https://orcid.org/0000-0003-4241-8408
![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
The Apaṇṇaka Sutta, one of the early recorded teachings of the Buddha, contains an argument for accepting the doctrines of karma and rebirth that Buddhist scholars claim anticipates Pascal’s wager. I call this argument the Buddha’s wager. Does it anticipate Pascal’s wager and is it a good bet? Contemporary scholars identify at least four versions of Pascal’s wager in his Pensées. This article demonstrates that the Buddha’s wager anticipates two versions of Pascal’s wager, but not its canonical form. Like Pascal’s wager, the Buddha’s wager presents a decision problem between two opposing theses in an epistemic context that lacks evidence of their truth or falsity. Like Pascal, the Buddha also tries to solve this problem using dominance, superdominance or ‘superduperdominance’ reasoning. The Apaṇṇaka Sutta likely provides the earliest textual example of such reasoning. While the Buddha’s wager does not exhibit the expected utility reasoning of the best-known form of Pascal’s wager, the article suggests a reformulation that parallels Alan Hájek’s (2018) vector-value reformulation. Is it a good bet? This article argues that it is not if this means we are rationally required to accept its recommendation. This is because, while it avoids two of the major objections levelled against Pascal’s wager, it succumbs to one and has at least two problems of its own.
Acknowledgements
This article was inspired by Roger Jackson’s (Citation2022) claim that the Apaṇṇaka Sutta may have anticipated Pascal’s wager. The article was also improved by conversations with Alan Hájek, whose careful scholarship, generous feedback, and overall assistance in understanding and articulating the reasoning involved in Pascal’s wager was invaluable. I am also grateful to the detailed comments, criticisms, and suggestions of two excellent anonymous reviewers, as well as for the helpful feedback of Szymon Bogacz, Graham Priest, and Koji Tanaka.
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Funding Information
This article was supported by a Discovery Early Career Research Award (DECRA) funded by the Australian Research Council (Grant: DE180100001).
| Abbreviations | ||
| AN | = | Aṅguttara Nikāya of The Buddha (2012) |
| DN | = | Dīgha Nikāya of The Buddha (1995) |
| MN | = | Majjhima Nikāya of The Buddha (1995) |
| P. | = | Pāli |
| Skt. | = | Sanskrit |
| SN | = | Saṃyutta Nikāya of The Buddha (2005) |
Notes
1 Although I express these assertions and arguments as made by the Buddha (even describing the underlying argumentative structure as the Buddha’s wager) it is to be understood that these claims are attributed to the Buddha in the context of the sutta. The Apaṇṇaka Sutta is part of the Majjhima Nikāya, one of five collections containing some of the earliest recorded teachings of the Buddha. According to Buddhist scholarship, the Buddha’s teachings were oral and transmitted orally from one generation of his disciples to the next for a century or more after his death before being collected, redacted, and eventually written down in languages not spoken by the Buddha. As a result, it is widely agreed that ‘these texts undoubtedly reflect the outcome of a long process of redaction, so that even in the case of the Sūtra and Vinaya material, we cannot be confident that what appears in them “is what the Buddha taught”’ (Jackson Citation2022: 28)
2 I am, here, using Ñāṇamoli and Bodhi’s translation. The Pāli-English dictionary (PED) takes apaṇṇaka to have the sense of ‘certain, true, absolute’. Thanissaro Bhikkhu translates it as ‘safe-bet’ or ‘cover your bets’ (Citation2008); Jayatilleke translates it as the ‘infallible dhamma’ (1963: 405); and in translation Buddhagoṣa is represented as defining it as ‘unopposedly leading to what is doubtless, holding onto certainty’ (MA 3:116, cited in Gamage Citation2013).
3 This is my interpretation, but consistent with Bodhi’s overall treatment of this thesis as concerned with karmic rebirth. Bodhi, himself, interprets ‘no mother, no father’ to mean that there are no karmic consequences of conduct towards one’s mother and father (Bodhi Citation1995: 234). I find this interpretation unmotivated and less plausible than the one offered here.
4 The Buddha also considers two further opposed thesis pairs: the affirmation and denial of (d) whether it is possible to experience the immaterial realms (the highest sphere of rebirth), and (e) whether one can achieve nibbāna, the cessation of suffering, and thus be liberated from the cycle of rebirth. He presents a slightly modified argument for (d) and (e) that I call the subsidiary argument. It makes no mention of wagering but draws the same structural conclusion: that one should believe their affirmation rather than denial. The subsidiary argument shares premises 1-3 of the Buddha’s wager but is aimed at the ‘wise man’ who is thinking of undertaking Buddhist practice (a potential recluse) but is unsure which thesis to take as the goal of their practice. While interesting, I will not analyse the subsidiary argument here.
5 This appears to contrast with the advice the Buddha offers to the householders in the Kālāma Sutta to whom, in similar epistemic circumstances and when presenting the same set of issues, he remarks: ‘know for yourselves … then you should live in accordance with [those ideas]’ (AN 65.5) There might be ways to reconcile this apparent inconsistency. It might be argued, for instance, that the Buddha in the Kālāma Sutta articulates an epistemic ideal, realisable if one abandoned the householder life and became a recluse. Whatever the explanation, the Buddha’s wager in the Apaṇṇaka Sutta is best understood as a decision-making strategy for householders who want to live a householder life that accords with one or other of the theses in question.
6 This term was coined by McClennen (Citation1994) but there is some debate about whether his formulation best represents Pascal’s reasoning in this first wager (see Hájek Citation2012, Citation2018).
7 This formulation of superdominence aligns with Hájek’s (Citation2012) notion of ‘superdominance+’ rather than McClennen’s original formulation, which does not include the conjunct. As Hájek argues: ‘Pascal’s case is better than [McClennen’s original proposal]: winning everything is better than (not merely at least as good as) each of the outcomes associated with wagering against God’ (Citation2018: 125). While Hájek identifies several other versions of ‘(super)dominance’ reasoning, they do not matter for the present argument. I do, however, distinguish superdominance from what Hájek goes onto call ‘superduperdominance’ reasoning. Many thanks to an anonymous reviewer for pressing me to clarify the reasoning involved here, and to Alan Hájek for helping me to think through and articulate the position adopted in this paragraph.
8 Jayatilleke (Citation1962) presents a simpler decision matrix, with only social outcomes for the state where karmic rebirth is not, and karmic outcomes for the state where there is karmic rebirth. This inaccurately represents the Buddha’s wager by removing the sense in which accepting can be a lucky throw ‘on two counts’ (and vice versa for accepting
). It also matters for replies to objections that there are two utilities in each state.
9 I will go onto argue that the Buddha’s wager also exhibits superduperdominance reasoning.
10 There is some debate about how best to represent the utility of ‘misery’; whether some negative infinite, representing the idea of eternal damnation, or some finite amount of punishment (Martin Citation1983; Sobel, Citation1996; Saka Citation2002; Hájek Citation2022).
11 Strictly speaking, these two classes of outcomes need not be exclusive, nor do they necessarily divide into pre- and post- rebirth. Karmic outcomes can occur in the present life and can include social outcomes. The claims that are relevant to the Buddha’s wager, however, concern karmic rebirth: the possibility of rebirth given the laws of karma. The Buddha also offers social outcomes in the present life as a contrast to karmic outcomes since he treats karmic and social outcomes as two distinct ‘counts’ with respect to which a wager can be lucky and unlucky.
12 While the numbers of how long these lives last are not consistent or accurate, they all express a very long duration of time (Braarvig Citation2009). As an example, consider the following exchange in the Kokālita Sutta:
Venerable sir, how long is the life span in the Paduma hell? The life span in the Paduma hell is long, bhikkhu. It is not easy to count it and say it is so many years, or so many hundreds of years, or so many thousands of years, or so many hundreds of thousands of years. Then is it possible to give a simile, venerable sir? It is possible, bhikkhu. Suppose, bhikkhu, there was a Kosalan cartload of twenty measures of sesamum seed. At the end of every hundred years a man would remove one seed from there. That Kosalan cartload of twenty measures of sesamum seed might by this effort be depleted and eliminated more quickly than a single Abbuda hell would go by. Twenty Abbuda hells are the equivalent of one Nirabbuda hell; twenty Nirabbuda hells are the equivalent of one Ababa hell; twenty Ababa hells are the equivalent of one Aṭaṭa hell; twenty Aṭaṭa hells are the equivalent of one Ahaha hell; twenty Ahaha hells are the equivalent of one Kumuda hell; twenty Kumuda hells are the equivalent of one Sogandhika hell; twenty Sogandhika hells are the equivalent of one Uppala hell; twenty Uppala hells are the equivalent of one Puṇdarīka hell; and twenty Puṇdarīka hells are the equivalent of one Paduma hell.” (SN 1.6.10)
13 Thank you to an anonymous reviewer for suggesting this parallel.
14 In actual fact, good karmic outcomes are not all equivalent but can be comparatively ranked since one can be reborn into one of a variety of heavenly realms with their own distinct modes and durations of pleasure (and the equivalent for bad karmic outcomes; there are numerous hell realms into which one can be reborn, each with its own distinct mode and duration of excruciating suffering). Precise information about karmic outcomes is not available in the context of the Buddha’s wager, however, and so I set this complexity aside.
15 The disjunction about how best to represent bad karmic outcomes (0 or -1) parallels the same ambiguity as in Pascal’s wager about how best to represent misery; whether as the negative counterpart of the alternative (gain all or good karmic outcomes) or as some finite value.
16 As an aside, the Buddha and later Buddhist philosophers were atheists about a creator God, and provide rational arguments against his/her existence (Patil Citation2009). While many Buddhists believe in a cosmology of deities, gods, and other divine beings, they deny any rational need to posit a God as first cause of all existing things. These arguments were not available to the householders that are the target audience of the Buddha’s wager, however.
17 Hájek (Citation2012) makes an analogous point in relation to Pascal’s wager.
18 It might be argued that this objection could be blocked if the consequences of believing in reincarnation (or rebirth) are the same on each of these conceptions (Buddhist, Jain, Sikh). Without closer inquiry, however, it is not obvious that adherents to these different traditions have the same assumptions about what counts as a good action and thus of the karmic consequences of those actions. Thank you to Swami Medhananda for pressing this point.