I've not read BR, but the way you present the two statements I'd have to say that KA's statement is more generally accurate. All nucleated cells, including macrophages, express MHC I. If the cell is invaded by a virus, undergoes malignant transformation, or for some other reason starts producing abnormal proteins, these are broken down into fragments and are attached to MHC I and transported to the cell membrane. Activated effector cytotoxic T-cells with corresponding receptors then can bind to the cell and destroy it by various means. That's the usual way cells bind antigen to MHC I, and it doesn't involve phagocytosis. The phagocytosis of extracellular antigens and their internal processing and cell-surface presentation of these antigens by specialized antigen presenting cells (APC's) mostly involves binding to MHC II. Dendritic cells, however, are known to be able, in vivo, to phagocytize exogenous pathogens and present them attached to MHC I so as to activate cytotoxic T-cells. They do this by a slightly different pathway than the one used by cells to present endogenous pathogenic antigens, in a process called "cross-presentation." Macrophages usually do not do this in vivo but they have been observed to do so in tissue culture experiments. This is just my limited understanding of this stuff, and if I've gotten something wrong I'm sure someone will chime in.