Read part 3 here.
4 Fides in Boethius’ Topical Works
Of the works in the above mentioned codex, those most important for determining Anselm’s use of fides are Boethius’ In Ciceronis Topica and De differentiis topicis. We turn to these now.
The most important place where fides shows up in Cicero’s topics is in the definition of a topic itself. Cicero defines locus, the Latin translation of the Greek τόπος, thus: ‘Therefore we may define ‘topic’ as the seat of an argument, and ‘argument’ as the ratio that grants fides to a doubtful matter.’ Boethius comments on the passage as follows:
‘There are many things that grant fides, but since they are not rationes, neither can they be argumenta: for instance, sight grants fides to things seen, but because sight is not a ratio, neither can it be an argumentum. He assumes one difference, that which grants fides, since every argumentum grants fides. If, then, we were to join the genus to the difference, and call this an argumentum… would the complete nature of an argumentum be made clear? Hardly…for an argumentum is what establishes something (quod rem arguit) – that is, what proves it – and nothing can be proven unless it is doubtful… Adding, then, another difference, that is, to a doubtful matter, the definition of argumentum is made complete, consisting of a genus and two differences: the genus, ratio; the first difference, granting fides; the other, to a doubtful matter.’
Here, we must not construe ratio as ‘reasoning’, but as what is reasoned about, since the term is constitutive of the meaning of argumentum. Further, as Toivo Holopainen has shown, the term argumentum need not refer to the complete set of premises constituting a valid deduction along with their conclusion, but may refer solely to the middle term of such a deduction. If this is so, the meaning of argumentum need not be extensionally differentiated from that of ratio, but may be so intensionally: a ratio is an argumentum when it plays the role of a middle term in a certain kind of deduction. We may, then, construe ratio as ‘idea’ or ‘concept’, provided we understand it as the essence of a thing qua received by the intellect, and not as a representational medium for understanding the world spontaneously drawn up by the intellect.
To construe argumentum as ‘argument’, or even ‘middle term of an argument, in the above passage would be too broad: only a middle term successfully establishing its conclusion can be an argumentum. An argumentum must be the subject or predicate of a true sentence, the major term taking up the sentential position not filled by the argumentum, serving to deduce the conclusion relating the minor term to the major term. Furthermore, it must be more immediately credible that the argumentum relates to the minor term than that the major term is predicable of the minor. Thus, redundant or circular proofs, such as:
B is A
A is B
A is A
will not have a middle term that is an argumentum; nor will a deduction attempting to establish a conclusion equally or less doubtful than its major premise. Thirdly, the conclusion must, à la contemporary relevant logic, genuinely follow from the premises.
In the above passage, we are told: 1) fides can be brought about by an argumentum, but 2) also by other means (e.g. sight); 3) every argumentum effects fides; 4) fides can be provided for doubtful things 5) as well as things not in doubt. Furthermore, we can deduce that 6) fides is not antithetical to ratio, since some rationes—namely, argumenta—grant it; 7) fides is granted by something fitting the object granted it – just as visus grants fides to visible things, so rationes grant fides to intelligible things; and 8) fides is not granted to the skeptical inquirer, but to the thing inquired about. This last point is of vital importance: for Anselm and the Boethian tradition leading up to him, the granting of fides essentially concerns the matters themselves, not subjective mental states. Faith as a state of belief must follow from this either derivatively, from grasping the object at hand as it is really presented; or alternatively, the faith of a believing subject may be construed as a special case of this object-level participation.
From the enumeration of the above criteria, one might construe the fides granted to a doubtful matter, from which a predicate’s holding/not holding of some subject follows, as analogous to the role played in contemporary logic by the assertibility conditions of a sentence: an argumentum is an idea making a matter previously in doubt assertible. But since the granting of fides is nothing linguistic, what we have granted here is rather a condition for assertibility. A good translation of fides, then, in this context, would be something like reliability, or even – bringing out the ontological tenor a bit more – groundedness.
Putting this together, we can rephrase Cicero’s definition thus: a topic is the seat of an argumentum, and an argumentum, an idea serving as a medium grounding a doubtful matter.
 I leave several key terms untranslated in the passage. The purpose of this is to leave the term at least somewhat unfamiliar, and thereby encourage the reader to think these terms from the manner and context in which they are employed, rather than bringing a prior notion to the reading of the text.
 ‘Itaque licet definire locum esse argumenti sedem, argumentum autem rationem quae rei dubiae faciat fidem’ Topica Chapter, 8, lines 27-28.
 BTC 1048BC:
Multa enim sunt quae faciunt fidem, sed quia rationes non sunt, ne argumenta quidem esse possunt, ut visus facit fidem his quae videntur, sed quia ratio non est visus, ne argumentum quidem esse potest. Differentiam vero unam sumpsit, eam quae faciat fidem, omne enim argumentum facit fidem. Si igitur iunxerimus genus ac differentiam, et id esse argumentum dicamus… num tota argumenti natura monstrata sit? Minime … argumentum namque est quod rem arguit, id est probat, nihil vero probari, nisi dibium, potest… Addita igitur alia differentia quae est rei dubiae, facta est integra definitio argumenti, ex genere et duabus differentiis constans, genere quidem, ratione: una vero differentia, quod faciat fidem; altera vero, quod rei dubiae est.
 See BDT 1174C, given in a prior footnote.
 See Holopainen 2007.
 Cf. Klima 2013, sec. 7:
So, a common nature or essence according to its absolute consideration abstracts from all existence, both in the singulars and in the mind. Yet, and this is the important point, it is the same nature that informs both the singulars that have this nature and the minds conceiving of them in terms of this nature.
 In Aristotelian syllogistic, the major term of a syllogism is the term that also serves as the predicate in the conclusion, while the minor term is the term that serves as the subject of the conclusion.
 However, the argumentum need not hold in this way of the minor term: the sentence joining these two need only be probabilis, i.e. worthy of esteem, something believed by the multitude or by the wise. Note that a topical argument does not primarily seek to establish the credibility of the sentence concluded to by the syllogism, but the credibility of the predicate of the conclusion’s holding of the subject. Though in any sensible logic, an assertible predication of a subject should entail the assertibility of the statement wherein the predicate is predicated of that subject, it seems that the earlier Medieval analysis took dyadic predicates like ‘true of’ and ‘credible of’, taking a subject and predicate as arguments, to be in some sense prior to the monadic ‘true’ or ‘credible’, taking (the name of) a sentence as argument. On the meaning of probabilis in Boethius’ theory of the topics, see BDT 1180C-1182C.