THE DISTINCTION BETWEEN FACTS AND VALUES REFUTED
By Christopher Eddy
The distinction between statements of fact and judgments of value has been central to moral philosophy for the last 250 years, ever since it was first formulated by David Hume. The distinction is between “factual” or “descriptive” statements, i.e., statements which make claims about what is the case (e.g., about what a person actually did) and “evaluative”, “prescriptive” or “normative” statements, i.e., statements about what ought to be the case (e.g., about what he ought or ought not to have done).
Hume claims that prescriptive conclusions can never validly be inferred from merely descriptive premises alone, which is probably true. For example, it may be a fact that everyone does tell lies from time to time, but from this descriptive premise we cannot validly infer the prescriptive conclusion that everyone ought to tell lies from time to time.
But Hume also implicitly claims that the distinction between descriptive and prescriptive statements is absolute, so that there is no kind of descriptive statement that could ever have the force of a prescription, and no kind of value-judgment that could ever have the status of a fact. This doctrine has had enormous importance in moral philosophy because it has been taken as implying that moral judgments can have no value in terms of truth, but must be regarded ultimately as expressions of subjective will rather than descriptions of objective reality.
Hume’s claim seems convincing as long as we consider only the explicit logical content of statements and ignore the relations of authority and responsibility they implicitly invoke between interlocutors in real exchanges, - relations recognized through recent developments in the philosophy of language involving the commitments assumed and entitlements claimed and conferred when one speaker makes a statement, whether descriptive or prescriptive, in the presence of others. In this connection I would mention particularly the work of Robert Brandom, which I think provides a basis for refuting the claim that the distinction between descriptive and evaluative statements is absolute, as follows.
If I say to you that “Everyone tells lies from time to time”, my doing so is an exercise of authority. I exert my authority as a speaker in calling on you to agree to that statement or at least not contradict it, and in doing so I implicitly make myself responsible, if you challenge me, for justifying that statement to you, i.e., giving reasons why you ought to agree to it or at least not contradict it.
We both take it for granted that, when I make a statement, I am exerting authority and assuming responsibility in relation to you: it is the basic understanding that all speakers have of what is implicit in relations between them. This is the grammar of interlocutory relations which we employ intuitively, as an essential part of our linguistic competence, just as we intuitively employ our grammatical sense of the difference between questions, statements and commands.
Thus we can see that the descriptive statement, “Everyone tells lies from time to time”, is not properly recognizable as the meaningful utterance of one speaker to another unless it is understood as implicitly establishing a range of commitments and entitlements between the interlocutors. In saying that everyone tells lies, I implicitly commit myself to justifying that statement if asked to do so, i.e., to giving you reasons for agreeing with it, so that, if I cannot supply a meaningful reason when asked, I am implicitly committed to withdrawing the statement, and, if I then refuse to withdraw it, I stand in breach of my implicit commitments as a speaker. The fact that I uttered the statement implicates me necessarily in commitments and entitlements from which certain conclusions follow which are unmistakably evaluative, prescriptive or normative. In other words, Hume’s distinction is not absolute: there are normative facts.
This analysis may seem trivial when applied to descriptive statements, but it has very powerful implications when applied to prescriptions. If I propose a course of action, I implicitly ask for your consent to it, which means that there are certain types of action which I could not propose to you without implicit self-contradiction, namely, those which could occur, by definition, only without your consent, e.g., that you be “raped, enslaved, tortured or murdered”; and this entails that these are types of action which can never meaningfully be justified, because there could never be a reason for your consenting to a proposal which could be acted on only without your consent. I call this the Consent Principle.
It is a logical fact that acts of rape, enslavement, torture and murder are, in this sense, absolutely unjustifiable, but “unjustifiable” is an evaluative, prescriptive or normative term, so it is clear that there are evaluative, prescriptive or normative statements which are also statements of descriptive fact, but ones from which prescriptive conclusions can validly be drawn, and therefore that the absolute distinction between facts and values stands refuted.
Various kinds of objection have been advanced to counter the Consent Principle, as follows:
First Objection: The use of punishment, - i.e., the compulsory enforcement of legal penalties for breaches of the law, - is necessary for the maintenance of social order and is therefore justifiable in every sense that matters, but, if that is so, then it cannot be the case that the contradiction involved in consenting to be penalized without one’s consent would render punishment absolutely unjustifiable.
This objection fails because the use of punishment is not necessary to the preservation of social order. A person convicted of an offence could be asked whether he consented to the penalty provided for it by law: it would then be open to him either to consent to it and freely accept it as a voluntary penance or to refuse it. If he didn’t consent, - like a player in a game who refused a penalty awarded against him and left the field of play, - he would be placing himself in an attitude of hostility to the law, which would then entitle society to act in self-defence by interning him, as if he were a prisoner of war, until the cessation of hostilities, i.e., until either his conviction or sentence was quashed on appeal or his refusal was withdrawn.
This procedure presents no practical problems and would have more than one advantage. It would enable the state to acknowledge three important kinds of possibility without prejudging them: that the conviction might be false, even though the appeals procedure had been exhausted; that the constitutional order defended by the law might be unjust, e.g., by excluding women or some other group from the franchise; or that the law might be driven by the irrational prejudices of a majority, as with laws against gay sex. Also, if the convict is given the option to accept or reject his sentence, then, if he accepts it, he takes the first step, - however reluctantly, - towards reconciliation with society.
Finally, the American Declaration of Independence asserts that the authority of government, which necessarily includes the right to impose legal penalties, derives only from “the consent of the governed”, and the arrangement I have proposed enables the state to make that principle explicit every time a convict is sentenced and thereby to acknowledge that punishment, - penalizing without consent, - is an exercise not of responsible authority, but only of arbitrary power.
Of course it is true that being “interned” under Geneva Convention conditions is very similar to being “imprisoned”, but it is very different from being compelled to suffer death, mutilation, physical pain or the mental torments of humiliation or solitary confinement, and having it as an option completely destroys the claim that punishment is necessary to the preservation of social order on which this objection depends.
Second Objection: Soldiers defending their country against invasion are justified in killing enemy soldiers intentionally, even though those enemy soldiers do not consent to be killed, and if it is justifiable to kill people intentionally without their consent in defensive war-fighting, then it may be justifiable in a defensive campaign against crime to kill intentionally and without their consent people convicted of murder; so consent cannot be the basis of a rigid distinction between what is and what is not justifiable.
This challenge can succeed only if we ignore the glaringly obvious difference between the intentions of a soldier on the battlefield in a defensive war and those of the executioner when he carries out the sentence of death on a convict, - intentions which are made explicit in the orders on which each is authorized to act. The soldier is ordered to “advance” or “retreat”, to “take”, “hold” or “bombard” particular positions, and he is trained to use his weapons in defence of himself and his comrades when acting on those orders, but he is never ordered specifically to “injure” or to “kill” anyone. It may be that his actions will result in death or injury to enemy soldiers, but, on condition that he is acting on legal orders in a genuinely defensive campaign, he cannot be said to intend those deaths and injuries, and he cannot therefore be asked to justify proposing to cause them.
As conclusive proof that defensive military action is not intended to cause death, we may observe that, when an army has prevailed on the battlefield in a defensive action, the medics will be sent in to save as many lives as possible on both sides; indeed, it is conceivable that a defensive military action might be successful without anyone on either side being actually killed or injured, e.g., if one side surrendered or ran away.
In vivid contrast, the executioner receives an explicit order from the court to kill a specified individual at a specified time on a specified day. When the executioner sets out to comply with that order, the convict presents no threat to him against which he could claim to be acting in self-defence, because the convict is bound hand and foot and helpless in his power, - as is also true of the relation between a torturer and his victim. Furthermore, if the executioner fails in his first attempt to kill the convict, - for example, by electrocution, - he is required to administer repeated charges of electricity until the convict is dead. If a soldier treated a prisoner of war the way the executioner treats the convict, he would be guilty of a war-crime and would stand in danger of a court-martial.
It follows that the way deaths are caused in defensive war-fighting provides no moral precedent or parallel justification for the use of capital punishment, and this is so perfectly clear that anyone who maintains the contrary must be suspected of willful self-deception.
Third Objection: There are certain circumstances in which our consent, once given, cannot subsequently be retracted, so that, even if we subsequently declared that we withdrew our consent, the subsequent declaration would have no meaning, - no logical force, - because we would remain logically bound by our initial consent. Any subsequent attempt to retract would be merely apparent and therefore could not in reality contradict the initial consent, so that no contradiction would be involved in consenting initially to be disposed of subsequently, e.g., punished, without our consent; and the circumstances in which this applies occur when our consent is given contractually, i.e., when each party gives his consent irrevocably on condition that every other party also gives his consent in a similarly irrevocable manner.
This objection must fail because it entails the contradiction that we could responsibly renounce in advance the responsibility to choose thereafter, abolishing our status as responsible agents by turning ourselves into automata logically programmed by contract. This idea is so obviously absurd that those who have espoused it must ask themselves what kind of motive could have been strong enough to lead them into such an egregious error.
Fourth Objection: Actions of rape, enslavement, torture or murder could be justified as the “lesser of two evils” if refraining from them would risk consequences which were disproportionately bad, e.g., in terms of the numbers of individuals harmed, but, if this is so, then consent cannot be the basis of a rigid distinction between what is justifiable and what is not.
This challenge merely states the well-known “consequentialist” position, and its credibility derives from the fact that it is common sense to evaluate actions in terms of their foreseeable consequences: if one action is likely to cause less harm than the alternatives, common sense says that we ought to choose it. It is common sense to reason consequentially, but the central point at issue is whether there is a logical limit to the application of consequential reasoning, and, on the theory for which I am arguing, there is just such a logical limit in the facts of interlocutory grammar, the rules which all speakers intuitively recognize as implicitly governing the relations between them.
It is a fact of interlocutory grammar that whoever proposes a course of action implicitly solicits consent to it and therefore implicitly makes the factual claim that it is the kind of proposal to which everyone could without self-contradiction give their consent. Because this is a claim about logical facts, - i.e., about whether consenting to it would be self-contradictory or not, - it is open to conclusive refutation by demonstrative argument, and, if it is refuted, the proposer is thereby implicitly committed to withdrawing the proposal. This sets a limit to what can meaningfully be proposed and therefore to the kinds of proposal that can meaningfully be justified.
Fifth Objection: The so-called facts of interlocutory grammar are open to question. It is reasonable to believe that proposals are addressed not to everyone, but only to those for whom consenting would not entail a contradiction: to what we might call “coalitions of the willing”. There is then no contradiction entailed in some consenting to punish others without having to seek their consent.
If this were so, then any group which had the power to do so in a particular situation might be justified in subjecting to its collective will any other group from which it was distinguished in terms of sex, class, race, language, religion or legal status without recognizing any requirement to seek the consent of the members of the subjected group; but this perspective, which would make genocide justifiable, is in violent conflict with the universally recognized demands of justice, which require that what is done to any person be justifiable to that person, and “justice” is simply another word for the implicit rules of interlocutory grammar.
Sixth Objection: Taxation is necessary to the functioning of a complex society, so taxation must be justifiable; but some people pay their taxes not by consent, but only under compulsion: they deny that the state is justified in confiscating their property by taxation. So, if taxation is justifiable, justifiability cannot be subject to the Consent Principle.
This objection fails because, as David Hume also pointed out, property is not a natural fact, but a legal convention with rules which specify the conditions under which an individual may claim property in goods of any kind, whether it be land, livestock or money; and the laws under which people are taxed are a part of those rules, so that, to claim “property” in anything is to invoke the rules which determine what shall be regarded as the claimant’s property, and these include the tax laws. To repudiate the tax laws is to repudiate the conventions of property as such and thereby forfeit any basis for claiming that what the taxman demands of you is “yours”. We consent to be taxed by consenting to the conventions of property ownership, so taxation is always by consent.
Seventh Objection: Under the Consent Principle, a proposal would be absolutely unjustifiable only if there were at least one speaker who could not meaningfully consent to it, but this means that some kinds of proposal which most people would regard as absolutely unjustifiable, for example, torturing non-linguistic animals or pre-linguistic infants, could be justified to all speakers, because there would be no speaker for whom consenting to it involved a contradiction, and such actions would therefore have to be permitted This suggests that what most people mean by “justifiable” has no necessary relation to consent.
The objection is correct that the proposal to torture non-linguistic animals or pre-linguistic infants is not absolutely unjustifiable, but this does not mean that it must therefore be permitted by law. For example, no contradiction is involved in proposing that someone be assisted to commit suicide, but it does not follow necessarily from this that the act of assisting suicide ought to be permitted by law, because, although there are reasons for permitting it, there are other reasons which would justify prohibiting it, as there are for prohibiting cruelty to animals: in matters like this, consequential arguments must be allowed to decide the issue.
Conversely, the fact that proposing to “insult” someone would be absolutely unjustifiable (since nobody could meaningfully consent to be insulted) does not mean that insulting people ought absolutely to be prohibited by law: of course there are situations in which a certain (e.g., racial) type of insult could have socially disruptive consequences, and the law might be used to prevent that, because that is the law’s raison d’etre. The law’s purpose is to minimize harm, not to make people virtuous; but even in minimizing harm, the law is subject to the limits of the Consent Principle.
Eighth Objection: To propose a law which prohibits a type of act, like assisting a suicide, which is not absolutely unjustifiable under the Consent Principle, and which many people do passionately advocate, is to ask those advocates to consent to what they do not consent to, i.e., a law which prohibits what they advocate, and doing that seems to violate the Consent Principle; but, if that is the case, and proposing such a law is justifiable, then the Consent Principle must be rejected. Indeed, the very idea of a legal system raises the possibility that individuals will be subject to constraints to which they do not consent, so that law as such becomes unjustifiable.
This objection fails because to prohibit a type of act is not to do anything which anyone is logically unable to consent to. Even if I advocated a law which permitted assisted suicide, there would be nothing self-contradictory in my consenting to obey one which prohibited it; because, as David Hume pointed out, none of the reasons I could offer in favour of doing that or any other type of action could ever irresistibly compel me to do it and thereby force me to endure without my consent the consequences of violating a legal prohibition; besides, consenting to obey a law which prohibits a particular act does not entail consenting to be punished if convicted of breaking it.
The idea that systems of law as such could be unjustifiable makes no sense: to propose a legal system is to ask others to agree with you on a set of laws, but to propose that there be no legal system is to ask others to agree that there be no agreement, which is absurd.
This concludes my defence of the Consent Principle and my refutation of the distinction between facts and values.
Copyright © Christopher Eddy 2011
(3237 words, including title)
In my paper, A Theory of Moral Intuition, I make the claim, "THAT HUMAN SPEAKERS CAN GRASP MORAL QUALITIES, I.E., GOOD AND EVIL, INTUITIVELY.” The argument may be summarized as follows:
1.1. A quality is grasped "intuitively", by definition, when the quality is self-evident and the grasp is direct (but not necessarily infallible).
1.2. For a quality to be "self-evident" it is necessary that, like sensory qualities such as colour or taste, it not be determined by the preferences of the individual subject.
1.3. A quality is grasped "directly" (but not infallibly) when, as in seeing or tasting, the grasping is not mediated by a reasoning process.
2.0. INTENTIONS & PROPOSALS.
2.1. "Moral" qualities are qualities of actions, by definition.
2.2. Actions are, by definition, events characterized by intentions, both positive and negative.
2.3. The intention that characterizes an action can be explicitly stated, before or after the event, in the form of a proposal.
2.4. To justify an action, before or after the event, reasons must be given why others should consent to that proposal.
2.5. So moral qualities are qualities of the intentions-expressible-as-proposals whereby events are recognized as actions.
3.0. RECOGNIZING CONTRADICTIONS.
3.1. If a proposal is self-contradictory, its being so is not determined by the preferences of the individual subject: self-contradictoriness is in this sense self-evident.
3.2. Reasoned argument can be used to show that a proposal is self-contradictory, but the absurdity of proposing a contradiction is not demonstrable by argument: it can only be recognized directly.
3.3 If the self-contradictoriness of a proposal is recognized, it can be only by that kind of intuition which is involved in the recognition of linguistic meaning.
4.0. MORAL QUALITIES.
4.1. The meanings of the words "good" and "evil" derive from their role in the linguistic practices of proposing and justifying.
4.2. A type of action which absolutely cannot be recommended as "good" is evil, by definition.
4.3. "Good" and "Evil" are words used to distinguish moral qualities, i.e., the qualities of self-evidently distinct and mutually exclusive types of intention.
5.0. CONSENT & IMPLICIT SELF-CONTRADICTION.
5.1. It is self-evident that the act of proposing a particular course or general rule of action expresses the agent's intention to obtain for it the consent of other speakers.
5.2. There are types of action, such as rape, which can occur, by definition, only in the absence of the consent of the person on whom they are inflicted.
5.3. Proposing actions of that type, i.e., seeking consent to what can occur only without it, is implicitly self-contradictory and therefore absurd: this is “the Consent Principle”.
5.4. A type of action which cannot be proposed without absurdity cannot without absurdity be justified.
6.1. The self-evident difference between actions which can and actions which cannot be proposed and justified without implicit self-contradiction can be grasped directly, and therefore intuitively, though not infallibly, by all human speakers through the exercise of that faculty which makes them able to understand speech.
6.2. That difference is between actions which can without absurdity be called "good" and those which cannot, and which can therefore only be called "evil".
6.3. THEREFORE: Human speakers can grasp moral qualities, i.e., good and evil, intuitively.