Abstract
Adaptation paradigms are frequently used as a means of assessing the tuning characteristics of underlying neural mechanisms. Here we used the “double-pulse” paradigm developed by Musselwhite and Jeffreys (1983) and Jeffreys (1996) to examine adaptation of early visual responses at posterior and more anterior brain sites using magnetoencephalography (MEG). Our results show a marked difference in recovery rates between the C1, a posterior response at 100 ms arising in occipital cortex, and the M170, a component at 170 ms commonly thought to reflect face perception at occipitotemporal sources. While the C1 re-emerges within a 150 ms SOA of adaptation, the more anterior M170 remains reduced even by 400 ms post-adaptation. This reduction in signal is modulated by the duration of both the adapting stimulus and the inter-stimulus interval, and appears largest for faces immediately preceded by faces, even though the identities of the adapting and test stimuli were never the same. However, a control experiment which reduced the amplitude of the face M170 through the addition of noise suggests that the response, rather than reflecting “face-selective” adaptation, may instead be modulated in a more general manner by the magnitude of the adapting stimulus: when the amplitude of faces and house stimuli were equalized, depression of similar magnitude was found for face responses preceded by house images, which are presumed to activate different neural populations. This finding suggests that, at least at early stages of object processing, the rapid temporal adaptation seen in the double-pulse paradigm may have surprisingly widespread cortical consequences.