The first demand to be made of a proposed scheme of subdivision is that it shall employ characteristics that may actually be recognized, so that the scheme is capable of being carried out. A mere "ideological" subdivision, under which it remains doubtful whether a thing falls under one class or the other, is a metaphysical monstrosity and should be banished from sciences claiming to be practical and exact.

39 This being so, it is surprising that some legal authors speak of "development of the contents of the statute" ("Entwicklung des Rechts-inhalts") as a separate mental operation, without making plain why the single mental operation directed towards finding, from the text of the statute, what the legal rule is, should be divided into two distinct operations. Commonly this "development" is disposed of very briefly by giving a few illustrative forms of reasoning. (See linger's "System des osterreichischen Privatrechts," section 3, sub lit. B.) What is included, is substantially the following:

1. Inference of the means intended from the purpose declared.

2. "Conclusio a majori ad minus."

3. "Conclusio a minori ad majus."

Possibly one might add the principle which prohibits "in fraudem legis agere."

Opinions differ whether the "argumentum a contrario" should be counted as "development" or interpretation. However, the division into interpretation and "development" lacks clearness and should not be adopted. It originated with the circumstance that certain observations regarding the process of interpretation, which were accurate but stood all by themselves, would not fit into any compartment of the scheme set out above, partly because the scheme is meaningless, partly because it is not practicable. So they made a new category for these observations, because they did not know what else to do with them, and gave this category the name "development." However, it is very hard to see why it is not interpretation (let us say extensive interpretation), when a lawyer holds that a party who under a statute has the right of taking water from a certain spring also has the right of access to the spring (a case of inference regarding the means from the purpose). It is hard to see where there is a difference in principle between this and ordinary interpretation.

Even to this simple requirement the schematic subdivision of the results of juridical thinking into seven classes, as set out in the preceding section, does not altogether conform. Let us try to measure the actual world by this scheme of seven classes.

It is striking to find that the very textbooks in which it is claimed that the various classes, such as interpretation by analogy and "interpretatio extensiva," are "strictly distinct" or "cannot be confounded," are contending among themselves as to which class the illustrations selected by them belong. For instance, Savigny40 offers as an illustration of "interpretatio extensiva" the case where a statute requires witnesses in order to make some transaction valid, but does not name their number. Unger41 and Thol42 severally argue otherwise and hold that the number of witnesses necessary can in this case be discovered scientifically in no way but by supplementing the legislative will, i.e. by analogy. The word "father" in section 150 of the Austrian Civil Code43 is understood by Pfaff-Krainz44 to include the mother, and this act of thinking he describes as an instance of extensive interpretation. Not a few minds would object to that and call this a case of the purest analogy.

40 "System des heutigen romischen Rechts"; Wachter is of the same mind.

41 "System," Sec. 10, note 31.

42 "Einleitung in das deutsche Privatrecht."

43 The section reads: "Where the usufruct is granted to the father by the party from whom the child derives his estate, the rents and profits remain liable for the support of the child in accordance with his station in life and cannot be taken in execution by the creditors of the father to the detriment of such support of the child." - The Supreme Court decisions of January 28, 1884 (GU 9853) and' March 7, 1865 (GU 2132) adopt the doctrine of Pfaff-Krainz.

44 "System des Privatrechts," Sec. 13, note 21.

To appreciate the full significance of such controversies one must not forget that the illustrations selected are naturally typical ones, that is, cases which exhibit the characteristics in question with particular clearness. Now, if it is possible to doubt even such examples, it must appear that the grounds on which the subdivisions are made are but specious ones. This is also brought out by Pfaff and Hoffmann, when in their Commentaries on the C'vil Code45 they speak of the point of indifference between broad interpretation and mere analogy; in fact many defects of the traditional doctrine of interpretation are shown up by these authors.

Nevertheless it must be admitted that on the whole a line may really be drawn between those cases where the result of legal thinking is contained in the meaning of the text, and on the other hand those where that is not the case,46 and that the traditional division into seven classes may be carried out in so far as it agrees with the above division into two groups. In this I am anticipating what I intend to treat more fully in the next section.

Yet even if one may, on the whole, carry out the subdivision into seven classes that does not prove that it is justifiable. It would be quite possible to classify all animals according to color, or all plants according to size, but nobody would think of such a thing. Every subdivision must be not only feasible but also, as the saying goes, productive; otherwise it is a meaningless cutting-up of the subject. To be productive means that it is possible to connect particular ideas of importance

45 Ad Sec. Sec. 6,7.

46 Or, to use the accepted terminology, cases of "interpretatio lata," either "declarativa" or "stricta." for the purpose in view with each particular class. Speaking juridically, each subdivision must be subject to a particular rule.

As a matter of fact, attempts used to be made to establish rules of this sort for those cases where a narrow interpretation was to be proper, and those in which words were to be taken in their broadest or their narrowest meaning. For instance: amendatory statutes are to be strictly construed47; grants in favor of a particular person must be strictly construed where, in a concrete case, they act unfavorably towards him48; special privileges according to some are to be liberally and according to others strictly construed49; declarations made before an officer are to be more liberally interpreted than those made in private50; particular rules derogatory of general ones must not be construed broadly nor extended by analogy. With the single exception of the rule that a particular rule shall not be extended by analogy,51 one may well say that all such maxims lack any vigorous, illuminating power of convincing. They are still carried along in the introductory chapters of textbooks, only to be disregarded later on. Whenever they are appealed to in practice, they are merely offered as ostensible reasons, while consciously or unconsciously the true reasons of the person using them are entirely different.

47 "Correctoriae leges sunt stricte interpretandae."

48 Codex de legibus, 1, 14: "Quod favore quorundam constitutum est, quibusdam casibusad laesionem eorum nolumus inventum videri."Pfaff-Hoffmann still acknowledge this rule.

49 "Beneficium imperatoris, quod a. divino scilicet ejus indulgentia proficiscitur, quam plenissime interpretari debemus." (Corpus juris, Tib. de auctoritate etc. 1, 4.) Exactly the opposite in Codex Theresianus, v, 90: ".....It is to be the general rule that no privilege is to be construed broader or more extensively than the plain text of the instrument granting it, but is to be most strictly interpreted."

50 Avenarius, "Interpretationum libri v," book 2, chap. 30: "In judiciis latior sit interpretatio quam in contractibus."

51 "Not a productive principle," says Jhering, "Geist des romischen Rechts," sect. 4, note 18.

As a matter of fact they are falling more and more into desuetude.

These detailed distinctions based on the relation between text and sense have sometimes been used in legislation as well as by legal science, especially often by way of prohibiting a liberal or a strict interpretation in particular connections.52 These prohibitions have proven impracticable and difficult of enforcement, and are now generally abandoned.53