2. that overly broad rights will chill speech, criticism, or scientific progress;
3. that it will restrict access in ways that discourage "follow- on" innovation;
4. that it will lead to industry concentration in a way that hurts consumers or citizens while being less subject to antitrust regulation precisely because the monopoly or oligopoly rests on intellectual property rights;
5. that it will establish strong "network effects" which cause the market to tip over to some inefficient technology; and
6. that it will give the rights holder control over some technology outside the range of the monopoly but closely linked to it. 52
The list of dangers goes on and on, and so does the list of exceptions, limitations, and restraints designed to prevent them. We restrict the length of intellectual property rights. (At least, we used to. The framers thought it so important to do so that they put the need to have a limited term in the Constitution itself; nevertheless both Congress and the Supreme Court seem to have given up on that one.) We restrict the scope of intellectual property rights, so that they cannot cover raw facts or general ideas, only the range of innovation and expression in between. (At least, we used to. Developments in database protection, gene patents, and business method patents are clearly eroding those walls.) As with fair use, we impose limitations on the rights when we hand them out in the first place. The exclusive right conferred by copyright does not include the right to prevent criticism, parody, classroom copying, decompilation of computer programs, and so on. (Though as the next chapter shows, a number of recent legal changes mean that the practical ability to exercise fair use rights is seriously threatened.) 53
These limitations on intellectual property do not fit a single theory, unless that theory is "avoiding the multiple and evolving dangers of intellectual property itself." Even a single limitation such as fair use clearly responds to many different concerns about the dangers of intellectual property rights. Indeed it will evolve to fit new circumstances. When computer programs were first clearly covered by copyright law, software engineers wondered if this would cripple the industry. Why? Anyone who wishes to compete with a dominant program needs to "decompile" it in order to make their program "interoperable," or simply better. For example, a new word processing program, no matter how good, would be dead on arrival unless it could read all the files people had created with the old, dominant word processing software. But to do this, the engineers at the upstart company would have to take apart their competitor's program. In the process they would have to create temporary copies of the old program, even though the final product—the hot new software—would be completely different from the old. Would this be a violation of copyright law? 54
In a series of remarkable and far-seeing cases involving such issues, the courts said no.22 "Decompilation" was fair use. The law of fair use had evolved in the context of expressive, nonfunctional, stand-alone works such as books, poems, songs. Now it was being applied to a functional product whose economics depended strongly on "network effects"—many types of programs are useful only if they are widely used. Without interoperability, we could never take our existing documents or spreadsheets or datasets and move to a new program, even if it was better. One program would not be able to read the files created by another. It would be as if language itself had been copyrighted. To have said that the incidental copies created in the process of decompiling software were actually infringements of copyright would have turned the law on its head because of a technological accident (you needed temporarily to "copy" the programs in order to understand how they worked and make yours work with them) and a legal accident (copyright was now being used to regulate functional articles of commerce: "machines" made of binary code). The difference between copying and reading, or copying and understanding, had changed because of the technology. The context had changed because the law was being stretched to cover new types of products, whose economics were very different from those of novels. Rather than let the dominant software companies use copyright to stop others from making interoperable software, the courts used an escape hatch—fair use—to prevent that danger and to uphold the basic goal of copyright: encouraging progress in science and the useful arts. 55
This long story is told to make a simple point. The variegated and evolving limitations on intellectual property are as important as the rights they constrain, curtail, and define. The holes matter as much as the cheese. 56
What does this have to do with the Sony case? In that case, remember, the Supreme Court had said that copying TV shows in order to time-shift was fair use. The Court could simply have stopped there. It could have said, "since most of what consumers do is legal, there can be no claim of contributory or vicarious infringement. Sony is not contributing to infringement since consumers are not infringing copyright by copying shows in the first place." Interestingly, though this is the heart of the ruling, the court went further. It quoted some seemingly unrelated patent law doctrine on contributory infringement: "A finding of contributory infringement does not, of course, remove the article from the market altogether; it does, however, give the patentee effective control over the sale of that item. Indeed, a finding of contributory infringement is normally the functional equivalent of holding that the disputed article is within the monopoly granted to the patentee." Clearly, the Justices were concerned that, by using copyright law, the movie studios could actually get control of a new technology. 57