§19. The Canonic of an Epicurean

Epicurus introduced an idea of epilogismos that became important in med­ical empiricism. Epicurean epilogismos is an appraisal or assessment, espe­cially by comparison, evaluating whether candidates are more or less evident, clear or less clear.

What Epicurus wants from epilogismos is the vindication of ordinary procedures of appraisal. These are good enough for natural philosophy and ethics, which have no need for Aristotle’s syllogistic or the logical formalities of the Stoics. That may be what drew medical empiricists to Epicurus. For these doctors, epilogismos was an unobjectionable ordinary sort of reasoning that they contrasted to a technical, objectionably rationalistic analogismos (§6). Modifying Epicurean procedure, they apply the term epilogismos to in­ference, emphasizing its character as “the reasoning common to all human beings,” rather than peculiar to mathematicians or dialecticians. Elucidating such inferences and explaining scientific knowledge on their basis is the task of the Epicurean discipline of canonic, the logic of reasoning about nature on the evidence of signs.

The centerpiece of canonic is the concept of a criterion, an Epicurean technicality and another idea with long legs in philosophy. A criterion is a rational license to infer the unseen. If a concept satisfies a criterion, then it may be validly used in natural philosophy. The analogy that apparently in­spired the idea is the artisan’s kanon, a standard length used in architecture and sculpture. The technician lays down the kanon and measures one thing by another, injecting rationality into the combination of elements. Epicurean reasoning compares the concept to the criterion, and when they match, the concept is allowed in reasoning about the causes of perception. Epicureans acknowledge three criteria: perceptions (aisthesis), common notions (pro- lepses), and affections (pathe). Of criterial aisthesis, perception, Epicurus wrote, “We must keep all our investigations in accord with our perceptions and in particular with the immediate apprehensions... in order that we may have indications whereby we may judge both the problem of sense percep­tion and the unseen.”¹⁵¹

Our best example of Epicurean canonic is On the Methods of Inference by Philodemus, of the Epicurean school at Naples in the first century bce. He says that all the arts and sciences agree in an empirical unity of method. All the technai, even mathematics, have their principles established by the em­pirical method of inference from the perceived to the unperceived, relying on what doctors call indicative signs. The relation of sign to its unseen ref­erent is founded on systematically confirmed perception. Epicurean canonic ignores the lauded Stoic teaching, which was also Aristotle’s, on the triadic sign: sign (S) signifying concept (C); concept denoting object (O). Epicurean signs signify not concepts but objects of experience or analogical extensions of experience, a relation mediated by perception independent of concepts. Philodemus relates how his teacher, Zeno of Sidon (also a teacher of Cicero), explained the inference from appearances to nature.

This empirical concept of knowledge entered Epicurean tradition early, even before the founder, with Nausiphanes, Epicurus’s Democretian teacher, who called it the method of “rhetoric.” Nausiphanes seems to have learned it from medicine, the Hippocratic method of conjectures (stochasmos). Ancient Medicine and Celsus both say medicine is conjectural. Conjectures are fal­lible inferences or predictions from our own observation and records of the past, made by means of signs. Hermogenes of Tarsus, a prominent rhetori­cian of the second century ce, said, “Conjecture is the substantial proof of an unknown fact from some apparent sign.” Such conjectural inferences are valid if and only if it is inconceivable (adianoeton) that the sign exist and the signified thing not exist, a rigor that bends under a commodious conception of the inconceivable.¹⁵³

Philodemus’s explanation of canonic is also a defense against criticism by Stoics. They say that knowledge of nature is impossible without the a priori assumption of nature’s divine rationality. Only then can we rationally ex­pect nature to be uniform even where unperceived, an unwanted dogma Epicureans warmly repudiate. They understand empirical knowledge em­pirically, which is the distinction of their canonic in contrast to the logic of Aristotle or the Stoics. If Democritus was an empiricist, Epicurus is a rad­ical empiricist. Stoic method favored reasoning by contraposition—the an­tecedent is impossible if the consequent is denied: If not q, then not p; but p is evident; therefore q. Ifp is the given evidence and q the hypothetical cause, then the fact ofp is proof of q, since by contraposition, if not q, then not p.

Stoics challenge the Epicurean use of analogical reasoning in natural phi­losophy. Analogical inference from perceived to unperceived is cogent, they say, only on the assumption of the similarity of the unperceived to the per­ceived, which is impossible to establish empirically. Philodemus replies that analogy is reasonable because it is inconceivable that phenomena would not be resembling. “An object deprived of all similarity with appearance is inconceivable.” The method of analogy is an alternative to the Stoic dogma of nature’s rationality. “Appealing to similarity, we declare that in respect to mortality the men outside our experience are similar to those within our ex­perience.” Philodemus says the method of analogy is “a sound method of inference, with this condition, that no other appearance or previously dem­onstrated fact conflicts with the inference.”¹⁵⁵

For another example, Philodemus explains that motion evinces void be­cause what we see suggests it, experience confirms it, and no disconfirma­tion is known after searching. If we systematically follow the evidence of the senses and are not led astray by false belief, we cannot fail to come to think that motion requires void. Having right belief is not a matter of reasoning but rather of examining thoughts prompted by what we observe. The belief is as natural as belief in gods or the goodness of pleasure, and should be accepted as reasonable on the same grounds. The belief is reasonable, but not “ra­tional,” according to the tendentious usage established by Plato and Aristotle in their schools.

Deductive reasoning, whether Aristotelian or Stoic, is not the nerve of Epicurean canonic, which is instead to test experience by experience.

This empiricism is developed with some energy in a line of rhetorical phi­losophy that little resembles the impression of rhetoric we get from Plato. This rhetoric practices an empirical method that it shares with strategy, his­tory, medicine, astronomy, and navigation, expressing an empirical tradi­tion of late antiquity in which the arts and sciences were drawn together. The conjectures this rhetoric credits have logical qualities that would not endear them to Platonists or Aristotelians, the inferences being fallible, revisable, merely probable. Classical thought was not nimble with probability, tending to think merely probable, therefore philosophically false. But later tradition, beginning with the forensic experience of the Roman courtroom, took the probable seriously and developed the first qualitative idea of scalar proba­bility. It was an achievement to establish that not all probable judgments have the same quality, that there are better and worse. Sextus says the idea began in medicine, from which it passed into this empiricist-rhetorical philosophy, especially with Carneades, who inspired its modern revival in Gassendi and Locke (§51).¹⁵⁷