ON LEGAL THEORY AND SOCIOLOGY

Max Weber

Max Weber, “Diskussionsrede zu dem Vortrag von H. Kantorowicz ‘Rechtswissenschaft und Soziologie’,” Gesammelte Aufsätze zur Soziologie und Sozialpolitik (Tübingen: Mohr, 1924), 477–81. Originally appeared in Verhandlungsberichte über den 1. Deutschen Soziologentag (Glashütten im Taunus: Auvermann, 1911), 323–30.

We can view a specific legal precept [Rechtssatz]—for example, a clause of the Civil Code—in two very different ways, or rather, a specific legal precept is something very different according to the question we bring to it. We can, on the one hand, ask about the “meaning” of the precept; that is, assuming its existence as a general, hypothetical norm, we can ask: Can it, given its meaning, be applied to cases X, Y, and Z in such a way that a judge, if he wants to decide “correctly,” must decide thus and so? This is a question of doctrine, not of fact, not a sociological question in any sense of the word, but a purely legal question. On the other hand, we can view the same legal precept sociologically; immediately, it not only takes on a new meaning, but is something completely different. What “is” the legal precept, sociologically? The meaning of the precept is that there exists a specific factual probability, a “chance” that once those fact situations X, Y, and Z are present, factual consequences of a specific sort will result, factual pressure in a specific direction will be exercised in favor of those who turn in a specific way to specific institutions set up by the state—to the “courts,” who are capable of spending the

---

money it costs and willing to let themselves in for the difficulties connected with it. This chance—that the economic or other interests involved, given the average, common “interpretation” of a printed sentence in a law code, will actually be backed by the protection of the state—this chance is, in principle, a possibility whose “probability” can be calculated—in principle, not in fact—as much as any possible event in dead or living nature. The claim of legal doctrine that a legal precept with a certain content “is valid” means only, in the language of sociology, that a specific probability exists that specific factual circumstances lead to a specific coercive interference by the state. …

Now, whether in individual cases these legal precepts are actualized through a judgment that is “correct”—if we look at the meaning of the legal precept, that is, if we ask a very different question from the sociological one—well, that depends on an enormous number of sociological circumstances and quite concrete things. Certainly, in certain circumstances, even on whether the judge had a lot to drink that morning. It depends upon the lawyer's education, it depends on thousands of concrete circumstances that, whether social in nature or not, are in any case pure facticity. The “validity” of a legal precept in its sociological meaning is a factual statement about empirical probability; the validity in its legal meaning is a logical “ought,” and these are two very different things. What I have demonstrated here with a clause of the Civil Code, perhaps somewhat unclearly, may become clearer if I choose a different example. … When we look at the following sequence of sentences: “The United States has the right, as against its individual states, to sign trade treaties”—first sentence; second sentence: “accordingly, the United States has signed a trade treaty with Mexico”; third sentence: “this trade treaty is not consonant with the interests of the United States”; fourth sentence: “as the balance of payments of the United States is influenced unfavorably by it”; “the interests of the United States would have lain more in such and such a direction”; “it is the fault of the Constitution of the United States that something like that could occur”; “thus the mood in the United States is such and such,” etc. If you take these individual sentences together and ask yourself what is meant in each case by “the United States,” you would, I say, come to the conclusion: each time something different but never the legal concept, United States. The legal concept “United States” is, namely, a complex of legal norms whose meaning must be interpreted by legal scholars, while “United States” in the sense with which we deal with it in economics, sociology, politics, everywhere except legal scholarship, is a practically endless complex of parliamentarians of every possible stripe, of presidents and bureaucracy, the military, of coal mines and gold mines and blast furnaces and iron, what is or could be produced there, of workers and I don't know what else—perhaps something different in each of the cases mentioned, conceptualized differently from different points of view. But the

---

legal concept “United States” has the enormous advantage over the sociological concept “United States” that its content is in principle logically clear, and for that reason the sociological concept and the collective concepts of other disciplines regularly take the legal concept as an orienting principle, although the legally conceptualized system “United States” is a purely ideal thought-construct, something that has no empirical reality in life as such, but instead is something that is, as we used to say, “valid.” It is something that “exists,” exists empirically, only to the extent that it tends to be thought by lawyers in accordance with its valid meaning—thus in accordance, more or less precisely, with the ideal norm of legal thought—not because it is valid as this ideal norm, but because a certain chance exists that people, especially judges, act in a specific manner that accords with it.

Thus: the doctrinal view—that is, the view of the meaning of constitutional law and state law—and the view of a legally ordered commonwealth, for which, viewed doctrinally, they are supposed to be “valid”—are two entirely different things.

Now the question arises: how is it logically possible that sociological findings can nevertheless gain significance for legal considerations [Hermann] Kantorowicz mentioned first, the unavoidable gaps in the legal system in the logical sense. It is, he said, not a logically closed system. Granted! However, can we infer from this alone that the sociological view, which is so heterogeneous, is suited to help out? That is certainly not what Kantorowicz meant to say; he must have meant only that from knowledge of the factual structure of the society—or, as I quoted, of the commonwealth, which is a quite heterogeneous legal term that we use for the sake of brevity—that from this knowledge, under certain circumstances, the only possible—because the only reasonable—purpose of legal norms may be reckoned. He himself has offered, as a classic example of what he believes to be logically correct, the words of the Swiss Code, under which the judge is to decide the individual case as he would craft the legal norm were he the legislator. This is clearly not a sociological but a strictly Kantian principle, taken almost word for word from the Critique of Practical Reason. …

Now, gentlemen, what results the recognition of this principle could have for our jurisprudence is, in turn, a question that is very difficult to answer sociologically. The position of a judge in England is different from that of a judge here, and this is not at all a matter of practical indifference for the consequences that would result if we would put this, at least apparently and from the judge's perspective, very great power in the hands of the German judge, whose social position is very different. However, this question would lead us into legal politics, and we are staying away from that. But as far as facts are concerned, we should remember that gaps in the law are already today not the only instances, brought about by legislation itself, of adjudication

---

apart from law or contrary to it. For if it is correct that there can be two ways of finding law, “formal justice” and “Kadi justice,” and if [Rudolf von] Ihering could say of formal justice that form is the enemy of arbitrariness, the twin sister of freedom—we won't ask whether he was right or not—then we should recall that as a matter of fact the institution of the jury trial opens the door here through which Kadi justice apart from and contrary to law enters. Often, jurors render a finding of manslaughter because they lack the courage to accept the consequence that someone found guilty of murder on circumstantial evidence will be condemned to death. Often jurors find a man innocent of the rape of a girl because the girl previously had sexual intercourse—in both situations, contrary to law. But jurors are not obliged to give reasons; there is no institution for rectification; despite all law, and despite the fact that the law's intent, of course, is to bind the jurors to its meaning, they evade it de facto. …

And now I come to legal history. … Essentially, all that my colleagues have talked about is that sociology is to serve legal history and that legal history is to be pursued sociologically—that is, that legal history should have as its subject the facts of legal life, the way in which law actually comes to life, and not what can be construed from some past legal norms. I would like to say one thing on that: The crucial question of what is relevant to legal history—and should therefore become its object—can be decided only from a systematic perspective. Research in legal history can only be conducted in such a way—it actually is only conducted in such a way—that, if I have before me a “legal source,” by which I mean a source of knowledge of the law—be it a legal code, ancient legal sayings, a judgment, a private document, or whatever—I must necessarily first get a picture of it in legal doctrine, the validity of which legal precept it logically presupposes. I find this out by transporting myself back as far as possible into the soul of a judge of the time; and by asking how a judge of the time would have to decide in a concrete case presented to him, if this legal precept which I am construing doctrinally were taken by him as the basis of his decision. As soon as one looks at the actual process of research in legal history, this cannot be denied, even though you won't perhaps believe it at first glance. Only on the basis of these doctrinal considerations do I become capable at all of noticing that—as often enough happened—in such and such case the actual legal consciousness did not function this way, not in accordance with the ideally construable meaning that I had found. And only then does my eye really open to how the living law of the period de facto looked, that is, how it was actually expressed in real coercion; perhaps—indeed, probably—extremely contradictory, and differing from court to court. In other words, as a heuristic principle a legaldoctrinal construction is indispensable even for the law of the past, even for legal history. Thus I would consider it unjustifiable to make a distinction—

---

to view law that no longer applies only as a fact and not as a “norm,” and the law that still applies not as a fact, but as a norm. The meaning of both can be viewed in two different ways.