See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/360562830

Conventionalising rebirth: Buddhist agnosticism and the doctrine of two

truths

Chapter · December 2023

CITATIONS READS

0 223

1 author:

Bronwyn Finnigan
Australian National University
23 PUBLICATIONS   82 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Buddhist Meta-ethics and Moral Psychology View project

All content following this page was uploaded by Bronwyn Finnigan on 23 November 2022.

The user has requested enhancement of the downloaded file.

Forthcoming in Global Dialogues in the Philosophy of Religion: From Religious Experience to the Afterlife,
Yujin Nagasawa & Mohammad Saleh Zarepour (eds.) Oxford University Press

Conventionalising rebirth:
Buddhist agnosticism and the doctrine of two truths1

Bronwyn Finnigan
Australian National University

What should the Buddhist attitude be to rebirth if it is believed to be inconsistent with current
science? This chapter critically engages forms of Buddhist agnosticism that adopt a position of
uncertainty about rebirth but nevertheless recommend ‘behaving as if’ it were true. What does
it mean to behave as if rebirth were true, and are Buddhist agnostics justified in adopting this
position? This chapter engages this question in dialogue with Mark Siderits’ reductionist
analysis of the Buddhist doctrine of the two truths, conventional and ultimate. Richard Hayes
(1998) characterises talk of rebirth as a useful fiction. Siderits characterises talk of persons as a
useful fiction and explains and justifies statements that involve it as conventionally true
despite persons not featuring in our final or ultimate ontology. Does rebirth satisfy the same
criteria to count as conventionally true, and does thinking of it in these terms help explain and
justify what it might mean to behave as if rebirth were true? This chapter will defend a
conditional yes to these questions. In the process, it will clarify what is distinctive about the
traditional Buddhist approach to rebirth, provide an analysis of how the concept of rebirth
might relate to practical outcomes, and address some limitations of this approach.

Keywords: rebirth, Buddhism, conventional truth, pragmatism, agnosticism, fiction,

agency, personhood, karma, reductionism

1Many thanks to Szymon Bogacz, Roger Jackson, Mark Siderits, and the editors of this volume for helpful
comments on a previous draft of this chapter.
Introduction

Until modern times, the idea of rebirth was widely accepted and asserted by Buddhists. The

Pāli Canon, which includes some of the earliest recorded teachings of the Buddha, describes

a cosmology of five (or six) ‘realms’ of existence into which sentient beings are born, die, and

are reborn in a continuous cycle; those of gods, humans, animals, hungry ghosts, and hell

beings (see DN8. MN130, KN10,7). The cycle of rebirth is known as saṃsāra.2 Where or how

one is reborn, as well as some of the auspicious and inauspicious events that occur in that

life, are said to be determined by the laws of karma which relate to the ethical quality of

one’s actions, good and bad.

Much Indian Buddhist discourse about rebirth focuses on explaining how it is

consistent with core Buddhist claims, such as the Buddha’s teaching of no-self (anātman).

While these explanations assume rebirth rather than attempt to prove it, they nevertheless

provide a distinctively Buddhist interpretation. Some arguments are also offered to justify

rebirth. Dharmakīrti (7th CE) presents what has come to be known as the standard Buddhist

argument for rebirth against materialism (see PV 1.34-119).3 Rebirth, Buddhists traditionally

assumed, involves a causal series of immaterial mental events transcending the boundaries

of a single lifetime; there is assumed to be a causal link between a subtle form of

consciousness present at the time of death and the first moment of consciousness in the next

life, and between the volitional mental entities in the present life (e.g. intentions and reactive

attitudes) and the psychophysical elements that constitute that person in the next life.

Dharmakīrti argues that these immaterial mental events (moments of consciousness and

volitions) cannot be sufficiently caused by the body, a material entity, because they are too

different in kind to be substantively related as cause and effect. A growing number of

2The italicised words in this chapter are in Sanskrit.

3For elaboration and discussion of this argument, see Franco (1997), Hayes (1993), Jackson (1993, 2022) and
Nagatomi (1957).

2
contemporary Buddhist philosophers find this argument unpersuasive, however, arguing

that it fails to refute reductive physicalism in a way that would convince a modern cognitive

scientist or philosopher of mind (Willson 1987, Jackson 1993, 2022, Hayes 1993, Batchelor 1997,

Arnold 2012, Thompson 2015). The dominant view amongst these scholars is that the

traditional Buddhist view of rebirth is inconsistent with modern science.

If we were to grant this point, what should the modern Buddhist attitude be towards

rebirth? Some Buddhist scholars affirm the inconsistency, arguing that Buddhism is a

radical cultural critique of the scientific viewpoint (Lopez 2012). Many argue, however, that

modern science provides the best evidenced set of theories we currently have about reality,

and treat its inconsistency with rebirth as a problem for rebirth. Of those scholars, some

respond by rejecting rebirth as an unnecessary cultural relic. Others ignore it or put its

discussion into abeyance while engaging other issues. This chapter will investigate whether

belief in rebirth, both that there is continuity after death and that it is explained by karma,

can be pragmatically justified as conventionally true despite being assumed to be inconsistent

with current science.

There is historical precedent of pragmatic arguments for rebirth in early Buddhism.

In the Apaṇṇaka Sutta (MN60), the Buddha offers an argument for belief in rebirth that

anticipates Pascal’s wager (Jayatilleke, 1962; Jackson, 2022; Finnigan, forthcoming). He

contends that if one is not in an epistemic position to directly ascertain whether claims about

karma and rebirth are true or false, it is better to believe their affirmation rather than their

denial because this belief has better consequences, irrespective of its truth or falsity.4

Modern Buddhist agnostics offer similar pragmatic arguments for rebirth in the face of

epistemic uncertainty. Stephen Batchelor (1997) coined the term Buddhist agnosticism to

denote an attitude of epistemic uncertainty about rebirth (“The only honest position I can

arrive at is: ‘I actually don’t know’”). We might query whether this attitude is warranted for

4 For elaboration and critical assessment of this argument, see Finnigan (forthcoming).

3
a Buddhist who holds that rebirth, traditionally construed, is inconsistent with the reductive

materialism of current science. Roger Jackson (2022) elaborates, however, that the term

applies to “any thinker who finds the traditional, rational, empirical, or faith-based

arguments for rebirth to be problematic but does not reject the idea outright, admitting

that—with our present limitations—we simply do not know whether past and future lives

are real.” (p. 267). If we take ‘present limitations’ to refer to the current incompleteness of

science, it might warrant some degree of agnosticism. Some philosophers and scientists

contest the assumed reductive physicalism of science, for instance, arguing that an

irreducibly causal conception of consciousness might be consistent with the basic laws of

physics if it can be established as a kind of non-physical energy (or physical energy, on some

broadened definition of the physical). If these arguments succeed, then rebirth need not be

denied for assuming this idea.5 These arguments have yet to be established by the methods

of current science, however. While the probability might be quite low that rebirth will turn

out to be consistent with a future completed science, few would be so dogmatic as to assert

this with certainty.

Unlike Western agnostics, Buddhist agnostics do not suspend judgment about the

object of their agnosticism (rebirth) but offer pragmatic reasons to justify retaining the idea

in some practical form. Jackson (2022) endorses Lati Rinpoche’s recommendation that even

if we judge the truth of rebirth to be unestablished, we should ‘behave as if it were true”

(p.267, my italics). Stephen Batchelor similarly remarks that we can, at least, “try to behave as

if there were infinite lifetimes in which [we] would be committed to saving beings.” (Tricycle

5It might be denied on other grounds, however. Buddhist philosophers offer complex analyses of how karma
operates. For some, it operates by merit-generation; good actions generate karmic merit (puṇya) which persists
in some form until causes and conditions are suitable for it to effect a good outcome in this life or the next. An
argument that established the causal efficacy of consciousness would not necessarily establish the existence
and efficacy of merit. Buddhist agnostics tend not to detail which conception of karma they take rebirth to
assume. They do assume, however, that there are complex causal links between ethical conduct and
consequences, good and bad, in this life or the next. This chapter will focus on this more general assumption
about karma. Thanks to Szymon Bogacz for pressing me on this issue.

4
1997, my italics). What does it mean to behave as if rebirth were true and is this approach

justified?

This chapter will approach this question in dialogue with the Buddhist doctrine of

the two truths, conventional and ultimate. It is inspired by Richard Hayes (1998), who

characterizes rebirth as a useful fiction; fictional because unestablished by logical proof or

empirical observation as real within a scientific framework, but useful nevertheless because,

as Jackson (2022) elaborates, it motivates individuals to “live ethically and compassionately.

In that way, they will generate happiness for themselves and others in this life, and if there

are future lives they will be happy ones” (p.267). Mark Siderits influentially characterizes

talk of persons as a useful fiction and elaborates its fictional status by appeal to the two

truths. On his account, talk of persons is conventionally true, and persons are

conventionally real, even though persons do not exist in the Buddhist final ontology and are

thus not ultimately real. Siderits also justifies the conventional truth of statements about

persons in terms of their utility or pragmatic value; persons are useful fictions. Does rebirth

satisfy the same criteria as persons to count as useful fictions in conventionally true beliefs,

by Siderits lights? And does thinking in these terms help clarify and justify what it might

mean to behave as if rebirth were true? This chapter will defend a conditional yes to both

questions. In the process, it will clarify what is distinctive about the traditional Buddhist

approach to rebirth, provide an analysis of how the concept of rebirth might relate to

practical outcomes, and address some limitations of this approach.

Conventional persons

The distinction between conventional truth and ultimate truth resolves an apparent

inconsistency between the Buddha’s teaching of no-self and his commitment to karma and

rebirth. How is rebirth possible if there is no self to be reborn into another life? How can

5
karma function if there are no agents to experience the karmic consequences of ‘their’

actions?

While Buddhists debate the meaning and entailments of the Buddha’s teaching of

no-self, most agree that he is not asserting that no one and nothing exists. At the very least,

he is rejecting the idea that there is an eternally existing substance (me!) that persists through

time and that grounds our diachronic identity in the face of psychological and physical

change. According to early Buddhism, when we subject persons to empirical and conceptual

analysis, all we find is a dynamic and complex causal system of psychophysical elements.

While the Buddha offers several classifications of these elements, the most well-known is

that of the five-aggregates (skandhas): (1) material bodily elements (rūpa), (2) feelings (vedanā),

(3) discriminative cognitions such as perceptions, thoughts, and recollections (saṃjñā), (4)

volitional elements such as intentions and reactive attitudes (saṃskāra), and (5) events of

consciousness (vijñāna). This analysis is assumed to be exhaustive; there is nothing else that

constitutes a person. All elements in these person-systems depend on causes and conditions

for their existence and (so) are impermanent; none have independent and permanent

existence. Moreover, the unification of these elements as a ‘whole’ system is not considered

to be a real substance with causal properties.6 Siderits characterizes this view as a

mereological reduction and endorses it as Buddhist Reductionism.7

If there are laws of karma, they must concern the psychophysical elements in these

causal series. But which elements in these series do they target? The Buddha famously

remarked that karma targets intentions,8 which are volitional elements in the reductive

analysis. Siderits (2003) clarifies that the laws of karma are not rules that are decreed or

enforced by some cosmic being and obeyed or broken by agents. Rather, they causally

6 Later Abhidharma Buddhist philosophers provided sophisticated arguments in defence of this view, the
most prominent being the neither-same-nor-different argument, see Siderits (2013: 74).
7 Siderits’ preferred position is more nuanced and takes on board the Madhyamaka critique of a final level of

description. I will return to this.

8 More specifically, “it is volition…that I call kamma [P. karma, Skt.]” (AN 6.63)

6
describe the way the world works akin to the so-called natural laws of science. Actions

caused by good intentions produce good karmic outcomes (that are pleasurable) and actions

caused by bad intentions produce bad karmic outcomes (that are painful). “If we could keep

track of enough persons over enough successive lives, we could find out what the laws of

karma are in the same way that science discovers what the laws of nature are: our

observations would disclose the patterns of regular succession that show causation at work.”

(p.9) Rebirth occurs when the set of psychophysical elements that make up a person in this

life causes a new set of psychophysical elements to come into existence in the next life.

Siderits argues this is similar in kind to what regularly occurs in a single lifetime; the set of

psychophysical elements that make up an infant, for instance, are not identical to but are

causally related in the right kind of way to the set of psychophysical elements that constitute

the adult later in life.

What should we make of talk of persons if all that really exists are complex causal

systems of psychophysical elements? Buddhists invoke the distinction between two truths

to explain or justify such talk. On Siderits’ analysis, the concept of a person conveniently

designates a whole causal system of psychophysical elements, past, present, and future.

While the elements in this series (at a time) are real, the whole system is not. Statements

involving the concept of a person are, at best, conventionally true but ultimately

meaningless since there are no persons at the ultimate level of analysis.

Criterion for conventional truth: common sense

What explains and justifies a statement as being conventionally true? Siderits offers several

accounts in his writings. In (2007) “a statement is conventionally true if and only if it is

acceptable to common sense and consistently leads to successful practice.” (p.56) Kris

McDaniels (2019) points out that these conjuncts come apart. The idea that conventional

truth is what is acceptable to common sense is central to the Prāsaṅgika Madhyamaka of

7
Candrakīrti (Cowherds, 2011). On this view, conventional reality is whatever ‘the world

acknowledges’ to be the case without subjecting it to rational or epistemic analysis

(Candrakīrti PPMV 18.8, in Cowherds 2011: 151). Prāsaṅgikas ascribe this conventional reality

a diminished status, judging it to be illusory, mistaken, and bound up with ignorance

(Tillemans 2016). Common sense might thus explain why we talk about persons, but it does

not thereby justify it if by justification we mean providing good reasons for continuing the

practice.

Siderits (2007) allows that rebirth was “part of the common-sense conception of the

world for most Indians for most of the time that Buddhism existed in India” but denies that

it is “part of our common-sense world-view” (p.10). Does this mean that statements about

rebirth were conventionally true in ancient India but not conventionally true for modern

Buddhists who accept the modern scientific viewpoint? If so, it seems arbitrary, relativising

truth to beliefs that just happen to be common in some local historical and cultural context.

Cultural contexts can also overlap. A modern Buddhist might both accept a broadly

scientific framework and yet believe in rebirth. Does rebirth count as conventionally true

for them?

Siderits (2007) denies that common sense is arbitrarily tied to mere belief in this way.

He argues that statements acceptable to common sense are those that “consistently lead to

successful practice” and if they fail this criterion then they are conventionally false. He

illustrates the point by reference to traditional flat-earth theory,9 arguing that while this was

once a widespread and common-sense viewpoint, “the statement that the world is flat was

never conventionally true” (p.57, n.10) Why? Because desires that are relevantly informed by

this belief, such as the desire to reach the edge of the world, consistently fail to be satisfied.10

9He also uses the example of fairies (2005: 92) and disease caused by demon possession (p.94)
10The claim is not that all desires of a flat-earther fail to be realised, merely those that are informed by beliefs
that are directly entailed by the belief that the world is flat, such as the belief that “if you were to sail far enough
in the same direction you will reach the edge.” (p.57, n.10)

8
 The definition of conventional truth in Siderits (2003) makes explicit this assumed

dependence of common sense on successful practice: “a sentence is said to be conventionally

true if and only if it is assertible by the conventions of common sense, where these are

understood as standards based on utility.” (p.7) The conventions governing common sense

both explain and justify talk about persons: such talk is explained because it is part of our

sociolinguistic practices and justified as useful and good because it leads to successful

practice. Tillemans (2016) objects that Siderits “over interprets” Buddhist sources and that

this justification of conventional truth in terms of utility is his own philosophical addition

(p.7). If we nevertheless grant ‘leading to successful practice’ as a pragmatic criterion for

what statements should be accepted as conventionally true,11 what does it amount to and can

it be satisfied by belief in rebirth?

Criterion for conventional truth: leading to successful practice

What does it mean for a useful fiction to consistently lead to successful practice? Siderits’

view is complex, and its various aspects have been contested. It assumes, for instance, that

there can be a final ontology that reflects ultimate reality, understood as the objective way

things really are, independent of our interests and cognitive limitations. Mādhyamika

Buddhists reject this view, reserving the term ‘ultimate’ for emptiness (śūnyatā) which is

taken to entail that there are no ultimately real entities and so no privileged ultimate

discourse about how things really are. Siderits’ preferred position accepts the Madhyamaka

critique but treats it as consistent with a broadly reductionist approach to metaphysical

explanation.12 I shall set these issues aside for now and focus on explicating Siderits’ view on

11In expressing this point in normative terms, I set aside the empirical question of whether Siderits is right to
claim that common sense folk ontology is, in fact, governed by considerations of utility.
12 For instance, Siderits (2019) defends a contextualist semantics that admits multiple levels of reduced

description as grounds for conventionally true claims depending on what counts as explanatory in a given
context.

9
what it means to lead to successful practice. For convenience, I analyse it into the following

five aspects.

Unit of utility: Siderits’ analysis of Buddhist Reductionism offers a consequentialist

justification of conventionally true statements. What are useful fictions useful for? They are

useful for minimising pain or suffering (2003: 37, 58; 2005: 113) and maximizing pleasure or

welfare impersonally construed (2003: 37, 57; 2005: 105). Pleasure and pain, impersonally

construed, are psychophysical elements of ultimate reality. Why is this the relevant measure

of utility? Because, according to Siderits, Buddhists take it to be ultimately true that pain is

bad and to be prevented (2003: 46, 58).

Desire-generation: How does the concept of a person help achieve this outcome?

Siderits argues that it motivates ‘us’ (viz. the present set of psychophysical elements) to take

an interest in this unit of utility as it relates to the future, to adopt it as an object of desire,

and to choose actions that help bring it about. He appears to grant that ‘we’ are naturally

averse to pain when it arises, a fact evident in the behaviour of small children, but argues

that we only anticipate and have an interest in preventing future pain when we are socialised

into the personhood convention (2003: 9). Siderits thus considers the concept of a person to

play a crucial role in the process of desire-satisfaction; it converts a natural aversion into an

interest in future events. But how does it perform this function?

Aggregation: Siderits analyses the person-concept as having an aggregate function.

What we call a person, according to Buddhist Reductionism, is a massively complex causal

series of psychophysical elements that are hard to track and perhaps impossible to

completely describe. The concept of a person “lightens our cognitive load” (2013: 5) and

“eases communication” (2005: 99) by aggregating the entire causal series as a singular whole.

In this respect, it functions like the concept of a chariot (2003: 40), forest (2007: 55), and water

(2013: 5); unifying certain kinds of particulars when arranged in certain kinds of way. Given

that the relevant particulars are causally related, these concepts have a temporal dimension;

10
they relate future states of affairs to the present as parts of the same thing. By relating future

psychophysical events to the present as events that will happen ‘to me’, the person concept

facilitates such inferences as: just as ‘I’ can experience pain now, just so ‘I’ can experience

pain in the future, and just as I don’t want to experience pain now, just so I don’t want to

experience pain in the future.

Deliberation and counterfactual reasoning: This is not enough for the person concept to

count as a useful fiction. That ‘I’ desire to prevent pain occurring to ‘me’ in the future does

not yet result in successful practice. For this, the desire must lead to actions that actually

produce the desired outcome. It does so by informing deliberations which result in choice

of action. Such deliberation involves counterfactual reasoning.

Consider Siderits’ (2005) example of flossing one’s teeth or getting a flu shot:

Neither action is particularly pleasant, so the present elements receive no reward for
performing them. But if they do not get performed, eventually there will be a great deal
of pain that might have been prevented. The best way to prevent that pain turns out to
be by getting the elements in a causal series to identify with and appropriate the past
and future elements in that series. (p.95)

“Socialisation into personhood”, Siderits contends, “involves learning to act on the basis of

the outcome of deliberation. And deliberation involves, in the first instance, seeing oneself

as an enduring entity having a variety of interests that might be served in various ways.”

(p.106). Deliberation also involves counterfactual reasoning about the possible

consequences of courses of action, such that if some action were (not) performed then some

desired outcome would (not) come to be.

Success: This is still not enough for persons to count as useful fictions. It must also be

the case that the actions which result from these deliberations, informed by the person

concept, actually succeed in generating the desired outcome. A fiction that makes no

difference to how things are, or that informs desires that cannot be satisfied, is not useful in

the relevant sense. The usefulness of fictions in conventionally true statements is justified

both in relation to what happens at the ultimate level of description as well as in relation to

11
what is ultimately valuable at this level; namely, minimising pain and suffering and

maximising pleasure and wellbeing, impersonally construed. It is this relation to ultimate

reality that marks the difference between fictions that are conventionally true or false.

Putting all this together, we can generate an account of what it means to behave as if

there were persons, in a sense that involves conventionally true beliefs. To behave as if there

were persons (such as ‘me’) in the relevant sense is (1) to believe that ‘I’ will experience future

events such as pain, (2) to desire to promote or prevent these future events occurring ‘to me’,

(3) to engage in deliberation about how to achieve this desired outcome, which involves

counterfactual reasoning about possible consequences of possible actions, and (4) to choose

an action that reliably achieves the desired outcome. The beliefs on which this behaviour is

conditioned are justified as conventionally true when the choices they inform do in fact

produced the desired outcome, describable at the level of ultimate reality. Persons are a

‘useful’ fiction because although there is, in fact, no ‘me’ that experiences (e.g.) the

anticipated pain, the action performed does in fact prevent pain, impersonally construed,

and so is justified as good by the lights of the ultimate truth.

Does belief in rebirth count as conventionally true?

Does the rebirth concept satisfy the same complex criterion to count as a useful fiction that

leads to successful practice and thus for belief in rebirth to count as conventionally true? I

will defend a conditional yes.

The Buddhist agnostic approach to rebirth aligns with Siderits’ unit of utility.

Jackson claims that to behave as if rebirth were true will “generate happiness for [yourself]

and others” (2022: 267). He takes this to align with the Buddha’s claim that belief in rebirth

will “lead to your welfare and happiness for a long time.” (MN60.4) This can be rendered

consistent with Siderits’ impersonal construal and interchange of maximising pleasure and

welfare with minimising pain and suffering.

12
 What about desire generation? Does belief in rebirth help convert a natural aversion

into a desire for some future outcome? Siderits’ explanation depends on aggregation; the

person concept enables this conversion by aggregating past, present, and future

psychophysical elements as one thing (me!). If we grant this point, does the rebirth concept

have the same aggregate function? It does insofar as it presupposes the concept of a person,

i.e., it is a person that is reborn again after death.13 What it relevantly (but not exhaustively)

adds is an extension of the causal series of psychophysical elements that count as me beyond

the boundaries of a single lifetime. If the aggregate function of the person concept is what

converts a natural aversion to a desire to prevent future states of affairs (as happening to me),

then it should make no difference how far into the future these states of affairs occur,

whether in this life or the next. The rebirth concept is thus consistent with desire generation.

The rebirth concept can also have a role in deliberation and counterfactual reasoning.

When we are deciding what to do, we consider the possible consequences of various courses

of action. The Buddhist concept of rebirth presupposes karma, the idea that acting well and

badly have corresponding consequences, good and bad, in and for the next life. Buddhist

philosophers provide complex analyses of how karma operates but claim that only a Buddha

knows its exact mechanics. Buddhist agnostics do not analyse karma but tend to accept the

general claim that a good karmic consequence is a good rebirth, meaning a life that involves

more pleasure than otherwise,14 and a bad karmic consequence is a bad rebirth, meaning a

life that involves more pain and suffering than otherwise. The rebirth concept thus informs

expectations about what possibilities might occur in and for the next life if we intentionally

act in certain ethical or unethical ways. This gives a deliberative sense to what it means to

behave as if rebirth were true; it is to give expectations about the possible karmic consequences

Williams (2000) contests this point but see Siderits (2000: 414).
13

According to Buddhism, all realms of existence involve suffering to some degree, even the blissful heavenly
14

pure lands. Different explanations are offered of this apparent fact, however.

13
of our actions weight in our decision-making processes, to treat them as reasons counting in

favour or against certain kinds of action.

That leaves success. Do actions which result from choices informed by the rebirth

concept actually succeed in generating their desired outcome and thereby make an

ultimately real difference in the world? And does this difference align with the ultimately

true values of minimising pain and suffering and maximising pleasure and welfare? I will

initially argue no but conclude that they might.

Buddhist agnostics tend to justify behaving as if rebirth were true by reference to its

capacity to “generate happiness for [oneself] and others in this life” (Jackson 2022: 267, my

italics) According to Lati Rinpoche, for instance, being compassionate and helping others

feels good, makes others feel good, and results in others loving and thinking highly of you

and being willing to help you in return (see Hayes 1998: 79). We are invited to suppose that

acting ethically creates a reciprocal and ramifying network of pleasure and wellbeing in this

life, irrespective of whether it has consequences for the next. These present-life

consequences can align with the unit of utility. Insofar as belief in rebirth provides reason for

actions that do, in fact, minimise pain and maximise welfare, impersonally construed, it

might seem that these beliefs are justified as useful and good by the lights of what counts as

ultimately valuable.

It does not thereby count as conventionally true, however. For Siderits, a useful

fiction informs desires which are actually satisfied at the level of ultimate reality. Person talk

is useful because it motivates me to both desire a certain effect (that pain of a certain kind

not occur in future) and to choose actions that actually bring it about (pain of that kind does

not occur). In the case of rebirth, the desired effect (that I obtain a good rebirth) is not the

same as the effects in the world used to justify its usefulness (happiness for myself and others

in this life). While I might very well desire happiness for myself and others in this life, it was

not my reason for action (which was to obtain a good rebirth) and so this does not count as

14
a case of desire-satisfaction in the relevant sense. It could thus be argued that the fiction of

rebirth does not thereby lead to successful practice in the right kind of way for belief in

rebirth to count as conventionally true.

This conclusion would not follow, however, if the action did in fact produce the

desired effect; good karmic consequences in the next life. From the point of view of current

science, it cannot have this outcome since rebirth does not hold at this level of analysis.

Buddhist agnostics do not deny rebirth, however. They maintain a position of epistemic

uncertainty which, we have suggested, might be warranted by reference to the

incompleteness of science. The kind of Buddhist agnostic we are considering believes that

rebirth is inconsistent with current science but maintains that it might nevertheless turn out

to be true for all we currently know. If we are willing to grant this position of epistemic

uncertainty about rebirth, it provides grounds for a conditional defence of its conventional

truth. Desire-satisfaction is a matter of fact, grounded in how things are at the ultimate level

of description. If a desire is satisfied, by causing actions that lead to effects that fit the

description of what is desired, then it is satisfied irrespective of whether we know it or not. I

might not know whether my desire is satisfied that you, the reader of this chapter, agree with

my arguments, but if you do agree with my arguments then you do, irrespective of my

epistemic position. Similarly, if the desired karmic consequences of an action do, in fact,

follow from certain kinds of action, then the rebirth concept can be said to lead to successful

practice in the relevant sense to count as a useful fiction in conventionally true beliefs. The

Buddhist agnostic allows the possibility that actions can cause karmic effects but doesn’t

know whether they do in fact. Belief in rebirth might thus count as conventionally true for a

modern Buddhist agnostic but they cannot say for sure.

Putting all this together, to behave as if rebirth were true is thus (1) to believe that I will

experience the karmic effects of my actions in the next life, (2) to desire to promote or

prevent these future events occurring to me, (3) to engage in deliberation about how to

15
achieve this desired outcome, which involves counterfactual reasoning about possible

consequences of possible actions, and (4) to choose an action that (a) achieves a desirable

outcome (justified as good or bad by the lights of the unit of utility), and (b) reliably achieves

the desired outcome (the intended karmic effects). The beliefs on which this behaviour is

conditioned are justified as conventionally true when the choices they inform do in fact

produce the desired outcome, describable at the level of ultimate reality. The Buddhist

agnostic is in a position of epistemic uncertainty about whether this desired outcome

obtains. But if it does obtain then belief in rebirth is a useful fiction that leads to successful

practice and so is conventionally true.

Some limitations and closing remarks

There is much more to say about this analysis of behaving as if rebirth were true. I will close

by addressing three potential limitations.

The deliberative role of useful fictions.

This chapter has provided a deliberative analysis of what it means to behave as if persons and

rebirth were true. By this I mean that it analysed these concepts as having a functional role

in deliberation or practical rationality. It did not exhaustively analyse the functional role of

these concepts for deliberation, however. Take the person concept. It could be argued that

the fact that ‘I’ engage in deliberation at all assumes that I am an agent whose choices can

make a real difference in the world; it would not make sense to deliberate about (e.g.)

whether to floss my teeth to prevent tooth decay if I did not believe that my decision could

actually result in an action that prevented tooth decay. Similarly, it would not make sense

to deliberate about whether to leave a room if I believed that the only exit is locked.15

15 This example is taken from Van Inwagen (2002)

16
Deliberation seems to presuppose that one is an agent in control of one’s actions and with a

genuinely open future such that one’s choices can have a real causal impact on the world.

Is the person concept therefore a necessary assumption of practical rationality? It

might seem so. Siderits argues no. This is due to its consequentialist justification. While

Siderits allows that deliberation informed by the person concept might effectively get us to

perform actions that, in fact, prevent pain and maximise wellbeing, he argues that it is in

principle possible that some other concept could be equally, if not more, efficacious (2000:

415, 2o05: 106). The reason why the person concept is a commonplace in so many historical

and cultural contexts, in Siderits’ view, is because it just so happens to be the most effective

strategy for minimizing overall suffering we currently have (2005: 113).

The concept of rebirth presupposes karma. As such, it also assumes that we are

agents whose choices can make a real difference in the world. Actions produced by good

intentions are assumed to lead to good karmic outcomes and actions produced by bad

intentions are assumed to lead to bad karmic outcomes in this and the next life. In contrast

to the person concept (and closely related possibilities),16 deliberation need not assume the

concepts of karma or rebirth. It is also not obvious that these concepts provide the most

effective reasons for choice of ethical conduct. While it is hard (but not impossible) to

conceive of viable alternatives to the person concept for deliberation, there are plenty of

alternative reasons an individual might have to choose to act ethically.

This suggests a potential difference in scope between the usefulness of the person

fiction and that of rebirth. Person talk is useful (but not necessary) for all rational beings

(that is, all beings with a capacity to choose actions for reasons); we should all use this

concept when decision-making insofar as it is the most effective cognitive strategy to achieve

our desired outcomes, including those deemed valuable by the ultimate truth. Rebirth talk is

16Siderits offers Punctualism and Weitgeistism as possibilities and gives reason to think they would be less
effective at minimising overall suffering than the person concept (see 2005: 106; 2003: 37).

17
useful, we might argue, for those rational beings that may not otherwise have sufficient

reason to act ethically or who are egoistic and would only be moved by self-interested

reasons to perform actions that benefit others.17

Is this deliberative analysis of rebirth self-defeating?

On this deliberative analysis, rebirth contributes self-interested reasons to decision-making;

one treats the possible karmic consequences of possible actions for you as reasons to choose

those actions (or not). These reasons are self-interested because, on Siderits analysis, desire

for certain future outcomes results from a process of aggregating future possibilities to the

present, via the person concept, as things that will happen to oneself. My reason for acting

is not merely that certain outcomes obtain but that they obtain for me.

It might be objected that self-interested reasons are self-defeating when related to

karmic consequences. Karma tracks intentions. An action chosen for self-interested reasons

rather than a genuine concern for others, one might argue, is not a good intention. It might

even be considered selfish and so constitute a bad intention that will have bad karmic

consequences. If this is right, then behaving as if rebirth were true might be self-defeating;

actions chosen for the reason that they produce good karmic consequences for me do not,

in fact, produce those consequences for me precisely because they were chosen for this

reason.

To settle this, we need some explanation of what counts as a good intention. Siderits’

analysis of utility suggests a possible solution. A good karmic outcome is one that minimises

pain and suffering and maximises pleasure and wellbeing, the unit of utility. Insofar as my

17 A potentially more nuanced answer might be derived from distinguishing the ‘continuity after death’ and
the ‘explained by karma’ dimensions of belief in rebirth. Some care would be needed with the detail, however.
The idea that one continues after death, for instance, might seem to have distinct practical utility by motivating
choice of actions aimed at more long-term or semi-idealistic goals unachievable in a single lifetime. Buddhists
assume, however, that it is unlikely that one will be reborn again as a human being, let alone with a functional
ability to recall decisions made in a previous life. While ‘I’ will experience the pleasure and pain in the next
life that result from ‘my’ actions in this life, ‘I’ may very well do so in the form of a cow (or a goddess!) with no
awareness of the goals or desires ‘I’ had in my previous mode of existence.

18
intention is to bring about these effects, it counts as good. While I might also intend for these

outcomes to happen ‘to me’, there are no persons at the ultimate level of analysis and so the

difference between whether they are experienced by me or you is meaningless. Self-

interested reasons are thus justified as good on consequentialist grounds and are not self-

defeating.

Is this deliberative analysis of rebirth consistent with Buddhist soteriology?

Self-interested reasons involve a sense of ‘I’ or ‘self’. A central feature of the Buddha’s

teachings is that the idea of self is a cause of suffering and has bad soteriological effects. The

Buddha is reported to have taught that the idea of self conditions craving and attachment

and thus suffering when the objects of our attachment (ourselves, most pertinently)

inevitably change given the fact of impermanence. The idea of self is also considered to

inform actions (of body, speech, and mind) that keep us in saṃsāra, the cycle of rebirth. It

might be argued that the deliberative analysis of rebirth provided in this chapter is

inconsistent with the soteriological framework of Buddhism.

There is a lot to be said about this objection. Here are three brief responses. First, we

might defend a developmental approach to Buddhist practice, and argue that behaving as if

rebirth were true, in a deliberative sense, is a stage on the path for the egoistic person who

needs reason to act ethically. It is a stage because, if such a person were to regularly choose

actions that help others rather than harm, it might lead them to habituate these actions, as

dispositional modes of response, and cultivate reactive attitudes such as compassion, which

produce the same kinds of ethical conduct but no longer via processes of self-interested

reasoning. This would be consistent with a gradual extirpation of the sense of self or ‘I’ from

our psychological processes.18

18 Thanks to Mark Siderits for this suggestion. I offer a similar strategy to Dharmakīrti to account for the agency

of a Buddha who does not engage conceptuality in the mode of deliberative choice (see Finnigan 2010-11)

19
 Second, we might defend a deliberative approach to rebirth, but deny that it must be

self-interested. Consider a Mahāyāna Buddhist who has taken the bodhisattva vow to

remain in the cycle of rebirth in order to relieve the suffering of all sentient beings but who

has learnt from Śāntideva to construe this impersonally (see BCA 8.101-103). It is arguable

that their reason for acting is that it would prevent pain and promote pleasure, overall and

impersonally construed, rather than specifically for themselves.19 If this is plausible, it

appears to contradict Siderits claim that ‘we’ only anticipate and take an interest in the unit

of utility as it relates to the future when we unify it with the present as something that will

occur to me. There is good reason to contest this empirical claim, or at least restrict its scope

of application. But whether a Mahāyāna Buddhist can eschew all self-interested reasons and

still achieve viable practical outcomes is a matter of considerable debate (see Williams, 1998;

Cowherds, 2015; Finnigan, 2018)

Third, and lastly, behaving as if rebirth were true need not only be analysed

deliberatively. The concept of rebirth includes, for instance, the idea of being reborn into a

different mode of being to that of one’s present existence; (e.g.) as a god or a hell being, a

cow or a cockroach. Hayes (1998) suggests that the idea of rebirth can inspire the creative

imagining of what it is like to live a different kind of life, which might facilitate an openness

to different perspectives and the cultivation of compassion towards others (p.79).20 This is

not inconsistent with the deliberative analysis, for it can be justified on the same grounds. It

is nevertheless distinct and suggests that there is more than one way to cognitively and

psychologically analyse what it means to behave as if rebirth were true.

Conclusion

What should the Buddhist attitude be to rebirth if one accepts its inconsistency with current

science? Buddhist agnostics adopt a position of uncertainty about rebirth but nevertheless

19 Thanks to Roger Jackson for raising this example and issue.

20 Sonam Kachru (2021) derives similar insights from his reading of Vasubandhu.

20
recommend ‘behaving as if’ it were true. This chapter investigated what this might mean

and whether it is justified in dialogue with the Buddhist doctrine of the two truths, as

analysed by Mark Siderits. To behave as if rebirth were true, it argued, is to treat possible

karmic consequences as reasons counting in favour or against certain kinds of action when

deliberating about what to do. These reasons need not be decisive. A modern Buddhist

might give these possibilities very low credence given the improbability (but not

impossibility) that rebirth will turn out to be consistent with a completed science. These

reasons might also be unnecessary for motivating the agent to choose to act ethically; the

agent might already have sufficient reason to act ethically without needing to consider

possible karmic consequences as well. But treating possible karmic consequences as reasons

for action might reliably contribute to ethical living by helping motivate agents to choose

actions that minimise pain and suffering and promote pleasure and well-being. “The trick”

to modify a remark by Siderits, “is to recognize that there ultimately is no such thing as a

person [and might not be such a thing as rebirth] and yet at the same time recognize that

much of the time it can be very useful to act as if there were.” (2005: 94 my italics).

Abbreviations
AN Aṅguttara Nikāya of The Buddha
BCA Bodhicaryāvatāra of Śāntideva
DN Dīgha Nikāya of The Buddha
KN Khuddaka Nikāya of The Buddha
MN Majjhima Nikāya of The Buddha
PV Pramāṇavārttika of Dharmakīrti

References
Arnold, D. (2012). Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and
Cognitive-Scientific Philosophy of Mind. New York: Columbia University Press.
Batchelor, S. (1997). Buddhism Without Beliefs: A Contemporary Guide to Awakening. New York:
Riverhead Books.
Cowherds, (2011). Moonshadows: Conventional Truth and Buddhist Philosophy, Oxford: Oxford
University Press
Cowherds, (2015) Moonpaths: Ethics and Emptiness, Oxford: Oxford University Press

21
 Finnigan, B. (2010-11) Buddhist Metaethics, Journal of the International Association of Buddhist Studies
(JIABS) 33 (1-2): pp.267-97
Finnigan. B. (2018) Madhyamaka Ethics, In Oxford Handbook of Buddhist Ethics, D.Cozort & J.Shields
(eds.) Oxford: Oxford University Press, pp.162-183
Finnigan, B. (forthcoming) The Buddha’s Lucky Throw and Pascal’s Wager, Australasian Journal of
Philosophy
Franco, E. (1997), Dharmakīrti on Compassion and Rebirth, Wiener Studien zur Tibetologie und
Buddhismuskunde heft 38, Wien: Arbeitskreis für Tibetische und Buddhistische Studien
Universität Wien
Hayes, R. (1993). “Dharmakītri on punarbhava.” In Studies in Original Buddhism and Mahāyāna
Buddhism. Egaku Maeda (ed.), I, 330. Kyōto: Nagata Bunshodo.
Hayes, R. (1998). Land of No Buddha: Reflections of a Sceptical Buddhist. Birmingham, UK: Windhorse
Publications.
Jackson, R. (1993). Is Enlightenment Possible? Dharmakīrti and rGyal tshab rje on Knowledge, Rebirth, No-
Self and Liberation. Ithaca: Snow Lion Publications.
Jackson, R. (2022). Rebirth: A Guide to Mind, Karma and Cosmos in the Buddhist World, Boulder:
Shambala
Jayatilleke, K.N. (1962). Early Buddhist Theory of Knowledge, George Allen & Unwin Ltd.
Kachru, S. (2021) Other Lives: Mind and World in Indian Buddhism, Oxford: Oxford University Press
Lopez, D. (2012). The Scientific Buddha: His Short and Happy Life. New Haven: Yale University Press.
McDaniels, K. (2019). Abhidharma Metaphysics and the Two Truths, Philosophy East and West 69(2):
439-463
Nagatomi, M. (1957). A Study of Dharmakīrti’s Pramānavārttika: An English Translation and
Annotation of the Pramānavārttika Book I. PhD diss., Harvard University.
Siderits, M. (2000). Reply to Paul Williams, Philosophy East 7 West, 50 (30): 453-459
Siderits, M. (2003). Empty Persons: Personal Identity and Buddhist Philosophy. Aldershot, Burlington,
Ashgate.
Siderits, M. (2005). Freedom, Caring, and Buddhist Philosophy, Contemporary Buddhism, 6:2, 87-116
Siderits, M. (2007). Buddhism as Philosophy. Aldershot, Burlington, Ashgate.
Siderits, M. (2013). Causation, ‘Humean’ Causation and Emptiness, Journal of Indian Philosophy, 42
(4):433-449
Siderits, M. (2019) The Prapañca Paradox,
Thompson. E. (2015). Waking, Dreaming, Being: Self and Consciousness in Neuroscience, Meditation, and
Philosophy. New York: Columbia University Press.
Tillemans, T. (2016). How do Mādhyamikas Think? And other essays on the Buddhist Philosophy of the
Middle, London: Wisdom Publications
Tricycle. (1997). Reincarnation: A Debate: Batchelor v. Thurman. Tricycle (Summer 1997).
https://tricycle.org/magazine/reincarnation-debate/. Accessed May 4, 2022.
Van Inwagen. P. (2002) Metaphysics, Westview Press
Williams, P. (1998) Studies in the philosophy of the Bodhicaryāvatāra: altruism and reality, Delhi: Motilal
Baranasidas
Williams, P. (2000) Response to Mark Siderits’ Review, Philosophy East 7 West, 50 (30): 424-535
Willson, M. (1987). Rebirth and the Western Buddhist. London: Wisdom Publications

22

View publication stats