Experiments play a central role in science. The role of experiments in computing is, however, unclear. Questions about the relevance of experiments in computing attracted little attention until the 1980s. As the discipline then saw a push towards experimental computer science, a variety of technically, theoretically, and empirically oriented views on experiments emerged. As a consequence of those debates, today's computing fields use experiments and experiment terminology in a variety of ways. This paper analyzes experimentation debates in computing. It presents five ways in which debaters have conceptualized experiments in computing: feasibility experiment, trial experiment, field experiment, comparison experiment, and controlled experiment. This paper has three aims: to clarify experiment terminology in computing; to contribute to disciplinary self-understanding of computing; and, due to computing's centrality in other fields, to promote understanding of experiments in modern science in general.

Via electrophysiological methods, early neural responses to musical chords have been, on one hand, associated with feature encoding, feature change discrimination, rule violation processing, and conscious updating of musical expectations. On the other hand, late neural responses have been related to affective evaluation of sounds. The chronometric succession of these neural processes and their underlying psychological mechanisms related to cognitive and affective aspects of music listening have thus far remained unexplored. Here, the neural correlates of affective and cognitive processing of musical chords were contrasted by means of the event-related potential (ERP) technique and behavioral ratings. Adult subjects (N = 24) performed an emotion judgment (affective) task and a correctness judgment (cognitive) task while listening to chord sequences ending in various major and minor final chords, which were either correctly tuned or mistuned. Enhanced negative ERPs during cadence listening preceding the affective ratings, relative to the cognitive ratings, suggest different neural preparation for these tasks. Furthermore, negatively rated (sad or incorrect) cadence endings in both tasks elicited early negative ERPs and later positive ERPs. These positivities, peaking at 500 ms, differed in scalp distribution between sad and incorrect stimuli. The present findings suggest a neural chronometry of music listening in which feature encoding and sensory memory processes are followed at a medium latency by affective classification, after which an evaluative stage takes place. This study provides a first look at the chronometric succession of electrophysiological brain responses in relation to emotion judgments of musical pitch as opposed to nonaffective correctness judgments.

No abstract

The task-irrelevant spatial location of a cue stimulus affects the processing of a subsequent target. This "Posner effect" has been explained by an exogenous attention shift to the spatial location of the cue, improving perceptual processing of the target. We studied whether the left/right location of task-irrelevant and uninformative tones produces cueing effects on the processing of visual targets. Tones were presented randomly from left or right. In the first condition, the subsequent visual target, requiring response either with the left or right hand, was presented peripherally to left or right. In the second condition, the target was a centrally presented left/right-pointing arrow, indicating the response hand. In the third condition, the tone and the central arrow were presented simultaneously. Data were recorded on compatible (the tone location and the response hand were the same) and incompatible trials. Reaction times were longer on incompatible than on compatible trials. The results of the second and third conditions are difficult to explain with the attention-shift model emphasizing improved perceptual processing in the cued location, as the central target did not require any location-based processing. Consequently, as an alternative explanation they suggest response priming in the hand corresponding to the spatial location of the tone. Simultaneous lateralized readiness potential (LRP) recordings were consistent with the behavioral data, the tone cues eliciting on incompatible trials a fast preparation for the incorrect response and on compatible trials preparation for the correct response.

A classical observation in neurology is that aphasic stroke patients with impairments in speech production can nonetheless sing the same utterances. This preserved ability suggests a distinctive neural architecture for singing that could contribute to speech recovery. However, to date, these structural correlates remain unknown. Here, we combined a multivariate lesion-symptom mapping and voxel-based morphometry approach to analyse the relationship between lesion patterns and grey matter volume and production rate in speech and singing tasks. Lesion patterns for spontaneous speech and cued repetition extended into frontal, temporal and parietal areas typically reported within the speech production network. Impairment in spontaneous singing was associated with damage to left anterior-posterior superior and middle temporal gyri. Preservation of grey matter volume in the same regions where damage led to poor speech and singing production supported better performance in these tasks. When dividing the patients into fluent and dysfluent singers based on singing performance from demographically matched controls, we found that preservation of left middle temporal gyrus was related to better spontaneous singing. These findings provide insights into the structural correlates of singing in chronic aphasia and may serve as biomarkers to predict treatment response in clinical trials using singing-based interventions for speech rehabilitation.