Adding a challenge to a patent application

A patent application is held confidential by the patent office that receives it, until such time as it is granted. By the time anyone but the patent office and the applicant knows what the patent does, all details of the patent are published: consequently, the applicant can argue, against anyone who alleges that their idea was obvious (to practitioners of the relevant art), that the challenger can easilly call it obvious, now that they've seen the applicant's non-obvious idea published, but they didn't publish it themselves, so it can't have been so obvious previously.

It would thus seem sensible to add a section, to patent applications, which requires the applicant to describe what problems they solve and for whom, suitable for publication at the time of application. The patent office would then call for persons of reasonable competence in relevant arts to submit, to them, proposed solutions to the problems. The patent office would keep these solutions in confidence (and might ultimately return each to its submitter with a suggestion that they either publish it, in the public interest, or make it the basis of a patent application of their own) but would let the applicant know who had submitted them: the applicant can then legitimately argue that the submitter is of significantly greater than normal competence in the relevant art or else has obtained the solution improperly (e.g. by consulting someone of greater competence, or by obtaining improper access to trade secrets). They should be obliged to raise any such objection – with supporting evidence – before the patent office rules on whether the offered solution is sufficiently similar to claims in their patent to render the latter unpatentable (through obviousness). While this challenge / response / rebuttal process would likely complicate the patent application process, it would eliminate much of the present scope for absurdites after a patent has been granted – in particular, the Catch 22 situation of the presumptive non-obviousness of anything once patented.

Naturally, some refinements are needed. The applicant must be deprived of any opportunity to obfuscate the statement of problems, yet must be allowed to use pertinent technical jargon. To allow them technical jargon, one must require them first to stipulate whose jargon it is – which then implicitly restricts them to terms of art among the stipulated folk. Since laymen often lump together practitioners of several related arts, who distinguish quite nicely among themselves, specifying whose jargon they need to use may itself involve a first layer of specifying the general class of arts and then, within the common terms used by practitioners of those arts, the particular art within it. Crucially, however, they must use plain language for the first layer of specification and a layman, given that first layer, must be enabled thereby to identify the kinds of person to whom they could take the statement of problems to obtain clarification as to who to go to next. A layman should thus be able to find practitioners of the relevant art. If such practitioners cannot make sense of the statement of the problem, then that problem shall be ignorable in the application – the applicant has failed to state the problem in terms intelligible to those for whom they propose to solve it.

In specifying whose problem they are solving, the applicant would necessarily be limiting the scope of the patent: if folk other than the practitioners of the specified art have the same problem but would not understand the statement of the problem given with the patent application, their use of the methods claimed in the patent would not be covered – so the applicant would need to either state the problem, and describe its audience, entirely in plain language, or do the research to enumerate the classes of folk apt to want to solve the problem.

The applicant must likewise stand to lose out by failing to state relevant problems: but that is easilly implemented. It suffices that it shall be a valid defense, against alleged infringement, that one's use of the patented method was neither to solve one of the problems listed nor to solve a problem reasonably encompassed by the range of problems listed. The scope of encompassed should be a matter for courts to decide; but applicants would have a clear incentive to make plain what the most prominent likely applications are, so as to avoid needing to argue this point.

The plain language requirements here might plausibly be extended to the claims themselves: if the claims' description of how to solve a problem depends so heavilly on jargon specific to some particular art, alleged infringers who could not plausibly have been expected to understand that jargon could perhaps be allowed this as a defence – had they read the patent, they could not have recognised their solution as being the one it describes. If, after publication of the patent, someone publishes an account of its claims in more broadly intelligible terms than the patent applicant used, it would make sense for that to annul this defense (for anyone able to understand the new account, in so far as they could reasonably have been expected to encounter this account); but it may also make sense to consider anyone who notionally infringed prior to this clearer publication, that would have enjoyed this defense, as evidence for the obviousness of the patent.

There are general public policy arguments for discouraging patent applicants from using unclear language, at least when clear language is available: allowing a defense based on it being unreasonable to suppose the alleged infringer could understand the patent claims would provide such discouragement. A demand for clarity can reasonably be linked to the requirement for full disclosure: in an important sense, if the language of the claims is not clear, then it is not clear whether the claims actually detail everything a normally-competent practitioner of the art would need to know to implement the claimed invention.

Note that

• Problems need not be original – several patents may solve the same problem (indeed, this would foster competition, which is desirable) and a patent may solve a problem which has previously been solved in the prior art.
• The target audience of a problem need not (at the time of the application) have a problem with it – the patent effectively says only if these folk were to want to solve this problem, my claims describe a way they could do so – and, indeed, may be surprised when they discover how useful to them it is to have a solution to that problem.
• In practice, since submitters of solutions to the problems are apt to be accused of greater-than-normal competence, one way they must be able to counter this is by citing prior art and explaining how the given solution is obvious to them, as practitioners of the art; in such a context, even if the submitter is (either self-admittedly or alleged to be) of greater than normal expertise in the relevant field, the patent office can still recognise the submission in so far as it makes a clear case that the solution would be obvious to a practitioner of normal competence.
• A problem needs to be stated with enough specificity that the claims do in fact solve it: if large swathes of problems matching the offered description are not solved by (any of) the claims, then the problem was not stated clearly enough, so it would be stricken from the application, thus restricting the scope of the patent. In some cases, it may be hard to so phrase the problem as to only specify the cases actually solved by the claims: so the applicant should be allowed to state, along with the problem, some constraints without which their claimed solution would not work. (The inventor of a steam locomotive could then stipulate that their solution for a vehicle which propels itself and can pull others is limited to applications where it is no problem to require a person on board the vehicle and the primary vehicle weighing at least several hundredweight; in stating the problem, there is no need to mention that this weight is due to a furnace and the person on board is there to fuel it.)

To take an example, under such a régime, the Amazon one click patent would have had to state, as one of its problems, that it enables anyone building a web-site to ensure that visitors to that site can build up a list of items (e.g. for subsequent purchase) by clicking just once to add each item to the list. Their patent claims would then have specified implementation of this via cookies. However, any dumb schmuck familiar with the cookie spec would have been able to submit (having seen the problem but not the claimed solution) a response pointing out that this can easilly be done using cookies, and it's even an example of the class of things the original authors of the cookie spec had in mind when writing it. The examiners would then have found such submissions to be so similar to the claimed method as to render it unpatentable.

One pivotal factor here is the separation of problem from solution: one of the forms of bad patent that has been emerging in recent years amounts to a claim on the problem, asserting control over any solution of that problem. There is so much scope for obfuscation in a patent application that it may not be obvious to the patent examiners that this is what it's doing; requiring the applicant to state the problem clearly, and separately from its solution, would make it easier for the patent examiners to recognise such tricks.