…claim to be an expert if you’re not one
…provide facts, evidence, and statistics to support your position
To embrace a purely formal approach to analogy and to abjureformalization entirely are two extremes in a spectrum ofstrategies. There are intermediate positions. Most recentanalyses (both philosophical and computational) have been directedtowards elucidating general criteria and procedures, rather than formalrules, for reasoning by analogy. So long as these are notintended to provide a universal ‘logic’ of analogy, thereis room for such criteria even if one accepts Norton's basicpoint. The next section discusses some of these criteria andprocedures.
…use strictly moral or religious claims as support for your argument
If analogical reasoning is required to conform only to a simpleformal schema, the restriction is too permissive. Inferences areauthorized that clearly should not pass muster… The naturalresponse has been to develop more elaborate formaltemplates… The familiar difficulty is that theseembellished schema never seem to be quite embellished enough; therealways seems to be some part of the analysis that must be handledintuitively without guidance from strict formal rules. (2012: 1)
oopen-mindedness regarding divergent world views
Hesse (1966) offers a sharpened version of Aristotle's theory,specifically focused on analogical arguments in the sciences. Sheformulates three requirements that an analogical argument must satisfyin order to be acceptable:
oflexibility in considering alternatives and opinions
Most theories of analogy agree with Aristotle and Hume on thisgeneral point. Disagreement relates to the appropriate way ofmeasuring overall similarity. Some theories assign greatestweight to material analogy, which refers to shared, andtypically observable, features. Others give prominence toformal analogy, emphasizing high-level structuralcorrespondence. The next two sub-sections discuss representativeaccounts that illustrate these two approaches.
ounderstanding of the opinions of other people
Wherever you depart, in the least, from the similarity of the cases,you diminish proportionably the evidence; and may at last bring it to avery weak analogy, which is confessedly liable to error anduncertainty. (1779/1947: 144)