Noncanonical sentence structures pose comprehension challenges because they require increased cognitive demand. Prosody may partially alleviate this cognitive load. These findings largely stem from behavioral studies, yet physiological measures may reveal additional insights into how cognition is deployed to parse sentences. Pupillometry has been at the forefront of investigations into physiological measures of cognitive demand during auditory sentence comprehension. This study offers an alternative approach by examining whether eye-tracking measures, including blinks and saccades, index cognitive demand during auditory noncanonical sentence comprehension and whether these metrics are sensitive to reductions in cognitive load associated with typical prosodic cues. We further investigated how eye-tracking patterns differ across correct and incorrect responses, as a function of time, and how each related to behavioral measures of cognition. Canonical and noncanonical sentence comprehension was measured in 30 younger adults using an auditory sentence–picture matching task. We also assessed participants' attention and working memory. Blinking and saccades both differentiate noncanonical sentences from canonical sentences. Saccades further distinguish noncanonical structures from each other. Participants made more saccades on incorrect than correct trials. The number of saccades also related to working memory, regardless of syntax. However, neither eye-tracking metric was sensitive to the changes in cognitive demand that was behaviorally observed in response to typical prosodic cues. Overall, these findings suggest that eye-tracking indices, particularly saccades, reflect cognitive demand during auditory noncanonical sentence comprehension when visual information is present, offering greater insights into the strategies and neural resources participants use to parse auditory sentences.

The neural basis of language has been studied for centuries, yet the networks critically involved in simply identifying or understanding a spoken word remain elusive. Several functional–anatomical models of critical neural substrates of receptive speech have been proposed, including (1) auditory-related regions in the left mid-posterior superior temporal lobe, (2) motor-related regions in the left frontal lobe (in normal and/or noisy conditions), (3) the left anterior superior temporal lobe, or (4) bilateral mid-posterior superior temporal areas. One difficulty in comparing these models is that they often focus on different aspects of the sound-to-meaning pathway and are supported by different types of stimuli and tasks. Two auditory tasks that are typically used in separate studies—syllable discrimination and word comprehension—often yield different conclusions. We assessed syllable discrimination (words and nonwords) and word comprehension (clear speech and with a noise masker) in 158 individuals with focal brain damage: left ( n = 113) or right ( n = 19) hemisphere stroke, left ( n = 18) or right ( n = 8) anterior temporal lobectomy, and 26 neurologically intact controls. Discrimination and comprehension tasks are doubly dissociable both behaviorally and neurologically. In support of a bilateral model, clear speech comprehension was near ceiling in 95% of left stroke cases and right temporal damage impaired syllable discrimination. Lesion-symptom mapping analyses for the syllable discrimination and noisy word comprehension tasks each implicated most of the left superior temporal gyrus. Comprehension but not discrimination tasks also implicated the left posterior middle temporal gyrus, whereas discrimination but not comprehension tasks also implicated more dorsal sensorimotor regions in posterior perisylvian cortex.

Broca's area has long been implicated in sentence comprehension. Damage to this region is thought to be the central source of “agrammatic comprehension” in which performance is substantially worse (and near chance) on sentences with noncanonical word orders compared with canonical word order sentences (in English). This claim is supported by functional neuroimaging studies demonstrating greater activation in Broca's area for noncanonical versus canonical sentences. However, functional neuroimaging studies also have frequently implicated the anterior temporal lobe (ATL) in sentence processing more broadly, and recent lesion–symptom mapping studies have implicated the ATL and mid temporal regions in agrammatic comprehension. This study investigates these seemingly conflicting findings in 66 left-hemisphere patients with chronic focal cerebral damage. Patients completed two sentence comprehension measures, sentence–picture matching and plausibility judgments. Patients with damage including Broca's area (but excluding the temporal lobe; n = 11) on average did not exhibit the expected agrammatic comprehension pattern—for example, their performance was >80% on noncanonical sentences in the sentence–picture matching task. Patients with ATL damage ( n = 18) also did not exhibit an agrammatic comprehension pattern. Across our entire patient sample, the lesions of patients with agrammatic comprehension patterns in either task had maximal overlap in posterior superior temporal and inferior parietal regions. Using voxel-based lesion–symptom mapping, we find that lower performances on canonical and noncanonical sentences in each task are both associated with damage to a large left superior temporal–inferior parietal network including portions of the ATL, but not Broca's area. Notably, however, response bias in plausibility judgments was significantly associated with damage to inferior frontal cortex, including gray and white matter in Broca's area, suggesting that the contribution of Broca's area to sentence comprehension may be related to task-related cognitive demands.