バイリンガリズムと新約聖書ギリシャ語の特徴（by モイセス・シルヴァ）〔英文〕

モイセス・シルヴァ。１９４５年生まれ。ウェストミンスター神学校（１９８１－１９９６）、ゴードン・コーンウェル神学校（１９９６－２０００）で聖書学を教鞭。ＮＡＳＢ（新米国標準訳聖書）およびＥＳＶ（英語標準訳聖書）の翻訳委員。

Moisés Silva, "Bilingualism and the Character of New Testament Greek," Bib 61 (1980):198-219.

目次

Prologue（プロローグ）

In his justly famous article on 'Grec biblique' Jean Vergote dismissed as absurd Albert Thumb's view that the Greek of Jewish speakers was not significantly affected by their native speech.*1

Precisely because it is absurd, however, it may well be that these two authors are not talking about the same thing. In effect, I wish to argue that much of the contemporary debate on this issue suffers from the use of imprecise language, and, more specifically, that the failure to distinguish clearly between 'language' (langue) and 'speech' (parole) lies at the root of the disagreements.*2

I shall further seek to demonstrate that recent criticisms of so-called 'Deissmannism' are largely misdirected and have failed to overturn it.

1. Deissmann and His Critics（ダイスマンと彼の批評家たち）

The history of the controversy is well known and frequently alluded to in the literature.*3

We cannot, however, altogether dispense with a survey of the various positions advanced, and that for two reasons. In the first place, the views of Adolf Deissmann and others have received less than fair press; indeed, one suspects from time to time that contemporary writers may not always take the time to read the works of the scholars they criticize.

Thus, a striking statement by Deissmann may be quoted without any clear attempt to understand its significance in his total presentation. For example, the standard (and almost wearisome) characterization nowadays is that Deissmann made a major contribution to the field but that he took his views to an extreme and so they need considerable modification; yet when concrete instances of modifications are given, they often turn out to be items that had been readily admitted by Deissmann himself.

In the second place, the viewpoints offered by the scholars involved are not limited to a single, well-defined issue but rather include a good number of separate questions. We may list the most important of these as follows.

Regarding the Koine in general:
1. The place of the Koine in the history of the Greek language;
2. The possibility of dialectal differentiation within the Koine.

Regarding the Koine in Alexandria:
3. The nature of the Greek spoken by native Egyptians;
4. The nature of the Greek spoken by Alexandrian Jews;
5. The nature of 'Septuagintal' Greek (insofar as it is a unity);
6. The relation between 'Septuagintal' Greek and the language
of Alexandrian Jews.

Regarding the Koine in Palestine:
7. The general linguistic situation in Palestine (more specifically:
which language did Jesus speak?);
8. The nature of the Greek spoken by Palestinian Jews;
9. The nature of NT Greek (insofar as it is a unity);
10. The relation between NT Greek and the language of Palestinian
Jews (including the possibility of 'translation Greek');

11. The influence of the living Aramaic (or Hebrew) substratum on the NT writers;
12. The influence of the Septuagint on the NT writers.

These and other questions need not of course be dealt with one by one in mechanical fashion, yet we must perceive them as distinct issues.

When a scholar advances an opinion, for example, on the character of Alexandrian Jewish Greek, one cannot tacitly assume that the same opinion holds for Palestinian Greek.

To begin with, let it be openly admitted that Deissmann's earlier formulations, framed as they were in the excitement of discovery, overstated the facts. Deissmann himself admitted as much.*4

We should therefore move to his later expositions. In one of the clearest, he argues against the view that biblical Greek 'must be sharply distinguished from the rest of what people have been so fond of calling "profane Greek"'. *5

He complains that before the study of the papyri 'we had greatly over-estimated the number of Hebraisms and Aramaisms in the Bible', but that 'not one of the recent investigators has dreamt of denying the existence of Semiticisms'. *6 The exposition ends with this paragraph:

"What we do deny is merely this: that the Semiticisms, particularly those of the New Testament, are sufficient reason for scholars to isolate the language of the sacred texts. Our opinion of the biblical language is reached by considering its innumerable coincidences with the cosmopolitan language, not its numerable differences from it. The Semiticisms do not place the Bible outside the scope of Greek philology; they are merely birthmarks. They show us that in this great cosmopolitan Book the Greek cosmopolitan language was spoken by men whose home lay in the East." *7

The italicized words in these quotations make Deissmann's position crystal-clear: while certain peculiarities of NT Greek give it away as having been written by Semites, they are not so many that the language should be 'sharply distinguished' from non-biblical Greek or 'isolated' from the normal tasks of Greek philology. We may also note that Moulton summarized both his own views and those of Thumb by admitting freely that some types of dialectal differences must have existed in the Koine, with the qualification that

"writings like the Greek Bible, intended for general circulation, employed a Durchschnittsprache which avoided local peculiarities. . . For nearly all the purposes of our own special study, Hellenistic Greek may be regarded as a unity, hardly varying except with the education of the writer, his tendency to use or ignore specialities of literary language, and the degree of his dependence upon foreign originals. "*8

The point to notice in this quotation is Moulton's emphasis on the unity of the language, not in some absolute sense, but for the specific purposes of grammatical description.

The opposition to 'Deissmannism' took different forms.*9Some scholars argued, for example, that the language of the Egyptian papyri might itself be semiticized as a result of the large Jewish population in Alexandria. Vergote, following the lead of L.-T. Lefort, admitted the improbability of this argument and argued instead that the peculiarities of the papyri were due to the native Egyptian language.

Since the Hamide and Semitic languages are closely related, these peculiarities should be construed as 'Copticisms' analogous to the Semitisms of biblical Greek. Vergote subsequently devoted his professional interests to the study of Coptic itself rather than to the character of 

NT Greek, but Francis T. Gignac, who is publishing A Grammar of the Greek Papyri of the Roman and Byzantine Periods*10 uses the same approach to the problem. It should be noted, however, that many of Deissmann's examples were not restricted to the Egyptian papyri but also came from inscriptions, not particularly those in Asia Minor.

Furthermore, many of Vergote's and Gignac's Coptic examples come by necessity from biblical texts, that is, from documents that are translations of the very language for which supposedly independent data are being adduced; this obstacle of course does not completely invalidate the argument, but it certainly restricts its application.

Another line of opposition has come from Scandinavian scholars. One thinks especially of Albert Wifstrand, who argued that, although 'we cannot discover any special Greek dialect spoken by the hellenized Jews', nevertheless 'the stylistic home' of the NT writers was the
synagogue.*11

A third approach, and the one which has stirred most of the recent controversy, is that of Nigel Turner. In one of his earlier contributions *12Turner stated that in spite of the impressive evidence brought forth by Deissmann and Moulton, 'we may still legitimately feel that in numerous uses an Hebraic idiom has popularized and extended one which was already fairly familiar in Greek'. This is an exceedingly curious statement, for the point is precisely the one made by Moulton when he argued that 'the ordinary Greek speech or writing of men whose native language was Semitic. ..brought into prominence locutions, correct enough as Greek, but which would have remained in comparatively rare use but for the accident of their answering to Hebrew or Aramaic phrases'. *13

Turner appears to be saying nothing more than Moulton did, yet he inexplicably opposes his viewpoint to that of his predecessor. Turner entitled another article, 'The Unique Character of Biblical Greek',*14 clearly indicating that he wished to go beyond Deissmannism.

His best piece of evidence is the fact that biblical Greek prefers overwhelmingly the patterns πάς άνθρωπος and πας ό άνθρωπος, whereas the papyri prefer ό άνθρωπος πάς and  ό πας άνθρωπος. (We should note that, in contrast to Vergote's approach, Turner assumes that the papyri may indeed be used as evidence of the general Koine.)

Turner's own charts, however, show that the former constructions are not at all rare (let alone nonexistent) in the papyri! In other words, even if we accept Turner's analysis, we still have not moved in substance from Moulton's position.

Turner's later and comprehensive work on syntax*15offers evidence which is not really different in character. One gathers, instead, that Turner has been impressed by the amount of evidence; that is, his exposure to the papyri has persuaded him of what he considers a major or even radical difference between their language and that of the NT. Now if our evaluation is accurate, if Turner's judgment has been moulded by the total impression which NT Greek makes, then he did not need to go beyond Wifstrand's position referred to earlier.

For some reason, however, he felt he must confront Deissmannism head-on. Thus, with regard to the NT writers' use of prepositions he argues that the standards must be looked at from 'outside the sphere of classical Greek, even outside secular Greek altogether, although the living Koine must be kept in mind always'.*16Re-opening the possibility that biblical Greek reflects a spoken Jewish Greek, he suggests that perhaps it is not all that bad to speak of a 'Holy Ghost language'.

'We now have to concede that not only is the subject matter of the Scriptures unique but so also is the language in which they came to be written or translated.'*17

Matthew Black has welcomed Turner's new evaluation, commenting (perhaps with deliberate allusion to Deissmann's concern not to isolate NT Greek) that the language of the Greek-speaking synagogue, 'like the Hebrew of the Old Testament which moulded it, was a language apart from the beginning; biblical Greek is a peculiar language, the language of a peculiar people.''*18

2. The Concept of Dialect（方言の概念）

Perhaps the biggest obstacle in the way of resolving our problem is the freedom with which the word 'dialect' (or related expressions like 'unique language', etc.) is used by the various parties. To be sure, even among professional dialectologists the term is somewhat ambiguous (though only in the sense that no hard and fast criteria have been established for distinguishing between a 'dialect' and a 'language').

Still, it makes very little sense to affirm or deny or even question whether NT Greek represents a dialect before the parties involved define precisely what meaning they are attributing to the word. As things stand, some writers appear to use the term in the vaguest possible way, based on the subjective impression made by the data; to the extent that this is true, one may be forgiven for responding, de gustibus non disputandum.

We may begin by referring to Thumb's discussion. Thumb himself does not give us a concise definition of 'dialect', but his extensive treatment makes plain that he uses the term in the established sense it has in classical philology. When classical scholars speak of ancient Greek dialects, the meaning is fixed by the kind of material differentiating Ionic from Attic, both of them from Doric, and so on.

By and large, however, the data come from the areas of phonology, morphology, those aspects of syntax that are most closely linked to morphology, and vocabulary (the last of which, however, does not normally include specialized terms, but only the presence of different words with common meanings).

Thus, the primary criteria differentiating Ionic-Attic from other groups are such features as the shift of α to η, the disappearance of p, etc.; again, northwest dialects are characterized by such conservative features as -μες (Attic μεν), -οντι (Attic-ουσι), etc. Now even a cursory examination of Thumb's argument shows that these are the kinds of factors that inform his view of dialect. We can hardly be surprised, therefore, not only that he denied dialectal differentiations in the Koine, but that the leading classical philologists support him*19

All of this means that when Turner re-opens the question, he is either rejecting the established conclusions of classical scholarship (in which case we may fairly ask for an extensive, scientific refutation of those conclusions) or, what seems most likely, he is using his terms in a different way. Could it be, for example, that Turner understands 'dialect' as that term is used in modern linguistics?

Dialectologists use the term when describing very small phonological variations: even the two pronunciations of caught will serve to distinguish American dialects of English. But Turner surely does not mean that, since it would lead him to speak of the 'uniqueness' of, say, Philadelphia English; besides, Thumb never denied the existence of such differentiations.*20

In short, we find ourselves at an impasse because of the failure of writers generally (not Turner alone) to define the very terms which stand at the center of the debate. Perhaps some further considerations will help to resolve this dilemma.

3. Bilingualism — Some General Principles（バイリンガリズムーいくつかの一般原則）

Vergote's article on 'Grec biblique', to a large extent a harsh criticism of Thumb and Deissmann for failing to understand the nature of bilingualism, offers modern linguistics as something of a panacea for our problem.*21

When Vergote wrote these words, however, 'modern linguistics' was in its infancy, and bilingualism in particular had not been subjected to systematic and scientific examination.

In the 1950s several scholars, principally Einar Haugen and Uriel Weinreich, gave a strong impulse to the study of bilingualism, and the more recent emphases on psycholinguistics and sociolinguistics have contributed enormously to our understanding of this phenomenon. It is all the more remarkable, then, that this field of research, although directly relevant to the question of Palestinian Greek, has remained practically untouched in the contemporary discussion. Instead, the very frequent comments made regarding bilingualism are usually based on haphazard personal observations or even on purely speculative assumptions.

There can be no question, in this article, of treating the field in comprehensive fashion; at any rate, the present writer can hardly pass for an authority on the subject. Still, it is not necessary to dig too deeply into the literature in order to discover some basic principles that can shed considerable light on the discussion.

First of all, it is essential to recognize the complexity of the field. One cannot afford merely to observe the language of, say, Portuguese immigrants in the US and assume that the same situation, mutatis mutandis, obtained for Palestinian Jews. For one thing, English and
Portuguese are genetically related, whereas Greek and Aramaic belong to different families. Secondly, the native language of a bilingual (in our illustration Portuguese or Aramaic) is not affected in the same way as his or her second language (English or Greek).

Thirdly, the radically different sociological milieus (immigrants faced with the pressures of the dominant language versus speakers who remain in their homeland) will also affect the bilingual's linguistic behavior. Fourthly, one must allow for various levels of competence within each group. These are only four out of a large number of distinguishing features which characterize various forms of bilingualism.*22

Perhaps the most important distinction, for our purposes, is whether we are dealing with the mother tongue of the bilingual or with the language he subsequently learns. The importance of this distinction rests on the observation (which goes back to the last century) that

"it is the language of the learner that is influenced, not the language he learns. English is hardly influenced at all by the immigrant languages, but these are all influenced by English; in Latin America the Indian, languages acquire material from Spanish, but the Spanish shows very little influence from Indian. The reason for this is that the social pressure in such cases is all in one direction, because of the difference in prestige of the speakers of the two languages." *23

On the basis of that observation, Thumb and Deissmann could have argued that Palestinian Aramaic and Hebrew were more likely influenced by Greek than vice versa;*24indeed, the strong influence of Greek on Syriac and on Coptic is admitted by all. It is interesting that Haugen speaks of English as 'hardly influenced' by immigrant languages in spite of the many loanwords and foreign phrases present in English; no doubt he recognized that these isolated elements do not materially affect the structure of the language.

Similarly, Deissmann readily granted the presence of Latinisms and Semitisms in Greek, but in his view these elements simply constituted 'booty' captured by the conqueror.*25

However, Haugen warns us that what may be true for the language in general does not necessarily hold for the individual bilingual.

"Those learners with whom we are most familiar in our foreign language classes or even adult immigrants do maltreat the language they learn. In their case there is bilateral influence between the languages. But the innovations they make in the language they learn do not spread to the native speakers ofthat language, while the innovations they make in their own language do spread?"*26

One can hardly exaggerate the significance of this last observation for our discussion. Without saying so explicitly, Haugen is in fact calling our attention to the well-known Saussurean distinction between langue and parole: the former refers to the (abstracted) linguistic system in the consciousness of a community, whereas the latter designates the actual speech utterances of individual speakers.*27

A slip of the tongue, for example, is part of parole, not of langue. Similarly, an individual will make mistakes (parole) when speaking a foreign language due to 'interference' from his native tongue, but these mistakes are not regularized, do not become part of the system (langue).

We could suggest, then, that whereas Thumb's views referred to langue, they appeared absurd to Vergote only because the latter was
concerned with parole. We may even argue that the scholars concerned were at least partly conscious of this distinction. In his article for Herzog-Hauck's Realen-cyclopädie (written before the publication of Saussure's seminal work) Deissmann drew on H. Paul's distinction between 'usuelle' and 'momentane Anomalien' to argue that the syntactical Semitisms of the Septuagint were occasional rather than usual (therefore, not part of the system *28

For his part, Vergote emphasized the 'caractère individuel de la langue des bilingues'.*29Although it would be foolhardy to suggest that the differences between the two parties are only a matter of semantics, perhaps we can expect some progress if we recognize that the discussion may be taking place at two distinct levels.

4. Alexandrian Bilingualism（アレクサンドリア地方のバイリンガリズム）

source

It should be clear from previous remarks that the bilingual situation in Palestine cannot be simply identified with the situation in Alexandria. The Jews in Alexandria were immigrants; furthermore, Greek was the dominant language in a sense that it certainly was not in Palestine. What can we say about the Greek of Alexandrian Jews?

Henry S. Gehman wrote an article in 1951 which has been used by Turner in support of his thesis. Entitled 'The Hebraic Character of Septuagint Greek', *30it sought to show that 'we can hardly avoid speaking of a Jewish-Greek, which was in use in the synagogues and in religious circles'.*31 Surveying a number of well-known syntactical and lexical Semitisms in the LXX, Gehman argues that 'it is often difficult to obtain the sense without comparing the Hebrew text'. *32

But since 'the LXX must have been read in most instances by itself and not by making continual references to the Hebrew', it follows that the LXX reflects the language of Greek-speaking Jews. *33 In support of this statement he says: 

"In a bilingual area a few individuals may speak both tongues perfectly, but the masses do not keep the idioms of the two apart, as may be abundantly observed in linguistic islands in this country. There is always a difficulty in passing from one language to another; in the transitional period a generation has a smattering of the tongue of the forefathers without having become thoroughly immersed in the new vernacular."*34

The following criticisms may be offered,

(a) Gehman, while rejecting the expression 'Jewish-Greek jargon' (a term which he does not define), speaks of 'a Jewish Greek which was understood apart from the Hebrew language'.*35 But this terminology is so vague that, in my judgment, it serves no useful purpose.

(b) The author throughout the article refers to 'LXX Greek' as though this were a well-defined entity. Now Gehman, himself one of our leading LXX scholars, was hardly ignorant of the wide divergences in translation technique—and therefore in the character of the resulting language—among the books in the Greek OT. Nevertheless he can jump in the same paragraph from Genesis 4 to 1 Kings 1, that is, from a fairly respectable Koine style to the very literalistic 'kaige section' of Samuel-Kings. Of one thing we may be sure: if 'LXX Greek' reflected the spoken Jewish Greek of Alexandria, we will need to specify which LXX style we are referring to.

(c) Gehman seems to argue that the strange idioms of the LXX could only be understood either by someone who referred to the Hebrew or by someone familiar with a spoken Jewish Greek. The truth is, however, that many of his examples, such as the use of the infinitive with a preposition, can be understood even if translated quite literally into English. In fact, Gehman himself (inconsistently) shows us how a Greek reader could understand Jhe subordinate use of καί without either checking the Hebrew or being familiar with some Jewish Greek: 'Even though a Hellenistic Jew [but a Gentile as well!] would not know Hebrew or Aramaic, it is probable that for the most part the context would lead him to the correct interpretation of καί in passages of this nature'. *36

As for the really difficult passages in some LXX books, we may readily concede that a Greek speaker would not have understood them, but sometimes we cannot be sure that the translator himself understood what he wrote!*37

(d) Although Gehman appropriately uses the bilingualism of US immigrants as a parallel to Alexandrian bilingualism, his actual description is completely unacceptable. I have seen no evidence in the literature of a 'transitional period' during which a generation of immigrants are less than competent in both their native tongue and the foreign language.*38

My first-hand contact with Cuban immigrants in Florida, for example, suggests a picture quite different from Gehman's. With rare exceptions, Cubans who migrated during their childhood (and children born to Cuban families after migrating) have, completely mastered the English language: only minor phonetic difficulties and a few turns of phrase could ever give some of them away.

Although when speaking with each other in English they may introduce words or phrases peculiar to Cuban culture, this is done quite deliberately and usually for humorous purposes. On the other hand, their Spanish is contaminated at all levels by the influence of English, even to the extent of using such a 'loanblend' as loquear (< 'to lock').*39

And we should emphasize that this has taken place in a matter of fifteen to twenty years—no time for a transitional generation of the kind envisioned by Gehman. Of course, not all groups develop linguistically in the same way, but unfortunately it is difficult to obtain reliable information, since most studies focus on the changes that take place in the native language, not in English (precisely because English does not change significantly).

One valuable exception is a study of 'Spanish-English Bilingualism in San Antonio, Texas', by Janet B. Sawyer.*40

It appears from her description that Mexican-Americans in San Antonio have been much slower than Cubans in Florida to adopt English culture and language; perhaps their situation is closer to that of Alexandrian Jews, who surely sought to preserve their identity as much as possible. Interestingly, Sawyer tackles the very question whether these Mexican Americans speak a dialect of English. In her opinion, to merit the term 'dialect',

"a particular variety of language should be fairly stable in its structure so that it can be learned by succeeding generations in the speech community. Nothing that could be called a Mexican-American dialect of English was found in San Antonio, Texas. The English spoken by the bilingual informants was simply an imperfect state in the mastery of English."

She continues:

"What does have significance is the fact that the relatively unskilled bilinguals.. .did not pass on their imperfect English to their children. . It was clear that the linguistic norm was not the English of their relatives or neighbors, but rather that of the members of the prestige, English-speaking community. From generation to generation, the second language was in a fluid state, becoming more and more expert."*41

I am not aware of any study that, in opposition to Sawyer's, has established the existence of an English 'dialect' among immigrant groups. In other words, the possibility of a Jewish-Greek dialect (or whatever we care to call some unified and stable speech form) in Alexandria appears to receive no support from modern research into bilingualism. Even apart from these considerations, however, it must be said that the very existence of the LXX militates against Gehman's position: the fact that Alexandrian Jews needed their Bible in Greek *42 is virtually conclusive proof that they had indeed 'become thoroughly
immersed in the new vernacular'.*43

5. Palestinian Bilingualism（パレスティナ地方のバイリンガリズム）

As we move to a consideration of the linguistic situation in Palestine we must recognize not only that it should be clearly distinguished from that in Alexandria, but also that we have fewer scientific treatments of this type of bilingualism.

To be sure, modern linguistic research has frequently examined bilinguals who remain in their homeland, but this is usually with a view to assessing the state of the native tongue, not that of a foreign lingua franca. We may, however, briefly note two modern situations roughly parallel to that of firstcentury Palestine.

In the fairly large northeastern region of Spain known as Catalonia nearly six million people speak Catalan even though Spanish remains the official language.*44

A.M. Badia-Margarit has paid attention to the effect of this situation on the Spanish of bilinguals and reports that 'with the superimposition of Spanish (the language of culture) on Catalan (the natural language), cultured Catalans cannot generally prevent a series of characteristic features of their natural language from appearing in their Spanish'.*45

We may deduce from these observations what was a priori likely, namely, that Palestinians could not prevent features of their Aramaic or Hebrew from interfering in their Greek. However, two qualifications are necessary.

First of all, the extremely close relationship of Spanish and Catalan presumably facilitates confusion; thus, for example, a Catalan speaker might unconsciously modify in some way the Spanish cuando vuelvas ('when you return') under the influence of his native quan tornaràs, whereas such structural similarities are not to be found between Greek and Semitic languages.

Secondly, and more important, Badia-Margarit reports that Catalan speakers do avoid interference if they hesitate: the structures of the languages do not become fused.*46

In other words, the influence is purely at the level of parole and, we might add, would not likely manifest itself in the written form. A closer parallel to the situation in Palestine is what we find today in Wales.

Although Welsh and English are of course related, the separate evolution of these languages has resulted in drastic differences at all linguistic levels; further, in contrast to Catalan (which has from time to time suffered considerable political pressure), Welsh has enjoyed uninterrupted prestige and vitality.

It is interesting, therefore, that Moulton drew his illustrations precisely from the English spoken by native Welsh. He concluded that this English was hardly a dialect to be isolated—it simply contains a superabundance of features which are possible and comprehensible in English.*47

Although I have not found a modern scientific study that confirms Moulton's judgment, neither am I aware of any data that conflict with it. But now, precisely what was the linguistic situation in Palestine?

Against Moulton, we should recognize that the Palestinian population in general did not enjoy a Greek education comparable to Spanish education in Catalonia or English education in Wales. In other words, we may well assume that Semitic interference in the spoken Greek of Palestine was much more noticeable than Moulton supposed *48

On the other hand, we must remind ourselves of the massive evidence, much of it unavailable to Moulton, confirming the vitality of Greek in Palestine. J.N. Sevenster, who maintains that the difference in the use of Greek between Diaspora and Palestinian Jews was relatively minor, has put it most strongly:

It has now been clearly demonstrated that a knowledge of Greek was in no way restricted to the upper circles, which were permeated with Hellenistic culture, but was to be found in all circles of Jewish society, and certainly in places bordering on regions where much Greek was spoken, e.g. Galilee.*49

This view has received support from the evidence that even as early as the third century BC Greek was widespread in Palestine *50

Now it is true that Turner supports the possibility that Jesus might have done much of his teaching in Greek,*51but he fails to appreciate that this argument damages his position. The greater the use of Greek in Palestine, the greater the evidence of Hellenistic influence among the Jews and the lesser the likelihood that they failed to master the common language (as opposed to speaking some 'hybrid' form).

5. Parole and Style（パロールと様式）

One of the distinctive developments in the linguistics of the last two decades is its concern with the analysis of parole, in contrast with the almost exclusive preoccupation of earlier scholars with langue. Now since 'style' may be defined roughly as the variations (parole) that grammar (langue) leaves out,*52we are not surprised to note, also during the past two decades, new interests in the linguistic study of style.

It is at this level, I believe, that the discussion of 'biblical Greek' must take place. Much of the debate has, sometimes explicitly, assumed that the differences between the two parties can be resolved arithmetically, depending upon whether there is a large enough number of Semitisms.

But such a resolution could only take place (and even then with great difficulty) if the parties were dealing with the same linguistic phenomena.

I wish to argue that, in fact, they are dealing with two distinct levels of linguistic description. Deissmann, concerned with grammatical rules (langue), insisted rightly that NT Greek cannot be isolated from the Hellenistic form.

Turner, who has devoted his efforts to syntactical phenomena—an area of grammar that constantly 'infringes' on stylistics (parole)*53 —sees an undeniable distinctiveness in the Biblical language.

There is little originality about this suggestion that style may be the key to the problem. Thumb himself noted, without, however, perceiving its full implications, that the Semitic element in the Greek Bible 'äussert sich mehr im Stil und in der Denk- und Anschauungsweise als in der Sprache im engern Sinn'.*54More explicit is Wifstrand:

"We cannot discover any special Greek dialect spoken by the hellenized Jews; in phonology, accidence, syntax, word formation and many significations of words their language was ordinary koine;... but in phraseology, in the formation of sentences, in preferences when equivalent expressions were at hand, in all such things to which the authors of New Testament grammars give less attention, the real foundation is, to a great extent, the Hebraic and Aramaic mode of thought."*55

We may note especially the role of the LXX in this connection. Although it is quite proper and necessary to emphasize the great influence of the Greek OT on the NT writers, we should specify exactly where this influence manifests itself.*56

In particular, we need to remember that ancient literary documents seldom affect the linguistic habits of a community. A good example is the influence of the King James Version on the English language. That influence is evident in idioms, phrases and allusions, not in linguistic structure (whether grammatical or lexical); further, it is much more frequently found in formal speech, such as sermons, than in colloquial conversation.

C.S.ルイス

C S. Lewis argues that the impact of this translation is less than generally thought, as may be shown by the fact that we rarely use its characteristic features without the awareness that we are quoting it.*57

Particularly interesting in this regard is J. Trend's investigation of OT influence on the French language.*58

His work is 649 pages long and it seeks to show just how strongly medieval French writings give evidence of that influence. However, the bulk of the book (pp. 243-599) consists of a treatment of 'expressions'. Indeed, apart from some examples of Hebraisms which are not merely idioms but approximate syntactical adaptation (cf. pp. 600-49), he gives hardly any evidence of influence on the structure of the language.

Since by and large Trend's material consists of religious authors who deliberately imitated the biblical style, it is all the more surprising that his evidence deals almost exclusively with form and not with structure. Indeed, perhaps this rough distinction between form and structure uncovers the real nature of LXX influence on the NT. From a somewhat different perspective, David Tabachovitz has already made the same point:

"Es kommt noch hinzu—was im Prinzip zwar anerkannt, tatsächlich aber selten in Rechnung gezogen wird—dass die neutestamentliche Koine zum Teil aus der Septuaginta übernommen ist. Mit einer freilich groben Schematizierung könnte man die Sache auch so ausdrücken: das Wort, als isolierte Einheit betrachtet, ist im NT allgemein hellenistisch, der Stil aber ist durch das alttestamentliche Griechisch bedingt."*59

We may in conclusion seek specifically to answer the question: is it proper to speak of a special Jewish (or Christian) Greek?

N. Fernandez Marcos points out that, although the Stoics, for example, had their own specialized vocabulary, no one speaks of a 'Stoic language'. He goes on to suggest, however, that the number of more or less technical lexical items in the Greek of Christian writings may be reason to treat it as a special case.*60

In my opinion, the answer to our question depends on whether such a designation is used to oppose Deissmannism or whether it is used in the sense in which a contemporary linguist speaks of various styles in each community.

Of course there is a semiticized Greek style or a Christian Greek style or even a Christian English style. We may take the matter to its logical and valid conclusion and remind ourselves that there are even individual Pauline and Johannine styles (better, 'idiolects'). However,
such descriptions should in all fairness be dissociated from the kinds
of issues with which Thumb and his collaborators were concerned. They did their work and they did it responsibly and well.

*1:To anyone who is familiar with the phenomenon of bilingualism, Vergote claimed, 'il paraît absurde que Thumb, et d'autres avec lui, aient pu nier l'existence
d'une langue judéo-grecque parce que Josephe et Philon ont écrit un grec à peu près correct' (DBSup 3 [Paris, 1938], col. 1367). For Thumb's views, see Die griechische Sprache im Zeitalter des Hellenismus. Beiträge zur Geschichte und Beurteilung der Κοινή (Strassburg, 1901), pp. 125-26. 

*2:I hinted at this possible solution in 'New Lexical Semitisms?', ZNW 69(1978), pp. 253-57. Much of the research for this article goes back to my PhD thesis, Semantic Change and Semitic Influence in the Greek Bible (University of Manchester, 1972), portions of which are quoted here. 

*3:Perhaps the clearest brief surveys are by N. Turner, 'The Language of the NT', in Peake's Commentary on the Bible (ed. M. Black and H.H. Rowley; London,1962), pp. 659-62; and E.V. McKnight, 'Is the NT Written in "Holy Ghost"Greek?', BT 16 (1965), pp. 87-93.

*4:With particular reference to the Greek of translation literature he said: Ί myself have formerly been less reserved in expressing my opinion on this point than I should be now' (The Philology of the Greek Bible: Its Present and Future [London, 1908], p. 51).

*5:Deissmann, Philology of the Greek Bible, p. 44 (my emphasis). 

*6:Deissmann, Philology of the Greek Bible, pp. 62-63 (my emphasis).

*7:Deissmann, Philology of the Greek Bible, p. 65 (my emphasis). The main weakness in this formulation is the exclusive reference to numerical comparisons; he should have emphasized that the coincidences and the differences do not refer to the same type of material. 

*8:J.H. Moulton, Prolegomena, vol. 1 of A Grammar of NT Greek (Edinburgh, 3rd edn, 1908), p. 40. In addition to the writings of the scholars already mentioned, note F. Büchsei, 'Die griechische Sprache der Juden in der Zeit der Septuaginta und des Neuen Testaments', ZAW 60 (1944), pp. 132-49.

*9:Vergote has documented this in 'Grec biblique', cols. 1352ff. It should be noted, however, that he refers to a number of works that merely seek to refine
Deissmann's work. It is certainly inaccurate, for example, to suggest that A.T. Robertson was part of a 'réaction contre Deissmann-Thumb'. Similarly, we should note that although Ludwig Radermacher refers to NT Greek as (in some ways) 'eine Art von Judengriechisch', this statement is made after a careful distinction between grammatical and stylistic phenomena (Neutestamentliche Grammatik [HNT, 1; Tübingen, 1925], pp. 28-29).

*10:Volume 1 on phonology has appeared (Milan, 1976) [vol. 2 on morphology appeared in 1981—ed.]. Note also his text, An Introductory NT Greek Course
(Chicago, 1973).

On p. 169 he argues: 'The establishment of Bilingualism as an operative factor in the Greek of the papyri from Egypt excludes the validity of an appeal to parallels in these papyri to show that a suspected Semitism in biblical Greek is nothing but a pure Greek spoken throughout the Mediterranean world'.

But this is an overstatement on two scores. First, the mere possibility of Coptic interference does not exclude outright all the papyrological evidence: only that evidence can be excluded for which such interference can be demonstrated (that is, not every peculiarity in the papyri can be explained in the same way).

Secondly, Gignac's expression, 'pure Greek', erects a straw man—more than that, it is a regrettable misrepresentation of the view he opposes, since neither Deissmann nor his collaborators ever thought of biblical Greek as pure (whatever that could mean).

*11:'Stylistic Problems in the Epistles of James and Peter', ST 1 (1947), pp. 170-82, esp. 180-82. Wifstrand's views, along with those of David Tabachovitz, Lars Rydbeck [see Rydbeck's essay in this collection—ed.] and others will occupy our attention towards the end of this article.

*12:'The "Testament of Abraham": Problems in Biblical Greek', NTS 1 (1954-55), pp. 219-23, esp. 222.-23.

*13:Prolegomena, p. 11. 

*14:In VT 5 (1955), pp. 208-13.

*15:Syntax, vol. 3 of Moulton's A Grammar of NT Greek (Edinburgh, 1963). 

*16:Syntax, p. 3. 

*17:Syntax, p. 9. Although in this volume Turner did not commit himself on whether biblical Greek reflected a spoken Jewish vernacular, we should note his Grammatical Insights into the NT (Edinburgh, 1965), p. 183: 'Biblical Greek is so powerful and fluent, it is difficult to believe that those who used it did not have at
hand a language all ready for use. This, I submit, was the normal language of Jesus, at least in Galilee—rather a separate dialect of Greek than a form of the Koine, and distinguishable as something parallel to classical, Hellenistic, Koine and Imperial Greek' [Turner's essay is reprinted in this collection—ed.]. Note, more recently, N. Turner, 'Jewish and Christian Influence in the NT Vocabulary', NovT 16 (1974), pp. 149-60, esp. his conclusion.

*18:The Biblical Languages', in The Cambridge History of the Bible (vol. I; ed. P.R. Ackroyd and CF. Evans; Cambridge, 1970), p. 11 (my italics). 

*19:A.Meillet,Aperçu d'une histoire de la langue grecque (Paris, 8th edn, 1975), p. 329, regarding 'la grande κοινή ionienne-attique, qui était la seule langue de civilisation': 'L'observation du grec moderne montre que toutes les classes de la population, par des adaptations successives, ont fini par la parler et que, une à une, les particularités locales ont été presque partout éliminées'. Similarly, A. Debrunner and A. Scherer speak of the 'Dialektlosigkeit der Koine', in Geschichte der griechischen Sprache, II (Sammlung Göschen 114/114a; Berlin, 2nd edn, 1969), p. 92. Note also R. Browning, Medieval and Modern Greek (London, 1969), p. 55. For a valuable survey of research, note J. Frösen, Prolegomena to a Study of the Greek Language in the First Centuries AD: The Problem of Koine and Atticism (Helsinki, 1974), esp. pp. 71-80, 85-88, 177.

*20:Die griechische Sprache, pp. 166-67. 

*21:In col. 1364 Vergote claims that 'Thumb appartenait à l'école linguistique ancienne'. Towards the end of the article he suggests: 'Il n'y a pas de doute que l'étude du Nouveau Testament à la lumière de la linguistique moderne peut fournir encore des résultats remarquables' (col. 1367). For a cursory application of linguistics to our topic, see Christine Mohrmann, 'General Trends in the Study of NT Greek and of Early Christian Greek and Latin', in Classica et Iberica (ed.P.T. Braunan; Worcester, MA, 1975), pp. 95-105.

*22:M. Beziers and M. van Overbeke devote virtually a whole book to classification: Le bilinguisme: Essai de définition et guide bibliographique (Cahiers de l'Institut des Langues Vivantes, 13; Louvain, n.d.). The authors use three basic criteria (the relationship between the languages used by the bilinguals, the manner in which the languages were acquired, the degree of mastery), which are then subclassified for a total of over a dozen possible categories; yet even this classification is not exhaustive. See also the excellent survey of research by Els Oksaar, 'Bilingualism', in Current Trends in Linguistics 9 (The Hague, 1972), pp. 476-511,and the methodological comments by Andrew D. Cohen, 'Assessing Language Maintenance in Spanish Speaking Communities in the Southwest', in ElLenguaje de los Chicanos: Regional and Social Characteristics Used by Mexican Americans (ed. Eduardo Hernândez-Châvez et al:, Arlington, VA, 1975), pp. 202-19.

*23:Ε. Haugen, The Norwegian Language in America: A Study in Bilingual Behavior (Philadelphia, 1953), II, p. 370. The concept of prestige, however, needs to be used with caution; see T.E. Hope, Lexical Borrowing in the Romance Languages: A Critical Study of Italianisms in French and Gallicisms in Italian from 1100 to 1900 (Oxford, 1971), II, pp. 722-23.

*24:For example, C. Rabin suggests that probably 'it was in fact Greek which influenced both Hebrew and Aramaic' in the use of the tenses ('Hebrew and Aramaic in the First Century' [CRINT, I, 2; ed. S. Safrai and M. Stern; Assen, 1976], p. 1024).

*25:'Hellenistisches Griechisch', in Herzog-Hauck's Realen-cyklopädie für protestantische Theologie und Kirche (ed. A. Hauck; Leipzig, 1899), VU, p. 638: the Greek Bible's Okkasionellen Semitismen sind Kuriositäten, aber kein sprachhistorisches Moment; ihre usuellen Semitismen jedoch können das sprachwissenschaftliche Urteil so wenig verändern, wie etwaige Latinismen oder andere Beutestücke aus dem siegreichen Eroberungszug des Griechischen durch die Welt der Mittelmeerländer' [Deissmann's essay is translated in this collection—ed.].

*26:The Norwegian Language, II, p. 371 (my italics). 

*27:Although discussed at length and subjected to criticism at various points, this basic dichotomy informs all major approaches. J. Lyons puts it this way: 'Linguistswill argue about the degree of abstraction and idealization involved in the postulation of an underlying relatively uniform language-system; and many of them will deny that the system they postulate is internalized, as such, in the brains of the native speakers of the languages they are describing. But most linguists do nowadays draw some kind of distinction between language-behavior and the system of units and relations underlying that behavior'(Semantics [2 vols.; Cambridge, 1977], I,p. 239).  

*28:'Hellenistisches Griechisch', p. 637 (see also three notes above); it appears, incidentally, that Deissmann was not as ignorant of linguistics as Vergote seems to imply, and one should also remark that Moulton was a profound student of the linguistics of his day.

*29:'Grec biblique', col. 1366.

*30:It appeared in VT 1 (1951), pp. 81-90 [this article is reprinted in this collection—ed.]. 

*31:Gehman, 'Hebraic Character', p. 81.

*32:Gehman, 'Hebraic Character', p. 81.

*33:Gehman, 'Hebraic Character', p. 90.

*34:Gehman, 'Hebraic Character', p. 90. 

*35:Gehman, 'Hebraic Character', p. 90. 

*36:Gehman, 'Hebraic Character', p. 82 (my emphasis).

*37:For a superb article on this general question, see C. Rabin, "The Translation Process and the Character of the Septuagint', Textus 6 (ed. S. Talmon; Jerusalem, 1968), pp. 1-26, esp. his comments on 'semantic tolerance' (pp. 9-10) and on 'translations of embarrassment' (pp. 23-24); on p. 25 he argues that the features of the LXX 'simply have no direct bearing on' the question whether Alexandrian Jews spoke a Jewish Greek. Note also S.P. Brock, 'The Phenomenon of the Septuagint',in The Witness of Tradition (ed. A.S. van der Woude; OTS, 17; Leiden, 1972), pp. 11-36, esp. 31ff.; without making explicit reference to the langue-parole distinction, Brock in effect alludes to it when he states that the very inconsistency of Hebraisms in the LXX argues against a spoken Jewish Greek.

*38:Even in the case of widely different languages, children adapt very rapidly. Cf. E.C.Y. Kuo, 'Bilingual Patterns of a Chinese Immigrant Group in the US', Anthropological Linguistics 16 (1974), pp. 128-40. 

*39:Note also A.G. Lozano, 'Grammatical Notes on Chicano Spanish', The Bilingual Review 1 (1974), pp. 147-51. On pp. 149-50 he reports that in the southwest of the US we find traces of English ranging from soné de (< con) ella = Ί dreamed of her* to the almost incredible loan translation hice mi mente pa' arriba (instead of llegué a una decision) = Ί made up my mind'.

*40:In Texas Studies in Bilingualism: Spanish, French, German, Czech, Polish,Serbian, and Norwegian in the Southwest (Studia Linguistica Germanica, 3; Berlin,
1970), pp. 18-41.

*41:Sawyer, 'Spanish-English Bilingualism', p. 19.

*42:This is the generally accepted view. For a different interpretation of the evidence, see Charles C. Torrey, The Apocalypse of John (New Haven, 1958), p. 8 note; and Elias Bickerman, Studies in Jewish and Christian History (part one; AGJU, 9; Leiden, 1976), pp. 171ff.

*43:Cf. Gehman's own description, quoted above. Gehman wrote a later, morenarrowly conceived article on 'Hebraisms of the Old Greek Version of Genesis', VT 3 (1953), pp. 141-48. His conclusions are milder: 'Although the Greek rendering of Genesis has a number of Hebraisms, for the most part it could be understood by one whose native language was Hellenistic Greek' (p. 148). Further, he shies away from the position 'that the Alexandrian generation which had lost Hebrew and Aramaic, spoke a Greek influenced by Semitic idioms as that found in the LXX'. This is a significant concession, apparently not recognized by Turner, but one wonders what it really means for Gehman to add that 'there was a Greek with a decided Hebrew cast that was understood in religious circles'. This final statement is either a contradiction of what he has just said or else it is a mere truism.

*44:I have no precise figures regarding what proportion of this population learned Catalan as their mother-tongue. Although my first-hand acquaintance with this bilingual situation is practically non-existent, it may be instructive to note that the manager
of a hotel in Barcelona where I stayed a few years back, though perfectly fluent in Spanish, resorted to Catalan when doing arithmetical operations, a clear indication that Catalan was her first language. Presumably, she was typical of the population in general.

*45:A.M. Badia-Margarit, 'Some Aspects of Bilingualism Among Cultured People in Catalonia', in Proceedings of the Ninth International Congress of Linguistics(ed. H.G. Lunt; The Hague, 1964), p. 367.

*46:Badia-Margarit, 'Some Aspects of Bilingualism', pp. 372-73.

*47:Moulton, Prolegomena, pp. 7, 10-11. From another perspective, A.W. Argyle has also alluded to the situation in Wales ('Greek Among the Jews of Palestine in NT Times', NTS 20 [1973-74], pp. 87-89).

*48:Palestinian speakers must have made mistakes in pronunciation (not evident in the written form because of standardized orthography) and in morphology (though someone who knew Greek well enough to write it was not likely to fail in this area). Since Vergote argued ('Grec biblique', col. 1364) that interference affects primarily the semantic domain, the reader may be referred to my article, 'Semantic Borrowing
in the NT', NTS 22 (1975-76), pp. 104-10 (on p. 109 of this article, lines 11 and 14, 'former loans' should read 'latter loans' and vice versa).

*49:J.N. Sevenster, Do You Know Greek? How Much Greek Could the First Jewish Christians Have Known? (NovTSup, 19; Leiden, 1968), p. 189. For a capable survey of the data and a sober evaluation, see G. Mussies, 'Greek in Palestine and the Diaspora', CRINT, pp. 1040-54. Several items could be added to his bibliography, such as J.A. Fitzmyer, 'The Languages of Palestine in the First Century AD', CBQ 32 (1970), pp. 501-31 [reprinted in this collection—ed.], and
R.M. Mackowsky, Spoken Greek of the First Century AD (dissertation, Hebrew University, Jerusalem, 1971 [Hebrew]) (unavailable to me). 

*50:M. Hengel,Judaism and Hellenism: Studies in Their Encounter in Palestine During the Early Hellenistic Period (2 vols.; Philadelphia, 1974), I, pp. 58ff. Most recently, see E. Schürer, HJPAJC (2 vols.; Edinburgh, 1979), II, pp. 74-80, which however plays down the use of Greek by the populace.

*51:See pp 211-12 n. 5. 

*52:Cf. G.W. Turner, Stylistics (Baltimore, 1973), p. 19. On p. 21 he states that the grammatical set of rules in a language 'is prior to style. It is given by the language, leaving no choice, and .. . an element of choice seems to be basic to all conceptions of style.' However, N.E. Enkvist (Linguistic Stylistics [Janua linguarum, series critica, 5; The Hague, 1973], p. 37) argues rightly against too facile an equation of style with parole.

*53:He himself tells us that in his view style 'involves the same considerations as syntax' (see Style, vol. 4 of Moulton's A Grammar of NT Greek [Edinburgh, 1976], p. 1). More properly, style cuts across all levels of linguistic description, although it manifests itself most clearly in lexical choices and in those syntactical constructions not determined by 'grammar'. Cf. my article, 'The Pauline Style as Lexical Choice: Γινώσκειν and Related Verbs', Pauline Studies (FS F.F. Bruce; ed. D.A. Hagner and M.J. Harris; Grand Rapids, 1980), pp. 184-207.

*54:Thumb, Die griechische Sprache, p. 121. Note also Deissmann's reference to 'birthmarks' (quoted above, p. 208) and Radermacher's words (quoted above, p. 209 n. 2).

*55:'Stylistic Problems', pp. 181-82. L. Rydbeck has also emphasized this point in 'What Happened to NT Greek Grammar After Albert Debrunner?', NTS 21 (1974-75), pp. 424-27. (On Rydbeck's Zwischenschichtsprosa, see Frösen, Prolegomena, p. 93, and E. Pax, 'Probleme des neutestamentlichen Griechisch', Bib 53 [1972], pp. 557-64.) Perhaps Rabin takes the same approach, although his use of the term 'diglossia' ('Hebrew and Aramaic', p. 1008) to describe the status of Greek in Palestine seems to me unfortunate.

*56:I have argued this point at greater length in 'Semantic Change in the Greek Bible', forthcoming in a Festschrift.

*57:CS. Lewis, The Literary Impact of the Authorized Version (London, 1958), pp. 1 Iff.; he states that in the specific area of vocabulary the AV has indeed influenced English, but unfortunately fails to elaborate the point. Büchsei ('Die griechische Sprache', p. 142) argues that 'die Umgangsprache der Juden und die Sprache ihres heiligen Buches waren zwei verschiedene Dinge, und man wird nie ein Verständnis weder der einen noch der andern erreichen, wenn man diesen
Unterschied unterschätzt'.

*58:L'Ancien Testament et la langue française du moyen âge (VIIIe-XVesiècle)(Genève, 1968 [1904]).

*59:D. Tabachovitz, Die Septuaginta und das Neue Testament: Stilstudien (Skrifter Utgivna av Svenska Institutet i Athen, 8 IV; Lund, 1956), p. 18. Cf. also Brock, 'The Phenomenon of the Septuagint', pp. 35-36, and Ε. Schwyzer, Griechische Grammatik (Handbuch der Altertumswissenschaft, II.l.l; Munich, 1953), p. 126.

*60:Έn torno al estudio del griego de los cristianos', Emérita 41 (1973), pp. 45-56, esp. 56. His opinions are based on G.J.M. Bartelink, Lexicologisch-semantische Studie over de Tool van de apostolische Voders (Utrecht, 1952). We should note that any linguistic group with specialized interests (even a small family) develops its own specialized vocabulary. A modern linguist would surely treat the Stoic writings as a special style.