How much cognition is in action? Using dual-tasks to explore the involvement of cognitive processes in action planning, programming and motor learning (2015-2018; 2018-2021)
Team
Prof. Dr. Thomas Schenk Principal Investigator | | Frederic Göhringer PhD Candidate | |
Miriam Löhr Collaborative PhD Candidate |
Abstract (2018-2021)
Are visuomotor acts immune to interference from a secondary task? We asked this question in our first funding period and concluded that there is no such immunity. Instead we and others found that action and cognition interact in sometimes surprising ways. While a concurrent, cognitive task will typically render the visuomotor behavior less efficient, not all visuomotor behaviors are affected by this curse, instead there seems to be a continuum of vulnerability to dual-task costs and cognitive penetrability. We are planning a series of experiments to explore this continuum. Some visuomotor tasks can be complex, demanding that the movement is adjusted to a distal goal, that its environmental consequences are anticipated or that the proper balance between task complexity and physical effort is found. For such tasks we expect high mental load and concomitantly high vulnerability to resource limitations due to secondary task demands. On the other hand we have acts which appear to be so engrained that the actor will perform them without even becoming aware of the stimulus that triggered the act. To introduce such situations and provide the opportunity to study the dual-task costs in this type of behavior we will exploit the phenomenon of saccadic suppression using gaze-contingent displays. The same continuum that is found in motor control is also at work in motor learning. The role of cognition in different aspects of motor learning will be examined in a sample of patients with cerebellar lesions. Finally, concurrent cognitive activity leads not in all instances to a general decrement of motor performance. Sometimes the effects can be more specific, causing a processing bias, or can be even helpful. Our last aim in this proposal is to understand when this biasing effect occurs and to test whether in some conditions mental load can facilitate motor learning.
Abstract (2015-2018)
The present project aims to evaluate the perception-action model (PAM) established by Goodale and Milner (1992) using multitasking.
According to the PAM, the visual system is divided into two anatomically and functionally distinct streams, namely the dorsal and the ventral streams. The dorsal stream, running from the primary visual cortex (V1) into the parietal cortex, is assumed to process visual information used for action (e.g. grasping an object), while the ventral stream, leading from V1 into the temporal cortex, is suggested to process visual information for perception (e.g. estimating the size of an object). This functional distinction attributes several properties to each stream. The dorsal stream codes spatial information with regard to the observer’s body (egocentric), while the ventral stream considers spatial relations of objects to one another (scene-based or allocentric). Apart from that, the dorsal stream is ideal for online corrections during the course of a movement, whereas the ventral stream is involved in planning movements before their actual onset (offline). Furthermore, the dorsal stream has no memory and functions in an immediate way, while the ventral stream takes care of a stable representation over time, involving information about the object identity. Finally, the PAM assumes automaticity in the dorsal stream and suggests that the cognitive system exclusively relies on information from the ventral stream which, in contrast to the dorsal stream, has conscious access to visual information.
Dual-task costs mainly evolve due to sharing of the same capacities by two simultaneously performed tasks and overload of the cognitive system. Since the dorsal stream is assumed to be divorced from the cognitive system (Clark, 2009), dorsal stream tasks should not be affected by a secondary task, or at least be much less prone to dual-task interference than ventral stream tasks. Former research already investigated this idea (e.g. Singhal et al., 2007), but some of the studies suffered from methodological problems like e.g. rather insensitive measurement techniques or discrepancies in relevant concepts.
Therefore, in the present project, we systematically explore the effect of a secondary task on dorsal and ventral stream tasks, respectively, employing a broad range of paradigms. Very sensitive 3D motion registration and eye-tracking will be applied during all experiments in order to allow the detection of subtle effects. The secondary task will be a rapid serial visual presentation task, in which digits are flashed in quick succession and the participant is asked to report the identity of a target stimulus (ventral stream task). Simultaneously, five work packages (WP) will address the previously explained distinctions between the dorsal and the ventral stream under single- and dual-task conditions.
With the first WP, we will investigate the influence of the secondary task on egocentric vs. allocentric tasks. For example, participants will perform pro-pointing vs. anti-pointing under single- and dual-task conditions. While the pro-pointing task consists of direct pointing to a target stimulus, in the anti-pointing condition, subjects have to point to an imagined target in the mirror-symmetric position of a presented stimulus (allocentric condition). According to the PAM, one would expect a larger effect of the secondary task in the anti-pointing condition. In addition to the pointing paradigm, we will also use grasping and obstacle avoidance vs. line-bisection paradigms.
The second WP addresses the distinction between online vs. offline processes under single- and dual-task conditions. As an example, we will introduce a shift into a classical obstacle-avoidance paradigm. In the stable condition, participants simply guide their hand through two obstacles, while in the shift condition, one of the obstacles will change its position upon motion-onset, such that participants will have to correct their movement online (action programming). According to the PAM, performance in the shift condition should suffer more from the simultaneously performed secondary task. Within this WP, we will also employ pointing paradigms with and without shifts and various grasping paradigms.
With the third WP, we will compare immediate vs. delayed actions and their sensitivity to dual-task interference. For example, subjects will perform a grasping task involving immediate grasping of a seen object and a delayed grasping, in which visual object presentation and grasping will be separated by 2500ms. According to the PAM, one would predict that the delayed grasping is more prone to dual-task interference than immediate grasping. Other paradigms used in this WP are immediate vs. delayed pointing and immediate vs. delayed obstacle-avoidance.
WP four will focus on the assumption that dorsal stream tasks should be immune to perceptual interference, employing mainly the Garner interference paradigm.
WP five contains cooperation projects with Kunde & Janczyk and Bublak, Klingner and Witte, respectively. Together with Kunde & Janczyk, we will explore an alternative account for evidence which has so far been interpreted as strong evidence for the PAM. More in detail, we will investigate whether an enhanced effect of multitasking on delayed compared to immediate action can be explained in terms of interference due to shared resources during action monitoring and initiation of the subsequent action. The cooperation project with Bublak, Klinger and Witte will address the influence of working memory based secondary tasks on postural controls in healthy and clinical populations.
Project Output
Löhr-Limpens, M., Göhringer, F., Schenk, T., & Hesse, C. (2019). Grasping and perception are both affected by irrelevant information and secondary tasks: New evidence from the Garner paradigm. Psychological Research. doi:10.1007/s00426-019-01151-z
Göhringer, F., Löhr-Limpens, M., & Schenk, T. (2018). The visual guidance of action is not insulated from cognitive interference: A multitasking study on obstacle-avoidance and bisection. Consciousness and Cognition, 64, 72-83. doi:10.1016/j.concog.2018.07.007