Logical Fallacy in Patent Law: Analysing Abolkheir’s Challenge to the Soundness of Non-obviousness Test
SpicyIP
Logical Fallacy in Patent Law: Analysing Abolkheir’s Challenge to the Soundness of Non-obviousness Test
By Bhuwan Sarine
In a recently published editorial in the Journal of Intellectual Property Law and Practice, Bristol-based philosopher Dr. Mo Abolkheir made a big claim about there being an (unnoticed) logical fallacy at the core of patent law. Questioning the logical foundations of patent laws, he argues that defining ‘inventive step’ in terms of ‘non-obviousness’ shifts the focus of inquiry to the inventor, rather than the invention itself. This happens when the invention is assessed for non-obviousness with respect to the standard of ‘person skilled in the art.’ Abolkheir has an issue with this because what ends up being examined is not the invention, but the inventor’s imaginative and cognitive skills. He even goes so far to call this inventio ad hominem fallacy, a variant of the well-known ad hominem fallacy (for the uninitiated, ad hominem is a logical fallacy in which someone’s argument is sought to be rebutted using personal attacks on his/her character, instead of addressing the argument itself).
In this post, I will reflect upon the claim made by Dr. Abolkheir. After laying the ground, I will assess the strength of his claim by addressing points for and against it. I will finally conclude with the questions this claim gives rise to.
The Argument
Across jurisdictions, ‘inventive step’ is assessed on the basis of the inventive concept’s non-obviousness, i.e., whether or not the core inventive concept is obvious to a person skilled in the art. This, as Dr. Abolkheir argues, gives rise to a logical fallacy since the focus is on the imaginative capacity of the inventor, rather than the invention itself. It confuses ‘invention’ with ‘person.’ His argument proceeds thus: for being patentable, (i) the core inventive concept must be beyond the grasp of those with average imaginative powers (person skilled in the art), (ii) the inventor has grasped the core concept of invention, and therefore (iii) the inventor’s powers of imagination are above average.
Illustration of the ‘non-obviousness’ test using India’s case:
Under s. 2(1)(ja) of the Patents Act, 1970, ‘inventive step’ is defined in terms of three main components, i.e., technical advance, and/ or economic significance, and non-obviousness. Judicial dicta has laid down the test for non-obviousness. In Hoffman-La Roche Ltd. v. Cipla Ltd., the Delhi HC held that if the invention is obvious to a skilled worker, in the field concerned, in the state of knowledge concerning at the date of patent to be found in the literature available to him, patent cannot be granted. It has to be seen if the differences between the claimed invention and the prior art are such that the claimed invention would have been obvious.
Now, who is this ‘skilled worker’ or ‘person skilled in the art?’ Recently, in Rhodia Operations v. Assistant Controller of Patents and Designs, the Madras HC elaborated on this expression. It held that this is a hypothetical person created by law, and possesses a skill level which is greater/good than an average person. Because most disciplines/arts require a range of skills or skill set, this person needs to possess the skill set to do the job well. Outstanding imaginative skill, however, is not required.
Coming back to Abolkheir’s claim
To sum up from India’s example, the question to be asked is whether, on using the literature (prior art), the person skilled in the art is able to arrive at the claimed invention. If the answer is yes, the threshold of ‘inventive step’ is not met.
In his piece, Abolkheir points to this very comparison between the imaginative capacities of the inventor and the person skilled in the art. He says that an invention is patentable only if the former’s imaginative powers are above average, i.e., beyond that of a person skilled in the art. In his words, this analysis (focusing on the inventor instead of the invention) falls prey to the inventio ad hominem fallacy. He calls it a variation of the well-known ad hominem fallacy.
What is the Consequence of this Fallacy?
Abolkheir suggests that patent laws have been indirectly testing the inventor’s powers of imagination indirectly for all these years. According to him, patent laws are not dual examination systems to assess both invention and inventor, but are designed for the former only. They end up examining the latter unintentionally. This has given rise to a system of second-hand examination, which according to him is similar to assessing the parents of a student by examining his/her homework. Another, and perhaps bigger, consequence is that this kind of assessment has led to an increased amount of subjectivity in the system. This causes differing assessments of patentability by controllers and judges at different levels. As a result, while getting patents becomes increasingly difficult, granted patents also become vulnerable to post-grant invalidation and revocation.
Strengths of the Piece
Abolkheir is correct in saying that analysis goes beyond the objective assessment of the invention alone while assessing non-obviousness. He therefore suggests limiting it to compare the invention and the relevant prior art. This is also because unlike schools and universities, patent offices have limited expertise to be able to go into cognitive assessments of this kind.
Another strength of his piece is that it highlights the issue with this kind of assessment with the increasing use of AI. Presently, there is a lot of indeterminacy around the non-obviousness analysis. In his article, Gregory Mandel notes the factors producing indeterminacy. One of them is the failure to define the baseline level of ordinary skill against which an invention would be assessed. With the increasing use and huge potential of AI to combine information from completely different sources, the threshold of ‘person skilled in the art’ will inevitably increase in the future. Patent applicants are therefore going to find it very difficult to judge and satisfy this indeterminate threshold. The present non-obviousness test was conceived in a completely different era. If the same is continued to assess applications in the present (and future), it will double the burden for the applicants. Abolkheir’s piece does well to highlight the desirability of a reform.
Evaluating Abolkheir’s Claim: Is there a Logical Fallacy?
A major criticism of Abolkheir’s piece is that when looked at closely, the claim of this analysis being similar to ad hominem fallacy is far-fetched. Under ad hominem, the response (of a personal nature) is irrelevant to the argument made/question asked. But under non-obviousness test, the analysis is still limited and relevant. There is no broad examination of the person’s overall cognitive abilities. What is examined is his/her imaginative skills with respect to the particular invention.
One of the points that Abolkheir seems to have missed is that a number of actors are involved in the entire patenting process. It is not limited to the invention in isolation. Keeping actors aside, the analysis also has to look at novelty, utility, public interest and pre-grant opposition, amongst other aspects. Because of the nature of inquiry being such, different people will feature at different stages, and it might not be practicable to examine the invention solely for these aspects. While advocating a restricted mode of assessment for non-obviousness, he overlooks the fact that the entire process otherwise is inherently multifaceted.
Another argument against this piece is that the inventor and his invention are not so distinct and separable as Abolkheir makes it look like. The latter is an expression of the former’s abilities, and both are not strictly two entities. His example “examining the parents of a student by examining his/her homework” appears to be a misfit to the case at hand because a student and his parents are two distinct entities.
Abolkheir highlights and criticises the subjectivity introduced into the system by this kind of analysis and even hints that an objective mechanism is missing. However, he fails to point out why objectivity is desirable and how it would solve the problem of differing assessments by controllers and judges, in the backdrop of growing use of AI.
Issues to Ponder Over
With the increase in AI usage, the threshold of ‘person skilled in the art’ is going to increase. This piece has underscored the necessity of a different sort of analysis, keeping in mind the changed circumstances. The question arises, what are the alternatives available? How would an objective mechanism be arrived at and how the same would address the challenge posed by AI? A related issue to think over is the possibility and feasibility of adopting different systems of assessment for AI and non-AI related patents.
Among alternatives, Abolkheir talks about the half-causation framework which he devised in 2019. However, it is highly technical in nature and would require specialised training. In addition to a rigorous examination of the invention in question, it entails dividing the entire process of working out the invention into phases. But such strict compartmentalisation might not be feasible for all inventions. If we extend the expertise argument here, it would not be practicable to expect the patent offices or Courts to apply half-causation. While it addresses the logical flaw highlighted, it is hit by lack of expertise.
Conclusion
Even though there are some missing pieces in his argument, Dr. Abolkheir has stirred a discussion on the possibility of reform in patent law. Towards the end, he hinted at coming up with a follow up paper presenting his alternatives to the logical fallacy including the half-causation framework. It would be interesting to see what he comes up with, since he claimed it to offer more objectivity as well as control the AI risk. The overall question still looms, are the foundations of the non-obviousness test as solid as has been assumed till date? Further research is required to answer this.
[Thanks to Swaraj for asking me to look at the paper, and the discussion session attended by Swaraj, Praharsh, Sunidhi Das, Samridhi Chugh and Yukta Chordia for the comments and questions which helped frame and write this post.]