Elsevier

Cognition

Volume 83, Issue 2, March 2002, Pages B25-B34
Cognition

Brief article
Tool-use changes multimodal spatial interactions between vision and touch in normal humans

https://doi.org/10.1016/S0010-0277(02)00003-3Get rights and content

Abstract

In a visual–tactile interference paradigm, subjects judged whether tactile vibrations arose on a finger or thumb (upper vs. lower locations), while ignoring distant visual distractor lights that also appeared in upper or lower locations. Incongruent visual distractors (e.g. a lower light combined with upper touch) disrupt such tactile judgements, particularly when appearing near the tactile stimulus (e.g. on the same side of space as the stimulated hand). Here we show that actively wielding tools can change this pattern of crossmodal interference. When such tools were held in crossed positions (connecting the left hand to the right visual field, and vice-versa), the spatial constraints on crossmodal interference reversed, so that visual distractors in the other visual field now disrupted tactile judgements most for a particular hand. This phenomenon depended on active tool-use, developing with increased experience in using the tool. We relate these results to recent physiological and neuropsychological findings.

Introduction

When we wield a long tool, we extend our possible reaching space. We can touch distant objects with the tool, so that tactile information felt at the hand can now relate to visual information from distant objects (Berti and Frassinetti, 2000, Farnè and Ladavas, 2000, Iriki et al., 1996, Maravita et al., 2001). Moreover, a hand located on, say, the right side of space may contact visual objects in the left visual field with the tool, or vice-versa, given that tools can be wielded in innumerable different postures. This means that the spatial mapping between tactile stimulation at the hand, and any related visual information near the end of the tool, alters as a tool is used. A wielded tool may become incorporated into the ‘body schema’, such that the end of the tool effectively becomes an extension of the effector wielding it. Here we examined whether experience in actively wielding a long tool can modulate automatic aspects of tactile–visual spatial integration for normal human subjects. We discuss later how such a phenomenon might relate to recent physiological data on plasticity induced by tool-use (Iriki et al., 1996, Iriki et al., 2001), and to recent neuropsychological studies on this issue (Berti and Frassinetti, 2000, Farnè and Ladavas, 2000, Maravita et al., 2001).

We exploited a visual–tactile interference paradigm previously used to study tactile–visual spatial integration in humans (Pavani et al., 2000, Spence et al., 2001). Subjects had to judge whether tactile vibrations were delivered to the thumb or index finger (equivalent to upper or lower locations here) on either hand. Visual distractor lights were fixed in vertical pairs at the far end of each of two “tools” grasped with each hand (Fig. 1). On each trial, a vibration from one of four possible locations (a finger or thumb, on the left or right hand) was presented concurrently with one of the four possible distractor lights (upper or lower, on left or right). All of the possible crossmodal pairings were equiprobable, so the visual distractors gave no information about the concurrent tactile target, and were simply to be ignored.

Previous work using visual distractors located on the hands (Pavani et al., 2000, Spence et al., 2001) showed that judgements of such tactile stimuli are slower and/or less accurate when the concurrent visual distractor is incongruent (i.e. upper vibration with lower light, or vice-versa). Importantly, this crossmodal interference has repeatedly been found to be stronger when visual and tactile stimuli appear on the same side of space (i.e. right visual field stimulation paired with vibration on the right hand, or left visual field stimulation with left hand vibration). Here we tested whether actively wielding a long tool can alter this crossmodal mapping plastically, such that when the tools are held crossed (connecting the right hand to the left visual field, and vice-versa), crossmodal interference might now become larger from visual distractors in the opposite visual field to the tactually stimulated hand.

Subjects held two long tools straight on some trials (Fig. 1a), while on other trials they crossed the far tips of the tools, while keeping the locations of the hands unchanged (Fig. 1b). Tool posture was actively changed by the observer every four trials. We hypothesized that prolonged active use of the tools in this way might alter the spatial pattern of crossmodal integration observed. Crossmodal interference should now be strongest for visual distractors currently connected to the hand by the tool. Critically, this should be observed not only in the straight posture, where distractors and hands are on the same side of space (cf. Pavani et al., 2000, Spence et al., 2001), but now also in the crossed posture where distractors are on the opposite side of external space relative to the hand.

Section snippets

Participants

Twenty healthy participants (mean age 25 years) took part.

Apparatus

The ‘tools’ were toy golf-clubs, 75 cm long, with a vertical peg incorporated into the far end. A pair of Oticon-A bone-conduction vibrators served as tactile stimulators on the handle of each tool, one below the thumb and the other under the forefinger pad on each hand (Fig. 1a). With this arrangement, each vibrator occupied either an upper (thumb) or lower (forefinger) position on the tool handle. Two red turbo LEDs (60 cd/m2) served

Procedure

Participants sat in a sound-proof booth with chin on a rest. On each trial, a tactile stimulus from one of the four vibrators (200 Hz sine-wave signal, in three successive 50 ms bursts, each separated by a 50 ms gap) plus a visual stimulus from one of the four distractors (same duration and frequency as the vibration) were presented. Onset of the visual distractor occurred 30 ms before the tactile target, as pilot work showed this maximizes crossmodal interference. The positions of vibrations

Results and discussion

Fig. 2a,b shows mean results for RTs and error rate in terms of the critical crossmodal interference effect, i.e. difference between incongruent minus congruent stimulus/distractor combinations, as defined above. When the tools were held straight (left pair of bars within Fig. 2a,b), crossmodal interference was stronger for visual distractors in the hemifield of the stimulated hand, confirming previous results (e.g. Pavani et al., 2000, Spence et al., 2001). The critical new result was that

Control study: Experiment 2

To further test the importance of active tool-use for the above results, we conducted a ‘passive’ control study with 20 new participants (mean age 26 years), who did not actively wield the tools. Now the position (straight or crossed) for the tools only changed at the end of each block of 48 trials, rather than every four trials. Moreover, all such changes were now made by the experimenter rather than by subjects.

Results (Fig. 2c,d) revealed that crossed and straight tool positions no longer

General discussion

Our results show for the first time in normal adult humans that prolonged active use of tools (over several minutes) can modify visual–tactile spatial integration. The data are consistent with a progressive functional change in the ‘body schema’, so that with experience in actively using the tools, they became treated as extensions of the effectors wielding them. This produced a stronger integration between visual stimulation at the current location of the far end of the tool, and tactile

Acknowledgements

This research was supported by a Medical Research Council (UK) Programme Grant. J.D. holds a Royal-Society/Wolfson Research Merit Award.

References (12)

There are more references available in the full text version of this article.

Cited by (275)

  • Action does not drive visual biases in peri-tool space

    2024, Attention, Perception, and Psychophysics
View all citing articles on Scopus
View full text