CLARK, Circuit Judge.
Plaintiff is the publisher and copyrighter of two paper-bound booklets, "High Points French Two Years" and "High Points French Three Years," compiled especially to facilitate review by New York high school students of their two or three years of French study in preparation for examinations given by the New York Board of Regents. The books are also sold generally for the use of students in preparation for college entrance board examinations. They are of 127 and 160 pages respectively, and include lessons and exercises in grammar and composition, sample copies of previous Regents' examinations, and lists of commonly used French words, with appropriate articles and English translations. Plaintiff hired for the work of compilation and arrangement two New York high school French teachers, who chose 1,590 French words for the two-year list from the Regents' syllabus of 1,623 elementary French words, and 775 for the three-year list from the Regents' syllabus of 778 more advanced words. They then supplied articles and translations from their own knowledge and from reference works.
Shortly after the publication of these books, defendant prepared for publication "French Two Years" and "French Three Years," books similarly designed and to be sold to the same trade at a price much lower than that of plaintiff's books. These, too, in addition to other material, contained word lists, compiled by Mr. Albert Beller, defendant's president. The lists in plaintiff's and defendant's books are so closely similar, according to plaintiff's contention, as to pass the bounds of coincidence and to show copying by Beller of plaintiff's lists, comprising French words, articles, and translations.
Beller admits that he owned copies of plaintiff's books at the time he arranged his own, and that he used them to some extent. He claims, however, that he relied on the Regents' syllabi for his choice of French words; on a Heath's and a Longman-Green's French-English dictionary for French articles; and on a list of English translations supplied by a French teacher, a Mr. Louis Goodman, and on the two dictionaries for translations. The similarities, he says, are only to be expected in the compilation of such stereotyped matter. The district court agreed, and dismissed the complaint. D.C.S.D.N.Y., 33 F.Supp. 276.
We think, however, that Beller's lists themselves are unimpeachable evidence that
Beller's choice of articles offers much stronger evidence of copying, for here he was helped by neither Goodman's work nor the syllabi. Of course, the dictionaries would show whether or not a word may be used as a noun, and if so, which gender it bears; but that is at most only a beginning. Thus, many French words may be properly used either as nouns or as other parts of speech. Of at least 9 such words in the three-year books — translated by Goodman by an English word equally ambiguous — 2 were treated in both plaintiff's and defendant's books as nouns by prefixing the article, while the other 7 were treated in both as adjectives by omitting the article. Likewise, 39 such words in the two-year list were accorded identical treatment by plaintiff's compilers and by Beller, a certain 9 of them as nouns, the rest as adjectives or adverbs.
Again, if a word is to be used as a noun, and its gender is known, it is still necessary to choose between the definite and the indefinite article. A fairly common practice in works of this kind is to use the definite article before all words beginning with a consonant, and the indefinite article before those beginning with a vowel or an aspirate "h." But in several instances, plaintiff's lists show an inadvertent departure from this rule; and in every such instance the departure was reproduced in defendant's lists. Both use the indefinite article before the first four nouns beginning with "B." In both, "le," instead of "l'," or the indefinite article, is erroneously used before "homage." In both, the article is entirely omitted before "arriere," translated as "back." In both, "le" is erroneously used, instead of "les," before "frais," translated as "expense"; and in both, there was a failure to indicate, either by an article or by express reference, the gender of the same four words: "midi," "les vacances," "minuit," "les guillemets."
In the case of many abstract nouns, however, such as "l'argent," translated as "money" or "the money," use of the indefinite article would be less usual than the definite, or even meaningless. Consequently plaintiff's compilers departed from the above practice in the case of 31 out of a total of 84 abstract nouns beginning with a vowel, and Beller did the same in exactly the same instances. Of course, a certain degree of similarity might be explained by Beller's ability, along with plaintiff's compilers, to choose the better usage; but in the case of not over half of these words assigned the "l'" is the necessity of it clear. It is unbelievable that the complete identity of choice was a matter of coincidence. Moreover, many of these words before which both used the indefinite article might be thought by an independent compiler to be more frequently used with the definite article.
It was also a common practice to include the English article as a part of the translation, but in the same 14 instances this rule was ignored in both plaintiff's and defendant's lists, although the French article was given in every such instance. A case could hardly be made out that it is less customary to use the article with these particular words than with many other abstractions before which the article was used.
Equally strong evidence can be derived from Beller's choice of translations. In well over 100 instances in his three-year list of 778 words, he departs from Goodman's suggestion and uses exactly the translation or translations given in plaintiff's list. Most of these he might conceivably have obtained from his two dictionaries; but the dictionary translations average six to a dozen words, Beller had not the experience or the qualifications to identify the most common usage, and he did not regularly pick the first or second meaning. In 9 instances his translations, identical with plaintiff's, are not to be found at all in Goodman's list or in the dictionaries which he claims to have used. Of his translations in the two-year lists, 29 are not to be found in the dictionaries; but as Goodman's translations for this list were never produced, no definite conclusion is possible. For 5 words in the three-year list which required one translation if used as one part of speech, another translation if used as an alternate part,
In all, 96 per cent of the 1,590 items in plaintiff's two-year list and 82 per cent of the 775 items in its three-year list were exactly reproduced in defendant's lists — as to the French word, whether or not an article was used, what it was, the English translation or translations, whether or not an article was used with it, what it was, and whether or not "to" was used before English verbs. This accumulation of evidence convinces beyond any doubt that Beller copied plaintiff's lists, with only a comparatively trivial number of changes. Since the evidence itself was uncontradictable, the inferences drawn from it by the court below cannot be sustained.
Nor do we believe that any of the other reasons advanced by defendant can support the judgment. "All the writings of an author" are entitled to be copyrighted, 17 U. S.C.A. § 4, and this includes "composite and cyclopedic works, directories, gazetteers, and other compilations." 17 U.S.C.A. § 5. There was some originality even in plaintiff's choice of the French words to be included, still more in its choice of articles, and quite a considerable amount in its choice of translations appropriate for the purpose. This last step, at least, required ingenuity equal to that used in the arrangement of a "handwriting chart," Deutsch v. Arnold, 2 Cir., 98 F.2d 686; in the compilation of a list of jewelers and their trade-marks, Jeweler's Circular Pub. Co. v. Keystone Pub. Co., 2 Cir., 281 F. 83, 26 A.L.R. 571, certiorari denied 259 U.S. 581, 42 S.Ct. 464, 66 L.Ed. 1074; or in supplying case annotations to statutes, W. H. Anderson Co. v. Baldwin Law Pub. Co., 6 Cir., 27 F.2d 82. In United Dictionary Co. v. G. & C. Merriam Co., 208 U.S. 260, 28 S.Ct. 290, 52 L. Ed. 478, it was not questioned that a dictionary might be copyrighted.
Although the word lists constitute only a small proportion — less than 15 per cent — of the printed matter included in the present editions of both parties' publications, it is evident that they are of real importance (defendant's original edition comprised only a word list and sample examinations), and entitled to copyright protection. Oxford Book Co. v. College Entrance Book Co., 2 Cir., 98 F.2d 688 (cartoons in a school history text); King Features Syndicate v. Fleischer, 2 Cir., 299 F. 533 (one character of a cartoon strip).
Both plaintiff's and defendant's books met exactly the same demand on the same market, and defendant's copying was unquestionably to avoid the trouble or expense of independent work. This is an unfair use. West Pub. Co. v. Lawyers' Co-operative Pub. Co., 2 Cir., 79 F. 756, 35 L. R.A. 400; see Lawrence v. Dana, Fed.Cas. No. 8,136. There is nothing to the contrary in Oxford Book Co. v. College Entrance Book Co., supra, 98 F.2d at page 691, which reversed in part a judgment for plaintiff because "no sound reason remains for saying that the accused book in its text is a copy of any substantial part of the copyrighted book."
The judgment of dismissal is therefore reversed and the action is remanded to the district court for such relief against the offending word lists by way of injunction, damages, and an accounting as it shall find appropriate in the light of this opinion. Any issue concerning an attorney's fee in this Court, as well as below, may be settled by the district court in its final judgment.