The official story is that copyright encourages creators by giving them temporary monopoly rights. The unofficial story is that copyright prevents the diffusion of art and knowledge, and nowhere is it as frustrating as with academic publishing. Commercial publishers sell the research others paid for, and can extract substantial rents because researchers have to publish in established outlets for reputation, tenure and promotion.
Giovanni Ramello remarks that there is another unfortunate consequence of copyright in academic publishing: having been granted some market power, the monopolist will seek to extend this market power through acquisitions and thereby obtain even more dominance. The obvious example is Elsevier, which has reached now a market share that should trigger anti-trust investigations along with profit margin in the order of 30%. The situation is quite bad in Economics, as scholarly societies have done little to prevent Elsevier taking hold of the major field journals, thereby making it essential to any tenure file. And given this, research libraries have no choice but subscribe to those journals, falling in the trap of the monopolist.
In other sciences, I hear the situation is not much better. And I have reported previously about horror stories that still seem to have little impact (1, 2).
In any case, journals are dead to me, for reasons cited above and also because the publishing process is broken, starting with refereeing.