The discovery of law
[T]he source of a "new rule" is the Constitution itself, not any judicial power to create new rules of law. Accordingly, the underlying right necessarily pre-exists our articulation of the new rule.
This contradicts the modern academic dogma of "positivism" in which the common law is made, not discovered, and therefore cannot be said to exist in any sense until the decision comes down. This is associated with another dogma, "realism", in which judicial decisions are largely just, as with legislation, exercises in political power.
Legal academics make far too much out of the "made" vs. "discovered" dichotomy. We can see that by looking at the supposed distinction between "discovering" a law of science and "inventing" a new piece of technology. If it's possible to make a strong and important distinction between "discovering" and "making", we should be able to do so here. Yet our patent system would not have been able to transition so smoothly to being able to patent new breeds of plant or new strands of DNA if this distinction were so clear. In patent law the distinction between a "discovered" law and a "made" invention can be as trivial as the distinction between a "discovery" that a high level of a certain hormone correlates with a high risk of a certain disease and the "invention" of a test that allows one to detect the high hormone level and algorithmically (or perhaps even mentally -- it's unresolved law) conclude that one is at serious risk of the disease. In other words, no serious difference at all -- the test apparatus and the mental or algorithmic step was how the discovery was made. To make the discovery is to make the invention, and vice versa, and the idea that there is an strong and important distinction between the two here is risible.
There are thus cases in which the distinction between making and discovering is trivial. But we can readily and fruitfully think of all invention as discovery. What was more important to inventing the first automobile: the physical making and combining of its parts, or the discoveries of ways make and combine them (in this case, discovering what are the better and worse ways to combine an internal combustion engine with a carriage)? One can discover better and worse ways of making things just as one can discover new continents or planets. That, and not whether the physical machine has actually been built, and thus can be said to now physically "exist", lies at the heart of invention. Furthermore, there is a Platonic sense in which these better or worse designs already exist, because the basic laws by which they operate already exist: the designs be deduced from the laws of physics and the history of the part of the universe in which it is desired that they operate. In practice, it this process of discovery requires its own empirical tests: build the machine, try it in the various desired environments, tweak the machine to make it work better. As with discovery those laws themselves, inventors hypothesize which designs are better or worse and test them against reality. They don't exercise their will and arbitrarily decide which machines will work better and which will not.
So too, common law is a painstaking way of discovering and making better law, case by case, dispute by dispute, piece of evidence by piece of evidence. For machines there is a Platonic "environment space", the set of environments we desire those machines to work in. As we explore new environments we discover new ways to make machines work better in them. In common law, there is a "facts space", a Platonic space of possible unique disputes, each with its own unique set of facts. It is not possible to explore more than a miniscule fraction of the facts space.
Instead of the hubristic exercise of waving one's hands calling for "Change" and then drafting broad legislation, the ultimate consequences of which are usually unknown, common law judges tweak the common law. In each case they are discovering a better way in which the law can be made. Just as there are empirically better and worse machines, so there are empirically better and worse laws. If the law applied to the facts of this case is unjust, it needs to be tweaked. Not repealed or overruled, as the absolutist minds of positivists would have -- but tweaked -- by a shade of meaning or an exception for a new situation. To repeal a rule of common law insults the combined wisdom of several previous courts and a variety of cases with their own large bodies of evidence not before the current court. It's like an octopus declaring that, based on his discovery that wheeled vehicles don't work on the ocean, that they won't work on land either, despite many landlubbers who've reported that wheeled vehicles work on land just fine. The principles of precedent or stare decisis allow the law to evolve towards worker better in an ever wider variety of disputes, even though only the facts of one case can be observed at a time.
So there is no dichotomy -- judges both make law and discover law. By not doing so in broad brushstrokes, they can avoid doing the broad unintended harms that legislation usually causes. Legislation purposefully disrupts previously existing legal rights and duties wholesale. Proper common law precedent -- any precedent that should be respected -- never should do this. Instead it should be clarifying the nature of previously unexplored territories of the facts space, the space of possible disputes, and tweak the law so that it works better in the newly explored part of the facts sapce. Incremental discovery of better law justifies the prohibition against ex post facto laws applying to legislation in a way that it does not apply to common law making.
N.B.: the fact that scientific discoveries are called "laws", by metaphor to human laws, should clue us in as to what people generally thought about the nature of law at the time of the scientific revolution, i.e. from the Renaissance through the Enlightenment, and including the drafting and ratification of the U.S. Constitution.