by Rev. Dr. Eric Phillips, Ph.D
At certain places in their exchange on the freedom of the will, Erasmus and Luther speak as if the topic is specifically soteriological and in need of being treated solely through Scriptural argumentation. Erasmus opens his introduction with these words: “Since Luther does not acknowledge the authority of any writer, of however distinguished a reputation, but only listens to the canonical Scriptures, how gladly do I welcome this abridgment of labor….”1 Even though he immediately proceeds to claim that all the fathers of the Church are on his side, he does not offer a patristic florilegia, but instead ties the importance of these authorities to their presumed reliability as interpreters of Scripture.2 Then a little later, when it comes time for him to set the rules of the debate by defining “free choice,” he ties it to salvation as closely as possible: “By free choice in this place we mean a power of the human will by which a man can apply himself to the things which lead to eternal salvation, or turn away from them.”3 Unsurprisingly, however, neither writer actually confines himself to these ground rules.
In Erasmus’s case, his objection to Luther goes deeper, to a more fundamental philosophical stratum. He is writing in response to Luther’s Assertion, he says, meaning the Assertio Omnium Articulorum M. Lutheri per Bullam Leonis X of 1520, in which Luther responds publically to the articles of condemnation listed in Exsurge Domine, Leo X’s bull of excommunication. Article 36 of the bull condemned Thesis 13 of Luther’s Heidelberg Disputation (“Free will, after sin, is a matter of a label only, and when one does what is in himself, he sins mortally”),4 and it is Luther’s response to this article that Erasmus attacks. All the Old Testament passages he treats in part III of the Diatribe (“Examination of Luther’s Arguments”) come from it, and he handles them in exactly the same order as Luther introduces them in Assertion 36.5 He gives special attention, though, to a passage that appears in the Assertion right after Luther’s last argument from the Old Testament, using it as a sort of exclamation mark to close part I of his critique. Luther wrote:
Wherefore it is needful to retract this article. For I was wrong in saying that free choice before grace is a reality only in name. I should have said simply, ‘free choice is in reality a fiction, or a name without reality.’ For no one has it in his own power to think a good or bad thought, but everything (as Wyclif’s article condemned at Constance rightly teaches) happens by absolute necessity.6
“These are Luther’s actual words,” he solemnly informs the reader.
Erasmus might have chosen Assertion 36 as his target even if it did not contain this passage, but this passage is the bull’s eye, the weakest point in Luther’s armor for several reasons. Most obviously, he finds Luther explicitly agreeing with the anathematized proposition of a condemned heretic, but there is more to it than that. As Erasmus has already said in his introduction, “From the time of the apostles down to the current day, no writer has yet emerged who has totally taken away the power of choice, save only Manichaeus and John Wyclif.”7 Associating Luther with Wyclif is one thing; he still had supporters. A charge of Manichaeanism is the heavy artillery. Erasmus drops this little land mine in passing, with his deft satirist’s touch, hidden — but artfully badly — in an argument that is masquerading as a simple observation that Luther does not have many authorities on his side. It’s like saying, “Only two men could be found to vouch for his character: Jack the Ripper and Charles Manson. That’s not very many, is it?” Then, at the end of his first section, when he has just wrapped up his positive argument (“Passages that Support Free Choice”) and feels himself to be in his strongest position vis-à-vis Luther, he detonates the bomb. “All things before and after grace, good equally with ill, yes even things indifferent, are done by sheer necessity. Which opinion Luther approves.” Can this be believed? Is it possible that the charge of Manichaean determinism is true? Erasmus understands your doubt; it is incredible, but regrettably true. “That no one may suppose me to have invented the charge, I will cite his own words…. These are Luther’s actual words.”8 This is the core, the strongest thrust, of Erasmus’s attack on Martin Luther.
Luther is not impressed. He does not even try to explain or defend himself:
And here, lest anyone should doubt whether Luther ever said anything so absurd, Diatribe quotes my own words, which I frankly acknowledge. For I take the view that Wyclif’s article (that all things happen by necessity) was wrongly condemned by the Council, or rather the conspiracy and sedition, of Constance. What is more, Diatribe herself defends the same position along with me when she asserts that free choice by its own powers can will nothing good but is necessarily in bondage to sin, although in the process of proving this she establishes the direct opposite.9
Luther is being cavalier, and maybe disingenuous as well.10 The statement he is proudly claiming goes well beyond the assertion “that free choice by its own powers can will nothing good.” Erasmus said that free choice needed divine aid to will the good;11 Luther denied that there was such a thing as free choice. In fact, Luther is muddying the waters simply by suggesting that “bondage to sin” has anything to do with Wyclif’s proposition. He is conflating a philosophical position based on divine foreknowledge and/or predestination with a theological position based on the fallen status of the human race. In addition to this, he is confusing absolute necessity, the kind that Wyclif was condemned for (allegedly) teaching, with relative necessity (or “necessity by supposition”), which even Erasmus endorsed, and as a result of that confusion is failing also to distinguish properly, here and everywhere else in his treatise, between “the necessity of the consequent” and “the necessity of consequence.” I will deal with each of these criticisms in turn.
The Argument from Absolute Necessity vs. the Argument from the Fall
In his original thesis at Heidelberg, the one he was defending against the Pope when he voiced his support for Wyclif’s condemned article, Luther argued against the freedom of the human will on strictly theological grounds, carefully attributing it to the preexisting sinfulness of mankind. “Free will, after sin, is a matter of a label only.”12 The next thesis followed with, “Free will, after sin, is capable of good by a passive power, but capable of evil always by an active one.”13 When he endorsed the language attributed to Wyclif in his Assertion, however—perhaps inspired to do that by the fact that Exsurge Domine mentions Wyclifites, Hussites, and the Council of Constance—this explanation no longer made sense. If “everything… happens by absolute necessity,”14 then the fall must have too, and if the fall happened necessarily, then its necessity cannot be attributed to preexisting human sin. Worse, if something other than human sinfulness necessitated the fall, and the fall in turn bound the human will, then the bondage of the will was really made necessary not by the fall (except in a redundant way), but rather by that first something. And as that original necessitating force could only be God, the question “What kind of necessity are we talking about?” becomes acute. Is this just the kind of necessity implied by divine foreknowledge, where if God by virtue of His eternity has already seen you taking a walk tomorrow, there is no way that can fail to happen? If so, absolute necessity is the wrong category. With the word absolute in play, Erasmus is not out of place in raising the specter of Mani.
Absolute vs. Relative Necessity
This distinction requires some understanding of scholastic theology. In the formulation of Thomas Aquinas, “There are two ways in which a thing is said to be necessary, namely, absolutely, and by supposition.”15 Absolute necessity is necessity by definition, i.e. “when the predicate forms part of the definition of the subject: thus it is absolutely necessary that man is an animal” (because man is defined as “a rational animal”) or “when the subject forms part of the notion of the predicate: thus it is absolutely necessary that a number must be odd or even” (because by definition the concepts “odd” and “even” divide all numbers between them).16 Suppositional necessity, on the other hand, is based on a state of affairs. Socrates is able to sit or to stand, but “granted that he is sitting, he must necessarily sit, as long as he is sitting.”17 Both kinds of necessity derive from tautology: given these definitions, or this state of affairs, this proposition has to be true, because it has done nothing but restate the givens. The difference between them lies in the nature of those givens. Absolute necessity is not only necessary in itself, but derives from a necessary given. Suppositional necessity is necessary in itself, but derives from a given that could have been otherwise. Metaphysical definitions are necessary givens for Thomas because he understands all Being to be defined by the Supreme Being that is God; since God is who He is, man must be a rational animal. States of affairs, on the other hand, are contingent givens. Unlike God Himself, or anything metaphysically defined by God, they could in theory (that is, without contradiction) have been different than they are.
This is not the same thing as saying that man must actually exist. Thomas rejects simple emanation as an explanation of creation. “The knowledge of God, joined to His will is the cause of things. Hence it is not necessary that what ever God knows, is, or was, or will be; but only is this necessary as regards what He wills to be, or permits to be.”18 God could have chosen not to make the human race, but given the fact that He did, it must be as He knows it to be, and He knows beings by knowing His own Being. “So we say that God sees Himself in Himself, because He sees Himself through His essence; and He sees other things not in themselves, but in Himself; inasmuch as His essence contains the similitude of things other than Himself.”19 So that man exists in actuality is contingent on God’s free will, but what he must be, whether in actuality or only in potentiality, is defined by God qua Being, and hence is just as necessary as He is. This is why it is such a problem to claim that “everything happens by absolute necessity.” It means that everything — not just the definitions of created essences but also their actualization and everything that happens to them in the world — is not only necessary but just as necessary as God Himself, which can only be true if it all flows by logical consequence from God’s Being. The statement as it stands does not even recognize the mediating contingency of God’s Will, whether in creation or by providence. It teaches pure emanation, which does in fact result in Manichaean-style determinism unless it is wedded to a robust definition of Evil as privation.20
Wyclif himself did not explicitly maintain this position. In fact, he explicitly denied it. In De Dominio Divino I.13, the passage that the editors of Erasmus’ Diatribe in The Collected Works of Erasmus, point to as the origin of this condemned article,21 Wyclif actually argues the opposite position. Wyclif’s editor, Reginald Lane Poole, succinctly summarizes his argument: “The position that the world is an eternal emanation from God would lead to the destruction of free will a) in God, who would then have no choice as to the creation of the world; and b) in man who could in that case have no merit, contrary to Scripture, making God the author of Sin and sin absolutely necessary, with other impossible inferences.”22 In chapter 14 of the same work, we find a statement from Wyclif that looks quite similar to the condemned thesis, but with the crucial difference that it does not contain the word absolute. “It is necessary that all things happen, which will happen.”23 Two sentences later, he adds:
I assume… the abovementioned distinction of necessity, that one thing is necessary from supposition and another is necessary absolutely. Necessity from supposition is any truth having an eternal cause from which it follows formally that it is so, as for example it follows formally from God willing me to run, that I run, and likewise with any truth, even the most contingent…. [Even sins, he goes on to say]. The second branch of necessity, the absolute, is conceded by all, because when something is absolutely necessary, when a truth is described which is unable not to be, it pertains to the divine essence, although any number of such truths may be formally distinct from one another in God.24
Now, when Wyclif calls God’s will concerning some contingent event “an eternal cause” of that event, he does seem to be implying more than just that God permits it to happen. There are problems with the consistency of his system. The appraisal offered by The Stanford Encyclopedia of Philosophy is that Wyclif’s arguments do not work. “He thought that such distinctions enabled him to maintain simultaneously the necessity of all that happens and human freedom… and many times he affirms that it would be heretical to say that all things happen by absolute necessity; but his attempt failed in achieving its goal.” 25 In other words, there are good reasons why the thesis from Constance was attributed to him even though he explicitly denied it. But the point remains that even though the logic of his own system pushed him to affirm it, he knew that it had to be denied. Wyclif himself would condemn Luther’s support for this proposition.
There are significant hints in Bondage of the Will that even Luther knew he had gone too far with the word “absolute.” When he answers Erasmus’ accusation, he does not give an inch rhetorically: “Diatribe quotes my own words, which I frankly acknowledge.”26 But when he restates Wyclif’s supposed thesis, he paraphrases it and omits the word absolute. “Omnia de necessitate absoluta eveniunt” becomes “omnia necessitate fieri.“27 This could be an unintended coincidence of the sort of offhand “quoting” that Luther is known for, but if so it is a consistent one, because the word is also missing from both sections of the work in which he argues at length that everything happens necessarily.28 Thus it seems likely that Luther was silently revising his previous statement. And even if this is not the case, the absence of the term “absolute” from so much talk about “necessity” must prove that it was not essential to his position.
Another thing to consider in Luther’s defense is that his own philosophical training came from the Via Moderna, the Ockhamist school that took issue with Aquinas and the Realist tradition in general on several fundamental points, one of which was the relationship between Being and Will in God. The Nominalists began with the model described above, but pushed the dividing line between metaphysical necessity and divine freedom far to the freedom side of the spectrum. If St. Thomas’s parsing of potentiality and actuality amounts to selective emanation, the Nominalist version is logically-consistent selection: “God’s absolute power refers to the initial set of possibilities which are open to divine actualisation, which is limited only by the condition that their actualisation does not involve contradiction…. Their actualisation results in the present order as we know it, which is defined as the realm of God’s ordained power.”29 In this understanding, metaphysical definitions stem not from the necessary Being of God, but directly from His free Will. As a result, the sphere of absolute necessity becomes much smaller. God Himself is still absolutely necessary, but there is no detail in creation about which we may legitimately say, “It absolutely could not have been otherwise.” If one was used to thinking in this way, i.e. that absolute necessity in the proper sense cannot be predicated of anything but God, he could conceivably miss the significance of the word and take it as a simple intensifier when it showed up in a proposition such as, “all things happen by absolute necessity.” That Luther did think along these lines is demonstrated by one of his recurring themes, one which figures prominently in Bondage of the Will under the famous heading of “Deus Absconditus“: namely, the unknowability of God apart from special revelation. But it also lies, significantly, at the heart of his argument against the distinction between the Necessity of the Consequent and the Necessity of Consequence.
Necessity of the Consequent vs. Necessity of Consequence
Aristotle raised the question whether the foreknowledge of God implies fatalism, and debated it inconclusively. Boethius answered Aristotle’s question by means of this distinction.30 Thomas Aquinas adopted Boethius’s solution and used the names by which Luther and Erasmus refer to the two kinds of necessity: “If each thing is known by God as seen by Him in the present, what is known by God will then have to be. Thus, it is necessary that Socrates be seated from the fact that he is seen seated. But this is not absolutely necessary or, as some say, with the necessity of the consequent; it is necessary conditionally, or with the necessity of the consequence.”31 As Thomas uses it, this is the same distinction as the one discussed above. Observe how he mixes terms in the following quotation: “But when he says the reason why God foreknows some things is because they are future, this must be understood according to the cause of consequence, and not according to the cause of essence. For if things are in the future, it follows that God knows them; but not that the futurity of things is the cause why God knows them.”32 It deserves separate notice, however, because of its ability to survive and be recognized as a valid distinction outside of a Realist philosophical system. This is because it does not matter whether there is such a thing as absolute necessity. All that matters is the recognition that necessity of supposition does not have causal force. The conditional statement “If God knows that someone will sit, then he will sit” expresses a necessary consequential relationship. Given the first half (the supposition), the second is unavoidable. That is Necessity of Consequence. The consequent event, though—the actual sitting—is not necessitated by the knowledge. Erasmus puts it this way:
Prescience is not the cause of things which happen, for it befalls us to foreknow many things which do not happen because we foreknow them, but rather we foreknow them because they are going to happen. Thus the eclipse of the sun does not happen because astrologers predict its occurrence, but they predict its occurrence because it was bound to happen.33
He admits that the interaction of divine foreknowledge and human free will is more complicated than this, because the sun cannot choose to avoid an eclipse, whereas Judas could have chosen not to betray Christ, but if Judas had chosen otherwise, Erasmus explains, “The foreknowledge of God would not have been falsified, nor His will hindered, since He Himself would have foreknown and intended beforehand that Judas should change his mind.”34
Luther takes several different tacks in opposing this argument. He argues against the eclipse example that God’s foreknowledge must not just see the future, but guarantee it, or else it could be thwarted before the event, making Him unreliable.35 This does not address Erasmus’ argument, though, because the eclipse is “bound to happen,” and because even if something completely unforeseeable spoiled the astronomers’ prediction (e.g. the unexplained disappearance of the moon, so that there was nothing left to eclipse the sun) what he said about the case of Judas would still stand: God’s foresight is of the event itself, so it cannot be deceived. Whatever was going to happen would be what He foresaw. Luther argues against the Judas example by accusing Erasmus of maintaining contradictory positions at the same time: “How, I ask you, do these two statements harmonize: ‘Judas can will not to betray’ and ‘It is necessary that Judas should will to betray.’”36 But of course what Erasmus actually said was that if Judas were going to change his mind, that would have been the content of God’s foreknowledge, meaning that there would have been no necessity of his betraying Christ at all. These feeble objections lose their way in the details of the examples, so they do not even get to the point of challenging the distinction between consequence and the consequent.
There is one section where he tackles it head-on, though, and since it came earlier in his treatise and he considered it conclusive, he probably did not feel the need to revisit this argument in the sections we have just discussed.
What they call the necessity of consequence means broadly this: if God wills anything, it is necessary for that thing to come to pass, but it is not necessary that the thing which comes to pass should exist, for God alone exists necessarily, and it is possible for everything else not to exist, if God so wills. … But what do they achieve by this playing with words? … This is no different from saying that the thing done is not God Himself.37
What has Luther done here? First, as I intimated above, he has understood the distinction in a Nominalist fashion, according to which God is the only necessary Being. But should it not still work, even divorced from a Realist framework? The real problem here is that Luther has misconstrued the distinction, changing it from “If God knows that I will sit, it is necessary that I sit, but His knowing does not necessitate my sitting” to “If God knows that I sit, it is necessary that I sit, but that act of sitting is not absolutely necessary.” Why has he made this mistake?
A probable explanation suggests itself when we notice that the two distinctions, different as they seem, are actually equivalent if we interpret them with Thomist assumptions. For Aquinas there is only one way in which the knowledge of God can necessitate a specific consequent: by combining with His Will to actualize one of the definitions implied by His own Essence. That is why he equates the necessity of the consequent with “the cause of essence,”38 i.e. with absolute necessity. Luther is getting stuck in the crack between two different philosophical systems. He knows from the Realist tradition that the necessity of the consequent is the same thing as absolute necessity, and from the Nominalist tradition that nothing is absolutely necessary except God. The problem is, under Realist assumptions there are lots of necessary things besides God (because every actualized essence shares in His absolute necessity), and under Nominalist assumptions, the necessity of the consequent is not the same thing as absolute necessity (because the things God makes are defined by His voluntary ordained power, not His natural absolute power). Luther is combining two mutually exclusive principles.
In conclusion, however convincing Bondage of the Will may be on the question that really mattered to Luther in his debate with Erasmus, in the Assertion, and in his early Disputations—the role of the human will in salvation—it is seriously deficient on the broader question, whether everything happens by [absolute] necessity. Erasmus knew what he was doing when he chose that battlefield. Now I need to read the Hyperaspistes and see if he calls Luther on any of these problems, or whether he himself is too disdainful of “the Sophists” to argue these things out.
1 Erasmus, De Servo Arbitrio, E. Gordon Rupp, Trans. and Ed., in Luther and Erasmus: Free Will and Salvation, E. Gordon Rupp and Philip S. Watson, Trans. and Ed. (Philadelphia: The Westminster Press, 1969), p. 42.
2 “I say it rather of certain others better known to me who, if there is any controversy concerning the meaning of the Scriptures, when we bring forward the authority of the Early Fathers, chant at once, ‘Ah! But they were only men.’ And if you ask them by what argument the true interpretation of Scripture may be known, since both sides are men, they reply, ‘By the sign of the Spirit.’ If you ask why the Spirit should rather be absent from those who have illuminated the world by their published miracles than from themselves, they reply as though for thirteen hundred years there had been no gospel in the world.” Ibid., p. 45.
3 Ibid., p. 47.
4 “Liberum arbitrium post peccatum res est de solo titulo, et dum facit quod in se est, peccat mortaliter.” WA I.354 and VII.142.
5 Compare WA VII.143-45 and Rupp/Watson pp. 74-78.
6 Rupp/Watson, p. 64; WA VII.146. “Omnia de necessitate absoluta eveniunt.”
7 Ibid., p. 43.
8 Ibid., p. 64.
9 Ibid., p. 219. Watson’s translation has quotation marks around the statement, “all things happen by necessity,” but I have removed them because Luther is not actually quoting the debated proposition (see p. 9 below).
10 See p. 9 below.
11 Erasmus states this “probable” position (Luther uses this word throughout his treatise to mock the formulation he is referring to here) this way: “At the other hand, those who, at the other extreme from Pelagius, attribute most of all to grace and practically nothing to free choice, yet do not entirely remove it, for they deny that many can will the good without peculiar grace, they deny that he can make a beginning, they deny that he can progress, they deny that he can reach his goal without the principal and perpetual aid of divine grace. Their view seems probable enough in that it leaves man to study and strive, but it does not leave aught for him to ascribe to his own powers.” Ibid., p. 53.
12 The Heidelberg Disputation, Thesis 13. See note 4 above. In the American Edition, the translation runs, “Free will, after the fall, exists in name only.” Luther’s Works, vol. 31, Harold J. Grimm Ed. (Philadelphia: Muhlenberg Press, 1957), p. 48.
13 The Heidelberg Disputation, Thesis 14: Liberum arbitrium post peccatum potest in bonum potentia subiectiva, in malum vero active semper. WA I.354.
14 Omnia de necessitate absoluta eveniunt. Luther quotes exactly from the decrees of the Council of Constance. Decrees of the Ecumenical Councils, vol. I. Norman P. Tanner, Ed. (Washington: Georgetown University Press, 1990), p. 412.
15 Summa Theologica I.19.3. The Summa Theologica of St. Thomas Aquinas, translated by Fathers of the English Dominican Province, Second and revised edition (London: Burns Oates and Washbourne, 1920), accessed online at http://www.newadvent.org/summa/index.html.
18 ST I.14.9
19 ST I.14.5
20 Erasmus rejects this definition, referring in passing to “that long-exploded idea that sin is nothing” (Rupp/Watson, p. 79). Luther, although he leaves no stone unturned when it comes to criticizing Erasmus’ errors, lets this pass without comment. As an argument from silence, this does not prove that Luther agreed with Erasmus that the idea had no merit, but it does demonstrate at least that he didn’t think it was relevant to the debate.
21 “His doctrine of the absolute necessity of all events (which he expressed in De dominio divino I.13) was condemned by the Council of Constance….” Collected Works of Desiderius Erasmus: Controversies, vol. 76, Peter Macardle, Trans., Charles Trinkhaus, Ed. (Toronto: University of Toronto Press, 1993), pp. 12-13, n. 35.
22 Ioannis Wycliffe: De Dominio Divino Libri Tres, Reginald Lane Poole, Ed. (London: Trubner & Co., 1890; reprinted 1966 by Johnson Reprint Corp., New York and London), pp. 111-115 (spaced out in marginal notes).
23 “[suppono… quod] omnia que evenient sit necessarium evenire.“ Cf. “Omnia de necessitate absoluta eveniunt” from the decrees of Constance and Luther’s Assertion (see note 13 above).
24 “Et [tercio] suppono… supratactam distinccionem de necessario, quod aliquod est necessarium ex supposicione et aliquod est necessarium absolute. Necessarium ex supposicione est quecunque veritas habens causam eternam ex qua sequitur formaliter ipsam esse; ut me currere sequitur formaliter ex Deum velle me currere, et ita de qualibet eciam contingentissima veritate. … Secundum membrum de necessario absolute a cunctis conceditur, cum necessarium absolute, cum sit descriptive veritas que non potest non esse, est in re divina essencia, licet sint quotlibet veritates tales in Deo formaliter condistincte.” Ibid. I.14, p. 116.
25 Conti, Alessandro, “John Wyclif”, The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.), forthcoming URL = <http://plato.stanford.edu/archives/win2011/entries/wyclif/>.
26 Rupp/Watson, p. 219.
27 WA 7:146 vs. WA18:699.
28 These are the sections on contingency and necessity (Rupp/Watson, pp. 117-124) and on how God’s foreknowledge imposes necessity (239-246).
29 Alister McGrath, Iustitia Dei, 3rd ed. (Cambridge: Cambridge University Press, 2005), p. 151.
30 Aristotle in On Interpretation 9, Boethius in Consolation of Philosophy 5.
31 Summa Contra Gentiles I.67, Anton C. Pegis, Trans. http://dhspriory.org/thomas/ContraGentiles.htm.
32 ST I.13.8.
33 Rupp/Watson, p. 68.
34 Ibid., p. 68.
35 Ibid., p. 241.
36 Ibid., p. 248.
37 Ibid., p. 120.
38 See the quotation from ST I.13.8 on p. 11 above.