Risch on America's First Patents

America's First Patents had been posted on SSRN by Michael Risch, Villanova University School of Law.  It is forthcoming in the Florida Law Review, Vol. 64, 2012.  Here's the abstract:

Courts and commentators vigorously debate early American patent history because of a spotty documentary record. To fill these gaps, scholars have examined the adoption of the Intellectual Property Clause of the Constitution, correspondence, dictionaries, and British and colonial case law. But there is one largely ignored body of information – the content of early patents themselves. While many debate what the founders thought, no one asks what early inventors thought - and those thoughts are telling. This article is the first comprehensive examination of how early inventors and their patents should inform our current thoughts about the patent system.

To better understand our early patent history, we read every available patent issued prior to the institution of the “modern” examination system in 1836, totaling nearly 2,500 handwritten patents. For good measure, we also read the first 1,200 patents issued after 1836, the last of which issued in the middle of 1839.

Part I discusses how vague and ambiguous patents are relevant to early judicial discussion of "principles." In conjunction with misplaced reliance on English law, the patents suggest a different interpretation of “principles” in these cases. In short, patentable subject matter jurisprudence developed in a way that was not necessarily intended by the first Congress.

Part II discusses some noteworthy patents, including asbestos and lead paint, milk of magnesia, many business methods, and a programmable loom that predated Babbage's Analytical Engine. This might lead us to reconsider how we view technological change in the patent system.

Part III presents a surprising rebuttal to those who believe that the machine-or-transformation test is engrained in American inventive ethos. This test requires that, to be patentable subject matter, a claimed process must be performed by a machine or transform matter to a different state. Though the Federal Circuit formally introduced this test in 2008, courts and scholars present it as a “historical” limitation on patentable subject matter. Examination of the first fifty years of patents shows that forty percent of patented processes would have failed the machine-or-transformation test, whether or not the patents were tested by the Patent Office. Many method patents did not involve a machine and did not transform matter to a different state or thing.

The article concludes with some suggestions about how we might rethink patentable subject matter in light of America's first patents.