Uncertainties - discrepancies in immunology

The many immunological observations and results from in-vitro or in-vivo experiments vary, and their interpretations differ enormously. A major problem is that within a normal distribution of biological phenomena, which are measurable with many methods, virtually anything is possible. Within a coevolutionary context, the definition of biologically relevant thresholds is an important key to improve our understanding of weaknesses and strengths of the immune system. This review is a personal view, comparing textbook rules and experiments using model antigens with observations on immunity against infections or tumors to critically evaluate our perception and understanding of specificity, affinity maturation, antigen presentation, selection of the class of the immune response, immunological memory and protective immunity, positive selection of T cells and self/nonself discrimination.