When a gap was found in the 1993 attempt by
Andrew Wiles to prove the semistable case of the
Taniyama-Shimura conjecture for elliptic curves, Wiles asked Taylor to help him clarify what was required to complete the proof. Wiles eventually realized that work of
Ehud de Shalit could be generalized to bypass the lacuna of 1993. Wiles and Taylor proved that a constructed
Gorenstein ring that arose from this approach was also a
complete intersection. This ring theoretic result essentially completed the proof of the semistable case of Taniyama-Shimura, which Wiles expounded in the same issue of the
Annals of Mathematics. This proof strategy has been dubbed "Taylor-Wiles patching". In subsequent work, Taylor (along with
Michael Harris) proved the
local Langlands conjectures for
GL(n) over a
number field. A simpler proof was suggested almost at the same time by
Guy Henniart, and ten years later by
Peter Scholze. Taylor, together with
Christophe Breuil,
Brian Conrad and
Fred Diamond, completed the proof of the
Taniyama–Shimura conjecture, by performing quite heavy technical computations in the case of additive reduction. In 2008, Taylor, following the ideas of Michael Harris and building on his joint work with
Laurent Clozel, Michael Harris, and
Nick Shepherd-Barron, announced a proof of the
Sato–Tate conjecture, for
elliptic curves with non-integral
j-invariant. This partial proof of the Sato–Tate conjecture uses Wiles's theorem about modularity of semistable elliptic curves. ==Awards and honors==