The drawings we collected above are just one way to demonstrate the spatial deficit in people with WS. This deficit can also be seen in other tests of visuo-spatial construction ability such as the block construction test which is a component of many IQ tests. Examples of WS adult performance on this test are shown here. In general, adults with WS score in the first percentile on this test. We have also collected eye movements from children with WS while they attempted to solve a computer-based version of the block construction test in order to determine if their difficulties could be attributed to a failure to fixate the model pattern. You can view a video showing a playback of these fixations from one WS participant. This video illustrates a couple of important points. First, WS children did fixate the model; in fact, they fixated it as often as mental age-matched control children for models containing 4 or fewer blocks. Second, WS children recognized that their constructions were incorrect but didn't know how to fix them (see Hoffman, Landau, and Pagani, 2003).

The fact that the WS children seemed to recognize that their constructions did not match the model suggests that their perception of the world may be fine and that the basis of their visuo-construction deficit may lie in some other cognitive process. We have subsequently examined several task that indicate that their ability to see how parts are organized into wholes is intact. A good example of this sort of task is biological motion perception in which lights are attached to various  joints of a person who is filmed while walking (a so-called point-light walker or PLW), dancing or performing various other actions on a dark stage. When the first frame of such a movie is viewed, the pattern of lights appears to be just that, a collection of individual lights. However, once the movie is presented, the lights suddenly cohere into the unmistakable percept of a person walking. This task requires the observer to integrate the motion of multiple lights into a coherent percept of and if people with WS have difficulty in perceptually  organizing parts into wholes, we might expect that they would have trouble correctly seeing these stimuli. In our lab, Jordan et al. asked children with WS to determine whether a point light walker was walking towards the left or right. We added various amounts of noise to the stimuli to make the task difficult (press here to see an easy condition and here to view the most difficult one). Surprisingly, WS children were as good or better than controls at this task. Reiss, Hoffman, and Landau (2005) extended these findings to other motions tasks showing that perception of motion coherence was also preserved in children and adults with WS although they also found that performance on a so-called 2D form-from-motion task was impaired in WS, possibly due to difficulty in segmenting targets from backgrounds.

Landau, Hoffman, and Kurtz (in press) reported that identification of common objects was also preserved in children with WS. In this task, the difficulty of object identification was varied by showing the object in unusual or noncanonical orientations and by blurring it (see below; press here to see the whole set of objects). Overall, WS and control groups performed quite well in this task with control children showing a slight advantage for noncanonical orientations and WS children having a comparable edge in identifying blurred objects.  A follow-up task used line drawings in place of full color pictures in order to to eliminate cues to identity based on color or texture. The results were similar.

One might question whether identification of objects, even those presented as line drawings, depends on the correct perception of the arrangement of object parts. In order to address this, we (Hoffman and Landau, in preparation) examined performance in a matching task requiring the correct perception of the spatial arrangement of parts within an object. To eliminate the effects of prior experience, we used novel objects which were constructed by rearranging and retexturing the parts of real objects (see example below). Participants were required to match a target object to one of three alternatives. Crucially, all three alternatives had the same parts but in different spatial arrangements. In the perception condition, the target object remained on the screen with the choices until response. In the memory condition, the target  was removed and the alternatives were presented one second later. In both conditions, WS and controls had the same accuracy suggesting that people with WS can perceive and remember (at least for a short time) the spatial arrangement of parts.

Very different results were obtained however when the nature of the discrimination was changed. In this series, participants were required to discriminate the "handedness" of the object rather than the spatial arrangement of it's features (see example below). For most object recognition tasks in the real world, handedness is largely irrelevant. We don't care whether the cup is seen with the handle on the left or right; it is still a cup and we need to recognize it as such. In fact, one of the difficult issues in modeling object recognition is how to ignore information about orientation. Other research indicates that an object's orientation is indeed separable from the object's shape. For example, damage in certain areas of the parietal lobe can result in a selective inability to discriminate handedness while preserving the ability to recognize the object (Davidoff & Warrington, 2001). Results showed that WS participants had trouble making this discrimination particularly when holding the target object in memory, in which case performance was close to chance.

Our current hypothesis is that spatial deficits associated with WS may be due to delayed and arrested development of certain structures in the dorsal stream of the visual system, particularly regions of the parietal cortex that are concerned with a wide variety of visual functions including mirror image discrimination and  visually guided action. In contrast, the ventral visual stream, which is concerned with recognition of objects and faces, appears to be largely intact in people with WS, consistent with their ability to identify faces and common objects. Our current research (supported by NINDS RO1 NS 050876) is designed to track the development of visuo-spatial functions of these two streams over the course of development to see if this hypothesis continues to provide a good account of the nature of spatial deficits in this fascinating population.