Introduction

By the 1880's, it was well known from experiments performed by David Ferrier [1] and Hermann Munk [2] that large portions of the primate and mammalian cerebral cortex were specifically devoted to visual and auditory functions. By the 1990's, extensive neurophysiological research [3]–[5] had identified and localized an abundance of sensory cortical areas: in the macaque monkey, for example, there were at least 25 distinct cortical areas devoted to vision, at least 8 devoted to somatosensory functions, and at least 5 devoted to audition. The evidence for these functional subdivisions of cerebral cortex is compelling. At the same time, however, this division of functionality is somewhat of an oversimplification. More recent research [6]–[12] has proven that the cerebral cortex is plastic and that significant changes in neuronal responsiveness to sensory information occur with long-term visual deprivation (e.g., blindness). For example, Kahn and Krubitzer [6] surgically removed both eyes from opossum pups at 4 days of age (it is important to keep in mind that opossums are mammals, with a cerebral cortex that is similar to other mammals). When the enucleated opossums' cortical functioning was assessed at adulthood, neurons in what would have been their primary visual cortex responded either exclusively to tactile input, exclusively to auditory input, or to both tactile and auditory input. Similarly, Lewis, Saenz, and Fine [7] have demonstrated that neurons within the occipital cortex of early-blind humans respond to both tactile and auditory stimuli. The occipital response to non-visual stimulation was much greater for their blind participants and was substantially reduced in sighted participants. It is clear that long-term visual deprivation (e.g., blindness), at least, is associated with substantive changes in cortical functioning.

Many studies have now found that blindness is associated with enhancements in behavioral performance for non-visual sensory and perceptual tasks [13]–[22]. Our own laboratory [13], for example, has recently demonstrated that blindness is associated with enhancements in both tactile acuity and haptic solid shape discrimination. The cortical changes in functionality that accompany blindness probably contribute to such enhancements. Blindness-associated improvements in performance occur for tactile and haptic tasks [13]–[17], as well as for those involving audition [18]–[20] and olfaction [21], [22]. If long-term visual deprivation (months or years) can enhance sensory performance in non-visual modalities, what length of deprivation is required? The evidence acquired to date suggests that the duration of deprivation can be surprisingly short. Medium-length visual deprivation of 5 days to a week has been shown to produce enhancements in performance for tactile tasks [23]–[25]. Given that a period of 5 days is probably insufficient for significant changes in patterns of neuronal connectivity per se, Merabet et al. [24] suggested that the improvement obtained for their tactile task was likely caused by “an unmasking of latent multimodal connections underlying the recruitment of occipital cortex for tactile processing”. In this view, there are pre-existing tactile inputs to “visual” cortex in all of us, but in everyday life visual input dominates and “masks” the tactile input. When visual deprivation occurs (for example, by blindfolding for hours or days), the tactile input to visual cortex is “unmasked” – the “visual” cortex then becomes responsive to tactile input and contributes to the performance obtained for tactile behavioral tasks.

Evidence now exists to indicate that performance for some tactile tasks can be facilitated by only a few hours of visual deprivation. For example, Weisser et al. [26] found that participants who had been blindfolded for two hours exhibited superior performance (11.9 percent higher accuracy) relative to non-visually-deprived participants for a tactile task involving the discrimination of 2-D shape (i.e., discriminating between upside-down letters “T” and “V”). Whether similar improvements in tactile acuity can occur following short-term visual deprivation is not yet clear. Facchini and Aglioti [27] found that their participants' grating orientation discrimination thresholds decreased by about 21 percent after a 90-minute period of visual deprivation – this decrease did not occur for their non-deprived participants. Leon-Sarmiento et al. [28] similarly found grating orientation thresholds to decrease (by 23 percent) after 45 minutes of visual deprivation. Concern has been expressed [29] about the validity of the findings of both Facchini and Aglioti [27] and Leon-Sarmiento et al. [28]. Wong et al. [29], for example, pointed out that the “non-deprived” participants in the study by Facchini and Aglioti [27] were actually blindfolded and visually deprived whenever their tactile acuity was being measured. These researchers [29] argued that the performance of Facchini and Aglioti's “non-deprived” participants was not therefore necessarily representative of how they might have actually performed under non-deprived (i.e., sighted) conditions (which was never tested). Wong et al. [29] also pointed out that there was no control group in the study by Leon-Sarmiento et al. [28]. The tactile acuity of their visually deprived participants did improve after 45 minutes of blindfolding, but because there was no control group of non-deprived participants with which to compare, this improvement could have occurred simply as a result of practice and increasing experience with the task. In their own study (which did employ control groups of non-deprived participants, who were tested under sighted conditions), Wong et al. [29] found no improvement in grating orientation discrimination following up to 110 minutes of visual deprivation. In the current experiment, we sought to determine whether there is or is not an effect of short-term visual deprivation on tactile acuity. In addition, we evaluated the possible effect of short-term deprivation on the haptic discrimination of solid object shape. No study to date has simultaneously investigated the potential effects of short-term visual deprivation upon both tactile acuity and haptic shape discrimination.

Materials and Methods

Experimental Stimuli

The experimental stimuli for the tactile acuity task were tactile gratings (JVP Domes, Stoelting, Inc.), used previously by both our laboratory and others [13], [15], [24], [30]–[33]. The experimental stimuli used for the solid shape discrimination task were solid (plastic) copies of eight natural bell peppers, Capsicum annuum, that have been regularly used in previous investigations [13], [34]–[37]. Photographs of these stimulus objects are shown in Figure 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. A photograph of the eight natural objects (bell peppers, Capsicum annuum) used as experimental stimuli for the solid shape discrimination task.

Starting from the bottom left (going clockwise), the objects depicted are 1, 2, 3, 5, 7, 8, 11, and 12. These objects are a subset of those developed by Norman et al. [30].

https://doi.org/10.1371/journal.pone.0112828.g001

Results

The participants' tactile acuities are plotted in Figure 2. According to the Shapiro-Wilk test for normality [40], the sighted and blindfolded participants' grating orientation thresholds were distributed normally for both the initial and second assessment of tactile acuity (i.e., the participants' thresholds did not deviate from normality: sighted, initial assessment, W(14)  = 0.94, p>.05; sighted, second assessment, W(14)  = 0.92, p>.05; blindfolded, initial assessment, W(14)  = 0.96, p>.05; blindfolded, second assessment, W(14)  = 0.93, p>.05). As can be readily seen from an inspection of Figure 2, there was a significant decrease in the participants' thresholds from the initial assessment to the second assessment (F(1, 26)  = 12.9, p = .001, partial  = .33). This improvement in performance for grating orientation discrimination is similar to that obtained for a different tactile task (grating groove width and ridge width discrimination) by Sathian and Zangaladze [41]. In their study, the participants' improvement in performance occurred for a dynamic tactile task (participants actively scanned grating surfaces), while our participants' improvement occurred for static touch. As Figure 2 clearly shows, there was a significant improvement in tactile acuity (i.e., lowered thresholds during the second assessment) for the blindfolded participants (open bars); however, there was an equal improvement in tactile acuity for the sighted control participants (filled bars). This equivalence in performance change for both groups of participants is reflected by a non-significant interaction between group and measurement session (F(1, 26)  = 0.007, p = .93, partial  = 0.0). When considering this result, it is important to keep in mind that there was sufficient power to detect an effect of the visual deprivation had one existed: we had, for example, just as many sighted and visually-deprived participants as in the study by Facchini and Aglioti [27]. Furthermore, given the means and inter-participant variability shown in Figure 2, a power analysis [42] revealed that to have a 90 percent chance of detecting an effect this small (i.e., a difference in tactile acuity between the sighted and visually-deprived participants in the second assessment), we would need 8450 participants per group (16,900 total participants). It is clear that even if the extremely small difference in tactile acuity we observed between the sighted and deprived participants had been statistically significant (following a 90 minute period of light deprivation for the blindfolded participants), it would not be meaningful or of practical importance.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Experimental results.

The participants' tactile acuity (grating orientation thresholds) measured both before and following a 90-minute period. Half of the participants were visually deprived of light (i.e., were blindfolded) during this 90-minute period. The filled bars indicate results obtained for the sighted participants, while the open bars indicate results obtained for the blindfolded (i.e., visually deprived) participants. The error bars indicate ±1 SE.

https://doi.org/10.1371/journal.pone.0112828.g002

The results for the solid shape haptic discrimination task are shown in Figure 3. While the blindfolded participants' shape discrimination accuracies were normally distributed (i.e., did not deviate from a normal distribution, W(14)  = 0.97, p>.05), the sighted participants' shape discrimination accuracies were not normally distributed (W(14)  = 0.76, p<.05). The deviation of the sighted participants' shape discrimination accuracies from a normal distribution largely occurred because of a single outlier: one sighted participant obtained a d′ score of 0.924, which was 2.96 standard deviations lower than the mean of the other 13 sighted participants (2.021). One, of course, could potentially exclude the outlier. However, we prefer to include all of the participants' data. Even though it has been demonstrated that parametric t-tests are ordinarily quite robust to violations of their assumptions [43], we chose to employ a distribution-free nonparametric analog of the t-test, the Wilcoxon Rank-Sum Test [44], [45], to test for any potential difference between the sighted and blindfolded groups of participants. It is clear from a visual inspection of Figure 3 that there was no significant difference between the shape discrimination abilities of the visually-deprived and sighted participants, and this was verified by the Wilcoxon Rank-Sum Test (W_x  = 200, p = .91, 2-tailed). A power analysis [42] was conducted upon the results (difference in d′ scores) depicted in Figure 3. The analysis revealed that (in the best case) we would have needed a total sample of almost three million participants (2,835, 654) in order to have a 90 percent chance of detecting a difference this small (the average d′ scores for haptic shape discrimination in our experiment were 1.942 and 1.941 for the sighted and visually-deprived participants, respectively).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Experimental results.

The participants' solid shape discrimination accuracies are plotted for both the sighted and blindfolded (i.e., visually deprived) participants. The error bars indicate ±1 SE.

https://doi.org/10.1371/journal.pone.0112828.g003

Discussion

Prior research by four different laboratories, including our own [13]–[16], has demonstrated that blindness is associated with increased tactile acuity. It seems clear, therefore, that long-term visual deprivation can produce significant enhancements in tactile acuity. Fifty years ago, Zubek et al. ([23], p. 1592) reported that a week of visual deprivation produced “a pronounced increase in tactual acuity of the palm”; in particular, the 2-point thresholds of their participants declined by about 24 percent over the course of seven days. Recent investigators [46], [47] have concluded, however, that the measurement of 2-point thresholds does not constitute a valid measurement of tactile acuity. Therefore, we do not necessarily know at the present time whether a week of visual deprivation truly produces enhancements in tactile acuity (five days, however, of visual deprivation is sufficient to improve performance on a tactile Braille character recognition task [24], [25]).

In 2008, Leon-Sarmiento et al. [28] reported that the grating orientation discrimination thresholds of their “neurologically normal subjects” decreased by 23 percent following 45 minutes of visual deprivation. It is very important to note, however, that these investigators did not compare the performance of their visually deprived participants with similar control participants who were not visually deprived. It is thus likely that their participants' tactile acuity improved during the second measurement session as a result of practice (just as our participants' tactile acuity improved through practice and increasing experience with the task, see the current Figure 2) and not as a result of the 45 minutes of deprivation. Facchini and Aglioti [27] did compare the tactile acuity of visually deprived participants with a group of “non-deprived” participants; these investigators found a decrease in grating orientation discrimination thresholds from the first to second measurement session for their blindfolded participants, but not for the “non-deprived” control participants. The report of a more recent study [29], however, noted that Facchini and Aglioti visually deprived their “non-deprived” participants whenever their tactile acuity was being assessed – these authors (Wong et al.) then argued that the performance of Facchini and Aglioti's “non-deprived” participants was not necessarily representative, because these participants had not been tested under sighted (i.e., non-deprived) conditions. One puzzling aspect of the data obtained by Facchini and Aglioti (see their Figure 2) is that the tactile acuity of their non-deprived participants at the baseline initial measurement was already nearly as good as that (very similar, since the two error bars overlap substantially) exhibited by the deprived participants after 90 minutes of blindfolding. If 90 minutes of total light deprivation was needed for the deprived participants' tactile acuity to improve to a particular level, how could the non-deprived participants perform at a statistically similar level with no light deprivation? The results shown in Figure 2 of Facchini and Aglioti certainly indicate that 90 minutes of visual deprivation is not needed to obtain superior tactile acuity (since the “non-deprived” participants at baseline performed similarly to the deprived participants after 90 minutes of blindfolding).

The results of the current study were very different from those of Facchini and Aglioti [27]; in our experiment, the grating orientation discrimination thresholds became significantly lower during the second measurement session for both the blindfolded and sighted participants (see the current Figure 2). Our current results are, however, entirely consistent with those of Wong et al. [29]. Both our study and that of Wong et al. found that short-term visual deprivation (e.g., 90 minutes) does not lead to significant improvements in tactile acuity over and above the improvement exhibited by non-deprived control participants. Our investigation extended the study of Wong et al. by additionally evaluating the potential effects of short-term visual deprivation upon haptic solid shape discrimination. No previous psychophysical study has investigated deprivation and solid shape discrimination. As Figure 3 clearly indicates, there was no effect – blindfolded and sighted participants possessed equal solid shape discriminability.

As reviewed earlier, Weisser et al. [26] found that two hours of visual deprivation was sufficient to produce significantly higher performance (than that obtained for non-deprived control participants) for a tactile task involving the discrimination of 2-D shape (discriminating between upside-down letters “T” and “V”). The most robust difference in task-related cortical activity between their deprived and control participants occurred in visual extrastriate areas V3A and vIPS. Neurons within V3A are responsive to sharp edges or discontinuities in depth [48] (for their stimuli, Tsao et al. [48] used random-dot stereograms that depicted “checkerboard” patterns; prominent discontinuities in depth occurred at the linear boundaries between pairs of checks). In this context, it is probably important to note that the “V” and “T” stimuli used by Weisser et al. had polygonal shapes with distinct depth edges (there was a discontinuity in depth, 3.5 mm, between the V and T shapes and their background). The 3-D (solid) shapes that were used in the current experiment, bell peppers, were quite different – the surfaces of these objects are smoothly curved and continuous everywhere, even at the outer boundaries (i.e., the orientations of local surface regions change smoothly at the outer boundaries). It is possible that our smoothly-curved 3-D objects (lacking sharp discontinuities in depth at the outer boundaries) do not alter neuronal activity in V3A in the same way as the 2-D stimuli used by Weisser et al. (thus leading to differences in behavioral performance). Another obvious difference between the current study and that of Weisser et al. is that in the current experiment, our participants actively manipulated the bell pepper stimuli in order to make their judgments (a haptic judgment), whereas the tactile stimuli in the Weisser et al. study were manually pressed into the participants' immobilized right index fingerpad (passive tactile judgments). It is likely that active haptic manipulation and passive tactile judgments differentially activate many cortical regions, potentially leading to differential performance. In any event, it seems clear that short periods of visual deprivation can heighten the tactile discrimination of 2-D shape. Brief periods of visual deprivation also appear to produce enhancements for a number of auditory tasks, including the auditory ability to estimate distances to sound sources [49] and sound localization [50]. The results of the current study and that of Wong et al. [29], however, indicate that tactile grating orientation discrimination and haptic solid shape discrimination tasks are different, and are not appreciably affected by short periods of visual deprivation.

Supporting Information