Even when we have identified what kinds of evidence are preferable (such as firsthand experience over hearsay), we still need to figure out the characteristics of evidence of a given kind that enable it, in a context, to move us from disbelief to belief or from opinion to solid knowledge. Special contexts suggest a special necessity for holding fast to tried-and-true standards.
Good evidence is typified by the presence of a reasonable number (though not necessarily all) of such characteristics as relevance, non-equivocation and non-circularity, plus replicability and testability, with controls for limiting conditions, all articulated in a cogent theoretical framework. Replicability and testability are why scientists publish their results.
Limiting conditions are disabilities in the apparatus or the observer that interfere with reliable results (e.g. colour blindness, bias or vested interest, uncontrolled outside variables). One explanation is not "just as good as another" because one theory is not "just as good as another.
How much good evidence does one need? There are no guarantees but we leave the door open for corrections "in light of further evidence." One had best not need enough to secure "certainty." This is why, on matters of substance, one must hedge one's bets and be "ready to revise."
No comments:
Post a Comment