Understanding why massive early-type galaxies are red and dead has proved to be difficult. This has been the source of tension with hierarchical formation models, such as the Cold Dark Matter model, in which massive objects are formed from mergers of smaller ones which existed at early times. The problem is to arrange for star formation to occur at higher redshift than the actual assembly of the stars into a single massive galaxy. The most recent galaxy formation models arrange for this to happen by a combination of two processes : dry mergers and AGN feedback.
I will discuss a number of results on massive galaxy formation that are based on analyses of Brightest Cluster Galaxies and other extreme objects in the SDSS and other recent astrophysical datasets. These suggest that at z < 2, in addition to being dry, the mergers should have been minor.