Just to recap on this series, since I’ve been so busy with so many other things and so lazy about my blog, you may want to get a refresher at what we’re discussing in the introductory post here: Linguistic adequacy and Greek grammars and here: Linguistic adequacy and BDF. I really need to finish this series, because I need to write at least two blog posts as prefatory material for sharing my thesis.

Of course, Robertson’s ‘Big Grammar’ refers to his Grammar of the Greek New Testament in the Light of Historical Research (Logos) (Amazon).[1] The primary challenge in discussing examples form Robertson’s grammar arises from the great difficulty of finding excerpts that are both complete in their discussion of a topic while also being fairly brief in nature. One of the shorter sections for this purpose is Robertson’s discussion of οὐ in the indicative mood and in main clauses. Nevertheless, the discussion is nearly 1500 words long.[2] Often times, it feels as if Robertson’s larger discussions create more questions than they answer. Before looking at the content itself, the structure of this excerpt is interesting. Robertson’s structure here is very well thought out. The sections flow nicely from one to the next: origin to history and then meaning to usage.

1. The Objective Οὐ and its Compounds

(a) Origin. This is unknown. Hübschmann⁴ sees a connection with the Latin haud as do other scholars.⁵ Fowler⁶ takes it as an original intensive particle like pas in the French ne pas and -χί(Indo-Ger. -g̑hi) in οὐ-χί. The Zend ava is also noted and the Latin au (au-fero).⁷ But there is no doubt that οὐ in the Greek took the place of the Sanskrit ná, Latin nĕ- (ne-que, ne-scio; the relation of nē nē-quidem, nē-quam to this nĕ is not known), Gothic ni. The use of the Greek οὐ corresponds to the Sanskrit ná.

(b) History. As far back as Greek goes we find οὐ, but οὐ did not hold its own with μή in the progress of the language. Within the past century οὐ has become obsolete in modern Greek outside of a few proverbs save in the Laconian and the Pontic dialects.¹ The Pontic dialect uses κί from Old Ionic οὐκί. But modern Greek has οὐδέ and οὔτε (Thumb, Handb., p. 200). In the Bœotian dialect, it may be noted, οὐ never did gain a place. We have seen οὐδέν used as an adverb, an idiom that goes back to Homer.² Jannaris³ explains that the vernacular came to use οὐδέν and μηδέν for emphasis and then on a par with οὐ and μή. Then οὐδέν dropped οὐ and μηδέν lost δέν, leaving δέν and μή for the modern Greek. At any rate this is the outcome. Δέν is the negative of the ind. in modern Greek except after νά and final clauses when we find νὰ μή (Thumb, Handb., p. 200). And δέν is the regular negative in the protasis of conditional sentences both with ind. and subj.⁴ The distinction between οὐ and μή did become more or less blurred in the course of time, but in the N. T., as in the κοινή generally, the old Greek idiom is very well preserved in the main. Buttmann⁵ even thinks that the N. T. idiom here conforms more exactly to the old literary style than in any other point. Δέν may represent μηδέν (Rendel Harris, Exp., Feb., 1914, p. 163).

(c) Meaning. Οὐ denies the reality of an alleged fact. It is the clear-cut, point-blank negative, objective, final.⁶ Jannaris⁷ compares οὐ to ὅτι and μή to ἵνα, while Blass⁸ compares οὐ to the indicative mode and μή to the other modes. But these analogies are not wholly true. Sometimes, indeed, οὐ coalesces with the word as in οὔ φημι=not merely ‘I do not say,’ but ‘I deny.’ So οὐκ ἐάω (Ac. 16:7)=‘I forbid.’ Cf. οὐ θέλω (Mk. 9:30); οὐκ ἔχω (Mt. 13:12); οὐκ ἀγνοέω (2 Cor. 2:11). See also τὸν οὐ λαόν in Ro. 9:25 (LXX) where οὐ has the effect of an adjective or a prefix. Delbrück⁹ thinks that this use of οὐ with verbs like the Latin ne-scio was the original one in Greek. In the LXX οὐ translates לֹא.

(d) Uses. Here it will be sufficient to make a brief summary, since the separate uses (pp. 917 f., 929 f., etc.) are discussed in detail in the proper places. The point here is to show how all the varied uses of οὐ are in harmony with the true meaning of the particle.

(i) The Indicative. We meet οὐ with the indicative in both independent and dependent clauses.

(α) Independent Sentences. Here the negative οὐ is universal with the indicative in declarative sentences. The force of οὐ (οὐκ before vowels, οὐχ before aspirate) is sometimes very powerful, like the heavy thud of a blow. Cf. οὐκ ἐδώκατε, οὐκ ἐποτίσατε, οὐ συνηγάγετε, οὐ περιεβάλετε, οὐκ ἐπεσκέψασθε (Mt. 25:42 f.). The force of all these negatives is gathered up in the one οὐ in verse 44. In verse 45 οὐ and οὐδέ are balanced over against each other. See οὐκ ἔπεσεν in Mt. 7:25. Cf. οὐ παρέλαβον in Jo. 1:11. In Mt. 21:29 see the contrast between ἐγώ, κύριε and οὐκ ἀπῆλθεν. Note the progressive bluntness of the Baptist’s denials till οὔ comes out flat at the last (>Jo. 1:21 f.). In the N. T. οὐ alone occurs with the future indicative used as a prohibition, though the classic idiom sometimes had μή. Cf. οὐ φονεύσεις (Mt. 5:21); οὐκ ἔσεσθε ὡς οἱ ὑποκριταί (6:5), etc. Still, Blass¹ quotes μηδένα μισήσετε in Clem., Hom., III, 69. The volitive subjective nature of this construction well suits μή, but οὐ is more emphatic and suits the indicative. In Mt. 16:22, οὐ μὴ ἔσται σοι τοῦτο, we have οὐ μή in the prohibitive sense. When οὐ occurs alone=‘no,’ as at the end of a clause, it is written οὔ as in οὔ, μή ποτε (Mt. 13:29); τὸ Οὔ οὔ (2 Cor. 1:17).

But in interrogative (independent) sentences οὐ always expects the answer ‘yes.’ The Greek here draws a distinction between οὐ and μή that is rather difficult to reproduce in English. The use of a negative in the question seems naturally to expect the answer ‘yes,’ since the negative is challenged by the question. This applies to οὐ. We may leave μή till we come to it. Οὐ in questions corresponds to the Latin nonne. Cf. Mt. 7:22, οὐ τῷ σῷ ὀνόματι ἐπροφητεύσαμεν κτλ., where οὐ is the negative of the whole long question, and is not repeated with the other verbs. See further Mt. 13:55; Lu. 17:17; 1 Cor. 14:23. In 1 Cor. 9:1 we have οὐ four times (once οὐχί). The form οὐχί is a bit sharper in tone. Cf. Mt. 13:27; Lu. 12:6. In Lu. 6:39 we have μή with one question, μήτι δύναται τυφλὸς τυφλὸν ὁδηγεῖν; and οὐχί with the other (side by side) οὐχὶ ἀμφότεροι εἰς βόθυνον ἐμπεσοῦνται; There is a tone of impatient indignation in the use of οὐ in Ac. 13:10, οὐ παύσῃ διαστρέφων τὰς ὁδοὺς τοῦ κυρίου τὰς εὐθείας; In Ac. 21:38, οὐκ ἄρα σὺ εἶ ὁ Αἰγύπτιος; the addition of ἄρα means ‘as I supposed, but as I now see denied.’¹ In Mk. 14:60 note the measured use of οὐ and οὐδέν in both question, οὐκ ἀποκρίνῃ οὐδέν; and the description of Christ’s silence, καὶ οὐκ ἀπεκρίνατο οὐδέν. In Lu. 18:7, οὐ μὴ ποιήσῃ—καὶ μακροθυμεῖ ἐπʼ αὐτοῖς; we come near having οὐ μή in a question with the present indicative as well as with the aorist subjunctive. In a question like μὴ οὐκ ἔχομεν; (1 Cor. 9:4) οὐ is the negative of the verb, while μή is the negative of the sentence. Cf. Ro. 10:18, 19. In 1 Cor. 9:8 we have μή in one part of the question and οὐ in the other, μὴ κατὰ ἄνθρωπον ταῦτα λαλῶ, ἢ καὶ ὁ νόμος ταῦτα οὐ λέγει; In Mt. 22:17 (Lu. 20:22; Mk. 12:14) we have ἢ οὔ; as the alternative question, and Mark adds ἢ μή. Babbitt² holds that “οὐ is used in questions of fact, while in other questions (e.g. questions of possibility) μή is used.” I doubt the correctness of this interpretation.

In declarative sentences the position of οὐ is to be noted when for emphasis or contrast it comes first. Cf. οὐ and ἀλλά in Ro. 9:8. So οὐ γάρ—ἀλλʼ ὄ in 7:15. In 7:18 f. note οὔ• οὐ side by side. Cf. also position of οὐ in Ac. 1:5; 2:15; Ro. 11:18 (οὐ σύ—ἀλλά). So ἀλλʼ οὐκ ἐγώ in 1 Cor. 6:12.

⁴ Cf. Das indoger. Vokal-System, p. 191.
⁵ Cf. Gildersl., Am. Jour. of Philol., XVIII, pp. 4, 123 f.; Horton-Smith, ib., pp. 43 ff.; Brug., Griech. Gr., p. 528.
⁶ The Negatives of the Indo-Europ. Lang., 1896. Cf. Delbrück, Grundr., IV, p. 519.
⁷ But Draeger (Hist. Synt., p. 133) says that this connection with the Lat. haud cannot be shown.
¹ Farrar, Gk. Synt., p. 182; Jann., Hist. Gk. Gr., p. 425.
² Monro, Hom. Gr., p. 259.
³ Hist. Gk. Gr., p. 426.
⁴ Thumb, Handb., p. 194 f.; Jebb, in V. and D., p. 339.
⁵ Gr. of the N. T. Gk., Thayer’s Transl., p. 344.
⁶ Cf. Thouvemin, Les Négations dans le N. T., Revue de Philol., 1894, p. 229.
⁷ Hist. Gk. Gr., p. 427.
⁸ Gr. of N. T. Gk., p. 253.
⁹ Synt. Forsch., IV, p. 147.
¹ Gr. N. T. Gr., p. 254.
¹ W.-Th., p. 511.
² Harv., Stu. in Class. Philol., 1901, The Use of Μή in Questions, p. 307.

The general pattern seems to be this: the more Robertson says on a topic, the more unanswered questions appear. To his credit, Robertson, in terms of actual descriptive content (as opposed to simply observations of data points), is far superior to BDF. At the same time, Robertson has a tendency to trade useful descriptive statements for rhetorical flourishes that are not overtly meaningful. Consider: “Here the negative οὐ is universal with the indicative in declarative sentences. The force of οὐ (οὐκ before vowels, οὐχ before aspirate) is sometimes very powerful, like the heavy thud of a blow.” If οὐ is essentially universal in indicative declarative sentences, then is it οὐ in and of itself that is so “very powerful, like the heavy thud of a blow”? Could it simply be the larger context of the negation? In the case of this particular quote here, I am not convinced that the references backing up the claim. They all seem to be simply negated statements of fact.

This is, in fact, a common problem. Later on Robertson writes: “The form οὐχί is a bit sharper in tone. Cf. Mt. 13:27; Lu. 12:6” and also, “There is a tone of impatient indignation in the use of οὐ in Ac. 13:10, οὐ παύσῃ διαστρέφων τὰς ὁδοὺς τοῦ κυρίου τὰς εὐθείας.” In these cases, the claims about the tone of the statements (whether “sharper” or “impatient indignation” are ambiguous. Is Robertson making a claim about the negator? Or is he making a claim about the context in which the negator is used? Whichever it is, we are given no account for it one way or the other.[3]

On other occasions, Robertson’s discussions are more compelling.[4] Consider the following portion:

In the N. T. οὐ alone occurs with the future indicative used as a prohibition, though the classic idiom sometimes had μή. Cf. οὐ φονεύσεις (Mt. 5:21); οὐκ ἔσεσθε ὡς οἱ ὑποκριταί (6:5), etc. Still, Blass¹ quotes μηδένα μισήσετε in Clem., Hom., III, 69. The volitive subjective nature of this construction well suits μή, but οὐ is more emphatic and suits the indicative. In Mt. 16:22, οὐ μὴ ἔσται σοι τοῦτο, we have οὐ μή in the prohibitive sense.

This section here, while still not exactly clear in its expression, has a statement of an observable facts, but also provides a motivation for those facts: μή is motivated in Clem., Hom., III, 69 by the volitional nature of the construction, while οὐ is used for future indicative that are prohibitions. Robertson does not say it, but we could extrapolate here that he perhaps believes οὐ prefers deontic modal contexts, while μή prefers epistemic modal contexts. This is a common issue: often his descriptive statements require guessing as to his meaning. Trying to reinterpret Robertson’s statements in contemporary linguistic terminology is often quite difficult. Occasionally while reading Robertson (and in fact, many old grammars), I feel a bit like Blinkin…

This situation is also the case in the following quote:

“In a question like μὴ οὐκ ἔχομεν; (1 Cor. 9:4) οὐ is the negative of the verb, while μή is the negative of the sentence. Cf. Ro. 10:18, 19. In 1 Cor. 9:8 we have μή in one part of the question and οὐ in the other, μὴ κατὰ ἄνθρωπον ταῦτα λαλῶ, ἢ καὶ ὁ νόμος ταῦτα οὐ λέγει.”

This appears to be an observation about to the scope of the negation? But what is the actual claim being made? Is Roberson saying that there a consistent division of the scope of negation between οὐ and μή in general (where μή is used for sentence negation and οὐ is not)? Or is he only talking about negation scope situations where the two negators appear together (μὴ οὐ)? And if the latter, is it just true for questions or true in general? Or is this simply a one off instance that Robertson thought worth noting in passing? It is not even clear how such a claim could be tested, especially if it is limited solely to instances of οὐ and μή together in polar questions.

Conclusion

Robertson is still descriptively superior to BDF. He certainly provides more meaningful information. Nevertheless, the fact that Robertson’s grammar was written in an era when formal or technical terminology was just nascent for grammar/linguistics causes significant problems. It often feels like while Robertson clearly knows what he wants to communicate conceptually, he does not know how to communicate it. This leaves us as readers with fairly common subjective statements about meaning with little to no real substance or justification. The formalization of grammatical terminology at this point in history only existed for a handful of grammatical categories and for the rest, grammarians were left grasping for the right words to communicate.

---

[1] That’s the edition that Logos Bible Software digitized. It’s the 3^th edition. I also have a physical 1923 4^th edition printed by Hodder & Stoughton. The two editions are essentially identical in their text, so far as I have found.

[2] Note indented footnotes below are the original footnotes & numbering from Robertson. Each page begins afresh with footnote 1. For convenience, my own footnotes are formatted differently than Robertson’s.

⁴ Cf. Das indoger. Vokal-System, p. 191.

⁵ Cf. Gildersl., Am. Jour. of Philol., XVIII, pp. 4, 123 f.; Horton-Smith, ib., pp. 43 ff.; Brug., Griech. Gr., p. 528.

⁶ The Negatives of the Indo-Europ. Lang., 1896. Cf. Delbrück, Grundr., IV, p. 519.

⁷ But Draeger (Hist. Synt., p. 133) says that this connection with the Lat. haud cannot be shown.

¹ Farrar, Gk. Synt., p. 182; Jann., Hist. Gk. Gr., p. 425.

² Monro, Hom. Gr., p. 259.

³ Hist. Gk. Gr., p. 426.

⁴ Thumb, Handb., p. 194 f.; Jebb, in V. and D., p. 339.

⁵ Gr. of the N. T. Gk., Thayer’s Transl., p. 344.

⁶ Cf. Thouvemin, Les Négations dans le N. T., Revue de Philol., 1894, p. 229.

⁷ Hist. Gk. Gr., p. 427.

⁸ Gr. of N. T. Gk., p. 253.

⁹ Synt. Forsch., IV, p. 147.

¹ Gr. N. T. Gr., p. 254.

¹ W.-Th., p. 511.

² Harv., Stu. in Class. Philol., 1901, The Use of Μή in Questions, p. 307.

[3] Not to mention the fact that the so-called tone is not necessarily itself even clear from the examples themselves!

[4] At least in appearance anyway—this discussion is not interested in evaluating the strength of the claims themselves, but only the clarity of the claims and the extent to which the grammar justifies or explains them.

¹ Gr. N. T. Gr., p. 25.