Discussion

Experiment 2 was designed to test the influence of the continuous presence of tinnitus on attention capture. On the one hand, data revealed that a deviant stimulus captured attention in tinnitus participants. Indeed, they responded less accurately and less fast to categorize the target-stimulus after the presentation of a frequency-deviant stimulus than after the presentation of a standard (i.e., homogeneous) stimulus. This suggests that the preattentive system underlying the automatic detection of deviance is effective in tinnitus participants. However, these effects were modulated by several factors. In particular, the attention capture disappeared when the interval between the to-be-ignored stimulus S1 and the target S2 was the longest. This is in favor with the idea that the detrimental effects of deviants are short-lived [Schröger, 1996]. In addition, the results showed a greater difference between response times to S2 after the presentation of a deviant or a standard S1 when the task had to be performed in the right ear than when it had to be performed in the left ear. It seems likely that the tinnitus patients were better to categorize S2 in the right ear. Very few studies were designed to explore differences between reaction times to sounds presented to the right or the left ear, nevertheless Simon [1967] demonstrated that the reaction times to the onset of sounds presented to the right ear were shorter than the reaction times to the onset of sounds presented to the left ear. Taken as a fact, this result could account for the better performance of our participants when the target-stimulus was presented to their right ear.

On the other hand, the comparison between performance in tinnitus and non-tinnitus ear showed that tinnitus participants categorized more accurately the target stimulus when it was presented in the ear where tinnitus was perceived. In other words, performance were better when the irrelevant stimulus, to-be-ignored, was not presented to the tinnitus ear, i.e. when it was presented to the non-tinnitus ear. This result suggests that the presence of tinnitus could affect the automatic detection of deviance. However, it is difficult to know whether performance of tinnitus participants are enhanced in the tinnitus ear or disturbed in the non-tinnitus ear. We had hypothesized that the presence of tinnitus could enhance categorization performance by facilitating attention focalization on the tinnitus ear, while ignoring the non-tinnitus ear. However, in view of the results of Experiments 1 and 2, tinnitus participants might have overall worse performance than control participants. For statistical comparison, see the “Comparison between groups” section. Thus, we cannot decide among the two above-interpretations. Nevertheless, it is worth noticing that attention capture could be affected by the presence of tinnitus. It seems like tinnitus may not only disturb globally automatic attention processes, but also prevent the deviance detection system from working correctly.

Finally, we could wonder whether the effect of tinnitus on attention processes might be equivalent to the effect of a continuous noise present in one ear. In order to determine whether the results of tinnitus participants could be interpreted in terms of interference between two competitive auditory stimuli (i.e., the target-stimulus and the tinnitus signal) or not, we carried out a third experiment where we « simulate » tinnitus in healthy controls. We reasoned that performance observed in tinnitus-simulated participants should only reflect perceptual differences due to the presence of an interfering stimulus and consequently should be different from those observed in tinnitus participants. Thus, tinnitus-simulated participants should manifest attention capture, but also some differences according to the side of the artificial-like tinnitus signal.