About the Author - Steve Schwartz earned his Ph.D. in philosophy from Cornell in 1971. Keith Donnellan was his mentor. He has spent virtually his entire career teaching at Ithaca College. Prof. Schwartz has published numerous articles on vagueness and other subjects. This article is a revised version of an article that originally appeared in The American Philosophical Quarterly

Discuss this Article

Submit a Rebuttal

Recommend this site

 

Search The Web

Philosophy Humanities History Classics Science Math

Social Sciences Literature Art Education Linguistics Religion

Free! Examined Life Philosophy Resource Browser

Free! Philosophy Quotations Program

Worldwide Philosophy Events Database

Philosophy Resource Search Engine

 

Click here for Printer Friendly Format

Page [ 1 ] [ 2 ]

IV

At this point we should consider in more detail the principle of equality, since it plays such a key role. We have seen that the principle of equality, the principle of differential treatment, and the fact of NTS are an inconsistent triad.  To salvage morality the defender of morality must deny one of these.  But which one?  To deny the principle of differential treatment or the fact of NTS is a hopeless strategy.  The only promising avenue is to revise or reinterpret the principle of equality.  The inconsistency arises because of  the vagueness of the features that confer moral considerability.  Suppose, attempting to avoid trouble with vagueness, we do not refer to "similarity," which is non-transitive, and instead express the principle of equality as follows:  “If two things are exactly alike with respect to all morally relevant features, then they should be treated alike,” or perhaps being even more careful and precise we can say “If two things are exactly alike with respect to feature R, then they should be treated exactly alike in all respects relevant to feature R.”   Let us call this "the strict principle of equality."  Isn’t this an unassailable principle and doesn’t it avoid the problems with NTS that I have raised?  Unlike similarity, "exactly alike" seems to be transitive.

Mark Platts urges something like this response in defending his version of moral realism.  Platts is arguing against a different point than the one I have raised but his remarks are relevant.  In arguing against the use of a "similarity" principle of equality in moral reasoning Platts urges the following:

Doubtless, for the pressing purposes of everyday life, considerations of consistency with previous judgments on the basis of crudely observed similarities and of rules for making specified kinds of judgments in crudely described states of affairs are useful, perhaps indispensable, rules of thumb; but for the realist, they are no more than that.  In ordinary life, moral situations do not repeat themselves; only insensitivity can suggest that they do.[8]

No one wants to be thought to be insensitive, but it seems obvious to me that ‘moral situations’ do constantly repeat themselves--perhaps not monumental ones like the holocaust or world communism, but everyday ones like our interactions with our families, friends, and colleagues.   Platts must not be attending to the quotidian and thinking only of heroic occasions.  Moral situations constantly repeat themselves; only inattention can suggest that they do not.

In any case, adopting the strict principle of equality is not much help to morality.  If we adopt the strict principle of equality where things must be exactly alike with respect to morally relevant feature R, the principle cannot be used to extend moral considerability to animals for example, since animals are not exactly like humans in any relevant moral respect.  Humans are not even exactly like each other.  Given this, it is hard to see how the strict principle of equality could be of any use at all in extending moral considerability or in any moral argument for that matter.  We could always apparently correctly claim that we were not committed to likeness of treatment, since the items were not exactly alike with respect to any morally relevant feature.  No moral judgments or rules could be extended beyond a single case.  To employ only the strict principle of equality in morality would be to eviscerate the use of rules and principles in morality. 

A defender, such as Platts, of the strict principle of equality might argue that animals (at least higher animals) are exactly like humans with respect to a morally relevant feature R, say the capacity to feel pain, and that this is the feature we appeal to when extending moral considerability to animals.  Are animals exactly like humans with respect to the capacity to feel pain?  Are humans even exactly like each other?  I cannot see that we are.  Nor will it solve our problems to say that what is relevant is that we are exactly alike in that we have the capacity to feel pain although that capacity may be realized differently in various individuals.  The capacity to feel pain I suppose is like being a chair in terms of NTS.  It would hardly solve our problems with the series of items in the Museum of Applied Logic to appeal to the property of being of chair, so that we only call those items chairs that are exactly like the Chippendale in that they also have the property of being a chair.  Many of the items in Black’s series are exactly alike, not just similar, in that they are chairs. This hardly helps solve our problems.  It just ignores them.[9]

But isn't it the case that something either has the capacity to feel or pain or it lacks it?  And moral considerability can consistently ride on this capacity.  This seems simple enough but consider:  The capacity to feel pain is a high-level property just as being a chair is.  They supervene on other underlying natural properties in complex ways which admit of all sorts of multiple realizability and vagueness.  We could create an NTS series of items, at least counterfactually, at one end of which is a completely insensitive rock and at the other is a being with exquisite sensitivity to pain and such that no item is distinguishable with respect to the capacity to feel pain from its neighbor.  If we require that two items be exactly alike in all the underlying natural properties, then since no two items will ever be exactly alike in the underlying properties or at least very few will be, we will not be able to extend moral considerability beyond a few items.  If you are prone to doubt whether "the capacity to feel pain" is vague in this way, consider how vague the term "pain" is.  There is no sharp dividing line between painful and non-painful sensations.  Our own sensations allow of an NTS series from pleasurable to painful such that no sensation in the series is distinguisable from its neighbor.  "The capacity to feel pain" cannot be any less vague than "pain."  To invoke a real example, do you suppose that there is an exact moment when a human fetus develops the capacity to feel pain?

Platts might welcome the outcome that no moral judgments or rules could be extended beyond a single case, since he describes his view as “realistic intuitionism.”  Nevertheless, he holds that moral facts supervene on non-moral facts in such a way that he cannot avoid the problems I have been raising.  Indeed, I would hold that arguments similar to my argument about the capacity to feel pain show that any reasonable view which holds that moral facts supervene on non-moral facts is subject to problems with NTS. 

Platts says:

The realistic intuitionist holds that, while non-moral facts fix moral facts such that two circumstances cannot differ in moral respects while being alike in all non-moral respects, still moral judgments are not analyzable(or translatable) into non-moral terms;...[10]

I am here more interested in the supervenience than the non-analyzability thesis.  We must impose some sort of reasonability conditions on this supervenience.  To make any sense at all, this supervenience of the moral on the non-moral must be orderly and uniform, but now problems with NTS arise.  The underlying circumstances can be transformed, if not actually at least counterfactually, in such a way that NTS applies--just like in Black’s series of chairs.  Transform the circumstances so minutely that at each step there is no reasonable way that the change could affect the moral facts.  So at each step the moral facts must remain the same.  Yet after enough, but only finitely many, such steps the underlying circumstances will have changed significantly.

Any moral theory that views morality and justice as supervening on vague natural features must face the problems with NTS that I have raised.  There is no way to revise the principle of equality to avoid these problems and still have it serve a purpose in moral argument.

V

The only way appeal to the strict principle of equality would avail at all would be if there were a precise set of possible objects that have the relevant moral feature, say the capacity to feel pain.  Although I just now dismissed this notion there are philosophers who heroically hold such a view.  The view that there is such a precise set is called precisionism.  The precisionist holds that for each possible object either it does or does not have the capacity to feel pain, either it is or is not a chair. Since precisionism is a general response to problems generated by NTS, the precisionist holds that vagueness is only an epistemic matter.  There is a sharp dividing line between those possible items that are chairs and those that are not.  The existence of borderline cases merely reflects our ignorance of the sharp dividing line.  The precisionist holds that in Max Black's series going from the Chippendale to the lump, there is a specific last chair and a first non‑chair. For the precisionist, there really is no vagueness, just ignorance.[11]

Unfortunately precisionism is simply an incredible view, despite the many attempts to make it plausible. The idea that there is a last definite chair in Black’s series or indeed in any such series is unbelievable.  Remember there is no limit to the number of items in the series. (Although the number is always finite.)  Consider a person developing along normal lines.  The precisionist would have us believe that there is a last second of someone’s life when he is a child.  And if one second is too coarse make it a nanosecond, or one hundredth of a nanosecond.  And likewise for any vague concept.  If morality is only to be salvaged from problems of NTS by appeal to precisionism, morality is in a weak position indeed. 

In any case, even if precisionism could somehow become believable, it is of no avail in practice. According to the precisionist the cutoff point between those items that are morally considerable and those that are not is absolutely precise and unknowable. Clearly such an unknowable difference cannot be appealed to to justify a difference in treatment between adjacent items.  By the march argument, unless difference of treatment between adjacent items at some point can be justified, all items must be treated the same. This violates the principle of differential treatment.

Granted we have shifted our ground now from morally-required-to-treat-similarly to inability-to-justify-difference-of-treatment, a rather more epistemic notion, but note how serious and unavoidable the quandary still is for the precisionist.  It will not do to try to stop the march argument by pointing out that inability to justify is not transitive. From the fact that I cannot justify treating A and B differently (because there is no detectable difference between them) and I cannot justify treating B and C differently (likewise) it does not follow that I cannot justify treating A and C differently (because it does not follow that there is no detectable difference between A and C).  All this may be reasonable enough, but still leads to a violation of the basic principles of morality. Suppose I can justify treating A and C differently from each other, but I cannot justify treating A differently from B, nor B differently from C. If I then do justifiably treat A and C differently, I must be treating either A or C differently from B, but by hypothesis this difference in treatment cannot be justified.  We have a straightforward violation of Hacking’s dictum:  “Every difference, every case of privileged status or inferior status, demands a reason.”  When I treat either A or C differently from B, I can give no reason.

The problem for the precisionist is not just that there are some inscrutable cases.  A moral theory can tolerate a certain amount of inscrutability.  A utilitarian can admit that in some cases it is hard or impossible to calculate utilities far into the future.  A divine command theorist can allow that the angels of the Lord sometimes speak cryptically.  The inscrutability for the precisionist on the other hand is systematic and pervasive, and I claim leads to insurmountable difficulties.  The precisionist admits there is no detectable difference between items n and n+1, so no difference in treatment between them can be justified.  Likewise there is no detectable difference between items n+1 and n+2, so no difference in treatment between them can be justified, and so on.  Unless the precisionist is willing to give up something basic to morality as we conceive it, such as Hacking’s dictum or the principle of differential treatment, he is just as stuck as everybody else.[12]

VI

Admittedly my argument against morality has an air of paradox about it, and in particular the sorites paradox.  The sorites paradox, which is unsolved and may be unsolvable by us, is a general infection that seems to plague any attempt to be rational outside the formal domains of mathematics and deductive logic.  It is present wherever there is vagueness, which is almost everywhere.  My arguments against morality remind us of Peter Unger's in his famous and admirable papers "There are no Ordinary Things" and "Why there are no People."[13] 

The problem is, the defender of morality may assert, that we know, pace Unger, that there are chairs, refrigerators, and people.  So why not morality?  The sorites against chairs and refrigerators fails, we suspect, because there is some flaw in the sorites reasoning.  Since the sorites against morality is of the same form, it must also be flawed.  If there is such a flaw, then I have not shown that the principle of equality leads to contradiction.  Perhaps not; but no successful solution to the sorites paradox is even on the horizon.  The argument of section III demonstrating the contradiction at the heart of moral reasoning is so simple, and pellucid that it seems desperate at best to pit against it some unknown and as yet undiscovered solution to the sorites paradox.  Such a move on the part of the defender of morality is tantamount to an admission of defeat--while hollowly declaring victory.

Still, the defender of morality could respond, your arguments really are not against morality per se, but against rationality in general‑‑against the possibility of making any rational decisions.  As Hacking points out, the principle of equality is not a specifically moral principle but a requirement of reason.  It is an expression of Aristotle's dictum that we must treat similar cases similarly (rather than of Leibniz's principle of sufficient reason, contrary to Hacking).[14]  Aristotle's dictum applies to any rational action, not just morality.  Unless my arguments are specifically directed against morality, it seems perverse to claim on their basis that morality is impossible.

It is true that the principle of equality taken as a general dictum about treating similar cases similarly is not a specifically moral principle.  But it is deeply troubled as I have shown.  How much damage this trouble does depends on the use that is made of the principle.  Most ordinary affairs make little or no use of it, whereas the principle of equality is indispensable to morality.

A crucial difference between morality and other endeavors is that morality specifically invokes the principle of equality. Satisfying the principle of equality is a necessary condition of morality, the principle of equality is openly used and strictly applied in moral argument, and indeed is indispensable in moral argument.  Marcus Singer emphasizes this in his classic study Generalization in Ethics (he calls the principle of equality the "generalization principle").

The generalization principle, I shall argue, is involved in or presupposed by every genuine moral judgment, for it is an essential part of the meaning of such distinctively moral terms as "right," "wrong," and "ought," in their distinctively moral senses.  It is also an essential feature of moral reasoning, for it is presupposed in every attempt to give a reason for a moral judgment.[15]

I will not stop to summarize Singer's arguments here, but merely note as an example that he follows Sidgewick in arguing that the Golden Rule is a form of the principle of equality. 

Anyone who judges an action to be right for himself implicitly judges it to be right for anyone else whose nature and circumstances do not differ from his own in certain important respects (that is, for any similar person in similar circumstances).[16]

If what Singer says is right, morality is inextricably entwined with the principle of equality.

Other activities and practices have nothing like this kind of involvement with the principle of equality, and when they are involved with it, it is treated as overridable and dispensable. For example, in the law, precedence can be overridden.  Similar cases can be treated dissimilarly when there is good reason or even not so very good reason to do so.  In our legal and other practices we confront the principle of equality by drawing somewhat arbitrary sharp lines where none in reality exist.  In New York State an individual cannot legally buy alcoholic beverages one second before midnight on his twenty-first birthday.  Thereafter he can.  That one second makes no relevant difference to one's capacities, but a line must be drawn.  If you get your grant application in one hour after the deadline, the selection committee will reject it.  That one hour can make no difference to the committee's ability to deal with the application.  But there has to be a deadline.  We all know and understand the practical pressures that necessitate having these precise lines.  The point is that we cannot have such arbitrary lines in morality, and indeed we sense the unfairness implicit in the sharp lines the law must draw, although the unfairness is mitigated by our ability to know and understand the regulations and control our compliance.  Such regulations cause inconvenience and we are willing to tolerate this inconvenience because we recognize the practical necessity.  Moral distinctions are not matters of convenience and inconvenience.  They cannot be based on somewhat arbitrary sharp lines that do not mark real distinctions.  If they were, we would be in the position of treating A, who falls on one side of the line, differently from B, who falls on the other, even though we could not justify the difference in treatment by appealing to any morally relevant difference between A and B.  The role of the principle of equality in morality is unique and compelling.  And therein lies the problem.  It cannot be gotten around. 

The principle of equality has been shown to be a troubled principle that leads to contradiction or at least to apparently irresolvable confusion.  Any practices or arguments that specifically rest on it are troubled to the extent that it is troubled.  We cannot simply and in good conscience continue to invoke the principle of equality in the absence of answers to the objections I have raised against it.  I do not believe that there are answers to these objections, thus if I am right the principle of equality may not be invoked in morality.  To the extent that morality rests on the principle of equality, morality is impossible.[17]