The patent statute creates a general set of legal rules that govern a widevariety of technologies. With only a few exceptions, the statute does notdistinguish between different technologies in setting and applying legalstandards. In theory, then, we have a uniform patent system that providestechnology-neutral protection to all kinds of innovation. Technology,however, is anything but uniform, and displays highly diversecharacteristics across different sectors. A wealth of empirical evidencedemonstrates deep structural differences in how industries innovate.Industries vary in the speed and cost of Research and Development ("R&D"),in the ease with which inventions can be imitated by others, in the needfor cumulative or interoperative innovation rather than stand-alonedevelopment, and in the extent to which patents cover entire products ormerely components of products. We show that there is no reason to assumethat a unitary patent system will optimally encourage innovation in thewide range of diverse industries that it is expected to cover.Despite the appearance of uniformity, however, patent law is actually asvaried as the industries it seeks to foster. Closer examination of patentlaw demonstrates that it is unified only in concept. In practice the rulesactually applied to different industries have shown increasing divergence.As a practical matter, it appears that although patent law istechnology-neutral in theory, it is technology-specific in application. Thedifferential application of patent standards to different industriescorrelates with a larger theoretical confusion in patent law. While mosttheorists agree on the general utilitarian framework of patent law - thatis, they agree on the goals the patent statute is intended to achieve -they have offered radically different ideas regarding how patent law shouldbe interpreted to achieve those goals. We examine the various differenttheoretical approaches to patent law. We suggest that none of thesetheories is entirely correct. Neither are they entirely wrong. We show howvarious different theories of patent law succeed in explaining theapplication of patent law to particular industries, but fail when takenoutside the narrow context of those industries.The fact that economic evidence, patent doctrine, and legal theory all varyby industry leads us to question whether patent law should explicitlyattempt to tailor protection to the needs of specific industries. We pointout a number of risks inherent in such a technology-specific approach,particularly one administered by Congress. In particular, concerns aboutrent seeking and the inability of industry-specific statutes to respond tochanging circumstances lead us to conclude that we should not jettison ournominally uniform patent system in favor of specific statutes that protectparticular industries. Nonetheless, there are other ways the law can takeaccount of the needs of different industries. We argue that it makes senseto take economic policy and industry-specific variation explicitly intoaccount in applying general patent rules to specific cases. Patent lawgives the courts substantial freedom to do this by means of flexible legalstandards we call "policy levers." We identify ten sets of policy leversthat already exist in patent law, and the ways in which they implicitly orexplicitly permit the courts to take account of different types ofinnovation in different industries. We also identify a variety of otherplaces where the statute grants the courts substantial discretion, andsuggest ways that those discretionary standards could serve as policylevers. Finally, having identified certain policy levers and the method oftheir employment, we consider the economic characteristics of innovation infive different industries that appear to be likely candidates forindustry-specific rules: chemistry, pharmaceuticals, biotechnology,semiconductors, and software. We offer concrete suggestions as to how thecourt can and should apply particular policy levers to help encourageinnovation in these very different industries.