Most evidence tends to be pretty clear even in the field of anthropology. You can publish a speculative theory, but more often a paper is going to say “we found these bones with these markers here, {type of dating} indicates age of X with a margin of error of Y”; “technology X was found at Y at a depth of Z, this matches/does not match technology A in aspects B and C, but not P, Q, etc.” Ethnologies are a bit more suspect, but you can check who visited when and observed what, and see if the observations are consistent. And as you might imagine, genetic studies tend to be fairly clear cut.
When you make a more general statement about ‘human nature’, you start to move into frequency counts of observed societies, which mean that your sampling frame is very limited, and much more likely to give you exceptions than rules. Much of what you see in informal writing is broadly extrapolated from comparison with animals and broad assumptions both about the environment and about humans (assumed lack of) ability to adapt without genetic change.
As a shorthand, as in most fields, if a claim is made and a peer reviewed paper is not cited, assume that this is not the proper source for this information.
Seems that this is key. The question is what kind of sampling is broad enough to support what kind of assertion. I’m not sure if that can always be neatly determined, so you might have two sets of claims on a continuum between well-supported and totally speculative with a muddy stretch in the middle.
Yes—and if authors gave an indication of what sort of evidence they were looking at, it would not be a fallacy. It is fine to report that ‘5/5 of the X that we looked at are Y’, but the claim that ‘X are Y’ is not so fine. Most educated people (for example, science writers) seem to understand this for most cases, but drop their critical thinking when it comes to humans…
Do you know of a clean criterion for deciding when you’re dealing with mere conjecture and when you’re dealing with evidence?
Most evidence tends to be pretty clear even in the field of anthropology. You can publish a speculative theory, but more often a paper is going to say “we found these bones with these markers here, {type of dating} indicates age of X with a margin of error of Y”; “technology X was found at Y at a depth of Z, this matches/does not match technology A in aspects B and C, but not P, Q, etc.” Ethnologies are a bit more suspect, but you can check who visited when and observed what, and see if the observations are consistent. And as you might imagine, genetic studies tend to be fairly clear cut.
When you make a more general statement about ‘human nature’, you start to move into frequency counts of observed societies, which mean that your sampling frame is very limited, and much more likely to give you exceptions than rules. Much of what you see in informal writing is broadly extrapolated from comparison with animals and broad assumptions both about the environment and about humans (assumed lack of) ability to adapt without genetic change.
As a shorthand, as in most fields, if a claim is made and a peer reviewed paper is not cited, assume that this is not the proper source for this information.
Seems that this is key. The question is what kind of sampling is broad enough to support what kind of assertion. I’m not sure if that can always be neatly determined, so you might have two sets of claims on a continuum between well-supported and totally speculative with a muddy stretch in the middle.
Yes—and if authors gave an indication of what sort of evidence they were looking at, it would not be a fallacy. It is fine to report that ‘5/5 of the X that we looked at are Y’, but the claim that ‘X are Y’ is not so fine. Most educated people (for example, science writers) seem to understand this for most cases, but drop their critical thinking when it comes to humans…