One was a similar spatial ‘preference’ task, with no right or wrong answer,

but employing non-face stimuli, namely greyscale gradient rectangles (see Fig. 3C). In analogy with the chimeric face preference task, in this greyscale gradient task the patients were presented with pairs of identical Src inhibitor left-right mirror-reversed greyscale rectangles, ranging from pure white at one end to pure black at the other end and were asked to indicate which one (upper or lower) seemed ‘darker’ to them. This task has been previously used to assess spatial biases in both normal subjects and neglect patients (e.g., Mattingley et al., 1994, Mattingley et al., 2004 and Loftus et al., 2009). Just like for the chimeric face lateral preference task, neglect patients tend to show a strong rightward bias in this greyscale task and normals tend to show a mild bias towards the left. Of particular relevance here is that this well-established greyscale task should presumably

not involve any face-specific or emotional processing mechanisms. The final task implemented here used chimeric face stimuli, but now requiring ‘explicit’ identification of the relationship between the left and right sides of the chimeric face tasks (objective discrimination between ‘chimeric’ and ‘non-chimeric’ face stimuli, see Fig. 3B). Unlike the greyscale or face lateral preference tasks, this task is unambiguous in having a single objectively correct response (rather Rapamycin manufacturer than merely requiring a choice between left/right mirror-imaged enough pairs) and in explicitly measuring awareness for the contralesional side, rather than indirectly via spatial preferences. We note also that it does not require any emotional assessment of the stimuli. If there is something special about prism adaptation effects on face-specific processing mechanisms, we might find a prism benefit on neglect for the greyscale lateral preference task, but not for the other two tasks that do employ faces (expression lateral

preference or chimeric versus non-chimeric discrimination). Alternatively, if prism adaptation is ineffective only in tasks that involve emotional processing in particular, we should again expect no prism benefit for the chimeric expression task, but we should find a benefit for the other two tasks (greyscale lateral preference, and chimeric/non-chimeric discrimination of faces), since they do not require emotional processing of the stimuli. Finally, if prism therapy can influence face-related mechanisms, but does not affect spatial preference biases, we should expect no prism benefit in either of the two lateral preference tasks (face expressions or greyscale gradients), yet could potentially find some prism benefit for the chimeric/non-chimeric face discrimination task. A series of eleven consecutive right-hemisphere stroke patients with left neglect were recruited for this experiment (7 males).