I am grateful to Professors Driediger-Murphy, McKaughan, and Howard-Snyder for reading Roman Faith and Christian Faith so carefully, thinking about it so interestingly, and challenging me to think further about some important issues. This article responds broadly in the order of their articles, though some of their interests, and my responses, overlap.

Professor Driediger-Murphy raises a number of questions based on what she rightly identifies as three key aims of the book: to shift our understanding of what Greeks, Romans, Jews, and Christians did and did not have in common; to help us rethink what is distinctive about Christian pistis/fides; and to resolve a disconnect between scholars in different disciplines. Professor Driediger-Murphy's questions share significant points of interest with those of Professor Howard-Snyder, Professor McKaughan, and others who are currently seeking to develop our understanding of faith from the perspective of analytical philosophy. In particular, all three seek to refine our understanding of the cognitive aspect of faith. They ask what role cognition plays in faith (whether in Greek and Roman religions or Christianity), whether cognition's role changes over time, and if it does, how and why. What is involved in cognition: does faith, for example, necessarily involve commitment to propositions? If it does, do ancient writers, including writers of the New Testament, require any particular attitude (such as belief, hope, or trust) towards such propositions?

Professor Driediger-Murphy's first question is whether non-Christians from the second century onwards were right to see Christians as depending more heavily than themselves on fideism, in the sense of a ‘leap of faith’ that does not depend on practical demonstrations or logical proofs of the thing that is believed. If so, what caused the change? This is a crucial question which goes well beyond Roman Faith. It has not been asked, let alone answered systematically for the patristic period (for which there are remarkably few studies of pistis/fides or ‘faith’ in general). To do it justice one would need to examine the full range of literary, documentary, visual, and material evidence for the period; to consider the possibility that understandings and practices of faith varied in different regions, communities, or contexts within communities; to investigate how and why certain forms of cognition, together with the concept of faith as a state of mind or heart, became more important over time (Morgan (Reference Morgan2015), 11–12, 14); and to consider what is involved in cognition in a wide range of authors who differ from one another.

These and related questions are the subject of my forthcoming monograph, The Invention of Faith, which seeks to explain how and why Christian pistis/fides evolved (as it clearly did) from the first century to the fifth, and what impact evolved Christian understandings and practices had on the Christianized later Roman empire. Among other themes, I will argue that during this period Christian faith developed many of the cognitive and affective aspects with which we still associate it, including a greater emphasis on fideism. For now, I will try to offer the beginnings of an answer to Professor Driediger-Murphy's question.

Non-Christians undoubtedly were, to some extent, justified in observing that Christians of the immediately post-apostolic era did not rely heavily on either logical proofs or demonstrations (though through later antiquity into the mediaeval world, logic would play a growing role in theology). In this, Christianity seems to have evolved somewhat since its earliest years. In the New Testament we regularly find writers appealing, for instance, to the apostles’ witness to the resurrection, the evidence of Jesus's followers’ own eyes, and converts’ experience of the power of the spirit as a basis for pistis towards God and Christ (Morgan (Reference Morgan2015), 241–246, 354–365). From the early second century, we find writers appealing increasingly to tradition rather than witness, and report rather than direct evidence (though personal experience of the spirit remains a powerful basis for faith).

It is worth noting that in attacking Christians for not relying on proofs or demonstrations, non-Christians were being more than a little disingenuous. Tradition and report, as bases of pistis/fides towards the gods, are also invoked regularly and unproblematically by non-Christians (ibid., 145–151). In some respects the evolution of Christianity was bringing it closer to other cults.

Where Christianity was becoming genuinely distinctive was in its increasing use of tradition and report specifically to support propositional beliefs about God and Christ. (I use ‘propositional belief’, here and throughout, as in Roman Faith to refer to the conviction that certain propositions are true, without necessarily specifying whether the attitude involved is, for instance, trust, hope, or what philosophers would call an attitude of belief. We will return below to the question what attitudes are involved in these beliefs.) Some non-Christians found this challenging, probably because they found propositional belief in the divine challenging in general. Roman Faith shows (ibid., 142–145) that although pistis/fides in the sense of trust in the gods is strong and widespread in the world of the first century, propositional beliefs tend to be problematized. It suggests that this is because ‘[p]ropositional belief demands grounds and reasons. But when one ponders one's reasons for believing something about the gods, they soon begin to seem fragile’ (ibid., 145). Arguments and proofs for the existence or nature of the divine are usually problematized by intellectuals, so it is not surprising that there are signs of tension in this period between the relative ease with which worshippers pursued good relations with the gods, and the difficulties philosophers encountered in thinking propositionally about the divine. It suggests, however, that when non-Christians accused Christians of irrational fideism, the real issue was less Christians’ reliance on tradition and report as such, than their willingness to use both as a basis for propositional belief, and, in general, their increasing interest in propositional belief at a time when some educated Greeks and Romans were struggling with it.

The beginnings of this change are already detectable in some of the later books of the New Testament. The undisputed letters of Paul, for example, describe community members as putting their trust in God and Christ on the basis of Paul's and the other apostles’ direct experience of Christ (e.g. 1 Cor. 15.1–5, 2 Cor. 12.1–10), the power of the spirit which he communicates because he has been entrusted with the gospel (e.g. 1 Thess. 1.5, 2.3), and their own experience of the spirit when they heard the gospel (e.g. 1 Thess. 1.9, 2.10–11, 5.2). The Pauline letter to Titus offers a slightly different account. Titus is told that a community elder should be someone who holds fast to the pistos logos, the ‘trustworthy word’ or ‘discourse’, according to the teaching he has received (1.9), so that he himself will be able to exhort others with ‘sound teaching’ and ‘refute opponents’ (cf. 3.8).

Titus’ emphasis is less on the experience of the apostle who preached the good news, or Titus’ experience of the power of receiving it as the basis of pistis, than on the word itself. Trusting or believing in a word or discourse, however, has a more strongly cognitive, even propositional aspect than trusting or believing in a person. The same shift is visible in relation to the content of the good news. Compare what Titus says with the way Paul refers to the content of his preaching when it has come into question. Criticizing the Corinthians for listening to other preachers, Paul says, ‘It has been reported . . . that there are rivalries among you . . . each of you is saying, “I belong to Paul,” or “I belong to Apollos”. . .’ (1 Cor. 1.11–12). Chastising the Galatians for the same reason, he says, ‘I am amazed that you are so quickly forsaking the one who called you by grace’ (Gal. 1.6). The focus in these passages is on the Corinthians’ and Galatians’ abandonment of a relationship, not on their wrong interpretation of a discourse (though no doubt something of both was involved). In Titus, the emphasis is reversed.

Titus is written as if by Paul, but I share the view of many scholars that it is pseudepigraphical, and this helps to explain its understanding of the nature and basis of pistis. After the deaths of the first generation of apostles, who knew Jesus in life and/or witnessed the resurrection, community leaders needed to find new bases for members’ pistis and their own authority. What survived of the apostles’ missions were memory, oral tradition, and (increasingly) written reminders of what they said. These became a new source and focus of authority. Belief in words, though, is not the same as belief in a person; something of the relationality of pistis has been lost.

The Letter to Titus also hints at another cause of the shift towards propositionality: disputes between Christian groups. This theme emerges even more clearly in other late biblical epistles, notably in the Johannine letters.

1 John has a strikingly propositional character. Unlike any other New Testament letter, it begins with a proclamation rather than a greeting and proclamations recur throughout. The importance of believing certain things about Jesus Christ is summed up twice at the start of chapter 5: ‘Everyone who trusts/believes that (pas ho pisteuōn hoti) Jesus is the Christ is begotten by God’ (5.1); ‘Who is the victor over the world but the one who trusts/believes that Jesus is the Son of God?’ (5.5). All the author's statements about the importance of believing or affirming certain things about Jesus Christ come in the context of one of this letter's main preoccupations: battling the Antichrist and identifying those who are inspired by him. Propositional language comes to the fore when what is at issue are the differences between communities, and who is right and who wrong.

The earliest non-testamental letters reveal a little more of this evolution. Some examples are discussed in Roman Faith (Morgan (Reference Morgan2015), 512–514) and I will not repeat them here, but they underline that it is often in the context of conflict between Christians and groups or between Christians and (those identified by others as) non-Christians that propositionality to the fore.

These examples reveal the beginnings of developments in Christian religiosity which would lead to major changes in the tradition. I have yet to complete the work which will trace the paths of change through the next three centuries, but wherever we find concerns about authority or doctrinal disputes, we can expect to find evolutions in understandings of pistis/fides which do not abandon the relational altogether but which bring the propositional increasingly into prominence.Footnote ¹

So far I have not tackled the question whether pistis/fides necessarily involves a commitment to propositions, and if it does, whether ancient writers, including writers of the New Testament, require any particular attitude (such as belief, hope, or trust) towards propositions. I will return to this, but first I turn to another question from Professor Driediger-Murphy: was the idea, for which I have argued, that early Christians are told to place pistis/fides in God but not in one another as community members, one reason why Christians seemed periodically so threatening to the community at large? This is a very attractive hypothesis. It cannot be the only reason (nor does Professor Driediger-Murphy suggest it is). Christians put their trust only in the ‘living and true God’ and are therefore, in relation to all the other gods, what contemporaries call atheists, and they proclaim the coming kingdom of God, which sounds potentially politically subversive. In the Graeco-Roman world, however, pistis and fides are by no means due only to rulers: they are also fundamental to good social relations of all kinds. When, therefore, Christian apologists feel compelled to claim, as they often do (e.g. Justin, Apol. 1.7, Tert. Apol. 1–3), that they are not bad citizens, it is surely possible that they seek to present themselves, against what might seem to be their interpretation of pistis/fides, not only as loyal subjects of the emperor, but also as good participants in civil society.

I turn next to Professor McKaughan's article, and its central concern: whether early Christian discourses represent or demand an attitude of belief. Do writers of the New Testament represent an attitude of belief as a part of believers’ response to the gospel? Do they go further and indicate that it must be involved, alongside trust and other relational aspects of pistis?

1 Corinthians 15.1–5 is an important text for many reasons, not least because it is perhaps the most strongly propositional passage involving pistis language in Paul's letters.

These verses evidently have a propositional aspect, in the sense in which Roman Faith used the term. Verses 3–5 read naturally as a series of propositions, and Paul talks of the logos which he received, using a verb (paralambanein, vv. 1, 3) which is regularly used of the transmission of stories and traditions. Paul therefore places himself here, unusually, more in the position of the author of the Letter to Titus than that of a first-generation apostle (though in vv. 8–9 he will identify himself more closely with the first apostles). As noted above, belief in words handed down tends to be more propositional and less relational than belief in a person who is communicating their own experience of Christ.

Roman Faith argued that this passage, and the verses which come after it, are less straightforwardly propositional than they look at first sight (Morgan (Reference Morgan2015), 229–232). They also describe a relationship with God and Christ which is based on grace (cf. 15.10) and which leads to release from sin and the creation of a new divine–human community in God's kingdom. Pistis is therefore relational and situational as well as cognitive. (I observed that this is consistent with accounts of the kerygma elsewhere, and argued, in general, that trust and propositional belief are always intertwined in pistis/fides language (e.g. Morgan (Reference Morgan2015), 19–22), so there is nothing unusual about this in its first-century context.)

Can we go further and say that Paul demands an attitude of belief (as opposed, for instance, to one of trust or hope) to these propositions? I cannot see anything in Paul's language which allows us to argue one way or the other: nothing indicates that he was thinking in terms of such fine distinctions. If we ask whether Paul himself had an attitude of belief to these propositions, again, his language does not tell us, though it would take a dramatic – surely, to most scholars, an implausible – shift in our understanding of him and his mission to imagine that he did not. To determine whether Paul ever distinguishes between attitudes would require investigation of his writings as a whole, which is beyond the scope of this article. I am doubtful that a wider analysis would show that Paul ever makes such fine distinctions, but the question remains open for now.

If we widened the search, could we identify passages elsewhere in the New Testament in which an attitude of belief demonstrably either was or was not involved? A thorough investigation is again beyond the scope of this article, but I have not yet identified a New Testament passage in which I would argue that belief is demonstrably uninvolved. One possible candidate might be the mysterious affirmation of the father of the boy with a demon at Mark 9.24: ‘pisteuō; help me in my apistia’. One might choose to interpret that outcry as meaning, ‘I hope (or ‘trust’) that you can help my son [without having an attitude of belief]; help me to believe it.’ It could, though, equally well mean ‘I believe that you can exorcise my son; help me to trust you’, or ‘I believe and hope/trust that you can help my son [but not as strongly as I might]; help me to believe and hope/trust more.’ Any of these would make sense in the immediate context and none, as far as I can see, would be impossible in Markan theology, but we have no way of deciding between them.

This in itself may be significant. Mark and other New Testament writers, like Paul (and writers of this period in general), simply do not use pistis/fides language with the precision of modern analytical philosophers, and distinguish clearly between possible attitudes implied in pistis/fides. This also suggests an answer to Professor McKaughan's question whether New Testament writers assume that propositional belief must be involved, alongside trust and other relational aspects of pistis. If New Testament writers provide no firm evidence of whether belief is involved in faith, a fortiori they cannot tell us whether it must be.

We may be able to make a little more headway with a slightly different question which deserves attention in its own right. Is it, for New Testament writers, possible to have an appropriate, effective pistis, for instance towards Christ, without having right beliefs about him? If it is possible to have faith without understanding Christ rightly, then we may suspect (though this does not in itself prove) that an attitude of belief is not always essential in New Testament pistis, because common sense suggests that it is unlikely that a New Testament writer would think it essential for the faithful to have an attitude of belief in Christ in the event that the content of that belief was wrong.

In Roman Faith (Morgan (Reference Morgan2015), 359–361, 365) I argued that the synoptic gospels sometimes use pistis language to mark moments when someone, such as the woman with a haemorrhage or the Roman centurion, puts his or her trust in Jesus, believing or hoping that he is a healer, and encounters not only physical healing but salvation. We cannot be sure whether pistis here involves an attitude of belief, but if it does, the gospels strongly suggest that the content of the belief is not right, or not entirely (there is no sign that the woman or the centurion recognizes Jesus as the Messiah). Nevertheless, the pistis involved is clearly powerful and effective: in Mk 5.34 Jesus tells the woman with the haemorrhage that it has saved her. That such stories are told several times in the gospels suggests that pistis such as that of the woman or the centurion is acceptable and appropriate – even exemplary – even if it involves little or nothing of what a modern philosopher would identify as an attitude of belief in Christ as saviour.

The gospel of John offers an account of the content of what people believe which is famously elusive. As Bultmann put it, in a much-quoted passage of his commentary, the ‘central intuition’ of the gospel

When John's Jesus says, for instance, ‘Believe in God, believe also in me’ (14.1), the propositional content of this belief remains a riddle. In Roman Faith I argued that John plays constantly with what looks like propositional language in order to underline that knowing Jesus is in fact a matter of relationship (Morgan (Reference Morgan2015), 429), and that those who seek to know something about Jesus, like the Pharisees, simply show by that token that they are not among the elect (ibid., 425–432). What the elect believe, or know (the two lexica are often interchangeable) is above all God and Jesus themselves, not anything specific about them. Whether or not an attitude of belief is involved in Johannine pisteuein, therefore, John's deliberate vagueness about its content and criticism of those who want to know about Jesus strongly suggests that it is less important than his readers might expect.

In these examples, an attitude of belief may be involved in pistis, but it may be wrong; it may be changeable or elusive; and all the gospel writers, in different ways, suggest that the relationship between God, Christ, and the characters concerned does not depend on it. We cannot say of any of these passages that an attitude of belief is definitely excluded, but they do hint that for some early Christian writers, at least, it is not always a sine qua non of faith, central to it, or even essential to it.

We must, of course, distinguish between characters who appear in the gospels and the early Christians who listened to them. It is possible that a form of pistis which was portrayed as acceptable in Jesus’ lifetime, while his Messianic identity was not widely recognized or was even a secret, would not have been acceptable in early Christian communities. Against this is the apparently exemplary character of so many of those who encountered Jesus in his lifetime. It would surely be odd and confusing to early Christians to find a form of faith represented and validated in the gospels that was forbidden to them.

We must also recognize variations between New Testament writers. To judge by 1 Cor. 15.1–5, Paul is more concerned with propositional belief than any of the gospel writers. (In light of the earlier point about the post-apostolic generation, this may be because he did not encounter Jesus in life; Paul is both, as he calls himself, the last of the apostles and the first of the post-apostolic generation. The writers of the gospels presumably did not encounter Jesus in life either, but I have argued (Morgan (Reference Morgan2015), 362) that one of the aims of the gospels is to recreate for later followers something of the experience of encountering Jesus in his lifetime.) At the least, however, the passages we have briefly discussed suggest that having an attitude of belief towards propositions is not, for New Testament writers, central to pistis, and it may not always be essential.

Professor McKaughan also raises the question whether a study like Roman Faith which focused more on the Hebrew Bible and ancient Israel would show that they emphasized even more strongly the ‘fidelity/faithfulness/loyalty/allegiance side of the register’. His view, developed at greater length elsewhere (McKaughan (Reference McKaughan2017) ) is that it would. Up to a point I agree, but I would follow John Barton's recent, magisterial study Ethics in Ancient Israel in drawing a slightly different picture. Barton observes that he'emin lacks a strong dimension of ‘belief’, the nearest Hebrew comes to expressing propositional belief being probably the phrase ‘to say in one's heart’ (Barton (Reference Barton2014), 165–166), and notes that faith in the fideist sense is ‘not very important in the Hebrew Bible’ (ibid., 167). (In this he underlines two of the respects in which Christianity departs furthest from its Jewish inheritance.) He also discusses the verb ’mn and its relatives (ibid., 118, cf. 116–120, 241–242), noting the importance of their aspects of faithfulness and loyalty, but also their meanings of reliability, solidity, confidence, and trustworthiness. God is reliably faithful and the Israelites can rest on his solidity in confidence (or, as Isa. 7.9b puts it, challengingly, ‘If you do not stand firm in faith (ta'aminu), you will not stand (ta'amenu) at all’).

In Roman Faith I noted how often ’emunah language is translated in the Septuagint with the alētheia rather than the pistis lexicon, especially with reference to God (e.g. Morgan (Reference Morgan2015), 188 n. 37). This testifies to a strong sense not just that God speaks truth (though, of course, God does), but of the reality of God and the rightness of worshipping God (cf. 1 Thess. 1.9 on the ‘living and true’ God). New Testament writers occasionally use the language of alētheia of God where pistis might seem equally appropriate, but they use pistis language much more often. In these passages, Christians appear to be more interested in marking God's fidelity and faithfulness than God's truth. On this showing, trust and faithfulness are fundamental both to New Testament pistis and to the Hebrew Bible. Beyond this central range of meanings, the Jewish scriptures, in Hebrew and Greek, and the New Testament part company in several directions.

Professor Howard-Snyder is also concerned, in this and other recent articles (e.g. Howard-Snyder (Reference Howard-Snyder2013), (Reference Howard-Snyder2017), Rice, McKaughan, & Howard-Snyder (Reference Rice, McKaughan and Howard-Snyder2017) ), with the attitude of belief. His first question probes the relationship between attitudes of belief and doubt in pistis. For New Testament writers, specifically, must pistis consist in an attitude of belief towards certain propositions, or can it coexist with a kind and degree of doubt or scepticism that precludes such belief? Professor Howard-Snyder rightly notes that Roman Faith claims that pistis/fides is almost always cut with fear, doubt, mistrust, and scepticism. He also notes that on occasions in the New Testament pistis seems to be presented as precluding mistrust or disbelief.

We have already seen that faith may not always, for New Testament writers, involve an attitude of belief. It would not be surprising if fear, doubt or scepticism existed in some situations in which a person did not believe or did not have pistis. But can pistis coexist with fear, doubt, or scepticism? If it can, can we establish whether or not pistis then involves an attitude of belief?

There is abundant evidence that persons in New Testament writings who are described as having pistis can, in some contexts, be described as fearing or doubting, or as needing to progress in pistis (which might mean that they have emotions or attitudes which are ideally incompatible with pistis: so e.g. Morgan (Reference Morgan2015), 222–223). Roman Faith argues, for instance, that in the synoptic gospels, Jesus implicitly recognizes that his followers’ pistis is a work in progress. He occasionally rebukes them for apistia or oligopistia, but in much milder terms than those in which he attacks people who have not put their trust in him at all, and failures of pistis never cause him to expel anyone from the group (ibid., 356–357).

(We might, incidentally, wonder why it is implicitly acceptable that the disciples’ pistis is imperfect. One possibility is that these stories recall genuine fears and uncertainties of Jesus’ followers during his lifetime. Another is that the gospel stories offer models for the encouragement of later followers; a third that it is acceptable because listeners know that these stories have a triumphant ending. Apart from Judas, the disciples do eventually, beyond the end of the gospels themselves, become exemplars of unshakeable faith, even to their own deaths, and this may make stories of their early struggles less problematic to later Christians.)

Where pistis appears in a story alongside fear or doubt, do they coexist or do they, for instance, alternate? Again, a full investigation is beyond the scope of this essay, but Matthew's story of Peter's attempt to walk on water (Mt. 14.22–33), which uses the language of faith, fear, and doubt in close proximity, offers a promising case study. The disciples are at sea when they see Jesus walking towards them over the water. They are terrified (etarachthēsan, v. 26), thinking that he is a ghost, and cry out in fear (phobos, v. 26). Jesus tells them not to be afraid and Peter says, ‘Lord, if it is you, command me to come to you on the water’ (v. 28). Jesus does; Peter steps out of the boat and walks, but becomes frightened (ephobēthē, v. 30) and starts to sink. Jesus says, ‘O you of little faith (oligopiste, v. 31), why did you doubt (edistasas, v. 31)?’. He catches Peter before he sinks, whereupon the other disciples in the boat bow to him, saying, ‘Truly, you are the Son of God’ (v. 33).

In the first part of the story, fear dominates and pistis does not appear. In vv. 28–31, Peter's confidence (whether cognitive or affective) seems to give way to fear, causing him to sink. At v. 31, however, Jesus treats Peter's doubt as simultaneous with – even a description of – his oligopistia. Matthew's Jesus uses oligopistos in contexts where the disciples’ pistis is imperfect, but not non-existent (Morgan (Reference Morgan2015), 336–370), so it seems likely that for Matthew, doubt (and probably fear) coexist here with pistis, if not with perfect pistis. Moreover, Peter's failure of pistis involves a doubt either that Jesus is real, or that the real Jesus can enable him to walk on water, or even, if the other disciples’ words affirm what Peter should have believed throughout, that Jesus is the Son of God, or more than one of these, so his pistis evidently has a propositional dimension.

Does Peter's pistis in this passage involve an attitude of belief? At v. 28 Peter says, ‘Lord, if it is you, command me to come to you on the water.’ No attitude of belief is required here: Peter is not sure whether the figure is Jesus, and if it is, he implies that Jesus can enable him to walk on water simply by commanding it, whatever Peter's attitude. What makes Peter sink is said to be fear, not the state of his attitude of belief. Jesus then accuses Peter of doubting. Distazein can mean ‘to doubt’ or ‘to hesitate’; it can be used of cognitive doubt, but also of something more instinctual or emotional. At no point in this passage, therefore, can we be certain that an attitude of belief is involved. Nor, however, can we be certain that it is not.

Is Paul as tolerant as the synoptic gospels of doubt, fear (excluding appropriate fear of God), or scepticism among the faithful, and does he represent any of them as coexisting with an attitude of belief? Paul can praise communities for their faithfulness (e.g. Phil. 1.3–6, 1 Thess. 1.3, 7) while recognizing that human pistis is not necessarily perfect and has scope to grow (e.g. 1 Thess. 1.2–3, 3.10). We cannot, though, be sure what kinds of pistis are involved in these passages or why in the second case it is not perfect (are the Thessalonians fearful, doubtful, ignorant, inconsistent in practice or something else?). Paul also confesses to being imperfect in pistis himself. He came to the Corinthians ‘in weakness and fear and much trembling’ (1 Cor. 2.3) despite his spiritual power (v. 4), so he recognizes that fear and unspecified imperfections can alternate, if not coexist, with pistis, but again, we cannot be sure in what sense of pistis.

Unlike the synoptic gospels, Paul keeps the language of apistia for non-community members and does not use the language of oligopistia at all,Footnote ² but in a few passages (e.g. 1 Cor. 1.11–12, Gal. 1.6–9), as noted, he does criticize community members sharply for listening to different apostles. These passages are interesting because, though they do not feature pistis language, it seems clear that the Corinthians and Galatians are being misled into doubt or even (if they have already forsaken Paul) scepticism about Paul's preaching.

The language of disturbance, misrepresentation, and anathema is strong and unambiguous: Paul is not only vexed that the Galatians are listening to other apostles; he is certain that what they are hearing is wrong. A few verses later (1.23), Paul reports other Christians as saying of him, ‘The one who was persecuting us is now preaching the pistis he once tried to destroy.’ I have argued (Morgan (Reference Morgan2015), 265–267) that pistis here must mean ‘the relationship of trust between God, Christ and human beings’ rather than the dogmatic content of the gospel. But even if what is at issue is a right or wrong relationship (rather than, for instance, a right or wrong view of the nature of Christ), and leaving aside how the Galatians were misled (which we do not know), Paul's claim that one can distinguish these relationships as right or wrong suggests that there is a cognitive and propositional aspect to choosing a particular pistis, and only those who make the right choice are adhering to the gospel.

Must those who choose the right pistis, though, have an attitude of belief towards it rather than, for instance, one of hope? As I noted above in relation to 1 Cor. 15, it will be hard for most biblical scholars to imagine that Paul does not himself have an attitude of belief towards his gospel and the pistis it preaches, nor to assume that he does not demand the same of the Galatians. But here as in 1 Corinthians, Paul himself does not make the distinction and his language is not fine grained enough to reveal one, so we cannot be sure.

These examples are necessarily brief, and they point to the need to do more work, but I am grateful to Professors McKaughan and Howard-Snyder for prompting me think further about these passages and these issues.

Professor Howard-Snyder's third question concerns Saint Teresa of Calcutta, and the nature of her faith.

I will raise and set aside here an issue which is important but which this discussion does not tackle. Belief in a narrative is not the same thing as belief in propositions, and it is the latter that is the focus of these articles. When Professor Howard-Snyder refers to the ‘basic Christian story’, I therefore take him for present purposes to be talking about Saint Teresa's belief, for instance, that Christ rose from the dead.

The answer to his question in very early churches would surely depend on which writer one referred to, and how Teresa's story was told and heard. If, for instance, Teresa were writing to Paul about her doubts, then on the basis of the argument above, he might chastise her under option (i) as lacking faith (even though she does not report switching her adherence to a different gospel). If a gospel writer were writing her story as that of a follower of Christ, then she might appear under option (ii), as someone who had faith, but not perfect faith, or even under option (iii), as someone whose life became a triumphant exemplar of faith despite her fears and doubts.

That generosity and tolerance of complexity and human imperfection which narrative allows is one of the reasons why people of faith continue to find the gospels inspiring and enriching. It prompts a further question: what pastoral implications there might be for the practice of faith by the Church in the modern world, if it were to reclaim the ideas and practices of faith that dominated early churches. That, and other questions of contemporary significance, however, is beyond the scope of this discussion.Footnote ³