Gigi Sohn proposes six points to reform copyright in a speech yesterday link here. The complete text is available here. Teeing off from the fact that the scope and duration of copyright has been steadily expanded, limiting innovation, she suggests fair use changes, limits on secondary liability, protection against copyright abuse, better licensing, orphan works reform, and better notification of technological and contractual restrictions on digital media. Most of our readers would prefer abolition of copyright but it isn't going to happen any time soon, so these ideas are worth a lot of discussion.

As a starting list. I think her ideas a pretty good.

I would also emphasize limiting the scope of rights to "derivative works" which have expanded so broadly in such a way as to threaten the very idea/expression dichotomy that copyright law suppoedly adheres to. In other words people should be free to create derivate works based on an original work as long as they do not incorporate an actual 'clone' of the original work into the newer/derivative work.

(Off-topic)

There is something I don't understand about fixed costs vs sunk costs in the creation of intellectual work.

I read somewhere, I believe in an essay or paper on dklevine.com, that the costs involved in creation of intellectual work are sunk rather than fixed.

In my understanding, sunk costs are costs that have already been made and should not be involved in decision making, and fixed costs are costs still open to decision, but already known.

Are costs involved in the creation of intellectual work really sunk? They may not be fixed; one cannot in advance be sure of the amount of time and effort you need to complete your work, but your estimate of the cost and the estimate of your revenue influence your decision to start the project.

So, don't these costs involve themselves in decision making? An argument often made for intellectual monopoly is that it increases the incentive to produce ideas, because it increases expected revenue. Without the extra incentive, producers of ideas will not produce many socially desirable ideas because they can recover only a fraction of the social value of the idea.

I'm not saying this argument is correct, but I thought the assertion that these costs were sunk intended to take the force from this argument.

Obviously I have missed something, somewhere. What is it?

It took me a while to find it, it seems to be here. In any case - after the fact the cost is sunk; it may or may not be a fixed cost before the fact (some sunk costs are variable costs before they are actually incurred). And of course before the fact costs may be uncertain. And there is no doubt that by whatever name before the fact the costs that must be faced whether certain or uncertain affect the decision to innovate. The point is that "The cost of coming up with (at least) one sample of the idea may be a lot smaller than the quasi-rents the same producer will collect when his creation is priced competitively." In other words, after the fact the producer will earn a return on her investment - regardless of whether she is promised a monopoly. Some of this return may come from first mover advantage, for example with patentable ideas. Some of this return may come from sale of complementary goods such as services, as in the open source software industry.

Before the fact the innovator will weigh the cost of innovating and the degree of uncertainty against the possible return. If ideas were perfectly divisible and there were no spillovers the competitive system would deliver exactly the right amount of innovation. Spillovers in fact seem to be small in practice, but ideas are not perfectly divisible, so - regardless of whether monopolies are granted or not - some socially beneficial ideas are not be produced. In general, a system that grants monopolies impacts the decision to innovate in two ways: it raises the return and it raises the cost. The evidence suggests that these effects tend to cancel out so that overall monopoly awards have little impact on innovation - that is, the system doesn't work.

Thank you very much.

There appears to be at least some divisibility in ideas, at least in the Free Software world.

A large number of people make very small improvements to an existing idea for very little return (ego-boost, recognition?). Each might produce 1/1000th of the new idea, but because they are working on the same idea, they can produce a fully working idea over time.

Software can be improved without limit, and such a community seems to continue for as long as there is interest in the work produced. The only prerequisite for such a community seems to be the recognition of some common goal. Moreover, since both effort and return is small, non-monetary motivations dominate.

It might be a little harder to produce works of literature or music in this manner, but those might also be more divisible than commonly believed. Who would have expected the degree of divisibility demonstrated in community ("bazaar") Free Software development?

Fan fiction shows signs of divisibility. A quite long series of Star Wars stories, the "New Jedi Order" ones, exists that was written by quite a diverse collection of authors. Fair numbers of single books are written collaboratively by two (and sometimes more) authors. Nearly all have some effective collaboration between the author(s) and at least one editor, regardless.

That's just books. Movies of course are made by huge numbers of individuals, e.g. actors, stage hands, SFX wizards, lighting technicians, screenwriters (sometimes plural), etc.; if all of those people were offered a percentage instead of paid up-front some of the high initial cost of filmmaking would disappear. A movie could also be made a scene at a time as money became available, or similarly. Movies look fairly divisible to me.

Music is highly divisible down to the unit of a single song, generally not more than five minutes long (with occasional exceptions). Even that single song has several subtasks: compose lyrics, record vocals, compose music, record instrument A, record instrument B, etc., and mix all the recordings. This divides the job into several tasks, some for musicians that play various instruments, at least one for a technician, one for a singer, one for a songwriter, and one for a composer.

In the case of music and movies, giving the individual job doers significant creative latitude divides the creative part of the job as well. At least one recent major motion picture, The Core, had a fair amount of ad-libbed dialog instead of every last word and nuance being dictated by the scriptwriter; the "creativity distribution" among the scriptwriter and the actors was thus balanced more than typical. (Now if they'd also had better science consultants...)

A book written by two author may prove that two half-baked ideas in some cases equal one. However, in most of those cases, I assume the full product gets assembled, sold, and the authors divide amongst themselves the profits.

The point I was trying to make is that in the Free Software world, each author produces and is compensated for one minuscule individual contribution, which by itself is leagues from "fully functional." In this case, the compensation is not for the "fully functional" product, but for the contribution itself.

(Because the effort is so small, it can be compensated by non-monetary means)

Actually I probably just don't understand what divisibility means.

If the optimal amount of cookies is 250/week, but we can only buy factory improvements in increments of 100 extra cookies/week, then we will be stuck at 200 instead of 250. Because factory improvements are quantized, we cannot have the optimal output. So far so good?

However, it is hard for me to go from factories to ideas here. I can understand how an unfinished idea can be valuable (because of the time it saves you when trying to make the fully-functional idea, part of the work is already done) and therefore socially desirable.

But, how can an idea that is impossible to create be socially desirable? Or, is the idea possible to create, but is there no incentive to create it because.... ? Because what? Because the producer cannot recoup all of the costs of production?

In that case, dividing the work among many people each of which makes a contribution small enough to be sufficiently compensated by "thanks," might be the solution.