Reading Joshua with Augustine and Sommer: Two frameworks for interpreting theophany narratives

1. Introduction—a fragmented theophany?

This discussion finds its starting point in a ‘strange story’ from Joshua. Theophany texts are intrinsically strange, Josh. 5.13-6.5 perhaps doubly so. The ambiguities that are typical of all theophany narratives are pronounced in this narrative, to the point that there is a serious question whether Josh. 5 does in fact report a theophany. The narrative begins rather abruptly when Joshua, apparently by himself outside the camp, looks up and sees a man standing before him with drawn sword. Joshua asks, we might gloss, ‘who goes there, friend or foe?’ ¹ The ‘man’ answers rather cryptically ‘No. But I am the commander of Yhwh’s army. Now I have come’ (5.14a). Joshua then falls to his face and worships, asking ‘what does my lord (אדני‎) say to his servant’ (5.14b). The commander responds by echoing the words of Exod. 3.5, instructing Joshua to take off his sandals, ‘for the place where you are standing is holy’ (Josh. 5.15).

Since at least Martin Noth, it has been commonly thought that Josh. 5.13-15 is a fragment of an earlier, extended episode and that the original message from the commander to Joshua is now missing. ² Evidence for the fragmentary nature of this narrative is seen in apparent vestiges of its earlier form: the narrative begins when Joshua is ביריחו‎ which we would normally read as ‘in Jericho’; only the context leads us to read the phrase with the Vulgate, Syriac, and Arabic versions as ‘while in the vicinity of Jericho’. ³ Similarly, the reference to a ‘holy place’ suggests that the narrative was perhaps originally connected with Gilgal or a now forgotten cultic site near Jericho. ⁴ Finally, the narrative begins very much like the commissioning reports found in Exod. 3 and Judg. 6 but then apparently breaks off before Joshua, who at this point has already been leading Israel for five chapters, is actually commissioned. ⁵

Especially if Josh. 5.13-15 is read as a discrete narrative fragment, then what is to be made of the figure of the commander? Origen argues that Joshua must have known that the commander was in fact God himself, since he would not have worshipped (וישתחו‎) a mere angel. ⁶ The lxx simply deletes the offending clause. But the term is used elsewhere to describe obeisance to a human superior and Joshua refers to the commander simply as אדני‎ (‘lord’) rather than יהוה‎ (‘Yhwh’). Read as a fragmentary and perhaps non-theophanic encounter, 5.13-15 describes ‘a brief and somewhat cryptic’ encounter ⁷ with ominous overtones as the drawn sword of the commander appears menacing and his ambivalent answer is less than reassuring. ⁸

On the contrary, if 5.13-15 is read in continuity with 6.1-5, it fits the pattern of Exod. 3 and some of the ambiguity is resolved. While many commentators do not even ask if 6.2 might be read as a continuation of the dialogue of 5.15, Sarah Hall provides a compelling syntactic and literary argument for reading 6.1 as a parenthetical rather than disjunctive clause and 6.2 as a continuation of the dialogue. ⁹ She suggests that syntactically, 6.1 is parenthetical since the exclusive use of participles would be an odd beginning to a narrative unit. Moreover, the mt marks a division between 6.1 and 2 rather than between 5.15 and 6.1, where it would be expected on the ‘fragment’ hypothesis. Finally, Hall argues, the parallels between Josh. 5.13-15 and Exod. 3 lead the reader to expect some instructions to be given to Joshua—precisely what is found in Josh. 6.2-5.

Accepting Hall’s argument, what is needed is a model that makes sense of the simultaneous difference and identity between the commander of the Yhwh’s army and Yhwh himself who speaks in 6.2. In the remainder of this article, I seek to show how Benjamin Sommer and Augustine suggest two models that account for this identity in difference.

2. Sommer’s ‘fluidity’ model

In Sommer’s fascinating book, The Bodies of God and the World of Ancient Israel, he proposes a ‘fluidity’ model of divinity which provides a possible framework for interpreting theophany texts in (and outside) the Hebrew Bible. ¹⁰ Initially, note how Sommer defines some of the key terms in his argument: a ‘self’ is ‘a discrete conscious entity that is conscious of its discrete nature’ (12) while a ‘body’ is ‘something located in a particular place at a particular time, whatever its shape or substance’ (2). ¹¹ Sommer takes for granted that ‘the God of the Hebrew Bible has a body’ (1) which means, given his definitions, that the biblical text depicts a ‘divine body … located at a particular place at a particular time’ (2). Although the fact that God is depicted as having a body in the Hebrew Bible has often been overlooked or denied, Sommer believes the evidence is overwhelming and so ‘asserting the carnal nature of the biblical God should not occasion surprise’ (1).

Thus, Sommer’s thesis is ‘not that God has a body—that is the standard notion of ancient Israelite theology—but rather that God has many bodies located in sundry places in the world that God created’ (1). Simply put, in at least some strands of the Hebrew Bible, God is depicted as simultaneously physically present in multiple earthly and heavenly locations.

Sommer first develops a model of divine fluidity from ancient Near Eastern texts, demonstrating that it was a significant factor in the world of ancient Israel. This involves several components since several kinds of fluidity of selfhood can be seen. One variety of fluidity is fragmentation. In this mode of fluidity, a deity may be ‘fragmented’ into several localized selves who share the same name and somehow both are and are not the same god (13). For example, a multiplicity of Baals can be named for different locations even within the same document—Baal Shamayim, Baal of Peor, Baal of Ṣidon, Baal of Ugarit, and so forth. ¹² Baal appears to have fragmented into a numerous local ‘baal-gods’ who were addressed and worshipped in various discrete locations (25). A second variety of fluidity of selfhood is seen in the overlap of identity among gods in ancient Near Eastern literature. In Canaanite literature, the goddess Tannit can be referred to as the ‘face/presence’ of Baal (KAI 14.18) and Astarte as the ‘name’ of Baal (KTU 1.16.vi.54-5). These goddesses who elsewhere appear as discrete selves can also be spoken of as aspects of another god.

Third, not only can a fluidity of selfhood be seen in ancient Near Eastern deities but also a multiplicity of divine embodiment—‘divine presence could inhabit multiple physical objects on earth without diminishing the heavenly body of the god’ (28). Thus, for example, in the ‘mouth opening’ (pīt pî) and ‘mouth washing’ (mīs pî) ritual activations of divine images (ṣalmu), the cult statue is taken up as an embodiment of the deity. ¹³ The god’s embodiment in the ṣalmu, however, does not mean that the god had left heaven or could not simultaneously inhabit other ṣalmu. Multiple images may form bodies of a god but they do not, either individually or taken together, exhaust that god’s being. Thus, in the ancient Near East, a basic contrast between gods and humans is that a god can exist in multiple bodies and this multiplicity of embodiment presupposes a fluidity of selfhood. ¹⁴

Sommer argues that the model of divine fluidity was, at least partially, shared by ancient Israel. Since the Hebrew Bible is monotheistic in its final form, there can be no overlap of identity among gods (54). Nevertheless, fluidity of divine selfhood is seen in ancient Israel and the Hebrew Bible in a number of ways. Something like fragmentation is seen in inscriptions that refer to Yhwh of Samaria or Yhwh of Teman or in 2 Sam. 15.7, where Absalom reports that he made a vow to worship Yhwh of Hebron (cf. Ps. 99.2). Similarly, argues Sommer, the multiplicity of divine embodiment can be seen in the view that God is present in wood (especially the אשׁרה‎) and stones (such as the מצבה‎ and ביתאל‎ stones). ¹⁵

Sommer considers passages featuring ‘a messenger’ (מלאך‎) to provide some of the strongest evidence of fluidity in the Hebrew Bible and it is this argument that is of particular relevance to the interpretation of theophany narratives. Sommer’s argument is that the fluidity thesis provides a basis for understanding the ambiguities of many of the narratives that feature ‘messengers’ of God. In many cases, Sommer argues, the מלאך‎ is ‘a small-scale manifestation of God’s own presence, and the distinction between the messenger and God is murky’ (40). Thus, Sommer wonders if ‘avatar’ may in fact be a better translation of מלאך‎ than ‘angel’ since the

mal’akh in these cases is not a separate being from Yhwh whom Yhwh sent on a mission; rather, it is part of the deity that can act on its own. Alternatively, it is possible that Yhwh temporarily overlaps with some heavenly beings who do God’s bidding. (40)

Sommer cites as an example Judg 6, where there is a sustained interchange in the identification of the speaker as Yhwh and ‘messenger’ (cf. Gen. 18; 32; Exod. 3.2; 33; Hos. 12.4-6). This speaker speaks sometimes in the first person as God (6:14, 16) and sometimes in the third person (6.12). Although it might be argued that Yhwh is located in heaven throughout the narrative and merely speaks through a messenger, ‘it is specifically Yhwh who turns His face toward Gideon’ in 6.14 (43). But then it is Yhwh’s ‘messenger’ which Gideon is said to have seen (6.22), not Yhwh himself:

The text seems self-contradictory only if one insists that an angel is a being separate from Yhwh. On the other hand, if one can understand an angel as a small-scale manifestation of God or even as a being with whom Yhwh’s self overlaps, the text coheres perfectly well. (43)

Sommer’s fluidity model can be succinctly summarized: a deity differs from humans insofar as the deity’s body and self are completely unbounded (124). Thus, the deity can produce small-scale manifestations that can to some degree act independently but without becoming a distinct deity. Thus, Yhwh manifests himself in a figure who

clearly is and is not identical with Yhwh; more precisely, He is Yhwh, but not all of Yhwh or the only manifestation of Yhwh; rather, He is an avatar, a ‘descent’ of the heavenly God who does not encompass all of that God’s substance. (41)

The strength of Sommer’s model is that passages that appear baffling or self-contradictory can be read coherently when we stop trying to ‘pin down something fluid’ (25) and recognize these passages ‘as examples of the fluidity of divine selfhood so common in the ancient Near East’ (43). In the Hebrew Bible too, God ‘could be present in a body (or perhaps several bodies) resembling that of a human, but this was not [His] only body’ (43). Sommer’s assertion that God has a body, however, appears to raise several basic and unresolved issues. First, it is not apparent that God having a body which is his makes it easier for him to inhabit additional bodies. If anything, it seems to me, his having a body would make it more difficult for him to take on additional bodies. Second, I suppose what we ought to say, given Sommer’s definitions, that God’s ‘self’ is not precisely identical with his ‘body’ and so can inhabit multiple bodies. But is this ‘self’ which is not wholly identical to God’s bodies visible or invisible, material or spiritual? My point is that even if one wants to affirm with Sommer the prominence of the theme of God’s ‘embodiment’ in the Hebrew Bible, several metaphysical questions about God’s nature arise naturally from a consideration of the development of the theme. Although the actual conclusions that early Jewish and Christian commentators arrived at regarding God’s nature might be rejected, at least some of the questions that they have put to the text are not merely foreign impositions based on Hellenistic concerns but rather arise naturally from a close reading of the Hebrew Bible. I now want to suggest that Augustine’s account of theophany narratives, which does ask some metaphysical questions about seeing God, might helpfully supplement and correct Sommer’s account at precisely this point.

3. Augustine’s ‘creature-control’ model

Two observations heuristically situate Augustine’s interpretation of theophany narratives. First, in The Trinity and elsewhere, Augustine argues that the goal of the Christian life is to ‘see God’

this contemplation is promised us as the end of all activities and the eternal perfection of all joys … what we shall contemplate as we live for ever is what he told his servant Moses: I am who I am. (trin. 1.9.17) ¹⁶

The goal of seeing or contemplating God structures the whole of Augustine’s The Trinity. ¹⁷

Second, we must note how the theophany narratives were interpreted in the Christian tradition prior to Augustine. Justin Martyr’s reading of Gen. 18 is typical. God is ‘sent’ with two angels

by another, who forever abides in the super-celestial regions, who has never been seen by any man, and with whom no man has ever conversed, and whom we call Creator of all and Father … There is mentioned in [Gen. 18] another God and Lord under the Creator … who is called an Angel. ¹⁸

Thus, the Father remains in heaven and the Son is sent to be manifest in the theophanies, as is proper given his role as mediator of the Father. This reading supports Justin’s theological claims about the Son’s preexisting identity and, Justin thinks, provides a convincing argument in his Jewish-Christian dialogue. This trend in interpreting theophanies dominated in the early church up to the time of Augustine. The problem that emerged, however, was that Arians, such as Palladius, capitalized on this interpretation to drive a wedge between the Father and the Son: the Father’s distinguishing feature is his invisibility and the Son’s his visibility. ¹⁹ In the theophanies, the Son is seen according to his nature while the Father remains invisible, according to his nature. Their natures fundamentally differ.

Augustine rejects this line of interpretation based on his trinitarian theology but in doing so, he also shifts the traditional understanding of how Christ unites the two testaments of the Christian Bible. ²⁰ For Augustine, the unified witness of all ‘catholic commentators … on the divine books of both testaments … has been to teach that according to the scriptures Father and Son and Holy Spirit in the inseparable equality of one substance present a divine unity’ (trin. 1.4.7). In this equality, Father, Son, and Spirit share one nature. And by nature, God—Father, Son, and Spirit—is invisible; he does not have a body (ep. 147 20.48; trin. 2.8.14). ²¹ This applies to the Son as well as the Father; the Son is not the ‘visible’ God. ²²

Augustine seems, then, to have worked himself into a corner—the goal of the life of faith is to ultimately see God. But God is invisible, so how is this possible? Matthew 5.8—‘Blessed are the pure in heart for they shall see God’—is axiomatic in Augustine’s discussion. Here is assurance that we will see God, not now but at the end of all things. How then is the invisible God seen? Not by his nature but by his will: ‘whoever saw God saw God because he appeared to whom he willed, when he willed, and in that form his will chose, while his nature remained hidden’ (ep. 147 7.19). Note, at least in Augustine, how the claim that God is ‘invisible’ by nature is construed: in the first instance, it is about who has the initiative in seeing God. If God had a body and was visible by nature, then he would be visible in a location. Seeing God would depend on human initiative in finding the location where God is. If God is invisible by nature, then he can only be seen as he wills to be seen. God is not seen in a location, says Augustine following Ambrose, but rather through the purification of our hearts and minds.

Augustine recognizes that the incarnation plays a central role in this process such that it is fundamentally distinct from prior theophanies. The goal of the incarnation is to lead those who live by faith, that is, who believe what is not seen, to the contemplation of God (trin. 1.9.17). In the incarnation, flesh is eternally joined to the Son (trin. 2.6.11). ²³ The Son is not visible or bodily by nature, but rather takes human form to himself in the incarnation in order that God might be ‘seen’ by humans. Thus, on Augustine’s view, the incarnation is fundamentally distinct from previous theophanies although even in the incarnation, the Son does not reveal the divine in ‘any direct, available-to-the-senses way’. ²⁴ The vision of God is eschatological, the incarnation is a step in inaugurating that eschaton.

How then was God ‘seen’ in the Hebrew Bible? The theophanies were not ‘christophanies’—appearances of the pre-incarnate Christ. After all, the Son is not intrinsically visible (trin. 2.8.14) nor had he yet taken on ‘the form of a servant’ in the incarnation (trin. 2. 11.20). No person of the Trinity has shown himself to human eyes except through created bodily substance (trin. 2.9.16). This appearance through created bodily substances can happen in two ways: either created things which already existed, such as an angel, were given a symbolic significance in order to represent God (trin. 2.6.11) or created things, such as the fire of Exod. 3, come into being in order to signify God (trin. 2.6.11). ²⁵ Thus, the theophanies were ‘produced through changeable creation subject to the changeless God’ (trin. 2.17.32). Augustine pulls these various strands together:

By nature … God is invisible, not only the Father but also the Trinity itself, one God, and because he is not only invisible but also immutable, he appears as he wills in what form he wills so that his invisible and immutable nature remains wholly within him. (ep. 147 8.20)

Later, Augustine will state that ‘God’s will is the first and highest cause of all physical species and motion. Nothing happens visibly … which does not issue as either a command or as a permission from the inmost invisible and intelligible court of’ God (trin. 3.4.9). By affirming that God appears by his will, Augustine correlates theophanies with creation and providence more generally: all these follow from God’s will and signify God in differing ways. God is always being revealed through the world in its regularity, but since we are not sensitive to this, God uses nature irregularly in the theophanies to signify himself in a peculiar way (trin. 3.5.11). Thus, Augustine argues, the theophany narratives often describe the subject who appears in shifting terms—for example, the ‘angel’, the ‘lord’, ‘God’—because of God’s ability to inhabit a creature not because of Christ’s functional identity as the angel of the Lord.

Finally, Augustine asks, who was seen in the theophanies? Since no person of the Trinity is intrinsically visibly, ‘we should not be rash in saying which person of the three appeared in any bodily form or likeness to this or that patriarch or prophet, unless the whole context of the reading provides us with probable indication’ (trin. 2.18.35). Some have understood Augustine as claiming that we cannot determine if God in his unity or a specific person of the Trinity is being revealed in any given theophany. ²⁶ I want to suggest that the ‘whole context’ be read as referring to what we today might call the ‘literary-canonical context of the two testament Christian Bible’. After all, Augustine does identify the cloud in Exodus with the Spirit and the ‘back’ of God that is seen by Moses in Exod. 33 as prefiguring the incarnate Son (e.g. trin. 2.17.28-29). ²⁷ But, I think, ‘prefigure’ is precisely the right term—the identification is figural, based on intertextual associations created by reading Hebrew theophany narratives as Christian scripture. ²⁸ Augustine’s interpretation, then, at least opens up the possibility of reading Hebrew scripture in a way that is consistent with his trinitarian theology but does not Christianize the text. ²⁹

We can call Augustine’s model for reading theophany narratives the ‘creature-control model’. Whenever God is said to appear in the Hebrew Bible, this vision is mediated by created reality, especially the angels (trin. 3.8.27). We have been exiled from God, argues Augustine, and so God has sent sights suitable to our wandering state (trin. 4.1.2). These sights are not God in himself but signify God by pointing forward to God’s taking on flesh in the incarnation. For Augustine, the theophanies illustrate in clear form how believers come to know God ‘in the tension between revelation and hiddenness that is present in mystery’. ³⁰ Although Augustine severed the traditional tie between Christology and theophanies, through the language of ‘signification’, he reorients and reconnects the theophanies to the incarnation by interpreting the theophanies (and the incarnation) on a forward-looking trajectory. ³¹ Thus, concludes Augustine,

Everything that has taken place in time … has been designed to elicit the faith we must be purified by in order to contemplate the truth, [and] has either been testimony to this mission or has been the actual mission of the Son of God. (trin. 4.19.25)

4. Comparisons

Here I highlight two significant differences between Sommer and Augustine. First, we must note the differing contexts that the two read within. Sommer, in the portion of his book that I have engaged, primarily sets forth a descriptive argument that reads specific sources within the Hebrew Bible independently of each other, within the larger ancient Near Eastern historical context. Augustine, on the contrary, approaches the Hebrew Bible primarily within a literary-canonical context. This means that Augustine seeks to read, for example, Exod. 33 as coherent in its final form while for Sommer the chapter is a prime witness to the ongoing debate in ancient Israel between two opposing views regarding divine embodiment. Furthermore, this means that for Augustine, the Hebrew scriptures are approached only insofar as they have been recontextualized by the New Testament and thus he seeks to find a way of reading that does justice to both the appearances of God in the Hebrew scriptures and the assertions in the New Testament that God is ‘invisible’ (1 Tim. 1.17) and that ‘no one has ever seen God … but the Son has made him known’ (John 1.18, 1 John 4.12).

Second, following from this, an obvious difference is that Sommer’s starting point is that God has a body in the Hebrew Bible while Augustine’s starting point is the exact opposite. What we make of this difference largely depends on our goals in reading the Hebrew Bible and the context in which we do so although, as Maimonides and Kaufmann both illustrate, the difference is not simply between Jews and Christians. ³² But the views of Sommer and Augustine are not entirely irreconcilable. It seems to me that Augustine could be read as providing one strategy, an explicitly Christian one, for dealing with the diversity in the Hebrew Bible that Sommer has revealed and the questions that arise from his account. Or, to put it the other way around, Sommer’s important historical argument is supplemented by the sorts of concerns which Augustine addresses.

5. Conclusion: rereading Joshua

Finally, I want to briefly return to the Joshua narrative. How do the models proposed by Sommer and Augustine help us read Joshua? First, both models, taken either independently or together, account for the simultaneous distinction between the commander and the Lord whose army he commands and the apparent identity between two such that it makes sense that 6.2 begins ויאמר יהוה‎ (‘and Yhwh said’). The commander is a distinct creature that Yhwh ‘overlaps with’ or assigns a special signification of his presence to in this encounter with Joshua. Reading 5.13-6.5 together thus brings the narrative more closely in line with the account of Exod. 3, further depicting Joshua as a ‘second Moses’. ³³

Reading Joshua 5.13-6.5 as a whole in light of Sommer’s argument draws particular attention to the theme of divine presence. The commander, as an embodiment of God, declares ‘I am now here’. Then, in the instructions given in 6.2-5, the ark (another way in which God makes himself present in the midst of Israel) ³⁴ is central to the plans for the liturgical assault on Jericho. By this connection, the presence of God that is made known to Joshua in the figure of the commander is ‘externalized’ through the ark’s signification of God’s presence in the midst of the marching army. ³⁵ If the commander’s ambivalent response in 5.14 is a refusal to be coopted into Israel’s agenda, the instructions of 6.2-5 make the same point in symbolic terms: victory is achieved through observing Yhwh’s instructions.

Reading Joshua 5.13-6.5 with Augustine highlights the progressive discernment of God in the created means that he has used to signify his presence. ³⁶ The term ‘behold’ (והנה‎) in 5.13 indicates that the narration has adopted Joshua’s point of view while the narrative genre entails casting the theophany within a temporal framework. ³⁷ Along with Joshua, our vision is limited. Initially, we simply see a ‘man’. Then after the commander’s first piece of dialogue, Joshua addresses him as ‘lord’ but not as ‘Yhwh’ and bows before him. Perhaps only with the echoes of Exod. 3.5 in 5.15 does Joshua recognize the full significance of the encounter. But at no point is it simply obvious to Joshua that this is in fact an appearance of Yhwh (nowhere in 5.13-6.5 does Joshua use this name). ³⁸

Finally, the narrative of Josh. 6 takes on a new light in this context. Surely, the commands given in 6.2-5 do not represent particularly good military strategy, in any obvious sense. The act of obedience thus exhibits a level of trust that it is in fact Yhwh who has revealed himself in the commander and given these unorthodox military commands. The significance of the narrative of Joshua 6.6-27 for the characterization of Joshua is thus further illuminated. ³⁹ Joshua is depicted as a fit commander (or sub-commander) of Israel not in the first instance through his military prowess but rather through his discernment of Yhwh in the ambiguous figure of the commander and his faithfulness to the command of Yhwh in the midst of difficult circumstances. ⁴⁰