JBL 113/2 ß994) 265-285

ROMANS 1:17-
A CRUX INTERPRETUM FOR
THE ΠΙΣΤΙΣ ΧΡΙΣΤΟΥ DEBATE
DOUGLAS A. CAMPBELL
University of Otago, Dunedin, New Zealand

An important debate that has been gathering momentum in Pauline circles

1
of late is the πίστις Χρίστου dispute. Since an opening round on the Continent
2
at the turn of the century, this has been a largely postwar phenomenon of
3
the English-speaking world, and lately a notably North American preoccupa­
4
tion. Interest there has culminated recently in an exciting "head-to-head"
debate between Richard Hays and James D. G. Dunn in the Pauline Theology
Group of the Society of Biblical Literature in Kansas City in November 1991,
which elicited definitive statements of both sides of the issue from those two

1
Good summaries of the debate, and bibliographies, may be found in Richard B. Hays, The
Faith of Jesus Christ (Chico, CA: Scholars Press, 1983) 158-62; idem, "ΠΙΣΤΙΣ and Pauline
Christology: What Is at Stake?" in SBL 1991 Seminar Papers (ed. David J. Lull; Atlanta: Scholars
Press, 1991) 714-29, esp. nn. 2-4, pp. 714-15; and George Howard, "Faith of Christ," ABD 2. 760;
see also my The Rhetoric of Righteousness in Romans 3:21-26 (Sheffield: JSOT Press, 1992) 58-60.
2
Prompted by Johannes Haußleiter's study "Der Glaube Jesu Christi und der chrisdiche Glaube,"
NKZ 2 (1891) 109-45, 205-30; initially endorsed by Gerhard Kittel ("πίστις Ίησοΰ Χρίστου bei
Paulus," TSK 79 [1906] 419-36) and G. A. Deissmann (Paul: A Study in Social and Religious History
[1912; 2d ed. trans. W. E. Wilson; New York: Harper & Row, 195η 161-65). A strong German
counterattack effectively closed off the debate, however, to the point that Rudolf Bultmann scarcely
mentions it in his article on πκιτ- words in TDNT (6. 204 n. 230 —Hays lists the opposition of
W. H. P. Hatch, O. Schmitz, E. Wissman, and W. Mundle [Faith of Jesus Christ, p. 185 n. 88]).
3
A. G. Hebert and T. F. Torrance introduced the debate to English-speaking scholarship, prob­
ably via their exposure to Karl Barth, who was sympathetic to the view (see Hebert, "Faithfulness
and Faith," Reformed Theological Review 14 [1955] 33-40, repr. in Theofogy 58 [1955] 373-79; and
Torrance, "One Aspect of the Biblical Conception of Faith," ExpTim 68 [195η 111-14). This also
excited a strong counterattack (see esp. C. F. D. Moule, "The Biblical Conception of Faith," ExpTim
68 [195η 157; and James Barr, The Semantics of Biblical Language [London: Oxford University
Press, 1961] 161-205), but one that failed to choke off discussion completely.
4
Hays notes the spate of recent North American discussion (see "ΠΙΣΤΙΣ and Pauline
Christologyf nn. 2 and 3, pp. 714-15), prompted probably in large measure by his own monograph
on the subject (Faith of Jesus Christ), and also by his continued advocacy of the view, which
influences SBL discussions of Paul. Earlier North American proponents (like George Howard
and Richard N. Longenecker) seem to have picked up the position from the Scottish advocates
like the Torrances (for their most recent statements, see Howard, "Faith of Christ," ABD 2. 758-60;
Longenecker, Galatians [Dallas: Word, 1990] 87-88).

265
266 Journal of Biblical Literature

5
protagonists, as well as a host of oral reflections and brief responses by
interested members of the group. Thus, it is clearly an appropriate moment
to ask (after the dust has settled a little) whether the debate has run its
course— or whether the course has only just begun. Actually I would suggest
that, notwithstanding its significance to date, the debate is becoming increas­
ingly misdirected. Furthermore, only if certain important navigational errors
are corrected will it continue to progress, ultimately to make a truly significant
contribution to our understanding of Paul. As things stand, there is a danger
that discussion is running into a cul-de-sac.
The danger of irrelevance stems from two problems with the present
discussion. On the one hand, the broader implications of the issue for Paul's
understanding of πιστ- language, and for his basic understanding of salvation,
have not been clearly articulated. These are, of course, fundamentally theo­
logical issues—hence, perhaps, the reluctance of NT scholars to engage or
even to acknowledge them. But they are undeniably present and mold the
current debate powerfully.6 Only if these agendas are exposed and analyzed
critically can the data of the debate be handled accurately. Moreover, it should
then be possible to see how this localized debate concerning various genitive
constructions has direct and massive implications for the whole shape of
Pauline (and, indeed, NT) theology.7
On the other hand, the debate has —rather incredibly—failed to isolate
the critical texts sufficiently and tends to merge discussion of a half dozen
or so genitive constructions with quite sweeping characterizations of Paul's
argument and theology, particularly as found in Galatians and Romans (and
in which, for example, Abraham tends to figure quite prominently, although
none of the disputed genitives actually refers to him).8 Given the difficulty
of, and contentions surrounding, these letters, it is not surprising that such
discussions fail to reach much resolution — if a particular reading of Gal
2:15-4:10 must first be granted, then we will be waiting some time for a

5
See Hays, "ΠΙΣΤΙΣ and Pauline Chnstology", and James D G Dunn, "Once More, ΠΙΣΤΙΣ
ΧΡΙΣΤΟΥ," m SBL 1991 Seminar Papers (ed David J Lull, Atlanta Scholars Press, 1991) 730-44
6
These forces are plainly at work in Dunn's SBL paper (and response), for example, he states
"on Hays' thesis we have no clear reference to the 'faith' of believers This is nothing short
of astonishing It now appears that a text (Galatians), which has provided such a powerful charter
of justifying faith' for Christian self-understanding, nowhere clearly speaks ofthat 'faith'"— Dunn
sounds positively Reformational here' ("Once More," 736)
7
For example, Hays states that " the emphasis in Paul's theology lies less on the question
of how we should dispose ourselves towards God than on the question of how God has acted
m Christ to effect our deliverance" ("ΠΙΣΤΙΣ and Pauline Chnstology," 715-16)
8
Hays simply states this "Our interpretative decision about the meaning of Paul's phrase,
therefore, is inevitably going to be governed by larger judgments about the shape and logic of
Paul's thought concerning faith, Christ, and salvation" ("ΠΙΣΤΙΣ and Pauline Chnstologyf 717)
Dunn is at times similarly programmatic, e g , "Paul's problem was to demonstrate that non-Jews
could be counted Abraham's children" ("Once More," 738)
 Campbell: Romans 1:17 267

definitive pronouncement on Paul's embedded πίσπς Χρίστου genitives! Such

9
discussions will only convince those who already believe.
If these sweeping observations are in fact correct, then further substantive
progress in this very important debate will be possible only as both its broader
and its more specific contexts are isolated and addressed. I will attempt only
the latter, more specific, task here, in the hope that it will be decisive for at
10
least the meaning of the various πίστις Χρίστου genitives in Romans: such
11
an objective is certainly sufficient for this short study. I will in fact argue
that Rom 1:17 is the programmatic—if usually unnoticed—text for this debate
in Romans. In particular, I will suggest that this text clearly deploys the critical
phrase εκ πίστεως as an intertextually motivated allusion to the faithful death
of Christ—a deployment that includes, perhaps surprisingly, Hab 2:4. Needless
to say, such a christological reading of Rom 1:17 has powerful implications for
Paul's repeated use of this phrase—and πίστις itself—in the famous arguments
that follow.

Ι. Πίστις Χρίστου Constructions in Romans

Discussion of the contentious genitives in Romans has previously con­
centrated largely on 3:22 and 26, with a nod in the direction of 3:3-5, 25,

9
Another symptom of this configuration is probably the constant attempt by both sides to
resolve the question on grammatical grounds, that will stand independently of any context. Although
the discussion gets fairly torrid at times, Hays, M. Hooker, and I concur that both grammatical
cases are invalid: see George Howard ("Notes and Observations on the 'Faith of Christ,'" HTR
60 [196η 459-65; "Rom. 3.21-31 and the Inclusion of the Gentiles," HTR 63 [1970] 223-33; "The
'Faith of Christ;" ExpTim 85 [1974] 212-25; "Faith of Christ," ABD 2. 758-59) on behalf of the
christological reading, arguing on grounds of the general incidence and construal of the con­
struction in the Greek of the period; and Arland J. Hultgren (The PISTIS CHRISTOU Formulation
in Paul," NovT 22 [1980] 248-63) and various comments by Dunn ("Once More," 731-35, 744)
on behalf of the traditional reading, arguing on grounds of the incidence of the article (see Hays,
Faith of Jesus Christ, 164; idem, "ΠΙΣΤΙΣ and Pauline Christology," 716, esp. n. 8; Morna Hooker,
"ΠΙΣΤΙΣ ΧΡΙΣΤΟΥ," NTS 35 [1989] 321-22, esp. n. 2 [p. 321] and n. 1 [p. 322]; Campbell, "Appendix
2: The Objective Genitive Reading of πίστις Τησου Χρίστου," in Rhetoric of Righteousness, 214-18;
see also Luke T. Johnson, "Again Pistis Christou" CBQ 49 [198η 431-47).
10
I have often been criticized when presenting this case orally for concentrating on one of
the two critical letters, but reject these objections because (1) Romans is the more promising
and decisive letter, as many acknowledge (e.g., Hays, "ΠΙΣΤΙΣ and Pauline Christology," 716, 717;
see also the focus of Haußleiter's original inquiry [see n. 2]): in Romans Paul is clearly being
more systematic, while in Galatians one has to take into account his previous teaching and that
of his opponents, while the argument of the letter itself is notoriously convoluted (for the previous
reasons?); and (2) Galatians has received the most attention lately in any case (see Hays, "ΠΙΣΤΙΣ
and Pauline Christology," nn. 2 and 3, pp. 714-15).
11
I hope to complete studies on the various other aspects of the debate at some point, namely,
its cultural backdrop and theological implications, while much has already been done. On Gala­
tians, Hays's work is exemplary (Faith of Jesus Christ). On cultural analogues, see Greer M. Taylor,
"The Function of ΠΙΣΤΙΣ ΧΡΙΣΤΟΥ in Galatians," JBL 85 (1966) 58-76.
268 Journal of Biblical Literature

30-31, and 4.16.12 Thus, at first glance, we appear to be occupied with basically
one chapter of Paul's discourse and with two phrases, with perhaps two to
three more being affected in context. The traditional importance of 3:21-26
for Pauline theology increases the significance of these phrases slightly, but
it is hard to escape the mathematical innocuousness of two to five instances.
This configuration of the data, however, is incorrect.
The disputed genitive phrase in 3:22 is constructed with the preposition
Sta, while the sections disputed concluding phrase in v. 26 follows an incidence
of εκ; a more unusual preposition in terms of the debate. Discussion has
generally concentrated on the διά genitives, of which there are about five
throughout Paul.13 Έ κ is used again only in Gal 2:16 and 3:22.1 have argued
elsewhere, however, that this unexplored parallelism between διά and έκ phrases
is crucial, recurring in numerous other contexts in Romans and Galatians
without the attached substantive Χριστός.14 For example, Rom 3:30 clearly
deploys the two phrases in parallel—and in some relation to Χριστός, although
it is unstated. When the rather extraordinary statistical profile of these occur­
rences is examined, it becomes clear that the πίστης Χρίστου dispute functions
within the broader context of a set of antithetical phrases in Paul that con­
trasts πίστις Χρίστου (whatever that means) with έργα νόμου (whatever that
means), often in an allusive and "sloganizing" fashion. Moreover, the dominant
phrase in the interlocked paradigms is clearly έκ πίστεως. Critical for the entire
issue, however, is the observation that the occurrence of this phrase correlates
perfectly with Paul's citation of Hab 2:4. Paul uses έκ πίστεως twenty-one times
in his extant letters, but only in those two that also cite the scriptural text
containing that exact phrase (in fact, a scriptural text trimmed by Paul to that
exact phrase). This cannot be mere coincidence — there is almost certainly
some causal connection between these two phenomena, and the authoritative
text's motivation of the other phrases seems the more likely. Consequently,
it is to Hab 2:4 that we should turn if we wish to understand further PauPs
flurry of έκ/διά πίστεως phrases, including his πίστεως Χρίστου phrases, in
Romans and Galatians.
In Romans, our current concern, Paul cites this text early in the letter.
It is of course his first scriptural citation, and it follows immediately on his
12
See Johnson, "Again Pistis Christou"; Dunn, "Once More," 740.
13
The numbers vary depending on one's position on the authenticity of Ephesians, and also
the textual variant in Gal 3:26 suggested by P 4 6 . The "undisputed" genitives, whether using έκ
or διά, are Rom 3:22, 26; Gal 2:16 (2x), 20; 3:22, 26 (if accepted, which it usually is); and Phil
3:9. Eph 3:12 is often introduced, at which point the invariably overlooked 4:13 should also be
considered, along with (perhaps) 3:17. Of course, the interpretation of many other instances of
πιστις in the contexts of these genitive constructions is also affected by one's decision, e.g., Rom
3:25, 27, 28, 30 (2x), 31; Galatians chap. 3 passim; Phil 3:9; and perhaps Eph 2:8 and 4:4.
14
This study was originally part of a broader discussion, but it proved too long to publish
in article form. Consequently, what is essentially the previous step in my case must basically
be presupposed here: see my "The Meaning of ΠΙΣΤΙΣ and ΝΟΜΟΣ in Paul: A Linguistic and
Structural Perspective," JBL 111 (1992) 91-103.
 Campbell: Romans 1:17 269

first use of the phrase έκ πίστεως, which occurs in 1:17. This verse therefore
gives us the luxury of two instances of our critical phrase, the second time
actually in a scriptural text, within a carefully crafted section replete with
contextual information that also functions programmatically for the rest of the
15
letter (or, at least, for chaps. 1-4). Thus, it seems that Rom 1:17 may well
hold the key to our interconnected issues. If it can be proved that έκ πίστεως
takes a certain meaning here, then it follows that the rest of these phrases
in Romans, including the disputed πίστις Χρίστου genitives (whether using έκ
or διά), will almost certainly conform to this—to argue otherwise would be
to ignore the deliberate, programmatic function of 1:17. Consequently, 1:17
seems to be "the Thermopylae" of the πίστις Χρίστου debate in Romans — the
strategic pass through which all else must travel, to be held (or taken) with
blood and tears.16 But before turning specifically to the exegesis of έκ πίστεως
in Rom 1:17, we must first touch briefly on an important preliminary issue.

II. The Meaning of δικαιοσύνη θεοΰ in Rom 1:17

Verse 17a begins with a much-disputed phrase in Paul, δικαιοσύνη θεοΰ.17
Fortunately, for our purposes it is not necessary to resolve fully the meaning
of all the instances of this word group, or even of this phrase, in Paul. We say
this because the implications of the immediate context of Rom 1:17 push the
meaning of the phrase at this point in Paul in certain directions strongly.18 The

15
I do not endorse the architechtonic understanding of Hab 2:4 as a device structuring the
theological argument in Romans (esp. chaps. 1-8), as suggested, e.g., by A. Feuillet ("La citation
d'Habacuc 2:4 et les huit premiers chapitres de 1-Epître aux Romains," NTS 6 [1959-60] 52-80)
and A. Nygren (Commentary on Romans [trans. C. C. Rasmussen; Philadelphia: Muhlenberg, 1949]
85-87), although I do recognize its thematic importance.
16
Hence I was extremely puzzled when Hays, in his 1991 debate with Dunn, simply con-
ceded to Dunn's alternative reading here: see his "ΠΙΣΤΙΣ and Pauline Christologyf 718, par.
1—this is to allow one's flank to be turned well and truly!
17
Historically, scholars have pondered the precise meaning of δικαιοσύνη, particularly when
it is related to God, and they have tried to determine whether the genitive construction is subjective,
objective, or some combination of the two (e.g., a genitive of origin). Surveys of the history of
the debate can be found in P. Achtemeier, IDB 4. 99; Μ. T. Brauch, "'God's Righteousness' in
recent German Discussion," in E. P. Sanders, Paul and Palestinian Judaism (Philadelphia: Fortress,
1977) 523-43; G. Klein, IDBSup, 754; and J. A. T. Ziesler, The Meaning of Righteousness in Paul
(Cambridge: Cambridge University Press, 1972) 217-30.
18
In earlier commentaries, an objective genitive reading is almost universally favored, as W.
Sanday and A. C. Headlam note (The Epistle to the Romans [2d ed.; Edinburgh: T. & T. Clark,
1896] 24). No doubt this tendency is in large measure attributable to Luther's "great awakening"
by this verse (see his Lectures on Romans [trans. W. Pauck; Philadelphia: Westminster, 1961] 18—the
lectures were written by Luther in 1515). This reading finds a few modern supporters among
the commentators, e.g., Black, C. E. B. Cranfield, O'Neill, and H. Schlier. A broad scholarly
consensus may be emerging here, however, following a "Hebraic" genitive and a salvific reading
of δικαιοσύνη, as Dunn suggests (Romans 1-8 [Dallas: Word, 1988] 40-42), and E. Käsemann
laments (Commentary on Romans [trans. G. W. Bromiley; Grand Rapids: Eerdmans, 1980] 24, 27).
270 Journal of Biblical Literature

clear reference in the previous verse, within a parallel construction, to the

gospel of salvation, suggests that δικαιοσύνη must be carrying at least a conno­
tation of salvation for Paul at this point. 19 The link with the gospel also suggests
that this salvation is at least in part eschatological, since it is clear that Paul
does not expect a further salvation from God beyond Christ, and God's salva­
tion is by definition an eschatological event.20 Moreover, the loading of the
21
context with subjective genitives suggests that the genitive relationship must
also be subjective here, although, as Ernst Kasemann has noted in a famous
study, the concept is not for this reason static: δικαιοσύνη is best understood
in a "Hebraic" (or, better, "non-Western") sense that rather transcends subject-
object distinctions.22 It seems more like a dynamic power flowing from God
to the world just as, once again, the context suggests with its use of δύναμις
and, later on, οργή. In sum, δικαιοσύνη in v. 17a is best understood as a reference
to an eschatological saving power, both of and from God. Further specifica­
tion of the phrase beyond this is not necessary for our present purposes.

III. Two Apocalyptic Models in Rom 1:17a

To focus now directly on our question: as is well known, the section is
not without its difficulties, but it does present us with a great deal of impor­
tant contextual information:
[1:16a] Ου γαρ έπαισχύνομαι το εύαγγέλιον,
δύναμις γαρ θεοΰ έστιν
[16b] εις σωτηρίαν παντί τω πιστεύοντι,
Ίουδαίω τε πρώτον και 'Έλληνι-
pL7a] δικαιοσύνη γαρ θεοΰ εν αύτω αποκαλύπτεται
έχ πίστεως εις πίστιν,
[17b] καθώς γέγραπται,
ó δε δίκαιος έκ πίστεως ζήσεται.
It is here that traditional interpretation has tended to overlook the obvious.
Despairing of interpreting the series έκ πίστεως εις πίστιν in v. 17a, it has gen­
erally given this clause a paraphrastic rendering and then concentrated on

19
See also Rom 3 21-26 and 9 30-1010, clearly salvific discussions, in which δικαιοσύνη features
20
So particularly Kasemann "The operation of God's power corresponds to its eschatological
nature" (Commentary, 22, see also 24-26)
21
δύναμις . . . θεοδ m 116, and οργή θεού in ν 18 Note, the first occurrence of δικαιοσύνη
after 117, m 3 5, is also within a subjective construction, see Otto Glombitza, "Von der Scham
des Glaubigen Erwägungen zu Rom 114-17," NovT 4 (1960) 79
22
Kasemann, "The Righteousness of God in Paul," m New Testament Questions of Today (trans
W J Montague, London SCM, 1969 [1965]) 168-93, see also his Commentary, 23-30 As Dunn
concludes " is it an attitude of God or something he does? Seen as God's meeting of the claims
of his covenant relationship, the answer is not a strict either-or, but both-and, with the emphasis
on the latter" (Romans 1-8, 41)
 Campbell: Romans 1:17 271

the complete scriptural citation that appears in v. 17b. But this quotation is
capable of various readings, and one wonders if the most important contextual
and theological clues to its use have already been passed by.23
At this point, however, as we try to press further into the meaning of 1:17a,
we encounter an important, if usually unnoticed, interpretative decision. The
meaning of v. 17a alters fundamentally depending on the meaning accorded
to the verb αποκαλύπτω and, in close relation to this first interpretative deci­
sion, the function accorded to the prepositional phrase εν αύτω. Commen­
tators tend to overlook or, if they do note the alternative, to misunderstand
and obscure the ambiguities that lie here, no doubt impelled by the stature
of the traditional reading, supported by well-worn renditions of the verse in
the vernacular.24 It is critical, however, to free the interpretative process from
this weight of tradition and to address the semantic possibilities suggested
by the Greek.
Fundamentally, there are two possible readings of Rom 1:17a, that we may
characterize broadly as the traditional anthropocentric reading, and the cosmic
eschatological reading. We will address these in turn.

The Traditional Anthropocentric Reading

For this interpretation, the act of revelation or disclosure denoted by
αποκαλύπτεται is assumed to be one directed to the generic individual. Paul
is held to be saying that the saving righteousness of God, which is already
present within the gospel, is being revealed to individuals as they grasp it by
faith (and such a reading comes very naturally to a good Protestant!). We might
translate this as, "The saving power of God within the gospel is being revealed

23
The series in v. 17a is usually rendered by something like "by faith from first to last." Most
such alternatives, however, paraphrase the disputed expressions, often using the prepositions in
a sense that loyalty to their normal meaning will not permit. For example, έκ πίστεως είς πίστιν
means "from the faith of the preachers to the faith of hearers" or "from the faith of the OT to
the faith of the NTT Such readings, however, tend to loosen the prepositional series from its gram­
matical moorings and to interpret it within a theological framework: see C. Ε. B. Cranfield's very
full discussion of the phrase (A Critical and Exegetical Commentary on the Epistle to the Romans
[2 vols.; Edinburgh: T. & T. Clark, 1975, 1979] 1. 99).
Specialized studies on the question are surprisingly rare: I know only of Anton Fridrichsen,
"Aus Glauben zu Glauben, Rom 1:17," ConNT 12 (1948) 54; and Glombitza, "Von der Scham des
Gläubigen," 74-80, esp. 78-80—a rather Bultmannian reading!—but neither of these studies notes
the implications of ambiguity that will concern us subsequently.
J. Hugh Michael resolves the problem in a novel fashion by excising the first instance of
έκ πίστεως as a vertical dittographical error ("A Phenomenon in the Text of Romans," JTS 39 [1938]
151), but such a solution both misses Paul's point and removes all hope of finding it.
24
E.g., Kasemann: faith is definitely a "personal responsibility" but at the same time "God's
self-manifestation is decisive for all history" (Commentary, 23)! How are these two things related,
since Paul speaks of them in one breath? Characteristically, even if the commentators grasp the
eschatological dimension of Paul's statement, they miss its important implications. Note, however,
those who read δικαιοσύνη in v. 17a in terms of God's faithfulness are not vulnerable to this criticism.
272 Journal of Biblical Literature

by means of faith." Of crucial importance for our purposes is the grammatical

observation that within this reading έν αύτω functions adjectivally—and hence
not very significantly—as it modifies δικαιοσύνη θεοΰ: "the saving righteousness
of God—which is within the gospel—is being revealed . . . ." It is also impor­
tant to note that here the righteousness of God is already assumed to be within
the gospel. That identification, according to this reading, is not something
that actually requires disclosure: what requires disclosure is the gospel's oifer
of salvation to individuals, where faith is the critical mediator. As a result of
this, an act of apprehension is central to the reading (albeit an important
one!), and it has, as a result, a distinctly anthropocentric (and even Cartesian)
cast.

The Cosmic Eschatological Reading

This reading, however, does not take the presence of God's saving righ­
teousness within the gospel for granted. It suggests that the revelation of God s
eschatological saving righteousness within and by the gospel is actually Paul's
main point. Here God's eschatological saving righteousness functions dynam­
ically, like the OT "Word of God," breaking into a chaotic or rebellious order
from above but, crucially for Paul, here definitively within the Christ-event.
The primary relationship presupposed by this reading is therefore that between
God and the gospel in the context of the cosmos, not that between the gospel
and the individual (or nation). Moreover, the revelation being spoken of is the
disclosure of the previously unseen righteousness of God to the world within
the gospel (unseen presumably because it is in the eschatological future), not
the disclosure of salvation from the gospel to the individual (although these
points are theologically related in that the first disclosure is the presuppo­
sition of the second): hence our characterization of the reading as cosmic and
eschatological.
Within this reading the functions of both the verb αποκαλύπτω and the
short prepositional phrase έν αύτω are rather different from that of the previous
reading. The prepositional phrase functions here as the agent or instrument
of the verb αποκαλύπτεται, and hence denotes a different disclosure altogether.
Consequently, within this reading the prepositional phrase έν αύτω is extremely
significant and quite central to Paul's meaning: it bears the brunt of the verb —
"God's eschatological salvation is being revealed in it, that is, in the gospel";
and Paul's basic point is, as we have said, that the eschaton is being realized
in the content of the gospel.
Why is a clear distinction between these two readings so important for
our interpretation of έκ πίστεως, the phrase that immediately follows these
words in v. 17a? Because a traditional reading of this phrase in terms of
anthropocentric faith in v. 17a (and hence also in v. 17b) is possible only if
the first reading is the correct one, that is, if the traditional anthropocentric
reading of v. 17a is adopted. If the second reading is adopted then an inter-
 Campbell: Romans 1:17 273

pretation of έκ πίστεως in terms of anthropocentric faith is utterly impossible.

Why does a cosmic eschatological reading of v. 17a exclude an interpreta­
tion of έκ πίστεως in terms of anthropocentric faith? If Paul is saying in Rom
1:17a that "the eschatological, saving righteousness of God is being revealed
within the gospel," then clearly this event is accomplished independently of
the individual's faith. To make the eschatological disclosure of God's saving
power conditional upon the believer's faith would be to press the role of anthro­
pocentric faith rather too far—even as some would define it within Paul's
theology, and particularly as it seems to be defined by Paul within Romans.
In short, it would be to make the coming of the eschaton dependent on
individual faith, and this is theologically (and practically [!])ludicrous. For Paul
this revelation is clearly grounded in God's grace and sovereignty— and love
(so especially Rom 5:8-9; 8:28-39).
But the phrase έκ πίστεως must be read in some relation to this previous
cluster of ideas—after all, it is in the same sentence! 25 So at this point it must
be concluded that if a traditional anthropocentric construal of Rom 1:17 is
not in fact suitable, then έκ πίστεως must refer to something else altogether,
like the faithfulness of God or of Christ. Rather significant interpretative con­
sequences for πίστις and for Romans flow from this decision. Thus, clearly
a great deal turns on the comparative suitability of these two possible readings
of Rom 1:17a.
In evaluating their comparative merits, as we have seen, it is absolutely
critical to assess the function of the prepositional phrase and the reference
of the verb. There are suggestions both from the immediate context and from
Paul's wider linguistic usage about how we should in fact read these two
semantic units, and these must now be carefully assessed.

IV. Clues from the Immediate Context

Paul has consistently emphasized the gospel throughout the letter open­
ing of Romans, to the extent that he breaks into the standardized opening
formulas of the Greco-Roman letter (in his christianized version) with a care­
fully crafted statement concerning the gospel's content. Whatever this text's
origin,26 it clearly signals Paul's focus on the substance of the gospel: the gospel
of God that Paul preaches "concerns his Son. . . ." Paul feels constrained to
make this clear before he has even designated the recipients of the letter or
greeted them! This concern is reiterated in v. 9 within a parenthetical but
suggestive comment: "God is my witness—whom I serve in my spirit in the

25
Nygren (On Romans, 80) and Käsemann, for example, both really ignore the sentence's actual
grammar: "logically, then, it [the phrase] is related only loosely to the preceding statement" (Com-
mentary, 31). Koine prepositions were extremely fluid but not grammatically disembodied.
26
Many argue that this is an early church creed or confession: for discussion and bibliography,
see Dunn, Romans 1-8, 5-6, 22-24.
274 Journal of Biblical Literature

gospel of his Son. . . ."27 The suggestion is that Paul's preoccupation with the
gospel and its christological content has not subsided during the proem, where
he is primarily seeking to establish a good relationship with the Roman Chris­
tians who will be listening to his letter.
But perhaps most importantly, the theme of the substance of the gospel
is placed centrally in the immediate context of v. 17a by v. 16. Both parts of
v. 16 contribute to this emphasis. Verse 16a states: "For I am not ashamed of
the gospel," thereby establishing Paul's focus on the gospel clearly. Both v. 16b
and v. 17a are then grammatically dependent and symmetrically structured
developments of this statement. In v. 16b Paul states δύναμις γαρ θεοΰ έστιν . . . ,
and v.l7a follows this construction closely: δικαιοσύνη γαρ θεού έν αύτω αποκα­
λύπτεται. In both of these verses genitive constructions combine θεός with
some quality, link this to v. 16a with the conjunction γαρ, and complete their
statement with purposive εις constructions using πιστις. Moreover, v. 16b also
(as we would expect from 16a) clearly states what the gospel is and then directs
this content teleologically toward the salvation of the believer. The import of
the verb έστιν could hardly be clearer: thus v. 16b continues to speak of what
the gospel "is." We would therefore also expect the stylistically and grammati­
cally parallel construction in v. 17a to follow this emphasis on the substance
of the gospel. This in turn suggests stressing the reading that most success­
fully achieves this, that is, the cosmic eschatological reading, which speaks
of the righteousness of God being in the gospel. In sum, the context of 1:17
quite strongly suggests reading the prepositional phrase έν αύτω in relation
to the verb, not the noun phrase: "the righteousness of God is being revealed
in it (that is, in the gospel)'.'
Conversely, a traditional, anthropocentric reading shifts the focus of Paul's
discussion at this point from the gospel itself and its content to how the gospel
is appropriated. These are not the same thing, particularly as Paul has just
defined them in the letter opening and proem—but what justification does
the context supply for such a move away from, and then immediately back
to (cf. v. 18), a cosmic horizon? It seems undeniably more correct to continue
to emphasize the cosmic and eschatological dimension throughout this text
(as long as it is grammatically possible to do so).
In my view, these accumulating contextual indicators suggest the following
critical interpretive decision: έν αύτω should be interpreted in v. 17a in the
way that emphasizes the content of the gospel most strongly, and this sug­
gests an instrumental function in relation to the verb within a cosmic eschato­
logical reading. 28 But we should also consider the verb αποκαλύπτω in more

27
Significantly, faithfulness overlaps semantically with the Hebraic understanding of sonship
(see E. Schweizer and E. Lohse, "υιός," TDNT 8. 343, 349-53, 357-62), and Paul-unusually -
uses "son" seven times in Romans with reference to Jesus.
28
The only semantic alternative open to the anthropocentric reading at this point is to claim
that some notion of appropriation is introduced into the discussion after αποκαλύπτεται as an
 Campbell: Romans 1:17 275

detail before making a final decision on the reading of the phrase. If Paul uses
this word elsewhere in a fairly consistent manner (and there is no guarantee
that he will), then it might be possible to extrapolate a probable contextual
connotation from it.

V. Indications from Paul's Customary Use of the Verb

The verb αποκαλύπτω appears nine times in the undisputed Pauline letters
(Rom 1:1, 18; 8:19; 1 Cor 2:10; 3:13; 14:30; Gal 1:16; 3:23; Phil 3:15) and also
in Eph 3:5 and 2 Thess 2:3, 6, and 8, with the result that exactly one half of
the verb's twenty-six NT occurrences are Pauline. Paul's usage is not, as we
would expect, completely uniform,29 but a basic trajectory is apparent. The
word usually denotes dramatic, revelational events, such as Paul's own
encounter with the gospel, that are always unconditional and are almost
invariably eschatological.30 Here Gal 1:15-16 seems particularly significant.
Although these verses are from a different letter (and one possibly written
rather earlier than Romans31), the theological content and theme are extremely
close to Rom L16-17:32 οτε δε εύδόκησεν . . . άποκαλύψαι τον υίον αύτοΰ έν έμοί
ινα εύαγγελίζωμαι αυτόν έν τοις εθνεσιν. . . , 3 3 The critical point here is that
this revelation of the gospel is unconditioned, as Paul's preceding argument
makes quite clear. The event is grounded in God's choice to set apart Paul
from his mother's womb, and hence in God's grace (cf. v. 15). Also significant
is Paul's failure to mention "faith" at all. If Paul uses the verb αποκαλύπτω here
with reference to the breaking of the gospel into his own life and, further­
more, does not mention his own πιστις, how much more should we expect
the eschatological dawn of the gospel in the world, as described by Rom 1:16-17,
to be unconstrained and unconditioned!? 34 Two of the instances in 1 Corin-

unstated presupposition. Such an introduction is certainly theologically possible for Paul, given
his later discussion, but the semantic shift is a noticeable one that should probably have been
indicated linguistically by the use of a participle, for example, νοούμενα, which functions analogously
in 1:20.
29
The verb sometimes describes charismatic and "occasional" revelations (e.g., 1 Cor 14:30) —
although the seriousness with which Paul treated these revelations should not be underestimated:
in Phil 3:15 such an event is supposed to confirm the heart of the gospel.
30
Excepting Rom 1:18, but see our discussion below.
31
If one follows the early, South Galatian provenance the letter is written ca. 49 CE, with
Romans ca. 57 CE (for a definitive statement of this position, see Longenecker, Galatians, lxi-c).
Others place the letters much closer together in Paul's letter sequence, and often in the same year.
32
Perhaps the importance of this event for Paul minimizes the problem of distortion creeping
into his account over time, in any case.
33
This statement is corroborated by 3:23: Προ του δέ έλθεΐν την πίστιν ύπο νόμον έφρουρούμεθα
συγκλειόμενοι εις τήν μέλλουσαν πίστιν άποκαλυφθήναι. The frequent use here of την πίστιν makes
an appeal to this text on this question impossible. But it is interesting to note that where υιός
functions in chap. 1, τήν πίστιν functions in 3:23 (see n. 27): see also 1 Cor 15:8 and context.
34
Eph 3:5, in the context of w. 2-12, creates a picture identical to this one, except that the
implications are perhaps even stronger.
276 Journal of Biblical Literature

thians (2:10 and 3:13) also especially confirm this pattern found in Galatians,
both being strongly eschatological and unconditioned. 35
The two occurrences of the verb in Romans outside 1:17 reinforce this
impression. To take them in reverse order, Rom 8:19, although it does not refer
to cosmic and salvific events in the past, does refer to those eschatological
events still to come. Thus, it reinforces the usual eschatological connotation
of the verb.36 In Rom 1:18 Paul is clearly playing on his previous use of the
verb, so he employs it in this subsequent verse even though the scenario is
not fundamentally eschatological—but it is certainly cosmic in scope.37
Thus it would seem that there is, broadly speaking, a consistent picture
in the Hauptbriefe. Paul tends to use the verb αποκαλύπτω (and related words)
to describe cosmic eschatological disclosures and, in particular, the primary
eschatological disclosure that is the gospel. These revelations are usually
dramatically powerful and unconditioned as, for example, in Gal 1:15-16, where
one bursts upon the zealous but unsuspecting Paul. Such disclosures seem
to be driven solely by the purposes and power of God, and this seems quite
appropriate for Rom 1:17 as well. Käsemann summarizes the implications of
such a picture in relation to this verse with customary force:

The gospel is . .. the epiphany of God's eschatological power pure and

simple.. .. God's self-manifestation is decisive for all history . .. [and t]he
apostle's theology involves a definite salvation-historical perspective. The
interpreter who radically denies this is forced to reduce the Lord of history
to the creator of the particular moment and hence to do violence to the Pauline
doctrine of God [emphasis added].38

Given the presence of this scenario behind Rom 1:17, it remains only to
draw the proper consequences from it for our reading of έκ πίστεως—con­
sequences that commentators seem so often to have overlooked. It seems clear
that έκ πίστεως cannot be translated here in terms of anthropocentric faith,

35
Our difficulties finding an anthropologically conditioned use of αποκαλύπτω in Paul within
this eschatological sphere are not relieved when the rest of the NT— and indeed the OT and
the literature of Second Temple Judaism —is scanned. The word group and the concept seem
to be too dominated by the sovereignty of God. A. Oepke's dated but still useful analysis discerns
an "eschatological" use in the OT (e.g„ Isa 56:1) and states concerning Paul: "Revelation . . . is
a divine act, the unveiling of what is hidden [and]... its true locus is eschatology" (TDNT3. 563-93,
esp. 577, 583-84). These observations certainly corroborate our conclusions.
36
Άποκάλυψις has a similar meaning in Romans: cf. 8:19 (and 16:25). Here a relationship with
the φανερ- word group is also apparent (cf. 1:17 and 3:21; cf. also the probably non-Pauline 16:25
and 26). Φανερόω does not exhibit as strong an eschatological connotation as αποκαλύπτω, but
an added semantic emphasis seems to be that the revelation is clear. Note that it is clear not
because of human πίστις but because God has made it clear (cf. esp. 1:19).
37
Although humans deserve the wrath of God, they do not in any sense condition or control
that disclosure, as the solemn threefold reiteration of ó θεός παρέδωκεν suggests (w. 24, 26, 28).
38
Of course, he is a well-known advocate of eschatology (properly defined), but our survey
of the evidence suggests that his comments on this text are accurate (Commentary, 23; see also
Glombitza, "Von der Scham des Gläubigen," 78-79).
 Campbell: Romans 1:17 277

since the theological consequences verge on the absurd.39 But the phrase still
has two appropriate and hence possible referents, namely, the faithfulness of
God (i.e., God "the Father"), and the faithfulness of Christ (it seems to me
rather less likely that it refers to some combination of meanings). These alter­
natives must be carefully weighed before Rom 1:17 will yield up its mysteries
fully.

VI. A Reference to God, or to Christ?

An important exegetical alternative to the traditional anthropocentric
reading of Rom 1:17 (but one that can ultimately save the theme elsewhere!)
is an interpretation of έκ πίστεως in v. 17a in terms of the faithfulness of God.
This is a much more contextually sensitive reading, and it integrates well with
the cosmic and eschatological dimension that we have emphasized is present
within Paul's discussion at this point. Hence, drawing this distinction here
unfortunately cannot help us by arbitrating between a theocentric and a chris­
tological interpretation: both are equally eschatological and cosmic in their
force. Dunn is probably the best-known and most powerful recent exponent
of the theocentric position, although it has had important earlier advocates,
like Karl Barth.40
Dunn argues quite correctly that a theocentric reference for έκ πίστεως
in v. 17a makes good sense in context: "the savingrighteousnessof God is being
revealed in the gospel, (that salvation springing) out of (his) faithfulness. . . Γ
This is sound Jewish theology, and it receives additional support from various
other contextual considerations. In Rom 3:3 Paul clearly speaks of the faith­
fulness of God, and this, as Dunn points out, is the first occurrence of πίστις
in Romans after 1:17. Moreover, the reading correlates v. 17a with v. 18 precisely
in that both now refer to cosmic acts springing from God and proceeding to
the earth. It also draws the phrase into alignment with an eschatological reading
of the immediately preceding δικαιοσύνη θεοΰ. Thus, the eschatological and
cosmic suitability of the interpretation is plainly apparent. One can also cite
the identical prepositional progression in 2 Cor 2:16 in support of a more
generalized, rhetorical interpretation of the critical phrases.41

39
We may note here, however, that εις πίστιν undoubtedly refers to the faith of the Christian.
In v. 16b the purposive είς of v. 5 reappears: είς σωτηρίαν παντί τω πιστεύοντι. These earlier
references to the faith of the believer are clearly being recapitulated by είς πίστιν in v. 17a; thus
it cannot be doubted that Paul considers the πίστις of the Christian (note: in some relation to
υπακοή) to be important.
40
See Dunn, Romans 1-8,44,48; Dunn cites Karl Barth, G. Hebert, T. W. Manson, and Lloyd
Gaston in support; see also Wilber Β. Wallis, "The Translation of Romans 1:17—A Basic Motif
in Paulinism,"/ErS 16 (1973) 17-23; and Torrance, "One Aspect," 113. Note that έν αύτω is best
read with the verb on this reading.
41
Dunn himself notes point (1) here, adds six more (!), but overlooks point (2) (Romans 1-8,
44). He argues that such a reading is good rhetorical style; έκ as source follows a verb of revelation
278 Journal of Biblical Literature

These are strong supports for the reading, and it is in many ways a pro­
foundly attractive one. Certainly it is decisively superior to the traditional
anthropocentric reading. However, it also suffers from a series of disadvantages
that, in my opinion, ultimately completely undermine its cogency. The Achilles'
heel of the reading is the systematic equation that exists in Romans (and Gala­
tians) between the phrase έκ πίστεως and the text of Hab 2:4. Unfortunately,
this is not a relationship that we have had time to establish fully within the
bounds of this short study. As has already been briefly stated in section one,
however, I have argued elsewhere that Hab 2:4 and the phrase έκ πίστεως are
closely related: indeed, this text seems to be the template for Paul's frequent
deployment of the isolated phrase, so that the two must really, if it is at all
possible, be interpreted in parallel. They are intimately related linguistic units
in Paul. This close relationship between the phrase and the scriptural text,
however, forces theocentric interpreters like Dunn onto the horns of an un­
enviable dilemma.
Theocentric exponents, initially, can only really deny that there are any
firm connections here and argue that έκ πίστεως does not mean the same thing
wherever it is deployed in Paul and that the instance in v. 17a does not have
any close relation to the intertext from the prophets quoted in v. 17b. This
allows the phrase έκ πίστεως in v. 17a to refer to the faithfulness of God, while
the text in v. 17b can still function as an authoritative scriptural legitimation
of the individual's saving faith in the gospel — and the phrase έκ πίστεως con­
sequently functions rather malleably, taking two quite different senses in the
same verse.
No one would deny that language —and Paul's use of it!—is flexible, but
this reading takes linguistic plasticity too far. It runs squarely into two obser­
vations in particular that suggest that a relationship does exist between these
two semantic units and that they should be interpreted in parallel:
(1) If the phrase in v. 17a and the text in v. 17b are interpreted diversely,
the verse is actually fractured into separate and unrelated statements. Not only
is this puzzling in its own right, but it cuts across the apparent function of
Hab 2:4, which is to support scripturally what has just been said. Why would

"naturally"; έκ . . . etc is "somewhat odd" if both refer to human faith (an argument from
redundancy?); Hab 2:4 is probably intended by Paul to be understood ambiguously; faithfulness
is a theme in Romans; and the righteousness of God is synonymous with his covenant faithfulness.
Regarding these arguments, the first seems indecisive — a christological reading would be equally
good rhetorical style; the second is also indecisive, because it fails to establish that έκ always
takes and hence must take such a meaning; the third applies only to the traditional reading and
hence is irrelevant to our christological interpretation; the fourth contention seems quite specula­
tive—Where does Paul tell us that "scripture is ambiguous"?! And the last two considerations,
although true, are also indecisive: as merely possible readings, they similarly fail to establish a
particular meaning here—while the evident redundancy generated by Dunn's last observation
clearly harms his own case! In sum, although initially impressive, Dunn's case is not particularly
powerful.
 Campbell: Romans 1:17 279

Paul cite a supporting text that does not in fact support anything but makes
a different statement—and that with an identical phrase? Such a procedure
is simply incoherent.
(2) The perfectly correlated distribution of the phrase and the text
throughout Paul's letters as a whole, to which we have alluded, also suggests
such a close relationship. Fundamentally, it seems better to accept that this
remarkable distribution says something about a relationship between the two
phenomena rather than to attribute it to mere chance. Paul only uses έκ πίστεως
when he has cited Hab 2:4 at some point (and here in Romans, notably after
his first use of the phrase!), and yet when he has done so it is a—and perhaps
the—leading motif in his arguments. Clearly, to my mind, a close relationship
exists here.
The theocentric interpreter's only alternative at this point is to bow to
the (really undeniable) presence of a parallelism, but to continue to read έκ
πίστεως in v. 17a as a reference to the faithfulness of God and to follow this
through consistently for Hab 2:4 in v. 17b, reading this text as a reference to
the faithfulness of God as well.
To my knowledge, no one has ever actually argued this reading of Hab
2:4, for reasons that should nevertheless be made explicit:
(1) Although understandable, in Romans such a deployment seems un­
necessary and even puzzling: Why would Paul so emphasize the faithfulness
of God? As his diatribe in 3:1-9 demonstrates, although this idea is important
to him, it is clearly shared by his audience. That is, it is not a point that is
in dispute; it is a shared presupposition. Hence, to emphasize it seems clumsily
redundant, while to make it the heart of a thesis paragraph seems positively
otiose.42
(2) Most commentators would want to read Hab 2:4 in Galatians simi­
larly to its apparent meaning in Romans — although, once again, it is formally
possible to argue that Paul has radically changed his use of this text between
these two letters (something of a bonus for his opponents, should they get
wind of it!). But a reading of Hab 2:4 —along with its associated instances of
έκ πίστεως—in Gal 3:11 in theocentric terms leads to extraordinary herme-
neutical acrobatics. Such a reading really cannot integrate the elements in
this compact discussion—Abraham and Gen 15:6, Christ's purchase of sinners,
and the torah and its observance—in a way that makes much sense at all. And,
once again, to my knowledge no commentator actually attempts this.
(3) One of Hab 2:4's enduring puzzles has been its textual form: Paul is
following no known textual variant when he speaks simply of έκ πίστεως. But
if Paul really wishes to speak of God's faithfulness in using Hab 2:4, why has
he ignored those versions of the verse, in well-known textual traditions that

42
Käsemann notes the redundancy generated by the reading (Commentary, 23); see also, most
clearly, Rom 3:3, 5.
280 Journal of Biblical Literature

he himself is familiar with, that make this explicit? The tradition represented
by MSS S and Β are unequivocal: the righteous one lives έκ πίστεως μου.
These considerations alone would probably be enough to weaken seriously
a reading of Hab 2:4 in Rom 1:17 in terms of the faithfulness of God. A fourth
consideration is utterly fatal to it, but it presupposes once again the equation
between the phrase έκ πίστεως and the text Hab 2:4 in Paul's discussions. If
this equation holds, then the interpreter of Rom 1:17 faces exegetical conse­
quences far more extensive than those merely of v. 17 itself. This verse—as
far as possible—should now be interpreted in a fashion that is sustainable for
the rest of Paul's uses of this phrase in any ensuing arguments.
Ironically, the traditional, anthropocentric interpreter has far less difficulty
with this requirement than the theocentric interpreter (and hence much of
the tenacity of the traditional exegesis of Rom 1:17: it is imported back from
this wider reading). An advocate of anthropocentric concerns in Paul can read
most of his arguments in Romans that use πίστις in a coherent fashion. The
theocentric interpreter in many places finds this difficult, however. Moreover,
the reading is impossible in 3:26 and 4:16, that is, wherever έκ πίστεως or its
equivalent is combined in a genitive construction with someone other than
God—a not infrequent occurrence. Usually the phrase is combined with
Χριστός in these constructions, but in 4:12 and 16 it is joined to Abraham.
Clearly the faithfulness of God cannot be present in these texts! However, a
christological reading can suggest that Abraham is a typological anticipation
of Jesus' messianic faithfulness, while the traditionalist generally construes
these genitives objectively, as is well known, and argues that Abraham is a
type of the Christian believer (although his genitives are subjective!).43 Thus,
a criterion of consistency in the interpretation of έκ πίστεως throughout the
argument of Romans proves utterly embarrassing to the theocentric reading.44
The second horn of our dilemma has now been duly sharpened, and it
seems that the theocentric interpreter has nowhere left to sit: one must either
be exposed as contextually insensitive or thematically redundant and incon­
sistent—or abandon the reading altogther. Really, this last course is the only
exegetically satisfactory option.
The final possible reading of the phrase έκ πίστεως in v. 17a is the seldom-
considered alternative of the faithfulness of Christ 45 —indeed, at this point it

43
Furthermore, it is an extremely difficult reading in Rom 9:30-10:8 (and in Galatians, where
in 2:16 it is impossible, as in 2:20; 3:22, 26).
44
This is not to disparage the faithfulness of God, which is clearly an important theme for
Paul (cf. 3:3), but it is to suggest that Paul does not express this idea with the phrase έκ πίστεως—and
probably not with Hab 2:4 either.
45
Faithfulness in the sense of his obedience and perseverance to the will of God, culminating
in Calvary (cf. Rom 1:5; 3:25; 5:1, 10; 5:18; and esp. Phil 2:5-11). No doubt Jesus believed in the
existence of God, but as the author of James remarks, "even the demons do that!" (Jas 2:19). Some
opponents of the christological reading of πίστις rather vulgarize it by suggesting that it speaks
of Christ's "bélier—and also demand verbal equivalents elsewhere in Paul for this idea (which,
 Campbell: Romans 1:17 281

is the only semantic alternative left standing. Within the broader cosmic
eschatological perspective it suggests the following translation: 'The eschato­
logical saving righteousness of God is being revealed in the gospel by means
of faithfulness (namely, the faithfulness of Christ), with the goal of faith/fulness
(in the Christian)." On this reading Paul is stating that the eschatological salva­
tion of God is actually being revealed now in the gospel, the substantive point
of that revelation being the life and death of Christ. In particular, it was the
faithful obedience of Christ through Calvary that revealed God's salvation and
also created the possibility of individual salvation through belief and per­
severance until the eschaton (an identification that becomes clearer later on
in Romans: see 3:25; 5:1, 9; etc.). Thus, for Paul, Christ's obedience and
obedient death are both the content of the gospel and the focus of the eschaton.
To our knowledge, this reading generates no contextual problems whatso­
ever. It smoothly continues the opening emphasis of the letter and the focus
of w. 16-17 on the substance of the gospel, and integrates with the sense of
cosmic disclosure suggested by v. 18. It integrates with the cosmic, escha­
tological connotation Paul usually evokes when employing the verb αποκα­
λύπτω. It also integrates neatly with a particular construal of Hab
2:4— necessarily, a christological or messianic one—but this question of Paul's
messianic deployment of Hab 2:4 in Rom 1:17b should be considered in a little
more detail.

VII. Some Observations concerning 1.17b

As we have already intimated, έκ πίστεως in Hab 2:4, when it is quoted
in Rom 1:17b, should take the same sense as the immediately preceding
instance of the phrase in v. 17a—to argue otherwise leads to unacceptable
interpretative convolutions. Our suggestion concerning 17a therefore neces­
sarily involves reading Hab 2:4 as a messianic proof-text.46 Certainly, if it cannot
be so read, this would falsify our suggested reading of v. 17a. Various scholars,
however, have already proposed this, although their suggestion has usually fallen
on deaf ears.47 Nevertheless, they have proved that such a reading is perfectly

not surprisingly, do not exist: see Dunn, "Once More," esp. 734-35, 736-37). The preceding distinc­
tion should avoid much of this confusion.
46
Most of the previous discussion of Hab 2:4 has heatedly debated whether έκ πίστεως modifies
the noun or the verb: see the literature listed by Dunn (Romans 1-8, 36-37). On our christological
reading, έκ πίστεως must be an agent or instrument of the verb.
47
So C. H. Dodd, According to the Scriptures (London: Nisbet, 1952 [Fontana, 1965]) 51; and
Anthony T. Hanson, Studies in Paul's Technique and Theohgy (Grand Rapids: Eerdmans, 1974)
39-45; endorsed by Hays, The Faith of Jesus Christ, 150-57; see also T. W. Manson, "The Argu­
ment from Prophecy,'7TS 4 (1945) 129-36; J. A. Sanders, "Habakkuk in Qumran, Paul, and the
Old Testament," JR 39 (1959) 233; B. Lindars, Old Testament Apolo&tic: The Doctrinal Significance
of the Old Testament Quotations (London: SCM, 1961) 230-32; and the only suggestion from a
recent commentator, Brendan Byrne, Reckoning with Romans: A Contemporary Reading ofPauVs
Gospel (Wilmington, DE: Michael Glazier, 1986) 48. Hays's more recent case is again definitive
282 Journal of Biblical Literature

acceptable; it simply hasn t been possible to show previously that it is necessary,

and so the traditional reading has continued to be preferred. 48 With our
exposure of the traditional reading's previously unnoticed difficulties in v. 17a,
however, the quoted text no longer has to bear this burden of demonstration:
it need only be shown how suitable Paul's citation is as a messianic witness,
without actually making any claims as to necessity. This reading can then re­
inforce, as against disqualify, our suggestion concerning v. 17a. We will only
summarize the evidence for this reading here, since the case has already been
well made elsewhere.
Our first cluster of observations focuses on the adjective δίκαιος, which
is present in Hab 2:4 in an arthrous, substantive form. A reference to Christ
as ó δίκαιος integrates smoothly with Paul's elevated use of this word else­
where — and perhaps integrates more smoothly than a reference to humanity,
since Paul so sternly and repeatedly characterizes humanity in the early
chapters of Romans as άδικος and αδικία.49 ó δίκαιος may also be functioning
in Hab 2:4 as a more stereotyped, titular reference to Jesus as the Messiah.
The phrase ό δίκαιος is found fairly frequently in the NT as an early Jewish
Christian christological title, and certainly makes good sense as Jewish mes­
sianic nomenclature (if it is not actually attested in pre-/non-Christian Jewish
sources).50 Paul's evident fondness for articular substantives as titles for Christ—
again, particularly well attested in Romans — lends still further support to a
titular use of ό δίκαιος in 1:17 as a reference to Christ.51 Moreover, for Paul to use

on this point: " T h e Righteous One as Eschatological Deliverer: A Case Study in Paul's Apocalyptic
Hermeneutics," in Apocalyptic and the New Testament (ed. J. Marcus and M. L. Soards; Sheffield:
JSOT Press, 1988) 191-215.
48
Hays admits candidly that "[t]hese reflections about the possibility that Paul understood
Hab. 2.4 as an apocalyptic testimonium to the coming of an eschatological deliverer are of course
not probative in their force . . ." ("The Righteous One' as Eschatological Deliverer," 209, 211).
49
The evidence is complicated by Paul's occasional use of δίκαιος to denote simply a "good"
or "upright" person, often in a legal context, over against rightness as a divine attribute that
approaches perfection (see BAGD, 195-96: for the "good" man, see Matt 1:19; 1 John 3:7; and
Rev 22:11). A shift in usage has been noted by, among others, G. Schrenk, TDNT 2. 182-91. That
Paul would be using δίκαιος in a legal or common sense in Rom 1:17b is, in any case, unlikely:
Is Paul's gospel concerned with mere "goodness"?
50
See Acts 3:14; 7:52; and, most significantly, 22:14, where Luke actually places it on the lips
of Paul (although its use is ascribed to Ananias). This of course may not testify to Paul's knowledge
of the title, but it does suggest that Luke thought the title a standard one in the context of Jewish
Christian proclamation. The title also occurs in Jas 5:6, anarthrously in 1 Pet 3:18; 1 John 2:1
(and 2:29 and 3:7?), and possibly at 2 Tim 4:8. It is also in 1 Enoch 38:2 and 53:6 (see R. N.
Longenecker, The Christology of Early Jewish Christianity [London: SCM, 1970] 466-47; and
the now-definitive statement by Richard B. Hays, " T h e Righteous One as Eschatological Deliv­
ered—although I feel that Hays could perhaps have strengthened his case further by giving greater
attention to the significance of the presence [or absence] of the definite article [see n. 51 below]).
51
This stylistic tendency seems evident in Romans 6, where Paul discusses the relation between
Christ's death, Christian baptism, and sin. Robin Scroggs argues that here ό αποθανών refers to
Christ ("Rom 6:7 ό γαρ αποθανών δεδικαίωται από της αμαρτίας," NTS 10 [1963] 104-8; see also
8:34 and 14:9). A parallel tendency to describe Christ in terms of articular substantives is apparent
 Campbell: Romans 1:17 283

a text to support a christological point would not be unusual, especially in

Romans, where he does so frequently.52
In close relation to the foregoing, a case can be made for the use of Hab
2:4 in the early church as part of a messianic testimonium. C. H. Dodd first
suggested that the practice of collating testimonia was widespread in the early
church. 53 We know that Hab 2:4 itself was used in pesher exegesis at Qumran,
which suggests that it was at least discussed within late Second Temple
Judaism.54 Moreover, that Hab 2:4 was so used by the early church (and perhaps
parts of Judaism) is suggested directly by Heb 10:37-38, which almost certainly
uses a messianic form of Hab 2:4 within the Septuagintal textual tradition (cf.
MS A etc.). The crafted parallelism between ό ερχόμενος and ό δίκαιος suggests
this (cf. Isa 35:4; Matt 3:11; 11:3; John 1:15, 27; cf. also Hab 2:3, which is included
in the citation), as does the inversion of the pronoun to produce the distinctly
messianic phrase ό δίκαιος μου. Admittedly, chap. 11 of Hebrews discusses
anthropocentric faith rather extensively, but what is less often noticed is that
this long discussion is grounded in chap. 12 on the faithfulness of Christ; . . . τον
της πίστεως άρχηγόν και τελειωτήν Ίησοϋν (Heb 12:2).55

elsewhere in the letter, e.g., "the son" (Rom 1:3, 4, 9; 5:10; 8:3, 29, 32), "the Christ" (Rom 9:3,
5; 14:18; 15:3, 7, 19; 16:16), and "the one" (Rom 5:15, 17, 18, 19).
52
Paul cites scripture extensively throughout Romans (at leastfifty-fourinstances), and, almost
without exception, these texts are either theocentric or christocentric when used positively, or
involved in a negative critique of humanity and Israel (references from R. N. Longenecker, Biblical
Exegesis in the Apostolic Period [Grand Rapids: Eerdmans, 1975] 110-11; Nestle-Aland lists sixty-
five explicit references and numerous allusions: 899, 901-11). The one exception to these trends
is Gen 15:6, which, of course, uses πιστεύω. Hence, if Paul is using Hab 2:4 with reference to
Christ in 1:17, this function is certainly supported by that of numerous other such citations in
the rest of the letter; see Isa 8:14/28:16 in 9:33; Deut 9:4/30:12-14 in 10:6-7; Isa 59:20-21/27:9
in 11:26-27; Ps 69:9 (68:10 LXX) in 15:3; and Isa 11:10 in 15:12.
53
See Dodd, According to the Scriptures; a more sophisticated rehearsal of this project may
now be found in Richard B. Hays's Echoes of Scripture in the Letters of Paul (New Haven: Yale
University Press, 1990).
54
See W. H. Brownlee, "Messianic Motifs of Qumran and the New Testament," NTS 3 (1956-57)
209; and J. A. Fitzmyer, "Habakkuk 2:3-4 and the New Testament," in To Advance the Gospel
(New York: Crossroad, 1981) 23-24.
55
The commentators do not, however, often notice this—perhaps because they are overly
influenced by the traditional reading of Hab 2:4 in Paul! On the point, see Hays ("The Righteous
One' as Eschatological Deliverer," 202-6), who also cites the more extensive discussions of A.
Strobel (Untersuchungen zum eschatolog^schen Verzögerungsproblem auf Grund der spät jüdisch-
urchristlichen Geschichte von Habakuk 2,2ff. [NovTSup 2; Leiden: Brill, 1961]), Hanson (Studies
in Paul, 44-45), and Dietrich-Alex Koch ("Der Text von Hab 2:4b in der Septuaginta und im
Neuen Testament," ZNW 76 [1985] 68-85) in support of the point. Hays accepts that the auctor
ad Hebraios does not apply the text in a messianic fashion, although the text is cited as such.
I am not so sure: the genitive constructions of 10:39 may allow for an interpretation within which
the letter's recipients are drawn into —and hence are "of—the faithfulness, as against the shrink-
ing back, of the messianic protagonist of w. 37-38, instead of the more nonmessianic application
of merely imitating the qualities themselves. The dialectic between the faithfulness of both "the
elders," Jesus himself, and the letter's recipients, through chaps. 11-12 to my mind support this.
284 Journal of Biblical Literature

Taken together, these observations suggest that a case can at least be made
for the early church's messianic reading of Hab 2:4. If Paul shared such a
reading with the rest of the early church, then it is unlikely that he would
have departed from it (and certainly not without clearly indicating it in the
context). Alternatively, if Hebrews was written significantly later than Paul,
it may be that its messianic usage was derived directly from him; the auctor
ad Hebraios seems to be part of the Pauline circle (see 13:23).56
In view of these arguments it seems reasonable to read Hab 2:4 in Rom
1:17b with reference to the faithfulness of Christ. Certainly many features of
the text, in combination with Paul's predilections elsewhere, make this a pos­
sible interpretation. And at least this reading can go some way toward explaining
Paul's puzzling textual variant.57 Furthermore, if our contextual and lexical
arguments concerning the interpretation of έκ πίστεως in v. 17a hold (and only
if), then it becomes necessary to read Hab 2:4 in this way. Such a reading would
mean that Hab 2:4 is attesting scripturally to Paul's presentation of the gospel
of Christ specifically through Jesus' death on the cross, which is (argues the
rest of v. 17) the perceptible historical point where God's eschatological salva­
tion has finally become apparent. Thus, "the prophets" do indeed testify to
the gospel concerning Gods son—even if latter-day exegetes would not approve
of Paul's hermeneutical procedure. In view of this, however, the text is not a
scriptural attestation of how the gospel is appropriated, as has usually been
thought: its focus is substantive rather than instrumental, and christocentric
rather than anthropocentric.
With the establishment of this equation between έκ πίστεως, Hab 2:4,
and the faithfulness of the Messiah in 1:17, much in the rest of Paul's argu­
ment within Romans is both resolved and clarified. Certainly the contentious
πίστις Χρίστου constructions in chap. 3 are christological, while much of the
rest of Paul's discussion in chaps. 3-4 and 9-10 is tilted in a more christological
direction — although to discuss these interpretative consequences here as they
58
deserve is unfortunately not possible. Probably all that can be claimed at

56
Ben Witherington III suggests such a dependence but does not perceive a shared or derived
messianic interpretation ("The Influence of Galatians on Hebrews," NTS 37 [1991] 146-52, esp.
148-49). The arguments of those listed in the preceding note rather overwhelm this reticence,
however.
57
The absence of the pronoun specifically liberates the meaning of έκ πίστεως from either
a precise reference to God (as in the LXX Β or S), or to humanity (as in the MT). Two conclu­
sions are then possible: (1) Paul intended the phrase to include all three possible objects (God,
Christ, and humanity—or at least two of these: so Hays, but with the primary emphasis on Christ;
see his Faith of Jesus Christ, 15-57). But this seems an excessively kaleidoscopic reading—it rejects
nothing and so, in a way, decides nothing. Or (2) Paul intended the text to refer specifically to
Christ, in which case he had to excise the pronoun to make this possible (that is, to avoid point­
ing the faithfulness to someone else). Note, this excision also makes it possible to deploy έκ πίστεως
as a slogan throughout more complex arguments, which would not really be possible if the phrase
was encumbered with a possessive pronoun.
58
For some beginnings, see Hays, "ΠΙΣΤΙΣ and Pauline Christology," 720-24.
 Campbell: Romans 1:17 285

this point is that the burden of proof now rests firmly on those who would
read many of these genitives and arguments in a way that does not link πίστις
with Χριστός in a subjective fashion, that is, as an elliptical reference, mediated
by the terminology of Hab 2:4, to his death on the cross. Yet even this minimal
concession opens up the possibility of a major réévaluation of Paul's argument
in Romans (especially chaps. 1-4), and of his theology as a whole.
 ^ s
Copyright and Use:

As an ATLAS user, you may print, download, or send articles for individual use
according to fair use as defined by U.S. and international copyright law and as
otherwise authorized under your respective ATLAS subscriber agreement.

No content may be copied or emailed to multiple sites or publicly posted without the
copyright holder(s)' express written permission. Any use, decompiling,
reproduction, or distribution of this journal in excess of fair use provisions may be a
violation of copyright law.

This journal is made available to you through the ATLAS collection with permission
from the copyright holder(s). The copyright holder for an entire issue of a journal
typically is the journal owner, who also may own the copyright in each article. However,
for certain articles, the author of the article may maintain the copyright in the article.
Please contact the copyright holder(s) to request permission to use an article or specific
work for any use not covered by the fair use provisions of the copyright laws or covered
by your respective ATLAS subscriber agreement. For information regarding the
copyright holder(s), please refer to the copyright information in the journal, if available,
or contact ATLA to request contact information for the copyright holder(s).

About ATLAS:

The ATLA Serials (ATLAS®) collection contains electronic versions of previously

published religion and theology journals reproduced with permission. The ATLAS
collection is owned and managed by the American Theological Library Association
(ATLA) and received initial funding from Lilly Endowment Inc.

The design and final form of this electronic document is the property of the American
Theological Library Association.