The argument from ignorance, or argumentum ad ignorantiam, infers that a proposition is true from the fact that it is not proven to be false (or alternatively, that a proposition is false because it is not proven to be true).
The old argument that “the absence of evidence is not evidence of absence” is a form of this logical fallacy, because the absence of evidence can be evidence of absence if substantial attempts to find evidence have proven negative.
This fallacy also asserts that a proposition is true because it has not yet been proven false, or it is “generally accepted” (or vice versa).
Furthermore, this argument presumes that there are only two choices — true or false. In fact, there are other choices, including “not enough investigation has been completed to choose between true or false.” So a non-fallacious argument may be made that a proposition is not false because insufficient testing has been done to show it false. That is a reasonable argument.
Appeals to ignorance are used to shift the burden of proof to the other side. However, the burden of proof should be on the side that is making the assertion, not on the side that disputes the assertion.
Examples of the argument from ignorance
Almost all the claims from the anti-science movement revolve around some form of argument from ignorance.
Proponents of the anti-science movement will usually pick some aspect of a currently accepted scientific theory and argue that it must be wrong because they do not believe it explains some aspect of the natural world. Common examples of this are such claims as “you can’t prove global warming is caused by humans,” “I don’t see how evolution could increase the complexity of an organism,” “material properties of the brain cannot presently explicitly explain consciousness so it must be caused by non-materialist processes,” or “I don’t know how this alternative medicine works, but it does.”
The argument from ignorance is used widely in the anti-vaccine movement. I constantly hear someone say “you cannot prove that vaccines don’t cause XYZ (whatever the adverse event of the week is), so it’s possible that it does cause it.” Actually, since they are asserting the adverse event, they need to provide the evidence (and anecdotes don’t count as evidence).