Resolving sentence ambiguity with planning and working memory resources: Evidence from fMRI Susana Novais-Santos, a James Gee, a,b Maliha Shah, c Vanessa Troiani, c Melissa Work, c and Murray Grossman c, ⁎ a Department of Bioengineering, University of Pennsylvania, USA b Department of Radiology, University of Pennsylvania, USA c Department of Neurology-2 Gibson, University of Pennsylvania School of Medicine, 3400 Spruce St., Philadelphia, PA 19104-4283, USA Received 16 December 2006; revised 24 February 2007; accepted 14 March 2007 Available online 13 May 2007 We used functional magnetic resonance imaging (fMRI) to test com- peting claims about the role of executive resources during the disam- biguation of a sentence featuring a temporary structural ambiguity. Written sentences with a direct object (DO) structure or a sentential complement (SC) structure were shown to 19 healthy, right-handed, young adults in a phrase-by-phrase manner. These sentences contained a main verb that is statistically more likely to be associated with a DO structure or an SC structure. Half of each type of sentence also contained an extra phrase strategically located to stress working memory prior to disambiguating the sentence. We found that sentences featuring a less consistent verb-structure mapping recruit greater dorsolateral prefrontal cortex (dlPFC) activation than sentences with a more consistent verb-structure mapping, implicating strategic on-line planning during resolution of a temporary structural ambiguity. By comparison, we observed left inferior parietal cortex (IPC) activation in sentences with an increased working memory demand compared to sentences with a low working memory load. These findings are consistent with a large-scale neural network for sentence processing that recruits distinct planning and working memory processing resources as needed to support the comprehension of sentences. Published by Elsevier Inc. Keywords: Sentence comprehension; Executive; fMRI; Prefrontal; Parietal Introduction Unraveling the complex neural substrate underlying the human language comprehension system has been a major goal of neu- roscience during the past century. Recent observations emphasize that executive resources play a substantial role during sentence comprehension. This has led to a two-component model of sentence processing that distinguishes between core sentence processing mechanisms and executive resources like working memory and strategic planning that support sentence comprehension (Friederici, 2002; Wingfield and Grossman, 2006). From this perspective, the core sentence processing mechanism in left peri-Sylvian cortex recruits resources such as working memory (WM), planning/ switching and selective attention as needed. These resources are supported in other, non-peri-Sylvian brain regions. The present study uses functional magnetic resonance imaging (fMRI) to exa- mine this model during the comprehension of sentences with a temporary structural ambiguity. Functional neuroimaging studies of syntactic processing have shown activation of core peri-Sylvian language areas, namely left inferior frontal (IFC) and posterolateral temporal (PLTC) cortices, during tasks such as grammaticality judgments and answering probes about complex sentences (Ben-Shachar et al., 2004; Cooke et al., 2006; Cooke et al., 2001; Kuperberg et al., 2003; Ni et al., 2000). Experimental factors such as the modality of sentence pre- sentation (visual or auditory) (Caplan et al., 1999; Friederici et al., 2000; Michael et al., 2001), the type of response (comprehension or expression) (Indefrey et al., 2001) and the age of the subjects (Caplan et al., 2003; Grossman et al., 2002) introduce subtle differences in the location of peak neuroanatomic activation, but over-all the results consistently emphasize the contribution of peri- Sylvian frontal and temporal regions of the left hemisphere during sentence processing. The mental representation of an unfolding sentence appears to be constructed incrementally, moreover. As each successive word is encountered in this first-pass analysis, it is rapidly integrated into an evolving interpretation. Some have referred to this as the principle of “minimal attachment” (Ferreira and Henderson, 1991; Frazier and Rayner, 1982). However, this is not a straightforward task, as the syntactic and thematic relationships between a word and the remainder of the sentence are often undefined when that word is www.elsevier.com/locate/ynimg NeuroImage 37 (2007) 361 – 378 ⁎ Corresponding author. Fax: +1 215 349 8464. E-mail address: mgrossma@mail.med.upenn.edu (M. Grossman). Available online on ScienceDirect (www.sciencedirect.com). 1053-8119/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.neuroimage.2007.03.077