3.1 Not “Trust Itself” but the Nature of Trusting

Eyal et al. accuse prior trust conceptualizations of wanting to isolate “trust itself” and effectively take it away from the actors. Trust scripts and the leap of faith are seen as ruling out that it is essentially the actors who accomplish trusting. This conclusion does not hold, because the research they cite is not reductionist in the way Eyal et al. claim. Models of trustworthiness à la Mayer et al. (1995), for example, and the survey items derived from them, do not deny the situatedness of trusting in practice. Leap-of-faith accounts do not reduce trusting to making leaps of faith, but include them as an important element of a comprehensive model (Möllering, 2006). Note that trustworthiness is treated as perceived by individual actors and leaps of faith are taken by individual actors and not happening to them. No one would think of trying to separate trust(ing) from the actors. Even Simmel’s (1990) “surrender of the ego” (p. 179) and Abraham’s sacrifice (see Möllering, 2006, pp. 117–118) are expressions of agency — “to believe in someone” — and not of resignation. Hence, there is no search for “trust itself”, no idea of a medium detached from actors, and no assumption of a kind of ether.

However, yes, many scholars including myself have been looking to identify the “essence” or “nature” of trust(ing). Eyal et al. do this as well and propose that trusting is essentially skillful situated practice. This is a hollow insight, though. We could come to the same conclusion for infinite other concepts: love, hate, friendship, happiness, violence, peace, power and so on. All of them can be framed as situated practices along the lines of a Bourdieusian logic of practice and it would be insightful to do so. But what are they essentially about?

Eyal et al. make no effort in their essay to consider what trusting is about. When they say that trust is about being seen as responsible, that is a derived issue, not the core issue. Responsibility needs a purpose. The same holds for the problem of navigating between mistrust and blind faith. There has to be something behind that problem, making it relevant. So what is trusting as a skill or method for? What is the problem trust solves? In contrast to Eyal et al., the work they criticize and many other authors do look for answers to these questions. They conceptualize trust(ing) as a way of dealing with irreducible uncertainty and vulnerability in social relations (e.g. Baier, 1986). This is the core issue which makes trust “trust” and which models of trust(ing) try to capture.

That is, when Mayer et al. (1995) synthesize ability, benevolence and integrity as three dimensions of perceived trustworthiness, they offer a characterization of what matters specifically for trusting as distinct from other behaviors. And when I end up identifying “suspension” (which can be imagined as leaps of faith or other metaphors) as the essence of trusting, I point first and foremost to the problem of radical uncertainty that trust relates to. There is no “trust itself”, but there are distinct concepts of “trusting”. Hence, we need to investigate the nature of trusting in terms of both what and how.

3.2 Not “Scholastic Fallacy” but Making Sense of Trusting

Eyal et al.’s portrayal of other work on trust as falling for a scholastic fallacy is unduly polemic and ultimately backfires anyhow when they take the leisure of constructing their own model of trusting and their own interpretations of their survey respondents’ trust methods. I propose that we do not dismiss each other’s attempts at making sense of trusting. The only fallacy we should worry about is that of claiming exclusive and conclusive truths about trust(ing).

This means that instead of trying to debunk trust-as-attitude and trust-as-leap-of-faith research, Eyal et al. could have discussed which part of trusting as practice these works might be able to capture. For example, work on attitudes or personality could capture predispositions as elements of situations in which trusting happens (e.g. Patent & Searle, 2019). And work on leaps of faith could help to capture those most agentic moments when people do go ahead, which Eyal et al. already integrate to some extent. Put differently, trust research at large is a rich resource for making sense of what the skillful practice of trusting entails.

For example, I have been very critical of rational choice accounts of trust as introduced by prominent scholars such as James Coleman and Russell Hardin (Coleman, 1990; Hardin, 2002). However, in the spirit of sensemaking, I would not say they have been wrong, but their conceptualizations have been simplistic and incomplete. I would accept that the interests of trustors and trustees, i.e. how they see their potential gains and losses (Coleman’s Bayesian calculation) and the way each other’s interests are structured (Hardin’s “encapsulated interest”) are part of the story, more or less, but not the whole story. I have also been very critical of Oliver Williamson’s (1993) attempt to sideline “trust” by claiming that “calculativeness” is a more appropriate concept for what Coleman and Hardin call trust. Yet, Williamson does have a point when he challenges us to see trusting as something that is more than calculation. Eyal et al.’s situated practice view, I think, equally allows for some form of “weighing the odds” by the actors, but would never reduce trusting to just that.

I can think of numerous authors, not cited by Eyal et al., who have offered eye-opening accounts of what trusting is about, what is special about it and how it works (e.g. Barber, 1983; Barbalet, 2009; Dibben, 2000; Frederiksen, 2014; Kroeger, 2017; Mizrachi et al., 2007; Nooteboom, 2002; Uslaner, 2002). They all present ideas that underpin, rather than undermine, Eyal et al.’s approach. For reasons of space, I will highlight two sources here that appear to be particularly relevant for trusting as practice:

First, Lucien Karpik (2014) counters Williamson (1993) and frames trust as an operation of “judgement”, whereby actors evaluate singular issues by constructing equivalences for idiosyncratic comparisons instead of following a generalized calculation They create “judgement devices” in the face of “singularities”. Strikingly similar to Eyal et al., Karpik points out real-life multiplicity and particularism. Trust as judgement closely resembles trusting as skilled situated practice.

Second, not least, because he also challenges other leap-of-faith accounts to some extent, I would like to recommend Adam Seligman’s Dewey-inspired framing of “Trust, Experience and Embodied Knowledge” (Seligman, 2021). For Seligman, unlike confidence, trusting means to be able to deal with ambiguity by relying on experience rather than abstract knowledge. Seligman’s perspective on trust, already contained in his much earlier writings, too, is concerned with how people actually trust, similar to Eyal et al., but he includes a clear analysis of what trusting is about, which is even indicated by the title of his very influential book The Problem of Trust (Seligman, 1997).

Situatedness of practice, then, is not just a general fact of all social life. Authors like Seligman or Karpik would contrast forms of interaction that are fairly standardized and can be handled with “confidence” from those interactions that require idiosyncratic “trust” in the face of social uncertainty and vulnerability. Both Seligman and Karpik see actors as capable of accomplishing this. Trusting, one might say, is a familiar way of dealing with the unfamiliar, not following a path but finding a way.

Overall, it does not help to label scholarly attempts at conceptualizing trust as “scholastic fallacy”. Perhaps Eyal et al. will agree to seeing research, their own included, as situated practice, too. If so, then we need not evaluate if a particular perspective on trust is right or wrong, but rather what it adds to making sense of trusting as a practice that can take many concrete forms but always relates to uncertainty and vulnerability as generalizable challenges of social relations.

3.3 Not “Trust Scripts” but Active Trusting

The previous section already indicates that when Eyal et al. position their notion of situated practice in opposition to “pregiven scripts” (pp. 181–182) they draw a misleading caricature of the state of trust research, given that Karpik, Seligman and others have in fact made the same point a long time ago. More importantly, ruling out scripts, Eyal et al. actually make it more difficult for themselves to maintain a meaningful practice perspective whereby trusting is at the same time idiosyncratic and embedded. Similarly, a temporal analysis of trusting fits well with prior research on open-ended trust dynamics that also does not envisage inevitable executions of scripts, but recognizes agency as well as emergence (e.g. Klein Woolthuis et al., 2005; Wright & Ehnert, 2010). Once again, Eyal et al. could strengthen their approach by treating script-like accounts of trusting as partial insights into what is supposed to be going on and where creative or even deviant agency occurs. When they reference Garfinkel, they actually do this to some extent already.

As a final conciliatory move, I therefore suggest that Eyal et al.’s approach might resonate very well with Anthony Giddens’s notion of “active trust” (Giddens, 1994). Considering the implications of the nature of trusting as maintaining cooperative social relations in the face of uncertainty and vulnerability, many authors do see trustors as active, not passive. Incidentally, Zucker’s (1986) theory of trust production, like Garfinkel (1963), emphasizes shared expectations and notes a historical trend towards more impersonal bases for trust. However, the rise of “institutional-based trust” can be seen already as a creative response to the practical difficulties of taking social relations for grated in modern societies. The institutionalized frames that Eyal et al. also refer to are needed but at the same time not necessarily as stable and reliable as one would hope. Giddens (e.g. 1990, 1991, 1994), as part of his larger project of theorizing late modern societies, refers to “active trust” in order to highlight the very fact that shared expectations — of mutual commitment and care in trustful relationships — have to be constantly worked upon by the actors involved (see also Beckert, 2002). This emphasizes “trust” as verb: what people do to maintain “trust” as an attribute of their relationship or even as a principle of interaction in groups, organizations and societies.

Active trusting, as we might call it then, can be studied from the individuals’ point of view. In fact, Eyal et al.’s analysis of Long COVID patients is strikingly individualistic. In this, it matches what I have referred to elsewhere as a “trusting as becoming” perspective (Möllering, 2013). Trusting shapes, and is shaped by, social identity. Trusting is an ongoing project of the self (e.g. Giddens, 1991), which is why, indeed, Eyal et al.’s focus on responsibility and accountability is highly meaningful. Whom and how actors trust reflects who they are and who they want to be. By implication, trusting as a skillful practice is not just about how trustors (and trustees) “read” the situation, as it were, but also how they “create” the situation and the relational dynamics for themselves and others (e.g. Henslin, 1968). Social uncertainty and vulnerability, we need to recall, are not just background conditions to this, but the actual issue addressed by trusting.

Active trusting, as a practice, also needs to be studied beyond individuals or, put differently, while trusting can never be detached from actors, trusting draws on context(s), as Eyal et al. would probably agree. Perhaps it is useful to imagine that trustors “activate” context when they accomplish trusting and their trusting actually reflects back into context. Now, this means that research about contexts of trusting is actually very valuable, too. Again, I would recommend being open and lenient to research that, at first, appears to be far away from a practice perspective. For example, measurements of “trust” in “most people” or in certain institutions might be able to capture significant contextual elements, especially in comparative applications (e.g. Uslaner, 2015). As a small illustration, Scandinavians are often very mindful that, according to surveys, they live in high-trust societies and they use this to make sense of their own and others’ behavior.

Moreover, it is valuable to study social systems not primarily in terms of whether they can be trusted as such, but in terms of how they support practices of trusting (e.g. Sydow, 1998). How much space and which resources to do they actually give to actors so that they can be responsible trustors? Finally, system features, like Eyal et al.’s institutionalized frames, do not equal pregiven scripts, because trusting in action is situated and because, in return, trusting in action shapes the system reflexively, as I would argue from a simplified structuration theoretical perspective (Möllering, 2013).

In sum, if research theorizes trust as scripted, it cannot be the whole story. Empirically, any script-like elements are actually quite interesting as evolving tools for trusting. However, trusting remains an active, not passive practice, also in relation to the definition and application of any scripts.