Sentence production in latent aphasia (Salis & DeDe, 2022)

Purpose: The purpose of this study was to improve our understanding as to which factors determine online, spoken sentence production abilities of adults with latent aphasia in a discourse context.Method: Discourse samples of the story of Cinderella elicited from AphasiaBank were analyzed with speech analysis software. Participants comprised people with latent and anomic aphasia as well as neurotypical controls (10 per group). Durations of pauses (silent and filled) were analyzed according to (a) the location they occurred (between or within sentences), (b) the syntactic complexity of sentences (simple, complex), and (c) sentence length (number of words). Statistical comparisons were conducted using mixed-effect models.Results: The two clinical groups (latent and anomic) differed from controls in the duration of pauses, both between and within sentences. Syntactic complexity did not exert an effect on either of the two clinical groups as compared with controls. As compared with controls, both clinical groups paused more before long in comparison with short sentences.Conclusion: Reduction in processing speed, which affects the ability to simultaneously maintain multiple linguistic and other cognitive demands associated with planning and monitoring of utterances, is a major factor that compromises sentence production in spoken discourse in latent aphasia.
Supplemental Material S1. Number of utterances, utterance length, and duration of pauses between and within utterances separately for simple and complex utterances.
Salis, C., & DeDe, G. (2022). Sentence production in a discourse context in latest aphasia: A real-time study. American Journal of Speech-Language Pathology. Advance online publication. https://doi.org/10.1044/2022_AJSLP-21-00232

研究目的：本研究旨在深入厘清话语语境下，潜在失语症（latent aphasia）成人在线口语句子生成能力的影响因素。
研究方法：本研究对从失语症银行（AphasiaBank）中获取的《灰姑娘》故事话语语料，采用语音分析软件（speech analysis software）开展分析。研究参与者包括潜在失语症患者、命名性失语症（anomic aphasia）患者以及神经典型对照者（neurotypical controls，每组各10名）。本研究从以下三个维度分析停顿（无声停顿与填充停顿）的时长：(a) 停顿发生位置（句间或句内）；(b) 句子句法复杂度（简单句、复合句）；(c) 句子长度（单词数量）。统计比较采用混合效应模型（mixed-effect models）进行。
研究结果：两组临床患者（潜在失语症患者与命名性失语症患者）在句间与句内的停顿时长均与对照组存在显著差异。句法复杂度对两组临床患者的停顿时长均无显著影响，与对照组相比无明显差异。相较于对照组，两组临床患者在长句前的停顿时长显著多于短句场景下的停顿时长。
研究结论：加工速度降低是损害潜在失语症患者口语话语句子生成能力的核心因素；该缺陷会削弱患者同时处理多项语言及其他认知需求的能力，而这些需求与话语的计划与监控过程紧密相关。
补充材料S1：分别针对简单话语与复合话语，统计列出话语数量、话语长度以及话语间与话语内的停顿时长。
参考文献：Salis C, DeDe G. (2022). 话语语境下潜在失语症患者的句子生成能力：一项实时研究. 《美国言语语言病理学期刊》(American Journal of Speech-Language Pathology) 预印在线版. https://doi.org/10.1044/2022_AJSLP-21-00232