Ty, Inc. v. GMA Accessories, Inc.

• Decision Date: 19 December 1997
• Docket Number: Nos. 97-2153,97-2356,s. 97-2153
• Citation: 132 F.3d 1167
• Parties: 1998 Copr.L.Dec. P 27,712, 45 U.S.P.Q.2d 1519 TY, INC., Plaintiff-Appellee, v. GMA ACCESSORIES, INC. and Paul HARRIS, Defendants-Appellants.
• Court: U.S. Court of Appeals — Seventh Circuit

Page 1167

132 F.3d 1167

1998 Copr.L.Dec. P 27,712, 45 U.S.P.Q.2d 1519

TY, INC., Plaintiff-Appellee, v. GMA ACCESSORIES, INC. and Paul HARRIS, Defendants-Appellants.

Nos. 97-2153, 97-2356.

United States Court of Appeals Seventh Circuit.

Argued Dec. 5, 1997.

Decided Dec. 19, 1997.

James P. White (argued), Welsh & Katz, Chicago, IL, for Plaintiff-Appellee.

Marc S. Cooperman, Charles W. Shifley (argued), Banner & Witcoff, Chicago, IL, John P. Bostany, Bostany Law Firm, New York City, for Defendants-Appellants.

Before POSNER, Chief Judge, and BAUER and FLAUM, Circuit Judges.

POSNER, Chief Judge.

Ty, the manufacturer of the popular "Beanie Babies" line of stuffed animals, has obtained a preliminary injunction under the Copyright Act against the sale by GMA (and also a retailer, but we can disregard that aspect of the injunction) of "Preston the Pig" and "Louie the Cow." These are bean-bag animals manufactured by GMA that Ty contends are copies of its copyrighted pig ("Squealer") and cow ("Daisy"). Ty began selling the "Beanie Babies" line, including Squealer, in 1993, and it was the popularity of the line that induced GMA to bring out its own line of bean-bag stuffed animals three years later. GMA does not contest the part of the injunction that enjoins the sale of Louie, but asks us on a variety of grounds to vacate the other part, the part that enjoins it from selling Preston.

We have appended to our opinion five pictures found in the appellate record. The first shows Squealer (the darker pig, actually pink) and Preston (white). The second is a picture of two real pigs. The third and fourth are different views of the design for Preston that Janet Salmon submitted to GMA several months before Preston went into production. The fifth is a picture of the two bean-bag cows; they are nearly identical. A glance at the first picture shows a striking similarity between the two bean-bag pigs as well. The photograph was supplied by GMA and actually understates the similarity (the animals themselves are part of the record). The "real" Preston is the same length as Squealer and has a virtually identical snout. The difference in the lengths of the two animals in the picture is a trick of the camera. The difference in snouts results from the fact that the pictured Preston was a manufacturing botch. And GMA put a ribbon around the neck of the Preston in the picture, but the Preston that it sells doesn't have a ribbon.

The two pigs are so nearly identical that if the second is a copy of the first, the second clearly infringes Ty's copyright. But identity is not infringement. The Copyright Act forbids only copying; if independent creation results in an identical work, the creator of that work is free to sell it. Selle v. Gibb, 741 F.2d 896, 901 (7th Cir.1984); Grubb v. KMS Patriots, L.P., 88 F.3d 1, 3 (1st Cir.1996). The practical basis for this rule is that unlike the case of patents and trademarks, the creator of an expressive work--an author or sculptor or composer--cannot canvass the entire universe of copyrighted works to discover whether his poem or song or, as in this case, "soft sculpture" is identical to some work in which copyright subsists, especially since unpublished, unregistered works are copyrightable. 17 U.S.C. § 104(a); Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539, 548, 105 S.Ct. 2218, 2224, 85 L.Ed.2d 588 (1985). But identity can be powerful evidence of copying. Gaste v. Kaiserman, 863 F.2d 1061, 1068 (2d Cir.1988); Ferguson v. National Broadcasting Co., 584 F.2d 111 (5th Cir.1978). The more a work is both like an already copyrighted work and--for this is equally important--unlike anything that is in the public domain, the less likely it is to be an independent creation. As is generally true in the law, circumstantial evidence--evidence merely probabilistic rather than certain--can confer sufficient confidence on an inference, here of copying, to warrant a legal finding.

The issue of copying can be broken down into two subissues. The first is whether the alleged copier had access to the work that he is claimed to have copied; the second is whether, if so, he used his access to copy. CMM Cable Rep, Inc. v. Ocean Coast Properties, Inc., 97 F.3d 1504, 1513 (1st Cir.1996); Fisher-Price, Inc. v. Well-Made Toy Mfg. Corp., 25 F.3d 119, 123 (2d Cir.1994). It might seem that access could not be an issue where, as in this case, the allegedly copied work is a mass-produced consumer product purchasable for $5. But we shall see that GMA has attempted to make an issue of access.

Obviously, access does not entail copying. An eyewitness might have seen the defendant buy the copyrighted work; this would be proof of access, but not of copying. But copying entails access. If, therefore, two works are so similar as to make it highly probable that the later one is a copy of the earlier one, the issue of access need not be addressed separately, since if the later work was a copy its creator must have had access to the original. Selle v. Gibb, supra, 741 F.2d at 901; Gaste v. Kaiserman, supra, 863 F.2d at 1068; Ferguson v. National Broadcasting Co., supra. Of course the inference of access, and hence of copying, could be rebutted by proof that the creator of the later work could not have seen the earlier one or (an alternative mode of access) a copy of the earlier one. But unlike the court in Towler v. Sayles, 76 F.3d 579, 584-85 (4th Cir.1996), and the authors of 4 Nimmer on Copyright § 13.02[B], pp. 13-24 to 13-25 (1997), we do not read our decision in Selle to hold or imply, in conflict with the Gaste decision, that no matter how closely the works resemble each other, the plaintiff must produce some (other) evidence of access. He must produce evidence of access, all right--but, as we have just said, and as is explicit in Selle itself, see 741 F.2d at 901, a similarity that is so close as to be highly unlikely to have been an accident of independent creation is evidence of access.

What troubled us in Selle but is not a factor here is that two works may be strikingly similar--may in fact be identical--not because one is copied from the other but because both are copies of the same thing in the public domain. In such a case--imagine two people photographing Niagara Falls from the same place at the same time of the day and year and in identical weather--there is no inference of access to anything but the public domain, and, equally, no inference of copying from a copyrighted work. Id. at 904; Gracen v. Bradford Exchange, 698 F.2d 300, 304 (7th Cir.1983); Warren Publishing, Inc. v. Microdos Data Corp., 115 F.3d 1509, 1516 n. 19 (11th Cir.1997); Key Publications, Inc. v. Chinatown Today Publishing Enterprises, Inc., 945 F.2d 509, 514 (2d Cir.1991). A similarity may be striking without being suspicious.

But here it is both. GMA's pig is strikingly similar to Ty's pig but not to anything in the public domain--a real pig, for example, which is why we have included in our appendix a photograph of real pigs. The parties' bean-bag pigs bear little resemblance to real pigs even if we overlook the striking anatomical anomaly of Preston--he has three toes, whereas real pigs have cloven hooves. We can imagine an argument that the technology of manufacturing bean-bag animals somehow prevents the manufacturer from imitating a real pig. But anyone even slightly familiar with stuffed animals knows that there are many lifelike stuffed pigs on the market, and whether they are stuffed with beans or other materials does not significantly affect their verisimilitude--though here we must emphasize that any factual assertions in this opinion should be treated as tentative, since the case is before us on an appeal from the abbreviated record of a preliminary-injunction proceeding and a full trial may cast the facts in a different light.

Real pigs are not the only pigs in the public domain. But GMA has not pointed to any fictional pig in the public domain that Preston resembles. Preston resembles only Squealer, and resembles him so closely as to warrant an inference that GMA copied Squealer. In rebuttal all that GMA presented was the affidavit of the designer, Salmon, who swears, we must assume truthfully, that she never looked at a Squealer before submitting her design. But it is not her design drawing that is alleged to infringe the copyright on Squealer; it is the manufactured Preston, the soft sculpture itself, which, as a comparison of the first with the third and fourth pictures in the appendix reveals, is much more like Squealer than Salmon's drawing is. And remember that the manufactured Preston in the photograph is a sport, with its stubby snout and its ribbon. Interestingly, these are features of Salmon's drawing but not of the production-model Preston, suggesting design intervention between Salmon's submission and actual production.

It is true that only a few months elapsed between Salmon's submission of the drawing to GMA and the production of Preston. But the record is silent on how long it would have taken to modify her design to make it more like Squealer. For all we know, it might have been done in hours--by someone who had bought a Squealer. The Beanie Babies are immensely popular. They are also, it is true, sometimes hard to find (though not this Christmas, in Chicago at any rate). Ty's practice, apparently, is to create a shortage (that is, to price its bean-bag animals below the market-clearing price) in order to excite the market. But it is unbelievable that a substantial company like GMA which is in the same line of business as Ty could not have located and purchased a Squealer if it wanted to copy it. A glance at the last picture in the appendix shows an identity between Louie the Cow and Ty's Daisy that is so complete (and also not explainable by reference to resemblance to a real cow or other public domain figure) as to compel an inference of copying. If GMA thus must have had access to...