Bhatt and Quinn (2011) review the substantial evidence that learning constrains perceptual organization in infants. With those findings as a foundation, they discuss five kinds of experiences that engender learning in infants and propose that attention and unitization mediate infant learning. Bhatt and Quinn's article is informative--the ideas regarding kinds of learning and mechanisms are set against an excellent review of the historical background and the contemporary experimental research--and provocative--the proposals are certain to stimulate additional experiments and theory. In this commentary, the author will focus on the deep theoretical implications of Bhatt and Quinn's (2011) demonstrations of the effects of variability on perceptual organization. In what follows, she will first describe the basic findings. Next, she will discuss a conundrum posed by these results and propose a solution. The solution points to facilitation and weight changes rather than either attention or unitization as the mechanism for effects of variability on perceptual organization. She will argue that two other types of learning discussed by Bhatt and Quinn--learning from temporal context and exposure to holistic images--are also likely to be instances of facilitation. A final note will ask whether all instances of learning from variability are open to the alternative interpretation she proposes. (Contains 3 figures and 1 footnote.)