Action II: SR-Binding

The Wisconsin Card Sorting Test (WCST) is a gold standard for the neuropsychological assessment of cognitive flexibility. Perseveration errors (repeating a sorting category following negative feedback), are typically treated as indicating reduced cognitive flexibility. Findings from recent research suggested that perseveration propensity is reduced when perseverative responding implies repeating responses. This finding led to the hypothesis that associative learning between cards and responses contribute to behavioral WCST performance. In our first study, we tested the associative learning hypothesis against non-associative accounts such as the avoidance of cards or responses due to negative feedback on the previous trial. We examined 40 participants on a computerized WCST, and we trial-wise manipulated the spatial arrangement of cards (they occurred at fixed or at variable spatial positions). Perseveration propensity was unaffected by pure card repetitions (when punished cards were associated with altered responses) and by pure response repetitions (when punished responses were associated with altered cards). Perseveration propensity was exclusively reduced when the punished cards and their previously associated responses co-occurred, revealing that reduced perseveration propensity does only occur when cards and responses remain bound together. A second study investigated to what degree these card-response associations are effector-specific by manipulating effector sequence (effector switching vs. repetition). Overall, we show that associative learning contributes to behavioral WCST performance, and that the formed associations should be described as stimulus-response bindings. The interpretation of behavioral performance on the WCST as indicating cognitive flexibility should be expanded toward instrumental learning.

In typical dual-task paradigms, the stimulus-response (S/R) pairings typically chosen are visual/manual and auditory/vocal pairings. In the visual/manual pairings, manual responses are exclusively implemented by pushing a button with the dominant hand. These specific S/R pairings are chosen in order to allow for optimal reaction times with minimal dual task costs. But are these S/R-pairings generalizable to other sources of response-input as well? In our study, the participants performed three 2-AFC tasks at the same time. Crucially, the visual/manual S/R pairings were switched in a block-wise fashion. Participants had to respond to a color discrimination task with their foot (condition A) or their hand (condition B), to a cued direction task with their hand (condition A) or their foot (condition B) and to a tone pitch task with their voice. Half of the participants started with condition A, the other half started with condition B. We found that there was a significant difference in re-learning after all switches: the group starting with condition A had significantly bigger problems in recombining the new S/R pairings after the switch compared to the group starting with condition B. This implies that when responding to a visual direction cue with the hand, a stronger S/R bond is formed, compared to responding to a color with the hand. This means that careful consideration is needed in order to select the ideal stimulus-response mappings in basic psychophysical experiments because the strength of the S/R bond directly affects performance.

The present work is based on a theoretical framework explicating the mechanism of how verbal thoughts translate into behavior. The investigated question is whether verbally encountered stimulus-response contingencies result in a response-priming effect upon encountering the stimulus – even if it is irrelevant for the task. In four studies (N = 206), participants memorized a verbal if-then action plan that linked a stimulus ("If I see an *apple*") either to a response that implicated an elbow extension (push) movement ("then I will point at it!") or an elbow flexion (pull) movement ("then I will point at [touch] my chest!"). In a subsequent letter categorization task, images of fruits – including the critical stimulus – were presented for 150 ms before the target letter. Categorization responses were performed by pushing (elbow extension) or pulling (elbow flexion) a joystick. Following the "push" plan, I predicted facilitated joystick push responses when primed with the critical stimulus as compared to the control stimuli (vice versa for the "pull" plan). Whereas descriptive statistics are mostly in the predicted direction, statistical analyses of each study individually fail to reach conventional significance levels. However, analyses of the combined studies – comparing the push versus pull plans – provide evidence that the results are in line with the predictions. The results are discussed in light of the hypothesis that encountering stimulus-response contingencies in a verbal format and by the actual perception of the stimulus and execution of the responses results in similar behavioral consequences.

According to action control theories, responding to a stimulus leads to the integration of stimulus features and the accompanying response. When stimulus features and the response fully repeat in a subsequent trial, the previous information gets retrieved, leading to faster reaction times and lower error rates. However, partial repetitions cause interference, resulting in increased reaction times and error rates. Even distractors irrelevant for task execution can be bound to a response. Although assumed to occur in most intentional actions, it is unclear how eye movements are affected by such distractor-response bindings. In an eye-tracking study, participants had to discriminate appearing target letters by looking at one of two locations on the right side of the screen. Crucially, a shape irrelevant for task execution, that is, a distractor, framed the letter. Whereas reaction times were unaffected, we observed a distractor-response binding effect in saccadic landing positions: Initiated saccades landed closer to the correct location, if response and distractor fully repeated. Partial repetitions caused saccades to land further away, that is, in the direction towards the incorrect location. This is the first study to observe distractor-response binding in eye-movements, showing that eye movements underlie the same action control processes as manual movements do.