This dense, difficult new book sees Raimo Tuomela expand his pioneering ‘we-mode approach’ to the philosophy of sociality. At its heart is Tuomela’s feted distinction between two ways of operating as an intentional subject: there is the I-mode, in which one thinks and acts ‘as a private person’ (2), and the we-mode, in which one thinks and acts ‘as a member of a group’ (2). Here, as in previous work, Tuomela seeks to delineate and support the distinction, and draw out its implications for various facets of the social world.
The book contains a healthy mix of consolidation and extension of Tuomela’s views, and so will prove rewarding both for those familiar with his previous work and those not. The opening chapter presents an instructive basic description of the I-mode/we-mode distinction, which gets fleshed out via detailed technical analyses in the two succeeding chapters. Apart from their discussion of group agency these first three chapters do not mark any significant change of mind, focus or direction from Tuomela’s previous work, and the same can be said of Chapter 5, which recapitulates Tuomela’s ‘bulletin board view’ of group attitude formation (cf. Tuomela 2007). By contrast, the discussion of collective reasoning in Chapters 4 and 7 provides crucial new insight into this core mechanism of we-mode activity. Also new and important is Tuomela’s effort (in Chapters 6 and 8) to accommodate ‘non-autonomous’ groups within the we-mode framework, by relaxing paradigmatic we-mode requirements such as voluntary entrance/exit and collective acceptance of a group ethos. Chapter 9, on solidarity, elucidates the ‘togetherness’ (or ‘we-ness’) that prevails in those groups which are paradigmatically we-mode.
Tuomela claims that the book’s ‘new insights and improvements have to do with a stronger emphasis on group agency and group reasoning’ (14). As I read him, these topics form two sides – one ontological, one conceptual – of the same coin, and in what follows I focus on them solely.
Right up front Tuomela declares that ‘the social world cannot be adequately explained or rationally understood without postulating groups as intentional agents’ (2). This is because some we-mode activity is necessary to explain the social world, and all such activity conceptually presupposes group agency. If we must draw our ontology from the posits of our best-explaining theories, this appears to straightforwardly imply a form of realism about group agents.
But the matter is not quite so simple, for Tuomela is also reticent to count groups as rational agents in the same sense, or to the same degree, as individual humans. Drawing on what he calls ‘our ordinary framework of agency’ (21) – which includes emotions and qualia ‘with their bodily accompaniments’ (21), and according to which individuals’ mental states count as intrinsically intentional ‘in virtue of their biological nature’ (23) – he insists that ‘groups can never be full-blown agents […] in the flesh-and-blood sense, but at best entities that share some functional features with intentional human agents’ (23).
Since Tuomela is emphatically neither a functionalist nor an instrumentalist about group agency, and he describes the mental properties of groups as ‘fictitious’, one might wonder in just what sense he thinks groups really do count as agents. Here we need to read his claim that there is a ‘conceptual’ connection between we-mode activity and group agency with care. What he means, I take it, is that for an individual to engage in some we-mode activity she must have a concept of a group agent, and must conceive of herself and certain others as members of it. She then adopts, and takes those others to likewise adopt, the rational perspective of the group agent she has postulated.
To this Tuomela adds the familiar claim that mental acts of postulation like this are sufficient to ‘construct’ a group agent. Just as (in Tuomela’s example) squirrel pelts will count as money when they are collectively accepted as such in the relevant money-using community, so too, when a set of individuals base their thought and action on what they take to be the rational perspective of the group, they ensure that the group really has such a perspective and can act on it. In this way their we-mode activity both presupposes and ‘entifies’ the group as an agent.
But is it coherent to develop such an account of group agency against the backdrop of the ‘ordinary framework of agency’? Tuomela appears to take this framework as a non-negotiable dictate of ‘common sense’, even though it is not obviously the dominant philosophical analysis of agency, and adopting it eschews the most direct approach to group agency, which involves conceptually disentangling the phenomenon of agency from the way it gets manifested in the normal case of individual adult humans. A group might count as a genuine agent if being just such an agent were a matter of, say, being functionally characterisable in certain ways (List and Pettit 2011), or of being usefully and voluminously predictable from the intentional stance (Tollefsen 2002), or of forging a rational point of view (Rovane 1998). But if agency is what the ordinary framework says it is – a particular kind of biological achievement – then the prospects for recognising flesh-and-blood-less groups as agents are slim to none. It is not clear that introducing a completely different account of agency, based on collective acceptance, can help to remedy this problem.
So much for the ontological aspect of group agency. In much of the book Tuomela is occupied instead with the conceptual aspect, that is, with the way in which the concept of group agency features in we-mode activity. This is most evident in Chapter 7, where Tuomela appropriates Michael Bacharach’s formal ‘team reasoning’ framework to support three key claims: (1) that the we-mode is irreducible to the I-mode; (2) that the we-mode is indispensable to any adequate understanding of the social world; and (3) that in many situations individuals ought to operate in the we-mode.
The irreducibility claim is supported by a Bacharach-inspired demonstration that ‘we-mode we-reasoning’ can generate different action recommendations from I-mode reasoning, even when the latter is done for the benefit of the group (when the individual is a ‘team benefactor’ in Bacharach’s parlance or a ‘Pro-Group I-moder’ in Tuomela’s). This is illustrated with reference to the Hi-Lo game:
|Hi||3, 3||0, 0|
|Lo||0, 0||1, 1|
Traditional decision theory sees the players in this scenario as addressing the question, ‘What should I do?’ by first making a prediction about what the other will do, then figuring out the best reply to that. As Tuomela indicates, even were both players to privately intend to maximize group utility, this way of reasoning leads to an impasse in the Hi-Lo game. For, if you play Lo, then, in order to maximize group utility, I should play Lo. But, if you play Hi, then, in order to maximize group utility, I should play Hi. And I can have no reason to expect you to play Hi rather than Lo (or vice versa) since I can only assume that you are predicting my action in the same way that I am yours. Traditional decision theory is thus unable to vindicate what appears to be the rational choice – Hi.
Bacharach sought to overcome this problem by introducing the idea that players might be ‘framing’ these scenarios in a non-individualistic way, asking ‘What should we do?’, where ‘we’ is not meant in the distributive sense but in the sense of what Tuomela calls ‘togetherness-we’. In the Hi-Lo game this allows the parties to transcend the strategic ‘best reply’ approach and simply select the joint outcome which maximizes group utility, viz., HiHi. And once each selects HiHi as her preferred joint outcome, it will be clear to each that she should do her part – that is, play Hi – in order to bring that outcome about. In this way, we-mode reasoning generates different action recommendations from I-mode reasoning, and this ‘functional difference’ shows that, at least as far as reasoning goes, the we-mode is irreducible to the I-mode.
What then of the indispensability claim, that we-mode reasoning is necessary to adequately understand and explain the social world? This need not detain us, since it is at least partly an empirical claim and Tuomela is careful here not to over-reach in attempting to support it. (He does mention experimental work on group reasoning which is ‘compatible’ with his views, but, of course, this is a far cry from what would be needed to support the indispensability claim.)
It is the normative claim which demands further scrutiny. According to Tuomela, we-mode reasoning ‘tends to create more collective order’, and this can form the basis of a ‘rationality-based argument […that] at least in some specifiable situations agents should reason and act in we-mode rather than the in the I-mode’ (193–194).
Here Tuomela parts company with Bacharach. While Bacharach insisted it is only a brute fact that certain decision problems prompt a shift from the I- to the we-frame, Tuomela holds that mode selection can be a matter of rational choice (195). But this, I think, generates a problem, for, if questions of what is rational are always agent-relative, then agency transformation cannot be a process governed by rationality. Team reasoning allows us to make sense of a choice being rational from within the I-frame, or from within the we-frame, but we cannot then see the shift from an I- to a we-frame as itself being a matter of rational choice, unless we imagine some meta-frame encompassing both. But Tuomela would clearly not accept that the I- and the we-mode are options available within some larger rational perspective, and yet holds that the we-mode is sometimes rationally preferable to the I-mode. This just invites the question: rational for whom, from which perspective?
This question might of course be answered. Perhaps, just as a different sense of agency (from the ‘flesh-and-blood sense’) was needed in order to make sense of the agency of groups, so too a different sense of ‘rational’ (from agent-relative rationality) must be invoked in order to make sense of the rational advantages of the we-mode over the I-mode. That is, perhaps adoption of the we-mode is rational (or ‘optimal’ or ‘beneficial’, terms Tuomela sometimes prefers in this context) in some agent-neutral sense – but then it would be helpful if that sense was more fully articulated.
Though I have focused here only on two of its central elements, it bears repeating that Tuomela’s approach to understanding the social world is a vast, many-splendored project. Besides group agency and group reasoning, this new book discusses collective intentions, group attitudes, co-operation, authority, social institutions and solidarity, all through the lens of the I-mode/we-mode distinction. Remarkably, this grand, synoptic vision is not painted with broad brushstrokes but with immensely fine-grained exactitude and analytical rigor, each chapter teeming with detailed technical analyses, exhaustive taxonomies, precisely stipulated definitions and formalized arguments. It is undoubtedly demanding, but for anyone with an interest in the themes designated in the book’s title – social ontology, collective intentionality, and group agency – the effort is surely worth making.
Thanks to Katharine Browne, Jennifer Hornsby, Olav Gjelsvik and Arto Laitinen for helpful comments and conversations.
List, Christian and Philip Pettit (2011): Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press.
Rovane, Carol (1998): The Bounds of Agency: An Essay in Revisionary Metaphysics. Princeton: Princeton University Press.
Tollefsen, Deborah (2002): “Organizations as True Believers”. In: Journal of Social Philosophy 33. No. 3, p. 395–410. [Web of Science]
Tuomela, Raimo (2007): The Philosophy of Sociality: the Shared Point of View. Oxford: Oxford University Press.