Abstract
For hearing adults, signed language processing increases the salience of iconicity and motor system involvement relative to spoken language processing. Nevertheless, it is unclear how embodied action, mental imagery and iconicity influence their acquisition of signed language. The current study examines the impact of these factors on sign acquisition by manipulating how signs are learned, as well as their semantic and phonological relatedness. The results of Experiment 1 demonstrated that American Sign Language (ASL) signs are learned more effectively via enactment than via referent visualisation and meaningless hand motion, and that iconic signs are learned more effectively than other types of signs. The results of Experiment 2 demonstrate that, when learned via enactment, semantically related ASL signs are recalled more accurately than phonologically related ASL signs. These results indicate that hearing adults' sign language acquisition can be enhanced via a learning method that combines mental imagery and meaningful embodied action (i.e., enactment), strengthening connections between the forms of signs and their referents.
I thank Aimée Kidd Boumiea and Kaitlyn Schuessler for modelling ASL signs for the stimuli of Experiments 1 and 2, respectively. Additionally, I thank Ashlie Caputo for serving as a second coder to establish reliability. Finally, I thank Jana Iverson for equipment support and Avniel Ghuman for support for the writing of this study.
No potential conflict of interest was reported by the author.
I thank Aimée Kidd Boumiea and Kaitlyn Schuessler for modelling ASL signs for the stimuli of Experiments 1 and 2, respectively. Additionally, I thank Ashlie Caputo for serving as a second coder to establish reliability. Finally, I thank Jana Iverson for equipment support and Avniel Ghuman for support for the writing of this study.
No potential conflict of interest was reported by the author.
Notes
1 Unequal numbers of arbitrary (4), iconic (8) and metaphorical (8) signs were included because we were interested in exploring possible differences in sign learning as a function of grammatical class (noun vs. verb) in the latter two categories. No such differences were found, however.
2 It was discovered following data collection that the model's pronunciation of at least two signs (TO-HAMMER and MESSAGE) was incorrect. Because participants were unfamiliar with ASL, the accuracy of their pronunciation of all signs was assessed against the model's pronunciation.
3 This signer is the daughter of a deaf father who communicates via ASL, and she serves as an ASL interpreter at major events multiple times per year.
4 In this research, ASL signs were produced in response to English prompts (forward translation) rather than the more commonly used task of producing English words in response to signs (backwards translation) due to ceiling-level performance on the latter task in pilot testing.
5 Although the primary coder (L.M.M.) was aware of the experiment's design and hypotheses, she was blind to the learning conditions to which signs were assigned for specific participants during coding.
6 To account for differences in the number of signs of each type, Bonferroni corrections were applied to alpha levels for tests examining the influence of sign iconicity.
7 All values analysed using sign as a fixed factor in Experiment 1 are expressed proportionally due to unequal numbers of arbitrary, iconic and metaphorical signs, as well as unequal numbers of signs learned in each condition due to assignment across participants.
8 This study originally tested sign recall via comprehension (i.e., asking participants to provide the English gloss for each sign) in addition to production. However, this measure was dropped due to ceiling effects in pilot testing.
9 This signer is currently studying to become a speech-language pathologist, and learned ASL beginning as an undergraduate as part of her training.
10 Without including sign iconicity as a covariate, no significant difference was found for sign class by sign either, F2(1, 22) = 1.71, p = .20, = .07.
11 Without including sign iconicity as a covariate, no significant difference was found for sign class by sign either, F2 < 1.
12 Without including sign iconicity as a covariate, no significant difference was found for sign class by sign either, F2< 1.