As such, there are some things that we can say with some certainty are very accurate, some that are on more shaky ground, others that are even less likely to be historical, and some that are cheerfully and self-consciously modern inventions.
Thus, things break down in terms of historicity something like this:
- We know this for a fact. First-hand accounts by unbiased observers, many names of deities, major myths that are represented in multiple locations and manners, etc. Into this category would go things like the myth of Thor fishing for the Midgard serpent, which is found in written form, in carvings, etc. Also the basics of the Roman Lupercalia ritual, which were described in detail by the Roman author Plutarch, who had seen them first-hand.
- We strongly suspect this is so. Things that are known from archaeology without any direct contemporary textual support, thus requiring interpretation, written accounts from contemporary, but flawed or biased, observers, etc. This includes things like the dates and broad themes of holidays, more detailed aspects of deities, carvings and tapestries with recognizable figures doing unrecognizable things, recurring symbols that might mean something due to context, etc.
- We're not sure, but this fits some of the evidence. Here we really start moving into the realm of speculation. A lot of material derived from later folklore falls into this category, which can be compelling in the aggregate, but for which the connections are still not certain.
- This is a stretch, but could fit some of the evidence if you squint. Things that take a piece of contested or ambiguous evidence and give it significance with no real certain backing. I would put the practice of runic divination into this category, which is based on an interpretation of a single ambiguous word ("notae") in Tacitus and assumes he meant runes, without any other mention of a connection between runes and divination in any other source.
- We made this part up. This includes almost all of the actual text of modern Asatru rituals (there are a handful of exceptions, and even then they mostly involve re-Heathenizations of songs and charms that are believed to have been Christianized). It should be noted that there is nothing at all wrong with doing so, as long as one acknowledges that that is what one is doing. The trouble starts when invention is passed off as genuine ancient practice.
Of course, for those who do not follow a reconstructionist faith, such as Wiccans, these considerations are pretty well irrelevant. But there are quite a number of us who follow a more reconstructionist faith, such as Asatru, the Religio Romana, Hellenismos, Celtic Reconstruction, and more, and for us the question of the historicity of some part of our practice can be very important. For one thing, it allows us to "triage" things; when presented with contradictory ideas or practices, for instance, I will always defer to something that's higher, and more historical, on the scale.