This shows higher sensitivity to relational than non-relational information, consistent with hierarchical incrementality. Fast encoding of an “easy” agent before 400 ms was then followed by a shift of attention to the patient around 400 ms. In find more other words, speakers did not continue fixating the subject character after 400 ms as predicted by the strong version of linear incrementality (Gleitman et al., 2007), but systematically shifted their gaze back to the patient. This type of character-by-character encoding is consistent with a weaker version of linear incrementality, where speakers do attempt to encode information about both

characters early in the formulation process but, crucially, they encode this information sequentially. Thus the rise and fall of fixations to the agent after 400 ms was also predicted by Agent codability: fixations to agents were generally delayed after 400 ms if agents attracted more attention before 400 ms, and vice versa. Specifically, formulation in events with “harder” agents showed that there is a benefit to distributing attention more evenly between the two characters before 400 ms: formulation after 400 ms continued with rapid,

preferential encoding of the agent. Importantly, shifts of gaze to the agent after 400 ms and away from the agent after 1000 ms RG7420 mw were also influenced by Event codability: as predicted by hierarchical incrementality, speakers began fixating the patient earlier in higher-codability

events than lower-codability events. As expected, the lexical primes produced analogous effects to Agent codability: within 400 ms of picture onset, speakers directed more fixations to the agent after agent primes than after patient primes and neutral primes. This shows Cediranib (AZD2171) that the agent primes selectively influenced encoding of the agent character and that they increased the likelihood of speakers prioritizing encoding of this character (non-relational information) shortly after picture onset. A shift of gaze away from the agent was then observed after 400 ms, confirming the tendency to encode sentences character by character after priming non-relational information. Finally, after 1000 ms speakers looked at the agent for less time after agent and patient primes than neutral primes, and thus shifted their gaze to the patient earlier when either character had been primed than when the primes mentioned an unrelated character. Taken together, the results show a direct link between the ease of encoding non-relational and relational information and the timecourse of formulation, both during the early deployment of attention to the subject character and then the deployment of attention to the object character around speech onset.