Hans Blumenberg, Paradigms for a Metaphorology: Chapter 8, "Terminologization of a Metaphor: From 'Versimilitude' to Probability'"

Blumenberg now considers the phenomenon of transition from a metaphor to a concept. He focuses on a transition where both the original metaphor and the succeeding concept use the same German word, Wahrscheinlichkeit. The literal meaning of the word (which is incorporated in the old metaphorical usage) is the resemblance of truth. In modern usage it has come to mean probability.

Plato's eikos -- likelihood or what looks like being true -- from the Phaedrus and Timaeus -- taken up by 'truthlikeness' in Aristotle's Rhetoric -- translated as verisimile by Cicero. A reliance on such a concept in a Platonic tradition isn't as odd as it first appears, which is to say that the distance from Platonism to Skepticism isn't that great. The heightened demands for truth to be transcendent in the later Plato makes it ever more difficult to come to grips with it, to the point where in the Timaeus Plato admits eikos as a sufficient guide. In the Critias, Plato contrasts the certainty and exactness which characterizes our everyday activities with the faintness of our knowledge of the transcendent. (82-83)

In Academic Skepticism, the probable is held up specifically as a guide in practical human concerns, without being seen as evidence of theoretical truth. Cicero associates the probable with verisimilitude, the appearance of truth, but the actual closeness to truth implied is only given weight pragmatically. (83-84)

In Epicureanism, by contrast, the appearance of truth is granted to all manner of theoretical explanations of celestial phenomena, but this serves only to neutralize the claims of all the contending theories, thus securing man's peace of mind. (84)

For Lactantius, verisimile is seen only negatively, as the false appearance of truth which misleads men. In this, he is reclaiming the Stoic view of probability that Cicero had left aside -- that what is merely probable should not be assented to, and that instead judgment should be suspended. (85-86)

Augustine revives the metaphorical content of verisimile in arguing against the Academic tradition. He contends that it is impossible to make sense of verisimile without first having possession of the truth itself -- for how can one know if something resembles the truth without knowing the truth it refers to? (86-87)

Late medieval theology again severs the true from what seems to be true. With the growing theological insistence on the transcendent nature of truth, a new skeptical tendency affirmed faith even while conceding its apparent absurdity. In so doing, it reversed the Platonic assessment (from the Critias) of the relationship between knowledge of the divine and worldly -- now knowledge of the divine is held to be certain while knowledge of nature is held to be merely probable. This separation of faith and the appearance of truth provided autonomous space for speculation about the probable -- at least up to a point. Thus, the Church's response to Copernican and Galilean doctrines depended on its willingness (ultimately rejected) to tolerate probable claims which conflicted with revealed truth. (88-89)

Blumenberg links the development of mathematical probability with a desire to give an account for the world without reference to teleology or theology. Chance could only provide a rational alternative to these forbidden premises if it could be made predictable and quantifiable -- a labor taken up first and foremost by Pascal. Blumenberg notes that Leibniz endorsed the study of probability as a legitimate science, but cast doubt on its usefulness for addressing religious questions. Here, Blumenberg detects a residue of the metaphorical meaning of verisimilitude, for Leibniz's objection is that religious mysteries have the appearance of being false even when they are true, so an argument from probability cannot successfully confirm their truth. (where the seeming false captures the metaphorical content of the term) (89-91)

Mendelssohn adopts the Wolffian notion of probability as incomplete truth (I think the significance of this is that it puts probability on a continuum with truth), and he holds that probability applies to judgments about the past as well as the future. He objects to the idea that there are some probabilities about which it is impossible even for God to have knowledge that goes further than the probability (which some saw as a requirement of free will). (92-94)

Mendelssohn's defence of Copernican theory represents a benchmark in the transition from metaphorical to logical verisimilitude. His argument on behalf of Copernicanism is probabilistic: it reduces the number of independent factors needed to explain the movements of the heavenly bodies, and thus the reliance on an apparent coincidence of separate causes; reducing the number of required coincidences makes the explanation more probable. (94-95)

Maupertuis and Diderot object to the consequence of seeing probability as an incomplete truth. Truth cannot be established from probability -- in particular, for Maupertuis, the existence of a designer for the solar system cannot be deduced from the improbability of a close coincidence between the plane of the planetary orbits. With Diderot, the point is amplified by bringing together probability with a new metaphysical background assumption of infinite time. Given infinite time, the universe has unlimited chances for anything to occur, so no outcome is truly improbable. The explanatory function of a divine creator is thus replaced by chance. but since the creator had already become arbitrary and capricious in the development of Christian theology, this was a gain rather than a loss for a feeling of security about the world. Chance at least has its own laws. (95-97)

Blumenberg briefly considers how the the term verisimilitude had been transferred from classical to modern aesthetics. Classical aesthetics based on mimesis put a higher value on verisimilitude than truth in art. It was more important for a work of art to resemble an unrealized ideal rather than faithfully portray an existing object. Lessing (the modern example to whom Blumenberg restricts himself) also puts higher value on verisimilitude than faithful representation of reality in art. He conceives of verisimilitude in art as an ability to appeal to our inner sense of probability, to persuade us of its reality. (97-98)