Intangible machines: patent protection for software in the United States

Intellectual property law has been interacting with software for over sixty years. Despite this, the law in this area remains confused and uncertain: this is particularly evident in patent law. Focusing on U.S. patent law from the 1960s through to the mid-1970s, this article argues that a key reason for this confusion relates to the particular way that the subject matter was construed. While the early discussions about subject matter eligibility were framed in terms of the question "is software patentable?", what was really at stake in these debates was the preliminary ontological question: what is software? Building on work that highlights the competing ways that software was construed by different parts of the information technology industry at the time, the article looks at the particular way that the law responded to these competing interpretations and how in so doing it laid the foundation for the confusion that characterizes the area. When engaging with new types of subject matter, patent law has consistently relied on the relevant techno-scientific communities not only to provide the law with a relatively clear understanding of the nature of the subject matter being considered; they have also provided the means to allow the law to describe, demarcate, and identify that new subject matter. The inherently divided nature of the nascent information technology industry meant that this was not possible. As a result, the law was forced to develop its own way of dealing with the would-be subject matter