In a survey of textbook definitions of phrase and cadence, Ann Blombach (1987) finds what she calls a "fatal logical flaw: phrase and cadence are being defined circularly." Prompted by this flaw, Blombach calls for more precise definitions of these two terms. Later, Caplin (2004) takes up Blombach's plea by asking the musicological community to "take seriously the idea of perceiving closure in a wide variety of ways...." Despite these calls, recent textbook definitions of phrase and cadence maintain this sense of circularity. This paper responds to Caplin (2004) by reframing the phrase/cadence issue and its implications in formal analysis into a broader discussion of the perception of closure that relies on recent work on cadence. We consider the perception of closure as having three primary components -- boundary detection, process conclusion, and confirmation -- and we show how these three aspects of closure are at play in a variety of examples.