Art History: A Preliminary Handbook
by Dr. Robert J. Belton
Three Important Acronymns
a. APOLO [sic]:
Always Pull Out Legitimate Observations. I.e., don't leave material out if it is pertinent to the argument you are constructing, and don't take it for granted that the reader will make connections without your explicit help. (Essays which require the reader to guess at the writer's meaning earn a failing grade. See Grades.
b. DAFFNIE [sic]:
Don't Argue From Facts Not In Evidence. This is truism in a court of law. Why should it be any different here? If you are making a point of something, be sure you have genuine and clearly stated evidence. Evidence consists of facts about a case that can be introduced as premises to determine a reasonable conclusion. There are various types of evidence in legal proceedings: autoptic (the thing itself, like a murder weapon), character (information about the offender, victim, etc.), exculpatory (tending to prove innocence), exemplars (forensic and other material evidence, like fingerprints), expert (reliance on authority figures in a given field), inculpatory (tending to prove guilt), material (objects, substances, measurable data), oral (testimony of witnesses), rebuttal (arguing against the relevance or reliability of another interpretation), and so on. Some of these are admissible only in certain circumstances. A well-known example is hearsay evidence (testimony that so-and-so said such-and-such, produced at second or third hand), which must be examined for reliability. All of these have rough analogies in art historical evidence: autoptic (the work), character (artist- and/or patron-directed context), exculpatory/inculpatory (tending to prove or disprove intentions), exemplars/material (characteristic brushwork, signatures, x-rays, letters, diaries, public statements, etc.), expert (previous research), oral (testimony), rebuttal, etc.
Relevant, Sufficient, Valid Propositions. Apart from woozily subjective musings, any critique of a work of art is an argument, in effect. An argument is defined in informal logic as "reasons (premises) given in support of a conclusion." A good argument must meet the criteria of relevance, sufficiency and validity.
Relevance means that premises must increase the probability of the conclusion being true. Because the conclusion is about spiritual matters in the Van Eyck example, conventional Christian symbolism is relevant, whereas Van Eyck's years of quasi-diplomatic service for the Duke of Burgundy is likely not. Note that "relevance" does not mean "truth." Let's say for the sake of argument that the sign "dog" actually had some other meaning at the time. The point is still structurally relevant, so if you want to disprove the point it you must counter it with rebuttal evidence (see DAFFNIE above). In this case, the point is both relevant and true. On the other hand, it would certainly not be relevant that Van Eyck was a cribbage-player or liked pickles, although it might have been true.
Sufficiency means that an analysis, definition, description, or explanation must be full enough that it applies only to the matter at hand and cannot be replaced by some other analysis, definition, etc. In a definition, for example, the proposition "a chair is something to sit on" is insufficient because we could replace "chair" with "stool," "bench," or even "committee." To increase the sufficiency, one has to distinguish between such alternates by genus and differentia (i.e., by general category and differences from other members of that category). The definition would thus become something like "a chair is a piece of furniture (genus, thus eliminating "committee") which usually consists of a flat surface on four legs (differentia, eliminating "stool") to support the buttocks of one person (eliminating "bench"), and providing support for the back and sometimes arms. The principle is the same for longer arguments, but a demonstration would take up too much space here. For the moment, note only that this means "don't leave anything out that helps your case," as in APOLO above.
Validity is a structural principle of informal logic and is not to be confused with truth either. Validity is determined by whether or not the conclusion of an argument follows necessarily from its premises. An argument may be structurally valid even if one of the premises is untrue. The syllogism "Van Eyck was a painter; painters are wild and irreverent; therefore, Van Eyck was wild and irreverent," is perfectly valid, even though it is not true that all artists are wild and irreverent. Similarly, an argument that is invalid may happen to be true, as in "Kenyan artists speak Swahili; Van Eyck was not a Kenyan artist; therefore, Van Eyck did not speak Swahili." The conclusion does not follow necessarily from the premises because it's possible a non-Kenyan artist could speak Swahili, although we happen to know in this instance that he did not. Obviously, the best argument is going to be both valid and true.
© Copyright 1996 Robert J. Belton
Last reviewed 12/19/2012 11:13:29 PM