Expressing Ideas: A Reply to Roger A. Shiner

I am pleased to have this opportunity to engage on matters of copyright and aesthetics and am grateful to Professor Shiner for providing it. As I pointed out in a recent piece in the American Society for Aesthetics Newsletter, copyright has been surprisingly overlooked as a topic in philosophical aesthetics and deserves much more attention from philosophers of art than it has so far gained.¹ However, it is perhaps this sentiment more than anything else that serves to draw Shiner's fervor.

i. points of confusion

Before moving on to what I take to be the more interesting issues in Shiner's discussion, I want to take a moment to clear up some points of confusion. First, Shiner takes my argument to be roughly the following: plots are neither ideas nor expressions, but, being distinct elements of already protected works, plots should be thus protected by copyright.² Much of Shiner's objections are directed at this understanding of my argument. However, this is not my argument—as a matter of law, this would be a very poor argument, indeed. Not all distinct elements of copyrighted works are protected by copyright: literary works are composed of words, but individual words are not protected by copyright, and paintings are made up of shapes and colors, but one cannot own these either. Rather, for some element of a work to merit copyright protection, that element itself must pass the minimum bar for copyrightability—among other things, that element must itself qualify as an original expression. Shiner rightly points out that this is why melodies may qualify for copyright protection, and the same, I argue, is true of plots.³ Like melodies, I contend that plots are expressions—unusual expressions in that they are neither immediately apparent nor selfsupporting, but expressions nevertheless. What do they express? What I suggest is that a literary plot expresses a chronology.⁴

In reconstructing my argument, Shiner restates my initial objection to Judge Learned Hand's “pattern test” as follows: “Intuitively, plots should be in principle protected by copyright. But they cannot be. If they are too sparse, they will be ideas, and so not protected by copyright; if they are too rich, they will be expressions, and so protected by copyright because they are expressions, not plots.”⁵

Shiner has the first part of my objection right, but not the second. To correct: if we were to accept, as Hand suggests, that the difference between plotasidea and plotasexpression is a matter of degree of detail, then an impoverished plot is only an idea, and so not copyrightable. However, if by adding detail we reach a rich enough plot to qualify as an expression, then the idea has become an expression, and so the only idea this expression can reasonably be said to express is itself. As such, it seems reasonable to suggest, idea and expression will have “merged,” and as a matter of law an expression that has merged with its idea cannot be copyrighted.⁶ Contrary to Shiner's reconstruction, I simply do not know what it would mean for something to be too richly detailed to qualify as a plot.

ii. issues of terminology

Much of my argument focuses on what ideas are, what expressions are, and what plots are. Determining whether—and how—plots may merit copyright protection, I argue, depends first upon a proper understanding of these matters. This is the central issue Shiner takes with my view. He says that, within the law, ‘idea’ and ‘expression’ have technical extensions and should not be taken necessarily to track ordinary (or even “enlightened”) usage of the terms.⁷ We cannot, in other words, rely upon a nonlegal analysis of the terms if we are looking to discuss them in a legal domain.

Certainly, the law employs ordinary daytoday terms in technical ways, and the technical usage may match to a greater or lesser degree that of ordinary usage. This should be nothing new to analytic philosophers, who make a living at doing very much the same thing. When an otherwise ordinary term is being used in the law in a technical way, however, legislators tend to go to great lengths to provide a precise technical definition to guide practice. After all, as Shiner stresses, the law is first and foremost a practical arena, and in copyright law millions of dollars can be made or lost on how some term is interpreted.

So, where might one find such a technical definition? Since we are discussing the particulars of American law, a reasonable starting place would seem to be the Copyright Act itself. The first section of the current Act sets out the central technical terminology of copyright law, and contains some fiftyone definitions—including technical definitions for ‘children’ and ‘widow.’⁸ Elsewhere in the Act, definitions are given for other terms. Nowhere in the Act, however, are definitions provided for either ‘idea’ or ‘expression.’ Nor are the terms defined in any earlier iteration of the Act. This certainly seems surprising given the central role they play in copyright law, that is, if they are to have technical definitions to be distinguished from ordinary usage.

Where else might we look? If there is an authoritative source on legal terminology, it is surely Black's Law Dictionary, the world's most cited law book. We have no problem finding Shiner's exemplars of legal terminology here. The entries on ‘harassment’ and ‘equality before the law’ are quite detailed. The entry on ‘murder’ (including all of its subtypes) takes up nearly a page, and ‘theft’ half a page. As for ‘expression,’Black’s jumps from ‘expressio falsi’ to ‘expression, freedom of’—there is no definition given for ‘expression.’ Nor is a definition provided for ‘idea.’ Rather, all we are given is a definition for ‘ideaexpression dichotomy’: “The fundamental rule that copyright law protects only specific expressions of an idea, not the idea itself.”⁹ Of course, this was the very problem we were trying to make sense of, so this is of little help.

Shiner says, “‘Idea’ and ‘expression’ … are legal terms, constructed and used by courts for legal purposes. So if anyone can confidently and accurately use those terms, it will be members of the legal profession.”¹⁰ Of course, while the courts have gotten a lot of miles out of Hand's test, the same courts continue to use the terms in their ordinary senses, so this only seems to strengthen my position. One way to take Shiner's claim, however, is that Hand's test is meant to be definitional—to itself determine what ‘idea’ and ‘expression’ mean in the law. On the letter of it, Hand treats ideas and expressions as existing on a continuum, and suggests that the line between ‘idea’ and ‘expression’ may well be impossible to fix—that it is like the vague boundary between surf and shore. Shiner may be suggesting that, within the law, this is all that ideas and expressions are: the former are things so impoverished in detail as to not qualify for protection, and the latter are things detailed enough to so qualify. Taking Hand's test as definitional, if a plot is too impoverished in detail to qualify for copyright protection, then it is merely an idea; if it is detailed enough to qualify for copyright protection, then it is an expression. But without some independent definition, it seems expressions just are things that qualify for copyright protection (passing certain minimum bars for originality and fixity). So a plot is copyrightable if it is an expression, and it is an expression if it is copyrightable. This is terribly uninformative. Hand's test is meant to be a practical means of determining copyrightability, but employing it seems to require already knowing if a thing is copyrightable. So, it would seem Hand must be operating on some independent understanding of what ideas and expressions are. But since Hand is not forthcoming with any technical definition, I cannot see any option (for the philosopher, legal scholar, or courts) but to appeal to ordinary usage.

What is most likely, of course, is that Hand's test is not meant to set new meanings for the terms, but is rather meant simply to track the distinction between ideas and expressions without thus explaining it. On this understanding, Hand's test would be a practical “rule of thumb,” a rough guide not meant to have explanatory value. (Shiner seems to say this much toward the end of his article.) This would not be unusual in copyright law; the courts have employed all manner of tests for the distinction, for example, between ‘function’ and ‘design’—what are called “conceptual separability tests”—in an attempt to track their distinction without the assumption that they are setting new meanings for the terms.¹¹ However, I would contend that, just as a number of these conceptual separability tests have lost favor in not actually tracking the distinction they are meant to track, so too does Hand's test fail to properly track the distinction between ‘idea’ and ‘expression.’ That a test in the law persists for decades in no way indicates that it is not flawed, nor does it imply any new technical meanings of the concepts involved.

iii. plots, ontology, and the law

Toward the end of my article, given their unusual nature, I admit to being unsure about how to practically go about comparing plots so as to settle claims of copyright infringement. Shiner considers expert testimony, but quickly dismisses it, concluding that “[t]he only expert decision needed is that of the court, as the decision is a legal decision.”¹² In fact, however, copyright law allows for two sorts of substantial similarity tests—“intrinsic” ordinary observer tests, which require only appeal to whether the manonthestreet would find two works to be substantially similar in “total concept and feel,” and “extrinsic” tests, which may or may not rely on expert testimony.¹³ Either or both of these tests may be employed when something other than wordforword copying is presumed to have taken place, such as in cases of plots, characters, and melodies.¹⁴ Expert testimony is used time and again in matters of musical copyright infringement, and given that the problems of comparing plots mirror those of comparing melodies, there seems little reason that the same principle could not be followed here.¹⁵ Shiner's worry seems to be that experts on literary works per se will be ill equipped to contribute to a discussion of what are essentially legal objects.

Running through Shiner's argument is the view that since the things protected by copyright (in the United States, “authored works”) are legally defined, it is thus the legislators and legal practitioners who are best suited to understanding, discussing, and comparing them. Philosophers, critics, and others concerned with such matters should stick to their own sandboxes. Shiner contends, “It misunderstands the nature of legal discourse to suppose it has exposure to the results of philosophical ontology.”¹⁶ Ontology, he seems to suggest, is a sort of idle but harmless theoretical game, while legal matters are essentially practical. I would suggest, rather, that it misunderstands the nature of ontology to suppose that it is without practical (and, indeed, legal) import. Copyright law is rife with metaphysical assumptions about its objects—beginning with the principle that authored works are abstract rather than material objects.¹⁷ The law goes further in suggesting that ideas are things themselves embodied in authored works.¹⁸ These are ontological distinctions, and in opening the door to ontology, the law invites in the philosopher. Introducing into copyright law a central distinction between ideas and expressions is like embossing the invitation in gold. And when the law is conceptually confused, whether about its own technical concepts or those of ordinary usage, I would argue not only that the door is open, but also that it is the philosopher's duty to step through it.

Footnotes