St. Thomas makes frequent use of the principle that “every agent makes a like to itself”. He uses the principle in quite significant ways: it frequently is used to show prove various divine attributes, and it is the basis of our knowledge of God (see ST q 4 a 2 and 3, which are the basis of question 13). The principle is not always stated in those exact words- somethimes it is said that an agent communicates its likeness.
The principle can also be partially justified by pointing out that nothing can give what it does not in some way have. I say that it “partially justifies” because it is not clear, on the face of it, that every operation involves giving and taking. Does falling to the ground involve giving and taking? What about eating asprin or smoothing boards?
But if the principle receives no firmer grounding than “nothing gives what it does not have”, we shouldn’t have too much confidence in it: one can easily imagine thinging: “An agent makes a like to itself! That’s ridiculous! How does this explain aspirin?” One can multiply out examples like this forever: sandpaper makes things smooth because it is rough; men make houses; eggs make animals. Trying to find a likeness here in the most obvious sense of “likeness” (the way a horse sires another horse) is probably a hopeless task, and at any rate, it would only seem like an ad hoc justification of a principle.
This is one of the great dangers in trying to study St. Thomas right away: one runs into principles which, taken in themselves, can seem like hasty inductions, medieval superstitions, slavish devotions to Aristotle, or products of limited experience. The truth of the matter is that these principles in medieval thought are always taken from that very limited field of knowledge about which human beings can have absolute certainty, even though this absolute certainty usually takes many years of meditation (let it be said right away that if a man can imagine that he knows what he does not know, for the same reason a man can imagine that he does not know what in fact he does).
The principle that an agent makes a like to itself is grounded on the first ideas we have of causing, which involves causing motion, or at least being involved with motion in one way or another. Now what is in motion first and primarily is a mobile, or a thing that is able to move. By “able to move” we mean it is receptive of something, or that it is moved. If “receptive” is too metaphysical or strange, we at least have to account for the different things that “it moves” can mean: and in the case of the mobile, what it means is that it is able to get something from another. The mobile might actually cause motion in another too, but insofar as it does this, it is a mover, not a mobile.
Now our ideas of motion are a bit more general than the use of the word “motion” in English. We don’t tend to say that the freezing of a lake is a motion, but we do understand it as a sort of motion: which is why we say it goes from water to ice; or it turns from liquid water to ice. Now there is no single genus of everything that goes from this to that, and so no single account of a mobile and a mover. There are, however, some common things that can be said of all movers and mobiles, even tough all these common things are-just like the mobiles- all in different genera.
A mobile is whatever is able to go from this to that, while a mover is whatever is determining a mobile to that. The word “whatever” is important here: we frequently have no idea what the mobile or the mover is. The problem may even be unsolvable. Something moved, so something is mobile, and something moves.
(We might note here that one needs little more than to meditate on this for a while to see the necessity of “everything in motion is being moved by another” but this is a principle we’ve spoken about in other places.)
Because of all this, it follows that the mover and the mobile are involved in one operation (motion). In fact, the one operation is the first thing we understood, and we then were forced to distinguish two principles involved in the motion. The mover can be understood as what inflows into the other, hence making it “like itself”. Are we being too quick in saying this? Not at all. If something is an agent because it freezes, then it acts by making something frozen; if something is a mover because it cuts, then it acts by making something cut; if something is an agent because it smoothes things (like sandpaper) then it acts by making something smooth; and if something is an agent because it relieves pain, then it acts by making something pain-free. This is the basis upon which “every agent makes a likeness to itself” rests. In all of this, we see the necessity of always remembering the distinction between the per se and the per accidens: a thing acts to take away pain only qua pain relieving agent, not qua medicine or qua chemical, even though every pain relieving agent is a medicine and certainly a chemical.