Book Review: Math You Can’t Use

MathMath You Can’t Use
by Ben Klemens

$19 from Amazon

Klemens is a scholar, but doesn’t really write like one — rather he tells a story and wraps the story around his argument.  His bottom line argument is that software should not be patentable. Here are some of his arguments:

1) Software is simply math (Church-Turing thesis) and math is not patentable.  His conclusion — that software should therefore be unpatentable — strikes me as odd.  Klemens leads us down the parade of horribles of granting 15 years of exclusivity to mathematical algorithms.  The bad results are (a) mathematicians being forced to pay royalties when a patent is valuable and (b) the Church-Turing thesis itself being patented.

2) The PTO is bad at granting patents and therefore should not grant software patents.

3) Patents only work well in industries with a few major players.  The software industry is decentralized (with lots of small players) and therefore should not be covered by patents. . .

Of course, Klemens may gain more traction with his explanation of how aggressive patent enforcement could kill the internet as a whole.


8 thoughts on “Book Review: Math You Can’t Use

  1. 7

    I am a software engineer. I want to affirm the position and amplify the remarks made by Lion Kimbro.

    First, I take it as a fact that virtually everything connected to the internet would be considered patentable by today’s standards. For example, the browser was clearly a patentable invention, as were the routing protocols, and the even the basic ability to pass ‘related multi-media elements back and forth between computers’. On the desktop, the Explorer tree of files and folders was clearly a patentable invention, as was virtualy every other UI widget.

    If the possible degree of patenting had been realized, it would have made it impossible for individual programmers and start-ups to function; the licensing fees and lawyering would have stopped them. ATT-“type” corporations would have a stranglehold on the internet and whatever business advantages it conferred, the larger playing field of computing generally would have been constrained and we would be hearing about how expensive it is to maintain the internet and why it must remain the sole province of mega-corporations.

    Just because every novel and important advance, not to mention every twit and twiddle, (“one click”) was not patented, the internet and the huge expansion of wealth and productivity, business models, tax revenue and worldwide employment was made possible.

    The argument against software patents is they should not be granted since they neither spur nor incentivize innovation, but rather constrict it. Patent rights not being natural rights, Congress therefore has a duty to reject them for the practical reason that they work exactly against the purpose of permitting the patenting of creative work in the first place.

    The ability to patent is a tool we wield to benefit society, not an Ayn Randian inalienable right.

    This is obvious to everyone who is not a patent lawyer or directly or indirectly benefiting from the activity around software patents.

    Within the software community, software patents are seen as yet another example of the cyncial twisting of our law by special interest to benefit the well-heeled few at the expense of society’s greater good. For example, Bezeo’s “one click” is sneered at and derided without exception, and the system that permitted it has only bred further disrespect for the law.

    Congress and the USPTO have a basic choice to make here. Either they can do what is right, and obviously so, and start to reverse the perception of themselves and the law as nothing more than unprincipled lackeys for the powers of the day or they can continue down the road they’re on, breeding contempt for Congress and the entire lawmaking process itself.

    Sooner or later, if truly innovative software developers come to believe that they will be hounded into poverty by this sad misapplication of monopoly power they will do one of two things- they will desist in trying to write software or they will emmigrate. Since both of those outcomes are detrimental to the US, lawmakers may want to sober up long enough to ban software patents as the EU has had sense to do.

  2. 6

    I’m against software patents.

    I agree; The argument “software is algorithms is math” is crude. I may as well argue that my loom is math as well, or any other invention. That is a lame argument.

    Good arguments.

    The space of inventors is ENORMOUS. There is no lack of innovation. In fact, it’s really hard to say what’s an invention, and what’s not. Programmers just regularly, habitually, systematically, “invent.” We hardly even call it invention, we just call it programming. The flow of invention is so regular, so continuous, so unmeasurable and unquantifiable.

    Now, I could see a case for software patents, if it were not this way. If it were like some arcane branch of biology, and there were only, say, 1,000 people qualified to perform this invention, to solve these technical problems, then yes: Patents would make sense there.

    You do remember that patents have a purpose, right? They’re not “natural rights?” They are government granted monopolies, put in place to make sure that science and the arts are moving ahead? Do remember that; It seems like a lot of times, your community forgets that.

    So, one argument against software patents, is that the pool of inventors is so vast, the likelihood of independent invention is likewise increasing, and the likelihood that you’re actually speeding up the development of science and the arts is going down down down WAAAY down.

    I hope you know that- companies are patenting stuff, just to knock out competitors. And this isn’t because, “Oh no, we’re going to go broke, if other people can make software that does something ours can do.” This is because: “Oh, we want to spectacularly insanely wealthy, rather than just really rich.” Because companies can definitely survive by making good software, good user interfaces, etc., etc.,. If you’re the first to make a good user interface, and a competitor starts to knock off it after a couple years, most people are still going to use the first companies software, though the price will be pulled down (as is right and proper.) Innovation will still happen.

    In fact, I think that the IP lawyers in general have a LOT of answering to do, about the whole “promoting the sciences & the arts” thing.

    For example, let’s take a different field: Cartoon characters.

    In Japan, if someone makes a comic, it’s okay for you to redraw the characters, and make a story about those characters doing different things. This has been going on for A LONG TIME, and if there’s been any damage, whatsoever, to the Japanese arts, there’s no sign of it. Anime is BOOMING, it’s doing incredibly well, and I don’t see a clear case that the ability of people to make their own comics using someone else’s invented characters is actually harming a single thing.

    So, see, this is the kind of thing where: We don’t trust you. You aren’t looking out for our interests. You aren’t looking out for our rights. You don’t believe in the constitution, you ignore the whole “for the promotion of the science and the arts” thing, and are instead just saying: “This is property. This is mine.”

    Argh! We are so angry with you!

    We need, seriously, a bunch of software developers, a bunch of IP lawyers, some constitutional scholars, people working in the patent office, maybe a congress person or two, all these different guys- we need a big room, a bunch of stairs in circles, or a bunch of coffee tables with note paper, and we just need to all TALK, for about 5 days straight, with all the different sides.

    We need to say what our needs our, what our views are, and we need to figure out how everybody’s needs are going to get met.

    Because right now, we’re feeling pretty robbed, us software developers. We’re seeing you guys patenting things that are plainly obvious, terribly stupid, and we see you dragging back the industry. We see the big boys, with the big bucks, playing patent battles with one another, (“you little guys don’t matter anything, you see,”) and working to lock us out, and the whole thing is rediculous. It’s lawlessness, that’s what it is. It’s a lawlessness, wild wild west style, that happens to live on top of a layer of law.

    And this just can’t go on. We can’t call ourselves civilized, and watch this going on.

    This is how I feel, take it for what you will.

  3. 5

    I’ve been looking forward to this text since I saw the publication notice on Amazon. Unfortunately, statements like this one completely eliminate my interest level:

    “Here are some of his arguments: 1) Software is simply math (Church-Turing thesis) and math is not patentable.”

    Anyone who so fundamentally misunderstands software is not well-positioned to write a treatise on its patentability. _Some_ software is purely mathematical – and is not patentable, even today, even after State Street. _Other_ software is an applied solution to a real-world problem, and is patentable _in that context._ So sayeth the CAFC in Diamond v. Diehr. Perhaps Mr. Klemens should have read that case before adopting an outdated and untenable generalization of software as the cornerstone of his treatise. It’s really a shame.

    Perhaps the worst result of this ineptitude is that the anti-software-patent movement will undoubtedly cite this text as some kind of authority on the subject – thereby further propagating the out-of-touch position of that group. As long as it continues to cite such balderdash, the credibility of this group will remain at zero. It’s too bad that this crowd is too busy making vacuous arguments to adopt the myriad stronger arguments against software patents.


    David Stein, M.S., J.D.

  4. 4

    The fact that software innovations are incremental is largely a non-arguement. Almost every innovation is incremental. It’s built on what went before it. There are lots of “silly” patents in almost all arenas–not just software.

    It seems to me that many of the problems w/respect to software patents boil down to 2 issues:
    1) There’s prior art for many issued software patents. Thus, these shouldn’t have been issued in the first place.
    2) Inventions are not appropriately valued by the court system. Most software programs have thousands of innovations in them. Yet, many software patents ends up being valued at several orders of magnitude over their actual inventive contributions. If the “real” value of most issued software patents were recognized, most of these issues would disappear, as we’d be arguing about licenses of $10 K, rather than $10 M.

    Steve Guttman

  5. 3

    Alex, software can be patented. It’s not a question of whether it could be. Apparently you are saying that the economy has done well even though software can be patented.

    Gregory, you can patent software in Europe. It’s business methods that can only be patented in the US. Your incremental argument is interesting, but would apply to _all_ inventions, not just software.

    Alun Palmer, US Patent Agent

  6. 2

    I’m not sure why so many patent attorneys believe that the idea of software not being patentable is odd. Given that the United States is the only country which currently allows them and given that much of Europe doesn’t agree. Perhaps it’s because they generally get rich off of bad patents, such as many of those on software.

    I have put into the URL section of this posting a petition containing roughly 13,000+ names of people who also believe that patenting software is a bad idea. So it’s not so uncommon after all.

    The reason that software shouldn’t be patentable is quite simple. It’s incremental nature, by definition, means that it’s not possible to build the next innovation without first using one which already exists. This makes the potential costs of licensing all of the technology which might be patented in a particular piece of software prohibitively expensive and the end result is that customers would be asked to foot the bill.

    So, as you can see, in the case of software, patents might actually stifle innovation as opposed to spur it.

    Later, GJC

  7. 1

    Interesting arguments against software patents and I am in favour of keeping software as free as possible of constraints like patenting. The benefits of network externalities, productivity improvements and speed of innovation derived by the US economy over the past 15-20 years could be lost if software could be patented. I believe that the cost of the patenting system exceeds the benefits of patenting when it comes to software.

Comments are closed.