Supramodal language comprehension: Role of the left temporal lobe for listening and reading

Abstract

In this fMRI study, we aimed at identifying the cortical areas engaged in supramodal processing of language comprehension. BOLD changes were recorded in 19 healthy right-handed subjects reading or listening to a story. During the visual control tasks the volunteers attended to a series of continuous letterstrings or a fixation cross, while during the acoustic control tasks either a reversed text or white noise were presented. The conjunction of the visual and acoustic story processing yielded left-dominant activations which in comparison to language-like stimuli focused to the left middle temporal gyrus as well as to the supramarginal gyrus. We conclude that the core structure representing supramodal language comprehension is the left temporal lobe at both banks of the superior temporal sulcus.

Introduction

Since the postmortem studies on aphasic patients by Broca and Wernicke, the left inferior frontal and parieto-temporal cortex have been implicated as critical cortical areas for language production and language comprehension (Broca, 1861, Graves, 1997, Lichtheim, 1885, Wernicke, 1874). The advent of functional neuroimaging methods has provided major modifications and refinements to this “Wernicke–Lichtheim model”. Furthermore, these methods yielded evidence for the representation of differentiated linguistic sub-functions in areas of the frontal and temporal lobes. However, with regard to the inferior frontal cortex, functional ascriptions of linguistic functions appear to be inconsistent in terms of function and exact localization (e.g., Dronkers, Wilkins, Van Valin, Redfern, & Jaeger, 2004; Grodzinsky, 2000, Muller, 2000; Muller & Basho, 2004). In addition, non-language functions were observed to map to the inferior frontal cortex including music perception (Maess, Koelsch, Gunter, & Friederici, 2001; Patel, 2003), the performance, imitation and observation of movements (Binkofski et al., 2000, Buccino et al., 2001; Carr, Iacoboni, Dubeau, Mazziotta, & Lenzi, 2003; Hamzei et al., 2003, Iacoboni et al., 1999; Leslie, Johnson-Frey, & Grafton, 2004; Rizzolatti & Arbib, 1998; Rizzolatti, Fogassi, & Gallese, 2001), working memory processes (Smith & Jonides, 1999), and the sequential processing of sounds (Gelfand & Bookheimer, 2003).

In the temporal lobe, lesion studies in primates and in humans, cortical mapping and functional imaging revealed areas related to processing of acoustic stimuli (Griffiths, Warren, Scott, Nelken, & King, 2004; Nelken, 2004; Ojemann, Schoenfield-McNeill, & Corina, 2002; Scott, 2005). Moreover, functional imaging studies underlined the role of temporal association areas in the perception of auditory verbal stimuli (Binder et al., 1999, Binder et al., 2000, Démonet et al., 1992; Démonet, Price, Wise, & Frackowiak, 1994; Narain et al., 2003, Perani et al., 1996, Price et al., 1996; Scott, Blank, Rosen, & Wise, 2000). That corresponds to the finding that the superior temporal sulcus accommodates specific representations for the processing of human voices (Belin, Zatorre, Lafaille, Ahad, & Pike, 2000). However, Hickok and co-workers presented first evidence that the planum temporale – comprising the perisylvian cortex in the occipital vicinity of the primary auditory area – seems to represent phonematic aspects of speech production as well (Hickok, 2001, Hickok et al., 2000). Thus, it was suggested that the planum temporale serves as an “auditory-to-meaning interface” for the construction of acoustically represented speech (Hickok & Poeppel, 2000).

The circuit involved in the comprehension of visually presented language stimuli is known to consist of widespread temporal and frontal regions (Crozier et al., 1999, Jessen et al., 1999, Joubert et al., 2004; Keller, Carpenter, & Just, 2001). With regard to the inferior frontal cortex, recent reviews hypothesized that this region comprises a higher-order interface rather than being responsible for distinct language-related functions (Bookheimer, 2002; Kaan & Swaab, 2002; Muller & Basho, 2004). In contrast, other authors tried to demonstrate separable functional, anatomical sub-divisions, or both (Fiebach, Schlesewsky, Lohmann, von Cramon, & Friederici, 2005; Friederici, Ruschemeyer, Hahne, & Fiebach, 2003; Poldrack et al., 1999; Suzuki & Sakai, 2003; Vigneau et al., 2006). With respect to the temporal lobe, previous studies demonstrated that it is involved in lexico-semantic as well as syntactical processing of written language comprehension (Binder et al., 2000; Mummery, Shallice, & Price, 1999; Rossell, Price, & Nobre, 2003; Vandenberghe, Price, Wise, Josephs, & Frackowiak, 1996).

Taken together, studies dealing with representations of either auditory or visual language comprehension suggest an involvement of lateral temporal and inferior frontal cortex. In addition, neurons in the surrounding area of the superior temporal sulcus have been shown to account for verbal memory functions independent of the input modality (Ojemann et al., 2002). From comparative anatomical studies it is known that the primate's temporal lobe receives input from auditory as well as visual primary and association cortices (Belin, Fecteau, & Bedard, 2004; Seltzer & Pandya, 1989).

Furthermore, there are a few recent neuroimaging studies on supramodal language comprehension. Like in the fMRI experiment we present in this article, Spitsyna and co-workers focused on the convergence of processing of spoken and written language in a PET study. They report distinct local representations of across-modality responses for intelligible language in both anterior and posterior cortical regions of the left temporal lobe (Spitsyna, Warren, Scott, Turkheimer, & Wise, 2006). Previously, an fMRI study was conducted to investigate the influence of sentence complexity and input modality on language comprehension (Constable et al., 2004). In accordance with the PET study's results, modality-independent language comprehension appeared to be strongly left-lateralized. But the shared circuit of the visual and acoustic domain did not map exclusively to the temporal lobe and its posterior vicinity as in the PET study. In addition, Constable and co-workers describe more extensive perisylvian regions including the inferior frontal gyrus.

Comparable to the PET study mentioned before, we used fMRI to examine whether there is an actual intersection for receptive language processing of both input modalities in the human brain. To test for the existence of such modality-independent brain regions, we combined visually and acoustically presented verbal stimuli in one paradigm to map the supramodal language processing area(s). According to the studies described above, we hypothesize this intersection to be found in the left temporal lobe.