Theory of Legal Interpretation-by Oliver Wendell Holmes, Jr.


Harvard Law Review (1899) Vol-12

THE paper upon the Principles of Legal Interpretation by Mr. F. Vaughan Hawkins, reprinted in Professor Thayer’s recently published and excellent Preliminary Treatise on Evidence, induces me to suggest what seems to me to be the theory of our rules of interpretation,—a theory which I think supports Lord Wensleydale and the others whom Mr. Hawkins quotes and disapproves, if I correctly understand their meaning and his.

It is true that in theory any document purporting to be serious and to have some legal effect has one meaning and no other, because the known object is to achieve some definite result. It is not true that in practice (and I know no reason why theory should disagree with the facts) a given word or even a given collocation of words has one meaning and no other. A word generally has several meanings, even in the dictionary. You have to consider the sentence in which it stands to decide which of those meanings it bears in the particular case, and very likely will see that it there has a shade of significance more refined than any given in the word-book. But in this first step, at least, you are not troubling yourself about the idiosyncrasies of the writer, you are considering simply the general usages of speech. So when you let whatever galvanic current may come from the rest of the instrument run through the particular sentence, you still are doing the same thing.

How is it when you admit evidence of circumstances and read the document in the light of them? Is this trying to discover the particular intent of the individual, to get into his mind and to bend what he said to what he wanted? No one would contend that such a process should be carried very far, but, as it seems to me, we do not take a step in that direction. It is not a question of tact in drawing a line. We are after a different thing. What happens is this. Even the whole document is found to have a certain play in the joints when its words are translated into things by parol evidence, as they have to be. It does not disclose one meaning conclusively according to the laws of language. Thereupon we ask, not what this man meant, but what those words would mean in the mouth of a normal speaker of English, using them in the circumstances in which they were used, and it is to the end of answering this last question that we let in evidence as to what the circumstances were. But the normal speaker of English is merely a special variety, a literary form, so to speak, of our old friend the prudent man. He is external to the particular writer, and a reference to him as the criterion is simply another instance of the externality of the law.

But then it is said, and this is thought to be the crux, In the case of a gift of Blackacre to John Smith, when the donor owned two Blackacres and the directory reveals two John Smiths, you may give direct evidence of the donor’s intention, and it is only an anomaly that you cannot give the same evidence in every case. I think, on the contrary, that the exceptional rule is a proof of the instinctive insight of the judges who established it. I refer again to the theory of our language. By the theory of our language, while other words may mean different things, a proper name means one person or thing and no other. If language perfectly performed its function, as Bentham wanted to make it, it would point out the person or thing named in every case. But under our random system it sometimes happens that your name is idem sonans with mine, and it may be the same even in spelling. But it never means you or me indifferently. In theory of speech your name means you and my name means me, and the two names are different. They are different words. Licet idem sit nomen, tamen diversum est propter diversitatem personæ.[1] In such a case we let in evidence of intention not to help out what theory recognizes as an uncertainty of speech, and to read what the writer meant into what he has tried but failed to say, but, recognizing that he has spoken with theoretic certainty, we inquire what he meant in order to find out what he has said. It is on this ground that there is no contract when the proper name used by one party means one ship, and that used by the other means another.[2] The mere difference of intent as such is immaterial. In the use of common names and words a plea of different meaning from that adopted by the court would be bad, but here the parties have said different things and never have expressed a contract. If the donor, instead of saying “Blackacre,” had said “my gold watch” and had owned more than one, inasmuch as the words, though singular, purport to describe any such watch belonging to the speaker, I suppose that no evidence of intention would be admitted. But I dare say that evidence of circumstances sufficient to show that the normal speaker of English would have meant a particular watch by the same words would be let in.

I have stated what I suppose to be our general theory of construction. It remains to say a few words to justify it. Of course, the purpose of written instruments is to express some intention or state of mind of those who write them, and it is desirable to make that purpose effectual, so far as may be, if instruments are to be used. The question is how far the law ought to go in aid of the writers. In the case of contracts, to begin with them, it is obvious that they express the wishes not of one person but of two, and those two adversaries. If it turns out that one meant one thing and the other another, speaking generally, the only choice possible for the legislator is either to hold both parties to the judge’s interpretation of the words in the sense which I have explained, or to allow the contract to be avoided because there has been no meeting of minds. The latter course not only would greatly enhance the difficulty of enforcing contracts against losing parties, but would run against a plain principle of justice. For each party to a contract has notice that the other will understand his words according to the usage of the normal speaker of English under the circumstances, and therefore cannot complain if his words are taken in that sense.[3]

Different rules conceivably might be laid down for the construction of different kinds of writing. In the case of a statute, to turn from contracts to the opposite extreme, it would be possible to say that as we are dealing with the commands of the sovereign the only thing to do is to find out what the sovereign wants. If supreme power resided in the person of a despot who would cut off your hand or your head if you went wrong, probably one would take every available means to find out what was wanted. Yet in fact we do not deal differently with a statute from our way of dealing with a contract. We do not inquire what the legislature meant; we ask only what the statute means. In this country, at least, for constitutional reasons, if for no other, if the same legislature that passed it should declare at a later date a statute to have a meaning which in the opinion of the court the words did not bear, I suppose that the declaratory act would have no effect upon intervening transactions unless in a place and case where retrospective legislation was allowed. As retrospective legislation it would not work by way of construction except in form.

So in the case of a will. It is true that the testator is a despot, within limits, over his property, but he is required by statute to express his commands in writing, and that means that his words must be sufficient for the purpose when taken in the sense in which they would be used by the normal speaker of English under his circumstances.

I may add that I think we should carry the external principle of construction even further than I have indicated. I do not suppose that you could prove, for purposes of construction as distinguished from avoidance, an oral declaration or even an agreement that words in a dispositive instrument making sense as they stand should have a different meaning from the common one; for instance, that the parties to a contract orally agreed that when they wrote five hundred feet it should mean one hundred inches, or that Bunker Hill Monument should signify Old South Church.[4] On the other hand, when you have the security of a local or class custom or habit of speech, it may be presumed that the writer conforms to the usage of his place or class when that is what a normal person in his situation would do. But these cases are remote from the point of theory upon which I started to speak.

It may be, after all, that the matter is one in which the important thing, the law, is settled, and different people will account for it by such theory as pleases them best, as in the ancient controversy whether the finder of a thing which had been thrown away by the owner got a title in privity by gift, or a new title by abandonment. That he got a title no one denied. But although practical men generally prefer to leave their major premises inarticulate, yet even for practical purposes theory generally turns out the most important thing in the end. I am far from saying that it might not make a difference in the old question to which I have referred.

Oliver Wendell Holmes.

  1. Bract. 190 a.
  2. Raffles v. Wichelhaus, 2 H. & C. 906. See Mead v. Phenix Insurance Co., 158 Mass. 124; Hanson v. Globe Newspaper Co., 159 Mass. 293, 305.
  3. In Nash v. Minnesota Title Insurance & Trust Co., 163 Mass. 574, I thought that this principle should be carried further than the majority of the court were willing to go.
  4. Goode v. Riley, 153 Mass. 585, 586.

%d bloggers like this: