Recently, turn-taking gaps, or unfilled pauses, have been viewed as a symptom or by-product of predictive planningmechanisms in speech production (Levinson & Torreria, 2015). Other works has shown that gaps can take signalingfunctions and that this is governed by politeness (Bgels, Kendrick, & Levinson, 2015). Two mouse-tracking experimentsexamined when gaps are interpreted as a symptom of processing or as a signal. This was tested by examining how gapsare interpreted in tandem to scalar implicatures (Bonneferon, Dahl, & Holtgraves, 2015). Experiment 1 found that longergaps slightly reduce implicature rates at longer gaps and these longer gaps do not lead to faster implicature responses.Experiment 2 found that filled and unfilled pauses (gaps) both signal hesitation, though filled pauses signaled hesitation atshort gaps. Overall, these experiments show that gaps lengths can have signaling functions beyond politeness and questionbias.