PART III
SYNTAX
CHAPTER IX
THE MEANING OF SYNTAX ( SUNTAXIS)
I. Backwardness in the Study of Syntax. What the Germans call Laut- und Formenlehre has received far more scientific treatment than has syntax. In 1874 Jolly1 lamented that so little work on syntax of a really valuable nature had been done. To a certain extent it was necessary that the study of the forms should precede that of syntax.2 The full survey of the words and their inflections was essential to adequate syntactical investigation. And yet one can but feel that syntax has lagged too far behind. It has been the favourite field for grammatical charlatans to operate in, men who from a few examples drew large inductions and filled their grammars with "exceptions" to their own hastily made rules. Appeal was made to logic rather than to the actual facts in the history of language. Thus we had grammar made to order for the consumption of the poor students.
Others perhaps became disgusted with the situation and hastily concluded that scientific syntax was impracticable, at least for the present, and so confined their researches either to etymology or to the forms. In 1891 Muller3 sees no hope of doing anything soon for modern Greek syntax except in the literary high style on which he adds a few remarks about prepositions. Thumb4 likewise has added a chapter on syntax to his Handbuch. If you turn to Whitney's Sanskrit Grammar, you will find no separate syntax, but merely some additional remarks on the "uses" of the aorist, the present, the subjunctive, etc. Monro in his Homeric Grammar follows somewhat the same plan, but with much more attention to the "uses" of cases and modes. Brugmann5 in his Griechische Grammatik devotes far more space to Formenlehre,
380 A GRAMMAR OF THE GREEK NEW TESTAMENT
even in the third edition, which chiefly differs from the second in the increased attention to syntax. Giles in his Manual of Comparative Philology, even in the second6 edition (1900), kept his discussion of the uses of the noun and verb apart and did not group them as syntax. When he wrote his first7 edition (1895) nothing worthy of the name had been done on the comparative syntax of the moods and tenses, though Delbruck had written his great treatise on the syntax of the noun. When Brugmann planned his first volume of Kurze vergleichende Grammatik (1880), he had no hope of going on with the syntax either with the "Grundriss" or the "Kurze," for at that time comparative grammar of the Indo-Germanic tongues was confined to Laut- und Formenlehre.8 But in the revision of Kuhner the Syntax by B. Gerth has two volumes, as exhaustive a treatment as Blass' two volumes on the Accidence. In the Riemann and Goelzer volumes the one on Syntax is the larger. Gildersleeve (Am. Jour. of Philol., 1908, p. 115) speaks of his convictions on "Greek syntax and all that Greek syntax implies." No man's views in this sphere are entitled to weightier consideration. May he soon complete his Syntax of Classical Greek.
As to the dialectical inscriptions the situation is still worse. Dr. Claflin9 as late as 1905 complains that the German monographs on the inscriptions confine themselves to Laut- und Formenlehre almost entirely. Meisterhans in Schwyzer's revision (1900) is nearly the sole exception.10 Thieme11 has a few syntactical remarks, but Nachmanson,12 Schweizer13 and Valaori14 have nothing about syntax, nor has Dieterich.15 The same thing is true of Thumb's Hellenismus, though this, of course, is not a formal grammar. A few additional essays have touched on the syntax of the Attic inscriptions16and Schanz in his Beitrage has several writers17 who have noticed the subject. The inscriptions do indeed have limitations as to syntax, since much of the language is official and formal, but there is much to learn from them. Thackeray has not yet published his Syntax of the LXX. nor has Helbing.
THE MEANING OF SYNTAX ( SUNTAXIS) 381
We are somewhat better off as to the papyri as a result chiefly of the work of Dr. James Hope Moulton, who has published his researches in that field as applied to the New Testament.18 Cronert in his Mem. Graeca Hercul. has a good many syntactical remarks especially on the cases,19 but no formal treatment of the subject. Volker20 has not finished his good beginning. No syntax has come from Mayser yet who stopped with Laut- und Formenlehre, though he is at work on one. Moulton does not profess21 to cover all the syntactical points in the papyri, but only those that throw light on some special points in the N. T. usage.
II. New Testament Limitations. It is evident therefore that the N. T. grammarian is in a poorer plight when he approaches syntax. And yet, strange to say, the N. T. grammars have largely confined themselves to syntax. Winer-Moulton, out of 799 pages, has only 128 not syntax. Buttmann, out of 403 pages (Thayer's translation), has only 74 not syntax. In Winer-Schmiedel syntax is reached on p. 145. Blass begins syntax on p. 72, out of 305 pages. Moulton in his Prolegomena starts syntax on p. 57 (232 in all). The present book has given the discussion of the forms more space at any rate. It is at least interesting to note that N. T. grammarians have reversed the example of the comparative philologists. Is it a case of rushing in where angels fear to tread?
One may plead in defence that the demands of exegesis are great and urgent, not to say more congenial. The distinctive character of the N. T. teaching is more closely allied to lexicography and syntax than to mere forms. That is very true, but many a theologian's syntax has run away with him and far from the sense of the writer, because he was weak on the mere forms. Knowledge of the forms is the first great step toward syntax. Deissmann even complains of Blass for assuming too much in his Syntax and not making enough comments "to rouse up energetically this easy-going deference of the youthful reader" (Expositor, Jan., 1908, p. 65).
Blass22 urges, besides, that it is just in the sphere of syntax that
382 A GRAMMAR OF THE GREEK NEW TESTAMENT
the N. T. variations from the ancient Greek can be best observed, in this and the change in the meaning of words. This is true, but just as much so of the koinh, in general. This is just the opposite of Winer's view,23 who held that the N. T. peculiarities of syntax were very few. The explanation of the difference lies partly in the undeveloped state of syntax when Winer wrote, though he wrote voluminously enough himself, and partly in the wider conception of syntax that Blass24 holds as being "the method of employing and combining the several word-forms and 'form-words' current in the language."
On the other hand attention must be called to the fact that the study of the forms is just the element, along with vocabulary, mainly relied on by Deissmann in his Bible Studies to show the practical identity of the vernacular koinh, in the papyri and in the N. T. Greek. Burton25 puts it rightly when he says of the N. T. writers: "The divergence of their language from that of classical writers in respect to syntax is greater than in reference to forms of words, and less than in respect to the meaning of words, both the Jewish and the Christian influence affecting more deeply the meanings of words than either their form or their syntactical employment." Deissmann26 readily admits that Christianity has a set of ideas peculiar to itself, as has every system of teaching which leads to a characteristic terminology.
But one is not to think of the N. T. as jargon or a dialect of the koinh, in syntax.27 It is not less systematic and orderly than the rest of the vernacular koinh,, and the koinh, is as much a real language with its own laws as the Greek of Athens.28 As remarked above, the koinh, showed more development in syntax than in forms, but it was not a lawless development. It was the growth of life and use, not the artificial imitation of the old language of Athens by the Atticists. Blass29 properly insists on the antithesis here between the artificial Atticist and "the plain narrator of facts or THE MEANING OF SYNTAX ( SUNTAXIS) 383 epoch, all the outward devices of rhetoric. In grammar, vocabulary, syntax and style it occupies a place in the midst of the people and draws from the inexhaustible soil of the popular element to which it was native a good share of its youthful strength." This is largely true. Men of passion charged with a great message do strike forth the best kind of rhetoric and style with simplicity, power, beauty. It is blind not to see charm in Luke, in John, in Paul, James and the writer of Hebrews, a charm that is the despair of mere "devices of rhetoric" or artificial rules of style and syntax. It is not surprising to find variations in culture in the N. T. writers, men who had different antecedents (Jew or Greek), different environment (Palestine, Asia Minor and possibly Egypt), different natural gifts and educational advantages, as seen in Peter and Paul. These individual peculiarities show themselves easily and naturally in syntax and style. See chapter IV, The Place of the N. T. in the koinh,, for a larger discussion of this matter of the peculiarities of the N. T. writers. But even in 2 Peter and the Apocalypse one has no difficulty in understanding this simple vernacular koinh,, however far short these books come of the standard of Isocrates or Demosthenes. The study of N. T. syntax is a worthy subject and one entirely within the range of scientific historical treatment so far as that subject has advanced. III. Recent Advance by Delbruck. Just as Brugmann is the great name in the accidence of comparative grammar, so Delbruck is the great name in syntax. Brugmann gladly recognises his own indebtedness to Delbruck. He has sought to follow Delbruck in the syntax of his Griechische Grammatik30 and in the Kurze vergleichende Grammatik.31 It is not necessary here to recount the story of how Delbruck was finally associated with Brugmann in the Grundriss, and the Syntax by Delbruck brought to completion in 1900. Brugmann tells the story well in Kurze vergl. Gr. (pp. v ff.) and Delbruck in the Grundriss itself. It is a great achievement and much led up to it. Delbruck has recounted the progress of comparative grammar in his Introduction to the Study of Language (1882). In 1872 he had published Die Resultate der vergleichenden Syntax. In 1879 he brought out Die Grundlagen der griechischen Syntax ("Syntaktische Forschungen,"
384 A GRAMMAR OF THE GREEK NEW TESTAMENT Bd. IV). That marked him as the man to do for syntax what Brugmann would do for forms. Delbruck does not claim all the credit. Bernhardy in 1829 had published Wissenschaftliche Syntax der griechischen Sprache, but Bopp, Schleicher and the rest had done much besides. The very progress in the knowledge of forms called for advance in syntax. In 1883 Hubner wrote Grundriss zu Vorlesungenuaber die griechische Syntax. It is not a treatment of syntax, but a systematized bibliography of the great works up to date on Greek syntax. It is still valuable for that purpose. One can follow Brugmann32 and Delbruck, Vergl. Syntax, Dritter Teil, pp. xvi-xx, for later bibliography. As the founders of syntax Hubner33 points back to Dionysius Thrax and Apollonius Dyscolus in the Alexandrian epoch. The older Greeks themselves felt little concern about syntax. They spoke correctly, but were not grammatical anatomists. They used the language instead of inspecting and dissecting it. Delbruck (Vergleichende Syntax, Erster Teil, pp. 2-72) gives a lucid review of the history of syntactical study all the way from Dionysius Thrax to Paul's Principles of the History of Language. He makes many luminous remarks by the way also on the general subject of syntax. I cannot accent too strongly my own debt to Delbruck. Syntax, especially that of the verb, has peculiar difficulties.34 Not all the problems have been solved yet.35 Indeed Schanz so fully appreciates the situation that he is publishing a series of excellent Beitrage zur historischen Syntax der griechischen Sprache. He is gathering fresh material. Many of the American and European universities issue monographs by the new doctors of philosophy on various points of syntax, especially points in individual writers. Thus we learn more about the facts. But meanwhile we are grateful to Delbruck for his monumental work and for all the rest. IV. The Province of Syntax. (a) THE WORD SYNTAX ( su,ntaxij). It is from sunta,ssw and means 'arrangement' (constructio).36 It is the picture of the orderly marshalling of words to express ideas, not a mere medley of words. The word syntax is indeed too vague and general to express clearly all the uses in modern grammatical discussion, but it is
THE MEANING OF SYNTAX ( SUNTAXIS) 385 too late to make a change now.37 Gildersleeve (Am. Jour. of Philol., 1908, p. 269) says that some syntacticians treat "syntax as a rag-bag for holding odds and ends of linguistic observations." (b) SCOPE OF SYNTAX. But the difficulty is not all with the term, for the thing itself is not an absolutely distinct province. What the Germans call Lautlehre ('teaching about sounds') is indeed quite to itself. But when we come to define the exact line of demarcation between syntax or the relation of words on the one hand and single words on the other the task is not always so easy. Ries38 indeed in his very able monograph makes the contrast between syntax (or construction) and single words. His scheme is this: Under Wortlehre ('science of words') he puts Formenlehre ('theory of forms') and Bedeutungslehre ('meaning of words').39 He also subdivides syntax in the same way. Syntax thus treats of the binding of words together in all relations. Brugmann40 follows Delbruck41 in rejecting the special use of syntax by Ries. Brugmann42 considers the breaking-up of the sentence by Ries into single words to be wilful and only conventional. It is indeed true that single words have a teaching both as to the word itself (form-word, as prepositions) and the form (inflection).43 That is to say, two things call for consideration in the case of single words: the facts as to the words and the inflection on the one hand and the meaning of these facts on the other. Now Ries refuses to give the term syntax to the meaning of these facts (words, inflections, etc.), but confines syntax to the other field of word-relations. One is bound to go against Ries here and side with Delbruck and Brugmann. (c) CONSTRUCTION OF WORDS AND CLAUSES. We use syntax, therefore, both for construction of the single word and for clauses. But one must admit the difficulty of the whole question and not conceive that the ancients ran a sharp line between the form and the meaning of the form. But, all in all, it is more scientific to gather the facts of usage first and then interpret these facts. This interpretation is scientific syntax, while the facts of usage are themselves syntax. Thus considered one may properly think of syntax in relation to the words themselves, the forms of the
386 A GRAMMAR OF THE GREEK NEW TESTAMENT words, the clauses and sentences, the general style. Clyde makes two divisions in his Greek Syntax, viz. Words (p. 126) and Sentences (p. 193). But this formal division is artificial. Here, as usual, Delbruck has perceived that syntax deals not only with words (both Wortarten and Wortformen), but also with the sentence as a whole and all its parts (Vergl. Syntax, Erster Teil, p. 83). How hard it is to keep syntactical remarks out of accidence may be seen in Thackeray's vol. I and in "Morphology" in Thumb's Handbook as well as in Accidence of this book. (d) HISTORICAL SYNTAX. But this is not to fall into the old pitfall of the Stoic grammarians and apply logic to the phenomena of grammar, using the phenomena of various grammatical categories previously laid down. Plato indeed first applied logic to grammar.44 The method of historical grammar and comparative grammar has had a long and a hard fight against the logical and philosophical method of syntax. But it has at last triumphed. "They sought among the facts of language for the illustration of theories," as Dr. Wheeler45 so well puts it. We still need logic and philosophy in syntax, but we call these two agents into service after we have gathered the facts, not before, and after the historical and comparative methods have both been applied to these facts. Thus alone is it possible to have a really scientific syntax, one "definitely oriented" "as a social science" dealing with the total life of man.46 (e) IRREGULARITIES. We shall not therefore be surprised to find many so-called "irregularities" in the use of syntactical principles in various Greek writers. This is a point of the utmost importance in any rational study of syntax. The personal equation of the writer must always be taken into consideration. A certain amount of elasticity and play must be given to each writer if one is to understand human speech, for speech is merely a reflection of the mind's activities. If a tense brings one to a turn, perhaps it was meant to do so. This is not to say that there are no barbarisms or solecisms. Far from it. But it is unnatural to expect all speakers or writers in Greek to conform slavishly to our modern grammatical rules, of most of which, besides, they were in blissful ignorance. The fact is that language is life and responds to the peculiarities of the individual temper, and it is to be remembered that the mind itself is not a perfect instrument. The
THE MEANING OF SYNTAX ( SUNTAXIS) 387 mind is not always clear nor logical. The ellipses, anacolutha, etc., of language represent47 partially the imperfections of the mind. "It often depends on the writer which of the two tenses he will use," Winer48 remarks about the aorist and the past perfect. It always depends on the writer which tense and which everything else he will use. Pray, on whom else can it depend? The writer happens to be doing the writing. He decides whether he will conform to the usual construction or will give added piquancy by a variation. This assumes, of course, that he is an educated writer. If he is not, he will often have the piquancy just the same without knowing it. "Syntactical irregularities are numerous in Greek," Clyde49 observes, and, he might have added, in all other living languages. Greek is not, like "Esperanto," made to order by any one man. In point of fact what we call idioms are the very peculiarities ( ivdiw,mata) which mark it off from other languages or at least characterize it. Some of these idioms spring out of the common intelligence of men and belong to many tongues, others mark the variations of certain minds which gain a following. Compare the rapid spread of "slang" to-day, if it happens to be a "taking phrase." Hence rules of syntax ought not to be arbitrary, though many of them are. Those that really express the life of language are in harmony with the facts. In general I would say that the fewer rules one gives the better for the student and for the facts. V. The Method of this Grammar. (a) PRINCIPLES, NOT RULES. As far as possible principles and not rules will be sought. The Greek grammarian is an interpreter of the facts, not a regulator of the facts. This point calls for special emphasis in syntax where the subjective element comes in so largely. (b) THE ORIGINAL SIGNIFICANCE. The starting-point therefore in the explanation of any given idiom is to find the original significance. This is not always possible, but it generally is. Historical and comparative grammar lend strong help in this endeavour. Always the best place to begin is the beginning if you can find it. (c) FORM AND FUNCTION. I would not insist that form and function always correspond. One does not know that the two did so correspond in the beginning in all instances. It is hard to prove a universal proposition. But certainly one is justified in beginning with one function for one form wherever he finds it to
388 A GRAMMAR OF THE GREEK NEW TESTAMENT be true. Burton50 says: "It is by no means the case that each form has but one function, and that each function can be discharged by but one form." Certainly the same function can come to be discharged by various forms, as is the case with the locative and dative infinitive forms ( labei/n, avkou/sai). But that is not to say that originally the locative and dative verbal substantive were identical in idea. The Sanskrit completely disproves it. It may very well be true that each form had one function originally, whereas later the same function came to be expressed by various forms. As a starting-point, therefore, one may assume, till he learns otherwise, that form and function correspond. The necessity of getting at the ground-idea of an idiom is rightly emphasized by Delbruck (Grundlagen, p. 1). It may indeed come to pass as in the English "but," that the one form may be used for most of the parts of speech (Giles, Man. of Comp. Philol., p. 237 f.). On the whole subject of the agreement of form and idea see KuhnerGerth, I, pp. 64-77. (d) DEVELOPMENT. But the beginning is not the end. The actual development of a given idiom in the Greek language up to the N. T. time must be observed. Each idiom has a history. Now it cannot be expected that the space can be given to the actual working-out of each idiom in history as Jannaris has done in his Historical Grammar, or minute comparison at every point by means of comparative grammar. What is essential is that the grammarian shall have both these points in mind as he seeks to explain the development from the etymological basis. This is the only secure path to tread, if it can be found. Burton51 indeed distinguishes sharply between historical and exegetical grammar and conceives his task to be that of the exegetical grammarian. For myself I regard exegetical grammar as the last stage in the process and not to be dissociated from the historical. Indeed how a Greek idiom is to be represented in English is a matter of little concern to the Greek grammarian till the work of translation is reached. The Greek point of view is to be observed all through the process till translation comes. It is Greek syntax, not English. (e) CONTEXT. There is one more stage in the interpretation of the Greek idiom. That is the actual context in any given instance. The variation in the total result is often due to the difference in the local colour of the context. The same idiom with a given etymology may not have varied greatly in the long course of history save as it responds to the context. In a word, etymol-
THE MEANING OF SYNTAX ( SUNTAXIS) 389 ogy, history, context are the factors that mark the processes in the evolution of a Greek idiom in a given case. These are the things to keep constantly in mind as we approach the idioms of Greek syntax. We may not always succeed in finding the solution of every idiom, but most of them will yield to this process. The result is to put syntax on a firmer scientific basis and take it out of the realm of the speculative subjective sciences. (f) TRANSLATION. This is the translation of the total result, not of the exact Greek idiom. Translation crisply reproduces the result of all the processes in harmony with the language into which the translation is made, often into an utterly different idiom. It is folly to reason backwards from the translation to the Greek idiom, for the English or German idiom is often foreign to the Greek and usually varies greatly from the original Greek. English is English and Greek is Greek. Syntax is not translation, though it is the only safe way to reach a correct translation. Exegesis is not syntax, but syntax comes before real exegesis. The importance of syntax is rightly appreciated by Gildersleeve.52 (g) LIMITS OF SYNTAX. After all is done, instances remain where syntax cannot say the last word, where theological bias will inevitably determine how one interprets the Greek idiom. Take u[dati in Ac. 1:5, for instance. In itself the word can be either locative or instrumental with bapti,zw. So in Ac. 2:38 eivj does not of itself express design (see Mt. 10:41), but it may be so used. When the grammarian has finished, the theologian steps in, and sometimes before the grammarian is through. 1 Schulgr. und Sprachw., p. 71.
2 Riem. and Goelzer, Gr. Comparee du Grec et du Lat., Synt., p. 7.
3 Hist. Or. der hell. Spr., p. 172.
4 Handb. der neugr. Volksspr., 1895; Handb. of Mod. Gk. Vernac., pp. 179-206.
5 P. vii.
6 P. xi.
7 P. viii f.
8 Kurze vergl. Gr., 3. Lief., 1904, p. iii f.
9 Synt. of the Bceot. Dial. Inscr., p. 9.
10 Gr. der att. Inschr. But even he has very much more about the forms,
11 Die Inschr. von Magn. etc., 1906.
12 Laute und Formen der magn. Inschr., 1903.
13 Gr. d. perg. Inschr., Beitr. zur Laut- und Formenl. etc., 1898.
14 Der delph. Dial., 1901.
15 Unters. etc., 1898.
16 Claflin, Synt. of the Boeot. Dial. Inser., p. 10.
17 Dyroff, Weber, Keck.
18 See Cl. Rev., Dec., 1901, pp. 436 ff.; Apr., 1904, p. 150; Exp., 1904, series on Charact. of N. T. Gk.; Prol., 1906.
19 Pp. 159
20 Synt. der griech. Pap., I, Der Art., 1903.
21 Cl. Rev., Dec., 1901, p. 436. Debrunner (p. xi of his 4. Aufl. of Blass' Gramm. d. N. Griech., 1913, which he has kindly sent me as I reach this point in the galley proof) laments: "Par die Studien der hellenistischen (und der mittel- und neugriechischen) Syntax gilt leider noch das Wort polu.j me.n o` qerismo,jà oi` de. evrga,tai ovli,goi."
22 Gr. of N. T. Gk., p. 72.
23 W.-M., p. 27.
24 Gr. of N. T. Gk., p. 72; cf. p. 3 also.
25 Notes on N. T. Gr., 1904, p. 22.
26 B. S., p. 65.
27 Thumb, Die sprachgeschichtl. Steil. des bibl. Griech., Theol. nu., 1002, p. 97.
28 Blass, Gr. of N. T. Gk., p. 3.
29 Ib., p. 72.
30 P. vii.
31 P. ix. He feels "als Schiller unseres Begrunders and Meisters der vergleichenden Syntax."
32 Griech. Gr., p. 363.
33 Grundr. zu Vorles., p. 3.
34 Giles, Comp. Philob; pp. 404 f., 475.
35 Riem. and Goelzer, Synt., p. 7.
36 Farrar (Gk. Synt., p. 54) quotes Suetonius as saying that the first Gk. gr. brought to Rome was by Crates Mallotes after the Second Punic War.
37 Brug., Griech. Gr., p. 364.
38 Was ist Syntax? 1894, p. 142.
39 Ib., p. 142 f. Ries calls it a "naive misuse of the word syntax" not to take it in this sense. But he is not himself wholly consistent.
40 Griech. Gr., p. 363 f.; Kurze vergl. Gr., III, p. vii.
41 Grundr., V, pp. 1
42 Kurze vergl. Gr., III, p. vii.
43 Brug., Griech. Gr., p. 363.
44 Sandys, Hist. of Cl. Scholarship, vol. I, p. 90.
45 The Whence and Whither of the Mod. Sci. of Lang., p. 97.
46 Ib., p. 107.
47 Clyde, Gk. Synt., p. 4 f.
48 W.-Th., p. 276.
49 Synt., p. 5.
50 N. T. Moods and Tenses, p. 1.
51 Ib., p. 3.
52 Synt. of Class. Gk., p. iv. C. and S., Sel. fr. the LXX, p. 22, observe that the life of a language lies in the syntax and that it is impossible to translate syntax completely. The more literal a translation is, like the LXX, the more it fails in syntax.