The Predictability Effects and Language Processing

Prior to 2001, the dominant view was that fluctuations in comprehension difficulty were driven by working memory demand. And then there was this alternative called surprising theory, which is that the dominant cost of language processing may actually be a allocating activation among possible interpretations of the unfolding sentence. So for example, as I'm reading like, you know, the cat, a thought, like there's some potential activation allocated to a totally unrelated interpretation about elephants or whatever. But when it comes to cats and eating, we get a lot more information from cat than anything else. That is quantified with the negative log probability of a word in context.