Regarding this problem of simplification by Stigler and later textbooks 
mentioned by several folks,
I think that this is a problem inherent in the nature of property rights and 
transactions costs.  I think
that Coase was less clear about this in his 1937 paper, where he laid out 
the initial idea of a firm
choosing the scope of its property control in order to minimize transactions 
costs, but is more evident
in his more complicated 1960 paper.  It also becomes much clearer in the 
work of Coase's follower,
Oliver Williamson, now the most cited economist of all time, according to 
Geoffrey Hodgson, whose
oeuvre has been overwhelmingly dedicated to analyzing the many 
sub-categories and varieties of
transactions costs and how they interact with property rights.  In his 
awareness of these complications
he long cited the old institutionalist, John R. Commons, even as some new 
institutionalists at Chicago
derided Commons for his reputed "socialist" tendencies.

The problem is that neither "transactions costs" nor "property rights" are 
nearly as clear cut as the
textbooks would have it.  Both involve many sub-categories and contingencies 
and varieties.  The
term transactions costs was used so broadly that it came to mean almost 
anything that was not a
direct production cost, thereby nearly emptying it of real meaning.  This is 
what has fed Williamson's
long research agenda, to straighten out this mess.  As for property rights, 
they are not absolute.  They
have many aspects: rights to receive income from, rights to alter, rights to 
change the use of, rights to
resell, and on and on.  Many of these are limited in arbitrary ways by laws 
and regulations, think of
zoning and real estate.  It is the complications of these concepts and how 
they interact that becomes
much more the story and modifies it considerably in real practice.  Again, I 
think that both Coase and
Williamson are aware of these complications, even as they disappear in many 
textbook presentations.

Barkley Rosser