1.1 Doing Things Together
In everyday contexts we do numerous things together. We walk together, we cook, eat, and drink together, do chores together, and engage in more complex activities such as building complex machines, work in highly structured organizations, and play in bands and orchestras together. Whether we cycled to work together or by sheer coincidence cycled to work almost at the same time and only a few meters apart, makes a difference in the way we prepare for the cycling, do the cycling, and experience the cycling. What this difference between parallel and joint action consists in is hotly debated. This debate on joint action is shaped by both philosophers and psychologists. Their traditions give them different starting points and different worries, leading to different theories. In recent scholarship, there is a tendency to distinguish between emergent coordination and planned coordination. I will argue that this pushes us to a position of dualism, a strict division between two kinds of joint action, which I deem problematic. Such a push towards two processes or conceptualizations through which to understand joint action has several consequences. For example, certain phenomena become understudied because they do not fit nicely with the distinction. The transition and relation between the âtwoâ kinds of joint action become harder to conceptualize because we formulate the conceptualization of the one in contrast to the other.
The tendency to introduce and use binary distinctions, so I will argue, is not unique to the study of joint action. In fact, in this book I will point at several different binary distinctions, which are all intuitively appealing but fail to give us the tool to conceptualize many everyday phenomena. I will also argue that these binary distinctions relate to one another. The four binary distinctions that I will focus on are:
-
Automatic versus controlled processes
-
Mere bodily movements versus full-blown intentional actions
-
Emergent coordination versus planned coordination
-
Joint action versus parallel action.
These binary distinctions stand in complex relations to one another. Take, for example, distinction 2. The distinction between mere bodily movements and (full-blown) intentional action has a strong relation to whether the process can be controlled by the agent or happens to the agent. Reflexes have often been understood as mere bodily movements. Mere bodily movements are thought to occur automatically and therefore to be out of reach for the agent to control. I will argue, however, that âautomaticâ and âcontrolledâ should not be understood as two opposites: there are processes that can be described as both controlled and automatic. This has an impact on the distinction between mere bodily movements and intentional actions. A similar relation also exists between distinctions 1 and 3, which makes it harder to conceptualize the space between emergent coordination and planned coordination. In all of these cases I will argue that it is a mistake to think about the distinctions as clear-cut dichotomous distinctions. Instead, the two opposites should be understood as ends of spectrums. In the case of the binary distinction between automatic processes and controlled processes, I will also argue that two distinctions are mixed up, namely the spectrum from automatic to non-automatic processes, and that from controlled to not-controlled processes.
A better understanding of the reason why I want to move away from a binary distinction between joint action and parallel action becomes available once we have gone through the first three binary distinctions and their problems. By moving towards more gradual distinctions and multiple factors that weigh into the success of a joint action, we also open up the intermediate space that lies between joint action and parallel action.
I develop two claims: (1) There are several functions that help human agents coordinate, which entails that joint action can be best understood by understanding the differences but also the interrelatedness of these different types of coordination. (2) In order to understand the different forms of coordination and their contribution to join action we need a multidimensional conceptual space. This multidimensional space will allow us to understand the jointness of an action in a more encompassing way.
The notion of control and how it relates to the dichotomy between automatic and non-automatic processes will be central in the book. I argue that the concept of control offers us a possibility to a better, more gradual, understanding of emergent coordination and planned coordination and a richer and more varied understanding of the jointness of our actions. This allows me to step away from the dichotomy between joint action and parallel action. I will use a provisional three-way distinction of levels of control that we find in skillful action (Fridland, 2014; Christensen, Sutton, & McIlwain, 2016; Christensen, Sutton, & Bicknell, 2019).1 These three levels of control provide an understanding of different ways in which we can coordinate our actions, leading to different ways of acting together. Control and coordination are two core elements that I will explore in detail in the second part of the book. A third element, relating to âwhat it feels like to act togetherâ, will also be explored. This element of togetherness or sense of togetherness, I argue, can arise from the different forms of coordination, as well as from situational and historical factors.
When we ask what we mean when we say that a behavior is automatic or a process is automatic, the answer will be different depending on the specific capacity we are interested in. For example: decision-making, language processing, or social cognition. Traditionally, four features were assumed to coincide when a process was said to be automatic (see also chapter two).
Table 1:Following Barghâs (1994) âfour horsemen of automaticityâ.
| Automatic | Non-automatic |
| Absence of (conscious) intention | Intentional |
| Absence of conscious awareness | Conscious |
| Uncontrolled (inability to disrupt) | Controlled (ability to disrupt) |
| Efficient operation (fast and not/hardly needing working memory capacity) | Time consuming and using cognitive resources (slow and consuming working memory capacity) |
The four features were thought of as a package deal in which you always get all four together. When a process was said to be automatic this implied, for example, that it also was unintentional and effortless. Studies have shown, however, that the four different features can be dissociated, creating problems for the traditional view (Bargh, 1994; Moors & De Houwer, 2006, 2007). Some processes can be characterized as efficient and intentional, and that would mean they are both automatic and non-automatic. This requires us to be more specific when we talk about whether, and in what sense, something is automatic. A process that is said to be automatic could be further described as either efficient, unconscious, non-intentional, uncontrolled, or any combination of these features. Moreover, when studying these four features, it seems that at least some of them are better understood as polar oppositions rather than as exclusive dichotomies. Take for example the feature âawarenessâ: it might be that we are aware to some extent without all of our attention focusing on a certain process, or we might become aware of something we are doing after some time. The dissociation of the different features of the package deal and the gradual understanding of some of the individual features render inadequate the traditional, dichotomous model (automatic/non-automatic) for covering many phenomena we encounter on a day to day basis. The perniciousness of this model can only be explained by its intuitive appeal. Chapter two will give an overview of the different features that we typically associate with automatic processes, providing reasons to adopt a model where the distinction between automaticity and non-automaticity is understood through a multidimensional space with several features. Secondly, it will lead us to a model where these features of automaticity are partly understood as gradual. Overall, this will allow us to move away from a strong form of dualism and create room for phenomena that combine automatic and non-automatic processes or are best described as in-between either categorization.
Of the four features of (non-)automaticity I will focus specifically on intentions and control (chapter three). Philosophers of action and joint action have tried to conceptualize the (joint) action through an understanding of the intention(s) involved. There is an important distinction to be made between control as a more biological notion and control as a normative, agentic notion. The latter is often thought to have an intimate relation with intentions, whereas this is not necessarily true for the first. After arguing that the binary distinction between automatic and non-automatic causes a gap in our conceptual tools for analyzing the middle ground, I will turn to current developments in the debate of skill acquisition and skillful action. Our understanding of the notion of control has implications for the way we can think about joint action (chapters five and six). At the same time, the way in which emergent coordination is spelled out in psychology also bears implications for our philosophical discussion of control and agency.
Chapter three will briefly report on the distinction between intentional and unintentional bodily movement. This distinction appears at first to be a dichotomy. I will point at a discussion where âintermediateâ phenomena are said to cause problems for this dichotomy and review some proposals to conceptualize this intermediate category. The result will be a first conceptualization of the space in between, which can later be used to define such a space for cases of joint action. If such an intermediate category makes sense, or if we should think about intentional/unintentional as a gradual distinction, then we most likely also act together on such âintermediate actionsâ and we need an adequate theory on how this could work and what consequences this has for our theories. The idea of an intermediate space will be further developed in chapter four, where I focus on recently developed accounts of skillful action. Current views in the debate in skill acquisition and skillful action might give us an understanding of the intermediate level, as well as a way to integrate the different levels. Moving from being a novice to an expert was long thought to reflect a move from reflective action to automatic action. This is much in line with the discussion on automaticity, following the idea that a process is either automatic or non-automatic. We now have data that suggests that this picture is inaccurate. Several recent proposals distinguish three âlevelsâ of control, with different degrees of automation (Pacherie, 2008; Fridland, 2014; Christensen, Sutton, & McIlwain, 2016; Christensen, Sutton, & Bicknell, 2019). In order to see the richness of this proposal, I will first discuss current proposals that are (but should not be, as I argue) conceptualized as two distinct ways of understanding how we can act jointly.
In my review of Bratmanâs influential theory of planned coordination (chapter five) and the four most developed current theories of emergent coordination (chapter six), I will stick, at first, to this binary picture, although I will also point at parts in these proposals that actually show the openness and the need for a richer spectrum. I do so by linking the discussion on three levels of control and the proposal of functions that allow us to act jointly, also proposing a threefold distinction of coordination.
With all this material in hand, we will be well equipped to discuss the overarching binary distinction between parallel and joint action. Philosophers of collective intentionality have wondered how we can distinguish parallel cases from cases where we act together. Often their theories argue in favor of one characteristic, feature, or process that differentiates the two. An often-found methodology in the debate in collective intentionality is the use of examples as a way of specifying the subject, where we use sets of contrasting cases that point at distinctions.2 The examples in collective intentionality debates distinguish parallel actions from joint action. Consider the following key examples (Table 2):
Table 2:Archetypical examples of collective intentionality.
| Walking together versus walking in parallel | Margaret Gilbert |
| Painting a house together versus pa... |
