After his father and other close relatives died in a short period, singer, composer and multi-instrumentalist Asa Horvitz wanted a musical language to evoke our metaphysical and tangible experiences with loss. Horvitz and a team of friends assembled a dataset of over 150 texts from throughout history that dealt with grief and fed them through a custom Natural Language Processing AI system. Based on the cryptic text that resulted, they created a music-theatre production which toured in Europe and is presented here as a series of standout recordings from the project’s long genesis. GHOST’s compositions started as vocal improvisations led by Horvitz and bassist/singer-songwriter Carmen Quill, accompanied by Ariadne Randall and Bryan West (processing, synths, viola da gamba, etc). Later, Horvitz’s uncle—esteemed pianist and composer Wayne Horvitz—interpolated additional passages. Recalling (at turns) Robert Ashley’s spoken-word operas, Arthur Russell’s vaporous pop song, medieval music, outre hip-hop, and the undulations of Einstein on the Beach, the music remains soulful despite its partially systematized means of production. GHOST derives its strength from its empathetic perspective, tracking the tidal patterns that underpin our grief without attempting to fabricate their logic. – Winston Cook Wilson