"In matters of patent law, Justice Bradley was one of the most knowledgeable Justices ever to sit on the Court. He authored over 30 patent opinions for the Court in his 22 year tenure -- more than one patent majority opinion for each year he sat. At the time he was considered a very wise jurist in matters of business law generally and in patent law especially.

"I think you are right in praising his insight from the Atlantic Works v. Brady case. There are several reasons why it is not cited much right now. First, the doctrine of nonobviousness was not codified until 1952. Prior to that time, the doctrine was part of a judicially developed concept of 'invention,' which was a prerequisite of a valid patent. Generally speaking, the Supreme Court developed this doctrine on its own in the second half of the nineteenth century. In the early to middle part of the twentieth century, however, the Justices got a bit too carried away with the doctrine. Anti-patent Justices, especially Black and Douglas, were quick to invoke the doctrine to strike down seemingly every patent that came before the Court. When Congress codified the doctrine in 1952, the intent was almost certainly to ease the more extreme edges of this doctrine. Of course, the Court does not like to have its handiwork overturned (and the Court had also suggested that the invention doctrine was constitutionally required!). Thus, when the Court first interpreted the 1952 codification of nonobviousness, it stated that Congress meant only to codify existing precedents. But the Court also stated that its pre-1952 precedents had perhaps been misunderstood and were not as draconian as people thought. In other words, the Court waffled, and so the precedential value of pre-1952 Supreme Court decisions is unclear. A case like Atlantic Works v. Brady doesn't get cited much because it predates congressional codification of the nonobviousness doctrine and, despite the good sense in the opinion, the precedential value of the case is murky.

"Second, the Supreme Court hasn't cited the case recently because it has not heard a case involving the nonobviousness doctrine in more than a quarter century. I recently gave a presentation at a patent reform conference in which I gave reasons why this drought may end very soon. Many people are not satisfied with the state of the law concerning nonobviousness. This dissatisfaction is evident in both the FTC patent reform report and the National Academies' report, both of which call for a reinvigoration of the nonobviousness doctrine. These reports, coupled with other legal developments, make it likely that the Supreme Court would grant certiorari on a nonobviousness case if the parties to a case file a halfway decent petition for cert.

"Third, Atlantic Works will not be cited by the Federal Circuit because that court has constructed a very lax nonobviousness doctrine. If the pre-1952 Supreme Court erred on the side of an overly stringent nonobviousness doctrine, the modern Federal Circuit has gone wildly in the other direction. Also, Atlantic Works v. Brady is not the only Supreme Court case on nonobviousness that the Federal Circuit doesn't cite. The three most recent Supreme Court decisions concerning nonobviousness (which all are now, as noted above, more than 25 years old) are virtually never acknowledged by the Federal Circuit. The law of nonobviousness is today largely ruled by the Federal Circuit, and it owes very little debt to the teachings of the Supreme Court.

"I agree with you that Justice Bradley had some good insight in this field. The key is to realize that nonobviousness is a particularly important doctrine in an area experiencing rapid change. This change can be brought about by exogenous technical developments, or by merely changed economic or cultural conditions. But whatever the cause of change, such fields will produce many, many novelties. All will be patentable unless the nonobviousness requirement is taken seriously. I made this point in a recent article in which I say:

" ' The value of the nonobviousness requirement can be seen most directly by considering the effects if the law did enforce patent rights on new but obvious ideas. By "obvious," I mean capable of being created with little or no effort by anyone who has the standard level of skill in the art. In other words, the economic cost of producing the idea (that is, the underlying information) is very close to zero. Though trivial to produce, the idea may still have a large economic value if one could obtain a monopoly on it.

" ' For example, consider the idea of creating a streamlined "one-click" method for internet purchasing. This idea is very likely obvious, but nonetheless monopoly rights to the idea gained substantial economic value during the late 1990's due to the rise of internet commerce. For simplicity, let us assume that this idea is trivial to produce and has substantial economic value beginning around 1996 or so. If the patent system were to enforce patent rights on such an idea and the rivalry to obtain the rights were to function well (more about this second assumption in the next paragraph), then the theory presented here suggests that competition would push patenting back to a time long before 1996. Indeed, if rivalry functions well, then patenting should be pushed back so far that the patent would expire just shortly after 1996 -- in other words, it would expire just shortly after it begins to have any economic effect on the market. The royalties realized by the patentee would be just enough to cover the administrative and legal costs of obtaining the patent. The harm to society would not be great; the market would not have to endure 20 years of monopoly distortion. But there would be some harm. Society would bear the administrative costs of defining and enforcing the patent but would reap no benefit because persons of ordinary would have generated this obvious streamlining of internet check-out systems even without the incentive of patent protection.

" ' The actual history of the "one-click" example also points to another important point: The obviousness doctrine may be most important where the temporal rivalry for patents has not functioned well due to an unexpected development. For example, assume that the dramatic rise of internet commerce in the late 1990's was an unexpected development in the sense that it could not have been accurately forecast. In that case, the temporal racing of the sort assumed in this Article could not have occurred. Rather, when the unexpected development occurred, only then would individuals in the nascent industry have realized the value of streamlined internet commerce systems or, for that matter, thousands of other obvious ideas relating to internet commerce. To the extent that the patent system is willing to enforce patents on such new but obvious ideas, then the unexpected development could trigger a flood of patent applications trying to secure monopoly rights to the new but obvious ideas that have suddenly come to have evident economic value. If the nonobviousness requirement is not enforced, then society could pay a particularly heavy price. Patent racing will not be able to drive the time of patenting earlier, and the market might very well have to endure approximately two decades of monopoly distortions due to patents on ideas that would have been generated without patent incentives.'