David Pozen and Thomas Schmidt of Columbia Law School have produced a groundbreaking and—in all the intriguing ways—unsettling study of the constitutional amendment process elucidated in Article V. After reading the abstract, I went into it expecting to disagree. I walked away convinced of their underlying analysis. I’m still ruminating on some of their conclusions.

Pozen and Schmidt point out multiple ambiguities in the text of Article V. Does a Congressional amendment resolution require presidential presentment, for example? Does the margin required for passage in Congress need to be achieved in each chamber or in the aggregate? Does Article V’s reference to Congress finding it “necessary” to amend the Constitution require a formal statement of necessity?

But a larger question lurks behind these: So what? We’ve figured it out in practice, right? Article V may be imprecise, but the process of Constitutional amendment has been “liquidated and ascertained by a series of particular discussions and adjudications,” has it not?

Well, actually, the answer turns out to be “no.” The process of enacting the 27 amendments to the Constitution, going all the way back to the Bill of Rights, has been so ambiguous and varied that Pozen and Schmidt conclude a plausible legal case could be made that nearly every one of them is invalid.

I can’t do justice to the breadth and care of this analysis in this short summary. But it is convincing. The authors note that the ultimate test of whether an amendment is in force is whether society has accepted it. Moreover, they write, ambiguity can serve a constructive purpose, leaving “play in the joints” for the amendment process to work itself out. But strikingly, they also say there may be enough play for the process to be far more responsive to public opinion than we typically expect.

Endorsing “Article V Thayerianism,” Pozen and Schmidt argue that preferably Congress or, failing that, the other branches should take a highly deferential approach to the validity of Constitutional amendments, impeding them only in the case of what James Bradley Thayer called “a clear mistake.”

The implications could be dramatic: More than three-quarters of states have ratified the Equal Rights Amendment. They just did not do so within Congress’ seven-year window. But there is a case, Pozen and Schmidt say, for Congress having the authority to repeal the seven-year limit, which was not part of the text of the amendment. In fact, Congress may not have had the authority to impose that limit in the first place, in which case the ERA may already be the 28^th Amendment. Conversely, the 27^th Amendment—whose sole authority rests on a Justice Department opinion that, contrary to Supreme Court precedent, a transgenerational ratification was permissible—may not be in force.

Radical implications of Constitutional arguments make me antsy, which is why I’m not yet wholly sold on the paper’s recommendations. But it is a terrific read and—literally and healthily—unsettling.