TESTING ETHICAL THEORIES:

THE RIGHT AND THE GOOD AGAIN

KAI NIELSEN
The University of Calgary

W. D. Ross argues that right like good is a unique non-natu-

ral property that is known immediately and intuitively. He
rejects Moore's claim in Principia Ethica that it is immedi-
ately evident that right means productive of the greatest
possible good. The right' and 'the optimific' are not iden-
tical in meaning. Right is not even to be determined in terms
of good alone. To say 'this is a right act' means, according to
Ross, 'this act is the act that ought to be done'.1 Obligations
are not entirely dependent on axiology for there are prima
facie duties (conditional duties) that are directly 'suitable to
a situation' even if they are not productive of as much good
as some other act open to the agent.
I would like to argue that the deontologists are justified
in maintaining that 'right' does not mean 'morally good', or
'productive of the greatest possible good', but that the deon-
tologists have not succeeded in refuting the ideal utilitarian
claim that what makes an act right is that which, under the
circumstances, produces the greatest possible amount of good
and that in trying to sort out that issue we are led into what
appears at least to be intractable issues of what Rawls calls
'moral methodology'. We are forced to ask very difficult
questions about what finally we appeal to in testing an ethical
theory and, indeed, even in accepting an ethical claim.
Let us first consider the meaning of right. Ross says that
'this is a right act' means 'this act is the act that ought to
1 W. D. Ross, "The Meaning of Right", in Readings in Ethical Theory,
W. Shelters and J. Hospers (eds.), First Edition, p. 164.

15
be done'.2 Ross's refutation that a 'right act' does not mean
a 'morally good act' runs as follows: morally good acts
proceed from a good motive. But ought implies can and I
cannot by choice produce a motive. I either have or do not
have the motive. I cannot, then, be obligated to act from a
good motive. I can only have a duty to perform certain acts
not a duty to perform them from a sense of duty. If this is
the case then 'right' is synonymous with 'what is my duty'
and is not synonymous with 'morally good'.3
In criticizing Ross's view that there are sometimes special
obligations that make a given act right even if it is not the
most probably optimific act, I shall limit myself to one cri-
ticism. It could be argued, on prior epistimic grounds, that
Ross's frequent appeals to self-evidence and to synthetic a
priori truth for particular obligations is illicit or that his
appeal to the moral consciousness of the 'best people' as the
ethical foundation on which we must build is ethno-centric.
But I shall not press these objections, but shall content my-
self with pressing one that Ross himself has tried hard to
meet, and thus the argument will stay on grounds that would
seem relevant to Ross.
It is certainly true, as Ross contends, that ought (at least
in an ethical sense) implies can and that it is our obligation
only to make a serious self-exertion to perform an act and
not necessarily actually to perform an act from a good mo-
tive.4 This, however, rids us only of the contention that what
makes an act right is that it is morally good, but good (basic
value) is used in a wider sense than morally good. Granted,
at least for the sake of this discussion, that a 'right act' means
'an act that ought to be done', it still remains at least reason-
able to believe that if we ask which of two or more acts
ought to be done, we must - unless we wish to be arbitrary -
appeal to some criterion in deciding that. The deontologist
with his specific obligations can offer us no criterion, but
2 ibid.
3 This argument has not hit rock bottom here because it could conceivably
be argued that 'optimific' entails in its meaning 'possibility*.
4 W. D. Ross, Foundations of Ethics, p. 160.

16
an axiological ethics which argues that an act is right which
produces the maximum amount of good (basic value) gives
us such a criterion. Unless we appeal to this criterion or
some alternative criterion (Ross offers us none), we have
no ground to judge which of two acts ought to be done.
However, Ross carries the argument a step further. Ross
argues that the contention that the right act is optimific and
the optimific is the right act is neither apprehended imme-
diately, as Moore thinks, nor is it an empirical generaliza-
tion that can be supported by inductive evidence. A crucial
test, that Ross actually utilizes in conducting his case here,
consists in getting us to reflect on the following schematic
ethical situation involving two alternative acts: an act A,
that involves keeping a promise, and an act B, that involves
breaking a promise, where the total resultant good of act B,
however, is slightly greater than the total resultant good of
act A. Ross agrees with the utilitarians that where the dis-
parity in terms of resultant total good between two such acts
is considerable, breaking the promise, if it results in a much
greater good, is the morally obligatory thing to do, but Ross
thinks in the above case, where the good of the act B is only
slightly greater, he has a genuine instance of a right act that
is not optimific. Yet utilitarians have responded that this
view arises from an overly narrow view of the utilitarian
weighing of good consequences. In determining what act
should, everything considered, be done, a thorough utilitar-
ian would certainly consider the weigth of promise-keeping
on individual character and its value to the whole of the com-
munity. These so-called special obligations that Ross draws
our attention to are valid factors in determining what would
be the most probably optimific. Ross's crucial test then col-
lapses.5
However, Ross is well aware that such a response will be
made and he has a reply to it. We need to see whether we
can ferret out whether it is an adequate response. He argues,

« W. D. Ross, "What Makes Right Acts Right?" in Readings in Ethical

Theory, W. Sellars and J. Hospers, (eds.), First Edition, p. 190.

17
against such a utilitarian response, that cases can arise
where, even after we have carefully weighed the utilitarian
value of keeping one's promises and the whole practice of
reinforcing promise-keeping, the value of integrity and the
like, still a very slightly greater total resultant utility
(greater resultant total value) will accrue to the act involv-
ing the breaking of the promise than to the act involving the
keeping the promise. The consistent utilitarian must there-
fore claim that in that situation we should break the promise.
Ross responds that where only slightly greater utility would
result we should not. We, in such a situation, appreciate that
it is fitting to be for the act that involves keeping the prom-
ise. Nothing else squares with our reflective moral experi-
ence, our considered specific moral convictions. And it is
these moral judgements, rooted in our experience of life,
which are more reliable as a guide to action than any ab-
stract and general moral theory. It is more reasonable and
more morally appropriate to so commit ourselves than to
allow the abstract moral theories to so commit us. Where
utilitarian theory is so at odds with our actual moral experi-
ence, it is the theory which should give way. It is our most
firmly embedded moral convictions which should be our
final test of theory and not our theory which should by the
final arbiter of the legitimacy of our specific moral convic-
tions.

II

It is surely natural to respond that in different cultures, in

different historical periods and even among different classes
within the same culture there will be different and some-
times conflicting firmly embedded considered moral convic-
tions. When we recognize this and further recognize as well
that these differences in even deeply embedded considered
convictions do not turn on any at least very plain differences
in beliefs about empirical matters of fact, it is reasonable to
be perplexed about how we can so confidently continue
to appeal to our own firmly embedded moral convictions as

18
the final test for the adequacy of our moral theorizing where
our theoretical moral conceptions are (1) coherent, (2) do
not imply false factual beliefs and (3) mesh together nice-
ly with our other conceptions. Suppose I have a deeply
embedded moral conviction S but I have a moral theory P
which I recognize is coherent, does not imply any false
factual beliefs and - S aside - fits in nicely with my other
conceptions and convictions. Suppose I also recognize that
there are other people with moral theory P or with moral
theories Q, Z, or R who do not have moral conviction S and
in some instances have a deep and firmly embedded moral
conviction not-S. If I recognize that such a situation obtains
why should I stick with S through thick and thin and modify
P or even abandon P? Why is it that I should continue to
regard as fundamental, S (one of my firmest considered
moral convictions) and be prepared to modify my theory to
match my conviction? For Ross, and for Broad and Prit-
chard as well, I should stick with my conviction, particular-
ly when it matches, as it does in this case, with my other
specific moral convictions? On such an account such moral
theories (normative ethical and metaethical) as we have
must match, must square with, such reflective considered
convictions. If they do not, the theories, not the convictions,
should be modified or abandoned.
Why should we go in that direction? Why not stick with
the theories? How, particularly given my recognition that
not every reflective person will share my conviction con-
cerning S, or indeed many of my other firm convictions,
including perhaps some of my firmest, can I feel so confi-
dent about sticking with these convictions, trying to require
the moral firmament to turn on them? The answer that is
usually forthcoming is that such convictions play in moral
theorizing a role analogous to that which observation plays
in theorizing in science. Since there is no independent appeal
in ethics to the facts, to that to which a moral statement
would correspond, there is no way of directly checking out
the truth of an appeal to our moral convictions such as there

19
is for checking out the truth of our belief that Jones is being
tortured. Such a factual belief can be checked out in a way
moral belief cannot. There is no such an observational test
for the truth of even our very firm moral belief that it is
vile to torture the innocent. That Jones is being tortured
can be, apart from any appeals to convictions, established
to be true or false by observation; that it is vile to so tor-
ture Jones cannot. Because that is true, deeply embedded
considered convictions come to have this central role in any
quest for justification in ethics. They provide, it is claimed,
the only reliable constraints on our theorizing. They are to
moral theorizing what an appeal to observation is to factual
theorizing.
Must we, or should we accept, such a methodological
claim? Why, we can ask, not replace them in their role as
a final court of appeal either (a) by an appeal to what
we like or dislike or want or do not want, or (b) by what
answers to our interests or does not answer to our interests?
I shall consider (a) first. Suppose my S (my firmly em-
bedded moral conviction) is that I should keep my promises.
Suppose one of those promises is that I should read a cer-
tain student's paper and return it to him the next day. Sup-
pose late that evening I remember that I promised, but that
by then I am rather tired and would, instead, rather watch
television. I also know that I could plausibly put him off
with an excuse. Why should I not do what I like rather than
stick with my conviction S? To say because I feel it would
be wrong only reasserts that I am going to stick by my con-
viction. But why is that, or is that, the rational thing to do
or the thing I should do? Why not test the viability of firm-
ly embedded moral convictions against human wants: against
those things we like and dislike or at least reflectively like
or dislike?
An ambiguity in 'wants' may cause confusion. Having
promised to look at my student's paper, I may, perfectly in-
teligibly, say to my wife 'I can't watch the idiot box, I want
to finish a paper before I go to bed'. That want, unlike 'I

20
want a hot chocolate before I go to bed', does not connote
anything about what I would enjoy doing or having. In that
sense it has nothing to do with what I would like (enjoy).
Rather, 'want9 in that context has a moral connotation. To
so use 'want' is to give to understand that it is something I
believe I ought to do. The sentence in which it occurs is it-
self an expression of a considered conviction; it is not some-
thing, as in the cases where it connotes enjoyment, which
may connote something distinct from and prehaps in conflict
with considered convictions.
With this ambiguity cleared up, we can see that what is
at issue with (a) is a matter of setting considered convic-
tions against what we enjoy or like having. Why test our
moral theories against such convictions rather than against
likes and dislikes? Why be so anti-hedonistic? What could
be better than that as many people as possible could have
as much as possible of whatever it is that they enjoy? We
should not give such a pride of place to considered convic-
tions but should try to ascertain what people enjoy and maxi-
mize that. But there is a catch in the above formula with
6as many people as possible' and with 'as much as possible'.
Sometimes it is empirically impossible to give everyone
what they want. X's desire to have a quiet cross-country ski-
ing trail at Bingo Pass is not compatible with Y's desire to
run his snowmobile there. Not everyone's desires can always
be satisfied - they are not always compossible - and in de-
ciding what then to do, we need principles, seemingly rest-
ing on moral convictions, which do not simply appeal to
what we enjoy. I know I would enjoy more going to bed or
listening to music than reading that term paper as I prom-
ised and I may be quite unsure what would maximize the
greatest amount of pleasure all around, but I give indepen-
dent weight to my considered convictions, in this case to my
reflective sentiments about promise-keeping. (This indeed is
what we - or most of us - do do, but how do we know, or
do we know, that that is what we should do?)
Isn't resistence to this a replay of hedonism and isn't it

21
plain enough that we have little reason for believing that
only pleasure and enjoyment are intrinsically good or, even
if such are the sole intrinsic goods, isn't it even more evident
that we do not have good grounds for believing that they
are the sole ends that are worth achieving for themselves?
Perhaps it is just fitting that certain things be done even if
they are not the most pleasure or utility maximizing alter-
natives? A sense of one's self -identity or self-knowledge or
integrity may also be intrinsically good and something that
we should prize independently of whatever utility may ac-
crue to them. That does not mean that maximizing considera-
tions are not relevant in appraising total situations. (Recall
Nozick's cases where people who are less rationally self-
aware are still happier than people who have a greater
awareness. Is enjoyment or happiness the only thing we want
for its own sake?6) If human beings lived in thorough self-
deception about themselves and yet were still happier than
they would otherwise be, is that all right? Is that a state of
affairs that we think ought to obtain? If unknown to you
someone photographs you in a moment of intimacy and then,
again unkown to you, sells the photographs to the porno
dealer in the Upper Volta, providing widespread pleasure
there, is that all right? I think most people's considered
convictions, firm bedrock moral attitudes, would be such
that they would not think those things right. But again why
accept as finally decisive considered convictions rather than
likes and dislikes, what provides enjoyment and its opposite?
What harm, a hedonistic utilitarian would ask, do such
things do? Harm or no harm, the pluralist deontologist will
reply, such things are wrong. And it isn't that they are wrong
because they would undermine self-respect, for, given the
lack of knowledge involved, they do not undermine self-
respect. To respond that people, where possible, should have
full knowledge of their life situation is to introduce a con-
sidered conviction, not that firmly shared, and it is, as well,

6 Robert Nozick, "On the Randian Argument", The Personalise Vol. 52

(1971), pp. 297-8.

22
to introduce considerations other than likes and dislikes and
thus in efect to capitulate to those who would say that ap-
peal to considered convictions is bedrock.
Ill

Let us next consider alternative (b). Instead of appealing

to wants, why not appeal to interests as bottom line in moral
argument? Let us not, except as a sometimes useful heuristic
device, appeal to considered convictions but let us appeal to
what answers to human interests instead. Interests are not
to be identified with wants. I may plainly be mistaken about
my interests in a way it is not so clear that I can be mis-
taken about my wants: 6I don't want to exercise but it clear-
ly is in my interests to' and 'I want another drink but it's
clearly not in my interests to have it' indicate that in talk-
ing about interests we are not talking about wants or even
about what are our strongest wants. So (b). is not (a) all
over again in disguise.
How do we decide what is in our interests? To say that
they are things we genuinely need only puts off the bad day,
for how do we decide what we genuinely need or even what
it is that we just plain need? Needs and interests are closely
linked and in some way are distinct from wants and desires,
though what exactly they are is hard to say.7 Perhaps we
can rightly say that what is in our interests is what we would
continue to want were we vividly aware of the causes of
our wanting it, what it means to have it and the consequences
of satisfying those desires. Interests are in short important
wants, where 'important' would be cashed out as I just have.
Interests may not be things - like that second helping of des-
sert - that we particularly want at the moment, but we would
want them if we had full knowledge under vivid recall. (We
may need to add - and surely if we do it makes for prob-
lems - something about being rational.)
7 Kai Nielsen "On Human Needs and Moral Appraisals", Inquiry, Vol. 6
(1963) and "Morality and Needs", The Business of Reason, J. J. Macintosh
and S. Coval (eds.). See also the collection edited by Ross Fitzgerald, Human
Needs and Politics.

23
 What so answers to human interests rather than what
squares with our considered convictions should be our de-
cisive test for the adequacy of moral theories or abstract
moral claims. But suppose, a critic will surely ask, that it
genuinely answers to the interests of a given community to
run rough-shod over the rights of a small, weak and despised
minority, e.g. Indians who have treaty rights to land where
the white community wants to engage in land development.
Does it follow that such individuals are so dispensible sim-
ply because we have established that human interests will
be maximally satisfied by so acting? Is that taking rights
seriously? Again our considered convictions conflict with
such an appeal to interests, though the pluralistic deontolo-
gist needs to face the counter that he should remember that
calculations of interest remain relevant because we should
acknowledge (1) that rights are defeasible and (2) that it
is questionable if an individual, after the fashion of Michael
Kohlhaas, should insist on his rights though the heavens fall.
However, Ross could, and indeed should, retort that, be-
cause action A satisfied interests to a slightly greater extent
than action B, where A also violated someone's rights
and B did not, that simply the fact of such a marginally
greater utility maximizing would not settle the issue con-
cerning whether A was the right thing to do tout court. Given
the acceptance of our ordinary moral thinking it is very
questionable indeed whether such utilitarian thinking would
be accepted. But again this just comes back to the fact that
the person making such an appeal gives this overriding
weight to firmly embedded moral convictions, but what is
being questioned is why do that when doing so does not
answer to our interests - to the interests of people generally?
We seem to be in a deadlock here and we will not escape
it by saying 'It would never be in our genuine interests to
run roughshod over the interests of anyone'. That will not
do because that is in effect to determine what is in our genu-
ine interests or true interests by appealing to our considered
moral convictions, but someone, sceptical of so treating an

24
appeal to our considered convictions, could resist such a
persuasive and in effect rethorical implicit definition or de-
lineation of interest by the emotive qualifiers 'genuine' and
'true' and claim that plainly some things which did not
square with the considered convictions of moral agents did
all the same answer to their interests.
So the question remains: why give such weight to deeply
embedded considered convictions when they conflict with
general moral theories systematically elaborated and in many
ways very plausible? Why abandon utilitarianism, where
reasoning in accordance with it does adequately answer to
our interests because utilitarianism does not square with
some of our very firmest considered convictions? Perhaps
we are too affected by early 'deontological socialization'
here - perhaps elements of a moral ideology enter in and
irrationally, or at least non-rationally, persuade us to ac-
cept principles that are both groundless and arbitrary.8 But,
ideology and the false consciousness and distorted under-
standing of social reality that goes with it, could attach to
the abstract utilitarian principles too. Two can play that
game and cry wolf about ideology. Whether we stick with con-
sidered convictions or accept interests as an even more fun-
damental court of appeal, what needs to be shown is that
either appeal must involve a mistaken understanding of so-
cial reality. It is not very evident what would show that or
even how such a conception applies here. Moreover, in any
domain justification must have an end or it wouldn't be jus-
tification and in morals the end point does not consist in

8 We should not forget that there are problems, linked with questions about
the 'genetic fallacy', about appealing to such considerations here. John An-
derson's remarks about Marx should be kept in mind in this context. John
Anderson, Studies in Empirical Philosophy, pp. 314-327. (For a sympathetic
general placement of Anderson that would help make his striking remarks
here more accessible, see John Passmore "The Anderson Legend", Times Lit-
erary Supplement, April 9, 1976, p. 415.) However, on the other side, the
work of Jiirgen Habermas, and particularly his Knowledge and Human In-
terests, should make us cautious about talk of the 'genetic fallacy'. Moreover, we
should not conclude that because a principle is groundless that it is there-
fore irrational, arbitrary or reflectively unacceptable. Not everything we reason-
ably believe we believe for a reason.

25
simply observing certain fundamental moral utterances to be
true as we can observe that the pencil is red and so ascer-
tain the truth of the claim 'The pencil is red'. There is no
such test for 'Abortion is evil', 'Capitalism ought to come
to an end' or even for 'Pleasure is good' and 'Promises are
to be kept'. I think our common moral thought - the reason-
ing implicit in our actual practices of life - is likely to in-
cline us to acceptance of pluralistic deontology, but should
we give it such weight in the face of the greater simplicity
of utilitarianism, particularly when utilitarian thinking does
have a very considerable plausibility and does not conflict
with, and indeed very well accounts for, many of our deeply
embedded moral convictions? Classical and repeatedly con-
tested claims of moral theory hang on this issue, but, as far
as I can see, the issue remains thoroughly contested with
little more that the hint, and that uncertain, that the bed-
rock appeal to our considered convictions appears impossible
reasonably to bypass, in spite of its cultural variability, and
no matter how persistent and genuine our fears of moral
ideology.

26
RESUMEN

En este articulo, Kai Nielsen se propone discutir la refutacion hecha

por W. D. Ross a la suposicion de Moore de que un acto justo es lo
mismo que un acto que "produce el mayor bien posible". Para Ross
un acto justo es aquel que "debe Uevarse a cabo" independientemen-
te de la cantidad de bien que produzca. Lo que Nielsen pretende ar-
giiir es que, si bien Ross tiene razon en pensar que "acto justo" no
es lo mismo que "acto moralmente bueno", no logra, sin embargo,
refutar la idea utilitarista de que lo que hace que un acto sea justo
es aquello que nos proporciona, bajo ciertas circunstancias, la mayor
cantidad de bien.
La preocupacion de Nielsen esta, pues, en encontrar un criterio
util para discernir, entre uno o mas actos, cual debe ser o no ejecu-
tado. Este criterio ha de ser proporcionado, piensa Nielsen, princi-
palmente por los axiologos eticos y no por los deontologistas, que se
encuentran absortos en otro tipo de obligaciones.
El criterio que Ross utiliza para discernir entre dos o mas accio-
nes es aquel que apela a las convicciones morales que se basan en la
experiencia concreta de los hombres. Por ejemplo : si tenemos que de-
cidir entre la accion A, que consiste en sostener una promesa, y la
accion B, que consiste en romperla, y siendo el caso de que el bien
acarreado por B es solo ligeramente mayor al bien acarreado por A,
tendriamos, segun Ross, que decidirnos por ejecutar A. Esta conclu-
sion, no obstante, esta en desacuerdo con la teoria utilitarista (y con
la conception de Moore) segun la cual debemos ejecutar siempre la
accion que nos proporcione el mayor bien.
Ross basa su punto de vista en la idea de que, siendo nuestras con-
vicciones morales mas arraigadas las que nos hacen inclinarnos por
A, debemos seguirlas, pues ninguna teoria abstracta de la moral pue-
de guiarnos mejor para la accion que este tipo de juicios. Sin em-
bargo, el criterio de Ross parece no resultar tan solido cuando consi-
deramos la cantidad de grupos sociales y culturales que sostienen
diferentes convicciones morales, tantas veces contradictorias entre si.
Esto lleva a Nielsen a preguntarse lo siguiente: Si tengo una teoria
P suficientemente coherente y adecuada a los hechos, la cual resulta
compatible con todas mis convicciones, excepto con una conviccion
S firmemente arraigada en mi, <?que debo hacer? <;Debo modificar
y aun deshacerme de mi teoria P para seguir siendo fiel a mi con-
viccion S? Parece absurdo, sin embargo, permanecer aferrado a una
creencia que no encaja ni con la teoria satisfactoria P ni con muchas

27
otras teorias morales perfectamente articuladas y plausibles que sos-
tienen otras personas razonables. <; Por que no buscar entonces - pre-
gunta Nielsen - otro tipo de criterio para probar nuestras teorias
morales? <?Por que no poner, en el lugar de nuestras convicciones
firmemente arraigadas, nuestros gustos y deseos, por ejemplo, o nues-
tros intereses?
El articulo de Nielsen se propone pues, considerar estos dos crite-
rios alternatives y enfrentarlos con la proposicion de Ross de que
cualquier teoria moral que este en conflicto con nuestras conviccio-
nes morales concretas debe de ser modificada o abandonada.

[Beatriz Quintero]

28