Rand had it that saying well what a thing is, in its definition, is saying what is its essential nature. What is essential in a thing’s nature is a characteristic of the thing on which the greatest number of its distinctive characteristics depend. That fundamental characteristic is explanatory of the distinct kind of thing a thing is.
Rand noted that when it comes to artifacts, such as furniture, the function of the object is part of the thing’s essential nature. It is essential to a life-jacket that it be buoyant and insoluble in water, for example. Those requirements driven by a certain function would explain why certain materials in certain assemblies can constitute a life-jacket and why certain others cannot.
Similarly, continued life and reproduction of a redbud tree will be fundamental functions with implications for what constitution and conditions are required for it to be a redbud tree.
Things not living and not a human artifact do not have functions as part of their fundamental nature. A rock can be used to throw at a critter, a rock affords the possibility of being used as a missile, but such an affordance is peripheral to its essential nature. Our capture of the essential nature of an igneous rock will not include its usefulness to man. It will include how the rock material is generated. We’ll want to say that igneous rock is rock formed through cooling and solidification of magma or lava. To define what is a solid or what is cooling, we’ll not need to bring any sort of function or genesis into our capture of their essential nature.
An explanatory characteristic in the definition of a thing, in Rand’s account, may be explanatory by cause, but it could be explanation of other sorts. Final causes figure in right definitions of artifacts and living things, but not into definitions of types of rock. Efficient causes (both in Aristotle’s generative sense and in our modern making-happen sense) figure in right definitions of types of rock, but not into definitions of solidity or of cooling.
Material causes, or anyway, specific range of material conditions, figure into all right definitions, at least by implication from essential nature. Formal causes, or anyway formal structure, I say, figure into all definitions. Genus-species is formal structure. Class inclusion and exclusion are formal structure. That is, class and sets are formal structures—tools of ours for better comprehending the world. Measurement and measurement omission in concepts and their inter-relations are formal structures.
Formal structure, it seems, figures fundamentally in artifacts. A lever has the formal structure that is the law of the lever. The buckyball (molecule of 60 carbon atoms) was invented by humans with formal specifications in view. However, it was later discovered that buckyballs have also been formed by nature in outer space. So I’d incline to not confine “formal cause” (which in our modern parlance is no cause at all, only a factor of formal structure) to designed objects.