Aprendizaje de las matemáticas

The same cluster pattern was also found for the correlations of latencies. Correlation coefficients for ocular and manual latencies reported in the literature vary between .5 (Prablanc et al. 1979) and .8 (Herman et al. 1981). We found a correlation of r=.6 and higher for intentional conditions. If hand and eye movements rely on a shared common final motor command, a high correlation independent of the paradigm used would be expected. However, our results suggest that the temporal coupling of eye and hand movements is different for reflexive and intentional tasks.

This pattern is confirmed by an analysis of reflexive ocular and manual “pro”- movements (i.e., prosaccades and hand movements towards the physically present cue) versus intentional ocular and manual “anti”-movements (i.e. antisaccades and hand movements towards the imaginary target) occurring within the same antisaccade paradigm. Again, the correlation for correct anti-movements was significantly higher than for reflexive, wrong pro-

movements obtained in the same condition. This implies that signals to initiate movement rely more on the same information for intentional movements than is the case for reflexive movements.

This interpretation agrees with a conclusion that Frens and Erkelens (1991) drew from temporal and spatial data in their study. They suggested that saccades can be generated by two different mechanisms: one relies only on visual information and is used exclusively to generate eye movements; the other relies on visual as well as cognitive information and is used to generate eye and hand movements. Our results suggest that the first mechanism is responsible for the generation of reflexive saccades, while the second operates in both intentional eye and hand movements.

Another observation in our data can be interpreted in terms of the above separation: in the antisaccade paradigm there was a considerably larger number of trials in which only the eye wrongly made a reflexive prosaccade (N=224) compared to trials in which this was the case for eye as well as hand movements (N=35). Trials with a wrong prosaccade only for the hand were absent. This fact might indicate that there is indeed a separate mechanism for the generation of reflexive eye movements.

As to the nature of a common mechanism for generating intentional eye and hand movements, there are at least two possible explanations. From a physiological perspective, the same structures may be involved in the generation of intentional eye and hand movements, e.g., the basal ganglia (as suggested by Frens and Erkelens 1991). The basal ganglia have been shown to play a role in the generation of intentional saccades, because this type of eye movement is impaired by basal ganglia disorders such as Parkinson’s syndrome (e.g., Crawford 1989). Equally, the role of the basal ganglia in the generation of internally triggered arm movements has been demonstrated in non-human primates (van Donkelaar et al. 1999) and in humans (e.g., Georgiou 1997). However, as the respective circuits within the basal ganglia are separate and work in parallel (Alexander et al. 1986), a higher correlation of latencies is difficult to explain. In a recent study comparing human brain areas active during anti-saccades and anti- hand movements a parietal network was found to be active during both eye and hand movements (Connolly et al. 2000). The authors suggest that these areas may be involved in the transformation of visual stimulus location into the location of the anti-target within a common frame of reference.

From a functional point of view, the generation of both intentional eye and hand movements may involve a synchronising process. Intentional tasks require delaying of movement initiation. Given that the delay is large enough so that motor planning is already

completed for eye and hand, both systems remain in a kind of “standby mode”. The go-signal then initiates the movements, thereby synchronising both motor systems.

Alternatively, both motor systems may have been already initiated, but the actual execution may depend on common information not yet available, e.g. about target location retrieved from working memory. Thus, waiting for the common information to be available would necessarily result in a synchronising effect.

2.5.1.2 Movement accuracy

While ocular and manual latencies were organised in the same pattern, errors of both systems were not. An interaction of movement type with task demonstrated that directional and variable errors of eye and hand change differently with the task. Ocular directional error was largest (undershoot) under the conditions memory and antiGap. This might be due to the reduced availability of visual target information. A visual target does not exist for antisaccades, while visual information about target location has already started to fade when the reaction is made for memory saccades. For hand movements, however, the directional error was largest (overshoot) under the conditions proGap and antiGap. This finding cannot be explained by the absence of visual target information, as this is not the case for the condition proGap. Probably it can be attributed to the existence of a temporal gap. A closer inspection of the data revealed that the increased directional error under the condition proGap was mainly due to two subjects who consistently showed a systematic overshoot under this condition.