Response latencies in auditory sentence comprehension: effects of linguistic versus perceptual challenge.

Understanding spoken language is such a common everyday activity that we often fail to appreciate the complexity of a task that requires a rapid dynamic interplay of both perceptual and cognitive operations. These operations include decoding the speech signal into phonemes, determining lexical identity, and syntactically parsing the speech input to yield a semantically meaningful representation in working memory. These operations, furthermore, must be conducted rapidly as the speech is arriving in real time. Because of the limits in processing speed and attentional capacity associated with adult aging (Bopp & Verhaeghen, 2005; Fisk & Sharp, 2004; Salthouse, 1991, 1996), it should thus not be surprising that older adults lose efficiency in sentence comprehension (Burke & Shafto, 2008). In addition to age-related limits in cognitive factors, it has been estimated that more than 80% of adults over the age of 70 have some degree of hearing loss (e.g. Cruickshanks, et al., 1998; Davis, 1990; Gordon-Salant, 2005). Hearing loss may not only cause a listener to miss words in speech or result in an impoverished representation in memory (Surprenant, 1999; Surprenant, Neath, & Brown, 2003). There are also suggestions that perceptual effort, whether caused by the presence of background noise (Murphy, Craik, Li & Schneider, 2000; Rabbitt, 1968) or hearing loss (Rabbitt, 1991; Wingfield, Tun & McCoy, 2005), can have a negative impact on memory even when the to-be-recalled words themselves could be correctly identified. This latter finding can be characterized in terms of an effortfulness hypothesis: the suggestion that losses in auditory acuity may require the listener to invest extra effort in early stage perceptual processes, thus draining resources that might otherwise be available for comprehension and encoding of what has been heard in memory (Pichora-Fuller, 2003; Rabbitt, 1991; Tun, McCoy & Wingfield, 2009; Wingfield et al., 2005). Consistent with this hypothesis, work in our laboratory has shown that, relative to older adults with good hearing, an age-matched group with mild-to-moderate hearing loss had poorer recall for words heard in a running memory task, even though a perceptual check insured that the words could be correctly identified at the intensities at which they were presented (McCoy et al., 2005). It has also been shown that even a relatively mild hearing loss can compound the effects of rapid speech rates and syntactic complexity in older adults’ comprehension of spoken sentences (Wingfield, McCoy, Peelle, Tun & Cox, 2006). In the above-cited studies, the effects of perceptual effort associated with age-related hearing loss were evidenced in poorer comprehension and/or memory performance relative to younger and older adults with good hearing. This raises the question of whether one may observe downstream effects of effortful perception even in a case where comprehension has nevertheless been successful. In order to investigate this question we selected a sentence comprehension task that would be made to vary in both perceptual and linguistic difficulty while still allowing for correct comprehension. Our question was whether conditions of increased early-stage perceptual difficulty might slow the production of these comprehension responses, relative to conditions representing a reduced perceptual challenge. For this experiment we manipulated cognitive demands by presenting sentences of equivalent length, but with the sentence meaning expressed with either a relatively simple or a more complex syntactic structure. The stimuli, sentences with subject-relative or object-relative clause structures, have been shown in numerous studies to produce reliable differences in ease of comprehension and recall in both young and older adults (e.g., Fallon, Peelle, & Wingfield, 2006; Just, Carpenter, Keller, Eddy, & Thulborn, 1996; Stewart & Wingfield, 2009; Waters & Caplan, 2001). Perceptual difficulty was manipulated by varying the sound level of the stimulus sentences within a range for each individual that still allowed words to be identified. Our key questions were first, whether older adults with hearing loss would show differentially longer latencies in their successful comprehension responses relative to age-matched controls with better hearing. Second, to the extent that this result is obtained, we wished to determine whether this effect of hearing acuity would be further impacted by added perceptual demands (represented by lower presentation levels) and increased cognitive load (represented by increasing syntactic complexity). Although our focus is on the effects of hearing acuity in older adults, we included a group of good-hearing young adults to illustrate maximal performance that might be expected without the impediments of age-related cognitive change and hearing loss.