Did Anselm Define God? Against the Definitionist Misrepresentation of Anselm’s Famous Description of God

Anselm of Canterbury’s so-called ontological proofs in the Proslogion have puzzled philosophers for centuries. The famous description “something / that than which nothing greater can be conceived” is part and parcel of his argument. Most commentators have interpreted this description as a definition of God. We argue that this view, which we refer to as “definitionism”, is a misrepresentation. In addition to textual evidence, the key point of our argument is that taking the putative definition as what Anselm intended it to be – namely a description of a content of faith – allows getting a clear view of the discursive status and argumentative structure of Proslogion 2–4, as well as making sense of an often neglected part of the argument.


Exposition: The Definitionist Versus the Thesist View
The so-called ontological proofs in Anselm of Canterbury's second major work, the Proslogion [= P], are among the most frequently discussed arguments of all time in the history of philosophy. Anselm himself considered the arguments so important that, when he prepared the second edition of P, he decided to repeat the text of chapters 2-4 between the full text and the additional material from his famous discussion with Gaunilo. 1 As he put this repetition under the headline that gap, God and the/a famous object must be linked or even identified in some way. The Anselmian belief that God is that than which nothing greater can be conceived seems to be such a bridging principle. But what is its argumentative status in the context of the ontological proofs? How is it justified?
Most Anselm interpreters have a natural inclination to take this belief to be a definition. Depending on whether the word "God" is taken as an individual constant or a predicate, 7 a more precise form of such a definition will probably look like this: (D1) God := that than which nothing greater can be conceived. (D2) x is God :↔ x is something than which nothing greater can be conceived.
Most of the commentators who suggest such a definition rely on the fact that when Anselm prays for insight into what faith holds about God's existence and nature (P2.1), he states that according to faith God is something than which nothing greater can be conceived: P2.2: "Et quidem credimus te esse aliquid quo nihil maius cogitari possit." "And indeed we believe You to be something than which nothing greater can be thought." Hence, they read this sentence as (a textual basis for) defining God, the addressee of "You," as the/a famous object. 8 Reading P2.2 as a definition of God is not the only way to bridge the gap between God and the/a famous object. Bear in mind that Anselm follows a distinctive procedure with respect to God's existence: First, he prays for insight into this element of faith, then he provides his rational proofs, and finally he thanks God for the illumination granted in the form of rational insight. Why not suppose that Anselm pursues the same procedure-prayer for understanding, rational argumentation, thanksgiving-for the second element of faith included in his prayer as well? He has explicitly included the bridging principle, that God really is "what we believe" [him to be] (hoc es quod credimus), in his prayer for understanding. One would therefore expect divine illumination to be reflected in a proof of this principle as well. And, in fact, we will see that there are candidates for such a proof in a frequently neglected part of the Sumptum. According to this view, then, the bridging principle is a thesis, not a definition. 7 The literal text of P provides evidence for both readings. In P2.2 ("And indeed we believe You to be something than which nothing greater can be conceived"), the personal pronoun "You" refers to God. Substitution of "God" for "You" thus strongly suggests that "God" should be taken as an individual constant, since that is by far the most common usage of "you". For an alternative usage, see footnote 49. In contrast, in P2.3 it is asked whether there is no such nature (aliqua talis natura) because the fool says in his heart, "there is no God"-which suggests that "God" is a predicate which expresses a nature. We will not decide on this issue here. 8 If "God" is read as an individual constant, the "something" in P2.2 must change into a "that" in order to arrive at a formally correct definition of "God" such as (D1). If "God" is read as a predicate, the "something" can be kept, and the definition will read: "Something is (a) God :↔ it is something than which nothing greater can be conceived." We will take up this discussion in Sect. 3 below. Hence, there are two opposing views with respect to the interpretation and reconstruction of the Sumptum arguments. For convenience, we call them "definitionism" and "thesism." Definitionism bridges the gap between God and the/a famous object by reading P2.2 as a definition; thesism holds that the bridging principle is a thesis Anselm argues for.
In this paper, we set out to show that definitionism is mistaken. The most powerful of our arguments is that reading P2.2 in the definitionist way precludes getting the overall structure of Anselm's arguments in P2 and P3 right. Hence, in the end, the somewhat academic-seeming question of whether or not P2.1-2 is to be read as a definition provides a key to the centuries-old puzzle of what Anselm's argument really is. 9 In what follows, we first describe the definitionist view in some detail and present the main arguments in its favor (Sect. 2). Then, we criticize definitionism by confronting it with some counterevidence and by questioning the strategies used to make it plausible ("plausibilizations" for convenience, Sect. 3). Next, we refute definitionism by showing how abandoning it in favor of thesism helps to understand what Anselm's argument really is (Sect. 4). If thesism is right, there should be arguments for the thesis that God is the/a famous object. Hence, as a touchstone for thesism, we offer reconstructions of such arguments from a neglected passage of P2-4 (Sect. 5). Finally, we turn to an explanation of why so many knowledgeable commentators have been inclined toward the definitionist misrepresentation (Sect. 6). The last section of the paper summarizes our findings and outlines newly emerging problems that invite follow-up (Sect. 7). In the Appendix, the text of the Sumptum is printed for convenience.

Definitionism and the Meaning of "Definition"
As already mentioned, the second sentence of the opening prayer of the Sumptum, P2.2, reads: P2.2: "Et quidem credimus te esse aliquid quo nihil maius cogitari possit." "And indeed we believe You to be something than which nothing greater can be thought." This sentence is part of a prayer in which God is addressed by the personal pronoun "You." It reproduces an article of faith about God, namely that God is 9 We think that this is indeed the question philosophers have discussed for centuries now: what Anselm's argument really is. To be sure, pursuing that question does not presuppose that there is one and only one best possible interpretation of his text. What one considers to be Anselm's real argument depends inter alia on the hermeneutical decisions one makes in the course of interpretation. Is there even such a thing as his real argument? something than which nothing greater can be thought. According to definitionism, this article of faith is to be read as a definition of "God": God is defined as that/something than which nothing greater can be thought. Definitionism "is almost universally held amongst recent commentators and interpreters" of Anselm. 10 It is the majority view.
Two senses of "definition" must be kept apart: the modern sense and the Aristotelian-Boethian [= AB] sense. 11 Definitions in the modern sense are clarifications of meaning. They are descriptions of what a term de facto means in a given context, stipulations or prescriptions of what it shall mean in some context, or a mixed form, (conceptual) explications, which declare the meaning of a term by considering and modifying its de facto use. In contrast, to define something in the AB tradition means to integrate a thing into a hierarchy of genera and species by referring to a genus proximum and a differentia specifica. 12 In medieval logic, definitions in the AB tradition have come to be called "real definitions", while definitions in the modern sense are roughly what the medievals called "nominal definitions." How are the two senses of "definition" related? Is a real definition in the AB sense a definition in the modern sense at all? We hold that in rare cases real definitions can be used as definitions in the modern sense, though in most cases they cannot. In most cases, real definitions do not fix the meaning of a term but express a thesis about the position of an entity in a given hierarchy of genera and species, somewhere between the first substances and the highest genera. Only in exceptional cases may someone point to a specific position in such an ontological hierarchy in order to fix the meaning of a term. Similarly, in exceptional cases definitions in the modern sense may also serve as real definitions, but usually they do not, since modern definitions do not usually presuppose an objective hierarchy of genera and species. So, real definitions and definitions in the modern sense are two quite distinct conceptions with only a small overlap.

The Main Advantage of Definitionism
The main advantage of definitionism is that it allows for quick finalization of the P2-3 arguments. The alleged definition is the bridge from the (necessary) existence of the/a famous object to the (necessary) existence of God. In case of P2, the final steps read roughly as follows. Note that, obviously, not all instantiations of the following schema constitute valid arguments 13 : (1) Since A/the famous object exists in the intellect and in re.
(Preliminary concl. of P2) Taken as an argument for the definitionist reading of Anselm's P2.2, we call this the finalization argument: reading P2.2 as a definition allows for quick and easy finalization of the arguments in P2 and P3 as proofs of the existence of God.

Examples of Definitionism in the Literature
We have already mentioned that definitionism is the majority view and that we are going to reject it. We will now illustrate definitionism with some examples from the literature, if only to dissolve any suspicion of its being a straw man.
When definitionist commentators on Anselm's argument read the bridging principle in P2.2 as a definition, some of them seem unaware of the need to make this reading plausible. Others offer at least some plausibilizations. To start with, here are some examples of the first group, who do not defend the plausibility of their definitionist reading. 13 Whether they result in valid arguments depends on several choices: whether "God" is taken as an individual constant or a predicate, whether the famous description is used in the definite (the/id) or the indefinite (a/aliquid) version, and whether certain presuppositions of definitions are fulfilled or not (for example, whether a non-conditional definition of an individual constant presupposes existence and uniqueness).
Thomas Aquinas refers to Anselm's argument inter alia in Summa theologiae I.2.1, when he discusses whether God's existence is known through itself (per se notum). 14 The second objection of this articulus is that the name "God" signifies that than which something greater cannot be conceived. In the corresponding second resolution, Aquinas counters that not everybody understands "God" in this way, for some even hold that God is a body. One might read Aquinas as opposing the thesis that the famous description correctly determines the meaning of the term "God." In doing so, he presupposes a (common? well-known?) definitionist reading of Anselm in the modern sense of "definition." The definitionist account is also dominant in most accounts of the history of philosophy. Here are some examples. In their standard work on the history of scholastic philosophy, Etienne Gilson and Philotheus Böhner recapitulate Anselm's P2 argument. Perfectly in line with the fides quaerens intellectum program, they interpret the opening passage P2.1-2 as a petition for understanding of faith, pointing out that, according to faith, God is so great that nothing greater can be conceived. They then proceed by calling this a "definition bestowed on us by faith." 15 In his account of medieval philosophy, Paul V. Spade begins his description of the Sumptum argument with "(1) By 'God' we mean 'the conceivably greatest'" and then comments on this, saying: "(1) is just a definition which expresses a-specific-property of what we customarily understand by 'God'." 16 Kurt Flasch wants to restrict his discussion of the ontological argument in his [1982] to "a few hints helping to avoid the worst misunderstandings." In giving his second hint, he expresses the definitionist (mis)understanding by saying that Anselm's proof "starts out from a meaning of the word believers and atheists can agree upon in order to produce proofs or disproofs of the existence of God. Anselm proposes to explain the word 'God' as 'that than which a greater cannot be thought'." 17 In their history of logic, William and Martha Kneale discuss the proof of the modal reality theorem (P3. [1][2][3][4][5] and claim right at the beginning of their discussion: "God, he [= Anselm, GS/CT] tells us, may be defined as that than which nothing greater can be conceived, and according to this definition God exists so truly that he cannot even be thought not to exist." 18 Definitionism is so established that it is almost the exclusive perspective in many well-known handbooks of philosophy. It seems to be the predominant view among philosophers. 19 It has also been the majority view of Anselm researchers. It is held by Anselm editors 20 as well as interpreters. 21 Morscher, for instance, presents an 18 Kneale/Kneale [1986], 201. 19 For instance, the entry "Definitions" in the Stanford Encyclopedia of Philosophy (Gupta [2021]) begins with a list of allegedly undisputed prominent examples of defining. There the author asserts: "The key step in Anselm's 'Ontological Proof' for the existence of God is the definition of 'God'". 20 Charlesworth [1965], 54-55. 21 For instance, see Visser/Williams [2009], 88, who do not explicitly call the bridging principle a "definition" but a "characterization of God" which they use as a premise of their reconstruens argument; Anscombe [1985], 32, says explicitly: "Descartes defines God as supremely perfect being. Anselm, as: that than which nothing greater can be conceived." 14 Other places include: In Sent I.3.1.2, arg. 4; SCG I.10-11; I.43; De ver. 10.12. 15 Gilson/Böhner [1937], 281, our transl. 16 Spade [1995], 93, our transl. 17 Flasch [1982], 207-208, our transl. informal reconstruction of the P2 argument, step 6 of which reads: "God is (is identical to) that than which nothing greater can be conceived (Definition)." 22 Further examples of the definitionist position are presented by Oppy. 23

Arguments for (the Plausibility of) Definitionism
Since there is no indication of defining vocabulary in P2.2, reading it as a definition needs to be made plausible. The need for such a plausibilization appears even more pressing insofar as the definitionist reading leads to differing treatments of the two beliefs mentioned in P2.1-2. Prima facie, it is implausible to treat one belief as a thesis and the other as a definition which is then used as a reason in the proof of the thesis.
After his [1998], Morscher seems to have felt the need to make the definitionist view plausible. In his [2013], he correctly remarks that the identity statement is needed as what we have called a bridging principle. Morscher assures the reader that Anselm has already established the identity statement before making the argument that he, Morscher, reconstructs in his paper, i.e., before P2.10. The only place where the identity statement could have been established before that is P2.2. Hence, Morscher does not read P2.2 as a thesis requiring further argument, although he does not explicitly call the identity statement a definition. 24 Morscher offers a plausibilization of his definitionist account by hinting at a parallel to the five ways of Aquinas. Each way ends preliminarily with the claim that a first mover, a first efficient cause, etc., exists; and each one reaches its final end by identifying one of these things with God by means of famous clauses such as "et hoc omnes dicunt Deum." The use of vocabulary like "dicunt," "nominant," etc., suggests that these identifications adduce definitions. If one reads Anselm's proofs as proofs for the (necessary) existence of God even though "God" does not occur in their conclusions, it is quite natural to consider their endings as preliminary, too. They too require a finalization step leading from the (necessary) existence of a famous object to the (necessary) existence of God. The reader must take this step on his own by making use of an appropriate bridging principle. The suggested parallel between Aquinas and Anselm may be regarded as an argument that makes definitionism plausible. Taken as such, we call it the Aquinas parallel argument. Barnes [1972] offers four strategies to make his definitionist reading of P2.2 plausible. The first strategy is the finalization argument discussed above: Reading P2.2 as a definition allows a quick and easy finalization of the P2 and P3 arguments for the (necessary) existence of God. 22 Morscher [1998], 65, our transl. 23 See Oppy [1995], 236-238, see also 207 et seq. 24 Morscher [2013], 95-109. Barnes' second strategy is to point out the parallel to chapter 80 of Anselm's Monologion [= M]. M offers an advanced and comprehensive doctrine of God. The term "God," however, does not appear before M80, the last chapter of M (with one negligible exception). 25 Throughout M, Anselm refers to God only by descriptions like "the highest essence" (summa essentia). In M80, then, God is introduced as a substance that is higher (in ontological rank) than everything else, adorable, and to be invoked by human beings.
M80: "Quippe omnis qui Deum esse dicit, sive unum sive plures, non intelligit nisi aliquam substantiam, quam censet supra omnem naturam quae Deus non est, ab hominibus et venerandam propter eius eminentem dignitatem, et exorandam contra sibi quamlibet imminentem necessitatem." "Indeed, everyone who affirms that a God exists (whether one God or more than one) understands [thereby] nothing other than a Substance which he believes to be above every nature that is not God-[a Substance which] men are to worship because of its excellent worthiness and which they are to entreat against lurking misfortune." 26 With the help of this definition, Anselm then proves that the highest essence, which all of M has been talking about up till now, is God. The Monologion parallel argument, then, is that in both M and P Anselm first proves the existence of something and then (in P as in M) uses a definition to conclude that that something is God.
Barnes' third strategy argues from the observation that Anselm claims twice "that the fool who denies the existence of God does not properly understand the meaning of the word 'God'." 27 As the fool is a disbeliever, there is an obvious contrast between his problems with understanding the meaning of "God" and the faithful description of God as something than which nothing greater can be thought in P2.2. Hence, this description suggests itself for consideration as a definition taken from faith. This we call the incomprehension argument for definitionism.
The fourth strategy combines interpretational matters with historical insights and conjectures. Anselm was probably familiar with Seneca's Naturales quaestiones, two copies of which were kept in the monastic library at Bec. In the preface to this work, we find a version of the famous formula: 25 In M1, the word "God" is used in passing to characterize several theses about the highest essence as "what we believe about God and His creatures". The famous and careful Anselm editor Franciscus Salesius Schmitt seems to have overlooked this use in passing when he stated that the name "God" does not appear before the last chapter of M (Schmitt [1964], 15); see also Holopainen [2020], 25, note 45. 26 Hopkins [2000], 86. 27 Barnes [1972] "What is God? The mind of the universe; everything that you see and everything that you do not see. His greatness, than which nothing greater can be imagined (qua nil maius cogitari potest), is only attributed to him if he alone is everything, if he holds his work from the inside and from the outside." 28 As Anselm's use of the formula depends on Seneca, and Seneca defines God by means of the famous formula concerning his greatness, Anselm must do the same. We call this the dependence argument.
The plausibilization strategies presented here will be critically assessed in Sect. 3.3 below.

The AB Objection and the Two Senses of "Definition"
Definitionism is not held without opposition. One line of opposition to it concerns the Aristotelian-Boethian [= AB] theory of definitions. According to Ian Logan: "It [=P2.2, GS/CT] is not a definition of God, and could not be, since God is not susceptible of definition in the dialectical tradition in which Anselm is operating, i.e. God is not a differentiated species of a higher genus." 29 Logan is alluding to the conception of so-called real definitions in the AB tradition. As we saw in Sect. 2.1 above, real definitions integrate a given thing into a hierarchy of genera and species. But God cannot fall into a genus, because then he would be an intermediate thing in the hierarchy of beings. Hence, God cannot be given a real definition. We call this the AB objection to definitionism.
A variant of the AB objection is the following. In P3.11 Anselm undoubtedly considers as established the statement that God is something/that than which nothing greater can be conceived. If this were a definition in the AB tradition, it would have to provide the nature (substantia) of the definiendum. But the famosa descriptio is a relative description, and relative descriptions can never provide the nature of something, as Anselm argues in some detail in M15-16. Hence, it cannot be a definition. And Anselm confirms this view when, in the headline of P5, he asks "What therefore are You, than whom nothing greater can be thought?" (P5.0). If 28 Barnes [1972], 7; the translation is Barnes', and the Latin text is from Seneca: Naturales Quaestiones I, 13. 29 Logan [2009], 91. the relative clause were a definition in the AB sense, a further "what is…" question about it would make no sense. 30 What force does the AB objection have? The answer depends on the sense of "definition" in the definitionist thesis (see Sect. 2.1 above).
If "definition" is taken in the sense of the AB tradition, according to which God cannot be defined, it would be extremely implausible to interpret Anselm's sentence P2.2 as a definition, because this would insinuate that he was trying to do something he himself held to be impossible. Hence, the AB objection is successful against definitionism in the AB sense.
If, in contrast, "definition" is taken in the modern sense-the sense we have in mind, together with nearly all other commentators, when definitionism is in focusthen the force of the AB objection depends on how the two senses of "definition" are related. Generally speaking, real definitions and definitions in the modern sense are two quite distinct conceptions with only a small overlap, as we saw in Sect. 2.1 above. Their distinctness also applies in the special case of God. While the proposition that a real definition of God cannot be given is standard in medieval philosophy and theology, because God does not belong anywhere in the ontological hierarchies of created things, the medievals had no problem with defining God in the modern sense, namely by determining the meaning of the term "God." The modern sense of "definition" was present in medieval philosophy when it came to clarify the significatio of a term, even though the medievals did not use a word like "definition" for it. Thomas Aquinas, for instance, held that a proof for the existence of God is a basic prerequisite for any rational theology. He explicitly said that such a proof would have to take the meaning of the term "God" (quid significet hoc nomen Deus) as a means (pro medio) to such a proof. 31 So Aquinas had no problem with defining God in the modern sense while holding that one cannot define Him in the real sense (as he puts it, one cannot grasp the quid est of God). 32 The same is true for Anselm. That he stands in the dialectical AB tradition does not preclude him from defining God in the modern sense, while it might well imply that he was convinced that a real definition of God is impossible. In fact, in Monologion 80, Anselm does present a definition of God in the modern sense (see the Monologion parallel argument in Sect. 2.4 above). So, Anselm was certainly not convinced that it is impossible to define God in the modern sense. Though, in 30 We owe this line of argument to Richard Campbell's written communication. -A third variant of the AB objection might refer to God's ineffability. While a definition in the AB tradition always aims at determining the given thing (compare Victorinus, De definitionibus (ed. Pronay [1997]) 3.12f, 7.1-4, 7.10-11.7), God cannot be determined in that way. Anselm deals with this problem in several places and makes clear that the knowledge he seeks and is able to attain is not of this kind (P1, P14-16, Responsio Anselmi 9, Monologion 26, Monologion 64-65). The clearest evidence is perhaps the formula used in P15 that God is the "greater than can be thought" (maius quam cogitari possit); for a general description of Anselm's enterprise between faith and immediate vision of God, see above all Holopainen [2020], 171-178. 31 Thomas Aquinas, STh I.2.2 ad 2. 32 Thomas Aquinas, STh I.1.7 ad 1; I.2.1 resp.; SCG I,3.
Proslogion 15, he characterizes God as a being greater than can be thought, thereby exempting God from the possibility of a definition in the AB sense.
Hence, it is by no means an anachronism to say that Anselm may have intended to define or can be read as defining God in the modern sense. And neither can definitionism in the modern sense with respect to P2.2 be rejected as easily as definitionism in the AB sense.
In the end, the AB objection is successful against definitionism when "definition" means "real definition." If, in contrast, "definition" in "definitionism" is taken in the modern sense, the AB objection has no force. Therefore, we will focus on arguing against definitionism in this modern sense in the remainder of this paper.

The id/aliquid Problem in the Context of Definitionism
Up until now, we have passed over an interpretive problem that, as it turns out, has some implications for definitionism. The problem is known as "the id/aliquid problem." It arises from the fact that the famous formula "something/that than which nothing greater can be thought" appears in the Sumptum in a variety of wordings, which cannot be explained completely as mere stylistic variations. 33 One feature of the variations is alternation between "something" and "that" (in Latin, "aliquid" and "id"). The resulting oscillation of the famous formula between a description with a definite pronoun and a description with an indefinite pronoun may be logically significant. It can be dealt with in different ways. One way is to presume that Anselm was a little sloppy at this point. This presumption provides Anselm interpreters with a degree of freedom: they can either decide to be sloppy as well or settle on a fixed reading (always "id" or always "aliquid"). The second way is to take Anselm at his word and to consider the alternation between definite and indefinite formulations a deliberate one that should be respected. Call this second response literalism with respect to the id/aliquid problem. As it turns out, there are literalist reconstructions that pay heed to the id-aliquid oscillation. So literalism is a live option: the difference between the id and aliquid versions cannot be ignored, and a reconstructor must decide whether to take the literalist approach, one of the fixed options, or some other option (such as freedom to be sloppy) on the basis of further reasons.
In the context of definitionism, this problem is linked to the problem of the syntactic category of the word "God" (see footnotes 7 and 8 above): Is "God" to be read as a singular term (a proper name, an individual constant) or a general term (a predicate constant)? In both cases, there are requirements for well-formed definitions. 34 One group of requirements is particularly important to us: according to definitiontheoretic orthodoxy, an explicit definition of an individual constant g must have the form of an equation (g = t, for some term t) or a biconditional (∀x (x = g ↔ ϕ(x)), for some formula ϕ. 35 Furthermore, uniqueness of whatever instantiates the defining term or formula is required for the definition of an individual constant, be it as a prerequisite for setting up the definition or as an explicit condition mentioned in the definition.
All these matters-the reading of P2.2 as an explicit definition, the id/aliquid problem, and the syntactic category of "God"-interact with each other. We can best illustrate this interaction with the following outline of a deductive argument: (1) Since Explicit definitions of individual constants have the form of equations or biconditionals and must observe certain uniqueness requirements.
(Definition-theoretic orthodoxy) (2) Since "God is something than which nothing greater can be thought" (P2.2) is to be read as a definition.
(Literalism w.r.t. the id-aliquid problem) (4) Hence "God" in the definition must be a predicate constant.
Conclusion (4) follows from (1)- (3): if the aliquid in P2.2 is to be taken literally (3) and P2.2 is to be read as a definition (2), then this definition reads: "God is (by definition) something than which nothing greater can be thought." Now, "God" must be either an individual constant or a predicate constant. If it is an individual constant, then according to definition-theoretic orthodoxy (1), an explicit definition of it must have the form of a biconditional or a term equation. Obviously, the definiens does not have the form of a (standard) term. 36 Hence, the definition can only be a biconditional (∀x (x = g ↔ ϕ(x))). Accordingly, uniqueness must either be given as a prerequisite to the definition or be built into the definition in form of a condition (formally, ∀x(∃!yϕ(y) → (x = g ↔ ϕ(x)))). Neither is the case: in the context of P2.2, there is absolutely no indication of a uniqueness condition. There is neither a uniqueness theorem about famous objects nor a mention of uniqueness in the (alleged) definition. Hence, under the assumptions (1)(2)(3), the definition of "God" cannot be the definition of an individual constant but only that of a predicate (4). This conclusion appears problematic, though, since the definitionist arrives at a definition of "God" only by replacing the second-person pronoun "You" in P2.2 with "God," i.e., by using "God" as a singular and not as a general term. 37 35 See, for instance, Siegwart [1999]. 36 One might think of the definiens as having the form of a non-standard term for an indefinite description, such as an η or an ε term (ηx.FO(x) or ɛx.FO(x)), such that the definition of the individual constant g (for God) would be g = ηx.FO(x) or g = ɛx.FO(x). Anyhow, we will not consider the possibility of nonstandard terms any further, because it is doubtful whether definitions of individuals having the form "… is (identical to) some F" make sense at all. According to Hans Reichenbach, η terms "must not be used in definitions," "since the indefinite description represents an existential operator in the definiendum" (Reichenbach [1947], 266).-Note, though, the possibility of implicit definitions, e.g., by meaning postulates or axioms. Hinst [2014], 23-28, reads P2.2 as such a meaning postulate for the individual constant deus; see footnote 40. 37 On this point, see also footnote 49.
To avoid this conclusion, one of the three premises of the argument must be dropped: definition-theoretic orthodoxy (1), definitionism (2), or id-aliquid literalism (3). As we are assuming definition-theoretic orthodoxy for the purpose of Anselm exegesis, the alternative boils down to dropping either definitionism or literalism.
When it comes to interpreting argumentative texts, a multiplicity of different factors and maxims influence concrete decisions about how to read a given passage. Faithfulness to what the author literally says is one such maxim; reconstructing the author as correct in as many claims and inferences as possible is another. In many cases, such maxims so conflict with each other that the interpreter must seek a reflective equilibrium, doing justice to as many of the hermeneutical maxims as possible in the highest degree possible for each. Often such considerations are highly non-trivial. 38 Regarding the id/aliquid problem, we hold that it is not decisively clear that literalism is compulsory. However, there are several good reasons for literalism, and given the argument above, each reason for literalism is a reason against definitionism. Reasons for literalism include: 1. Ceteris paribus, what the author literally says takes priority over the assumption that he meant something different. 2. 1. is especially true when a coherent and plausible reconstruction of the literal text is possible. There are ways of reconstructing the Sumptum proofs that take the id-aliquid variation into account and pari passu do not display additional major weaknesses when compared with reconstructions that allow a sloppy reading or choose a fixed alternative (always id or always aliquid). 3. 1. and 2. are especially true in cases where the author of the text is known to be well-versed in the area in focus. Anselm is considered to be well-versed in logical matters (in a traditional, wide sense of "logic"), 39 the id/aliquid problem is a logical problem (in that sense), and, finally, the history of interpretations of the Sumptum arguments suggests that Anselm actually used his skills here.
We call this line of argument the id/aliquid argument against definitionism. 40 38 See, for instance, Brun [2014], Löffler [2014], and Reinmuth [2014]. 39 This traditional logic-as far as it was explicitly developed-is essentially identical to the Porphyrio-Boethian tradition of commentary on Aristotle's Organon, also known as logica vetus. Implicitly, however, several techniques were successfully used even before the corresponding logical systems had been developed. Depending on interpretational decisions, one might accept as examples of such techniques the transition from aliquid to id for existential elimination (or introduction of a parameter) procedures and the characterization of the positive element of the contradiction at the end of the P2 argument-i.e. that nothing greater than (something/that) than which nothing greater can be conceived can be conceived-as true by a (later called "Russellian") logic of definite descriptions (Morscher [2013], p. 101). 40 Sensu strictu, our argument in Sect. 3.2 is only an argument against common definitionism, i.e., definitionism in the sense of an explicit definition of "God." Excluding definitionism in the sense of implicit definitions such as meaning postulates as well (as in Hinst [2014]) requires a little more work. We close this small argumentative gap in Sect. 4 below, where we argue that P2.2 is presented as the content of a religious belief which can be understood by way of proof and for which Anselm in fact argues in the Sumptum. Understanding some ϕ by way of proof and arguing for ϕ would not make sense if ϕ were a meaning postulate.

Against the Proposed Plausibilizations of Definitionism
The id/aliquid argument provides some evidence against definitionism. But what about the plausibilizations of definitionism? Can they withstand the rising pressure?
The strength of Morscher's Aquinas parallel argument depends entirely upon the degree of similarity between Anselm's and Aquinas' argumentation. When Aquinas identifies the things he has proven to exist (first mover, first efficient cause, etc.) with God, he explicitly uses definitional vocabulary like "nominant," "dicunt," etc., which Anselm does not. Nor does Aquinas make the identification an object of insight requested in prayer, as Anselm does. Hence, the dissimilarities between Aquinas and Anselm are too great for the parallel argument to have force. 41 Barnes' first strategy, the finalization argument, is surely correct in saying that Anselm wanted to prove the (necessary) existence of God and that, to this end, the arguments for the simple and the modal existence theorems require finalization by means of a bridging principle. A bridging principle can, however, fulfill its function independently of whether it has the status of a definition or that of a (proven) thesis. Therefore, the finalization function is not decisive for the argumentative status of the bridging principle. Thus, the finalization argument is a non-starter.
The Monologion parallel argument (Barnes' second strategy) is extremely weak in that nothing prevents an author from drawing one and the same conclusion from a theorem and a definition in one place and from two theorems in another. The Monologion parallel argument would be stronger in the presence of more evidence of a parallel argumentative structure, a similar wording of theses, and other forms of parallelism and similarity. However, there is barely any such evidence. To the contrary, there are many dissimilarities: the bridging principle in P2.2 is formulated without definitional vocabulary (whereas in M80 definitional vocabulary is present); it is mentioned before the existence arguments (whereas in M80 it is mentioned after them); and it is explicitly argued for in the pars neglecta (whereas in M80 it is not argued for at all). Hence, reasoning from similarities and dissimilarities between M and P does not favor definitionism either.
The incomprehension argument argues from the contrast between the fool's (a disbeliever's) difficulties in comprehending "God" and the insightful definition informed by faith in P2.2. As Barnes rightly observes, Anselm says twice (namely in P4.5 and in the last sentence of Responsio 7) that the fool does not properly understand the meaning of "God." The latter passage does indeed deal with comprehension problems. However, it would be absurd to infer from the comprehension problems of one person in one place that the word must be used by another person in another place as a definition. The context of the former passage is an explanation of why the fool can say in his heart what cannot be thought. In this explanation Anselm refers back to what God is, namely (enim) that than which nothing greater can be thought (P4.6). The fact that this refers back to the bridging principle is inconclusive regarding whether what is referred back to is a definition or a (proven) thesis.
Barnes' fourth strategy, the dependence argument, is based on the fact that Seneca had used the famous formula as an answer to the question "Quid est Deus?" (What is God?). However, not every answer to a what-question is a definition. The fact that in Seneca's original text (as opposed to Barnes' quotation) the question "Quid est Deus?" is asked several times and with different answers makes it probable that the answers were not intended as definitions. And in the unlikely case that one of the answers was intended as a definition, some justification is needed for selecting the answer concerning the famous formula (out of all the answers given) as the definiens. Furthermore, even if Seneca did intend to give a definition in this passage, it does not follow that Anselm used the formula with the same performative intent. Finally, all these considerations depend upon the unverified assumption that Anselm took the famous formula from Seneca in the first place, and not, e.g., from Augustine, 42 Boethius,43 or Cicero. 44 In sum, the dependence argument rests on several doubtful premises.
In the end, all the plausibilization strategies for definitionism are unsuccessful. They turn out to be inconclusive or to point to evidence that is outweighed by counterevidence. The id/aliquid argument, in contrast, has put definitionism under additional pressure. These facts strongly call for an alternative to definitionism.

The Alternative: a Thesist Reading-and its Advantages
The alternative to definitionism is thesism, according to which the bridging principle that allows Anselm to proceed from the (necessary) existence of the/a famous object to the (necessary) existence of the/a God is not a definition but a thesis. We will defend the view that it is, in fact, a thesis Anselm argues for in the Sumptum. Thesism also fits with the fact that the bridging principle is an object of Anselm's petition for insight in P2.1-2, which Anselm thanks God for having granted in P4.9.
The main advantages of the thesist reading are that it allows one to get the argumentative macrostructure of the Sumptum (P2-4) right and that it takes the different forms of speech in the Proslogion into account. It also circumvents the difficulties with the definitionist reading, namely that the need to make it plausible is not successfully met and that the id/aliquid argument speaks against it. 42 Augustine, Confessiones VII.4.6: "Neque enim ulla anima umquam potuit poteritue cogitare aliquid, quod sit te melius"-"For never soul was, nor shall be able to conceive any thing which may be better than Thou" (Latin text from Augustine [1981], 95, English text from Augustine [1909], 106); De moribus II.11.24: "Summum bonum omnino et quo esse aut cogitari melius nihil possit, aut intelligendus aut credendus est deus, si blasphemiis carere cogitamus"-"That God is the supreme good, and that than which nothing can be or can be conceived better, we must either understand or believe, if we wish to keep clear of blasphemy" (Latin text from Augustine [1992], 109; English text from Augustine [1872], 65). 43 Boethius, De consolatione philosophiae III.10: "cum nihil deo melius excogitari queat, id, quo melius nihil est, bonum esse quis dubitet?"-"For nothing can be thought of better than God, and surely He, than whom there is nothing better, must without doubt be good" (Latin text from Boethius [1934], 65; English text from Boethius [1902]). 44 Cicero, De natura Deorum, II.7.1: "Atqui certe nihil omnium rerum melius est mundo nihil praestabilius nihil pulcrius, nec solum nihil est sed ne cogitari quidem quicquam melius potest."-"Yet beyond question nothing exists among all things that is superior to the world, nothing that is more excellent or more beautiful; and not merely does nothing superior to it exist, but nothing superior can even be conceived" (Cicero [1961], 140-143); on Anselm's possible sources with respect to the famous description see Logan [2009], 92-93, andDemetracopoulos [2012].
In this section, we will show how a thesist reading fits into the structuring of the text provided by the different forms of speech it contains (Sect. 4.1). Then, we will briefly survey some of the arguments in favor of thesism (4.2). One of these arguments will subsequently be substantiated in Sect. 5.

Getting the Forms of Speech and the Argumentative Structure Right
In the Sumptum (P2-4) we find three different forms of speech. Two forms immediately catch the eye: prayer and discourse. The prayer form can most easily be recognized by the appearance of singular second-person pronouns referring to God ("You"), by the insertion of invocations of God ("o my Lord"), and by verbs expressing petitions, laud, and thanks, which are typical of prayers. The discursive texts can be recognized by their discursive vocabulary, argumentative structure, and/ or explanatory nature. The speech forms of prayer and discourse are not mutually exclusive: there are passages in which Anselm argues and prays at the same time. Hence, there is also a mixed form of speech, which is often overlooked: discursive prayer, or praying discourse. Hence, we distinguish three forms of speech: pure discourse, praying discourse, and pure prayer.
The overall structure of the Sumptum is that of a prayer framing a discursive text (P2. 3-4.8). In the opening part of the prayer (P2.1-2), Anselm asks God for understanding of what he believes about God. In the closing part of the prayer (P4.9), Anselm thanks God for His illumination, by virtue of which he, Anselm, has now come to understand what he first believed. The obvious reference in this second prayer to the prior petition strongly suggests reading the two prayers as constituting a single frame within which the understanding has arisen. The message is clear: God has enlightened Anselm by enabling him to find proofs of what he believes.
What is it that is believed? According to P2.1, what is believed is: that God is (i.e. exists) and that he is what he is believed to be (namely, something than which nothing greater can be thought, P2.2). Hence, from the structure of the text, the forms of speech, and their meaning, one can conclude that in the argumentative part of P2-4 there should be rational insight into God's existence and into what he is believed to be. Hardly anybody doubts that the argumentative part somehow deals with God's existence. Shouldn't one naturally expect it to deal as well with what God is believed to be?
Three discursive passages are easily identifiable and in fact identified by almost all commentators: the argument for the existence of a famous object "et in re" (the simple reality theorem, P2. [3][4][5][6][7][8][9][10][11][12][13][14], the argument that it cannot be thought not to exist (the modal reality theorem, P3.1-5), and the longer explanatory passage devoted to explaining how the fool could say in his heart that there is no God even though, according to the modal reality theorem, this cannot even be thought (and thinking is saying-in-the-heart, P3.12-4.8).
While the three discursive passages and the two framing prayers are clear cases, there remains a further text passage, P3.6-11. This passage is so often neglected in discussions of the Sumptum that we like to call it pars neglecta. 45 It deals with the bridging principle (God is the/a famous object) and displays characteristics both of prayer (God is invoked and addressed by the singular second-person pronoun) and of argumentative discourse (argument-indicating words like "enim" and "igitur" are used along with phrases like "et merito"-"rightly so"-to confirm what was said and announce that justification is about to follow). Hence, it belongs to the third, mixed, form of speech-discursive/argumentative prayer, or praying discourse.
In sum, the textual structure of the Sumptum is as follows: P2.1-2 Prayer: petition for understanding P2. [3][4][5][6][7][8][9][10][11][12][13][14] Argument for the simple reality theorem P3. [1][2][3][4][5] Argument for the modal reality theorem P3.6-11 Praying argument for bridging principle ("pars neglecta") P3.12-4.8 Explanation of thinking something unthinkable P4.9 Prayer: thanks for understanding This textual structure of the Sumptum is in perfect coherence with its argumentative macrostructure as displayed in Fig. 1 below. First, in P2.3-9, Anselm argues that a famous object exists in the understanding. Next, in P2.10-14, he proves the simple reality theorem, which states that the famous object exists et in re. Then, in P3.1-5, the modal reality theorem is shown. It says that a famous object so truly exists that it cannot be thought not to exist. Finally, in the pars neglecta (P3.6-11), Anselm argues for the bridging principle inter alia on the basis of the modal reality theorem and what might be called an exclusive existence thesis. In the end, he has  purportedly provided all the building blocks for a proof that a famous object (necessarily) exists and that it can somehow be identified with God. This is exactly what Anselm begins by asking God to illuminate and ends by thanking God for having illuminated. 46

In Favor of Thesism
The thesist view is the minority position with respect to interpreting P2.2, but it does have some adherents. One of the earliest is Henry of Ghent (1217-1293). In his Summa Quaestionum Ordinariarium Theologiae (tom. I art. XII q. 1), he explicitly refers to Anselm's argument, saying: "Probat autem minorem sic: quoniam si Deus non esset quo maius cogitari non posset, tunc intellectus noster posset excedere Deum, et creatura posset cogitare aliquid melius Deo et maius, quod falsum est: ergo etc." 47 "He proves the minor thus: If God were not [something] a greater than which cannot be thought, then our intellect could exceed God, and the creature could think something better and greater than God, which is false: ergo, etc." Hence, Henry clearly recognizes that Anselm argues for the bridging principle, even if he explicitly considers only one of Anselm's two arguments for it.
One of the first modern thesists is Arthur C. McGill [1967]. He does not neglect the pars neglecta, but sees Anselm as trying "to prove that the subject under discussion-namely, that which has been shown not only to exist, but also to exist in such a way that it cannot even be conceived not to exist-is actually the Lord God, Creator of heaven and earth" (p. 40).
An early critic of definitionism is Richard La Croix. In his [1972], 14-16 and 36, he advances several reasons to reject definitionism. Those reasons are then taken up by Richard Campbell [1976], 27-29, and many others. 48 Summing up, the arguments in favor of the thesist reading of P2.2 are: 5. Thesism is suggested by the interpretational perspective which takes seriously the three forms of speech (prayer, discourse, and a mixed form). Furthermore, it is indirectly supported by the fact that this perspective provides a key to recognizing the textual structure of the Sumptum. 6. This textual structure coheres well with the argumentative macrostructure. Reading P2.2 as a thesis, argued for in the pars neglecta by making use of the modal reality theorem, is the key to recognizing the argumentative macrostructure of the whole Sumptum (P2-4). 7. There are (one or two) arguments for a bridging principle in the pars neglecta (P3.6-11). 8. One of these arguments reduces the redundancy of the two existence arguments in P2-3 by making use of the modal reality thesis. 9. Thesism gives rise to a logical reconstruction of the neglected arguments from P3.6-11 that fits neatly into logical reconstructions of the whole argument of the Sumptum.
The last three points will need substantiation. For reasons of space and complexity, we will provide it in the next section.

A Touchstone for Thesism: The Neglected Arguments for the Bridging Principle
We take the arguments for thesism to be decisive. However, we consider a touchstone for any perspective on an argumentative text to be whether that perspective allows for a logical reconstruction of the text. This motivates us to address the following question: Can the arguments for the bridging principle in the pars neglecta be integrated into a reconstruction of the Sumptum argument as a whole-and, if they can, how?

Readings of the Bridging Principle
In the literature, the arguments for the simple and the modal reality theorems in P2 and P3 are logically reconstructed in a variety of ways. The reconstructions vary not only with respect to the logical means they use (formal languages, logical axiom/inference systems, etc.) but also with respect to some very basic hermeneutical decisions that influence the reading of the bridging principle. The bridging principle results from P2.2 by substituting "God" for "You." It reads: "(A) God is (something/that) than which nothing greater can be conceived," or, for short: "(A) God is (a/the) famous object." Two hermeneutical decisions having implications for this principle have already been mentioned. The first decision is whether "God" (in Latin "deus") should be construed as a proper name or a predicate. The proper-name reading results in formally rendering "God" as an individual constant g, the predicate reading in rendering it as a predicate constant G(..). The Latin text permits both readings, the proper-name reading being slightly more plausible. 49 The second decision concerns the question whether to settle on the definite (id quo nihil maius…) or the indefinite (aliquid quo nihil maius…) version of the famous formula. In the context of P2.2, the definite version fits slightly better with the proper-name reading of "God," while the indefinite version fits slightly better with the predicate reading of "God." The crossed-over combinations can also be embraced. Other things being equal, given that the Latin text of P2.2 has "aliquid," the indefinite version is more plausible because it stays closer to the text.
A further decision concerns the "is" in "God is a/the famous object": Does the "is" express (first-order) identity, predication, extensional equivalence, or subsumption? 50 This decision is tied to the previous two decisions, since the first two possibilities-identity and predication-must be considered if "God" is taken as a proper name, the second two possibilities-extensional equivalence and subsumptionif "God" is taken as a predicate. If "God" is taken as a predicate and the famous formula is taken in its literal, indefinite sense, the "is" will be read as extensional equivalence or subsumption: "A God is a famous object" will be read as "Something is a God if and only if it is a famous object" or as "If something is a God, then it is a famous object." If "God" is taken as a proper name instead and the famous formula is taken in the definite sense, then the "is" will probably be read as (first-order) identity: "God is identical to the famous object." Besides these combinations of decisions, there are other combinations that can be considered to be more or less plausible. For example, taking "God" as a predicate can also be combined with the definite version of the famous formula. This will require amending the predicate with a uniqueness condition: "Something is a God if and only if it is the famous object" will result in "Something is a God if and only if it is a famous object and all famous objects are identical with it." 49 The proper-name reading is slightly more plausible because God is addressed by the singular secondperson pronoun "you." However, imagine that somebody is introduced to you as a heart surgeon. When you ask her, "Do you also operate on children?" she might answer: "Yes, we as heart surgeons do also operate on children, but I personally do not because I specialize in geriatric heart diseases." This completely ordinary conversation shows that we sometimes use the singular second-person pronoun for a particular person, but also sometimes use it for the function, profession, or office of a person, i.e., in lieu of a predicate. 50 If P and Q are predicate terms, "P is Q" may express extensional equivalence: "For every x: x is P if and only if x is Q" (as in: "Bachelors are unmarried men"), or it may express subsumption: "For every x: if x is P, then x is Q" (as in: "Planets are interstellar objects"). If p is an individual term, "p is Q" expresses predication: "Q(p)" (as in: "Jupiter is a planet").
Combinatorially, the following eight options result. The formal readings of the bridging principle use as vocabulary: g individual constant for "God" G(..) predicate symbol for "..is (a) God" FO(..) famous formula saying, informally, "..is such that nothing greater can be thought," or, even more informally, "..is a famous object." ɩx._ logical operator turning a predicative description (formula) _, in which x is a free variable, into a definite description (term) ɛy._ logical operator turning a predicative description (formula) _, in which y is a free variable, into an indefinite description (term) "God" "is" expresses Famous formula Reading of bridging principle 1 Singular name Identity Definite "God is identical to the famous object." g = ɩx.FO(x) 2 Singular name Identity Indefinite "God is identical to (an example of) a famous object." g = ɛx.FO(x) 3 Singular name Predication Definite "God is the famous object." FO(g) ∧ ∀y (FO(y) → y = g)

Singular name Predication
Indefinite "God is a famous object." FO(g) 5 General name Extensional equivalence Definite "To be a God is to be the one and only famous object." ∀x (G(x) ↔ (FO(x) ∧ ∀y (FO(y) → y = x))) 6 General name Extensional equivalence Indefinite "To be a God is to be a famous object." ∀x (G(x) ↔ FO(x)) 7 General name Subsumption Definite "A God is a unique famous object." ∀x (G(x) → (FO(x) ∧ ∀y (FO(y) → y = x))) 8 General name Subsumption Indefinite "A God is a famous object." ∀x (G(x) → FO(x)) Which option one takes for a logical reconstruction of parts of or the whole Sumptum is a truly hermeneutical decision: it is a choice influenced by the textual data, other interpretational choices, and the desire to produce valid derivations; and it, in turn, influences other choices and the coherence of the reconstruction as a whole. While the other options are also worth further consideration, we take options no. 1 and 6 to be most preferable. And, in fact, they are the most discussed options in the literature. Hence, we will choose them as examples.
The next subsection offers the text of the pars neglecta (Sect. 5.2). In the following two subsections, we show what possible reconstructions may look like, given chosen combinations no. 1 (Sect. 5.3) and 6 (Sect. 5.4).

The Reconstruendum: Anselm's Arguments for the Bridging Principle
Here is the text of the pars neglecta of the Sumptum: (3.6) Et hoc es tu, domine deus noster.
(3.6) And You are this thing, O Lord our God.
(3.7) Therefore, O Lord my God, You truly exist in such a way that You cannot be thought not to exist.
(3.9) For if some mind could think something better than You, the creature would ascend above the Creator, and would pass judgement on the Creator, which is quite absurd.
(3.10) And indeed whatever is other than You alone can be thought not to exist.
(3.11) Therefore, You alone have being most truly of all, and thus most greatly of all, because whatever is other [than You] truly does not exist in this way, and for this reason has less being.
In this passage, we find two arguments for the bridging principle. The first argument is taken from theology of creation: God is such that nothing greater can be thought because otherwise the creature thinking the thing greater than God would thereby somehow elevate itself above the creator (3.7-9). We call this the theological argument. The second is an argument operating with the exclusive existence thesis: Every being except God can be thought not to exist. Given this strong thesis, some version of the bridging principle follows from the modal reality theorem (3.10-11): the famous object cannot be thought not to exist, but everything except God can be thought not to exist, so God is the/a famous object. We call this the metaphysical argument.
Generally speaking, one can take two perspectives on the pars neglecta: either it consists in one two-step argument or in two more or less independent arguments for the same conclusion. Call these the one-argument view and the two-arguments view. 51 According to the one-argument view, the theological and the metaphysical arguments are somehow combined to reach the conclusion that God is the/a famous object. According to the two-arguments view, they are independent arguments that both yield that conclusion.
In one or the other way, the arguments from the pars neglecta establish the bridging principle, which in turn finalizes the whole proof for the existence of God. The exact role of the bridging principle in this finalization depends on its precise syntax, alternative assessments of which were developed in Sect. 5.1 above. We will now briefly discuss how the main alternative readings of the bridging principle fare with respect to the reconstruction of the pars neglecta and its contribution to Anselm's overall argument in the Sumptum. 51 The two-arguments view is championed by Campbell; see his [1976], 126, and [2018], 218-280.
The main advantages of this reading are: -It finalizes the macro-argumentation smoothly.
-It makes use of both reality theorems in one place.
-Reading "God" as a singular name is well in line with its being the replacement for "You" in P2.2 and P3.6-11.
The main disadvantages are: -If one uses constants or definite descriptions like ɩx.FO(x), one is faced with hallenging questions concerning existence and uniqueness. Given what is derivable from what the other parts of the argumentation presuppose or show, uniqueness can be shown only for ¬T¬EiR(x), not for ¬T(∃y Gr(y,x)) (i.e., FO(x)), and this particular reconstruction does not address existence either. -All aliquids are read non-literally as ids.

Second Exemplary Reconstruction: The Bridging Principle as an Extensional Equivalence Between Two Predicates
Our second exemplary reconstruction is based on set of choices no. 6. In some respects, this set of choices is antipodal to no. 1: "God" is (not a singular but) a general term, the famous description is (not definite but) indefinite, and the "is" expresses (not first-order identity of objects but) extensional equivalence of predicates.
The bridging principle then reads: "To be a God is to be a famous object." Using the same assignment of reconstruens (formal) and reconstruendum (informal) language expressions as above, this formally comes down to P2.2 ∀x (G(x) ↔ FO(x)).
A possible reconstruction of the macro-argumentation of the Sumptum according to this reading has it that P2 results in P2.14 ∃x (FO(x) ∧ EiR(x)) and that the first part of P3 results in P3.5 ∃x (FO(x) ∧ ¬T¬EiR(x)).
For reasons of space, we will not elaborate further on these matters in this paper. However, they had to be mentioned to forestall rejecting this line of reconstruction merely on the grounds that the formalization of the theistic thesis does not include an "EiR(x)" clause.
Step (4) of the reconstruction adduces the thesis, "∀z (G(z) → (∀x (C(x) → ¬Th(x,∃y Gr(y,z))) → ¬T(∃y Gr(y,z)))" which says that, for a God, if no creature can think something greater than the God, nothing greater than it can be thought. This thesis has the form: If no creature can think φ, then φ is (creaturely?) unthinkable. This transition from a two-place relation "..can think that__" to a one-place relation "it can be thought that__" is not completely straightforward, but it is strongly suggested by the text: While P3.9 says that "some mind could think..", P3.10 confirmingly infers therefrom that "indeed.. can be thought." The main advantages of this reading are: -The argument from creation in the pars neglecta plays a role on its own.
-No uniqueness problems arise.
-All aliquids and ids are read literally. 52 The main disadvantage is: -Reading "God" as a general name is in tension with its being the replacement for "You" in P2. 2, P3.6-11, and P4.6.
In all of our suggested formalizations, further advantages and disadvantages may appear when the formalizations are elaborated and the premises are examined (e.g., that there are quantifications into propositional-attitude contexts). For reasons of space, we will not go into such details here, as our focus was only to show that there are ways of formal reconstruction that assign the pars neglecta an argumentative function in establishing the bridging principle.

Conclusions in Light of the Reconstructions
As we noted before, choices in arriving at a more detailed reconstruction of the pars neglecta arguments depend holistically on hermeneutical decisions and data from other parts of the Sumptum. Establishing the exact nature of the bridging principle depends on all of this, too. By exploring two options in some detail, we hope to have made it plausible that even very different hermeneutical choices, such as options no. 1 and 6, allow for a convincing integration of the arguments presented in the pars neglecta as parts necessary to bridge the gap from the (necessary) existence of a/the famous object to the (necessary) existence of a/the God.
Quite naturally, the definitionist ignores these arguments because, in his eyes, the alleged definition already bridges the gap. The thesist, however, does not have to read as a definition what Anselm presents as an insight he prayed for. She can take Anselm's arguments in the pars neglecta to be important parts of his overall argument for the existence of God. Reading the pars neglecta as it stands, it is almost certain that what the thesist takes to be arguments are indeed intended by Anselm as arguments. Crucial evidence includes, for example, the use of the signaling word "enim" in P3.9 and the fact that premises can be distinguished from conclusions. The final conclusion is that God (addressed in the context of prayer by the singular second-person pronoun) most truly and most greatly has being, i.e. that, as the modal reality theorem says, God has being in such a way that He cannot be thought not to exist. The conclusion that God so truly is that he cannot be thought not to exist is explicitly noted (in P3.7) before the argumentation in the second half of P3 and is alluded to in that argumentation.
When commentators have overlooked these arguments, as many have, it has probably been for one or more of the following three reasons: (1) the arguments are (really or allegedly) weak; 53 (2) the arguments for the two reality theses in P2 and P3 are so intriguing that they absorb all the interpreter's attention; (3) the form of speech in the pars neglecta is neither pure argumentation nor pure prayer, so interpreters, not expecting a mixed form, fail to recognize its argumentative role.

Explaining the Misrepresentation
Given that there is overwhelming evidence against a definitionist and for a thesist reading of P2.2, it remains to explain why so many commentators-the vast majority-have misinterpreted P2.2 as a definition. The need for such an explanation is especially pressing for three reasons. Firstly, because of the sheer size of the group of definitionists: it is not only the majority but the vast majority. Secondly, because most of these commentators are very knowledgeable, well-trained, bright scholars whose opinion cannot be dismissed as the opinio apud vulgares. Thirdly, because definitionists are so heterogeneous in their professional training, their attitudes and their aims: they include philologists, logicians, historians, theoretical philosophers, natural theologians, denominational theologians, atheists, and agnostics. Why do so many qualified interpreters with such diverse backgrounds all read P2.2 as a definition without further ado? 53 Can the weakness of an argument be a reason for misclassifying it as a non-argument? Some explanation may be in order here. It is, in point of fact, sometimes hard to tell whether a particular portion of text is intended as a thesis, an argument, a repetition of or allusion to an argument, an illustration, or something else. When this happens, it is often the case that what is in fact an argument either is being presented in a quite elliptical way or is almost trivially weak. For example, if an interpreter takes an author to have argued "p, therefore p" in a passage of his text, his opponent has a good prima facie reason to question whether there really is an argument in that passage at all, as opposed to a definition or a thesis to be justified elsewhere. Argumentative weakness is probably only a minor factor in the final assessment of an interpretation, but it may explain why certain extremely weak or otherwise atypical arguments (elliptical, trivial, …) are not recognized as such even though the interpreter is rational, attentive, and otherwise competent. "You consider that an argument? That argument is so weak that you cannot mean to impute it to an author like xy!".
The first element in our explanation relates to the finalization argument. The arguments in P2 and P3 are described by Anselm and have been known for centuries as arguments for the existence of God. Every logically trained reader of the arguments easily recognizes that their conclusion is the (necessary) existence of a famous object and that P2.2 offers a bridge between such a famous object and God. The bridging principle supplies just what the arguments lack in order to be complete arguments for the existence of God. This situation invites readers to cross the bridge and end up with the desired conclusion-God's (necessary) existence. However, using the bridging principle in that way requires it to be cognitively established. Reading P2.2 as a definition offers a short and simple solution.
Further, many readers are familiar with Aquinas' five ways, which really do call for a two-line completion (a first mover exists, and that is what everyone calls God, hence God exists-remember the Aquinas parallel argument above). Hence, these readers naturally anticipate a similar two-step completion of Anselm's arguments.
A third explanatory element is the unusual stylistic nature of the Proslogion, with its mixture of prayer and discourse. This mixture invites people interested in arguments to divide the text into passages of prayer and passages of discourse and to focus their interest on the discursive passages. (There is plenty of evidence of this approach, for example when commentators exclude the prayer passages completely as more or less irrelevant to the arguments.) From there it is only a small step to a disjointness presumption, namely, a presumption that the two kinds of text are disjoint in that each passage of the text is either prayer or discourse but not both. Given the disjointness presumption, one can well understand how P3.6-12 became a pars neglecta: this passage obviously has characteristics of prayer, and therefore (so the thinking goes) it is not relevant to the discursive argumentation. 54 Finally, readers of intellectually demanding argumentative texts are accustomed to uses of language that vary from author to author. Instead of just presupposing that a word is used in the sense in which they themselves use it, they are attentive to evidence in an author's writing for how he or she understands that word. 55 Hence, able philosophical readers not only respect definitions but also exercise a habit of carefully collecting evidence for what an author means by his words. When there is no explicit determination, such as a definition-like description, the reader enters something like a Davidsonian process of interpretation. He gradually approximates what the author believes and what she means by her words by initially ascribing his, the reader's, own beliefs and meanings to the author and then modifying these ascriptions step by step in the light of how the author in fact uses the words in her assertions. Given that one frequently has to rely on such laborious and fallible interpretive processes to determine what an author means, one naturally develops a strong inclination to take everything that comes close to a definition as a definition. 54 See also Campbell [2018], 18-21. 55 According to Otto Muck, a minimal aim of academic courses (in philosophy) is to be able not to confuse the first association a reader has in his mind when reading a philosophical text with the opinion of the author (Muck [1999], 471).
Hence, there are several good reasons why even well-trained and knowledgeable readers (including both authors of this paper in earlier stages of their Anselm research) could be taken in and misinterpret the identity thesis in P2.2 as (a textual basis for) a definition.

Summary and Outlook
We hold that in P2.2 Anselm does not provide a definition of God. Definitionism is decidedly false. According to its fides quaerens intellectum program, the Proslogion aims at rational insight into elements of faith. It produces intellectually demanding arguments for the (necessary) existence of a famous object. But the man of prayer has asked for insight into the existence of God and into the fact that God is as He is believed to be, namely, such that nothing greater can be conceived. The missing link is the bridging principle. It allows the arguments in P2 and P3 to be finalized as arguments for the (necessary) existence of God. And it is confirmed as true through the arguments in the Sumptum, which constitute the insight the author recognizes as the answer to his prayer. The bridging principle-roughly, that God is the/a famous object-is introduced as an element of faith in P2.1-2, reaffirmed as such in P3.6-7, and included in the prayer of thanksgiving for rational insight in P4.9. Anselm is firmly convinced that God exists and that He is that than which nothing greater can be conceived. He wants to prove this in full by means of argument. According to his own assessment, he is successful. He understands his success as a fruit of divine grace that has granted him the understanding he prayed for.
To be sure, one can define God as that than which nothing greater can be conceived and even call this an Anselmian definition, since it is Anselmian in spirit and since this understanding of God is a theorem of Anselm's doctrine of God. But one should be clear that Anselm himself did not define God in this way.
In addition to arguing for thesism, we have outlined several hermeneutical options for understanding Anselm's famous arguments. It should have become clear that reading the most famous passages, P2.10-14 and P3.1-5, with their function in the entire Sumptum in mind leads to a more appropriate interpretation of them than not doing so. Our next task will be to present defensible reconstructions within the framework developed above. This task will require taking into account the (initially confusing) spectrum of previously offered interpretations. Here, too, we expect to end up with several options. For each of these reconstructions, there will be substantial theological reasons. Interpreters deeply interested in Anselm's theology will next want to reconstruct the considerations behind these reasons. For the most promising reconstructions, this work will have to be continued up to presumably ultimate reasons. In this way, step by step, there will emerge a picture of Anselm's understanding of God that is detailed and rich in alternatives. It will then finally be possible to judge the (2.9) Thus, even the fool is convinced that there is in the understanding at least something than which nothing greater can be thought, because he understands this when he hears it, and whatever is understood is in the understanding.
(2.10) And certainly that than which a greater cannot be thought cannot be in the understanding alone.
(2.11a) For if it is in the understanding alone, it can also be thought to exist in reality, (2.11b) which is greater.
(2.12) If therefore that than which a greater cannot be thought is in the understanding alone, that same thing than which a greater cannot be thought is [something] than which a greater can be thought.
(2.13) Sed certe hoc esse non potest. (2.13) But this cannot be the case.
(2.14) Therefore without doubt something than which a greater cannot be thought exists, both in the understanding and in reality. (3.1) Quod utique sic vere est, ut nec cogitari possit non esse.
(3.1) Surely this truly exists in such a way, that it cannot be thought not to exist.
(3.2a) For there can be thought to exist something that cannot be thought not to exist; (3.2b) [and] this is greater than what can be thought not to exist.
(3.3) Quare si id quo maius nequit cogitari, potest cogitari non esse: id ipsum quo maius cogitari nequit, non est id quo maius cogitari nequit; (3.3) Therefore if that than which a greater cannot be thought can be thought not to exist, this same thing than which a greater cannot be thought is not that than which a greater cannot be thought, (3.4) quod convenire non potest. (3.4) which cannot be logically consistent.
(3.5) Therefore something than which a greater cannot be thought truly exists in such a way that it cannot be thought not to exist. (3.7) Sic ergo vere es, domine deus meus, ut nec cogitari possis non esse.
(3.7) Therefore, O Lord my God, You truly exist in such a way that You cannot be thought not to exist.
(3.9) For if some mind could think something better than You, the creature would ascend above the Creator, and would pass judgement on the Creator, which is quite absurd.
(3.10) And indeed whatever is other than You alone can be thought not to exist.
(3.11) Therefore, You alone have being most truly of all, and thus most greatly of all, because whatever is other [than You] truly does not exist in this way, and for this reason has less being.
(3.12) Why then has the fool said in his heart, "There is no God," when it is so obvious to a rational mind that You exist most greatly of all?
(3.13) Cur, nisi quia stultus et insipiens? (3.13) Why, except that he is stupid and foolish? PIV [(4.0 (4.2) If he has truly-or rather because he has truly-both thought it, since he has said it in his heart, and not said it in his heart, since he could not think it, something is not said in the heart or thought in only one way.
(4.3) For a thing is thought in one way when the word signifying it is thought, and in another when the thing itself is understood.
(4.4) Accordingly, God can be thought not to exist in the first way, but not at all in the second.
(4.5) Indeed, no one who understands what God is can think that God does not exist, although he may say these words in [his] heart; be it without any, or with some extraneous, meaning.
(4.6) For God is that than which a greater cannot be thought (4.7) Quod qui bene intelligit, utique intelligit id ipsum sic esse, ut nec cogitatione queat non esse.
(4.7) Whoever understands this [that than which a greater cannot be thought] properly, understands at least that this same thing exists in such a way that not even in thought can it not exist.
(4.8) Therefore, whoever understands that God exists in the same way [as that than which a greater cannot be thought], cannot think that He does not exist.
(4.9) I give thanks to You, good Lord, I give thanks to You, because what I first believed by Your gift, I now understand by Your illumination in such a way that even if I should not want to believe it, I would not be able not to understand that You exist.
The Latin text is taken from Schmitt's edition of P; the enumeration roughly follows Scherb [2000].
The English text is taken from , 32-35, except the punctuation and the enumeration, which follows the Latin text.
Funding Open Access funding enabled and organized by Projekt DEAL.

Conflict of Interest
The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.