In the contemporary literature on collective mental states, it is often argued that such states present a distinctive intentional structure, irreducible to individual forms of intentionality. This view has been mostly defended in accounting for the distinctiveness of collective actions (Searle 1990). For most authors, such actions are sustained by a specific type of agentive intention, namely “we-intentions”, where the “we” is classically ascribed either to their content (Bratman 1999), mode (Tuomela 2007) or subject (Gilbert 1989). More recently, it has been argued that collective emotions possess a “we-mode” intentionality (Salmela 2012). Similarly, it is often argued that joint attention and mutual recognition should be considered as collectively intentional (Gilbert 2011).

In addition, shared mental states are often described as involving a “sense of us” or as being experienced “as ours”, hence possessing a distinctive phenomenal property. This is especially salient in phenomenological descriptions of collective emotions. For instance, Thonhauser (2018) describe collective emotions as involving a “sense of togetherness”, that Léon et al. (2019) take to be what distinguish them from other forms of affective alignment such as emotional contagion. Again, this type of phenomenological description is not restricted to collective emotions. For instance, Pacherie (2014) defend that collective action imply a feeling of “we-agency”, and Campbell (2005) describe the transition from solitary to joint attention as involving an “experiential shift”.

The fact that shared mental states possess both a distinctive intentional structure and a distinctive phenomenology raises a straightforward question: what is the link between these two types of properties? Both of them have classically been studied separately, within a division of labor roughly corresponding to the division between phenomenologists and analytic philosophers. And this separatist tendency is not surprising giving the long-standing tradition in philosophy of mind of distinguishing qualitative states, considered as inherently non-intentional, from intentional states, generally considered as inherently non-conscious (See for instance Ryle 1949). Nevertheless, since the 90’s, this view has been largely challenged in favor of a more intentionalist picture of the mind (for an overview see Siewert 2002), according to which consciousness and intentionality are intimately related. Endorsing this intentionalist perspective, it seems that the sense of us and collective intentionality must be somehow related. In this paper, I will defend one way of accounting for this relation, that I will call consciousness-first approach to collective intentionality.

Even if the question of the relation between collective intentionality and phenomenology haven’t been directly addressed, we can identify two antagonistic approaches among authors who tried to account for the sense of us. Some seems to endorse an Intentionality-first approach, and try to ground this experiential property in a collective intentional state it is supposed to emerge from. For instance, Salmela (2022), argues that the “sense of togetherness” emerges from the shared concerns underlying collective emotions. On that approach, phenomenal properties play no role in accounting for collective intentional states, but should be rather considered as a mere by-product of those. Others, however, plead for a Consciousness-first approach, according to which collectively intentional properties are actually grounded in phenomenal ones. For instance, Schmid (2014, 2023) suggests that collective intentionality arises from a plural pre-reflective self-awareness, or Crone (2018, 2021) argues that a sense of us pre-structure explicit we-attitudes. On this second approach, the sense of us is what grounds collective intentionality, and is considered as necessary for collective mental states. Furthermore, the sense of us plays for these authors a central theoretical function in allowing them to avoid infinite regress in accounting for the collectivity of shared mental states.

While I am more sympathetic to consciousness-first approaches, I suggest that it is by no means clear in what sense an experiential property can be said to ground collective intentionality. In this paper, I propose to defend such views by clarifying the relation between consciousness and intentionality, and suggesting that collective mental states obtain their intentional structure in virtue of their phenomenal character. More precisely, I’ll argue that the intentional specificities of collective mental states fall into what Horgan & Tienson (2002) call “phenomenal intentionality”, in other words intentional properties that are constitutively determined by phenomenology, such that the sense of us, to rephrase Kriegel (2014), is what injects collective intentionality into the world. I will first present phenomenal intentionality theories and the three main arguments in their favor (Woodward 2019). First, assesability for accuracy arguments (Siewert 1998) suggest that a subject can be assessed for accuracy in virtue of the phenomenal character of his or her perceptual states. Second, phenomenal duplicates arguments (Loar 2003) suggest that we cannot conceive a pair of perceptual states that are phenomenologically identical but differ in all their intentional properties. And third, intentional contrast arguments (Siegel 2010) suggest that we cannot conceive a pair of perceptual states that differ intentionally without differing phenomenologically. According to their defender these three facts are best explained by perceptual phenomenal intentionality.

I will then show that all of these three arguments can be applied to the sense of us. In other words, I will defend that one can be assessed for accuracy in virtue of his or her experience of togetherness, that two identical experiences of togetherness that nevertheless differ in the presence or absence of co-participants both count as collectively intentional, and that a transition from individual to shared mental states cannot happen without a phenomenal switch to an experience of togetherness. I will argue that the sense of us should therefore be considered as a phenomenal-intentional property that grounds the intentional distinctiveness shared mental states. Finally, I will discuss two potential objections to this view: one regarding the nonconscious character of plural subjects, and a second concerning the absence of qualitative character in shared propositional attitudes. I will then conclude by stating two consequences of this view, one regarding the necessity of taking the sense of us as a distinguishing criterion for shared mental states, and a second regarding the constrains that phenomenological description of collective mental states should impose on accounts of collective intentionality.

Propositional attitude accounts of imagination (PA-I) currently fall short of explaining an important aspect of imagination. They do not explain why we attribute imaginativeness to a person when they use their imagination, nor do they explain that imaginativeness can come in degrees. We often attribute imaginers as being imaginative when they engage in pretense, fiction, and the creation of ideas and achievements. I provide cases that show that being imaginative picks out more than just the fact that one mentally represented some content that is neither perceptually available nor truth-directed; it refers to the particularity of how and what someone imagines. I argue that propositional attitude accounts can make sense of this by supplementing the view with a concept of flexibility.

A representative PA-I account comes from Nichols and Stich (2000), for whom imagination, conceived of as the representation of i-beliefs, is a propositional attitude that represents information as it would be given certain assumptions. The goal of their account is to explain how the imagination fits into our mental economy among other folk psychological states. Nichols and Stich offer i-beliefs to explain how we produce pretense behavior. Since we often attribute imaginativeness to pretenders, we should expect an account like this to explain what grounds such attributions.

I demonstrate that PA-I accounts fail to explain what makes someone imaginative. The representation of an i-belief is insufficient because it does not explain how one comes to have the imagining they do, and this is what is picked out by the concept of imaginative. PA-I only describes the resultant state as one of imagination, whereas we ought to describe the process that underlies it as being imaginative. A similar point was made by Wiltsher (2022) and Stokes (2014, 2016), who argue for a view of imagination as a process or as the ability to manipulate, respectively. However, neither Wiltsher nor Stokes describe the kind of process or manipulation required to do this job.

The concept of flexibility can provide us with the answer. We are attributed as being imaginative when we are able to control how we move information among settings. Cognitive scientists and psychologists describe this as ‘cognitive flexibility’, a type of executive function. They use flexibility to explain creativity (for example, Zabelina et al. 2010, 2018), but it can also explain the range of behaviors to which imaginativeness is attributed (including non-creative ones). Flexibility explains that we can organize information based on different principles, and that such organization requires moving information among settings. Moreover, one can organize information in highly flexible ways or in less flexible ways.

I look at two experiments in psychology that have demonstrated a relationship between imagination and flexibility. I argue their findings support supplementing accounts of imagination with a concept of flexibility. Veraksa et al. (2022) hypothesized that if one demonstrated a higher development of imagination, then one has a higher development of executive functioning. The authors asked 206 Russian school-aged children to complete ten drawings. They presented the students with cards, on which was a partial drawing. Students were asked to create a “whole” drawing, with no time limit, but were required not to erase or change anything once they started. They measured imagination development in terms of uniqueness, detail, and strategy, and compared these measures to executive functioning through card sorting tasks. The authors used general linear models to demonstrate that higher degrees of imagination co-occurred with higher degrees of executive functioning, in particular cognitive flexibility.

White (2018) compared 52 American college students with and without diagnosed ADHD on a similar drawing condition to examine how creativity and knowledge structures are related. Students were asked to construct drawings of alien fruit, and White determined that those with ADHD drew fruit that was based less on known fruit. In the analysis, White concludes that those with ADHD are more flexible in their response to task demands than those without ADHD.

From these experiments, we can see evidence that psychologists already see imagination and flexibility as related. We can conclude that the way in which one organizes information relates to how imaginative they are, and that the better one is able to move information from one setting to another, the more imaginative they are determined to be. This supports supplementing PA-I accounts with the concept of flexibility to explain the processes that lead to the production of an imagination state. As a result, we can make sense of attributions of imaginativeness in reference to this process of being flexible.