Brandon ([1982] 1984, 1990) has argued that Salmon's (1971) concept of screening-off can be used to characterize (i) the idea that natural selection acts directly on an organism's phenotype, only indirectly on its genotype, and (ii) the biological problem of the levels of selection. Brandon also suggests (iii) that screening-off events in a causal chain are better explanations than the events they screen off. This paper critically evaluates Brandon's proposals.

Recently, Fred Gifford attempted to explicate the meaning of the term genetic as applied to phenotypic traits. He takes as his primary goal the explication of how the term is used and tries to avoid conclusions about how it should be used. He proposes two independent criteria (DF and PI) which together capture much of what biologists mean when they describe traits as genetic. Although Gifford's approach is extremely insightful in many ways, I argue that his analysis is not sufficiently (...) critical concerning the adequacy of common usage.In particular, while DF is a perfectly legitimate and useful measure of heritability in populations, it is not necessarily a genetic one and should not be labeled as such. PI on the other hand, although very intuitive, depends on an extremely problematic distinction between causes and mere conditions (e.g., genes and epigenetic factors). Both criteria will be highly relative and both, via what I term the new problem of genetics, will inspire contradictory analyses based on the same data. (shrink)