9.5 Hume Principle

Discussion of the problems of substantiation of estimates and norms should take into account the fact that the descriptions are logically not deducible assessments or norms, and from the estimates or norms are not deducible descriptions.

Descriptive statements are usually formulated with the “is” link, in the evaluative and normative statements the “must” is often used. Therefore, the idea of ​​the non-deductibility of assessments (norms) from descriptions and descriptions from assessments (norms) is also expressed in the form of a provision, which cannot be “must” from “is”, but “should” be passed from “should”.

The English philosopher D. Hume was the first to emphasize the impossibility of a logical transition from “is” to “should” and rebuked all previous ethics in that it did not take into account this important circumstance. “In every ethical theory that I have met until now,” wrote Hume, “the author for some time argues in the usual way, establishes the existence of God or sets forth his observations on human affairs; and suddenly, to my surprise, I find that instead of the usual bundle used in sentences, namely “is” or “not is”, I do not meet a single sentence that would not have “must” or “ should not. " This substitution occurs unnoticed, but, nevertheless, it is of the utmost importance. Since this “must” or “should not” expresses some new attitude or statement, the latter must be taken into account and explained, and at the same time the basis of what seems completely incomprehensible, namely how this new a relationship can be deducted from others completely different from it ... I am sure that this minor act of attention would disprove all the usual ethical systems and would show us that the distinction between vice and virtue is not based solely on the relationship between objects and does not etsya mind. "

This excerpt from Hume’s "Treatise on Human Nature" is very popular. The provision on the impossibility of a logical transition from actual assertions to assertions of obligation was called the Hume principle.

This principle served as a starting point for important methodological conclusions concerning ethics and other sciences, establishing or substantiating any statements about the obligation. It was argued, in particular, that if moral conclusions could not logically follow from non-moral assumptions, then it is impossible to substantiate moral principles, going beyond the limits of morality itself. This provision, which, it seems, asserts the independence of morality from facts, was called the “principle of autonomy of morality” and caused great controversy.

The French mathematician and physicist A. Poincaré, using this argument, tried to show the impossibility of a scientific substantiation of morality or ethics: all scientific proposals are indicative, and all moral proposals are imperative; from indicative sentences, only indicative sentences can be obtained using logical inference; therefore, it is impossible to derive a moral proposition from scientific propositions. In the last century, the philosopher K. Popper attached great importance to the non-imposition statement of evaluative statements from facts. “Our decisions,” he wrote, “are never derived from facts (or statements of facts), although they have something to do with facts.”

Popper considers two examples.

The decision to fight slavery does not depend on the fact that all people are born equal and free and no one is born in chains. Indeed, even if everyone is born free, most likely, there will be people trying to chain others and even believing that they should do it. If a person notices that a certain fact can be changed - for example, the fact that many people suffer from illnesses - then in relation to this fact he can take completely different positions: decide to do everything possible to change this fact, decide to fight with any an attempt to change it or decide not to take any action against it at all. The action to make a decision, to introduce a norm or standard is a fact, but the norm or standard itself is not a fact.

The fact that most people follow the “Do not steal” norm is a sociological fact. However, the rule “Do not steal” is not a fact, and it cannot be inferred from statements describing facts. In relation to a particular fact, different and even opposite solutions are always possible. So, knowing the sociological fact that most people obey the “Do not steal” norm, we can decide to either obey this norm or fight it; we can either greet those who obey her, or scold them, urging them to obey another rule.

It is impossible to derive a proposal approving a norm, a decision, or, say, a political recommendation, from a proposal approving a fact; in other words, it is impossible to deduce norms, decisions, project proposals or recommendations from facts, Popper concludes.

Thus, the principle of Hume, which asserts the non-derivability of assessments, in particular norms and decisions, from facts, has an important theoretical and cognitive significance. Its significance for the methodology of science is also undoubted: if this principle is true, every moral, legal, etc. should be recognized as untenable. proof, in which the estimated thesis is supported by actual (descriptive) arguments.

Hume himself did not give any arguments in support of the idea of ​​the non-deducibility of "must" from "is." He referred to the fact that it would be erroneous to introduce into conclusion some relation or statement missing in the premises, and pointed out that the relation or statement expressed by means of “must” or “should not” is clearly different from the relation or statement expressed by "there is".

Of course, “must” is different from “is”, but Hume is mistaken, thinking that this remark is enough to disqualify the logical transition from premises from “is” to conclusion with “must”. The meaning required to refute this transition is as follows: the term A is clearly different from the term B, if and only if the statement containing A cannot be derived from the premises containing B and not containing A. In other words, to show that “should "Is clearly different from" is ", it is necessary to show that the statement with" must "is not deducible from the statement with" is ". But this is precisely the problem, the solution of which suggests a reference to the “clear difference” of one connection from another.

How to justify the principle of Hume, without falling into a vicious circle?

Two arguments can be made to support this principle. Firstly, all attempts to refute it have come to nothing. The failure of falsification serves as an argument in favor of the adoption of an assertion that has withstood the pressure of criticism. Secondly, the principle of Hume can be given a theoretical justification by incorporating it into a theoretical system, within the framework of which it will be the result of other, more fundamental provisions. In particular, such a system is opposed to descriptions and assessments as two polar uses of language.

Numerous counterexamples were proposed for the Hume principle, in which the evaluative (normative) conclusion was deductively derived from the premises that seemed purely descriptive. However, a more careful analysis showed that pi, one of the conclusions proposed as a counterexample, did not reach its goal: either its premises contained an implicit estimate (norm), or there was no logical consequence connection between the premises and the conclusion. It can be said that no one has been able to demonstrate the logical transition from “is” to “must” and thereby refute the principle of Hume.

A more significant theoretical argument. The description must fit the world; the task of evaluation is ultimately to bring the world in line with the assessment. These two opposite tasks are not reducible to each other. Obviously, if value is interpreted as the opposite of truth, the search for a logical transition from “is” to “must” is meaningless. There is no rational conclusion that would lead from premises containing only descriptive statements to a conclusion that is an estimate or a norm.

Although the principle of Hume seems reasonable, some methodological conclusions drawn from it need to be clarified.

Thus, Poincaré, Popper and others believed that due to the lack of a logical connection between assessments and norms with descriptions, ethics cannot have any empirical basis and, therefore, is not a science. Hume's principle often still plays a central role in the methodology of ethics and other sciences, seeking to substantiate any values ​​and requirements. Sometimes it is even claimed that, by virtue of this principle, ethics is not capable of moving from observing moral life to codifying it, and since the weight of the system of (normative) ethics is not based on facts, in this sense they are autonomous and equivalent.

Despite the fact that the principle of Hume is fair, the principle of autonomy of ethics is erroneous. Neither the logic of norms, nor the logic of ratings authorize conclusions leading from purely factual (descriptive) premises to evaluative or normative conclusions. Of course, a discussion of the specifics of the substantiation of moral norms requires consideration of this logical result. At the same time, it is clear that he does not predetermine the solution of the methodological problems of the substantiation of ethics, just as the impossibility of transitioning only logic from facts to scientific laws does not prejudge the answer to the question about the validity of theoretical knowledge.

Scientific laws do not follow logically from facts, but this does not mean that experience is indifferent for them. The transition from an empirical description to a law is not a logical conclusion; it is always a leap into uncertainty, due to the fact that the law has a dual, descriptive-essay value. The law not only summarizes known facts, but also acts as a criterion for evaluating new facts and other laws. Of course, the duality of scientific laws does not mean that every science is autonomous and does not depend on empirical material.

The fact that moral statements cannot be deduced by the rules of logic from descriptive statements is of particular interest due to the fact that in the philosophy of morality there are many concepts that substantiate the norms of morality, referring to some of their compliance with certain realities of the external world: the laws of nature, direction natural evolution, an objective course of history, etc. All these concepts are incorrect because they imply a violation of the Hume principle.

For example, the need for history prevails in history and the transition from one stage in the development of society to another is accomplished naturally, but this does not at all mean that every person is morally obligated to promote historical necessity and even try to speed up the transition dictated by it. Moral norms do not follow from sociological laws in the same way as from the law of nature, that all people are mortal, it is not a moral duty to contribute to this outcome.

The reduction of morality to historical or natural necessity is not only methodologically untenable, but also dangerous.

The situation is simpler with the principle that descriptive statements are logically deducible from evaluative statements. Neither the evaluation logic, nor the logic of norms, consider the transition from “must” to “is” to be justified. If value is understood as the opposite of truth, then not only the search for a logical transition from “is” to “should” is lost, but also the search for transition from “should” to “is”.

As an illustration, consider the so-called "Kant principle": if a person is obliged to do something, he is able to do it. If this principle is interpreted as an assumption of the possibility of a logical transition from a judgment of a duty (that is, an assessment) to a description (namely, to a description of a person’s abilities), it is incorrect. But this principle is rather a advice addressed to regulatory authority: there should be no standards set that do not go beyond (normal) human abilities. This advice, like any other, is an assessment. The “Kant principle” can be interpreted in the same way as the description: from a statement about the existence of a norm one can with some certainty deduce that the action prescribed by this norm lies within the limits of human capabilities. But this descriptive interpretation, of course, is not a counterexample to the non-deductibility of descriptions from estimates.

Values ​​are not given in direct experience, so estimates are not able to have direct empirical evidence. The evaluative statements do not imply empirical consequences, the evidence of which in experience could be interpreted as evidence in support of such statements. This means that, firstly, not only direct confirmation, but also indirect empirical confirmation is not applicable to estimates, and, secondly, that estimates cannot be not only confirmed, but also disproved through experience. Thus, the principle that denies the possibility of deriving descriptive statements from evaluative statements is just as important as the principle of Hume. These two principles speak directly to the fact that assessments should be justified quite differently from descriptions.

At the same time, the principles under consideration only limit the ways of substantiating appraisal statements, but do not exclude the very possibility of their substantiation.

The principle of Hume speaks of the absence of a logical connection between descriptions and evaluations. However, in a real situation, the matter is complicated by the fact that pure descriptions and pure assessments are not often found in the humanities and social sciences. Dual, descriptive and evaluative statements are much more frequent. Such expressions blur the boundary between descriptions and grades, as a result of which Hume’s principle loses the clarity that it achieves when juxtaposing pure descriptions and net grades.

The abundance in the sciences of culture of dual expressions intertwining “is” and “must” is the source of many obscure arguments about the relationship between factual and evaluative statements in these sciences.