An Adjunct to the Graphic Method

Abstract

t 1 s O THOSE who have used the graphic method of instrumental investiga1 tion in the field of phonetics its limitations and its advantages are well known; to the wider circle of readers whose contact with instrumental phonetics is through the written word, the disadvantages of this method are probably better known than its virtues. Those who use the expensive oscillographic technique seem to feel called upon to demonstrate, or at any rate to allege, the inadequacy of the graphic method. The application of a modicum of that homely virtue 'horse sense' will suffice to indicate that there are things which the graphic method should not attempt to do, and that there are other things which it can very well accomplish; as Menzerath aptly said: 'Das Problem bestimmt stets die Methode: mit Kanonen schiesst man nicht nach Spatzen.'2 It cannot be too strongly asserted that a graphic record, by whatever method it be obtained, proves nothing until it is adequately interpreted, and it should be clear that no graphic record is easy to interpret. Such records require prolonged study and manifold experience on the part of the experimenter before they yield adequate interpretations. The basic objection to any form of graphic record, whether it be obtained by the graphic method or by the oscillograph, is the fact that this record cannot be made to sound. There is no direct evidence from the record of what sounds were recorded or of what the recorded utterance sounded like. To meet this objection I have for two years been making phonographic recordings simultaneously with my graphic records. This is accomplished by the simple expedient of building a small microphone into the mouthpiece used for kymographic recording and leading the current from this microphone to an apparatus for recording on aluminum discs. I use aluminum rather than acetate because it is adequate to my needs and much cheaper. All this requires no special mechanical skill beyond that necessary for any sort of phonographic recording, and it requires no assistance from another person. The records thus obtained are, of course, not so good qualitatively as they would be if the mouth were not covered by the mouthpiece; a certain amount of 'muffling' is unavoidable. But at least one may thus hear what goes on inside the mouth-