Structural priming is determined by global syntax rather than internal phrasal structure: Evidence from young and older adults.

Abstract

Structural priming refers to the tendency of speakers to repeat syntactic structures across sentences. We investigated the extent to which structural priming persists with age and whether the effect depends upon highly abstract syntactic representations that only encompass the global sentence structure or whether representations are specified for internal constituent phrasal properties. In Experiment 1, young and older adults described transitive verb targets that contained the plural morphology of the patient role ("The horse is chasing the frogs/ The frogs are being chased by the horse"). While maintaining the conceptual and global syntactic structure of the prime, we manipulated the internal phrasal structure of the patient role to either match (plural; "The king is punching the builders/ The builders are being punched by the king") or mismatch (coordinate noun phrase; "The king is punching the pirate and the builder/ The pirate and the builder are being punched by the king") the target. In both age groups, we observed limited priming of onset latencies, but robust effects of choice structural primingparticipants produced more passive targets following passive primeswhich critically did not vary dependent on whether the internal constituent structure matched or mismatched between the prime and target. Experiment 2 replicated these findings for the agent role: choice structural priming was unaffected by age or changes to the prime noun phrase type. This demonstrates that global, not internal, syntactic structure determines syntactic choices in young and older adults, as predicted by residual activation and implicit learning models of structural priming.

Structural priming is supported by different components of nondeclarative memory: Evidence from priming across the lifespan.

“…If our hypothesis about the role of conceptual and perceptual memory is accurate, we should observe that, as the age of the participant increases, their long-term priming magnitude declines whereas their short-term priming magnitude remains unaffected. Reviewing the structural priming literature does not lead to any clear conclusions on how priming ability changes with age: The studies that show intact priming in older adults (Ferreira et al, 2008; Hardy, Messenger, & Maylor, 2017; Hardy, Wheeldon, & Segaert, 2018; Hartsuiker et al, 2008) are in contrast with studies showing a decline in priming ability in older adults (Heyselaar, Segaert, et al, 2017). Additionally, these studies have focused on short-term priming, with little to no results indicating how long-term priming may change with age.…”

Cognitive constraints on advance planning of sentence intonation

“…In line with earlier studies, active sentences uttered in response to complex visual scenes are assumed to be produced less rather than more incrementally [ 15 , 32 , 35 , 65 , 76 ]. Oppermann et al [ 76 ] show for German that sentences with a frequent subject-verb-object ordering (e.g., ‘Die Maus frisst den Käse’, ‘The mouse is eating the cheese’) can be successfully planned in full even at the phonological planning stage.…”

“…They must be native speakers of Estonian, and between 18 and 50 years old. The upper age limit is applied because the WM capacities are known to decay in ageing speakers (see in [ 65 ]) and due to the COVID-19 pandemic, in which the older participants belong to the higher risk group. Speakers eligible for the study should have normal or corrected to normal vision (soft contact lenses included, glasses excluded), no diagnosed language impairments and normal hearing ability.…”

“…The aim of the first block is to establish the tonal differences between short and long sentences when no additional cognitive load is imposed on speech production. For this block, participants are asked to describe monochromatic line drawings depicting an action involving three actors (Fig 2, for similar materials see [ 65 ]). Either one or two of the three actors can function as the initiator(s) or the recipient(s) of an action, i.e., agent or patient respectively.…”

“…As the length effect is sometimes reported to occur between the sentences consisting of short or long sentence-initial subject noun phrases (e.g., ‘ The man is pulling the donkey’ vs ‘ The man and the woman are pulling the donkey’), but to be absent between the sentences containing short or long sentence-final noun phrases (e.g., ‘The man is pulling donkeys ’ vs ‘The man is pulling the donkey and the goat ’) [ 4 ], the proposed experiment will control the sentence length at both of these syntactic positions. However, the methodological assumption is that the conceptual planning processes do not differ between the two conditions of multiple actors (two agents vs two patients) because the complex visual scenes are similar in that they require an early higher-level conceptual decision for encoding the relationships between the actors depicted [ 32 , 65 ].…”