Comprehension cueing strategies in elderly: a window into cognitive decline?

Comprehension cueing strategies in elderly: a window into cognitive decline? Abstract Language abilities gradually decline as we age, but the mechanisms of this decline are not well understood. The present study investigated comprehension of subject vs. object who and which direct questions (DQs), embedded questions (EQs) and relative clauses (RCs) in 39 cognitively healthy native speakers of Spanish. The elderly participants (n = 21) were further classified according to their scores on a general cognitive test, Montreal Cognitive Assessment (MoCA), into a group with low MoCA scores, LM (n = 10), and a group with normal MoCA scores, NM (n = 11). A mixed-model, repeated-measures analysis of variance (ANOVA) showed that the elderly participants achieved significantly worse accuracy and speed than the young participants (Y) in all tasks. Accuracy was significantly lower and reaction times significantly longer in the LM group compared to the NM group in DQs and RCs. Accuracy in comprehension of EQs was also worse in LM compared to NM, with no significant difference in RTs between the two groups. The results are explained within the competition model and reliance on a language-specific cueing strategy. Reliance on cueing strategies in sentence comprehension may be an effective indicator of cognitive decline associated with aging. Keywords: comprehension; wh-dependencies; aging. Introduction Cognitive aging is typically associated with a decline in speed of processing and deterioration of memory and attention (Salthouse, 2009). Language abilities also gradually decline as we age, which is reflected in decreased vocabulary, smaller mean number of clauses per utterance, simplified syntactic structure of produced sentences, reliance on optimization strategies when choosing referring expressions as well as difficulty in comprehension of complex sentences (Kemper, Thompson & Marquis, 2001; Grossman, Cooke, De Vita, Chen, Moore et al., 2002; Hendriks, Englert, Wubs & Hoeks, 2008). Older adults’ language comprehension decline appears to be due not to sensory, but cognitive demands of spoken language, with complex syntax slowing down the comprehension even when sentence understanding is accurate (Tun, Benichov & Wingfield, 2010). Research on English has shown that comprehension of structures that require a syntactic operation of movement and involve a longer gap between a moved element and its trace (t), such as object relative clauses (e.g., The cat i that the dog chased t i is black), is impaired in elderly adults, while comprehension of subject relative clauses, in which this gap is smaller (e.g., The cat i that t i chased the dog is black), is spared (e.g., Zurif, Swinney, Prather, Wingfield & Brownell, 1995; Stine- Morrow, Ryan & Leonard, 2000). One explanation of this finding is that the object relative clauses require allocation of more working memory (WM) resources than subject relative clauses, and WM limitation is one of key features of cognitive aging (Zurif et al., 1995; Caplan & Waters, 1999; Stine-Morrow et al., 2000; Grossman, Cooke, De Vita, Alsop, Detre et al., 2002). Furthermore, neuroimaging research has shown that when processing complex sentences, healthy seniors compared to young participants show reduced activation in the core language areas (e.g., inferior frontal regions), while showing additional activation of some areas that are not considered the “core” sentence processing network as well as difference in the coherence of connectivity of the involved brain areas (Peelle, Troiani, Wingfield, & Grossman, 2010; Tyler, Shafto, Randall, Wright, Marslen-Wilson et al., 2010). Activation of the brain regions that are not typically involved in language processing has been interpreted as an indicator of compensatory processes (Grossman et al., 2002; Wingfield & Grossman, 2006; Tyler et al., 2010). Better understanding of the earliest changes in typical cognitive aging is also an important step towards better understanding of the Alzheimer’s disease (AD) continuum. Structural and metabolic changes in AD brain occur long before cognitive symptoms become apparent (Dubois et al., 2007, 2010; Sperling et al., 2011). Crucially, even small metabolic and structural alterations in the brain may affect the dynamics enabling cognitive function (Buckner, Snyder, Shannon, LaRossa, Sachs, et al., 2005). Thus, it is important to understand the brain’s ability to engage alternate networks and rely on cognitive strategies compensating for a deteriorating cognitive function. One goal of the present study was to determine whether elderly native speakers of Spanish rely on compensatory strategies in sentence comprehension. We chose to study comprehension of wh-structures (i.e., structures formed by wh-words, such as what, who, which, etc.): direct and embedded questions introduced by interrogative pronouns que (“what, which”) and quien (“who”) and relative clauses introduced by que. Like in English, the distance between a moved element and its gap is longer in object than in subject wh-structures, as shown in (1-2): (1) ?Quien i t i comio una naranja? (2) ?A quien i mordio j el perrito t j t i ? However, in Spanish preposition a marks object wh- questions and therefore it could serve as a processing cue. Since it appears before the moved wh-word, it signals an object structure, allowing the parser to assign a temporary thematic role before encountering the gap. Thus, reliance on this cue would facilitate comprehension of object structures, resulting in their good comprehension, even though they are syntactically more difficult than subject structures and require more WM resources.