverence. In other words, The Sutra of Reverence should have recorded the titles of The Messiah Sutra and On One God, provided that these two documents had been made by Aluoben and sanctioned by Tang Taizong.

However, The Sutra of Reverence does not give the titles of the two manuscripts in question. Among the thirty-five titles listed, two can be matched with the headings of existing sources. One is The Mysterious Bliss Sutra 志玄安樂 經. The other is The Origin of the Origins of the Daqin Jingjiao 宣元至本經, which appears to be the shortened heading of the surviving text The Sutra of Origins of the Daqin Jingjiao. Given that the scribe knew some very specific details about their Tang predecessors, one may wonder why he excluded the titles of the two longest and 'earliest' manuscripts from the list of sources used by the Tang Christians but archived the titles of other shorter and 'later' religious sources.

Admittedly, the above speculation awaits to be substantiated by hard evidence. However, there is no good reason for today's researchers to use the conventional time frame of the 640s. Since a more reasoned timeline of the two manuscripts discussed in this article is a crucial matter to anyone interested in the early history of Christianity in China, particularly given that no other religious texts made in the seventh century have come down to us, the most urgent job for us is to establish more reliable dates. Hopefully, the recent exhibition and new photos, as well as this article, can inspire more researchers to investigate the chronology of The Messiah Sutra and On One God, which in all likelihood are not the earliest statements of Christian faith in China.