Did Amram have a choice? Divine determinism and human choice in 4QVisions of Amram

An introduction to VA

The beginning of the composition, including what must be considered the title, is attested in 4Q543 1, 1: ]פרשגן כתב מלי חז̇ו̇ת ע̊.מרם בר̊‎. In 4Q545 1a i, 1, the words before Amram’s name are more or less lost. The genealogical information, on the other hand, is partially legible: פרש̇ג̊[ן כתב מלי חזו]ת̊ עמרם בר̊ קה̊ת̊ ב̇ר לוי‎. Thus, what we have here is a “copy of the writing” called “the words of the vision (or visions) of Amram, son of Qahat, son of Levi.” ¹⁴ Amram is mentioned only briefly in the canonical Old Testament, in Exod 6:18, 20 (twice); Num 3:19; 26:58, 59 (twice); 1 Chr 5:28, 29; 6:3; 23:12, 13; 24:20 (in Exod 2:1 he is introduced without his name being mentioned). In VA, he plays the leading role.

The very first word of the composition, פרשגן‎ (Masoretic פַּרְשֶׁ֫גֶן‎), is an Iranian loanword also found in the Aramaic part of the Book of Ezra (4:11, 23; 5:6). In Hebrew (in the form פַּתְשֶׁ֫גֶן‎), the word occurs in the Book of Esther (3:14; 4:8; 8:13; in the Hebrew part of Ezra, the r-form occurs in 7:11). ¹⁵ Jurgens suggests that פרשגן‎ signifies the “permanent authoritative value” of a writing so designated and, in the case of VA, the use of the word shows that the content was presented as “inherited from the original, sacred words of Amram, providing the reader exclusive access to secret wisdom derived from the exceptional experiences and vision of this particular ancient.” ¹⁶ Similarly, Perrin states that פרשגן‎ as used here and elsewhere “frames the work or section thereof as a ‘copy’ of an authoritative record, be it a royal edict or alleged patriarchal discourse.” ¹⁷

The following words further explicate the contents of the work. It includes “everything that he (Amram) told (אחוי‎) his sons and that he commanded (פקד‎) them on the day of his death” (4Q543 1, 1–2; 545 1a i, 1–2). The verbs clearly indicate a didactic and instructional purpose, ¹⁸ and they emphasize the authoritative claims that seem to be inherent in the first part of the incipit. The testamentary setting (“on the day of his death”) and the final part of the incipit, which indicates the precise year of Amram’s death in relation to the exile of Israel in Egypt, also point in the same direction. ¹⁹

Even though the manuscripts are fragmentary, the general structure of the composition seems to be reasonably clear. Before his death, Amram arranges for his daughter Miriam to be married to his younger brother Uzziel. After a brief description of the wedding feast (4Q543 1; 4Q545 1a i; 4Q546 1), we hear that Amram summons his sons (4Q545 1a i, 8–9). ²⁰ The introductory part presents Amram in the third person but in the rest of the text, Amram seems to be the speaker. The first short passage after the introduction appears to refer to the future deeds of Moses and/or Aaron (4Q543 2; 4Q545 1a i). ²¹ Then Amram relates the events of a journey he undertook earlier in his life to Canaan, as participant in an expedition to bury the bones of the forefathers (4Q543 3–4; 4Q544 1; 4Q545 1a–b ii; 4Q546 2; 4Q547 1–2). ²² War and border closure forced Amram to stay in Canaan for 41 years; while he was there, he had a vision that is detailed in the final part of the work, including predictions about the future of his children and the events of the exodus (4Q543 5–9; 4Q544 1; 4Q547 1–2). ²³

As noted above, the title of the composition includes the word “vision” (or “visions”), which seems to point to the vision account as the central concern of the work. Furthermore, the centrality of the vision account is emphasized in the passage preserved in 4Q547 9, 8 according to which Amram awoke from his dream (מן שנת עיני‎, “from the sleep of my eyes”) and immediately wrote down the vision (וחזוא כתב̊[ת‎). In the next line of the same fragment, we are informed that what Amram had been told in the vision came to pass (]והוא לי כדי אמר̇‎, lit. “and it was/became for me according to what he had said”). As Perrin remarks, “authority and authenticity derive principally from the claim that the text is a record” of this vision. ²⁴ Accordingly, much scholarship has centered on the contents and meaning of the vision.

In the vision, Amram sees two beings, usually described as angels or “watchers” (עירין‎) even though the extant text never uses those terms. ²⁵ The preserved fragment (4Q544 1, 10) merely says that Amram saw תרין דאנין עלי‎, “two (who were) arguing over me.” Further on in the passage, the two beings describe themselves as the rulers of all humanity (4Q544 1, 12 combined with 4Q543 5–9, 2–3 and 4Q547 1–2, 11), and the appearance of each is described. One is sinister and dark, the other smiling (4Q543 5–9, 4–8; 4Q544 1, 13–15; 4Q547 1–2, 13). According to Milik, Puech, and others, the first being is compared to a serpent (reading כפ]תן] in 4Q544 1, 13), but this interpretation has been contested. Perrin, for instance, argues for reading this and the preceding word as ד̇ח̊[י]ל[ ואימ]תן‎ (“dre[ad]ful[ and terrify]ing”). ²⁶ Milik interpreted both parts of the description as concerning the same being with the word אחרנה‎ meaning “again.” ²⁷ However, Puech notes that the word means “the other,” marking a shift from the description of the first to the second being, who is “smiling” (problematically spelled חעכון‎ in 4Q543 5–9, 7 and העכן‎ in 4Q544 1, 14). ²⁸ Amram seems to continue the conversation with the latter being, who describes himself as having authority over all that is light, while the other being is described as having authority over everything that is dark (4Q544 2). The sinister being is seemingly called מלכי רשע‎ in 4Q544 2, 13. This—along with the mention of “three names” in 4Q544 3, 2—has given rise to the view that each being had three names, reflecting their opposite domains: מלכי רשע‎ (Melchireša‛, “King of wickedness” or “My king is wickedness”), בליעל‎ (“Belial”), and שר חשוכא‎ (“Prince of darkness”) for the dark one, and מלכי צדק‎ (Melchizedek, “King of justice” or “My king is justice”), מיכאל‎ (“Michael”), and שר נהורא‎ (“Prince of light”) for the being associated with light. However, only one of the six names is actually there. ²⁹

The preserved text, the reconstruction, and the immediate context

The passage in which the two beings—according to the prevalent interpretation—ask Amram to choose between them is part of their initial interaction with Amram. The question that they pose survives in its most complete form in 4Q544 1, 12, introduced by ואמרו לי‎ (“and they said to me”): ]במן מננא̇ א̇נ̊ת̇‎. ³⁰ These words must be interpreted as the preposition ב‎, the interrogative pronoun (“who”), the preposition מן‎ with a first plural suffix (“of/from/among us”), and a nearly illegible second singular masculine personal pronoun. So far, nothing indicates any kind of free choice. Milik, however, not only reconstructed the ending of the second person pronoun but also a following word: ה בחר]‎. ³¹ The reason for this particular reconstruction (with the verb “choose”) is not clear from Milik’s article, although Perrin suggests that Milik took the verb from his reading of 4Q547 1–2, 12, which seems to preserve a tiny part of the same passage. ³² Kobelski added the infinitive (שלט לאשתלטה‎ ithpeel) to Milik’s reconstruction, giving the translation “By which of us do yo[u choose to be ruled?”—even though he added that the idea of “choosing” is unexpected in the setting as he sees it. ³³ Fragment 2 from 4Q547 indeed seems to belong to the same passage (there are several overlaps between this fragment and 4Q544 1), but as noted by Puech, the verb attested in the fragment must be a form of the root בעי‎ (“seek,” “ask,” “want”), not בחר‎. ³⁴ Puech reads תה̇]אנ‎ ב̇ע̇ה̇‎, arguing that ה is to be preferred as the final letter of the verb rather than א‎. ³⁵ This reading, with Kobelski’s infinitive, is the basis for the usual interpretation that Amram is allowed to seek to be ruled by either of the two beings and for the widespread view that VA stands for human choice or “free will.”

In his critical assessment of the evidence, Perrin notes the uncertain character of the reconstruction, and—concerning the fragment from 4Q547—he states that even “a quick glance at PAM 43.567 shows that the smudged block of text at this point in fragment 2 is nearly illegible, rendering both proposals [i.e. בחר‎ and בעה]‎ speculative.” ³⁶ However, the image now available is more helpful. ³⁷ It is possible to see the end of the personal pronoun and even though the first letter of the verb is not very clear, ע‎ and probably א‎ (rather than ה‎?) are discernible. Hence, while Perrin’s cautious remarks remain important, some kind of “seeking,” “asking,” or “wanting” on the part of Amram does seem to be implied. For this reason, we cannot follow Beyer’s minimalistic reading and translation (במן מננה̊‎ אנ̊ת̊ [ה‎, “Zu wem von uns gehörst du?”). ³⁸ Furthermore, for the expression of belonging to someone, we would probably expect ל‎ rather than ב‎.

The infinitive לאשתלטה‎ proposed by Kobelski, on the other hand, is pure conjecture—and from a syntactic point of view, it may be problematic. In the reconstructed clause במן ממנא אנתא בעא לאשתלטה‎, the preposition ב must indicate the agent of the passive infinitive (“which of us do you seek to be ruled by”). The problem is that ב‎ seems to be rare in this function, and passive constructions with clear expression of agent are in fact generally uncommon. Carmignac notes a few occurrences in Qumran Aramaic, with ב‎ and other prepositions, but they are not convincing examples of the expression of real agents in passive constructions. ³⁹ Cook explicitly states that a word introduced by ב‎ “is never the agent of the action.” ⁴⁰ Hence, it may be worthwhile to consider other possible reconstructions, keeping in mind that we are in the realm of speculation. With the verb בעא‎ we would not expect a direct complement introduced by the preposition ב, ⁴¹ but the construction could possibly be understood as “through/by which of us do you seek X” with a lost noun indicating the direct object (the possibilities are numerous since nothing has been preserved; for instance “shelter” or “protection”?). In such a connection, however, the expected preposition would probably be מן‎ rather than ב‎.

Alternatively, we may accept Kobelski’s reconstruction with an infinitive following בעא‎, which seems plausible in light of other similar occurrences (as in 1QapGen XX, 9). However, instead of לאשתלטה‎, which would require ב‎ to be the marker of the agent, an alternative may be found in Cook’s list of verbs that take complements introduced by ב. ⁴² Verbs that could conceivably fit the context would be the following: שמע‎ (“listen to, heed, obey”; not often with ב‎, but see 11QtgJob XXXVIII, 2); ידע‎ (“know about”; see 4Q196 2, 1; 4Q204 4, 4; 4Q204 5 ii, 26); סכל‎ (hithpaal: “consider, look at”; see 11QtgJob XXIX, 5); בין‎ (hithpolel: “pay attention to, observe”; see 1QapGen XIII, 14); רחץ‎ (hithpeel: “trust”; see 11QtgJob XXXII, 10); תקף‎ (aphel: “hold on to”; see 4Q542 1 i, 8; 1QapGen VI, 6); אחד‎ (“hold on to”; see 4Q542 1 i, 7–9; 1QapGen VI, 6); מלל‎ (“speak with”; we expect the preposition עם‎ rather than ב, but a possible counterexample may be found in 4Q197 4 ii, 3, in Cook’s translation). ⁴³

Some of these verbs would function on the level of the conversation (which being does Amram want to have a closer look at or know about?), while others would seem to function in a way similar to the reconstruction with לאשתלטה‎ (which being does he want to obey or hold on to?). The context may support the latter possibility. As mentioned above, the vision begins with Amram’s statement that the two beings were דאנין עלי‎ (4Q544 1, 10). The verb דין‎ (or דון‎) can have the meaning “judge,” but with the preposition על‎, the meaning seems to be “debate about” or “argue over.” ⁴⁴ In the next line, the beings begin a great תגר‎, “strife” or “dispute,” over Amram. ⁴⁵ The context, then, would appear to indicate that at the time of the vision nothing has been settled with regard to Amram’s future. This, of course (along with the occurrence of other forms of the root שלט‎ in the preceding passage), seems to be the reason for the widespread acceptance of the reconstruction with לאשתלטה‎ and the idea that Amram is allowed to choose. ⁴⁶

However, the setting with the dispute between the beings has been interpreted differently. According to some scholars, the vision refers to the time of Amram’s death, and if that is the case, the beings are not arguing in order to convince Amram to follow one or the other but in order to decide which side he belongs to, based on what he has done throughout his life. Berger, for instance, compared the passage with other traditions concerning the struggle of angels over the soul of a deceased person. ⁴⁷ Kobelski interprets the vision in a similar manner, “In this vision, ‛Amram saw the two spirits who rule over humanity engaged in a struggle to take possession of him at his death.” ⁴⁸ He states that the use of the verb דין‎ “suggests the background of a legal procedure in which a judgment will be made on the merits of ‛Amram’s life and a claim made on him by one of the spirits.” ⁴⁹ An argument against this approach would be that the structure of VA as a whole—as far as the fragmentary state of the manuscripts allows us to see—seems to indicate that Amram saw the two beings and their struggle over him while he was in Canaan and not at the time of his death. ⁵⁰ However, as a counter-argument, it could be claimed that the vision refers to something that will happen in “real life” when Amram is about to die. In any case, if we were to accept this reading of the setting, it would make the notion of active, conscious choice on the part of Amram unlikely, but it could still be argued that Amram’s position relative to the realms of light and darkness has been decided by himself, through the way he has lived his life.

Such speculation demonstrates that the matter remains open. As was correctly stressed by Perrin, there is no certain evidence that the text stands for a type of dualism that emphasizes free choice on the part of human beings. ⁵¹ We cannot exclude the possibility that Amram in some way was asked about his preference with regard to the two beings and their respective realms but other interpretations are possible. Furthermore, even if Amram is asked to choose to be ruled by one of the beings, to seek something from one of them, or to seek to obey or hold onto one of them, does that necessarily mean that his answer actually decides his future? If (in the “judgment at the time of death”-reading) the dispute is settled by Amram’s actions throughout his life, does that mean that Amram had full control over those actions? In the present passage, nothing suggests the type of divine determinism known from some other Qumran texts (most notably 1QS with the so-called Treatise on the Two Spirits), but it will be of interest to examine the other surviving parts of VA in order to ascertain how the question of choice and determinism is treated elsewhere.

Human choice and determinism in VA, in 4Q548–549, and in 4Q542

According to Popović, VA “does not contain deterministic thought, as far as can be gauged from the extant text.” ⁵² To be sure, the material from 4Q543–547 that has survived does not discuss the question of determinism explicitly—but some of the fragments do contain passages that may suggest a deterministic point of view, that is, the notion that events are fundamentally determined by God. In contrast, apart from the contested passage discussed above, nothing in the fragments seems to support the idea of human choice as the decisive factor in the relation between man and the divine realm.

VA contains much prediction. Amram’s first words to his sons (in 4Q543 2 and 4Q545 1a i) as well as several passages from the vision account seem to present future events related to Moses and Aaron, the exodus, and the future priesthood. This is the case in 4Q543 2, 4: ומלאך אל תתקרה‎; ⁵³ 4Q543 2, 5:]תעבד בארעא דא ודין ח̊סין ת̊‎; ⁵⁴ 4Q543 15, 1: תתוב לעמך̇‎; ⁵⁵ 4Q543 16, 2: תעבד לעמך‎; ⁵⁶ 4Q545 4, 16: א]ח̇וה לכה רז עובדה̇]‎; ⁵⁷ 4Q545 4, 18: שביעי באנוש רעות̊[ה ית]קרה‎; ⁵⁸ 4Q545 4, 19: יתבחר לכ̊הן עלמין̇‎; ⁵⁹ 4Q546 6, 2: ] ו̇תטבע̇‎; ⁶⁰ 4Q547 9, 4: בהר סיני‎; ⁶¹ 4Q547 9, 6: ב]ר̊ה יתרם כהן מן כול בני עלמא באח̇[רי‎/והי‎. ⁶² In some of these examples (those with second person active verbs), Amram may be stating what he wants his sons to do, not what he knows will happen, but the general impression is that these are actual predictions. Even though there is no mention of a fixed history in set periods, a typical deterministic trait in apocalyptic texts, ⁶³ the notion of prediction and the use of words like “will be chosen,” “[his] favor/will,” and “the secret” seem to imply some kind of divine determinism. The testament-like setting and the emphasis on the authoritative value of the writing (פרשגן‎) appear to underline the trustworthiness of the predictions. ⁶⁴

In addition to the predictive passages, the following fragments include words or expressions that may indicate a deterministic point of view or some kind of divine control. However, the lack of context precludes clear conclusions: 4Q543 16, 4: עדבך‎; ⁶⁵ 4Q545 2, 2: ש]מ̊ו ט̊[ע]מ̇א בקשט̇ לעלם ‎; ⁶⁶ 4Q545 9, 5:ו̊מנ˙י עדניהו̇ן[‎; ⁶⁷ 4Q546 12, 4: ו̇רז .מרים‎. ⁶⁸

So far, only the manuscripts universally acknowledged to belong to VA have been considered, but some scholars consider 4Q548 and 4Q549 part of the composition as well. Milik followed Starcky’s count of only five manuscripts (543–547), but mentioned a manuscript (4Q548), which could be another copy. ⁶⁹ In Puech’s edition of the text, both 4Q548 and 4Q549 are part of the presentation of 4QVisions de ‛Amram, but with a degree of uncertainty. Puech notes that there are no overlaps with the other manuscripts, yet the dualistic distinction between light and darkness in 4Q548 could fit into the framework of VA, and 4Q549 could be the ending of a testament-like work (with a meal, somebody dying, a genealogy). ⁷⁰ Several scholars have expressed reluctance about including the two manuscripts. For instance, Duke sees the identification of either manuscript with VA as very unlikely, ⁷¹ as do Jurgens and Perrin, ⁷² while Goldman accepts 4Q549, but not 4Q548. ⁷³ In the present context, 4Q549 is of little interest since nothing in the two fragments seems relevant to the discussion of human choice and determinism. In 4Q548, however, we have a fragmentary description of what appears to be a prediction about the eschatological end for people on either side of the dualistic divide. As was the case with the predictions mentioned above, this in itself may be interpreted as a form of determinism, but there is no information about the mechanism deciding one’s place in this scenario, whether it is personal choice or divine decree. Yet, in the text of 4Q548 presented by Beyer and by García Martínez and Tigchelaar, one fragment includes the words בחיר לקשט‎ (“chosen for truth”), which may indicate that the latter concept (divine decree) is decisive. ⁷⁴

How are we to interpret the probable presence of deterministic ideas in VA (and in a fragment that may belong to it)? Should it affect our understanding of the passage with Amram’s supposed choice? Divine election, God’s decisive action in—and control of—history, and the reality of prediction and prophecy (and hence some kind of pre-determined future) are common notions in much of the Old Testament and related texts. In this sense, most ancient Jews must have subscribed to a basic layer of determinism. ⁷⁵ However, it is clear that these notions often occur alongside statements that seem to presuppose human responsibility and the possibility of choice. For instance, the Book of Proverbs contains many paraenetical passages that point in the latter direction, but at the same time some of the most clearly deterministic ideas in the Old Testament are found in the same book (Prov 16:1, 9, 33; 19:21; 20:24; 29:26). A similar tension can be observed in the Qumran material. The Community Rule, as preserved in 1QS, contains the most well-known and explicit deterministic statements in the Qumran corpus, primarily in the so-called Treatise on the Two Spirits (1QS III, 13–IV, 26; for instance 1QS III, 15: מאל דעות כול הויֿה ונהייה ולפני היותם הכין כול מחשבתם‎, “from the God of knowledge is everything that is and comes into being, and before their coming into being he has established all of their design/plan”), but also in the final part of the scroll (1QS XI, 11: ובדעתו נהיה כול וךול הוֿיֿה במחשבתו יכיֿנו ומבלעדיו לוא יעשה‎, “and by his knowledge everything has come into being, and everything that is, by his plan/in its design he establishes it, and apart from him it is not done”; cf. XI, 17–18). ⁷⁶ Yet, in other places, scholars have noted statements that seem to imply personal responsibility with reward and punishment and the possibility of choice. Especially the passages that describe the entry into the community (such as 1QS V, 8.10.22) and the use of the term המתנדבים‎ might give the impression of voluntarism and conscious, personal choice. For instance, 1QS V, 1 (המתנדבים לשוב מכול רע‎) is translated by García Martínez and Tigchelaar as “who freely volunteer to convert from all evil.” ⁷⁷ Such passages, along with the penal code (1QS VI, 24–VII, 25), have led some scholars to question the degree of determinism in the Rule. However, Stauber argues convincingly that the meaning of מתנדבים‎ should not be equated with the notion of “free will” or personal volition, but rather with having been “incited/inspired to do” something. ⁷⁸ In support of this view, one could cite 1Q14 8–10, 7 (ולכ[ו]ל המתנד̇בים לוסף על בחיר̊י [אל), which combines מתנדבים‎ with the notion of “joining the chosen ones of [God].” ⁷⁹ This shows that the use of מתנדבים‎ does not contradict the notions of determinism and chosenness. Further, Stauber argues that a penal code does not imply belief in “free will”—“since laws imply only that actions are taken, not that agency is present.” ⁸⁰ In general, passages that appear to imply free choice and human agency ought to be interpreted in light of the fundamental theological assumptions inherent in clearly deterministic passages rather than the other way around. ⁸¹

Applying this line of thought to VA seems to provide a warning that we should not conclude from one very dubious passage that choice and “free will” were central concerns of a text that may in fact have been more concerned with divine determinism. ⁸² Admittedly, the examples referred to above are mostly indicative of a kind of underlying determinism that is never stated in explicit terms in the extant material. However, if we follow Machiela’s recent proposal to see 4Q542 (Testament of Qahat) as a part of VA, we suddenly have a passage that presents a much clearer and very explicit statement of divine determinism. Connecting 4Q542 and VA is not a new idea, ⁸³ but Machiela argues not only that 4Q542 and 4Q547 were written by the same scribe and were part of the same scroll, but that they can plausibly be seen as parts of the same work, that is, “what scholars have hitherto called the Testament of Qahat was, in fact, an integral sub-section of the Visions of Amram.” ⁸⁴

In 4Q542 1 i, 2–3, God is described as follows: הוא אלה עלםיה ומרא כול םעבדיא ושליט בכולא למעבד בהון כרעותה‎ (“He is the God of the ages/eternity and Lord of all the deeds and ruler of everything/everyone, to do with them according to his will”). ⁸⁵ This description seems to leave little room for personal choice on the part of human beings. If it was in fact part of VA, it ought to make us even more careful not to interpret the (possible) reference to Amram’s choosing in terms of “free will” and human agency in relation to the divine realm. On the other hand, if we let the standard reading of the contested passage in VA be decisive, the notion of human choice expressed in that interpretation could be used as an argument against seeing 4Q542 as a part of VA. However, since the actual text provides only very limited support for the “free choice” reading, as explained above, this argument is hardly conclusive. ⁸⁶

The Iranian connection? VA and Yasna 30:3

An interesting offshoot of the standard interpretation of VA as upholding some kind of “free choice” theology is the attempt to find an external parallel to the point of view ascribed to VA. Some scholars see an Iranian connection and have drawn attention to perceived similarities between VA and Yasna 30, one of the Avestan Gathas (the oldest part of the Avesta, often uncritically stated to derive from Zarathustra himself). ⁸⁷ Popović, in particular, has argued that there is a striking agreement between the passage from VA with Amram’s “choice” and verses 3–4 of Yasna 30. ⁸⁸

The history of scholarship has seen many attempts at finding parallels between various aspects of ancient Judaism and Zoroastrian or—expressed more generally—Iranian religious thought. Dualistic and eschatological ideas, the resurrection of the dead, the development of angelology and demonology, periodization of history, etc. have all been claimed to derive from Iranian influence. Instances of influence have been pointed out in canonical Old Testament books, apocrypha, and pseudepigrapha—and the discovery of the Qumran material provided added impetus for comparison and claims of influence. ⁸⁹ For a long time, Judaism developed as part of the Persian empire and there must have been numerous opportunities for contact between Iranians and Jews, in the Persian period and beyond; hence, the idea that some kind of influence must have occurred ought to be uncontroversial. ⁹⁰ In VA, the very first word (פרשגן‎) is an Iranian loanword (see n. 15), as is the word רז‎, which was mentioned twice above. However, in the realm of ideas and religious motives, it is more problematic to provide tangible proof of influence in specific cases due to the oft-noted uncertainties regarding the dating of the Iranian material and the possibility of deriving new developments in Jewish writings from “internal” factors. Shaked rightly rejects the two extreme positions of either denying Iranian influence altogether or attributing practically everything new in post-Biblical Judaism to such influence. ⁹¹

In his discussion of Yasna 30 and VA, Popović does not go further into the historic connection between the two texts, but focuses on the similarities as a means for understanding the position taken in VA with regard to determinism and choice. Here, I want to argue briefly that the similarities are not as clear as they seem to be in Popović’s presentation.

Popović discusses the passage based on a translation of Yasna 30:3–4 taken from Alexander:

The two primeval Spirits who are twins were revealed [to me, Zarathustra] in sleep. Their ways of thinking, speaking, and behaving are two: the good and the evil. And between these two [ways] the wise men have rightly chosen, and not the foolish ones. And when these two Spirits met, they established at the origin life and non-life and that at the end the worst existence will be for the followers of Falsehood and for the followers of Truth the best thinking. ⁹²

Clearly, the two spirits may remind us of the two beings in VA, the notion of being revealed in sleep agrees with Amram’s vision/dream, and the notion of choice fits well with the standard reading of the passage in VA, as argued by Popović. However, things are a bit more complicated. Unfortunately the Gathas are notoriously difficult to understand. According to Shaked, the “difficulties inherent in the text of the Gathas have given it the reputation of being one of the most obscure texts in religious literature.” ⁹³ In particular Yasna 30:3 (the first part of the passage quoted above, called the “Twins Stanza”) has been the subject of much discussion. Numerous competing interpretations and translations have been proposed. ⁹⁴ For our purposes, especially the word translated “in sleep” above is crucial. The Avestan is x^vafənā, which in the interpretation behind the translation used by Popović is parsed as an instrumental singular. This is possible, but seeing the word as a nominative dual may be a more plausible solution, as argued by Humbach, who gives the following (annotated) translation of verse 3:

These (are) the two spirits (present) in the primal (stage of one’s existence), twins who have become famed (manifesting themselves as) the two (kinds of) dreams, the two (kinds of) thoughts and words, (and) the two (kinds of) actions, the better and the evil. Between these two, the munificent discriminate rightly, (but) not the miserly. ⁹⁵

Humbach criticizes translations such as “have been revealed by a vision” for not accounting for the meaning of the verb in the sentence (asruuātəm), which literally means “they (du.) have been heard.” ⁹⁶ He adds that “visions are usually seen, not heard,” but it might be argued that dreams (the basic meaning of the root in x^vafənā is “sleep”) can surely include aural impressions as well. However, he also notes that with his interpretation, x^vafənā fits better into the immediate context as the first of four dual forms describing the two spirits. ⁹⁷ Skjærvø gives a similar translation and explains: “The origins of good and evil are two Spirits (Avestan manyu ‘mental impulse, inspiration’), depicted as ‘twin ‘sleeps’ (= sleeping things),’ presumably twin fetuses embodying contrasting and irreconcilable potentials for good and evil.” ⁹⁸

It bears repeating that the meaning of the passage is contested and uncertain, and it seems difficult to decide whether Humbach’s interpretation is correct. But if it is (or if Insler’s is, see n. 98), the similarity between the passages in Yasna 30 and VA is less striking. Still, the general situation with a human being choosing between two antagonistically opposed figures from the spiritual or divine realm could be argued to be the same. According to Popović, “Both texts emphasize the choice individuals have in determining their own destiny.” ⁹⁹ However, in light of the possibility that there may be no choice in VA (as argued in the preceding sections), we ought to be very cautious when making such comparisons. The possible parallel between the two texts is interesting, but as this brief discussion has attempted to show, it cannot be used uncritically.

Further reflections on human choice and divine determinism

The standard reading of Amram’s initial interaction with the two beings does not find clear and unambiguous support in the actual text of VA, nor does the proposed Iranian parallel provide independent support for that interpretation. In contrast, we found that VA contains several passages that seem to include deterministic thoughts rather than notions of human choice. Does that mean that we can conclude that the standard interpretation of the problematic passage cannot be right? Unfortunately, things are not that simple. As noted above, passages referring to divine control and determinism often occur side by side with passages that seem to presuppose human agency. For instance, the lines after the highly deterministic words quoted above from 4Q542 1 i, 2–3 contain exhortations to be careful with the inheritance and to hold firm to the commands of the ancestors, walking in uprightness (4Q542 1 i, 4–10: ואזלין בישירותא ;אחדו בממר יע̇קוב אבוכון ;אזדהרו בירותתא‎). Conversely, even in a well-known text that clearly speaks about human choice (Deut 30:15–20) we find an undercurrent of determinism if we look at the passage in its broader context. Moses, on behalf of YHWH, presents the Israelites with two radically opposed paths, described as life and death respectively; in 30:19, he directly admonishes them to choose life (ובחרת בחיים‎). Goldman notes the importance of these verses in relation to VA, suggesting that the passage in VA may have been modeled on the biblical command. ¹⁰⁰ Since Amram is Moses’ father and probably used the vision in his final words to his sons, this connection indeed seems important—it may imply the view that Moses later reused his father’s description of the fundamental dualistic divide when he made his famous speech on the plains of Moab. However, the possible connection to Deuteronomy 30 does not constitute clear evidence for a “free choice” reading of VA. In Deuteronomy, the wider context is characterized by divine control of events (getting the Israelites into Egypt, then out again, leading them through the desert) and the immediate context includes two intriguing statements, viz. 29:3 (“YHWH has not given you a heart to understand and eyes to see and ears to hear until this day,” ולא נתן יהוה לכם לב לדעת ועינים לראות ואזנים לשמוע עד היום הזה‎) and 30:6 (“and YHWH your God will circumcise your heart and the heart of your offspring to love YHWH your God,” ומל יהוה אלהיך את־לבבך ואת־לבב זרעך לאהבה את־יהוה אלהיך‎). ¹⁰¹ This suggests that divine control and determinism are somehow more basic than actual human agency.

When discussing determinism and human choice, it could be argued that the notion of determinism applies on the general, historical level, while individual human beings make their own choices. However, this does not seem to fit the all-encompassing statements on divine control made in texts such as 1QS and 4Q542. To be sure, within the Treatise on the Two Spirits in 1QS it is stated that each of the opposing spirits are present in every human being in various proportions and that until now the spirits of truth and deceit/injustice are fighting in the heart of man (1QS IV, 23, עד הנה יריבו רוחי אמת ועול בלבב גבר‎). This interior struggle in 1QS may be relevant for the interpretation of the “great dispute” that the beings are having over Amram in VA. Popović is surely right when he states that, as far as can be told from the preserved remains, VA “does not locate the strife of the angelic beings somehow within the human self.” ¹⁰² Yet, on a symbolic level, the text may still be making this point (which does not mean that it could not be describing the “real” world at the same time as well). What the two beings represent is not immediately clear from the text. According to Goldman, the dualism in VA “relates to life and death rather than to righteousness and sinfulness, . . . its purpose lies in emphasizing the necessity of ‘choosing Life’ on a personal, familial, and national level.” ¹⁰³ She states that VA “contains no traces of moral judgment in its treatment of the forces of darkness.” ¹⁰⁴ Similarly, Frey notes the “almost complete lack of specific ethical precepts,” arguing that the pattern of dualistic thought in the document should be classified as “merely cosmic.” ¹⁰⁵ This is true, as far as the surviving fragments allow us to see, but it seems possible that Amram would derive some lessons from the cosmic dualism when addressing his children at the time of his death. Jurgens suggests that Amram’s vision “depicts his choice between a righteous or corrupted priesthood.” ¹⁰⁶ The extant text certainly seems to attach importance to the priesthood and the need to keep the priestly line pure through endogamy. However, what the two beings say about their role as rulers of all mankind could indicate a broader reference for the dualism represented in the vision and in the struggle between the two beings.

How one describes the mechanism deciding the outcome of that struggle in VA depends on one’s acceptance or rejection of the standard reading and interpretation of 4Q544 1, 12 and 4Q547 1–2, 12, and on one’s understanding of the deterministic passages discussed above, including the one from 4Q542, which seems to indicate that God has decided everything. In 1QS, we must surely interpret the internal struggle in light of the broader context (for instance III, 15 quoted above and III, 16, ואין להשנות‎, “and nothing can be changed”). That is, God has determined the outcome of the fight in advance. However, this outcome cannot be known with complete certainty until the appointed time for God’s visitation (מועד פקודה‎, 1QS IV, 18–19). Until that time, a person’s standing may be partially discerned through the actions that he performs and the association of such actions with either of the two fundamental spirits (1QS IV, 2–11) and through yearly scrutiny of the behavior and spiritual condition of each person (1QS V, 24). ¹⁰⁷ Distinguishing between the completely predetermined plan of God and the limited human insight into the exact status of individual persons may in fact be the key to interpreting the co-occurrence of deterministic thoughts with statements that seem to imply personal choice, both in 1QS and other texts where a similar situation obtains. A person’s final destiny as well as the entire path of his life, his career of apostasy and/or (re)conversion, his response to exhortation, and so on may be determined and unalterable, but since we do not know the details of God’s plan, exhortation and the threat of punishment or promise of reward still make sense, hence the need to treat people as if they had freedom to make choices. ¹⁰⁸ The predetermined plan has to be played out by the actions of actual human beings, who are—for the most part—not aware of that plan.

A final reflection regarding the notion of choice in VA has to do with the description of the two beings as they appear to Amram. The fact that one is frightening and the other one smiling may indicate that the choice is not truly free. Goldman refers to “Malchiresha’s fearsome appearance” as a “warning sign” that would aid Amram in making the right decision. ¹⁰⁹ In this connection, she notes the similarity with the figure of Death in chapter 17 of the long recension of the Greek version of the Testament of Abraham. ¹¹⁰ Although the details are probably less similar than Goldman imagines (since the being in VA is unlikely to be serpent-like like Death in the Testament of Abraham, cf. above), it might be significant that Death in the Testament of Abraham appears differently to sinners and the righteous. If something similar is at play in VA, the notion of Amram’s “personal choice” seems to become even less clear than it already was.

Concluding remarks

So, did Amram have a choice? We do not know for sure. The fragmentary state of VA forces us to be content with stating that it may have been the case that Amram had a chance to seek to be ruled by one or the other of the two beings that he saw in his vision. I have argued that the first part of the usual reading of the fragmentary passage in which the beings ask Amram a question is most likely correct (with the verb בעא‎) but that the second, wholly reconstructed, part is more problematic. Other possible reconstructions do not imply the same degree of choice for Amram in relation to the divine realm as in the usual reconstruction. Furthermore, even if we were to accept the reconstruction with לאשתלטה‎, it does not necessarily mean that the composition as a whole advocates a “free choice” or “free will” theology. First, the fragments do not answer the question whether the supposed choice is an opportunity for every human being or only for Amram. ¹¹¹ Second, it is doubtful that the issue of “choice” is a central concern in the composition—in fact, it was argued that nothing else in the surviving fragments focuses on human choice. Rather, several passages suggest that an underlying notion of determinism is more prevalent in the composition. If 4Q542 belongs to VA, the text contains an explicit statement of far-reaching divine determinism. Even if the standard reading of the passage under discussion is accepted, the comparison with other texts that seem to present human beings with a choice suggests that the situation may be more complex and that a choice (if indeed there is a choice) is not always “free.” In short, there is good reason to treat Amram’s supposed choice in a much more cautious manner than has, with few exceptions, hitherto been the case.