Much has been written about the nonspatial functions of the parietal lobe, but these nonspatial functions are rarely evaluated as to whether they are also nonmotoric or reflect some covert form of spatial attention. Establishing whether the parietal lobe has truly nonmotoric and nonspatial functions is essential for understanding why parietal cortex appears to be involved in so many different tasks. Here I’ve attempted to evaluate whether the parietal lobe truly has nonspatial, and perhaps also nonmotoric functions by reviewing the relevant literature…
Magnitude processing – the size congruity effect. The size congruity effect refers to facilitation and interference observed when subjects must name the physically larger of two digits with different numerical values appearing in different size fonts. Facilitation is observed in the form of speeded reaction times when physical size and numerical value are congruent between the two digit pairs (e.g., a large “5″ and a small “2″), and interference is observed when the two are incongruent (e.g., a small “5″ and a large “2″) relative to a neutral condition (e.g., “5″ vs “2″ in the same size font). Neuroimaging shows that the intraparietal sulcus is more active for congruent pairs than incongruent pairs, despite the time-on-task difference between conditions, indicating that IPS may be a common locus for magnitude processing (whether of physical size or of numerical value), in line with Walsh’s hypothesis for domain-general representations of magnitude in the parietal cortex. Recently, Kadosh et al. have shown that IPS doesn’t merely correlate with magnitude processing but is in fact critical for it: disruption of right (but not left) IPS activity through TMS reproduces the deficits observed in dyscalculic populations, in which no facilitation is observed when size and magnitude are congruent. This latter result rules out a purely motoric role for parietal cortex in the size-congruity effect: it is not the case that more parietal activity merely speeds responding in a fashion that is independent of magnitude, since the facilitation effect was not observed after TMS.
Numerical processing – the number line. Even when subjects are given no task, the appearance of numbers triggers activity in IPS, and IPS activity appears to habituate to the particular values used, as indicated by a notation-independent but distance-dependent rebound effect observed in IPS in response to numbers that are numerically deviant (shown by Piazza et al). This data again rules out purely spatial or motoric roles for parietal cortex, since subjects are not given a task. Similarly, as reviewed by Hubbard et al., patients with parietal damage are impaired at reporting the numerical midpoint between two numbers: with numbers like 3 and 12, these patients tend to respond with numbers larger than 7.5, whereas with smaller numbers and inter-number intervals (e.g., 1 vs 2) these patients tend to respond with numbers smaller than the actual numerical midpoint. These symptoms are observed only among patients with parietal damage who show spatial neglect, and like spatial neglect, it can be remediated when leftward-adapting prismatic glasses are worn. Again, this rules out a purely spatial role for parietal cortex, since there is no apriori reason for a spatial manipulation (like the wearing of prismatic glasses) to influence magnitude processing – unless there is some intrinsic connection between how the two are represented in the brain.
Detection of salient items in a sequence – Attentional Blink and Oddball tasks. The temporo parietal junction, inferior to the intraparietal sulcus, has been consistently activated in studies of the attentional blink (in which subjects cannot report the second of two rapidly presented stimuli, if the second follows the first within a window of ~100-500ms). Similar activity is observed in oddball paradigms, which also require detection of one event embedded in a sequence of other events – even when the tasks require no eye or limb movements nor any (spatial) shifts of attention (according to Husain & Nachev). Likewise, Wojciulik & Kanwisher showed that both posterior & anterior sections of the intraparietal sulcus were active in an RSVP task).
Working memory maintenance. The superior parietal lobe, as well as the intraparietal sulcus, have been implicated in the maintenance of information over delays; in the most compelling finding from this literature, individual differences in working memory capacity can be predicted by activity in this region of the brain. Kawasaki et al have shown that SPL activity also occurs regardless of whether the information maintained in working memory is color, shape, or motion-related, and that it is memory-load dependent in a way that cannot be explained by motor intentions or preparation (using a single-item change detection paradigm). Xu has also shown that unlike the inferior parietal lobe, the superior intraparietal sulcus shows activation profiles that closely match the total amount of feature information being maintained in memory (e.g., color & shape information) whereas inferior parietal activity more closely tracks the total number of integrated objects that are being maintained in memory (e.g., Xu & Chun 2006).
Selective Attention and Orienting. Closely related to the “oddball” section above, Humphreys et al. examined patients with parietal lobe damage (and signs of Balint’s syndrome) and showed that they are impaired not only at detecting stimuli in their contralesional visual field, but also at perceiving words in the presence of pictures: regardless of whether the words were presented above or below the pictures, they cannot be detected when pictures are also present. This holds for pictures of real life objects as well as novel shapes, which Humphreys et al interpret as reflecting the parietal lobe’s capacity for remediating nonspatial attentional biases. (It is possible, though, that parietal cortex overcomes these nonspatial biases with a spatially-based representation).
There are no doubt many more examples of nonspatial, nonmotoric functions of the parietal lobe. The crucial question is, however, why the parietal cortex should show such flexible information processing given that it is clearly so important for spatial and motoric functions. In other words, what aspect of the parietal cortex’s capacity for motoric & spatial processing yields it a capacity for active maintenance, and the processing of magnitude and order information?