The title of the work refers to the Harvestworks media center in New York City; I am grateful for a residency there that enabled the realization of Part 1 of this composition. It also refers obliquely to a messy ex-housemate of mine, in whose house many of the sounds in this work were recorded.

The speakers are divided into 4 groups (pairs, in an octophonic setting) surrounding the listeners. (In the stereo version, each pair is simply mixed in as a stereo file.) Taking after Carter's poly-ensemble conceptions, each group is treated as a separate instrument or ensemble, with its own musical materials (rhythmic, source and processual). During the course of the work, the speaker groups appear separately on occasion, but mostly combine with each other to form polyensembles: duos and trios, and a few tuttis.

As in Part 1, the speakers in front are called the “Solo” pair, generally featuring the most complex and detailed material, the speakers to one side are called “Earth”, typically featuring arpeggiated ‘mountain sonorities’ or Varese-like “skyscraper chords”. The speakers to the other side are called “Human”, and typically feature vocally-derived sound sources, march-rhythms, or "musical" sounds (like held chords). The rear speakers represent "Water" and often feature highly-filtered, "submerged"-sounding sound materials.

In the Harvest Kitchen pieces, I continued to make use of the “musique concrete gesture engine” that I had used in Sand and Divertimento. A composer enters data about sounds in a sound collection and then specifies musical gestures abstractly by creating parametric profiles called models. Models, realized as sequences of actual soundfiles via database queries, can be written out as mix-files which can then be modified, processed, and mixed. In this way, the submixes of the composition were created; these were then stitched together using ProTools and later, Reaper. You can read more about the engine here: christopherbaileymusic.com/sand/ConcEngine.pdf