Control of task sequences: what is the role of language?

Abstract

It is almost a truism that language aids serial-order control through self-cuing of upcoming sequential elements. We measured speech onset latencies as subjects performed hierarchically organized task sequences while "thinking aloud" each task label. Surprisingly, speech onset latencies and response times (RTs) were highly synchronized, a pattern that is not consistent with the hypothesis that speaking aids proactive retrieval of upcoming sequential elements during serial-order control. We also found that when instructed to do so, subjects were able to speak task labels prior to presentation of response-relevant stimuli and that this substantially reduced RT signatures of retrieval-however, at the cost of more sequencing errors. Thus, while proactive retrieval is possible in principle, in natural situations it seems to be prevented through a strong "gestalt-like" tendency to synchronize speech and action. We suggest that this tendency may support context updating rather than proactive control.

Figure 1

Sequence of events in the task-span procedure. At the beginning of each block, participants memorized the sequence of response rules at leisure and then initiated the series of stimulus presentations to which the sequence had to be applied repeatedly (i.e., 7 times per block). In the Early Speaking condition, participants were asked to speak the task label during the 1380 ms response-stimulus interval.

Figure 2

Response times since the preceding response as a function of the six sequence positions (positions 1–3 = chunk 1, positions 4–6 = chunk 2) and low versus high positional interference for all conditions. The horizontal dashed line indicates the stimulus onset. The horizontal endpoints of the shaded rectangles represent average speech onsets and offsets for the speaking conditions. The small circles and squares with dashed lines in the Early Speaking condition represent RTs in trials for which the speech onset occurred prior to stimulus onset.

Figure 3

Errors as a function of sequence position and high vs. low positional interference for all conditions. For the two speaking conditions, we distinguished between errors that result from speaking the incorrect task label and response errors that occur despite speaking the correct label.

Figure 4

Individual subject's RTs as a function of speech onset RTs (both residualized with regard to all condition effects) and linear fit statistics (R²) in the Free Speaking condition.

Figure 5

Individual subject's RTs as a function of speech onset RTs (both residualized with regard to all condition effects) and linear fit statistics (R²) in the Early Speaking condition.