Renaissance Quarterly, Vol. 67, No. 4 (Winter 2014), pp. 1337-1340
http://www.jstor.org/stable/10.1086/679809 (toll access)

Excerpts:

"Most revealing is Bredekamp’s contribution: his refusal to recognize his own methodological failure is not only saddening, but also counterproductive. To put it bluntly, as has been done in the German press: if one of the world’s foremost art historians is incapable of seeing the difference between a 1610 drawing by Galileo and a 2005 forgery, what does that say about art history (or art historians)? [...]

Are we, then, at a methodological impasse, where forgeries are unidentifiable?

The answer is a definite no. As the volume shows, what matters is the right approach. Nicholas Pickwoad, a new addition to the group, shows in his illuminating and characteristically brilliant chapter on the book’s structure that this analysis alone would have been sufficient to cast deep doubt on SNML’s authenticity. We now have a clear methodological directive: analyze bindings first, as this is where mistakes are most visible to the trained eye. It’s a short-term solution, though, that will disappear once the knowledge gap between binders and Pickwoad diminishes."

See also
http://archiv.twoday.net/search?q=galileo