The science of sound at the beginning of 17th century

Madalina Giurgea

Ghent University

The main focus of this paper is the move from a qualitative to quantitative explanation of sound at the beginning of the 17th century. More precisely I show that the process of measurement in the study of sound and therefore the quantification by means of experimental practice is problematic. The largely accepted working hypothesis, on which the whole 17th century study of sound is built up, is that sound should be associated with the vibratory motion of the medium through which it moves until it reaches the ear. The next step in the process of quantifying sound was to establish specific one-to-one correlations between the qualitative aspects of sound and quantitative aspects of the sonorous body. These correlations were established either by stipulation (Galileo, Beeckman) or by varying the parameters of the sonorous object in such a way that the audible properties of sound changed in controlled circumstances (Mersenne). The latter alternative offered a way to measure the qualitative properties of sound, such as pitch, in terms of the numbers of vibration made by a string per unit time. The scholarship has often argued that the experimentally oriented approach to sound in the 17th century makes the establishment of the correlations between qualitative and quantitative aspects unproblematic and that by means of instruments such as the monochord, measurement of the properties of sound is made possible. By contrast, I show that there are at least three reasons for revising this interpretation. The first regards the nature of the object under study; there will always be a gap between the sensorial input and the physical effect. The second concerns the inability of experiments in acoustics to isolate a specific sound; Mersenne’s experiment with a long piece of string allows him to visualize the frequency of the string and even to count the number of vibrations, but the sound attributed to that string is inaudible. The third is that, after establishing one-to-one correlations between the physical and audible aspects, the lack of a) a proper measuring instrument and b) a widely recognised unit of measurement impede accurate measurement of the properties of sound. Even though the monochord allows the establishment of correlations between the physical properties of the string and the audible properties of the sound produced by that string (e.g. a change in frequency of string vibration determines a change in pitch of sound), the numerical data attributed for example to an octave were stipulated, rather than empirically determined. For example, Mersenne’s table of consonances is constructed by means of multiplying the pre-established ratios of the musical intervals. Moreover, the results of the measurements varied considerably, because of the imprecise measure of time. The significance of my paper is that it shows that the common view that the account of sound in the 17th century is experimentally orientated (which is supposed to rely on accurate measurements of sound’s properties) should be revised. Thus, mathematical thinking and metaphysical assumptions rather than experimental practice played the central role for natural philosophers, such as Mersenne, Beeckman or Galileo, interested in the science of sound.