How does a historian or an amateur determine what legitimate methods can be used and what legitimate sources are? Obviously there are many methods to read texts and speculate about political, social, cultural and individual past experiences; but how are legitimate and illegitimate methods marked? Some sources are obviously very good and very useful, others very bad but very useful, yet others are very good or very bad but not at all useful. How is this differentiation made in historiography?
When assessing sources, one of the warning signs to watch for is a tight loop of self-references amongst a small group of authors, often even a single author, with no citations into the group from outside it. External citations from within the group will be made to mimic the sheen of respectability. This often signals an extreme viewpoint well outside the mainstream, because it is very rare for a single researcher or small group to be well advanced of the mainstream.
Of course, not all such small tight groups are kooks. Continental Drift was ridiculed for two generations before becoming a consensus and accumulating hard geologic evidence. The terms Black Hole and Big Bang were first coined as terms of ridicule because the constructs seemed so unlikely, but are now mainstream physics. However one must be particularly wary in using such sources.