The classification treated in the preceding sections is a subdivision of the accomplished results of juridical thinking, with an eye especially on the relation of text and sense. It is a classification made from the standpoint of the systematizing theorist. If we start from the standpoint of the practical lawyer who is not dealing with finished results in the shape of decisions already rendered but is looking for the right way of deciding his cases, we shall be more interested in a classification of methods than results. Then we shall discover the customary subdivision into grammatical and logical methods of interpretation. The former attempts to attain its object by a consideration of the language or words of the text; the latter employs different means. The idea underlying this classification of methods is the same as in the above classification of results. There are here two equations representing in different forms the same function and therefore necessarily in strictest dependence on each other. The more independently, pregnantly, decisively the grammatical method of interpretation is contrasted with the logical method, the more justified shall we be to distinguish results contained within the literal meaning of the text from those without. The best test of a proper classification is precisely this: that the various classes established are genetically distinct, and that they have a difference of origin. And precisely because genetically a classification according to the meaning of the text can produce but two subdivisions I cannot admit that the scheme of seven classes treated of in the preceding section is in reality anything more than a division into two classes. Now, if we consider the division of interpretation into grammatical and logical interpretation more closely, we shall discover that logically this classification is built on the importance of the linguistic factor, but goes no farther. This is just what is meant by speaking of "grammatical" interpretation; but the other class has really no significance at all. It is put down merely to make the scheme complete, is merely the opposite of the first class, but has no positive value at all. For it is certainly no positive and valuable contribution to knowledge, no characteristic of some specific method, to assert that we must never lose sight of logic, and especially not while we are conducting a scientific inquiry. Grammatical interpretation likewise needs to be logical. It is generally admitted that this second class is vague and meaningless, but sometimes the attempt is made to represent the trouble as one of terminology merely. The expression "logical" interpretation is too vague, it is said. Nobody, however, has yet found a better term. We might as well admit that it is not a question of finding a fitting name for a definite conception, but the second class simply does not represent a definite conception.

52 E.g., Codex Theresianus, v. 81, line 2: "Nobody is permitted to extend or narrow the application of a statute on pretext of a difference between the text and meaning." There is a trace of this attempt in OBGB, section 771, where we read: "within the letter and the meaning of the statute."

53 In penal law the prohibition of analogies is still in force, as regards the construction of criminality or penalties out of states of fact. The same effect, however, is often obtained by excessive projection. For instance when the usury laws had been repealed [in Austria], facts formerly constituting usury were treated as frauds.

That is why one has to resort to such all-embracing terms as "logical," or sometimes "philosophical," or, according to Thol, even "juridical."

Savigny was the first to make an attempt at substituting something definite for the vagueness of these general expressions.54 He brought to emphasis, by the side of the linguistic element, the systematic and the historical factors. Thus he obtains the grammatical, historical, and systematic methods of interpretation. Recognizing, however, that there are still other means of interpreting, he holds the notion of "logical" interpretation in reserve as a fourth category. This quadruple classification clearly enough recognizes three important factors in interpretation : the linguistic element; the tendency to reduce all legal rules to a single, unified, and self-consistent will; and finally the element of conservatism.55 Savigny's way of expounding the subject has become a permanent part of the theory of interpretation, although his most eminent pupils have rejected his classification. It is characteristic of the scholasticism still prevalent in jurisprudence that they did so, by no means because they claimed it to be erroneous but because it was not exhaustive.56 They said that there were still other factors in interpretation - in other words, his classification was rejected by them because it offended their sense of symmetry. Is it likely that in the natural sciences anybody would reject a new truth merely because it did not completely exhaust the subject?

54 "System des heutigen romischen Rechts," vol. i, pp. 212 seq.

55 Savigny called this the historical method, and while he emphasized it strongly, he defined it too narrowly. Fixing his eyes, as lawyers are apt to do, on the regulative feature, he defines the historical element in interpretation exclusively as regard for the condition of the legal rules referring to the point in question at the time when the statute to be interpreted was adopted. However, there are still other phenomena due to the conservative factor. Comp. chapter iv (Equity And Law: Judicial Freedom Of Decision), last section, infra.

56 E.g., Thol, "Einleitung in das deutsche Privatrecht," Sec. 59, note 2.

However, Savigny himself did not entirely realize the value of his classification. For instance, in his discussion57 of the means of construing defective statutes, he abandons his own division, and proposes the following list:

1. Intrinsic consistency of legislation.

2. Connection of the statute with its reason.

3. Intrinsic (?) value of the meaning arrived at by construction.

The first of these means is identical with the systematic factor. The third and last lacks all scientific definiteness. About the second we must say a few words.