Galatians 5.11: Evidence of an Early Law-observant Mission by Paul?

Abstract

Galatians 5.11 refers to Paul ‘proclaiming circumcision’—a proposition that he is concerned to refute because he constructs two compact but powerful inferences designed to falsify it. One argues from present persecution, the other from the cross. Following a precise reconstruction of these it can be shown that the three main previous interpretations of the reference of Paul's ‘proclamation of circumcision’ are dubious, whether in terms of a blatantly false charge by opponents, a phase in Paul's pre-Christian Jewish life, or an occasional apostolic mission to Jews. A fourth, embarrassing reading is more likely, especially when other comparable missionary work is considered. Early on Paul proclaimed a fully law-observant gospel to pagans that included circumcision, but then later revised his praxis.

Galatians 5.11 is a difficult, contested, but potentially highly significant text that merits careful scrutiny.¹ It will be discussed here in the following terms:

[5.11a²] Ἐγὼ δέ, ἀδɛλφοί, ɛἰ πɛριτομὴν ἔτι κηρύσσω,

[5.11b] τί ἔτι διώκομαι;

[5.11c] ἄρα κατήργηται τὸ σκάνδαλον τοῦ σταυροῦ.

At first glance, 5.11a suggests that Paul at some point ‘preached circumcision’, which would be a deeply interesting if not astonishing fact if it applied to his apostolic ministry. It is not surprising, however, that this reading has been widely marginalized in favor of other construals. Meanwhile, some exegetes simply despair of ever fully understanding Paul's pithy rejoinders.³ I suggest in what follows, however, that the verse's difficulties can be negotiated, and that the embarrassing ‘apostolic’ reading is the most likely construal of Paul's claims in 5.11a, and to some degree for just this reason.

1. The Original Text

As we turn to detailed investigation our first question concerns the text's original form since the Western tradition omits the first instance of ἔτι, from 5.11a.⁴ If this variant is original, the majority text could be explained as the eye of a scribe generating a doubling error by traveling forward along the line to the second ἔτι and including another instance earlier. But omissions are usually easier to explain than additions since scribes, paid by quota, tended to hurry and to compress. Moreover, it seems unlikely that this addition would then have influenced so many subsequent manuscripts in different major families, while our earliest MSS also include it. So the Western reading that omits the first ἔτι is almost certainly secondary. This variant might have been caused by a careless omission that reversed the foregoing error—a scribal eye jumping forward to the second ἔτι and omitting the first, or simply getting confused during aural and/or mental repetition. Also possible, however, is the deliberate omission of an embarrassing word, which eliminates the obvious further problem that Paul now never preached circumcision. (‘How could this be?!’) Reconstructing textual variants is not a precise science, but the possibility of deliberate scribal omission on grounds of embarrassment will complement our later arguments nicely.⁵

In due course we will see that a majority reading of this text exists, along with two minority readings, but that these approaches all possess debilitating problems. So I will contend that another, less popular approach does solve these difficulties. Somewhat counter-intuitively, however, it will be clearest to consider one of the minority positions first.

2. J. L. Martyn's Non-temporal Reading⁶

Most construals of 5.11 grapple with the usual temporal sense of ἔτι in terms of ‘still’ (‘…if I still preach circumcision then…’). Martyn, however (partly in dependence on Mussner⁷), argues for a rather different translation, and if he is successful then all the biographical implications of the verse are eliminated at one fell swoop, not to mention its difficulties. But I am not convinced that he has made his case.

The NT uses the adverb ἔτι almost one hundred times, Paul using it fifteen or sixteen times. The most usual sense is one of continuation or extension, whether through time or space, some activity, or by way of analogy (e.g., extending inclusion within a group to ‘yet/still one or two more’; cf. Matt 18.16), suggesting the translations ‘still’, ‘yet’, ‘further’, ‘even’, and occasionally ‘additionally’.⁸ The best translation in all Paul's material, barring question-begging instances, is ‘still’.⁹ However, it is possible that Paul is using one of its less common senses in 5.11. And Martyn has suggested the meaning ‘additionally’ for its occurrence in 5.11a¹⁰ (urging the same meaning for its occurrence in 1.10¹¹). ‘If I were on occasion advocating the circumcision of Gentile converts, the persecution of me…would cease’.¹² But this is of course a paraphrase. A stricter translation would be: ‘And brothers, if I preach circumcision by way of addition, then why am I still being persecuted?’¹³

The broader implication of this reading, and its main strength, is the suggestion that a blatantly false claim is being made by Paul's opponents—that elsewhere (i.e., outside Galatia), and only from time to time,¹⁴ Paul proclaims a gospel that includes a commitment to circumcision.¹⁵ This reading then removes the awkward implication from the usual translation that Paul may at some earlier point have preached circumcision himself. Paul's opponents could plausibly then have gone on to suggest that the apostle's original proclamation to the Galatians that omitted circumcision could fairly be supplemented by the Teachers' addition of the practice, while the apostle seems to have been an inconsistent and untrustworthy figure (cf. 1.10; 2.11-14).

Martyn's suggested translation is typically insightful. But it is unlikely.

(1) The second ἔτι, in 5.11b, cannot be rendered ‘additively’. How could Paul be persecuted ‘in addition’ or by way of extension except temporally, as something ‘still’ happening? So the adverb takes its usual sense in 5.11b of ‘still’. And in such a short, pithy set of arguments it seems more likely that Paul would hold the meaning of the adverb constant than that he would shift the meaning of ἔτι in successive clauses. (Martyn concedes that this is a problem.¹⁶)

(2) The word order supports an adverbial interpretation, ἔτι invariably modifying the material that follows it elsewhere in Paul (and almost always in the NT¹⁷), here ‘preaching’.¹⁸ Hence, ἔτι must be functioning out of sequence for Martyn's construal of 5.11a to work, which is not impossible but neither is it ideal.

(3) Martyn's reading views the charge behind 5.11a as an outright falsehood, a sheer fabrication, at which point it seems puzzling that Paul does not simply deny it. Paul frequently repudiates lies under oath in his letters (cf. 2 Cor 11.31), and even does so earlier on in Galatians (cf. 1.19-20). Moreover, it seems an odd charge to fabricate. (Martyn admits to being puzzled by this.¹⁹) If it is blatantly untrue then it would seem to risk rebounding onto the heads of its fabricators quite directly; they are liars.

These counter-arguments do not amount to a definitive repudiation of Martyn's suggestion, but if we can formulate a less problematic reading based on more typical usage then it should be preferred.

As we turn to consider the remaining interpretative options I suggest that we understand the syntax as follows.

5.11 is a complex first class conditional argument in which two apodoses are supposed to falsify in coordination the ostensible facticity of a single protasis.²⁰ We are led to this syntactical judgment because—as will become apparent shortly—it is difficult to supply a coherent line of argument to the text if the claim of 5.11c is read directly on from 5.11b, and yet both 5.11b and 5.11c function well as attempted refutations of 5.11a. (It is also impossible to assign 5.11b a parenthetical function.) So the most plausible reconstruction views 5.11b and 5.11c as functioning together, in coordination, to refute the charge of 5.11a.²¹ This leaves us with two related but distinguishable arguments—5.11a-b and 5.11a-c—that can be traced through in more detail. Both are highly compressed but powerful contentions.

3. The Argument of 5.11a-b

In 5.11b Paul suggests in a conditional argument that he is not ‘still’ preaching circumcision (so 5.11a) because he is ‘still’ being persecuted. The apodosis therefore supplies a narrative and biographical contention. Moreover, the overarching refutation of 5.11a seems to rest on a true claim. (The argument in 5.11c will work rather differently.) But we need to supply quite a bit of additional information to Paul's argument in order to grasp it fully.

For Paul's argument in 5.11a-b to hold, his gospel and its revolutionary stance on circumcision must have elicited at least some of his previous persecutions, and this information must have been clear to his Galatian auditors. (Note that we are not considering here why a circumcision-free gospel generates such opposition, but just the brute fact of it.) That is, the Galatians must know that Paul's circumcision-free gospel and any persecution of him exist in a directly correlative relationship. And the converse must also hold: that when Paul's—or anyone else's—gospel endorses circumcision it does not seem to entail persecution. Moreover, these two associations—of circumcision-free preaching and persecution, and pro-circumcision preaching and no persecution—must be mutually exclusive. Put more formally then: A entails Y, and a converse situation, B, entails Z; A-Y and B-Z are mutually exclusive. Paul states in 5.11b explicitly that Z is present (persecution), therefore A is not the case (an inoffensive pro-circumcision gospel; so 5.11a); rather, B is (which is implict; Paul continues to preach a circumcision-free gospel). So here the truth of ‘Z’, supplied specifically in 5.11b, refutes the ostensible truth of ‘A’, stated in 5.11a.

It will follow then from the Galatians' knowledge of these largely biographical equations that any roughly current persecution of Paul—so 5.11b—suggests fairly directly that any purported abandonment by him of a circumcision-free gospel in favour of one endorsing circumcision must be false, thereby negating 5.11a. Without this narrative, however, Paul's argument does not really make sense. That is, it is not merely invalid but incoherent.²² But it is likely that the Galatians were aware of this correlation.

Scholars have tended to debate in the past whether the persecution evident in relation to Paul came from Jewish Christians in particular or from outside the early church from Jewish groups.²³ Certainly, however, there is sufficient evidence in Galatians alone to posit its existence and thereby a narrative in which Paul's circumcision-free gospel has been opposed strongly, at times to the point of persecution.²⁴ And if Paul had recently traveled through Galatia in person, that is, shortly before writing the letter, as both 1 Cor 16.1 and Gal 4.13 might suggest, then he has even personally recounted to the Galatians a story of his recent troubles in Jerusalem precisely over his circumcision-free approach to mission.²⁵ So the correlation implicit in Paul's argument in 5.11a-b seems plausible. He has been persecuted in the past for his circumcision-free gospel; these two phenomena exist in a direct relationship. Moreover, the converse also holds. Others do not seem to have been persecuted for preaching a gospel that includes circumcision (cf. esp. 6.12), a trajectory that included Paul at some stage. Hence the refutation of 5.11a by 5.11b holds. If Paul is still preaching circumcision then he clearly would not be suffering such persecution. But he is suffering such persecution so equally clearly he is not still preaching an inoffensive gospel that affirms circumcision but its converse. And we have now explained Paul's second use of the word ἔτι in 5.11b.

At first glance this word does not seem necessary in argumentative terms. Paul only needs instances of relatively recent and appropriate persecution to suggest that he is not therefore ‘still’ preaching an inoffensive gospel of circumcision. And these could well have been drawn (and probably were) from events in Jerusalem and Judea—the pressures he and Barnabas have just suffered there. The second use of the adverb ἔτι in 5.11 in a temporal sense is not then really necessary. ‘If I am being persecuted, then I am not still preaching circumcision’ would suffice. But the use of this word a second time achieves at least one, and possibly two, things.

First, it has a noticeable rhetorical contribution to Paul's argument. Because it resumes the ‘still’ of the protasis in 5.11a it is this particular notion in 5.11a that is refuted by the sharp riposte in 5.11b. This emphasis is clearest in reverse. ‘That I am still being persecuted suggests that I am not still preaching circumcision’. And this is a highly revealing emphasis for the Pauline biographer since it points explicitly to the concession lying within Paul's argument that he did in fact at some stage ‘preach circumcision’. The apostle even seems to emphasize here that he is not ‘still’ doing this, thereby conceding that once admittedly he did. He is just denying—emphatically—that he is doing this now.

Martyn adds a second interpretative layer.²⁶ He observes that the second instance of ἔτι pulls Paul's narrative of persecution right into the present, thereby neatly invoking Paul's present difficulties at Galatia as a further instance of his history of persecution by his opponents—opponents, that is, of his apostolic ministry to the pagans in liberative terms. (Paul could not of course have included a narrative of his difficulties in Galatia to the Galatians if he had passed through recently; they would not yet have occurred.) The word suggests that these situations all belong together in one trajectory. The Galatians' own situation therefore confirms Paul's refutation of the shocking charge implicit in 5.11a—and they are drawn into his side of the battle: ‘If I were on occasion advocating circumcision of Gentile converts, the persecution of me that was commenced by the False Brothers at the Jerusalem meeting—and that continues to this day in the activity of the Teachers—would cease. In fact, as you yourselves know very well, that persecution has by no means come to an end…’²⁷

4. The Argument of 5.11a-c

A second conditional argument is effected by 5.11c that begins with the inferential particle ἄρα.²⁸ But whereas 5.11b introduced a biographical contention, 5.11c makes a more overtly theological claim in relation to the cross. And whereas 5.11b used a true claim to refute the ostensible truth of the protasis, assuming that the associations of the two claims were mutually exclusive, 5.11c uses a false, and even absurd, claim to indicate the falsity of 5.11a. So here the two claims are correlative, not mutually exclusive, and the obvious falsity of the second proves the falsity of the first. (I suspect that it is Paul's shift between these two different types of argument, coupled with their compressed, allusive developments, that has led to most of the interpretative difficulties surrounding Gal 5.11 in the past.) Hence the argument of 5.11a-c fills out a lacuna in the previous argument and in doing so fashions a brief but effective rejoinder.

We need first to recall the two correlations present in the argument of 5.11a-b that are doubtless still resonating when 5.11c is read: preaching a pro-circumcision gospel is inoffensive, but preaching a circumcision-free gospel is offensive. Indeed, some are so offended by it that on occasion they persecute its advocates. In the light of these assumptions Paul now runs a further argument:

1. (1) The cross of Christ is a shameful event—a σκάνδαλον—and hence is offensive.²⁹

2. (2) The cross of Christ cannot, however, (obviously!) be nullified in this sense (at least without massive reinterpretation and long cultural reorientation). The cross remains irreducibly offensive.

3. (3) And Paul still preaches—and even embodies—the cross and its obviously offensive implications (cf. 3.1; 4.13, 14; 6.17). Far from obscuring this dimension, he glories in it!

4. (4) Paul has not therefore abandoned his offensive gospel.

5. (5) By implication, however, he still preaches an offensive gospel in the sense of a gospel free of circumcision (that is, for male pagan converts).³⁰

6. (6) It follows from this that Paul has abandoned his inoffensive gospel of circumcision—so 5.11a.

7. (7) Note that if he has not, then the converse would apply—that he has abandoned the offensive cross, which is self-evidently not the case (see 2 and 3 above)—so 5.11c.

Paul's specific reasoning in 5.11a-c runs more formally, if A is true (an inoffensive pro-circumcision gospel) then F would also be true (an inoffensive cross)—which we understand is the case from a chain of implicit connected claims, B through E—but F is obviously false (a cross emptied of offence) therefore A is false as well (that Paul preaches an inoffensive pro-circumcision gospel). And it seems that the exact meaning and force of all of Paul's contentions in 5.11 have now been settled—a double refutation of the truth of the claim that Paul ‘still preaches [a gospel including] circumcision’. This is denied, on two counts. But we have yet to determine exactly when Paul proclaimed circumcision.

5. Identifying when Paul Preached Circumcision

We noted earlier that a majority opinion is apparent in the literature, followed by two minority suggestions, and the view that will ultimately be urged here. This yields four main options. (Other more idiosyncratic views with obvious weaknesses will not be treated in detail.³¹)

1. (1) 5.11a refers to preaching activity by Paul before his call;

2. (2) 5.11a refers to missionary work to Jews by Paul after his call;

3. (3) 5.11a refers to an early phase in Paul's missionary work to pagans, after his call, when he did expect his male converts to be circumcised (a phase that has now passed); and

4. (4) 5.11a utilizes a non-temporal meaning of ἔτι and is simply a fabricated charge—that Paul occasionally preaches circumcision, i.e., when it suits him (so Martyn).

But the fourth reading has already been challenged so only three remain to be discussed in detail.

5.1. A pre-Christian Mission by Paul?

Perhaps 5.11a speaks of a mission on behalf of Judaism by Paul before he was called to the apostolate, a position well-represented by Terence L. Donaldson but dating back much earlier in the modern interpretative tradition, and really dominating it.³² However, this reading suffers initially from the increasingly apparent flaw that there is little evidence of overt missionary work by Jews before the time of the early church, so it needs to be reformulated.³³ Rather more likely then is the sense that 5.11a refers to a vigorous ‘proclaiming’ of circumcision (and presumably of related matters in the Torah) by Paul in some teaching capacity within the synagogue or some similar forum as suggested by Acts 15.21, perhaps along with Rom 2.21. And this is a possible, although distinctly minority, construal of κηρύσσω.³⁴ However, this somewhat domesticated interpretation has one principal difficulty—of rhetorical opacity.

It is difficult to reconstruct a meaningful argumentative (i.e., rhetorical) function for 5.11a in these terms. Understood as a reference to pre-call activity, on the one hand, Paul's critics would seem to gain little if anything by introducing it into discussion. And neither would Paul. On the other hand, Paul would not really need to refute it, and certainly not in these terms. So in all respects the argumentative function of the pre-call view is problematic. In a little more detail:

It is next to impossible to explain the presence of this assertion in the Galatian situation if it refers to pre-call activity on Paul's part. Paul had no need to raise this activity himself (that is, other than in the contextually irrelevant terms of reversal that we will discuss momentarily). It seems almost certain then that his opponents have raised this matter³⁵ as part of their counter-biography.³⁶ But it is not especially damaging to charge the apostle to the pagans with proclaming Moses—and even proselytizing Jews, if this activity is granted—before he was called to Christ! Indeed, this seems singularly otiose. Moreover, easy refutations of any such suggestion lie just to hand that for some reason are not deployed.

Paul's usual strategy in relation to his background is reversal (see esp. Gal 1.13–17 and Phil 3.4–10). He uses zealous but misguided activity prior to his call as evidence of the divine nature of that call since it cut diametrically across those previous activities.³⁷ And that strategy would clearly work here as well. So it is puzzling that he does not make this move simply stating that in the light of Christ his prior teaching zeal—or his missionary ardour—has been abandoned: ‘Even as I used to proclaim circumcision, God revealed his son to me, and now I preach him!’ But he does not. (Indeed, he does not even mention any such missionary activity in his autobiographical texts.) But Paul could in any case just reply to this charge with the response ‘so what’ (τί γάρ or some such; cf. Rom 3.3). It signifies nothing in relation to his current Christian activities. By this point in the letter the misguided nature of his pre-call life is old news (cf. 1.13–14). Moreover, in view of these effective and reasonably straightforward possibilities there is no need for Paul to make the weaker and more complex inferential and concessive arguments that he does make in 5.11b and c—the arguments associated precisely with his use of the word ‘still’ that have so often baffled his later readers.

In short, the pre-call reading cannot explain what this text is doing in the letter—why the charge is present in the text at all, why it is weighty enough to call for argued rejection, and why it is rejected in the subtle, convoluted, and concessive fashion that it is. Hence it seems that if another reading can deal with these exigencies it will enjoy a significant advantage over the pre-call reference.

5.2. A Post-call Mission to Jews?

Partly in view of these difficulties Francis Watson has argued cleverly that Paul is referring in 5.11a to previous missionary work when he was first called to be an apostle, but to Jews.³⁸ Indeed, 1 Cor 9.20, supported by various statements in Romans 11, indicates in Watson's view a turning by Paul to the pagans only after the failure of this mission to the Jews. Watson appeals further to Paul's testimony that he received a Jewish lashing five times (cf. 2 Cor 11.24) as evidence of extensive contact with Jews in a missional setting, and many scholars have cited Paul's circumcision of Timothy in this relation, as reported by Acts 16.3. But I am not convinced by this case, for a number of reasons.

We really have no idea when Paul received his synagogue lashings. These might fall into an early period of proclamation, but equally well might not. And even if these took place in an early period, we do not know exactly what gospel elicited them (cf. 1 Thess 2.14–16). Hence, although Watson's thesis is consistent with the lashings it is by no means necessitated by them.

More importantly, significant positive evidence stands against Watson's claims. All the references to Paul's call suggest that pagans were intrinsic to it, and this evidence cannot simply be dismissed. Galatians 1.15–16 are particularly clear, but the rest of the narrative in Galatians is consistently problematic for Watson as well. Paul tells us explicitly that his apostolic call and ministry were to pagans from its first moment: ῾Ότɛ δὲ ɛὐδόκησɛν [ὁ θɛὸς]…ἀποκαλύψαι τὸν υἱὸν αὐτοῦ ἐν ἐµοί, ἵνα ɛὐαγγɛλίζωμαι αὐτὸν ἐν τοῖς ἔθνɛσιν [κ.τ.λ.]. However, not all scholars advocating this view would necessarily follow Watson at this point, and they can soften the reading with the claim that Paul preached circumcision to Jews at times even as he was principally taken up with a mission to pagans; and this is the position of, among others, James D. G. Dunn.³⁹

But this modified reading faces the obvious problem that Paul would have had no need to deny his occasional proclamations of Christ in tandem with circumcision to Jews in the Galatian setting, which explicitly concerns converts from paganism (cf. esp. Gal 4.8; 5.2). Indeed, to a charge couched in these terms in 5.11a Paul could have replied just as he does in 1 Cor 9.20. By 5.11 the letter to the Galatians has conceded the existence of a different gospel to Jews that includes circumcision (see 2.7–8), and just the same considerations apply to any invoking of the circumcision of Timothy, who had a Jewish mother (Acts 16.3). At bottom, Paul does not need to reject a charge concerning an orthopractic evangelism of Jews, and has already implicitly suggested as much. Hence, to cite 1 Cor 9.19–23 in this relation is not to explain the charge in Gal 5.11a so much as to problematize the text, which is clearly talking about something different. Paul's vociferous rebuttals of 5.11a in 5.11b and c are now simply opaque. But this leads to a still more awkward point for this reading.

Paul states in 5.11a that at some point he ‘proclaimed circumcision’. Ideally, in Dunn's construal, πɛριτομήν denotes the constituency to whom he was preaching in an accusative of respect or something grammatically analogous. But it seems highly unlikely that Paul is speaking of a constituency here on analogy to 2.8 (which actually uses a genitive); he is referring to the content of his apostolic proclamation, as the following refutations make clear. And it makes little sense to suggest that Paul is speaking here in some way of the need for Jews to become circumcised. This is simply absurd.

In short, this approach to the charge underlying the protasis in 5.11—in terms of an occasional mission by Paul to Jews—seems to collapse in all its variations into incoherence. That Paul had an occasional ministry to Jews is something that does not need to be denied (although that he ever had a sustained apostolic mission to Jews that preceded his mission to the pagans seems doubtful). But Paul is denying the content of 5.11a, and emphatically, so clearly we need a reading that can explain this. Only one option remains.

5.3. A Post-call Mission to Pagans?

A possibility sometimes noted but rarely advocated is a reference by Paul to an earlier phase in his apostolic ministry when he did in fact ‘proclaim circumcision’ (i.e., in the sense of expecting any male converts from paganism to adopt full Jewish law-observance).⁴⁰ This implies further that Paul made an important transition at some point in his missionary praxis to a gospel that did not require circumcision of its converts from paganism, a shift that his present opponents clearly deplore but find rhetorically useful.⁴¹

This reading has been widely dismissed, but on grounds that turn out on closer examination to be quite fragile; in particular, they tend to beg the question. Moreover, a range of evidence, drawn both from the primary text and broader church history, stands positively in the reading's favour. It seems then that the interpretative road less travelled turns out to be strangely workable. We will begin a more detailed advocacy by addressing past objections and their weaknesses.

There are two distinguishable types of protest, both of which reject the post-call reading on essentially a priori grounds.⁴² For some, it seems that the principal problem is simply one of inconsistency.⁴³ Interpreters are offended by Paul ‘flip-flopping’ on the basic question of circumcision.⁴⁴ But for others, especially those influenced by loyalties to a particular construction of the Reformation, it is inconceivable that Paul could have undergone this sort of development given what we know from elsewhere of his basic theological convictions—in particular, of justification by faith.

(a) Theological inconsistency versus development. Inconsistency always creates rhetorical vulnerabilities in a contested situation. But it is still ubiquitous. Indeed, in a postmodern interpretative setting, the challenge is to find a text (or figure) that (or who) is not inconsistent in some significant sense.⁴⁵ Church tradition is consequently thickly populated with what are more charitably called ‘developmental’ thinkers. Tertullian berated Montanism, and then converted to it.Augustine ridiculed predestinarian Manichaeanism for years, then campaigned vigorously against Pelagius on behalf of predestination. A recent benchmark biography of Martin Luther has been written in three volumes corresponding to the three main developmental phases in his life (1483–1521, 1521–32, and 1532–46 CE).⁴⁶ John Henry Newton converted to Christianity in 1748 but continued to work in the slave trade until 1754, only writing his famous denunciation of slavery, ‘Thoughts upon the African Slave Trade’, in 1788, forty years later. And so it goes on.

Consequently it is simply not compelling to claim that Paul could not have changed his mind or undergone a major transition. He may well have, like many famous Christians subsequently, at which point we should explore its implications and not (implausibly) deny its possibility. Indeed, to deny it is to beg the question concerning important possible evidence that he did change his mind—that he was a developmental thinker in this respect.

(b) Theological impossibility versus missiological development. Various post-Reformation interpreters have of course denied Paul's post-call gospel of law-observance on apparently stronger grounds. But such denials are invariably based on what is known as ‘the Lutheran reading’ of Paul.⁴⁷

The Lutheran reading is not so much a reading of Paul strictly in terms of Luther—although it has important points of contact with both Luther and subsequent Lutheran tradition—as a construal of his soteriology or ‘gospel’ in terms of a particular theory. The theory is well known and tends to be articulated in relation to what we might call Paul's justification texts where he deploys a set of interrelated motifs—works of law, judgment, justification, righteousness, faith, and the figure of Abraham. The ‘Lutheran’ theory understands Paul to be giving a universal account of conversion and salvation in these texts, describing how the generic convert (including himself) progresses from an agonized experience of condemnation under the law to a blessed experience of salvation and forgiveness through faith alone, an account consequently centered on the individual's introspective journey to salvation. So Paul's shift from Jewish persecutor of the church to Christian apostle must correlate exactly with the shift in the Lutheran theory from an agonized existence in relation to attempting works informed by the law to the blessedness of forgiveness in relation to Christ and faith alone. Paul, according to this theory, simply had to have left all works of law behind on the road to Damascus; he began his journey not knowing Christ but struggling unhappily and anxiously to fulfill the demands of the law, and ended it knowing both of Christ and of justification by faith alone, the demands of the law having been entirely abandoned. And herein lies the impossibility of a post-call reading of Gal 5.11a.

If an interpreter is already convinced on the basis of Paul's justification texts that the apostle's account of salvation is ‘not through works of law but through faith alone’ then it is essentially inconceivable that Paul would ever ‘proclaim circumcision’ as a Christian apostle.⁴⁸ This would constitute a paradigmatic embarrassment.⁴⁹ But clearly this is not actually a good reason for rejecting this reading of Gal 5.11a. Such a denial simply begs the (admittedly important) question.

The Lutheran construal of Paul's justification texts does not account for more than a small portion of his extant writings on salvation and is, moreover, deeply contested as the correct interpretation of those texts.⁵⁰ Its cogency has been disputed by important scholars throughout the entire modern interpretative period, and mainstream Pauline scholarship has been significantly preoccupied with its debate for almost fifty years. Hence any denial of the post-call reference of Gal 5.11a in the name of the Lutheran construal of Paul's gospel is merely petitio principii. Read with a post-call reference, Gal 5.11a might be important evidence of the falsity of the Lutheran construal of Paul. And with this objection removed we can turn to some evidence that stands in favour of the reading.

The principal difficulty in the past with the post-call reading has of course been the supposed embarrassment caused by its inconsistency—that the apostle to the pagans who so famously advocated circumcision-free conversion included at some stage full Jewish law-observance in his missionary ethic. Hence the history of interpretation of this text, perhaps from its first scribal transmitters through to its recent historical critical interpreters, has largely been one of marginalization. But I suggest that this very embarrassment within the tradition attests to the probable truth of this reading.

It is precisely its embarrassment that delivers the reading's rhetorical force in the polemical Galatian situation, and that consequently explains both its presence in the argument and Paul's cryptic but powerful responses. Paul's opponents are, after all, trying to embarrass him (or worse). Its scandal is therefore actually one of its strengths. We know exactly why this charge has been made and this behaviour brought up! So we must now ask not so much whether the reading is too embarrassing to be entertained, as whether the further local data support this reading, and the missionary transition implicit in the embarrassment is plausible.

In fact, a reference by 5.11a to post-call, apostolic activity is suggested positively by Paul's language of proclamation in 5.11a that integrates most obviously with his widespread use of a diplomatic discourse to describe his apostolic work.⁵¹ It is clearly not implausible to suggest that Paul, in obedience to his divine commission, ‘preached’ or ‘proclaimed’ Christ to pagans—that is, acted as a Christian missionary. (He could do this whether the gospel's associated ethic was law-observant or not.) So we know why the text is freighted with his own language; it refers to Paul and so, in a sense, his opponents are merely throwing his own position back at him.⁵²

Moreover, we now know why Paul treats this issue in the anomalously concessive way that he does—that he is not ‘still proclaiming circumcision’. He cannot reverse this activity narratively and rhetorically at his call because it was not reversed then; it was only reversed later on. The best he can do is to supply the compact set of inferences in 5.11b-c arguing that evidence abounds that he is not still doing this.⁵³ Hence we have explained his repeated use of the word ἔτι. Some broader—and rather fascinating—evidence drawn from church history further corroborates this reading.

While certain theological expectations have in the past led to an overly static treatment of Paul, and to the concomitant dismissal of evidence of his development as suggested by texts like Gal 5.11a, data drawn from historical missions suggest that Paul's missiological development is anything but exceptional. Significant shifts in missiological praxis are evident throughout the history of Christian mission—such shifts frequently getting their advocates into trouble with central headquarters just as Paul experienced. That is, missionaries made perceived concessions in their specific missionary locations to local culture, which involved simultaneously shifting away from the traditional Christian ethic practiced originally by the missionaries and still practiced elsewhere by the rest of the church. This disjunction then caused conflict when it became known back at the mission's headquarters in the home church and country/ies. These narratives are uncannily parallel to Paul's story of fully endorsing then shifting away from sacred Jewish practices and toward more indigenous pagan practices in certain respects, a shift that later had to be vigorously defended against offended Jewish Christian traditionalists.⁵⁴

Two classic cases of this from later church history are Matteo Ricci in China in the late sixteenth century and Roberto de Nobili in India in the seventeenth, both of whom contextualized their Christian proclamations in astonishingly integrated ways, incurring the wrath of the Vatican in the process.⁵⁵ But China remains the missionary context where these dynamics have been most visible—probably because of its astonishingly rich indigenous culture that has repeatedly captivated missionaries. Indeed, a debate began in Ricci's day over ancestor veneration that continued for centuries known as ‘the Chinese rites controversy’.⁵⁶

Missionaries in China sensitive to Chinese culture often argued that traditional Chinese ancestor rites were merely an honouring of ancestors assisted by visual and practical prompts and so acceptable (cf. Exod 20.12)—a contextualized development in missionary praxis. But most Christian missionary movements viewed it in essentially Jewish terms as idolatry and false worship, and hence unacceptable, and so banned it. (Arguably the cause of Chinese missions was set back dramatically by this intransigent stance.⁵⁷) But many other points of tension became apparent in subsequent missionary work in China.

Timothy Richard, a legendary Welsh Baptist missionary in the nineteenth century, got into difficulties with his employers for advocating famine relief in addition to preaching, aphoristic wall posters instead of street evangelism and tracts (which appealed to Chinese literary and poetic culture), and a full engagement with Buddhist literature by Christian scholars (and so on)—each of these a development, if not a shift in direction, from standard missionary policy that caused controversy and offence.⁵⁸ But this missionary narrative is not confined to China.⁵⁹

In the light of these scattered narratives of missiological development—which by no means exhaust the phenomenon—the shift in Paul's praxis denoted by Gal 5.11a from an ethic demanding the circumcision of male converts, probably along with the observance of Jewish calendrical and dietary customs, to one abandoning such practices, seems unremarkable. Indeed, it appears to be a classic instance of contextualization by an intelligent and empathetic missionary at work on the ground within his designated constituency. Far from being unthinkable or impossible then, the narrative glimpsed by a post-call reading of Gal 5.11a is understandable if not commendable. This does not remove the story's negative value in a polemical situation; it remains embarrassing for Paul in Galatia that he shifted his praxis at some earlier point. But it is by no means impossible as the putative reference behind the text, while its very embarrassment is useful evidence in the reading's favour in its later polemical setting.

In short, it seems that the post-call construal of Gal 5.11a is superior to its strongest interpretative alternatives on all counts. There are no decisive objections to it, and much can be said directly in its support, all of which seems to suggest that its long interpretative marginalization should cease. And as a result of this we can conclude that Paul did spend part of his apostolic career proclaiming a gospel that included a commitment to circumcision, which is to say that he did not always relax this expectation for male converts to his movement from paganism. At some auspicious moment a significant transition in Paul's missionary praxis seems to have taken place from a gospel merely for the uncircumcised to one that allowed them to remain so following their conversion. But we must leave it to other studies to establish exactly when and where this transition took place, and to reconstruct its rationale. It suffices for now to conclude that Gal 5.11, suitably interpreted, points toward the fact that such a transition did take place. At an early point in his missionary career, and by his own admission, Paul the apostle preached a law-observant gospel, and his modern interpreters must learn to interpret him in a way that comprehends this intriguing transition, rather than ignores it.