The Acquisition of Virtue-Power (de) and the Marginality of Hell

Abstract

The idea of hell entails a type of extramundane retribution, and such extramundane retribution is useful as a deterrent to antisocial behaviour. This functionalist view of extramundane retribution was, in fact, explicitly countenanced during pre-imperial China. Also, like many roughly contemporaneous pre-Christian cultures in western Eurasia, pre-imperial China had a notion of an “underworld”. For early China, then, the more relevant “problem of hell” might be this one: why does hell appear to be so marginal? This paper surveys the idea of hell in the pre-Qin and Han periods, and concludes that an answer may be found in the Ruist appropriation of Zhou ideas about acquisition of virtue-power (de) and the afterlife, which were promulgated as state orthodoxy during the Han. With the fall of the Han, however, this state orthodoxy crumbled, and the culturally-adaptive memetic power of hell reasserted itself in the interdynastic and Tang periods. During these periods, a mélange of Buddhist and Daoist ideas of hell more strongly informed popular belief, and the idea of hell was arguably thereafter marginal in appearance only.

1. Introduction

Belief in a site of postmortem punishment for wrongdoing committed during one’s mortal lifetime is a phenomenon that became widespread amongst various cultures, and particularly amongst large-scale post-agrarian civilisations of significant technological complexity and population density. From the infinite set of possible beliefs a culture could entertain about what happens to a person after death, it is prima facie surprising that this particular type of belief is one upon which so many cultures tended, historically, to converge. This calls out for an explanation. Evolutionary approaches to the explanation of religious phenomena such as these can be sorted into two schools of thought: the non-functionalism of classical memetics and the functionalism of cultural evolution.

Richard Dawkins is a well-known exemplar of the former approach (Dawkins 1976, pp. 192–93), that of classical memetics. For him, beliefs such as these are part of memeplexes hosted by human brains, and—like viruses—these memeplexes propagate simply because they can. They become well-adapted to human cognitive biases and contain their own mechanisms of propagation through generational indoctrination, but are not seen as evolving towards any functional end from the perspective of their host populations. They are just selfish memes, spreading purposelessly through their particular host substrates like viral jingles or catchy melodies.

Proponents of cultural evolution exemplify the latter approach (Wilson 2002; Norenzayan et al. 2013; Johnson 2016). For such theorists, bodies of religious doctrine co-evolve with their host populations, and—regardless of their veridicality as descriptions of some mind-external reality—tend in the long-term towards some adaptive functionality which confers a benefit on the host population. The evolutionary logic is easy to see here. Sincere belief in extramundane retribution will, in many cases, effectively deter types of antisocial behaviour which elude the regular mechanisms of mundane retribution practiced by some population, and this will redound to the benefit of this population. This is, in fact, borne out by the empirical evidence (Purzycki et al. 2016). As we shall see in the next section, a connection between widespread belief in extramundane retribution and the conferral of a functional benefit upon a society espousing such beliefs was explicitly recognised in pre-imperial China. Importantly, in evolutionary terms, just as aversion to pain helps steer organisms into heritable adaptive behavioural patterns which give them an advantage over competitors, so aversion to extramundane retribution helps steer populations into heritable adaptive behavioural patterns which give them an advantage over other populations, in turn ensuring the further propagation of those behavioural patterns.

The prima facie case for the cultural evolution of belief in extramundane punishment seems, then, to be on solid grounds. But what about the case for the cultural evolution of a belief in the type of postmortem punishment entailed by hell? Belief in hell has been theorised to be one of the strongest and most functionally efficacious forms of extramundane punishment produced by cultural evolution (Johnson 2016, esp. chaps. 1, 3, 4). As Johnson notes, postmortem punishments issued by a high god can often be significantly more effective than premortem punishments in that “they are certain, possibly worse than death, and infinite” (Johnson 2011, p. 79). Belief in postmortem punishment specifically results in stronger condemnation of immoral behaviour (Atkinson and Bourrat 2011). It correlates negatively with societal crime rates, and is much more effective at promoting prosocial behaviours than is belief in heaven (Shariff and Rhemtulla 2012; Johnson 2016, chap. 2). The prevalence of a belief specifically in postmortem consequences correlates positively with economic growth, whereas mere religious observance in isolation from this belief correlates negatively (McCleary and Barro 2006). Moreover, there are reasons to believe that these are causative rather than merely correlative relationships (see Johnson 2016, pp. 64–67), and likely to have been favoured by the same cultural evolutionary processes which favoured premortem extramundane punishment to begin with. Finally, and importantly for our case in this paper, the type of punishment thought to be meted out by the extramundane agents believed in by small-to-medium-scale societies has been found to be overwhelmingly premortem, whereas it is in large-scale societies that belief in postmortem punishment becomes common (Hartberg et al. 2014). All this leads one to wonder why—seemingly contrary to theoretical expectations—hell was so marginal in the Qin and Han periods.

Let us recall that during the Western (1045–771 BCE) and Eastern (770–256 BCE) Zhou periods, through the Qin (221–207 BCE) unification, and indeed for the duration of the Han period (206 BCE–220 CE), the large-scale Chinese civilisation had attained to a relatively advanced level of technological sophistication and complexity, in many areas very much the equal or even superior of other large-scale civilisations of western Eurasia and the Mediterranean. Many of these other civilisations converged upon the notion of a punitive hell. The emergence of “hell” and “Gehenna” in (respectively) Christianity and Judaism, for instance, was contemporaneous with the Han period, and these conceptualisations of hell had an unmistakably moralistic and retributive character (Launonen 2022), as did those that emerged later in the Islamic world. As foreshadowed above, when we move to civilisations of diminished technological complexity, conceptions of the afterworld tend rather to be more amoralistic and somewhat ambiguously retributive.1 The Aztec and Incan civilisations are instructive counterpoints here. Though one may marvel at the ingenuity of their social and governance structures and communications networks, their societies were not as technologically and socially complex as the Han,2 with a lower volume of anonymous interactions and long-range economic transactions, and neither had a strongly punitive notion of postmortem retribution, certainly not one that was moralistic in character.3 In precociously cosmopolitan, large-scale and technologically advanced ancient Egypt,4 however, by at least its twentieth dynasty (1189–1077 BCE, somewhat predating the Western Zhou)—and possibly as far back as its eighteenth dynasty (1550–1292 BCE) or even earlier—there was most certainly a strong notion of extramundane retribution in the afterlife (Van Dijk 2002, pp. 161–64).

For early China, then, there arises the following “problem of hell”: why is this notion of hell qua site of postmortem retribution, such a marginal one during the pre-Qin and Han periods? This question can be seen as an important subsidiary of a broader problem in the evolutionary anthropology of religion, namely, the problem of an apparent Chinese exceptionalism (Sarkissian 2015) in having bucked a trend towards “high gods”. As upscaling societies increase in population density and technological complexity (Henrich and Muthukrishna 2021), with a higher frequency of exchanges between mutually anonymous interaction partners, it is statistically likely that this will be accompanied by belief in a so-called “high god” (Norenzayan 2013), that is, one who takes an interest in human adherence to moral codes, and rewards or punishes accordingly. It would, of course, be a grave error to expect the specific form of postmortem punishment in every society to converge on something similar in every detail to that meted out in the Abrahamic hells. It has, for instance, been demonstrated that the omniscient and omnipotent (high) god/s of the various monotheisms are not a necessary part of this picture, and that “broad supernatural punishment” (Watts et al. 2015) of humans by a range of different morally-concerned extramundane agents (including but not limited to high gods) better correlates with societal upscaling.5 But given that, as we shall see in subsequent sections, such extramundane agents and conceptions of an underworld were already part and parcel of the standard ensemble of pre-Qin and Han mythology, we are quite entitled to ask: why did we not end up with a punitive underworld during the late Zhou, Qin, and Han periods? Why was hell so marginal?

The answer we proffer in this paper concerns the specific ways in which the Ruist (historically mistranslated as “Confucian”6) tradition construed and elaborated upon Zhou ideas about the relationship between “virtue-power” (de 德) and extramundane agents. Again, what we are trying to do here is precisely to explain the uniqueness of the Chinese pattern of extramundane punishment, identifying and explaining the reasons that the Chinese conception of hell followed the trajectory it did rather than expecting some necessary default to an Abrahamic model of hell. To briefly foreshadow the argument somewhat: the structure of Zhou-style ancestor worship and its connection to the idea of de virtue-power, for various reasons to be dwelt upon later, mitigated the need for the sort of punitive hell to be found in Buddhism and the later monotheisms. Then, subsequent to the social structure which typically undergirds ancestor worship (Sheils 1975), having given way to the more technologically complex cosmopolitanism of the fragmented and geopolitically fractious late Eastern Zhou period, the unifying Qin and Han imperia had their own reasons to give strong “artificial” support to ancestor worship. This resulted in the continuance of widespread beliefs about the afterlife which were not entirely amenable to postmortem punishment, beliefs which—if they did emphasise any sort of cosmic retribution or personal consequences based on one’s moral standing at all—emphasised that it was to take place in the here and now, and was (perhaps ironically) carried out precisely by disgruntled ghosts and spirits who themselves emerged from hell to do so.

It may also be salutary to clarify why we prefer the cultural-evolutionary approach over the classical memetics approach in interpreting the propagation of the cultural variant, which is “hell”. Firstly, if the “hell” cultural variant was the view of the afterlife which was just the best adaptive fit for essentially unchanged Pleistocene psychology shared by all Homo sapiens—as the classical memeticists would have it—then we would expect it to have spread a lot more broadly from its possibly second millennium BCE Egyptian origins into (for instance) Africa than it in fact did. The cultural-evolutionary approach, by contrast, allows that religious beliefs are responsive to the particular features of local social ecologies (Bendixen et al. 2024),7 and that the emergence of some theoretically-predicted cultural variant can often be countervailed by other features of the cultural environment. That is, in fact, precisely the wager of this paper with respect to “hell”.

Secondly, we need not think that cultural evolution can only be thought to be operative in cases of random variation (Henrich et al. 2008, pp. 129–31) such that only the ab initio appearance of the cultural variant constitutes “cultural evolution” whereas any further spread is “memetic transmission”. Cultural variants which benefit their host populations can equally evolve and spread through conquest—made easier by the putative benefits the variant entails—and also through prestige-biased copying and imitation. Of course, the correlation of morally-concerned broad supernatural punishment with social upscaling and technological complexity becomes more compelling in the absence of a linkage in cultural phylogeny, a linkage which may well indeed have underlain the traversal of the “hell” cultural variant across Eurasia. But strict independence and cultural isolation are not necessary features of the cultural-evolutionary model.8

The case we put together will hopefully be instructive for the functionalist view of both the evolution of hell specifically, and for this theoretical approach to the evolution of a package of doctrines regarding extramundane retribution more broadly, helping contribute to an explanation of why China seemed to have “bucked the trend”—if only temporarily. This perspective will be further sharpened by a brief examination of later developments in the evolution of the hell concept after the deposition of Ruism from its erstwhile position as state orthodoxy during the Han. The Ruist resistance to its evolution having being weakened, a punitive conception of hell subsequently moves back out from the margins of the popular imagination and reverts to the more typical pattern: the social utility of widespread belief in hell as a site of postmortem punishment drives it back in the direction of fixation in the post-Han intellectual climate.

2. Precis of Pre-Buddhist Hell

Let us sketch out a picture of the autochthonous conceptualisations of “hell” which were present in China prior to the introduction of its heterochthonous Indic counterparts towards the end of the Han period. Was there any pre-Buddhist notion of “hell” in China? Were any such notions firmly impressed on minds of the time as something to avoid at all costs, or were they somewhat marginal? In this special issue, Bony Schacter (Schacter 2025) has called upon Jonathan Kvanvig’s helpful list of theses, the satisfaction of which amounts to a “strong view” of hell. They are as follows:

(H1) The Anti-Universalism Thesis: some persons are consigned to Hell;

(H2) The Existence Thesis: Hell is a place where people exist, if they are consigned there;

(H3) The No Escape Thesis: there is no possibility of leaving Hell, and nothing one can do, change, or become in order to get out of Hell, once one is consigned there; and

(H4) The Retribution Thesis: The justification for and purpose of Hell is retributive in nature, Hell being constituted so as to mete out punishment to those whose earthly lives and behavior warrant it.

(Kvanvig 1993, p. 19)

Judged against this list of theses, the pre-Qin and Han conceptions of hell are “weak” indeed. The only one on this list which they satisfy is the extremely sparse H2. It will not help to get into a semantic debate about necessary and sufficient conditions for the word “hell”,9 nor will I organise this section around enumeration of the evidence for and against each of Kvanvig’s theses. But it may be helpful right at the beginning to be aware that at this stage in Chinese cultural history: contra H1, all (not “some”) deceased humans will be consigned to hell; contra H3, there is the possibility of leaving hell—some souls return to the mortal world as spirits or ghosts, whereas others merely fade away, being reassimilated in some indeterminate fashion back into the qi-flow of the world at large; contra H4, hell is not punitive.

It is this latter feature of hell as non-punitive, along with a certain structural continuity characterised by the “bureaucratic” nature of hell, that we will wish to highlight in this precis. Let us now begin by considering the following assertion made by sinologist Laurence G. Thompson:

It must be admitted that the Chinese hardly contributed anything to the structure or the punishments of purgatory, but simply took over literally the ideas that had arrived from India.

(Thompson 1989, p. 35)

Post-Han Buddhist scriptures contributed a wealth of new motifs and ideas on the concept of hell, not to mention a systematic theological framework. This much is true. There is also a degree of overlap between Thompson’s assertion above, and a conclusion drawn by the celebrated cultural historian Yu Yingshi that the “idea of Heaven and hell as opposing sites of reward and punishment in the afterlife was not fully developed in Chinese thought until the coming of Buddhism” (Yu [1987] 2016, p. 71). But Thompson’s assertion is—at best—too strong, and—at worst—misleading. To imply, as both Thompson and Yu do, that there was little autochthonous notion of the underworld as a place of punishment seems quite fair. But as we will see presently, however, there is in fact plenty of evidence that the structure of the pre-Qin and Han conceptualisations of the underworld contributed much to later Buddhist ideas of hell. Let us examine these two aspects of the pre-Buddhist underworlds in turn.

Possibly the most well-known of several underworlds is that of the Yellow Springs (Huang Quan 黃泉). In one sense, the pre-Qin Yellow Springs just referred prosaically to a physical location under the earth.10 In another sense, however, they referred to a dark underworld through which people11 were to pass after death regardless of the uprightness or otherwise of their conduct.12 In one early anecdote from the Zuo Commentary, for instance, Lord Zhuang of the state Zheng (in modern day Henan) is challenged with an armed rebellion led by his younger brother, and fomented and abetted by his mother, the Lady Jiang. After defeating the rebellion, he dispatches his mother to another city, swearing an oath as he does so: “we will not meet again until we’re at the Yellow Springs” (“First Year of Duke Yin”, Zuo Zhuan Yizhu 1998, p. 3).13

There are two pertinent points to notice about this passage. Firstly, though dating the Zuo Commentary is tricky,14 it is not implausible that what the Commentary recorded was in fact the verbatim utterance of Lord Zhuang from 722 BCE—if this were true, it would push the terminus ante quem for the establishment of these ideas about the Yellow Springs back to the very early Eastern Zhou. Secondly, though Lord Zhuang presumably had good reason to think that his mother had done him a grave moral wrong, the passage itself in its full context gives no reason to think that he conceived of the Yellow Springs as a place in which either of them were to be punished. Nor is this true of, say, Guanzi in the eponymous text in which he kowtows to Duke Huan of Qi and rhetorically and ceremonially invites the Duke to kill him and thereby send him to the Yellow Springs;15 nor of the Lord of Anling, when he offers to join the King of Chu in death as a human sacrifice, following him to the Yellow Springs;16 nor, for that matter, is it true of any of the other references to the Yellow Springs in the pre-Qin textual record.

To proceed from this to a further abductive inference that there was no conception of a punitive aspect to hell until very late in the Han period is, of course, an argumentum ex silentio which could potentially be overturned by some explicit assertion in the next tranche of excavated manuscripts, but it seems to be a reasonably compelling17 one at this point. This inference is made even more compelling by the Western Han datum provided by the painted banner draped upon the coffin of the Marquise of Dai, excavated at Mawangdui 馬王堆 and now displayed in the Hunan Provincial Museum. The tripartite division of the painting depicts the realms of heaven, the mundane human world, and the underworld. As emphasised by Eugene Wang, the way the underworld is depicted therein is not at all horrifying, but is rather a scene of joyful copulation between various creatures (Wang 2011, pp. 48–49). Not until the Scripture of Great Peace 太平經 in the late Eastern Han do we find any punitive aspect to the conceptualisation of hell, and in this case it only applies to those whose “evil deeds were unceasing” in life,18 with a tone that tends much more towards conciliation than the lurid fire-and-brimstone of much later19 texts like the Jade Records 玉曆寶鈔.

So much for the lack of any punitive aspect to the pre-Qin and Western Han underworld. What about the structure of the underworld, though? Is it true that, with respect to structure, autochthonous pre-Qin and Han notions of the underworld “hardly contributed anything” to the later post-Han conceptualisations of hell? It is not true. Various excavated manuscripts have shown that the conceptualisation of hell as a sort of underworld bureaucracy—which really took hold several centuries later during the Tang period with the authorship of the Scripture on the Ten Kings 十王經, in which it was elaborated upon in significant detail—was in fact present much earlier even than the Han,20 dating back to at least the Warring States period. A wooden document excavated at Mawangdui in 197321 and a bamboo manuscript excavated at Fangmatan 放馬灘 in 198622 both demonstrate detailed conceptions of an underworld bureaucracy in the pre-Qin.

So, we have highlighted two features of the pre-Buddhist conceptions of hell in China—the lack of a punitive component, and the presence of a pre-existing bureaucracy. Why is it important to underline these two features? Let us recall the overall goal: we are trying to understand why—right at the end of the Warring States when the functional approach to the evolution of religion might have theoretically predicted the Chinese underworld to evolve in the direction of a terrifying tool of extramundane retribution—what we rather observe is several more centuries of the continued marginality of hell. Well, we could say that the lack of postmortem punishment meant that no-one particularly feared hell, over and above any baseline fears about one’s own eventual departure from the present world, and thus it was marginal. But this is just to beg the question: why was there no postmortem punishment? My contention going forward is that hell and all of its byzantine bureaucratic apparatus did in fact have a functional role at this time, but this role was less about securing adaptive behaviours through fear of incurring a postmortem penalty, and more as a necessary ontological posit to support a complex system of extramundane but premortem retribution which was already in place. We will come back to this system of premortem retribution in Section 4.1. Far from the thanatological vacuum suggested by Thompson’s “hardly contributed anything” assertion, it was with these “weaker” but nonetheless well-fleshed-out autochthonous conceptions of hell with which any “stronger” conceptions (i.e., satisfying more of Kvanvig’s theses) would have to compete, and eventually oust or incorporate.

3. The Functionalist Approach to Extramundane Retribution

Before proceeding to examine reasons for the marginality of hell in the next section, it will be worthwhile highlighting the fact that there is a wealth of evidence to suggest that since quite early in recorded history, Chinese theorists have embraced functionalist explanations of beliefs in the extramundane. Western theologies tend to be discursive, appealing to what are (putatively) true facts about the universe and the reasons for its being created in some particular way. These theologies can be mutually exclusive in the trite but straightforward sense that there are facts about the external world that has been created—such as whether it is Muhammed or Jesus Christ who is God’s true earthly representative—facts about which either Islamic theology is correct, or Christian theology is correct, but not both simultaneously. Or, to use the example most pertinent to this special issue, there is (for instance) a fact about whether or not H3 obtains and punishment in hell is in fact eternal, and one’s eschatology either correctly or incorrectly represents that fact. Chinese theorists, however, have been much more content to focus on whether or not beliefs are functionally useful, while very much decentering the question of whether such beliefs are accurate representations of some mind-independent external reality, often maintaining a strategic ambivalence toward the question. With respect to attitudes towards the conscious existence (or otherwise) of the ghosts and spirits who emerge from the underworld as the potential arbiters of extramundane punishment, this ambivalence is characteristic of Confucius, of later Ruists who appropriated his sagely imprimatur, and of Guanzi:

“Sacrifice as if [they were] present” means that, when sacrificing to the spirits, you should comport yourself as if the spirits were present.

The Master said, “If I am not fully present at the sacrifice, it is as if I did not sacrifice at all.” (Analects 3.12, Slingerland 2003a, pp. 21–22).

祭如在，祭神如神在。

子曰：“吾不與祭，如不祭。”

(Analects 3.12, p. 27)

Zigong asked the following of Confucius: “Are the dead conscious? Or not conscious?”

子貢問於孔子曰：“死者有知乎？將無知乎？”

Confucius replied: “I should like to say that there is consciousness in death, but would worry that filial sons and obedient grandsons would hamper the living in sending off the dead; I should like to say that there is no consciousness in death, but would worry that unfilial sons would abandon their parents and not bury them. Ci, you wish to know whether the dead are conscious or not, but at present this is not urgent, in the future you will know for yourself.”

子曰：“吾欲言死之有知，將恐孝子順孫妨生以送死；吾欲言死之無知，將恐不孝之子棄其親而不葬。賜欲知死者有知與無知，非今之急，後自知之。”

(School Sayings of Confucius, p. 92)

Guanzi entered and reported as follows to Duke Huan: “Heaven has sent its envoys, who have arrived at the outskirts of your lordship’s [domain], I would recommend you have your great officials and close advisers don dark robes and send them out [to meet the envoys]. These are envoys from Heaven!” When this was noised abroad, people exclaimed: “Just incredible, this Duke Huan of Qi—to have Heaven sending envoys to the outskirts [of his domain]!” Without even having raised an army, eight feudal lords came to his court [to pay homage]. This is the dao of co-opting the awesome aegis of Heaven to influence [the people] under it. Thus the wise put ghosts and spirits to their own uses, while [only] the ignorant believe in them.

管子入復於桓公曰：“天使使者臨君之郊，請使大夫初飭、左右玄服，天之使者乎！”天下聞之曰：”神哉齊桓公，天使使者臨其郊。”不待舉兵，而朝者八諸侯。此乘天威而動天下之道也。故智者役使鬼神而愚者信之。

(Guanzi, p. 1487)

In all of these examples, there is at play a certain kind of strategic aporia with respect to whether or not ghosts and spirits actually exist, a certain doctrinal flexibility and equivocation about what is factual with respect to these extramundane punishers, and a willingness to act as if they exist regardless of whether or not—in some mind-independent way—they actually do. There are several other examples of this type of functionalist explanation for beliefs about the supernatural cropping up itself within the Chinese tradition—Mozi23 and Wei Yuan24 魏源 are two further illustrative examples. For Guanzi, the objective was political expediency and advantage rather than descriptive accuracy. For Confucius and his Ruist followers, it was inculcation of “solemn” (yan 嚴) and “reverential” (jing 敬) habits of mind supportive of filial (xiao 孝) behaviour rather than epistemic veracity. This mode of ritual behaviour operating in a register of “as if” is what Michael Puett has astutely labelled the “subjunctive mode” (Puett 2013, passim), which operates very differently from the “discursive mode” of Christian theology.

Now, it might be intuitive to think that such doctrinal flexibility on the part of elite theorists in the late Eastern Zhou and early Han period would have been advantageous in allowing for the development of a much “stronger” (and therefore more adaptive) conception of hell at precisely the time it would be theoretically predicted, that is, right as increasing social mobility and technological advance was undermining those “older elite communities [with] the social structure of economically corporate kin groups or clans (zu 族) spread over a homeland region” (Cook 2009, p. 241). But on the contrary, quite the opposite seems to have been true: there was already some latent doubt over whether the “extramundane punishers from the underworld” narrative was true,25 and without these doctrines being “preached from the pulpit” by the cultural elite, and with the Ruists operating in their “subjunctive mode” and their figurehead Confucius avoiding all discussion of the paranormal and extramundane (see Analects 7.21), it was difficult for such beliefs to get significant traction. Attempts to propagate such beliefs with reference to (for instance) their functional utility in improving the efficiency of legal systems and transportation networks26 (rather than with reference to their descriptive truth) were bound to smack of ulterior motives. Especially for beliefs that are inherently aversive (such as the belief in a punitive hell), in the absence of strong social and didactic support for their veracity, individual humans will be biased against assenting to them.

In sum, just as us theorists of the evolution of religion stand outside a religious belief system and “unmask” it by pointing to functional utility as the reason for its propagation while never actually assenting to its discursive truth, so I propose that the “self-unmasked” functionalist orientation of the “subjunctive mode” of the Warring States and early Han was—perhaps counterintuitively—a positive hinderance to the uptake of “strong” conceptions of hell. There is also one further parenthetical point that seems worth making here: as this section shows, when we look to functionalist explanations for beliefs about extramundane punishment, we are not imposing some completely alien category of interpretation on the Chinese tradition, but rather are advancing more elaborated versions of arguments that can be found in the Chinese tradition itself, only now undergirded with the theoretical apparatus of cultural evolution27 to better explain how the functionality arises.

4. Virtue-Power (de) and the Ancestral Cult

If, as I have proposed above, Ruism’s unique “subjunctive mode” of ritual activity and its didactic reticence with respect to the extramundane was indeed an influential causal factor in China having bucked a trend towards stronger conceptions of hell amongst civilisations of comparable scale and technological complexity, it is unlikely to have been the only causal factor. We must also avoid the obvious pitfall of overemphasising the causal role of Ruism. While Ruist thought does indeed have an outsized importance given that it was able to install itself as the imperial orthodoxy, it was far from the “only game in town”: this would certainly have been true amongst the commoners, and was even true at the court—where the same promiscuously ecumenical Han emperors and “Ruist at court, Daoist at home” officials responsible for permitting this installation were also enthusiastically pursuing various Daoist prescriptions for becoming an immortal (xian 仙) (See Yu [1987] 2016, pp. 24–36). As such, we must look to other factors plausibly unique to China which may have militated for the continued marginality of hell and against the evolution of stronger versions of postmortem retribution, especially where there are inherent assumptions common to both Ruism and other schools of thought. I present two such factors here, and then a third tied more specifically to Ruism. All three are related to the notion of virtue-power (de) and the ancestral cult.

4.1. De as a Credible Bodily Signal of Self-Monitoring

Implicit in the definition of a high god is that it acts as an extramundane monitor responsible for policing and punishing transgression from a shared moral code. It has been argued previously that such a high god (variously named Di 帝, Shangdi 上帝, Tian 天) was part of the Chinese imagination during the Shang (1600–1046 BCE) and Zhou periods (Clark and Winslett 2011). By the end of the Zhou period, however, this deity was becoming increasingly non-anthropomorphic, such that eventually tian came to mean something more like “the workings of the cosmos in their totality”. Earlier in the Zhou period however, the Zhou kings worshipped Tian—who they supposed had bequeathed a mandate to rule to their ancestral lineage. They also sacrificed to their lineage ancestors, who were minor deities also capable of punishing and rewarding, but were much more tractable to propitiation than was Tian. Other zu clans, of course, had their own lineage ancestors, and de was the commodity that was shared between living descendants and their ghostly ancestors, who were ritually invited back into the ancestral temple to be presented with sacrificial offerings, often of food and alcohol.

In the Shang and early Zhou period, de was a thaumaturgical power bequeathed on the living by the ancestors, and it was decidedly substantive rather than purely adjectival (hence the translation as “virtue-power” rather than simply “virtue”). For the living, it could be accumulated through excellence in “modeling” (xing 型) one’s conduct after the ancestors and thereby “grasping their virtue-power” (bing de 秉德), and by provisioning them with food and other ritual items, such as the spectacular bronze vessels which provide a rich source of information about the ancestral cult specifically, and about this period more broadly. The morally-tinged power of action on the world which was the de of an ancestral lineage was very credibly signaled by the ability to commission these bronze vessels which were “quite probably, the most accomplished, expensive, labor-intensive, and beautiful human-made things their owners and handlers had ever seen.”28 Elsewhere, I have argued (Martin forthcoming) that de was the ability to credibly signal this morally-tinged power of action, that it was recognised as such in philosophical works of the Warring States and Han periods, and that we can profit by recognition of this aspect of it. For the ancestor spirits, their standing and comfort in the afterlife and indeed their continued existence as a ghost or spirit was determined by how sumptuous were their mortuary rites and subsequent sacrifices received.29 As we foreshadowed at the end of section two, these ancestor spirits were very likely to emerge from the underworld and engage in premortem punishment of the unfilial if such rites were not undertaken with the necessary gravitas.

Towards the end of the Zhou period, however, the ancestral cult was weakened somewhat as the zu clans were broken up by military upheaval, increasing technological complexity, and social mobility, and there were concomitant changes in the way that de was conceptualised. The thaumaturgical connection of de to the ancestral temple weakened, and while de continued to credibly signal one’s morally-tinged power of action on the world, it was now thought of more in terms of a difficult-to-fake effortlessness and charisma indicating the genuine internalisation of certain moral qualities, and necessarily associated with one’s bodily presence.30 To use a term from the etymological dictionary Shuowen Jiezi, it was conceived as being “obtainable from within oneself” (nei de yu ji 內得於己) rather than in large part mediated by extrahuman ancestors.

Having rehearsed this background on de and the ancestral system, we are now positioned to show how this bears on the marginality of hell, advancing an argument which builds on one originally sketched out by Hagop Sarkissian (Sarkissian 2015) in doing so. While the Ruists of the Spring and Autumn/Warring States periods maintained a studied aversion to theorising about ghosts, spirits, and the afterlife (Analects 11.12), they were, on the contrary, extremely enthusiastic about didacticising on the need to self-monitor. It was extremely important to them that the abundance of one’s human-heartedness, uprightness, ritual propriety, and wisdom be displayed bodily in a “wordless illustration” (Mencius 7A21; Bloom 2009, p. 148), that “as wealth adorns a house, so de adorns the body” (“Great Learning”, in Liji Zhengyi, p. 1593). Importantly, if one was to be assured of possessing this coveted and respected quality of de, one would need to be extremely assiduous in self-monitoring one’s thoughts and behaviours, even when one is totally alone (shendu 慎獨), because “what is genuinely present within will take form externally” (i.e., in one’s body language; “Great Learning”, in Liji Zhengyi, p. 1593). Consummate de conduct borne of habituated self-monitoring must either come in an effortless way which eo ipso announces itself as genuine, or not at all: importantly, this was an idea shared not just by the Ruists, but also by the Daoists (see Slingerland 2003b), and many other schools of thought.31

Let us dwell a little further on how de could possibly have become a motivating force effective enough to drive genuine behavioural modification in the way that hell and the threat of postmortem retribution later did. Without spelling this out carefully, we might seem to be making the rather extraordinary claim that the Chinese were just more predisposed to morality than other peoples. The ancestral system featured extramundane monitoring and broad supernatural punishment of laggardly and unfilial sacrificers unpossessed of de by ancestral “ghosts and spirits” who emerged from the underworld for the task. As spheres of interpersonal interaction expanded during the late Zhou, these ghosts and spirits came to be conceived as particularly responsive to morally-tinged de virtue-power in particular rather than just acting to benefit (or punish) their living descendants for (dis)loyalty to the clan,32 and as such people were motivated to acquire this de. Though the ancestral system slowly went defunct during the Warring States period (before being revived during the Han, as we discuss below), it was replaced by a form of self-monitoring motivated by the desire to safeguard one’s own de virtue-power. For those skeptical about the descriptive reality of ghosts, spirits, and the afterlife, de virtue-power was indeed valued for its own sake. For the more credulous, however, the possession of de virtue-power enabled one to avoid premortem punishment by ancestor spirits, and was also a potential avenue to the postmortem reward of veneration by the living, a veneration which itself was thought to sustain one’s postmortem existence. In conclusion, it is not that the Chinese were necessarily more predisposed to morality than other peoples, but rather that they had a relatively sophisticated written discourse regarding the premortem functional value of de virtue-power,33 a system of beliefs about de which already encompassed premortem punishment by ancestral spirits, and a view of postmortem existence by which individuals potentially stood to gain from its acquisition. These factors, in tandem with the others we discuss in this paper, to some degree obviated the need for postmortem punishment, and are plausible causal factors in the marginality of hell in the Chinese intellectual landscape of the period.

4.2. Shared Assumptions of Impermanence of the Individual

In the previous subsection, the main factor we put forth as part of the causal picture of the marginality of hell (and the moralistic high gods who might consign one there) in Zhou and Han period China was an idea about de which was shared broadly across a few different schools of thought. The same is true of the factor to which we turn now, viz., a broadly shared assumption of the ultimate impermanence of the individual subject. To explicitly spell out the relevance of this assumption to the “marginality of hell” claim: if there was no assumption that one’s hunpo 魂魄 (soul) or shen 神 (spirit) could exist permanently, that would rule out a “strong” conception of hell satisfying Kvanvig’s H3 (The No Escape Thesis),34 which would in turn take a lot of the sting out of the threat of hell, even if a conceptualisation of hell including H4 (The Retribution Thesis) was to evolve.

For the Daoists, though later religious variants sought to use various potions to extend the human premortem lifespan indefinitely, or to the point where one became an immortal xian (neither of which would merit a fear of hell), the default was nevertheless postmortem oblivion: early texts such as the Zhuangzi were quite sanguine about the idea that after humans had passed from initial oblivion into amorphous qi, from amorphous qi into qi with form, from qi with form into form with life, then this process would inevitably be reversed like the cycling seasons, with a return to oblivion.35

For the Ruists, though they did in fact avoid all discussion of the afterlife in favour of copious discussion of self-cultivation, their ritual praxis was nevertheless aimed at reviving the Zhou ancestral system. In this system—as the anthropologist Maurice Freedman put it—“dead ancestors rely for their perennity on the ritual memory of their descendants” (Freedman 1967, p. 85). The ontological status of postmortem ancestors in this ancestral system is still open to several interpretations, though, all of which have some textual support in the historical record. These interpretations can be split into two broad categories36—interpretations in which the ancestors exist purely in the minds of the living, and interpretations in which despite having their own existence outside the minds of the living, they are nevertheless reliant on ritual provisioning by their descendants. Presuming the former interpretation, it is obvious that the ancestors could not possibly stand to fear hell. Presuming the latter interpretation, it seems clear from certain passages that ancestors who failed to obtain the sustenance from their living descendants would then be faced with oblivion as they faded away,37 often becoming the agents of punishment, wreaking vengeance on the living before doing so. On the contrary, the way to extend the postmortem “lifespan”38 of the ancestors was to ritually memorialise them in the ancestral temple. As the generations passed, however, the wooden tablets representing the ancestors furthest from living memory would be removed from the ancestral temple, upon which they would recede to oblivion. The only exceptions were for lineage founders or sagely ancestors of particularly exceptional de—in these cases, their tablets would be kept in perpetuity, becoming “irremovable ancestors” (bu tiao zhi zu 不祧之祖). On neither interpretation, however, was there a presumption of a hunpo existing into the future independently of the ritual actions of the living, and this very much undermined the possibility of a strong hell satisfying the No Escape Thesis.

It may be wondered why the eternal soul concept didn’t evolve ab initio in China, given that it is a precondition for stronger (and therefore more socially functional) conceptions of hellish postmortem punishment?39 Our answer to this question runs along similar lines to that which we proposed towards the end of the introduction: while ab initio appearances of religious beliefs in the absence of direct civilisational contact would indeed be strong confirmation of the functionalist view, such data are not absolutely necessary to the case. Using a standard conceptualisation of cultural fitness as involving a stepwise walk through a fitness landscape, we would expect the cultural evolutionary process to often get “stuck” in local maxima (i.e., at the apex of smaller peaks), and as such for it to take quite some time—even under conditions of optimal selection pressure for strong hell concepts—to break out from such local maxima. Indeed, our task in this paper has been to show some of the countervailing cultural forces holding the Chinese on the smaller adaptive peak defined by impermanence of the soul. Strongly held pre-existing ideas about the subjunctive (rather than descriptive) reality of extramundane agents, the transience of qi, and the connection of de to the ancestral system were all, to a greater or lesser degree, in tension with the notion of an eternal soul. This leads us now to a further factor: the extra level of support for this package of beliefs about the afterlife, which it was to the advantage of the Han imperium to provide.

4.3. The Political Utility of the Ruist Doctrine of Xiao for the Han Imperium

The final factor we propose as having contributed to the marginality of hell (in this case, in the Han period in particular) is the political utility for the Han imperium of revamping the ancestral system.40 This was because (a) the ancestral system underscored the political legitimacy of the Han regime, and (b) it instilled the value of xiao (filial piety), which could be co-opted to ensure the obedience of the populace—this was certainly how it was conceived by the Ruists.41

As we noted in Section 3 and Section 4.1, the ancestral system was weakened during the Warring States by the emergence of precisely those social conditions which would generally predict the obsolescence of ancestor worship and the turn towards high gods in many other world civilizations. For the Han, however, their reinstatement of the ancestral system—along with all the powerfully symbolic physical infrastructure that strengthened the imperial presence outside the capital42—allowed them to cast themselves in the mould of the Zhou dynasty, triumphing over the forces of chaos and receiving the mandate of Heaven/Tian through their ancestral lineage, replete even with their own emperors Wen and Wu to parallel the celebrated emperors of the early Western Zhou, and most importantly, with their emperor acquiring the status of “Son of Heaven/Tian”. This provided sufficient motivation to exploit the unique opportunity afforded by this dynastic turning point, and “artificially” turn back the tide of the macro technological and social trends working against ancestor worship.43

Finally, there was a deep connection between the ancestral cult and reverential obedience toward the patriarchal system entailed in the moral quality of xiao. This is particularly well captured by a phrase from the excavated bamboo manuscript The Way of Tang and Yu 唐虞之道, which states that the sages “served personally in the ancestral temples, thereby teaching the populace filial piety”.44 It is worth noting here that xiao had originally meant the act of making offerings to deceased ancestors in the ancestral temple, although the Ruists augmented the concept to encompass a rich set of emotions which far exceeded mere “provision” (see Analects 3.7). The Classic of Filial Piety ordained the Son of Heaven/Tian at the apex of a hierarchy of xiao-mediated relations which were necessarily to be ritually affirmed in the ancestral temples, and for these reasons it was accorded special status by the Han court. Again, any rival conceptions of hell and the afterlife would need to compete at the fringes of this state-sponsored ideology, which, as mentioned earlier, entailed that the only individuals existing indefinitely in the afterlife were the “irremovable ancestors” and sages of exceptional de. This made it difficult for such rival conceptions to be anything but marginal.

5. Conclusions

In this paper, I claimed that hell was a marginal notion in pre-Qin and Han China, and that the Ruist inheritance and modification of Zhou ideas about de virtue-power played a role in suppressing the development of a more punitive conception of hell, which according to the predictions of the cultural-evolutionary functionalist view of religion should have been expected to arise earlier than it in fact did in a civilisation like that of ancient China. Ruist teachings—with their emphasis on xiao, correct ritual adherence to the ancestral system, and self-monitoring in order to “obtain [de] from within oneself”—offered a suite of attractive and legitimacy-endowing features to the incoming Han emperors, who installed Ruism as the state orthodoxy. The ancestral system was characterised by a form of premortem extramundane retribution carried out by temporarily emancipated ghosts rather than postmortem retribution in a hellish underworld, and this plausibly resulted in suppression of the emergence and/or propagation of such “strong” conceptions of hell.

The collapse of the Han dynasty, however, was accompanied by a concomitant decline in the influence of Ruism, and this allowed Indic conceptualisations of a “stronger”, more moralistic, more punitive, and more memetically potent hell to gain a foothold during the interdynastic and Sui-Tang periods in the absence of a unified and well-functioning ideological framework. Terrifying frescoes depicting judgement and torment appeared in many diverse locations, as did Buddhist texts such as the Scripture on the Ten Kings, which nevertheless incorporated some autochthonous notions—in particular the notion of hell as an underworld bureaucracy—into their portrayal of hell. This reversion to a more adaptive attractor state (“strong” hell) once the countervailing factors militating for a marginal hell had weakened or disappeared plausibly strengthens their standing as hypotheses about why hell remained so unexpectedly marginal despite China having already become a civilisation of significant scale and technological complexity. These hypotheses may be applicable mutatis mutandis to broader questions about where the seemingly exceptional religious history of China fits into the cultural-evolutionary picture of the evolution of religion.