I study how people process multiword expressions! My undergrad thesis was on whether people represent multiword expressions holistically or whether they are only built up compositionally during processing. My data suggested that people store highly frequent multiword expressions as holistic units. Now I'm looking at a couple of related aspects, like whether the frequency of a multiword expression influences how people process the individual words inside it. For my dissertation though, I think I'm leaning towards doing some more mechanistic stuff, like how do people establish representations of multiword expressions in the first place? How do we know which information will be useful and which information to discard? etc.