Suppose a rock at the top of a cliff and a feeble little old lady at the bottom of the cliff. It is difficult to lift the rock, it takes a great strain to stop it returning to the ground if it is lifted. It looks easy to knock the little old lady over. The rock is pushed off the cliff and there is nothing to stop it falling on the little old lady below, who isn't quick enough to move. Inevitably, necessarily, the little old lady is crushed by the rock. This is causal necessity; there are elements of a situation sufficient to produce a result and there is nothing in the situation sufficient to stop that result.
But how do you know, given that the rock appeared heavy at the top of the cliff, or just once, that it will continue to feel heavy? And how do you know, given the FEELING of 'being heavy' what the rock will do, which is a completely different thing?
To take the first question;
Just as, by comparison we can see, that if the rock continues to behave as it did at the top of the cliff, as it is pushed off the cliff, it would be sufficient to crush the little old lady, and, apparently, there is nothing to stop it, so the result is inevitable, what is there in the situation to stop it behaving like that? If the rock stopped half way down the cliff, all of a sudden, this would be astonishing, a miracle--how could that happen? we would ask. Where is there in the situation something sufficient to produce this change? But this also applies to the situation of the rock if it behaves in a particular way on one occasion, 'what is there to stop it behaving in that way?'
But we might ask instead 'what is to keep the rock behaving like that?' This last question seems to require some cause sufficient to sustain the rocks behavior, but the way a cause can be sufficient to produce some result is if it is seen to continue, as the weight of our rock was seen, or was supposed to continue, so that it crushed the old lady. So the question 'what keeps the rock behaving like that?' seems to miss understand the basis for sufficient factual explanation and tries to ask for a sufficient factual explanation where the basis for sufficient factual explanation is made the point at issue.
Instead we might have phrased the question 'what reason do you have to suppose the rock will continue to behave like that?' Meaning that we want some logical deduction, or some logical guarantee, or some sort of basis for a probability of what will happen. But the reason involved is not logical deductive in form. The reasons are the same as the answer to the question how the heaviness of the rock could be sufficient to crush the little old lady; we compare the behavior of the rock in one situation with what the result of that behavior would be if placed in another situation, and we can see how that rock with that behavior could get into this other situation, if there is nothing sufficient to stop it.