On what basis do these changes take place?
On what basis do these changes take place? It might seem that the division of everything in the world into "substances" (jawahir) and "accidents" ( 'awarid), which most Mutakallimun eventually embraced, already answers the question. Accidents are attributes that bodies acquire, or of which they are deprived: as accidents replace each other, a body's "state" ( hal) changes. From this point of view, the instability of accidents is the cause of the world's transformation.
However, the question of change in the world may be rephrased in that case: what is the cause of the constant coming-and-going of accidents? Even those Mutakallimun who argued that any body always exhibits all of the possible classes of accidents, had to provide an explanation for why the given - and not its opposite - accident is found in the body at a particular moment.
This question was formulated with respect to the "priority" ( awlawiyya) that the existence of one of the two opposite accidents has over the existence of the other. For example, "motion" and "rest" are opposite accidents that equally "deserve" or "have the right" (istihqaq) to be manifested in the body: why then is it one and not the other that gains existential preference at some moment, later giving way to its counterpart?
It is hardly an exaggeration to say that the Mutakallimun advanced almost every possible answer to this question. The variety of their theories is rivalled only by their incompatibility. Some of them reproduced the scheme that explained changes in bodies, to supply a reason for the presence of accidents. There is something that accounts for the existence of the given, as opposed to its opposite, accident, they argued.
This is called ma`na ("meaning"; the term is sometimes translated as "nature" or "idea": see Chittick, 1983, pp. 15, 352; Wolfson, 1965). Motion outweighs rest and exists in the given body because there is the "meaning of motionability" (ma'na al-harakiyya) in that body. The Ash'arite school later expressed this as a general rule: "Any change of attribute*(wasf)* in being is due to some meaning*(ma`na)* that takes place in it" (al-Baghdadi, 1981, p. 55).
Certainly, this way of reasoning provides no final explanation, since it initiates an infinite regress. If any foundation, any "meaning," has to be justified by its own foundation, the resulting chain of principles is unending.