Magnetic Stimulation of the Right Visual Cortex Impairs Form-specific Priming Gorana Pobric1, Stefan R. Schweinberger2, and Michal Lavidor1 Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/19/6/1013/1756719/jocn.2007.19.6.1013.pdf by guest on 18 May 2021 Abstract & Recent evidence suggests that priming of objects across dif- cipital cortex in form-specific priming [Kroll, N. E. A., Yonelinas, ferent images (abstract priming) and priming of specific images A. P., Kishiyama, M. M., Baynes, K., Knight, R. T., & Gazzaniga, of an object (form-specific priming) are mediated by dissociable M. S. The neural substrates of visual implicit memory: Do the neural processing subsystems that operate in parallel and are two hemispheres play different roles? Journal of Cognitive predominantly linked to left and right hemispheric processing, Neuroscience, 15, 833–842, 2003]. Compared to no TMS and respectively [Marsolek, C. J. Dissociable neural subsystems un- sham TMS, rTMS of the right occipital cortex disrupted derlie abstract and specific object recognition. Psychological immediate form-specific priming in a semantic categorization Science, 10, 111–118, 1999]. Previous brain imaging studies have task. Left occipital rTMS, on the other hand, had no converse provided important information about the neuroanatomical re- effect on abstractive priming. Abstract priming may involve gions that are involved in form-specific and abstract priming; deeper semantic processing and may be unresponsive to mag- however, these techniques did not fully establish the functional netic stimulation of a single cortical locus. Our TMS results show significance of priming-related changes in cortical brain activ- that form-specific priming relies on a visual word-form system ity. Here, we used repetitive transcranial magnetic stimulation localized in the right occipital lobe, in line with the predictions (rTMS) in order to establish the functional role of the right oc- from divided visual field behavioral studies [Marsolek, 1999]. & INTRODUCTION plementary to this is that the form-specific subsystem Priming is a nonconscious form of memory that involves enables the identification of specific letter cases, fonts, a change in a person’s ability to identify, produce, or and scripts. There is disagreement over the neuroana- classify an item due to a previous encounter with that or tomical substrates of this form of priming that we aimed a related item (Schacter, Dobbins, & Schnyer, 2004). For to resolve here by using magnetic stimulation. example, priming in the lexical decision task is observed Recent evidence suggests that priming of objects across when participants are faster and more accurate in cate- different images and priming of specific images of an gorizing letter strings as words when these words were object occur in dissociable neural processing subsys- studied in an earlier episode (Bowers, 1996). One type tems that operate in parallel (Marsolek, 1999). Marsolek, of visual priming is form-specific priming, where facili- Kosslyn, and Squire (1992) first reported that case- tation results from visual similarity or identity between specific visual priming for words is greater when test prime and target. Together with abstract visual form stimuli are presented initially in the right cerebral hemi- priming, these two systems are fundamental to visual sphere (RH). In contrast, case-specific explicit memory recognition, as we need both the ability to recognize the for words is not greater when stimuli are presented initial- abstract category to which an input shape corresponds ly in the left hemisphere (LH). Marsolek and colleagues (e.g., keyboard vs. pen), as well as the ability to recog- (Marsolek, 1995, 1999, 2004; Burgund & Marsolek, 1997; nize a specific image or exemplar to which that same Marsolek et al., 1992) suggested that an abstract repre- input shape corresponds (e.g., a specific pen). For sentations subsystem operates more effectively than words, the abstract format of orthographic knowledge the specific representations subsystem in the LH, and is the ability to map perceptually dissimilar words the specific subsystem operates more effectively than the (READ/read) onto a common abstract orthographic abstract subsystem in the RH. representation (Bowers, Vigliocco, & Haan, 1998). Com- Although the work of Marsolek (1999, 2004) and Burgund and Marsolek (1997) continued to support the lateralization of abstract and form-specific systems, 1University of Hull, United Kingdom, 2Friedrich-Schiller-University, others have reported completely different patterns. Germany Koivisto (1996), for example, showed that for uppercase D 2007 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 19:6, pp. 1013–1020 Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/jocn.2007.19.6.1013 by guest on 30 September 2021 targets, there was evidence for form-specific priming 2002; Cowey & Walsh, 2000; Kammer, 1999; Kastner, that was exclusive to the RH. However, lowercase test Demmer, & Ziemann, 1998; Amassian et al., 1989, 1993, items did not reveal any hemispheric asymmetries in 1998). Furthermore, we recently showed that processing form-dependent priming, indicating that in some con- left visual field (LVF) targets was significantly impaired ditions form-specific representations are computed in with repetitive TMS (rTMS) to the right occipital cor- the LH as well. Kroll et al. (2003) have shown that the tex, whereas no significant impairment was observed for RH is involved in form-specific priming, but only when right visual field (RVF) target processing. The comple- using a specific memory test (stem completion). Kroll, mentary pattern of rTMS effects was obtained with LH Rocha, Yonelinas, Baynes, and Frederick (2001) re- stimulation, which significantly impaired lexical decision Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/19/6/1013/1756719/jocn.2007.19.6.1013.pdf by guest on 18 May 2021 ported that in the fragment-completion task both hemi- performance to RVF but not LVF targets (Lavidor, Ellison, spheres were capable of supporting form-specific visual & Walsh, 2003). implicit memory. In addition, Koivisto did not find any The use of TMS therefore allowed us to investigate asymmetries in the abstract component of priming, in lexical processing in the RH and LH under conditions of line with more recent findings (Bowers & Turner, 2005). TMS-induced impairment of the left and right occipital Patient data gave more consistent support of the areas. We assessed the ability of participants to conduct lateralized abstract/specific duality. A case study re- semantic categorization of animal and professions words ported by Samuelsson, Bogges, and Karlsson (2000) that were primed by identical words, by the same words suggested that the RH is necessary for normal visual that differed in letter case, or by unrelated words. rTMS priming. More specifically, Vaidya, Gabrieli, Verfaellie, was applied to the LH and RH occipital areas as partic- Fleischman, and Askari (1998) have examined visual ipants made semantic categorization responses to words priming in patients with global amnesia and patient that represented animals or professions. All the target M.S., who had a right occipital lobectomy. Word stems words were presented in lower case and were primed appeared in the same or different font as study words. by the same or unrelated words (repetition priming), Amnesic patients showed normal font-specific priming where half of the same primes were in lower case (form- despite impaired word-stem cued recall; however, pa- specific priming), and half were in upper case (abstract tient M.S. failed to exhibit font-specific priming despite priming). We expected to observe form-specific priming preserved declarative memory. Therefore, Vaidya et al. and abstract priming in the control conditions. Crucially, concluded that perceptual specificity in visual priming we expected rTMS of the right occipital region to impair depends on visual processes mediated by the right form-specific priming. Lack of such impairment would occipital lobe rather than medial temporal and dience- imply that either the right occipital cortex is not involved phalic regions involved in declarative memory. In ac- in form-specific priming (in line with Dehaene et al., cordance, Kroll et al. (2003) have shown that the right 2001), or that both hemispheres can benefit from same- occipital cortex is involved in form-specific priming, but case repetition priming (in accordance with Bowers & only when using a specific memory test (stem comple- Turner, 2005). In addition, if rTMS over the left occipital tion). However, Yonelinas et al. (2001) argued that the cortex would yield selective effects for targets preceded right occipital lobe does not play a necessary role in with uppercase primes, that might support the com- visual implicit memory, and that the isolated LH can plementary hypothesis of Marsolek (1999), that is, the support normal levels of visual priming in a variety of existence of an abstract priming mechanism that is lat- tasks. In addition, a different priming region in the eralized to the LH. We should note that this prediction RH, the right extrastriate cortex, was suggested as the is less likely, as the identification of abstract letter iden- locus for same-format specific system (functional mag- tities (i.e., the mapping from A ! a) relies heavily on netic resonance imaging [fMRI]: Dehaene et al., 2001; memory processes and differs from perceptual