Intellectual Property and Open Source

Chapter 2. The Patent Document

Imagine you are a programmer learning a new computer language. When you are given a program in the new language, the syntax is usually obscure even if the overall constructs are familiar. Repeated exposure and study may alert you to reserved words and give you an idea of their meaning, but fully understanding the program requires you to know the syntax and semantics of the language as well as the problem domain addressed by the code.

Patents are the type of intellectual property that most closely resemble code in this context. A well-written patent document is highly structured, with required sections, definitions, reserved words, and “program flow” constructs.

As a result, patent documents tend to be very boring, somewhat ungrammatical, and only semi-intelligible to an ordinary competent English speaker. Even when you understand the problem domain addressed by a particular patent (i.e., the area of technology described within the patent) you do not fully understand the patent until you also have a handle on the specificities of the patent language.

In fact, patents are specifically like pattern-matching code such as regular expressions. Instead of matching text, however, patents match technology. As anyone who has used regular expressions can tell you, though, very complex regular expressions don’t always match what you think they should when you first run them. Patents are similar; in truth, nobody (not even patent lawyers) knows exactly what a patent ...