COMMENSAL ISSUE 90


The Newsletter of the Philosophical Discussion Group
Of British Mensa

Number 90 : January 1998

ARTICLES
1st January 1998 : Theo Todman

'IS', 'OUGHT' and THE VOLUNTARISTIC FALLACY

What follows is a critique of an article from the October 1997 issue of Philosophy, the Journal of the Royal Institute of Philosophy.

The original article is by Oswald Hanfling, Visiting Research Professor at the Open University, whose book The Quest for Meaning (Blackwell 1987) I remember reading with pleasure when it first came out.

I apologise to those members of the Philosophical Discussion Group of British Mensa who haven’t seen the article in question and who therefore may find what follows rather opaque. Those wishing to be enlightened further can request a copy from me.

Note : I have now remedied this deficiency, and the article in question can be viewed by clicking on the following hyperlink.

May I just exhort you all to join the Royal Institute of Philosophy ? It is open to all on presentation of the £25 annual fee, for which you receive some 600+ pages of Philosophy amongst other benefits. Write to :-

THE SECRETARY, ROYAL INSTITUTE OF PHILOSOPHY,
14 GORDON SQUARE, LONDON WC1H 0AG.


Response to

‘Is’, ‘Ought’ and the Voluntaristic Fallacy

(Oswald Hanfling; Philosophy, October 1997)


The above article commences as follows :-
"The view that ‘ought’ cannot be deduced from ‘is’, credited to Hume as a major insight into the nature of morality, is surprisingly easy to refute : (1) what they are doing is evil (2) therefore they ought not to do it. Here we have a case of deducing ‘ought’ from ‘is’. The conclusion follows, because ‘ought not’ is analytic to ‘evil’."
In the article, Professor Hanfling quotes excerpts from the last paragraph of Section 1 of Book 3 ("Of Morals") of David Hume’s 1739 work A Treatise of Human Nature. The essence of the quotation is that Hume held the view that since an ought expresses some new affirmation or relation than is contained in an is, it therefore cannot be deducible from it, because the conclusion of an argument must not contain material that is not in the premises. Professor Hanfling then provides a critique of this view, along the lines that the is has to be joined to a predicate in order to make sense, so there is already more than a mere is to lead us on to the ought. He also reviews the writings of many other philosophers (eg. J. L. Mackie, R. M. Hare, Bernard Williams and others) who have discussed the issue. I would love to enter into all these detailed arguments, but fear I would get lost in the morass, not being an academic and being substantially ignorant of the works quoted and even more derelict of the time to become acquainted with them.

What follows is nevertheless a critique of Professor Hanfling’s article. It is derived from a query as to whether or not the discussion has forgotten what the motivation behind the is / ought question really is. Surely, it has nothing specifically to do with analytic relationships between words and ideas. Rather, it is connected with the search (or the presumed failure of any such search) for a set of moral ideas that does not require underpinning by any particular conception of the deity but which could be universally assented to, based on reason alone, in all societies. If we can find a state of affairs that necessitates a particular moral response, irrespective of the society in which it occurs, then this search is hopeful of achieving its object; else not. Again, I don’t want to get caught up in what particular authors in the past did or did not intend by what they wrote. I want to do philosophy not discuss literature. The quest for a universal morality is the import of the is / ought question that is of interest to me today.

All that follows is predicated on the assumption that the search for a universal morality is what we are talking about. Otherwise, the sequence of extracts & critiques that follows is all beside the point. I will repeatedly use counter-examples in what follows. On the assumption that we are heading in the right direction, on we go.

Professor Hanfling’s first footnote quotes another example that is alleged to defeat the is / ought principle :-

Smith wants (is desirous) to get X. The best way to get X is by doing Y. Therefore, Smith ought to do Y.

What is meant by ‘best’ and ‘ought’ here ? If ‘best’ is without moral sense (ie. equivalent to ‘expedient’) then we have sanction for mugging unless ‘ought’ is also taken amorally. On the presumption that Professor Hanfling is not a crypto-mugger, we can but presume that ‘best’ is a morally circumscribed word, and therefore we do not have a pure ‘is’.

Intrinsic to an ought is the concept of duty. Now, clearly, we can conceive of specific states of affairs without number that, in our society, entail particular duties. Otherwise all talk of duties would be meaningless. For example, given that that child over there is drowning and you are a qualified lifeguard, you ought to seek to rescue it, whether or not you are ‘on duty’. We could examine quite why this ‘ought’ works in this case. It would doubtless involve the value placed by our society on children, reciprocity principles, social pressures, costs to the rescuer versus benefits to the rescued. Some would feel a strong ‘ought’ to do something in this situation even if they could do nothing, and there are tragic cases of people being drowned failing to rescue others failing to rescue drowning animals.

However, this isn’t what the is / ought principle is denying. It’s talking about the impossibility of defining a set of ethical principles necessarily valid in every society. For instance, it is possible to envisage societies, eg. in a computer game, where, taking up Professor Hanfling’s paradigmatic example of an ‘ought not’, namely murder, the state of affairs : "A kills B to take his money, which he wants to improve his standard of living", far from constituting murder would be a smart move. The reason it wouldn’t constitute murder in that society would be, as Professor Hanfling points out, because the term ‘murder’ (in our vocabulary) has moral, indeed condemnatory, overtones which in that society would be absent.

How could this state of affairs arise ? Well, maybe people in such a society would be trained from youth to do this very thing. Legalised piracy, as, for instance, that practised by the English against the Spanish in Elizabethan times, is akin to this. It just happens that in some societies, these facts (ie. killing to enhance a standard of living) constitute murder, a term which already has moral judgements and prohibitions embedded in it; whereas in others, they don’t. Again, if such people did use the term ‘murder’ in this context, they would (again as Professor Hanfling acknowledges) be using a word with merely the same sound or form. Indeed, it is possible, thankfully rarely, to hear people referring to an act of murder (as in the film Pulp Fiction), though without calling it such, without any of the connotations that we would normally use when referring to such an act. They don’t use the term because for them ‘illegal killing’ is a job. The only issue for them is to execute the job (and the victim) efficiently and avoid retribution.

Similarly, is it possible to imagine societies in which propensities such as frugality, industry and so on don’t carry the undertone of virtue ? Well, easily; for both in plutocratic society and in criminal society (as viewed by the law-abiding) such traits might be considered as vices.

The proposition "‘Frugality’ is a virtue, therefore you ought to be frugal" fails :-

  1. in case of very special considerations (eg. ‘the world will end tomorrow’).
  2. if the frugality is otherwise inappropriate (eg. in a billionaire) or misdirected (eg. in excessively restricted philanthropy).
  3. generally, because prescriptions are less binding than prohibitions (in our society), ie. we allow freedom of action (in this case, to be profligate) provided we don’t interfere too much with others’ freedom.
  4. if we generally disagree that this is so, ie. if we inhabit a different moral culture or have a fundamentally different outlook on life.
Whether a violent response to a state of affairs is deemed chivalry, rashness, pride or wickedness is society-dependent. The preservation of honour is to some an excuse for a panoply of (to us) vices. Note that duelling was once a legal, if maybe not moral, method for the strong to get what they wanted by force from the weak. Today, we’d view this as murder rather than chivalry. Consider, for instance (and allowing for some folkloreic embellishment), the "setting up" of the mathematician Evariste Galois by political opponents with a dual over a woman as a pretext for elimination.

Treating courage as a virtue may not be a personal preference, or not so for all persons at all times, but societies do have preferences. The question is, whether acting in a way in which we normally use the term ‘courageous’ could ever be separated from the approbative feelings we normally apply. Such a state of affairs is, maybe, difficult to envisage, but is nevertheless already happening to moral concepts such as patriotism or monogamy that now, instead of being purely positive concepts, tend, in some quarters at least, to indicate a narrowness of vision. The issue is not merely whether words (in our vocabulary) can have different values (in our moral structure), as in the "Dutch courage" example given by Professor Hanfling, but whether different societies might differently evaluate the same act. What we are considering is a society that has no use for the concept because it doesn’t agree with the prescription inherent in it; much as we have no word in our vocabulary for ‘(virtuously) offering one’s first-born in sacrifice to Baal’. Are there some virtues that are virtues in all conceivable societies, or vices that are universally vicious ?

In some societies, matters of no account in our society are raised to pivotal levels. For instance, failure to keep holy the Sabbath day or the commission of adultery have each been capitally proscribed; and failure to obey the prohibition has led the recalcitrant to be rapidly submerged beneath a heap of stones.

Thinking up analytic relationships (smuggling in the ‘ought’ into the ‘is’) isn’t the point. But, what is the point ? If we reject analytic connections between the ‘is’ and the ‘ought’, we are stuck with there being no necessary connection, as Hume said; rather a boring discussion all round. So we end up with logical trivialities - either we allow analytic connection or we do not. If we do, the is / ought idea is trivially falsified as Professor Hanfling demonstrates. If we disallow analyticity, Hume seems to be right.

The real issue seems to be, as previously stated, whether there exist states of affairs which, in any society conceivable, demand a consistent moral response. We’ll continue to examine a few of the test cases that Professor Hanfling raises.

Firstly, does having promised to do such and such necessarily entail that one ought always unconditionally to fulfil that promise ? Well, no. We can think of games of bluff where this would not be the case. Also, where one society doesn’t consider itself bound by promises to another. Countries confiscate or nationalise the assets of foreign institutions without compensation in the name of self-determination. People borrow money to finance risky ventures & then go bankrupt without a qualm. On the positive side, there are amnesties where promises of retribution are not fulfilled. Taking up Professor Hanfling’s point, not only can we imagine tribes in which the idea of a promise is unknown, but also those in which the breaking of promises is allowed. We can, by convention, restrict the use of the term ‘promise’ to those promises that cannot be broken, but by importing into the definition of ‘promise’ the analytic relation ‘it must not be broken’ we should not be surprised that the ought follows from the is.

Secondly, with respect to lying; some people have seen this as a universal prohibition. Others have truth-telling in an ethical hierarchy, the prohibition on lying overridden if more pressing priorities arise. As with promise-keeping, we can envisage games in which deception is allowed and in which declarative statements, normally construed as factual statements, could be intended to mislead and should therefore not be relied on - and all in the course of the normal run of the game. The question is, could we run coherent societies like that ?

As previously noted, what we are in search of is a pan-societal ethical system. We might add, also, that there is a difference between the fact that particular moral principles have been doubted and whether they should have been. In the old imperialist days, it was assumed that one’s own system of morality, even one’s own culture, was the one to which other societies ought to conform. Today, in most societies, this is not the presumption, and clearly it cannot be true for all, because ethical systems differ. But, despite the fact that societies run their affairs differently, is there no right way ?

What is important is that the vocabulary in the ‘is’ statement should be value-free. If it is value-laden, it will entail prescriptions or prohibitions. What we want is for it to deal solely with objective facts.

What, then, is an ‘objective fact’ ? A fact (assumed to be related to a general state of affairs) may be said to be objective if it would be irrational for any suitably informed person in any society to deny. Are there any such ‘facts’ ? Well, some - Pythagoras’ theorem in Euclidean geometry, for instance. However, we may be mistaken about some ‘facts’ that are currently generally agreed upon (eg. the Big Bang theory). We are left with an awkwardness concerning the ‘suitably informed’, but I can’t see how to avoid this restriction. Any attempt to propose objective ethics must be predicated on the possibility of objective knowledge, however, so we can ignore these difficulties for our present purposes. The question remains whether moral judgements are ever correctly categorised as objective facts.

If what we are saying is that, given a state of affairs and given a value system, certain actions ought to follow, we are saying something trivial; but not, I assume, what Hume meant. The question is, is there a state of affairs that morally demands the same action in any coherent value system (not just one we approve of).

My contention is that a full description of a state of affairs has to be constructed from elementary propositions, which in practise (in moral systems) leads to a value-laden proposition that entails an ‘ought’. This leads to two issues :-
  1. factual considerations that affect the moral evaluation of states of affairs (eg. stealing to escape starvation, the latter circumstance being a mitigation) and
  2. moral differences (differences in the evaluation of what’s right and wrong).

Professor Hanfling seems to deal with (1), the hierarchy of values, but not (2).

However, Professor Hanfling is correct when he says that the law can be criticised on moral grounds and therefore that the law, or institutionalism, cannot be the basis of moral systems. However, moral criticism (or approbation) of institutions varies. The evaluation of the institution of slavery is a classic case; defended on moral grounds by some as a Biblical institution, rejected by others on humanitarian grounds.

The central question in all this is whether such a precept as that above, ie. the acceptance of slavery, is fundamentally immoral in a society-independent way. Also, whether moral codes are merely societies’ constructs or whether there are any over-arching issues beyond this that might constrain systems of morality so that we could endeavour to develop a common or universal morality that would be necessary to the maintenance of any conceivable society. Only in the latter case would the is / ought principle be non-trivially falsified.

Professor Hanfling states that ‘The truth is that our moral values, and the language in which they are expressed, are given and not invented’. But, are they ? Further, he suggests that ‘Moral discussions are undertaken in terms of more fundamental values, which are not themselves candidates for acceptance or rejection’. Well, any discussion has to start from certain premises shared between the protagonists, else all we get is a slanging match. But, these shared premises change over time, and may individually each be subject to acceptance or rejection. Take the Ten Commandments, for instance; once, they’d have been an ‘all or nothing’ collection, with the order of the commandments indicating a motivational structure and relative importance. Now we disagree both as to content and sequence.

Towards the end of his essay, Professor Hanfling draws a parallel between the acceptance of moral systems and the acceptance of the rules of logic, incited by some who have denied the universal objectivity of either. He states, "there is no conceivable situation in which it would be up to us, collectively or individually, to make our own ‘value commitments’ with respect to valid reasoning." Well, it is meaningful to say that one should be bound by reason or logical argument. However, is the ‘should’ here a moral constraint or one of prudence ? Is it related to doing what is right or what is sensible ? People are not always swayed by reason - is this behaviour immoral or irrational ?

What, then, is the ‘voluntaristic fallacy’ ? According to Webster, voluntarism is :-

  1. the principle or system of doing something by or relying on voluntary action or volunteers
  2. a theory that conceives will to be the dominant factor in experience or in the world.
Professor Hanfling doesn’t (it seems to me) define precisely what he means by the voluntaristic fallacy, but it seems that, while he accepts that individuals may volunteer to support various causes or play their part in various value systems, "the existence of the relevant values is not a matter of personal commitment, any more than the existence of ‘valid’ and ‘invalid’ in the case or arguments." To believe otherwise is, it seems, to commit the voluntaristic fallacy.

While I am hopeful that Professor Hanfling is correct, I don’t believe that he has made out his case; because, to do so is dependent on him developing a universally valid system of ethics, which he hasn’t attempted in this place. Nor have I, come to that, though I have previously made a first stab at it, and may well inflict these musings on Philosophical Discussion Group members should other contributions run dry !

Theo


Response by Professor Hanfling

Original Article by Professor Hanfling

Previous Article (Editorial in Commensal 90)