We used phonological priming and ERPs to investigate the organization of the lexicon in American Sign Language. Across go/no-go repetition detection and semantic categorization tasks, targets in related pairs that shared handshape and location elicited smaller N400s than targets in unrelated pairs, indicative of facilitated processing. Handshape-related targets also elicited smaller N400s than unrelated targets, but only in the repetition task. The location priming effect reversed direction across tasks, with slightly larger amplitude N400s for targets in related versus unrelated pairs in the semantic task, indicative of interference. These patterns imply that handshape and location play different roles during sign recognition and that there is a hierarchical organization for the sign lexicon. Similar to interactive-activation models of word recognition, we argue for differentiation between sublexical facilitation and lexical competition. Lexical competition is primarily driven by the location parameter and is more engaged when identification of single lexico-semantic entries is required.

我们使用语音启动和 ERP 来研究美国手语词汇的组织。在 go/no-go 重复检测和语义分类任务中，共享手形和位置的相关对中的目标比不相关对中的目标引出更小的 N400，这表明处理更容易。与手形相关的目标也比不相关的目标引起更小的 N400，但仅限于重复任务。位置启动效应在任务之间反转方向，语义任务中相关对和不相关对中的目标的 N400 幅度稍大，表明存在干扰。这些模式意味着手形和位置在符号识别过程中扮演着不同的角色，并且符号词典存在分层组织。与单词识别的交互激活模型类似，我们主张词汇下促进和词汇竞争之间的区别。词汇竞争主要由位置参数驱动，并且当需要识别单个词汇语义条目时更加激烈。