On the role of response conflicts and stimulus position for hemispheric differences in global/local processing: an ERP study
Introduction
It has widely been assumed that the left and right cerebral hemispheres (LH/RH) are specialized for processing the local and global levels of hierarchically structured visual stimuli, respectively (e.g. Robertson & Lamb, 1991). For testing this hemispheric specialization compound letters, similar to those depicted in Fig. 1 (cf. Navon, 1977), are often used as stimuli, and the participants have to detect or identify the letter at a pre-specified level. Such stimuli can be applied in combination with several experimental techniques.
For example, in neuropsychological studies it is investigated how unilateral brain lesions affect global and local processing. Their results suggest that damage to the RH impairs the processing of global information, whereas damage to the LH leads to a deterioration of local processing (Delis, Robertson, & Efron, 1986; Lamb, Robertson, & Knight, 1990; Robertson, Lamb, & Knight, 1988). However, unilateral lesions did not always produce the expected neuropsychological impairments. For instance, Polster & Rapcsak (1994) report about two patients who showed normal local processing in spite of massive LH lesions. Also, the results of neuro-imaging studies are inconclusive. Of three studies where positron emission tomography (PET) was used to asses brain activity during global/local processing, only one found effects in the expected direction (Fink, Halligan, Marshall, Frith, Frackowiak, & Dolan, 1996). A second study failed to replicate this finding (Heinze, Hinrichs, Scholz, Burchert, & Mangun, 1998), whereas in a third study the pattern of neural activity was actually reversed to that observed in the first one (Fink, Marshall, Halligan, Frith, & Frackowiak, 1997).
The large variability of these results shows that the conditions under which hemispheric differences for global/local processing occur are still largely unknown. Therefore, the general aim of the present study is to isolate possible favorable factors for this effect. The results should improve our understanding of the mechanisms underlying the hemispheric difference, and could ultimately lead to more precise predictions about the occurrence of hemispheric differences in future investigations.
In the remaining part of this section, we will consider the results of the two most popular experimental techniques. In reaction-time studies, hierarchical letters are either presented to the left visual field (LVF) or to the right visual field (RVF) and the speed and accuracy of the responses are measured. If the proposed hemispheric specialization exists, then local processing should be faster and more accurate for RVF stimuli, which are initially projected to the LH, than for LVF stimuli projected to the RH. The opposite should hold for global processing. We will call respective differences in reaction times and accuracy VF-effects.
Another method that has often been applied for the investigation of hemispheric specialization is the recording of event-related brain potentials (ERPs), which arise from the synchronous activity of neural populations that are engaged into the actual task. If the hemispheres differ in the proposed way, then it can also be expected that ERP amplitudes recorded from the LH are larger in the local compared to the global condition, whereas the reverse pattern should occur in the RH. Such amplitude differences will be called hemispheric asymmetries.
Although VF-effects for global/local processing have been reported in several experiments (Blanca, Zalabardo, Garcia-Criado, & Siles, 1994; Evert & Kmen, 2003; Hübner, 1997), taken together, reaction-time studies provide not much evidence in favor of hemispheric specialization. This conclusion can also be drawn from two meta-analyses showing that reaction-time studies with positive results were nearly balanced (Van Kleeck, 1989) or even outnumbered (Yovel, Yovel, & Levy, 2001) by those where no VF-effects were observed. In contrast, the results obtained in ERP studies are more conclusive. Negative results were reported only in a few studies with atypical geometrical figures as stimuli (Han, Fan, Chen, & Zhuo, 1997; Han, He, Yund, & Woods, 2001; Johannes, Wieringa, Matzke, & Münte, 1996). On the other hand, nearly all published ERP studies with compound letters report hemispheric asymmetries in the expected direction (Heinze & Münte, 1993; Heinze, Johannes, Münte, & Mangun, 1994 (Experiment 1); Heinze et al., 1998; Malinowski, Hübner, Keil, & Gruber, 2002; Yamaguchi, Yamagata, & Kobayashi, 2000).
How can the discrepancy between VF-effects and hemispheric asymmetries be explained? One possible factor that could be responsible for the difference is the stimulus position. In contrast to reaction-time studies, a lateral stimulus presentation is not necessary in ERP studies. Consequently, the stimulus is usually presented to the central visual field (e.g. Heinze et al., 1998, Yamaguchi et al., 2000). That stimulus position might be crucial has also been suggested by Han, Weaver, Murray, Kang, Yund, and Woods (2002). These authors compared ERP studies in which the stimuli were presented centrally with those in which they were presented laterally. If stimulus position matters for the occurrence of hemispheric asymmetries, then the corresponding results should differ. This is indeed the case. Whereas uniform positive results occurred in ERP studies with stimuli in the central visual field (Heinze & Münte, 1993; Heinze et al., 1994 (Experiment 1); Heinze et al., 1998; Malinowski et al., 2002, Yamaguchi et al., 2000), only two out of four studies with laterally presented stimuli found hemispheric asymmetries (Han, Fan, Chen, & Zhuo, 1999; Proverbio, Minniti, & Zani, 1998).
In order to test the effect of stimulus position directly, Han et al. (2002) conducted an experiment in which a compound letter was either presented in the center of the display or in one of the lateral visual fields. As predicted, hemispheric differences showed up only for centrally presented stimuli. The authors explained their result by hypothesizing that the hemispheres compete for the processing of a centrally presented stimulus because both have simultaneous access to the same visual information. This leads to the assignment of more resources to a given target level within the specialized hemisphere. On the other hand, if a stimulus is presented laterally, the hemisphere contralateral to the stimulated hemifield receives the visual information first, whereas the ipsilateral hemisphere receives it only after its transmission through the corpus callosum. The time difference may eliminate the competition between the hemispheres and, accordingly, reduces the associated differences between them (see also Han, Yund, & Woods, 2003). We will call this account the competition hypothesis.
Although formulated for the explanation of hemispheric asymmetries, the competition hypothesis might also be used to account for the absent VF-effects. That is, lateral stimulation might generally be unfavorable for observing effects indicating hemispheric specializations. This would explain the generally weak and equivocal outcomes obtained in reaction-time experiments. However, there is also evidence that reliable VF-effects can be produced if the information on the stimulus levels is conflicting (for a meta-analysis see Van Kleeck, 1989). It seems that congruent stimuli, i.e. stimuli where both levels activate the same response, are less likely to produce a VF-effect than incongruent stimuli, i.e. stimuli where the levels activate different responses. Supporting results have also been reported in a recent paper by Hübner & Malinowski (2002). To explain the effect of response conflict on VF-effects these authors proposed that an elaborated stimulus representation is necessary for resolving the conflict induced by incongruent stimuli. They further assumed that the hemispheres differ with respect to creating such representations, according to the known specialization. Consequently, hemispheric differences mainly show up in the incongruent situation.
Thus, there are two different factors that have been proposed as being favorable for observing effects of hemispheric specialization. On the one hand, a central stimulus position seems to have a positive effect for hemispheric differences. On the other hand, it has been hypothesized that incongruent stimuli are favorable for VF-effects. However, it seems reasonable to assume that each factor should affect measures obtained in the respective other domain as well. That is, stimulus congruency should also affect hemispheric asymmetries. That this is indeed the case has been shown in an ERP study by Malinowski et al. (2002), where the corresponding effects were most pronounced at temporal-parietal electrodes in the early P3 component (320–400 ms after stimulus onset). Since the stimuli were presented to the central visual field and hemispheric asymmetries were nevertheless found only for incongruent stimuli, it follows that a central stimulus presentation cannot be sufficient for obtaining reliable results. Rather, response conflict seems to be the more important factor. This hypothesis would even be further supported by an experiment which shows that a central stimulus presentation is not necessary either.
Such an experiment is reported in the present article. Congruent and incongruent stimuli were either presented to the LVF or RVF and ERPs, reaction times, and error rates were registered. According to the competition hypothesis (Han et al., 2002, Han et al., 2003) no hemispheric asymmetries should occur under these conditions. Furthermore, if we apply this hypothesis to behavioral data, then VF-effects should also be absent. On the other hand, if congruency is the crucial factor, then hemispheric asymmetries as well as VF-effects should be observed, but only (or at least to a larger extent) for incongruent stimuli.
Section snippets
Subjects
Sixteen students of the Universität Konstanz participated in the experiment. Some of them fulfilled course requirements; others were paid a small fee for participation. After a preliminary data analysis, four subjects were excluded because of strong artifacts in the EEG data. The remaining subjects (10 females, two males, mean age: 22 years) were right-handed by self-report, had normal or corrected-to-normal vision, and gave informed consent on the experimental procedure.
Stimuli
The stimuli consisted
Behavioral data
The data were evaluated by means of repeated-measures analyses of variance (ANOVAs) with latencies of correct responses and error rates as dependent variables. The factors were target level (global and local), visual field (LVF and RVF) and congruency (congruent and incongruent).
For the reaction times there was a significant main effect of target level [F (1, 11) = 11.70, P < 0.01], indicating that the responses to the global level (635 ms) were faster than those to the local level (680 ms). Also
Discussion
From the observation that hemispheric asymmetries occurred more often in ERP studies with centrally presented stimuli than in those with laterally presented stimuli, Han, Weaver, Murray, Kang, Yund, and Woods (2002) concluded that a central stimulus position is favorable for these asymmetries. Such an account could also explain that only some of the behavioral studies found VF-effects for global/local processing, because these studies depend on a lateral stimulus presentation. On the other
Acknowledgements
This research was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG) to the second author as part of a research group (Ro 805/11-1). We thank Franka Glöckner and Anja Nörenberg for the data acquisition, and Andreas Keil for his help concerning the handling of the electrophysiological equipment and the ERP-data analyses.
References (29)
- et al.
Hemispheric differences in global and local processing dependent on exposure duration
Neuropsychologia
(1994) - et al.
Hemispheric specialization of memory for visual hierarchical stimuli
Neuropsychologia
(1986) - et al.
Hemispheric asymmetries for global and local processing as a function of stimulus exposure duration
Brain and Cognition
(2003) - et al.
Attentional selection in the processing of hierarchical patterns: an ERP study
Biological Psychology
(2001) - et al.
Hemispheric asymmetry in global/local processing: effects of stimulus position and spatial frequency
Neuroimage
(2002) - et al.
An ERP study of the global precedence effect: the role of spatial frequency
Clinical Neurophysiology
(2003) - et al.
Electrophysiological correlates of hierarchical stimulus processing: dissociation between onset and later stages of global and local target processing
Neuropsychologia
(1993) - et al.
Hierarchical visual stimuli: electrophysiological evidence for separate left hemispheric global and local processing mechanisms in humans
Neuroscience Letters
(1996) - et al.
Hierarchical stimuli and hemispheric specialization: two case studies
Cortex
(1994) - et al.
Electrophysiological evidence of a perceptual precedence of global vs. local visual information
Brain Research: Cognitive Brain Research
(1998)
Neuropsychological contributions to theories of part/whole organization
Cognitive Psychology
Hemispheric differences in global versus local processing of hierarchical visual stimuli by normal subjects: new data and a meta-analysis of previous studies
Neuropsychologia
Where in the brain does visual attention select the forest and the trees?
Nature
Hemispheric specialization for global and local processing: the effect of stimulus category
Proceedings of the Royal Society of London (B)
Cited by (79)
Spatial frequency selection and integration of global and local information in visual processing: A selective review and tribute to Shlomo Bentin
2016, NeuropsychologiaCitation Excerpt :It is not the absolute frequencies, but the relative frequencies that drive the asymmetries. Findings from ERP studies (e.g., Flevaris et al., 2014; Han and Chen, 1996; Han et al., 2000; Heinze et al., 1998; Volberg and Hübner, 2004) corroborate the hypothesis of lateralization emerging later in visual processing. Hemisphere asymmetries in global compared to local perception have most often been reported in mid-level ERP components such as P2 (190–240 ms after stimulus onset), N2 (260–350 ms) and P3 (320–400 ms) rather than in earlier components that reflect more automatic visual processing such as C1 (55–80 ms), P1 (80–150 ms) and N1 (130–210 ms).
The visual encoding of tool-object affordances
2015, NeuroscienceCitation Excerpt :Both our study and that of Dickinson and Intraub (2009) required participants to main fixation at center prior to the appearance of an image with two random objects/object clusters, one in each hemifield. It is unclear why this ‘artificial laboratory setting’ (Dickinson and Intraub, 2009) would reveal a leftward bias in attention, with plausible hypotheses including right-hemispheric specialization for attention or the fact that English is read from left to right (Volberg and Hübner, 2004; Spalek and Hammad, 2005; Dickinson and Intraub, 2009). We found the leftward bias to be augmented in this study when compared to (Dickinson and Intraub, 2009), possibly due to the much smaller sample size in this study when compared to (Dickinson and Intraub, 2009).
Switching between forest and trees: Opposite relationship of progesterone and testosterone to global-local processing
2014, Hormones and BehaviorCitation Excerpt :Thus, one third of stimuli in this condition did require a left mouse response and two thirds a right mouse response. Designs using two targets have been successfully employed previously during both divided and focused attention (Hubner et al., 2007; Muller-Oehring et al., 2007; Volberg and Hubner, 2004). The divided attention condition does not allow for a decision about which letter has been seen, but requires the identification of a target, as both levels need to be attended (compare e.g. Plaisted et al., 1999; Roalf et al., 2006).