Nuclear energy and obligations to the future


Nuclear energy and obligations to the future


V. Routley


Antipodean Antinuclearism: (Re)constructing Richard Routley/Sylvan's Nuclear Philosophy


January 1, 1978


This item was identified for digitisation at the request of The University of Queensland's 2020 Fryer Library Fellow, Dr. N.A.J. Taylor.


For all enquiries about this work, please contact the Fryer Library, The University of Queensland Library.




Journal Article


Inquiry, 21, 133-79

Nuclear Energy and Obligations
to the Future
R. and V. Routley
Plumwood Mountain, Braidwood, Australia

The paper considers the morality of nuclear energy development as it concerns
future people, especially the creation of highly toxic nuclear wastes requiring longterm storage. On the basis of an example with many parallel moral features it is
argued that the imposition of such costs and risks on the future is morally unacceptable. The paper goes on to examine in detail possible ways of escaping this
conclusion, especially the escape route of denying that moral obligations of the
appropriate type apply to future people. The bulk of the paper comprises discussion
of this philosophical issue, including many arguments against assigning obligations
to the future drawn both from analyses of obligation and from features of the future
such as uncertainty and indeterminacy. A further escape through appeal to moral
conflict is also considered, and in particular two conflict arguments, the Poverty
and Lights-going-out arguments are briefly discussed. Both these escape routes are
rejected and it is concluded that if the same standards of behaviour are applied to
the future as to the present, nuclear energy development is morally unacceptable.

I. The Bus Example
Suppose we consider a bus, a bus which we hope is to make a very long
journey. This bus, a third world bus, carries both passengers and freight.
The bus sets down and picks up many different passengers in the course of
its long journey and the drivers change many times, but because of the way
the bus line is managed and the poor service on the route it is nearly always
full to overcrowded, with passengers hanging off the back, and as in
Afghanistan, passengers riding on the roof, and chickens and goats in the
freight compartment.
Early in the bus's journey someone consigns on it, to a far distant
destination, a package containing a highly toxic and explosive gas. This is
packaged in a very thin container, which as the consigner well knows is
unlikely to contain the gas for the full distance for which it is consigned,
and certainly will not do so if the bus should encounter any trouble, for
example if there is a breakdown and the interior of the bus becomes very
hot, if the bus should strike a very large bump or pothole of the sort
commonly found on some of the bad roads it has to traverse, or if some

134 R. and V. Routley
passenger should interfere deliberately or inadvertently with the cargo or
perhaps try to steal some of the freight, as also frequently happens. All of
these things, let us suppose, have happened on some of the bus's previous
journeys. If the container should break the resulting disaster would probably kill at least some of the people and animals on the bus, while others
could be maimed or contract serious diseases.
There does not seem much doubt about what most of us would say about
the morality of the consigner's action, and there is certainly no doubt
about what the passengers would say. The consigner's action in putting
the safety of the occupants of the bus at risk is appalling. What could
excuse such an action, what sort of circumstances might justify it, and
what sort of case could the consigner reasonably put up? The consigner
might say that it is by no means certain that the gas will escape; he himself
is an optimist and therefore feels that such unfavourable possible outcomes should be ignored. In any case the bus might have an accident and
the passengers be killed long before the container gets a chance to leak; or
the passengers might change to another bus and leave the lethal parcel
He might say that it is the responsibility of the passengers and the driver
to ensure that the journey is a smooth one, and that if they fail to do so, the
results are not his fault. He might say that the journey is such a long one
that many of the passengers may have become mere mindless vegetables
or degenerate wretches about whose fate no decent person need concern
himself, or that they might not care about losing their lives or health or
possessions anyway by that time.
Most of these excuses will seem little more than a bad joke, and certainly would not usually be reckoned any sort of justification. The main
argument the consigner of the lethal parcel employs, however, is that his
own pressing needs justify his actions. He has no option but to consign his
potentially lethal parcel, he says, since the firm he owns, and which has
produced the material as a by-product, is in bad financial straits and
cannot afford to produce a better container or to stop the production of the
gas. If the firm goes out of business, the consigner says, his wife will leave
him, and he will lose his family happiness, the comfortable way of life to
which he has become accustomed and sees now as a necessity; his employees will lose their jobs and have to look for others; not only will the
firm's customers be inconvenienced but he, the consigner, will have to
break some business contracts; the inhabitants of the local village through
loss of spending and cancellation of the Multiplier Effect will suffer finan-

Obligations to the Future 135
cial hardship, and, worst of all, the tiny flow of droplets that the poor of the
village might receive (theoretically at any rate) as a result of the trickling
down of these good things would dry up entirely. In short, some basic and
some perhaps uncomfortable changes will be needed in the village.
Even if the consigner's story were accepted at face value - and it would
be wise to look critically at his story - only someone whose moral sensibilities had been paralysed by the disease of galloping economism could see
such a set of considerations, based on 'needs', comfort, and the goal of
local prosperity, as justifying the consigner's action.
One is not generally entitled to thus simply transfer the risks and costs
arising from one's own life onto other uninvolved parties, to get oneself
out of a hole of one's own making by creating harm or risk of harm to
someone else who has had no share in creating the situation. To create
serious risks and costs, especially risks to life or health for such others,
simply to avoid having to make some changes to a comfortable life style, or
even for a somewhat better reason, is usually thought deserving of moral
condemnation, and sometimes considered a crime; for example, the action
of a company in creating risks to the lives or health of its workers or
customers to prevent itself from going bankrupt. What the consigner says
may be an explanation of his behaviour, but it is not a justification.
The problem raised by nuclear waste disposal is by no means a perfect
analogy to the bus case, since, for example, the passengers on the nuclear
bus cannot get off the bus or easily throw out the lethal package. In many
crucial moral respects, however, the nuclear waste storage problem as it
affects future people, the passengers in the bus we are considering, resembles the consignment of the faultily packaged lethal gas. Not only are
rather similar moral principles involved, but a rather similar set of arguments to the lamentable excuses the consigner presents have been seriously put up to justify nuclear development, the difference being that in the
nuclear case these arguments have been widely accepted. There is also
some parallel in the risks involved; there is no known safe way to package
the highly toxic wastes generated by nuclear plants that will be spread
around the world if large-scale nuclear development goes ahead.1 The
wastes problem will not be a slight one, with each one of the more than
2,000 reactors envisaged by the end of the century, producing on average
annual wastes containing one thousand times the radioactivity of the
Hiroshima bomb.2 The wastes include not merely the spent fuels and their
radioactive by-products, but also everything they contaminate, from fuel
containers to the thousands of widely distributed decommissioned nuclear

136 R. and V. Routley
reactors which will have to be abandoned, still in a highly radioactive
condition, after the expiry of their expected lifetimes of about thirty years,
and which have been estimated to require perhaps one and a half million
years to reach safe levels of radioactivity.3 The wastes must be kept
suitably isolated from the environment for their entire active lifetime; for
fission products the required storage period averages a thousand years or
so, and for the actinides (transuranic elements) which include plutonium,
there is a half-million to a million-year storage problem.4
Serious problems have arisen with both short-term and proposed longterm methods of storage, even with the comparatively small quantities of
waste that have been produced over the last twenty years.5 With present
known short-term surface methods of storage there is a continued need for
human intervention to keep the material isolated from the environment,
while with proposed longer-term methods such as storage in salt mines or
granite to the risk of human interference there are added the risks of
leakage, e.g. through water seepage, and of disturbance, for example
through climatic change, earth movements, etc. The risks are significant:
no reasonable person with even a limited acquaintance with the history of
human affairs over the last 3,000 years could be confident of safe storage
by methods involving human intervention over the enormous time periods
involved. No one with even a slight knowledge of the geological and
climatic history of the earth over the last million years, a period which has
seen a number of ice ages and great fluctuations in climate for example,
could be confident that the waste material could be safely stored for the
vast periods of time required. Much of this waste is highly toxic; for
example, even a beachball sized quantity of plutonium appropriately
distributed is enough to give every person on the planet lung cancer - so
that a leak of even a small part of this waste material could involve huge
loss of life, widespread disease and genetic damage, and contamination of
immense areas of land.6
Given the enormous costs which could be involved for the future, it is
plainly grossly inadequ'ate'to merely speculate concerning untested, but
possibly or even probably, safe methods for disposal of wastes. Yet none
of the proposed methods has been properly tested, and they may prove to
involve all sorts of unforeseen difficulties and risks when an attempt is
made to put them into practice on a commercial scale. Only a method that
could provide a rigorous guarantee of safety over the storage period, that
placed safety beyond reasonable doubt, would be acceptable. It is difficult
to see how such rigorous guarantees could be given concerning either the

Obligations to the Future 137
geological or future human factors. But even if an economically viable,
rigorously safe long-term storage method could be devised, there is the
problem of guaranteeing that it would be universally and invariably used.
The assumption that it would be, especially if, as seems likely, such a
method proved expensive economically and politically, seems to presuppose a level of efficiency, perfection, and concern for the future not
previously encountered in human affairs, and certainly not conspicuous in
the nuclear industry.7 Again, unless we assume continuous and faultless
guarding of long-term storage sites through perhaps a million years of
possible future human activity, weapons-grade radioactive material will
be accessible, over much of the million-year storage period, to any party
who is in a position to retrieve it.
Our behaviour in creating this nightmare situation for the future is
certainly no better than that of the consigner in the bus example. Industrialized countries, in order to get out of a mess of their own making essentially the creation of economies dependent on an abundance" of
non-renewable energy in a situation where it is in fact in limited supply opt for a 'solution' which may enable them to avoid the making of uncomfortable changes during the lifetime of those now living, at the expense of
passing heavy burdens on to the inhabitants of the earth at a future time burdens in the shape of costs and risks which, just as in the bus case, may
adversely affect the life and health of future people and their opportunity
to lead a decent life.8
It is sometimes suggested that analogies like the bus example are defective; that morally they are crucially different from the nuclear case, since
future people, unlike the passengers in the bus, will benefit directly from
nuclear development, which will provide an abundance of energy for the
indefinite future. But this is incorrect. Nuclear fission creates wastes
which may remain toxic for a million years, but even with the breeder
reactor it could be an energy source for perhaps only 150 years. It will do
nothing for the energy problems of the people of the distant future whose
lives could be seriously affected by the wastes. Thus perhaps 30,000
generations of future people could be forced to bear significant risks,
without any corresponding benefits, in order to provide for the extravagant energy use of only five generations.
Nor is the risk of direct harm from the escape or misuse of radioactive
materials the only burden the nuclear solution imposes on the future.
Because the energy provided by nuclear fission is merely a stop-gap, it
seems probable that in due course the same problem, that of making a

138 R. and V. Routley
transition to renewable sources of energy, will have to be faced again by a
future population which will probably, again as a result of our actions, be
very much worse placed to cope with it.9 For they may well have to face
the change to renewable resources in an over-populated world not only
burdened with the legacy of our nuclear wastes, but also in a world in
which, if the nuclear proponents' dream of global industrialization is
realized, more and more of the global population will have become dependent on high energy consumption and associated technology and heavy
resource use, and will have lost or reduced its ability to survive without it.
It will, moreover, probably be a world which is largely depleted of non-renewable resources, and in which such renewable resources as forests and
soils as remain, resources which will have to form a very important part of
the basis of life, are in a run-down condition. Such points tell against the
idea that future people must be, if not direct beneficiaries of nuclear fission
energy, at least indirect beneficiaries.
The 'solution' then is to buy time for contemporary industrial society at
a price which not only creates serious problems for future people but
which reduces their ability to cope with those problems. Just as in the bus
case, contemporary industrial society proposes to get itself out of a hole of
its own making by creating risk of harm, and by transferring costs and
risks, to someone else who has had no part in producing the situation and
who will obtain no clear benefit. It has clear alternatives to this action.
That it does not take them is due essentially to its unwillingness to avoid
changing wasteful patterns of consumption and to its desire to protect the
interests of those who benefit from them.
If we apply to the nuclear situation the same standards of behaviour and
moral principles that we acknowledge (in principle if perhaps often not in
fact) in the contemporary world, it will not be easy to avoid the conclusion
that the situation involves injustice with respect to future people on a
grand scale. It seems to us that there are only two plausible moves that
might enable the avoidance of such a conclusion. First, it might be argued
that the moral principles and obligations which we acknowledge for the
contemporary world and the immediate future do not apply because the
recipients of our nuclear parcel are in the non-immediate future. Secondly,
an attempt might be made to appeal to overriding circumstances; for to
reject the consigner's action in the circumstances outlined is not of course
to say that there are no circumstances in which such an action might
possibly be justifiable, or at least where the case is less clearcut. It is the
same with the nuclear case. Just as in the case of the consigner of the

Obligations to the Future 139
package there is a need to consider what these justifying circumstances
might be, and whether they apply in the present case. We turn now to the
first of these possible escape routes for the proponent of nuclear development, to the philosophical question of our obligations to the future.

II. Obligations to the Distant Future
The area in which these philosophical problems arise is that of the distant
(i.e. non-immediate) future, that is, the future with which people alive
today will make no direct contact; the immediate future provides comparatively few problems for moral theories. The issues involved, although of
far more than academic interest, have not received any great attention in
recent philosophical literature, despite the fact that the question of obligations to future people presents tests which a number of ethical theories fail
to pass, and also raises a number of questions in political philosophy
concerning the adequacy of accepted institutions which leave out of
account the interests of future people.
Moral philosophers have predictably differed on the issue. But contrary
to the picture painted in a recent, widely read, and influential work
discussing it, Passmore's Man's Responsibility for Nature, a good many
philosophers who have explicitly considered the question have come
down in favour of the same consideration being given to the rights and
interests of future people as to those of contemporary or immediately
future people. Other philosophers have tended to fall into three categories
- those who acknowledge obligations to the future but who do not take
them seriously or who assign them less weight, those who deny, or who
are committed by their general moral position to denying, that there are
moral obligations beyond the immediate future, and those like Passmore
and Golding who come down, with admirable philosophical caution, on
both sides of the issue, but with the weight of the argument favouring the
view underlying prevailing economic and political institutions, that there
are no moral obligations to the future beyond those to the next generation.
• According to the most extreme of these positions against moral obligations to the future, our behaviour with respect to the future is morally
unconstrained; there are no moral restrictions on acting or failing to act
deriving from the effect of our actions on future people. Of those philosophers who say, or whose views imply, that we don't have obligations to
the (non-immediate) future, i.e. those who have opted for the uncon-

140 R. and V. Routley
strained position, many have based this view on accounts of moral obligation which are built on relations which presuppose some degree of temporal or spatial contiguity. Thus moral obligation is seen as grounded on or as
presupposing various relations which could not hold between people
widely separated in time (or sometimes in space). For example, obligation
is seen as grounded in relations which are proximate or of short duration
and also non-transitive. Among such suggested bases or grounds of moral
obligation, or requirements for moral obligation, which would rule out
obligations to the non-immediate future are these: First, there are those
accounts which require that someone to whom a moral obligation is held
be able to claim his rights or entitlement. People in the distant future will
not be able to claim rights and entitlements as against us, and of course
they can do nothing to enforce any claims they might have for their rights
against us. Secondly, there are those accounts which base moral obligations on social or legal convention, for example a convention which would
require punishment of offenders or at least some kind of social enforcement. But plainly these and other conventions will not hold invariantly
over change in society and amendment of legal conventions and so will not
be invariant over time. Also future people have no way of enforcing their
interests or punishing offenders, and there could be no guarantee that any
contemporary institution would do it for them.
Both the view that moral obligation requires the context of a moral
community and the view that it is contractually based appear to rule out
the distant future as a field of moral obligation, as they not only require a
commonality, or some sort of common basis, which cannot be guaranteed
in the case of the distant future, but also a possibility of interchange or
reciprocity of action which cannot apply to the future. Where the basis of
moral obligation is seen as mutual exchange, the interests of future people
must be set aside because they cannot change the past and cannot be
parties to any mutual contract. The exclusion of moral obligations to the
distant future also follows from those views which attempt to ground
moral obligations in non-transitive relations of short duration such as
sympathy and love. There are some difficulties also about love and sympathy for (non-existent) people in the far distant future about whose
personal qualities and characteristics one must know very little and who
may well be committed to a life-style for which one has no sympathy. On
the current showing in the case of nuclear energy it would be easy to
conclude that contemporary society lacks both love and sympathy for
future people; and it would appear to follow from this that contemporary

Obligations to the Future 141
people have no obligations to future people and can harm them as it suits
What all these views have in common is a naturalistic picture of obligation as something acquired, either individually or institutionally, something which is conditional on doing something or failing to do something
(e.g. participating in the moral community, contracting), or having some
characteristic one can fail to have (e.g. love, sympathy, empathy).10
Because obligation therefore becomes conditional, features usually
thought to characterize it, such as universality of application and necessitation (i.e. the binding features), are lost, especially where there is a choice
of whether or not to do the thing required to acquire the obligation, and so
of whether to acquire it. The criteria for acquisition suggested are such as
to exclude people in the distant future.
However, the view that there are no moral constraints with respect to
future people, that one is free to act as one likes with respect to them, is a
very difficult one to sustain. Consider the example of a scientific group
which, for no particular reason other than to test a particular piece of
technology, places in orbit a cobalt bomb which is to be set off by a
triggering device designed to go off several hundred years from the time of
its despatch. No presently living person and none of their immediate
descendants would be affected, but the population of the earth in the
distant future would be wiped out as a direct and predictable result of the
action. The unconstrained position clearly implies that this is an acceptable moral enterprise, that whatever else we might legitimately criticize in
the scientists' experiment, perhaps its being unduly expensive or badly
designed, we cannot lodge a moral protest about the damage it will do to
future people. The unconstrained position also endorses as morally acceptable the following sorts of examples: A firm discovers it can make a
handsome profit from mining, processing, and manufacturing a new type
of material which, although it causes no problems for present people or
their immediate descendants, will over a period of hundreds of years
decay into a substance which will cause an enormous epidemic of cancer
among the inhabitants of the earth at that time. According to the unconstrained view the firm is free to act in its own interests, without any
consideration for the harm it does to future people.
Such counterexamples to the unconstrained view might seem childishly
obvious. Yet the unconstrained position concerning the future from which
they follow is far from being a straw man; not only have a number of
philosophers writing on the issue endorsed this position, but it is the clear

142 R. and V. Routley
implication of many currently popular views of the basis of moral obligation, as well as of economic theory. It does not appear, on the other hand,
that those who opt for the unconstrained position have considered such
examples and endorsed them as morally acceptable, despite their being
clearly implied by their position. We suspect that when it is brought out
that the unconstrained position admits such counterexamples, that being
free to act implies, among other things being free to inflict pointless harm,
most of those who opted for the unconstrained position would want to
assert that it was not what they intended. What those who have put
forward the unconstrained position seem to have had in mind in denying
moral obligation is rather that future people can look after themselves, that
we are not responsible for their lives. The view that the future can take
care of itself also seems to assume a future causally independent of the
present. But it is not. It is not as if, in cases such as those discussed above
and the nuclear case, the future is simply being left alone to take care of
itself. Present people are influencing it, and in doing so must acquire many
of the same sorts of moral responsibilities as they do in causally affecting
the present and immediate future. The thesis seems thus to assume an
incorrect model of an independent and unrelated future.
Also, to say that we are not responsible for the lives of future people
does not amount to the same as saying that we are free to do as we like with
respect to them, that there are no moral constraints on our action involving
them. In just the same way, the fact that one does not have, or has not
acquired, an obligation to some stranger with whom one has never been
involved - that one has no responsibility for his life - does not imply that
one is free to do what one likes with respect to him, for example to rob him
or to pursue some course of action of advantage to oneself which could
seriously harm him.
These difficulties for the unconstrained position arise in part from the
failure to make an important distinction between, on the one hand, acquired or assumed obligations towards somebody, for which some act of
acquisition or assumption is required as a qualifying condition, and on the
other hand moral constraints, which require, for example, that one should
not act so as to damage or harm someone, and for which no act of
acquisition is required. There is a considerable difference in the level and
kind of responsibility involved. In the first case one must do something or
be something which one can fail to do or be, e.g. have loves, sympathy, be
contracted. In the second case responsibility arises as a result of being a
causal agent aware of the consequences or probable consequences of his

Obligations to the Future 143
action, and thus does not have to be especially acquired or assumed. Thus
there is no problem about how the latter class, moral constraints, can
apply to the distant future in cases where it may be difficult or impossible
for acquisition or assumption conditions to be satisfied. They apply as a
result of the ability to produce causal effects on the distant future of a
reasonably predictable nature. Thus also moral constraints can apply to
what does not (yet) exist, just as actions can cause results that do not (yet)
exist. While it may be the case that there would need to be an acquired or
assumed obligation in order for it to be claimed that contemporary people
must make special sacrifices of an heroic kind for future people, or even to
help them especially, only moral constraints are needed in order for us to
be constrained from harming them. Thus, to return to the bus example, the
consigner cannot argue in justification of his action that he has never
assumed or acquired responsibility for the passengers, that he does not
know them and therefore has no love or sympathy for them, and that they
are not part of his moral community, in short that he has no special
obligations to help them. All that one needs to argue in respect of both the
bus and the nuclear case is that there are moral constraints against harming, not that there are specially acquired obligations to take responsibility
for the lives of the people involved.
The confusion of moral constraints with acquired obligation, and the
attempt therewith to view all constraints as acquired and to write off
non-acquired constraints, is facilitated through the use of the term 'moral
obligation' in philosophy to indicate any type of deontic constraint, while
in natural language it is used to indicate something which has to be
assumed or acquired. Hence the equation and at least one root of the
unconstrained position, that is of the belief that there are no moral constraints concerning the distant future.
The unconstrained view tends to give way, under the weight of counterexamples, to a more qualified, and sometimes ambivalent position. Passmore's position in [1] is a striking example of the second ambivalent
position. On the one hand Passmore regularly gives the impression of one
championing future people; for example, in the final sentence of [1] he
says, concerning men a century hence:
My sole concern is that we should do nothing which will reduce their
freedom of thought and action, whether by destroying.the natural
world which makes that freedom possible or the social traditions
which permit and encourage it.

144 R. and V. Routley
Earlier (esp. pp. 84-85) Passmore appears to endorse the principle 'that we
ought not to act so as certainly to harm posterity' and claims (p. 98) that,
even where there are uncertainties, 'these uncertainties do not justify
negligence'. Nevertheless, though obligations concerning non-immediate
posterity are thus admitted, the main thrust of Passmore's argument is
entirely different, being in favour of the unconstrained position according
to which we have no obligations to non-immediate posterity. Thus his
conclusion (p. 91):
So whether we approach the problem of obligations to posterity by
way of Bentham and Sidgwick, Rawls or Golding, we are led to
something like the same conclusion: our obligations are to immediate
posterity, we ought to try to improve the world so that we shall be able
to hand it over to our immediate successors in a better condition, and
that is all.11
Passmore's position is, to all appearances, simply inconsistent. There are
two ways one might try to render it consistent, but neither is readily
available to Passmore. The first is by taking advantage of the distinction
between moral constraints and acquired obligations, but a basis for this
distinction is not evident in Passmore's work and indeed the distinction is
antithetical to the analyses of obligation that Passmore tries to synthesize
with his own analysis in terms of loves. The second, sceptical, route to
consistency is by way of the argument that we shall consider shortly, that
there is always gross uncertainty with respect to the distant future, uncertainty which relieves us in practice of any moral constraints regarding the
distant future. But though Passmore's writing strongly suggests this uncertainty argument (especially his sympathetic discussion of the Premier
of Queensland's argument against conservationists [p. 77]), he also rules it
out with the claim that uncertainties do not justify negligence.12
Many of the accounts of moral obligation that give rise to the unconstrained position are fused in Passmore's work, again not entirely consistently, since the different accounts exploited do not give uniform results.
Thus the primary account of obligation is said to be in terms of loves though the account is never satisfactorily formulated or developed - and it
is suggested that because our loves do not extend into the distant future,
neither do our obligations. This sentimental account of obligation will
obviously lead to different results from utilitarian accounts of obligation,
which however Passmore appeals to in his discussion of wilderness. In yet
other places in [1], furthermore, social contract and moral community

Obligations to the Future 145
views are appealed to - see, e.g., the treatment of animals, of preservation, and of duties to nature. In the case of obligations to future people,
however, Passmore does try to sketch an argument - what we call the
convergence argument - that all the accounts lead in the end to the
unconstrained position.
As well as the convergence argument, and various uncertainty arguments to be considered later, Passmore appears to endorse several other
arguments in favour of his theme that there are in practice no obligations to
the distant future. In particular, he suggests that such obligations would in
practice be otiose. Everything that needs to be accounted for can be
encompassed through the chain picture of obligation as linking successive
generations, under which each generation has obligations, based on loves,
only to the succeeding generation. We outline three objections to this
chain account. First, it is inadequate to treat constraints concerning the
future as if they applied only between generations, as if there were no
question of constraints on individuals as opposed to whole generations,
since individuals can create causal effects, e.g. harm, on the future in a
way which may create individual responsibility, and which can't necessarily be sheeted home to an entire generation. Secondly, such chains, since
they are non-transitive, cannot yield direct obligations to the distant
future. But for this very reason the chain picture cannot be adequate, as
examples again show. For the picture is unable to explain several of the
cases that have to be dealt with, e.g. the examples already discussed which
show that we can have a direct effect on the distant future without
affecting the next generation, who may not even be able to influence
matters.13 Thirdly, improvements for immediate successors may be
achieved at the expense of disadvantages to people of the more distant
future. Improving the world for immediate successors is quite compatible
with, and may even in some circumstances be most easily achieved by,
ruining it for less immediate successors. Such cases can hardly be written
off as 'never-never land' examples, since many cases of environmental
exploitation might be seen as of just this type, e.g. not just the nuclear case
but also the exhaustion of non-renewable resources and the long-term
depletion of renewable resources such as soils and forests through overcropping. If then such obvious injustices to future people arising from the
favouring or exclusive concern with immediate successors are to be avoided, obligations to the future will have to be seen as in some way fairly
distributed over time, and not merely as accruing to particular generations
in the way the chain picture suggests.

146 R. and V. Routley
Passmore tries to represent all obligations to the distant future in terms
of heroic self-sacrifice, something which cannot of course be morally
required. But in view of the distinctions between constraints and acquired
obligation and between obligation and supererogation, this is just to misrepresent the position of these obligations. For example, one is no more
engaging in heroic self-sacrifice by not forcing future people into an
unviable life position or by refraining from causing them direct harm, than
one is resorting to heroic self-sacrifice in refraining from beating and
robbing some stranger and leaving him to starve.
Passmore's most sustained argument for the unconstrained position is a
convergence argument, that different analyses of obligations, including
his own, lead to the one conclusion. This style of argument is hardly
convincing when there are well-known accounts of obligation which do
not lead to the intended conclusion, e.g. deontological accounts such as
those of Kant and of modern European schools, and teleological accounts
such as those of Moore (in [8]). But such unfavourable positions are either
rapidly passed over or ignored in Passmore's historical treatment and
narrow selection of historical figures. The style of argument becomes even
less persuasive when it is discovered that the accounts of the main authorities appealed to, Bentham, Sidgwick, and Rawls,14 do not lead, without
serious distortion, to the intended conclusion. Indeed Passmore has twisted the historical and textual evidence to suit his case, as we now try to
Consider Bentham first. Passmore's assumption, for which no textual
evidence is cited, ls is that no Benthamite calculation can take account of a
future more extensive than the immediate future (cf. pp. 87-88). The
assumption seems to be based simply on the fact that Bentham remarked
that 'the value of the pleasure or pain to each person to be considered in
any estimate will be greater or less in virtue of the following circumstances'. '3. Its certainty or uncertainty. 4. Its propinquity or remoteness'
([10], p. 16). But this does nothing to show that future persons are discounted: the certainty and propinquity do not concern persons, but the
utilities of the persons concerned. As regards which persons are concerned in any calculation Bentham is quite explicit, detailing how
to take an exact account. . . of the general tendency of any act.. . .5 .
Take an account of the number of persons whose interests appear to
be concerned; and repeat the above process [summation of values of
pleasure and of pain] with respect to each. ([10], p. 16)

Obligations to the Future 147
It follows that Bentham's calculation takes account of everyone (and, in
his larger scheme, every sentient creature) whose interests appear to be
concerned: if the interests of people in the distant future appear to be
concerned - as they are in conservation issues - they are to be included in
the calculation. And there is independent evidence16 that in Bentham's
view the principle of utility was not temporally restricted: 'that is useful
which, taking all times and all persons into consideration, leaves a balance
of happiness' ([10], pp. 17-18, our italics). Thus the future cut-off that
Passmore has attributed to Bentham is contradicted by Bentham's own
The case of Sidgwick is more complex, because there is isolated oscillation in his application of utilitarianism between use of utility and of
(something like) expected utility (see [11], pp. 381,414): Sidgwick's utilitarianism is, in its general characterization, essentially that of Bentham:
the conduct which . . . is objectively right is that which will produce
the greatest amount of happiness on the whole; that is, taking into
account all those whose happiness is affected by the conduct. ([11], p.
All includes all sentient beings, both existing and to exist, as Sidgwick goes
on to explain (p. 414). In particular, in answer to the question 'How far are
we to consider the interests of posterity when they seem to conflict with
those of existing human beings?' Sidgwick writes ([11], p. 414, our italics):
It seems, however, clear that the time at which a man exists cannot
affect the value of his happiness from a universal point of view; and
that the interests of posterity must concern a Utilitarian as much as
those of his contemporaries, except in so far as the effect of his
actions on posterity - and even the existence of human beings to be
affected - must necessarily be more uncertain.
But Passmore manages, first of all, to give a different sense to what
Sidgwick is saying by adjusting the quotation, by omitting the clause we
have italicized, which equalizes the degree of concern for present and
future persons, and by italicizing the whole except-clause, thereby placing
much greater emphasis than Sidgwick does on uncertainty. For according
to Sidgwick's impartiality principle, 'the mere difference in time is not a

148 R. and V. Routley
reasonable ground for having more regard to the consciousness of one
-amount than to that of another' ([11], p. 381; see also p. 124). The apparent
tension in Sidgwick's theory as to whether uncertainty should be taken
into account is readily removed by resort to a modern distinction between
values and expected values (i.e. probability weighted values); utilitarian
rightness is defined as before in terms of the net happiness of all concerned
over all time without mention of uncertainty or probabilities, but it is
distinguished from probable rightness (given present information), in the
utilitarian sense,17 which is defined in terms of the expected net happiness
of those concerned, using present probabilities. It is the latter notion, of
probable rightness, that practical reasoning is commonly concerned with
and that decision theory studies; and it is this that Passmore supposes
Sidgwick is using ([1], p. 84). But it is evident that the utilitarian determination of probable rightness, like that of rightness, will sometimes take
into account the distant future - as Sidgwick's discussion of utilitarian
determination of optimum population (immediately following his remark
on uncertainty) does. So how does Passmore contrive to reverse matters,
to have Sidgwick's position lead to his own unconstrained conclusion?
The answer is: By inserting an additional assumption of his own - which
Sidgwick would certainly have rejected - that the uncertainties entitle us
to ignore the distant future. What Passmore has implicitly assumed in his
claim ([1], p. 85) that 'utilitarian principles [such as Sidgwick's] are not
strong enough' 'to justify the kinds of sacrifice some conservationists now
call upon us to make' is his own thesis that 'The uncertainty of harms we
are hoping to prevent would in general entitle us to ignore them.. .'. From
a decision-theory viewpoint this is simply irrational18 unless the probabilities of damage are approaching zero. We will deal with the essentially
sceptical uncertainty arguments on which Passmore's position depends
shortly: here it is enough to observe that Sidgwick's position does not lead
to anything like that which Passmore attributes to him - without uncertainty assumptions which Sidgwick would have rejected (for he thought
that future people'will certainly have pleasure and suffer pain).
We can also begin to gauge from Passmore's treatment of nineteenthcentury utilitarians, such as Bentham and Sidgwick, the extent of the
distortion which underlies his more general historical case for the unconstrained position which, so he claims,
represents accurately enough what, over the last two centuries, men
have seen as their duty to posterity as a whole. . . . ([1], p. 91)

Obligations to the Future 149
The treatment accorded Rawls in only marginally more satisfactory.
Passmore supposes that Rawls's theory of justice leads directly to the
unconstrained position ([1], p. 87 and p. 91), whereas Rawls claims ([5], p.
293) that we have obligations to future people just as to present ones. But
the situation is more complicated than Rawls's claim would indicate, as we
now try to explain in a summary way (more detail is given in the Appendix). For, in order to justify this claim on his theory (with its present
time-of-entry interpretation), Rawls has to invoke additional and dubious
motivational assumptions; even so the theory which thus results does not
yield the intended conclusion, but a conclusion inconsistent with Rawls's
claim. However, by changing the time-of-entry interpretation to an omnitemporal one, Rawls's claim does result from the theory so amended.
Moreover, the amended theory also yields, by exactly Rawls's argument
for a just saving rate, a resource conservation policy, and also a case
against nuclear development. Accordingly Passmore's other claims regarding Rawls are mistaken, e.g. that the theory cannot justify a policy of
resource conservation. Rawls does not emerge unscathed either. As on
the issue of whether his contract is a necessary condition for obligations,
so on obligations which the contract yields to the distant future, Rawls is
far from consistent. Furthermore, institutions such as qualified market
and voting systems are recommended as just though from a future perspective their results are far from that. Rawls, then, does not take obligations to the future with full seriousness.
In sum, it is not true that the theory of Rawls, any more than the theories
of the historical figures actually discussed by Passmore, unequivocally
supports the unconstrained position.

III. Uncertainty and Indeterminacy Arguments
Although there are grave difficulties for the unconstrained position, qualification leads to a more defensible position. According to the qualified
position we are not entirely unconstrained with respect to the distant
future: there are obligations, but these are not so important as those to the
present, and the interests of distant future people cannot weigh very much
in the scale against those of the present and immediate future. The interests of future people then, except in unusual cases, count for very much
less than the interests of present people. Hence such things as nuclear
development and various exploitative activities which benefit present

150 R. and V. Routley
people should proceed, even if people of the distant future are disadvantaged by them.
The qualified position appears to be widely held and is implicit in most
modern economic theories, where the position of a decrease in weight of
future costs and benefits (and so of future interests) is obtained by application over time of an (opportunity cost) discount rate. The attempt to
apply economics as a moral theory, something that is becoming increasingly common, can lead then to the qualified position. What is objectionable in such an approach is that economics must operate within the bounds
of moral (deontic) constraints, just as in practice it operates within legal
constraints, and cannot determine what those constraints are. There are,
moreover, alternative economic theories and simply to adopt one which
discounts the future is to beg all the questions at issue. The discounting
move often has the same result as the unconstrained position; if, for
instance, we consider the cancer example and consider costs as payable
compensation, it is evident that, over a sufficiently long period of time,
discounting at current prices would lead to the conclusion that there are no
recoverable damages and so, in economic terms, no constraints. In short,
even certain damage to future people could be written off. One way to
achieve the bias against future people is by the application of discount
rates which are set in accord with the current economic horizons of no
more than about fifteen years,19 and application of such rates would
simply beg the question against the interests and rights of future people.
Where there is certain future damage of a morally forbidden type the
whole method of discounting is simply inapplicable, and its use would
violate moral constraints.20
Another argument for the qualified position, which avoids the objections from cases of certain damage, comes from probability considerations.The distant future, it is argued, is much more uncertain than the
present and immediate future, so that probabilities are consequently lower, perhaps even approaching or coinciding with zero for any hypothesis
concerning the distant future.21 But then if we take account of probabilities in the obvious way, by simply multiplying them against costs and
benefits, it is evident that the interests of future people, except in cases
where there is an unusually high degree of certainty, must count for (very
much) less than those of present and neighbouring people where (much)
higher probabilities obtain. So in the case of conflict between the present
and the future where it is a question of weighing certain benefits to the
people of the present and the immediate future against a much lower

Obligations to the Future 151
probability of indeterminate costs to an indeterminate number of distant
future people, the issue would normally be decided in favour of the
present, assuming that anything like similar costs and benefits were involved. But of course it can't be assumed that anything like similarly
weighted costs and benefits are involved in the nuclear case, especially if it
is a question of risking poisoning some of the earth for half a million or so
years, with consequent risk of serious harm to thousands of generations of
future people, in order to obtain quite doubtful or trivial benefits for some
present people, in the shape of the opportunity to continue unnecessarily
high energy use. And even if the costs and benefits were comparable or
evenly weighted, such an argument would be defective, since an analogous argument would show that the consigner's action is acceptable
provided the benefit, e.g. the profit he stood to gain from imposing significant risks on other people, was sufficiently large. Such a cost-benefit
approach to moral and decision problems, with or without the probability
frills, is quite inadequate where different parties are concerned, or for
dealing with cases of conflict of interest or moral problems where deontic
constraints are involved, and commonly yields counterintuitive results.
For example, it would follow on such principles that it is permissible for a
firm to injure, or very likely injure, some innocent party provided the firm
stands to make a sufficiently large gain from it. But the costs and benefits
involved are not transferable in any simple or general way from one party
to another. Transfers of this kind, of costs and benefits involving different
parties, commonly raise moral issues - e.g. is x entitled to benefit himself
by imposing costs ony ? - which are not susceptible to a simple cost-benefit approach of the sort adopted by some proponents of nuclear energy,
who attempt to dismiss the costs to future people with the soothing remark
that any development involves costs as well as benefits. The transfer point
is enough to invalidate the comparison, heavily relied on by McCracken
[16] in building a case for the acceptability of the nuclear risk, between
nuclear risks and those from cigarette smoking. In the latter case those
who supposedly benefit from the activity are also, to an overwhelming
extent* those who bear the serious health costs and risks involved. In
contrast the users and supposed beneficiaries of nuclear energy will be
risking not only, or even primarily, their own lives and health, but also
those of others who may be non-beneficiaries and who may be spatially or
temporally removed, and these risks will not be in any direct way related
to a person's extent of use.
The transfer objection is essentially the same as that to the utilitarian's

152 R. and V. Routley
happiness sums as a way of solving moral conflict between different
parties, and the introduction of probability considerations does not change
the principles involved but merely complicates analyses. One might further object to the probability argument that probabilities involving distant
future situations are not always less than those concerning the immediate
future in the way the argument supposes, and that the outcomes of some
moral problems such as the bus example do not depend on a high level of
probability anyway. In some sorts of cases it is enough, as the bus example
reveals, that a significant risk is created; such cases do not depend critically on high probability assignments.
Uncertainty arguments in various forms are the most common and
important ones used by philosophers and others to argue for the position
that we cannot be expected to take serious account of the effects of our
actions on the distant future. There are two strands to the uncertainty
argument, capable of separation, but frequently entangled. Both arguments are mistaken, the first on a priori grounds, the second on a posteriori grounds. The first argument is a generalized uncertainty argument
which runs as follows: In contrast to the exact information we can obtain
about the present, the information we can obtain about the effects of our
actions on the distant future is unreliable, woolly, and highly speculative.
But we cannot base assessments of how we should act on information of
this kind, especially when accurate information is obtainable about the
present which would indicate different action. Therefore we must regretfully ignore the uncertain effects of our actions on the distant future. More
formally and crudely: One only has obligations to the future if these
obligations are based on reliable information; there is no reliable information at present as regards the distant future; therefore one has no obligations to the distant future.
The first argument is essentially a variation on a sceptical argument in
epistemology concerning our knowledge of the future (formally, replace
'obligations' by 'knowledge' in the crude statement of the argument
above). The main ploy is to considerably overestimate and overstate the
degree of certainty available with respect to the present and immediate
future, and the degree of certainty which is required as the basis for moral
consideration both with respect to the present and with respect to the
future. Associated with this is the attempt to suggest a sharp division as
regards certainty between the present and immediate future on the one
hand and the distant future on the other. We shall not find, we suggest, that
there is any such sharp or simple division between the distant future and

Obligations to the Future 153
the adjacent future and present, at least with respect to those things in the
present which are normally subject to moral constraints. We can and
constantly do act on the basis of such 'unreliable' information as the
sceptic as regards the future conveniently labels 'uncertainty'; for scepticproof certainty is rarely, or never, available with respect to much of the
present and immediate future. In moral situations in the present, action
often takes account of risk and probability, even quite low probabilities. A
good example is again the bus case. We do not need to know for certain
that the container will break and the lethal gas escape. In fact it does not
even have to be probable, in the relevant sense of more probable than not,
in order for us to condemn the consigner's action. It is enough that there is
a significant risk of harm in this sort of case. It does not matter if the
decreased well-being of the consigner is certain and the prospects of the
passengers quite uncertain; the resolution of the problem is still clearly in
favour of the so-called 'speculative' and 'unreliable'. But if we do not
require certainty of action to apply moral constraints in contemporary
affairs, why should we require a much higher standard of certainty in the
future? Why should we require epistemic standards for the future which
the more familiar sphere of moral action concerning the present and
adjacent future does not need to meet? The insistence on certainty as a
necessary condition before moral consideration can be given to the distant
future, then, amounts to an epistemic double standard. But such an
epistemic double standard, proposed in explaining the difference between
the present and the future and to justify ignoring future peoples' interests,
in fact cannot itself provide an explanation of the differences, since it
already presupposes different standards of certainty appropriate to each
class, which difference is in turn in need of justification.
The second uncertainty argument is a practical uncertainty argument,
that whatever our theoretical obligations to the future, we cannot in
practice take the interests of future people into account, because uncertainty about the distant future is so gross that we cannot determine what
the likely consequences of actions upon it will be and therefore, however
good our intentions to the people of the distant future, in practice we have
no choice but to ignore their interests. Uncertainty is gross where certain
incompatible hypotheses are as good as one another and there is no
rational ground for choosing between them. The second uncertainty argument can also be put in this way: If moral principles are, like other
principles, implicational in form, that is of such forms as 'if* has character
h then x is wrong, for every (action) x', then what the argument claims is

154 R. and V. Routley
that we can never obtain the information about future actions which would
enable us to detach the antecedent of the implication. So even if moral
principles theoretically apply to future people, in practice they cannot be
applied to obtain clear conclusions or directions concerning contemporary
action of the 'It is wrong to do x' type.
Many of the assumptions of the second argument have to be conceded.
If the distant future really is so grossly uncertain that in every case it is
impossible to determine in any way that is better than chance what the
effects of present action will be, and whether any given action will help or
hinder future people, then moral principles, although they may apply
theoretically to the future, will not be applicable in practice for obtaining
any clear conclusions about how to act. Hence the distant future will
impose no practical moral constraints on action. However, the argument
is factually incorrect in assuming that the future is always so grossly
uncertain or indeterminate. Admittedly there is often a high degree of
uncertainty concerning the distant future, but as a matter of (contingent)
fact it is not always so gross or sweeping as the argument has to assume.
There are some areas where uncertainty is not so great as to exclude
constraints on action, especially when account is taken of the point,
noticed in connection with the first argument, that complete certainty is
commonly not required for moral constraints and that all that may be
needed in some cases is the creation of"a significant risk. Again there is
considerable uncertainty about many factors which are not highly, or at
all, morally relevant, but this does not extend to many factors which are of
much greater importance to moral issues. For example, we may not have
any idea what the fashions will be in a hundred years in girls' names or
men's footwear, or what brands of ice cream people will be eating if any,
but we do have excellent reason to believe, especially if we consider 3,000
years of history, that what people there are in a hundred years are likely to
have material and psychic needs not entirely unlike our own, that they will
need a healthy biosphere for a good life; that like us they will not be
immune to radiation; that their welfare will not be enhanced by a high
incidence of cancer or genetic defects, by the destruction of resources, or
the elimination from the face of the earth of that wonderful variety of
non-human life which at present makes it such a rich and interesting place.
For this sort of reason, the second uncertainty argument should be rejected. While it is true that there are many areas in which the morally relevant
information needed is uncertain or unavailable, and in which we cannot
therefore determine satisfactorily how to act, there are certainly others in

Obligations to the Future 155
which uncertainty in morally relevant areas is not so great as to preclude
moral constraints on action, where we ascertain if not absolute certainties
at least probabilities of the same sort of order as are considered sufficient
for the application of moral principles in parallel contemporary cases,
especially where spatially remote people are involved. The case of nuclear
waste storage, and of uncertainty of the effects of it on future people,
seems to be of the latter sort. Here there is no gross indeterminacy or
uncertainty; it is simply not true that incompatible hypotheses about what
may happen are as good as each other. It is plain that nuclear waste storage
does impose significant risks of harm on future people, and, as we can see
from the bus example, the significant risk of harm is enough in cases of this
type to make moral constraints applicable.
In terms of the defects of the preceding uncertainty arguments, we can
see the corresponding defects in a number of widely employed uncertainty
arguments used to write off probable harm to future people as outside the
scope of proper consideration. Most of these popular moves employ both
of the uncertainty arguments as suits the case, switching from one to the
other in a way that is again reminiscent of sceptical moves. For example,
we may be told that we cannot really take account of future people because
we cannot be sure that they will exist or that their tastes and wants will not
be completely different from our own, to the point where they will not
suffer from our exhaustion of resources or from the things that would
affect us (cf. Passmore [1]). But this is to insist upon complete certainty of
a sort beyond what is required for the present and immediate future, where
there is also commonly no guarantee that some disaster will not overtake
those we are morally committed to. Again we may be told that there is no
guarantee that future people will be worthy of any efforts on our part,
because they may be morons or forever plugged into enjoyment- or other
machines (Golding [12]). Even if one is prepared to accept the elitist
approach presupposed - according to which only those who meet certain
properly civilized or intellectual standards are eligible for moral consideration - what we are being handed in such arguments as a serious defeating
consideration is again a mere outside possibility - like the sceptic who says
that the solid-looking desk in front of us is perhaps only a facade, not
because he has any particular reason for doing so, but because he hasn't
looked around the back, drilled holes in it, etc. Neither the contemporary
nor the historical situation gives any positive reason for supposing that a
lapse into universal moronity or universal pleasure-machine escapism is a
serious possibility, as opposed to a logical possibility. We can contrast

156 R. and V. Routley
with these mere logical possibilities the very real historically supportable
risks of escape of nuclear waste or decline of a civilization through destruction of its resource base.
The possibilities just considered in these uncertainty arguments of
sceptical character are not real possibilities. Another argument which may
consider a real possibility, but still does not succeed in showing that it is
acceptable to proceed with an action which would appear to be harmful to
future people, is often introduced in the nuclear waste case. This is the
argument that future people may discover a rigorously safe and permanent
storage method for nuclear wastes before they are damaged by escaped
waste material. Let us grant for the sake of the argument that this is a real
possibility (though physical arguments may show that it is not). This still
does not affect the fact that there is a significant risk of serious damage and
that the creation of a significant risk is enough to rule out an action of this
type as morally impermissible. In just the same way, future people may
discover a cure for cancer, and the fact that this appears to be a real and not
merely a logical possibility, does not make the action of the firm in the
example discussed above, of producing a substance likely to cause cancer
in future people, morally admissible. The fact that there was a real possibility of future people avoiding the harm would show that actions of these
sorts were admissible only if what was required for inadmissibility was
certainty of harm or a very high probability of it. In such cases, before such
actions could be considered admissible, what would be required is far
more than a possibility, real or not22 - it is at least the availability of an
applicable, safe, and rigorously tested, not merely speculative, technique
for achieving it, something that future people could reasonably be expected to apply to protect themselves.
The strategy of most of these uncertainty arguments is fairly clear then,
and may be brought out by looking yet again at the bus example, where the
consigner says that he cannot be expected to take account of the effect of
his actions on the passengers because they may find an effective way to
deal with his parcel or some lucky or unlucky accident may occur, e.g. the
bus may break down and they may all change to a different bus leaving the
parcel behind, or the bus may crash, killing all the passengers before the
container gets a chance to leak. These are all possibilities, of course, but
there is no positive reason to believe that they are any more than that, that
is they are not real possibilities. The strategy is to stress such outside
possibilities in order to create the false impression that there is gross
uncertainty about the future, that the real possibility that the container will

Obligations to the Future 157
break should be treated in the same way as these mere logical possibilities,
that uncertainty about the future is so great as to preclude the consigners'
taking account of the passengers' welfare and of the real possibility of
harm from his parcel, and thereby excuse his action. A related strategy is
to stress a real possibility, such as finding a cure for cancer, and thereby
imply that this removes the case for applying moral constraints .This move
implicitly makes the assumptions of the first argument, that certainty, or at
least a very high probability, of harm is required before an action can be
judged morally inadmissible, and the point of stressing the real possibility
of avoidance of damage is to show that this allegedly required high degree
of certainty or probability cannot be attained. That is, the strategy draws
attention to some real uncertainty implying that this is sufficient to defeat
the application of moral constraints. But, as we have seen, this is often not
An argument closely related to the uncertainty arguments is based on
the non-existence and indeterminacy of the future.23 An item is indeterminate in a given respect if its properties in that respect are, as a matter of
logic, not settled (nor are they settlable in a non-arbitrary fashion). The
respects in which future items are indeterminate are well enough known
for a few examples to serve as reminders: all the following are indeterminate: the population of Australia at 2001, its distribution, its age structure,
the preferences of its members for folk music, wilderness, etc., the size
and shape of Wollongong, the average number of rooms in its houses and
in its office blocks, and so on. Philosophical discussion of such indeterminacy is as old as Aristotle's sea battle and as modern as truth-value gaps
and fuzzy logics, and many positions have been adopted on the existence
and determinacy of future items. Nevertheless theories that there are
obligations to the future are not sensitive to the metaphysical position
adopted concerning the existence or non-existence of the future. Any
theory which denied obligations to the future on the metaphysical grounds
that the future did not exist, and did not have properties, so that the
present could not be related to it, would be committed to denying such
obvious facts as that the present could causally influence the future, that
present people could be great-grandparents of purely future people, and so
on, and hence would have to be rejected on independent grounds. This is
not to say that there are not important problems about the existence or
non-existence of future items, problems which are perhaps most straightforwardly handled by a Meinongian position which allows that items
which do not exist may have properties. The non-existence of the future

158 R. and V. Routley
does raise problems for standard theories which buy the Ontological
Assumption (the thesis that what does not exist does not have properties),
especially given the natural (and correct) inclination to say that the future
does not (now) exist; but such theories can adopt various strategies for
coping with these problems (e.g. the adoption of a platonistic position
according to which the future does now exist, or the allowance for certain
sorts of relations between existents at different times), although the satisfactoriness of these strategies is open to question (cf. [4]). Thus whether or
not the Ontological Assumption is assumed and however it is applied, it
will be allowed that future items will have properties even if they do not
have them now, and that is enough to provide the basis for moral concern
about the future. Thus the thesis of obligations to the future does not
presuppose any special metaphysical position on the existence of the
If the non-existence of future items creates no special problems for
obligations to the future, the same is not true of their indeterminacy.
Whether the indeterminacy of future items is seen as a logical feature of
the future which results from the non-existence of purely future items or
whether one adopts a (mistaken) platonistic view of the future as existing
and sees the indeterminacy as an epistemological one resulting from our
inability to know the character of these entities - that is, we cannot
completely know the future .though it exists and has a definite characterwhichever view we take indeterminacy still creates major difficulties for
certain ethical theories and their treatment of the future.
The difficulties arise for theories which appear to require a high level of
determinacy with respect to the number and character of future items, in
particular calculus-type theories such as utilitarianism in its usual forms,
where the calculations are critically dependent on such information as
numbers, totals, and averages, information which so far as the future is
concerned is generally indeterminate. The fact that this numerical information is typically indeterminate means that insofar as head-count utilitarianism requires determinate information on numbers, it is in a similar
position to theories discussed earlier; it may apply theoretically to future
people, but since the calculations cannot be applied to them their interests
will be left out of account. And, in fact, utilitarianism for the most part
does not, and perhaps cannot, take future creatures and their interests
seriously; there is little discussion as to how the difficulties or impossibility of calculations regarding the open future are to be obtained. Non-platonistic utilitarianism is in logical difficulty on this matter, while platonis-

Obligations to the Future 159
tic utilitarianism - which faces a range of other objections - is inapplicable
because of epistemic indeterminacy. We have yet another case of a theory
of the sort that applies theoretically but in practice doesn't take the future
seriously. But far from this showing that future people's interests should
be left out of account, what these considerations show are deficiencies in
these sorts of theories, which require excessive determinacy of information. This kind of information is commonly equally unavailable for the
accepted areas of moral constraint, the present and immediate future; and
the resolution of moral issues is often not heavily dependent on knowledge
of such specific determinate features as numbers or other determinate
features. For example, we do not need to know how many people there
will be on the bus, how intelligent they are, what their preferences are or
how badly they will be injured, in order to reach the conclusion that the
consigner's action in despatching his parcel is a bad one. Furthermore, it is
only the ability of moral considerations to continue to apply in the absence
of determinate information about such things as numbers that makes it
possible to take account of the possible effects of action, as the risks
associated with action - something which is quite essential even for the
present if moral considerations are to apply in the normal and accepted
way. For it is essential in order to apply moral considerations in the
accepted way that we consider alternative worlds, in order to take account
of options, risks, and alternative outcomes; but these alternative or counterfactual worlds are not in so different a position from the future with
respect to determinacy; for example, there is indeterminacy with respect
to the number of people who may be harmed in the bus case or in apossible
nuclear reactor melt-down. These alternative worlds, like the distant
future, are indeterminate in some respects, but not totally indeterminate.
It might still be thought that the indeterminacy of the future, for example
with respect to number and exact character, would at least prevent the
interest of future people being taken into account where there is a conflict
with the present. Since their numbers are indeterminate and their interests
unknown, how can we weigh their competing claims against those of the
present and immediate future where this information is available in a more
or less accurate form? The question is raised particularly by problems of
sharing fixed quantities of resources among present and future people,
when the numbers of the latter are indeterminate. Such problems are
indeed difficult, but they are not resolved by ignoring the claims of the
future, any more than the problems raised by the need to take account in
decision-making of factors difficult to quantify are resolved by ignoring

160 R. and V. Routley
such factors. Nor are such distributional problems as large and representative a class of moral problems concerning the future as the tendency to
focus on them would suggest. It should be conceded then that there will be
cases where the indeterminacy of aspects of the future will make conflicts
very difficult or indeed impossible to resolve - a realistic ethical theory
will not deliver a decision procedure - but there will equally be other
conflict cases where the level of indeterminacy does not hinder resolution
of the issue, e.g. the bus example which is a conflict case of a type. In
particular, there will be many cases which are not solved by weighing
numbers, numbers of interests, or whatever, cases for which one needs to
know only the most general probable characteristics of future people.
Moreover, even where numbers are relevant often only bounds will be
required, exact numerical counts only being required where, for instance,
margins are narrow; e.g. issues may be resolved as in parliament where a
detailed vote (or division) is only required when the issue is close. It is
certainly not necessary then to have complete determinacy to resolve all
cases of conflict.
The question we must ask then is what features of future people could
disqualify them from moral consideration or reduce their claims to it to
below those of present people? The answer is: in principle None. Prima
facie moral principles are universalizable, and lawlike, in that they apply
independently of position in space or in time, for example.24 But universalizability of principles is an outcome of those ethical theories which are
capable of dealing satisfactorily with the present; in other words, a theory
that did not allow properly for the future would be found to have defects as
regards the present, to deal unjustly or unfairly with some present people,
e.g. those remotely located, those outside some select subgroup such as
(white-skinned) humans, etc. The only candidates for characteristics that
would fairly rule out future people are the logical features we have been
looking at, uncertainty and indeterminacy; what we have argued is that it
would be far too sweeping to see these features as affecting the moral
claims of future people in a general way. These special features only affect
certain sorts of cases (e.g. the determination of best probable or practical
course of action given only present information). In particular they do not
affect cases of the sort being considered, the nuclear one, where highly
determinate or certain information about the numbers and characteristics
of the class likely to be harmed or certainty of damage are not required.
To establish obligations to the future a full universalizability principle is
not needed: it is enough to require that the temporal position of a person

Obligations to the Future 161
cannot affect his entitlement to just and fair treatment, to full moral
consideration;25 inversely that it is without basis to discriminate morally
against a person in virtue of his temporal position. As a result of this
universalizability, there is the same obligation to future people as to the
present; and thus there is the same obligation to take account of them and
their interests in what we do, to be careful in our actions, to take account of
the probability (and not just the certainty) of our actions' causing harm or
damage, and to see, other things being equal, that we do not act so as to rob
future people of what is necessary for the chance of a good life. Uncertainty and indeterminacy do not free us of these obligations. If, in a closely
comparable case concerning the present, the creation of a significant risk
is enough to rule out an action as immoral, and there are no independent
grounds for requiring greater certainty of harm in the future case under
consideration, then futurity alone will not provide adequate grounds for
proceeding with the action, thus discriminating against future people.
Accordingly we cannot escape, through appeal to futurity, the conclusion
tentatively reached in our first section, that proposals for nuclear development in the present state of technology for future waste disposal are

IV. Overriding Consideration Arguments
In the first part we noticed that the consigner's action could not be justified
by purely economistic arguments, such as that his profits would rise, the
firm or the village would be more prosperous, or by appealing to the fact
that some possibly uncomfortable changes would otherwise be needed.
We also observed that the principle on which this assessment was based,
that one was not usually entitled to create a serious risk to others for these
sorts of reasons, applied more generally and, in particular, applied to the
nuclear case. For this reason the economistic arguments which are thus
most commonly advanced to promote nuclear development - e.g. cheapness, efficiency, profitability for electricity utilities, and the need otherwise for uncomfortable changes such as restructuring of employment,
investment, and consumption-do not even begin to show that the nuclear
alternative is an acceptable one. Even if these economistic assumptions
about benefits to present people were correct (and there is reason to doubt
that most of them are),26 the arguments would fail because economics
must operate within the framework of moral constraints, and not vice

162 R. and V. Routley
What one does have to consider, however, are moral conflict arguments, that is arguments to the effect that, unless the prima facie unacceptable alternative is taken, some even more unacceptable alternative is
the only possible outcome, and will ensue. For example, in the bus case,
the consigner may argue that his action is justified because unless it is
taken the village will starve. It is by no means clear that even such a
justification as this would be sufficient, especially where the risk to the
passengers is high, as the case seems to become one of transfer of costs
and risks onto others; but such a moral situation would no longer be so
clearcut, and one would perhaps hesitate to condemn any action taken in
such circumstances.
Some of the arguments advanced to show moral conflict are based on
competing duties to present people, and others on competing obligations
to future people, both of which are taken to override the obligations not to
impose on the future significant risk of serious harm. The structure of such
moral conflict arguments is based crucially on the presentation of a genuine and exhaustive set of alternatives (or at least practical alternatives),
and upon showing that the only alternatives to admittedly morally undesirable actions are even more undesirable ones. If some practical alternative which is not morally worse than the action to be justified is overlooked, suppressed, or neglected in the argument-for example, if in the bus
case it turns out that the villagers have another option to starving or to the
sending off of the parcel, namely earning a living in some other way - then
the argument is defective and cannot readily be patched. We want to argue
that suppression of practicable alternatives has occurred in the argument,
designed to show that the alternatives to the nuclear option are even worse
than the option itself, and that there are other factual defects in these
arguments as well. In short, the arguments depend essentially on the
presentation of false dichotomies.
The first argument, the poverty argument, is that there is an overriding
obligation to the poor, both the poor of the third world and the poor of
industrialized countries. Failure to develop nuclear energy, it is often
claimed, would amount to denying them the opportunity to reach the
standard of affluence we currently enjoy and would create unemployment
and poverty in the industrialized nations.
The unemployment and poverty argument does not stand up to examination either for the poor of the industrial countries or for those of the third
world. There is good evidence that large-scale nuclear energy will help to
increase unemployment and poverty in the industrial world, through the

Obligations to the Future 163
diversion of great amounts of available capital into an industry which is not
only an exceptionally poor provider of direct employment, but also helps
to reduce available jobs through encouraging substitution of energy use for
labour use. 27 The argument that nuclear energy is needed for the third
world is even less convincing. Nuclear energy is both politically and
economically inappropriate for the third world, since it requires massive
amounts of capital, requires numbers of imported scientists and engineers,
and creates negligible employment, while politically it increases foreign
dependence, adds to centralized entrenched power and reduces the
chance for change in the oppressive political structures which are a large
part of the problem.28 The fact that nuclear energy is not in the interests of
people of the third world does not, of course, mean that it is not in the
interests of, and wanted by, their rulers, the westernized and often military elites in whose interests the economies of these countries are usually
organized; but it is not paternalistic to examine critically the demands
these ruling elites may make in the name of the poor.
The poverty argument then is a fraud. Nuclear energy will not be used to
help the poor.29 Both for the third world and for the industrialized countries there are well-known energy-conserving alternatives and the practical option of developing other energy sources,30 alternatives which are
morally acceptable and socially preferable to nuclear development, and
which have far better prospects for helping the poor.31
The second major argument advanced to show moral conflict appeals to
a set of supposedly overriding and competing obligations to future people.
We have, it is said, a duty to pass on the immensely valuable things and
institutions which our culture has developed. Unless our high-technological, high-energy industrial society is continued and fostered, our valuable
institutions and traditions will fall into decay or be swept away. The
argument is essentially that without nuclear power, without the continued
level of material wealth it alone is assumed to make possible, the lights of
our civilization will go out.32 .
The lights-going-out argument raises rather sharply questions as to
what is valuable in our society, and of what characteristics are necessary
for a good society. These are questions which deserve much fuller treatment than we can allot them here, but a few brief points should be made.
The argument adopts an extremely uncritical position with respect to
existing high-technology societies, apparently assuming that they are
uniformly and uniquely valuable; it also assumes that technological society is unmodifiable, that it can't be changed in the direction of energy

164 R. and V. Routley
conservation or alternative energy sources without collapse. Such a society has to be accepted and assessed as a whole, and virtually unlimited
supplies of energy are essential to maintain this whole.
These assumptions are hard to accept. The assumption that technological society's energy patterns are unmodifiable is especially so - after all, it
has survived events such as world wars which have required major social
and technological restructuring and consumption modification. If western
society's demands for energy are totally unmodifiable without collapse,
not only would it be committed to a programme of increasing destruction,
but one might ask what use its culture could be to future people who would
very likely, as a consequence of this destruction, lack the resource base
which the argument assumes to be essential in the case of contemporary
There is also difficulty with the assumption of uniform valuableness; but
if this is rejected the question becomes not: what is necessary to maintain
existing high-technological society and its political institutions? but rather: what is necessary to maintain what is valuable in that society and the
political institutions which are needed to maintain those valuable things?
While it may be easy to argue that high energy consumption is necessary to
maintain the political and economic status quo, it is not so easy to argue
that it is essential to maintain what is valuable, and it is what is valuable,
presumably, that we have a duty to pass on to the future.
The evidence, e.g. from history, is that no very high level of material
affluence or energy consumption is needed to maintain what is valuable.
There is good reason in fact to believe that a society with much lower
energy and resource consumption would better foster what is valuable
than our own. But even if a radical change in these directions is independently desirable, as we believe it is, it is not necessary to presuppose such
a change, in the short term at least, in order to see that the assumptions of
the lights-going-out argument are wrong. No enormous reduction of wellbeing is required to consume less energy than at present, and certainly far
less than the large increase over present levels of consumption which is
assumed in the usual economic case for nuclear energy.33 What the nuclear strategy is really designed to do then is not to prevent the lights going
out in western civilization, but to enable the lights to go on burning all the
time - to maintain and even increase the wattage output of the Energy
In fact there is good reason to think that, far from the high energy
consumption society fostering what is valuable, it will, especially if energy

Obligations to the Future 165
is obtained by nuclear-fission means, be positively inimical to it. A society
which has become heavily dependent upon an extremely high centralized,
controlled, and garrisoned, capital- and expertise-intensive energy
source, must be one which is highly susceptible to entrenchment of power,
and one in which the forces which control this energy source, whether
capitalist or bureaucratic, can exert enormous power over the political
system and over people's lives, even more than they do at present. Very
persuasive arguments have been advanced by civil liberties groups and
others in a number of countries to suggest that such a society would tend to
become authoritarian, if only as an outcome of its response to the threat
posed by dissident groups in the nuclear situation.34
There are reasons to believe then that with nuclear development what
we would be passing on to future generations would be some of the worst
aspects of our society (e.g. the consumerism, growing concentration of
power, destruction of the natural environment, and latent authoritarianism), while certain valuable aspects would be lost or threatened. Political
freedom is a high price to pay for consumerism and energy extravagance.
Again, as in the case of the poverty arguments, clear alternatives which
do not involve such unacceptable consequences are available. The alternative to the high technology-nuclear option is not a return to the cave, the
loss of all that is valuable, but the development of alternative technologies
and life-styles which offer far greater scope for the maintenance and
further development of what is valuable in our society than the highly
centralized nuclear option.35 The lights-going-out argument, as a moral
conflict argument, accordingly fails, because it also is based on a false
dichotomy. Thus both the escape routes, the appeal to moral conflict and
to the appeal to futurity, are closed.
If then we apply the same standards of morality to the future as we
acknowledge for the present - as we have argued we should - the conclusion that the proposal to develop nuclear energy on a large scale is a crime
against the future is inevitable, since both the escape routes are closed.
There are, of course, also many other grounds for ruling it out as morally
unacceptable, for saying that it-is not only a crime against the distant future
but also a crime against the present and immediate future. These other
grounds for moral concern about nuclear energy, as it affects the present
and immediate future, include problems arising from the possibility of
catastrophic releases of radioactive fuel into the environment or of waste
material following an accident such as reactor melt-down, of unscheduled
discharges of radiation into the environment from a plant fault, of proli-

166 R. and V. Routley
feration of nuclear weapons, and of deliberate release or threat of release
of radioactive materials as a measure of terrorism or of extortion. All these
are important issues, of much moral interest. What we want to claim,
however, is that on the basis of its effects on the future alone, the nuclear
option is morally unacceptable.

Passmore's Treatment ofRawls,
and What Really Happens on Rawls's Theory
Passmore takes it that Rawls's theory yields an unconstrained position
but, according to Rawls, the theory leads to quite the opposite result;
persons in different generations [and not merely neighbouring generations] have duties and obligations to one another just as contemporaries do. The present generation cannot do as it pleases but is bound
by the principles that would be chosen in the original position to
define justice between persons at different moments of time.. . . The
derivation of these duties and obligations may seem at first a somewhat far-fetched application of the central doctrine. Nevertheless
these requirements would be acknowledged in the original position
[where the parties do not know to which generation they belong], and
so the conception of justice as fairness covers these matters without
any change in its basic idea. ([5], p. 293; the second insert is drawn
from p. 287)
Through judicious use of the veil of ignorance and the time of entry of
parties to the original contract position, Rawls's contract theory, unlike
simpler explicit contract theories, can yield definite obligations to distant
future people,36 for example, we ought to save at a just rate for future
But, as Rawls remarks (p. 284), 'the question of justice between generations . . . subjects any ethical theory to severe if not impossible tests'. It is
doubtful that Rawls's theory as formulated passes the tests; for the theory
as formulated does not yield the stated conclusion, but a conclusion
inconsistent with the thesis that there are the same obligations to future
people as to contemporaries. Exactly how these obligations arise from the
initial agreement depends critically on the interpretation of the time of

Obligations to the Future 167
entry of the parties into the agreement. Insofar as Rawls insists upon the
present-time-of-entry interpretation (p. 139), he has to introduce supplementary motivational assumptions in order to (try to) secure the desired
bondings between generations, in particular to ensure that the generation
of the original position saves for any later generation, even their immediate successors ([5], p. 140 and p. 292). Rawls falls back on-what is as we
have seen inadequate to the task, since it does not exclude one generation
damaging another remote generation in a way that bypasses mutually
successive generations - 'ties of sentiment between successive generations' (p. 292): to this limited extent Passmore has a point, for such a social
contract on its own (without additional assumptions about the motives of
the parties to the agreement) does not furnish obligations even to our
immediate successors. This is indicative also of the unsatisfactory instability of Rawls's theory under changes, its sensitivity to the way the original
agreement is set up, to the motivation of parties, their time of entry, what
they can know, etc.
To arrive at a more adequate account of obligations to the distant future
under Rawls's theory, let us adopt, to avoid the additional, dubious and
unsatisfactory, motivational assumptions Rawls invokes, one of the alternative - and non -equivalent - time-of-entry interpretations that Rawls lists
(p. 146), that of persons alive at some time in simultaneous agreement. Let
us call this, following Rawls's notation (on p. 140), interpretation 4b (it is
perhaps unnecessary to assume for 4b any more than 4a that all people
need be involved: it may be enough given the equivalencizing effect of the
veil of ignorance that some are, and as with the particular quantifier it is
quite unnecessary to be specific about numbers). Then of course the
parties, since they are, for all they know, of different generations, will
presumably agree on a just savings rate, and also to other just distribution
principles, simply on the basis of Rawlsian rationality, i.e. advancing their
own interests, without additional motivation assumptions. This more
appealing omnitemporal interpretation of time of entry into the agreement,
which gives a superior account of obligation to the future consistent with
Rawls's claim, Rawls in some places puts down as less than best (p. 292)
but in his most detailed account of the original position simply dismisses
(p. 139):
To conceive of the original position [as a gathering of people living at
different times] would be to stretch fantasy too far; the conception
would cease to be a natural guide to intuition.

168 R. and V. Rout ley
This we question: it would be a better guide to intuition than a position
(like 4a) which brings out intuitively wrong results; it is a more satisfactory
guide, for example, to justice between generations than the present-timeof-entry interpretation, which fails conspicuously to allow for the range of
potential persons (all of whom are supposed to qualify on Rawls's account
for just treatment, cf. § 77). Moreover, it stretches fantasy no further than
science fiction or than some earlier contract accounts.36 But it does
require changes in the way the original position is conceived, and it does
generate metaphysical difficulties for orthodox ontological views (though
not to the same extent for the Meinongian view we prefer); for, to consider
the latter, either time travel is possible or the original hypothetical position
is an impossible situation, with people who live at different times assembled at the same time. The difficulties - of such an impossible meeting help to reveal that what Rawls's theory offers is but a colourful representation of obligations in terms of a contract agreed upon at a meeting.
The metaphysical difficulties do not concern merely possible people,
because all those involved are sometime-actual people; nor are there
really serious difficulties generated by the fact that very many of these
people do not exist, i.e. exist now. The more serious difficulties are either
those of time travel, e.g. that future parties relocated into the present may
be able to interfere with their own history, or, if time travel is ruled out
logically or otherwise, that future parties may be advantaged (or disadvantaged) by their knowledge of history and technology, and that accordingly fairness is lost. As there is considerable freedom in how we choose to
(re)arrange the original position, we shall suppose that time travel is
rejected as a means of entering the original position. For much less than
travel is required; some sort of limited communicational network which
filters out, for example, all historical data (and all cultural or species
dependent material) would suffice; and in any case if time travel were not
excluded essentially the present-time-of-entry interpretation would serve,
though fairness would again be put in doubt. The filtered communicational
hook-up by which the omnitemporal position is engineered still has - if
fairness is to be seen to be built into the decision making - to be combined
with a reinterpreted veil of ignorance, so that parties do not know where
they are located temporally any more than they know who they are
characterwise. This implies, among other things, limitations on the parties' knowledge of factual matters, such as available technology and world
and local history; for otherwise parties could work out their location,
temporal or spatial. For example, if some party knew, as Rawls supposes,

Obligations to the Future 169
the general social facts, then he would presumably be aware of the history
of his time and so of where history ends, that is of the date of his
generation, his time (his present), and so be aware of his temporal location. These are already problems for Rawls's so-called 'present-time-ofentry interpretation' - it is, rather, a variable-time-of-entry interpretation
- given that the parties may be, as Rawls occasionally admits (e.g. p. 287),
of any one generation, not necessarily the present: either they really do
have to be ofthepresent time or they cannot be assumed to know as much
as Rawls supposes.37 There is, however, no reason why the veil of ignorance should not be extended so as to avoid this problem; and virtually any
extension that solves the problem for the variable-time-of-entry interpretation should serve, so it seems, for the omnitemporal one. We shall
assume then that the parties know nothing which discloses their respective
locations (i.e. in effect we write in conditions for universalizability of
principles decided upon). There are still gaps between the assumptions of
the omnitemporal position as roughly sketched and the desired conclusion
concerning obligations to the future, but (the matter is beginning to look
non-trivially provable given not widely implausible assumptions and) the
intuitive arguments are as clear as those in [5], indeed they simply restate
arguments to obligations given by Rawls.
Rawls's theory, under interpretation 4b, admits of nice application to
the problems of just distribution of material resources and of nuclear
power. The just distribution, or rate of usage, of material resources38 over
time is an important conservation issue to which Rawls's theory seems to
apply, just as readily, and in a similar fashion, to that in which it applies to
the issue of a just rate of savings. In fact the argument from the original
position for a just rate of saving - whatever its adequacy - can by simply
mimicked to yield an argument for just distribution of resources over
generations. Thus, for example:
persons in the original position are to ask themselves how much they
would be willing to save [i.e. conserve] at each stage of advance on
the assumption that all other generations are to save at the same rate
[conserve resources to the same extent]. . . . In effect, then, they
must choose a just savings principle [resources distribution principle]
that assigns an appropriate rate of accumulation to [degree of resource conservation at] each level of advance. ([5], p. 287; our
bracketed options give the alternative argument)

170 R. and V. Routley
Just as 'they try to piece together a just savings schedule' (p. 289), so they
can try to piece together a just resource distribution policy. Just as a case
for resource conservation can be made out by appeal to the original
position, since it is going to be against the interests of, to the disadvantage
of, later parties to find themselves in a resource depleted situation (thus,
on Rawlsian assumptions, they will bargain hard for a share of resources),
so, interestingly, a case against a rapid programme of nuclear power
development can be devised. The basis of a case against large-scale
nuclear development is implicit in Rawls's contract theory under interpretation 4b, though naturally the theory is not applied in this sort of way
by Rawls. To state the case in its crude but powerful form: people from
later generations in the original position are bound to take it as against their
interests to simply carry the waste can for energy consumed by an earlier
generation. (We have already argued that they will find no convincing
overriding considerations that make it worth their while to carry the waste
can.) Thus not only has Passmore misrepresented the obligations to the
future that Rawls's theory admits; he is also wrong in suggesting (p. 87 and
p. 91) that Rawls's theory cannot justify a policy of resource conservation
which includes reductions in present consumption.
There is, in this connection, an accumulation of errors in Passmore,
some of which spill over to Rawls, which it is worth trying to set out. First,
Passmore claims ([1], p. 86; cf. also p. 90) that 'Rawls does not so much
mention the saving of natural resources'. In fact the 'husbanding of natural
resources' is very briefly considered ([5], p. 271). It is true, however, that
Rawls does not reveal any of the considerable power that his theory,
properly interpreted, has for natural resource conservation, as implying a
just distribution of natural resources over time. Secondly, Passmore attempts ([1], pp. 87 ff.) to represent the calls of conservationists for a
reduction in present resource usage and for a more just distribution as a
call for heroic self-sacrifices; this is part of his more general attempt to
represent every moral constraint with respect to the non-immediate future
as a matter of self-sacrifice. 'Rawls's theory', Passmore says (on p. 87),
'leaves no room for heroic sacrifice', and so, he infers, leaves no room for
conservation. Not only is the conclusion false, but the premiss also:
Rawls's theory allows for supererogation, as Rawls explains ([5], p. 117).
But resource conservation is, like refraining from nuclear development,
not a question of heroic self-sacrifice; it is in part a question of obligations
or duties to the distant future. And Rawls's theory allows not only for
obligations as well as supererogation, but also for natural duties. Rawls's

Obligations to the Future 171
contract, unlike the contracts of what is usually meant by a 'contract
theory', is by no means exhaustive of the moral sphere:
But even this wider [contract] theory fails to embrace all moral
relationships, since it would seem to include only our relations with
other persons and to leave out of account how we are to conduct
ourselves towards,animals and the rest of nature. I do not contend
that the contract notion offers a way to approach these questions
which are certainly of the first importance; and I shall have to put
them aside, (p. 17)
The important class of obligations beyond the scope of the contract theory
surely generates obligations between persons which even the wider contract theory likewise cannot explain. The upshot is that such a contract
account, even if sufficient for the determination of obligations, is not
necessary. To this extent Rawls' s theory is not a full social contract theory
at all. However, Rawls appears to lose sight of the fact that his contract
theory delivers only a sufficient condition when he claims, for example (p.
298), that 'one feature of the contract doctrine is that it places an upper
bound on how much a generation can be asked to save for the welfare of
later generations'. For greater savings may sometimes be required to meet
obligations beyond those that the contract doctrine delivers. In short,
Rawls appears to have slipped into assuming, inconsistently, that his
contract theory is a necessary condition.
Although Rawls's theory caters for justice between generations and
allows the derivation of important obligations to people in the distant
future, the full theory is far from consistent on these matters and there are
significant respects in which Rawls does not take justice to the future
seriously. The most conspicuous symptoms of this are that justice to the
future is reduced to a special case, justice between generations, and that
the only aspect of justice between generations that Rawls actually considers is a just savings rate; there is, for example, no proper examination of
the just distribution of resources among generations, though these resources, Rawls believes, provide the material base of the just institutions that
he wants to see maintained. In fact Rawls strongly recommends a system
of markets as a just means for the allocation of most goods and services,
recognizing their well-known limitations only in the usual perfunctory
fashion ([5], pp. 270 ff.). Yet market systems are limited by a narrow time
horizon, and are quite ill-equipped to allocate resources in a just fashion

172 R. and V. Rout ley
over a time span of several generations. Similarly Rawls's endorsement of
democratic voting procedures as in many cases a just method of determining procedures depends upon the assumptions that everyone with ah
interest is represented. But given his own assumptions about obligations
to the future and in respect of potential persons this is evidently not the
case. Catering in a just fashion for the interests of future people poses
serious problems for any method of decision that depends upon people
being present to represent their own interests.
Some of the more conservative, indeed reactionary, economic assumptions in Rawls emerge with the assumption that all that is required for
justice between generations is a just savings rate, that all we need to pass
on to the future are the things that guarantee appropriate savings such as
capital, factories, and machines. But the transmission of these things is
quite insufficient for justice to the future, and neither necessary nor
sufficient as a foundation for a good life for future generations. What is
required for justice is the transmission in due measure of what is valuable.
Rawls has, however, taken value accumulation as capital accumulation,
thereby importing one of the grossest economic assumptions, that capital
reflects value. But of course the accumulation of capital may conflict with
the preservation of what is valuable. It is for this sort of reason (and thus,
in essence, because of the introduction of supplementary economistic
theses which are not part of the pure contract theory) that Rawls's theory
is a reactionary one from an environmental point of view; on the theory as
presented (i.e. the contract theory plus all the supplementary assumptions) there is no need to preserve such things as wilderness or natural
beauty. The savings doctrine supposes that everything of value for transmission to the future is negotiable in the market or tradeable; but then
transmission of savings can by no means guarantee that some valuable
things, not properly represented in market systems, are not eliminated or
not passed on, thereby making future people worse off. It becomes evident
in this way, too, how culturally-bound Rawls's idea of ensuring justice to
future generations through savings is. It is not just that the idea does not
apply, without a complete overhaul, to non-industrial societies such as
those of hunter-gatherers; it does not apply to genuinely post-industrial
societies either. Consideration of such alternative societies suggests that
whafis required, in place of capital accumulation, is that we pass on what
is necessary for a good life, that we ensure that the basics are fairly
distributed over time and not eroded, e.g. that in the case of the forest
people that the forest is maintained. The narrowness of Rawls's picture,

Obligations to the Future 173
which makes no due allowance for social or cultural diversity (from the
original contractual position on) or for individual diversity arises in part
from his underlying and especially narrow socio-economic assumption as
to what people want:
What men want is meaningful work in free association with others,
these associations regulating their relations to one another within a
framework of just basic institutions. ([5], p. 290)
This may be what many Harvard men want; but as a statement of what
men want it supplies neither sufficient nor even necessary conditions.39

1 Thus according to the Fox Report ([2], p. 110; our italics).
There is at present no generally accepted means by which high level waste can be
permanently isolated from the environment and remain safe for very long periods.
. . . Permanent disposal of high-level solid wastes in stable geological formations is
regarded as the most likely solution, but has yet to be demonstrated as feasible. It is
not certain that such methods and disposal sites will entirely prevent radioactive
releases following disturbances caused by natural processes or human activity.
The Fox Report also quoted approvingly ([2], p. 187; our italics) the conclusion of the
British (Flowers) Report [6]:
There should be no commitment to a large programme of nuclear fission power until
it has been demonstrated beyond reasonable doubt that a method exist s to ensure the
safe containment of long-lived, highly radioactive waste for the indefinite future.
Although the absence of a satisfactory storage method has been conceded by some
leading proponents of nuclear development, e.g. Weinberg ([3], pp. 32-33), it is now
disputed by others. In particular, the headline for Cohen [15], which reads 'A substantial
body of evidence indicates that the high level radioactive wastes generated by U.S.
nuclear power plants can be stored satisfactorily in deep geological formations', has
suggested to many readers - what it was no doubt intended to suggest- that there is really
no problem about the disposal of radioactive wastes after all. Cohen presents, however,
no new hard evidence, no evidence not already available to the British and Australian
Commissions ([2] and [6]). Moreover the evidence Cohen does outline fails conspicuously
to measure up to the standards rightly required by the Flowers and Fox Reports. Does
Cohen offer a commercial-scale procedure for waste disposal which can be demonstrated
as safe? Far from it:

174 R. and V. Routley
The detailed procedures for handling the high-level wastes are not yet definite, but
present indications are that. . . . (Cohen [1], p. 24; our italics)
Does Cohen 'demonstrate beyond reasonable doubt' the long-term safety of burial of
wastes, deep underground? Again, far from it:
On the face of it such an approach appears to be reasonably safe. . . . (p. 24; our
Cohen has apparently not realized what is required.
At issue here are not so much scientific or empirical issues as questions of methodology, of standards of evidence required for claims of safety, and above all, of values, since
claims of safety, for example, involve implicit evaluations concerning what counts as an
acceptable risk, an admissible cost, etc. In the headline 'a substantial body of evidence
. . . indicates that. . . wastes . . . can be stored satisfactorily' the key words (italicized)
are evaluative or elastic, and the strategy of Cohen's case is to adopt very low standards
for their application. But in view of what is at stake it is hardly acceptable to do this, to
dress up in this way what are essentially optimistic assurances and untested speculations
about storage, which in any case do little to meet the difficulties and uncertainties that
have been widely pointed out as regards precisely the storage proposals Cohen outlines,
namely human or natural interference or disturbance.
2 See [18], pp. 24-25.
3 On all these points, see [14], esp. p. 141. According to the Fox Report ([2], p. 110):
Parts of the reactor structure will be highly radioactive and their disposal could be
very difficult. There is at present no experience of dismantling a full size reactor.
4 See, in particular, The Union of Concerned Scientists, The Nuclear Fuel Cycle, Friends
of the Earth Energy Paper, San Francisco 1973, p. 47; also [3], p. 32 and [14], p. 149.
5 As the discussion in [14], pp. 153-7, explains.
6 Cf. [17], pp. 35-36, [18] and, for much detail, J. R. Goffman and A. R. Tamplin, Poisoned
Power, Rodale Press, Emmau Pa. 1971.
7 On the pollution and waste disposal record of the infant nuclear industry, see [14] and
The record of many countries on pollution control, where in many cases available
technologies for reducing or removing pollution are not applied because they are considered too expensive or because they adversely affect the interests of some powerful group,
provides clear historical evidence that the problem of nuclear waste disposal would not
end simply with the devising of a 'safe' technology for disposal, even if one could be
devised which provided a sufficient guarantee of safety and was commercially feasible.
The fact that present economic and political arrangements are overwhelmingly weighted
in favour of the interests and concerns of (some) contemporary humans makes it not
unrealistic to expect the long-term nuclear waste disposal, if it involved any significant
cost at all, when public concern about the issue died down, would be seen to conflict with
the interests of contemporary groups, and that these latter interests would in many cases
be favoured. Nor, as the history of movements such as the Campaign for Nuclear
Disarmament shows, could generalized public concern in the absence of direct personal
interest, be relied upon to be sustained for long enough to ensure implementation of costly
or troublesome long-term disposal methods - even in those places where public concern
exists and is a politically significant force.
It must be stressed then that the problem is not merely one of disposal technique.
Historical and other evidence points to the conclusion that many of the most important
risks associated with nuclear waste disposal are not of the kind which might be amenable

Obligations to the Future


to technical solutions in the laboratory. A realistic assessment of potential costs to the
future from nuclear development cannot overlook these important non-technical risk
8 Of course the effect on people is not the only factor which has to be taken into consideration in arriving at a moral judgment. Nuclear radiation, unlike most ethical theories, does
not confine its scope to human life. But since the harm nuclear development is likely to
cause to non-human life can hardly improve its case, it suffices if the case against it can be
made out solely in terms of its effects on human life in the conventional way.
9 Proponents of nuclear power often try to give the impression that future people will not
just bear costs from nuclear development but will also be beneficiaries, because nuclear
powerprovides an 'abundant' or even 'unlimited' source of energy; thus Weinberg ([3], p.
34): 'an all but infinite source of relatively cheap and clean energy'. A good example of an
attempt to create the impression that 'abundant' and 'cheap' energy from nuclear fission
will be available to 'our descendants', i.e. all future people, is found in the last paragraph
of Cohen [15]. Such claims are most misleading, since fission power even with the breeder
reactor has only about the same prospective lifetime as coal-produced electricity (a point
that can be derived using data in A. Parker, 'World Energy Resources: A Survey', Energy
Policy, Vol. 3 [1975], pp. 58-66), and it is quite illegitimate to assume that nuclear fusion,
for which there are still major unsolved problems, will have a viable, clean technology by
the time fission runs out, or, for that matter, that it ever will. Thus while some few
generations of the immediate future may obtain some benefits as well as costs, there is a
very substantial chance that tho se of the more distant future will obtain nothing but costs.
10 These feelings, of which Smith's and Hume's sympathy is representative, are but the
feeling echoes of obligation. At most, sympathy explains the feeling of obligation or lack
of it, and this provides little guide as to whether there is an obligation or not - unless one
interprets moral sympathy, the feeling of having a obligation, or being obligated, itself as
a sufficient indication of obligation, in which case moral sympathy is a non-explanatory
correlate in the feelings department of obligation itself and cannot be truly explanatory of
the ground of it; unless, in short, moral sympathy reduces to an emotive rewrite of moral
11 Elsewhere in [1] Passmore is especially exercised that our institutions and intellectual
traditions - presumably only the better ones - should be passed on to posterity, and that
we should strive to make the world a better place, if not eventually an ideal one.
12 This is not the only philosophically important issue in environmental ethics on which
Passmore is inconsistent. Consider his: 'over-arching intention: to consider whether the
solution of ecological problems demands a moral or metaphysical revolution' (p. x),
whether the West needs a new ethic and a new metaphysics. Passmore's answer in [1] is
an emphatic No.
Only insofar as Western moralists have [made various erroneous suggestions] can
the West plausibly be said to need a 'new ethic'. What it needs, for the most part, is
not so much a 'new ethic' as adherence to a perfectly familiar ethic.
For the major sources of our ecological disasters - apart from ignorance - are
greed and shortsightedness, which amount to much the same thing. . . There is no
novelty in the view that greed is evil; no need of a new ethic to tell us as much. (p. 187)
'The view that the West now needs. . . a new concept of nature' is similarly dismissed (p.
186, cf. p. 72). But in his paper [1*] (i.e. 'Attitudes to Nature', Royal Institute of
Philosophy Lectures, Vol. 8, Macmillan, London 1975), which is said to be an attempt to
bring together and to reformulate some of the basic philosophical themes of [1], Passmore's answer is Yes, and quite different themes, inconsistent with those of [1], are

176 R. and V. Routley
[T]he general conditions I have laid down . . . have not been satisfied in most of the
traditional philosophies of nature. To that degree it is true, I think, that we do need a
'new metaphysics' which is genuinely not anthropocentric. . . . A 'new metaphysics', if it is not to falsify the facts, will have to be naturalistic, but not reductionist.
The working out of such a metaphysics is, in my judgement, the most important task
which lies ahead of philosophy. ([1*], pp. 260-1)
A new ethic accompanies this new metaphysics.
The emergence of new moral attitudes to nature is bound up, then, with the emergence of a more realistic philosophy of nature. That is the only adequate foundation
for effective ecological concern. ([1*], p. 264)
This is a far cry from the theme of [1] that ecological problems can be solved within the
traditions of the West.
13 Put differently, the causal linkage can bypass intermediate generations, especially given
action at a temporal distance: the chain account implies that there are no moral constraints in initiating such causal linkages. The chain picture accordingly seems to presuppose an unsatisfactory Humean model of causation, demanding contiguity and excluding
action at a distance.
14 Golding we shall concede to Passmore, though even here the case is not clearcut. For
Golding writes towards the end of his article ([12], p. 96):
My discussion, until this point, has proceeded on the view that we have obligations
to future generations. But do we? I am not sure that the question can be answered in
the affirmative with any certainty. I shall conclude this note with a very brief
discussion of some of the difficulties.
All of Passmore's material on Golding is drawn from this latter and, as Golding says,
'speculative' discussion.
15 There is no textual citation for Bentham at all for the chapter of [1] concerned, viz. Ch. 4,
16 As Passmore himself at first concedes ([1], p. 84):
If, as Bentham tells us, in deciding how to act men ought to take account of the
effects of their actions on every sentient being, they obviously ought to take account
of the pleasure and pains of the as yet unborn.
17 Neither rightness nor probable lightness in the hedonistic senses correspond to these
notions in the ordinary sense; so at least [13] argues, following much anti-utilitarian
18 On this irrationality different theories agree: the rational procedure, for example according to the minimax rule for decision-making under uncertainty, is to minimize that
outcome which maximizes harm.
19 Discount, or bank, rates in the economists' sense are usually set to follow the market (cf.
P. A. Samuelson, Economics, 7th ed., McGraw-Hill, New York 1967, p. 351). Thus the
rates have little moral relevance.
20 Cf. Rawls [5], p. 287: 'From a moral point of view there are no grounds for discounting
future well-being on the basis of pure time preference.'
21 What the probabilities would be depends on the theory of probability adopted: a Carnapian theory, e.g., would lead back to the unconstrained position.

Obligations to the Future 177
22 A real possibility is one which there is evidence for believing could eventuate. A real
possibility requires producible evidence for its consideration. The contrast is with mere
logical possibility.
23 Thus, to take a simple special case, economists dismiss distant future people from their
assessments of utility, welfare, etc., on the basis of their non-existence; cf. Ng ('the utility
of a non-existent person is zero') and Harsanyi ('only existing people [not even "nonexisting potential individuals"] can have real utility levels since they are the only ones
able to enjoy objects with a positive utility, suffer from objects with a negative utility, and
feel indifferent to objects with zero utility') (see Appendix B of Y. K. Ng, 'Preference,
Welfare, and Social Welfare', paper presented at the Colloquium on Preference, Choice
and Value Theory, RSSS, Australian National University, August 1977, pp. 24, 26-27).
Non-existent people have no experiences, no preferences; distant future people do not
exist; therefore distant future people have no utility assignments - so the sorites goes. But
future people at least will have wants, preferences, and so on, and these have to be taken
into account in adequate utility assessments (which should be assessed over afuture time
horizon), no matter how much it may complicate or defeat calculations.
24 There are problems about formulating universalizability satisfactorily, but they hardly
affect the point. The requisite universalizability can in fact be satisfactorily brought out
from the semantical analysis of deontic notions such as obligation, and indeed argued for
on the basis of such an analysis which is universal in form. The lawlikeness requirement,
which can be similarly defended, is essentially that imposed on genuine scientific laws by
logical empiricists (e.g. Carnap and Hempel), that such laws should contain no proper
names or the like, no reference to specific locations or times.
25 Such a principle is explicit both in classical utilitarianism (e.g. Sidgwick [11], p. 414), and
in a range of contract and other theories from Kant and Rousseau toRawls ([5], p. 293).
How the principle is argued for will depend heavily, however, on the underlying theory;
and we do not want to make our use depend heavily on particular ethical theories.
26 See esp. R. Lanoue, NuclearPlants:The MoreTheyBuild, The More You Pay, Centerfor
Study of Responsive Law, Washington DC 1976; also [14], pp. 212 ff.
27 On all these points see R. Grossman and G. Daneker, Guide to Jobs and Energy,
Environmentalists for Full Employment, Washington DC 1977, pp. 1-7, and also the
details supplied in substantiating the interesting case of Commoner [7]. On the absorption
of available capital by the nuclear industry, see as well [18], p. 23. On the employment
issues, see too H. E. Daly in [9], p. 149. A more fundamental challenge to the poverty
argument appears in I. Illich, Energy and Equality, Calder & Boyars, London 1974,
where it is argued that the sort of development nuclear energy represents is exactly the
opposite of what the poor need.
28 For much more detail on the inappropriateness see E. F. Schumacher, Small is Beautiful,
Blond & Briggs, London 1973. As to the capital and other requirements, see [2], p. 48, and
also [7] and [9].
For an illuminating look at the sort of development high-energy technology will tend to
promote in the so-called underdeveloped countries, see the paper of Waiko and other
papers in the Melanesian Environment (ed. by J. H. Winslow), Australian National
University Press, Canberra 1977.
29 This fact is implicitly recognized in [2], p. 56.
30 A useful survey is given in A. Lovins, Energy Strategy: The Road Not Taken, Friends of
the Earth Australia, 1977 (reprinted from Foreign Affairs, October 1976); see also [17],
[6], [7], [14], pp. 233 ff., and Schumacher, op. cit.
31 This is also explained in [2], p. 56.
32 An argument like this is suggested in Passmore [1], Chs. 4 and 7, with respect to the
question of saving resources. In Passmore this argument for the overriding importance
of passing on contemporary culture is underpinned by what appears to be a future-

178 R. and V. Routley
directed ethical version of the Hidden Hand argument of economics - that, by a coincidence which if correct would indeed be fortunate, the best way to take care of the future
(and perhaps even the only way to do so, since do-good intervention is almost certain to
go wrong) is to take proper care of the present and immediate future. The argument has
all the defects of the related Chain Argument discussed above and others.
33 See [14], p. 66, p. 191, and also [7].
34 For such arguments see esp. M. Flood and R. Grove-White, Nuclear Prospects. A
Comment on the Individual, the State and Nuclear Power, Friends of the Earth, Council
for the Protection of Rural England and National Council for Civil Liberties, London
35 For a recent sketch of one such alternative which is outside the framework of the
conventional option of centralized bureaucratic socialism, see E. Callenbach's novel,
Ecotopia, Banyan Tree Books, Berkeley, California 1975. For the outline of a liberation
socialist alternative see Radical Technology (ed. by G. Boyle and P. Harper), Undercurrents Limited, London 1976, and references therein.
36 Some earlier contract theories also did. Burke's contract (in E. Burke, Reflections on the
Revolution in France, Dent, London 1910, pp. 93-94) 'becomes a partnership not only
between those who are living, but between those who are living, those who are dead and
those who are not yet born'. Thus Burke's contract certainly appears to lead to obligations to distant future generations. Needless to say, there are metaphysical difficulties,
which however Burke never considers, about contracts between parties at widely separated temporal locations.
37 Several of the preceding points we owe to M. W. Jackson.
38 Resources such as soil fertility and petroleum could even be a primary social goods on
Rawls's very hazy general account of these goods ([5], pp. 62, 97): are these 'something a
rational man wants whatever else he wants'? The primary social goods should presumably be those which are necessary for the good and just life -which will however vary with
39 We have benefited from discussion with Ian Hughes and Frank Muller and useful
comments on the paper from Brian Martin and Derek Browne.

[1] J. Passmore, Man's Responsibility for Nature, Duckworth, London 1974.
[2] Ranger Uranium Environmental Inquiry First Report, Australian Government Publishing Service, Canberra 1977.
[3] A. M. Weinberg, 'Social Institutions and Nuclear Energy", Science, Vol. 177 (July 1972),
pp. 27-34.

[4] R. Routley, 'Exploring Meinong's Jungle II. Existence is Existence Now', Notre Dame
Journal of Formal Logic (to appear).
[5] J. Rawls, A Theory of Justice, Harvard University Press, Cambridge, Mass. 1971.
[6] Nuclear Power and the Environment. Sixth Report of the British Royal Commission on
Environmental Pollution, London 1976.
[7] B. Commoner, The Poverty of Power, Knopf, New York 1976.
[8] G. E. Moore, Principia Ethica, Cambridge University Press, Cambridge 1903.
[9] B. Commoner, H. Boksenbaum and M. Corr (Eds.), Energy and Human Welfare - A
Critical Analysis, Vol. III, Macmillan, New York 1975.

Obligations to the Future 179
[10] J. Bentham, The Works of Jeremy Bentham, Vol. 1, published under the superintendence of J. Bowring, with an intro. by J. H. Burton; William Tait, Edinburgh 1843.
[11] H. Sidgwick, The Methods of Ethics, Macmillan, London 1962 (reissue).
[12] M. P. Golding, 'Obligations to Future Generations', Monist, Vol. 56 (1972), pp. 85-99.
[13] R. and V. Routley, 'An Expensive Repair Kit for Utilitarianism', presented at the
Colloquium on Preference, Choice and Value Theory, RSSS, Australian National
University, August 1977.
[14] R. Nader and J. Abbotts, The Menace of Atomic Energy, Outback Press, Melbourne
[15] B. L. Cohen, 'The Disposal of Radioactive Wastes from Fission Reactors', Scientific
American, Vol. 236 (June 1977), pp. 22-31.
[16] S. McCracken, 'The Waragainst the Atom' .Commentary (September 1977), pp. 33-47.
[17] A. B. Lovins and J. H. Price, Non-Nuclear Futures: The Case for an Ethical Energy
Strategy, Friends of the Earth International, San Francisco 1975.
[18] A. Roberts, "The Politics of Nuclear Power', Arena, No. 41 (1976), pp. 22-47.



R. Routley and V. Routley, “Nuclear energy and obligations to the future,” Antipodean Antinuclearism, accessed June 15, 2024,

Output Formats