Institut Jean Nicod

Accueil > Séminaires/Colloques > Archives > Colloques > 2016-2017 > Skilled Action > Presentation




"Executive and automatic control of skilled action:

how they interface and interact"

May 11-12, 2017



Venue: Ecole Normale Supérieure, 29 rue d'Ulm, 75005 Paris

Convenors: Myrto Mylopoulos (Carleton University) and Elisabeth Pacherie (Institut Jean Nicod)





Thursday, May 11, 2017, afternoon - Salle Théodule Ribot (Ground floor)


2:00-3:00 Stephen Butterfill (University of Warwick)
The Interface Problem Strikes Again: Is There a Role for Action Experience in Fine-Grained Executive Control of Very Small Scale Actions?

3:00-4:00 David Papineau (King's College London)
The Intelligence of Basic Actions

4:00-4:30 Coffee Break

4:30-5:30 Corrado Sinigaglia (University of Milan)
The Three Lives of Motor Representations

5:30-6:30 Myrto Mylopoulos (Carleton University) & Elisabeth Pacherie (Institut Jean Nicod)
Skilled action and dynamic interfacing



Friday, May 12, 2017, morning - Salle 235A (second floor)


9:30-10:30  Wayne Christensen (University of Warwick)
Multiple interface problems in the control of skilled action

10:30-10:45 Coffee Break

10:45-11:45 Ellen Fridland (King's College London)
Whenever the twain shall meet: the many joints of skilled action control

11:45-12:45 Joshua Shepherd (University of Oxford)
Skilled action and the double life of intention

12:45-2:00: Lunch break


Friday, May 12, 2017, afternoon /Salle 235B (second floor)


2:00-3:00 John Michael (University of Warwick)
Executive and automatic control of skilled social interaction: how they interface and interact

3:00-4:00 Nura Sidarus (Institut Jean Nicod)
The role of action selection fluency in the sense of agency

4:00-4:30 Coffee break

4:30-5:30 General discussion




The Interface Problem Strikes Again: Is There a Role for Action Experience in Fine-Grained Executive Control of Very Small Scale Actions?

Stephen Butterfill (University of Warwick)

Sometimes when performing a very small scale purposive action such as pressing a switch, a goal is both represented motorically and specified by an intention. When things go well, the goal specified by the intention is, or matches, the goal represented motorically; and this is no accident.  But this does not always happen. After all, it is possible to perform goal-directed actions contrary to anything one intends. How is it possible that the goal specified by an intention ever nonaccidentally matches a goal represented motorically? Answering this question is a problem--the ‘Interface Problem’---because at least two obstacles stand in our way.  One is that intentions and motor representations are not inferentially integrated. Another is that an answer must specify ‘not just how motor representations are triggered by intentions, but how motor representations’ continue to match intentions as circumstances change in unforeseen ways ‘throughout skill execution’ (Fridland, 2016 p. 19). This second obstacle is illustrated by cases of fine-grained intentional control over very small scale actions, as when making a slight period adjustment in performing a rhythmic action. Existing attempts to solve the Interface Problem (e.g. Mylopoulos & Pacherie, 2016) are arguably at best only partially successful. One approach is to consider the nature of experiences associated with motor representation and the role these might play in fine-grained executive control of very small scale actions. Accordingly this talk will introduce and evaluate a view on which motor representations structure experiences of events.


Multiple interface problems in the control of skilled action

Wayne Christensen (University of Warwick)

Butterfill and Sinigaglia (2012) argue that an ‘interface problem’ confronts accounts of intentional action: cognitive and motor processes each represent action outcomes, but because they do so using different representational formats it is difficult to see how they can be appropriately aligned. Mylopoulos and Pacherie (2016) have proposed a solution which holds that the relation between intentions and action outcomes is mediated by ‘executable action concepts’ and motor schemas. I endorse the main features of this approach but reinterpret the problem, arguing that there are multiple interface problems and they are not primarily based on representational format per se, though format plays a role. The Mylopoulos and Pacherie solution effectively claims that the intentions and higher level motor planning and control processes involved in proximal action control share a common (conceptual) representational basis. Arguably, this can help to explain rapid, flexible action abilities. But the representations that participate in online action control face substantially different organisational demands to those that participate on offline cognitive processes, and this results in an interface problem between distal and proximal intentions that presents practical difficulties for experts. In addition, while the Mylopoulos and Pacherie account helps to explain how lower level motor processes can be appropriately shaped by proximal intentions, a fuller understanding of this interface is needed. Lower level forms of predictive motor control, such as anticipative postural adjustment and endstate planning, develop in the course of skill learning and can enhance the fluency of action. However, they are not directly responsive to conceptual action goals and can conflict with them. A poor practice approach in skill learning (excessively rapid, repetitive practice with low attention) allows considerable interference to arise, and produces inferior performance ability with weaker intentional control over action. Better practice technique (slower, varied practice with focused attention) helps to build a stronger interface between proximal intentions and lower level motor processes, and results in superior action abilities.


Butterfill, S. A., & Sinigaglia, C. (2012). Intention and Motor Representation in Purposive Action. Philosophy and Phenomenological Research.

Mylopoulos, M., & Pacherie, E. (2016). Intentions and Motor Representations: the Interface Challenge. Review of Philosophy and Psychology, 1–20.


Whenever the twain shall meet: the many joints of skilled action control

Ellen Fridland (King's College, London)

To give an account of how personal-level control states interface with low-level motor representations to generate skilled, fluid action, we must first identify when or at which points, control states and processes connect to lower-level procedures. My goal in this talk will be to present a comprehensive picture of the different points of contact between propositional and subpersonal states, which any theory of skill must account for. I will forward both philosophical and empirical reasons supporting the claim that semantic and subpersonal states interface at a variety of action joints and temporal scales.


Executive and automatic control of skilled social interaction: how they interface and interact

John Michael (University of Warwick)

Humans are experts in social interaction. To a large extent, this is likely to be the result of evolutionary processes which have endowed us with dedicated mechanisms for social learning and social interaction (e.g. gaze following, imitation, interactive alignment). Does this mean that human social expertise comes for free, without the need for strategic thinking and deliberate control superfluous? No – drawing upon Christensen and colleagues’ (2015) ‘Mesh’ theory of skilled action, I will point out several ways in which strategic thinking and deliberate control are crucial to the acquisition and exercise of social expertise. I first discuss the implications of these insights for the interface challenge (Butterfill & Sinigaglia, 2014; Mylopoulos & Pacherie, 2016) and then present a case study on social expertise in Möbius Syndrome (MS). MS is a form of congenital, bilateral facial paralysis resulting from maldevelopment of the sixth and seventh cranial nerves. Since people with MS are unable to produce facial expressions, and are thus deprived of a central medium for the automatic communication of emotional information and the exchange of social cues, it is unsurprising that many people with MS experience difficulties in their social interactions and in terms of general social well-being. However, some people with MS have cultivated expert compensatory strategies, such as using more hand gestures and prosody to express themselves, and do not report having any difficulties in social interaction. On the basis of interviews with these individuals, we developed a social skills workshop designed to train individuals with MS to adopt alternative strategies to compensate for the unavailability of facial expression in social interactions (e.g. expressive gesturing and prosody), and have implemented this workshop in Denmark, the US and the UK. In order to evaluate the effectiveness of the workshop, participants with MS engage in a battery of interactions before and after the workshop with partners who do not have MS. I will discuss our observations against the theoretical backdrop of Mesh and the interface challenge.


Skilled Action and Dynamic Interfacing

Myrto Mylopoulos (Carleton University) & Elisabeth Pacherie (Institut Jean Nicod)

An adequate theory of skilled action must address two central explananda. First, central to a full understanding of skilled action is some account of the intelligence that it exhibits. We take flexibility to be at the heart of this phenomenon.  Skilled behaviour is highly sensitive to the nuances of a given action context and involves robust and fine-tuned interaction among different psychological states and processes—what the skilled action theorist must therefore explain is how this flexibility gets trained up and implemented by action control mechanisms. This understanding of flexibility directly gives rise to the second explanandum facing the skilled action theorist: how it is that the different psychological states and processes involved in the implementation of the flexible control of skilled behaviour manage to interact in a coordinated way given the different forms they take? Following Butterfill and Sinigaglia (2014), we call this "the interface problem". We propose that three dimensions of flexibility characterize skilled motor behaviour – sensory sensitivity, situational sensitivity, and strategic sensitivity – and that the latter two require a continuous interplay between cognitive and motor control processes, giving rise to a dynamic situated interface problem. We offer a two-tiered solution to this problem, appealing to motor schemas to explain how intentional goal states can hook up with motor representations and to structured action representations to explain the dynamic and situation-sensitive nature of this interface.


The Intelligence of Basic Actions

David Papineau (King's College London)

I shall argue that unconscious sub-personal motor control systems are fully capable of representing goals, and of adjusting behaviour in response to variable circumstances. This means that there is no ‘interface’ problem, apart from the general issue of understanding how conscious long-term intentions of any kind manage to affect behaviour.”


Skilled action and the double life of intention

Joshua Shepherd (University of Oxford)

This talk concerns what Butterfill and Sinigaglia (2014) have called the Interface Problem. The problem surrounds the puzzling idea that though intentions appear to possess a propositional representational format, and motor representations appear to possess a proprietarily motoric representational format, these states appear to interact in rich ways during the execution of action. I discuss Butterfill and Sinigaglia’s proposed solution, Mylopoulos and Pacherie’s (2016) proposed solution, and I offer a novel solution of my own. Along the way I reflect on the importance of the Interface Problem for any account of the role of intelligent processes in skilled action.


The role of action selection fluency in the sense of agency

Nura Sidarus (Institut Jean Nicod)

Human voluntary action is typically accompanied by an experience of being in control of one’s actions and of their outcomes, that is, a sense of agency. Research has shown that the sense of agency relies on retrospectively comparing observed outcomes with predictions, or expectations. Our work has shown that the sense of agency is also prospectively informed by the fluency of action selection processes. I will discuss the cognitive and neural mechanisms underlying the contribution of action selection fluency to the sense of agency, and how this is integrated with outcome-related information.


The Three Lives of Motor Representations

Corrado Sinigaglia (University of Milan) & Steven Butterfill (University of Warwick)

Motor representations live a kind of double life. Several studies suggest that, although paradigmatically involved in performing actions, they also occur when merely observing others act and sometimes influence thoughts about the goals of observed actions.  Much less studied is whether motor representations may live a further, third life when people are not alone, but together. What happens motorically when people are acting together or observing others acting together? In this talk I shall draw on recent psychological and neuroscientific research to argue that fully explaining acting together may involve identifying a certain interagential structure of motor representations. I also explore the conjecture that this interagential structure of motor representations might be involved also in observing other people acting together, providing some preliminary supporting evidence.