Introduction

In my previous nine papers, I have looked at the testimony of nine ancient exegetes—I do not like to call them ‘fathers’ or their study ‘patristics’ (Matt 23:9), but ‘Christians’, ‘churchmen’, ‘exegetes’, ‘scholars’, ‘experts’—concerning whether those familiar with Greek took the referent of the neuter demonstrative τοῦτο (‘this’) in Ephesians 2:8 as the preceding feminine noun in the genitive πίστις (‘faith’), located two words before, in spite of the fact that these two words do not agree in gender.

I found that eight of the nine were demonstrably familiar with Greek grammar, none of the nine say that feminine nouns cannot be the antecedent to a neuter demonstrative, eight of the nine took the antecedent of the neuter demonstrative as a feminine noun, seven of the nine took the antecedent to be πίστις (‘faith’, a feminine noun), one of the nine took the antecedent to be χάρις (‘grace’, also a feminine noun), one of the nine took the antecedent to be χάριτί ἐστε σεσῳσμένοι (the conceptual ‘salvation by grace’), and another one of the nine suggested that the conceptual ‘salvation by grace’ is a second and alternative interpretation after giving a first interpretation of ‘faith’ as the antecedent. As far as I know, these 9 are the extant commentators on this passage of Scripture who have applied their mind to this question arising from the passage.[1] All of this justifies Abraham Kuyper’s assertion that:

Nearly all the church fathers and almost all the theologians eminent for Greek scholarship judged that the words “it is the gift of God” refer to faith. 1. This was the exegesis, according to the ancient tradition […] 2. Of those that spoke the Greek language and were familiar with the peculiar Greek construction. 3. Of the Latin church fathers, who maintained close contact with the Greek world.[2]

These facts seem to suggest that the onus rests on those who say that such an understanding is grammatically incorrect to show it to be so, rather than merely assert what all who can read the Greek text can see—that the demonstrative and the preceding nouns do not agree in gender—and then think that the work of exegesis is done. The key text is Ephesians 2:8-10.

⁸For [it is] by grace[3] you[4] have been saved[5] through faith[6], and this [thing][7] [is] not from you[8], [it is] the gift[9] of God, ⁹not from works, so that no-one may boast. ¹⁰For we are his handiwork, being created in Christ Jesus for good works, which God prepared beforehand, so that in them we might walk. (my translation)

⁸Τῇ γὰρ χάριτί[10] ἐστε σεσῳσμένοι[11] διὰ πίστεως[12]· καὶ τοῦτο οὐκ ἐξ ὑμῶν, θεοῦ τὸ δῶρον· ⁹οὐκ ἐξ ἔργων, ἵνα μή τις καυχήσηται[13]. ¹⁰αὐτοῦ γάρ ἐσμεν ποίημα, κτισθέντες ἐν Χριστῷ Ἰησοῦ ἐπὶ ἔργοις ἀγαθοῖς οἷς προητοίμασεν ὁ θεός, ἵνα ἐν αὐτοῖς περιπατήσωμεν.[14]

This paper argues that it is not only an ancient exegetical view that the referent of ‘this’ (τοῦτο) in Ephesians 2:8-9 is ‘faith’ (πίστεως), even though they do not agree in gender and therefore faith is both οὐκ ἐξ ὑμῶν (‘not from ourselves’), and θεοῦ τὸ δῶρον·(‘the gift of God’), and it is also οὐκ ἐξ ἔργων, ἵνα μή τις καυχήσηται (‘not from works so that no one can boast’), but also that this ancient exegetical view is quite acceptable in terms of what we know about ancient Greek grammar.

Moreover, in this paper, I seek to do more than to just show the evidence for the existence of such a rule from the fact of the ancient Greek and Latin exegetes approach to interpreting Ephesians 2:8-9. Here I want to find, in relation to the Greek demonstratives, a statement of the syntactical rule together with positive examples that illustrate the rule in other literature, preferably the New Testament, or in the LXX, or other Greek literature.

To put the matter another way, finding similar constructions in Greek literature prior to and contemporaneous with the writing of the New Testament is evidence that the demonstrative can refer back to an antecedent of a different gender. The majority ancient exegetical opinion concerning Ephesians 2:8-9 is evidence that the construction does mean what it can mean.

The Modern Consensus

The ‘gift of God’ on the ‘majority’ understanding is the concept ‘salvation by grace through faith’. This view has, of the ancient exegetes, at least implicitly, Marius Victorinus as an adherent, and Theophylact explicitly suggests it as an alternative to his first option, which follows the main stream of the Eastern exegesis, that ‘faith’ is the denominated gift.

Furthermore, it must be granted (against Kuyper), that though generally and from other passages Calvin holds that faith is and must be a divine gift, and he is clearly a monergist, when Calvin comes to consider Ephesians 2:8-9 in his commentary, he likewise holds that the conceptual antecedent is the correct understanding:

And here we must advert to a very common error in the interpretation of this passage. Many persons restrict the word gift to faith alone. But Paul is only repeating in other words the former sentiment. His meaning is, not that faith is the gift of God, but that salvation is given to us by God, or, that we obtain it by the gift of God.[15]

Notice what Calvin has added to the exegetical tradition we have evidenced thus far. Calvin has said something that none of the nine ancient exegetes have said. Calvin has asserted that the main ancient exegetical view is ‘a very common error’. Now, Marius Victorinus doesn’t say that the view Calvin criticizes is an error. He simply understands the passage in the way that Calvin prefers. And Theophylact does not consider the main Eastern view as an error: he in fact gives it as his first alternative.

The fact is that those modern exegetes who cite Theophylact almost always misrepresent his view, because they do not acknowledge that Theophylact gave the Chrysostom-Theodoret-Œcumenius-Jerome-Augustine-Fulgentius view as the incumbent interpretative option, as it was for Theophylact himself. The fact that so many of the ancient Greek and Latin exegetes support the view that ‘faith’ is the gift and not from ourselves in Ephesians 2:8-9 raises serious questions about Calvin’s claim, and that of the moderns that follow him, that it is an error at all.

It is true that ‘the great majority of modern commentators’[16] argue that the antecedent of neuter demonstrative τοῦτο is not πίστις (‘faith’), but to the whole concept, τῇ γὰρ χάριτί[17] ἐστε σεσῳσμένοι[18] διὰ πίστεως (salvation by grace through faith).[19] This view is so predominant amongst modern scholars and commentators that it is easier to speak about the exceptions among them—and none of these are modern.[20] As it is a current standard Greek Syntax for New Testament studies, I will give this extended quote from Wallace, who, after giving Ephesisans 2:8, says:

This is the most debated text in terms of the antecedent of the demonstrative pronoun, τοῦτο. The standard interpretations include (1) “grace” as antecedent, (2) “faith” as antecedents, (3) the concept of a grace-by-faith salvation as antecedent, and (4) having an adverbial force with no antecedent (“and especially”).

The first and second options suffer from the fact that τοῦτο is neuter while χάριτί and πίστεως are feminine. Some have argued that the gender shift causes no problem because (a) there are other examples in Greek literature in which a neuter demonstrative refers back to a noun of a different gender, and (b) the τοῦτο has been attracted to the gender of δῶρον, the predicate nominative. These two arguments need to be examined together.

While it is true that on rare occasions there is a gender shift between antecedent and pronoun, the pronoun is almost always caught between two nouns of different gender. One is the antecedent; the other is the predicate nom[inative]. In Acts 8:10, for example (οὗτός ἐστιν ἡ δύναμις τοῦ θεοῦ), the pronoun is masculine because its antecedent is masculine, even though the predicate nom[inative] is feminine. In Matt 13:38 inverse attraction takes place (the pronominal subject is attracted to the gender of the predicate nom[inative]): τὸ δὲ καλὸν σπέρμα οὗτοί εἰσιν οἱ υἱοὶ τῆς βασιλείας (“the good seed, these are the sons of the kingdom”). The construction in Eph 2:8, however, is not parallel because δῶρον is not the predicate nom[inative] of τοῦτο, but of the implied “it” in the following clause. On a grammatical level, then, it is doubtful that either “faith” or “grace” is the antecedent of τοῦτο.[21]

Wallace also considers the article by Countess.

In particular, note R H Countess, “Thank God for the Genitive!” JETS 12 (1969) 177-22. He lists three examples from Attic Greek, arguing that such a phenomenon occurs frequently in Greek literature (120). His approach has weaknesses, however, for not only does he cite no NT examples, but two of his classical illustrations are better seen as referring to a concept than to a noun. Further, the usage is not at all frequent in and every instance requires explanation.[22]

Wallace is quite right to point out that the failure of the antecedent and the pronoun to agree occurs ‘on rare occasions’. The relative paucity in the LXX and New Testament testifies to this, as we shall see. He is also right to show that the two issues needing to be distinguished and examined are (a) the other examples in Greek literature where a neuter demonstrative refers back to a noun of a different gender, and (b) that τοῦτο has been attracted to the gender of δῶρον, the predicate nominative.

It is a shame, however, that Wallace only had recourse to the article by Countess as his conversation partner and the three examples there cited, because Countess is only depending on Kuyper (as Countess makes clear) and he doesn’t even cite all of Kuyper’s examples. Moreover, there are also other 19^th century grammars with equally long lists of examples as Kuyper (of which the most relevant ones immediately related to the neuter τοῦτο referring to an antecedent are quoted and fully dealt with below). This does not mean it is a frequently found construction. It is not. But as we shall see, the Classical Greek corpus and the LXX certainly testifies to the existence of the construction.

The Grammatical Issues

Modern New Testament Greek grammars do not seem to cite the rule that the older Classical grammars have posited. My search for a statement of such a rule in relatively modern intermediate Biblical Greek syntax has so far been unsuccessful.

The rule Kuyper cites by quoting Dinant, that of attraction to what follows rather than agreement with the antecedent which precedes, is well known in relation to the relative pronoun. The rule is that the relative pronoun may be attracted by the gender to the gender of the following noun in apposition and epexegetical with it. But we are now speaking about demonstratives, not relative pronouns. Is there such a rule of Greek grammar as Dinant and Kuyper contend that applies to the Greek demonstrative? Or is there another rule that accounts for the exegesis we have seen in eight of the nine ancient exegetes?

The fact is that even the modern commentators acknowledge that there are examples of neuter pronouns such as τοῦτο taking up feminine nouns.[23]

An initial consideration for our enquiry is that οὗτός (‘this’) refers to the proximate person or thing, that which is ‘comparatively near at hand’ in the discourse material, the writer’s mind, or space and time—and that this can be contrasted with ἐκεῖνος (‘that’), which tends to have more remote references by comparison.[24] Thus it seems that the proximity of the possible referents of the demonstrative is a relevant factor in determining what is the demonstrative’s antecedent. This is based on the meaning of the lexeme itself. We are talking about ‘this’, a near thing, rather than ‘that’, something pointed out but a little further away.

It is also important to appreciate the following statements by Greek grammarians. There are three Classical grammars from the 19^th century that cite the rule—or a similar one—to which Kuyper refers. The first from the German Greek Grammar by Raphael Kühner, which Kuyper translates as follows:

Very common is the use of a neuter demonstrative pronoun to indicate an antecedent substantive of masculine or of feminine gender when the idea conveyed by that substantive is referred to in a general sense.[25]

The other 19th century grammars, whose authors also have a strong background of classical Greek, cite a similar rule, are English works. The second one that cites the rule, is, not surprisingly, that of Jelf’s grammar based on Kühner’s:

The neuter demonstrative also is joined with a masculine and feminine substantive when this expresses a general notion, as is most frequently the case in abstract substantives.[26]

The third is from Gildersleeve.

Neuter pronouns referring to masculine or feminine substantives. Not an exception to the rule of concord is the case of a neuter pronoun referring to a masculine or feminine substantive regarded as a thing. Cf. 126. In most of the instances cited, however, the reference is rather to some thought implied in the substantive than to the substantive itself.[27]

Gildersleeve’s rule applies to other pronouns also, for which he provides examples.[28]

Stating a supposed rule is one thing. Demonstrating it is another. What is the evidence in the Greek corpus for this rule?

As we turn to the classical material, the question might be raised whether this is sound methodologically. After all, the New Testament is commonly said to be an example of Koiné (‘common’ or ‘hellenistic’) Greek, while the classics were generally written in the Attic (or Athenian) dialect. Does this difference affect the conclusions we draw from the data? Is the Koiné Greek a different dialect with different rules?

Buttmann asserts that the Koiné ‘never can be considered as a particular dialect’ for it ‘continued in the main to be Attic, and hence Atticism is the principal object of every Greek grammar’.[29] In other words, when we look at the rules of the Attic speakers, we are looking at the progenitor to Koiné, of which Koiné is a type. Koiné Greek evolved from Attic Greek, and Attic Greek with a few adjustments became Koiné Greek. Testimony to this is born by the references to the LXX and the Greek New Testament scattered through LSJM, a classical Greek dictionary. Classical Attic Greek and New Testament Greek are not two separate languages or dialects, but share an organic unity.

Methodologically, ‘confirmation bias’ is a risk for any inductive process, and it is appropriate to set adequate controls. One such control is finding testimony in a secondary secondary source—either in a grammar, commentary, or English translation—that affirms my reading of a particular text. All of the examples below have been culled from secondary sources, the 19^th century grammars I’ve cited above and LSJ. So every instance has at least one secondary source citing it for authority. Thus I will adopt the following ranking for all of the instances I analyse.

★★★★ Cannot be anything but testimony to the rule.

★★★ Highly probable that this is testimony to the rule.

★★ Probable that this is testimony to the rule.

★ Possible that that this is testimony to the rule.

✪ The number of secondary source asserting that it is an example of the rule. (All of the examples from the Classical corpus below are derived from secondary sources. I have not searched TLG to find them. Such a study would be useful but beyond my scope.)

However, I am not the only with ‘confirmation bias’, for almost all the modern commentary writers or grammarians are committed to the ‘majority view’. I can only hypothesize as to why this change has occurred. The 19^th century Greek grammarians were extremely well-versed in the Classical corpus, as well as the LXX and the NT. By contrast, the grammars read by students of the NT are written by those whom I might call ‘super-pastors’—not Classicists, but Evangelical Christians who have learnt Greek generally as part of their theological education as a young adult in their mid to late 20s for the purpose of understanding the New Testament as the Word of God. As brilliant as they are in learning the rudiments of Greek for understanding the NT and having picked it up when they did, their corpus tends to be severely reduced to 27 books (many of them tiny). By contrast, the 19^th century grammars reflect a Classical education system, where Greek and Latin rudiments were learnt in primary and high school, and students studied the Classics first.[30]

Greek Classical Usage

The following 13 examples from the Classical corpus illustrate the rule that the above cited grammars have asserted. These 13 examples are gathered from LSJ, Kuyper, Jelfs, Gildersleeve, and Kühner, cited above. I take 6 to have no other explanation, 4 having a high probability, and 3 being possible but there are other more tenable explanations. I have not included every example from these sources, but only those I thought most appropriate to the task of illustrating this rule.

1. Herodotus (c. 484–c. 425 BC), The Histories, 3.82[31] (★★★★✪✪✪)

Here, Godley (1✪) correctly translates the neuter demonstrative τοῦτο as referring to the previous masculine noun μουνάρχου two words before. It cannot be a referent to a conceptual antecedent, because the speaker is choosing between three options (★★★★). It is the nearest proximate candidate, and that is decisive in determining which of the three nouns is the referent, though on the modern reasoning from New Testament grammars, the speaker should have chosen to use the masculine form of the demonstrative, οὗτος. Gildersleeve (2✪) regards τοῦτο as natural because of the τριῶν which precedes and the οὐδέν which follows. [36] Likewise, Jelf (3✪) cites this as an example of his rule above, and takes τοῦτο to be a reference to ‘μονάρχον εἶναι’.[37] This is a clear example of our rule.

2. Herodotus, The Histories, 4.23 (★★★★✪✪)

The first instance of τοῦτο in the passage above is cited in LSJ (1✪) in the entry for οὗτος for the proposition that ‘the neuter may also refer to a masculine or feminine noun’.[40] Likewise Godley (2✪) rightly takes the neuter demonstrative τοῦτο to be referring to the masculine noun καρπὸν, fruit. I don’t know what else the neuter demonstrative can refer to (★★★★). This confirms the rule.

3. Plato (c. 428–c. 347 BC), Protagoras, 352b (★★★★✪)

The question in this passage is clearly concerning knowledge (ἐπιστήμη), a feminine noun. The neuter demonstrative τοῦτό has this noun as its antecedent. There really doesn’t seem to be another alternative. This is a clear example of the rule ★★★★✪.

4. Plato, Protagoras, 357c[45] (★★★★✪✪✪)

It is clear that the translator of Plato, Lamb (1✪), here has correctly taken the neuter demonstrative τοῦτο (‘this’) as referring to the antecedent genitive feminine singular noun ἐπιστήμης (‘knowledge’), by replacing the demonstrative with the substantive noun in his translation. Furthermore, the following commentary (2✪) on Plato takes this view, which reads here ‘τοῦτο (sc. ἐπιστήμη, 357c)’, where ‘sc.’ abbreviates Latin contraction scilicet, here meaning ‘namely’ or ‘as if to say’.[56] This is a clear example of the rule.

5. Plato, Republic, 9.583e (★★★✪✪)

The translator of Plato’s Republic (✪) seems to have taken the antecedent of the neuter demonstrative τοῦτό to be the feminine τὴν ἡσυχίαν translated as ‘quietude’ by translating it with the English demonstrative ‘this’ in the attributive position, ‘this quietude’. This rendering is supported by the proximity of the antecedent—the demonstrative immediately follows the articular noun. I take this as an example of the rule. However, it is possible that the antecedent might be the concept of ‘the intermediate state between the two [i.e. pleasure and pain]’, so I consider that it is highly probable that this is testimony to the rule.

6. Plato, Phaedo, 61a (★✪)

In the final clause of the last sentence the neuter demonstrative τοῦτο either refers back to the feminine singular noun φιλοσοφίας as antecedent, or it refers to a conceptual antecedent, being the adjectival participial phrase, οὔσης μεγίστης μουσικῆς which is in apposition to φιλοσοφίας and epexegetical to it. But each form in that phrase is also feminine. If φιλοσοφίας is the referent, this is a reference to the literal description of the ‘greatest music’ that Socrates had given himself in his life to working at. If οὔσης μεγίστης μουσικῆς is the referent, this too is in the feminine. It may well be that μουσικῆς is the antecedent, because later Socrates says that he understood his dream differently, and so started laboring at literal ‘music’. If both the single word and the adjectival participial phrase are the referent, we are moving toward a conceptual antecedent here, which would give an alternative explanation for the neuter τοῦτο. However, in this case, the use of the feminine demonstrative αὕτη would not have distinguished between either μουσικῆς or φιλοσοφίας as to which one was the antecedent. Given that both ‘philosophy’ and ‘being the greatest music’ are two different ways of referring to the same thing, if the demonstrative refers to one, it refers to the other. So I tend to think that this example provides a possible instance of testimony to the rule.

7. Plato, Theaetetus, 145d (★★★★✪✪)

In this example, that the antecedent of the neuter demonstrative τοῦτο is the feminine σοφίᾳ is confirmed in the course of the dialogue by Socrates himself, because he is directly asked by Theaetetus a clarifying question, in answer to which Socrates then says ‘ἡ σοφία’, which clarifies the referent of τοῦτο as being the antecedent lexeme in the dative (σοφίᾳ), which he had previously emphasized by placing it at the beginning of the sentence. The antecedent really cannot be anything else. This is a very clear example of the rule.

8. Xenophon (c. 430–354 BC), Hiero, 9.9 (★★★✪✪)

The English translator of Xenophon (1✪) has correctly taken the feminine noun ἐμπορία as the antecedent to the neuter demonstrative τοῦτο—located in a participial phrase literally translated, ‘honouring the one doing this [i.e. commerce] best’— and has translated the demonstrative by using an English synonym for ‘commerce’ (i.e. ‘business’), which he had used to translate the Greek word ἐμπορία previously in the sentence. This has a high probability that this illustrates the rule.

9. Demosthenes (384-322 BC), Against Aphobus for Phanos 3 [Oration 29 c. 362/361 BC] [11] (★★★★✪)

The antecedent of the neuter demonstrative τούτου is feminine μαρτυρίας. It cannot be a reference to the ‘whole contest centering about the deposition’, because the contest is not at issue in terms of its truth or falsity, but the ‘deposition’ (★★★★). It confirms the existence of the rule.

10. Demosthenes, Orations 2: Olynthiac 2 [15] (★★★✪)

A literal translation of the phrase for consideration, ἀλλ᾽ ὁ μὲν δόξης ἐπιθυμεῖ καὶ τοῦτ᾽ ἐζήλωκε, is ‘Instead he [Philip of Macedeon] desires glory, and has zealously pursued this’. One commentator takes the antecedent of τοῦτο in the phrase τοῦτ᾽ ἐζήλωκεν (‘has made this his passion’) as τὸ δόξης ἐπιθυμεῖν (‘to desire glory’)[93], and thus is a conceptual antecedent. Another takes the antecedent as δόξαν λάμβανειν (‘to take glory’).

But there are two problems with this view of the construction. The first is that there are two similar and perhaps synonymous verbs: present ἐπιθυμεῖ (I desire, lust) and perfect ἐζήλωκε (I zealously pursue, am jealous for). Because of this, to take the antecedent to be a conceptual one has the intended result that Phillip of Macedon ‘passionately pursues’ (ἐζήλωκεν) the concept of ‘his desire for glory’ (ὁ […] δόξης ἐπιθυμεῖ). This redundancy of a passion for ‘a desire for glory’ rather than simply ‘a passion for glory’ strongly suggests that the antecedent is simply the feminine noun δόξης. The second is that in the subsequent text, the focus is actually on the glory, δόξαν, which Philip of Macedon is seeking beyond any other king before him. So I take this as confirming the rule at the level of high probability (★★★).

11. Demosthenes, Against Leptinesm, 20 [140] (★★★✪)

The nominative masculine noun, ὁ φθόνος, is the antecedent of the neuter demonstrative τοῦτο in the adjectival participial clause ὁ τοῦτο πεπονθώς. This likewise is Gildersleeve’s opinion (✪), who takes the demonstrative τοῦτο as referencing the underlying verbal concept τὸ φθονεῖν (‘to envy’) which is expressed by the noun form ὁ φθόνος. The antecedent being ὁ φθόνος is confirmed by the subsequent discourse, wherein to seem to be envious, φθονερὰ [adjective] δοκεῖν εἶναι, is again criticized. Whether this is an example of the neuter demonstrative referencing the masculine noun itself ὁ φθόνος or the underlying verbal idea expressed as a masculine noun, either way the neuter demonstrative takes as antecedent the masculine noun, and so this example clearly illustrates the rule at the level of high probability (★★★).

12. Isocrates (436–338 BC), Archidamus, 6[15] (★✪)

Either the phrase λόγους ἀγαπήσας (‘love of words’) or the noun λόγους (‘words’) is a reference to oratory, or the ability to speak well to motivate others. It would seem that the antecedent is the idea or concept expressed by either the single word or the phrase as an idiom.

Gildersleeve (✪) takes περὶ τοῦτο as a more general reference to ‘this sort of thing’ and the reference is ‘rather to some thought implied in the substantive than to the substantive itself’.[98] It may be that the noun λόγους conveys this, or it may be the phrase λόγους ἀγαπήσας. While this example proves the broader rule of Gildersleeve, it is not a clear example of the demonstrative referring to the ‘substantive itself’, though it is possible that it is (★).

13. Xenophon, Anabasis (c. 370 BC), 1.5.10 (★✪)

This example is taken from LSJ, 1276 (✪) as an example for the rule that neuter τοῦτο can refer to either a masculine or feminine antecedent. Here the demonstrative possibly refers to the genitive feminine noun μελίνης, for it is more likely that this is that which is very abundant in the country as the thing which is grown (rather than milled) in the country. However, it is probable that the antecedent could refer to the conceptual antecedent of ‘grain of millet’, a specific type of grain. However, if the grain was the focus, the masculine could have been used. Following LSJ, it possibly is an example of the rule, but it is more probable that the antecedent is conceptual.

Septuagint (LXX: 3^rd Century BC) Examples

The Septuagint or LXX, the 3^rd Century BC translation of the Hebrew Scriptures, which includes some other inter-testamental books, is an example of Koine Greek just as the NT is.

The Septuagint is written in Koine, that is, the common Greek of the Hellenstic age, a form of the language that had developed from the Classical Greek of fifth-century Athens. For studnets of the Greek language during the Hellenistic period, the Septuagint is a major source of information.[103]

In fact, it is arguable that the LXX more than any other book constitutes the background to the world-wide mission of Jesus and his disciples that the NT represents.

I gathered the following from browsing the electronic search results of the form τοῦτο across the LXX in the BibleWorks database. While there are many examples of conceptual antecedents of the form τοῦτο in the LXX, I include the following as potential examples of the rule, with the ranking applied. In the 12 passages below, 1 passage certainly has an example of the rule which could not be explained by anything other than the rule, 5 contain highly probable instances of the rule, 2 are probable instances, and 2 are possible examples. Two do not prove the rule, but are otherwise interesting in the light of Ephesians 2:8-9.

1. Genesis 2:23 LXX (★★★)

It is unlikely that the demonstrative τοῦτο refers to the closest neuter noun, being plural dative τοῖς θηρίοις in verse 20. Rather, the most likely antecedent is γυναῖκα and αὐτὴν, both of which refer to Eve.

The Hebrew underlying the Greek might provide us with a guide to understanding the LXX. The Hebrew zot haphaam (literally, ‘this, the stroke’, ‘this, the hit’), which corresponds to Greek τοῦτο νῦν (‘this now’). Skinner argues that the Hebrew should be rendered ‘this, this time’, with the demonstrative זאת as subject of the sentence and the article which is prefixed to the noun (הפעם) being given full demonstrative force.[104] Keil and Delitizsch take הפעם as ‘lit. this time’, allowing the demonstrative זאת to refer to the antecedent ‘the woman’. John King says likewise that each of the words is meant to be emphatic: ‘This living creature (זאת) which at the present time (הפעם) passes before me, is the companion which I need’.[105] So too does the HCSB (‘This one, at last, is bone of my bones’) and NET (‘This one at last is…’).

If the LXX is intending to accurately represent this understanding of the Hebrew, then τοῦτο indeed does refer back to feminine pronoun αὐτὴν (‘her’), which itself refers back to feminine γυναῖκα (‘the woman’), and the adverb of time νῦν refers to the present instance. This is highly probable then that it is an example of the construction (★★★).

2. Genesis 14:17 LXX (★★★★)

This is an example of the rule, as τοῦτο refers back to feminine singular τὴν κοιλάδα τὴν Σαυη (the valley of Saby). The demonstrative cannot refer to anything else as shown by the predicate of the equitative clause, which explains what ‘the valley of Saby’ also is called or which distinguishes it: τοῦτο ἦν τὸ πεδίον βασιλέως || ‘the plain of the kings’ (★★★★). Notice also that τοῦτο as subject also agrees with its predicate neuter τὸ πεδίον βασιλέως (the plain of the kings) in the clause to which it belongs. Thus it has two points of contact with the reading being argued for regarding Ephesians 2:8-9: (1) the neuter demonstrative has a feminine antecedent, and (2) the neuter demonstrative as subject has potentially been attracted by the neuter predicate.

3. Genesis 28:17 LXX (★)

The decision required here is whether τοῦτο is retrospective or prospective. The comparable construction strongly suggests the referent is prospective—and the demonstrative does not agree with either the masculine or feminine nouns to which it could potentially refer. But at one level it doesn’t matter, because the possible antecedent or forward referents all refer to the same thing, viz, the place where Jacob fell asleep and saw his vision. At another level, it would appear that the neuter τοῦτο is ‘caught between’ both masculine and feminine referents, and in that case, it makes sense. However, it may be that the referent is an underlying concept, being ‘Bethel’. None of those possible referents agree with the neuter demonstrative although, they all refer to the same thing, Bethel. However, the passage also shows that the translators can use the demonstrative such that it agrees with the gender of the head noun when it is in a subject-predicate construction or it is in the attributive position. I suggest this is possibly an example of the rule (★).

4. Leviticus 11:4-7 LXX (x 4) (★★★)

In Leviticus 11:4-7, there are four examples of neuter demonstrative τοῦτο referring back to a masculine antecedent.

The four animals whose names take masculine grammatical gender and referred to by neuter τοῦτο are τὸν κάμηλον (‘the camel’) in verse 4, τὸν δασύποδα (‘the rabbit’) in verse 5, τὸν χοιρογρύλλιον (‘the hare’) in verse 6, and τὸν ὗν (‘the swine’) in verse 7. It is true that κτῆνος (‘beasts’) takes the neuter gender, but in reference to unclean beasts as a class is in the plural (vv. 2, 4), not in the singular.

The neuter singular might reflect the fact that the class as a whole is neuter and thus be a variety of the constructio ad sensum[106], but this does not do away with the fact that the clear referent of the neuter demonstrative is a masculine noun. Actual gender might displace grammatical gender, as in the case of neuter beasts. But if the principle of constructio ad sensum explains this passage and the use of the neuter demonstrative to refer to beasts whose names are masculine nouns, it can also apply to ‘faith’, an abstract feminine noun describing a human attitude or response. Such a proposition would then apply no less to Leviticus 11:4-7 than to Ephesians 2:8: as ‘faith’ as an abstract noun doesn’t have an actual gender, but only a grammatical gender. Moreover, if the neuter demonstrative is explained by the gender of the predicate nominatives in each of the clauses, then that can only serve to strengthen my argument, suggesting that attraction of the demonstrative subject to the neuter singular predicate is part of the explanation for Ephesians 2:8-9 also. It thus demonstrates the rule as a probable example (★★★).

5. Leviticus 25:34 LXX

The neuter reference here is either to the fields attached to the cities (rather than the cities themselves), or to the concept of the agricultural lands not able to be sold. Here I think the antecedent is the conceptual one because (1) the verbaI idea of πραθήσονται is prominent and part of the rationale, (2) the demonstrative is singular while the nouns are plural, suggesting the conceptual antecedent. So this is not an example of our rule.

6. Leviticus 7 (x 12)

Twelve records of gifts from each tribe of Israel in substantially the same form appear in Numbers 7. We will simply look at one.

The antecedent of the demonstrative is a concept—the sacrifice of the peace-offering and all of its constituent components. It is unlikely to refer to the feminine noun θυσίαν as the point of the passage is to outline what constitutes the sacrifice. However, it is interesting to see the construction τοῦτο τὸ δῶρον, which is fitting for two reasons: (1) it is a proper referent for the conceptual antecedent; and (2) it also happens to agree with τὸ δῶρον in number and gender. But it cannot be posited as evidence for the rule.

7. Deuteronomy 14:7-8 LXX (x 3) (★★★)

The three instances in verse 8 above are parallel to Leviticus 11:4-7 above, and the explanation is the same as it was for that passage. The neuter singular is best explained by the constructio ad sensum—and the referents of the three neuter demonstratives is a masculine noun. It is probably an example of the rule on the same basis as Leviticus 11:4-7 is (★★★).

8. 1 Chronicles 22:1 LXX (★★)

The antecedent here is either ἅλῳ, qualified by the then owner of the threshing floor in the genitive, or the concept of the same—what Gildersleeve describes as ‘some thought implied in the substantive’. The dictionary form of the feminine noun is ἅλως, which seems to be the antecedent of both the masculine (οὗτός) and neuter (τοῦτο) demonstratives. This verse also shows the principle of attraction of the demonstrative to the gender of the predicate. It is a probable instance of the rule. (★★)

9. 2 Chronicles 1:10-12 LXX (★)

It is more likely that this is an example of a conceptual antecedent, the referent being everything that was in Solomon’s heart, and not just ‘σοφίαν καὶ σύνεσιν’. However, given the prevalence of the couplet in God’s direct speech in response, and the fact that they are both feminine in gender, it may be that the demonstrative was referring to the couplet. This possibility suggests perhaps it is an example of the rule (★).

10. Zechariah 3:1-3 LXX (★★)

The Hebrew of the MT of verse 2 reads the subject of the sentence as masculine demonstrative זֶה (zeh, ‘this’) and the predicate as masculine noun אוּד (’ūḏ, fire brand, stick to stir the fire) in a verbless simple noun clause. There is no neuter form of the demonstrative in Hebrew. Normally, זֶה is prospective and refers to something that has not been introduced. This would explain the repetition of Joshua in verse 3, as the high priest shown in verse 1.

The text of the LXX of verse 2 has the predicate of the verbless clause as masculine singular δαλός ‘brand for the burning’, and the subject being neuter τοῦτο, which therefore has not been attracted to its predicate. The nature of the syntax of a verbless clause is not changed by the addition of adverb ὡς (‘as’, for which there is nothing in the Hebrew), but points out the metaphorical usage. But more germane to this investigation is that ὡς points out that the neuter demonstrative τοῦτο does not immediately refer to δαλός which follows it but something which preceded it: ‘this’ human being Joshua, mentioned in verse 1, is ‘as’ or ‘like’ a brand snatched from the fire. That is the only understanding that makes sense of the addition of the ὡς.

The antecedent again is most probably masculine ᾿Ιησοῦν τὸν ἱερέα τὸν μέγαν. However, the antecedent may be ᾿Ιησοῦν […] ἑστῶτα || ‘Jesus […] standing’. This is moving toward the conceptual, or what Gildersleeve describes as ‘some thought implied in the substantive’. Again, it hardly matters whether the demonstrative τοῦτο is prospective or retrospective, given that verses 1 and 3 describe the same person. For the retrospective antecedent in verse 1 is the introduction to the pericope where we read Καὶ ἔδειξέν μοι ᾿Ιησοῦν τὸν ἱερέα τὸν μέγαν ἑστῶτα—the emphasis on the ‘showing’ of Jesus/Joshua suggests that the demonstrative is retrospective rather than prospective.

I consider that there is a probability that this is an example of our rule, but it is possible that a conceptual antecedent is on view. (★★)

11. Isaiah 6:6-7 LXX (★★★)

While the Hebrew masculine demonstrative זֶה (zeh, ‘this’) in Isaiah 6:7 MT usually points forward, it really cannot be said to do that here—it must refer to that which has already been introduced in verse 2.

The Greek noun for coal is ἄνθραξ, a masculine noun, confirmed by the masculine relative pronoun which follows it and which anaphorically refers to it. This is undoubtedly an example of constructio ad sensum, but that does not mean it is not an example of the rule—it is a neuter demonstrative referring back to a noun with which it does not agree in gender when it clearly could. And the same could be said of Eph 2:8, that it is an example of constructio ad sensum for ‘faith’ as an abstract noun in reality has no gender, though the Greek form is grammatically feminine (★★★).

12. Ezekiel 16:49 LXX (★★★)

While this is not an example of an anaphoric but a kataphoric referent of the demonstrative, and a prospective not a retrospective τοῦτο, this is worthy of consideration. The postcedent reference here is to a feminine noun, ὑπερηφανία, arrogance. Despite the fact that it refers to a postcedent not antecedent, it confirms the rule—for a feminine demonstrative could certainly have been used (★★★). It is likely that the neuter noun τὸ ἀνόμημα in the predicate position had a role in attracting the gender of the demonstrative.

The New Testament

The New Testament is a relatively small corpus, even when compared to the extant koine literature (a word count of around 138,000 words). Moreover, there is a substantial element of overlap in the content and vocabulary—consider the Synoptic Gospels, for example. Nevertheless, we should expect to see some examples, certainly not as many as in the Classical and LXX corpora, but a commensurate proportion of instances. Wallace performs a search on καὶ τοῦτο, and of the 22 NT instances, regards only 1 (Phil 1:28) as ‘possible’, but prefers its taking of a conceptual antecedent as ‘probable’.[107]

It seems to me that apart from Ephesians 2:8-9, there is one instance of high probability, and five instances (four parallel), that illustrate the rule at the level of probability

1. Mark 14:22-24; cf. Matt 26:26-28; Luke 22:19-20; 1 Cor 11:23-24 (★★)

Compare Matt 26:26-28; Luke 22:19-20; 1 Cor 11:23-24, which at least with regards to this question present exactly the same issues.

The noun ἄρτον is masculine singular, and yet it appears to be the referent of neuter τοῦτό. Notice that what Jesus offers is not οὗτος ὁ ἄρτος but τοῦτο, this thing. It is possible that since a process occurs with the bread, viz, ἔκλασεν καὶ ἔδωκεν αὐτοῖς (he broke and he gave to them), that what is referred to is that which is given, not the whole of the ἄρτον, but a piece thereof, a fragment, which then might be impliedly the referent of the neuter demonstrative.

It is clear that in this case, the idea of a conceptual antecedent (ie. a fragment of bread the result of ἔκλασεν καὶ ἔδωκεν αὐτοῖς) is not really that different from Gildersleeves’ rule that ‘the reference is rather to some thought implied in the substantive than to the substantive itself’ (ie. a piece or part of the named whole ἄρτον). The different ways of categorizing what appears to be essential the same thing suggest that it is probably an example of the rule (★★).

2. 1 Corinthians 12:15-16 (★★)

This is pretty clearly an example of constructio ad sensum: the noun ὁ πούς, ‘the foot’, is grammatically masculine, but is referred to with the neuter demonstrative τοῦτο, almost certainly because it is a body part. The demonstrative also happens to concord with the gender of neuter σώματος, which may be an instance of attraction (★★).

3. Philippians 1:28-29 (★★★✪✪✪)

As mentioned above, Wallace regards this as possibly but not probably an example of our rule. The referent of the neuter demonstrative τοῦτο in verse 28 is either the feminine nominative noun ἔνδειξις,[108] the feminine noun in the genitive σωτηρίας[109] (neither of which agrees in gender with τοῦτο), or the entire preceding section from ἔνδειξις.[110] The first two understandings have strong support among particularly older commentators, and provide an illustration from within the Pauline corpus of an antecedent not agreeing with an anaphoric demonstrative in gender.

I think it is more likely that the antecedent is σωτηρίας. The first factor suggesting this identification of the antecedent is proximity, the two words being separated only by καὶ. The second factor is that the following explanation in verse 29 which commences ὅτι ὑμῖν ἐχαρίσθη […] confirms that Paul’s focus is on ὑμῶν […] σωτηρίας (your salvation) rather than including αὐτοῖς ἔνδειξις ἀπωλείας (to them a sign of destruction). This is shown by the second use of a 2^nd person plural pronoun (compare ὑμῶν in v. 28 with ὑμῖν in v. 29). That is, Paul’s explanation focuses on the ‘you’ but does not refer to the ‘them’. The third is the highly theological denotation of the lexeme σωτηρία, ‘salvation’, which could conceivably have prompted from Paul the reflections immediately following.

4. Luke 2:1-2

One would have expected a conceptual antecedent of the neuter demonstrative τοῦτο given that the cognate passive infinitive ἀπογράφεσθαι contains the concept or idea. However, we see Luke using the feminine form of the demonstrative αὕτη, no doubt on account and under the influence of the predicate noun ἀπογραφὴ. This suggests that that attraction may indeed be a factor.

Expository Considerations

The passage in context is Ephesians 2:1-10:

vv. 1-7. The trajectory of Ephesians 2:1-7 is not that the human will is merely impaired and able to co-operate with grace. Humans under sin are not sick or impaired but dead and enslaved:

vv. 1, 5: you being dead in your transgressions and sins (ὑμᾶς ὄντας νεκροὺς τοῖς παραπτώμασιν καὶ ταῖς ἁμαρτίαις ὑμῶν). Cf. v. 5: us being dead in transgressions (ὄντας[140] ἡμᾶς νεκροὺς τοῖς παραπτώμασιν).

v. 2: you once walked according to the age of this world, according to the ruler of the authority of the air, the spirit now working in the sons of disobedience (ποτε περιεπατήσατε κατὰ τὸν αἰῶνα τοῦ κόσμου τούτου, κατὰ τὸν ἄρχοντα τῆς ἐξουσίας τοῦ ἀέρος, τοῦ πνεύματος τοῦ νῦν ἐνεργοῦντος ἐν τοῖς υἱοῖς τῆς ἀπειθείας).

v. 3: we all once lived in the desires of our flesh, doing the wills of the flesh and mind (ἀνεστράφημέν ποτε ἐν ταῖς ἐπιθυμίαις τῆς σαρκὸς ἡμῶν ποιοῦντες τὰ θελήματα τῆς σαρκὸς καὶ τῶν διανοιῶν).

Humans in the state they all once endured are dead, and subject to the ruler of the air and their own sinful will. This implies and indeed requires that God must revivify them and enable faith. A special work of God is required to make them alive and enable them to believe. They cannot make the decision to trust in Christ unless they are made alive.

v. 5: συνεζωοποίησεν τῷ Χριστῷ—χάριτί ἐστε σεσῳσμένοι (‘He must us alive with Christ Jesus—by grace you have been saved’). In verse 5, God made us alive with Christ (v. 5a: συνεζωοποίησεν τῷ Χριστῷ), and this phrase stands in apposition and thus is said to explain and is explained by the parenthetic phrase phrase ‘by grace you have been saved’ (v 5b: χάριτί ἐστε σεσῳσμένοι). The verb points to being given life with Christ, in the context of being dead (vv. 1, 5), and this conditions and is conditioned by ‘by grace you have been saved’. If the action of the verb συζωοποιέω points to prevenient activating grace which enables faith, this strengthens the argument of this paper—that faith as a gift entails the regenerating and prevenient enabling of God. It is quite conceivable that ‘to be made alive with Christ’ for Paul is explained by ‘salvation by grace’: ‘making alive in Christ’ is in the same conceptual field as ‘salvation by grace’.

vv. 5-6: συνεζωοποίησεν τῷ Χριστῷ […] καὶ συνήγειρεν καὶ συνεκάθισεν ἐν τοῖς ἐπουρανίοις ἐν Χριστῷ Ἰησοῦ. ‘He [God] made us alive together with Christ […] and raised and sat us in the heavenly regions in Christ Jesus.’ The meanings of the concepts of making us alive with Christ, raising us with him, and seating us in the heavenlies with him, are not immediately obvious. They must be looked at together.

‘Being made alive with Christ’ (v. 5: συνεζωοποίησεν) might refer to (a) an act of God in contrast with being dead in sin (vv. 1, 5) which enables faith (i.e. ‘regeneration’ or special ‘prevenient grace’); (b) an act of God in contrast with being enslaved to sin and Satan which implants virtue in the soul, contrary to ποιοῦντες τὰ θελήματα τῆς σαρκὸς καὶ τῶν διανοιῶν (v. 3: i.e. ‘sanctification’, particularly ‘vivification’ and ‘living the risen life now’); (c) a judicial and relational act of God in which the sinner is transferred from the realm of Satan to the Kingdom of Christ in the same way as we are ‘already seated in the heavenlies with Christ’ (v. 6: i.e. positional sanctification, akin to justification); (d) related to (c), the resurrection of Christ applied to believers prior to the general resurrection as a realized eschatological event for and on behalf of his people—the general resurrection brought into the present and anticipated by the subjective appropriation of salvation through faith; (e) Any combination of the above. Possibly, all of the above are to be included in the connotation of the term.

‘Raised together with Christ’ (v. 6: συνήγειρεν). Clearly the general resurrection has not occurred. The word may nevertheless refer to all the options above. It differs from the above in its accent on ‘standing’ and its particular emphasis on Christ’s bodily resurrection on the first easter day, and that event signalling the in-breaking of the age of resurrection.

‘Seated with Christ in the heavenlies’ (v. 6: συνεκάθισεν ἐν τοῖς ἐπουρανίοις ἐν Χριστῷ Ἰησοῦ). Of the three verbs used, this is the most likely to be viewed as realized eschatology and positional. Origen, however, took it in a different way:

Origen: The one who understands the phrase, ‘raised us up together and seated us together’, rather simply will say it has been said of God’s foreknowledge and predetermination, as though what will be has already occurred.[141] But one who perceives that the kingdom of Christ is spiritual will not hesitate to say that just as the one who is already holy is not in the flesh even if he may be said to be in the flesh by those who are rather simple, so he is not on the earth even if he may be seen to be on the earth so far as sense perception is concerned. For he who is in the spirit is not on the earth, and no one of those who are in the heavenly places as ‘in the flesh but’ are already ‘in the spirit’ (Rom 8:9).

John of Damascus, however, approaches the ‘positional’ view.

John of Damascus: We are the flesh of Christ, and limbs and body, and the Father has given Him headship. Already we are in heaven, in so far as we are in Christ, and we will have the enjoyment of these good things at the last, when the good things in Christ go on ahead to be displayed, now being unspoken and innumerable.[142]

I suspect John’s view here is to be preferred over Origen’s.

vv. 5, 8: χάριτί ἐστε σεσῳσμένοι (x 2) (by grace you have been saved). The parallelism between verse 5 and verse 8 is suggestive. Verse 5 uses the same phraseology as verse 8.

The phrase ‘by grace you are saved’ is used to explain that which precedes it—in v. 5 being the making alive with Christ, and in v. 8 the statement that the riches of God’s grace are demonstrated by the act of salvation. In v. 5, the nature of the explanation is expressed by simple apposition, in v. 8 by the use of the explanatory and backward referring γὰρ. In v. 6, the narrative is carried forward by the two new verbs with connective καὶ. However, in v. 8, the explanation is carried forward by the introduction of a new element—of its very nature a human element—‘through faith’. The instrument of subjective human appropriation is mentioned. This is new.

It is then fitting and appropriate that the new element, ‘faith’, be further explicated by the retrospective τοῦτο. To put it another way, in the discourse, it would make less sense for Paul to be explicating χάριτί ἐστε σεσῳσμένοι in verse 8, as he has already done this in verses 5-7. If we want to know what ‘by grace you are saved’ means, verses 5-7 has abundant synonyms and antonyms to give us the necessary contextual clues. However, the new element in the discourse, ‘διὰ πίστεως’ is that which is in most need of explanation given what Paul has already spoken about.

v. 8: χάριτί […] σεσῳσμένοι […] πίστεως. None of the individual words—χάρις (‘grace’), σεσῳσμένοι (‘having been saved’), πίστις (‘faith’)—agree with τοῦτο (‘this’) in gender. The noun χάρις (‘grace’) is feminine singular, σεσῳσμένοι (‘saved’) is a masculine plural participle in a periphrastic construction, and πίστις (‘faith’) is feminine singular.

If one was to consider the individual words of verse 8, of all of them, πίστις (‘faith’) is arguably the most likely antecedent of any of them. It is the closest word. Moreover, there is a logical argument in favour of this that Ellicott has brought out.

Still it may be said that the clause καὶ τοῦτο was suggested by the mention of the subjective medium πίστις, which might be thought to imply some independent action on the part of the subject (comp. Theodoret); to prevent even this supposition, the Apostle has recourse to language still more rigorously exclusive.[143]

Of each of the three words, πίστις (‘faith’) is the only one that could conceivably be thought of as ‘from yourselves’. There would be no need for Paul to have spoken about ‘grace’ being not from us. Indeed, consider:

John of Damascus: Grace is the thing which lies with God; faith is the thing which lies with us. (Χάρις, τὸ παρὰ θεοῦ, πίστις, τὸ παρ᾽ ἡμῶν).[144]

Œcumenius: On the one hand faith is from yourselves, but the cause of it is God. (Ἐξ ὑμῶν μὲν ἡ πίστις · ἀλλὰ καὶ αὐτῆς πρὸφασις ὁ θεός).[145]

Nor is it easily conceivable that being ‘saved’ conceptually could be thought to be from us—those very concepts suggest that someone outside of us is doing the action. But there is a possibility that someone might think that faith originates from us. Indeed, that is the key to interpreting the verse in the Eastern tradition. This is because the Easterns adopt what cannot be phenomenologically denied, that ‘faith’ is from ourselves, and yet still see that Paul is saying something about this particular phenomenological reality.

Chrysostom: ‘he adds also our part in the work (ἔθηκε καὶ τὰ ἡμῶν), and yet again cancels it, and adds, ‘And that not of ourselves.’ (καὶ πάλιν αὐτὸ ἀνεῖλε, καί φησι · Καὶ τοῦτο οὐκ ἐξ ἡμῶν.) Neither is faith, he means, ‘of ourselves.’ (Οὐδὲ ἡ πίστις, φησὶν, εξ ἡμῶν.)[146]

Jerome: And this faith itself is not from yourselves (Et haec ipsa fides non est ex vobis), but from him that called you (sed ex eo qui vocavit vos). This [he] also [says], lest perhaps the secret thought might creep up on us, that if we have not been saved by our own works (per opera nostra), at any rate we have actually been saved by faith (per fidem), and in this way by a different method we are saved by ourselves. Accordingly, he [Paul], said in addition, and asserted, that even faith is not by our own will (non nostrae voluntatis esse) but is God’s gift (sed Dei muneris).[147]

In fact, many people do think that faith is what we bring to salvation. Pelagius did previously, John of Damascus did, and Arminians do today. Indeed, the semi-Pelagians also believe we contribute some element to our regeneration.

This paper has presented what I believe to be a strong argument for the referent of the demonstrative to be ‘faith’. However, even in the case of the conceptual antecedent, and that ‘salvation by grace through faith’ is the gift, that does not mean that faith is not a gift—even on the strength of the passage. It is only on the basis of an assumption that faith cannot be the gift because it is ours phenomenologically that it must be excluded. One could argue that as God gives the whole, so also God gives the parts: he gives each of the three separate items, and each item does not find its ultimate origin in the human will. And indeed, as I have shown with the statement from Ellicott, the one element that does not immediately suggest of its very nature that it is from God is ‘faith’. So even if the conceptual antecedent were correct, the only reason to think that ‘faith’ should be excluded is based on the presupposition that it is unfair or unworthy or unfitting of God to give faith, either as an individual element, or as part of a conceptual package.

But if we reason from the other direction, if ‘faith’ is the antecedent and therefore the gift of God, so too is ‘grace’ and ‘salvation’. It cannot be otherwise.

Œcumenius: ‘So that, for us to believe [is the] gift of God, and to be saved through faith [is the] gift of God’ (Ὥστε τὸ πιστεῦσαι ἡμάς, δῶρον θεοῦ, καὶ τὸ διὰ πίστεως σωθῆναι δῶρον θεοῦ).[148]

Œcumenius’ first statement re-iterates the point that belief itself is the gift of God. The second statement makes an extension based on the point he has just made, that faith would not be sufficient to save except that God has willed salvation through faith. So not only is faith itself a gift from God, but the scheme of salvation by faith is a gift from God. I doubt it is possible to hold that ‘faith’ is particularly denominated as ‘the gift of God’ in Ephesians 2:8-9, but that consequently ‘salvation’ and ‘grace’ are not gifts of God. If faith is the gift of God, it follows that salvation is the gift.

v. 8: σεσῳσμένοι (‘have been saved’). The question to be answered here is ‘what is the nature of the “salvation” conceived’. Paul in the letter to Ephesians does not use the category of ‘justification’. But in other places he does use ‘salvation’ to encompass and stand as a broad synonym for ‘justification’ (Romans 10:6-13). If justification is in antynomous relationship to ‘condemnation’, salvation is an antonym of ‘suffering wrath’ or ‘receiving punishment’. Salvation moves beyond the judicial and into the substance of the punitive. Further, salvation is broader than justification, in that it encompasses not merely the juridical aspects of salvation, but also the renovative and regenerative aspects, including sanctification. This is especially so given that Paul also mutually conditions a gracious ‘being made alive with Christ, raised with Christ, and seated with Christ’ by the statement ‘by grace you have been saved’ (v. 5). All this is to say that it is important to observe that Paul here says we have been saved by faith, rather than we have been justified by faith. The substance of the statement in Ephesians may be broader or differently conceived than what is suggested by statements we find in Romans and Galatians that ‘we have been justified by faith’.

v. 8: διὰ πίστεως (‘through faith’). This preposition with the genitive rightly denotes a means or instrumentality of the preceding verbal idea of ‘salvation’. The experience described as ‘we have been saved’ comes to a person through exercising faith in Christ as an instrument for receiving that salvation as a benefit. Human faith, rather than Christ’s faithfulness, is here invisaged.

vv. 8-9: καὶ τοῦτο οὐκ ἐξ ὑμῶν, θεοῦ τὸ δῶρον· 9οὐκ ἐξ ἔργων (‘and this [is] not from yourselves, [it is] the gift of God, [it is] not from works’). The syntax of the sentence is that of a complex subject-predicate with the verb ‘to be’ elided. The subject is τοῦτο, the predicate is positively stated as θεοῦ τὸ δῶρον, and in addition the predicate is specifically declared not to be two things (v. 8: οὐκ ἐξ ὑμῶν; v. 9: οὐκ ἐξ ἔργων). This then calls into question Wallace’s claim that ‘δῶρον is not the predicate nom[inative] of τοῦτο, but of the implied “it” in the following clause.’[149] What is the supposed ‘it’, but the subject implied, which is τοῦτο? The subject is almost certainly whatever τοῦτο refers to. So the distinction Wallace attempts to draw is unsound, for it is a distinction without a difference. The versification, which breaks up the clause at verse 9, and the Greek punctuation, are all editorial, and have some value in telling us where previous readers of Paul’s text have thought breaks in the flow of the text exist, but they do not in any way give us insight into Paul’s mind.

v. 8: καὶ can be conjunctive (‘and’) or ascensive (‘even’), but it would seem that the ascensive and emphatic role of καὶ fits the context.

v. 8: τοῦτο (‘this’, neuter), as I have argued is retrospective, referring to antecedent ‘πίστις’ (faith, feminine), on the basis of following factors:

(1) the progression of the discourse, which suggests that the new element introduced is διὰ πίστεως, both χάρις and σεσῳσμένοι having explicitly be mentioned and explained in vv. 4-7, and it is then πίστις which is now being described.

(2) the evidence from Classical Greek, the LXX, and the NT, which suggests that τοῦτο can and indeed does refer back to neuter antecedents, or otherwise showing that antecedent and demonstrative need not agree. The following are 15 other passages I regard as certain or highly probable instances of the construction:

1. δήμου τε ἀρίστου καὶ ὀλιγαρχίης καὶ μουνάρχου [masculine], πολλῷ τοῦτο [neuter] προέχειν λέγω || the best democracy and oligarchy and monarchy [masculine], I hold that monarchy [Lit. neuter ‘this’] is by far the most excellent (Herodotus, The Histories, 3.82[1])
2. καρπὸν [masculine] δὲ φορέει κυάμῳ ἴσον, πυρῆνα δὲ ἔχει. τοῦτο [neuter] ἐπεὰν γένηται πέπον || and bears a fruit as big as a bean, with a stone in it. When this fruit is ripe, they strain it through cloth (Herodotus, The Histories, 4.23[3])
3. πῶς ἔχεις πρὸς ἐπιστήμην [feminine]; πότερον καὶ τοῦτό σοι δοκεῖ ὥσπερ τοῖς πολλοῖς ἀνθρώποις || how are you in regard to knowledge [feminine]? Do you share the view that most people take of this [neuter]? (Plato, Protagoras, 352b)
4. ἤρεσθε δέ, εἰ μέμνησθε, ἡνίκα ἡμεῖς ἀλλήλοις ὡμολογοῦμεν ἐπιστήμης [feminine] μηδὲν εἶναι κρεῖττον, ἀλλὰ τοῦτο [neuter] ἀεὶ κρατεῖν, ὅπου ἂν ἐνῇ, καὶ ἡδονῆς καὶ τῶν ἄλλων ἁπάντων || You asked it, if you remember, when we were agreeing that there is nothing stronger than knowledge (ἐπιστήμης feminine), and that knowledge [literally, ‘this’, neuter], wherever it may be found, has always the upper hand of pleasure or anything else (Plato, Protagoras, 357c)
5. ὃ μεταξὺ ἄρα νυνδὴ ἀμφοτέρων ἔφαμεν εἶναι, τὴν ἡσυχίαν, τοῦτό ποτε ἀμφότερα ἔσται, λύπη τε καὶ ἡδονή || “What, then, we just now described as the intermediate state between the two—this quietude—will sometimes be both pain and pleasure.” (Plato, Republic, 9.583e)
6. Σωκράτης: σοφίᾳ [feminine] δέ γ᾽ οἶμαι σοφοὶ οἱ σοφοί. Θεαίτητος: ναί. [145ε] Σωκράτης: τοῦτο [neuter] δὲ μῶν διαφέρει τι ἐπιστήμης; Θεαίτητος: τὸ ποῖον; Σωκράτης: ἡ σοφία [feminine]. || Socrates: And the wise, I suppose, are wise by wisdom [feminine]. Theaetetus: Yes. [145e] Socrates: And does this [neuter] differ at all from knowledge? Theaetetus: Does what differ? Socrates: Wisdom [feminine]. (Plato, Theaetetus, 145d)
7. εἰ δὲ καὶ ἐμπορία [feminine] ὠφελεῖ τι πόλιν, τιμώμενος ἂν ὁ πλεῖστα τοῦτο [neuter] ποιῶν καὶ ἐμπόρους ἂν πλείους ἀγείροι || If commerce [feminine noun] also brings gain to a city, the award of honours for diligence in business [Lit. ‘this’, neuter] would attract a larger number to a commercial career (Xenophon, Hiero, 9.9).
8. [11] ἐγὼ γάρ, ὦ ἄνδρες δικασταί, περὶ τῆς μαρτυρίας τῆς ἐν τῷ γραμματείῳ γεγραμμένης εἰδὼς ὄντα μοι τὸν ἀγῶνα, καὶ περὶ τούτου τὴν ψῆφον ὑμᾶς οἴσοντας ἐπιστάμενος || I knew, men of the jury, that I should find the whole contest centering about the deposition [Lit. marturia, ‘testimony’] inserted in the record, and that it would be regarding the truth or falsehood of this [neuter] that you would cast your votes (Demosthenes, Against Aphobus for Phanos 3, Oration 29[11])
9. ἀλλ᾽ ὁ μὲν δόξης [feminine] ἐπιθυμεῖ καὶ τοῦτ᾽ [neuter] ἐζήλωκε || Instead he [Philip of Macedeon] desires glory, and has zealously pursued this (Demosthenes, Orations 2: Olynthiac 2 [15]).
10. ὅτι παντάπασι φύσεως κακίας σημεῖόν ἐστιν ὁ φθόνος [noun, masculine], καὶ οὐκ ἔχει πρόφασιν δι᾽ ἣν ἂν τύχοι συγγνώμης ὁ τοῦτο [neuter demonstrative] πεπονθώς || Because in every way envy is the mark of a vicious nature, and the man who is subject to it [lit, ‘this’] has no claim whatever to consideration (Demosthenes, Against Leptinesm, 20 [140]).
11. ²²καὶ ᾠκοδόμησεν κύριος ὁ θεὸς τὴν πλευράν [feminine sing], ἣν ἔλαβεν ἀπὸ τοῦ Αδαμ, εἰς γυναῖκα [feminine singular] καὶ ἤγαγεν αὐτὴν [feminine singular] πρὸς τὸν Αδαμ. ²³καὶ εἶπεν Αδαμ Τοῦτο [neuter] νῦν ὀστοῦν [neuter] ἐκ τῶν ὀστέων μου καὶ σὰρξ [feminine] ἐκ τῆς σαρκός μου·[note verbless clause] αὕτη κληθήσεται γυνή || ²² And God formed the rib which he took from Adam into a woman, and brought her to Adam. ²³ And Adam said, This now is bone of my bones, and flesh of my flesh; she shall be called woman (Genesis 2:22-23 LXX).
12. καὶ τῶν βασιλέων τῶν μετ᾽ αὐτοῦ— εἰς τὴν κοιλάδα τὴν Σαυη [feminine singular] (τοῦτο [neuter] ἦν τὸ πεδίον βασιλέως) || and the kings with him, to the valley of Saby; this was the plain of the kings (Genesis 14:17 LXX).
13. 6καὶ ἀπεστάλη πρός με ἓν τῶν σεραφιν, καὶ ἐν τῇ χειρὶ εἶχεν ἄνθρακα [masculine, acc sing], ὃν [masc rel pron] τῇ λαβίδι ἔλαβεν ἀπὸ τοῦ θυσιαστηρίου,7καὶ ἥψατο τοῦ στόματός μου καὶ εἶπεν ᾿Ιδοὺ ἥψατο τοῦτο [neuter] τῶν χειλέων σου καὶ ἀφελεῖ τὰς ἀνομίας σου καὶ τὰς ἁμαρτίας σου περικαθαριεῖ || 6 And there was sent to me one of the seraphs, and he had in his hand a coal, which he had taken off the altar with the tongs: 7 and he touched my mouth, and said, Behold, this has touched thy lips, and will take away thine iniquities, and will purge off thy sins (Isaiah 6:6-7 LXX).
14. 49πλὴν τοῦτο [neuter] τὸ ἀνόμημα Σοδομων τῆς ἀδελφῆς σου, ὑπερηφανία [feminine]·|| 49 Moreover this was the sin of thy sister Sodom, pride (Ezekiel 16:49 LXX).
15. ²⁸ […] ἥτις ἐστὶν αὐτοῖς ἔνδειξις [feminine] ἀπωλείας, ὑμῶν δὲ σωτηρίας [feminine], καὶ τοῦτο [neuter] ἀπὸ θεοῦ· ²⁹ὅτι ὑμῖν ἐχαρίσθη τὸ ὑπὲρ Χριστοῦ, οὐ μόνον τὸ εἰς αὐτὸν πιστεύειν ἀλλὰ καὶ τὸ ὑπὲρ αὐτοῦ πάσχειν || ²⁸ […] which is a sign of destruction for them, but of salvation for you, and that too, from God. ²⁹ For to you it has been granted for Christ’s sake, not only to believe in Him, but also to suffer for His sake (Philippians 1:28-29).

(3) Constructio ad sensum: The fact that ‘faith’ is only ‘feminine’ grammatically, but as a feminine abstract noun, it is quite acceptable to refer to it with the neuter demonstrative on the basis of constructio ad sensum.

(a) τὸν κάμηλον [masculine], ὅτι ἀνάγει μηρυκισμὸν τοῦτο, ὁπλὴν δὲ οὐ διχηλεῖ, ἀκάθαρτον τοῦτο ὑμῖν·|| the camel, because it chews the cud, but does not divide the hoof, this is unclean to you (Leviticus 11:4 LXX; Cf. vv. 5-7; Deuteronomy 14:7-8 LXX).

(b) ἐὰν εἴπῃ ὁ πούς· [masculine] ὅτι οὐκ εἰμὶ χείρ, οὐκ εἰμὶ ἐκ τοῦ σώματος, οὐ παρὰ τοῦτο [neuter] οὐκ ἔστιν ἐκ τοῦ σώματος || If the foot shall say, Because I am not the hand, I am not of the body; it is not therefore not of the body (1 Corinthians 12:15).

(4) The testimony of ancient exegetes who testify that not only can the neuter demonstrative refer back to a feminine antecedent, but that it does in fact do so. These ancient churchmen who so understood the construction were either native Greek speakers, or Latins who were among the best Greek Christian scholars of their generation (even if we were to exclude Augustine, but perhaps we ought not think his ability in Greek so wholly undeveloped). For all the Greek speaking Eastern exegetes, such a finding was a difficulty to explain, either because of their free-will soteriology, their synergism and emphasis on virtue, and their anthropological presuppositions:

1. Chrysostom: Then, that on the other hand, our free-will (τὸ αὐτεξούσιον) be not impaired, he adds also our part in the work (ἔθηκε καὶ τὰ ἡμῶν), and yet again cancels it, and adds, ‘And that not of ourselves.’ (καὶ πάλιν αὐτὸ ἀνεῖλε, καί φησι · Καὶ τοῦτο οὐκ ἐξ ἡμῶν.) Neither is faith, he means, ‘of ourselves.’ (Οὐδὲ ἡ πίστις, φησὶν, εξ ἡμῶν.) […] “It is the gift,” said he, “of God,” it is “not of works” (θεοῦ, φησὶ, τὸ δῶρον;· οὐκ ἐξ ἐργων).[150]

2. Theodoret: Now we bring only faith (Latin: solam fidem; μόνην τὴν πίστιν). But the divine grace has become fellow worker (συνεργός) even of this (ταύτης). For he continues with this: ‘even this (καὶ τοῦτο) [is] not from yourselves, it is the gift of God, not from works, so that no-one might boast’. For it is not from our own will [that] we have believed (Οὐ γὰρ αὐτόματοι πεπίστεύκαμεν), but having been called, we have come, and when we came, he did not demand purity of life, but accepting faith alone (Latin: sola fide; ἀλλὰ μόνην τὴν πίστιν δεξάμενος), he bestowed the forgiveness of sins.[151]

3. Jerome: And this faith itself is not from yourselves (Et haec ipsa fides non est ex vobis), but from him that called you (sed ex eo qui vocavit vos). This [he] also [says], lest perhaps the secret thought might creep up on us, that if we have not been saved by our own works (per opera nostra), at any rate we have actually been saved by faith (per fidem), and in this way by a different method we are saved by ourselves. Accordingly, he [Paul], said in addition, and asserted, that even faith is not by our own will (non nostrae voluntatis esse) but is God’s gift (sed Dei muneris) […] that we are able to believe (credere possumus), is from him[.][152]

4. Œcumenius: On the one hand faith is from yourselves (Ἐξ ὑμῶν μὲν ἡ πίστις), but the cause of it is God (ἀλλὰ καὶ αὐτῆς πρὸφασις ὁ θεός.). […]Therefore, he calls faith [the] gift of God (Διὸ δῶρον θεοῦ καλεῖ τὴν πίστιν). Doubtless it is also a gift for this reason ( Ἢ καὶ δῶρον διὰ τοῦτο), because faith would not be strong enough on its own to save, except if God wished to save through faith. So that, for us to believe [is the] gift of God (Ὥστε τὸ πιστεῦσαι ἡμάς, δῶρον θεοῦ), and to be saved through faith [is the] gift of God (καὶ τὸ διὰ πίστεως σωθῆναι δῶρον θεοῦ).[153]

5. Theophylact: ‘And this not from yourself, it is the gift of God.’ Again, he cancels it, and says, that neither is faith from ourselves (Πάλιν αὐτὸ ἀνεῖλε, καί φησιν, ὅτι οὐδέ ἡ πίστις ἐξ ἡμῶν) […] ‘For how’ he says, ‘will they believe, if they do not hear?’ (Romans 10:14: << πῶς γάρ >> φησὶ, << πιστεύσουσιν, ἐὰν μὴ ἀκούσωσιν; >>), so that also this is the gift of God (‘Ωστε καὶ αὕτη δῶρον θεοῦ).[154] While it is true that Theophylact offers a second interpretation as an alternative, he still offers the first and does not criticize it as impossible on grammatical grounds.

6. Augustine: ‘For he himself also says, “By grace you are saved through faith; and this not of yourselves; but it is the gift of God,” (Ephesians 2:8-9)—that is to say, “And in saying 'through faith,' even faith itself is not of yourselves, but is God's gift.”’ (On the Predestination of the Saints, Book 1 Chapter 12)[155]

7. Fulgentius: ‘The blessed Paul argues that we are saved by faith, which he declares to be not from us but a gift from God. Thus there cannot possibly be true salvation where there is no true faith, and, since this faith is divinely enabled, it is without doubt bestowed by his free generosity’ (On the Incarnation, 1)[156]

8. John of Damascus: I add John of Damascus here because, even though he says that the antecedent of the demonstrative is not the word for ‘faith’ but ‘grace’, the Greek word for ‘grace’ is still a feminine noun when he comments: ‘The gift of God’ Grace is the thing which lies with God; faith is the thing which lies with us. For this reason, then, for those for whom the fitness to receive [grace] may not be present, then neither does the grace come alongside to assist. It [grace] is not from us, therefore, but it is the gift of God.θεοῦ τὸ δῶρον · Χάρις, τὸ παρὰ θεοῦ, πίστις, τὸ παρ᾽ ἡμῶν. Ὄθεν οὔν οἷς μή πάρεστι τὸ δεκτικὸν, οὐδὲ ἡ χάρις παραγίνεται. Οὐκ ἐξ ἡμῶν οὖν, ἀλλὰ θεοῦ τὸ δῶρον ·[157]

vv. 8-9: οὐκ ἐξ ὑμῶν, θεοῦ τὸ δῶρον Ÿ οὐκ ἐξ ἔργων (not from yourselves, [it is] the gift of God, not from works). These three phrases constitute the predicate of the subject τοῦτο. It would seem that Wallace would disagree with this, as he says of the construction of Ephesians 2:8: δῶρον is not the predicate nom[inative] of τοῦτο, but of the implied “it” in the following clause’.[158] But on Wallace’s reasoning, what is the reference of implied ‘it’? It can only be τοῦτο at the beginning of verse 8. Wallace has failed to reckon with the reality we find, which is a complex predicate nominative, consisting of two negations as well as a positive assertion, with Paul’s quite usual abbreviated style—and all of these clauses both logically and grammatically part of the same idea with the same grammatical subject.

The relationship between the three phrases that constitute the predicate has been subject to debate. However, a progression of ideas is discernible. This progression rests on the denotation of the vocabulary and the adversertive or epexegetical nature of the clauses.

οὐκ ἐξ ὑμῶν >> θεοῦ τὸ δῶρον >> οὐκ ἐξ ἔργων

οὐκ ἐξ ὑμῶν… The first of three phrases is a negated (οὐκ) prepositional phrase, with ἐξ and the genitive denoting source or origin. The plural pronoun ὑμῶν refers to the Ephesian Christians who have been saved, and by implication, every Christian.

…θεοῦ… The genitive θεοῦ is immediately to be contrasted with ὑμῶν. It is almost certain that ἐκ should not be thought to govern θεοῦ but rather the genitive θεοῦ modifies τὸ δῶρον. However, that is not to say that the co-location of the two genitives next to one another (ὑμῶν θεοῦ)[159] did not also perform a rhetorical function. That is, they are juxtaposed by way of contrast. There is no adversative conjunction—no doubt a function of Paul’s abbreviated, condensed, and clipped style evidenced elsewhere—but it is clear that a strong contrast is required but the negative particle οὐκ, by the continued contrast of the actions and situation of the ‘we/you’ (ὑμᾶς: v 1; ἡμεῖς πάντες […]ἡμῶν […] ἤμεθα: v. 3; ἡμᾶς: v. 4) and ‘God’ (ὁ δὲ θεὸς: v. 4, who is the subject of the verbs and the one adjectivally described in vv. 4-7), and by the contrasting denotations of ὑμῶν and θεοῦ. That is to say, ‘God’ and ‘yourselves’ are strongly contrasted—it is a binary choice being presented. This thing, ‘faith’, is not from you, it does not originate with you, you cannot call it yours—it is of God.

…τὸ δῶρον… This articular noun is nominative predicate (contra Wallace) of the subject of the verbless clause, τοῦτο. The denotation of the word for gift (δῶρον) tells us something about the negation in the previous clause. It implicitly affirms something that we know phenomenologically about πίστις, that it is in one sense ours. This is because the negating statement οὐκ ἐξ ὑμῶν at one level is counter-intuitive—of course ‘faith’ is exercised by the Christians in Ephesus. This phenomenological fact is explicitly stated by several of the ancient exegetes:

• Chrysostom: ‘he [Paul, by Διὰ τῆς πίστεως] adds also our part in the work (ἔθηκε καὶ τὰ ἡμῶν)’.[160]
• Theodoret: ‘Now we bring only faith (Latin: solam fidem).’[161]
• John of Damascus: ‘faith is the thing which lies with us (πίστις, τὸ παρ᾽ ἡμῶν.).’[162]
• Oecumenius: On the one hand faith is from yourselves ( Ἐξ ὑμῶν μὲν ἡ πίστις).[163]
• Theophylact: ‘he brings forward that which pertains to ourselves, that is, ‘through faith’ (ἐπάγει καὶ τὸ ἡμέτερον, τὸ, << Διὰ τῆς πίστεως), […] For let us have it established that faith is our own (Ἔστω γὰρ, ὅτι ἡ πίστις ἡμέτερον).’[164]

In saying this, these ancient exegetes are only reflecting what Paul himself says in Ephesians:

• ‘you, when you heard the word of truth, the gospel of your salvation, in which also you believed’ (ὑμεῖς ἀκούσαντες τὸν λόγον τῆς ἀληθείας, τὸ εὐαγγέλιον τῆς σωτηρίας ὑμῶν, ἐν ᾧ καὶ πιστεύσαντες: Ephesians 1:13);
• ‘when I heard your faith in the Lord Jesus’ (ἀκούσας τὴν καθ’ ὑμᾶς πίστιν ἐν τῷ κυρίῳ Ἰησοῦ: Ephesians 1:15);
• ‘to you who believe’ (εἰς ἡμᾶς τοὺς πιστεύοντας: Ephesians 1:19)

In one sense, faith is indeed ‘from us’.

The function of the noun δῶρον here in the discourse is to explain the relationship of the exercise of faith by the Ephesian Christians and God’s activity. ‘Faith’ is a gift from God. The use of the word δῶρον and the concept of ‘gift’ draws on the ancient practice of benefaction. [165] It denotes the benefit itself that comes from the attitude of benevolence of the giver.[166] In Ephesians 2:1-10, the divine beneficence is described in vv. 4-7 (‘God being rich in mercy’ || θεὸς πλούσιος ὢν ἐν ἐλέει; ‘because of his great love with which he loved us’ || διὰ τὴν πολλὴν ἀγάπην αὐτοῦ ἣν ἠγάπησεν ἡμᾶς; ‘by grace’ || χάριτί; ‘the superabundant riches of his grace in kindness upon us’ || τὸ ὑπερβάλλον πλοῦτος τῆς χάριτος αὐτοῦ ἐν χρηστότητι ἐφ’ ἡμᾶς). Moreover, the content of gift referred to by δῶρον is informed by the verbs Paul uses to describe God’s actions upon the believers: συνεζωοποίησεν, συνήγειρεν, συνεκάθισεν τῷ Χριστῷ (Eph 2:5-6). It is a gracious making alive as opposed to allowing us to remain dead, and raising us up and seating us with Christ.

The nature of the divine origin of faith is disputed amongst the ancient exegetes. A strong theme in the Eastern expository tradition is to explain the divine cause of faith as lying in God’s initiative in the incarnation and the preaching of the gospel, per Romans 10:14 (Chrysostom; Œcumenius). This leaves a space for human free will (τό αὐτεξούσιον) in their soteriology. In this scheme, divine grace itself has become our fellow-worker’ (συνεργός: Theodoret). Thus it is synergistic.

Theodoret seems to go further than merely explaining the proposition that faith is a divine gift because God sent the incarnate Christ and issued an outward call. He says, ‘For it is not from our own will [that] we have believed, but having been called, we have come (Οὐ γὰρ αὐτόματοι πεπίστεύκαμεν, ἀλλὰ κληθένητες προσεληλύθαμεν)’. This seems to refer to an efficacious call. While any causal link between Theodoret and Augustine cannot be shown, Theodoret can be seen to be moving further towards Augustine, although he is still a synergist.

After his change of mind around AD 397, Augustine presented a significantly different understanding of the nature of the divine gift of faith than the Eastern theologians. Augustine held that the impulse by which we ask and seek and knock is itself given to us by God (Augustine, On the Gift of Perseverance, 64).^[167] He held ‘that we receive, without any merit of our own, that from which everything […] has its beginning— that is, faith itself.’ (On the Grace of Christ and Original Sin, Book 1 Chapter 34).^[168]

The mature articulation of Augustinian theology or prevenience and divine enablement of faith is found in the second generation Augustinian North African Bishop Fulgentius, the ‘little Augustine’: ‘and, since this faith is divinely enabled, it is without doubt bestowed by his free generosity’ (On the Incarnation, 1).^[169] In De Fide ad Petrum (ET: To Peter on the Faith), Fulgentius gathers this Augustinian teaching into the propositions that (1) the beginning of a good human will is not from a person, but is bestowed from God (§34 p81); (2) God must by his grace enable someone to believe in his heart (canon 32 §75 p102); and (3) neither the merit of having a good will or good works precede that faith (§43 p88). The key to the Augustinian position is not the denial of the reality of the human will, but the divine enablement of the human will so that it is a good will. Without secret divine enablement, the human will cannot respond in faith to God’s invitation. It is ‘dead in transgressions and sins’. The view of Augustine and Fulgentius is far more satisfying than the Eastern exegetes in explaining how ‘faith’ is a gift from God: it is not only the divine invitation to faith, but the enablement of it, that best accords with faith being ‘the gift of God’, in the context of Ephesians 2:1-10.

…οὐκ ἐξ ἔργων… (‘not from works’). Not only does θεοῦ τὸ δῶρον assist to illuminate οὐκ ἐξ ὑμῶν which precedes it, but it also illuminates οὐκ ἐξ ἔργων which follows it. Just as θεοῦ was placed in contrast with ὑμῶν, and δῶρον serves to explicate the seeming paradox how something so clearly ours is counter-intuitively not, so also δῶρον stands in contrast with ἔργων. The ‘this’, ‘faith’, which is a gift from God, is not from works. What does this mean?

The phrase οὐκ ἐξ ὑμῶν refers to the primary source, cause, and origin of faith, though there are also secondary causes. The nature of the contrast suggests that οὐκ ἐξ ἔργων refers to a different type of cause or origin—not to primary origin or source, but a denial of a potential meritorious cause.

The neuter noun δῶρον does not appear anywhere else in Paul’s writings. However, other gift words appear with ‘faith’ and in contrast with ‘works’ (albeit ‘works of the law’ but I don’t consider the phrases ‘works of the law’, ‘good works’ or ‘works’ really that different—as the law is ‘holy, righteous, and good’ (Romans 7:12). So consider Romans 3:20, 23-24, 27-28, 4:4-5:

• ch. 3 v. 20: ‘Therefore, from works of the law no flesh will be justified before him, for through the law is knowledge of sin (διότι ἐξ ἔργων νόμου οὐ δικαιωθήσεται πᾶσα σὰρξ ἐνώπιον αὐτοῦ, διὰ γὰρ νόμου ἐπίγνωσις ἁμαρτίας) […]
• ch. 3 vv. 23-24: For all have sinned and fallen short of the glory of God, being justified [as a] gift by his grace through the redemption which is in Christ Jesus (πάντες γὰρ ἥμαρτον καὶ ὑστεροῦνται τῆς δόξης τοῦ θεοῦ δικαιούμενοι δωρεὰν [adverb] τῇ αὐτοῦ χάριτι διὰ τῆς ἀπολυτρώσεως τῆς ἐν ΧριστῷἸησοῦ) […]
• ch. 3 vv. 27-28: So then, where is the boasting? It is excluded. Because of what law? Of works? No, but through the law of faith. For we reckon a person to be justified by faith apart from works of the law. (Ποῦ οὖν ἡ καύχησις; ἐξεκλείσθη. διὰ ποίου νόμου; τῶν ἔργων; οὐχί, ἀλλὰ διὰ νόμου πίστεως. λογιζόμεθα γὰρ δικαιοῦσθαι πίστει ἄνθρωπον χωρὶς ἔργων νόμου […].
• ch. 4 vv. 4-5: Now to the one working, his wages are not reckoned according to grace but according to obligation. But to the one not working but trusting upon the one justifying the ungodly, his faith is reckoned for righteousness (τῷ δὲ ἐργαζομένῳ ὁ μισθὸς οὐ λογίζεται κατὰ χάριν ἀλλὰ κατὰ ὀφείλημα, τῷ δὲ μὴ ἐργαζομένῳ πιστεύοντι δὲ ἐπὶ τὸν δικαιοῦντα τὸν ἀσεβῆ λογίζεται ἡ πίστις αὐτοῦ εἰς δικαιοσύνην).

Barclay helpfully and I believe correctly recognizes that in Ephesians 2:9 ‘works’ are ‘moral achievements’.[170] But I cannot accept other aspects of his thesis. Barclay regards the phrase ‘works of the law’ and its contrast with ‘faith in Christ’ in Romans and Galatians as formulated for the gentile mission, in which Paul declared ethnic distinction between Jew and gentile superceded because of the divine gift of Christ. According to Barclay, ‘works of the law’ means Jewish practices, and the ‘law’ in question is the torah. For Barclay, it is not that ‘law or ‘works’ are ‘problematic principles of soteriology’, but that torah and every other norm has been dethroned and replaced with the unconditioned gift of Christ.[171]

But Barclay’s conception fails to reckon with the fact that the law’s stipulations are ‘holy, righteous, and good’ (Romans 7:12). ‘Works of the law’ of necessity have to do with human achievement, because human achievement is central if a person is to be justified by the law (Romans 2:12-13). In this case, ‘the works’ (τὰ ἔργα: Romans 2:6) which are to be rewarded at the judgement under such provisions as required by the law are good works. Compare Romans 2:6, which re-iterates the retributive principle in that God will give back to each ‘according to his works’ (κατὰ τὰ ἔργα αὐτοῦ) and the description of those so justified according to that schema, which is ‘according to perseverance in good works’ (Romans 2:7: καθ’ ὑπομονὴν ἔργου ἀγαθοῦ). ‘Works of the law’ are ‘good works’ because the God who gave the Torah is good and only gives good things. The Gentiles in any case have the ‘work of the law’ written on their heart (Romans 2:15), and by God’s general revelation the Gentiles are without excuse (Romans 1:18-20).

Of course there is no-one who has fulfilled the conditions of these requirements, and Romans 2:13 remains an unfulfilled null-set (except in the case of Christ), because no-one is righteous (Romans 3:10-20, 23), and the righteousness of the law which promised life is not about hearing but about doing (Romans 10:5; cf. 7:10; Gal 3:10-14).

To emphasise the Jewishness of those works in the phrase ‘works of the law’, as Barclay does, is really not that different to Dunn’s delimitation of ‘works of the law’ to ethnic boundary markers—despite Barclay’s efforts to distinguish himself from Dunn.[172] This is why Barclay seems to regard Ephesians 2:8-10 as a decline from the true Pauline soteriology, such that Barclay regards the word ‘works’ in Ephesians 2:8-10, 2 Timothy 1:9; and Titus 3:5, as ‘refocussed as moral achievements’ and boasting as ‘pride in achievement’, because the terms ‘works’ was used in the context that the Gentile-Jewish division was no longer to the fore, and the ‘critical edge’ of Paul’s theology was ‘turned against believers’.[173]

However, rather than limiting the meaning of ἔργα νόμου, ‘works of the law’ should be understood as all works done pursuant to obedience to the law, whether the Torah as the Mosaic law or the law of the heart, and not limited to ceremonial law or identity markers such as circumcision and food laws. Against many proponents of or those influenced by the NPP, the phrase ‘the ones doing the law will be justified’ (οἱ ποιηταὶ νόμου δικαιωθήσονται: Romans 2:13) is functionally synonymous with the phrase ‘from the works of the law will be justified’ (ἐξ ἔργων νόμου […] δικαιωθήσεται: Romans 3:20).[174] The ἔργα νόμου are ‘actions performed in obedience to the law’.[175] ‘”Works of law”, then is a comprehensive expression that refers to the entirety of the actions and abstentions prescribed by the law’.[176] The reason that ‘works of the law’ (Romans 3:20, 28) or ‘works’ (Romans 4:3, 5) cannot justify is ‘because no one is able to perform works to the degree needed to secure such a standing’.[177] Thus, ‘Paul’s argument from 1:18-3:26 must be taken as a whole’ and ‘Paul does not envision Romans 2:13 to be a real possibility…’.[178]

Further, Paul conceives of the works that are excluded from Abraham’s justification (Romans 4:3, 5) as including those works done by Abraham after Abraham first had faith. Paul quotes Genesis 15:6, which occurred many years subsequent to Abraham’s entering the life of faith (Genesis 12:1, cf. Hebrews 11:8-9). Abraham as exemplified in Genesis 15:6 and as used in Romans 4 and Galatians 3 is not and cannot be an example of ‘Christian initiation’, because the declaration of Genesis 15:6 occurs decades after Abraham’s first evidence of faith (Genesis 12:1ff.), and thus cannot be Abraham’s ‘initiation’ into righteousness. Moreover, there is no way that Paul’s use of Psalm 32 can be so conceived as an initiation into righteousness. In Psalm 32:1-2, free justification is in no way limited to pre-conversion works, as Psalm 32 in its original context clearly refers to David as an existing member of the covenant community, not an outsider entering the covenant. Indeed, David is Anointed Christ within the community! You cannot get more within the covenant than Messiah David. Yet Paul speaks of David’s words, uttered as part of his ongoing life of faith, as illustrating the proposition ὁ θεὸς λογίζεται δικαιοσύνην χωρὶς ἔργων (Romans 4:6). Thus, the free justification ‘apart from works’ spoken of in Romans 3:21-4:25 applies not merely to ‘initial justification’ and works before conversion, but also to the ongoing application of the same free justification apart from works (however defined) to the very end of the believer’s life.[179] The enigmatic phrase ‘from faith for faith’ (ἐκ πίστεως εἰς πίστιν: Romans 1:17; cf. 2 Corinthians 2:16) likewise points this way, that faith and nothing but faith, with no admixture of works, can justify sinful humans.[180]

All of this is relevant to the meaning of οὐκ ἐξ ἔργων (Ephesians 2:9). The Pauline background above suggests that ‘works’ should not be distinguished from good works, or limited to pre-conversion works, or ethnic boundary markers, or ceremonial works, or Jewish works. Rather, ‘works’ are ‘human achievements’, ‘human effort’, ‘good works’. It is in this context that ‘gift’ helps us. The freeness and unconditioned nature of the gift of God—faith—means that the gift of faith is given by God not in response to human works or human efforts. Human faith is both enabled by God (and thus οὐκ ἐξ ὑμῶν), and is given freely as a gift (θεοῦ τὸ δῶρον), that is, unmerited by humans (οὐκ ἐξ ἔργων).

Chrysostom: He did not reject us as having works, but having been abandoned of works, He has saved us by grace (Oὐχὶ ἔχοντας ἔργα ἀπώσατο, ἀλλὰ προδεδομένους ἀπὸ τῶν ἔργων χάριτι ἔσωσεν) […][181]

Photius of Constantinople (c. 815 – 897) as quoted by Œcumenius, expresses an appreciation of why it is impossible to be saved by good works:

‘Not by works’ << Oὐκ ἐξ ἔργων. >> [Phot[ius] [ΦΩΤ. […] But even when we were created for good works, not only have we done nothing good (οὐ μόνον οὐδὲν ἠγαθουργήσαμεν), but we have even returned the very opposite, and when we returned (ie to God), it was not from ourselves that we have been turned back (nb passive ἀπεστράφημεν), but by the grace and gift of God (οὐκ ἐξ ἑαυτῶν ἀπεστράφημεν, ἀλλὰ χάριτι καὶ δωρήματι θεοῦ.). How great is the gratitude owed to God by us? But this, being created for good works, is at one and the same time both urging us forward to do [good deeds] and standing [us] apart from good deeds. So that He might say ‘You are debtors to do the good, and it is [your] obligation from the beginning’. Because it was for this reason that you were created, but even at your very births you were distressed by many things, and you grew, paying nothing back but even making additions through transgressions. Therefore, we have great need of haste and care to repay our debts.[182]

Photius points out that though we have been created for good works, we have done nothing good (οὐδὲν ἠγαθουργήσαμεν). Not only is this so but we have returned the very opposite (εἰς τούναντίον ἀπηνέχθημεν). Later Photius speaks of the fact that we have not rendered God his due (μηδὲν ἀποδιδόντες) as his workmanship, but we have also added to our indebtedness (χρέος) through our transgressions (ἀλλὰ καὶ διὰ παραβάσεως προστιθέντες). Moreover, when we were turned back (divine passive of ἀποστρέφω), this turning back was not from ourselves (οὐκ ἐξ ἑαυτῶν) but by the grace and gift of God. This can be read quite consistently with divine enablement and Augustinian prevenient grace.

The western interpretative tradition sees that what stands behind the denial of works as having a role is that the gift (whatever it might be) is unmerited. Marius Victorinus almost certainly takes the referent of the ‘gift’ as ‘salvation’. So the following should be regarded as a statement about the gracious salvation God has bought in Christ:

Marius Victorinus: Nor is it from your works, but it is the grace of God, it is the gift of God—not by your merit (meritum). Works are one thing, and our merit another, whence he has differentiated the not from you by saying not from works. Certainly, above and beyond works which are called for every day in our duties toward the poor and other good deeds (but also because one can obtain merit on the basis of duty and religious observance, on the basis of chastity and abstinence), it can be neither by your works[.] So he includes both, saying not from you, nor from works—and then he adds lest someone boast. For he who imagines that the reward (meritum) was merited by his works, wants the reward to be of his own doing (don’t ask me how) and not of the one who bestowed it—and this is boasting.

What is to be noted here is that Victorinus (1) distinguishes the phrases ‘not from you’ and ‘not from works’, and then (2) argues that what is not obtained from works is not merited. I think these are helpful exegetical observations, which can be modified in light of the Eastern exegetical and Western Augustinian view that it refers to ‘faith’.

What could ‘faith’ being ‘not from works’ mean? ‘Faith’ might be ‘not from works’ in that no works merited the divine granting of the gift of faith. Works are not a condition precedent (or subsequent) of the gift of faith. Or alternatively, ‘faith’ might be ‘not from works’ in that faith is an incompatible principle to works, and ‘faith’ is a synechdoche which describes ‘salvation by faith’—in this case, ‘faith’ stands for ‘salvation by grace through faith’, and in effect has the same meaning as if it was a conceptual antecedent. Given that most commentators effectively take this view, I will concentrate on the first one, which does not find any exposition apart from Augustine.

The first interpretation has affinities with that of Augustine. While for Augustine, verse 8 establishes that faith is a gift and not from ourselves, but originates from God (the question of the ultimate origin of faith), verse 9 for Augustine establishes that the gift of faith is not merited nor deserved (the question of whether God gives faith to those who merit or deserve it). Stanley summarizes Augustine this way:

The reason for the qualification in v. 9 (“not by works…”), [Augustine] claimed, was to preclude any notion that faith is merited.[183]

So on Paul’s words, ‘not from works, so that no-one may boast’, Augustine says,

And again, lest they should say they deserved so great a gift by their works, he immediately added, “Not of works, lest any man should boast”.[184]

So here, according to Augustine, verse 8 teaches that faith is a gift and from God and not from ourselves—expressly teaching that God enables faith—and verse 9, by saying it is not of works, teaches that the gift of faith enabled by God is not merited, earned, or deserved by faith or anything else God has foreseen in the person to be saved. I think that this is a helpful and cogent exposition of how each negated phrase in v. 8 and v. 9 can be distinguished.[185]

v. 9: ἵνα μή τις καυχήσηται (‘so that no one might boast’). The purpose clause (ἵνα + subjunctive καυχήσηται) indicates that the purpose of either the whole clause (οὐκ ἐξ ὑμῶν, θεοῦ τὸ δῶρον Ÿ οὐκ ἐξ ἔργων ), or at least the final phrase (οὐκ ἐξ ἔργων). I think the former is preferable, because of the inter-relationship between the three clauses: in which case, that saving faith does not originate from ourselves but comes to us as a gift from God, and that faith as a gift is not merited by our works, means that we cannot boast about our faith, and a fortiori, our salvation by grace.

v. 10: αὐτοῦ γάρ ἐσμεν ποίημα…

Post-positive γάρ marks an explanation of previous discourse. So verse 10 explains verses 8-9. That we are God’s workmanship explains how saving faith is not found to originate from us, but is God’s gift and is not merited by works. Being God’s ‘made items’ (ποίημα) explains how faith is from God not us. Possessive pronoun αὐτοῦ modifies ποίημα, meaning ‘anything made or done, a work, deed, act’, and the reference is to θεοῦ.

v. 10: κτισθέντες ἐν Χριστῷ Ἰησοῦ

Divine passive participle κτισθέντες is dependent on 1^st person plural ἐσμεν to which it refers, and the aorist marks attendant circumstances. The verb κτίζω denotes ‘to found, build, create, bring into being, make’, and serves to explain αὐτοῦ […] ἐσμεν ποίημα. God has created us in Christ Jesus. This creation is a reference to regeneration in Christ, effected through the divine giving of faith.

Chrysostom: Observe the words he uses: He here alludes to the regeneration (τὴν ἀναγέννησιν), which is in reality a second creation (‘Όντως κτίσις ἑτέρα ἐστίν[186] […]

Theodoret: The [word] ‘being created’ (κτισθέντες) here refers to the new birth (ἀναγεννήσεως).[187]

Jerome: He returns to the reasons why we are saved by grace through faith (quare gratia salvati sumus per fidem), and that this is not from ourselves (et hoc ipsum non ex nobis), but the gift of God (sed ex munere Dei), saying 'For we are his workmanship’ (factura), that is, that we live, (vivimus), that we breathe (spiramus), that we understand (intelligimus), and that we are able to believe (credere possumus), is from him, for He is our Creator.[188]

Note that v. 10 for Jerome, and the fact that we are of God’s making, also suggests that God has enabled our belief.

v. 10: ἐπὶ ἔργοις ἀγαθοῖς

Preposition ἐπὶ with the dative ἔργοις ἀγαθοῖς can signify both the purpose and the result or destination. Compare Galatians 5:13, Ὑμεῖς γὰρ ἐπ’ ἐλευθερίᾳ ἐκλήθητε || ‘For you were called to be free’.[189] While some distinguish ἔργοις ἀγαθοῖς in verse 10 from ἔργα in verse 9, it is best to see that there is no difference. That is, it is not as if the ἔργα in verse 9 are Jewish works, and in verse 10 ἔργοις ἀγαθοῖς denotes the moral works of a Christian. In Romans 9:11-12, ‘not from works’ (οὐκ ἐξ ἔργων) expounds ‘and not practicing anything good or bad’ (μηδὲ πραξάντων τι ἀγαθὸν ἢ φαῦλον). It would also be quite unusual to consider that the ἔργα which are excluded from meriting the gift of faith are limited to pre-baptismal works, or only those works done with a particular motive—legalistic works—as there is no such overt limitation in the text.

That there is no effective distinction between the ‘works’ of verse 9 and the ‘good works’ of verse 10 is shown by the fact that if saving faith was merited by ‘good works’ rather than ‘works’, it would still provide a ground for boasting. It is not the addition of the adjective ‘good’ that means that we cannot boast about them, but the ultimate source of these good works, which is God the creator and marker of us in Christ. But even if there is a distinction between ‘works’ in verse 9 and ‘good works’ in verse 10, it is to be found in the difference between works as a meritorious cause of God giving faith which brings salvation, and good works as the purpose of salvation. It is not that ‘works’ in v. 9 are ‘bad’ and ‘good works’ in v. 10 are good, but rather that when Christians are exhorted to ‘works’, they are almost always exhorted to ‘good works’ in distinction to ‘bad ones’. The law does not exhort us to do ‘bad works’ either—only to either good works, which we as Christians are bound to perform, or to ‘obsolete works’, which we should no longer act upon because of the coming of Christ and our place in salvation history.

We are not given saving faith by good works, but for good works. The works that we are not saved by, we are saved for, even though both are good. ‘Good works are never the cause of salvation but ought to be its fruit’.[190]

v. 10: οἷς προητοίμασεν ὁ θεός,

God has prepared beforehand these ‘good works’. That is, he has predestined and prepared the specific good works that those to whom he gives faith walk in. The pre-fix προ— is regularly used in Ephesians (1:4, 5, 9, 11) connoting a divine decree before the foundation of the world. The good works do not derive from the believer as they are planned and purposed by God, and therefore they cannot be said to merit salvation or faith. Thus Paul has moved back into the domain of election and predestination. Unsurprisingly, the word translated ‘prepared beforehand’ also appears in Romans 9:23, in a context which suggests divine predestination.

v. 10: ἵνα ἐν αὐτοῖς περιπατήσωμεν (‘so that in them we might walk’).

Theodoret takes the purpose clause as a command, but strictly the purpose clause is not a command. However, it is the undoubted result of the gift of faith and the creation in Christ Jesus that those saved by faith will indeed produce and walk in these good works. The ‘walking motif’ in verse 10 works as a foil against verses 1-3, where all of us were once ‘walking’ under the power of the Spirit of the air, Satan.[191]

[1] ‘Ambrosiaster’, or Hilary, the Latin writer of the ancient commentaries on Paul popular from the mid-4^th century in Rome, does not address his mind to the question at hand. This commentary is available in English: Ambrosiaster, Commentaries on Galatians-Philemon: ACT (ET: Gerald L Bray: Downers Grove: IVP, 2009), 40.

[2] Abraham Kuyper, The Work of the Holy Spirit, Chapter 39, ‘Defective Learning’ accessed at https://www.ccel.org/ccel/kuyper/holy_spirit.vii.vii.vi.html on 4 July 2017.

[3] Feminine singular noun.

[4] Plural pronoun.

[5] Masculine plural participle in periphrastic construction.

[6] Feminine singular noun.

[7] Demonstrative, Neuter.

[8] Plural pronoun.

[9] Neuter noun.

[10] Instrumental Dative with post-positive γὰρ.

[11] PAI2P εἰμί, I am + PfPPtcpNMP σῴζω, I save; constituting perfect periphrastic construction.

[12] διὰ + genitive, instrumental.

[13] ἵνα + Aorist subjunctive purpose clause.

[14] https://www.academic-bible.com/en/online-bibles/novum-testamentum-graece-na-28/read-the-bible-text/.

[15] Calvin, Comm Eph, 228-9. This view is supported by the silence of Calvin as to this point in his Sermons on Ephesians, and the fact that nowhere to my knowledge does Calvin use Ephesians 2:9 to teach prevenient grace.

[16] E Best, Ephesians: ICC (London/New York: T & T Clark, 1998), 226.

[17] Instrumental Dative with post-positive γὰρ.

[18] PAI2P εἰμί, I am + PfPPtcpNMP σῴζω, I save; constituting perfect periphrastic construction.

[19]

[20] The exceptions are Caird, Ellicott, Westcott, Hodge, who agree with me.

[21] D B Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996), 334-5.

[22] Wallace, op cit, 334 fn 51.

[23] E Best, Ephesians: ICC, 226. Robinson, who also agrees with Calvin on the point, states ‘the difference of gender is not fatal to such a view’ though he argues that a reference to salvation by grace through faith is required: J A Robinson, Ephesians, 157.

[24] BAGD, 596; Wallace, op cit, 325-8.

[25] The quotation is from the Kühner, Ausführliche Grammatik der Griechischen Sprache (Hanover: Hahn, 1870), Vol 2 Part 1 §361 p54.

[26] William Edward Jelf, A Grammar of the Greek Language Chiefly from the German of Raphael Kühner: Syntax (Oxford: John Henry Parker, 1842), p35 §381 Obs 2.

[27] Basil L Gildersleeve, Syntax of Classical Greek from Homer to Demosthenes (New York: American Book Company, 1900), 18.2.501.

[28] For example, Plato, Meno, 73c: ἐπειδὴ τοίνυν ἡ αὐτὴ ἀρετὴ πάντων ἐστί, πειρῶ [...] ἀναμνησθῆναι, τί αὐτό φησι Γοργίας εἶναι || Since the same virtue resides in all, try to recall what Gorgias says it (the thing itself, the essence of it) is.

[29] Dr Philip Buttmann, Intermediate or Larger Greek Grammar (3^rd Ed. ET: Charles Supf: London: Whittaker & Co, 1848), 5.

[30] At this point, some comments on modern social and educational changes and our understanding of Greek grammar seem necessary. Anyone who has had the opportunity to study and benefit from English biblical scholarship from the 19^th century will immediately notice the difference between their scholarship and ours. Often Greek and Latin grammars were written for ‘schools and colleges’. This reflects that an education in the classics was much more common in those days than today. One frequently finds in Dean Alford or Ellicott or Lightfoot, for example, untranslated Greek and Latin extracts from classical and patristic sources. 19^th century scholars produced editions of Greek patristic works without English translations—witness Migne! The readers of their books were ministers who were conversant with Greek and Latin. Indeed, the 1662 Prayerbook of the Church of England indicated that ordinands must be conversant in the theological lingua franca, Latin. Compare this with today’s theological students. I am most familiar with the situation in Sydney and amongst Protestants. Most people attending theological college or seminary are in their mid to late 20s or even 30s. My own college provided 3 weeks of elementary Greek study prior to the beginning of a bachelors level theological courses. Prior to studying Greek, most seminarians have had to learn a crash course of English grammar first, as English grammar was not taught in any meaningful way in Australian primary and high schools since the 1970s. The best students in Protestant Theological colleges manage to work serviceably in Greek and Hebrew by the end of 3 or 4 years full time study only because they have BibleWorks or Logos/Faithlife or other computer programs or a library of commentaries on the original biblical texts. This is so very different to the learned scholars of the 19^th century, at least judging by what they expected of their readers. The relative demise of classical studies has meant that the reason most people learn ancient Greek is because of their desire to read the Greek New Testament. Rarely have they learnt Greek in their formative years. Those who teach Greek in seminaries are what we might call ‘super-pastors’. That is, in theological college they were able to quickly pick up Greek for use in reading the New Testament, and they were also able acquire a serviceable facility in Greek for the purpose of New Testament exegesis. But the standard reference works they use to study Greek are specifically New Testament resources, and these exclusively use the New Testament to inductively determine its grammatical rules (e.g. BDAG, Mounce, Wallace). However, as helpful and indeed necessary as these works are, is it really correct to say that the Greek New Testament is an adequate corpus from which to inductively and exhaustively determine grammatical rules for New Testament Greek? The Koine Greek is derived from the Attic Greek. The New Testament is not the only Koine work. This makes the phrase ‘New Testament Greek’ misleading. The Greek of the New Testament is Koine, and New Testament Greek is not a dialect, but an application of Koine Greek by the foundational Christian writers to their missionary purposes. With the advent of computer analysis of texts and corpora, the field of ‘corpus linguistics’ is a bourgeoning one. As a matter of words, the corpus of classical works is much larger than the Septuagint Corpus, which is much larger than the New Testament. Indeed, the Koine corpus is much larger than the GNT alone. Yet, generally speaking, the same features of the Greek language can be illustrated from each of the three corpora. Thesaurus Linguae Graecae (TLG) has collected and digitized most surviving literary texts written in Greek from Homer to the fall of Constantinople in AD 1453. The total word count of TLG amounts to around 110 million (https://en.wikipedia.org/wiki/Thesaurus_Linguae_Graecae.) This can be compared with a word count in the GNT of around 138,000 and in the LXX of 587,000. That is, on this ‘rough’ figures, the GNT constitutes .0012% of the total TLG. It is therefore not very remarkable to find a GNT hapax legomenon or a rare or unique construction in the GNT because of the size of the corpus.

[31] Kühnher, Ausführliche Grammatik der GriechischenSsprache (Hanover: Hahn, 1870), Vol 2 Part 1 §361 p54.

[32] http://data.perseus.org/citations/urn:cts:greekLit:tlg0016.tlg001.perseus-eng1:3.82

[33] Noun, GSM δήμος people, commons, villages.

[34] Noun, GSF ὀλιγαρχία rule by a few families.

[35] Noun, GSM μόναρχος, monarch, sole ruler.

[36] Basil L Gildersleeve, Syntax of Classical Greek from Homer to Demosthenes (New York: American Book Company, 1900), 18.2.501.

[37] William Edward Jelf, A Grammar of the Greek Language Chiefly from the German of Raphael Kuhner: Syntax (Oxford: John Henry Parker, 1842), 35 §381 Obs 2.

[38] http://data.perseus.org/texts/urn:cts:greekLit:tlg0016.tlg001.perseus-eng1.

[39] http://data.perseus.org/texts/urn:cts:greekLit:tlg0016.tlg001.perseus-grc1

[40] LSJM, 1276

[41] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg022.perseus-grc1

[42] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg022.perseus-eng1

[43] ἐπιστήμη, ἡ, Noun, AFS, acquaintance, skill, knowledge

[44] which of the two.

[45] Plato, Plato in Twelve Volumes, Vol 3 (ET: W R M Lamb: Cambridge, MA: Harvard University Press; London: William Heinemann Ltd, 1967) accessed at http://www.perseus.tufts.edu on 14 August 2016.

[46] Accessed at http://data.perseus.org/citations/urn:cts:greekLit:tlg0059.tlg022.perseus-grc1:357d on 4 July 2017.

[47] Accessed at http://data.perseus.org/citations/urn:cts:greekLit:tlg0059.tlg022.perseus-eng1:357d on 4 July 2017.

[48] ImpfAI1P ὁμολογέω I agree with, grant, concede, say the same thing.

[49] ἐπιστήμη, ἡ, Noun, GFS, acquaintance, skill, knowledge.

[50] Adj, ASN, μηδείς, not one, nothing.

[51] Adj, ASN, comparative, κρείσσων, stronger, mightier.

[52] ἀεί adv, ever, always.

[53] PAInfin κρατέω I am strong, powerful.

[54] AASubj3S ἐνίημι I send in, throw in, launch.

[55] Plato, Platonis Opera, John Burnet (ed) (Oxford: Oxford University Press, 1903), accessed at http://www.perseus.tufts.edu on 4 July 2017.

[56] Hermann Sauppe and James A Towle, Commentary on Plato: Protagoras (Boston and London: Ginn & Co, 1889).

[57] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg030.perseus-grc1.

[58] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg030.perseus-eng1.

[59] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg004.perseus-grc1.

[60] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg004.perseus-eng1.

[61] The noun for ‘dream’ is neuter.

[62] Noun, AFS μουσική, singing poetry over music, any art over which the Muses presided.

[63] ImpAI1S πράσσω I pass over, pass through, experience, achieve, effect, accomplish, manage, transact, negotiate, practice.

[64] ImpfAI1S ὑπολαμβάνω I take up, seize, come suddenly upon.

[65] PM/PInfin παρακελεύομαι I exhort, encourage, recommend, prescribe.

[66] PAInfin ἐπικελεύω I exhort, encourage, cheer on.

[67] PM/PPtcpNPM διακελεύομαι I exhort, give orders, direct.

[68] Noun, GFS.

[69] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg006.perseus-grc1.

[70] http://data.perseus.org/texts/urn:cts:greekLit:tlg0059.tlg006.perseus-eng1.

[71] noun, DFS σοφία wisdom.

[72] PM/PI1S οἶμαι forebode, think, suppose, believe: LSJ.

[73] noun, NFS σοφία wisdom.

[74] http://data.perseus.org/citations/urn:cts:greekLit:tlg0032.tlg008.perseus-grc1:9.9.

[75] http://data.perseus.org/citations/urn:cts:greekLit:tlg0032.tlg008.perseus-eng1:9.9.

[76] N, NFS ἐμπορία, commerce.

[77] PAI3S ὠφελέω I help, aid, succour.

[78] Prn, ASN, τις any.

[79] PM/PPtcpNSM τιμάω I honour, revere, reverence.

[80] Adj, NSF superlative, πλεῖστος, most, greatest, largest.

[81] Adj APM/F ἔμπορος, merchant, trader, dealer.

[82] Adj, APM/F πλείων, comparative, more.

[83] PApt3S ἀγείρω I gather, collect, assemble, put together.

[84] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg029.perseus-grc1.

[85] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg029.perseus-eng1.

[86] PfAPtcpNSM οἶδα I know.

[87] Noun, ASM ἀγών contest.

[88] Demonstrative, GSM/Ntr, οὗτος ‘this’.

[89] Noun, AFS, ψῆφος pebble used in voting.

[90] FutAPtcpAMP φέρω I carry.

[91] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg002.perseus-grc1.

[92] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg002.perseus-eng1.

[93] H Sharpley, Demosthenes: Olynthiacs I, II, III (London: W Blackwood and Sons, 1900).

[94] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg020.perseus-grc1.

[95] http://data.perseus.org/texts/urn:cts:greekLit:tlg0014.tlg020.perseus-eng1.

[96] http://data.perseus.org/texts/urn:cts:greekLit:tlg0010.tlg016.perseus-grc1.

[97] http://data.perseus.org/texts/urn:cts:greekLit:tlg0010.tlg016.perseus-eng1.

[98] Basil L Gildersleeve, Syntax of Classical Greek from Homer to Demosthenes (New York: American Book Company, 1900), 18.2.501.

[99] http://data.perseus.org/texts/urn:cts:greekLit:tlg0032.tlg006.perseus-grc1.

[100] http://data.perseus.org/texts/urn:cts:greekLit:tlg0032.tlg006.perseus-eng1.

[101] Noun, ASM, σῖτος, grain.

[102] Noun, GFS, μελίνη, millet.

[103] Karen H Jobes and Moisés Silva, Invitation to the Septuagint, (2^nd Ed: Grand Rapids: Baker, 2000/2015), Introduction.

[104] John Skinner, Genesis: ICC (Edinburgh: T & T Clark, 1910), 69.

[105] John King (trs), Calvin Comm Gen 2:23 in CC, 1:135 fn 1.

[106] Some statements concerning construction ad sensum from 19^th and early 20^th century grammars: (1) ‘Pronouns, whether personal, demonstrative, or relative, not unfrequently take a different gender from the nouns to which they refer. This is called constructio ad sensum, the meaning, and not the grammatical gender of the word, being mainly considered. It is used particularly when some animate object is denoted by a Neuter or an abstract Feminine noun. The pronoun is then made to agree grammatically with the object in question.’: Georg Benedikt Winer, Grammar of the New Testament Diction (6th ed German, 1855; ET: Edward Masson, 1860) at 153.

(2) ‘The neuter gender is not unfrequently used in reference to persons, where the expression is designed to be of a general nature […] Note. Whenever constructio ad sensum takes place, the gender or number of the word employed is overlooked, and the verb, adjective, etc., accords with the real gender or number of the thing or person intended to be expressed […]: Moses Stuart, A Grammar of the New Testament Dialect (Andover/New York: Allen & Morrill/Dayton and Saxton, 1841) 158.

[107] Wallace, op cit, 335 fn 56.

[108] J B Lightoot, St Paul’s Epistle to the Ephesians (London: Macmillan, 1890), 106; H Alford, The Greek New Testament (5^th Ed: London/Oxford/Cambridge: Rivingtons/Deighton, Bell & Co, 1875) 3:164.

[109] ‘This is restricted to the last clause, that a taste of the grace of God may allay the bitterness of the cross’: Calvin, Comm Philippians, in CC, 21:48 (1✪). It is strange and inconsistent that Calvin would restrict the reference to the last clause when in Ephesians 2:8-9 he takes the referent of the demonstrative to be the concept of gracious salvation by faith. Also Fee: ‘’and this (i.e. your salvation/vindication) comes from none other than God himself’: G D Fee, Paul’s Letter to the Philippians (Grand Rapids/Cambridge: Eerdmans, 1995), 170 (2✪). And Heinrich Meyer (3✪), Critical and Exegetical Commentary on the New Testament (1832), ‘καὶ τοῦτο ἀπὸ θεοῦ […] It adds force to the encouragement conveyed by ὑμῶν δὲ σωτηρίας; for the context shows by the ὑμῖν which is emphatically placed first in Philippians 1:29,—without making the reading ὑμῖν necessary, however, in Philippians 1:28 (Hofmann); see the critical remarks,—that τοῦτο refers only to this second and main part of ἥτις κ. τ. λ. (Calvin, Piscator, Calovius, Flatt, and others, also Ewald and Hofmann), and not to both halves of ἥτις (Beza, Grotius, and many others, also Wiesinger, Weiss, and Ellicott)’: http://www.studylight.org/commentaries/hmc/philippians-1.html.

[110] John Eadie, Commentary on the Greek Text of the Apostle Paul to the Philippians, (New York: Robert Carter & Bros, 1859), 77; C J Ellicott, St Paul’s Epistles to the Philippians, the Colossians and Philemon (London: Longmans, Green & Co, 1875), 34; P T O’Brien, The Epistle to the Philippians: NIGTC (Grand Rapids: Eerdmans, 1991), 157. Chrysostom also seems to take this view: Homily 4 on Phillippians: http://www.newadvent.org/fathers/230204.htm.

[111] PAPtcpAMP εἰμὶ, I am.

[112] AAI2P περιπατέω I walk.

[113] ἀήρ, ἀέρος, ὁ, the air (particularly the lower distinguished from the higher and rarer ether or the atmospheric region.

[114] API1P ἀναστρέφω I return, live, dwell in a place, walk, conduct or behave oneself.

[115] PAPtcpNMP ποιέω I do, make.

[116] Mind, understanding, thinking, desiring.

[117] ImpfAI1P εἰμὶ I am.

[118] PAPtcpNMS εἰμὶ I am.

[119] AAI3S ἀγαπάω I love.

[120] PAPtcpAMP εἰμὶ I am.

[121] AAI3S συζωοποιέω I make alive together with. Elsewhere in NT/LXX only at Colossians 2:13.

[122] PAI2P εἰμὶ, I am.

[123] PfPPtcpNMP σώζω I save; periphrastic tense, being perfect.

[124] AAI3S συνεγείρω I raise together with; Used in the LXX of a donkey lifted up with help of a passerby (Ex 23:5) and of dead Kings, who are raised up together to meet the defeated King of Babylon as they rejoice in Babylon’s downfall (The rulers … are raised together for you: Isaiah 14:9). Also Col 2:12, 3:1.

[125] AAI3S συγκαθίζω I sit together with; used in LXX Gen 15:11, Ex 18:13, Nu 22:27; Jer 16:8, also in Lk 22:55.

[126] AMSubj3S ἐνδείκνυμι I show, demonstrate.

[127] PM(d)PtcpDPM ἐπέρχομαι I am coming to, upon.

[128] PAPtcpASNtr ὑπερβάλλω I go beyond, surpass.

[129] PAI2P εἰμὶ I am.

[130] PfPPtcpNMP σώζω I save; periphrastic tense, being perfect.

[131] AM(d)Subj3S καυχάομαι I boast.

[132] PAI1P εἰμὶ I am.

[133] Work, creation, deed, act.

[134] APPtcpNPM κτίζω I create, make, found, establish

[135] Of purpose and end, unto, for.

[136] AAI3S προετοιμάζω I prepared in advance, beforehand; cf. Rom 9:23.

[137] AASubj1P περιπατέω I walk.

[138] Seems to refer to Paul’s mainly Gentile readers in Asia Minor; compare ‘you Gentiles in the flesh’ (Eph 2:11) and see O’Brien, 156. Paul will quickly add ‘we also’ (v. 3), which probably denotes himself and his fellow Jews.

[139] O’Brien takes as causal, because of.

[140] PAPtcpAMP εἰμὶ I am.

[141] Ronald E Heine, The Commentaries of Origen and Jerome on St Paul's Epistles to the Ephesians: Oxford Early Christian Studies (Oxford: Oxford University Press, 2002), 125: cf. ‘Now indeed one who is rather simple will answer that he has asserted this because that which will be has been said as if it has already happened in accordance with the foreknowledge of God and that it is the custom of the Scriptures occasionally that they are inflected in the past for future time.’: ibid, 126.

[142] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 95, ‘John Damascene’ volume 2: (various), Sacra Parallela, retrieved from https://archive.org/details/patrologicursus50migngoog on 21 September 2016, Col 829A-832A; also at http://litteralchristianlibrary.wikifoundry.com/page/Commentary+on+Ephesians+by+St.+John+of+Damascus.

[143] Ellicott, Ephesians, 35.

[144] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 95, ‘John Damascene’ volume 2: (various), Sacra Parallela, retrieved from https://archive.org/details/patrologicursus50migngoog on 21 September 2016, Col 829A-832A.

[145] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 118, Oecumenius Vol 1: Commentary on Acts, Commentary on Paul’s letters, Commentary on the Catholic letters (Migne, PG), 118:1192B-1193B at http://www.archive.org/details/patrologicursus21migngoog accessed 1 October 2016. I have checked and occasionally emended Mignes text based on ‪Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu: Triccae episcopus Oecumenius (Athens, 1844), 108-9.

[146] Chrysostom, Homilies on Ephesians IV. Accessed at http://www.ccel.org/ccel/schaff/npnf113.iii.iv.v.html on 12 August 2016; Greek Text from J-P Migne (ed), Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca Volume 62 S. Joannes Chrysostomus Vol 11 (Paris, 1862), Col 33-34, accessed at http://books.google.com/books?id=E_gbZgKru-QC on 12 August 2016.

[147] J –P Migne, Patrologia Latina (Patrologiae Cursus Completus. Series Latina) Jerome Vol 26 Col 470A-B [577]- Col 471A-C [579] at http://books.google.com/books?id=UfkUAAAAQAAJ accessed on 1 October 2016.

[148] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 118, Oecumenius Vol 1: Commentary on Acts, Commentary on Paul’s letters, Commentary on the Catholic letters (Migne, PG), 118:1192B-1193B at http://www.archive.org/details/patrologicursus21migngoog accessed 1 October 2016. I have checked and occasionally emended Mignes text based on ‪Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu: Triccae episcopus Oecumenius (Athens, 1844), 108-9.

[149] D B Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996), 334-5.

[150] Chrysostom, Homilies on Ephesians IV, accessed at http://www.ccel.org/ccel/schaff/npnf113.iii.iv.v.html on 12 August 2016; Greek Text from J-P Migne (ed), Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca Volume 62 S. Joannes Chrysostomus Vol 11 (Paris, 1862), Col 33-34, accessed at http://books.google.com/books?id=E_gbZgKru-QC on 12 August 2016. Cf. W A Jurgens (Ed and Trans), The Faith of the Early Fathers: Post-Nicene and Constantinopolitan eras (Collegeville: Liturgical, 1979), Vol 2:120.

[151] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 82 Theodoret volume 3: Commentary on Paul’s letters, 82:521-2, accessed at http://www.archive.org/details/patrologiaecurs46migngoog and https://ia800307.us.archive.org/35/items/patrologiaecurs46migngoog/patrologiaecurs46migngoog.pdf on 13 August 2016. W A Jurgens, The Faith of the Early Fathers: 3 Volumes (Collegeville: Liturgical, 1970-79), 3:248-9 sec 2163, cited in T Oden, The Justification Reader, 44.

[152] J –P Migne, Patrologia Latina (Patrologiae Cursus Completus. Series Latina) Jerome Vol 26 Col 470A-B [577]- Col 471A-C [579] at http://books.google.com/books?id=UfkUAAAAQAAJ accessed on 1 October 2016. Cf. M J Edwards, Galatians, Ephesians, Philippians: ACCS NT 8 (Downers Grove: IVP, 2005), 126, citing Jerome (c 347-420), Epistle to the Ephesians, 1.2.8-9. Ronald E Heine, The Commentaries of Origen and Jerome on St Paul's Epistles to the Ephesians: Oxford Early Christian Studies (Oxford: Oxford University Press, 2002), 129-130.

[153] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 118, Oecumenius Vol 1: Commentary on Acts, Commentary on Paul’s letters, Commentary on the Catholic letters (Migne, PG), 118:1192B-1193B at http://www.archive.org/details/patrologicursus21migngoog accessed 1 October 2016; ‪Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu: Triccae episcopus Oecumenius (Athens, 1844), 108-9.

[154] As is well known, Theophylact also gives a second interpretation: ‘Or [we can] also [take it] in another way: He is not saying [that] faith is the gift of God, but that to be saved through faith, this is the gift of God. For let us have it established that faith is our own: How could it [ie faith] alone be sufficient to save [us], unless God is pleased to receive us through it? ( Ἤ καὶ ἄλλως · Oὐ τὴν πίστιν λέγει δῶρον θεοῦ, ἀλλὰ τὸ διὰ πίστεως σωθῆναι, τοῦτο δῶρόν ἐστι θεοῦ. Ἔστω γὰρ, ὅτι ἡ πίστις ἡμέτερον · πῶς ἄν ἴσχυσεν αὕτη μόνη σῶσαι, εἰ μὴ ὁ θεὸς ηὐδόκησε δεχθῆναι δι᾽ αὐτῆς ἡμᾶς;’ Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu PIaulu: Triccae episcopus Oecumenius (Athens, 1844), 108 fn (1); Also J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 124 Theophylact of Bulgaria Volume 2: Commentarius in Joannis Evangelium (contd.); Commentary on Paul’s letters, retrieved from http://books.google.com/books?id=lzURAAAAYAAJ on 7 October 2016, Col 1056B-1057A.

[155] P Schaff (ed), NPNF1 Vol 5, 504.

[156] Mark J Edwards (ed), Galatians, Ephesians, Philippians: Ancient Christian Commentary on Scripture, New Testament VI: (Downers Grove: InterVarsity Press, 1998), 133-134; also cited in Oden, Justification Reader, 48. Latin original: CCL 91:313.

[157] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 95, ‘John Damascene’ volume 2: (various), Sacra Parallela, retrieved from https://archive.org/details/patrologicursus50migngoog on 21 September 2016, Col 829A-832A; also at http://litteralchristianlibrary.wikifoundry.com/page/Commentary+on+Ephesians+by+St.+John+of+Damascus.

[158] D B Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996), 334-5.

[159] The versification and punctuation cannot be considered original for the purposes of exegesis but editorial.

[160] Chrysostom, Homilies on Ephesians IV. Accessed at http://www.ccel.org/ccel/schaff/npnf113.iii.iv.v.html on 12 August 2016; Greek Text from J-P Migne (ed), Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca Volume 62 S. Joannes Chrysostomus Vol 11 (Paris, 1862), Col 33-34, accessed at http://books.google.com/books?id=E_gbZgKru-QC on 12 August 2016.

[161] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 82 Theodoret volume 3: Commentary on Paul’s letters, 82:521-2, accessed at http://www.archive.org/details/patrologiaecurs46migngoog and https://ia800307.us.archive.org/35/items/patrologiaecurs46migngoog/patrologiaecurs46migngoog.pdf on 13 August 2016.

[162] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 95, ‘John Damascene’ volume 2: (various), Sacra Parallela, retrieved from https://archive.org/details/patrologicursus50migngoog on 21 September 2016, Col 829A-832A; also at http://litteralchristianlibrary.wikifoundry.com/page/Commentary+on+Ephesians+by+St.+John+of+Damascus.

[163] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 118, Oecumenius Vol 1: Commentary on Acts, Commentary on Paul’s letters, Commentary on the Catholic letters (Migne, PG), 118:1192B-1193B at http://www.archive.org/details/patrologicursus21migngoog accessed 1 October 2016. I have checked and occasionally emended Mignes text based on ‪Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu: Triccae episcopus Oecumenius (Athens, 1844), 108-9.

[164] Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu PIaulu: Triccae episcopus Oecumenius (Athens, 1844), 108 fn (1); Also J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 124 Theophylact of Bulgaria Volume 2: Commentarius in Joannis Evangelium (contd.); Commentary on Paul’s letters, retrieved from http://books.google.com/books?id=lzURAAAAYAAJ on 7 October 2016, Col 1056B-1057A.

[165] John M G Barclay, Paul & the Gift (Grand Rapids/Cambridge: Eerdmans, 2015), 562-571. Barclay helpfully notes that the ancient world had elements of mutuality and reciprocity that might be unconditioned (and this is reflected particularly in Paul by the context of human sin), but was not unconditional (because of the social obligations of mutuality the gift imposed). For example, we see this reflected in Romans 11:35 (or who has given to him and will receive back from him?; ἢ τίς προέδωκεν αὐτῷ, καὶ ἀνταποδοθήσεται αὐτῷ). Paul in particular accents the notion of incongruity of grace in that it is given to the undeserving, and yet this gift is not received without the reciprocal obligation of Christ-like living from the recipients. However, it is a pity that in his detailed exploration of the concept of ‘gift’ in Paul, Barclay, because he sees Ephesians as ‘Deutero-Pauline’, only deals with Ephesians 2:8-10 through the lens of the disciples of Paul who he assumes wrote Ephesians, and Augustine. So Barclay characterizes Ephesians 2:8-10 as the result of Paul’s ‘missionary theology […] turned inwards’. Ephesians 2:8-10 represents the view of later theologians who drew lines of division within the Christian community. The works-faith dichotomy, which Barclay does not believe truly belongs to Paul, ‘became a tool for the inner reform of the Christian tradition, its critical edge turned against believers’ (op cit, 570-1). Part of this, which I see to be an error, stems not only from accepting critical views of the authorship of Ephesians and the Pastorals, but also because he thinks that in the ‘undisputed Paulines’, ‘works of the law’ means Jewish practices: ibid, 567.

[166] Barclay, op cit, 577-8.

[167] Reginald Stewart Moxon (ed), The Commonitorium of Vincentius of Lerins (Cambridge: CUP, 1915), xxviii-xxix; Peter Holmes and Robert Ernest Wallis (trs), Benjamin B Warfield (rev), Philip Schaff (ed), NPNF1 Vol 5, (Buffalo, NY: Christian Literature Publishing Co, 1887), revised and edited for New Advent by Kevin Knight <http://www.newadvent.org/fathers/15122.htm>.

[168] NPNF1 Vol 5 at http://www.newadvent.org/fathers/15061.htm.

[169] Mark J Edwards (ed), Galatians, Ephesians, Philippians: Ancient Christian Commentary on Scripture, New Testament VI: (Downers Grove: InterVarsity Press, 1998), 133-134; also cited in Oden, Justification Reader, 48. Latin original: CCL 91:313.

[170] Barclay, op cit, 571.

[171] Ibid, 567-8.

[172] Ibid, 373-5.

[173] Ibid, 570-1.

[174] Sprinkle, Paul and Judaism Revisited, 186-7; Fitzmyer, Romans, 338-9, idem, ‘Justification by Faith in Pauline Thought: A Catholic View’ in D E Aune (ed), Rereading Paul Together: Protestant and Catholic Perspectives on Justification (Grand Rapids: Baker, 2006), 77-94 at 88-90; Cranfield, Romans, 1:198; D J Moo, ‘”Law,” “Works of the Law,” and Legalism in Paul’, WTJ, 45 (1983), 73-100; idem, Romans, 208-10; idem, ‘The Law of Christ as the Fulfillment of the Law of Moses: Modified Lutheran View’ in Five Views on Law and Gospel, (Grand Rapids: Zondervan, 1996), 319-376 at 328-333; Schreiner, Romans, 168-75; idem, Paul, 110-15; Contra Dunn, Romans, 1:153-5; Jewett, Romans, 266.

[175] Moo, ‘“Law”, “Works of the Law”’, 92.

[176] F Watson, Paul and the Hermeneutics of Faith (London: T & T Clark, 2004), 334.

[177] Moo, Romans, 217; cf idem, ‘”Law,” ”Works of the Law”’, 98.

[178] Sprinkle, Paul and Judaism Revisited, 187.

[179] J Calvin, Institutes of the Christian Religion, ed J T McNeill; tr F L Battles; Library of Christian Classics Vol XX (Philadelphia: Westminster, 1960), III, XIV, 11, (p778-9).

[180] Moo, Romans, 76.

[181] Chrysostom, Homilies on Ephesians IV. Accessed at http://www.ccel.org/ccel/schaff/npnf113.iii.iv.v.html on 12 August 2016; Greek: J-P Migne, Vol 62 Chrysostom Vol 11, Col 33-34, accessed at http://books.google.com/books?id=E_gbZgKru-QC on 12 August 2016.

[182] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Vol 118, Oecumenius Vol 1: Commentary on Acts, Commentary on Paul’s letters, Commentary on the Catholic letters (Migne, PG), 118:1192B-1193B at http://www.archive.org/details/patrologicursus21migngoog accessed 1 October 2016. I have checked and occasionally emended Mignes text based on ‪Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu. (Commentarius in apostolorum facta et epistolas Pauli.): 5. 6; ‪ Biblia Graece: He Kaine Diatheke 5.6 Volume 1 of Exegis eis tas praxeis ton apostolon kai eis tas epistolas tu Paulu: Triccae episcopus Oecumenius (Athens, 1844), 108-9.

[183] Alan P Stanley, Did Jesus Teach Salvation By Works?: The Role of Works in Salvation in the Synoptic Gospels: Evangelical Theological Society Monograph Series 4 (Eugene, OR: Wipf & Stock, 2006), 24.

[184] On Grace and Free Will, chapter 17 in NPNF1 Vol 5 at http://www.newadvent.org/fathers/1510.htm, accessed on 10 June 2017.

[185] However, in On Grace and Free Will, chapter 20, Augustine considers that ‘not of works, lest any man should boast’ means that Paul foresaw the possibility that ‘men would think from this statement that good works are not necessary to those who believe, but that faith alone suffices for them; and again, the possibility of men's boasting of their good works, as if they were of themselves capable of performing them.’ NPNF1 Vol 5 at http://www.newadvent.org/fathers/1510.htm, accessed on 10 June 2017. My first point is that for Augustine, Paul’s rejection of ‘works’ is only a rejection of ‘works’ that we think come from ourselves. “Not of works” is spoken of the works which you suppose have their origin in yourself alone; but you have to think of works for which God has moulded (that is, has formed and created) you. But since according to Augustine, any good works any regenerate person did would be from grace, the fact is that ‘not of works’ must be merely a statement of pre-baptismal works, together with wrong thinking that the regenerate apply to their particular actual good works. This is because there is no such good work that comes from ourselves, except our sin. Thus, on Augustine’s interpretation, the ‘works’ that are excluded from our obtaining faith or salvation are in fact works that are not found in the regenerate anyway. Such is tantamount to saying that ‘the saved are not saved by something that does not exist’. Well of course! That is self-evident. So the works we are not saved by, according to Augustine, are actually non-existent. But on Augustine’s reasoning, the regenerate do have good works which save them, but they are the result of grace, being God’s enabling. Or to put the same point another way, Augustine holds that the ‘works’ that we are explicitly said not to be saved by in verse 9 are different from the ‘good works’ we are explicitly said to be prepared for.

And this then leads us to the second point: that Augustine defines ‘grace’ as the ‘good works’ of the regenerate. For Augustine says, ‘It follows, then, dearly beloved, beyond all doubt, that as your good life is nothing else than God's grace, so also the eternal life which is the recompense of a good life is the grace of God’. A good life and good works are effectively synonymous. Living a good life and doing good works empowered by God is, for Augustine, God’s grace. Therefore, we are in fact saved by our good works, because we are saved by grace, and our good works or good lives ‘are’ God’s grace, and the result of and consequence of God’s prevenient and enabling grace. Good works in this way are included in our salvation—because they are in fact ‘God’s grace’.

So the ironic situation is that, for Augustine, though Paul says that faith is not from ourselves, a gift, and not by work, the fact is that Augustine says that our salvation is indeed by works because our works are in reality God’s ‘grace’. There is a problem with this approach, however, because Romans 11:6 says grace is not of works, lest grace not be grace (εἰ δὲ χάριτι, οὐκέτι ἐξ ἔργων, ἐπεὶ ἡ χάρις οὐκέτι γίνεται χάρις). This inconsistency in Augustine was remedied at the reformation by Calvin, who systematically distinguished justification from sanctification and regeneration.

[186] http://www.ccel.org/ccel/schaff/npnf113.iii.iv.v.html; J –P Migne, Vol 62, Chrysostom Vol 11, Col 34 accessed on 19 August 2016 at http://books.google.com/books?id=E_gbZgKru-QC.

[187] J –P Migne, Patrologia Graeca (Patrologiae Cursus Completus. Series Graeca) Volume 82 Theodoret volume 3: Commentary on Paul’s letters, 82:521-2, accessed at http://www.archive.org/details/patrologiaecurs46migngoog and https://ia800307.us.archive.org/35/items/patrologiaecurs46migngoog/patrologiaecurs46migngoog.pdf on 13 August 2016.

[188] J –P Migne, Patrologia Latina (Patrologiae Cursus Completus. Series Latina) Jerome Vol 26 Col 470A-B [577]- Col 471A-C [579] at http://books.google.com/books?id=UfkUAAAAQAAJ accessed on 1 October 2016.

[189] M J Harris, ‘Prepositions and Theology in the Greek New Testament’, in C Brown (ed), New International Dictionary of New Testament Theology (Grand Rapids: Zondervan, 1971), 3:1193.

[190] E Best, Ephesians: ICC (London/New York: T & T Clark, 1998), 231.

[191] Ibid, 232.