(i) Overview of the interpretative tradition

We can divide the interpreters of the Proslogion into a number of overlapping camps: (1) those interested in it as an instance of one or more ontological arguments; (2) those who make the same kind of objection to the argument of the Proslogion as Gaunilo; (3) those who argue that the Proslogion is a theological work which does not seek to deliver a philosophical proof; (4) those who seek to defend an improved version of Anselm's argument; and (5) those for whom the Proslogion is simply the starting point of their own arguments or proofs.

(1) It is generally agreed that Anselm's philosophical fame is based on three chapters of the Proslogion, i.e. chapters 2 to 4. The historian, R.W. Southern, pointed out that the rest of the text “has been, from the point of view of Anselm's own thought, unduly neglected”.Footnote ⁵ This is particularly true of those who classify the argument of the Proslogion as ‘ontological’. These ‘ontologisers’ constitute the most broadly based and influential group of interpreters. They have transferred to Anselm's argument Kant's ‘ontological’ label, originally applied to the arguments of Cartesian and Wolffian rationalism.Footnote ⁶ The result has been that the thoroughly unanselmian arguments of a later epoch continually determine the reading of the Proslogion and the concentration on chapters 2 to 4. In so far as Anselm's philosophical fame is based on one or more ontological arguments found in Proslogion 2 to 4, it is based on a misunderstanding. Small wonder, then, that there is little effort made to understand the text of the Proslogion as a whole. And yet, Anselm gives a clue as to the unitary character of the text by referring to the argument presented there as unum argumentum. Footnote ⁷

A classic case of the predisposition to misunderstand is to be found in Hans Küng's book, Does God Exist?. Having linked together the names of Anselm, Descartes, Leibniz and Wolff, Küng characterises the ‘ontological proof’ thus:Footnote ⁸

This may or may not be an adequate account of ontological proofs. It has virtually nothing to do with Anselm's argument. In fact, Küng like the other members of this group are not really interpreters of Anselm at all. They generally give no indication of being acquainted with the work of Anselm other than at second hand. They are interested in certain arguments concerning the existence of God. To this group Anselm's meaning is unimportant. Thus, one of the more recent commentators, Graham Oppy, in a wide-ranging survey of English language discussions of the ontological arguments, states openly with reference to the Proslogion: “I am not interested in pursuing these tiresome exegetical questions.”Footnote ⁹ However, Oppy does not leave it there, but goes on to make assertions about Anselm's position, which are dependent on exegesis.Footnote ¹⁰

(2) The followers of Gaunilo are found in all the categories. They, like Gaunilo, misunderstand or distort Anselm's argument. They choose to replace the formula, ‘that than which a greater cannot be thought’, by another phrase, such as aliquid omnibus maius (something greater than everything),Footnote ¹¹ which they take to be synonymous with Anselm's formula. Anselm had informed Gaunilo, the first to misunderstand him in this way, that such language did not convey what was meant by ‘that than which a greater cannot be thought’.Footnote ¹² There appears to be little excuse for later writers, such as Küng, who repeat the error. Yet, even a sympathetic interpreter such as T. Morris, is able to say: “The Anselmian conception of God is that of a greatest possible, or maximally perfect, being.”Footnote ¹³ Although Anselm's conception of God contains this idea, it is not identical with it, for God is greater than any greatest being I can possibly think: “Ergo domine non solum es quo maius cogitari nequit, sed es quiddam maius quam cogitari possit.” (Proslogion, 15) [My translation: “Therefore, Lord, not only are You that than which a greater cannot be thought, but You are [also] something greater than can be thought.”]

The characteristic feature of the followers of Gaunilo is that they do not understand the uniqueness of Anselm's argument. They see little problem in rewording it, so that it no longer concerns ‘that than which a greater cannot be thought’. This particular phrase is at the heart of the argument. They do not grasp that an argument concerning ‘that than which a greater cannot be thought’ does not work for ‘that than which a greater can be thought’. This is why, for Anselm, Gaunilo's objection, based on the Lost Island, misses the point.

(3) The theologisers include Stolz and Barth. They seek to rescue Anselm from the grip of philosophers and to reinforce their anti−philosophical viewpoint. Stolz seizes on the prayerful nature of the Proslogion as an indication that it is a work of mystical theology and thus not an argument intended to convince. He writes: “Nothing is more absurd than to see a philosopher in the author of the Proslogion.”Footnote ¹⁴ Stolz argues that there are two styles of writing to be found in Anselm, the doctrinal/dialectical and the prayerful, and concludes that since the former do not contain prayers, then the Proslogion must belong to the second group. Consequently it does not involve the dialectic of a philosophical proof of God's existence. Stolz's understanding of the structure and purpose of this work arises out of his belief that it does not contain a philosophical proof. It allows him to overlook the glaringly obvious, that Anselm clearly states he wishes to prove God's existence. It is a sobering thought that an otherwise serious reading of the text can overlook this fact.

Barth too overlooks the obvious, omitting Anselm's reference to the proof−nature of the argument when he quotes the relevant passage from the Preface of the Proslogion. He writes:Footnote ¹⁵

The passage Barth quotes should read:

Once we are aware of this, it is surprising to read:Footnote ¹⁷“What is set out in Prosl 2–4 is first described as a ‘proof’ (probare, probatio) by Anselm's opponent Gaunilo” and “[Anselm's] own particular description of what he is doing is not probare at all but intelligere”. According to Barth, the argumentum should not be identified with the proof of chapters 2 to 4. This is correct. But Anselm's argumentum is meant to be a proof in a way that cannot be accommodated by Barth's interpretation. The ‘proof’ of chapters 2 to 4 is not the argumentum, but according to Anselm the argumentum is a proof.

More recently, D. Hogg has argued that it “is somewhat unfair to Anselm to assume he is offering a proof here. After all, proving so many different things by one single argument, by five little words, seems just a bit ambitious.”Footnote ¹⁸ Hogg seeks support for this position in the view that the verb probare is not always to be taken as referring to proof, but is “commonly used to mean to probe, to search, to test or to prod”.Footnote ¹⁹ There is some truth in this as a general comment on the meaning of probare, although it is important to note that to refer to something as probatus does not simply mean that it has been probed or tested, but that it has also been found not wanting. More to the point, Anselm's use of the verb, probare, throughout his writings carries the clear meaning of ‘to prove’.Footnote ²⁰

Thus in Responsio, 1, he writes:

And again, in Reponsio, 6, he writes:

In both cases, the literal translation of probare as ‘to prove’ conveys most successfully Anselm's meaning. In the Proslogion, Anselm is setting out to prove the existence of the God of Catholic belief. He really is that ambitious.

(4) A diverse group consists of those who would defend versions of Anselm's argument in some form. These are generally philosophers with their own philosophical programme. Hartshorne, for instance, is concerned with ‘process’ philosophy and theology. He criticises the failure to read Anselm –“the extraordinarily poor scholarship, in which great scholars have indulged.”Footnote ²¹ It is his view that there are two arguments in the Proslogion: a weaker one in chapter 2, and a stronger one in chapter 3. He writes that Proslogion 2 “is but a blundering preamble or unlucky false start in the development of the Proof”.Footnote ²² These assertions, which overlook the significance of Anselm's commitment to unum argumentum, would suggest that Anselm did not understand what he was doing. In fact, much of the more recent scholarship has addressed these different ‘ontological’ arguments to be found in Anselm. The contributions of Hartshorne and other philosophers, such as MalcolmFootnote ²³ and Plantinga,Footnote ²⁴ who seek to set Anselm straight, are not directly relevant to Anselm scholarship, though they may be interesting in their own right.

(5) The final group is represented by those who use Anselm to develop their own positions. This is true in a sense of all of Anselm's interpreters. M. Blondel, in the late 19^th and early 20^th Century, was particularly interested in the argument of the Proslogion.Footnote ²⁵ However, his work, if viewed as an interpretation of Anselm, fails. It is not Anselmian, because it departs too far from the text and thought of Anselm. Blondel and the other French-speaking reflexive philosophers recognised a similarity between the argument of the Proslogion and their own arguments, but did not develop their arguments fully in relation to the text of the Proslogion. In his summary of such arguments, A. Forest writes: “the reflexive proof is authorised by St. Anselm, rather than in agreement with him”.Footnote ²⁶

The appeal to Anselm as an authority explains much of what has gone wrong with Anselm scholarship. Whatever the motives of these and other scholars in using Anselm as a channel through which to pass their own ideas, and as interesting as these ideas may be in their own right, they have done a disservice to students of Anselm. It seems impossible for the student to glimpse the real Anselm in the muddied waters of secondary sources.

I have tried to indicate, if only very briefly, that the history of the interpretation of Anselm's argument is littered with contributions from those who have not seen the obvious in the text. Thus, B. Hendley sees a response to the rational non-believer where Anselm claims to be responding to the catholic, rather than to the atheist, in the Responsio to Gaunilo. He argues that we need to understand Anselm's purpose and the audience he is addressing in order to evaluate his argument, stating:Footnote ²⁷“To do so we must first turn to the text.” Yet, he fails to see what is there in the text: “sufficere mihi potest respondere catholico.”[My translation: “may it suffice for me to reply to the Catholic.”]Footnote ²⁸ So, clearly, asking people to look at the text is no guarantee of progress.

(ii) The effect of interpretation on the translation of the Proslogion

It would be impossible to undertake a wholesale review of the history of the relation of translation and interpretation with reference to the Proslogion in this short paper. I will therefore take as a paradigm case a passage from Proslogion 4, “Quod qui bene intelligit … .”

There is in my view good cause for taking this particular passage as significant. It is a summary of the discussion contained in chapters 2, 3 and 4 of the Proslogion – those chapters, which have been the focus of attention of interpreters. The phrase, “whoever understands this properly”, is a clue to the significance of this summary. It tells us in a nutshell what Anselm has been doing in chapters 2 to 4.Footnote ²⁹ As such, we can expect different interpretations of what Anselm has been doing in these chapters to have an impact on the translator's reading of this passage. In fact, when we look at the original and compare it with the translations, we find a remarkable set of variations.Footnote ³⁰ The text in question (with my translation) is as follows:Footnote ³¹

The translations of Ward,Footnote ³² of McGill,Footnote ³³ of Hopkins and Richardson,Footnote ³⁴ of FairweatherFootnote ³⁵ and of ShannonFootnote ³⁶ introduce God into the second step, translating the neuter ‘id ipsum’ by ‘God’ or ‘he’. This makes a tautology of Anselm's reasoning: God cannot conceivably [in thought] not exist; therefore God cannot be thought not to exist. ‘Id ipsum’ is translated by ‘il’ in the French translation of Koyré,Footnote ³⁷ and by ‘er’ in the German translation of Schrimpf.Footnote ³⁸ Michel Corbin on the other hand translates ‘id ipsum’ as ‘cela même’, but is still forced to use ‘il’ to refer to what it is that cannot be thought in step 2.Footnote ³⁹ The translation of Thomas Williams, which seeks to be literal within the constraints of contemporary English idiom, translates ‘id ipsum’ as ‘this being’, rather than ‘this same being’.Footnote ⁴⁰ Both are acceptable in English idiom but the latter is more literal. Williams in fact loses some of the clarity of the argument, because he unnecessarily avoids being literal.

In all the English translations there is a blurring of the issue brought about by the literal translation of ‘quod’ as ‘this’ in Step 2. In the Latin, the referent of ‘quod’ is clearly ‘id quo maius cogitari non potest’. In the English translations, the referent of ‘this’ appears to be the statement ‘God is that than which a greater cannot be thought’, leaving the second step entirely ambiguous. Any translation that attempts to present Anselm's text as clearly as possible must make clear the antecedent of ‘quod’, but in doing so will need to move away from literal translation.

Further confusion arises from the treatment of the word, ‘sic’, in our passage. S. N. Deane translates the first occurrence as ‘so truly’, rather than as ‘so’ or ‘in such a way’.Footnote ⁴¹ Similarly, Dom Pouchet translates ‘sic’ as ‘si réellement’.Footnote ⁴² Charlesworth translates the first instance of ‘sic’ as ‘so’ and the second instance as ‘in such a way’.Footnote ⁴³ Dom Schmitt translates the first instance of ‘sic’ as ‘so’ and the second instance as ‘auf diese Weise’.Footnote ⁴⁴ These different translations of ‘sic’ are unnecessary. Furthermore, the use of ‘so’ here carries connotations of the superlative (see Deane and Pouchet), which distracts from Anselm's meaning.

Hopkins looks closely at the usage of ‘sic’ in this passage.Footnote ⁴⁵ He criticizes both CampbellFootnote ⁴⁶ and SchufreiderFootnote ⁴⁷ for failing to appreciate that, although ‘sic’ is being used adverbially in its first occurrence in our passage, it is used adjectivally in its other occurrence (since it is not correlated with an ‘ut’). He argues that, in English, the final sentence should read as follows:

and not (as Schufreider translates it),

In fact, Schufreider amends his translation to take account of the criticism of Hopkins.Footnote ⁴⁸ Hopkins’ reasons for translating as he does appear plausible. However, there are many possible adverbial uses of ‘sic’, which do not require that it be ‘correlated’ with ‘ut’, and here it is perfectly reasonable to translate ‘sic’ as “in such a way”, with the meaning of “in the manner just specified” (or “in the same way”– see my Translation 1). The meaning of this passage can only be resolved, if one understands its role as a summary of Anselm's argument so far, i.e. as an interim report on progress. One can get this right whether one follows Hopkins or Campbell and Schufreider in translating ‘sic’.

It takes Anselm the remainder of the Proslogion to establish Step 1, the identity of God with ‘that than which a greater cannot be thought’. This explains why Anselm does not leave the argument at this point. Step 3 should be understood as a conditional statement. If one understands that God is such a being or exists in the same way as such a being, then one understands that God cannot be thought not to exist. The point then is to establish the identity, which allows this conclusion to follow. This whole process is misunderstood if one introduces God into Step 2. The lack of rigour exhibited by translators of this passage is understandable, once we realise that their translations create no problems for the tradition of interpretation, because according to this tradition Anselm's argument is already complete by the time he reaches the end of chapter 4. Such translators have reinforced the position of the interpreters, who have looked for and found confirmation of their views in the place where their views have determined what is to be found.

Given the weight that Hopkins places on the translation and interpretation of the word, ‘sic’, it is interesting to note his criticism of Schmitt for concluding that Anselm, in writing the Proslogion, “weighed each word exactly”.Footnote ⁴⁹ In fact, an important hermeneutic principle underlies Schmitt's approach: that, in translating a work such as Anselm's, to be imprecise or free in translation is to do a disservice to the ideas being articulated there. Hopkins misses the point, in referring to Schmitt's position as a conclusion.Footnote ⁵⁰ The basis of Schmitt's approach is a spirit of empathy, and he is operating with the concomitant methodological assumption here. As Nida points out: “Even if the translator possesses all the necessary technical knowledge, he is not really competent unless he is also a truly empathetic spirit.”Footnote ⁵¹ This empathy makes a practical difference to the outcome of the translator's work. It is best expressed in the words of Karl Rahner:

So it is that even the translation of an apparently straightforward term such as ‘unum argumentum’ becomes problematical, if one does not want to take Anselm's claim as philosophically serious. It is possible to discern ‘one argument’ from which all that Anselm says of God follows. The argument is the proof that God and ‘that than which a greater cannot be thought’ are to be identified.Footnote ⁵³ Establishing the existence of ‘that than which a greater cannot be thought’ is the cornerstone of establishing this identity. (Otherwise, Anselm would not be addressing his argument to the God in whom he believes.) It is in this sense alone that it is possible to agree with Barth that Proslogion 2 to 4 represents the ‘Haupttext’.Footnote ⁵⁴ Nevertheless, chapters 2 to 4 are not to be identified with the unum argumentum. Hopkins takes Anscombe to task for misunderstanding what the unum argumentum is, and for assuming that it refers to a very short single argument in Proslogion 2.Footnote ⁵⁵ However, Hopkins appears to commit the error he ascribes to Anscombe, that a single argument must be short, taking the view that, since Anselm's argument is not short, it is not a single argument. Thus, he does not translate unum argumentum as ‘single argument’, but as ‘single consideration’. And yet, in the very same sentence, he translates the same noun as ‘argument’.Footnote ⁵⁶

(iii) Can there be a definitive interpretation or translation of the Proslogion?

Anselm is medieval, and so, of course, he predates modern philosophy and theology. However, it is important to note that he also predates all the other major medieval philosophers and theologians. He knew less about classical thought than his successors did or we do. His written works derive from middle age, although he had been a student and teacher from a young age. We do not see the development of his thought. Thus, we do not fully understand his relation to Augustine, even though the influence of Augustine is visible throughout his writings.Footnote ⁵⁷ McIntyre writes that Anselm's relation to Augustine “is too complex to merit for him the title of Augustinian”.Footnote ⁵⁸ Whilst De Lubac suggestsFootnote ⁵⁹ that there is a significant difference between Anselm and Augustine in their view of the “intelligence de la foi”. For Augustine it is “une étape sur la voie qui conduit de la foi obscure à la vision”. For Anselm, understanding does not bring with it the vision of God, nor bring him any nearer to it. Dom Schmitt arguesFootnote ⁶⁰ that Anselm de-platonises Augustine. And Heinzmann statesFootnote ⁶¹ that Anselm's anthropology is “spezifisch christlich und nicht neuplatonisch”.

My view is that we can refer to Anselm as an Augustinian. The problem is what we mean by that. We may be misled if we think that Anselm followed anyone in anything except in his own particular way. Anselm is driven by the exigencies of dialectic within the parameters of Catholic teaching. He is not opposed to dialectic, but to the ‘heretics of dialectic’.Footnote ⁶² The parameters and the conceptual vocabulary are set by the Church, Scripture and the Church Fathers, especially Augustine, but Anselm gets to the conclusion and employs the vocabulary in his own way. It is possible to argue that, perhaps because of his own originality, Anselm is in the process of breaking away from Augustinianism (understood as employing the assumptions and method, as well as the conclusions, of Augustine). He certainly marks a step in that direction.

In comparing the argument of Anselm in Monologion, 1, with that of Augustine in De Trinitate, VIII, III, 4, Katherin Rogers indicates the difficulties faced by someone who wants to show (as she seeks to) that Anselm “is a powerful spokesman for the persuasive system which is Augustinian Neoplatonism”.Footnote ⁶³ She writes:Footnote ⁶⁴“There are indubitable parallels between Anselm's argument and Augustine's … . Augustine speaks explicitly of participation, so it is reasonable to suppose that Anselm adopted the doctrine.” In fact, it is interesting how Anselm appears to have stripped his vocabulary of such language. And, on the few occasions he uses it, a Neoplatonic interpretation is not required to understand Anselm's meaning.Footnote ⁶⁵

Our understanding of Anselm is coloured by the fact that, in a sense, we know too much about philosophy and theology before and after Anselm's time and not enough about Anselm's own philosophical and theological experience. It is possible to surmise something about the books Anselm read prior to writing the Proslogion, from his writings and from later, 12^th Century library catalogues at Bec.Footnote ⁶⁶ But even when we can identify a possible source we are unable to be sure what Anselm made of it, how he assimilated it and how it worked out into his own ‘system’. Anselm is then a difficult thinker for us, because he operates in a world that is more closed to us than we assume.Footnote ⁶⁷ No interpretation or translation can be definitive, because we do not have the required level of access to Anselm.

(iv) Conclusion: some ground rules

Does all this matter? Yes, if we are interested in what Anselm was getting at in the Proslogion. Of course, it is legitimate to use Anselm to develop one's own ideas, and the above is not intended to suggest that so much effort has been in vain. However, if we claim to be making a contribution to Anselmian scholarship, then the rules that govern our reading of the text must be stricter than those that apply to readers for whom the accuracy of interpretation is secondary. For, although there may be no definitive interpretation or translation of the Proslogion, it is possible to avoid errors of interpretation and translation by remaining within the constraints of the text. We can draw some guidance from the above discussion to assist us in remaining within these constraints. To this end, and in conclusion, the following are offered as a minimal set of ground rules for the reading of Anselm, with which it is hoped any serious reader of Anselm would agree.

1 Read the (whole of the) Proslogion, before you decide that you understand what Anselm is saying there. It is a rare student who manages this feat. In order to support the reading of the entire text, teachers should avoid the almost universally adopted practice of producing handouts containing just chapters 2 to 4 of the Proslogion.

2 Assume that there is a single argument in the Proslogion, and that this notion is important to the understanding of Anselm's meaning in the Proslogion. But, do not assume that Proslogion 2 to 4 is the unum argumentum. It is certainly possible to discuss these chapters apart from the rest of the text, but this is because posterity (starting with Gaunilo) has produced a particular set of issues quite independently of Anselm's intentions. I suspect Anselm's otherwise praiseworthy decision to insist on the inclusion of Gaunilo's response in the manuscripts of the Proslogion along with his Responsio is partly to blame for this failure to address the whole of the text.Footnote ⁶⁸

3 Avoid anachronism. Place Anselm in his historical context. This means in practice accepting that he probably did not put forward an ontological argument, interesting as the subsequent debate may have been. Forming one's view of Anselm's argument by engaging with practitioners of later arguments and debates is likely to lead the reader astray.

4 Do not assume that Anselm was an Augustinian or a Platonist or the father of scholasticism or a rationalist or a mystic or a philosopher or a theologian in any sense that has accrued to these terms since Anselm's time. And do not base your reading on any such assumption.

5 Do not use a translation unless you have to. If you must, then use several, preferably in more than one language. This will provide an insight into the variety of ways the text can be understood or even misunderstood. And when you do translate, translate literally, unless it really is impossible to do so. As we have seen, even an apparently innocuous freedom in translation carries the risk of distorting the meaning of the text.

6 Finally, follow Rahner's heuristic principle. Assume there is a philosophical sense in what Anselm writes, and be open to the possibility that a failure to make sense of what he writes may indicate a failure in the reading of the text.