On the other hand, I'm also working my way through [parts of] Douglas Campbell's The Deliverance of God: An Apocalyptic Rereading of Justification in Paul (Grand Rapids: Eerdmans, 2009). Campbell's book was the subject of the December 2011 issue of JSNT, with very critical review articles by Barry Matlock and Grant Macaskill. Of the very many distinctive reading positions and arguments in this very lengthy book, Campbell argues that we should read Rom. 1.18–32 as "speech-in-character," as Paul taking up the voice of and speaking as a character other than himself. Stanley Stowers has argued convincingly that Paul employs the rhetorical device, speech-in-character, in Rom. 7.7–25, so such things are not impossible. However, as Stowers emphasizes and Campbell acknowledges, anyone who proposes this as a reading strategy has to explain how Paul might have expected his readers to recognize his rhetorical maneuver. In Romans 7 there are several clear signals (which I discuss in my notes on Romans; if there's interest I can post those comments on this site). But Romans 1 lacks such clear signals. So Campbell argues instead,
It is my contention that the initial auditors of Romans could have detected such a strategy relatively easily through a plethora of nonverbal signals—the types of signals with which performed texts abound. And it must be emphasized that at this stage of our discussion it is necessary only to establish this identification as a possibility. Ultimately, my suggested rereading of Romans 1:18–32 will rely on evidence that emerges "downstream," so we will affirm this identification strongly only in retrospect. But it should be noted that this retrospective judgment is an accident of the text's canonical preservation, of the resulting loss of its original performed context, and also of some of our modern hermeneutical assumptions. We will realize belatedly what the Roman auditors could recognize relatively quickly, through the text's appropriate performance. For this reason, my auditors need for the moment only to be open to this reading of 1:18–32 as a possibility. Could the Roman Christians have detected a satirical textual operation here if it was indeed present? (Deliverance of God, 530; my emphasis [see pp. 530–41])
As I have already admitted, I am not a Pauline specialist, so I'm a bit outside my field. But I am a bit more experienced with oral traditional scholarship, and I have spent eight years reflecting on and exploring how to apply the theories, methods, and results of oral traditional scholarship to written textual questions. In fact, oral traditional scholars themselves have spent decades exploring this issue, so there's a vast body of literature on the subject. And oral traditional scholars have long emphasized that nonverbal and/or extratextual factors such as intonation, gesture, voice inflection, pace/tempo, musical accompaniment, social and/or physical environment, facial expression and other body language, eye contact, directionality, etc. etc. etc. affect how a written text communicates its meaning to an audience. In fact, it's safe to say that the written text itself didn't communicate its meaning to its original audience; rather, a reader/lector/oral performer communicated the text's meaning, and s/he did so via a number of techniques that (i) cannot be found within the tradition's textual stratum (i.e., the written text itself) and so (ii) is no longer available to us. This is unfortunate, because these methods, as Campbell rightly recognizes, are often if not always determinative for how and what a written text means.
Anyone reading Campbell's book, however, needs to know, this is NOT the way to apply issues of oral performance and/or presentation to (written) textual analysis. The problem with Campbell's argument here in Romans 1 is that there's no way to falsify it. There's no way to prove that Campbell's rereading runs exactly counter to Paul's intentions for Rom. 1.18–32 precisely because there's no evidence in support of his rereading. Unlike Romans 7, no textual clues suggest that Paul is speaking "in character" in Romans 1, and so all Campbell can do is grin and ask us to bear it.
But, given that, first, Rom. 1.18–32 makes perfectly good sense within Paul's own authorial voice (i.e., in his undisputed letters, in the disputed [or deutero-Pauline] letters, and in Acts, Paul speaks of God's judgment against the gentiles) and that, second, nothing that follows 1.18–32 requires us to hear Paul's comments about gentile depravity as "speech-in-character," we need to read this text straightforwardly, as the author presents it to us. Indeed, Thorsteinsson, in the much more helpful reading of Romans that I mentioned above, argues persuasively against exactly the kind of retrospective reading strategies Campbell employs here. So while we might grant that we could read Rom. 1.18–32 as a "satirical textual operation," we—and Campbell, too—need to admit that Paul has utterly failed to encode that operation with the text itself. And that failure, I would argue, is the best evidence against the proposal that we should read this passage satirically (pace Campbell, 541: "[s]uch texts therefore deliver their ironic and subversive potential entirely performatively" [emphasis his]).