If an ancient understanding of faith, or of the response that early Christians took God to desire of humans, turns out to differ in significant ways from influential medieval and/or modern accounts of ‘faith’ and ‘belief’, what implications might this have for contemporary discussions about the nature, value, and rationality of faith?

Teresa Morgan is to be commended for her magnificent book, Roman Faith and Christian Faith (Reference Morgan2015) – a fascinating examination of pistis/fides discourse and associated practices encountered in ancient Graeco-Roman, Jewish, and early Christian texts. Professor Morgan approaches the study of faith with a truly admirable sensitivity to language and historical context. Her command of classical texts and culture and of current scholarship is evident throughout this thoroughly researched book and also in her judicious appraisals of the relevant secondary literature. Her study helps us to understand what it was that people in the Graeco-Roman world of the early principate so valued about pistis/fides in all manner of ordinary relationships, why pistis/fides was once widely thought to be a crucial part of the fabric of social relations, and why some ancient Hellenized Jewish and early Christian authors appropriated this language to describe the central response that they took God to desire of humans. Among other things, she carefully examines the role that the pistis lexicon plays in every book of the Greek New Testament, explaining its meaning in key passages, calling attention to the distinctive voices of various authors, and spotting subtle developments in patterns of faith-related discourse. She considers the significance of the fact that, whereas Paul uses the noun πίστις (pistis) about twice as often as the verb πιστεύω (pisteuō), the Johannine literature almost exclusively favours the verb. She attends to whether people are called to place faith directly in God or where faith is described as mediated in some more complex way by Jesus. And she presents a highly original and convincing case for an evolution in Paul's own way of conceptualizing faith between his early and late letters. Professor Morgan's book is now one of my favourite resources on faith and I would like to take this opportunity to draw further on her expertise and to invite her into some interdisciplinary conversation. The main questions that I want to raise, and also a further explanation of what I find philosophically exciting about her findings, will require a bit of stage setting.

By the first century, in the Graeco-Roman Mediterranean world, people thought of faith as a social bond that strengthens relationships and enables forms of mutually beneficial social cooperation. Although the lexicographical range of the virtually synonymous Greek pistis and Latin fides is such as to leave translators often with a range of options, their core meanings are picked out by ‘faith and faithfulness’ or interpersonal trust and loyalty – by ‘trust and trustworthiness—which stand equally at the heart of the meaning and operation of fides’ (Morgan (Reference Morgan2015), 5; see also: Campbell (Reference Campbell2005), ch. 9; Dunn (Reference Dunn and Sakenfeld2007) ).Footnote ¹ Morgan argues convincingly that trust is the centre of gravity for pistis/fides (15), while emphasizing that we should be alert to hearing both aspects of the faith(fulness) register: ‘While English speakers distinguish between “faith” and “faithfulness”, Greek speakers may not have done so’ (13). I anticipate – and I wonder if Professor Morgan would agree with me here – that a comparable study of faith focused more directly on the Hebrew Bible, its stories, and the culture of ancient Israel would display an even stronger emphasis on the fidelity/faithfulness/loyalty/allegiance side of the register.Footnote ²

Morgan presents evidence of pistis/fides playing a role in relationships ‘at every socio-economic level, between individuals and groups’ (6) – ‘relationships of wives and husbands, parents and children, master and slaves, patrons and clients, subjects and rulers, armies and commanders, friends, allies, fellow-human beings, gods and worshippers, and even fellow-animals’ (117–118). For example, the Romans promoted mutual good faith and loyalty in their propaganda and on coins that circulated in Roman territories from circa 275 bce to 270 ce. On a silver Roman denarius from c. 73–74 ce faith is symbolized by clasped right hands, which requires the setting down of swords, to invite common trust in and reliance on the emperor to provide goods of peace, some basic political protection (dependent on social standing), and greater prosperity symbolized with grain ears and poppies. Other coins paired phrases such as ‘FIDES PVBLICA’ with a depiction of Fides, goddess of trustworthiness and good faith, to spread a similar message. These practices, though of course often falling short of the ideals and propaganda in all sorts of ways, were very effective in establishing valued social relations that in a clear sense depended on mutual fides (trust and loyalty) to the benefit of conquered peoples and in directing loyalties towards Rome.

The picture of faith that emerges from Morgan's study is a relationship- and action-focused understanding of faith, from which cognitive and affective dimensions need not be absent but are not the focus. Pistis/fides ‘is, first and foremost, neither a body of beliefs nor a function of the heart or mind, but a relationship which creates community’ (14). Her careful engagement with the classical sources leads her ‘to downplay the interiority of pistis/fides, which becomes important to later Christians but which . . . does not attract much interest in the first century’ (503). As she observes, ‘propositional belief (secular or religious) is usually marked, in Greek and Latin, by the language of thinking (dokein, nomizein, putare, censere, etc.) rather than that of pistis or fides’ (30). In contrast to many medieval and modern treatments of faith, in the earlier documents that she is examining ‘Graeco-Roman, Jewish, and Christian sources alike take frustratingly little interest in exploring the nature and internal’ aspects of pistis/fides as an emotion or cognitive state (472). Instead, ‘Pistis and fides are fundamentally relational concepts and practices, centring on trust, trustworthiness, faithfulness, and good faith’ (503).

The lack of interest that ancient authors display in describing the interior aspects of pistis, both cognitive and affective, is of interest for recent philosophical literature exploring non-doxastic faith and whether there is room for such faith within the Judaeo-Christian tradition.Footnote ³ Later philosophical discussions of faith have tended to emphasize the cognitive aspect of faith and, in particular, propositional belief.Footnote ⁴ Proponents of what I call the Belief Plus view of faith insist that, in order for a person to have Christian faith at all, one must find themselves with a particular attitude towards the content of the kerygma: only believing will do.Footnote ⁵ In the period that Morgan is examining, however, ‘even in Christian thinking, propositionality is often less important than has often been assumed’ (30). As she notes, ‘the propositional content of pistis, though always implicit in its relationality and not infrequently articulated or alluded to, is not usually its main focus in the New Testament’ (444).

The complex story of how the Christian tradition transitions from a relational focus on what we might, at the risk of oversimplification, call ancient faith (in which acts of trust and fidelity to a person are central) to a medieval and modern faith (in which something propositional and interior, such as believing a set of propositions, is prominent) would require a second book. But I would be grateful for any further guidance or signposts that Professor Morgan might offer us here. Perhaps she can help me to understand, for example, why credere, rather than fidere, was selected in Jerome's Vulgate as the Latin translation for pisteuein?Footnote ⁶ I also wonder whether Professor Morgan would agree that part of the story about this transition should take into account the fact that belief talk has also undergone a complex historical evolution.

Although philosophers still don't all agree about how precisely to define belief (McKaughan (Reference McKaughan2007); McKaughan & Elliott (Reference McKaughan and Elliott2015) ), we might say that for you to believe that p is, roughly, for you to have a tendency for p to seem to be true when you consider it. Wilfred Cantwell Smith (Reference Smith1998a; Reference Smith1998b) argues that over the centuries Indo-European words for belief saw a shift in meaning. Here is one subtle difference. Nowadays, we tend to answer questions about whether someone believes a proposition by offering descriptive factual reports about people's psychological states (e.g. one's high level of confidence that p.)Footnote ⁷ For centuries Indo European belief talk was most often used as a performative act of avowal, commitment, or self-engagement in the context of interpersonal relations. The Latin verb credo (credere, ‘to believe’) – apparently a compound of cor, cordis ‘heart’ and -do, -dere, ‘to put’ – derives from the Proto-Indo-European *ḱred d^heh, and means literally ‘I give my heart, I set my heart on’, someone or something, or ‘I put [it] in my heart’.Footnote ⁸ It is perhaps especially understandable why first- and second-person declarations of allegiance, often taking grammatical forms that more closely track our notion of ‘believing in’, would be recruited to express commitment to God or to the cause of Christ and to shared teachings in connection with a relationally focused ancient understanding of pistis/fides in which acts of trust and fidelity to a person are central. Such a choice might also naturally coincide with increasing interest in interior aspects of faith. But the rich set of affective and value-laden connotations once prominent in terms for ‘believing’ eventually gave way to an understanding of believing as a dispassionate (auto-)biographical report about a person's opinion concerning the epistemic status of a proposition.Footnote ⁹

I have elsewhere advocated a view of relational faith which characterizes Judaeo-Christian faith as an active commitment to engaging in a relationship with God and/or with Jesus – a relationship centred on acts of (1) trust in or reliance on God and/or Jesus and (2) following Jesus and/or walking in God's ways (McKaughan (Reference McKaughan2016); McKaughan (Reference McKaughan2017) ). This seems to be very much in line with the picture of New Testament pistis that emerges in Morgan's discussion. The core of the human response of faith that the New Testament authors care about, what they are using pistis language centrally to describe and what they take to be worth having, is actively engaging in a relationship with God and/or with Jesus that perseveres.

There is a further question about what range of accompanying interior states a person would need for sensible engagement in such a response. While I locate what is central to faith in the realm of action, I also take paradigm cases of Christian faith to involve characteristically some sort of positive cognitive attitude and positive affective-evaluative attitude towards the person or content that is the object of one's faith. I am sympathetic to Dan Howard-Snyder's proposal that faith involves perseverance, or what he calls ‘resilience’, in the face of challenges, with respect to each of these three aspects (Howard-Snyder (Reference Howard-Snyder2013); Howard-Snyder (Reference Howard-Snyder2017); McKaughan (Reference McKaughan2016); McKaughan (Reference McKaughan2017) ).

Morgan does not, of course, deny that pistis/fides often, or even characteristically, involves cognitive and affective dimensions. Paul and other authors of the New Testament believed, for example, that Jesus was raised and that God was reconciling the world to himself in Christ, regarded these as vitally important truths, and took themselves to have good grounds for faith in this message. No doubt they also desired that others believe their proclamations (227). But is there any reason to think one couldn't respond in the way that they so value and call for if one's psychological attitude towards the content of the proclamation is something other than believing?

Even if believing is not itself part of the core of the pistis response, perhaps it is nevertheless required. Motivation for the Belief Plus view might stem, first, from the thought that the Judaeo-Christian tradition is constrained by the idea that believing is the distinctively acceptable cognitive accompaniment of faith or, second, from general considerations about the nature of faith or the relation between trust and belief.

Is there good reason to think that the authors of the Greek New Testament would have insisted, or that we should insist upon reflection, uniquely on believing as a necessary condition for having faith at all? Keep in mind, first, that these authors are also using pistis to describe the beginning of a relationship with God (e.g. an ongoing commitment inaugurated at the moment of conversion) (14, 204, 216, 224, 296, 298, 304, 336, 339, 382, 473) and having pistis also functions as an identity marker and as a criterion for community membership: hoi pisteuontes/pistoi (‘the faithful’, ‘those who trust in God/are faithful to God’) are employed as designations for followers of Christ (Christians) (234–241, 308). Second, as we have already seen, in using pisteuein they display little interest in exploring fine-grained distinctions between believing as we understand it today and a variety of mental states or acts – such as believing that p is more likely than not, believing that p is more likely any of the alternatives that one takes seriously, or trusting, accepting (voluntarily endorsing), championing, belieflessly assuming, hoping that p, or even the idea of setting one's heart on the gospel glossed above – that might serve as candidates for the positive cognitive response accompanying faith.

However, once we set out a more fine-grained array of positive cognitive attitudes towards the content of the kērugma that might be available, how would one establish that the New Testament authors intended to exclude non-doxastic faith accompanied by the sorts of positive cognitive responses mentioned above from the semantic range of πίστις (pistis) and πιστεύω (pisteuō) and that only believing will do?Footnote ¹⁰ Such an argument would depend, among other things, on making a clear distinction between believing (in the modern sense) and other relevant positive cognitive attitudes in the Greek from this period. Even this much risks anachronism. As Morgan notes:

This comment about the lack of attention to describing interior cognitive and affective aspects of pistis in the Septuagint also applies to ancient Graeco-Roman writings and to the texts of the New Testament (see chapter 11). Given the looseness, flexibility, and semantic range of Greek (and Hebrew) faith talk, and what believing has come to mean today, it will be hard to insist on ‘believe’ as a translation in contrast to or to the exclusion of other legitimate alternatives, including ‘to have faith’, ‘to trust’, ‘to rely on’, ‘to commit’, ‘to put or come to faith’, ‘to be faithful/loyal/obedient’, ‘to entrust’, ‘to give allegiance to’, and so on. Morgan's discussion will provide proponents of the Belief Plus view with very little comfort on this score. As she notes: ‘Wherever possible, I translate pisteuein with “trust/believe”; in places I use pisteuein as a placeholder for various forms of the English verbs’ (396).

But what about the idea that trust requires belief? On this point, proponents of the Belief Plus view may find Morgan more sympathetic. In some places, Morgan seems at least tacitly to endorse something like the following claim:

Although we do ordinarily have such beliefs when we trust someone, and often ‘defer’ trust to such grounds for trust in the ways that Morgan suggests, the ‘must’ in this formulation strikes me as too strong. In Homer's Odyssey, after the decade-long Trojan War, it takes Odysseus another ten years to journey home. Suppose that Penelope comes to have significant and reasonable doubts about whether he is still alive, such that she no longer believes that he is alive, but also does not believe that he is dead. Could Penelope persevere in faith? Clearly she could continue to rely on Odysseus, to give him her allegiance, and to honour her ongoing commitment to him. These responses take the form of actions that remain available to her even given significant doubt, risk, and uncertainty of a sort that precludes believing that Odysseus is alive.

Trust is, like belief, a more disputed notion and nowadays we often emphasize the interior psychological states involved. But, as we have seen, that is a modern rather than ancient preoccupation. There are forms of action-oriented trust available to Penelope which do not require belief (see Swinburne (Reference Swinburne2005), 143; McKaughan (Reference McKaughan2013); McKaughan (Reference McKaughan2016) ). Penelope could, for example, (1) act on the assumption that Odysseus is alive (and will remain faithful, acting in ways that are worthy of her trust and doing for her what he knows that she wants or needs), (2) when the evidence gives her some reason for supposing that Odysseus may not be alive (or may not be faithful), (3) where there will be bad consequences (and/or she will feel betrayed) if her assumption is false, and where (4) her commitment to such actions is contingent on Odysseus’ existence in such a way that learning that Odysseus is not alive would cause her to lose or abandon her faith.

There is room for disagreement here about how we should understand trust and how it relates to belief. But notice this: here what is doing the work in support of the Belief Plus view is an independent principle that we might or might not bring to our reading of the ancient sources, rather than the thought that certain New Testament texts require it. Moreover, it is an assumption that Morgan, or someone in almost all other respects sympathetic to her project, might readily give up as a friendly amendment that she could accept with very little revision.Footnote ¹²

Professor Morgan's book offers an important and substantial answer to her own central question – why was faith so important to early Christian communities? – and will no doubt continue to contribute to other discussions about faith for years to come.Footnote ¹³