Download PDFOpen PDF in browserExperimental Artefacts in Aphasia Research: How Experimental Variables Raise Semantic over Phonological Errors in Conduction AphasiaEasyChair Preprint 63273 pages•Date: August 17, 2021AbstractThe STEPS is an unexpected behavioral pattern found in the literature according to which PWA produce more phonemic errors with non-number words (tale → lale) but more semantic errors with numbers (42 → 13). This effect has been explained by the Building Blocks Hypothesis by which a lesion of the phonological output buffer (POB) would be the responsible for the production of semantic errors with numbers (Dotan & Friedmann, 2015). Recently, we showed evidence that STEPS was not necessarily related to POB damage (Garcia-Orza et al., 2020). Here we explore the STEPS from an interactionist perspective (Martin et al., 1996) that can easily explain the prevalence of semantic errors during number production; they are high-frequency items assessed in pure lists (of numbers) with a high memory load (2 to 4 digits). In specific, we compare the performance of two women with conduction aphasia (one input, ML, and one output, DNR) on 3 production tasks (naming, repetition, reading) with pure lists of (1) multidigit numbers and (2) high- (HF) and (3) low-frequency (LF) colors presented in sequences (2 to 4 elements). Results showed that both numbers and HF color sequences were produced with more semantic than phonemic errors (which were scarce), while these phonemic errors emerged significantly in LF color sequences. As the sequences became longer (increasing the memory load), semantic errors increased for both numbers and HF colors, when phonemic and formal errors only showed a slight tendency to increase (only for DNR). Our findings indicate that (a) lexical frequency has a key role in the occurrence of semantic (numbers, HF colors) vs. phonemic (LF colors) errors, and that (b) error occurrence is directly proportional to memory load, indexed by the number of elements in the sequences. Thus, they support that STEPS seems to be an experimental artifact driven by the interaction between lexical frequency, semantic context, and memory load during speech production. Keyphrases: Memory load, condution aphasia, frequency, number production, paraphasia, semantic context, sequences, word production
|