When people hear words for objects with prototypical colors (e.g., ‘banana’), they look at objects of the same color (e.g., lemon), suggesting a link in comprehension between objects and their prototypical colors. However, that link does not carry over to production: The experimental record also shows that when people speak, they tend to omit prototypical colors, using color adjectives when it is informative (e.g., when referring to clothes, which have no prototypical color). These findings yield an interesting prediction, which we tested here: while prior work shows that people look at yellow objects when hearing ‘banana’, they should look away from bananas when hearing ‘yellow’. The results of an offline sentence-completion task (N = 100) and an online eye-tracking task (N = 41) confirmed that when presented with truncated color descriptions (e.g., ‘Click on the yellow…’), people anticipate clothing items rather than stereotypical fruits. A corpus analysis ruled out the possibility that this association between color and clothing arises from simple context-free co-occurrence statistics. We conclude that comprehenders make linguistic predictions based not only on what they know about the world (e.g., which objects are yellow) but also on what speakers tend to say about the world (i.e., what content would be informative).

We present a replication of Levy, Bicknell, Slattery, and Rayner (2009). In this prior study participants read sentences in which a perceptually confusable preposition (at; confusable with as) or non-confusable preposition (toward) was followed by a verb more likely to appear in the syntactic structure formed by replacing at with as (e.g. tossed) or a verb that was not more likely to appear in this structure (e.g. thrown). Readers experienced processing difficulty upon fixating verbs like tossed following at, but not toward. Levy et al. argued that this suggests readers maintained uncertainty about previously fixated words’ identities. We argue that this finding has wide-ranging implications for language processing theories, and that a replication is required. On the basis of a Bayes Factor Design Analysis we conducted a replication study with 56 items and 72 participants in order to determine whether Levy et al.’s effects are replicable. Using Bayesian statistical techniques we show that in our dataset there is evidence against the existence of the interaction Levy et al. found, and thus conclude that this study is non-replicable.

An active question in psycholinguistics concerns how we mentally encode and retrieve linguistic information in memory. In particular, it remains unclear what information (“cues”) guide retrieval. Previous work has extensively tested retrieval of noun phrases, but less is known about retrieval of other constituents, including verb phrases (VPs). This study examines retrieval for VP ellipsis to allow for a more comprehensive theory of cues. Four experiments (acceptability, self-paced reading) used an interference paradigm to examine voice information (active, passive) in retrieval. Results revealed a selective profile: passive ellipsis shows interference, but active ellipsis does not. These results are aligned with the markedness asymmetry observed for agreement attraction, where marked features (plural, passive) trigger interference, but unmarked features (singular, active) do not. This analysis motivates a unified account of verbal dependencies where markedness plays a more fundamental role than previously assumed. Lastly, I use ACT-R to demonstrate how markedness effects might arise in a cue-based retrieval architecture and discuss the current findings with respect to the leading theories of interference effects.