They say:
“Physiognomy based on bodily features is itself reasoning by analogy, except that the common feature linking the primary and secondary cases is an indication of the cause, not the cause itself. This is what jurists call ‘analogy based on an indication’: one infers the existence of a cause from one of its effects, then infers its other effect from the existence of that cause. This rests on the premise that bodily temperament causes both an inward disposition and an outward feature. The outward feature thus indicates the temperament, which in turn indicates the inward disposition. An example is inferring courage from breadth of the upper body, on the grounds that both result from the same temperament, as is found in the lion.
“Establishing the cause in the primary case necessarily requires covariation or division, as explained above. Even if we assume that the two judgments in the primary case have a single cause, nothing prevents one of them from holding in the secondary case through a cause other than that operative in the primary case. In that event, the other judgment does not necessarily follow.”
Such is their argument.
The response is:
Their distinction between syllogistic reasoning and reasoning by analogy—that the former may yield certainty, whereas the latter yields only probability—is invalid. Rather, whenever either yields certainty, the other does too; and whenever either yields only probability, the other likewise yields only probability. Whether an argument yields certainty or probability does not depend on its taking one of these two forms rather than the other. It depends instead on whether it contains something that yields certainty. If either contains something that entails the judgment with certainty, certainty is attained through it. If it contains only something that establishes the judgment as probable, it yields only probability.
What is called the middle term in one is the shared attribute in the other. And the major premise, which states that the major term necessarily follows from the middle term, explains the efficacy of the attribute