It seems to me there is one basic question we are discussing from different angles, namely, what is the relationship of rules against sharing (RAS) to knowledge production (KP)?

I've presented some ideas on how one hypothesis that is often advanced, that decreasing RAS decreases KP, might be tested in a precise way.

If I understand correctly, you're suggesting that we should first test a different hypothesis, the hypothesis that decreasing RAS increases KP. You write that, on your read of the evidence, this hypothesis is more plausible, so it would be a more natural and fruitful starting place.

I'm agnostic as to which way the initial hypothesis is framed. I'm not familiar enough with the relevant details about how RAS operate in different contexts and how knowledge is produced to take a strong stance at a high level of abstraction.

I think my intuition was to start with testing and establishing the more minimalist claim - "it is not true that decreasing RAS decreases KP" - before building on that to move on to establishing "it is true that decreasing RAS increases KP." But either starting point will work.

Either way, if the goal is to understand how RAS affects KP, specifying as precisely as possible the causal pathways by which RAS is hypothesized to affect KP seems important. Precise subhypotheses about causal pathways make a systematic inquiry easier and enable focused gathering of evidence. Given the complexity of the phenomenon we're interested in, this type of reductionist methodology seems optimal, even essential. But I am open to and interested in suggestions about more effective strategies for addressing our shared question.

There are other, more complex empirical and normative questions you've raised about agenda-setting, distributive justice and unnecessary ignorance. I think those are important as well, even more important. I've just focused here on one initial empirical question.

-- DevinMcDougall - 14 Sep 2011