A note on inferences regarding network subgroups

The evidence presented in Bernard et al. (1980) does not warrant their conclusion. We comment on their lack of concern with the substance of social differentiation in the systems they use as evidence. The substance of social structure is argued to affect the scope, adequacy and power of conclusions drawn from a network analysis and is accordingly a factor to be explicitly considered in justifying those conclusions. With these ideas in mind, our reanaly sis of one system on which good data are available, ostensibly supporting the Bernard et al. conclusion, illustrates how their conclusion regarding network subgroups in that system is unwarranted