External location of touch is constructed post-hoc based on limb choice

Femke Maij; Christian Seegelke; W Pieter Medendorp; Tobias Heed

are an implicit indicator of precise tactile localization and, thus, tactile remapping. 537 This is a seemingly surprising conclusion given that limb crossing quite obviously 538 creates conflict between the anatomical body side and the side of space, and it is 539 a strong claim that seems to invalidate the experimental logic of numerous papers 540 that have applied this logic. Our conclusion is, however, in fact compatible with 541 that of another recent study in which participants performed TOJ of tactile stimuli 542 presented to the hands and feet (Badde et al., 2019) . In that study, stimuli were 543 presented to two of the four limbs (hands and feet), but it was unknown to the 544 participant which two limbs would be stimulated in a given trial. Accordingly, four 545 rather than just two response options were always available. In a percentage of 546 trials, participants systematically assigned the first touch to a limb that had not 547 received a tactile stimulus at all during the respective trial. For instance, after 548 stimulation of the left hand and the right foot, a participant may have indicated that 549 the first stimulus had occurred on the right hand. These errors depended on 550 whether the arms and/or legs were crossed. Importantly, neither the side of space 551 of the limb that had received the correct stimulus, nor the spatial distance between 552 stimulus and response limb, were indicative of the erroneously chosen, 553 unstimulated limb -again incompatible with the prevailing view that crossing 554 effects reflect conflict during the integration of anatomical and spatial stimulus 555 location. Instead, participants responded with unstimulated limbs that occupied the 556 space anywhere around the body side of the correct, first tactile stimulus, 557 independent of whether that limb was crossed or not. Thus, what mattered for limb 558 assignment was not the precise, spatial location of the stimulus, but much broader, 559 global spatial stimulus features such as an entire side of the body and its 560 surrounding space. Critically, these global features explained behavior without 561 . CC-BY-NC-ND 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/549832 doi: bioRxiv preprint having to recur to the use of external-spatial coding of touch location, consistent 562 with the present finding that the latter is only computed when necessary. 563