Alan P. Merriam’s definition of ethnomusicology as “the study of music as culture” has sometimes led to the erroneous opinion that we will gain no insights into other cultures by research into their musical practices and the specialized knowledge and conceptualizations that go along with them. To be sure, music is part of a given culture and, therefore, as worthy of scrutiny as any other aspect of a culture. When trying to elucidate emic conceptualizations, however, ethnomusicologists often have to deal with the fact that musical concepts are not always verbalized in indigenous discourse. Also, a strictly intracultural approach usually will not lend itself to communicating its findings cross-culturally. The research project “Virtual Gamelan Graz” can serve as a case in point. By employing an analysis-by-synthesis approach it aims at an evaluation of the knowledge and understanding we have gained regarding Central Javanese gamelan music (karawitan). The idea is to let Javanese music specialists listen to computer-generated renditions of pieces from the traditional repertoire which are idiomatically acceptable in regards both to the sound of the instruments, which is digitally emulated, as well as the performance of the various instrumental parts in the ensemble. Aspects of performance practice like the extent of variability of certain parts or musical “intangibles” like the sound aesthetics of idiophones can then be tackled by interactively modifying pertinent parameters in real-time and having the indigenous music specialists evaluate the sound result, thus avoiding the disadvantages of abstract verbal discourse.