Do people predict different aspects of a predictable word to the same extent? We tested prediction of phonological and gender information by creating phonological and gender mismatches between an article and a predictable noun in Italian. Native Italian speakers read predictive sentence contexts followed by the expected noun (e.g., un incidente: 'accident') or another plausible, but unexpected noun, either beginning with a different phonological class (consonant vs. vowel, e.g., uno scontro: 'collision'; phonological mismatch) or belonging to a different gender class (e.g., un'inondazione: 'flooding'; gender mismatch). Phonological mismatch articles elicited greater negativity than expected articles at posterior channels around 450-800 ms post-stimulus. In contrast, gender mismatch articles elicited greater negativity than expected articles at left posterior channels around 250-800 ms. Unexpected nouns showed an N400 effect followed by frontal positivity relative to expected nouns. The earlier effect for the gender mismatch articles suggests that people are quicker or more likely to pre-activate gender information vs. phonological information of a predictable word. We interpret the results with respect to production-based prediction accounts.

中文翻译：

---

语音和性别信息的预测：意大利语中与事件相关的潜在研究。

人们会以相同的程度预测可预测单词的不同方面吗？我们通过在文章和意大利语中的可预测名词之间创建语音和性别不匹配来测试语音和性别信息的预测。意大利语为母语的人会读预言性句子的上下文，后跟预期的名词（例如，偶然事件：“意外事件”）或另一个似乎合理但意外的名词，它们以不同的语音学类别（辅音对元音，例如uno scontro：“碰撞”）开头'；语音不匹配）或属于不同的性别类别（例如，un'inondazione：“泛洪”；性别不匹配）。音质不匹配的文章在刺激后约450-800毫秒的后通道上引起比预期文章更大的否定性。相比之下，性别不匹配的文章在左后声道约250-800 ms处比预期的文章引起更大的负面影响。意外的名词显示出N400效应，随后相对于预期的名词具有正面正性。性别不匹配文章的早期影响表明，与可预测单词的语音信息相比，人们更快或更可能预激活性别信息。我们根据基于生产的预测帐户解释结果。