Thanks for stopping by! A short video overview of our poster is provided below. Please join us for the Live Q&A. <br><br> #### **Project summary:** This project address the perception of **VP ellipsis** (marked as <> below) when the antecedent and elided clause do not match in the Voice of the verb. Previous research has reported an asymmetry in which **Passive-Active mismatches are rated as more acceptable than Active-Passive mismatches**: ----- **Regular, homophonous voice:** *Active-Passive:* * John met Sue, and then Mary was <met by Sue> too. *Passive-Active:* ? Sue was met by John, and then Mary did <meet John> too. ----- Here, we explore the idea that the form of the verb modulates this asymmetry. We present evidence that the **Passive-Active advantage only holds for regular verb paradigms**, where the Passive and Active voice sound the same. In contrast, we do not find a Passive-Active advantage for irregular verb forms, where the Passive has a different morphology than the Active: ----- **Irregular, non-homophonous voice:** *Active-Passive:* * John drove Sue, and then Mary was <driven by Sue> too. *Passive-Active:* * Sue was driven by John, and then Mary did <drive John> too. ----- We sketch for a lexical competition-based account of Voice mismatch in VP ellipsis. For regular, homophonous verbs: A Passive verb in the antecedent activates an Active form in the lexicon, increasing the perception of acceptability. However, irregular Passives do not activate their Active counterparts (as much) as they have different forms.