Context effects in a temporal discrimination task" further tests of the Scalar Expectancy Theory and Learning-to-Time models.

Arantes et al. (2008) tested two rival timing theories. One says animals carry a single internal clock. The other says they learn separate time rules for each context.

Pigeons pecked keys for food. Sessions alternated between red and green lights. Each color had its own time rule. The birds never saw both rules at once.

Researchers watched which color the birds preferred when the rules switched. They wanted to see if past red sessions changed choices during green sessions.

The birds acted like they kept two separate clocks. Choices in green sessions were swayed by what happened in past red sessions. This fits Learning-to-Time, not Scalar Expectancy Theory.

Even with no overlap, context history leaked into current timing. The birds did not use one master clock. They used local, context-bound clocks instead.

Alvarez et al. (1998) showed the same thing with concurrent chains. Pigeons followed local cues, not global rates. Joana extends that idea to pure timing tasks.

Preston (1994) already doubted Scalar Expectancy Theory in time-left procedures. Joana gives the theory a direct hit by showing context overrides scalar predictions.

Sherwell et al. (2014) later built a model that adds brief cues to sharpen timing. Their model grows out of the same insight: local context beats pure clock rules.

When you teach a client to wait two minutes before a break, the color of the table, the room, or the teacher’s voice can become part of the timing cue. If you later change rooms, the old context may still control the clock. Check for context drift before you blame non-compliance. Start new timing programs with fresh, salient cues and probe across settings to be sure the skill travels.