This thesis addresses sound production in bowed-string instruments from two perspectives: the physics of the bowed string, and bow control in performance. Violin performance is characterized by an intimate connection between the player and the instrument, allowing for a continuous control of the sound via the main bowing parameters (bow velocity, bow force and bow-bridge distance), but imposing constraints as well. In the four included studies the focus is gradually shifted from the physics of bow-string interaction to the control exerted by the player. In the first two studies the available bowing parameter space was explored using a bowing machine, by systematically probing combinations of bow velocity, bow force and bow-bridge distance. This allowed for an empirical evaluation of the maximum and minimum bow force required for the production of a regular string tone, characterized by Helmholtz motion. Comparison of the found bow-force limits with theoretical predictions by Schelleng revealed a number of striking discrepancies, in particular regarding minimum bow force. The observations, in combination with bowed-string simulations, provided new insights in the mechanism of breakdown of Helmholtz motion at low bow forces. In the second study the influence of the main bowing parameters on aspects of sound quality was analyzed in detail. It was found that bow force was totally dominating the control of the spectral centroid, which is related to the perceived brightness of the tone. Pitch flattening could be clearly observed when approaching the upper bow-force limit, confirming its role as a practical limit in performance. The last two studies were focused on the measurement of bowing gestures in violin and viola performance. A method was developed for accurate and complete measurement of the main bowing parameters, as well as the bow angles skewness, inclination and tilt. The setup was used in a large performance study. The analyses revealed clear strategies in the use of the main bowing parameters, which could be related to the constraints imposed by the upper and lower bow-force limits and pitch flattening. Further, it was shown that two bow angles (skewness and tilt) were systematically used for controlling dynamic level; skewness played an important role in changing bow-bridge distance in crescendo and diminuendo notes, and tilt was used to control the gradation of bow force. Visualizations and animations of the collected bowing gestures revealed significant features of sophisticated bow control in complex bowing patterns. / QC 20100809