Where the world comes to study the Bible

The Syntax of Romans 3:22-24 Part 1

Leon Morris called Rom 3.21-26 “possibly the single most important paragraph ever written,”1 a sentiment shared by many exegetes and theologians. There are several interpretive difficulties in this short paragraph, however, that give commentators some grief. One of them is the relation of v. 23 to v. 24; bound up in that issue is how the participle that leads off v. 24, δικαιούμενοι, should be taken. As for v. 24, Edwards has summed up its significance: “In all Scripture there is probably no verse which captures the essence of Christianity better than this one. Here is the heart of the gospel, the mighty Nevertheless, the momentous divine reversal. Everything in verse 23 was due to humanity; everything in verse 24 depends on God.”2

In the NET Bible, the paragraph read as follows:

(21) But now apart from the law the righteousness of God (which is attested by the law and the prophets) has been disclosed—(22) namely, the righteousness of God through the faithfulness of Jesus Christ for all who believe. For there is no distinction, (23) for all have sinned and fall short of the glory of God. (24) But they are justified freely by his grace through the redemption that is in Christ Jesus. (25) God publicly displayed him at his death as the mercy seat accessible through faith. This was to demonstrate his righteousness, because God in his forbearance had passed over the sins previously committed. (26) This was also to demonstrate his righteousness in the present time, so that he would be just and the justifier of the one who lives because of Jesus’ faithfulness.

The NET note on “But they are justified” at the beginning of v. 24 says, “Grk “being justified,” as a continuation of the preceding clause. Because of the length and complexity of the Greek sentence, a new sentence was started here in the translation.” This will be an important consideration as we examine this text in some detail. For now, it should just be noted.

There are of course several interpretive routes that the NET has taken on this passage that some exegetes would question (e.g., “faithfulness of Jesus Christ” in v. 22; “mercy seat” instead of “propitiation” in v. 25), but our focus in this brief essay is on the relationship of v. 23 to v. 24. Verse 22 has to be brought into the mix as well because it impacts the discussion.

The Greek text of Rom 3.21-26 reads as follows:

(21) Νυνὶ δὲ χωρὶς νόμου δικαιοσύνη θεοῦ πεφανέρωται μαρτυρουμένη ὑπὸ τοῦ νόμου καὶ τῶν προφητῶν (22) δικαιοσύνη δὲ θεοῦ διὰ πίστεως ᾿Ιησοῦ Χριστοῦ εἰς πάντας τοὺς πιστεύοντας. οὐ γάρ ἐστιν διαστολή(23) πάντες γὰρ ἥμαρτον καὶ ὑστεροῦνται τῆς δόξης τοῦ θεοῦ (24) δικαιούμενοι δωρεὰν τᾐ αὐτοῦ χάριτι διὰ τῆς ἀπολυτρῶσεως τῆς ἐν Χριστῷ ᾿Ιησοῦ· (25) ὃν προέθετο ὁ θεὸς ἱλαστήριον διὰ [ τῆςV πίστεως ἐν τῷ αὐτοῦ αἵματι εἰς ἔνδειξιν τῆς δικαιοσύνης αὐτοῦ διὰ τὴν πάρεσιν τῶν προγεγονότων α῾μαρτημάτων (26) ἐν τᾐ ἀνοχᾐ τοῦ θεοῦπρὸς τὴν ἔνδειξιν τῆς δικαιοσύνης αὐτοῦ ἐν τῷ νῦν καιρῷεἰς τὸ εἶναι αὐτὸν δίκαιον καὶ δικαιοῦντα τὸν ἐκ πίστεως ᾿Ιησοῦ.

What is the basic syntactical problem in Rom 3.23-24? It is how to take the participle δικαιούμενοι at the beginning of v. 24. And in terms of ‘how to take it,’ we mean whether it is independent or dependent, and if dependent, what it is dependent on. Theologically, much more seems to be at stake. For example, Morris notes that “Grammatically δικαιούμενοι should go with πάντες. But while it is certainly the case that all sin, it is not the case that all are justified.”3

Morris has identified two assumptions that are behind a lot of exegetical discussion of this text: First, the grammar of v. 24 naturally shows the participle δικαιούμενοι to be dependent on the πάντες of v. 23. Thus, “all have sinned… being freely justified” would be the natural sense to get from the construction. But second, such a sense, though grammatically proper, is theologically off-base. Why? Because not all sinners are saved. Or to put it otherwise, universal salvation is not in view in these verses. To be sure, some exegetes may want to see some sort of universal salvation, but that will simply not work with Paul’s prior restriction mentioned in v. 22 (“for all who believe”). That is, this justification, this salvation, is applied only to believers. Further, when Paul prays for his fellow Jews in Romans 9, he wishes that he could be sent to hell if that would save but one of them! Why would this even be contemplated if everyone is saved? Something is obviously amiss in a flat reading of Rom 3.24, even though the syntax seems to clearly point in this direction.4

How have exegetes resolved the issue? Three different options are generally offered.

(1) A prima facie reading of the syntax is acknowledged, but the force given to δικαιούμενοι is restricted. So, Morris: “The meaning appears to be that all who are justified are justified in this way.”5 That is, the participle is viewed as offering only potential, not actual, justification to all sinners. They have to believe if it is to be realized.

But this view is unlikely for at least two reasons. First, Morris has subtly shifted the syntax. In effect, he is saying that if one is to be justified, it will occur “freely,” “by his grace,” “through the redemption that is in Christ Jesus.” Or, to put it in Morris’ words, “all who are justified are justified in this way.” But the Greek does not restrict the ‘all’ of v. 23 to ‘all who are justified.’ Rather, it seems to plainly say that all have sinned and these all are justified ( πάντες ἥμαρτον... δικαιούμενοἵ. Only by treating the participle as a tendential or voluntative present is it possible to see the justification as only potential rather than actual. But even here, the tendential present does not quite fit because the tendential present indicates that “an attempt is about to be made or …is desired to be made in the present time (or, very near future tim . The action may or may not be carried out.”6 The participle in Rom 3.24 does not, in Morris’ view, represent an attempt or desire, just a potentiality. If we were to apply this approach to three verses earlier we would get a nonsensical idea: δικαιοσύνη... μαρτυρουμένη: “the righteousness of God was revealed, being potentially testified by the law and the prophets” (Rom 3.21)! What needs to be established before Morris’ suggestion can even be entertained is that a present adverbial participle, subordinate to an aorist indicative, can indicate only potentiality.

Second, and more broadly, to argue that the justification in v. 24 is only potential seems to be a desperate move. Not only does it not fit the syntax of the construction well, it also is simply too convenient. The prima facie syntax screams that all the sinners of v. 23 are justified—not potentially, but actually. We cannot resolve this problem by reinventing how language works.

(2) Verse 23 is seen as parenthetical; the participle in v. 24 thus reaches back to v. 22. This was Morris’ second option: “Or there may be a reference back to the πιστεύοντες of v. 22… If we ask, ‘Who are the δικαιούμενοι?’ we must answer, ‘The πιστεύοντες’.”7 What is overlooked in this explanation is that the participle in v. 22 is not πιστεύοντες; rather, it is πιστεύοντας. In other words, it is accusative rather than nominative. That is no small matter: the lack of concord with the participle in v. 22 underscores the ease with which we could take δικαιούμενοι back to πάντες in v. 23, and the difficulty with which we can simply pass over it in search of a more suitable subject—especially one that lacks complete grammatical concord with δικαιούμενοι. Yet, the parenthetical view is popular in exegetical literature on Romans, largely because it gives a much more satisfactory theological sense than subordinating δικαιούμενοι to πάντες. It is held by Käsemann, Michel, Moo, Murray, and several others.

Exegetes usually regard not just v. 23 but vv. 22b-23 (“for there is no distinction. For all have sinned and fall short of God’s glory”) to be parenthetical. If so, then we should read δικαιούμενοι as subordinate to πάντας in v. 22: “even the righteousness of God which comes through {Jesus Christ’s faithfulness/faith in Jesus Christ} to all who believe…being freely justified…”

But doing this is awkward. Bypassing a subject ( πάντες in v. 23) that has complete grammatical concord and is in much closer proximity to the participle ( δικαιούμενοι in v. 24) for one that lacks complete concord ( πάντας) and is much farther away, looks like an expedient intended to remove a theological difficulty. Sanday and Headlam admit, “Easier and more natural than any of these expedients seems to be…to make οὐ γάρὑστεροῦνται practically a parenthesis, and to take the nom. δικαιούμενοιas suggested by πάντες in ver. 23, but in sense referring rather to τοὺς πιστεούοντας in ver. 22.’ No doubt such a construction would be irregular, but it may be questioned whether it is too irregular for St. Paul.”8 In other words, it’s possible. But is it likely?

Indeed, if the theological difficulty were not present, few would consider this as a viable option.

(3) Because of what looks to be a natural connection of δικαιούμενοι with πάντες, some exegetes have built on the parenthetical view by suggesting that δικαιούμενοι functions like an indicative. For example, Dunn speaks of “the passive indicative participle”9 (which presumably is a participle that functions like an indicativ , without further comment.10 Michel elaborates, “An die negative Aussage von V 23 schließt sich jetzt eine positive ohne Verknüpfung an. Statt des Partizips δικαιούμενοι erwartet man eigentlich eine indicative Verbform (wie δικαιοῦνται δέ). Paulus will den Begriff δικαιοσύνη θεοῦ (V 21.22) durch die Partizipialaussage unseres Verses näher bestimmen, so daß der Gedankengang an V 21.22 anknüpft und V 23 wie eine Parenthese wirkt.”11

If the participle functions as an indicative, there is no need to find concord with what precedes. But there is usually little argument offered in behalf of this view. Several scholars translate the participle as an indicative without any justification at all (e.g., Bruce, Moo, Schreiner, Stott, etc.). The rationale for an indicative force is as follows: First, such usage, though rare, is not without examples in the New Testament. Second, Paul in particular uses it, especially in Romans (cf. 5.11; 12.6 ). Third, it avoids the problem of the simple parenthetical view which bypasses concord with the πανvτες of v. 23 for the πάντας of v. 22, creating the tension of the misfit on the cases. And fourth, it creates a satisfactory theological sense, if 22b-23 is considered parenthetical: “all who believe… they are justified…”

This view, however, is not without its problems. The major problem is simply that it looks like a means to get around the obvious exegetical difficulty created by subordinating v. 24 to v. 23. But it is doubly convoluted: not only does it posit a parenthesis for vv. 22b-23, but also elicits a rare usage of the participle in v. 24.12 There are probably no more than a couple dozen instances of participles functioning as indicatives in the New Testament. Though it is true that Paul does use this construction in Romans (5.11; 12.6),13 it is nevertheless quite rare. Further, its use is demanded in Rom 5.11 and 12.6: there is no verb that the participle could be dependent on in those texts. But that is not the case in Rom 3.24. The verb ὑστεροῦνται in v. 23 naturally presents itself as that to which δικαιούμενοι is subordinate. Brooks and Winbery offer the sober assessment that “Certainly no participle should be explained as an independent participle if there is any other way to explain it.”14 At bottom, the indicative use of the participle is possible only because the theological ramifications of taking as a dependent participle are unacceptable. But it is certainly too much to assume that the syntax of v. 24 must be “awkward” (to use Dunn’s vocabulary) just because the natural syntactical connection is theologically unacceptable.

In conclusion (of part 1), we have found the standard treatments of the syntax of Rom 3.23-24 to be less than satisfactory. If theological factors were not a consideration, most likely the parenthetical view would not have been suggested, nor especially the participle as indicative view. In part 2, we will examine the major underlying presupposition that leads exegetes to try a ‘work-around’ to the most likely syntactical connections here, and offer a solution that retains such connections.


1 Leon Morris, The Epistle to the Romans, The Pillar New Testament Commentary (Grand Rapids: Eerdmans, 1988) 173.

2 James R. Edwards, Romans, New International Biblical Commentary (Peabody, MA: Hendrickson, 1992) 102.

3 Ibid., 177, n. 113.

4 It is for this reason that many exegetes speak of the construction as puzzling. It is not that the syntax really is at all unusual; rather, if the normal rules of syntax apply here, then the theology causes difficulty. But the many discussions in which the syntax itself seems to be a conundrum really involve an ellipsis. For example, Sanday and Headlam assert that “The construction and connexion of this word [ δικαιούμενοι] are difficult, and perhaps not to be determined with certainty” (85). John Murray, Romans 113-14: “Commentators have encountered difficulty with the construction at the beginning of verse 24. The participle ‘being justified’ does not appear to stand in relation to what precedes in a way that is easily intelligible.” J. D. G. Dunn speaks of “the awkwardness of the syntax” (168) and that “the syntactical link with the preceding context is obscure” (168).

5 Morris, 177, n. 113.

6 D. B. Wallace, Exegetical Syntax, 535.

7 Morris, 177, n. 113.

8 Sanday and Headlam, Romans, ICC, 85-86.

9 J. D. G. Dunn, Romans 1–8, Word Biblical Commentary (Nashville: Thomas Nelson, 1988) 168.

10 Moo also seems to hold to this view, implied in his NIV Application Commentary on Romans (Grand Rapids: Zondervan, 2000) 128.

11 Otto Michel, Der Brief an die Römer, MeyerK (Göttingen: Vandenhoeck & Ruprecht, 1966 [5th Auflage]) 149.

12 Wallace, Exegetical Syntax, 653, lists 15 instances, most of which occur in the Apocalypse.

13 The participle also functions as an imperative in Rom 12.9, 11, 12, 13, 14, 16, 17, 18, 19.

14 J. A. Brooks and C. L. Winbery. Syntax of New Testament Greek (Washington, DC: University Press of America, 1979) 138 (italics in original).

Related Topics: Grammar