Portrait of a Community: Society, Culture, and the Structures of Kinship in the Mulan River Valley (Fujian) from the Late Tang through the Song

In this wonderful new book, Hugh Clark shows us the possibilities of a historical source that has not yet been fully exploited by Western students of Chinese society, namely the lineage genealogy. It has been used to some extent for studies of demographic history, but to my knowledge no one has used it to describe the history of local elites as a larger socio-historical phenomenon. On the basis of this remarkable resource in combination with local gazetteers, collected works, inscriptions and collected notes, Clark presents us with a detailed and fascinating description and analysis of the ways in which local families in one sub-region of southern Fujian from the late ninth to midthirteenth centuries used kinship constructed around ancestral worship and genealogies to form lasting social networks. A crucial aim of the author is to establish the local nature of kinship networks, even when elites established regional or court-level careers. Since this is an important question in itself and has drawn considerable scholarly attention in the field of Song studies, focusing on kinship issues is indeed fully justified. Here I wish to raise some supplementary questions and suggestions from another angle, not so much as criticism as in the spirit of taking the project further.

The focus of the book on the kinship networks of the Mulan River 木蘭溪 Valley in southern Fujian grew more or less naturally out of the author's research, since the most informative sources tended to deal with that particular area. This itself is also an important piece of information, since it implies that other regions around the valley not only provided less evidence on the issue of kinship networks but most likely also produced fewer such networks, since according to the author the production of genealogies was an important element of constructing lineages. The absence of information is a relevant historical fact that needs to be fully explained in connection with the material that forms the main body of this book.

The usual story about the opening up of Fujian, of which the present book tells an important part, is that it started relatively late. Although there was an aboriginal population, we know little of it, and from the Tang onwards immigrants from outside the province supposedly populated the region. The Mulan River Valley, too, was populated from outside, a process that started to speed up in the ninth century due to the rebellions and civil wars that plagued northern China in the second half of that century. Or this is the story that is told by local sources, including the lineage genealogies, and therefore also the story that is presented by the author.

At this point I think that the genealogies themselves should have received more explicit attention. They are, first of all, hardly uncontaminated Song sources, since we possess them only in Qing or even later editions. Rather than being reprints, they are new compilations. To what extent they reliably transmit earlier evidence or later editors have (re)constructed that evidence to fit later needs and interests could have been discussed more explicitly. People like Michael Szonyi or David Faure who have worked on the same type of materials in order to discuss the late imperial period are much more skeptical.¹

The genealogies studied by Clark tell, for instance, of the origin of many southern Fujianese families from the northern county of Gushi 固始, which is a type of migration story that we also find among the Yao 瑤 minority (who claim to come from the Nanjing region), the Hakka (who have claimed various origins, including the counties of the Yellow River plain from the nineteenth century onwards), the Pearl River delta elites (who claim to originate from Nanxiong in northern Guangdong), and so forth.² It is inspired by the historical events around the refugee band leaders Wang Chao 王潮, Wang Shenzhi 王審知, and others, who apparently came from Gushi during the political–military chaos of the late Tang and Five Dynasties period in order to try their luck in northern Fujian. Wang Shenzhi also became the object of a northern Fujianese temple cult, contributing further to the spread of local stories. Clark voices some skepticism about the larger claims (see pp. 46–47 and passim), but he still divides the lineages of immigrant families he studies into those that do belong in a category of people from the He'nan county of Gushi and those that do not (pp. 47–54, 54–66). Since the same type of lineage genealogies are used for both types of families, it is not clear to me why some of the claims are reliable and others are not.

To me, the story is largely the invention of southern Fujianese people who were inspired by the Gushi–origin of the former Kings of Min and used it to invent a respectably "Chinese" origin for themselves. I would have appreciated a more explicit discussion of the Gushi case in relation to other southern migration myths, all of which have to do with hiding the ethnic origins of southern groups by connecting them to the venerable counties of the Yellow River plain. The very need for such a myth shows that kinship is not the only story that binds, but that local families also found a need to connect to the north through even larger stories. It is also not the only myth, as is suggested by the equally unlikely and unproven story of one Li lineage from a collateral line of the Tang imperial family (pp. 39–40). If such stories are indeed, as I suspect, inventions, then Gushi families may just as well have been much newer than families without such a claim, since it is in the nature of such claims that they are intended to cover up doubts about the "respectable" origins of a family! Being an "old family" is first and foremost an ideological construction and not a simple historical fact.

Two criteria that might be introduced into such a discussion on the northern provenance of Fujianese elites are linguistic and genetic. Of course, such criteria—especially linguistic ones—can be and often have been tailored to fit one's prejudices. However, the languages spoken in Fujian surely do not reflect a northern Chinese language as it was spoken in the late Tang period. Similarly, the genetic evidence that is now available does not confirm the hypothesis that many Fujianese migrated to the south from northern China.³ If anything, all evidence points to aboriginal origins. This not only places the Gushi story in a different light but also suggests an answer to the possible question (not raised in this book) why lineage thinking was and still is especially popular in southern China, such as Fujian and Guangdong. After all, one of the more important local cultures that is still attested on and across the northern Fujian borders, as well as further eastward and in Guangdong, is the one nowadays known as Yao (or She north of Fujian). Their thinking is very strongly shaped by ideas of kinship—albeit with significant differences from the Han Chinese—and migration from the north.⁴ Even during the late Ming and early Qing, sources on the far southern reaches of Fujian and the north of Guangdong still speak of tribal peoples. We can therefore be sure that they were even more present in the late Tang and Song periods. The very fact that so few families claim to be of local descent is perhaps the most eloquent indication that there is something amiss with conventional migration accounts.

Clark has selected one specific region and one specific period to attain some unity in time and space. He does explain why the Mulan River Valley became the focal point of his investigation but never addresses why he ends his discussion with the late Southern Song. Especially for China below the Huai River, I have always felt that the succeeding Yuan period should be included when we deal with socio–economic historical topics. Instead, the late Yuan or early Ming is a much better cut–off point. A slightly longer period also means that one can see temporal trends more clearly and usually has better evidence—if the resulting period is coherent in terms of the aims of the investigation. The beginning of the period may be the immigration of outside families to the region (the dominant interpretation, shared by Clark) or just the acculturation of existing aboriginal groups into the Han Chinese framework. What is demonstrated in wonderful detail by Clark is that the newly forming elites were connected to commercialization and most of all to an improving agricultural infrastructure (for instance, pp. 75–77). This second connection should have received a bit more attention, since it is clear that several lineages played central roles in this respect, and it further undermines the story of a northern Chinese (Gushi) connection. After all, it is clear that people moved down from the hilly regions to the plains, which suggests some sort of control over and at least resistance to malaria (and anemia or the like), a disease that would otherwise have decimated them. Then they would have accumulated knowledge about irrigation systems in which water is regularly refreshed, mosquito larvae are eaten by the fish, and snakes and birds kept in or around the flooded fields. I find it highly unlikely that northern families coming from an agricultural system based on millet and irrigation from rain, wells, and maybe canals, could have been responsible for this type of improvement. The families around Wang Chao and other leaders from Gushi would have specialized in warfare to begin with, many of them would have died during the internecine warfare that followed their migration south, and the local population would have rapidly absorbed those who remained.

A second bone of contention seems to me to be the precise origins of lineage culture, which I would place in age–old Buddhist institutions and only secondarily in the creation of new Confucian traditions (which happened only after the rise of lineage thinking!). As Clark points out, the Minnan 閩南 region was called the "Land of the Buddha" in the twelfth century, at the very time that the type of lineage formation under investigation in this book was in full swing. Somewhat frustratingly (no doubt also for the author himself), most of the discussion in this chapter on Buddhist institutions really deals with nearby regions outside the Mulan River valley under investigation. The only truly local Buddhist institutions mentioned frequently are those where local lineages originated around some kind of religious worship (pp. 265–270, 277–281, 288–291). Clark also mentions Buddhist institutions that originated in someone's private house (pp. 175–176), usually belonging to a later lineage. In other words, there is a frequent early link between lineages and Buddhist institutions. Several instances were "grave cloisters," a type of institution that is so well–known from the same period in the Lower Yangzi region.⁵ I find this recurrent Buddhist connection even more significant because it appears in lineage genealogies, which are otherwise heavily prejudiced against Buddhist institutions!

The gradual trend away from lineages in a Buddhist context towards specialized institutions legitimated through the great classical texts definitely was not the transition towards secular institutions that Clark makes it out to be (pp. 273–274ff.). Ancestor worship is, after all, a thoroughly religious activity by whatever definition experts are wont to use.⁶ Clark classifies as secular a ritual in which graduates in the examination system announced their success in front of ancestors (p. 291). Furthermore, he downplays the fact that monks prepared the sacrifices and that the ritual took place in a Buddhist institution. However, the very same source that he quotes points out that monks prepared the sacrifices (vegetarian to boot!) and that the participants themselves had to burn incense and fast beforehand.⁷ When we consider that classicist texts always prescribe meat sacrifices and do not speak of burning incense, the Buddhist influence clearly remains significant in this instance.

In addition, I do not think that the new type of worship is "within" tradition and that the Buddhist context was in conflict with or even outside tradition (as stated on p. 272). After all, there simply was not yet any tradition of lineage organization and classicist ancestor worship before the Song period. On the contrary, I would argue that it is the Buddhist context that was common and, in fact, is precisely what made lineage thinking plausible and popular to most people during the Song period. In socio–historical terms, this Buddhist context would have constituted the existing tradition, and the new type of worship a divergence from it. Instead, the classicist innovations of the Song were outside tradition despite their use of old classical texts: in other words, they constituted a new development—and eventually a tradition—that made Buddhism seem a distortion of eternal practice. If classical texts had really been at the origin of these lineage customs, they should have given rise to lineage practice already in the Han, which they patently did not. It is therefore crucial to take the Buddhist context seriously in understanding the origin of lineage construction in the Song period. It is also only in a later period that lineages could divest themselves completely of their frequent Buddhist origins, but this was the reinvention of tradition and not tradition itself.⁸ We should beware of taking the Neo–Confucian claim of a classicist tradition going back to Confucius and Mencius and cut short by Buddhism as a serious historical analysis, rather than as the ideological and ahistorical statement that it really is.

To be honest, this Buddhist context is hardly surprising. Lineage thinking to support religious authenticity is well–attested in pre–Song Chan Buddhism and would have been the most widely available example for non–aristocratic families of the late Tang and onwards. This was especially true of Fujian, where there had never been such families. Buddhist practices similarly supported and furthered filial piety (think only of the Ghost Festival and the Mulian story). Buddhist funerary and commemoration rituals for the deceased, sometimes over several decades, strengthened the sense of relatedness over time. Donations for Buddhist charitable activities often benefitted one's parents and grandparents. Thus, core Buddhist practices served to strengthen rather than to weaken the kinship bond. The later antipathy of cultural elites for Buddhist institutions, partly enshrined in that wonderful, yet biased, resource of the lineage genealogy, should not blind our eyes to the crucial role of these institutions in earlier periods.

As Clark is no doubt aware, religious institutions have remained of crucial importance in the structuring of local society, as is demonstrated by the ethnohistorical studies carried out by Ken Dean on roughly the same region for the late imperial period. Although this no longer falls within the period selected by Clark, some reference to this project would have been in order for the benefit of the general reader.⁹ After all, Dean has demonstrated quite clearly just how important cultic institutions outside the lineage continued to be in developing the region agriculturally. It seems highly unlikely that this was different in the Song period, and using that insight to interpret the much more disparate evidence from the earlier period might have been a useful exercise. The ritual and quasi–historical construction of kinship was part of a larger repertoire for connecting with people, to which different types of religious institutions continued to belong. Even for elite families the lineage could not entirely replace the need to participate in other religious types of organization. Looking at the lineage therefore also requires us to look at other local socio–religious organizations. I therefore look forward with great anticipation to Hugh Clark's upcoming project of a longitudinal study of religious institutions in Fujian province.

Finally, I would like to conclude with the hope that more studies like this will be carried out for other regions. Even now, we still do not have one single sustained study of lineage institutions in the Lower Yangzi region in the same period—in any period for that matter. Furthermore, we sorely need a comparative analysis of the social histories (plural!) of the lineage in different parts of Fujian, Guangdong, and other regions.