This paper argues that the Renaissance philosopher Lorenzo Valla's De libero arbitrio (1439) appears to anticipate Luther's De servo arbitrio (1525) on several key issues.Footnote 1 This thesis was defended in the 1940s by the well-known Renaissance scholar Charles Trinkaus, who expressed regret over the fact that modern interpreters of Valla's dialogue had “failed to see its connection with the Reformation.”Footnote 2 As we shall see further on, however, Trinkaus's position does not seem to have had much impact on the subsequent literature.
In seeking to rehabilitate Trinkaus's thesis, I argue that Valla's De libero arbitrio appears to advocate a position on determinism and free will very similar to that of Luther in De servo arbitrio. Indeed, the parallels between what seem to be Valla's and Luther's respective positions on these topics are so striking that any attempt to reconstruct the historical-theological background of Luther's De servo arbitrio would do well to take Valla's De libero arbitrio into serious account.
An Overview of Valla's De libero arbitrio
De libero arbitrio narrates a dialogue between Valla (referred to by his first name, “Lorenzo”) and a certain Antonio. The dialogue is relatively short; in Maria Anfossi's Latin edition of the text the dialogue comprises a mere 846 lines.
To get an overview of the dialogue I shall divide it into seven parts, which I identify using the numbered line system provided in Anfossi's Latin edition of the text. I also provide quotations from some of the more important passages, which I label as passages A–H to facilitate subsequent referencing.
The seven “parts” of the dialogue are as follows. Part 1 (lines 1–79) contains an introductory exhortation to theologians “not [to] depend so much on philosophy,” along with a statement of the purpose of the dialogue, which is said to be
(A) to show that Boethius (for no other reason than that he loved philosophy excessively) argued incorrectly about free will in the fifth book of his Consolation of Philosophy.Footnote 3
Part 2 (lines 80–203) introduces the dialogue proper. Antonio gives his reason for not considering Boethius's solution to the free will/foreknowledge problem satisfactory:
(B) [Boethius], to whom all give the palm in explaining this question, is himself unable to complete what he undertakes. . . . For he says that God, through an intelligence which is beyond reason, both knows all things for eternity and holds all things present. But can I, who am rational and know nothing outside of time, aspire to the knowledge and intelligence of eternity? I suspect Boethius himself did not understand them, even if the things he said were true, which I do not believe.Footnote 4
Antonio then goes on to specify the topic that he wishes to discuss with Valla:
(C) Lo[renzo]. What do you ask me to explain? An[tonio]. Whether the foreknowledge of God stands in the way of free will and whether Boethius has correctly argued this question.Footnote 5
We see here that there are two questions Antonio wants to discuss, namely 1) whether divine foreknowledge is reconcilable with free will, and 2) whether Boethius has argued correctly on this question. These questions are dealt with, in this order, in parts 5 and 6 of the dialogue.
In part 3 (lines 204–29) Valla requests Antonio to “get into the very heart of the question.”Footnote 6 In response, Antonio puts forth the following argument:
(D) If God foresees the future, it cannot happen otherwise than He foresaw. For example, if He sees that Judas will be a traitor, it is impossible for him not to become a traitor, that is, it is necessary for Judas to betray, unless—which should be far from us—we assume God to lack providence. Since He has providence, one must undoubtedly believe that mankind does not have free will in its own power; and I do not speak particularly of evil men, for as it is necessary for these to do evil, so conversely it is necessary for the good to do good, provided those are still to be called good or evil who lack will or that their actions are to be considered right or wrong which are necessary and forced.Footnote 7
This argument contains an important assumption that Valla is to pick up and criticize in part 4: to wit, if God has foreknowledge of all things, then it is “necessary” that these things take place, in which case people lack free will.
Having presented the above argument Antonio proceeds to state what he takes to be a disturbing consequence of his argument, namely, that if people have free will, then God lacks foreknowledge, but that if God has foreknowledge and people consequently lack free will God is unjust in punishing sinners.Footnote 8
In part 4 (lines 230–435) Valla attacks an assumption of Antonio's foregoing argument on the basis of a distinction between what “can” and what “will” happen:
(E) Something that can happen and something that will happen are very different. I can be a husband, I can be a soldier or a priest, but will I right away? Not at all. Though I can do otherwise than will happen, nevertheless I shall not do otherwise; and it was in Judas’ power not to sin even though it was foreseen that he would, but he preferred to sin, which it was foreseen would happen. Thus foreknowledge is valid and free will abides.Footnote 9
After a debate Antonio finally concedes Valla's argument, remarking that his argumentative weapons have been “smashed” and that he will “object no further.”Footnote 10
Part 5 (lines 436–517) opens with Valla first noting that he has thus far not put forth “what I had decided to say” but only what “defense itself demanded.”Footnote 11 He then sets out to explain “what persuades me . . . that foreknowledge is no impediment to free will.”Footnote 12 This topic corresponds to the first of the two questions that Antonio raised in part 2, i.e., whether divine foreknowledge and free will are reconcilable.
In arguing that divine foreknowledge is reconcilable with free will, Valla makes use of a fable. In the fable, Sextus Tarquinius consults Apollo, who knows the future, about the remainder of his life, and Apollo replies that Sextus will commit many crimes and then be killed. Horrified, Sextus asks Apollo to repeal his prophecy, but Apollo responds that he merely “knows” fates and that Jupiter is the one who “decide[s]” them.Footnote 13 From Apollo's response Valla takes it to be evident that foreknowledge does not as such determine the future, and so Antonio's worries over whether foreknowledge and free will are reconcilable are taken to be “answered.”Footnote 14 Antonio concedes Valla's conclusion, saying that the matter is now “fully solved.”Footnote 15
Part 6 (lines 518–764) is concerned with the second of Antonio's two questions mentioned in part 2, i.e., “whether Boethius has correctly argued this question.” In addressing this question Valla returns to his above fable, and asks Antonio how he thinks Sextus might have responded to Apollo's claim that Jupiter decides fates. Valla suggests that Sextus might have protested the claim via an appeal to his own free will. But to this Valla has Apollo answer as follows:
(F) This is the way things are, Sextus. Jupiter as he created the wolf fierce, the hare timid, the lion brave, the ass stupid, the dog savage, the sheep mild, so he fashioned some men hard of heart, others soft, he generated one given to evil, the other to virtue, and, further, he gave a capacity for reform to one and made another incorrigible. To you, indeed, he assigned an evil soul with no resources for reform. And so both you, for your inborn character, will do evil, and Jupiter, on account of your actions and their evil effects, will punish sternly, and thus he has sworn by the Stygian swamp it will be.Footnote 16
At this point Antonio retorts that if this is the case Sextus might ask how Jupiter can be just in punishing him.Footnote 17 To this Valla replies:
(G) This is what I wished to say for my proof. For this is the point of my fable, that, although the wisdom of God cannot be separated from His power and will, I may by this advice of Apollo and Jupiter separate them. What cannot be achieved with one god may be achieved with two, each having his own proper nature—the one for creating the character of men, the other for knowing—that it may appear that providence is not the cause of necessity but that all this whatever it is must be referred to the will of God.Footnote 18
Antonio is baffled by this turn in Valla's reasoning, saying that this is to “annul free will” by “the will of God.”Footnote 19 Antonio accordingly asks Valla to “solve” the further issue of reconciling human free will with the will of God. Valla replies that this further topic is not part of what they initially agreed to discuss (see part 2). Nevertheless, Valla points Antonio in the direction he believes he should go to look for an answer to this question—to wit, chapter 9 of the apostle Paul's Romans. The apostle Paul there says, with respect to God's election of some and hardening of others, that “it is not of him that willeth, nor of him that runneth, but of God that showeth mercy,” and that God's reasons for electing and hardening are simply beyond human comprehension (which, however, does not prevent us from knowing that the reasons, whatever they are, are “most worthy,” inasmuch as God is “most wise and good”).Footnote 20
Lorenzo then proceeds to use Paul's statement in Romans 9 to attack what he takes to be the heart of Boethius's position:
(H) Listen, for this is what I wished to say: Paul first said, “So then it is not of him that willeth, nor of [man] that runneth, but of God that showeth mercy.” But Boethius in his whole argument concludes, not actually in words but in substance: It is not of God who forseeth, but of man who willeth and runneth. Thence it is not enough to dispute about the providence of God unless the will (of God) is also discussed.Footnote 21
Antonio finds Valla's reasoning persuasive, saying that Valla has argued “a most true opinion of Boethius.”Footnote 22
In part 7 (lines 765–846) Valla brings the discussion to a conclusion, returning to and reiterating his earlier exhortation (in part 1) not to depend too much on philosophy.
Interpretation 1: Valla as Proponent of Libertarian Free Will
Valla's De libero arbitrio has been interpreted in three main ways: 1) as advocating libertarian free will, 2) as urging no definite position on free will, and 3) as opposing libertarian free will. In sections 2–5, I survey these three interpretations, and argue that whereas the first two are inadequate the third one seems correct.
I begin with the view that Valla's dialogue advocates libertarian free will.Footnote 23 This interpretation can be found in the work of a number of scholars not primarily concerned with Valla but with the Luther-Erasmus debate. In Luthers Lehre vom unfreien Willen (1967), for example, Harry J. McSorley describes Valla's De libero arbitrio as “a work in which the natural freedom of man's will—in the traditional Catholic sense—is defended” (where free will in “the traditional Catholic sense” is taken to be inconsistent with determinism).Footnote 24 A similar view is expressed in the footnotes provided by Peter Macardle, Clarence H. Miller, and Charles Trinkaus in Macardle and Miller's translation of Erasmus's De libero arbitrio diatribe sive collatio: “Valla's own rhetorical theology was not discordant with that of Erasmus, and in his own terms Valla strongly favoured free will.”Footnote 25
Macardle et al. provide no textual support for their reading of Valla. McSorley, however, appeals to Valla's claim in part 4 of De libero arbitrio (see passage E) that divine foreknowledge is not inconsistent with free will:
Valla says: “. . . I deny that if it is possible for something to happen in another way than has been foreknown, it follows that the foreknowledge can err. . . .” Valla is certain both of man's free will and God's foreknowledge. But how they are to be reconciled, he says, cannot be known by reason—only by faith.Footnote 26
There are at least three problems with this interpretation.
First, the interpretation would only be plausible if Valla, in holding free will to be reconcilable with divine foreknowledge, uses the term “free will” in a non-libertarian sense in which free will is inconsistent not simply with coercion but with the possibility that people act otherwise than they do. However, McSorley provides no evidence for thinking that Valla understands the term “free will” in this sense. On the contrary, he concedes in a footnote that Valla “conceives of no other necessity than . . . ‘necessitas coactionis,’” i.e., necessity in the sense of absence of coercion.Footnote 27 Hence, Valla's claim that free will is reconcilable with divine foreknowledge does not show that he holds people to have free will in the libertarian sense.
Second, McSorley misinterprets Valla's claim about “what cannot be known by reason.” Valla's claim is not that the reconciliation of free will and divine foreknowledge is something that “cannot be known by reason.” Rather, Valla takes free will and divine foreknowledge to be rationally reconcilable on the basis of a distinction between what “can” happen and what “will” happen (see passage E). As we saw in connection with our survey of part 6 of the dialogue, what Valla regards as beyond rational comprehension pertains to God's reason for electing some and hardening others, which is something quite different. Accordingly, Valla's claim about “what cannot be known by reason” does not show that Valla concedes the existence of libertarian free will.
And third, the idea that Valla's dialogue advocates libertarian free will sits uncomfortably with Valla's affirmation in passage G that all things whatsoever—including human volitional action—take place of necessity by virtue of the will of God. Indeed, this affirmation militates against libertarian free will.
Interpretation 2: Valla as Indefinite on Free Will
We turn next to the view that Valla is not arguing either for or against free will in De libero arbitrio, but has some other aim in mind. An early variant of this view is found in Ernst Cassirer's classic study The Individual and the Cosmos in Renaissance Philosophy (1964), and a more recent variant is urged in a paper by Margaret Cameron.
According to Cassirer, Valla's aim in the dialogue is not to advocate a particular position on free will, but instead to pose a question to his readers and then leave it up to them to make up their own minds. Says Cassirer:
How is it possible that man, a creature, owes the whole of his being to God and yet is gifted with freedom of the will by virtue of which he is responsible for his decisions? . . . Neither here nor elsewhere does he [i.e., Valla] want to present us with final conclusions. He is satisfied with placing the question before us in its sharpest possible formulation; then he leaves us with the question.Footnote 28
Valla's alleged aim is taken to accord with a secularist and antidogmatic thrust in Valla's work. Indeed, Cassirer sees Valla as a forerunner of eighteenth-century Enlightenment rationalism:
The problem of freedom is taken before an entirely secular forum, before the bench of “natural reason.” . . . Valla created the form of dogma-criticism used in the seventeenth century by Bayle and even in the eighteenth century by Lessing. Valla leaves the decision to other instances; but he demands that the investigation be conducted exclusively from the point of view of reason and with its means.Footnote 29
How does Cassirer defend his view that Valla does not advocate a definite position on free will but seeks rather to pose a question? Perhaps the closest we come to a defense is what he says in connection with his exposition of the section of Valla's fable quoted above in passage F:
[Apollo] may summon Sextus before the tribunal of Jupiter, who gave him this tendency, this disposition of the will. And at this point, Valla's investigation breaks off. How is it possible that man, a creature, owes the whole of his being to God and yet is gifted with freedom of the will by virtue of which he is responsible for his decisions? This question, according to Valla, can no longer be answered by philosophy. Only resignation remains, only refuge in mystery.Footnote 30
Cassirer's claim, then, is that Valla's investigation “breaks off” after it has been said that Jupiter decides fates, and this is taken to indicate that Valla is not defending a definite position for or against free will.
As a defense of his interpretation this will not suffice, however. For, to begin with, it is not clear what Cassirer means when he says that Valla's investigation “breaks off” in passage F. If we suppose he means that Valla says nothing more on the topic of free will after passage F, his claim is simply false. After passage F, Valla goes on to say several important things pertaining to this topic, such as that the “point” of his fable is that all things “must be referred to the will of God” (passage G), which has a clear bearing on the topic of free will. So Cassirer fails to provide adequate support for his interpretation.
We turn next to Cameron's more recent variant of the present view. According to Cameron, Valla's aim in De libero arbitrio is to undermine Boethius's attempt to provide “consolation” by means of what is variously termed “rational transformation,” “reason,” and “rational argument”:
By rejecting the possibility of rational transformation, Valla aimed to eliminate reason as a source of consolation. Ultimately, what matters for Valla is the way in which consolation is sought and answers are given, and the allegorical delivery of the answer to the dilemma is supposed to be more satisfying than Boethius’ rational argument.Footnote 31
In place of Boethius's rationalistic means of consolation, Valla is said to have proposed “the art of persuasion”:
[Valla sought to] replace rational consolation with something more persuasive, namely to use rhetorical tools to show that only the love of God (caritas) could soothe the soul. . . . Valla turned to what he considered to be the correct approach to the conflict between divine foreknowledge and human freedom: the art of persuasion. Persuasion here comes in the form of pagan myth [i.e., the fable about Sextus].Footnote 32
Moreover, it is not part of the relevant “art of persuasion” to defend any definite position on the question of free will:
Without doubt, Valla used his dialogue as an opportunity to mock the twists and confusions of dialectical debate, but at the same time he tended to bow out of philosophical debate simply by objecting to its technical terminology. . . . Ultimately, Valla rejected Boethius's solution on the grounds that it absurdly requires the human mind to grasp something which is itself inaccessible to mere mortals; that is, limited reason is asked to comprehend the infinite, or at least to know that God's understanding is infinite, which Valla thought amounted to the same.Footnote 33
This interpretation is very problematic. As her textual evidence for the idea that Valla's “art of persuasion” (as manifested in the fable about Sextus) does not take any definite position on free will, Cameron appeals to the criticisms of Boethius's position found in passage B. What Cameron misses here, however, is that the criticisms in passage B are not Valla's but Antonio’s. Valla's criticisms of Boethius are given in passage H, and center on the quite different charge that Boethius puts the will of human beings in the place of the will of God.
There are also more general problems with Cameron's interpretation. For example, she takes Valla's dialogue to aim at offering “consolation,” where consolation is understood as a means of “sooth[ing] the soul.” But this does not square well with the textual evidence: as is made clear in passages A and C, the aim of the dialogue is to “show that Boethius . . . argued incorrectly about free will in the fifth book of his Consolation of Philosophy” by discussing “whether the foreknowledge of God stands in the way of free will” and “whether Boethius has correctly argued this question.” This is not aimed at providing “consolation” in the sense of a means for soothing the soul.
Interpretation 3: Valla as Opponent of Libertarian Free Will
We turn now to a third interpretation, according to which Valla's dialogue is concerned with opposing libertarian free will, or with espousing a form of determinism (which amounts to the same thing). This is the view of Trinkaus.Footnote 34 According to Trinkaus, modern commentators on Valla's dialogue have largely failed to see the dialogue's “connection with the Reformation.”Footnote 35 To make the Valla-Luther connection evident, Trinkaus points out that Erasmus singles out Valla as a precursor of Luther in the context of their famous free will debate,Footnote 36 and notes that Luther himself in a Tischrede remark praises Valla as “the best Italian I’ve ever seen or heard of in all my life,” and also says that Valla “argued well about free will.”Footnote 37 The connection is viewed as especially evident with respect to the irreconcilability of free will and divine predestination, which Trinkaus takes to be a major theme not only in Luther but also in Valla's dialogue (as is seen from the use that Valla makes of the predestinarian passage in Romans 9 in part 6 of the dialogue):
[Valla] denied that there was any way open to man for comprehending the contradiction of man's seeming possession of free will and God's powers exercised in “hardening” this man and “showing mercy” on that one. While divine foreknowledge might not contradict human freedom, divine predestination seemed to.Footnote 38
Trinkaus takes Valla's above position to amount to a sort of “determinism.”Footnote 39 When it comes to explaining this determinism in greater detail, however, he resorts to an idea, urged by the nineteenth-century Italian scholar Luciano Barozzi, according to which Valla is advocating a form of “psychological determinism” akin to nineteenth-century positivism: “Barozzi . . . seems to have justly seen that . . . like modern positivist philosophers, Valla solved the problem of human liberty by a psychological determinism.”Footnote 40 The textual evidence for this is taken from Valla's words in passage G:
just as God “created the wolf fierce, the hare timid . . . so he generated some men hard of heart, others soft, he generated one given to evil, the other to virtue, and further he gave a capacity of reform to one and made another incorrigible.” The individual's “inborn character” will determine voluntary behavior. Footnote 41
For reasons to be given shortly, I think Trinkaus's view that Valla's dialogue opposes libertarian free will is correct. I think Trinkaus goes wrong, however, in describing Valla's non-libertarianism as a form of psychological determinism. Trinkaus bases this view mainly on the part of Valla's fable in which Apollo says to Sextus that “you, for your inborn character, will do evil” (see passage F). But this will not do as evidence for psychological determinism, for the point of the fable is not that people do what they do on account of their “inborn character” but rather that “all things” (totum) are “necessary” (necessitates) by virtue of “the will of God” (voluntatem Dei). Moreover, Valla even concedes in passage H that humans act willfully although he maintains at the same time that their wills cannot prevail over the will of God (for in saying that “it is not of him that willeth” it is conceded that people “willeth”). Given this, it would seem more correct to characterize Valla's position not as psychological determinism but as theological determinism, and more specifically as a theological determinism that allows people to act willfully as opposed to under external coercion.
In Defense of Interpretation 3
I now proceed to adduce two main reasons in favor of interpretation 3, and also address one main objection to this view.
Two main reasons in favor of this interpretation are as follows. First, Valla is unequivocal in his affirmation in passage G that although “providence is not the cause of necessity” it is nevertheless the case that “all this whatever it is must be referred to the will of God.” As the context makes clear, by the term “all this” (totum) Valla is referring to all things that take place in God's creation, including in particular human volitional action. By saying that all things “must be referred to the will of God,” Valla seems to be saying that even human volitional action takes place of necessity by virtue of the will of God. Since this is inconsistent with libertarian free will, it seems that Valla is here denying libertarian free will.
And second, it is clear from passage H that Valla's main objection to Boethius is that Boethius puts the will of human beings in the place of the will of God. Now the view of Boethius that Valla is here criticizing is simply that people have libertarian free will and that this sort of free will can be seen to be consistent with divine foreknowledge via considerations pertaining to metaphysical relations between God and time. Hence, if Valla thinks that Boethius puts the will of human beings in the place of the will of God, this could only mean that Valla takes libertarian free will to be inconsistent with the all-powerful, all-determining will of God. Once again, then, Valla comes out as an opponent of libertarian free will.
One main objection to the present view is as follows. Valla argues (e.g., in passage E) that divine foreknowledge and free will are reconcilable, which might be regarded as an attempt on the part of Valla to avoid determinism. However, and as Valla makes plain in passage H, the point in opposing this idea is not to show that the doctrine of God per se is reconcilable with free will but rather that it is “not enough” (non satis) to consider this question only by considering divine foreknowledge and without taking into account also “the will” (voluntate) of God.
It seems, then, that Trinkaus's view that Valla's dialogue opposes libertarian free will is defensible. By way of summary, we can express two main points of De libero arbitrio in the following theses: firstly, that divine foreknowledge in conjunction with the divine will entails that all things take place necessarily by virtue of the will of God; and secondly, that human action is still something willful as opposed to being subject to external coercion. In what follows I shall refer to the combination of these two theses as “Valla-style determinism.”
Luther's Familiarity with Valla's De libero arbitrio
At an early stage in his theological career, in thesis 13 of his Disputatio Heidelbergae habita (1518), Luther defended the thesis that “after the Fall” (post peccatum, literally “after sin”) “free will . . . exists in name only.”Footnote 42 This thesis was condemned by Leo X in his papal bull Exsurge Domine (1520).Footnote 43 In responding to Leo X's condemnation, in article 36 of his Assertio omnium articulorum (1520), Luther revised and radicalized his earlier thesis, saying now that he “should have simply said ‘free will is a fiction among real things, a name with no reality’” and that “all things occur by absolute necessity.”Footnote 44 Erasmus criticized this radicalized thesis in his De libero arbitrio diatribe sive collatio (1524),Footnote 45 to which Luther replied with his De servo arbitrio, in which he reiterated and expanded on his view that free will is a fiction and that all things take place by necessity. So content was Luther with his response to Erasmus that he in later years spoke of De servo arbitrio as one of his very best theological books.Footnote 46
In the remainder of this paper I will seek to vindicate Trinkaus's previously mentioned thesis that Luther's position in De servo arbitrio was anticipated by Valla's De libero arbitrio on several key points. As was noted earlier, this claim does not seem to have had much of an impact in the literature. Thus, in well-known overviews of the historical-theological background of Luther's views on free will, Valla's dialogue is often not mentioned at all or else mentioned mostly in passing.Footnote 47 If we turn to specialist studies on De servo arbitrio, things are not much different. For example, in Gerhard Rost's major study, Der Prädestinationsgedanke in der Theologie Martin Luthers, Valla's work merits no attention at all.Footnote 48 In McSorley's meticulously researched Luthers Lehre vom unfreien Willen (quoted in the foregoing), the standard modern study of De servo arbitrio, Valla's De libero arbitrio is mentioned only very briefly, and is described as “a work in which the natural freedom of man's will—in the traditional Catholic sense—is defended.”Footnote 49 Linwood Urban's “Was Luther a Thoroughgoing Determinist?” (1971)—one of the most detailed studies of the historical-theological background of De servo arbitrio after McSorley's book—does not mention Valla's dialogue at all,Footnote 50 nor does Bengt Hägglund's oft-cited “Die Frage der Willensfreiheit in der Auseinandersetzung zwischen Erasmus und Luther.”Footnote 51 One of the more recent scholarly studies of De servo arbitrio, Robert Kolb's Bound Choice, Election, and the Wittenberg Theological Method (2005), contains a brief reference to “Valla's rejection of the freedom of the will,” but only for the purpose of supporting the claim that “the junior professor from Wittenberg did not invent the issue of the freedom of the will.”Footnote 52 Apart from this, there is no discussion of the relation between Valla and Luther in Kolb's study either.Footnote 53 Instead, accounts of the historical-theological background of Luther's views on free will typically focus on the anti-Pelagian writings of Augustine, which are taken to have exerted a major “positive” influence on Luther, and on Pelagian or semi-Pelagian ideas of medieval nominalists like Gabriel Biel, which are taken to have exerted a major “negative” influence on Luther, in the sense that they provided the position against which Luther reacted critically.Footnote 54 At the same time, however, it is recognized that while the anti-Pelagian writings of Augustine support Luther's attack on Pelagian or semi-Pelagian ideas, they do not urge anything like the deterministic denial of free will that (as we shall see shortly) we find in De servo arbitrio. For this reason the deterministic denial of free will in De servo arbitrio has often been viewed as a novel development in the theology of Luther, with little historical precedent.Footnote 55
As far as I can see, this seeming lack of attention to Valla in studies of the historical-theological background of De servo arbitrio can only be due to a failure to appreciate the extent to which Luther was acquainted with, and highly valued, Valla's De libero arbitrio. Accordingly, my first task in seeking to rehabilitate Trinkaus's thesis will be to provide evidence for Luther's acquaintance with and high regard for Valla's De libero arbitrio prior to his composition of De servo arbitrio in 1525.Footnote 56
Contrary to what Trinkaus suggests, the main evidence does not derive from Luther's Tischreden remark that “[Valla] argued well about free will,” for this remark is from 1532 and so could—for all we know—be based on a post-1525 reading of Valla. Nor does the evidence derive from Erasmus's claim in De libero arbitrio diatribe sive collatio that “[Valla] almost seems to agree” with Luther,Footnote 57 or from Luther's concession of this claim in De servo arbitrio (“on my side there is just Wycliffe and Laurentius Valla”Footnote 58); for neither Erasmus nor Luther reproduce any content from Valla's dialogue in the context of their debate and so it cannot be known on the basis of these remarks whether Luther was familiar with the argument of Valla's dialogue at this stage. What we need, rather, is evidence of an acquaintance, on the part of Luther, with Valla's De libero arbitrio at some point prior to 1525, and preferably also some indication that Luther understood Valla as opposing a libertarian doctrine of free will. Such evidence, I suggest, can be gleaned from the Loci communes (1521) of Philip Melanchthon, Luther's close collaborator. Melanchthon's Loci communes was based on Melanchthon's lectures at Wittenberg University in 1519–1520.Footnote 59 The first Lutheran dogmatics ever written, it quickly became immensely influential in Germany. It was also greatly admired by Luther, to such an extent that Luther, in De servo arbitrio, praises the book as an “unsurpassed volume on the doctrines of theology.”Footnote 60 Interestingly, the first chapter of this book is devoted to the question of “free will” (liberum arbitrium), and early on in his discussion Melanchthon states, boldly and triumphantly, that “Valla . . . refuted the position of the Schools on free will.”Footnote 61 This claim makes it very probable that not only Melanchthon but also other Wittenberg Reformers, including Luther, were acquainted with Valla's De libero arbitrio already in the early 1520s (and hence prior to Luther's composition of De servo arbitrio in 1525), that they esteemed it highly, and that they understood it as arguing against a libertarian doctrine of free will (i.e., against what they took to be the position of the “Schools”).Footnote 62
Valla-Style Determinism in De servo arbitrio
I now proceed to argue that Luther's De servo arbitrio endorses what I have previously termed Valla-style determinism, i.e., the theses that divine foreknowledge in conjunction with the divine will entails that all things take place necessarily by virtue of the will of God, and that human action is nevertheless willful rather than subject to external coercion.
I start with the thesis that divine foreknowledge in conjunction with the divine will entails that all things take place necessarily by virtue of the will of God. To see that Luther endorses this thesis in De servo arbitrio, I begin by noting that he endorses the subordinate claim that all things take place of necessity by the will of God, which we see, for example, in his claim that “[God] does all things according to His own immutable, eternal and infallible will,”Footnote 63 and in particular in his claim that “all we do, however it may appear to us to be done mutably and contingently, is in reality done necessarily and immutably in respect to God's will.”Footnote 64 Furthermore, that it is divine foreknowledge in conjunction with the divine will that enables us to see that all things take place of necessity by virtue of the will of God is also clearly affirmed by Luther, in particular in the following well-known passage:
If He wills what He foreknows, His will is eternal and changeless, because His nature is so. From which it follows, by resistless logic, that all we do, however it may appear to us to be done mutably and contingently, is in reality done necessarily and immutably in respect to God's will. . . . Since, then, His will cannot be impeded, what is done cannot but be done where, when, how, as far as, and by whom, He foresees and wills.Footnote 65
We see in this passage that Luther takes the necessity of events to follow not simply from the claim that “[God] foreknows,” but, rather, from the claims that “[God] wills what he foreknows” and that “[God’s] will cannot be impeded.” As is the case with Valla, then, Luther's idea appears to be that it is the conjunction of divine foreknowledge and the divine will that enables us to see that all things take place of necessity by virtue of the will of God.
Not all interpreters read the above passage in the presently proposed way, however. McSorley, for example, takes the passage to urge the necessity of events from divine foreknowledge apart from considerations pertaining to the divine will. More precisely, he takes Luther to be arguing that the necessity of events follows from the two premises “God foreknows all things” and “whatever God foresees and foreknows must happen necessarily.”Footnote 66 A comparison of this reading with what Luther actually says in the passage makes it clear that this reading does not get the argument right, however, for Luther clearly makes use of considerations pertaining to the divine will in making his inference. A second alternative reading, proposed by Urban, is sensitive to the fact that Luther's argument involves considerations pertaining not only to divine foreknowledge but also to the divine will, but seeks to make sense of this by taking Luther to be proposing two equivalent arguments in the passage, the first one consisting in the inference of the proposition “the event necessarily takes place” from the two premises “if God wills an event, then the event necessarily takes place” and “God necessarily wills the event,” and the second one consisting in the inference of the proposition “p is necessarily true” from the two premises “if God knows p, then p is necessarily true” and “God necessarily knows p” (for any proposition p).Footnote 67 But this reading fails to recognize that Luther argues for the necessity of events on the basis of the conjunction of the divine will and divine foreknowledge, not from divine foreknowledge or the divine will considered in isolation from each other. A third alternative reading, recently proposed by Kolb, involves distinguishing between what is expressed in the relevant passage and what Luther intended to express. Talk of the necessity of events in the passage is said to represent an “experiment” on the part of Luther “in expressing his views,”Footnote 68 but what Luther intended to express is said to be “what it means that God continues to be Creator with absolute and unconditioned power as befits a Creator.”Footnote 69 This reading need not contradict the present understanding of the passage, for it could be maintained that the necessity of events is part of what Luther took God's status as Creator to involve. On the other hand, this reading could also be understood as suggesting that Luther was not inferring the necessity of events from the conjunction of the divine will and divine foreknowledge, but was merely expressing what he took God's status as Creator to amount to. If understood this way, Kolb's reading does indeed conflict with the present understanding of Luther's argument, since Luther is presently being taken to be making a genuine inference in the relevant passage. There are at least two disadvantages to Kolb's interpretation (as presently construed), however. First, the interpretation is grounded in a claim for which no adequate support is provided (and, it seems, for which no adequate support can be provided), namely that in saying that all things take place of necessity Luther intended to express what it means for God to be the Creator. And second, the interpretation does not harmonize well with what Luther actually says in the passage, for in the passage Luther talks of the necessity of events as something that “follows” (sequitur) from the divine will in conjunction with divine foreknowledge, and nowhere suggests that this is a way of expressing what it means for God to be the Creator. I conclude, then, that an understanding of the relevant passage that accords with Valla-style determinism does better justice to what Luther actually says in the passage than do the alternative readings considered above.
I proceed next to argue that De servo arbitrio appears to endorse the further thesis of Valla-style determinism, i.e., that human action is willful rather than subject to external coercion. This thesis might at first seem to conflict with Luther's deterministic line of reasoning in connection with the argument quoted in the foregoing, especially since Luther, immediately after having urged this argument, goes on to claim that the argument “knocks ‘free-will’ flat, and utterly shatters it.”Footnote 70 The conflict appears to be only apparent, however, for in denying that people have free will, Luther makes it clear that he is not denying that human action is willful, but is using the term “free will” (liberum arbitrium) in what he calls the “natural force” (natura vocabuli) of the term, which he understands in a libertarian sense as “a power of freely turning in any direction, yielding to none and subject to none.”Footnote 71 It is accordingly in this specifically libertarian sense of “free will” that Luther claims that his argument knocks free will flat, and not in the sense of willful as opposed to externally coerced action.
That Luther does not intend to deny that human action is willful and not subject to external coercion is something about which Luther is explicit:
I said “of necessity”; I did not say “of compulsion”; I meant, by a necessity, not of compulsion, but of what they call immutability. That is to say: a man without the Spirit of God does not do evil against his will, under pressure, as though he were taken by the scruff of the neck and dragged into it, like a thief or footpad being dragged off against his will to punishment; but he does it spontaneously and voluntarily.Footnote 72
Some commentators have concluded, on the basis of this and similar passages, that De servo arbitrio does not mean to deny free will after all.Footnote 73 As Urban points out, however, this is a mistake, for Luther does not equate “free will” with acting willfully as opposed to under external coercion.Footnote 74
It seems, then, that we are justified in regarding Luther as in full agreement with Valla not only in that all things, including human actions, take place of necessity by virtue of the will of God, but also in that human action is willful rather than subject to external coercion.
The Relative Importance of Valla vis-à-vis Augustine and Wyclif
In the foregoing I have argued that Valla-style determinism surfaces in Luther's De servo arbitrio and that it is very probable that Luther was familiar with Valla's De libero arbitrio prior to his composition of De servo arbitrio in 1525. Against this it might be said that Erasmus claims in his De libero arbitrio diatribe sive collatio that not only Valla but also John Wyclif held Luther's view, and that Luther concedes this in De servo arbitrio, adding also Augustine to the list.Footnote 75 This, it might be thought, indicates that Augustine and Wyclif may have been equally important predecessors of Luther's determinism in De servo arbitrio as Valla. In this last section I seek to underscore the relative importance of Valla vis-à-vis Augustine and Wyclif by arguing that neither Augustine nor Wyclif espouse the distinctive claim of Valla-style determinism (later endorsed by Luther) that the necessity of all events follows from the divine will in conjunction with divine foreknowledge.
We start with Augustine. The locus classicus for theological determinism in Augustine is De civitate Dei 1.5.9–10.Footnote 76 Augustine here argues that divine foreknowledge (praescientia Dei) is compatible with human free will (libera hominis voluntate), and in so arguing he allows that divine foreknowledge entails that there is in God's mind “a definite pattern of causation” (omnium ordo causarum)Footnote 77 according to which all events must unfold, and, in line with this, that “it is necessary for something to be as it is, or happen as it does” (necesse esse ut ita aliquid vel ita fiat).Footnote 78 These things, Augustine thinks, are all compatible with free will. Although there is dispute as to the proper interpretation of De civitate Dei 1.5.9–10, many commentators—including myself—take Augustine here to espouse a version of theological determinism that carries commitments to a non-libertarian conception of free will.Footnote 79 Does Augustine's position, thus understood, differ in any significant way from Valla-style determinism? It does. For an important difference is that unlike Valla (and Luther), Augustine does not infer the necessity of all events from the conjunction of divine foreknowledge and the divine will, but from divine foreknowledge per se. Hence, Augustine's version of theological determinism is not as close to Luther's as is Valla’s.
We turn next to Wyclif. Although there is no evidence that Luther studied Wyclif's writings, we know that after John Eck at the Leipzig Debate in 1519 accused Luther of following the errors of Wyclif condemned by the Council of Constance, Luther immediately took to studying Wyclif's views as they are presented in the acts of this council.Footnote 80 In so doing, he would have come across a seemingly deterministic thesis attributed by the council to Wyclif: “all things that happen, happen from necessity” ([o]mnia, quae eveniunt, absolute necessario eveniunt).Footnote 81 In his Assertio omnium articulorum from the following year, Luther publically sided with this claim of Wyclif’s, saying that “no one has it within his control to intend anything, good or evil, but rather, as was rightly taught by the article of Wyclif which was condemned at Constance, all things occur by absolute necessity.”Footnote 82 It is accordingly clear that Luther regarded his view in Assertio omnium articulorum (and later in De servo arbitrio), that all things take place by necessity to have been anticipated by Wyclif. Moreover, Luther would have known from the acts of the Council of Constance that Wyclif's view was grounded in considerations pertaining to divine foreknowledge, for Wyclif is also said to have held the view that “if Paul is foreknown as damned, he cannot truly repent; that is, he cannot cancel the sin of final impenitence by contrition, or be under obligation not to have sin.”Footnote 83 Nevertheless, the distinctive claim of Valla-style determinism, that the necessity of events follows not from divine foreknowledge per se but from divine foreknowledge in conjunction with the divine will, is not part of Wyclif's determinism as put forth in the acts of the Council of Constance. So this is not a claim that Luther could have derived from Wyclif.Footnote 84
We see then that although Luther took Augustine and Wyclif (in addition to Valla) to be forerunners of his position in De servo arbitrio, he would not have found in either of these writers a position nearly as close to his own as that of Valla in De libero arbitrio. This appears to be strong evidence for the relative importance of Valla's De libero arbitrio, vis-à-vis Augustine and Wyclif, for an understanding of the historical-theological background of Luther's De servo arbitrio.
Summary
In the foregoing I have argued that Valla's De libero arbitrio anticipates Luther's De servo arbitrio on several key points. Like Luther, Valla appears to hold that divine foreknowledge in conjunction with the divine will entails that all things—including human actions—take place of necessity by virtue of the will of God. And like Luther, Valla at the same time appears to hold that human actions are willful rather than subject to external coercion. If my arguments for these conclusions are sound, it would seem that any satisfactory account of the historical-theological background of Luther's De servo arbitrio would do well to take Valla's De libero arbitrio into serious account.
