But there seems to be a possibility that is not considered by Casey, namely that scholars might not have enough relevant material of any kind to reconstruct sayings of Jesus with any confi dence. There are all sorts of aspects of what has happened in the past which are effectively shut off for us from investigation by lack of literary evidence. What if the Aramaic sayings of Jesus were often in that situation? This might seem a pity, but our mere desire that there be suffi cient evidence is not enough to generate evidence. Using problematic sources (such as Aramaic from centuries later) simply because we have no other evidence is not satisfactory.
However, Casey has a response. In several of his publications he insists on the stability of the Aramaic language in order to justify his use of later sources.
He says,
While it is clear that some changes such as the dropping of ) at the beginning of
#n) and the decline of the use of the absolute state of the noun did take place, it is equally clear that the basic vocabulary and structure of the language did not alter over a period of centuries. The semantic area of common words such as rm), b) and Mwq continued to include all the basic uses attested in earlier Aramaic, and idiomatic features such as the construct state of the noun and the uses of participles as fi nite verbs are also found in many different dialects over a long period of time.21
20. Casey, ‘The Aramaic Background of Mark 9:11: A Response to J. K. Aitken’, JTS n.s.
55 (2004), pp. 92–102 (101).
21. Casey, ‘General, Generic and Indefi nite: The Use of the Term “Son of Man” in Aramaic Sources and in the Teaching of Jesus’, JSNT 29 (1987), pp. 21–56 (23). A similar list can be found in Casey, ‘Aramaic Idiom and the Son of Man Problem’, p. 5, though four of the fi ve terms he cites also occur in Hebrew.
68
‘Who is this Son of Man?’
However, on the basis of the arguments presented in this quotation, it would be equally possible to claim that Modern Hebrew could be used as a reliable source for the language of Jesus since it possesses rm), b), and Mwq in approximately the right senses, uses the construct state and uses participles to express the present tense.
Casey’s argument is not merely problematic with respect to the Aramaic which he actually reconstructs, but also with respect to the Aramaic he does not reconstruct. Negative evidence is diffi cult to evaluate, but if we begin by asking, on the basis of our positive evidence, how much of all the evidence that has existed we actually have, we are starting from a legitimate point. It seems that the extant fi rst-century Aramaic represents a small fraction of the Aramaic that once existed. We therefore need to exercise extreme caution in making claims such as that Aramaic could have or could not have said a certain thing.
And yet Casey uses whether he can reconstruct a saying in Aramaic as one of his foremost criteria for whether or not a saying is authentic.22
From Aramaic into Greek
Casey’s theory of the translation of Aramaic into Greek is that the phrase was properly translated, but then misconstrued – not mistranslated. The bilingual speakers of Aramaic and Greek intended the phrase o9 ui9oj tou= a0nqrw&pou in a different sense from how it was understood by monoglot Greek speakers. At the same time the reason for the use of the fi rst article by the bilingual speakers was ‘interference’ from Greek. In a fascinating passage Casey even allows for the possibility that Jesus might have approved of the Greek expression o9 ui9oj tou= a0nqrw&pou himself. He says: ‘A bilingual Jesus might have accepted this translation [o9 ui9oj tou= a0nqrw&pou]. A Jesus who taught in Greek, however, would surely have expressed himself quite differently.’23
If Casey is prepared to envisage a bilingual Jesus, this raises a question as to why so much emphasis is put by him on the Aramaic. He may, of course, 22. In a similar way, Richard Bauckham, ‘The Son of Man: “A Man in My Position” or
“Someone”?’, JSNT 23 (1985), pp. 23–33 (23–24), notes how Lindars uses conformity with the idiom he has reconstructed as a criterion of ‘dominical authenticity’.
23. Casey, ‘Idiom and Translation: Some Aspects of the Son of Man Problem’, NTS 41
(1995), pp. 164–82 (178).
4. Expressing Defi niteness in Aramaic
69
merely be entertaining the possibility for the sake of argument, but there is a serious possibility that, even if Jesus did not teach in Greek, his sayings were discussed in Greek in his hearing contemporaneously. If so, the translation o9 ui9oj tou= a0nqrw&pou may be almost authorized by Jesus. This in itself would be a reason to put the emphasis on the Greek. Likewise, we would be wise not to put too much stress on the Aramaic if, as Barr and Emerton maintain, Hebrew was also widely used.24
But suppose we knew that Jesus taught only in Aramaic, would this not give us ground to begin with the Aramaic? As Casey says,
[i]f Jesus spoke Aramaic, it should be possible to see more clearly what he meant in Aramaic, rather than in translation Greek. In principle, sayings which cannot be reconstructed in feasible Aramaic cannot be authentic sayings of Jesus, though this criterion must be handled with care, because severe problems can be caused both by free translation and by the meagre remains of Aramaic from the time of Jesus.25
I would wish to qualify this statement: if both the Aramaic and Greek were extant, it would obviously be possible to see the intention more clearly in the original than in the translation. However, it is not evident to me that the same can be said when comparing a reconstructed original with an extant translation.
We proceed now to consider how to reconstruct the Vorlage. To quote Casey:
‘Reconstruction of a Vorlage is possible only from a literal translation.’26 I quite agree. The question is how we know that the translators did translate literally.
Here Casey comes to his principal piece of evidence: ‘Nothing gives clearer evidence of literal translation than o9 ui9oj tou= a0nqrw&pou. Neither this phrase, nor the anarthrous ui9oj a0nqrw&pou, is found in natural Greek.’27 Casey does cite other evidence, but, according to him, none is as strong as this. Yet, in claiming that ui9oj a0nqrw&pou is also not natural Greek, he is presumably saying that the Greek phrases with and without the article would both be able 24. Details in Randall Buth, ‘A More Complete Semitic Background for )#n)-rb, “Son of Man”’, in C. A. Evans and J. A. Sanders (eds), The Function of Scripture in Early Jewish and Christian Tradition (JSNTSup, 154; Sheffi eld; Sheffi eld Academic Press, 1998), pp. 181–86.
25. Casey, ‘Idiom and Translation’, p. 179.
26. Casey, ‘Idiom and Translation’, p. 171.
27. Casey, ‘Idiom and Translation’, p. 173.
70
‘Who is this Son of Man?’
to be cited as evidence of literal translation, which in turn challenges Casey’s view that one of the renderings can be used as evidence of literal translation.
Moreover, if this is the strongest argument, some might perceive a degree of circularity in Casey’s position. Even if we could establish that it was meaningful to use the word ‘literal’ of the translation of this phrase, this would not establish that the translation was consistently literal. A translation can be partially literal too. Nothing but a consistently literal translation will allow us to reconstruct its Vorlage.
But here I come to a deeper reservation about Casey’s method. There is an extensive literature on the translation technique of biblical versions. Those undertaking research in this area have compared extant texts of the translation with extant texts in the source language. In this way, analysis of translation technique has been based on evidence that is to a certain degree known.
Nevertheless, the procedure is complicated. There is often legitimate scholarly disagreement as to which source text gave rise to the text in the receptor language. To cite the most signifi cant parallel to the situation Casey envisages: there is often dispute as to which Hebrew text stands behind a translation of the Greek Old Testament. If this happens where texts in the source language are extant a
nd we have a considerable body of comparative material, it is surely even harder to know anything about translation technique from Aramaic into Greek in the fi rst century, from which we have no fi rm illustrations of translation technique based on knowledge of texts in both source and target languages and our best information is provided by the way parts of Ezra and Daniel were rendered into Greek. The method of using such a little known area as the foundation for study, as Casey does, is problematic.
We come then to more specifi c questions of how the translation that Casey proposes occurred. He proposes that more than one translator, specifi cally the translators of the Aramaic sayings material in Mark and the Aramaic material in Q, chose to represent cases of bar enash(a) by o9 ui9oj tou= a0nqrw&&pou when it contained application to Jesus, but not otherwise.28 I have three objections to this supposition:
28. Casey,
Solution, pp. 253–54. Casey seeks to account for the uniformity in an ad hoc manner on p. 265.
4. Expressing Defi niteness in Aramaic
71
1. Translators of this period were not consistent unless they chose to be consistent (as occurred to some degree with the kaige translations); consistency does not occur naturally without effort.
2. The process of differential translation, whereby translators rendered the same phrase in two different ways, lacks adequate parallels from ancient translations.
3. To suppose that the same striking differential translation method occurred in two translations independently with no sign of alternatives presupposes an implausible coincidence.
As for the specifi c mechanism he proposes which triggers the translation, he speaks of ‘interference’. I note here that if the translators were genuinely bilingual, then interference from Greek to Aramaic was just as likely as interference the other way round. But if we suppose that there was interference from Aramaic to Greek, what would this look like? Here I will quote Casey: A particularly important distorting factor is the double level of interference present when bilingual people translate texts. Bilingual people generally experience interference between the languages which they use. This is increased when they translate from one language into another, because the features of the host language are present before them to cause the interference. Neubert has noted the problems which arise when monoglot users of the target language do not share the presuppositions of the language community in which the text was written. These problems should not be confused with mistakes. . . . Perhaps the most famous cultural shift in our fi eld is the regular translation of hrwt with no/moj.29
This seems to be correct as far as interference goes. However, ‘generally experience interference’ allows for a range of possibilities and it would be appropriate to suggest that it could more helpfully be stated that bilinguals sometimes experience some interference. The question then arises as to how likely it is that interference has occurred in any given instance, but Casey supplies no indications of frequency. If it is not a frequently occurring phenomenon, or one that is frequently detectable, why should we prefer an explanation invoking 29. Casey, ‘Idiom and Translation’, p. 175.
72
‘Who is this Son of Man?’
it in any particular instance? I would maintain that an explanation involving interference should only be preferred if all comparable explanations without positing it are shown to fail. The positing of a relatively infrequent occurrence should thus only occur if it is not possible to explain the text in the receptor language without interference.
Expressing defi niteness in Aramaic
Back in 1967, Vermes said, ‘It might seem from the vast mass of “son of man”
literature that the linguistic aspect has been explored in such depth that nothing new can possibly emerge; but this is not so.’30 If I am not mistaken, there continue to remain open signifi cant avenues for linguistic research upon this question. Arguably there has been myopia concerning the sorts of linguistic questions that have been addressed in debates about the Son of Man. More particularly, debate has often assumed that the question of whether or not the phrase as used by Jesus was defi nite or indefi nite can only be settled by investigating whether the Aramaic defi nite state (originally, a sort of post-positive defi nite article) had continuing force at the time of Jesus. This is a signifi cant question, and I think there is much to be said for the view that it retained considerable force in Western Aramaic.31 Even if the Aramaic defi nite state was breaking down, this does not mean that it was breaking down at the same rate and in the same way for all nouns simultaneously. But even if I were to concede the unlikely position that the defi nite state had no force that still would not lead me to the view that defi niteness was not expressed. In fact, Aramaic had numerous ways of indicating whether or not an entity was defi nite. Moule intuitively sensed this, but could not provide specifi c relevant evidence: It has been asserted (though the assertion has not gone unchallenged) that in Palestinian Aramaic of this period bar naš and bar naša [sic] were indistinguishable in meaning, so that an unambiguous defi nite – ‘the Son of Man’ – could not have been expressed. But it is hard to conceive that it would have been beyond the 30. Vermes, ‘Use’, p. 311.
31. Casey, ‘Aramaic Idiom and the Son of Man Problem’, p. 6, uses ad hoc arguments to seek to deny the signifi cance of the distinction between Eastern and Western Aramaic which is generally made by scholars.
4. Expressing Defi niteness in Aramaic
73
capacity of the language to indicate an unambiguously defi nite or deictic sense by some means or other – perhaps by a periphrasis.32
Quite apart from the defi nite state, there are many ways in which relevant dialects of Aramaic could mark defi niteness. These are often at the discourse level or at the level of the sentence rather than the phrase. This should not be a problem since, after all, Casey has often extolled the virtue of reconstructing whole sentences or paragraphs in Aramaic,33 and has even spoken of the ‘dire effects . . . of treating sentences “in isolation”’.34 One demonstrative pronoun accompanying a specifi c phrase early on in a discourse may well help that phrase to be understood as defi nite through the following discourse. But we can also look at other clues at the phrase level. To begin with the fi rst example we can take the direct object marker. In Hebrew there is a signifi cant correlation between the object marker t) and defi niteness. Aramaic dialects have their own equivalents, l and ty. In Biblical Aramaic l marks a defi nite pronominal object and it seems probable that l or ty would have been used by Jesus at least before some defi nite common nouns.35 If Jesus used the phrase ‘son of man’
with any frequency in his teaching then it would be bound to occur in some object constructions and thereby indicate to Aramaic speakers whether or not it was defi nite. In fact, the ‘son of man’ sayings in the gospels are sometimes found to have the ‘son of man’ in object position. Verbs used include ‘see’
(Mk 13.26; 14.62; Mt. 16.28; 26.64; Lk. 21.27)36 or ‘betray’ (Lk. 22.48). It might be objected that some of these sayings are not authentic, but we must remember that in the work of Casey (and Lindars for that matter) the limits of what is authentic have been set by their own understanding of the Aramaic idiom. Even if one were to dispute the authenticity of all of these sayings, the overwhelming balance of probability rests with the view that Jesus would 32. C. F. D. Moule, ‘“The Son of Man”: Some of the Facts’, NTS 41 (1995), pp. 277–79
(278). For a consideration of possible ways of marking determination see James Barr,
‘“Determination” and the Defi nite Article in Hebrew’, JSS 34 (1989), pp. 307–35.
33. Casey,
Solution, p. 19.
34. Casey,
Solution, p. 41.
35. Cf. E. Y. Kutscher, ‘The Language of the Genesis Apocryphon’ in Kutscher, Hebrew and Aramaic Studies (Jerusalem: Magnes, 1977), pp. 20–21; G. Dalman, Grammatik des Jüdisch-Palästinischen Aramäisch (Leipzig: J. C. Hinrichs, 2nd edn, 1905), p. 110.
36. C
asey, ‘Idiom and Translation’, p. 167, even cites a Syriac text using the object marker l after the word ‘see’ in conjunction with a defi nite ‘son of man’.
74
‘Who is this Son of Man?’
sometimes have had cause to utter the phrase in question as an object in a sentence, and listeners would have known thereby whether or not he intended it to be understood as defi nite. An Aramaic example using the word ‘see’ is from 4Q214b 2, which reads [M]hrb)l tyzx ‘I saw Abraha[m]’. Another possible example of this from later Aramaic would be the example cited from Genesis Rabbah 38.13 by Vermes:37
)xwr lybsd )#n rbl dwgsn
‘Let us worship the man who bears the wind/spirit’.
The object marker is used and the phrase ‘son of man’ is understood as defi nite.
Note also the relative clause which we will comment on shortly.
A further way of marking defi niteness would involve an anticipatory object suffi x on the verb followed by an object marker as in Kwnxl hxk#) ‘[he found + him] [object marker + Enoch]’ ( 1QapGen ar 2.4.23).
There are other ways whereby defi niteness could have been indicated. With each of these we are hampered by our lack of knowledge of the details of the precise form of language Jesus spoke. It is diffi cult therefore to be sure which constructions may have been used by Jesus and which not, and also diffi cult to know whether, even if a construction was used by Jesus, it was brought into connection with the phrase ‘son of man’. However, together they indicate the many ways in which defi niteness may have been indicated.
1. Repetition of a noun within a discourse may indicate that later occurrences within the discourse are identifi able and thus defi nite.
2. If an indefi nite noun is introduced in a narrative and then referred to by use of a pronoun, a subsequent use of the full noun within a narrative may be judged to be defi nite.
3. Some nouns, such as proper nouns, may be inherently defi nite.
4. A preceding genitive phrase may use an anticipatory suffi x and thereby indicate that what follows is defi nite. It is conceivable that this could have occurred with phrases such as ‘the parousia of the son of man’ (Mt. 24.27, 37. Vermes, ‘Use’, p. 324.
Who Is This Son of Man Page 11