Heterophony--where multiple parts independently develop the same material--can challenge theories of texture. Psychological research on auditory streaming supports a theoretical model of texture created by David Huron. His texture space has two dimensions: onset synchrony and semblant motion. Heterophony would theoretically combine low onset synchrony with high semblant motion. However, I argue that some analytical methods cannot detect this combination and propose a new, less biased technique. I use this model to analyze performances from the Chinese Jiangnan sizhu tradition, where players improvise embellishments on a mother tune. Ultimately, I call for further theory and analysis of musical texture across diverse genres.