The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 6

Read part 5 here.

6 Conclusion

Anselm, then, tells us that he is searching for one notion, in contrast to the many of the Monologion, from which the many things believed of God could be derived. Thus, when Anselm prays ‘Therefore, Lord, who grants intellectus to fides, grant that I may understand that you are, as we believe, and that you are what we believe,’[1] he is certainly seeking that his faith be deepened by understanding; but his asking for this is simultaneously, and even primarily, his asking God to unravel the core sense (intellectus) of something making secure (faciens fidem), i.e. the notion ‘that than which a greater cannot be thought.’ For this reason, the Proslogion as a whole is a meditation on the substance of something worthy of belief (ratio fidei). The notion ‘that than which a greater cannot be thought’ itself serves as a medium leading to a fuller notion of God, thereby securing the divine attributes understood through this ratio. It is a consequence of this that the work also exhibits the noetic satisfaction of one holding to this faith – faith seeking understanding in the sense commonly understood. Anselm is searching for a single notion or description that can lead to its ground; he is searching for a title or name of God that can bring him closer to seeing God as he truly is. This role is filled by the notion id quo maius cogitari non potest.

Bibliography and Abbreviations

Abelson, Paul. The Seven Liberal Arts: A Study in Mediaeval Culture. New York: Teachers College, Columbia University Press, 1906.

Anselm of Canterbury, Saint. De Grammatico. In Henry 1974: 48-80 [DG].

            . Liber apologeticus contra Gaunilonem respondentem pro insipiente. [Resp.]

            . Proslogion. [Pros.]

Archambault, Jacob. ‘Aquinas, the A Priori/A Posteriori Distinction, and the Kantian Dependency Thesis.’ Religious Studies 50 (2014): 175-192.

  1. Lacombe et al., eds. Codices: Pars Prior, ed. Rome: La Libreria dello Stato, 1939. [Aristoteles Latinus]

Barth, Karl. Anselm: Fides Quaerens Intellectum. Translated by Ian W. Robertson. London: SCM Press, 1960.

Bekker, Gustav Heinrich. Catalogi Bibliothecarum Antiqui. Bonn: M. Cohen et filium, 1885.

Bencivenga, Ermanno. Logic and Other Nonsense: The Case of Anselm and His God. Princeton: Princeton University Press, 1993.

Boethius. De Differentiis Topicis. In PL 64, 1173-1216. [BDT]

            . In Categorias Aristotelis libri quatuor. In PL 64, 159-254. [BC]

            . In Librum Aristotelis De Interpretatione Libri Duo. In PL 64, 293-392. [BDIL]

            . In Topica Ciceronis Commentariorum Libri Sex. In PL 64, 1039-1174. [BTC]

            . Commentaria In Porphyrium a se translatum. In PL 64, 71-158. [BCP]

Cicero, Marcus Tullius. Topica. Edited and translated with an introduction and commentary by Tobias Reinhardt. Oxford: Oxford University Press, 2003. [Topica]

Eadmer. Vita Anselmi. In PL 158, 49-118. [VA]

Gasper, Giles. Anselm of Canterbury and his Theological Inheritance. Aldershot: Ashgate, 2004.

Gaunilo. Liber pro insipiente adversus S. Anselmi in proslogio ratiocinationem. [Pro Ins.]

Gilson, Etienne. Reason and Revelation in the Middle Ages. New York: Scribner, 1951.

Green-Pedersen, Niels J. The Tradition of the Topics in the Middle Ages: Commentaries on Aristotle’s and Boethius’ ‘Topics’. Munich: Philosophia Verlag, 1984.

Henry, Desmond Paul. Commentary on De Grammatico: The Historical-Logical Dimensions of a Dialogue of St. Anselm’s. Dordrecht: D. Reidel, 1974.

Holopainen, Toivo. ‘Anselm’s Argumentum and the Early Medieval Theory of Argument.’ Vivarium 45 (2007): 1-29.

Klima, Gyula. ‘The Medieval Problem of Universals,’ The Stanford Encyclopedia of Philosophy (Fall 2013 edition). Edited by Edward N. Zalta. URL =>.

            . ‘Two Summulae, Two Ways of Doing Logic: Peter of Spain’s ‘realism’ and John Buridan’s ‘nominalism’.’ in Methods and Methodologies: Aristotelian logic East and West, 500-1500, edited by Margaret Cameron and John Marenbon, 109-126. Leiden: Brill, 2011.

            . ‘Saint Anselm’s Proof: A Problem of Reference, Intentional Identity and Mutual Understanding,’ in Medieval Philosophy and Modern Times, edited by G. Hintikka, 69-87. The Netherlands: Kluwer, 2000.

            . ‘Ancilla Theologiae vs. Domina Philosophorum: Thomas Aquinas, Latin Averroism, and the Autonomy of Philosophy.’ In What is Philosophy in the Middle Ages? Proceedings of the Tenth International Congress of Medieval Philosophy, edited by J. Aertsen and A. Speer, 393-402. Berlin: Walter de Gruyter, 1998.

Leff, Michael C. ‘Boethius and the History of Medieval Rhetoric.’ Central States Speech Journal 25 (1974): 135-141.

Migne, J. P. (1841-55). Patrologia Latina [PL].

Priscian of Caesarea. Opera. Edited by August Krehl. Leipzig: Weidmann, 1819-20.

Remigius of Auxerre. Commentum in Martianum Capellam, Libri I-II. Edited by Cora E. Lutz. Leiden: Brill, 1962.

Scotus, John Duns. A Treatise on God as First Principle. Translated by A. B. Wolter. Chicago: Franciscan Herald Press, 1966.

Southern, R. W.. Saint Anselm and His Biographer: A Study of Monastic Life and Thought. Cambridge: Cambridge University Press, 1963.

Sweeney, Eileen C. Anselm of Canterbury and the Desire for the Word. Washington, DC: CUA Press, 2012.

            . ‘Anselm’s ‘Proslogion’: The Desire for the Word.’ The Saint Anselm Journal 1 (2003): 17-31.

Thomas Aquinas (1941). Summa Theologiae, Prima Pars. (Ottawa: Studii Generalis O. Pr.).

[1] Pros. 2: ‘Ergo, Domine, qui das fidei intellectum, da mihi, ut … intelligam quia es, sicut credimus, et hoc es, quod credimus.’

The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 5

Read part 4 here.

5 Fides Quaerens Intellectum.

We are now in a position to see how this account clarifies the meaning of the title Anselm originally gave to his Proslogion. In the prologue to the text, Anselm says his brother monks asked him for an ‘example of meditating on a ratio fidei.’[1] This could consist, he says, in a multitude of arguments, as did the Monologion, or it could consist in one argument. Anselm asks:

‘whether perhaps one argument (unum argumentum) could be found, which 1) would require nothing other than itself alone to be proven; and 2) would be sufficient to establish: a) that God truly is; and b) that he is the highest good needing no other; and c) is whom all need that they may be and be well; and d) whatever else we believe concerning the divine substance.’[2]

He goes on to call this same argumentum a ‘cogitatio I zealously embraced.’[3]

Like the Monologion, the Proslogion is an example of meditating on a ratio fidei – in this case, the notion that than which nothing greater can be thought. The dialectical aspect of the text is thus not in tension with it having a ‘meditative’ or ‘theological’ sense, since the very process of determining what can be deduced in that context is given via a rumination on, and unpacking of, the meaning of the middle term. Anselm is searching after some one thing, a ratio fidei, called a cogitatio in relation to his possessing it, to serve as an argumentum. The ratio is ‘that than which nothing greater can be thought’, taken up as an argumentum, i.e. a middle term in a series of syllogisms running as follows:

Major Premise: That than which nothing greater can be thought exists in reality, is the highest good, etc.

Minor Premise: God is that than which nothing greater can be thought

Conclusion: God exists in reality, etc.

Each of the above arguments is an example of the Themistian topic from a description (a descriptione), itself grouped among the topics from substance (a substantia).[4] Here, ‘that than which nothing greater can be thought’ is what later Medievals would call a nominal definition, and what philosophers today call a definite description: it captures the content of the idea of God in an exact manner, but since it does so by way of certain contingent features (i.e. being thinkable, being greater than other thinkables),[5] it cannot qualify as an example of the topic from a definition (a diffinitione). The topic is an intrinsic topic, one where the major term’s relation to the minor is elicited directly from the argumentum, usually in a straightforwardly syllogistic manner. Intrinsic topics contrast with extrinsic topics,[6] where the argumentum bears a synthetic relation – proportionality, contrariety, etc. – to the other terms of the argument; and middle topics, which as the name suggests, are in some sense between intrinsic and extrinsic.

The order in which Anselm presents the divine attributes in the Proslogion is gradual, designed to facilitate the spiritual progress of the reader. After establishing God’s existence, Anselm claims God is ‘whatever it is better to be than not to be.’[7] But this general claim is complicated immediately in chapters 6-8, which ascribe sensibility, omnipotence, and impassibility to God; and again, in 9-11, which highlight the conflict between justice and mercy in God. In these chapters, the protagonist’s process of discovering what predicates befit God is also a rarification of his concepts of ordinary goods, where aspects of the ordinary concept in play are also denied of God. This process reaches a peak in chapter 12, where the whole manner in which anything is predicated of God is distinguished from the predication of properties of ordinary objects on account of God’s simplicity; and again in chapter 15, where even the middle term ‘that than which nothing greater can be thought’ is qualified by the claim that God is greater than can be thought. After this, the succeeding chapters become decidedly more apophatic in character.

To support his major premises, Anselm uses reductio proofs. All of these are cases of the middle topic from division (ex divisione), which is similar to, but not quite the same as, disjunctive syllogism.[8] The topic from division begins with a partitioning of opposing predicates with respect to a subject. From here, it can proceed either directly or indirectly. Anselm proceeds indirectly:

That than which nothing greater can be thought is F or G

Assume it’s F

It’s not F

Therefore, it’s G.[9]

The proof is indirect because it assumes the horn of the dilemma it ultimately rejects. The topic is midway between intrinsic and extrinsic because though one of the predicates follows directly from the idea ‘that than which a greater cannot be thought’, this is only made evident by comparing the two predicates to each other as lesser to greater.

Though the Proslogion forms a unity relying on a single argumentum, it does not provide one argument in the modern sense, but several, identical in all but the major term. Since this is so, Gaunilo’s Pro insipiente is no longer merely a criticism of the argument of Proslogion 2-3: it is a criticism of the entire work, since if it prevents linking the predicate ‘exists’ to that than which nothing greater can be thought, it will a fortiori prevent any of the other things Anselm attempts to establish of God from being established.[10]

In each argument of the Proslogion, the middle term is taken up as something whose holding of the minor term is immediately credible or reliable: it ‘requires nothing besides itself alone to be established.’ This is why Anselm never attempts to justify the minor premise either in the Proslogion or in the response to Gaunilo: that the premise is to be assumed is simply part of the rules of the monastic disputation Anselm and Gaunilo are engaging in.[11] Thus, when in section 4 of the Liber Pro Insipiente Gaunilo allows his fool to deny that that than which nothing greater can be thought is God – thereby also reconstruing Anselm’s original first-figure proof as a third-figure one – Anselm bars this line of attack. ‘I use your faith and conscience for a most sure proof of how false this is.’[12]

This reading further reveals something about Anselm’s use of the term ‘fool’ (insipiens) as an appellative for the atheist. The minor premise of a topical argument must be probabilis, or reliable. Boethius tells us ‘the reliable is what seems to be either to all, or to many, or to the wise; and among these [last] either to all, or many, or to those most renowned or distinguished; or to the specialist concerning his own province.’[13] God’s being something than which a greater cannot be thought fits this in several ways. First, it would have been widely assented to in Anselm’s time, and hence ‘what seems to be either to all or to many’. Second, it would have been attributable to ‘those most renowned or distinguished among the wise’, since the claim is found in Seneca the Elder, who, not always having been clearly distinguished in the Middle Ages from his nephew, was an authority in philosophy on par with Plato and Aristotle. Thirdly, and perhaps most interestingly, the pair wise/fool is mapped onto that between the Catholic and the non-believer, and so the claim is in this sense something known to many among the wise. This sense thus underlines the opening lines of the Response to Gaunilo: ‘Since the fool himself, against whom I spoke in my little work, does not reprove me in these words, but rather one who is not a fool, and a Catholic on behalf of the fool, it suffices for me to reply to the Catholic.’[14] Here, the fool is explicitly excluded from the second round of debate as incapable of perceiving what is plain to Catholics.[15]

Given the importance Anselm ascribes to the notion ‘that than which a greater cannot be thought’ as the unum argumentum binding the Proslogion together, it should not be surprising that Gaunilo’s first and principal objection to Anselm’s argument attacks precisely the fittingness of taking up this ratio as an argumentum:[16]

‘Perhaps one can respond that this is now said to be in my understanding from nothing else besides that I understand what is said. Am I not likewise said to have certain false things, and even things existing in no way in themselves in mind when I would understand someone saying them, whatever he would say?’[17]

Here, the mode of cogitation for a false concept – i.e. one failing to be instantiated in a subject under discussion – or one ‘existing in no way in itself’ – i.e. an intrinsically incoherent one – is different from that of one known to belong to a subject. But if the concept ‘that than which a greater cannot be thought’ doesn’t fit into one of these categories, then the mode of cogitation proper to it with respect to God must be that of being understood, which is different still from the previous modes. But understanding only befits a thing established by proof. Hence, if God is understood in this way, then ‘that than which a greater cannot be thought’ does not serve as an argumentum, since it does not secure a matter in doubt, but rather something already established. Put more plainly, even if this concept of God is coherent, an argument relying on it to establish God’s existence will be sound only if it is not probative, i.e. provided it proves nothing new. This is why Gaunilo says the reasoner following this argument will not move from having God in mind as id quo maius cogitari non potest at a preceding time to understanding him to exist at a later time.[18]

But where Gaunilo insists the mode of being of the cogitatum ‘that than which a greater cannot be thought’ must be given beforehand, thereby blocking the temporal process of moving from the premises to the conclusion of the argument, Anselm insists on the partiality of his description to safeguard this temporality. We have already seen Anselm himself later qualifies his famous description of God in chapter 15 of the Proslogion. We can now also see that the minimality and incompleteness of the concept is vital to ensuring its accessibility to the reader meditating along with the text.

‘But if you say that what is not understood completely is not understood and is not in the intellect, say that one who cannot look into the purest light of the sun does not see the light of day, which is nothing besides the light of the sun.’[19]

For Anselm, the meditator certainly has this concept ‘that than which a greater cannot be thought’ in mind, even though neither the attributes nor the mode of being of what it signifies need be specified antecedently to or immediately in that concept. Just as the light of the sun is visible, albeit not directly, so too God, than whom a greater cannot be thought, is intelligible, even if a complete understanding of the divinity escapes us.

[1]Pros. proemium.

[2] Pros. proemium: Coepi mecum quaerere si forte posset inveniri unum argumentum, quod nullo alio ad se probandum, quam se solo indigeret; et solum ad astruendum quia Deus vere est, et quia est summum bonum nullo alio indigens, et quo omnia indigent ut sint et bene sint; et quaecunque credimus de divina substantia.


[4] See BDT 1187B-1187D.

[5] This needn’t imply the existence of real distinct accidents inhering in God. But this point has to wait until Proslogion 12, and so isn’t yet present at the beginning of Anselm’s treatise.

[6] On extrinsic topics, see BDT 1190B-1192B.

[7] Pros. 5.

[8] See BDT 1192C-1193D.

[9] This format is especially prominent in Scotus’ reformulation of the proof. See Scotus (1966).

[10] This may help explain Anselm’s sometimes acerbic tone in the Responsio: he rightly read Gaunilo’s praise at the end of the Pro Insipiente as tongue-in-cheek.

[11] Cf. Aquinas’ criticism of the argument at ST. Ia, q. 2, art. 1. There, Aquinas attacks the evidential status of the minor premise, and thereby the fittingness of taking it up into the context of a dialectical disputation: ‘It is possible that who hears the name ‘God’ does not understand it to signify something than which a greater cannot be thought, since some would believe God is a body.’

[12] Resp. 1: ‘Quod quam falsum sit, fide et conscientia tua pro firmissimo utor argumento’

[13] BDT 1180CD: ‘Probabile vero est quod videtur vel omnibus, vel pluribus, vel sapientibus, et his vel omnibus, vel pluribus, vel maxime notis atque praecipuis; vel quod unicuique artifice secundum propriam facultatem.’

[14] Resp. proemium.

[15] This strategy for excluding the fool from the disputation is retained as the environment for the proof shifts from monastic to scholastic. Witness Bonaventure: ‘the intellect has in itself … sufficient light to repel this doubt and to extricate itself from its folly. Whence the foolish mind voluntarily rather than by constraint considers the matter in a deficient manner, so that the defect is on the part of the intellect itself and not because of any deficiency on the part of the thing known.’ Trans. from Klima 2000, 75.

[16] However, I know of no secondary literature on the topic that has so much as mentioned this.

[17] Pro Ins. 2: ‘Respondere forsitan potest, quod hoc jam esse dicitur in intellectu meo, non ob aliud, nisi quia id quod dicitur intelligo. Nonne et quaecunque falsa, ac nullo prorsus modo in seipsis existentia, in intellectu habere similiter dici possem cum ea, dicente aliquot, quaecumque ille diceret, ego intelligerem?’

[18] Pro Ins. 2: ‘Sed si hoc est: primo quidem non hic erit aliud, idemque tempore praecedens, habere rem in intellectu; et aliud, idemque tempore sequens, intelligere rem esse.’

[19] Resp. 1: ‘Quod si dicis non intelligi et non esse in intellectu, quod non penitus intelligitur; dic quia qui non potest intueri purissimam lucem solis, non videt lucem diei, quae non est nisi lux solis.’

Part 6.

The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 4

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.[1]

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.’[2] 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.’[3]

Here, we must not construe ratio as ‘reasoning’, but as what is reasoned about, since the term is constitutive of the meaning of argumentum.[4] Further, as Toivo Holopainen has shown,[5] 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.[6]

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[7] taking up the sentential position not filled by the argumentum, serving to deduce the conclusion relating the minor term to the major term.[8] 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.

[1] 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.

[2] ‘Itaque licet definire locum esse argumenti sedem, argumentum autem rationem quae rei dubiae faciat fidem’ Topica Chapter, 8, lines 27-28.

[3] 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.

[4] See BDT 1174C, given in a prior footnote.

[5] See Holopainen 2007.

[6] 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.

[7] 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.

[8] 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.

Part 5

The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 3

Read part 2 here.

3 Intellectus in the Boethian Works of the Logica Vetus

The prominence of the subject-object dichotomy for our own understanding of the faith/understanding relation provides perhaps the greatest obstacle to correctly approaching the text of Anselm’s Proslogion. The main features of this dichotomy are 1) that the world of objects is conceived of after the fashion of a domain of objects in model theory, typically the minimal set of things necessary for the maintenance of some sufficiently canonical activity (usually given by the hard sciences); and 2) that anything outside of this basic ‘furniture of the world’ is thought of as added to it by the activity of subjects. Broadly, everything that is must be either a mere entity, the paradigmatic cases of which tend to be artefacts; or a thinking thing, or something pertaining to the activity of such a being. Particularly, meanings must either be platonic objects or mental constructions of some sort.

In contrast, earlier Medieval logic generally takes meanings to be neither objects nor impositions of subjects: meanings are something had by entities;[1] and because of this, we may by an appropriate transference ascribe that meaning to a term referring to the entity. Boethius makes the point as follows:

‘Whenever one thing partakes of another, this participation is given in the name as well as in the thing. For instance, a certain man, because he partakes of justice, draws near [to justice] really, and hence draws his name near as well: he is called just.’[2]

The same point can be made not only with respect to a thing and its name, but also with respect to a thing and its concept. In the De Grammatico, there is a point where the teacher admonishes the student saying:

‘When it is asserted that every man can be understood to be man without literacy, and no literate can be understood to be literate without literacy, doesn’t this mean that being a man does not require literacy, and being a literate requires literacy?’[3]

The above passages manifest two aspects of early Medieval thinking more generally. First, the Medieval analysis, unlike our own, bridges the gap between talk of the meaning of terms and the talk of meaning one finds, for example, in questions about the ‘meaning of life’.[4] Second, it denies thought the character of spontaneity. So, for instance, if Anselm can be truly described as fidelis/faithful, then i) our ability to call him such is a consequence of his being such, and ii) he is faithful by having faith, conceived as the principle of his faithfulness. Similarly, if Anselm understands something, he understands by understanding, i.e. whatever underlying principle it is that grants understanding to a subject. Such attitudes are ‘subjective’ merely the sense in that they are intentions of a subject subordinated to an intended ideal: it does not mean they are empty intentions, nor that the intention becomes subordinated to the intending by its being enacted.[5]

If we are faithful to this general pattern, we will conclude that for Anselm, the sense of understanding given in ‘x understands’ will be unpacked as ‘x has understanding’, just as Anselm unpacks ‘x is white’ as ‘x has whiteness’, and ‘x is grammatical’ as ‘x knows grammar’.[6] More generally, a paronymous term – and, correspondingly, the being of what it names – will yield its substantive corollary in its definition upon analysis, and in this sense is reducible to it.[7]

A passage from Boethius’ longer commentary on Porphyry’s Isagoge proves illuminating on this point. Where Porphyry states his intention to bypass questions concerning the ontological status of genus and species,[8] Boethius comments:

‘The first of these [questions] is of this sort: Whatever the mind understands it either receives by a core meaning (intellectu) constituted in the nature of things and explicates to itself by a reason, in the case of that which is; or it portrays to itself in an empty imagining, in the case of what is not. Therefore, we ask, concerning the meaning (intellectus) of genus and others like it, whether we so understand species and genus as those things that are, and from which we grasp a true meaning (intellectum), or whether we are deceiving ourselves, when we form for ourselves things that are not with hollowed out imaginings.’[9]

In the above passage, Boethius is attempting to sustain two contrasts: the first, between what is and what isn’t; the second, between intellectus and imaginatio. Implicit alongside these two contrasts are several others: those between the inner and the outer, the true and the merely apparent, and the full and the empty. Each first member of the above dichotomies is parallel with every other, as is each second member. So intellectus in the above passage must correspond to what is, as opposed to what is not; and as what is fully true as opposed to what is empty and apparent. The contrast between the faculties of intellect and imagination partially fulfils this role, but only because of what those faculties themselves are oriented towards. Imagination takes its bearings from the changing realm of sensible things: hence what it imagines need not be when it imagines it; in some cases, it need not be at all. Hence, it is possible for an imagination to be empty, hollow, unfulfilled. Understanding, by contrast, is oriented toward an essential, stable core in material natures. Hence, to highlight the primacy of this dichotomy, upon which that between the faculties depends, I have translated intellectus in the above passage as ‘meaning’.[10]

The same sense comes through in Boethius’ commentary on the Peri Hermenias. Commenting on the passage at the beginning of the work explaining the relations between writing, speech, concepts, and things, he writes:

‘There are three things from which every debate and disputation is composed: the matters at hand (res), the meanings (intellectus), and the spoken words (voces). The matters (res) are what we grasp by measure of mind and distinguish by understanding (intellectu). The meanings (intellectus) are that by which we come to know the matters themselves. Spoken words are that by which we signify what we grasp of the meaning.’[11]

Translated thus, this passage can be disengaged from the metaphysics of representational realism[12] within which it is typically ensconced. The scope of res used here is broader than that typically ascribed to ‘object’: it refers to anything that can be talked about as the subject of an inquiry. These subjects grant a sense (intellectus1) to minds (intellectus2) which achieve a state of understanding (intellectus3) in grasping the sense. These meanings do not mediate between the mind and world of objects: rather, the matters themselves are meaningful; and, since meaning is, qua meaning, granted to minds, these matters are meaningful for understanding subjects. Hence, though we do speak of ‘what we have in mind’ (id quod intellectu capimus), such a designation is opaque if it fails to note that what we grasp is the meaning of the matter at hand, and that what we do in speaking is constitute this meaning (constituere intellectum) for another.[13]

[1] See Henry 1974, 88: ‘Hence it here appears that for Boethius one could speak of things being asserted in a certain fashion (e.g. denominatively, paronymously). Hence the whole sentence with which we are now concerned is perfectly coherent with the Boethian pattern, and the modern compulsion to insert quotation marks around ‘grammatico’/‘literate’ thereby removed.’

[2] BC 167D-168A:

Atque ideo quotiescunque aliqua res alia participat, ipsa participatio sicut rem ita quoque nomen adipiscitur, ut quidam homo, quia iustitia participat et rem quoque inde trahit et nomen, dicitur enim iustus.

[3] DG V = Henry 1974, 52 (alt.):

Qui dicit: omnis homo potest intelligi homo sine grammatica; et nullus Grammaticus potest intelligi grammaticus sine grammatica, nonne hoc significat quia esse hominis non indiget grammatica, et esse grammatici indiget grammatica?

[4] Hence, in no way can Anselm’s project be construed, a la Wittgenstein, as giving a linguistic analysis of the meaning of the term ‘God’. Cf. Gasper 2004, 107-143.

This attitude towards names remains standard even for Aquinas. cf. Archambault 2014, 185-90.

[5] Boethius makes the point forcefully in the case of the relation between arguing and argument at BDT 1174C:

Non vero idem est argumentum et argumentatio: nam vis sententiae ratioque ea quae clauditur oratio cum aliquid probatur ambiguum, argumentum vocatur; ipsa vero argumenti elocutio, argumentatio dicitur. Quo fit ut argumentum quidem virtus, et mens argumentationis sit atque sententia; argumentatio vero, argumenti per orationem explicatio.


An argument is not the same thing as argumentation: for both the ground of a judgment and the reason contained in a speech when something uncertain is proven are called ‘arguments’; while the actual speaking of the argument is called an ‘argumentation’. By this, it happens that an argument is in a way the power and principle as well as the meaning of an argumentation; while an argumentation is the explication of an argument through speech.

[6] These examples are used by Anselm in DG XIV. Here, knowing is regarded as a way of having, the appropriate one for things like grammar.

[7] So the primary sense of intellectum in fides quaerens intellectum cannot, pace Bencivenga, be identified with the state of satisfaction achieved by the subject who understands (though normally, nothing prevents this from being present as well—as, for instance, in VA I, xix), but rather is what the intellect has when it is in this state. See Bencivenga 1993, 33.

[8] BCP 82AB:

Mox de generibus et speciebus illud quidem, sive subsistant, sive in solis nudis intellectibus posita sint, sive subsistentia corporalia sint an incorporalia, et utrum separata a sensibilibus an in sensibilibus posita: et circa haec consistentia dicere recusabo.

I translate Porphyry’s passage as follows:

Now concerning genera and species: I won’t say 1) whether they subsist [in the realm of nature] or 2) are placed in separated ideas, unobscured [i.e. by admixture with matter]; or 1a) if subsisting, whether corporeal or incorporeal; or 2a) if separate, whether set [forth] in sensibilia or from them; and [I won’t speak] about how such things belong together.

Note that when translated thus, the notion that universals exist only in the mind of a human subject is not even mentioned as an option. The main options laid out in the first clause are that genera and species are 1) natural/physical or 2) supernatural/metaphysical, i.e. existing apart from the realm of earthly things.

[9] BCP 64, 82BC:

Quarum prima [harum quaestionum] est huiusmodi: Omne quod intelligit animus, aut id quod est, in rerum natura constitutum intellectu concipit et sibimet ratione describit; aut id quod non est, vacua sibi imaginatione depingit. Ergo intellectus generis et caeterorum cuiusmodi sit quaeritur: utrumne ita intelligamus species et genera ut ea quae sunt et ex quibus verum capimus intellectum, an nosmetipsos eludimus cum ea quae non sunt nobis cassa imaginatione formamus.

[10] This translation is not entirely novel, even in Anselm scholarship. See Holopainen 2007, 18, 21.

[11] BDIL 297B:

Tria sunt ex quibus omnis collocutio disputatioque perficitur: res, intellectus, voces. Res sunt quas animi ratione percipimus, intellectuque discernimus. Intellectus vero quibus res ipsas addiscimus. Voces quibus id quod intellectu capimus, significamus.

[12] Representational realism is typically regarded as a position in the philosophy of mind. What I call the metaphysics of representational realism, though, is the understanding of being presupposed in that position, the dominant traits of which are 1) the treatment of the distinction between mind and world, the mental and the physical as a distinction between two distinct spheres of reality (or, if ‘reality’ has already been co-opted to refer to one half of the dichotomy, the reader may choose another term ad placitum), and 2) the corresponding search for a medium by which this chasm is to be traversed, typically found in the idea.

[13] See BDIL 429C-431C, where the phrase is used throughout. Anselm uses it at DG XIV.

Part 4.

The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 2

Read part 1 here.

2 The Trivium at Bec in the Early 12th c.

The following entry is found toward the end of a library list from Bec abbey from the early twelfth century:

  1. In alio Martianus Capella de nuptiis Mercurii et philologie lib. II et de VII artibus editis ab eo lib. VII et commentum Remigii super eumdem IX lib. Priscianus de VIII partibus et de constructionibus II. Utraque rethorica II. Dialectice III. Utrumque commentum super Porphirium. Primum super catheg. Primum, secundum super periermeneias. Commentum super topica Ciceronis.[1]

Both the wording and content suggest these works, given as follows, were bound in a single volume. The first is the de nuptiis Mercurii et philologiae, here listed as two separate entries—the first two books by the title of the whole and the remaining seven by their content, ‘on the seven liberal arts,’[2] accompanied by a commentary of Remigius of Auxerre on the entire work. The work attributed to Priscian listed as de viii partibus et de constructionibus ii is the Institutiones Grammaticae. [3] One of the rhetorics listed is likely also his. The other may be the fourth book of Boethius’ De Differentiis Topicis, or the Pseudo-Ciceronian Herennian Rhetoric, with which Anselm was familiar.[4] The ‘Dialectic in three books’ is likely the first three books of the De Differentiis Topicis.[5] The remaining works of the list are also by Boethius: his commentaries on Porphyry’s Isagoge, Aristotle’s Categories and On Interpretation, and Cicero’s Topics.

While it cannot be established definitively, there are reasons to think this codex was already at Bec during Anselm’s time, and perhaps slightly earlier. Broadly, there is Lanfranc’s reputation among his contemporaries as one ‘‘raised up as a guide and a light to lead the minds of the Latins to the study of the trivium and quadrivium, which had fallen into neglect and profound obscurity’.’[6] Further, Anselm shows familiarity with at least some of the works on the list: He mentions Aristotle’s Categories in the De Grammatico,[7] and examples and vocabulary used in the same work suggest familiarity with both Priscian’s Institutiones and Boethius’ Logica Vetus commentaries.[8] That he would be familiar with the others would hardly be surprising.

Where the codex is on the list gives us another clue. The codex listed immediately before ours contains the Quaestiones Naturales of Seneca the Elder. This work is likely the proximate source for the phrase id quo nihil maius cogitari potest, ‘that than which nothing greater can be thought,’ so central to the Proslogion.[9] It is easy to imagine Anselm perusing this section of the library, devoted in large part to works of philosophy and more systematic theology, putting away his trivium codex, and fortuitously picking up the codex next to it that would provide the Proslogion with its key phrase.[10]

[1] Bekker 1885, 266.

[2] This division of the work has some justification, since it is only the first two books of the de nuptiis that tell the myth of the marriage of the god Mercury to philology; while each of the remaining books introduces a personification of one of the liberal arts, expounding their content, in the following order: Grammar, Dialectic, Rhetoric, Geometry, Arithmetic, Astronomy, and Harmony, i.e. music.

[3] The eight parts of speech are 1) nouns, 2) pronouns, 3) verbs, 4) adverbs, 5) participles, 6) conjunctions, 7) prepositions, and 8) interjections. Cf. Martianus Capella, lib. III, par. 281-88. In Priscian, each of these parts (with the exception of the interjection) is introduced by the formulaic phrase ‘n est pars orationis…’, followed by a further specification: nouns in book two, verbs in book eight, participles in book eleven, pronouns in book twelve, prepositions in book fourteen, adverbs in book fifteen, conjunctions in book sixteen; the final two books of the work, books seventeen and eighteen, shift from treating parts of speech to treating constructions.

[4] See Henry 1974, 87.

[5] In the middle ages, the first three books of the De Differentiis Topicis were occasionally regarded as topically distinct from the fourth: the first three books were on dialectic, while the final book was on rhetorical theory (see Leff 1974). This distinction in how the material was viewed provided the impetus for treating the first three books as a single work, and occasionally for excluding the fourth from a codex (e.g. Orleans, Bib. Mun. 265; Tours, Bib. Mun. 678) or for circulating the fourth separately (e.g. Paris, Bib. Nat. Lat. 16709). See Aristoteles Latinus 1939. Furthermore, no extant commentary on the De Differentiis Topicis from the 12th century or earlier comments on book IV of the text (Green-Pedersen 1984, 124). Our scribe’s loose tendency to name books by their contents rather than their titles – as he does for Priscian and Martianus Capella – combined with the wide circulation of the De Differentiis Topicis during Anselm’s time and the absence of any other dialectic in three books with which the entry could be identified, thus makes it probable the work referenced at Bec under the simple title de dialectica is the aforementioned one of Boethius.

[6] Southern 1963, 14. The works in the volume listed are standard for studying the trivium through much of the middle ages. See Abelson 1906, 73-74.

[7] DG XVI.

[8] Henry 1974, 92, 100.

[9] Seneca’s passage reads as follows: ‘Quid est Deus? Quod vides totum et quod non vides totum. Sic demum magnitudo sua illi redditur, qua nihil maius excogitari potest.’ Quoted in Southern 1963, 59, n. 3.

[10] This does not entail Anselm was conscious of this influence: he could have found the phrase in Seneca, forgotten about it, and then had it return to him weeks or months later. In this way, the assumption of the argument’s origin is quite compatible with that given by Eadmer at VA I, xix.

Part 3

The trivium at Bec and its bearing on Anselm’s program of faith seeking understanding, 1

The following is the first of six posts on St. Anselm’s Proslogion (c. 1077-78). This series reintroduces Anselm’s short theological work by i) making the interpretive assumptions underlying the recent reception of the text explicit; and ii) instead situating the argumentation in the text against the backdrop of earlier medieval contributions to thinking about language and logic – particularly those of Boethius (c. 480-524). This first post reproduces in large part an earlier post, but is included here for completeness. Enjoy the text. More importantly, may it profit you in wisdom and in charity.


Abstract. How should the phrase ‘fides quaerens intellectum’ be understood as a characterization of Anselm’s Proslogion? I argue that ‘fides’ and ‘intellectum’ should be understood not in accordance with the standard meanings we assign to ‘faith’ and ‘understanding’ today, but in accordance with their meaning in works used to teach grammar and dialectic at Bec in and around Anselm’s lifetime.

I begin by making the structures behind the usual understanding of ‘faith’ and ‘understanding’ explicit. From here, I detail the works of the Trivium used at Bec according to a 12th century library list, and provide justification for thinking these same works were used slightly earlier, in Anselm’s time. Next, I show how the terms fides and intellectus functioned in these works. Finally, I turn to the Proslogion and the exchange with Gaunilo to show how these considerations illuminate Anselm’s method in these works.

Keywords: Anselm of Canterbury; Medieval dialectic; Topics, Medieval interpretations of; faith; understanding; reason.

1 The Current State of Anselm Scholarship and the Usual Understanding of Fides and Intellectus

            In Anselm of Canterbury and the Desire for the Word, Eileen Sweeney describes the current state of Anselm scholarship thus:

‘There are real divisions in the interpretation of Anselm. […] Philosophers and systematic theologians carry off parts of his corpus, while those interested in spirituality take others. […] One of the most vexed questions in Anselm scholarship is its disciplinary location. Though the question of whether Anselm’s work is philosophy or theology is ultimately anachronistic, the extreme positions that have been taken on this question reveal something about how incompatible the elements of Anselm’s corpus seem to modern sensibilities.’[1]

According to the operative conception behind the aforementioned divide, to study religion, theology, or spirituality is to take faith as the object of one’s inquiry; while philosophy is that academic discipline most characterized by its connection with reason – and if reason and understanding, ratio and intellectus, are not identical, then the former at least provides the paradigmatic case of the latter.

            Fides, or ‘faith’, is understood in two ways: primarily, as a subjective state, that of believing a matter not objectively verified; secondarily, as the matter assented to. In the first way, the study of faith is part of the study of human subjectivity, and thus subordinated to psychology. In the second, to study faiths is to study the structures of propositional content assented to by those having faith in the first sense. If one structure is privileged as correct, then one is doing theology; if one remains agnostic about the correctness of any one religious system and instead studies properties of these structures for their own sake, then one is engaging in comparative religion.

A divide between subjective and objective senses similarly governs the meanings of intellectus, typically translated as ‘understanding’. This term can refer to: 1) the faculty of knowing; 2) the state of knowledge attained by the proper exercise of that faculty; or 3) the knowledge attained by that exercise. The faculty of understanding is often taken to be identical to the faculty of reason, while the state of understanding occurs when one can give sufficient reasons (i.e. known true propositions taken as premises) for what is understood. As in the previous case, the faculty retains priority over both the state and the object attained, and is usually identified with the human mind. Understanding understanding is, then, the province of psychology. Correlatively, understanding in the second sense is taken to be a psychological state; understanding in the third sense is dubbed ‘mental content’, and is studied, most often in philosophy of mind, under that heading.[2]

The relation between these states is taken as follows. Faith qua act of belief is an act of the will, marked by the subjective indeterminacy of the truth value of propositional contents taken as its object. The state of faith is, compared with that of understanding, incomplete. And the content of religious belief relates to that of understanding according to one of two models: on the first, stemming from Aquinas, truths of faith are above reason; on the second, stemming from Latin Avërroism,[3] truths of faith may contradict reason. In the first, faith acts as a supervaluation function filling in truth values undecidable by the machinery of reason alone. In the second, the data given to faith and reason conflict, and one must choose between the two.

            Fides quaerens intellectum, then, is most often taken to mean ‘subjective, willed believing-without-evidence seeking definitive proof that what it believes unknowingly is in fact the case’. If one takes the content of reason and faith to conflict, then the ‘without’ in the above definition can be strengthened to ‘against’. This sense is not terribly flattering. This does not, of course, mean it is an incorrect reading of the phenomena. But we should hesitate to think this was the picture Anselm endorsed when he gave fides quaerens intellectum as the working title to his Proslogion.

This essay aims not so much to refute standard interpretations as to deepen them. Standard interpretations of the Proslogion and its key terms are not so much incorrect as superficial: the meanings we assign primarily to fides and intellectus are, on Anselm’s analysis, derivative. If reading Anselm differently on this point proves fruitful, perhaps doing so will provide us with a way of seeing the matters themselves differently as well.

[1] Sweeney 2012, 6-7. Sweeney cites Gilson and Barth as examples of ‘philosophizing’ and ‘spiritualizing’ interpretations of Anselm’s work. See Barth 1960, 55-59; Gilson 1951, 26.

While Sweeney should be commended for drawing attention to this division, I do not think her work successfully transcends it. Sweeney sees Anselm’s thought as a coincidentia oppositorum of logic and rhetoric, reason and desire. ‘Anselm’s project in the Proslogion is one that Anselm himself views as both necessary and paradoxical’ (Sweeney 2003, 17). The problem with such a reading is that it merely entrenches the dichotomies and assumptions brought to the table by the above mentioned rival groups; the attempt to balance these opposing elements without questioning their internal content effects a mere reconfiguration of concepts, instead of leading to an improved understanding of those concepts.

[2] To paint the situation a bit more accurately, what is studied is theories of mental content, thereby effecting a redoubling of the primacy of the subject putting forth these theories.

[3] At least, this is how the story is told. For reasons to think the story isn’t quite as simple as its retelling, see Klima 1998.

Go to part 2