Table 5 Predictions about our model.

From: The principle of anticipation in language use

Prediction #1

Falsification of prediction #1:

Anticipation should shape our non-linguistic understanding of events.

Anticipation should not shape our non-linguistic understanding of events.

Anticipation should also shape how we express this understanding in language implicitly (e.g. via sentence structures) and explicitly (e.g. via categories such as Discourse Markers).

Linguistic structures should lack implicit expressions (e.g. intonation patterns, context-sensitivity) of anticipation, or categories explicitly expressing anticipations (e.g. Discourse Markers).

Prediction #2:

Falsification of prediction #2:

Our language use can guide our ongoing actions and understanding of language in discourse (e.g. via speech acts). Categories explicitly expressing anticipations should match the anticipations that structures implicitly express, or discourse becomes uninterpretable.

Our language use cannot guide our ongoing actions and understanding of language in discourse (i.e. speech acts may not occur).

Our language use can guide our ongoing actions and understanding of language in discourse (e.g. via speech acts). Categories explicitly expressing anticipations should match the anticipations that structures implicitly express, or discourse becomes uninterpretable.

Categories explicitly expressing anticipations should not match the anticipations that structures implicitly express. Discourse is interpretable irrespective of these mismatches.

  1. Our model makes predictions about how anticipations can affect language. Prediction #1 involves the bottom-up projection of anticipation. If anticipation guides perceptual and cognitive systems, including language, then it should be a guiding principle in language use and comprehension, with some categories expressing anticipation types. Prediction #2 involves the top-down projection from language to other cognitive systems. If language includes anticipation as a guiding principle, then agents should shape non-linguistic anticipations via language (e.g. via speech acts), and categories expressing anticipation types should match this information. The possible falsification of these predictions emerges as a possibility that our model is incorrect.