Matthew Yglesias argues that the virtue of originalism–its reinforcement of the law’s predictability and stability–also sets a limit on originalism. To interpret a law according to its original understanding preserves the stability and predictability of the law. But when precedents have departed from the original understanding, and people have formed expectations based on those precedents, then a return to the original understanding is destabilizing.
There is some force to this argument, which is why almost every originalist would allow anti-originalist precedents to have some weight. I think it is right that the soundest normative argument for originalism has to do with the requirements of the rule of law. The Constitution lays out a process for its own amendment, which does not include amendment through reinterpretation. The trouble with Yglesias’s view is that it seems to dispense altogether with the idea that the Constitution really is a higher law than the Supreme Court’s case law.
Also, I’m not sure how O’Connor’s opinion in Lawrence, which Yglesias discusses, can really be squared with the idea that she trimmed back liberal precedents and expanded conservative ones.