“Like a ball and chain”: Altering locomotion effort perception distorts spatial representations

https://doi.org/10.1016/j.jenvp.2018.10.008Get rights and content

Highlights

  • Experiencing increased locomotion effort distorts walking speed perception but not actual walking speed.

  • Participants with low spatial ability tend to spend more time planning routes when wearing ankle weights.

  • Wearing ankle weights alters spatial memory retrieval.

  • Motor simulation and anticipated physical effort could play a role in spatial processing.

Abstract

This study aimed at demonstrating the role of sensorimotor information through the influence of walking effort perception on the process of route planning for navigation. Participants were asked to walk wearing weighted or non-weighted ankle belts prior to learning virtual environments. They then had to plan routes between landmarks and to estimate travel times and distances for each route. Results newly showed that loaded ankle weights made participants underestimate their own self-reported walking speed compared to the non-loaded participants. In addition, participants with low spatial ability tended to spend more time planning their route when wearing weighted belts than with non-loaded belts. Loaded ankle weights also negatively affected performances on a spatial memory task: landmark position retrieval on a map. These results provide support for embodied views of spatial cognition.

Introduction

Embodied views of cognition argue that cognitive processing is strongly intertwined with action-related mechanisms, instead of only amodal representations (Barsalou, 2008). A large body of literature has demonstrated the influence of body-related information on perception (Tucker & Ellis, 1998) or even language comprehension (e.g. Glenberg & Kaschak, 2002). In the field of spatial cognition, some studies also strongly suggest that space perception could be shaped by body experiences and motor actions (e.g. Franklin & Tversky, 1990), but they mainly focused on immediate “around the body” space (Tversky, 2009), and perceptual tasks that were only a part of spatial processes. Our aim in this study was to provide new insights into an embodied view of spatial cognition by studying the influence of effort perception in large-scale realistic spatial processing, which implies the use of locomotion. To do so, we focused on a specific type of complex spatial task which has not yet been studied in an embodied framework: route planning. This work also provides information about the processes of motor imagery, and especially how it can be influenced by effort perception. In the first part, we will review previous works on embodied spatial cognition dealing with the influence of motor-related processes during the encoding of spatial representations. We will then show how the manipulation of physical effort can influence spatial processing through motor imagery. Lastly, we will present the field of route planning, and how its study could add new elements to embodied spatial cognition.

Spatial representations have been traditionally described as images of previously visited places, including information about Euclidian metrics in a two dimensional space (Baird, 1979). This view is derived from the concept of cognitive maps (Tolman, 1948). The assumption of cognitive maps as a coherent spatial layout of landmarks and metric properties is debated nowadays, especially because it appears too costly in terms of cognitive resources. McNamara (1986) provided evidence for a hierarchical model, in which the spatial representation is fragmented into different regions for reasons of storage efficiency and economy of cognitive resources. Various authors also started to question the isomorphic and abstract nature of a spatial representation. Among alternative approaches, it has been described as a cognitive collage (Tversky, 1993). This view emphasizes that mental representations are not necessarily isomorphic, as substantiated by numerous studies showing that metric information is not well preserved in spatial representations (Maddox, Rapp, Brion, & Taylor, 2008). In addition, this approach considers a multimodal construction of spatial representations, in contrast with the abstract nature of cognitive maps. Within this framework, body-based information is among the material that can participate in the multimodal integration of spatial representations. It has been mainly studied through the comparison between “passive” and “active” learning of spatial material, contributing to a first step towards embodiment-inspired views on spatial cognition.

Active learning is often described as the combination of various components involved during the construction of spatial representation, in opposition to passive learning. Active components of spatial learning include decision-making during navigation, allocation of attention during learning, mental manipulation of spatial material, and body-based information (see Chrastil & Warren, 2012, for a review). Body-based information includes all the motor components of physically walking through an environment (Chrastil & Warren, 2012), among which is podokinetic information (Mittelstaedt & Mittelstaedt, 2001). This comprises two types of body-related information: efferent motor actions and proprioceptive feedback about the displacement of body segments and muscle activity. Several studies have shown that being active during spatial learning results in a better performance than the mere observation of a visual description of places, with podokinetic information contributing significantly more than decision making to learning. The availability of podokinetic information during the learning of a new environment helps individuals to retain metric information about distances, angles and landmarks (Waller, Loomis, & Haun, 2004). Podokinetic information is responsible for biomechanical control of locomotion through leg/feet action with the floors, and bottom-up sensory-motor control of locomotion (Weber, Fletcher, Gordon, Jones, & Block, 1998). The availability of podokinetic information during learning has been shown to improve survey knowledge, which is a map-like representation of distances and angles between landmarks (Ruddle, Volkova, & Bülthoff, 2011). All these findings speak in favor of the prominent role of podokinetic information during the construction of spatial representations. Though the link between podokinetic information and spatial processes has been assessed, the cognitive mechanisms underlying this relation are yet to be understood. A strong step towards this is provided by mental imagery studies, from classical mental scanning literature to new embodied insights on motor simulation.

Motor simulation can be described as a conscious or unconscious process of action simulation, which is “an internal representation of motor programs without overt movement” (Jeannerod, 2001). Early on in the field of mental imagery, a linear relation between the distances on a map and the time taken by participants to mentally travel these distances was shown (Kosslyn, Ball & Rieser, 1978). It is also known that imagined movements share the same biomechanical constraints as real movements (Jeannerod, 1999) and that motor imagery time is linked to movement execution time for simple as well as complex movements (McIntyre & Moran, 1996). Likewise, strong evidence has even shown that imagined walking speed closely matches actual walking speed (Papaxanthis, Pozzo, Skoura, & Schieppati, 2002). However, contradictory results showed some discrepancies between imagined and walking time. Kunz, Creem-Reghr and Thompson (2009) failed to replicate the results by Papaxanthis et al. (2002), conversely showing that individuals imagine walking speed faster than their actual walking speed. Decety, Jeannerod, and Prablanc (1989) showed that wearing a heavy load as a backpack does not decrease actual walking time, but decreases imagined walking time. These results can be explained by a recalibration process, as a result of the changing relationship between perceptual and podokinetic information during locomotion (Rieser, Pick, Ashmead, & Garing, 1995). Recalibration has an effect on actual walking behaviour, and is believed to also affect imagined walking (Kunz, Creem-Regehr, & Thompson, 2009). Its effects on motor simulation are not clearly established, and to date only a few factors that provoke this recalibration have been identified (e.g. optic flow). Another one is the anticipated physical effort, on which we will focus in this work.

Motor-related activity can be associated in most cases with experiencing physical effort to use podokinetic information during encoding. Physical effort is defined as the recruitment of physical and muscular resources in order to resist an opposing strength. In the example of walking, it lies in the muscular effort deployed towards the ground to lift the body and move it forward. From an embodied point of view, motor imagery and more generally motor simulation could contribute to constructing spatial representations by including information about walking effort management. Hence, during spatial processing, podokinetic information linked to locomotion might be mentally simulated while retrieving effort-related information, contributing to the encoding and retrieval of metric properties. For example, it has been shown that individuals tend to estimate slopes in front of them as steeper if they wear a heavy backpack (Bhalla & Proffitt, 1999), or overestimate small immediate distances in their field of view (Proffitt, Stefanucci, Banton, & Epstein, 2003). This can be associated with recalibration effects on perception from effort related information. Other work showed that individuals tend to choose preferentially south-oriented paths on a map compared to north-oriented routes (Brunyé et al., 2012). The authors interpreted this as evidence of the existence of an implicit association between North and climbing steeper landforms. Lessard, Linkenauger, and Proffitt (2009) also showed that applying additional walking effort to their participants influenced their immediate space perception. Participants wore either loaded ankle weights or non-loaded ankle weights. They were then submitted to an action capability task, where they had to assess whether they would be able to jump a certain gap. Finally, they had to estimate the metric extent of the gap by matching external distant markers with the gap. Results showed a recalibration effect in that wearing ankle weights made participants overestimate distances, but only for jumpable gaps; outside of their action capability range (non-jumpable gaps), no effect of ankle weights was observed.

These experiments suggest that physical effort and possibly motor simulation are involved in our interactions with our immediate environment. They also suggest that the effort component of motor execution can influence our itinerary choices, action capability estimations and short-extent distance estimations by recalibrating perceptions and actions. How individuals use motor simulation for large-scale spatial learning has still to be elucidated. A first attempt has been made in the field of reading comprehension: Brunyé, Mahoney, and Taylor (2010) found that route description readers implicitly adjust their reading speed to both metronome and footsteps, whereas readers of survey descriptions adjust their reading speed to metronome sounds only. Footstep rhythms might influence reading speed only when locomotion is primed by the route perspective text, possibly because first person spatial texts tend to stimulate mental simulation of the motor walking activity. Furthermore, only the sound of running footsteps led route description readers to overestimate distances. Brunyé et al. (2010) suggested that reading a first-person spatial text involves mentally simulating motor representations of locomotion. To date, only a few studies have focused on how podokinetic information linked with physical anticipated effort was used to learn the environment; no other studies linked this influence of anticipated effort on the construction of spatial representations with motor simulation, and even fewer when using real-like complex spatial tasks. In fact, it is difficult to control the parameters of mental imagery, which led authors to use only distance estimation tasks (Decety et al., 1989, Lessard et al., 2009) in a very narrow scale. These perceptual tasks hardly reflect all the complexity of spatial processing in real-life. In this study, we will attempt to fill this gap and test the recalibration of motor simulation caused by additional physical effort applied to podokinetic information during a route planning task.

Route planning is a spatial problem-solving task in which the objective is to find a path from an origin to a destination. The route planning task implies that origin and destination locations are known, and requires identifying the spatial relation between these two points, comparing different possible options and selecting the most suitable path (Golledge, 1995). As a result, route planning is considered a specific subset of navigation tasks, namely “unaided directed wayfinding tasks” (Wiener, Büchner, & Hölscher, 2009). The spatial skills required for route planning and environmental learning are different from the perceptual spatial tasks presented so far as they involve a large number of cognitive processes such as working memory (Gyselinck, De Beni, Pazzaglia, Meneghetti, & Mondoloni, 2007), various spatial abilities (Hegarty, Montello, Richardson, Ishikawa, & Lovelace, 2006), executive functions, and also other individual and contextual factors (Grison, Gyselinck, & Burkhardt, 2016). Contrary to distance estimation tasks, route planning is also much closer to a real-life task that better reflects spatial cognition in all its complexity.

Although no work has directly considered motor simulation during route planning, some studies suggest that spatial problem solving can benefit from podokinetic information: Grant and Magee (1998) for example, found that realistically walking in virtual reality allows participants to find their way faster compared to walking on the spot (altered podokinetic information). Besides, a large number of planning heuristics have been described, defined as a fast mental process aiming to find solutions minimizing computational cost, without guaranteeing optimality (MacGregor, Chronicle, & Ormerod, 2006). Most of these studies have dismissed motor influence by using a map-based methodology or abstract material such as points to connect on a map. To our knowledge, only Brunyé et al. (2012) explained planning heuristics by processes related to the embodied cognition framework, claiming that the preference for south routes on the map is derived from the “north = uphill” implicit association. Going further, planning heuristics might be used to minimize an anticipated navigational effort, as we never plan routes in real situations without the aim of walking the planned path. Consequently, we assume that motor simulation is used during route planning to anticipate the navigational effort of possible routes, and that motor simulation implicitly uses parameters derived from podokinetic information to take the physical effort into account through recalibration processes.

Our claim is that using more realistic material such as virtual reality devices to make individuals effectively learn places and plan routes in a natural way should make it possible to observe a wider variety of functional clues other than computational economy, including walking effort management. This idea draws on a large body of literature dealing with the feeling of presence in virtual reality, showing that individuals tend to feel that they are physically located and interacting with the virtual environment if it is immersive enough, hence implicitly reproducing natural behaviours (Sanchez-Vives & Slater, 2005).

The objective of our study was to show that the anticipated walking effort derived from podokinetic information has an impact on the complex process of route planning for navigation. As walking effort perception could play a role in the encoding of metric properties and influence the motor simulation of locomotion, it should also be involved in a complex spatial process such as route planning. This idea implies that motor simulation could be used to mentally travel routes during planning and to evaluate the cognitive and physical cost of each alternative, in order to reduce the anticipated effort. A route planning task using virtual reality was chosen here for its high level of complexity – mixing decision-making with spatial recall – but also for its realistic value.

Our first hypothesis is that increasing walking effort perception (using loaded ankle weights) should lead participants to implicitly modulate their perception of their own walking speed by recalibrating to a slower speed. This hypothesis deals with recalibration of walking speed (Decety et al., 1989, Kunz et al., 2009), and tests how additional physical effort applied to podokinetic information can affect imagined walking speed. We used an implicit measure of perceived walking speed by asking participants to estimate distances and travel times from their route planning phase. This should allow us to get rid of biases such as individual variability of estimations and explicit strategies (oriented responses about speed). Moreover, this measure is likely to provide additional information about metric properties of the spatial representation, as slower/higher implicit walking speeds can come from distorted distances.

Our second hypothesis is that individuals tend to mentally simulate their walk during route planning. Modulating walking effort should prompt them to perceive their own walking speed as slower, hence leading them to spend significantly more time planning their route. This would be in favor of an influence of walking effort perception on mental scanning time. Moreover, it would add new elements to the route planning heuristics literature: as already mentioned, the anticipation of navigational cost through planning heuristics has up to now only been hypothesized by Brunyé et al. (2012).

A third hypothesis is that individuals tend to simulate their walk during recall of spatial representations. The increase in perceived walking effort (and therefore also decrease in perceived walking speed) should then negatively affect parameters of motor simulation during a spatial recall task, resulting in lower scores in a spatial retrieval task. In order to test this hypothesis, we created a landmark position retrieval task on a map, where participants were asked to retrieve the spatial positions of previously encoded landmarks from their spatial representation. This should help us specify results from the active learning literature, showing that not only podokinetic information but more specifically effort perception has an effect on the acquisition of spatial representation.

Section snippets

Participants

50 undergraduate students (36 females and 14 males) were recruited from the Faculty of Psychology of Paris Descartes University for the experiment. Mean age was 23.77 (SD = 5.51). Among these 50 participants, 3 were excluded from the results for not following instructions. The 47 remaining participants were randomly assigned to one of the two independent groups formed by students wearing weighted ankle belts (Loaded group = 24 participants) or non-weighted ankle belts (Non-Loaded group = 23

Results

All analyses were performed using R software 3.4.0. In order to confront our hypotheses, we examined the effect of ankle weights on three different measures: implicit walking speed, route planning time and landmark position retrieval performance. All hypotheses were tested with multiple linear regression models, using weighting condition (loaded vs non-loaded ankle weights) as a first categorical predictor. As general sense of direction or spatial abilities can predict the quality of visual

Discussion

This study aimed at demonstrating the influence of walking effort perception on route planning, as a means to evidence the role of motor simulation in such a complex cognitive activity. The general idea was that motor simulation is used during route planning to anticipate the navigational effort of possible routes, implicitly using parameters derived from podokinetic information to take the physical effort into account through recalibration processes. Results showed that subjective walking

Conflicts of interest

None.

Acknowledgment

This work was funded by the French National Research Agency (Grant ANR-13-APPR-0009 to Valérie Gyselinck).

References (44)

  • M. Bhalla et al.

    Visual–motor recalibration in geographical slant perception

    Journal of Experimental Psychology: Human Perception and Performance

    (1999)
  • E.R. Chrastil et al.

    Active and passive contributions to spatial learning

    Psychonomic Bulletin & Review

    (2012)
  • L. Dutriaux et al.

    Cognition incarnée: Un point de vue sur les représentations spatiales

    L'Année Psychologique

    (2016)
  • L. Dutriaux et al.

    Learning is better with the hands free: The role of posture in the memory of manipulable objects

    PloS One

    (2016)
  • N. Franklin et al.

    Searching imagined environments

    Journal of Experimental Psychology: General

    (1990)
  • A.M. Glenberg et al.

    Grounding language in action

    Psychonomic Bulletin & Review

    (2002)
  • R.G. Golledge

    Path selection and route preference in human navigation: A progress report

  • S.C. Grant et al.

    Contributions of proprioception to navigation in virtual environments

    Human Factors

    (1998)
  • E. Grison et al.

    Exploring factors related to users' experience of public transport route choice: Influence of context and users profiles

    Cognition, Technology & Work

    (2016)
  • V. Gyselinck et al.

    Working memory components and imagery instruction in the elaboration of a spatial mental model

    Psychological Research

    (2007)
  • B. Hommel

    Goal-directed actions

    Handbook of Causal Reasoning

    (2017)
  • M. Jeannerod

    The 25th bartlett lecture

    The Quarterly Journal of Experimental Psychology: Section A

    (1999)
  • View full text