The Detection and Primed Production of Novel Constructions

Abstract

Situated within second language (L2) research about the acquisition of morphosyntax, this study investigated English L2 speakers’ detection and primed production of a novel construction with morphological and structural features. We report on two experiments with Thai (n = 69) and Farsi (n = 70) English L2 speakers, respectively, carried out an aural construction learning task that provided low type‐frequency input with the transitive construction in Esperanto—which is marked by accusative case marking (–n) and flexible word order (subject‐verb‐object and object‐verb‐subject)—followed by aural comprehension tests and a priming activity (20 primes and 20 prompts). Results of the aural comprehension tests showed that 23% of the Thai participants (16/69) and 50% of the Farsi participants (35/70) detected the target construction in the input. Results of the primed production task revealed that only those participants who detected the target construction were able to be primed. The findings are discussed in relation to the role of speakers’ previously learned languages in the detection and primed production of novel constructions.

Primed From the Start: Syntactic Priming During the First Days of Language Learning

“…We did not observe such an effect here; rather, syntactic priming effects in read‐aloud times were similar for all structures (but see Weber et al., , for differences in the neural signature of priming effects using a similar paradigm). Moreover, for the picture choices, a syntactic repetition effect was detected for both known and frequent structures but not the infrequent ones, indicating that at least a certain level of exposure has to be reached before structural priming can lead to a benefit in understanding the meaning of the sentences (McDonough & Fulga, ). The syntactic priming effect on the picture choices appeared to level off on Day 9.…”

Structural priming and the acquisition of novel form-meaning mappings

“…Prior construction learning studies (Boyd, Gottschalk, and Goldberg 2009;Casenhiser and Goldberg 2005;Goldberg and Casenhiser 2008;Goldberg, Casenhiser, and Sethuraman 2004;Goldberg, Casenhiser, and White 2007) have shown that English L1 speakers can comprehend and produce the novel construction of appearance (N 1 N 2 V), with the corresponding meaning of N 1 "appears in/on" N 2 (e.g., the spot the king moopoed) after relatively brief exposure. In contrast, construction learning studies with L2 English speakers reported greater difficulty in detecting a variety of constructions, including the appearance construction and the Samoan ergative construction (Nakamura 2012), English dative constructions (McDonough and Nekrasova-Becker 2014;Year and Gordon 2009), and Esperanto transitives (Fulga and McDonough 2014;McDonough and Fulga 2015;McDonough and Trofimovich 2013). For example, only about one-fifth of all Thai participants tested by McDonough and Trofimovich (2013) were able to detect the Esperanto transitive construction.…”

“…The findings of previous L2 research (Hopp 2010;Jiang 2004Jiang , 2007Papadopoulou et al 2011), including research on novel construction learning (Fulga and McDonough 2014;McDonough and Fulga 2015;McDonough and Trofimovich 2013), suggest that it might be difficult for L2 learners to detect and use a novel morphological cue for sentence interpretation. The current set of findings further qualifies this challenge by suggesting that structural priming, in the absence of any other input, feedback, or output learning opportunities, provides little evidence to help learners abandon a familiar cue in favor of a new one.…”

The Role of Statistical Learning and Working Memory in L2 Speakers’ Pattern Learning

“…The Esperanto materials consisted of a picture identification task divided into exposure and test phases based on items used in previous studies (Fulga & McDonough, ; McDonough & Fulga, ; McDonough & Trofimovich, , ). For the exposure phase, the participants listened to 24 simple Esperanto transitive sentences (12 SVO, 12 OVS) and chose one of two images (printed side‐by‐side on paper) that correctly depicted the meaning of each sentence.…”

“…CONSTRUCTION LEARNING STUDIES CARried out by Goldberg and colleagues have shown that first language (L1) speakers of English can comprehend and produce the novel construction of appearance (N 1 N 2 V), with the corresponding meaning of N 1 ‘appears in/on’ N 2 (e.g., the spot the king moopoed ) after relatively brief language exposure (Boyd, Gottschalk, & Goldberg, ; Casenhiser & Goldberg, ; Goldberg & Casenhiser, ; Goldberg, Casenhiser, & Sethuraman, ; Goldberg, Casenhiser, & White, ). However, construction learning studies with second language (L2) speakers have reported far greater variability with a variety of constructions, including the appearance construction and Samoan ergative constructions (Nakamura, ), Esperanto transitives (Fulga & McDonough, ; McDonough & Fulga, ; McDonough & Trofimovich, , ), and English datives (McDonough & Nekrasova–Becker, ; Year & Gordon, ). These studies have shown that L2 speakers experience considerable difficulty when learning new patterns.…”