Go to previous page, China Resources main page, Jordan's main page.
This page requires Unicode.

Content Created 861225

Xúnshòu 巡狩 and Jìnxiāng 進香:
Two Kinds of Chinese Religious Processions
& Their Sociological Implications

(Corrected Version)
David K. Jordan

Abstract. This article examines the general proposition that religious activities often have the social-structural consequence of contributing to peacekeeping. Using ethnographic data gathered in Taiwan, three kinds of regional festivals are discussed: (1) those centering on an temple organizing villages around it, (2) those that rotate responsibility through a fixed set of sectors (both of these associated with a tour of inspection by a god around his territory), and (3) those involving pilgrimages from one temple to another. The three types of festivals are all found logically to affect regional social integration, but in contrasting ways. The implications of the differences for the role of temples as local peacekeeping political forces in Taiwan history are discussed.


  1. The General Problem
  2. The Specific Problem: Táiwān
    1. Types of Festival Organization
      1. Type I: The Tour of Inspection From a Central Temple
      2. Type II: The Tour of Inspection Through Rotating Sectors
      3. Type III: The Pilgrimage
  3. Conclusions
  4. Appendix: Patron Gods of Emigrant Groups
  5. Bibliography

Publication History. This paper was presented at the Second International Conference on Sinology, at Academia Sinica in Nángăng 南港, Táiwān 臺灣, December 29-31, 1986. A version was published in Proceedings of the Second International Conference on Sinology. Taipei: Academia Sinica (pp. 255-270).

Changes in This Edition. The conference for which this paper was written occurred shortly after the Republic of China issued new government guidelines on the spelling of Chinese words in Roman letters. In the charter to the committee that created the system, it was provided that (1) it must not be the same as the Hànyŭ Pīnyīn 漢語拼音 system used on the Chinese mainland and internationally, and (2) it must not include tone in the spellings as had the former national standard (Gwoyeu Romatzyh). As finally enunciated, use of the new rules also included provision that Chinese place names would be frozen in the unsystematic "postal" spellings that were official in 1949, and that names of people would follow whatever spelling an individual preferred, or, failing that, should follow the older Wade-Giles system in preference to the new system.

Few people paid any attention to this absurd policy, but Academia Sinica is in the end a government agency, and my article, like others for the conference, was a victim of the policy, and was largely reedited to conform to a modification (!) of the new official standard. Hokkien terms were, happily, untouched, not being regarded as really Chinese, but Mandarin words were hashed to unrecognizability. I was sufficiently ashamed of the resultant idiotic mess that I never cited this paper or gave off-prints to colleagues.

I recently (2000) came upon an old computer file containing the original draft, with tonal spellings, although without characters, which had been written in by hand. The following version of the paper gives all Mandarin terms in Pīnyīn spellings. The original editors also deleted all Chinese characters from papers presented in languages other than Chinese. I have restored most of them here, although in a few cases I have lost track of them. I have made no other significant revisions and have not attempted to update the text or the bibliography.

Return to top.

Xúnshòu 巡狩 and Jìnxiāng 進香:
Two Kinds of Chinese Religious Processions
& Their Sociological Implications
David K. Jordan (Footnote 1)

1. The general problem.

1. Material for this paper was collected in field periods in 1966-68, supported by the (United States) National Institute for Mental Health, in 1976, partially supported by a grant from the Chinese Cultural Center of New York, and in 1984-85, during my tenure as a Language & Research Fellow of the Committee on Scientific & Scholarly Cooperation with the United States, Academia Sinica, Republic of China. I am most grateful for this financial assistance.

Because of its concern with a moral community, religion is often regarded as having peacekeeping as one of its by-products. Some religious systems directly teach that human cooperation and harmony are goals in themselves. However, there are indirect contributions that religion can also make to peacemaking that do not directly proceed from religious ideology. Although religious customs are presumably performed for their intended religious effects, it is nevertheless useful to examine their various unintended social effects. Human societies often maintain cooperative relations between and among different groups, in part because of customs that do not have peacekeeping as their intentional effect. In this paper I shall be concerned with the non-religious, unintended (but often well appreciated) local political effects of religious activity. (Footnote 2)

2. For a discussion of the theoretical implications of the distinction between an actor's intentions with respect to the effects of his actions and his cognizance of these effects, see Spiro 1961 and 1966.
I conceive of the general model this way:

1 Large-scale activity requires broad participation.

Large-scale religious activity, requiring participation of large numbers of people, presumes willingness on the parts of the participants to cooperate with each other at least minimally. (Differently structured activities require, of course, different degrees and kinds of cooperation.) I envision a large regional festival, with many villages participating, as a typical example.

2 The threat of non-cooperation produces anxiety.

The threat of non-cooperation by some participants causes the other participants anxiety. There are three reasons why I think this is so.

3 Anxiety leads to mediation.

Anxiety caused by threatened non-participation leads to attempts at intervention and persuasion to prevent the uncooperative groups or individuals from disrupting the activity by non-participation. These efforts, to be effective, often must confront the underlying dissatisfactions that led to defection in the first place. When those dissatisfactions are group conflicts of one kind or another, the intervention by co-religionists produces dispute mediation. (E.g., village A refuses to cooperate with village B because of a feud over the theft of a cow; and village C proposes mediation or supports one or the other party in order to gain or force the continued participation of both villages A and B in an annual harvest festival.)

4 Structure has effects.

The extent to which this peacekeeping effect is likely to occur depends upon the way in which a religious activity is structured. For example, a festival which requires participation of all villages in an area as a condition of its performance is likely to have more peacekeeping effects (because it is more susceptible to disruption), than a festival which is merely grander (or jollier) when it has more participants than when it has few.

5 Religion not the only peacekeeping custom.

Religion is not the only body of custom that probably has peacekeeping effects, and religion can also inspire inter-group conflict. Accordingly the above propositions should be understood as tendencies implicit in religious customs, not as inevitable or unexceptional.

Return to top.

2. The Specific Problem: Táiwān.

Following this logic, I shall explore the proposition that one unintended social effect of traditional local festivals in Táiwān may have been to reduce conflict and promote cooperation. This was not, however, their usual intended function (which was religious), and of three different types of religious festivals I shall discuss, only two arguably had significant peacekeeping results.

Terminology. The term bàibài 拜拜 (Hokkien: pàipài), as popularly used in Táiwān, has two quite different spheres of meaning: As a verb, it usually refers to worshiping, and everything pertaining to worshiping. As a noun, it normally serves as a generic term for a religious festival. A local religious festival, in other words, in addition to its religious significance, is often an opportunity to invite friends and relatives from outside the celebrating district to come and dine, and bàibài is therefore also used to refer to the broad-scale entertaining and feasting associated with a religious festival. It is possible to say, "Tomorrow we are going to eat bàibài." Traditionally at least three major social functions were served by such festivals, aside from their religious functions:

1. Reciprocal obligations were maintained with matrilateral and uxorilateral kin, who normally lived outside of the village or neighborhood. These ties could be called upon in times of economic or social necessity, (Gallin 1960) and their cultivation and maintenance was especially important in cases of village exogamy. (Footnote 3)

3. Village exogamy was more common than not in Táiwān, in part because of a desire to cultivate ties to a wider social network, and in part as an effect of the combination of surname exogamy and the well-known tendency of villages to include large numbers of people with the same surname

2. The social networks of individual guests were expanded as they met other guests whom they had not known before.

3. Because whole communities feasted at once, whole communities were in a position to gain and lose regional prestige, and were therefore (I maintain) motivated to cooperate in producing a good village or town bàibài, subordinating local disputes.

This third function --motivating broad community support to reduce conflict-- is what is of interest for present purposes. For several years I have been collecting material to try to see whether local festivals might indeed have had a significant "peacekeeping" function in traditional Taiwanese society, particularly during periods when regular government failed to provide sufficient peacekeeping resources. At first, it seemed to me that this was most likely in cases where a central temple was organizing festivals, and therefore clearly appointed temple "officials" had a direct responsibility for the organization of labor for a festival. However, the existence of large festival systems, particularly in northern Táiwān, which do not focus on large temples, convinced me that the key issue is not necessarily the temples and their administrators, but the distribution of responsibility to participate.

This paper deals with some aspects of my model for Táiwān, as it has been emerging from my case material over the past few years. My data are still very rudimentary, as is also my understanding of Táiwān local history. The present effort is therefore necessarily more in the nature of a progress report than of a finished statement solving the problem I have set myself.

Return to top.

Types of Festival Organization.

There were several different ways in which local feasting could be organized, some of which still operate in some areas. (Footnote 4) I shall differentiate three general types of festival organizing, which I name after the temple activities usually associated with them: One is referred to as a "tour of inspection" (xúnshòu 巡狩). The "tour of inspection," moreover, has two variants: the central-temple model, and the rotating-sectors model. The other type of festival organizing is the "pilgrimage" (jìnxiāng 進香). (Footnote 5)

4. Some local festival systems continue to function, although diminished in scale; a few have apparently expanded; many have been discontinued, but recently enough to be well remembered; and some vanished before the memory of living informants. In general I have used the past tense when I have reason to believe that the system I am describing has ceased or seriously declined, and the present tense when it continues apparently in good health. The distinction is not rigorous, obviously.

5. The usual referent of each of these Chinese terms is religious, not essentially organizational. I have used them here as convenient handles for types of popular organization of festivals, and to the best of my knowledge the forms of organization described are perfectly correlated with the types of religious activity that the terms label, but it is important to stress that for present purposes it is the way in which festivals are organized, not their religious significance that is at issue, despite the religious terms.

Return to top.

Type I: The Tour of Inspection From a Central Temple

Background example: Xīgăng Jiào rites. In the Xīgăng 西港 area of Táinán county 臺南縣, a triennial Taoist jiào rite is held, centered on the Qìng'ān Gōng 慶安宮, a large temple dedicated to the goddess Māzŭ 媽祖. Local convention has it that the temple is visited by three plague kings (wēnwáng 瘟王) at each festival, and to entertain them appropriately, a temporary pavilion, the wángfŭ 王府, is erected in the main hall of the temple. The statue of Māzŭ is moved to one side, and the central altar of the wángfŭ is used to display temporary statues of the visiting divinities. (Footnote 6)

6. The Xīgăng jiào has been studied by LIÚ Zhīwàn [Liu Chi-wan]. See LIÚ Zhīwàn 1983: 285-400. For other work on jiào rites in Táiwān in general, see other chapters of the same work and LIÚ Zhīwàn 1974. On the Qìng'ān Gōng itself, one of the oldest Māzŭ temples in Táiwān, see GUĀN Shānqíng 1980 (v.3): 290-293.
In addition to their display in the temple, emblems of the visiting divinities (and of Māzŭ) are carried in procession through the system of participating villages, in what is called a "tour of inspection on behalf of Heaven" (dài Tiān xúnshòu 代天巡狩).

Processions. Each day for three days the procession proceeds from the temple, through a group of villages, and back to the temple. Thus there are three groups of villages, one for each of the procession days. On the day when the procession passes through a village, that village provides a feast, normally attended by large numbers of people from villages passed by the procession on the other days, but also attended by friends and relations of the host families. (Footnote 7)

7. The feasts given on the same day are, of course, in competition with each other for guests, but I lack the data to examine the social results of this.

A feast on the day that the procession passes through a village is an integral part of village participation in the jiào. Other parts include offering a palanquin of josses for the procession itself, with bearers and attendants, placing village josses in the Qìng'ān Gōng, providing financial support to the Qìng'ān Gōng, etc. In principle there are seventy-two participating villages (not quite corresponding with official administrative units), but it depends how one counts. In most years the effective number is smaller than that. (Footnote 8)

8. In 1985, for example, there were processional elements from fifty-six numbered, distinctive communities. Furthermore in many years one or another village declines to participate A few satellite communities of emigrant families may also send processional palanquins without being counted among the privileged seventy-two.

Opting out. There may be several reasons for a village declining to participate. Perhaps the most common one given is that the village cannot afford it because of recent financial strains caused by some massive public works project (repaving the roads, say, or building a new temple). Another reason commonly given is dissatisfaction with the administration of the procession (such as dissatisfaction with the order of march or with the parade route). Both of these reasons for non-participation are widely viewed as mere pretexts. The real reason, it is always muttered, is rivalry between villages, rivalry between political leaders of different villages, or dissatisfaction with festival arrangements that do not seem to pay proper deference to a village. Placing the village in a less favorable position in the order of march than a rival village is a good example. Villages with good relations between them do not object to any particular order of march. Those with poor relations can become quite sensitive to order-of-march considerations, since the processional order, established by the central temple, is construable as a political ranking of regional villages.

Thus we see that in Xīgăng the system of inter-village feasting is intimately linked to a very flexible and constantly changing processional system, in which different villages participate in different years, and in which the general organization is under the control of a central temple. The jiào enjoys considerable prestige in the Táinán region, and people privileged to live within the processional circuit are usually proud of their participation.

Competition. There is another factor that must be considered when discussing Xīgăng's jiào. Other regional celebrations in adjacent areas are in competition with it, both for the prestige of being a prominent south Táiwān religious center, and potentially for the participation of some of Xīgăng's participating villages. The most noteworthy of these competing temples at the moment is the Shèngmŭ Miào 聖母廟, located in Tŭchéng 土城. (Footnote 9)

9. Tŭchéng 土城 (Hokkien: Thớ-siân-á) is now officially Chéngdōng Lĭ 城東里 of Ānnán 安南 district, Táinán city. Māzŭgōng 媽祖宮 is now officially Xiăngōng Lĭ 顯宮里 of Ānnán district). They are two rival communities within the historical settlement of Lù'ĕrmén 鹿耳門. Each of the two claims that its temple (respectively the Shèngmŭ Miào 聖母廟 and the Tiānhòu Gōng 天后宮) maintains the original cult (and holds the original statue) of Māzŭ, brought by the hero Zhèng Chénggōng 鄭成功 (Koxinga) to Táiwān in the seventeenth century when he landed at Lù'ĕrmén to attack the Dutch colonialists by land. In general, the claim of Tŭchéng is widely recognized as legitimate, while the claim of Māzŭgōng is, in my opinion, more convincingly documented.
Tŭchéng was formerly part of the Xīgăng jiào circuit, but, taking offense when their venerable statue of Māzŭ was treated with less respect than a statue borrowed from the prestigious Māzŭ shrine of Bĕigăng 北港, the people of Tŭchéng withdrew from the Xīgăng festival, founded their own copy of it, (which is celebrated a couple of months earlier) and proceeded to recruit nearby villages into their own festival.

Observations. The details of this fascinating conflict need not concern us here, but the point is that Xīgăng is always potentially at risk of "losing" villages into a competing festival system. It follows that it is in the interest of planners and elders of the Qìng'ān Gōng to try to mediate inter-village disputes that may lead to non-cooperation and withdrawal of any village. Not only is the prestige of the village potentially damaged by its inability or refusal to participate, but the prestige of the Xīgăng festival system itself is also damaged somewhat by non-participation, and would be damaged even more by the loss of such a village into another system.

In the Xīgăng local history one reads little of the ethnic conflict found in northern Táiwān. The principle axis of conflict in this area was surname groups, in particular an alliance of groups associated with the Hwang surname, in competition with people named Guō and their allies (Jordan 1972: 18-26). The Qìng'ān Gōng is officially neutral in this conflict. Neither Māzŭ nor the visiting plague kings are patron gods of the contending groups, and both Hwang villages and Guō villages participate in the jiào and apparently always have. Still Guō -Hwang surname loyalties represent the usual fracture line in local rivalry, as well as the basis of much political factionalization even today. Each triennial jiào represents, to some extent, the triumph of regional interests over inter-village squabbling, and the cooperation between Hwang and Guō villages, factions, and individuals in order to produce the jiào reveals a clear peacemaking function played by the Qìng'ān Gōng.

10. It is my impression that the triennial revision of the order of march and processional route by the Qìng'ān Gōng has served to recognize and sanction differences in the size, wealth, and political power of the various participant villages, keeping ritual precedence in accord with local realities. My data are not yet sufficient to demonstrate this clearly, however.
If we try to isolate the features of this system that relate it to peacekeeping, clearly the key issue is that concern with prestige is a key issue. The possible loss of face for a village that does not participate motivates some peacemaking voices in individual villages, and the possibility of non-participation by one or more villages, thereby diminishing the glory of Xīgăng as a whole, motivates peacemaking efforts by central temple officials, whose own prestige would potentially also be damaged if they could not succeed in maintaining the willing participation of most or all of the villages. (Footnote 10)

Return to top.

Type II: The Tour of Inspection Through Rotating Sectors

Background Example: Sānzhānglí Hamlet. A somewhat different arrangement occurs in the procession of the god Băoyí Dàifū 保儀大夫, who is worshipped in the eastern districts of the Táibĕi 臺北 basin. (Footnote 11)

11. Local informants insist that Băoyí Dàifū is not the same as Băoyí Zūnwáng. Folklorists more often suggest the opposite (e.g. ZHŌNG Huácāo 1979: 304-310). Zūnwáng is often associated with Xŭ Yuăn 許遠, Zhāng Xún 張巡 (709-75), martyred warriors who fought against the rebellion of ĀN Lùshān 安祿山, and one or both of whom who may or may not have been titled Băoyí Dàifū in the Táng dynasty, and retitled Băoyí Zūnwáng in the Sòng. See QIÚ Dézāi 1979:87, LÍNG Zhìsì 1985 (v.3): 119, WÚ Zhāomíng 1984: 117-121. For present purposes the two divinities must be differentiated, since their cults are distinct.
Băoyí Dàifū, like the Māzŭ of Xīgăng, also makes a tour of inspection (xúnshòu) of his territory. Yet the organization of festivity is slightly different. Let me take as an example a festival system located in a place called Sānzhānglí 三張犁 in the southern half of Sōngshān 松山 District. This region is divided into six festival sectors (Hokkien: kak-thâu). Each year a statue of Băoyí Dàifū is borrowed from a shrine in another region and brought to Sānzhānglí, where it is carried in procession through the six sectors. The center of worship rotates through the sectors each year, with each sector in turn taking responsibility for providing a place for the worship and providing a feast for the other five sectors and for such other guests as people in the host sector wish to invite. Informants insist that it makes no difference where in the host sector the festival is centered, so long as there is enough space for sacrifices to be set out. However, each sector has a land god temple (tŭdìgōng miào 土地公廟), which is normally commandeered for the purpose. (Footnote 12)
12. -Land god temples in this region are now generally quite large, comparable in size to much more "important" temples in the south twenty years ago. I do not know whether northern Táiwān in general, or Táibĕi in particular, has always tended to have larger land god temples than the south. With the exception of a Buddhist monastery, other temples are wanting in Sānzhānglí.

Ethnicity. Băoyí Dàifū is the particular patron of people from Ānxī 安溪 county in Fújiàn 福建 province. The vast majority of the inhabitants of Sānzhānglí are (or consider themselves to be) descended from Ānxī immigrants to Táiwān. In general, particularly in northern Táiwān, early immigrants tended to block themselves into alliances on the basis of place of origin on the mainland. In the literature on Táiwān, these differences are usually described, for better or worse, as differences of "ethnicity." Historians differentiate three large "ethnic" blocs: Zhāngzhōu 漳州, Quánzhōu 泉州, and Hakka or Kèjiā 客家. (See appendix.) Ānxī people tended to affiliate with the Quánzhōu group.

The six sectors (largely of Ānxī people) in the Sānzhānglí festival system were not the only worshipers of Băoyí Dàifū. Other similar festival systems also can be found in this general region. Some indeed are quite large, so that the whole of the present Dà'ān 大安 district, for example, was effectively one sector in a quite large system covering much of the center of the city (although the area was of course rural when the system came into existence). In this case, each of the sectors was itself subdivided into sub-sectors, and one sub-sector of each sector would take responsibility on behalf of the whole sector when the turn of the sector came around. Since different sectors were divided into different numbers of sub-sectors, the actual frequency with which an individual family had a financial or hostly responsibility was different in different parts of the system, although the financial burden in those years was heavy. (Footnote 13)

13. My data are not complete enough to establish the point, but my strong suspicion is that most of these sub-sectors, like the sectors in Sānzhānglí, had land god temples that served as focal points of the rites, and that the distribution of land god temples in Táibĕi today may still largely reflect the old system of festival sectors and sub-sectors, with larger temples preempting the land god temples in areas where larger temples existed.

14. The lower density of temples in northern Táiwān may help to account for the additional interesting fact that the touring gods were often not permanently enshrined in the festival districts themselves; rather they were borrowed from shrines of much wider clientele.

A significant aspect of most of these systems is that the people participating in them were self-consciously members of the same place-of-origin groups, and the patron divinities selected were often patrons the immigrants' home districts in Fújiàn. Thus the festival of Sānzhānglí was as much a celebration of common origin from Ānxī as it was a festival of common residence in Sānzhānglí. This stress on "co-ethnicity" rather than "co-locality" in northern Táiwān is not surprising when we reflect on the history of inter-ethnic feuding there a century and a half ago. (Footnote 14)

Other examples. The fact that these festival systems were based on ethnic or kinship links that are also of potential local political significance is underlined by some of the variations that occur in the general picture. Here are two examples:

Băoyí Zūnwáng. The cult of Băoyí Zūnwáng 保儀尊王 (as opposed to Băoyí Dàifū) was brought to northern Táiwān, perhaps not for the first time, by three immigrant families named Gāo , Zhāng , and Lín . The alliance of the three families was not tight, however, and eventually they split the cult, assigning cult paraphernalia by lot to the three surname groups. The censer and the cult of Băoyí Zūnwáng's wife went to the Zhāng and Lín surnames; the statue went to the Gāo surname group. The Jĭngmĕi 景美 district saw a separate Gāo surname temple for Băoyí Dàifū by 1867 (presumably the predecessor of the present one, dating from 1894), and Lín surname temples, in Mùzhà 木栅 district in 1884 LÍN Héngdào 1974: 78) and in Jĭngmĕi in 1921 (LÍN Héngdào 1974: 67), presumably the predecessors of the Lín-dominated temples to Băoyí Zūnwáng now in Pínglín 坪林 township of Táibĕi county. (Footnote 15)

15. Boundaries have shifted, making some older addresses hard to identify. For the cult of Băoyí Zūnwáng at Jĭngmĕi, see LÍNG Zhìsì 1985 (v.3): 117-119. The Gāo temple was eventually converted into a public one when the Gāo surname group found a tax advantage in doing so, and some individuals of other surnames are now prominent among its worshippers, but I was told that the registered congregation of 500 or so is made up entirely of Gāo's, and that the most prominent members of the board of directors are Gāo's. The Lin temple was apparently the Băoyí Zūnwáng temple of Pínglín, in Táibĕi county, also said to have been founded in 1923, or was its parent. Additional Băoyí Zūnwáng temples in Shídìng 石碇 township in Táibĕi and in Táizhōng 臺中 county appear to be later. See QIÚ Dézāi 1979:87-89. (His locating the Jíqìng Gōng 集慶宮 of Shídìng in Táizhōng 臺中 appears to be a printer's error.)

16. The temple to Kāizhāng Shèngwáng in Nèihú was founded in 1806 to commemorate a miracle. A group of recent immigrants from Zhāngzhōu, the story goes, were fleeing a group of local bandits and were in a desperate situation atop a hill. They prayed to Kāizhāng Shèngwáng, and the boulders began to fall down the hill, seeking out the bandits one after another, until the Zhāngzhōu immigrants were left in peace. Three stones on this site are worshipped as manifestations of the god and his two subordinates, Mă and Lĭ . A large temple and pilgrimage guesthouse have also been constructed on the front of the site.

What is of interest for present purposes is that the division of the cult also brought the establishment of completely separate festival systems, with every evidence being that the change was very methodically worked out.

Nèihú district. The general idea of rotating responsibility among sectors, with its implication of locking the sectors into the general commonweal, is not limited to festivity, but is a broader organizing principle in Táiwān society. In the Nèihú 內湖 district, north of the center of the city, two different systems dominate the Nèihú basin, including slightly different territory. One is a five-sector system associated with Kāizhāng Shèngwáng 開彰聖王, patron of the Zhāngzhōu group, and patron of the region's most prominent temple. (Footnote 16)

It includes only Zhāngzhōu people, and is used as a device to ensure the rotation of the office of lúzhŭ 爐主 or "censer master" among all geographical parts of the basin, while limiting it to Zhāngzhōu immigrants (which, in this district, traditionally meant excluding a small group of Tóng'ān 同安 immigrants and some Hakkas who had settled here). The same temple is at the center of a (recently organized) rotating feast system composed of six sectors and associated with the goddess Māzŭ, whose statue occupies a subordinate place in the same temple with Kāizhāng Shèngwáng; this system is slightly narrower in geographical spread than that for Kāizhāng Shèngwáng, but aims to be ethnically "blind." The sectors of the two systems do not correspond with each other, and presumably each system produces solidarities of a slightly different character from the other. (Both exclude the Tóng'ān immigrants of the region, who participate in a separate, much older Māzŭ festival centered on a Māzŭ temple patronized principally by Tóng'ān people living in two neighborhoods within the district. Even Māzŭ is not entirely above ethnic associations, even though the Kāizhāng Shèngwáng temple uses her that way.)

Thus we see that the same temple can be the center for more than one system of sectors, depending upon the cult, and that, as in the case of Băoyí Zūnwáng, cults may interpenetrate spatially but be distinguishable on the basis of the ethnic or kinship (surname) identification of the majority of the participants. (Footnote 17)

17. In my experience, few Táiwān organizations are really exclusively of one surname or ethnicity, since all are effectively open to "outsiders" friendly to the principal group. "Purely" Tóng'ān or Zhāngzhōu organizations almost certainly also included non-Tóng'ān and non-Zhāngzhōu individuals. I have argued that the same permeability is characteristic of village-based religious systems in the Xīgăng area (Jordan 1972: 1-26.

Some observations: We saw that in the central-temple model, there is no significant inter-village cooperation uniting the villages who constitute one day's procession as opposed to another's, and the villages themselves have no responsibility for the organization of the festival, so that the defection of a village or group of villages, while regrettable, does not seriously disrupt the festivity. Instead, the participation of individual villages or districts in the festival is a matter of concern for administrators of the temple, who are motivated to seek to mediate local disputes in the interest of maximizing participation in the festival region. Additional pressure is added by the possibility that dissatisfied villages may withdraw from the festival permanently and affiliate with a competing temple system.

The rotating-sector model does not depend upon a strong central temple --indeed a central temple is often missing. On the other hand, and compensating for this, the allocation of all responsibility to each sector in turn means that the withdrawal of a participating unit in the rotating-sector model potentially disrupts the entire festival. Each sector must perform as a close-knit unit, and the defection of such a unit (or its malfunction due to disruptions within it) would mean the destruction of the ritual until a new order of rotation could be established. I have so far been unable to discover any instances of such disruption. It appears that the sanction of public expectation effectively renders defection impossible. (Footnote 18)

18. When change does occur, it must be negotiated on a region-wide basis. Thus there was formerly a large Băoyí Dàifū rotation system among three sectors located in Sānchóng 三重, Xīzhĭ 汐止, and Mùzhà 木栅. This was eventually discontinued, and each district became an independent system. Both Xīzhĭ and Mùzhà are on nine-year, three-sector festival schedules, small-scale models of the original larger one. Quite separately from this, there is a yearly pilgrimage between the temples at Xīzhĭ and Mùzhà acknowledging the historical relationship between the them. (I have no data about the Sānchóng sector, which appears to have been discontinued.)

On the other hand, the passivity of each sector in the years when it does not have direct responsibility means that in many respects the rotating-sector festivals are less coordinated and require less cooperation in their performance. Beyond a certain willingness to produce as effective a festival this year as last, this year's sponsor need bother little about the preferences of people in other segments, and the "defection" of one or more communities in a non-sponsoring segment could effectively go unnoticed, whatever might motivate it, since it would probably be marked by little more than a lower rate of participation on the part of people in the "defecting" sector.

The rotating-sector model does not provide obvious outside mediators of local conflict; and it requires less between-sector cooperation, and hence logically less dampening of between-sector conflict; but on the other hand it may put even stronger pressure on the contending parties to suppress within-sector conflict themselves, since disruption of the system would focus unbearable criticism on them from participants in all of the other sectors. What is lacking in leadership from a central temple is made up for in the greater severity of the sanctions on the sponsoring sector, but peacekeeping between and among non-sponsoring sections is almost certainly less effective, at least on a day-to-day basis.

Broadly speaking, the social functions of these rotating-sector festivals were not dissimilar to those of the festivity associated with the central-temple type of organization, instanced by the Xīgăng triennial jiào: Uxorilateral and matrilateral kinship links were strengthened; social contacts were made across wide geographical areas; and local areas gained prestige. In this case, however, the strong ethnic associations of the patron gods made such festivals into celebrations of ethnic identity, tending to strengthen (or if unsuccessful to weaken) the coherence of self-conscious ethnic groupings, while emphasizing the lines between ethnic groups. (This would be so even when individuals of other ethnic groups were invited guests, since the patron gods themselves were associated with ethnic groups, and since the villages or districts taking responsibility for the banquets and processions were acting as ethnically self-conscious groupings.) It is legitimate to ask whether the association between section rotation and ethnicity is an inevitable (or even particularly common) one. My impression is that ethnographically there is indeed a high empirical correlation between ethnic festivals and rotating sectors, but my survey data are still too few to be certain what exactly the correlation is. (Footnote 19)

19. Partly in order to establish the extent of the correlation between rotating sectors and the prominence of ethnicity as a principle of collaboration, I did some brief interviewing in Miáolì 苗栗 and Yílán 宜蘭 counties in 1985. In Yílán festival activity of the same kind either did not exist or has much deteriorated in recent years, and informants were able to describe no functioning or recently remembered systems of broad-scale festival activity. (Some temples had festivals, but they involved no rotation or subordinate temples.) In Miáolì the system, both in towns and in the countryside, seemed virtually identical with the situation in Táibĕi, but Hokkien and Hakka settlements maintained separate systems, tending to extend the correlation between festival and ethnicity suggested by the Táibĕi data. Anecdotal data from other areas strongly suggests that the rotating sector method of organizing recurrent responsibilities is a general Chinese approach to the problem, however, and has no inherent relationship either to religion or to ethnic differentiation. Ethnographic data on non-religious Chinese organizing suggests the same thing.
Logically, it seems to me that the principal point of contrast between the two models is not ethnicity as a basis for organization however. More likely the contrast is one between northern and southern Táiwān, which happens also to correspond with differences in the salience of ethnic difference.

Return to top.

Type III: The Pilgrimage

Definitions. The Chinese word jìnxiāng can be translated "pilgrimage." It refers, however, to several only loosely related things. Three of these interest us here:

(1) Offering incense in a temple is referred to as jìnxiāng, especially if one travels a long distance to reach the temple, and especially on a festival day. Pilgrimages may be made to important religious festivals, such as the jiào at Xīgăng (when a god whose image is carried from another place to the Qìng'ān Gōng is said to be making a pilgrimage) or such as the annual tour of inspection of Băoyí Dàifū in Sānzhānglí. Such pilgrimages are rather rare, however. More often, a pilgrimage is undertaken to a well-known shrine (Footnote 20)

20. I use the word "temple" as a generic term for a place of worship. A "shrine" contrasts with this as being either (1) a very small temple, or (2) a temple of such great scale or religious importance that it frequently attracts pilgrims from outside its own area.
on occasion of the anniversary of a temple founding or on occasion of a god's birthday, and this is done without regard to any tours of inspection or local festivities at the visited temple itself. The most famous pilgrimage center is no doubt the Cháo Tiān Gōng 朝天宮, the shrine of Māzŭ at Bĕigăng 北港, in Yúnlín 雲林 county; Indeed, the pilgrimage from Dàjiă 大甲, in Táizhōng 臺中 county, to Bĕigăng each year has become the popular model of the classic pilgrimage. (See DÙ Róngzhé 1984.) But a number of other shrines also attract large numbers of pilgrims.(Footnote 21)

21. Figures quoted in a 1985 almanac (WÉI Duān 1984: 264), for example indicate that the shrine of the Five Kings at Mádòu 麻豆, near Xīgăng, attracted 1,579,784 visitors in 1982, and the shrine to the same gods in Nánkūnshēn 南鲲鯓 3,765,773. By comparison, the National Palace Museum attracted 1,778,790 and Sun-Moon Lake 1,102,755 in the same year.

(2) Carrying an image and incense pot from a temple to the temple from which it is founded to acknowledge and renew its links with the parent temple is also referred to as jìnxiāng. Traditionally this reinforced lasting links between communities with historically linked temples. This was of potential social importance if, as often happened, the visit was between an emigrant community and its home community. It was less significant in the case where noted shrines had thousands of daughter temples, or where daughter temples were founded by private individuals, so that there was no particularly logical or historical link tying the communities together except for their common cult.

(3) Carrying an image and other paraphernalia from its home temple to a temple which is unrelated to it is also referred to as jìnxiāng. In recent years, Táiwān has seen a rapid growth in the number of such expeditions, often carrying land gods (tŭdìgōng 土地公) (!) and other local divinities on pilgrimage tours to well-known shrines in distant parts of the province. (Many of the Táibĕi land god temples sponsor such tours annually, selecting different shrines in different years. Many of the pilgrims are women, for many of whom the pilgrimage is at least partly a pretext for a merry romp to visit parts of Táiwān they have not visited before, while escaping from heavy burdens at home. Because different shrines are visited each year, I shall call these "shifting pilgrimages." It is clear that the ability to attract such pilgrimage groups, and thus to be widely acknowledged as a major shrine, is a mark of prestige for a temple. Therefore many new and newly rebuilt temples provide pilgrimage guest houses to facilitate such traffic; sometimes the guest houses are larger than the temples themselves.

Shifting pilgrimages: Large shrines. With a few exceptions, shifting pilgrimages matter to the receiving shrine because of the number of pilgrims and the distances they come, more often than because of long association with the sending temples. Many a shrine sports a wall filled with flags left by visiting groups of pilgrims, and particular pride is taken in flags from distant spots. Accordingly, the particular visits today rarely have significant effects on social relations between temples or their communities.

Shifting pilgrimages: Small temples. The more important effect is generally upon the social relations of the sending community alone. In this regard, we can see shifting pilgrimages as a response to changing conditions: A combination of government pressure and changes attendant upon industrialization has weakened most of the festivals that centered on feasts and feast exchange, so that many have been discontinued, or exist today but attract very little participation. (This is particularly true in areas of Táibĕi city that were formerly villages but now contain block after block of apartment buildings.) Further, the solidarities that such feasting fostered have tended to decline in importance as Táiwān society has evolved into one in which individuals' personal social networks are island-wide, or often even international. In such a context, annual pilgrimages, with participation open to anyone in a given religious district, represent a new kind of community activity, providing pious fellowship centering on a temple that comes to redefine itself as a congregation of pious people rather than a parish of co-resident neighbors. However in this case there is no-one --village, family, or individual-- whose participation is religiously or traditionally necessary. Although pilgrimage is clearly a kind of festival, it motivates no obvious reason for mediation and peacekeeping of the same kind that we easily saw in the tour-of-inspection activities. The receiving shrine is pleased at all the attention it can get, but a visit from one sending shrine is, in most cases and for most purposes, as good as a visit from another. The sending temple is proud to present a popular and meritorious activity each year, but the trip is patronized by only a small proportion of the target population, and little or no importance is attached to exactly who does and who does not go. Festivity (bàibài) it is; peacekeeping it is not.

Return to top.

3. Conclusions.

The Táiwān case illustrates the force of the general perspective than an unintended consequence of religious festivals can indeed aid in peacekeeping, not just during the festival period, but between festivals as well.

But examination of the Táiwān data suggests that festivals organized in different ways seem to keep the peace in different ways. The rotation of total festival responsibility through a fixed sequence of sectors of the celebrating region seems to operate to require the cooperation of the sector charged with that responsibility in a given year, but offers little constraint on the behavior of the other sectors in their years of non-responsibility. (It also seems highly correlated in my data so far with temples involved in ethnically affiliated divinities.)

In the case of festivals centering on a coordinating central temple, on the other hand, the managers of the temple seem to be motivated to seek to limit inter-participant conflict. Although the ultimate pressure on a given locale to sustain cooperation may be slightly less, the constancy of it is greater.

Pilgrimages, at least in their most common modern form, would seem on the other hand to have less local peacekeeping potential, although they may make an important contribution to peaceful relations across broader regions.

Return to top.

Appendix: Patron Gods of Emigrant Groups

This table lists towns in Fújiàn 福建 province whose emigrants in Táiwān tended to affiliate with each other, and the patron gods they typically selected.

Zhāngzhōu 漳州, or (Lóngxī 龍溪), immigrants under the patronage of a god named Kāizhāng Shèngwáng 開彰聖王, includes groups from towns located in the basin of the Jiŭlóngjiāng 九龍江 river and its tributaries, as follows (reading upstream):

Place Patron Divinity

Jīnmén 金門 county Kāizhāng Shèngwáng 開彰聖王
Xiàmén 廈門 city
Shímă 石碼鎮zhèn
 (Lónghăixiàn 龍海縣)
Lóngxī 龍溪 county
 (Zhāngzhōu 漳州 prefecture)
Huà'ān 華盦 county
Zhāngpíng 漳平 county

Quánzhōu 泉州 or (Jìnjiāng 晉江) immigrants include immigrants from towns located in the basin of the Jìnjiāng 晉江 river and its tributaries, as follows:

Place Patron Divinity

1. "Sānyì" 三義 alliance,
 consisting of:
Guăngzé Zūnwáng 廣澤尊王
 & Băoyí Dàifū 保儀大夫
   A. Jìnjiāng 晉江 county
   (Quánzhōu 泉州 prefecture)
   B. Nán'ān 南安 county Guăngzé Zūnwáng 廣澤尊王
   C. Huì'ān 惠安 county Líng'ān Zūnwáng 靈安尊王
 (= Qīngshān Wáng 青山王)

2. "Xiàjiāo" (??) alliance,
 consisting of:
   A. Ānxī 安息 county Qīngshuĭ Zŭshī 清水祖師
 & Băoyí Dàifū 保儀大夫
   B. Tóng'ān 同安 county Băoyí Dàifū 保儀大夫
 & Xiáhăi Chénghuáng 霞海城隍

3. Yŏngchūn 永春 county  

4. Déhuà 德化 county  

The Hakkas were generally devotees of Sānshān Guówáng 三山國王. A scattering of other groups, apparently either composed of Hakkas or allied with Hakkas most of the time, came from Huìzhōu 惠州 prefecture, Cháozhōu 潮州 prefecture, Jiāyìng xxx district, and Dĭngjiāo xxx, and tended to be devotees of Guānyīn Púsà 觀音菩薩 when they sought a patron divinity over their alliances.

Return to top.


DÙ Róngzhé 杜榮哲
1984 論『大甲媽祖的巡禮運動』對信徒神學的啟示。(On "the pilgrimage activity of the Māzŭ of Dàjiă" as a revelation of believers' theology.) In DŎNG Fāngyuàn 董芳苑 (ed.) 臺灣民間宗教信仰。(Popular religious belief in Táiwān.) Revised edition. Táibĕi: 長青文化事業。 Pp. 331-395.

GALLIN, Bernard
1960 Matrilateral and affinal relationships of a Taiwanese village. American Anthropologist 62: 632-642.

GUĀN Shānqíng 關山情
1980 臺灣古蹟全集。 (Historical monuments of Táiwān.) Táibĕi: 户外生活雜誌社。

JORDAN, David K.
1972 Gods, ghosts, & ancestors: the folk religion of a Taiwanese village. Second edition: 1985 Táibĕi: Caves Books.
(The third edition of this book is available on this web site.)

LÍN Héngdào 林衡道
1974 臺灣寺廟大全。 (Compendium of temples in Táiwān.) Táibĕi: 青文。

LÍNG Zhìsì 凌志四
1985 臺灣民俗大觀。(Overview of Táiwān folklore.) Táibĕi: 大威。.

LIÚ Zhīwàn 劉枝萬
1974 中國民間信仰論集。 (Collected essays on Chinese popular belief.) Táibĕi: Institute of Ethnology, Academia Sinica (=Monograph No. 22).
1983 中國民間信仰論集。 (Collected essays popular belief in Táiwān.) Táibĕi: 聯經。

QIÚ Dézāi 仇德哉
1979 臺灣廟神傳。 (Histories of temples & gods in Táiwān.) Dŏuliù, Yúnlín: Privately published.

SPIRO, Melford E.
1961 Social system, personality, and functional analysis. In B. Kaplan (ed.) Studying personality cross-culturally. Evanston: Row, Peterson. Pp. 93-128.

1966 Religion: problems of definition and explanation. In Michael Banton (ed.) Anthropological approaches to the study of religion. London: Tavistock. Pp. 85-126.

WÉI Duān 韋端
1984 實用百科年鑑,1985 年版。 (Practical almanac for 1985.) Táibĕi: 故鄉。

WÚ Zhāomíng 吳昭明
1984 神話,話神。 (Myths and speaking of gods.) Táibĕi: 臺灣新生報。

ZHŌNG Huácāo 鐘華操
1979 臺灣地區神明的由來。 Táibĕi: 臺灣省文獻委員會。

Return to top.