All your Aramaic are belong to us
In an earlier post, we introduced the subject of Maurice Casey’s Aramaic monomania. His affliction led him not only to claim that he has revealed the original language behind significant parts of the New Testament but to insist that he has discovered the actual words of Jesus.
Casey directs our attention to particular sections of Mark’s gospel and the Matthean-Lukan double tradition (Q) as alleged examples of “interference” at work.
Some features of Mark’s Greek are characteristic of the work of bilinguals. For example, at Mark 9.43, 45, 47 we read καλόν [kalon] where a monoglot Greek-speaker would use a comparative. Aramaic has no comparative, so the use of καλόν [kalon] is due to interference in someone who was used to saying טב [tav]. (Aramaic Sources of Mark’s Gospel, p. 85., emphasis mine)
Other signs of interference include the use of certain words. For example, in the Lord’s prayer we are to ask God to forgive τὰ ὀφειλήματα [ta opheilēmata] (Matt. 6.12), literally our ‘debts’, but a metaphor for our ‘sins’, so a literal translation of the Aramaic חובינא [kobena]. (An Aramaic Approach to Q, p. 55, emphasis mine)
Accordingly, Mark did not mean that Jesus was angry. He was suffering from interference, the influence of one of his languages on another. All bilinguals suffer from interference, especially when they are translating, because the word which causes the interference is in the text which they are translating. (Jesus of Nazareth, p. 63, emphasis mine, incoherence Casey’s)
A correct understanding of interference is essential if we are to understand our Gospel translators, and consequently essential if we are to have any confidence in our Aramaic reconstructions. (Aramaic Sources of Mark’s Gospel, p. 55, emphasis mine)
What does Casey mean by “interference”?
Since Casey’s argument depends heavily on the concept of interference, you might think he would have defined the term for his readers. You would probably also expect that if he believes bilinguals have more interference when translating than when composing, he would back that idea up with research.
But as usual, Casey disappoints. He gives examples of interference, but he fails to define the term. That’s a shame, since the literature surrounding this idea is vast and fascinating, with no shortage of scholarly contention. So before we go any further we need to rectify this situation.
The term interference is now somewhat out of favor. In the literature we see several alternatives, including:
- Language transfer
- Linguistic interference
- The role of the mother tongue
- Native language influence
- Language mixing
- Crosslinguistic influence (which includes transfer and avoidance)
(See The Handbook of Second Language Acquisition, p. 436)
Many scholars now prefer transfer or “L1 transfer” over interference. Simply put, L1 transfer refers to the phenomenon in which people apply the rules and structures of their first language (L1) to their second language (L2). When Casey uses the term interference, he means negative L1 transfer. In some cases, transfer from L1 has a positive effect on a person’s ability to learn and understand L2. If the second language has similar case structures, syntax, and many true cognates, students have an advantage. However, applying L1 rules inappropriately can lead to errors.
Examples of negative transfer include word-for-word translations that are often unintentionally hilarious. You may be familiar with a 19th century book called English as She Is Spoke. What began as a Portuguese-English phrase book became a beloved source of comedy. It gave some much-needed comfort to Abraham Lincoln and William Seward during the trying days of the Civil War. Here are some unfamiliar “Familiar Phrases” from that book.
- Apply you at the study during that you are young.
- Dress your hairs.
- Sing an area.
- These apricots and these peaches make me and to come water in mouth.
- To craunch a marmoset.
- Burn the politeness.
Among the several manifestations of L1 transfer, Casey’s favorite appears to be lexical interference. This form of transfer occurs when people try to use an L2 word as if it were completely synonymous with a word in their native language. So, for example, Engish-speaking students might know that in German “Brief” is a letter (a piece of written correspondence), and incorrectly think it’s also the correct word for a letter of the alphabet. Their experience with English (L1) gives them a false expectation of semantic range.
The limits of transfer as an explanation for errors
When the scholarly tide in linguistics turned against behaviorism, some researchers rejected transfer as the primary source of L2 errors. They threw out transfer baby with the behaviorism bathwater.
According to the contrastive analysis hypothesis (CAH), errors were assumed to be the result of transfer from learners’ first language. Detailed analysis of learners’ errors revealed, however, that not all errors made by second language learners can be explained in terms of first language transfer alone. A number of studies show that many errors can be explained better in terms of learners’ developing knowledge of the structure of the target language rather than an attempt to transfer patterns of their first language (Richards 1974). Furthermore, some of the errors are remarkably similar to those made by young first language learners, for example, the use of a regular -ed past tense ending on an irregular verb.
Lightbown, Patsy M.; Spada, Nina (2013-03-21). How Languages are Learned 4th edition (Oxford Handbooks for Language Teachers) (Kindle Locations 1047-1053). Oxford University Press, USA. Kindle Edition. (Emphasis mine.)
One reason that some researchers rejected the hypothesis that ‘transfer’ or ‘interference’ would best explain a learner’s difficulties with the target language was the fact that contrastive analysis was closely associated with behaviourist views of language acquisition. In rejecting behaviourism, some researchers also discarded contrastive analysis. In doing so, they potentially lost an essential source of information about language acquisition.
Lightbown, Patsy M.; Spada, Nina (2013-03-21). How Languages are Learned 4th edition (Oxford Handbooks for Language Teachers) (Kindle Locations 1381-1384). Oxford University Press, USA. Kindle Edition.
Hence, we can see that transfer remains a viable option for explaining L2 errors, but it’s far from the only one.
On the other hand, Casey appears to be unaware of any of the research around language acquisition. Once again he has found an “explainer.” He discovered the term, thought that it fit nicely, and stopped looking. For if he really knew anything at all about the subject and cared about how (or whether) it bolsters his case, he wouldn’t simply invoke it as an incantation. Presto! Interference! Ephphatha!
Translated or composed?
Despite Casey’s insistence, I can find no linguistic research that would help us determine whether a Koine Greek document from the first century CE assuredly contains L1 transfer errors that can best be explained by translation from a source Aramaic document. In order to prove his thesis, Casey would have to demonstrate that a lexical item in Mark that looks like nonstandard Greek is:
- Not idiomatic Greek used in Palestine by Jews of the period.
- Not language that mimics the Septuagint.
- Not an error rooted in something other than L1 transfer.
- Best explained by transfer during translation and not during composition. (How would we ever know the difference?)
Casey likes to raise the specter of “double interference.” Hocus-pocus! Abracadabra! Talitha cum(i)! As with “single” interference, he does not fully explain what he means by this term. Casey seems to be saying that a writer would see the written Aramaic or Hebrew in a source document, which “interferes” with his translation of the target Greek, and he would also experience L1 transfer errors in his head that “doubly interfere” with his attempts to translate the document into good Greek. He writes:
There have also been a number of detailed studies of particular translators and of particular words and constructions. For example, in a detailed study of the translation of כי (ki) with ὄτι (hoti), Aejmelaeus showed that it is often used incorrectly by the standards of monoglot Greek speakers precisely because it is so often used correctly.192 This set up too close an association between the two words in the minds of translators who were suffering the double level of interference which is inevitable when translators translate texts. (Aramaic Sources of Mark’s Gospel, p. 56., bold emphasis mine)
That footnote (192) above refers to an essay by Anneli Aejmelaeus called “Oti [Hoti] Causale in Septuagintal Greek,” which you can find in On the Trail of the Septuagint Translators: Collected Essays. The author considers the problem of the Hebrew כי (ki), which has an “exceptionally wide range of usage in the most varied contexts and functions.” The LXX translators often applied (transferred) this wide range incorrectly and used ὄτι (hoti) when γάρ (gar) would have been more appropriate.
So we have what indeed looks like classic Casey-esque interference. In the minds of the translators, the range of the L1 word interfered with their understanding of L2. If you only skimmed the essay, you might think Casey is spot on.
What Casey neglects to tell us, however, is that Aejmelaeus discussed the improper use of ὄτι (hoti) not only in translation but also in composition. We must pay special attention here, because it gets to the heart of the matter. Aejmelaeus focuses on Luke and Paul as NT authors who used a causal ὄτι (hoti) not just in their quotations from the LXX, but in their narrative prose as well.
As for Luke, he had obviously already found frequent use of ὄτι [hoti] causale in the tradition. But, comparison with the synoptic gospels shows that Luke even made independent use of the Septuagintal type of ὄτι [hoti]. Exactly the same observation has been made in connection with other Hebraistic elements in Lk. In Acts the use of ὄτι [hoti] causale — both classical and Hebraistic — is concentrated mainly in the first half of the book.
Indeed, it has been maintained that Luke deliberately employs Semitizing expressions, particularly in the first half of Acts, with the intention of imitating Septuagintal Greek. On the other hand, he has the reputation of being a competent user of Greek. The use of ὄτι [hoti] causale may obviously be regarded as one of the features attributable to a conscious effort on his part to compose “biblical Greek”. The fact that in Acts it mostly occurs in the words spoken by the Apostles or by the Lord — and only seldom in the narrative parts — supports this view. A skilful imitator could even utilize a small feature like this to give solemn, Septuagintal colouring to his language. (p. 27, formatting and bold emphasis mine)
We’ve gone beyond interference here. New Testament writers clearly employed certain verbal constructions and deliberately chose certain words while composing, based on a particular style of Greek that they associated with holy writing. In fact, what Aejmelaeus has described is the opposite of what Casey is looking for. But he goes even further.
As far as these two examples, Paul and Luke, are concerned, ὄτι [hoti] causale in place of γάρ (gar) is a clear Septuagintism. Paul probably employed it unconsciously, being influenced by the Scriptures. The frequent reading of the Septuagint had an effect on his sense of correct linguistic usage. If this could happen to Paul, it obviously could happen to many other readers of the Septuagint.
It was possible that a translational Greek feature did not remain a peculiarity of translations only, but became part of “biblical” Greek and was further employed by Greek writers. In the Septuagintal use of ὄτι [hoti] causale we have evidence of the fact that the Septuagint could influence not only the religious vocabulary but even the structure of the language used by its readers. . . . In a limited sense one could even argue that ὄτι [hoti] causale in place of γάρ (gar) is Jewish Greek! (p. 27, formatting and bold emphasis mine)
Casey believes that improper Greek construction that looks like Aramaic construction is a sure sign that the author is confused by Aramaic text in a document that he is translating. That first-century author would, he presumes, suffer from the “double” interference introduced by his native understanding of Aramaic (his “L1”). However, as we see above with the usage of Hebraicisms (or Septuagintisms or perhaps “Jewish Greek“) in the NT, several factors will always hinder our ability to prove Casey’s thesis. We will always lean toward proven models of behavior against Casey’s unsubstantiated claims.
Moule on “Semitisms”
C.F.D. Moule never seems to have warmed up to the “ugly and rather jargonistic” word, “Semitism,” but he nonetheless devoted an entire chapter to the subject in An Idiom-Book of New Testament Greek. Concerning the presence of Aramaisms, Moule acknowledged the possible ramifications.
A demonstrable Aramaism  (as distinct from a Hebraism) may point to contact with some source from primitive Palestinian Christianity or go back to the actual words of Jesus; a proved Hebraism  will point only to contact, direct or secondary, with the Scriptures; while a Semitism  such as might be either Aramaic or Hebrew will be correspondingly vague in evidential value. (p. 172, emphasis mine)
The key word, of course is may, since it would only leave open the possibility. We need strict and verifiable ways to identify an unusual usage of Greek that can only be explained by the influence of Aramaic. However, I must repeat: An Aramaism in the text may also occur because of L1 transfer during composition and not translation.
Even if we believe we have discovered a true Aramaism, we have to consider something else — namely, the way in which the inhabitants of Palestine received Koine Greek in the Hellenistic period. Quoting Moule again:
It is not always possible to determine where to draw the line between a clear, alien ‘Semitism’ and a term or idiom which is indeed reminiscent of a characteristically Semitic equivalent but which is none the less good or tolerable Greek, and which may, therefore, owe little or nothing to Semitic influence. Sometimes it is only the frequency of its occurrence, and not its actual existence, that a term or phrase owes to its alien influence.
Obviously, too, this problem is complicated by the question of how far the generally understood, secular κοινή had unconsciously absorbed and, so to speak, naturalized what were originally alien elements from Semitic populations. (p. 171, emphasis mine)
In my day job I’ve often dealt with customers and coworkers from India, and at first I would sometimes hear an odd turn of phrase and assume it was just an idiosyncratic quirk. But then I would hear different people on the same team use the same expression, causing me to adjust my theory. Finally, it became clear that I wasn’t hearing an “error” in English, but instead a normal Indian English idiom that was hitherto unfamiliar to me.
If I receive an email from an Indian friend, and it contains some English construction that sounds “Indian” to me, it would be wrong for me to leap to the conclusion that he is translating Hindi or Urdu from a written document. He might just be experiencing negative L1 transfer during composition.
But it’s even more likely that he’s simply using English in a way that’s completely ordinary among the people he lives and works with. We shouldn’t be surprised — only dead languages are static. When a lingua franca takes hold in a society, it necessarily changes to meet the needs of the people who use it.
Case(y) study — Mark 1:7
Let’s take a look at Casey’s method in action. In Jesus of Nazareth, Casey describes an example “unidiomatic” Greek in the Gospel of Mark.
As well as simple mistakes, there are cases where Mark’s Greek is unidiomatic, because Mark was suffering from interference from an Aramaic source in front of him, as translators often do. For example, Mark transmits John the Baptist’s prediction of his successor in a form which may be literally translated as follows:
The one stronger than me is coming after me, of whom I am not worthy to bend down and undo the latchet of the sandals of him.
Here the meaning is unambiguous and correct, but ‘of whom’ followed by ‘of him’ is no more idiomatic in Greek than it is in English. The relative particle in Aramaic (de or di), the equivalent of the English ‘who’, ‘what’, etc., has to be picked up by another particle later in the sentence, in this case delēh, the equivalent of ‘of him’. In Greek, this is not the case, just as in English, and once Mark had correctly put ‘of whom’, there was no need for him to put ‘of him’ at the end of the sentence. He did so because he had delēh at the end of the sentence in the text in front of him. This is very valuable evidence, because it shows that Mark was translating a written Aramaic source. (p. 112, bold emphasis mine)
First, we need to establish whether Casey is correct in his declaration that the redundant, trailing αὐτοῦ (autou) is not the sort of Greek that a native speaker would ever use. I found a tantalizing but brief discussion about this verse in A Translator’s Handbook on the Gospel of Mark:
[T]his use of the relative hos ‘who’ and the pronoun autos ‘he’, both in the genitive case, is a construction which reflects Semitic influence (Black Aramaic, 75: “clear proof of its origin”), although, as Lagrange points out, it is not unknown in Greek itself. (p. 22, emphasis mine)
After some digging, I found the book by M.-L. Lagrange the authors were referring to. On page 8 of Évangile selon saint Marc he writes:
La tournure οὗ avec αὐτοῦ à la fin, est sémitique, mais c’est une négligence qui n’est pas tout à fait inconnue au grec classique (Bl.-Debr. §297). Cette tournure se retrouve abrégée dans Lc. iii, 16.
Or in English (roughly):
The turn of phrase οὗ (hou) with αὐτοῦ (autou) at the end is Semitic, but to say it was entirely unknown in classical Greek would be an act of negligence (Blass and Debrunner, §297). This turn of phrase is found in an abbreviated from in Luke 3:16. (p. 8, emphasis mine)
Wait a minute. I own a copy Blass and Debrunner’s A Greek Grammar of the New Testament. Have I been on a wild goose chase? Sure enough, at section sign 297 we find:
The pleonastic personal pronoun incorporated into the relative clause is a phenomenon especially suggested by Semitic usage (Hebrew אֳשֶׁר … לוֹ [asher . . . lo]; similarly דְּ ,דִּי [de or di, see Casey above]; . . .), but it is a slip not unknown in classical and later Greek. (p. 155, emphasis mine)
Among the many verses Blass and Debrunner cite as examples of the pleonastic (i.e., redundant) personal pronoun the very first is Mark 1:7.
So how do we score Casey? We said that for his argument to have any merit he first needed to demonstrate that the supposed unidiomatic phrase was unknown in “normal” Greek. Unfortunately, we find that Casey is wrong about that. We can point to scholars who know a great deal more about Greek than Casey who say it was “not unknown.” Fail.
Suppose we gave him a pass on the first point. Let’s say that perhaps Matthew Black (the author of An Aramaic Approach to the Gospels and Acts) and Casey were right, and it was unknown in Greek. Can we then determine that it is an Aramaism and not a Hebraism? No. All the sources agree that while we could call it a Semitism, it impossible to determine that it is specifically an Aramaism, meaning that it is “vague in evidential value” (see Moule above). Fail.
Suppose we gave him a pass on the first and second points. Even if it were not known in Greek (but it is) and even if we knew it was an Aramaism (but we don’t), could we prove that Mark was translating and not composing in Mark 1:7? No, we cannot. Casey has already argued persuasively that Mark’s first language was Aramaic. Hence, he would have “interference” from his first language if he was composing Greek or translating text into Greek. We have no way of telling the difference. Fail.
Finally, as we saw above, supposed Semitisms like this one may have crept into normal Jewish Greek in Palestine, or they may have gained a kind of holy aura — writing that sounded biblical, which even supposedly good Greek writers such as Luke tried to imitate. Therefore, contra Casey, Mark 1:7 is not “very valuable evidence,” because it proves absolutely nothing.
In this post we looked at the heart of Maurice Casey’s argument for the historicity of Jesus. We have weighed it in the balance and found it wanting. Despite his self-professed unmatched expertise in the matter, we have found that some of his supposed Aramaisms can only be safely termed Semitisms. Moreover, these Semitisms may have actually found their way into “biblical Greek,” which authors deliberately imitated to add authenticity and solemnity.
We found that his references to linguistic interference merely serve to demonstrate that he does not understand the term and is unfamiliar with the research. He maintains that Mark’s unidiomatic Greek can only be the result of interference during translation but he fails to explain why it cannot be interference during composition. He also fails to account for and eliminate other possible sources for Mark’s errors, if indeed they are errors.
Casey has been found wanting in every phase of his argument. He does not fully explain his methodology. He applies that methodology, such as it is, haphazardly and with no regard for alternate possibilities. He does not demonstrate; he asserts.
I would never claim that the study of Aramaic has no place in NT studies. Nor do I argue that it is impossible to find Aramaisms in the NT. My point in this (regrettably long-winded) post is that Casey has utterly failed to deliver. His method is muddled. His execution is faulty. He insists he has found proof of written Aramaic documents behind Mark and Q. He believes he has discovered some of the authentic words of Jesus. He has in fact proved nothing.
Verdict: Do not want.
If you enjoyed this post, please consider donating to Vridar. Thanks!