The problem of hindsight has been long recognized, and its perniciousness long understood. By not requiring documentation of anticipation, the appeals court has forgotten what was well understood over a century ago.

We live in an age where copious documentation for practically every technical discipline is relatively quickly available. The Internet affords ready access to incredible databases of practically all technical arts, and, at worst, provides leads to documented sources. Whereas hundreds labored in the same field a hundred years ago, now thousands do, and practically all technical accomplishment of patent-worthy note is documented. This professional prior art searcher states this with a robust sense of confidence, owing to repeated past success in invalidating patents.

Falling short of requiring documented evidence, case law has constructed a complex schema of determining obviousness that relies upon subjective determination. The recent case of Apotex v. Pfizer, where three district courts separately decided a patent valid, to be overturned by a rambunctious CAFC, is exemplary of the current legal caprice of an injudicious judiciary.

The underlying factual determinations made by the trial court that this court must review for clear error include (1) the scope and content of the prior art, (2) the level of ordinary skill in the art, (3) the differences between the claimed invention and the prior art, and (4) objective indicia of non-obviousness. Graham v. John Deere Co., 383 U.S. 1, 17 (1966).

The four-factor Graham has become a formula for chaotic application veering to impermissible hindsight. In recent rulings, like some drunken sailor fearing for the future, swearing out loud to whatever the Supreme Court may decide in the upcoming KSR v. Teleflex, lurching about, the CAFC has slid from requiring a "suggestion, motivation, or teaching" from documented prior art to a wafting of implicitness to "obviousness cannot be avoided simply by a showing of some degree of unpredictability in the art so long as there was a reasonable probability of success." Now it doesn't matter how much testing is required, only that someone could have guessed; talk about impermissible hindsight.

35 U.S.C. §103(a) - A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title [35 USC § 102], if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made.

§103(a) is, in its vagueness, insensible; case law has produced the mayhem of litigation crap shoot. Better to raise the threshold for patenting than to let uncertainty taint administration of the law. Veracity requires documentation: facts written down. Anything less is a formula for arbitrariness. Why require an expert to determine what level "ordinary skill" is? Consider the documented body of prior art the requisite level of skill. If the features matching all claim limitations are found in a combination of references, where the combination goes to the solving the same problem, consider it obvious & anticipatory; no guessing required as to skill in the art.

The court has already given broad latitude to what constitutes "publicly accessible" prior art: practically anything that passes between strangers qualifies as prior art - an abandoned figure filed with the Quebec patent office that was never included in the final patent makes the grade. Wonderful! If it was documented in the past, let's say that's good enough for what was known to one skilled in the art.

Let's also throw in, for §102, the certainty of invention: prior art to the date of filing as the basis for anticipation, with the patent filing date the certified date of invention; no more "swearing back" to an uncertain origin. The U.S. is the only country with an arcane "date of invention" left hanging as an unknown. To the degree practical, legislate certainty.

Statutorily exclude classes of technical art improvement, as contrasted with innovation, with a bright line. Discriminate against uncertainty.

In Apotex v. Pfizer, the court recounted the hoary axiom that adjusting the ranges used in a known process isn't patent news:

In re Aller, 220 F.2d 454, 456 (C.C.P.A. 1955), our predecessor court set forth the rule that the discovery of an optimum value of a variable in a known process is usually obvious. See also In re Boesch, 617 F.2d 272, 276 (C.C.P.A. 1980) (“[D]iscovery of an optimum value of a result effective variable in a known process is ordinarily within the skill of the art.”).

Fine. Render unpatentable by law such fiddling of known processes.

Already, by judicial precedent, new uses of known devices are unpatentable. Put that into law. But case law is on the fence about new uses of old processes, which may be patentable.

The application of an old process or machine to a similar or analogous subject, with no change in the manner of application and no result substantially distinct in its nature, will not sustain a patent even if the new form of result had not before been contemplated. - In re Pearson, 494 F.2d 1399, 1403 (C.C.P.A. 1974)

Is a new use of a known process really an invention? Better to say "no" than leave it to a threshold of subjective judgment. Worst case, put the burden on an applicant to document long-sought need for patenting a known process; as long as the criterion is objective evidence.

The fact that the invention seems simple after it is made does not determine the question; if this were the rule, many of the most beneficial patents would be stricken down. It may be safely said that if those skilled in the mechanical arts are working in a given field, and have failed, after repeated efforts, to discover a certain new and useful improvement, that he who first makes the discovery has done more than make the obvious improvement which would suggest itself to a mechanic skilled in the art, and is entitled to protection as an inventor. - Expanded Metal Co. v. Bradford, 214 U.S. 366, 381 (1909).

There's nothing wrong with codifying into statute, and thus clarifying, what is already known in case law, and otherwise open to judicial reinterpretation. Draw the borders of patentability into law, even if comprises compiling a list of what's in and what's out from a deterministic standpoint. For example, codify in statute the degree to which prior art must be enabling to be anticipatory.

For prior art anticipation, the best patent reform would be to require what is obvious to be objectively so, even if fewer patents are granted because of it. A patent regime should provide a reasonable degree of certainty, more than just a "reasonable probability of success" of anticipation or validity. Require that "one skilled in the art" be literate.

A hat tip to Joe Miller at The Fire of Genius for inspiring this entry.