Theories of implicature and focus both make use of alternatives in deriving inferences. However, little is known about how comprehenders use alternatives in real time. We conducted a visual-world eye-tracking study to investigate whether comprehenders draw their visual attention to the depicted alternative when interpreting QI and 'only'. In addition, we examine whether the use of alternatives differs between these two cases.
We found that participants who heard *Mia peeled the banana* / *Mia only peeled the banana*... (implying / asserting that *she did not cut the cake*) directed their attention to **the competitor** (cut cake, **relevant alternative**) before fixating the target (uncut cake). Whereas participants who heard *Mia also peeled the banana*... (presupposing that she cut the cake) didn't direct their attention to **the competitor** (uncut cake, **irrelevant state**) before fixating the target (cut cake, relevant alternative). These results suggest that comprehenders make use of alternatives in real-time when interpreting QI and 'only'.
In addition, we found a greater bias to the competitor (depicted alternative) in the QI condition compared to the 'only' condition. Since the competitor could be viewed as the representation of the context under discussion, this greater bias might be the result of participants paying more attention to context in the QI condition.