This document is in the public domain, and may be redistributed freely. 

Open Source software is in danger from patents primarily because the Open Source movement is not presently able to broker the kinds of compromises that commercial firms employ among themselves. It is therefore necessary that the Open-Source movement evolve its own defensive strategies.

Virtually all economically significant patents are spurious. Invention simply does not work the way the patent law says it works. There are large numbers of people who play around with ideas far in advance of technical feasibility. They are not motivated by the expectation of commercial profit. To the extent that they have an economic motivation, it rests in leaping into the future, and being seen to be leaping into the future. An example would be Vanevar Bush's description of hypertext in 1949, at a time when the most powerful existing computers had only several kilobytes of memory. That was an instance of a standard technique: look at an existing device, imagine that it was a gazillion times more powerful, never mind how, and in that case, think what could you do with it. Arthur Clarke's _Profiles of the Future_ is virtually a handbook for such thinking. There is a whole learned society of people who pursue this approach, not always with success: the World Future Society. Economically significant patents can always be broken if you put enough effort into searching for prior art, and sufficient effort into bringing this prior art to the attention of a judge. I do not think the MPEG patents, the latest point of controversy, are likely to differ substantially in this regard.

The problem is how to marshal sufficient resources to do the breaking job. The Mauchley computer patent was exposed by a combination of disgruntled computer companies, and one might add, the enraged John Vincent Atanasoff, who had had both his confidence and his hospitality abused by Mauchley. In the case of the MPEG patents, there is the additional problem of a patent pool. Given the true nature of invention and patents, a patent pool can never be a quite legitimate arrangement. A patent pool is in effect a collusive agreement to support each other's spurious patents in the hope of levying royalties from outsiders. The creation of a patent pool is an act of mutual bribery, suppressing the normal economic motivation to expose spurious patents.

The patent office is not going to be much use to us. It periodically promises reform after a particularly egregious case of improper patenting, but the promises inevitably become dead letters. The truth is that the vested interest of the patent office lies in the wholesale creation of spurious patents, and the patent office is continually slipping over into something approaching official corruption. This is not a new trend. Bruce Sinclair, in his _Philadelphia's Philosopher Mechanics: A History of the Franklin Institute, 1824-1865_ (1974), has documented the process in the early nineteenth century. As he points out, the Franklin Institute, the de facto premier engineering society of the time, had to mobilize to fight the patent office to make the patent office do its constitutional duty.

We need something like an Open Source Movement for patents. That is, we need a dispersed, acephalous movement of tens of thousands of people who would systematically collect, collate, and forward evidence of prior art. Exposing spurious patents will become easier in a few years, of course, as the Web expands and search engines improve. For the time being, we have to deal with certain important lacunas, most importantly, the technical literature of the 1960's. This was the period after people started to think about computers in a serious way, but before machine-indexing got going in a substantial way. It is the period when one is most likely to find patent-breaking prior art. We need large numbers of people to go and read 1960's technical literature, write up descriptions, and post them to public databases. Apart from technical literature per se, it is also important to read cognate literatures, such as mathematics, philosophy, linguistics, and even science fiction.

Ideally, of course, there would also be some kind of analog to the Gutenberg Project, only adapted to materials which are still under copyright. Such an organization might agree to mobilize large numbers of transcribers, provided the publisher agreed to certain access requirements, e. g. Web-posting, with copyright notice, but with no log-in requirements, and no obstruction of document linking, search engine use, or Web-archiving, by either legal or technical means, and no fee collection for any of these purposes. In effect, we can create our own code of Fair Use, analogous to the General Public Licence. If one could get a few publishers to participate on these terms, it would create pressure for other publishers to Web-post more of their own old materials.

Once we have enough of a movement in place, it will be actively imprudent for a patent holder to behave in a conspicuously blatant manner, because that would run the risk of angering a few thousand amateur prior-art searchers, and causing them to concentrate their efforts on all the patents held by that patent holder. Let us suppose, purely hypothetically, that there is a certain particularly obnoxious company. Let us suppose further that it suddenly emerges that eighty or ninety percent of this company's hundreds or thousands of patents are spurious, and that this fact is publicized in a major newspaper. The Department of Justice would have no real alternative, but to open an anti- trust investigation. The nature of anti-trust law is such that a giant corporation, simply by virtue of being a giant corporation, is almost axiomatically in breach of the law. Being investigated would result in the price of the stock falling, and since senior officers are compensated with stock options, and sometimes have to borrow money to exercise them, this would result in a financial catastrophe for the key officials.

The historian Craig Muldrew (_The Economy of Obligation: The Culture of Credit and Social Relations in Early Modern England_, 1998) has made an interesting argument for individuals in sixteenth-century England, which applies to modern corporations. Individuals were able to continue indefinitely in a state of debt which ought to have been legally untenable, provided they behaved as responsible members of the community. As long as they did so, they would be protected, but if someone stepped outside of the role of responsible community member, he would immediately be sued for his total outstanding debts, and hauled off to debtor's prison. The same reasoning can be applied to modern corporations. In Muldrew-ian terms, Microsoft got chewed up, not because of any strictly legal point, but simply because Bill Gates is an arrogant twit who chose to surround himself with other arrogant twits.

Andrew D. Todd
1249 Pineview Dr., Apt 4
Morgantown, WV 26505

adtodd@mail.wvnet.edu (formerly  U46A8@WVNVM.WVNET.EDU)