Demon Semen: Traditional and Metaphysical Assumptions in Early Lutheran and Reformed Treatments of Genesis 6:1-4
[This essay was originally published in Quodlibet Journal: Volume 5 Number 4, October 2003 ISSN: 1526-6575
Sixteenth- and seventeenth-century Protestant commentaries on Genesis 6:1-4 unanimously asserted that the “sons of God” (bĕnê-hā ĕlōhîm) were human, rather than angelic or semi-divine beings. Why, one might ask, did early Lutheran and Reformed exegetes insist upon this non-preternatural interpretation, while their theologies affirmed the existence and activity of angels and demons? The discrepancy here is especially striking since both the Fathers and modern scholars contend “sons of God” indicates fallen angels or members of the Divine Council. In seeking to understand the reasons behind magisterial Protestant interpretations, one finds that, though they affirmed the doctrine of sola Scriptura, they did not ignore tradition or metaphysical assumptions altogether. This essay will examine the various reasons why the early evangelicals chose their particular interpretations. It will conclude by asking, (1) whether objective exegesis of a difficult passage can arise apart from traditional and metaphysical assumptions, and (2) whether the assumptions held by early Lutheran and Reformed commentators can be reevaluated by Protestant interpreters today, using the hermeneutical and theological principles inherent in the evangelical traditions.
The “sons of God” and angelology in pre-Reformation tradition
To better understand the background of the early Protestant interpretations of Genesis 6:1-4, one must recall the exegesis and angelologies of pre-Reformation thinkers. In each period, one sees a connection between a theologian’s identification of the “sons of God” and his answer to the question, Do angels have physical bodies? Generally, the earliest Fathers both assumed an angelic interpretation of Genesis 6:1-4 and also affirmed that angels were somehow corporeal, with fiery or ethereal bodies. Many of these individuals even assumed that the would-be demons fell because they lusted after human women.
Angelology presented a major problem for the Fathers. Though some details emerge in John’s Apocalypse, Scripture in general left theologians with very little specific information about angels. Moreover, confusion existed regarding the Divine Council and Satan, as described in the book of Job. Because of this, some Fathers were tempted to indulge in a level of angelological speculation unparalleled by any other doctrinal locus. This fact may have led Luther and Calvin, temperamentally and doctrinally opposed to such speculation, to avoid anything resembling a fantastical approach to the Genesis text.
Two notable advocates of the belief in corporeal angels, Tatian the Assyrian (110-172) and Tertullian (160-225), also fell into heresy, though there is no necessary connection between their angelology andtheir subsequent heterodox associations. In his “Address to the Greeks,” Tatian used a corporeal view of angels to explain the nature of the Greek pantheon. They are, he argued, material spirits whom “having received their structure from matter and obtained the spirit which inheres in it, became intemperate and greedy . . . choosing what was inferior in matter.” The angelic interpretation of Genesis 6:1-4 fit wellwith Tatian’s purpose, which was to identify the promiscuous Greek gods with demons.
Tertullian, likewise, discussed angels in the context of his polemic against Greek philosophy. While the Reformed writer Francis Turretin would later say the corporeal view of angels stemmed from the Platonists, Tertullian contended that the opposite was true. Indeed, the bulk of Tertullian’s argument is aimed at the rejection of Platonic metaphysics and its opposition to the Christian doctrines of the Incarnation and Resurrection. Tertullian insists that proper Christology is related to proper angelology. Since angels have “solidity in their bodily substance” and can “change into human form,” he argues, Christians should not take the Gnostic approach and deny that Jesus—who is more powerful than angels—could possess a real physical body.
Medieval thinkers increasingly rejected the idea that angels had ethereal bodies, along with the idea that they could procreate with humans. Pseudo-Dionysius was important to medieval angelology, contributing a concern for the angelic hierarchy; he may also have been the first to contend that they were pure spirits.
While Aquinas and most of the Dominicans also believed angels were pure spirits, not all agreed. According to Keck, “Bonaventure and almost all of the Franciscans followed strict Aristotelian hylomorphism and argued that angels are made of both form and some kind of matter (albeit a spiritual matter).” Nevertheless, Bonaventure stopped short of allowing angelic procreation with human women. Keck explains: “For angels to create real human bodies would violate not only the divine economy but also the laws of nature.” It is at this point that the use of synthetic a priori reasoning asserts itself upon the interpretation of Genesis 6:1-4.
Consciously or subconsciously, many exegetes, including the early modern Protestants, would follow Bonaventure and assume that a supernatural identification of the“sons of God” would be metaphysically absurd, unless some kind of deception was involved. Bonaventure suggested that, though spiritual beings do not have biologically functional reproductive organs, they may assume the form of a woman (a succubus), collect human semen, and then assume the form of a man (an incubus) in order to impregnate a woman with the now-demonic seed. This complicated hypothesis demonstrates the lengths to which theologians would go to preserve their metaphysical preconceptions. Reformed and Lutheran thinkers appropriated the incubus-succubus hypothesis frequently, in order to maintain the proper distinction between organic and inorganic beings.
Some medieval thinkers, particularly those actively involved in parish ministry, focused on the disembodiment of demons for soteriological reasons. Demons had no bodies because bodies were the focus of penance, and penance is unavailable to fallen angels. The syllogism, according to Elliot, was asfollows: “reconciliation with God depends on a body; a demon does not have a body; ergo a demon cannot be saved.”
Medieval trends in angelology continued into early modernity. The trend away from an ethereal or corporeal view of angels continued until only a handful of Protestant theologians (e.g. Vorstius, Zanchi, and Grotius) held that view. Almost no one questioned the medieval perspective that angels are incapable of procreation. Thus, there is nothing novel about the early Protestant claim that the “sons of God” episode contained human characters alone. Given the late medieval perspective, the Protestant approach may be little more than the natural reception of this tradition. However, one wonders why, despite their reevaluation of Scripture in other places, they never seriously questioned the purely humanview of the “sons of God.” Hence, one must consider possible philosophical motivations for Lutheran and Reformed interpretations of Genesis 6:1-4.
A preternatural reading of the “sons of God” episode would have presented Lutheran and Reformedcommentators with some disturbing implications. Unlike many modern scholars, to admit the “sons of God” were angelic or semi-divine would entail the acceptance of an historical event that was prima facie repugnant to reason, and downright spooky. Nevertheless, these same theologians accepted otherpreternatural stories at face value. This discrepancy leads Van Gemeren to ask, “Why does the theologyin which creation, miracles, the miraculous birth and resurrection of Jesus have a place, prefer a rational explanation of Genesis 6:1-4?” If Van Gemeren’s fears are justified when he wonders whether “a philosophical theology” has “explained away the difficulties” of the text, a possible breach has occurred within the Protestant hermeneutic, which ostensibly eschewed the use of overtly metaphysical axioms in the context of exegesis.
Despite important differences, both Lutherans and Reformed exegetes insisted on the doctrine of sola Scriptura. This did not mean that rationality or logic was excluded from interpretation, for they often appealed to the analogia Scripturae to summon the logical implications of related scriptural passages. Likewise, it did not mean that tradition was useless as an interpretive guide, since they also appealed to the analogia fidei, when they used chief articles of faith as guides. It did mean, however, that the a priori axioms of reason and ecclesiastical pronouncements were not to take precedence over a clear reading Scripture. In the case of Genesis 6:1-4, it is not always clear what motivated the Lutheran and Reformed readings. At times, however, they made reference to philosophical principles, and indicate fundamental opposition to the possibility of angelic-human procreation. Some forms of this thinking appear to be grounded in synthetic a priori reasoning. The following discussion will focus primarily on the extent to which philosophical preconceptions may have determined Lutheran and Reformed interpretations of the “sons of God” episode.
For early modern commentators of all confessions, the episode of the “sons of God” was originally connected with the subsequent flood account, since they did not operate with a source-critic’s paradigm. Luther held that the “giant” offspring of the “sons of God” were not Sethites but tyrants, whoclaimed greater dignity than they deserved. They were “regarded as, and appeared to be, exceedingly wise and righteous, just as today our kings, princes, popes, bishops, theologians, doctors, jurists, and noblemen from the uppermost rank.” The telling phrase here is “just as today.” Luther saw “giants” inthe Vulgate and likely jumped on the chance to slay his own giants: oppressive religious and secular Tyrannen. He thus chose an interpretation that identified the pope and his political allies with the haughty tyrants who brought about the greatest example of divine wrath in human history. Luther makes the connection explicit when he states, “[Giants] usurped both the government and the priesthood. The pope does the same thing when he usurps both spiritual and temporal power.”
In a letter to George of Anhalt, he also connects the “sons of God” with promiscuous monarchs. Recalling that tyrants “took wives from whomever they desired,” he says, “I hear that some similar events are definitely occurring in secret,” as was the case with “the King of England” (Henry VIII) who was “an unhappy husband.” Likewise, commenting on Genesis 6:1-4, Luther writes:
From the marriage of the sons of God with the (wicked) daughters of the world came mighty men, or giants, that is, proud and arrogant people who appropriated to themselves the rule both in Church and state . . . Since they misused their rule in Church and state, God called them by theopprobrious name “giants,” for He cannot tolerate the arrogant, but scatters “the proud” and puts “down the mighty from their seats” (Luke 1: 51-2).
Luther, entangled in a Church-state crisis of his own, apparently found personal meaning in the reasons for God’s great demonstration of displeasure. He must have identified himself with Noah when he noted that these “giants” considered the latter a heretic, since he “endangered the rule both in Church and state.” With this reading, the whole of Genesis 6 made sense for Luther and provided immediate comfort; it was an attempt to warn the arrogant men who threatened the German monk:
[Events like the flood] are intended to frighten the proud and humble us, that we may learn thatour lives and all that we have depend on God’s approval, who is disposed to give grace to thehumble but to destroy the proud (1 Peter 5:5). But because the world neither understands nor does this, the kings, the mighty, and the righteous constantly fall, one after the other, until thereare abundant examples everywhere of God’s wrath and judgment.
Here, Luther views Genesis 6 as a coherent message to haughty leaders who would crush an earnest monk for rocking (or building) the boat. An angelic identification of the “sons of God” or semi-angelicidentification of the “giants” would not suit Luther’s purpose here.
Luther views Genesis 6 in the context of a spiral toward corruption in antediluvian society. He explainsthat the world, “rejecting God and His Word,” committed the greatest vices such as tyranny and oppression, fornication and adultery.” The themes he mentions form a common thread throughout the Old Testament, which would make his reading seem plausible. Luther spends little time with thepreternatural alternative, since he depicts it merely as a rabbinic error when he writes, “The Jews foolishly explain this expression [“sons of God,” which Luther renders “children of God”] to designate evil spirits from whom came the generation of the ungodly.” Luther’s commentary here is expositional rather thanexegetical. He has a point to make for his readers, and this point is not aided by philology or excessive digression. Nevertheless, his language made it extremely difficult for later Lutheran exegetes to evenconsider a “foolish” and “Jewish” reading (despite the fact that in the first century, Christians typically tookthe preternatural view and a common “Jewish” interpretation was that the “sons of God” were kings).
Perusal of Lutheran commentators after Luther on Genesis 6:1-4 becomes tedious, since most do littlemore than recapitulate Luther’s original interpretation. Orthodox Lutheran theologian Johann Gerhard (1582-1637), writing a century after Luther, demonstrates little change. For him, the “sons of God” areroyal Cainites, who took Sethite wives. He echoes Luther by contending the “son’s of God” were notdemons but tyrannical kings (homines oppressores), despite the perspective of the suspect book of Enoch and the mistaken Fathers. Similarly, echoing Luther, Abraham Calov (1612-1686) brushes the angelic interpretation of Genesis 6:1-4 aside by indicating that it was not the original view, but rather theaccretions of a false tradition; he contends that “Platonic Jews” confused things by saying foolishly that “demons have bodies.” This language, of course, indicates strong reliance on Luther.
Scholars familiar with Luther and his intellectual heirs will know they demonstrate no desire to downplay the role of demons in human affairs. Indeed, laymen aware of the themes of “Ein Feste Burg” are no lessconscious of this. Moreover, Luther admits the possibility that angels might engage in an illusory form of sexual intercourse with humans. Incubi and succubae may well exist, he says, since his pastoral experience has led him to hear of encounters with these entities, and since Augustine believed in them. Nevertheless, he insists that, though Satan “can delude us by taking on the appearance eitherof a young man or of a woman,” the idea “that anything can be born from the union of a devil and a human being is simply untrue.” At this point, Luther mentions a common hypothesis that allows him to maintain his metaphysical assumptions concerning angelic and demonic natures:
[W]hen the devil is in the bed, the young man imagines he has a girl, and the girl that she has a young man. But that anything could be born from this cohabitation, this I do not believe. Yet in many places sorceresses have been consigned to the pyre and burned for having commerce with a devil. If he is able to delude the eyes and ears to have you believe that you are seeing and hearing something that is not really there, how much easier it is for him to delude the sense of touch, which is the most obtuse in our nature!
Here, no physical intercourse takes place; it is a demonic hallucination. Characteristically, Luther refuses to continue with this discussion, since it is too speculative, and excuses himself for his digression bysaying, “We got into this discussion because of the silly ideas of the Jews.” Commentators following Luther provided even less explanation for their interpretations, choosing simply to rest on Luther’s final word.
Thus far, one may detect three potential motives for the Lutheran interpretation: the need to apply the text to contemporary problems, the desire to maintain the traditional approach when possible, and the philosophical assumption that angelic beings cannot engage in true sexual intercourse with humans, since their natures are dissimilar. Note that this third aspect is only implicit, and never serves as a conscious axiom for Luther or his intellectual heirs.
Calvin begins his treatment of Genesis 6 by setting the stage for his interpretation of the text: “theprinciple is to be kept in memory that the world was then as if divided into two parts; because the family of Seth cherished the pure and lawful worship of God, from which the rest had fallen.” Thus, Calvindivides humanity into the “sons of God” or descendents of Seth, and the fallen children of Cain. This perspective is in accord with his emphasis, evident throughout his commentaries, upon the covenantal lineage, culminating in the Nativity. It also coincides with his doctrine of election. He writes, “WhenScripture speaks of the sons of God, sometimes it has respect to eternal election, which extends only tothe lawful heirs.”
Believing the “sons of God” to be Sethites, Calvin vehemently denies the preternatural interpretation:
The ancient figment concerning the intercourse of angels with women is abundantly refuted by its own absurdity [sua absurditate abunde refellitur]; and it is surprising that learned men should formerly have been fascinated by ravings [deliriis] so gross and prodigious.
In this passage, Calvin provides no philological or exegetical evidence. Rather, he appeals to a logical or philosophical problem: the idea is described as absurd and self-refuting. Here and elsewhere, it becomes evident that, whereas Luther’s metaphysical assumptions were implicit in his interpretation, synthetic a priori axioms affect the angelology and exegesis of the early Reformed tradition explicitly.
This is nowhere more evident than in the orthodox Reformed theologian Francis Turretin, who demonstrates latent disdain for matter in his description of angels: “Their dignity is so much the greater as they approach nearer the divine nature and are free from all contact with matter.” This statement indicates that matter is, by definition, undignified. We are not surprised to find, then, that Turretin vehemently rejects the idea that angels could soil themselves with matter. How could he not, when, for him, in his very first paragraph concerning angels, he states his a priori assumption that, by definition,angels are immaterial? Claiming, “Spirit is diametrically opposed to body,” Turretin argues that Scripture attributes “spiritual faculties and operations” such as “intellect and will, and the acts elicited by them” to angels, which “cannot proceed from a corporeal nature.”
To be fair, Turretin would argue, as would most of his predecessors, that his axioms are not based on pure metaphysics, but upon the analogia Scripturae. Thus, he points to the standard passages showing that spirits are not flesh and blood (e.g. Luke 24:39), though one might note that such passages do not deny the potential for angels to take physical forms, only that they do not have flesh and bones. In any case, most of Turretin’s arguments indicate metaphysical concerns, such as his curious argument from symmetry:
[A]s God has made in it creatures merely corporeal which strike the senses (as bodies); others partly corporeal and visible, partly spiritual and invisible (as men); so it was fitting to the perfection of the world that the merely spiritual should be created (such as are angels).
One might argue that Turretin is more likely than others in his tradition to allow metaphysical assumptions to dictate his exegesis of a text. This may be true, but the concerns he makes explicit can be found throughout his tradition, and even on its fringes, as in the case of King James I of England.
Though not included in Reformed circles as a beacon of Reformed orthodoxy, King James’ famous Daemonologie illustrates a common early modern application of synthetic a priori knowledge concerning angelic beings: “So to have a quantitie, is so proper to a solide bodie, that as all Philosophers conclude, it cannot be any more without one, then a spirite can have one.” He does not believe that apparent spiritual bodies are real bodies, but rather “vaine impressions in the aire,” or “bodies of aire.” What of the so-called incubi and succubae? According to James, there are two possibilities, “The one, when the Deuill onelie as a spirite, and stealing out the sperme of a dead bodie, abuses them that way, they not graithlie seeing anie shape or feeling anie thing . . . . The other meane is when he borrowes a dead bodieand so visiblie, and as it seemes unto them naturallie as a man converses with them.” Further, “we know spirites hath no seede proper to themselues, nor yet can they gender one with another.” Most important is his philosophical point, arguing that for demons to have their own semen would be “utterly against all the rules of nature.” This last concern summarizes a standard early modern metaphysical objection to angelic procreation.
Recall again that to spend too much time on this subject would be, to the Reformers, to entertain excessive speculation. Thus, Calvin seems reluctant to accept the angelic interpretation of Genesis 6:1-4,because it tempts the theologian to neglect his task, which is “not to divert the ears with chatter, but to strengthen consciences.” What kind of chatter concerns Calvin? Likely, he is thinking of books outside the Protestant canon, like the Book of Enoch (6:1-6; 8:1-3) and Book of Jubilees (see 5:1). However, he seems even more concerned with any interpretation that would give credence to the idea that pagan mythology is contained in the Genesis narrative.
The connections between the Flood of Genesis and Near Eastern parallels were imperfectly known before the nineteenth-century rediscovery of the Epic of Gilgamesh. However, mythological accounts of divine-human sexuality were familiar to humanistic Protestants like Calvin. These scholars were well acquainted, for example, with Hesiod’s account of the divine motives for the Trojan War, and the creation of Pandora, both of which employ the concept of hēmitheoi, or semi-divine beings. Here, and concerning the wives of Cain and Seth, several late medieval and renaissance commentators discuss—without necessarily confirming them—legendary tales to fill gaps left by the Genesis narrative. Though Calvin comfortably alludes to classical pagan sources elsewhere in his writings, he is uncharacteristically apprehensive about the connections between mythology and Genesis 6, except for a final, passing comment, where he writes:
Moreover, as Satan is an ingenious contriver of falsehoods, by which he would corrupt the truth of God, and in this manner render it suspected, the poets have invented many fables concerning the giants . . . .
It is strange that Calvin does not mention parallel “fables” when they are immediately relevant, but only mentions them at the end of his discussion, in the form of a throwaway remark. When he does note the apparent connection, he indicates that it may cause people to doubt its authenticity, as is evident in the above quotation. Perhaps his anxiety over the issue inclined him to underemphasize the point, but also demanded that he fend off any preternatural readings in his closing remarks. Surely, as a humanist, he was aware of the possible connections, and knew others would be as well.
Not all contemporary commentators shared Calvin’s anxiety at this point, though most were familiar with the resemblance between the Roman and Greek myths and the Genesis account of antediluvian society. For instance, the Jesuit Benedict Pererius (c. 1535-1610) thought pagan mythology, as seen in Homer, Hesiod, Appolodorus, and Ovid, was derived from the Hebrew Scriptures. A similar approach is foundin Calvin’s successor Theodore Beza, who often stated that similarities between the Pentateuch and ancient myths were due to the spread of a Great Tradition, passed from Seth and the Patriarchs to the Egyptians, and then indirectly to the Greeks and Romans. In any case, part of the pagan corruption, the argument goes, was a misunderstanding of who the “sons of God” and Nephilim were. The point remains that early modern writers differentiated “myths” from the “sons of God” episode in such a way that the Genesis account does not appropriate myth but provides the material for pagan corruptions of the originally true story.
Thus, the Reformed writers had three potential motives for their interpretations of Genesis 6:1-4. First,and most pronounced, was the metaphysical “absurdity” of human-angelic procreation. Second, is the uneasy resemblance between pagan mythology and the Genesis passage, if one accepts the preternatural reading. Third, and only implicit, is the testimony of tradition.
The possibility of objective interpretation and reevaluation
Van Gemeren states, concerning modern evangelical exegesis of Genesis 6:1-4, “It is sobering to recognize how difficult it is to read the passage without any preconceived ideas.” This is no less true today than it was in the sixteenth and seventeenth centuries. For, in addition to the concerns already mentioned, scholars in the current evangelical tradition have a further motivation for a non-preternatural reading--they do not want to appear silly in the midst of modern scholars. Newman, describes the paradoxical climate that now exists:
On the one hand, liberal theologians, who deny the miraculous, claim the account pictures a supernatural liaison between divine beings and humans. Conservative theologians, though believing implicitly in angels and demons, tend to deny the passage any such import. The liberal position is more understandable with the realization that they deny the historicity of the incident and see it as a borrowing from pagan mythology. The rationale behind the conservative view is more complex: though partially a reaction to liberalism, the view is older than liberal theology.
In other words, evangelical scholars who adhere to a literal interpretation of Scripture must affirm an event that sounds mythological, if they are to accept a preternatural reading of Genesis 6:1-4. Scholars unwilling to take this route might choose (consciously or not) any interpretation that will let them off the hook.
Given what has been said thus far about the apparent traditional and metaphysical assumptions in early modern Protestant commentaries, one might abandon any attempt at objective exegesis as fundamentally doomed. Indeed, many postmodern thinkers today, such as Delwin Brown, argue that it is impossible for an interpreter to work independently of philosophical assumptions, entrenched traditions, and community values. To be sure, the preceding discussion tends to confirm the idea that commentators of both magisterial Protestant traditions have in fact allowed embedded preconceptions to influence their interpretations. However, this does not mean that an interpreter cannot strive for objectivity and allow the natural reading of a text to subvert metaphysical assumptions. It is also possible that an exegete might hold the “philosophical assumption” that Sacred Scripture will inevitably challenge one’s previous philosophical assumptions! In other words, if one expects Scripture to confront traditions and preconceptions, the ideal of objectivity is not so far off. Moreover, even if one resigns to the idea that all interpretation is bound to traditional and metaphysical assumptions, one may at least question whether the assumptions employed in the interpretation of a particular passage are consistent with a given tradition’s theological and hermeneutical principles. It is at this point that the traditional Lutheran interpretation appears inconsistent and in need of reevaluation.
To illustrate the possible inconsistency for the Lutheran tradition, and the possibly predictable approach of the Reformed tradition, one can look to the Christological and Eucharistic debates of early modern Protestantism. In these contexts, Calvin and the Reformed appealed implicitly to an a priori principle, finitum non capax infiniti (the finite is incapable of or cannot contain the infinite), in order to show that one should not predicate of the finite, human nature of Christ that which pertains to His infinite, divine nature. When applied to the Eucharist, the Lutheran tenet that Christ’s body was present “in, with, and under”sacramental bread and wine appeared not only unbiblical but metaphysically absurd. This a prioriassumption of what is and is not possible for Christ’s human nature was relatively common for Calvin and his tradition. For instance, concerning Christ’s entrance into a locked room after the Resurrection, Calvin writes:
For just as the water, like a solid pavement, provided Christ with a path as he walked upon the lake [Matt. 14:25], so it is no wonder if the hardness of the stone yielded at his approach. Yet it is more probable that the stone was removed at his command, and immediately after he passed through, returned to its place. And to enter through closed doors means not just penetrating through solid matter but opening an entrance for himself by divine power, so that he suddenly stood among his disciples clearly, in a wonderful way, although the doors were locked.
Here, Calvin demonstrates his obsession with maintaining proper boundaries for spirit and flesh. He does not rule out extraordinary possibilities, but he is most comfortable with a Christ who does not walk through stone, but rather miraculously moves it out of the way. Thus, returning to Genesis 6:1-4, if Jesus will not break the rules of human nature by strolling through solid objects, preternatural entities are even less likely to break the rules of their ontology by inseminating women. To be fair, Calvin and his successors abhorred the idea of importing a synthetic a priori into their exegesis; they thought their assumptions about Christ’s human nature were based on the definition of human nature as such. The problem here is that, according to the Lutherans, human reason cannot determine the limits of Christ’s nature without appealing to synthetic a priori axioms.
While epistemologically suspect, according to Lutherans, the Calvinist approach here retains a level of uniformity. The Lutherans, on the other hand, defended their entire Christology and Eucharistic theology with the maxim that human reason cannot make pronouncements about the limitations of Christ’s human nature. Because a complete definition of human nature is elusive, Lutherans could argue,  it is inappropriate to set arbitrary limits on the glorified and hypostatically united human nature of Christ. To do so would be to employ synthetic a priori reason, precisely what Luther called the Devil’s whore.
Understanding the philosophical issues behind Lutheran Christology and Eucharistic theology, one can more clearly see how the evangelical tradition ought not assume, a priori, what angelic and demonic natures can and cannot do. Likewise, it suggests they ought to use Ockham’s razor to slash through complicated hypotheses that were constructed to explain away the metaphysically troubling implications of Genesis 6:1-4. Thus, an evangelical interpreter’s a priori assumption that angels cannot mate with women is incompatible with fundamental principles of evangelical thought. Granted, Luther and his successors spent little time discussing possibilities, and for good reason. However, notable in the evangelical tradition is the (perhaps psychological) demand that spirit and flesh stay within their boundaries. Language condemning the possibility of a breach of these limits typically demonstrates visceral disgust on the part of the commentator. Ironically, of course, the Fathers saw this as precisely the problem of the narrative: Yahweh was angry that preternatural beings and humans did in fact transgress their ontological boundaries.
It should be clear by this point that the issue is only partially exegetical; it is at bottom a matter of religious epistemology. One may thoughtfully opt for any of several interpretations of Genesis 6:1-4, using parallel texts to make sense of what most admit is a perplexing passage. Evangelical dogmaticians and commentators do frequently list one or two proof texts to demonstrate their angelological perspectives, though the connections to Genesis 6 are seldom fleshed out. What evangelicals cannot do with epistemological and theological integrity is make a priori assertions as to what is and is not possible regarding divine, human, or angelic nature. To do so would be arrogant and a breach of fundamental hermeneutical principles valued by Luther and his descendents, even if it is done in an attempt to support the interpretation of the Wittenberg Reformer himself. Thus, to state with J. T. Mueller, “the claim that angels once mingled with men by marriage ... is as foolish as it is anti-Scriptural” is only half worthy of an evangelical dogmatician if by “foolish” one means contrary to the a priori axioms of reason.
Several lessons from this investigation may prove useful to exegetes and theologians in traditions that stem from the Reformation era, and those who value revelation as the source of theological knowledge. First, it is important to note that, despite their commitment to sola Scriptura, the Reformers were often inclined to maintain traditional interpretations, especially if those interpretations posed no threat to evangelical theology. In such cases as Genesis 6:1-4, it is of course important to acknowledge exegetical tradition. Nevertheless, one must subsequently reappraise the strengths and weaknesses of traditional arguments. Second, evangelical exegetes should avoid the tendency to accept the Gnostic idea that matter is evil and spirit is good in principle, and thus imply that angels are superior to humans because they are not “soiled” by matter. This tendency was apparent in some of the Reformed writers, but may also have been latent in Luther, at a visceral level. Third, evangelical exegetes should be aware that the intrusion of synthetic a priori axioms can encroach upon exegesis, even when the exegete is notconsciously employing them. For example, Luther and Calvin likely assumed that the definition of “angel” entailed incorporeality, and had no intentions of imposing rational limits upon them. However, Protestant exegetes today should note this issue and consider whether the metaphysical assumptions they bring to the text are supported by the analogia Scripturae or by some post-Enlightenment worldview. Fourth, while most will find it laudable that Luther applied Genesis 6 to his contemporary context, and thus applied it meaningfully to his audience, evangelical exegetes ought also to emphasize the original context of a text, and thus avoid clouding any interpretation with anachronism. Fifth, exegetes uncomfortable with parallels between Near Eastern and Hebrew texts cannot seek a remedy for their intellectual discomfort by transforming apparently mythological elements in the Hebrew texts into palatable and mundane accounts. It makes no sense to obscure the literal meaning of Scripture in order to make it easier to maintain faith in the literal meaning of Scripture! Sixth and finally, in no case should a confessional tradition that takes revelation seriously shy away from a preternatural interpretation of Genesis 6:1-4 on philosophical grounds. Rather, evangelical religious epistemology requires an interpreter to submit to Scripture even when the text challenges one’s comfortable conceptions of the created order.
 I lump multiple theories together under the identification of the “sons of God” as human or non- preternatural. Some have argued that they were Sethites, others that they were great pagan kings. There are numerous variations on these themes. I am not concerned with the nuances here, since these are all attempts at a non-preternatural view. Further, I am not concerned whether the “sons of God” are seen asangels or lesser divinities within the Divine Council, since in both cases they are something other thanhuman, and Greek and Hebrew terms for “angel” usually indicate role rather than nature.
 The perspective of Luther and Calvin continued through the Enlightenment and is presented as the standard position in Abraham Arnoldus van Toll, Dissertatio Philologico-Theologica de Nephilim (Harderwijk, Netherlands, 1713), 6: “Ita Lutherus, qui recte Genes. vi. 4 Nephilim vertarat tyrranen . . .Calvinus ibid. idem vocabulum explicans.”
 See Umberto Cassuto, Biblical and Oriental Studies, trans. I. Abrahams (Jerusalem: Magnes Press, 1973), 18-19; Gerhard von Rad, Genesis, rev. ed., trans. John H. Marks (Phildadelphia: Westminster/John Knox Press, 1972), 115; Claus Westermann, Genesis (Neukirchen: Neukirchener Verlag, 1976), 1:493-503.
 I use the term “evangelical” throughout this paper rather broadly, referring to traditions that stemmed from the magisterial Reformations of the sixteenth century, especially the Lutheran and Reformed traditions. I am not referring specifically to the conservative American Christian subculture.
 Tatian, “Address to the Greeks,” in Alexander Roberts and James Donaldson, eds., Ante-Nicene Fathers (Hendrickson Publishers, Peabody Massachusetts, 1995), 2:70.
 Francis Turretin, Institutes of Elenctic Theology, ed. J. T. Dennison, trans. George Giger (Phillipsburg, N.J.: Presbyterian and Reformed, 1992), 1:541.
 Tertullian, “On the Flesh of Christ,” in Ante-Nicene Fathers, 3:523.
 David Keck, Angels and Angelology in the Middle Ages (Oxford: Oxford University Press, 1998), 31. Keck, Angels, 32.
 The basic distinction between analytic and synthetic propositions is stated by Anthony Flew, “Analytic and Synthetic,” A Dictionary of Philosophy: Revised Second Edition (New York: St. Martin’s Press, 1984),
12: “A statement is an analytic truth if it is true in virtue of the meanings of the words it contains; astatement is a synthetic truth if it is true in virtue of the way the world is.”
 Dyan Elliot, Fallen Bodies: Pollution, Sexuality, and Demonology in the Middle Ages (Philadelphia: University of Pennsylvania Press, 1999), 141.
 Willem A. Van Gemeren, “The Sons of God in Genesis 6: 1-4 (An Example of Evangelical Demythologization?),” Westminster Theological Journal 43 (1981): 320-1.
 Note, however, that there is good reason to believe, even from a modern perspective, that the connection between the “mingling of gods with mortals and the procreation of the demigods was originally connected to the flood narrative and functioned as its motivation.” Ronald S. Hendel, “Of Demigods andthe Deluge: Toward an Interpretation of Genesis 6:1-4,” Journal of Biblical Literature 106.1 (1987): 16.
 Luther, Luther’s Works, J. J. Pelikan, H. C. Oswald, and H. T. Lehmann eds. (St. Louis: Concordia Publishing House, 1999, c1960), 2:5 [WA xlii, 265].
 Luther, Luther’s Works, 2:32-3 [WA xlii, 285]
 Luther, Luther’s Works, 50:309 [WA, Br xi, 292]
 Luther, Luther’s Commentary on Genesis (Grand Rapids: Zondervan, 1958), 130.
 Luther, Commentary on Genesis, 132. Elsewhere, (Luthers Works, 2:34; WA xlii, 285-6) Lutherderides the idea that the Hebrew root “to fall,” contained in the term “Nephilim,” implies that people “fell”before them, due to the size of their bodies.
 Luther, Luther’s Works, 2:4 [WA xlii, 265]
 Luther, Commentary on Genesis, 128.
 Ibid., 129.
 Johan Gerhard, Commentarius super Genesin in quo textus declaratur, questiones dubiae solvuntur, observationes eruuntur, et loca in speciem pugnantia conciliantur (Jena, 1637), 174.
 Gerhard, Commentarius super Genesin, 175.
 Abraham Calov, Commentarius in Genesin (Wittenberg, 1671), 642. Luther, Luther’s Works, 2:11; [WA xlii, 270]
 Ibid., 2:12; [WA xlii, 270]
 Calvin, Commentaries on the First Book of Moses Called Genesis (Grand Rapids: Baker Book House, 1979), 237-8;
 Note that throughout my book, Faith, Reason, and Revelation in Theodore Beza 1519-1605 (Oxford: Oxford University Press, 2003), I reject the simplistic view that Reformed exegesis was forced to conformto the “central dogma” of predestination. I also reject the idea that Beza and his tradition affirmed the priority of reason over revelation. Nevertheless, Reformed Christology, Eucharistic theology, and angelology represent inconsistencies within their system, since those loci tend to rely implicitly on metaphysical presuppositions the Reformed believe are analytic, but are in fact synthetic a priori axioms.
 Calvin, Genesis, 239.
 Calvin, Genesis, 238; idem, Corpus Reformatorum, eds. G. Baum, E. Cunitz, E. Reuss (Brunswick,1882), 51:111.
 Turretin, Institutes, 1:539.
 It is worth noting that Christ did not find it undignified to assume a human nature. For this reason, Lutheran dogmaticians refuse to equate Christ’s state of humiliation with the Incarnation, and affirm that it rests in His refusal to assert His divine prerogative.
 Turretin, Institutes, 1:541-2.
 Incidentally, the argument for the corporeal view of angels usually assumed they were “fiery” bodies, not mammalian. Ibid., 1:539.
 James I of England, Daemonologie (Edinburgh, 1597) [facsimile: The English Experience: Its Record in Early Printed Books Published in Facsimile 94 (Amsterdam/New York: Da Capo Press, 1969)], 40.
 Ibid., 52.
 Ibid., 67.
 Ibid., 68.
 Ibid., 68.
 Calvin, Institutes of the Christian Religion, 164. On the demonology of the Calvinist writers see Ursula Lange, Untersuchungen zu Bodins Demonomanie (Frakfurt: Klostermann, 1970), 144.
 Hendel, “Of Demigods and the Deluge,” 18-19.
 Arnold Williams, The Common Expositor: an Account of the Commentaries on Genesis 1527-1633 (Chapel Hill: University of North Carolina Press, 1948), 141. Calvin, Genesis, 246.
 Cited in Williams, The Common Expositor, 214.
 Van Gemeren, “Sons of God,” 332.
 Robert Newman “The Ancient Exegesis of Genesis 6:2, 4,” Grace Theological Journal 5.1 (1984): 13- 14.
 Delwin Brown, Boundaries of Our Habitations: Tradition and Theological Construction (New York: State University of New York Press, 1994).
 John Calvin, Institutes of the Christian Religion, ed. J. T. McNeill, trans. F. L. Battles (Philadelphia: Westminster, 1960), 2:1400.
 This was brought to my attention by David Andersen, “A Critique of John Calvin’s PhilosophicalAxiom finitum non capax infiniti” (paper presented at Westminster Seminary, May 4, 1999). The conclusion of his paper provides a good summary: “Calvin insists, as we have seen, that anyone whodenies certain limitations of the human body—limitations, of course, that he determines—commits the error of asserting A is not A (or, contradicts the meaning of the term body). However, we now see that Calvin has no basis for making such a sweeping statement. One who holds that Jesus’ human natureparticipates in the divine omnipresence has not contradicted himself. Strictly speaking, there is no contradiction at all in saying that a human body may be empowered with divine gifts. In fact, in order to say that it is self-contradictory for Christ’s human nature to participate in his divine nature, one wouldhave to be able to give a complete definition of both natures e.g., the limitations of each of theirproperties.”
 See David Andersen, “Martin Luther’s Understanding of Reason and Its Relation to Faith: A Reexamination in Light of the Epistemological, Logical, and Christological Issues” (Ph.D. diss., Coventry University in collaboration with Wycliffe Hall, Oxford, 1999).
 J. T. Mueller, Christian Dogmatics: A Handbook of Doctrinal Theology for Pastors, Teachers, and Laymen (St. Louis: Concordia, 1955), 198. Note that he defends his assertion that it is “unscriptural”merely by citing Matt. 22:30, though this common proof text seems unable to bear the weight of a developed angelology.
 In other words, I am not qualified to provide further evidence on whether the “sons of God” wereangels or semi-divine beings. However, if such evidence were available and conclusive, evangelicals should be the first to accept it, since, in theory, they ought to have no overriding metaphysical axioms to preserve.