Columbia Law School Columbia Law School
Scholarship Archive Scholarship Archive
Faculty Scholarship Faculty Publications
2021
The Puzzles and Possibilities of Article V The Puzzles and Possibilities of Article V
David E. Pozen
Columbia Law School
, dpozen@law.columbia.edu
Thomas P. Schmidt
Columbia Law School
, tschmidt@law.columbia.edu
Follow this and additional works at: https://scholarship.law.columbia.edu/faculty_scholarship
Part of the Constitutional Law Commons, and the Law and Politics Commons
Recommended Citation Recommended Citation
David E. Pozen & Thomas P. Schmidt,
The Puzzles and Possibilities of Article V
, 121 COLUM. L. REV. 2317
(2021).
Available at: https://scholarship.law.columbia.edu/faculty_scholarship/3161
This Article is brought to you for free and open access by the Faculty Publications at Scholarship Archive. It has
been accepted for inclusion in Faculty Scholarship by an authorized administrator of Scholarship Archive. For more
information, please contact scholarshiparchive@law.columbia.edu.
COLUMBIA LAW REVIEW
VOL. 121 DECEMBER 2021 NO. 8
2317
ARTICLES
THE PUZZLES AND POSSIBILITIES OF ARTICLE V
David E. Pozen* & Thomas P. Schmidt.**
Legal scholars describe Article V of the U.S. Constitution, which sets
forth rules for amending the document, as an uncommonly stringent and
specific constitutional provision. A unanimous Supreme Court has said
that a “mere reading demonstrates” that “Article V is clear in statement
and in meaning, contains no ambiguity, and calls for no resort to rules
of construction.” Although it is familiar that a small set of amendments,
most notably the Reconstruction Amendments, elicited credible challenges
to their validity, these episodes are seen as anomalous and unrepresenta-
tive. Americans are accustomed to disagreeing over the meaning of the
constitutional text, but at least in the text itself we assume we can find
some objective common ground.
This paper calls into question each piece of this standard picture of
Article V. Neither the language nor the law of Article V supplies a deter-
minate answer to a long list of fundamental puzzles about the amend-
ment process. Legally questionable amendments have not been the
exception throughout U.S. history; they have been the norm. After detail-
ing these descriptive claims, the paper explores their doctrinal and theo-
retical implications. Appreciating the full extent of Article V’s ongoing
ambiguity, we suggest, counsels a new approach to judging the validity
of contested amendments, undermines some of the premises of originalism
and textualism, and helps us to see new possibilities for constitutional
change. Because the success or failure of attempted amendments turns
out not to be exclusively or even primarily a function of following the
* Charles Keller Beekman Professor of Law, Columbia Law School.
** Academic Fellow, Columbia Law School. For helpful comments on earlier drafts,
we thank Richard Albert, Akhil Amar, Mark Barenberg, Will Baude, Seyla Benhabib, Jessica
Bulman-Pozen, Josh Chafetz, Mike Dorf, Katherine Franke, Maeve Glass, Jamal Greene,
Vicki Jackson, Jeremy Kessler, Sandy Levinson, Jane Manners, Tom Merrill, Gillian Metzger,
Henry Monaghan, Robert Post, Richard Re, Steve Sachs, Fred Schauer, Mike Seidman,
Ganesh Sitaraman, T.J. Smith, Larry Solum, Julie Suk, Seth Tillman, Larry Tribe, Mark
Tushnet, and John Vile, as well as participants at a Columbia Law School workshop. For
valuable research assistance, we thank Alak Mehta and Alexandra Nickerson. For archival
assistance, we thank David Langbart from the National Archives and Edith Sandler from the
Library of Congress.
2318 COLUMBIA LAW REVIEW [Vol. 121:2317
rules laid out in the canonical document, all constitutional amending in
an important sense takes place outside Article V.
I
NTRODUCTION ....................................................................................... 2319
I. “EXTERNAL” LIMITS TO ARTICLE V’S (OR ANY AMENDING CLAUSE’S)
RESOLVING POWER............................................................................ 2325
II. “INTERNAL” LIMITS TO ARTICLE V’S RESOLVING POWER: TEXT ....... 2328
A. Interpretive Puzzles ................................................................... 2328
B. Putting These Puzzles in Perspective ....................................... 2334
III. “INTERNAL” LIMITS TO ARTICLE V’S RESOLVING POWER:
P
RECEDENT ....................................................................................... 2339
A. The Bill of Rights ...................................................................... 2339
B. The Eleventh Amendment ....................................................... 2342
C. The Twelfth Amendment ......................................................... 2346
D. The Reconstruction Amendments ........................................... 2347
E. The Sixteenth Amendment ...................................................... 2351
F. The Seventeenth Amendment ................................................. 2353
G. The Eighteenth Amendment ................................................... 2354
H. The Nineteenth Amendment ................................................... 2356
I. The Twenty-First Amendment .................................................. 2358
J. The Twenty-Second Amendment ............................................. 2360
K. The Twenty-Seventh Amendment? .......................................... 2362
L. The Twenty-Eighth Amendment? ............................................ 2365
1. Article the First ................................................................... 2365
2. The Titles of Nobility Amendment .................................... 2367
3. The Equal Rights Amendment .......................................... 2368
M. An Article V Convention? ......................................................... 2370
IV. LIVING WITH ARTICLE V AMBIGUITY: JUDGING CONTESTED
AMENDMENTS .................................................................................... 2371
A. The Highly Incomplete Liquidation of Article V .................... 2372
B. Revisiting Coleman ..................................................................... 2377
V. EMBRACING ARTICLE V AMBIGUITY: LESSONS FOR INTERPRETERS AND
REFORMERS ....................................................................................... 2384
A. Originalism and Textualism ..................................................... 2384
B. Amendment Inside and Outside Article V ............................... 2387
C. Loosening the Constitutional Cage Through Article V
Thayerianism ............................................................................. 2389
CONCLUSION ........................................................................................... 2395
2021] PUZZLES AND POSSIBILITIES 2319
INTRODUCTION
One of the distinctive features of the original U.S. Constitution was its
capacity for lawful change. In George Washington’s words, the
Constitution “contain[ed] within itself a provision for its own amend-
ment.”
1
That provision was and is Article V, which instructs that an amend-
ment shall become “Part of this Constitution” when “propose[d]” by “two
thirds of both Houses” of Congress and “ratified by the Legislatures of
three fourths of the several States, or by Conventions in three fourths
thereof.”
2
The Supreme Court gave voice to the standard view of Article V
when it wrote in 1956 that “[n]othing new can be put into the Constitution
except through the amendatory process. Nothing old can be taken out
without the same process.”
3
Constitutional theorists have challenged or complicated this view in a
variety of ways. Akhil Amar has suggested that the Constitution may be
amended by a national popular referendum.
4
Some have proposed that
norms without a foothold in the canonical document may nevertheless
attain “constitutional” status.
5
And many have emphasized the extent to
1. George Washington, Farewell Address (Sept. 17, 1796), in 1 A Compilation of the
Messages and Papers of the Presidents, 1789–1907, at 213, 217 (James D. Richardson ed.,
1908); see also Carl J. Friedrich, Constitutional Government and Democracy: Theory and
Practice in Europe and America 138 (4th ed. 1968) (“[T]he clear recognition and deliber-
ate organization of the amending power was an achievement of the American revolution.”);
Sanford Levinson, Introduction: Imperfection and Amendability, in Responding to
Imperfection: The Theory and Practice of Constitutional Amendment 3, 4 (Sanford
Levinson ed., 1995) [hereinafter Responding to Imperfection] (“It was a fundamental
breakthrough in American constitutional theory . . . that the ‘rules of government’ would
be decidedly ‘alterable’ through a stipulated legal process.” (quoting Willi Paul Adams, The
First American Constitutions: Republican Ideology and the Making of the State
Constitutions in the Revolutionary Era 139–44 (1980))).
2. U.S. Const. art. V. Article V also permits “the Legislatures of two thirds of the sev-
eral States” to apply to Congress to “call a Convention for proposing Amendments,” which
then go to the states for ratification. Id. No such convention has yet been called. See infra
section III.M.
3. Ullmann v. United States, 350 U.S. 422, 428 (1956); see also, e.g., Akhil Reed Amar,
The Consent of the Governed: Constitutional Amendment Outside Article V, 94 Colum. L.
Rev. 457, 458 (1994) [hereinafter Amar, Consent of the Governed] (“The conventional
reading of Article V sees it as the exclusive mechanism of lawful constitutional amend-
ment . . . .”); Frederick Schauer, Amending the Presuppositions of a Constitution, in
Responding to Imperfection, supra note 1, at 145, 147 [hereinafter Schauer, Amending the
Presuppositions] (“Article V appears on its face to exhaust the possibilities for amending
the Constitution consistent with the Constitution itself . . . .”).
4. See Amar, Consent of the Governed, supra note 3, at 462–94; Akhil Reed Amar,
Philadelphia Revisited: Amending the Constitution Outside Article V, 55 U. Chi. L. Rev.
1043, 1044–76 (1988) [hereinafter Amar, Philadelphia Revisited].
5. See, e.g., Richard Primus, Unbundling Constitutionality, 80 U. Chi. L. Rev. 1079,
1129 (2013) [hereinafter Primus, Unbundling Constitutionality] (explaining that “small-c”
constitutional theorists insist that “the text of the written Constitution” is not a necessary
“basis for a rule’s constitutionality”); Ernest A. Young, The Constitution Outside the
Constitution, 117 Yale L.J. 408, 410 (2007) (“[T]he American ‘constitution’ consists of a
2320 COLUMBIA LAW REVIEW [Vol. 121:2317
which new propositions of supreme law, including propositions that
depart sharply from prior understandings, may emerge and become
entrenched in the absence of formal amendment.
6
Virtually all constitu-
tional lawyers, however, take as given that conformity with Article V’s
“amendatory process” has determined ever since the Founding what is and
is not “put into” the written Constitution, and therefore what its text does
and does not say.
7
This paper questions Article V’s capacity to perform that function.
8
It
is by now familiar that the perceived clarity of the constitutional text is
“constructed” to a significant degree by norms of legal argument and
other social practices.
9
We endeavor to show that what counts as the
constitutional text in the first place is also constructed to a significant degree
by such practices.
10
Part of the reason is that the ultimate rule of recogni-
tion in any system is a matter of ocial and popular acceptance, rather
much wider range of legal materials than the document ratified in 1789 and its subsequent
amendments.”).
6. See, e.g., Richard S. Kay, Formal and Informal Amendment of the United States
Constitution, 66 Am. J. Compar. L. 243, 260 (2018) [hereinafter Kay, Formal and Informal
Amendment] (“It is fair to say that most of what now goes under the caption ‘constitutional
law’ in the United States is attributable to extraconstitutional, ‘o-the-books’ develop-
ments.”); Reva B. Siegel, Constitutional Culture, Social Movement Conflict and
Constitutional Change: The Case of the De Facto ERA, 94 Calif. L. Rev. 1323, 1327 (2006)
(arguing that social “movements regularly succeed in changing the Constitution without
amending it”).
7. Even the most radical Article V revisionists do not necessarily dispute this. Alt-
hough Amar contends that the Constitution may lawfully be amended by a popular referen-
dum or comparable mechanism, he never suggests that such an amendment has in fact
occurred. See, e.g., Amar, Consent of the Governed, supra note 3, at 457–61. Likewise, alt-
hough Bruce Ackerman advances an elaborate theory of constitutional change outside
Article V, he never suggests that new words have been put into the Constitution’s text by a
procedure that does not purport to follow Article V. See, e.g., 2 Bruce Ackerman, We the
People: Transformations 15–31 (1998) [hereinafter Ackerman, We the People].
8. Although this piece is technically an “Article,” it will be referred to as a “paper”
throughout to avoid confusion with references to Article V.
9. See, e.g., Curtis A. Bradley & Neil S. Siegel, Constructed Constraint and the
Constitutional Text, 64 Duke L.J. 1213, 1216 (2015) (“The perceived clarity of the text . . .
is often partially constructed by [interpretive] practice.”); Peter Jeremy Smith, Commas,
Constitutional Grammar, and the Straight-Face Test: What if Conan the Grammarian Were
a Strict Textualist?, 16 Const. Comment. 7, 17 (1999) (discussing “all-too-common”
situations in which “we are confronted by [constitutional] text that, according to established
rules of grammar, does not actually say what we all know it to mean”).
10. Amar has made a conceptually similar, but empirically narrower, point about the
original Constitution. While most assume that the parchment Constitution in the National
Archives is the authoritative document, a printed version with minor dierences was trans-
mitted to and ratified by the states. See Akhil Reed Amar, America’s Unwritten Constitution:
The Precedents and Principles We Live By 63–68 (2012) [hereinafter Amar, America’s
Unwritten Constitution]. That latter version is probably the legally operative one. Id. But
there would be no way to ascertain which version is legally operative simply by studying the
documents themselves. Id.; cf. Laurence H. Tribe, The Invisible Constitution 6 (2008)
[hereinafter Tribe, The Invisible Constitution] (“[N]othing in the visible text can tell us
that what we are reading really is the Constitution . . . .”).
2021] PUZZLES AND POSSIBILITIES 2321
than constitutional design. Even an amending clause that looks itself like
the system’s “supreme criterion of law”
11
owes its ecacy, and indeed its
legality, to extratextual forces.
12
But another, more U.S.-specific part of the
reason—and the one on which this paper focuses—is that neither the
language of Article V nor subsequent constructions of Article V specify the
amendatory process in enough detail to establish in many cases which
amendments are valid and which are not. Article V continues to be
shrouded in a remarkable amount of legal uncertainty, which further
attenuates the link between its contents and the failure or success of any
given amendment eort.
13
Leading scholars have characterized Article V as an unusually clear
and constraining constitutional provision.
14
The Supreme Court has said
that a “mere reading demonstrates” that “Article V is clear in statement
and in meaning, contains no ambiguity, and calls for no resort to rules of
construction.”
15
Yet, as we explain, the text of Article V leaves open
numerous fundamental questions, from the time limits (if any) on an
amendment’s pendency to the substantive limits (if any) on an amend-
ment’s subject matter to the role (if any) of the President and state gover-
nors in the amendatory process to the respective roles (if any) of Congress
and the courts in deciding whether an amendment has been validly
adopted.
16
Debates during the drafting and ratification of the Constitution
11. Kent Greenawalt, The Rule of Recognition and the Constitution, 85 Mich. L. Rev.
621, 632 (1987) [hereinafter Greenawalt, Rule of Recognition].
12. See infra Part I.
13. A “successful” eort to enlist Article V, for purposes of this paper, is one that results
in a new amendment widely understood to have become part of the written Constitution.
In other words, we equate success with sociological legitimacy. And we contend that amend-
ment success is not exclusively or even primarily a function of following the rules laid out in
Article V.
14. See, e.g., Adrian Vermeule, Judging Under Uncertainty: An Institutional Theory
of Legal Interpretation 270 (2006) [hereinafter Vermeule, Judging Under Uncertainty]
(listing the “rules governing . . . constitutional amendment” as an example of “clear and
specific constitutional text[]”); Richard Albert, Constitutional Disuse or Desuetude: The
Case of Article V, 94 B.U. L. Rev. 1029, 1035 (2014) [hereinafter Albert, Constitutional
Disuse or Desuetude] (“The simplicity and clarity of Article V’s enabling clause allow us to
identify when the Constitution has been formally amended . . . .”); David R. Dow, The Plain
Meaning of Article V, in Responding to Imperfection, supra note 1, at 117, 117 (“The mean-
ing of Article V . . . is an example of yet another text the meaning of which is essentially
clear.”); John Harrison, The Lawfulness of the Reconstruction Amendments, 68 U. Chi. L.
Rev. 375, 457–61 (2001) (depicting Article V as privileging “clarity” and “certainty” over
other values).
15. United States v. Sprague, 282 U.S. 716, 730 (1931).
16. See infra section II.A. This is by no means the first work to observe that Article V is
vague or underspecified in certain respects. See, e.g., Richard B. Bernstein with Jerome
Agel, Amending America: If We Love the Constitution So Much, Why Do We Keep Trying
to Change It? 248 (1993) (“The procedures outlined in Article V pose a host of unresolved
diculties.”); Walter Dellinger, The Legitimacy of Constitutional Change: Rethinking the
Amendment Process, 97 Harv. L. Rev. 386, 432 (1983) [hereinafter Dellinger, Legitimacy
of Constitutional Change] (acknowledging that “[t]he spare language of article V leaves
2322 COLUMBIA LAW REVIEW [Vol. 121:2317
shed hardly any light on these questions.
17
More than two centuries later,
post-ratification practice has done little to resolve them, or to establish
much of anything concerning the never-used amendment-through-con-
vention procedure. As a result, the overwhelming majority of amendments
added to the Constitution since 1787 have faced credible challenges to
their validity—challenges that were beaten back by proponents at the time
but that in many respects have never been definitively dispelled—while
other amendments have plausibly satisfied Article V’s formal criteria yet
nevertheless failed to gain widespread acceptance.
18
Charles Black once advised Congress that “[f]undamental law should
be not merely of arguable, but of clear legitimacy,” and that accordingly
the “legitimization of constitutional amendments” is an area “where, per-
haps more than anywhere else, square corners should be cut.”
19
The actual
experience of constitutional amendment throughout U.S. history has
been far messier. Article V contains so many ambiguities and lacunae that
it can be expected to yield, and in fact has yielded, amendments of only
“arguable” legal legitimacy at the time of their adoption.
20
Consider, in
this regard, that no fewer than twenty-six of our twenty-seven recognized
amendments failed to comply with a requirement of presidential approval
that Black himself found “plain” on the face of the Constitution.
21
open critical questions,” yet contrasting Article V with “open-textured provisions of the
Constitution”). And “some scholars have cited diculties arising out of the ratification of a
handful of constitutional amendments.” Saikrishna Bangalore Prakash, Of Synchronicity
and Supreme Law, 132 Harv. L. Rev. 1220, 1270 n.267 (2019) [hereinafter Prakash, Of
Synchronicity]. So far as we are aware, however, this is the first work to document all of the
major unresolved legal questions raised by Article V and to explore their collective signifi-
cance for constitutional law, politics, and theory.
17. See Thornton Anderson, Creating the Constitution: The Convention of 1787 and
the First Congress 161 (1993) (explaining that “the consideration of amending procedures
was one of the least adequate of the Convention debates”); Carlos A. González,
Representational Structures Through Which We the People Ratify Constitutions: The
Troubling Original Understanding of the Constitution’s Ratification Clauses, 38 U.C. Davis
L. Rev. 1373, 1444 (2005) (“To the extent that Article V was discussed during the
Philadelphia drafting convention and in the subsequent ratification process, only Article V’s
federalism and entrenchment [implications]—whether Article V made amendment too
dicult—were mentioned.”).
18. See infra Part III.
19. Charles L. Black, Jr., Amending the Constitution: A Letter to a Congressman, 82
Yale L.J. 189, 190, 209 (1972).
20. Cf. Richard H. Fallon, Jr., Legitimacy and the Constitution, 118 Harv. L. Rev. 1787,
1794–801 (2005) [hereinafter Fallon, Legitimacy and the Constitution] (defining legal
legitimacy and contrasting it with sociological and moral legitimacy). As this paper uses the
term, an amendment enjoys “legal legitimacy” at the time of its adoption if the amend-
ment’s proposal and ratification complied with Article V. A constitutional amendment, or
for that matter an entire constitution, may come to be accepted as authoritative notwith-
standing defects under then-existing law in the process by which it was created. See id. at
1803–06.
21. Article I, Section 7, Clause 3 of the Constitution states that “[e]very Order,
Resolution, or Vote” of Congress “shall be presented to the President” and must be
2021] PUZZLES AND POSSIBILITIES 2323
Recent controversies over the Twenty-Seventh Amendment and the
Equal Rights Amendment (ERA) underscore just how many questions
about Article V remain unsettled at this late date.
22
The Twenty-Seventh
Amendment was “ratified” by the requisite number of states nearly 203
years after it was proposed by Congress alongside the amendments that
became the Bill of Rights. As the Supreme Court opined in the 1921 case
Dillon v. Gloss, there is a serious objection that this violates an implicit con-
dition of Article V that ratification take place within a reasonable time
frame.
23
The Dillon Court specifically stated that it was “quite untenable”
to think that what is now the Twenty-Seventh Amendment could be revived
“by some future generation.”
24
And the Justice Department’s Oce of
Legal Counsel disagreed with members of Congress over whether the
Archivist of the United States was required to refer the amendment to
Congress prior to its ocial certification.
25
Meanwhile, the ERA has not, at this writing, been accepted by the
legal community as part of the Constitution because several state
ratifications occurred after a deadline imposed by Congress, among other
complications, even though the amendment seems to have checked all of
the boxes for validity indicated on the face of Article V. In addition to the
ERA, five amendments have been proposed by Congress but never ratified
by a sucient number of states.
26
Or, at least, so goes the conventional
wisdom. A recent lawsuit contends that the other amendment proposed
by Congress alongside the current Twenty-Seventh and the Bill of Rights,
regarding congressional apportionment, did receive the requisite number
of ratifications in the eighteenth century.
27
These unsuccessful amend-
ments dwell in a kind of legal purgatory due to the apparent acceptance
approved by the President before it “take[s] Eect.” According to Black, this instruction
would seem to mean, “if plain words can have plain meaning,” that once passed by Congress
a proposed amendment must be approved by the President. Charles L. Black, Jr., On Article
I, Section 7, Clause 3—And the Amendment of the Constitution, 87 Yale L.J. 896, 899 (1978)
[hereinafter Black, On Article I]. Only one amendment has ever received a presidential
signature prior to state ratification: President Abraham Lincoln’s on the Thirteenth. See
Harrison, supra note 14, at 389 n.79; infra note 130. The so-called Corwin Amendment,
proposed by Congress in 1861 but never ratified, was “inadvertently presented” to and then
signed by President James Buchanan. Cong. Globe, 38th Cong., 2d Sess. 630 (1865) (state-
ment of Sen. Trumbull); see also Bernstein with Agel, supra note 16, at 91. The presentment
issue is discussed in detail infra section III.B.
22. For more extensive discussion of both amendments, see infra sections III.K, III.L.3.
23. 256 U.S. 368, 375 (1921) (“We conclude that the fair inference or implication from
Article V is that the ratification must be within some reasonable time after the proposal.”).
24. Id.
25. See Cong. Pay Amendment, 16 Op. O.L.C. 85, 99–105 (1992); Richard B.
Bernstein, The Sleeper Wakes: The History and Legacy of the Twenty-Seventh Amendment,
61 Fordham L. Rev. 497, 540–41 (1992).
26. See H.R. Doc. No. 110-50, at 29–31 (2007) (listing unratified amendments).
27. See LaVergne v. U.S. House of Representatives, 392 F. Supp. 3d 108, 114 (D.D.C.
2019) (three-judge court). The dispute hinges on whether Connecticut validly ratified the
2324 COLUMBIA LAW REVIEW [Vol. 121:2317
of the Twenty-Seventh Amendment after its long dormancy. Even more
striking, there is a colorable argument that enough states have “applied”
for a constitutional convention to obligate Congress to call one—even
though few, if any, members of Congress appear to realize this.
28
Beyond its intrinsic interest, Article V’s ambiguity carries significant
doctrinal and theoretical implications. On the doctrinal side, it points
toward a new defense of, and twist on, the Supreme Court’s ruling in
Coleman v. Miller that Congress has the power to “promulgate” or “pro-
claim” constitutional amendments after ratification.
29
Many commenta-
tors have criticized Coleman on textual and historical grounds, but the
persistent controversy over the validity of amendments suggests a distinct
prudential rationale for allowing one organ of government to resolve the
status of a new amendment more quickly and democratically than the
Court is capable of. Congress, we suggest, is the branch best suited for this
task—and could further bolster its comparative competence through the
use of subconstitutional mechanisms such as special commissions and
advisory referenda.
On the jurisprudential side, our account informs multiple debates
about constitutional change through and beyond Article V. Because the
success or failure of an attempted amendment bottoms on social
acceptance, which throughout U.S. history has not turned on punctilious
adherence to a set of rules, all constitutional amending in an important
sense takes place “outside” as well as “inside” Article V. Textual and extra-
textual considerations are entwined right from the start of the law-recog-
nition process. Article V, in consequence, may have more play in the joints
than is typically realized. Given the extreme antidemocratic potential of
the double-supermajoritarian Article V formula, there is a strong case for
what might be called Article V Thayerianism: an interpretive presumption
favoring ease of amendability on those (many) questions that Article V
does not clearly resolve. Vicki Jackson has warned that “sociocultural
beliefs in the diculty of amendment . . . may contribute to the diculty
of amendment today,” as claims about the impossibility of amendment
“can become self-fulfilling.”
30
Our descriptive analysis of Article V bears
out this warning, while our proposed adaptation of Thayerianism fur-
nishes a practical tool for breaking out of the vicious cycle that Jackson
identifies. At the same time, our showing of the constructedness of the
constitutional text holds lessons for constitutional interpretation more
amendment when each house of its legislature approved the amendment seven months
apart, during dierent legislative sessions. See infra notes 280–288 and accompanying text.
28. See infra section III.M.
29. 307 U.S. 433, 450 (1939); see infra section IV.B.
30. Vicki C. Jackson, The (Myth of Un)Amendability of the US Constitution and the
Democratic Component of Constitutionalism, 13 Int’l J. Const. L. 575, 576–77 (2015)
[hereinafter Jackson, The (Myth of Un)Amendability] (emphasis omitted).
2021] PUZZLES AND POSSIBILITIES 2325
generally, as it undermines some of the positivist premises of originalism
and textualism.
31
The paper proceeds in five parts. Part I sets the general jurispruden-
tial stage by explaining the inherent limits of Article V, or any constitu-
tion’s amending clause, to determine which eorts at constitutional
change will be seen as legally valid. Part II catalogs the many questions
about the amendment process that the text of Article V fails to answer, and
it explains that these uncertainties are striking both from a comparative
perspective and because of Article V’s unique function as the gateway to
the constitutional text. Part III provides a historical review of amendment
eorts, which reveals that legally plausible contestation over amendment
validity is the norm in U.S. practice, not the exception, and that many
important questions about the amendment process remain unsettled.
Parts II and III are the empirical centerpiece of the paper. Taken together,
they confound any notion that Article V is a clear
32
or “straightforward”
33
guide to amendment, even though there is arguably no more fundamental
issue in U.S. law than what is or is not inscribed in the constitutional text.
Moving from deconstruction to reconstruction, Part IV considers doctrinal
implications of this account and argues, in particular, that it provides a
stronger basis for Coleman than the reasons given by the Court. Finally, Part
V explores broader implications for constitutional theory and
interpretation.
I.
“EXTERNAL” LIMITS TO ARTICLE V’S (OR ANY AMENDING CLAUSE’S)
RESOLVING POWER
The main descriptive burden of this paper is to demonstrate that
Article V of the U.S. Constitution is significantly less clear and constrain-
ing, and therefore significantly less determinative of the success or failure
of attempted amendments, than is generally assumed. But before turning
to those claims, let us imagine a hypothetical country, Sovereignia, with an
amending clause in its written constitution, known as Article X, that is as
precise as a legal directive can be, spelling out in meticulous detail and
with meticulous care the requirements for any amendment to be added.
Would the requirements of Article X, in themselves, dictate which
attempted amendments are considered part of the supreme law of
Sovereignia?
They would not be capable of doing this for at least two reasons. First,
in the terminology of Friedrich Waismann, the sort of language used in
31. On all points previewed in this paragraph, see infra Part V.
32. See supra notes 14–15 and accompanying text.
33. See, e.g., Paul M. Schwartz, Constitutional Change and Constitutional
Legitimation: The Example of German Unification, 31 Hous. L. Rev. 1027, 1088 (1994)
(“Article V sets out a relatively straightforward process for changing the Constitution.”).
2326 COLUMBIA LAW REVIEW [Vol. 121:2317
any amending clause might be “open texture[d].”
34
Even the most “care-
fully delimited empirical terms,” according to Waismann, “might never-
theless produce uncertainty in the face of unforeseen and virtually
unimaginable instances.”
35
And because legal rules are written in and
dependent on “empirical” language, law might necessarily be open tex-
tured as well—not just in a system that treats legal rules as defeasible but
also in one that treats the literal meaning of rules as binding in all cases,
without exceptions or modifications to avoid absurd or unjust outcomes.
36
If this is right, then no matter how detailed the drafting of Article X or
how rigid Sovereignia’s interpretive culture, the potential for future
uncertainty based on unanticipated developments will remain. When such
developments arise, the plain meaning of Article X will run out; any result-
ing disputes as to the validity of an amendment will have to be resolved
with reference to something other than the internal resources of Article
X.
37
Second, and more fundamentally, even if no confusion ever arises as
to the meaning of Article X, nothing contained within Article X or any
other clause of Sovereignia’s constitution can explain why Article X is
treated as the authoritative means of amending the constitution—or
ensure that it will continue to be treated this way in the future. It is always
possible that the citizens and ocials of Sovereignia will decide one day to
interpret Article X in a less literal manner; or to recognize alternative
amendment rules as equally authoritative; or to deny the validity of an
amendment that indisputably satisfied the requirements of Article X; or to
ignore the requirements of Article X altogether (as the Founding genera-
tion of Americans eectively did with the amendment procedures in the
34. Friedrich Waismann, Verifiability, in Logic and Language: First and Second Series
117, 119–20 (Antony Flew ed., 1951).
35. Frederick Schauer, On the Open Texture of Law, 87 Grazer Philosophische
Studien 197, 198 (2013) [hereinafter Schauer, Open Texture] (discussing Waismann’s
ideas); see also Stewart Shapiro & Craige Roberts, Open Texture and Analyticity, in
Friedrich Waismann: The Open Texture of Analytic Philosophy 189, 192 (Dejan Makovec
& Stewart Shapiro eds., 2019) (describing open texture as “a kind of semantic indetermi-
nacy” that cannot be ruled out in advance “since we do not know where the indeterminacy
comes from”). Waismann was not entirely clear what he meant by associating open texture
with “empirical” terms and concepts, see Joost Jacob Vecht, Open Texture Clarified,
Inquiry, July 2020, at 4, but no one seems to dispute that codified legal language qualifies
as empirical in his sense.
36. See Schauer, Open Texture, supra note 35, at 202 (“If Waismann’s idea is sound,
then open texture is indeed an ineliminable feature of law . . . precisely because it is
ineliminable in language.”).
37. Although we find his formulation helpful, we do not mean to endorse all of
Waismann’s ideas or to wade into debates about open texture in the philosophy of language.
The point here is simply that even the clearest amendment rule conceivable may give rise
to legal underdeterminacy.
2021] PUZZLES AND POSSIBILITIES 2327
Articles of Confederation
38
); or to ignore their written constitution alto-
gether. That the people of Sovereignia treat Article X as the exclusive gate-
way to constitutional amendment is a contingent social fact, subject to
change whether or not Article X itself changes. All constitutions and all
constitutional provisions, as Frederick Schauer and others following
H.L.A. Hart have detailed, “owe their ‘constitutionality’ to logically and
politically antecedent conditions” that “are not themselves legal or consti-
tutional in any important sense.”
39
Article X may inform the rule of recog-
nition in Sovereignia with respect to the validity of constitutional
amendments.
40
But the ultimate rule of recognition determining what is
and is not considered the law of Sovereignia, or any other system, is
predicated on public acceptance.
41
None of the above is meant to be controversial. The practical-minded
reader might fairly ask, though, how much these conceptual caveats mat-
ter. In countries such as the United States, one particular amending clause
has been treated as the gateway to the constitutional text. While many
38. The Articles of Confederation required any amendment to “be agreed to in a
Congress of the United States, and be afterwards confirmed by the legislatures of every
State.” Articles of Confederation of 1781, art. XIII, para. 1. On the Constitution proposed
by the Philadelphia Convention as a violation of this requirement, see Bruce Ackerman &
Neal Katyal, Our Unconventional Founding, 62 U. Chi. L. Rev. 475, 479–80 (1995); Richard
S. Kay, The Illegality of the Constitution, 4 Const. Comment. 57, 67–70 (1987). Even a con-
stitution that purported to make itself “unalterable,” such as the constitution for the
Carolinas drafted by John Locke, Fundamental Consts. of Carolina of 1669, art. 120, could
in practice be amended if there were sucient public and ocial support.
39. Schauer, Amending the Presuppositions, supra note 3, at 160–61; see also, e.g.,
Larry Alexander & Frederick Schauer, Rules of Recognition, Constitutional Controversies,
and the Dizzying Dependence of Law on Acceptance, in The Rule of Recognition and the
U.S. Constitution 175, 192 (Matthew Adler & Kenneth Einar Himma eds., 2009) (noting
“the unavoidable dependence of law on the nonlegal environment in which it exists, not
simply to decide how law should be interpreted . . . but more broadly to determine just what
is to count as law and what is not”); Laurence H. Tribe, Taking Text and Structure Seriously:
Reflections on Free-Form Method in Constitutional Interpretation, 108 Harv. L. Rev. 1221,
1291 (1995) (“Ultimately, one must step outside the Constitution—as with any legal text—
to identify criteria for legitimating that body of law . . . .”).
40. But cf. Michael C. Dorf, How the Written Constitution Crowds Out the
Extraconstitutional Rule of Recognition, in The Rule of Recognition and the U.S.
Constitution, supra note 39, at 69, 73 (observing that “one must look outside the
Constitution itself for the rules and standards governing how amendments are recognized
as having satisfied the criteria” of an amending clause).
41. See Schauer, Amending the Presuppositions, supra note 3, at 150 (“The ultimate
rule of recognition is a matter of social fact, and so determining it is for empirical investiga-
tion rather than legal analysis.”); Young, supra note 5, at 422 (describing the ultimate rule
of recognition as “predicated on social acceptance”); see also H.L.A. Hart, The Concept of
Law 97–120 (1961) (discussing the “ultimate rule of recognition” in similar terms). We
bracket the questions of exactly how and by whom a norm must be “accepted” to qualify as
legally valid or sociologically legitimate. Cf. Fallon, Legitimacy and the Constitution, supra
note 20, at 1805 (“elid[ing]” the same questions).
2328 COLUMBIA LAW REVIEW [Vol. 121:2317
argue that the small-c constitution, or the “constitution in practice,”
42
has
been updated through judicial rulings, framework statutes, and more, no
one seriously suggests that the big-C or written Constitution has been
revised through a process that does not purport to comply with Article V.
43
The requirements of a clear and concrete amending clause, widely seen as
the definitive route to formal constitutional change, could go a very long
way toward determining which attempted amendments are accepted as
valid and which are not.
In principle, then, the success or failure of attempted amendments
rests unavoidably on extralegal or prelegal foundations. But in practice,
such success or failure might, in certain systems, be tightly tied to the rules
of an amending clause, so that the sociological legitimacy as well as the
legality of amendments is in eect a function of their fit with those rules.
Alternatively, an amending clause might contain vague language yet none-
theless give rise to a body of law that supplies determinate legal answers to
the vast majority of real world questions raised by amendment eorts.
44
Is either true of the United States?
II.
“INTERNAL” LIMITS TO ARTICLE V’S RESOLVING POWER: TEXT
The short answer is no. This Part first tours the many ambiguities and
lacunae in Article V’s text, before discussing why these underdetermi-
nacies are so significant.
45
The next Part then surveys the history of suc-
cessful constitutional amendments to show that the lived experience of
Article V has been, and continues to be, beset by legal uncertainty.
A. Interpretive Puzzles
Article V provides in full:
42. David A. Strauss, The Irrelevance of Constitutional Amendments, 114 Harv. L. Rev.
1457, 1459 (2001) [hereinafter Strauss, Irrelevance of Amendments] (noting the distinction
between “the small-‘c’ constitution—the fundamental political institutions of a society, or
the constitution in practice—and the document itself”); see also Primus, Unbundling
Constitutionality, supra note 5, at 1082 (associating the small-c constitution with “the web
of documents, practices, institutions, norms, and traditions that structure American
government”).
43. See supra notes 4–7 and accompanying text.
44. See Kent Greenawalt, How Law Can Be Determinate, 38 UCLA L. Rev. 1, 56–62
(1990) (discussing circumstances in which “open-ended” constitutional directives can yield
determinate legal answers).
45. We describe Article V as “underdeterminate,” rather than “indeterminate,”
because it does have a core of relatively clear meaning. Cf. Lawrence B. Solum, On the
Indeterminacy Crisis: Critiquing Critical Dogma, 54 U. Chi. L. Rev. 462, 473 (1987) (distin-
guishing between underdeterminacy and indeterminacy in law). Essentially every reader of
Article V agrees, for instance, that the “Houses” of Congress referenced therein are the U.S.
Senate and House of Representatives rather than, say, the personal residences of individual
members.
2021] PUZZLES AND POSSIBILITIES 2329
The Congress, whenever two thirds of both Houses shall deem it
necessary, shall propose Amendments to this Constitution, or, on
the Application of the Legislatures of two thirds of the several
States, shall call a Convention for proposing Amendments,
which, in either Case, shall be valid to all Intents and Purposes,
as Part of this Constitution, when ratified by the Legislatures of
three fourths of the several States, or by Conventions in three
fourths thereof, as the one or the other Mode of Ratification may
be proposed by the Congress; Provided that no Amendment
which may be made prior to the Year One thousand eight
hundred and eight shall in any Manner aect the first and fourth
Clauses in the Ninth Section of the first Article; and that no State,
without its Consent, shall be deprived of its equal Surage in the
Senate.
46
This run-on sentence immediately presents a host of puzzles that, the
next Part shows, are not clearly resolved by context. Bracketing for the
moment all issues regarding how “a Convention for proposing
Amendments” is supposed to work, consider the following questions
raised by each clause in order.
• “two thirds of both Houses”: Does this voting rule require the support
of two-thirds of each chamber, voting separately, or two-thirds of
the total membership of Congress, voting together?
47
And does it
require the support of two-thirds of all members of the House and
Senate or only the support of two-thirds of those present, assuming
there is a quorum? Or two-thirds of those present and voting, such
that abstaining members can contribute to the quorum without
46. U.S. Const. art. V. The text reproduced here and in other quotations of the
Constitution is drawn from the parchment version on display in the National Archives. As
noted above, a distinct version of the Constitution was printed and distributed to the states,
which may be more authoritative because it was voted upon by state ratifying conventions.
See supra note 10; see also Denys P. Myers, For a Master File of the Constitution, in S. Doc.
No. 87-49, at 67, 70–71 (1961) (noting “slight” discrepancies among the texts that were
voted on by the state conventions). These two versions contain numerous dierences in
capitalization and punctuation; for a list of all the variations in Article V, see Philip Hu,
The Constitution of the United States: A Variorum 21–22 (Apr. 15, 2017), https://ssrn.com/
abstract=2778049 [https://perma.cc/947X-52J9] (unpublished manuscript). In addition, a
third version of the Constitution was printed on September 18, 1787, “at the Philadelphia
Convention’s behest,” and “formed the basis for the earliest newspaper printings of the
Constitution.” Philip Hu, How Dierent Are the Early Versions of the United States
Constitution? An Examination, 20 Green Bag 2d 163, 165 (2017). That version contained a
significant typo in Article V. It said that Congress could not regulate the slave trade prior to
the year 1708 rather than 1808—rendering the provision inoperative. Id. at 171–72. The
typo was corrected in most, but not all, early newspaper and pamphlet printings
disseminated to the public as the debates over ratification began. See Leonard Rapport,
Printing the Constitution: The Convention and Newspaper Imprints, August–November
1787, Prologue, Fall 1970, at 69, 82–83. This all goes to show yet again that the literal identity
of “the constitutional text” rests on extratextual foundations.
47. See Dow, supra note 14, at 118 (noting this ambiguity); Kay, Formal and Informal
Amendment, supra note 6, at 244 (same). In several other places, the Constitution uses the
clearer phrase “Each House.” E.g., U.S. Const. art. I, § 5, cls. 1–3.
2330 COLUMBIA LAW REVIEW [Vol. 121:2317
counting in the denominator for the two-thirds calculation?
48
Must
both chambers vote on an amendment within a single session? If
not, does an amendment passed by one chamber remain pending
and available for approval by the other chamber indefinitely?
Once an amendment has cleared the two-thirds hurdle (however
understood) and been “propose[d]” to the states, may Congress
rescind it?
49
If so, does rescission likewise require two-thirds super-
majorities?
50
• “shall deem it necessary”: What does it mean for an amendment to
be “necessary”? Is Congress required to make any kind of finding
of necessity in the process of promulgating an amendment?
51
Is
such a finding, whether express or implicit, a legal determination
reviewable by a court?
• “Amendments”: Is there any limitation on the scope of permissible
constitutional reforms implicit in this term? Could an entirely new
Constitution be passed through Article V’s procedures, or would
that no longer constitute an “Amendment”? Does the choice of
the term “Amendment” suggest that constitutional changes pur-
sued through Article V must be incremental and limited to a “per-
fecting” role?
52
48. See Adrian Vermeule, The Constitutional Law of Congressional Procedure, 71 U.
Chi. L. Rev. 361, 409 (2004) (“The framers . . . blundered by leaving open the critical inter-
pretive question whether the express majority quorum for ordinary majority voting still
obtains where the Constitution requires a supermajority of the votes cast.”). An additional
set of puzzles, not specific to the amendment process, concerns how to calculate a quorum
if a large proportion of Congress is incapacitated or deceased. See Howard M. Wasserman,
Continuity of Congress: A Play in Three Stages, 53 Cath. U. L. Rev. 949, 959 (2004) (“The
Constitution does not define whether the quorum requires a majority of authorized seats in
a house or a majority of occupied seats, of living, selected, and sworn members of that
house.”).
49. Compare, e.g., Akhil Reed Amar, America’s Constitution: A Biography 456 (2005)
[hereinafter Amar, America’s Constitution] (yes), with John Alexander Jameson, A Treatise
on Constitutional Conventions: Their History, Powers, and Modes of Proceeding 634 (4th
ed. 1887) (no).
50. See, e.g., Michael Stokes Paulsen, A General Theory of Article V: The
Constitutional Lessons of the Twenty-Seventh Amendment, 103 Yale L.J. 677, 728–29 (1993)
[hereinafter Paulsen, A General Theory] (arguing in the armative while noting alternative
possibilities).
51. See 1 Annals of Cong. 430 (1789) (statement of Rep. Vining) (contending that
both houses must agree that an amendment is “necessary” before deliberating on a
proposal).
52. See Philip A. Hamburger, The Constitution’s Accommodation of Social Change,
88 Mich. L. Rev. 239, 300–01 (1989) (arguing that most Framers and ratifiers envisioned
amendments as limited to a “perfecting” role); see also Thomas M. Cooley, The Power to
Amend the Federal Constitution, 2 Mich. L.J. 109, 118 (1893) (“[A]n amendment . . . in the
very nature of the case, must be in harmony with the thing amended, so far at least as con-
cerns its general spirit and purpose.”); Walter F. Murphy, Merlin’s Memory: The Past and
Future Imperfect of the Once and Future Polity, in Responding to Imperfection, supra note
1, at 163, 177 (“The word amend . . . means to correct or improve; amend does not mean ‘to
2021] PUZZLES AND POSSIBILITIES 2331
• “ratified”: When and how must a state legislature “ratify” an
amendment for that ratification to be eective? Is there a time
limit? Must both houses of a bicameral state legislature approve an
amendment in the same legislative session? Must they adopt a sim-
ple majority voting threshold, or can they set their own voting pro-
cedures?
53
Does it matter if those procedures are contained in the
state constitution? And must each state legislature vote on the
identical text, or can dierences in punctuation or wording defeat
ratification of an amendment?
• “Legislature”: What is a “Legislature” for purposes of Article V? If
the people of a state participate directly in the process of making
laws, can they qualify as the “Legislature”?
54
May a state hold an
advisory or binding referendum on ratification? If the state gover-
nor plays a role in the normal legislative process, must the gover-
nor play the same role in the Article V process?
55
If the lieutenant
governor may break a tie in the state senate for other matters, like
the Vice President does at the federal level, may the lieutenant
governor cast a tiebreaking vote for ratification?
56
• “three fourths”: How should “three fourths” be understood in cir-
cumstances where the denominator is not cleanly divisible by
four?
57
If a state initially rejects a proposed amendment, can the
deconstitute and reconstitute,’ to replace one system with another or abandon its primary
principles.”); Roman J. Hoyos, Article V and the Law of Constitutional Conventions 28–29
(2020), https://ssrn.com/abstract=3723119 [https://perma.cc/74ZL-3FA9] (unpublished
manuscript) (maintaining that “the 1787 convention’s decision to use the term amendment
rather than alteration suggests that Article V does not envision a general revision power”).
53. See Dyer v. Blair, 390 F. Supp. 1291, 1304 (N.D. Ill. 1975) (three-judge court)
(“[T]he Constitution is totally silent with respect to the procedure which . . . each state leg-
islature . . . should follow in performing its ratifying function.”). This question is important
because of the wide variety of procedures followed by states to ratify an Article V amend-
ment. See Matt Gehring, Rsch. Dep’t, Minn. House of Reps., United States Constitutional
Amendment Process: Legal Principles for State Legislators 15–37, 52–55 (2016), https://
www.house.leg.state.mn.us/hrd/pubs/conamendlegal.pdf [https://perma.cc/AU3G-HQ6W]
(collecting state constitutional provisions and legislative rules).
54. Cf. Ariz. State Legislature v. Ariz. Indep. Redistricting Comm’n, 576 U.S. 787, 813–
24 (2015) (answering in the armative in the Elections Clause context).
55. See Herman V. Ames, The Proposed Amendments to the Constitution of the
United States During the First Century of Its History, in 2 Annual Report of the American
Historical Association for the Year 1896, at 5, 297 (1897) (“There has been a great lack of
uniformity in the actual practice by the governors of the States in this respect.”).
56. See Coleman v. Miller, 307 U.S. 433, 447 (1939) (dividing equally on whether this
question is justiciable).
57. In the First Congress, a Senate committee apparently took the position that “nine
states out of thirteen had suced to ratify the Bill of Rights,” even though three-fourths of
thirteen is 9.75. David P. Currie, The Constitution in Congress: The Federalist Period, 1789–
1801, at 208 (1997) [hereinafter Currie, The Federalist Period] (citing 6 Annals of Cong.
1537 (1797)). South Carolina Representative Robert Goodloe Harper countered with the
arresting claim that “there must be twelve ratifying States to be three-fourths [of fourteen],
as intended by the Constitution, because that number would be three-fourths of sixteen,
2332 COLUMBIA LAW REVIEW [Vol. 121:2317
state later change its mind and be counted toward three-fourths
(and, if so, how)? Conversely, once a state has ratified an amend-
ment, can it subsequently rescind its ratification (and, if so, how)?
What if only one house of the state legislature votes to rescind rat-
ification? If new states join the Union while an amendment is
pending, do they necessarily aect the numerator and
denominator?
• “Conventions”: If an amendment is ratified by state conventions
rather than legislatures, who chooses the process to be followed—
Congress or each state? Are there any limits or requirements
regarding how a convention is to operate? How are the delegates
to such conventions selected? Can a statewide referendum or
other plebiscitary process qualify as a “convention”?
• “one or the other Mode of Ratification”: Does Congress have complete
discretion in specifying which mode of ratification shall be fol-
lowed? Or are there certain sorts of amendments that must be rat-
ified pursuant to certain modes? When choosing a mode of
ratification, may Congress put a time limit on it (and, if so, how)?
Once in place, may such time limits be extended (and, if so, how)?
For example, does the imposition or modification of a deadline
require a two-thirds vote?
These are some of the uncertainties that are apparent on the face of
Article V and, as the next Part details, have given rise to legal controversy.
If one burrows into the text’s silences and elisions, the uncertainties
compound. For instance, is Article V the sole method of amendment con-
sistent with the Constitution, or does the Constitution preserve for “We
the People” the option to amend the text through other routes, such as a
national referendum?
58
Apart from the procedures used, are there sub-
stantive limits on acceptable amendments implicit in the constitutional
structure or in the concept of popular sovereignty?
59
For example, could
which was the nearest number to fourteen capable of four equal divisions.” 6 Annals of
Cong. 2281 (1797).
58. The most famous argument for an unenumerated constitutional “right” to amend
the document outside Article V belongs to Amar. See supra notes 4, 7 and accompanying
text. As Amar notes, Article V “emphatically does not say that it is the only way to revise the
Constitution.” Amar, Consent of the Governed, supra note 3, at 459. Amar concedes that
government ocials are stuck with Article V; his argument applies only to “People-driven”
amendment eorts. Id. at 460. For a prominent rebuttal of Amar on textual and historical
grounds, see Henry Paul Monaghan, We the People[s], Original Understanding, and
Constitutional Amendment, 96 Colum. L. Rev. 121, 130–73 (1996).
59. For illustrative discussions of this issue, see Richard Albert, America’s Amoral
Constitution, 70 Am. U. L. Rev. 773, 786 (2021) (arguing that “nothing in America’s mod-
ern Constitution is legally immune to change”); Douglas Linder, What in the Constitution
Cannot Be Amended?, 23 Ariz. L. Rev. 717, 733 (1981) (arguing that the amending power
remains limited by Article V’s express prohibition against restructuring the Senate and by
an implied prohibition against amendments that “create any new limitations on the amend-
ing power” itself); Walter F. Murphy, An Ordering of Constitutional Values, 53 S. Cal. L.
2021] PUZZLES AND POSSIBILITIES 2333
the First Amendment be repealed, or is it too fundamental to the republi-
can underpinnings of the constitutional project?
60
Could an amendment
change the amendment process or make itself unamendable? If not, does
it follow that the one unexpired substantive limit stated in the text of
Article V—the final clause providing that no amendment shall deprive a
state of “equal Surage” in the Senate without its consent—is void?
61
An additional set of uncertainties involves the legal role of the
President. There is no express mention of the executive branch in Article
V. But Article I, Section 7, Clause 3 of the Constitution seems to demand
that every order, resolution, or vote requiring the concurrence of the
House and Senate be presented to the President for approval.
62
Must every
Article V amendment, then, be presented to the President before being
transmitted to the states for ratification? If so, does the President have the
power to veto a proposed amendment?
Yet another fertile source of uncertainty is the provision for a state-
initiated convention to propose amendments to the Constitution, which
has never been successfully invoked.
63
Article V says virtually nothing about
how this process works. How does a state submit an “application” to
Congress?
64
May a state rescind an application once submitted? Do appli-
cations for a convention remain pending in perpetuity, or do they expire
after some period of time? Can the states call a “limited” convention to
craft amendments for a particular purpose, or even to propose a specific
Rev. 703, 754–57 (1980) (arguing that amendments must be consistent with human dignity
to be valid); John R. Vile, Limitations on the Constitutional Amending Process, 2 Const.
Comment. 373, 380–87 (1985) [hereinafter Vile, Limitations on the Amending Process]
(disputing Linder’s and Murphy’s positions).
60. See, e.g., John Rawls, Political Liberalism 239 (expanded ed. 2005) (suggesting that
an “amendment to repeal the First Amendment and replace it with its opposite” should be
deemed “invalid” by the Court); Amar, Consent of the Governed, supra note 3, at 505 (con-
tending that repealing the core of the First Amendment would be “unconstitutional . . .
despite formal compliance with Article V”).
61. See Linder, supra note 59, at 722–28 (reviewing arguments to this eect); see also
Vile, Limitations on the Amending Process, supra note 59, at 379 (explaining that some
members of Congress argued in the period leading up to the Civil War “that the equal suf-
frage provision is not legally binding” but rather “a mere declaration”). The other substan-
tive limit stated in the text of Article V (“no Amendment which may be made prior to [1808]
shall in any Manner aect the first and fourth Clauses in the Ninth Section of the first
Article”) has been a legal nullity since 1808.
62. U.S. Const. art. I, § 7, cl. 3; see supra note 21 and accompanying text.
63. For broad attempts to think through many of these uncertainties, see generally
Russell L. Caplan, Constitutional Brinksmanship: Amending the Constitution by National
Convention (1988); John R. Vile, Conventional Wisdom: The Alternate Article V
Mechanism for Proposing Amendments to the U.S. Constitution (2016); Arthur Jacobson,
John C. Maloney, Jr. & Kenneth F. Ripple, Article V and the Proposed Federal Constitutional
Convention Procedures Bills: Report and Recommendation to the New York State Bar
Association by the Committee on Federal Constitution, 3 Cardozo L. Rev. 529 (1982).
64. Cf. Harisay v. Clarno, 474 P.3d 378, 379 (Or. 2020) (determining, as “an issue of
first impression,” that Oregonians’ “initiative power” does not “authorize the people to
directly apply for a federal constitutional convention”).
2334 COLUMBIA LAW REVIEW [Vol. 121:2317
amendment, or would a convention have plenary power to devise any
amendments it wants?
65
If a limited convention is permissible, how similar
must the various state applications be to trigger such a convention? Does
Congress have any discretion in determining whether a convention should
be called or in specifying its procedures and the scope of its authority?
How would delegates to such a convention be apportioned and selected?
What internal procedures would the convention follow to draft and
approve amendments? Would the delegates vote by state, as in the
Philadelphia Convention, or according to some population-based for-
mula? Whatever the answer to these questions, note that the product of
any such convention would be submitted to the states, thus layering on top
of these uncertainties all of the uncertainties regarding the state ratifica-
tion process cataloged above.
The final lacuna in Article V, which Part IV revisits, is in some ways
the most fundamental. Given the myriad puzzles raised by the text, it will
often be unclear whether the requirements of Article V have been
satisfied. Who decides, then, whether an amendment has been validly
promulgated? If there is a disagreement among the branches, does any
branch have the ultimate say? And does an amendment become eective
as soon as the final state needed to reach three-fourths ratifies it, or only
when ratification has been confirmed by the appropriate federal organ?
B. Putting These Puzzles in Perspective
There are many provisions of the Constitution that are vague or
underdeterminate in certain respects and many questions of constitu-
tional law that have not yet been resolved by the Supreme Court. What is
special about the case of Article V? Two things, we think: the extent of the
underdeterminacy and the function of Article V in our constitutional order.
The sheer number of questions about constitutional amendment that
Article V leaves open is striking. The questions, moreover, are not limited
to subsidiary or minor matters. The silences and ambiguities of Article V—
How many members of Congress must vote on an amendment? How long
may an amendment remain pending? How is a state ratifying convention
to be constituted? Does the President have a role? Is there anything an
65. See Robert G. Natelson, Proposing Constitutional Amendments by Convention:
Rules Governing the Process, 78 Tenn. L. Rev. 693, 715 (2011) (“Perhaps no Article V ques-
tion has been debated so fiercely, on so little evidence, as whether applying states may limit
the scope of a convention for proposing amendments.”). Compare Michael B. Rappaport,
The Constitutionality of a Limited Convention: An Originalist Analysis, 81 Const. Comment.
53, 56 (2012) (contending “that the original meaning of the Constitution allows for limited
conventions” and “forbids runaway conventions”), with Michael Leachman & David A.
Super, Ctr. on Budget & Pol’y Priorities, States Likely Could Not Control Constitutional
Convention on Balanced Budget Amendment or Other Issues 2 (2017), https://
www.cbpp.org/sites/default/files/atoms/files/7-16-14sfp.pdf [https://perma.cc/22GZ-
M389] (“There is no guarantee that a convention could be limited to a particular set of
issues . . . .”).
2021] PUZZLES AND POSSIBILITIES 2335
amendment cannot change?—go to the heart of the amendment process
and the conception of popular sovereignty that it embodies.
It is dicult to nail down the extent to which Article V, or any legal
directive, is underdeterminate. But perhaps some headway can be made
by drawing two comparisons. The first is to other structural provisions of
the U.S. Constitution. The Constitution devotes more than six times the
number of words to presidential eligibility and elections as it does to con-
stitutional amendments, going into such niceties as what constitutes a
quorum and how electoral votes are to be transmitted to the capital.
66
The
Constitution devotes nearly three times as many words to the rules for deal-
ing with presidential vacancies and disabilities.
67
Both of these other pro-
cesses have generated high-profile legal controversies of their own,
including in and around the 2020 presidential race.
68
But much more
often, they have generated legally uninteresting compliance, whether
because their language is more determinate than that of Article V, because
frequency of usage enhances legal clarity over time, or both. At a mini-
mum, the Constitution’s treatment of presidential change, as compared to
its treatment of constitutional change, shows that its authors were capable
of greater procedural specificity when they wished.
Moreover, the legal uncertainty associated with Article V is not just a
product of a spare text and limited precedent. It also follows from the
absence of an authoritative interpreter. For most provisions of the
Constitution, as every U.S. law student learns, the federal judiciary is widely
understood (and understands itself) to enjoy interpretive supremacy.
69
When the judiciary has recognized exceptions to this rule, it is generally
on account of “a textually demonstrable constitutional commitment of the
issue to a coordinate political department.”
70
Thus, while the Constitution
is vague about many details of the impeachment process, the Supreme
Court accepts that the Senate has “the sole Power to try all
Impeachments.”
71
By contrast, the Court, Congress, and the executive
66. Compare U.S. Const. art. V (143 words), with id. art. II, § 1, cls. 1–5, (479 words),
and id. amend. XII (398 words).
67. Compare id. art. V (143 words), with id. amend. XXV (388 words).
68. See Ann Gerhart, Election Results Under Attack: Here Are the Facts, Wash. Post,
https://www.washingtonpost.com/elections/interactive/2020/election-integrity/ (on file
with the Columbia Law Review) (last updated Mar. 11, 2021) (compiling Electoral College
challenges pursued by the Trump campaign and allied groups); Brett Samuels, Pence
Rejects Calls to Invoke 25th Amendment to Remove Trump, Hill (Jan. 12, 2021),
https://thehill.com/homenews/administration/533957-pence-rejects-calls-to-invoke-25th-
amendment [https://perma.cc/YVL6-F9Q9] (describing House Democrats’ claims that
President Trump was unable to discharge the duties of the oce).
69. See David E. Pozen & Adam M. Samaha, Anti-Modalities, 119 Mich. L. Rev. 729,
769–70 (2021) (documenting this feature of the U.S. system). This is not to deny that “many
dierent government bodies and civil-society groups contribute to the long-run develop-
ment of constitutional law.” Id. at 770.
70. Baker v. Carr, 369 U.S. 186, 217 (1962).
71. U.S. Const. art. I, § 3, cl. 6; see Nixon v. United States, 506 U.S. 224, 229–38 (1993).
2336 COLUMBIA LAW REVIEW [Vol. 121:2317
branch have each, at times, claimed interpretive primacy over the question
whether an amendment has become part of the Constitution.
72
The second comparison that throws some light on the extent of
Article V’s underdeterminacy is to amending clauses in other democra-
cies’ written constitutions. Consider Canada. As Walter Dellinger has
observed, the “detailed provisions” of the “Canadian amendment proce-
dures . . . answer several perplexing questions that Article V of the
American Constitution has left to speculation.”
73
The Canadian text makes
clear that a “proposed amendment lapses unless ratified by the requisite
number of assemblies within three years of the adoption of the resolution
which initiated the amendment procedure.”
74
The approval of an amend-
ment may be revoked at any time before the amendment is proclaimed.
75
The amendment process is not itself amendable, except with the agree-
ment of every province and the federal parliament.
76
Pursuant to an
“intricate” framework elaborated across a dozen subsections, “[e]ach of
Canada’s five formal amendment procedures is specially designated for
amending specific constitutional provisions.”
77
Canada’s amendment scheme may be particularly well reticulated,
but many other constitutions around the world specify time limits for
ratification of amendments,
78
substantive limits on their content,
79
and
distinct procedures for dierent categories of amendments.
80
Within the
72. See infra notes 265–272 and accompanying text; infra section IV.B.
73. Walter Dellinger, The Amending Process in Canada and the United States: A
Comparative Perspective, Law & Contemp. Probs., Autumn 1982, at 283, 298 [hereinafter
Dellinger, A Comparative Perspective]. Canada’s amendment rules are laid out in
Procedure for Amending Constitution of Canada, Part V of the Constitution Act, 1982,
being Schedule B to the Canada Act, 1982, c 11 (U.K.).
74. Dellinger, A Comparative Perspective, supra note 73, at 299. This time limit applies
to one of Canada’s five amendment procedures. See Richard Albert, The Structure of
Constitutional Amendment Rules, 49 Wake Forest L. Rev. 913, 944–45 (2014) [hereinafter
Albert, Structure of Amendment Rules].
75. Dellinger, A Comparative Perspective, supra note 73, at 299.
76. Id. at 299–300.
77. Albert, Structure of Amendment Rules, supra note 74, at 921, 945.
78. See id. at 952 (explaining that “temporal limitations” on the amendment process
“are commonly entrenched in written constitutions”).
79. See Yaniv Roznai, Unconstitutional Constitutional Amendments—The Migration
and Success of a Constitutional Idea, 61 Am. J. Compar. L. 657, 660 (2013) (discussing “the
global trend . . . towards accepting the idea of limitations—explicit or implicit—on consti-
tutional amendment power”).
80. See Markus Böckenförde, Int’l Inst. for Democracy & Electoral Assistance,
Constitutional Amendment Procedures 13 (2d ed. 2017), https://www.idea.int/sites/
default/files/publications/constitutional-amendment-procedures-primer.pdf
[https://perma.cc/5AT7-JCFS] (“In an attempt to identify the right balance between rigid-
ity and flexibility in constitutional amendments, many constitutions oer dierent thresh-
olds for dierent parts of the constitution.”). There is a burgeoning literature comparing
the design of constitutional amendment rules. See, e.g., Albert, Structure of Amendment
Rules, supra note 74; Joel Colón-Ríos, Introduction: The Forms and Limits of Constitutional
Amendments, 13 Int’l J. Const. L. 567 (2015); Rosalind Dixon, Constitutional Amendment
2021] PUZZLES AND POSSIBILITIES 2337
United States, most state constitutions likewise have amending clauses that
are significantly more detailed and precise than Article V.
81
Other
democracies’ amending clauses, meanwhile, seem less perplexing than
Article V not because their procedures are more detailed and precise but
rather because they are simpler and more streamlined—employing what
Richard Albert calls a “comprehensive single-track framework.”
82
This is far from a complete survey, of course, and textual comparisons
across constitutions of dierent ages and lengths are vexed. Our point is
merely that the revision rules in many other constitutions seem on their
face to raise fewer interpretive puzzles than does Article V. Even the most
carefully crafted amending clause cannot foreclose future uncertainty, as
Part I explains, and disputes about the meaning of such clauses are bound
to arise sooner or later in Canada and elsewhere.
83
But we think it is fair to
say that, both globally and domestically, the U.S. Constitution stands out
in just how many questions about the amendment process it leaves to
“speculation.”
84
The fact that Article V is so underdeterminate would not necessarily
matter much if it were addressed to some obscure issue of governance. But
the topic of Article V could hardly be more important. Unlike other vague
provisions in the Constitution, Article V sets out the rules of formal consti-
tutional change and thereby constitutes the constitutional text itself. As
Bruce Ackerman explains, Article V can be seen as “the most fundamental
text of our Constitution, since it seeks to tell us the conditions under which
all constitutional texts and principles may be legitimately transformed.”
85
In Gordon Wood’s words, Americans “institutionalized and legitimized
revolution” by enabling sweeping social change through a preestablished
Rules: A Comparative Perspective, in Comparative Constitutional Law 96 (Tom Ginsburg &
Rosalind Dixon eds., 2011).
81. For a thorough inventory of every U.S. state’s constitutional amendment rules,
see Amending State Constitutions, Ballotpedia, https://ballotpedia.org/
Amending_state_constitutions [https://perma.cc/5TJ4-QRRS] (last visited Aug. 5, 2021).
The amending clause in our home state of New York, for example, is over five times as long
as Article V. N.Y. Const. art. XIX (807 words).
82. See Albert, Structure of Amendment Rules, supra note 74, at 937–39 (categorizing
ten of the thirty-six democracies in his study as adopting such a framework and noting that
it “has the virtue of clarity”).
83. Cf. Otto Pfersmann, Comparative Hermeneutics of Constitutional Revision Clauses
and the Question of Structural Closure of Legal Systems, 40 Cardozo L. Rev. 3191, 3194–96
(2019) (discussing “interpretative problems” common to all constitutional revision clauses).
84. Dellinger, A Comparative Perspective, supra note 73, at 298.
85. Bruce A. Ackerman, The Storrs Lectures: Discovering the Constitution, 93 Yale L.J.
1013, 1058 (1984) [hereinafter Ackerman, Discovering the Constitution].
2338 COLUMBIA LAW REVIEW [Vol. 121:2317
legal process.
86
Constitutions are often described as focal points that coor-
dinate expectations
87
and “channel disputes.”
88
The underdeterminacy of
Article V threatens these coordination and channeling functions by elicit-
ing reasonable disagreement not just about the Constitution’s meaning
but also about its very terms.
During the Constitutional Convention of 1787, James Madison made
a similar observation. He urged that the amendment process be made
clearer because “diculties might arise as to the form, the quorum &c.
which in Constitutional regulations ought to be as much as possible
avoided.”
89
This is not just an American instinct. The European
Commission for Democracy Through Law (known as the Venice
Commission), which advises the Council of Europe on constitutional mat-
ters, concluded in its 2009 report on constitutional amendment proce-
dures that “[r]ules and procedures on constitutional amendment should
be as clear and simple as possible, so as not to give rise to problems and
disputes of their own.”
90
It is an interesting question whether and under what circumstances
this conclusion should be qualified on account of the potential benefits of
uncertainty. Ackerman, for instance, has suggested that wise constitutional
drafters will follow the U.S. Framers in recognizing “the limited extent to
which they [can] legitimately specify the higher lawmaking procedures to
be followed by succeeding generations.”
91
As Part V discusses, some play
in the joints may help keep the amendment process from thwarting
change that is overwhelmingly desired by the people. Regardless, the point
remains that significant amounts of legal uncertainty concerning a pur-
portedly exclusive constitutional amendment clause raise very dierent—
and potentially more destabilizing—issues than such uncertainty
surrounding, say, a rights guarantee.
92
86. Gordon S. Wood, The Creation of the American Republic, 1776–1787, at 614
(1969). Note that Wood’s characterization of Article V implicitly rejects the idea that amend-
ments are limited to an incremental, “perfecting” role. See supra note 52 and
accompanying text.
87. See Pozen & Samaha, supra note 69, at 793 n.313 (collecting sources).
88. Robert Post & Reva Siegel, Roe Rage: Democratic Constitutionalism and Backlash,
42 Harv. C.R.-C.L. L. Rev. 373, 427 (2007).
89. 2 The Records of the Federal Convention of 1787, at 630 (Max Farrand ed., 1937);
see also supra note 19 (quoting Charles Black to similar eect). Madison was talking specif-
ically about the convention procedure, but his general point—that the amendment rules
ought to be as clear as possible—applies equally to the whole of Article V.
90. Eur. Comm’n for Democracy Through Law, Final Draft Report on Constitutional
Amendment Procedures 46 (2009), https://www.venice.coe.int/webforms/documents/
default.aspx?pdle=CDL(2009)168-e [https://perma.cc/9PUH-C8SZ].
91. Ackerman, Discovering the Constitution, supra note 85, at 1058.
92. Cf. Cong. Pay Amendment, 16 Op. O.L.C. 85, 95 (1992) (“The very functioning of
the government would be clouded if Article V, which governs the fundamental process of
constitutional change, consisted of ‘open-ended’ principles without fixed applications.”).
2021] PUZZLES AND POSSIBILITIES 2339
III. “INTERNAL” LIMITS TO ARTICLE V’S RESOLVING POWER: PRECEDENT
A skeptic may respond at this juncture: Even if one can conjure up an
array of interpretive puzzles when squinting at the text of Article V in iso-
lation, surely most of the important issues have been cleared up by a quar-
ter millennium of historical practice and judicial doctrine. The law of
Article V, in other words, might be considerably clearer than the language
of Article V. We turn in this Part to that body of law and demonstrate that
the skeptic’s response misses the mark. Ever since the Founding, amend-
ments of uncertain legal validity have been the norm in the United States,
not the exception. According to the conventions of mainstream constitu-
tional reasoning, the overwhelming majority of recognized amendments
had significant arguable legal infirmities at the time of their adoption.
93
And still to this day, the “gloss” of judicial and nonjudicial precedent has
not cleared up many of the uncertainties regarding the operation of
Article V that the previous Part discusses.
The best way to establish these points, we believe, is to review the his-
tory of debates over the validity of amendments that are widely seen to
have become part of the constitutional text. If anything, focusing on suc-
cessful eorts to amend the Constitution ought to bias our results toward
suggesting greater clarity and settlement about the Article V process than
really exists. But even the successful amendments, it turns out, have left a
legacy of legal contestation and confusion.
A. The Bill of Rights
The story of legally dubious amendments begins at the beginning,
with the Bill of Rights.
94
The most serious concern with the procedure used
for these amendments is that the First Congress did not present them to
President Washington for his consideration before they went to the states
93. A qualifier such as “arguable” is unavoidable here because one cannot adjudicate
all of these issues without a full-blown theory of constitutional interpretation, about which
people will disagree. Cf. infra note 323 (discussing limitations of this paper’s approach).
Past Article V infirmities do not necessarily mean that an amendment is presently infirm,
insofar as longstanding acceptance can cure or render irrelevant earlier legal concerns. See
Greenawalt, Rule of Recognition, supra note 11, at 640–42; see also supra note 20; infra note
424.
94. The ten amendments that comprise the Bill of Rights were not the first ten amend-
ments sent to the states. Congress’s first proposed amendment would have changed the
apportionment formula for the House, but it fell just short of ratification. See Akhil Reed
Amar, The Bill of Rights as a Constitution, 100 Yale L.J. 1131, 1137, 1143 n.52 (1991)
[hereinafter Amar, The Bill of Rights]. Congress’s second proposed amendment prohibited
a congressional salary increase without an intervening election. Id. at 1145. That amend-
ment also fell short in the eighteenth century and was left for dead—until revived in the
twentieth century as the Twenty-Seventh Amendment. Recent litigation has similarly sought
to revive the “first” first amendment. We come back to both of these developments later.
See infra sections III.K–L. We mention them here to highlight that legal controversy over
amendment validity dates back to, and indeed continues to involve, the very first amend-
ments proposed by Congress.
2340 COLUMBIA LAW REVIEW [Vol. 121:2317
for ratification, notwithstanding Article I, Section 7’s language on presi-
dential approval.
95
“The Annals reveal no discussion of this important
question” in either the House or Senate.
96
The next section considers this
issue in more detail in conjunction with the Eleventh Amendment, which
prompted Hollingsworth v. Virginia.
97
For now, it suces to say that there is
a serious argument that presidential presentment is mandated by the plain
terms of the Constitution. The most significant counterargument, even by
the time of Hollingsworth, is based on historical practice—but of course that
argument was not available the very first time Article V was enlisted.
A second concern involves the role of new states. Article V says noth-
ing about what to do with states that join the Union after an amendment
has been submitted to the states but before the amendment has been rat-
ified. Nor is the answer to this question obvious as a matter of policy or
political morality. Excluding new states from ratification might unduly
privilege their predecessors; including them might disrupt and delay an
ongoing process. When the Bill of Rights was proposed by Congress, the
Union contained eleven states. Nine ratifications were therefore required
to reach three-fourths.
98
By 1791, three more states had joined the Union,
increasing the necessary number of ratifications to eleven.
99
Thomas
Jeerson declared the Bill of Rights part of the Constitution in 1792,
counting the new states in both the numerator and denominator.
100
We
are not aware of any discussion as to why this position was taken. But if one
only counts states that are part of the Union at the moment an amendment
is proposed, constitutional history would look “dramatically dierent”:
The first amendment proposed by Congress, on congressional apportion-
ment, would be ratified, and the Bill of Rights would not have made it into
the document until the twentieth century.
101
A final legal objection to the Bill of Rights is that there were discrep-
ancies in the instruments of ratification sent by the states to the federal
95. See supra note 21 and accompanying text.
96. Currie, The Federalist Period, supra note 57, at 115. Delaware Representative John
Vining raised a separate objection: that, under the language of Article V, both Houses of
Congress had to concur by a two-thirds vote that a proposed amendment was “necessary”
before proceeding to consider it. 1 Annals of Cong. 430 (1789). This objection did not carry
the day and apparently was not pursued further. See Edward S. Corwin & Mary Louise
Ramsey, Constitutional Law of Constitutional Amendment, 26 Notre Dame L. Rev. 185, 191
(1951).
97. 3 U.S. (3 Dall.) 378 (1798).
98. See Cong. Pay Amendment, 16 Op. O.L.C. 85, 107 (1992).
99. Id. But cf. supra note 57 (noting controversy in the First Congress over how to
construe the “three fourths” requirement for state ratification).
100. See Jol A. Silversmith, The “Missing Thirteenth Amendment”: Constitutional
Nonsense and Titles of Nobility, 8 S. Cal. Interdisc. L.J. 577, 597 (1999).
101. See id. at 598–99. This assumes, perhaps implausibly, that such a change in count-
ing method would not otherwise have changed the course of amendment history. It also
elides the question whether Massachusetts and Virginia should have had to re-ratify because
the new states of Maine and West Virginia were carved out of their respective borders.
2021] PUZZLES AND POSSIBILITIES 2341
government. Take the Second Amendment as an example. In its 2008
opinion in District of Columbia v. Heller, the Supreme Court quoted the text
as follows: “A well regulated Militia, being necessary to the security of a
free State, the right of the people to keep and bear Arms, shall not be
infringed.”
102
This is the version of the Second Amendment that was pro-
posed by Congress and ratified by Delaware (though “arms” was uncapi-
talized).
103
But according to their ratification resolutions, New Jersey
ratified a Second Amendment with no commas;
104
New York, Pennsylvania,
Rhode Island, and South Carolina ratified a version with only the middle
comma;
105
and Maryland and North Carolina ratified a version with the
middle and last commas.
106
Thomas Merrill, in an unpublished paper, has
similarly documented an evolution in how the Fifth Amendment’s Takings
Clause is punctuated.
107
This may be more than constitutional flyspecking. After all, “punctu-
ation is a permissible indicator of meaning.”
108
Judge Laurence Silberman
102. 554 U.S. 570, 576 (2008).
103. Resolution of Congress Proposing Amendatory Articles to the Several States (Mar.
4, 1789), in 2 Documentary History of the Constitution of the United States of America,
1786–1870, at 321, 322 (1894) [hereinafter Documentary History]; Delaware Ratification
Resolution (Jan. 28, 1790), in Documentary History, supra, at 347, 349. This documentary
compilation was derived from ocial records held at the Department of State. We have not
independently confirmed that the compilation in fact reflects the original documents on
file.
104. New Jersey Ratification Resolution (Aug. 3, 1790), in Documentary History, supra
note 103, at 325, 327.
105. Rhode Island Ratification Resolution (June 15, 1790), in Documentary History,
supra note 103, at 363, 364; New York Ratification Resolution (Mar. 27, 1790), in
Documentary History, supra note 103, at 357, 359; Pennsylvania Ratification Resolution
(Mar. 10, 1790), in Documentary History, supra note 103, at 352, 354; South Carolina
Ratification Resolution (Jan. 19, 1790), in Documentary History, supra note 103, at 340, 342.
106. North Carolina Ratification Resolution (Dec. 22, 1789), in Documentary History,
supra note 103, at 335, 337; Maryland Ratification Resolution (Dec. 19, 1789), in
Documentary History, supra note 103, at 330, 332. There were also dierences in capitaliza-
tion in the various states. For an overview of all these dierences, see Ross E. Davies, Which
Is the Constitution?, 11 Green Bag 2d 209, 210–11 (2008).
107. See Thomas W. Merrill, The Constitution and the Comma 15–18 (on file with the
Columbia Law Review) (unpublished manuscript).
108. Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts
161 (2012). It is debatable what the Founding generation thought about the significance of
punctuation. On the one hand, Chief Justice John Marshall wrote that “the construction of
a sentence in a legislative act does not depend on its pointing [i.e., punctuation].” Black v.
Scott, 3 F. Cas. 507, 510 (C.C.D. Va. 1828); see also Caleb Nelson, Preemption, 86 Va. L.
Rev. 225, 258 n.102 (2000) (discussing the “relatively casual attitude toward punctuation”
taken by British and American legal authorities at the time of the Founding). On the other
hand, during the Constitutional Convention, the Committee on Style apparently tried to
eect a major change in congressional power through punctuation (the insertion of a sem-
icolon) until it was caught in the act. See Philip Hamburger, The New Censorship:
Institutional Review Boards, 2004 Sup. Ct. Rev. 271, 317 n.112 (describing an attempt to
create a separate spending power by adding a semicolon to the first paragraph of Article I,
Section 8).
2342 COLUMBIA LAW REVIEW [Vol. 121:2317
began the merits portion of his D.C. Circuit opinion armed in Heller with
a reference to the “provision’s second comma.”
109
Justice Antonin Scalia’s
opinion for the Court in Heller divided the amendment into two clauses
and cited a source that seemed to place weight on the same comma.
110
Merrill has argued that the punctuation of the Takings Clause could very
well aect its modern meaning.
111
If Congress proposes and some of the
states (but not three-fourths) ratify a text that most objective observers at
the time would understand to mean X, while other states ratify a slightly
dierent text that most would understand to mean Y, it is not at all clear
that either text should be seen as part of the Constitution.
112
And if punc-
tuation can change the meaning of a legal text, it does not seem far-
fetched to insist that states ratify an identical text.
113
B. The Eleventh Amendment
In the 1793 case Chisholm v. Georgia,
114
the Supreme Court held that
Georgia could be sued without its consent by a citizen of another state.
The day after the decision was announced, a constitutional amendment to
109. Parker v. District of Columbia, 478 F.3d 370, 378 (D.C. Cir. 2007).
110. Justice Scalia wrote that the “Second Amendment is naturally divided into two
parts: its prefatory clause and its operative clause.” District of Columbia v. Heller, 554 U.S.
570, 577 (2008). For support, he cited a “Linguists’ Brief” that had suggested the second
comma was significant, as it “marks the customary separation of an adverbial clause . . . from
a main clause.” Brief for Professors of Linguistics and English Dennis E. Baron et al. in
Support of Petitioners at 5–6, Heller, 554 U.S. 570 (2008) (No. 07-290).
111. Merrill, supra note 107, at 8–11.
112. William W. Van Alstyne, A Constitutional Conundrum of Second Amendment
Commas: A Short Epistolary Report, 10 Green Bag 2d 469, 475–76 (2007). This scenario is
distinguishable from, though related to, one in which states ratify an identical text with
divergent understandings of that text’s communicative content. Cf. Cong. Pay Amendment,
16 Op. O.L.C. 85, 95 n.13 (1992) (reserving the question whether ratification would be valid
if the public meaning of a proposed amendment changed while it was pending); David
Pozen, E.R.A. Puzzles, Balkinization (Feb. 27, 2020), https://balkin.blogspot.com/2020/
02/era-puzzles_27.html [https://perma.cc/XG6Z-7QHH] (discussing this issue with
respect to the ERA).
113. Another significant question that arose in Congress as it debated the Bill of Rights
concerned the placement of amendments. Specifically, should amendments be woven into
the text of the original Constitution, or should they appear as supplements at the end of an
unchanged original text? James Madison favored the former approach; Roger Sherman was
the leading proponent of the latter. See Jonathan Gienapp, The Second Creation: Fixing
the American Constitution in the Founding Era 177–90 (2018). Sherman prevailed, of
course. Id. at 189. According to Jonathan Gienapp, this decision “numbers among the most
important milestones in the entire sweep of American constitutional history” by fostering
the perception of “the original Constitution as a ‘sacred’ text,” fixed for all time, rather than
an “organic, evolving” project. Id.; see also Richard Albert, Constitutional Amendments:
Making, Breaking, and Changing Constitutions 231–33 (2019) [hereinafter Albert,
Constitutional Amendments] (reviewing the Madison–Sherman debate).
114. 2 U.S. (2 Dall.) 419 (1793).
2021] PUZZLES AND POSSIBILITIES 2343
overrule Chisholm was introduced in the House.
115
In its final form, the
amendment passed the House and Senate by large majorities in 1794, was
ratified by the requisite number of states in 1795, and was declared part of
the Constitution by President John Adams in 1798.
116
As with the Bill of Rights, however, President Washington had never
approved the amendment before it was sent to the states. This fact formed
the basis for a challenge to the amendment’s validity in Hollingsworth v.
Virginia, in which plaintis argued that the “amendment ha[d] not been
proposed in the form prescribed by the Constitution” because it “was
never submitted to the President for his approbation.”
117
The Supreme
Court tersely rejected this challenge. The report of the decision contains
only the following explanation: “The Court, on the day succeeding the
argument, delivered an unanimous opinion, that the amendment being
constitutionally adopted, there could not be exercised any jurisdic-
tion . . . .”
118
During the defendant’s oral argument, though, Justice
Samuel Chase had made the following statement: “The negative of the
President applies only to the ordinary cases of legislation: He has nothing
to do with the proposition, or adoption, of amendments to the
Constitution.”
119
Justice Chase’s statement is dubious on its own terms. According to
Article I, Section 7 of the Constitution, the veto power extends not only to
“ordinary cases of legislation” but also to “Every Order, Resolution, or
Vote to which the Concurrence of the Senate and House of
Representatives may be necessary (except on a question of
Adjournment).”
120
A resolution or vote proposing a constitutional amend-
ment seems to fall within the latter category. Perhaps Justice Chase meant
to suggest that Article V sets forth “its own separate higher-lawmaking
track” not subject to the presentment rules of Article I.
121
But Congress
and the Court have arguably taken the opposite position with respect to
115. See David P. Currie, The Constitution in Congress: The Third Congress, 1793–
1795, 63 U. Chi. L. Rev. 1, 35–38 (1996).
116. See John V. Orth, The Judicial Power of the United States: The Eleventh
Amendment in American History 12–29 (1987); see also 1 Charles Warren, The Supreme
Court in United States History 101 n.2 (1922) (discussing “the extremely informal and care-
less manner in which the ratification was promulgated” by President Adams).
117. 3 U.S. (3 Dall.) 378, 379 (1798).
118. Id. at 382.
119. Id. at 381 n.*. On why Justice Chase’s spoken statement is “clearly not” part of the
Court’s opinion, see Seth Barrett Tillman, A Textualist Defense of Article I, Section 7,
Clause 3: Why Hollingsworth v. Virginia Was Rightly Decided, and Why INS v. Chadha Was
Wrongly Reasoned, 83 Tex. L. Rev. 1265, 1285 (2005) [hereinafter Tillman, A Textualist
Defense].
120. U.S. Const. art. I, § 7, cl. 3.
121. Amar, America’s Constitution, supra note 49, at 594 n.7 (noting this theory without
endorsing it).
2344 COLUMBIA LAW REVIEW [Vol. 121:2317
the quorum rules of Article I,
122
and to call Article V a separate track risks
begging the question whether the Presentment Clause applies by its plain
terms to the amendment process.
Another possible view is that presidential presentment is superfluous
because Article V already requires a two-thirds vote, which is enough to
override a veto.
123
This view is not persuasive. When Presidents veto a bill
or resolution, they must state their “Objections,” and the bill or resolution
must be “reconsidered” by both chambers of Congress in light of those
objections.
124
As the Hollingsworth plaintis’ counsel pointed out, this
reconsideration requirement presumes a dialogic model of governance in
which the President’s objections might influence some members of
Congress.
125
Even a “Bill” that initially passes by a two-thirds vote in both
chambers still needs presidential approval or repassage after a veto before
it becomes law.
126
In addition, if the initial vote in either chamber was
taken with a quorum but less than full membership, a presidential veto
could cause the number of members present and voting to be dierent
upon reconsideration.
127
In short, the Court’s ruling in Hollingsworth is questionable as a matter
of constitutional text and structure. Charles Black has described it as “an
unreasoned decision, uttered in the teeth of plain constitutional language,
and with no really adequate reason even projectable.”
128
In 2000, the
Supreme Court of Wyoming invalidated a proposed amendment to its
state constitution on the ground that it had not been presented to the
governor, under a provision that is identical in relevant part to the federal
presentment requirement.
129
If Black and the Wyoming high court are cor-
rect, then Hollingsworth deserves little if any deference on the merits—and
all thoroughgoing textualists must face the prospect that twenty-six of our
twenty-seven ostensible amendments are void.
130
122. See Nat’l Prohibition Cases, 253 U.S. 350, 386 (1920); David P. Currie, The
Constitution in Congress: The Jeersonians, 1801–1829, at 61–63 (2001) [hereinafter
Currie, The Jeersonians].
123. See Amar, America’s Constitution, supra note 49, at 594 n.7 (describing this as the
other “main” theory that has been oered in support of Justice Chase’s contention).
124. U.S. Const. art. I, § 7, cl. 2.
125. Hollingsworth v. Virginia, 3 U.S. (3 Dall.) 378, 379 (1798) (argument by counsel).
126. See Prakash, Of Synchronicity, supra note 16, at 1254 n.180.
127. See Tillman, A Textualist Defense, supra note 119, at 1290 n.64.
128. Black, On Article I, supra note 21, at 898; see also Primus, Unbundling
Constitutionality, supra note 5, at 1110 n.78 (“Justice Chase’s dictum . . . makes not the
slightest gesture toward explaining how the text could be read to accord with [it].”).
129. Geringer v. Bebout, 10 P.3d 514, 523 (Wyo. 2000).
130. See supra note 21 (noting that only the Thirteenth Amendment received a presi-
dential signature). Some have suggested that these twenty-six amendments might be sal-
vaged on the ground that Congress provided copies to the executive branch, and the
subsequent transmittal of the amendments to the states for ratification constituted a kind of
constructive approval by the President. See Tillman, A Textualist Defense, supra note 119,
at 1277–83; Sopan Joshi, Note, The Presidential Role in the Constitutional Amendment
2021] PUZZLES AND POSSIBILITIES 2345
Process, 107 Nw. U. L. Rev. 963, 992–97 (2013). This suggestion contravenes the contempo-
raneous understanding of virtually every relevant ocial involved in the amendment pro-
cess. In Hollingsworth, the Virginia Attorney General (as well as Justice Chase) argued that
presentment was unnecessary, not that it had occurred. 3 U.S. (3 Dall.) at 381 & n.*. A few
years later, the Senate debated and specifically rejected a motion to “present” the proposed
Twelfth Amendment to “the President . . . for his approbation.” William Plumer’s
Memorandum of Proceedings in the United States Senate 1803–1807, at 79–80 (Everett
Somerville Brown ed., 1923). After President Lincoln was presented with and approved the
Thirteenth Amendment, the Senate passed a resolution “declar[ing] that such approval was
unnecessary” and “inconsistent with the former practice in reference to all amendments
heretofore adopted, and being inadvertently done, should not constitute a precedent for
the future.” Cong. Globe, 38th Cong., 2d Sess. 629–30 (1865). President Andrew Johnson,
who opposed the Fourteenth Amendment, noted that the amendment had not been
“submitted by the two Houses for [his] approval” and observed in a letter to Congress that
“the steps taken by the Secretary of State” to transmit the proposed amendment to the states
“are to be considered as purely ministerial, and in no sense whatever committing the
Executive to an approval.” Cong. Globe, 39th Cong., 1st Sess. 3349 (1866). This
understanding has persisted in Congress and the Supreme Court. See, e.g., INS v. Chadha,
462 U.S. 919, 955 n.21 (1983). Under these circumstances, we do not think it is persuasive
to say that a requirement of presentment and approval, if one exists, was constructively
satisfied.
For similar reasons, we do not think it is persuasive to say that the amendments took
eect on account of ten days of presidential inaction. See U.S. Const. art. I, § 7, cl. 2 (“If
any Bill shall not be returned by the President within ten Days (Sundays excepted) after it
shall have been presented to him, the Same shall be a Law, in like Manner as if he had
signed it, unless the Congress by their Adjournment prevent its Return . . . .”). First, it is not
clear, as a textual matter, that the approval-through-inaction mechanism carries over from
Clause 2 to Clause 3 of Article I, Section 7. See Tillman, A Textualist Defense, supra note
119, at 1320 (noting this ambiguity). Second, it seems a stretch to attach any significance to
ten days of inaction when neither Congress nor the President contends that the President
has a veto power. Finally, even if one were inclined to invoke this provision, thirteen amend-
ments (the Bill of Rights and the Fifteenth, Eighteenth, and Twenty-Seventh Amendments)
would have been pocket vetoed because they were passed fewer than ten days prior to the
end of a congressional session and therefore would have been ineective absent approval.
See Joshi, supra, at 994 tbl.1.
Seth Tillman defends the outcome of Hollingsworth based on a creative reading of
Article I, Section 7, Clause 3 (which he calls the “ORV Clause”). He takes the position “that
‘[e]very order, resolution or vote to which the concurrence of the Senate and House [is]
necessary’ refers to single-house action taken pursuant to prior authorizing or later ratifying
legislation.” Tillman, A Textualist Defense, supra note 119, at 1364. In other words,
Congress can “delegate lawmaking authority” to a single house, or possibly even a single
committee, and the ORV Clause ensures that such legislation must still be presented to the
President. Id. at 1334 n.144. As Tillman acknowledges, his proposal conflicts with the near-
unanimous understanding of the ORV Clause from the Founding to the present, including
the views of James Madison, Joseph Story, the Federalist Papers, and the Court in Chadha.
Id. at 1364–65. In any event, even if Tillman’s reading were otherwise correct, it is not clear
why the ORV Clause would not also apply by its plain terms to constitutional amendments.
See Gary Lawson, Burning Down the House (and Senate): A Presentment Requirement for
Legislative Subpoenas Under the Orders, Resolutions, and Votes Clause, 83 Tex. L. Rev.
1373, 1386 (2005) (“Does the ORV Clause also apply to amendment resolutions under
Article V? The answer seems to be yes.”).
2346 COLUMBIA LAW REVIEW [Vol. 121:2317
C. The Twelfth Amendment
Congress’s eorts to pass the Twelfth Amendment, which revamped
the nation’s system for selecting presidents, surfaced a number of novel
legal questions.
131
One group of senators argued that Article V could not
be used to “ingraft new principles into the Constitution which will destroy
the rights of individual States, without the consent of those States.”
132
The
Senate debated whether a two-thirds majority was required for all votes
respecting constitutional amendments, or only for the vote on the final
proposed amendment.
133
Most contentiously, the Senate debated whether
Article V requires two-thirds of the full House and Senate to propose an
amendment, or only two-thirds of a quorum. The issue arose because the
Senate had a total membership of thirty-four, and the amendment passed
the Senate by a vote of 22 to 10, with the votes of two senators
unrecorded.
134
The amendment thus crossed the two-thirds threshold
among those voting, but not among the whole body, and an objection was
raised on that ground.
135
The objection has real force. To revise the Constitution is an extraor-
dinary step, and Article V’s supermajority voting rules demand “a geo-
graphically broad and numerically deep consensus” to make it happen.
136
Yet, as Senator William Plumer of New Hampshire explained, if two-thirds
of a quorum suced to propose an amendment, “it would follow that
twelve senators . . . might propose an amendment contrary to the opinion
& against the will of twenty two senators,” and a “little more than one full
third of the Senate” could “be considered as constitutionally performing
the act that required the concurrence of two thirds.”
137
In contrast to
Article V, moreover, the Constitution’s impeachment and treaty clauses
expressly refer to “two thirds of the [senators] present.”
138
Of course,
Article I provides that “a Majority of each [House] shall constitute a
Quorum to do Business,”
139
and a quorum was present for the vote on the
131. As explained in the preceding footnote, the question of presidential presentment
also resurfaced in the Senate debate. Interestingly, it seems that no senators invoked the
Court’s recent decision in Hollingsworth, much less suggested that it had settled the question.
One senator wrote in his diary that the relevant “precedents [were] established without
debate, or without particular attention to the subject—& therefore they prove nothing.”
Currie, The Jeersonians, supra note 122, at 58. The Senate nonetheless resolved by a vote
of 23 to 7 that the proposed amendment need not be submitted to President Jeerson for
his approval. Id.
132. 13 Annals of Cong. 738 (1803).
133. See Currie, The Jeersonians, supra note 122, at 59–60.
134. Id.
135. Id. at 61.
136. Lawrence G. Sager, The Incorrigible Constitution, 65 N.Y.U. L. Rev. 893, 960
(1990).
137. Currie, The Jeersonians, supra note 122, at 61–62 (emphasis omitted) (quoting
a speech by Senator Plumer that was recorded in his diary but not in the Annals).
138. U.S. Const. art. I, § 3, cl. 6 (emphasis added); id. art. II, § 2, cl. 2 (emphasis added).
139. Id. art. I, § 5, cl. 1.
2021] PUZZLES AND POSSIBILITIES 2347
Twelfth Amendment. But the validity of the first eleven amendments had
apparently been premised on the proposition that the amendment process
was special and, therefore, not subject to another provision of Article I—
the requirement of presidential presentment.
Ultimately, the objection was overruled in the Senate and the pro-
posed amendment went to the House, where Federalist representatives
“essentially reiterated what Plumer had said” in arguing that the Senate
had not legally approved it.
140
The House voted 85 to 34 to take up the
Senate resolution and then 84 to 42—exactly two-thirds of a quorum—to
approve the Twelfth Amendment.
141
Over a century later, the Supreme
Court agreed that two-thirds of a quorum was sucient, in light of
longstanding congressional practice.
142
But David Currie, on whose schol-
arship we have drawn throughout this section, likely spoke for many when
he opined in 2001: “I still have trouble convincing myself that when the
Framers prescribed a two-thirds majority to ensure broad support for
constitutional amendments they meant it could be provided by fewer than
half the members.”
143
D. The Reconstruction Amendments
In 1861, Congress approved in its lame-duck session an amendment,
known as the Corwin Amendment, which would have protected slavery in
the South by prohibiting any future amendment giving “Congress the
power to abolish or interfere, within any State, with the domestic institu-
tions thereof, including that of persons held to labor or service by the laws
of said State.”
144
The Corwin Amendment failed to attain ratification, but
140. Currie, The Jeersonians, supra note 122, at 62. Some House Republicans coun-
tered with the argument that the first ten amendments had been approved only by two-
thirds of those present. See id. at 63; see also Mo. Pac. Ry. Co. v. Kansas, 248 U.S. 276, 281–
82 (1919) (making the same argument). But the historical record is hazy on this point; “it
was entirely possible that [those amendments] had in fact been endorsed by two thirds of
all the members.” Currie, The Jeersonians, supra note 122, at 63.
141. Currie, The Jeersonians, supra note 122, at 63–64; David E. Kyvig, Explicit and
Authentic Acts: Amending the U.S. Constitution, 1776–2015, at 116 (rev. ed. 2016). At the
time, the House had 141 members. Currie, The Jeersonians, supra note 122, at 63 n.186.
142. Nat’l Prohibition Cases, 253 U.S. 350, 386 (1920); Mo. Pac. Ry. Co., 248 U.S. at 281.
There is an additional wrinkle: Does Article V require two-thirds of those present, or only
two-thirds of those present and voting? See infra section III.J.
143. Currie, The Jeersonians, supra note 122, at 64. One might counter that the
Framers operated against a background presumption that a quorum was sucient for
legislative business, and Article V should be read in that context. But the very existence of
debate in Congress shows that such a presumption was not universally shared by the
Founding generation, at least as applied to amendments. Congress’s resolution of the issue,
moreover, did not quell contemporaneous debate. After Congress proposed the Twelfth
Amendment, the legislatures of three states, “in their resolutions rejecting the amendment,
reiterated the charge of unconstitutionality.” Ames, supra note 55, at 295.
144. See generally A. Christopher Bryant, Stopping Time: The Pro-Slavery and
“Irrevocable” Thirteenth Amendment, 26 Harv. J.L. & Pub. Pol’y 501 (2003) (reviewing the
history and legacy of the Corwin Amendment).
2348 COLUMBIA LAW REVIEW [Vol. 121:2317
it raised two novel Article V questions that were not resolved then and have
not been resolved since: whether an amendment can make itself una-
mendable,
145
and whether a state can choose to ratify an amendment by
convention even if Congress provides for ratification by legislatures.
146
Another oddity of the Corwin Amendment is that, Hollingsworth notwith-
standing, it was presented to President James Buchanan, who promptly
added his signature.
147
The next set of amendments to make it out of Congress, aimed at
dismantling rather than entrenching slavery, generated still fiercer legal
controversy as the nation emerged from the Civil War. Uniquely among
the Article V disputes recounted in this Part, the disputes associated with
the passage of the Thirteenth, Fourteenth, and Fifteenth Amendments
have received prominent attention in the contemporary constitutional
literature.
148
The most substantial objections clustered around three
issues: the composition of Congress, the legitimacy of the state legislatures
that ratified the amendments, and federal coercion of the states.
149
After the Civil War, Article V was pushed to the breaking point by
some stark numerical realities. There were thirty-seven states in the Union
in 1868, meaning twenty-eight states were needed to ratify a constitutional
amendment.
150
Eleven states had seceded, or purported to secede, and
formed the Confederacy.
151
Ten of those states could block an amendment
by themselves, assuming no defections from the Northern states—which
was not a safe assumption. The former Confederate states would also have
145. See id. at 530–34 (describing the March 1861 congressional debates on this
question).
146. See Philip L. Martin, Convention Ratification of Federal Constitutional
Amendments, 82 Pol. Sci. Q. 61, 65 (1967) (noting that Illinois purported to ratify the
Corwin Amendment through a convention that had met to draft a new state constitution).
147. See supra note 21.
148. Throughout this section, we cite what we take to be the leading works in this genre.
149. Other legal issues surfaced as well. One was whether states may rescind their ratifi-
cations. Both New Jersey and Ohio ratified the Fourteenth Amendment and then tried to
reverse course, but Congress, when it promulgated the Fourteenth Amendment, listed both
of them among the ratifying states. See Coleman v. Miller, 307 U.S. 433, 448–49 (1939)
(opinion of Hughes, C.J.). There were also quorum questions at the state level in addition
to the federal level. Indiana had a general rule in its state constitution requiring two-thirds
of the legislature to be present to conduct business, but the speaker of its House of
Representatives ruled that a majority of the total membership suced to act on the
Fifteenth Amendment. See Lester Bernhardt Orfield, The Amending of the Federal
Constitution 66 n.87 (1942).
150. See Thomas B. Colby, Originalism and the Ratification of the Fourteenth
Amendment, 107 Nw. U. L. Rev. 1627, 1643–44 (2013) This count assumes that former
Confederate states should be included in the Article V “denominator.” But see Christopher
R. Green, Loyal Denominatorism and the Fourteenth Amendment: Normative Defenses
and Implications, 13 Duke J. Const. L. & Pub. Pol’y 167, 169–70 (2017) (arguing that rebel
states should be excluded from the Article V denominator).
151. See Colby, supra note 150, at 1644; Harrison, supra note 14, at 422.
2021] PUZZLES AND POSSIBILITIES 2349
enough clout in Congress to block any amendment proposal.
152
Navigating
these obstacles demanded a series of bold legal maneuvers.
First, the Congresses that proposed the Reconstruction Amendments
excluded many of the representatives and senators sent from states that
had been part of the Confederacy, pursuant to each chamber’s power to
“be the Judge of the Elections, Returns and Qualifications of its own
Members.”
153
The Reconstruction Amendments were adopted by two-
thirds of a quorum composed mostly of Northern Republicans, and it is
exceedingly doubtful that all of the amendments could have passed had
the Southern representatives and senators been seated.
154
Second, in proclaiming the Thirteenth Amendment ratified by three-
fourths of the states on December 18, 1865, Secretary of State William
Henry Seward included several former Confederate states in the count.
155
Two weeks prior, Congress had refused to seat any senators or representa-
tives from those same states.
156
How could state “legislatures” validly ratify
constitutional amendments but not elect senators? Moreover, in the First
Military Reconstruction Act of 1867, Congress declared that there were
“no legal State governments” in the South and that existing “civil govern-
ments” were “provisional only.”
157
President Andrew Johnson argued in
his veto message that, because the bill “denies the legality of the
Governments of ten of the States which participated in the ratification of
the amendment . . . abolishing slavery,” the implication is that “the con-
sent of three-fourths of the States . . . has not been constitutionally
obtained” for the Thirteenth Amendment.
158
The Southern state governments that ratified the Fourteenth and
Fifteenth Amendments, meanwhile, were the products of military recon-
struction. The Reconstruction Acts divided the former Confederate states
into five military districts and instructed the Union Army to register voters,
with universal adult male surage, and to hold elections for constitutional
152. See Ackerman, We the People, supra note 7, at 102.
153. U.S. Const. art. I, § 5, cl. 1; see Amar, America’s Constitution, supra note 49, at
366–76; David P. Currie, The Civil War Congress, 73 U. Chi. L. Rev. 1131, 1218–22 (2006);
Harrison, supra note 14, at 380–89, 398–401.
154. See, e.g., Ackerman, We the People, supra note 7, at 102 (“Every student of the
period recognizes that, were it not for the purge of Southern Senators and Representatives,
the ‘Congress’ meeting in June would never have mustered the two-thirds majorities
required to propose the Fourteenth Amendment.”). It is less clear that the Fifteenth
Amendment would have been rejected if Congress were complete, see Kyvig, supra note
141, at 180, though opponents of the amendment eort did object on this ground, see David
P. Currie, The Reconstruction Congress, 75 U. Chi. L. Rev. 383, 454–55 (2008) [hereinafter
Currie, The Reconstruction Congress].
155. Currie, The Reconstruction Congress, supra note 154, at 397.
156. Amar, America’s Constitution, supra note 49, at 366.
157. Act of Mar. 2, 1867, ch. 153, pmbl., § 6, 14 Stat. 428–29.
158. 12 The Papers of Andrew Johnson, February–August 1867, at 91 (Paul H. Bergeron
ed., 1995).
2350 COLUMBIA LAW REVIEW [Vol. 121:2317
conventions.
159
Those conventions yielded ten new governments that
promptly ratified the Fourteenth and Fifteenth Amendments.
160
Many
argued then and afterward that these reconstructed governments lacked
legal authority to ratify amendments.
161
In addition, the Southern state legislatures were arguably coerced
into ratifying. The Reconstruction Acts provided that these states would be
entitled to representation in Congress only when the Fourteenth
Amendment “ha[d] become part of the Constitution.”
162
This pressure to
ratify, according to Ackerman, amounted to a “naked violation[] of Article
V.”
163
There is little doubt that the unreconstructed Southern states would
not have ratified voluntarily, as it “would be dicult to overstate the depth
and breadth of opposition to the Fourteenth Amendment” within their
white populations.
164
Southern states still under military supervision
“faced the same pressure to ratify” the Fifteenth Amendment.
165
These interrelated legal problems have elicited powerful responses.
John Harrison, for instance, has argued that the amendments were “legally
eective” (even if not strictly speaking “legal”) under the de facto govern-
ment doctrine, which recognizes that “a government de facto may bind
the state for which it acts despite defects in its claim to power.”
166
Amar
has defended the legality of the Reconstruction Amendments on the basis
of Congress’s authority to judge its members’ qualifications and to guar-
antee a republican form of government in the states.
167
And, stepping
back, it is notable that the Reconstruction Congresses went to such lengths
even to try to adhere to the forms of Article V, given the dire circum-
stances.
168
But wherever one comes out in these debates, the legal legitimacy of
the Reconstruction Amendments at the time of their adoption is at least
contestable—as countless scholars have recognized.
169
As Harrison recounts,
159. See Ackerman, We the People, supra note 7, at 110.
160. See Harrison, supra note 14, at 404–09.
161. See Ackerman, We the People, supra note 7, at 110–11; Colby, supra note 150, at
1654–56.
162. Act of Mar. 2, 1867, ch. 153, pmbl., § 5, 14 Stat. 429.
163. Ackerman, We the People, supra note 7, at 111.
164. Colby, supra note 150, at 1644–45.
165. Kyvig, supra note 141, at 181; see also Travis Crum, The Lawfulness of the Fifteenth
Amendment, 97 Notre Dame L. Rev. (forthcoming 2022) (manuscript at 31--38) (on file
with the Columbia Law Review) (detailing additional Article V problems “unique” to the
Fifteenth Amendment).
166. Harrison, supra note 14, at 379.
167. See Amar, America’s Constitution, supra note 49, at 370–80.
168. See Michael Les Benedict, Constitutional History and Constitutional Theory:
Reflections on Ackerman, Reconstruction, and the Transformation of the American
Constitution, 108 Yale L.J. 2011, 2027–31 (1999).
169. See, e.g., Kyvig, supra note 141, at 156; Colby, supra note 150, at 1629; Greenawalt,
Rule of Recognition, supra note 11, at 640; Douglas Laycock, Notes on the Role of Judicial
Review, the Expansion of Federal Power, and the Structure of Constitutional Rights, 99 Yale
2021] PUZZLES AND POSSIBILITIES 2351
“[a]ll those who participated in reconstruction, including those who were
paying attention to the process of constitutional amendment, knew that
something very unusual and legally doubtful was going on.”
170
“The
Republicans . . . got away with something Article V probably was supposed
to prevent.”
171
And yet, no one in their right mind would deny that the
Reconstruction Amendments are part of the Constitution today (though
the Georgia General Assembly denied this as late as 1957
172
), which illus-
trates our thesis in an especially dramatic fashion. The sociological legiti-
macy of the Reconstruction Amendments is not a function of their original
legal legitimacy; it does not derive from a judgment about whether Article
V’s rules were followed.
173
Rather, it derives from the fact that these
amendments have been accepted by most ocials since the 1860s and
have become deeply embedded in the nation’s laws, practices, and
ethos.
174
Their authoritative legal status is quite literally beyond dispute in
our political culture, just like the status of the Constitution itself.
175
E. The Sixteenth Amendment
After a long period of disuse, the machinery of Article V creaked back
into motion in the early twentieth century, yielding a spurt of four amend-
ments in a decade. The first of these was the Sixteenth Amendment, which
overruled the Supreme Court’s decision in Pollock v. Farmers’ Loan & Trust
L.J. 1711, 1729 (1990) (book review); see also supra note 20 (explaining this paper’s use of
the term “legal legitimacy”).
170. Harrison, supra note 14, at 409.
171. Id. at 458.
172. S. Res. 39, 129th Gen. Assemb., 2d Reg. Sess. (Ga. 1957); see also Colby, supra note
150, at 1661 n.209.
173. See Greenawalt, Rule of Recognition, supra note 11, at 641 (“The present authority
of these amendments may depend more on their acceptance for over a century than on
their actual adoption by a process that may or may not now be thought to conform to what
article V prescribes.”).
174. In this respect, the addition of the Reconstruction Amendments to the
Constitution resembles the addition of Texas to the Union. Both acts were constitutionally
dubious at the time—there is a plausible constitutional argument that new territory must be
annexed through the treaty power, not through congressional resolution—but are now
politically and practically settled. See Mark A. Graber, Settling the West: The Annexation of
Texas, the Louisiana Purchase, and Bush v. Gore, in The Louisiana Purchase and American
Expansion, 1803–1898, at 83, 83–103 (Sanford Levinson & Bartholomew H. Sparrow eds.,
2005).
175. Cf. Henry Paul Monaghan, Our Perfect Constitution, 56 N.Y.U. L. Rev. 353, 383
(1981) (“The authoritative status of the written constitution is . . . an incontestable first prin-
ciple for theorizing about American constitutional law.”); David E. Pozen, Constitutional
Bad Faith, 129 Harv. L. Rev. 885, 941–43 (2016) (explaining that mainstream U.S. political
culture eectively prohibits constitutional “atheism,” or a professed lack of faith in and
commitment to the Constitution).
2352 COLUMBIA LAW REVIEW [Vol. 121:2317
Co. and expressly authorized a federal income tax.
176
The amendment
“sailed through both houses of Congress” in 1909.
177
After hitting a few
speedbumps in the states, including the opposition of New York Governor
and future Chief Justice of the United States Charles Evan Hughes, it was
deemed ratified four years later.
178
The most persistent, but by no means the only,
179
set of legal chal-
lenges to the Sixteenth Amendment involves discrepancies among the
state ratification instruments. When Secretary of State Philander Knox cer-
tified the amendment in 1913, just four states had sent him instruments of
ratification with the language of the amendment exactly as Congress had
approved it.
180
The rest had variations in capitalization and punctuation,
and some even had dierences in wording. The instrument from
Oklahoma, for instance, said “from any census or enumeration” instead of
“without regard to any census or enumeration,” Illinois’s said “remunera-
tion” instead of “enumeration,” and Missouri’s said “levy” instead of
“lay.”
181
The Oce of the Solicitor of the Department of State prepared a
memorandum for Secretary Knox addressing these discrepancies. The
Solicitor’s Oce concluded that they were “probably inadvertent” and
that the legislatures in question had “intended . . . to ratify the amend-
ment proposed by Congress.”
182
The Oce also noted that similar discrep-
ancies had existed with earlier amendments.
183
It therefore recommended
that Knox certify the amendment.
184
A spate of “tax protester” lawsuits in
the 1980s placed heavy emphasis on this memorandum in challenging the
validity of the Sixteenth Amendment. Courts uniformly rejected these
176. 158 U.S. 601 (1895); see U.S. Const. amend. XVI (“The Congress shall have power
to lay and collect taxes on incomes, from whatever source derived, without apportionment
among the several States, and without regard to any census or enumeration.”).
177. Amar, America’s Constitution, supra note 49, at 409.
178. See Kyvig, supra note 141, at 204–07.
179. See generally Danshera Cords, Tax Protestors and Penalties: Ensuring Perceived
Fairness and Mitigating Systemic Costs, 2005 BYU L. Rev. 1515, 1537–40 (cataloging com-
mon tax protestor arguments that the Sixteenth Amendment was not properly ratified);
Christopher S. Jackson, Note, The Inane Gospel of Tax Protest: Resist Rendering Unto
Caesar—Whatever His Demands, 32 Gonz. L. Rev. 291, 301–07 (1996–1997) (same). Some
argued, for instance, that the amendment was invalid because certain state legislatures
lacked authority under their state constitutions to ratify such a taxation measure, or because
the amendment transgressed federalism-based substantive limits on Article V. See, e.g.,
Raymond G. Brown, The Sixteenth Amendment to the United States Constitution, 54 Am.
L. Rev. 843, 847–53 (1920).
180. Memorandum from O. of the Solic., Dep’t of State, Ratification of the 16th
Amendment to the Constitution of the United States 6 (Feb. 15, 1913) (on file with the
Columbia Law Review).
181. Id. at 7–8 (emphasis added).
182. Id. at 15.
183. Id. at 8–15.
184. Id. at 16.
2021] PUZZLES AND POSSIBILITIES 2353
claims,
185
often imposing sanctions along the way.
186
By and large, these
courts relied on the political question doctrine and some version of the
enrolled bill rule, without reaching the merits of the plaintis’ Article V
arguments.
187
We agree that this litigation is frivolous. The interesting question is
why. The basic answer, in our view, is that the long acceptance and socio-
logical legitimacy of the Sixteenth Amendment make it unthinkable that
a court in the 1980s would or should entertain a challenge to the amend-
ment’s validity—not because the merits of the underlying Article V ques-
tions are obvious.
188
F. The Seventeenth Amendment
Throughout the 1890s, “Congress was deluged with petitions and
published appeals for direct popular election of senators.”
189
Many
senators were wary of reforming their own institution, and Southern
senators were especially wary of increased federal involvement in elec-
tions.
190
But the public pressure continued to mount. Thirty-one states
petitioned Congress for change (one short of the number required to trig-
ger a convention);
191
and in 1910, after the Senate voted down a popular
election amendment, ten senators who had opposed the measure lost
reelection.
192
The Senate approved the amendment in 1911. The text
approved by the Senate, however, was dierent from the text that the
House had already approved. A conference committee wrangled for nearly
185. See, e.g., United States v. Sitka, 845 F.2d 43, 47 (2d Cir. 1988); United States v.
Stahl, 792 F.2d 1438, 1441 (9th Cir. 1986); Sisk v. Comm’r, 791 F.2d 58, 61 (6th Cir. 1986);
United States v. Thomas, 788 F.2d 1250, 1253–54 (7th Cir. 1986); United States v. Foster,
789 F.2d 457, 463 (7th Cir. 1986); United States v. Wojtas, 611 F. Supp. 118, 121 (N.D. Ill.
1985).
186. See, e.g., Miller v. United States, 868 F.2d 236, 242 (7th Cir. 1989); Lysiak v.
Comm’r, 816 F.2d 311, 313 (7th Cir. 1987). This wave of tax protester litigation was
fomented by a popular book, Bill Benson & M.J. Beckman, The Law That Never Was: The
Fraud of the 16th Amendment and Personal Income Tax (1985). See, e.g., Wojitas, 611 F.
Supp. at 119 (describing how the plainti’s attorney submitted the book to the court in
support of the plainti’s motion). One of the book’s authors was subsequently found to
have engaged in fraud. See United States v. Benson, 561 F.3d 718, 724 (7th Cir. 2009).
187. The enrolled bill rule holds that when “‘the presiding ocers’ of the House and
Senate sign an enrolled bill (and the President ‘approve[s]’ it), ‘its authentication as a bill
that has passed Congress should be deemed complete and unimpeachable.’” NLRB v. Noel
Canning, 573 U.S. 513, 551 (2014) (quoting Marshall Field & Co. v. Clark, 143 U.S. 649, 672
(1892)). While the enrolled bill rule limits the types of challenges “the judicial department”
may entertain, Clark, 143 U.S. at 672, it does not bear on what is required by Article V.
188. See supra notes 102–112 and accompanying text (explaining why it is at least plau-
sible to believe that states must ratify an identical text).
189. Kyvig, supra note 141, at 209.
190. Id. at 209–12.
191. Not all of the states formally requested a convention; some only asked that
Congress propose an amendment. See Caplan, supra note 63, at 63–65.
192. See Kyvig, supra note 141, at 212.
2354 COLUMBIA LAW REVIEW [Vol. 121:2317
a year, until the House finally caved and voted to send the Senate version
of the amendment to the states.
193
This intercameral wrangling led to a thorny Article V question. The
Seventeenth Amendment is one of only two recognized amendments to
have been passed by the House and the Senate in dierent legislative ses-
sions. Is that permissible? In a recent article, Saikrishna Prakash argues
that, as a matter of constitutional text and structure, it is not, as Article V
implicitly requires that amendments be proposed by both chambers of
Congress in the same legislative session.
194
“This reading of the
Constitution,” Prakash notes, “would suggest the rather immodest conclu-
sion that the Thirteenth and Seventeenth Amendments, neither of which
passed in a single session, failed to satisfy the standards of Article V.”
195
G. The Eighteenth Amendment
The Eighteenth Amendment, on Prohibition, raised many more legal
questions and led to some of the Supreme Court’s most significant pro-
nouncements on Article V. First, the amendment included a novel provi-
sion stating that it would “be inoperative unless it shall have been
ratified . . . within seven years.”
196
Opponents argued vigorously that
Congress could not impose a time limit on ratification when Article V says
nothing about the matter. In Dillon v. Gloss, however, the Court rejected a
challenge on this ground in concluding “that the fair inference or impli-
cation from Article V is that the ratification must be within some reasona-
ble time after the proposal.”
197
The Court continued:
Whether a definite period for ratification shall be fixed so that all
may know what it is and speculation on what is a reasonable time
may be avoided, is, in our opinion, a matter of detail which
Congress may determine as an incident of its power to designate
the mode of ratification.
198
In other words, Congress’s power to impose a time limit is either implicit
in Article V or perhaps grounded in the Necessary and Proper Clause.
Second, a wide range of scholars, states, and alcohol-industry groups
argued that the Eighteenth Amendment violated inherent individual-
193. See id. at 213. As Kyvig relates, the “racially fraught issue of state or federal control
of elections blocked agreement” in the conference committee. Id.
194. Prakash, Of Synchronicity, supra note 16, at 1268–71. “As a matter of constitutional
structure,” Prakash explains, “I believe that whatever the rule for the bicameral passage of
statutes, the same period ought to apply to bicameral proposal of an amendment.” Id. at
1270. And “[t]he Constitution, by incorporating the concept of ‘session,’ provided that bills
must be perfected into law in a single session.” Id. at 1264.
195. Id. at 1270.
196. U.S. Const. amend. XVIII, § 3.
197. 256 U.S. 368, 375 (1921).
198. Id. at 376.
2021] PUZZLES AND POSSIBILITIES 2355
rights and federalism limitations on Article V.
199
A district court in 1930
endorsed a procedural variant on the federalism objection, holding that
the amendment should have been ratified by state conventions rather than
state legislatures given the extent to which it transferred power to the fed-
eral government.
200
As far as we are aware, this is the only instance in which
a federal court has invalidated an Article V amendment. The Supreme
Court reversed.
201
Third, the Eighteenth Amendment raised the question whether ref-
erenda may play a part in the process of state ratification. In Ohio, citizens
concerned about malapportionment in the state legislature decided to
amend the state constitution to require that controversial federal amend-
ments be submitted to a popular referendum after approval by the legisla-
ture.
202
Soon afterward, the Ohio legislature approved the Eighteenth
Amendment by a wide margin, and the U.S. Secretary of State included
Ohio’s ratification in the ocial count when he proclaimed the
Eighteenth Amendment adopted.
203
The people of Ohio, however, nar-
rowly rejected the Eighteenth Amendment in a referendum after this proc-
lamation.
204
The Supreme Court held that this referendum was a
constitutional nullity in Hawke v. Smith (No. 1).
205
The Court said that the
“legislature,” as used in Article V, is “the representative body which
ma[kes] the laws of the people.”
206
The decision was excoriated in the
press,
207
and both its reasoning and its “elitist anti-populis[t]” tone con-
tinue to garner criticism.
208
Shortly after Hawke, the Court rejected an array of other challenges
to the Eighteenth Amendment, including that Congress had not properly
found the amendment “necessary”; that “two thirds” means two-thirds of
199. See Jason Mazzone, Unamendments, 90 Iowa L. Rev. 1747, 1815 n.364 (2005) (col-
lecting scholarly sources arguing that the Eighteenth Amendment violated inherent sub-
stantive limits on the amending power); Roznai, supra note 79, at 673 n.107 (same).
200. United States v. Sprague, 44 F.2d 967, 981 (D.N.J. 1930), rev’d, 282 U.S. 716
(1931).
201. Sprague, 282 U.S. at 734.
202. See Kyvig, supra note 141, at 242. Specifically, a ratification by the legislature would
not be eective for ninety days. Id. If, during that period, six percent of voters signed a
petition for a referendum, the legislature’s action would not be eective unless approved
by a majority of voters in the referendum. Id.
203. Id.
204. Id. at 243.
205. 253 U.S. 221, 230–31 (1920).
206. Id. at 227.
207. See Kyvig, supra note 141, at 246 (collecting prominent criticisms); see also id. at
245 (“Hawke v. Smith left a large and lasting impression that the Article V amending process
denied democratic choice in the case of national prohibition.”).
208. Vikram David Amar, The People Made Me Do It: Can the People of the States
Instruct and Coerce Their State Legislatures in the Article V Constitutional Amendment
Process?, 41 Wm. & Mary L. Rev. 1037, 1077–78 (2000).
2356 COLUMBIA LAW REVIEW [Vol. 121:2317
the whole membership of Congress, not a quorum; and that the amend-
ment transgressed substantive limits on the amending power.
209
The Court
heard an extraordinary five days of oral argument on these questions and
related ones involving the scope of the amendment and the National
Prohibition Act.
210
But the Court issued only a brief per curiam opinion,
stating its conclusions “without an exposition of the reasoning by which
they ha[d] been reached.”
211
H. The Nineteenth Amendment
The women’s surage amendment was ratified on the heels of
Prohibition and raised similar procedural issues. A companion case
decided the same day as Hawke held that the Ohio legislature’s ratification
of the Nineteenth Amendment could not be overturned by referendum.
212
When the case was argued, thirty-five of the thirty-six states needed to ratify
the amendment had done so. If Hawke had come out the other way, it
would have been a significant setback for the amendment eort, as there
were movements afoot to hold referenda in several states.
213
Hawke also led indirectly to the final ratification of the Nineteenth
Amendment, which unfolded in a dramatic scene in Tennessee.
214
The
Tennessee Constitution contained (and still contains) a provision
prohibiting the legislature from acting on a federal constitutional
amendment without an intervening election following Congress’s
proposal of the amendment.
215
Governor Albert Roberts refused to call a
special session of the legislature to consider the Nineteenth Amendment
before election day, thinking himself constrained by this provision. But
after Hawke and some arm twisting by President Woodrow Wilson, he
209. Nat’l Prohibition Cases, 253 U.S. 350, 386 (1920).
210. See Kyvig, supra note 141, at 246. Edward Corwin opined at the time that “[n]ot
since the Milligan case was argued in 1866 has a more notable array of counsel stood up
before the court.” Edward S. Corwin, Constitutional Law in 1919–1920. I: The
Constitutional Decisions of the Supreme Court of the United States in the October Term,
1919, 14 Am. Pol. Sci. Rev. 635, 651 (1920). For a colorful eyewitness account, see 2 Philip
C. Jessup, Elihu Root 479–80 (1938).
211. Nat’l Prohibition Cases, 253 U.S. at 388 (White, C.J., concurring); see also id.
(expressing “profound[] regret” over this absence of explanation). Justice Oliver Wendell
Holmes wrote elliptically in a letter to then-Professor Felix Frankfurter that there were
“good reasons” for “not giving reasons,” but he did not elaborate. Alexander M. Bickel &
Benno C. Schmidt, Jr., The Judiciary and Responsible Government 1910–21, at 546 (1984)
(quoting Letter from Oliver W. Holmes to Felix Frankfurter (June 22, 1920)). To this day,
the “reasons for not giving reasons that Holmes alluded to remain obscure.” Id. at 546–47.
212. Hawke v. Smith (No. 2), 253 U.S. 231, 232 (1920).
213. See Kyvig, supra note 141, at 244.
214. On this scene, see generally Robert B. Jones & Mark E. Byrnes, The “Bitterest
Fight”: The Tennessee General Assembly and the Nineteenth Amendment, 68 Tenn. Hist.
Q. 270 (2009).
215. Tenn. Const. art. II, § 32.
2021] PUZZLES AND POSSIBILITIES 2357
convened the legislature, which voted to ratify by a razor-thin margin.
216
Opponents of the amendment immediately filed a motion to reconsider,
only to realize that their motion would fail unless they could buy time to
persuade more colleagues. Thirty-seven antisurage legislators therefore
decamped to Decatur, Alabama, to deprive the House of a quorum.
217
Governor Roberts certified that Tennessee had ratified the amendment,
and President Wilson’s Secretary of State in turn certified the Nineteenth
Amendment.
218
But that did not stop the Decatur contingent from
returning; ordering the sergeant-at-arms to amass a quorum; and, with
many prosurage members absent from the hall, voting to grant
reconsideration and reject ratification.
219
This mess came before the U.S. Supreme Court in Leser v. Garnett.
220
The Court held that the ratification by Tennessee (and several other
states) could be treated as valid even if obtained in violation of the state
constitution or state rules of legislative procedure.
221
Once a state legisla-
ture has transmitted a “duly authenticated” ratification resolution to the
U.S. Secretary of State, the Court suggested, federal judges should not look
behind the curtain.
222
Separately, the Court also rejected a claim that the
Nineteenth Amendment was invalid because “so great an addition to the
electorate, if made without the State’s consent, destroys its autonomy as a
political body.”
223
The Court noted that this theory, if accepted, would also
invalidate the Fifteenth Amendment, which “ha[d] been recognized and
acted on for half a century.”
224
And that proposition, whatever its legal
merits, simply “cannot be entertained.”
225
216. See Kyvig, supra note 141, at 238, 245.
217. A trial court also granted an injunction prohibiting the governor from certifying
the ratification vote to Washington, but the injunction was quashed by the Tennessee
Supreme Court. See Clements v. Roberts, 230 S.W. 30, 36 (Tenn. 1921).
218. See Kyvig, supra note 141, at 238.
219. Jones & Byrnes, supra note 214, at 286.
220. 258 U.S. 130 (1922).
221. Id. at 137.
222. Id.
223. Id. at 136. For a comprehensive summary of arguments made against the substan-
tive validity of the Nineteenth Amendment, which the Court briskly brushed aside in Leser,
see Reva B. Siegel, She the People: The Nineteenth Amendment, Sex Equality, Federalism,
and the Family, 115 Harv. L. Rev. 947, 997–1006 (2002).
224. Leser, 258 U.S. at 136.
225. Id.
2358 COLUMBIA LAW REVIEW [Vol. 121:2317
I. The Twenty-First Amendment
226
The Twenty-First Amendment, “repeal[ing]” Prohibition,
227
is the
only amendment in U.S. history ratified by state conventions rather than
state legislatures. That choice sprang from a lingering feeling, intensified
by Hawke, that ratification of the Eighteenth Amendment had not accu-
rately reflected public opinion.
228
Conventions would allow for a more
direct appeal to the people. As a result, the Twenty-First Amendment
brought up a host of new legal questions that have never been definitively
resolved or adjudicated.
The threshold question was whether Congress or the states should set
the rules governing the formation and operation of the conventions. Many
members of Congress, the bar, and the academy expressed the view that
Congress had the power to set these rules if it wished, although this view
was by no means unanimous.
229
But Congress never really had to bother.
226. We are not aware of any novel challenges to the validity of the Twentieth
Amendment, which moved up the dates of the presidential inauguration and the beginning
of the congressional term to shorten the lame-duck period. Presumably, this reflects the
larger lack of controversy over the amendment: Within about a year after being proposed
by Congress, it became the first amendment to be unanimously ratified by the states on the
first pass. See Kyvig, supra note 141, at 274 (“To say that the states welcomed the lame duck
amendment would be an understatement.”); Edward J. Larson, The Constitutionality of
Lame-Duck Lawmaking: The Text, History, Intent, and Original Meaning of the Twentieth
Amendment, 2012 Utah L. Rev. 707, 734 (“Because the states acted so quickly and with
virtually no dissent, there was often little debate in the state legislatures.”). The House’s
proposed version of the amendment included a provision requiring that “ratification shall
be by legislatures, the entire membership of at least one branch of which shall have been
elected subsequent to [the] date of submission.” 75 Cong. Rec. 4059 (1932). That provision
was removed by the conference committee when the House and Senate versions of the
amendment were reconciled, but it serves as an interesting precedent for building novel
conditions of ratification (beyond deadlines) into the text of proposed amendments. See
infra notes 377–380 and accompanying text.
227. U.S. Const. amend. XXI, § 1.
228. See Ratification of the Twenty-First Amendment to the Constitution of the United
States: State Convention Records and Laws 3 (Everett Somerville Brown ed., 1938)
[hereinafter Ratification of the Twenty-First Amendment] (“Much of the criticism of the
Eighteenth Amendment was based on the claim that its ratification had not properly
reflected the opinion of the people of the country.”). Defiance of Prohibition was wide-
spread. As Pauline Sabin, the head of a Prohibition reform organization, put it: Proponents
“thought they could make prohibition as strong as the Constitution, but instead they have
made the Constitution as weak as prohibition.” Sean Beienburg, Prohibition, the
Constitution, and States’ Rights 229 (2019).
229. For summaries of these debates, see Clement E. Vose, Constitutional Change:
Amendment Politics and Supreme Court Litigation Since 1900, at 111–12 (1972)
[hereinafter Vose, Constitutional Change]; Abraham C. Weinfeld, Power of Congress Over
State Ratifying Conventions, 51 Harv. L. Rev. 473, 474–75 (1938). The states themselves were
also in disagreement. California petitioned Congress to pass a law regulating the conven-
tions. See Ratification of the Twenty-First Amendment, supra note 228, at 515. New Mexico,
on the other hand, declared that any rules from Congress would be “null and void” and that
state ocers were “authorized and required to resist to the utmost any attempt to execute
any and all such congressional dictation and usurpation.” Id. In the end, twenty-one states
2021] PUZZLES AND POSSIBILITIES 2359
By the time Congress proposed the amendment, twenty-nine states were
already in the process of legislating how their conventions would take
place.
230
Many of the choices made by states in establishing these conventions
were contentious and challenged at the time.
231
“The widespread use of at-
large districts,” for instance, “disregarded both historical precedent and
the advice of legal and constitutional experts commissioned to make rec-
ommendations.”
232
Indeed, even though more than half the states used at-
large elections, the Maine Supreme Judicial Court held that this practice
violated Article V.
233
A number of states required delegates to vote in
accord with whatever position they had pledged to support when elected
or in accord with the result of a statewide referendum.
234
This, too, was
controversial, because it broke with the historical understanding of a con-
vention as an authentically deliberative body. As one leading scholar put
it, “a delegate must be free to decide how he individually will vote, to ratify
or not to ratify,” and “he must not be subject to any legal process either to
compel him to vote in a given way or to punish him if he does not so
vote.”
235
The Maine high court agreed.
236
The Alabama Supreme Court
came out the opposite way, upholding a law requiring delegates to pledge
that they would abide by the result of a statewide referendum on the
ground that a convention is “more truly representative when expressing
the known will of the people.”
237
In the end, the lawsuits challenging the ratification process did not
carry the day. But the U.S. Supreme Court never weighed in.
238
And the
included “escape hatch” provisions in their state convention laws, providing that any federal
law would take precedence if passed. Id.
230. Clement E. Vose, Repeal as a Political Achievement, in Law, Alcohol, and Order:
Perspectives on National Prohibition 97, 112 (David E. Kyvig ed., 1985).
231. See Vose, Constitutional Change, supra note 229, at 121–26 (describing litigation
filed by opponents of repeal).
232. Thomas F. Schaller, Democracy at Rest: Strategic Ratification of the Twenty-First
Amendment, Publius, Spring 1998, at 81, 88; see also Noel T. Dowling, A New Experiment
in Ratification, 19 A.B.A. J. 383, 384 (1933) (“Election at large, for all delegates, involves a
break-away from prior notions concerning the composition of conventions. The idea of local
representation permeates our political thinking and possibly is to some extent imported
into the Constitution by the term ‘convention.’”).
233. In re Opinion of the Justices, 167 A. 176, 179 (Me. 1933) (“[W]e do not deem it
permissible for the state, under the terms of article 5 of the Federal Constitution, to organize
a convention wherein the delegates entitled to participate are all elected at large.”).
234. Everett S. Brown, The Ratification of the Twenty-First Amendment, 29 Am. Pol.
Sci. Rev. 1005, 1013–14 (1935).
235. Dowling, supra note 232, at 385.
236. In re Opinion of the Justices, 167 A. at 180 (“A convention is a body or assembly
representative of all the people of the state. The convention must be free to exercise the
essential and characteristic function of rational deliberation.”).
237. In re Opinion of the Justices, 148 So. 107, 111 (Ala. 1933).
238. See Vose, Constitutional Change, supra note 229, at 121 (“No great cases resulted
and none ever reached the United States Supreme Court.”). In addition to the Alabama
2360 COLUMBIA LAW REVIEW [Vol. 121:2317
litigation did little to settle or clarify what constraints Article V might
impose on the process of ratification by state conventions.
239
J. The Twenty-Second Amendment
The Twenty-Second Amendment codified the unwritten norm, which
President Franklin Roosevelt had transgressed, against presidents serving
more than two terms.
240
From a procedural point of view, the most striking
thing about the amendment is the final vote tally in the House of
Representatives: 81 to 29.
241
The size of the House, then as now, was 435
members.
242
A quorum therefore required at least 218 members, and two-
thirds of a quorum required at least 146 members. How was the Twenty-
Second Amendment sent to the states with so few “yeas”—less than a fifth
of the total House membership?
The answer comes from the House Rules. As the Supreme Court has
observed, the “Constitution has prescribed no method” for determining
the presence of a quorum, “and it is therefore within the competency of
the house to prescribe any method which shall be reasonably certain to
ascertain the fact.”
243
The “method” the House has settled on is to presume
the presence of a quorum unless the fact is challenged.
244
Furthermore,
the Speaker of the House determined in 1890 that all members present in
the House would count toward a quorum, even if they did not vote.
245
The
and Maine cases discussed above, the high courts of Missouri and Ohio held that acts estab-
lishing their convention procedures did not have to be submitted to a referendum like other
laws. See State ex rel. Tate v. Sevier, 62 S.W.2d 895, 898 (Mo. 1933); State ex rel. Donnelly
v. Myers, 186 N.E. 918, 918–19 (Ohio 1933).
239. Cf. Recent Cases, 47 Harv. L. Rev. 126, 130 (1933) (observing that “[t]he opposing
views of [the Alabama and Maine] decisions indicate the lack of both authority and satisfac-
tory analogies” regarding how ratification of Article V amendments by convention is sup-
posed to work).
240. See Josh Chafetz & David E. Pozen, How Constitutional Norms Break Down, 65
UCLA L. Rev. 1430, 1435–37 (2018) (describing the Twenty-Second Amendment as a classic
example of an imperiled norm being “displaced by law”).
241. 93 Cong. Rec. 2392 (1947).
242. See Determining Apportionment, U.S. House of Representatives Hist.,
Art & Archives, https://history.house.gov/Institution/Apportionment/Determining-
Apportionment/ [https://perma.cc/S94V-XW4X] (last visited Aug. 5, 2021).
243. United States v. Ballin, 144 U.S. 1, 6 (1892).
244. See Charles W. Johnson, John V. Sullivan & Thomas J. Wickham, Jr., House
Practice: A Guide to the Rules, Precedents, and Procedures of the House 756 (2017) (“A
quorum is presumed to be present unless a point of no quorum is entertained and the Chair
announces that a quorum is in fact not present or unless the absence of a quorum is dis-
closed by a vote or by a call of the House.”). The Court has implicitly blessed this method
in a separate context. See NLRB v. Noel Canning, 573 U.S. 513, 553 (2014) (“Senate rules
presume that a quorum is present unless a present Senator questions it.”).
245. See Johnson et al., supra note 244, at 760. The Speaker made this change to defeat
the obstructionist “disappearing quorum” tactic, by which minority-party members would
prevent a quorum by refusing to vote. See Josh Chafetz, Unprecedented? Judicial
Confirmation Battles and the Search for a Usable Past, 131 Harv. L. Rev. 96, 112–14 (2017).
2021] PUZZLES AND POSSIBILITIES 2361
upshot is that the House can pass a bill—or apparently an amendment—
with fewer than half of its members voting, as long as no one questions
whether a quorum is present.
246
That is what happened with the Twenty-Second Amendment. The
House passed by a vote of 285 to 121 an initial version of the proposed
amendment that would have prevented anyone from being elected
President who had served any part of two terms.
247
The amendment was
then reworked in the Senate into its current form, which provides that
someone who serves as President for less than two years of another’s term
may still be elected twice.
248
This was clearly a “material change,”
249
requir-
ing a new vote in the House. The new vote, approving the Senate version,
was 81 to 29.
250
One representative “object[ed] to the vote on the ground
a quorum is not present.”
251
But he subsequently withdrew his objection
without explanation, and a count never took place.
252
While the final vote seems to have been eective under House rules,
there is reason to be skeptical. Even if one is willing to accept that a bare
quorum may vote on a constitutional amendment, the idea that a handful
of representatives could propose an amendment in the absence of an
actual quorum may be a bridge too far. The final House vote on the
Twenty-Second Amendment severely strained both the text (“two thirds of
both Houses”) and the supermajoritarian spirit of Article V. Moreover, the
Supreme Court has given added reason for skepticism. In the National
Prohibition Cases, the Court approved the application of the general
quorum rule to constitutional amendments, but it advised: “The two-thirds
vote in each house which is required in proposing an amendment is a vote
of two-thirds of the members present—assuming the presence of a
quorum.”
253
The Court did not say two-thirds of the members present and
voting. Perhaps this was a slip-up by the Court, but the Twenty-Second
246. See Thomas J. Wickham, Constitution, Jeerson’s Manual, and Rules of the House
of Representatives 83 (2019) (“The majority required to pass a constitutional amend-
ment . . . is two-thirds of those Members voting either in the armative or negative, a
quorum being present, and Members who only indicate that they are ‘present’ are not
counted in this computation.”).
247. See Stephen W. Stathis, The Twenty-Second Amendment: A Practical Remedy or
Partisan Maneuver?, 7 Const. Comment. 61, 67 (1990).
248. Id. at 68; see U.S. Const. amend. XXII.
249. 93 Cong. Rec. 2389 (1947) (statement of Rep. Michener).
250. Id. at 2392.
251. Id. (statement of Rep. Forand).
252. See Corwin & Ramsey, supra note 96, at 192–94.
253. 253 U.S. 350, 386 (1920). A half-century earlier, Senator Lyman Trumbull had put
the point similarly on the floor of Congress: “[I]t was decided” during debates on the
Corwin Amendment, he said, that “two thirds of the Senators present, a quorum being pre-
sent, was sucient to carry a constitutional amendment.” Cong. Globe, 40th Cong., 3d Sess.
1642 (1869) (emphasis added).
2362 COLUMBIA LAW REVIEW [Vol. 121:2317
Amendment emphatically did not receive a vote of two-thirds of a present
quorum.
254
To be clear, we are not suggesting that judges (or anyone else) today
could or should disregard the Twenty-Second Amendment, which has
been an entrenched feature of our constitutional order for more than
seventy years. As with the Sixteenth Amendment, courts would almost
certainly dismiss any lawsuit challenging the Twenty-Second Amendment
without reaching the merits.
255
But the amendment’s current status as part
of the Constitution is not due to painstaking, or even particularly
conscientious, adherence to the rules of Article V.
K. The Twenty-Seventh Amendment?
256
The Twenty-Seventh Amendment brings this story full circle. It was
proposed by Congress in the early days of the republic, but it is the most
recent amendment to have gained entry to the Constitution . . . if indeed
it did gain entry. The fact that leading constitutional scholars still refer to
254. We have not been able to ascertain whether those who voted on the Twenty-Second
Amendment were the only members present in the House, in which case there was not a
quorum, or whether there was in fact a quorum present in the House (in keeping with the
presumption in the House Rules) and the amendment failed to obtain a “vote of two-thirds
of the members present.” Nat’l Prohibition Cases, 253 U.S. at 386. Either way, Article V
would not be satisfied per the rule announced by the Court in 1920.
255. See supra notes 185–188 and accompanying text.
256. We are not aware of significant novel challenges to the Twenty-Third through
Twenty-Sixth Amendments. These “modest” amendments were “narrowly drawn reforms
that produced only marginal change.” Kyvig, supra note 141, at 349. The Twenty-Third
Amendment contained, for the first time, a ratification time limit in the resolution propos-
ing the amendment rather than in the text of the amendment itself. See Allison L. Held,
Sheryl L. Herndon & Danielle M. Stager, The Equal Rights Amendment: Why the ERA
Remains Legally Viable and Properly Before the States, 3 Wm. & Mary J. Women & L. 113,
126 (1997). The validity of this time limit was never tested, however, because the amend-
ment (which authorized presidential electors for the District of Columbia) was ratified
within a year. See Michael J. Garcia, Caitlain Devereaux Lewis, Andrew Nolan, Meghan
Totten & Ashley Tyson, Cong. Rsch. Serv., The Constitution of the United States of America,
S. Doc. No. 112-9, at 40 n.15 (2017). When the Twenty-Fifth Amendment was certified as
part of the Constitution, President Lyndon Johnson signed the certification, although
apparently only as a witness. See Certification of Amendment to Constitution of the United
States, 32 Fed. Reg. 3287, 3287–88 (Feb. 25, 1967); Remarks at Ceremony Marking the
Ratification of the Presidential Inability (25th) Amendment to the Constitution, 1 Pub.
Papers 217–18 (Feb. 23, 1967); Stephen W. Stathis, Presidential Disability Agreements Prior
to the 25th Amendment, 12 Pres. Stud. Q. 208, 212 (1982). The Twenty-Sixth Amendment,
which was the “most quickly ratified constitutional amendment in our history,” Tex.
Democratic Party v. Abbott, 978 F.3d 168, 186 (5th Cir. 2020), featured a reprise of the
debate in Tennessee over whether a state constitution could validly require an intervening
election before the legislature ratified the amendment. The Tennessee Supreme Court held
this state constitutional limitation invalid. See Walker v. Dunn, 498 S.W.2d 102, 106 (Tenn.
1972); supra notes 214–225 and accompanying text.
2021] PUZZLES AND POSSIBILITIES 2363
it as the “purported”
257
or “alleged”
258
Twenty-Seventh Amendment high-
lights how many questions about Article V have remained open in the two
centuries between its proposal and ratification.
What is now regarded as the Twenty-Seventh Amendment, which pro-
hibits a congressional salary increase from taking eect before an inter-
vening election, was proposed as the second of twelve amendments sent to
the states in 1789 (the third through twelfth were the Bill of Rights). By
the time the Bill of Rights was ratified in 1791, only six states had ratified
the so-called Congressional Pay Amendment, and it fell into a long state
of dormancy.
259
About a hundred years later, the Ohio legislature voted to
ratify the amendment to protest the “Salary Grab” Act of 1873, in which
Congress gave itself a large and retroactive salary increase.
260
But that was
an isolated maneuver, and the concomitant introduction in Congress of
numerous proposals for a new congressional pay amendment suggests that
many regarded the original version as dead.
261
Nonetheless, the
Congressional Pay Amendment was resurrected a century later, owing to
the dogged eorts of a University of Texas undergraduate and public dis-
pleasure at congressional salary increases.
262
Maine and Colorado ratified the amendment in 1983 and 1984. News
coverage of those ratifications prompted a Wyoming legislator to report
that his state had also ratified the amendment in 1977—a fact that had
somehow escaped notice in the nation’s capital.
263
After that, it was o to
the races. By 1992, forty-one states had ratified.
264
Since 1984, the Archivist of the United States has been assigned the
statutory responsibility to “publish[]” new amendments.
265
The Archivist
sought the advice of the Justice Department’s Oce of Legal Counsel
(OLC). OLC opined that there was no time limit on ratification, brushing
aside the Supreme Court’s contrary language in Dillon as “dictum,” and
257. E.g., Sanford Levinson, Authorizing Constitutional Text: On the Purported
Twenty-Seventh Amendment, 11 Const. Comment. 101, 101 (1994) [hereinafter Levinson,
Authorizing Constitutional Text]; Adam M. Samaha, Dead Hand Arguments and
Constitutional Interpretation, 108 Colum. L. Rev. 606, 636 (2008).
258. E.g., Amar, Philadelphia Revisited, supra note 4, at 1046 n.3; Michael J. Perry, What
Is “the Constitution”? (And Other Fundamental Questions), in Constitutionalism:
Philosophical Foundations 99, 133 n.1 (Larry Alexander ed., 2001).
259. See Bernstein, supra note 25, at 532–33.
260. Id. at 534.
261. Id.
262. See Kyvig, supra note 141, at 464–65; Bernstein, supra note 25, at 536–39.
263. Kyvig, supra note 141, at 465; Bernstein, supra note 25, at 537.
264. Bernstein, supra note 25, at 539. The ratification of Idaho raised a variant of an old
problem: whether the state could require that a proposed amendment be submitted to the
people in a referendum prior to ratification. The Idaho Attorney General opined that such
a requirement was unconstitutional, but following a favorable referendum vote, the
legislature ratified anyway. Id.
265. Pub. L. No. 98-497, § 107(d), 98 Stat. 2280, 2291 (1984) (codified at 1 U.S.C.
§ 106b (2018)).
2364 COLUMBIA LAW REVIEW [Vol. 121:2317
that the Congressional Pay Amendment had been validly ratified.
266
OLC
also opined that Congress has no role in proclaiming an amendment part
of the Constitution, despite the contrary views of seven Justices in
Coleman.
267
In keeping with OLC’s advice, the Archivist “certif[ied]” in the
Federal Register that the “Amendment has become valid, to all intents and
purposes, as a part of the Constitution of the United States.”
268
Congressional leadership was “stunned.”
269
Several members took to
the floor to reassert Congress’s primacy in judging the validity of an
amendment. Senator Robert Byrd, for instance, invoked the “firm histori-
cal understanding . . . that the Executive’s function with regard to certify-
ing constitutional amendments is purely ministerial” and insisted that
“Congress should have the opportunity to decide substantive questions”
of amendment validity.
270
Numerous members of Congress expressed con-
cern about whether ratification of the Twenty-Seventh Amendment would
furnish a precedent for reviving other seemingly lapsed amendments.
271
Nevertheless, in the end Congress voted overwhelmingly for a resolution
validating the Twenty-Seventh Amendment.
272
Yet even as countless pocket Constitutions and textbooks have flown
o the presses since 1992 with the Twenty-Seventh Amendment included,
“there is, at least at this time, no consensually agreed-upon, positivistic
‘given’ that allows us to say that we are simply engaging in description
when granting the amendment the status of ‘law.’”
273
Two of the foremost
266. Cong. Pay Amendment, 16 Op. O.L.C. 85, 97 (1992). The Court in Dillon, recall,
had unanimously described the idea that the Congressional Pay Amendment might still be
ratified “quite untenable.” Dillon v. Gloss, 256 U.S. 368, 375 (1921); see supra note 23.
267. Cong. Pay Amendment, 16 Op. O.L.C. at 99.
268. Certification of Amendment to the Constitution of the United States Relating to
Compensation of Members of Congress, 57 Fed. Reg. 21,187, 21,188 (May 19, 1992).
269. Bernstein, supra note 25, at 540.
270. 138 Cong. Rec. 11,654 (1992) (statement of Sen. Byrd); see also, e.g., id. at 11,779
(statement of Rep. Fish) (“Where . . . there may be lingering concerns as to the validity of
the amendment, it is appropriate for Congress to resolve such doubts . . . .”); id. at 11,860
(statement of Sen. Grassley) (“[T]he Supreme Court made clear in 1939 in the Coleman
decision that Congress has the authority to say whether the timeliness standard has been
met.”); id. at 11,871 (statement of Sen. Roth) (stating that “the functions of” the Archivist
“are ministerial only: To count the number of ratifications of an amendment and not to act
as a constitutional tribunal to ‘decide doubtful questions’ of law”).
271. See, e.g., id. at 11,655 (statement of Sen. Byrd) (“I do not intend our action with
regard to this amendment to serve as a precedent or model for any other amend-
ment . . . .”); id. at 11,780 (statement of Rep. Edwards) (“The House . . . should be clear
that this is an exception, not a precedent.”). Although one senator stated that the Senate
had “decided to declare that four [sic] other proposed and pending amendments . . . were
to be considered to have lapsed,” id. at 11,870 (statement of Sen. Sanford), neither the
Senate nor the House took formal action on that question. Senator Jesse Helms provided
the more accurate assessment: “I regret that some questions are left unanswered.” Id. at
11,871.
272. Id. at 11,869 (Senate); id. at 12,052 (House).
273. Levinson, Authorizing Constitutional Text, supra note 257, at 113.
2021] PUZZLES AND POSSIBILITIES 2365
academic authorities on the amendment process, Dellinger and Prakash,
maintain that the amendment is not part of the Constitution.
274
Justice
Scalia stated at a public event that he is inclined to agree.
275
Ackerman has
written that “the so-called twenty-seventh amendment should be treated
as a bad joke by sensible citizens.”
276
A unanimous Supreme Court,
including Justices Oliver Wendell Holmes and Louis Brandeis, suggested
that the amendment was unrevivable a century ago.
277
On the other hand,
eminent constitutional scholars such as Amar, Michael Stokes Paulsen,
and Laurence Tribe have taken the opposite position.
278
Perhaps the best
an editor of the Constitution can do is to adopt the approach suggested by
William Van Alstyne: Include the Twenty-Seventh Amendment, but with
an asterisk.
279
In light of the history of legal contestation and confusion recounted
above, an asterisk would be a fitting way for the written Constitution to
end.
L. The Twenty-Eighth Amendment?
Or maybe that is not quite the end. Supporters of other amendments
proposed by Congress over the years claim that these amendments, too,
have already become part of the canonical document. As a matter of
Article V law, it is not clear that all of these claims are wrong.
1. Article the First. — The very first amendment proposed by Congress,
denominated “Article the First” in the same package of amendments that
contained the Bill of Rights and the putative Twenty-Seventh, involved the
size of the House of Representatives.
280
The ratification history is obscure,
274. See Prakash, Of Synchronicity, supra note 16, at 1283; Paul M. Barrett & David
Rogers, A Timely Measure Gains Ratification After Two Centuries, Wall St. J., May 8, 1992,
at A10 (quoting Dellinger).
275. Tribe, The Invisible Constitution, supra note 10, at 4.
276. Ackerman, We the People, supra note 7, at 490 n.1.
277. See Dillon v. Gloss, 256 U.S. 368, 375 (1921).
278. See Amar, America’s Constitution, supra note 49, at 457; Paulsen, A General
Theory, supra note 50, at 682–83; Laurence H. Tribe, The 27th Amendment Joins the
Constitution, Wall St. J., May 13, 1992, at A15.
279. William Van Alstyne, What Do You Think About the Twenty-Seventh Amendment?,
10 Const. Comment. 9, 15 (1993).
280. See Resolution of Congress Proposing Amendatory Articles to the Several States,
in Documentary History, supra note 103, at 321, 321–22. Article the First would have set an
initial minimum number of representatives based on population, until the national popula-
tion reached a certain number, at which point it would have mandated a minimum of 200
representatives and a maximum of one per fifty thousand. Id.; see also Amar, The Bill of
Rights, supra note 94, at 1143–45. The language is convoluted enough that unless one
adopts some sort of “scrivener’s error” theory, its meaning is misrepresented on the Senate’s
ocial website. See Congress Submits the First Constitutional Amendments to
the States, U.S. Senate, https://www.senate.gov/artandhistory/history/minute/
Congress_Submits_1st_Amendments_to_States.htm [https://perma.cc/8TNE-QG55] (last
visited Aug. 5, 2021) (claiming that if Article the First were adopted, the House “would today
have more than 6,000 members”). A possible source of confusion is that a congressional
2366 COLUMBIA LAW REVIEW [Vol. 121:2317
but it appears to have fallen just short.
281
Of the states that voted to add
the Bill of Rights to the Constitution, only Delaware declined to ratify
Article the First, probably because, as a small state, its power would shrink
in a larger House.
282
This left the amendment one state shy of ratification.
And the failure of Massachusetts and Connecticut to ratify not only this
amendment but also the Bill of Rights “poses something of a mystery.”
283
In Massachusetts, the two houses of the legislature approved slightly dif-
ferent sets of amendments, and a joint committee never reconciled the
dierences.
284
In Connecticut, the lower house of the legislature appar-
ently approved eleven of twelve proposed amendments in 1789, including
Article the First, but the upper house tabled the issue until the next ses-
sion.
285
At the next session, the upper house approved all twelve amend-
ments, but the lower house only approved the Bill of Rights (and maybe
the Congressional Pay Amendment—the journal contradicts itself).
286
The
two houses could not reconcile their resolutions, so no notice of
ratification was ever sent to the capital.
A recent pair of lawsuits, however, alleges that Connecticut did ratify
Article the First. The basic claim is that the ratification was eective
conference committee at the eleventh hour “inexplicably” changed “less” to “more” in the
final clause, converting a minimum to a maximum and introducing a “technical glitch[]”
into the formula. Amar, The Bill of Rights, supra note 94, at 1143.
281. See Amar, The Bill of Rights, supra note 94, at 1143. As Amar explains, numerous
historians have bungled this ratification count in a variety of ways. Id. at 1143 n.52.
282. For the ratifying states, see Virginia Ratification Resolution (Nov. 4, 1791), in
Documentary History, supra note 103, at 377, 385; Vermont Ratification Resolution (Oct.
27, 1790), in Documentary History, supra note 103, at 371, 373–74; New Jersey Ratification
Resolution (Aug. 3, 1790), in Documentary History, supra note 103, at 325, 327; Rhode
Island Ratification Resolution (June 15, 1790), in Documentary History, supra note 103, at
363, 364; New York Ratification Resolution (Mar. 27, 1790), in Documentary History, supra
note 103, at 357, 358; New Hampshire Ratification Resolution (Jan. 25, 1790), in
Documentary History, supra note 103, at 345, 346; South Carolina Ratification Resolution
(Jan. 19, 1790), in Documentary History, supra note 103, at 340, 341; North Carolina
Ratification Resolution (Dec. 22, 1789), in Documentary History, supra note 103, at 335,
337; Maryland Ratification Resolution (Dec. 19, 1789), in Documentary History, supra note
103, at 330, 331. Pennsylvania first rejected the amendment, Pennsylvania Ratification
Resolution (Mar. 11, 1790), in Documentary History, supra note 103, at 352, 353, but then
ratified on September 21, 1791, Pennsylvania Ratification Resolution (Sept. 21, 1791), in
Documentary History, supra note 103, at 367, 369. “The files of the Department of State
contain no notices of ratification from the legislatures of Connecticut, Georgia, or
Massachusetts.” Denys P. Myers, Massachusetts and the First Ten Amendments to the
Constitution, S. Doc. No. 74-181, at 7 (1936) [hereinafter Myers, Massachusetts and the First
Ten Amendments]; see also Ames, supra note 55, at 320.
283. Alan P. Grimes, Democracy and the Amendments to the Constitution 28 n.40
(1978).
284. Id.; Myers, Massachusetts and the First Ten Amendments, supra note 282, at 11–
12.
285. Thomas H. Le Duc, Connecticut and the First Ten Amendments to the Federal
Constitution, S. Doc. No. 75-96, at 3 (1937).
286. Id.; see also Grimes, supra note 283, at 28 n.40.
2021] PUZZLES AND POSSIBILITIES 2367
because both houses approved Article the First, albeit during two dierent
legislative sessions, and that any state rule requiring approval by both
chambers during a single session is superseded by Article V.
287
Both law-
suits, unsurprisingly, were dismissed on standing and political question
grounds.
288
But the legal arguments they press are by no means untenable
on the merits.
2. The Titles of Nobility Amendment. — The so-called Titles of Nobility
Amendment, proposed by Congress in 1810, has had a similarly checkered
legal history.
289
It fell short of the necessary number of state ratifications.
But confusion about its status persisted for years. In an ocial 1815 com-
pilation of the laws of the United States, published with the authorization
of Congress, the Titles of Nobility Amendment appeared in the
Constitution as the Thirteenth Amendment.
290
The editors of the compi-
lation explained that there had “been some diculty in ascertaining”
whether it had been ratified, because the “evidence to be found in the
oce” of the Secretary of State was “defective.”
291
The editors “considered
[it] best” to include the supposed “thirteenth” amendment with that pref-
atory note.
292
The ocial edition of the Constitution printed for the
Fifteenth Congress also contained the amendment.
293
Secretary of State
John Quincy Adams, in a letter from 1817, seemed to be under the impres-
sion that the amendment was part of the Constitution.
294
Even today, his-
torians give conflicting accounts of how close the amendment was to
ratification.
295
287. See LaVergne v. House of Representatives, 392 F. Supp. 3d 108 (D.D.C. 2019)
(three-judge court).
288. LaVergne v. Bryson, 497 F. App’x 219, 221–23 (3d Cir. 2012); LaVergne, 392 F.
Supp. 3d at 117–19.
289. This proposed amendment provides in full:
If any citizen of the United States shall accept, claim, receive or retain any
title of nobility or honor, or shall, without the consent of Congress, accept
and retain any present, pension, oce or emolument of any kind
whatever, from any emperor, king, prince or foreign power, such person
shall cease to be a citizen of the United States, and shall be incapable of
holding any oce of trust or profit under them, or either of them.
Resolution Proposing an Amendment to the Constitution of the United States (Nov. 27,
1809), in Documentary History, supra note 103, at 452, 452.
290. See 1 Laws of the United States of America 74 (John Bioren & W. John Duane eds.,
1815).
291. Id. at ix.
292. Id.
293. See Ames, supra note 55, at 188; Silversmith, supra note 100, at 587 (citing 31
Annals of Cong. 530–31 (1817)).
294. Silversmith, supra note 100, at 587 n.62. Adams later corrected himself. Id.
295. Compare, e.g., Bernstein with Agel, supra note 16, at 176 (“one state short”), and
Kyvig, supra note 141, at 117 (one state short), with Silversmith, supra note 100, at 595–96
(“[I]t was never a single ratification short . . . .”). Silversmith appears to have the better of
the argument.
2368 COLUMBIA LAW REVIEW [Vol. 121:2317
Throughout the early to mid-1800s, “the general public continued to
think that [the Titles of Nobility] amendment had been adopted.”
296
The
amendment frequently appeared in ocial state codes, in printed editions
of the Constitution, and in textbooks.
297
By the end of the nineteenth cen-
tury, it was “commonly recognized” that the Titles of Nobility Amendment
had failed.
298
But the whole episode confirms once again that the consti-
tutional text is constituted to a significant extent by public attitudes and
that those attitudes can become detached from Article V.
3. The Equal Rights Amendment. — The final, and in our view strongest,
contender to be the twenty-eighth amendment is the ERA, which would
prohibit the denial or abridgement of “[e]quality of rights under the
law . . . by the United States or by any State on account of sex.”
299
The
odyssey of the ERA recapitulates many of the uncertainties already can-
vassed above.
300
It was proposed by Congress in 1972, and the accompany-
ing resolution set a ratification deadline of seven years.
301
As that deadline
approached, only thirty-five states (of thirty-eight needed) had ratified,
and several of those had purported to rescind their ratifications,
302
so
Congress voted by simple majority to extend the ratification deadline by
three years.
303
President Jimmy Carter signed the resolution extending the
deadline while disclaiming that it was constitutionally necessary.
304
No
296. Ames, supra note 55, at 188–89.
297. See Gideon M. Hart, The “Original” Thirteenth Amendment: The Misunderstood
Titles of Nobility Amendment, 94 Marq. L. Rev. 311, 366–67 & nn.302–316 (2010)
(collecting sources).
298. Silversmith, supra note 100, at 593; accord Hart, supra note 297, at 367. This has
not stopped certain “extremist” groups from claiming that the amendment was ratified and
then suppressed by unscrupulous lawyers. Silversmith, supra note 100, at 580–81.
299. Proposed Amendment to the United States Constitution, 92 H.R.J. Res. 208, 92d
Cong., 86 Stat. 1523 (1972).
300. For a broad overview of the legal and political fight for the ERA from the early
1900s to the present, see Julie C. Suk, We the Women: The Unstoppable Mothers of the
Equal Rights Amendment (2020) [hereinafter Suk, We the Women].
301. Joint Resolution Proposing an Amendment to the Constitution of the United States
Relative to Equal Rights for Men and Women, H.R.J. Res. 208, 92d Cong., 86 Stat. 1523
(1972).
302. Ratification of the Equal Rights Amendment, 44 Op. O.L.C., slip op. at 6–7 & nn.6–
8 (Jan. 6, 2020). Four state legislatures voted to rescind their ratifications, and one passed a
resolution providing that its ratification would be withdrawn if the ERA was not ratified dur-
ing the seven-year period. Id. at 7 & n.8. Further complicating matters, the acting governor
of Kentucky vetoed the state’s purported rescission while the governor was on vacation.
Acting Governor Vetoes Kentucky Rights Reversal, N.Y. Times, Mar. 21, 1978, at 20. A sixth
state, North Dakota, recently passed a joint resolution “clarifying” that its initial ratification
expired in 1979. S. Con. Res. 4010, 67th Legis. Assemb., Reg. Sess. (N.D. 2021).
303. Joint Resolution Extending the Deadline for the Ratification of the Equal Rights
Amendment, H.R.J. Res. 638, 95th Cong., 92 Stat. 3799 (1978).
304. Jimmy Carter, U.S. President, Speech at a Ceremony Commemorating the
Extension of the Ratification Deadline of the Equal Rights Amendment: Remarks on
Signing H.J. Res. 638, 2 Pub. Papers 1800 (Oct. 20, 1978).
2021] PUZZLES AND POSSIBILITIES 2369
additional states ratified before the expiration of the new deadline. Begin-
ning in 2017, however, Nevada, Illinois, and Virginia each ratified, pushing
the total number of ratifications over the three-fourths line if one counts
the states that later rescinded.
305
OLC opined in January 2020 that the ERA
had expired and was no longer pending before the states.
306
But the House
of Representatives passed a resolution the next month eliminating the
deadline, and Senators Ben Cardin and Lisa Murkowski have introduced
a similar measure in the Senate.
307
Meanwhile, a number of lawsuits
regarding the status of the ERA are pending at this writing.
308
There is a credible argument that the ERA is already part of the
Constitution based on Article V. To start, many ocials and scholars have
taken the view that purported rescissions are ineective, given, among
other things, the value of finality and the textual commitment to states of
the power to “ratif[y]” only.
309
Several nineteenth-century treatise writers
advised that rescissions were ineective.
310
And “[e]very state legislature
that passed a resolution rescinding a prior ratification of the ERA did so
under the cloud of an express opinion that such an action would be a legal
nullity.”
311
As for Congress’s attempt to impose a deadline, that too was
arguably invalid. After all, as OLC observed in 1992, “the plain language
of Article V contains no time limit on the ratification process.”
312
It is at
least arguable, then, that Congress has no power to create a ratification
time limit through a resolution that does not itself go through the Article
V process.
313
305. See Timothy Williams, Virginia Approves the E.R.A., Becoming the 38th State to
Back It, N.Y. Times (Jan. 16, 2020), https://www.nytimes.com/2020/01/15/us/era-virginia-
vote.html (on file with the Columbia Law Review).
306. Ratification of the Equal Rights Amendment, 44 Op. O.L.C., slip op. at 1–2, 12–24.
307. See Patricia Sullivan, U.S. House Removes ERA Ratification Deadline, One
Obstacle to Enactment, Wash. Post (Feb. 13, 2020), https://www.washingtonpost.com/local/
legal-issues/us-house-removes-era-ratification-deadline-one-obstacle-to-enactment/2020/
02/13/e82aa802-4de5-11ea-b721-9f4cdc90bc1c_story.html (on file with the Columbia Law
Review).
308. See, e.g., Complaint, Alabama v. Ferriero, No. 7:19-CV-02032-LSC (N.D. Ala. filed
Dec. 17, 2019), 2019 WL 6894418; see also Kentiya Orange, In the Know: Equal Rights
Amendment Lawsuits, League of Women Voters (Feb. 25, 2020), https://www.lwv.org/
blog/know-equal-rights-amendment-lawsuits [https://perma.cc/9BAS-PXJU] (summarizing
lawsuits).
309. See, e.g., Dellinger, Legitimacy of Constitutional Change, supra note 16, at 421–
24; John R. Vile, Permitting States to Rescind Ratifications of Pending Amendments to the
U.S. Constitution, Publius, Spring 1990, at 109, 115–20 (reviewing five “basic arguments . . .
against a state’s right to rescind amendments”). But see Vile, supra, at 113–15 (reviewing
counterarguments).
310. Dellinger, Legitimacy of Constitutional Change, supra note 16, at 423 & n.177.
311. Id. at 423.
312. Cong. Pay Amendment, 16 Op. O.L.C. 85, 88 (1992).
313. See, e.g., Danaya C. Wright, “Great Variety of Relevant Conditions, Political, Social
and Economic”: The Constitutionality of Congressional Deadlines on Amendment
Proposals Under Article V, 28 Wm. & Mary Bill Rts. J. 45, 77–92 (2019); Mason Kalfus,
2370 COLUMBIA LAW REVIEW [Vol. 121:2317
The Supreme Court did state in Dillon that Congress has the power
“to fix a definite period for” ratification.
314
But this statement rested
(again, arguably) on the logically prior conclusion that Article V requires
that “ratification must be within some reasonable time after the pro-
posal.”
315
That conclusion did not survive the Twenty-Seventh
Amendment, at least in the view of Congress and the executive branch.
316
Moreover, even if one believes that Congress has an implied power to con-
trol ratification deadlines and that the ERA’s deadline has expired, it may
be that retroactively extending such a deadline is a permissible exercise of
this power. Not only that, it may be Congress’s exclusive prerogative under
Coleman to judge whether its extension of the deadline is valid.
317
For now,
of course, the ERA does not appear in printed copies of the Constitution.
But the story may not be over, especially given recent shifts in power in
Washington.
M. An Article V Convention?
A brief coda: Our focus has been on the Article V “track” of congres-
sionally initiated amendments, because that is how every amendment in
U.S. history has been passed. There is also the separate and unused track
of amendments proposed by a federal convention. Article V is “strikingly
vague” on how such a convention would be triggered and how, if trig-
gered, it would operate.
318
Over the years, the states have sent hundreds of
Comment, Why Time Limits on the Ratification of Constitutional Amendments Violate
Article V, 66 U. Chi. L. Rev. 437, 446–67 (1999); cf. Clinton v. City of New York, 524 U.S.
417, 445–46 (1998) (holding that Congress cannot change by statute the constitutional pro-
cess for enacting legislation). It is also notable that the ERA’s time limit was worded dier-
ently than prior time limits. The resolution proposing the ERA stated that it would be valid
“when ratified . . . within seven years,” whereas prior limits had stated that the proposed
amendment would be “inoperative unless,” or valid “only if,” ratified within a certain time
period. Suk, We the Women, supra note 300, at 177 (emphasis added). In light of this con-
trast, the ERA’s time limit was arguably just hortatory.
314. Dillon v. Gloss, 256 U.S. 368, 375–76 (1921). It is debatable whether this issue was
properly before the Court in Dillon, given that the Eighteenth Amendment had been ratified
well within the deadline set by Congress. In addition, the deadline was in the text of the
proposed amendment itself, not in the accompanying resolution. Dillon is therefore “dubi-
ous” authority at best for the validity of the ERA’s limit. Robert Hajdu & Bruce E.
Rosenblum, Note, The Process of Constitutional Amendment, 79 Colum. L. Rev. 106, 126
n.75 (1979); see also Virginia v. Ferriero, 525 F. Supp. 3d 36, 57 (D.D.C. 2021) (noting that
whether Congress may impose a deadline by resolution “is a question of first impression”).
315. Dillon, 256 U.S. at 375.
316. See Prakash, Of Synchronicity, supra note 16, at 1226 (“While the Supreme Court
once claimed that the states had to ratify proposed amendments within a reasonable period
of time after their receipt, the political branches later decided otherwise.”).
317. See Daniel Hemel, Some Thoughts on the 28th Amendment, Medium (June 5,
2018), https://medium.com/whatever-source-derived/some-thoughts-on-the-28th-
amendment-fc4d8372ab14 (on file with the Columbia Law Review).
318. Laurence H. Tribe, Issues Raised by Requesting Congress to Call a Constitutional
Convention to Propose a Balanced Budget Amendment, 10 Pac. L.J. 627, 632 (1979); see
also Michael B. Rappaport, Reforming Article V: The Problems Created by the National
2021] PUZZLES AND POSSIBILITIES 2371
applications to Congress for an Article V convention. Some are limited to
a particular subject; some are unrestricted or “general.”
319
Paulsen, after
tabulating all of the applications that had arrived by 1993, concluded that
“there are, at present, forty-five states with their lights ‘on’ for a general
convention.”
320
In other words, “Congress is obliged to call a constitutional
convention and has been for some time.”
321
Paulsen’s conclusion rests on some questionable premises about how
to count state applications. Based on his belief that Article V permits only
unrestricted conventions for proposing amendments, he puts applications
that are limited to a certain subject, but not conditioned on a convention
adopting such a limit, into the “general” column.
322
Even so, the fact that
there is a colorable argument that recent Congresses have been obliged to
call a constitutional convention—all of the rules of which would have to
be invented more or less from scratch—is another vivid illustration of how
little is settled about the law of Article V.
IV.
LIVING WITH ARTICLE V AMBIGUITY: JUDGING CONTESTED
AMENDMENTS
The vast majority of Article V amendments, as Part III shows, have
faced credible challenges to their validity.
323
This Part first explains why
Convention Amendment Method and How to Fix Them, 96 Va. L. Rev. 1509, 1517 (2010)
(“Although the Constitution specifies that an amendment may be drafted by a national con-
vention, it unfortunately does not clearly answer various questions about this amendment
method.”).
319. See Paulsen, A General Theory, supra note 50, at 765–89 (compiling convention
applications as of 1993).
320. Id. at 756.
321. Id.; see also Robert G. Natelson, Counting to Two Thirds: How Close Are We to a
Convention for Proposing Amendments to the Constitution?, 19 Federalist Soc’y Rev. 50,
53–60 (2018) (arguing that Congress is one state application short of being obliged to call
a convention to propose a balanced budget amendment); Walker Hanson, Note, The States’
Power to Eectuate Constitutional Change: Is Congress Currently Required to Convene a
National Convention for the Proposing of Amendments to the United States Constitution?,
9 Geo. J.L. & Pub. Pol’y 245, 258 (2011) (“Even when considering the rescissions of Oregon,
North Dakota, and Wyoming as valid, those three rescissions combined with the rescissions
of the other eight states since 1993 leave a total of thirty-four states with valid applications
before Congress calling for an Article V convention . . . .”). Colorado recently rescinded its
outstanding applications for an Article V convention. H.J. Res. 21-1006 (Colo. Apr. 27,
2021).
322. See Michael Stokes Paulsen, How to Count to Thirty-Four: The Constitutional Case
for a Constitutional Convention, 34 Harv. J.L. & Pub. Pol’y 837, 839–55 (2011) (detailing
his methodology). In this 2011 essay, Paulsen concluded that the number of states with
“lights on” for a general convention had dropped from forty-five to thirty-three—just below
the two-thirds threshold—on account of post-1993 rescissions. Id. at 856–58.
323. More specifically, Part III shows that these challenges were credible in terms of the
interpretive conventions used by mainstream constitutional lawyers. We take no position in
this paper on the ultimate persuasiveness of the challenges in their day or on the appro-
priateness of the conventions. This synoptic approach means that we do not apply any par-
ticular interpretive method in depth; we have not, for example, attempted any sort of
2372 COLUMBIA LAW REVIEW [Vol. 121:2317
such challenges are likely to keep arising, given this history and the extent
of legal uncertainty that persists. It then asks how, as a matter of institu-
tional design, disputes over Article V compliance might best be resolved
in the future.
A. The Highly Incomplete Liquidation of Article V
The standard view of constitutional lawyers today is that at most a
“handful” of amendments, above all the Reconstruction Amendments,
have raised any meaningful legal diculties.
324
And these diculties can
be dismissed as outliers, not only because of their statistical rarity but also
because of the extraordinary upheaval wrought by the Civil War and the
“fluky” circumstances of the Twenty-Seventh Amendment.
325
We hope we
have turned this standard picture upside-down and shown that just about
every recognized amendment suers from a credible defect under Article
V. Even the five relatively modest amendments that did not (as far as we
know) face contemporaneous legal challenges built on questionable past
practices such as quorum rules and the absence of presidential
presentment.
326
The fact that amendments have generated significant legal contro-
versy in the past would not necessarily matter much in the present if, over
time, the law of Article V had become progressively clearer. Many under-
determinate provisions of the Constitution have spawned rich bodies of
case law interpreting their terms or otherwise had their meaning “liqui-
dated” and settled by practice.
327
But for the most part, as Part III illustrates
rigorous investigation into the original public meaning of Article V or into originalist theo-
ries of constitutional construction for resolving Article V underdeterminacies.
324. Prakash, Of Synchronicity, supra note 16, at 1270 n.267.
325. Strauss, Irrelevance of Amendments, supra note 42, at 1486.
326. Presidential presentment did not occur for the Twentieth Amendment or the
Twenty-Third through Twenty-Sixth Amendments. See supra note 21. The House votes on
the Twentieth Amendment, 75 Cong. Rec. 5027 (1932), Twenty-Third Amendment, 106
Cong. Rec. 12,570–71 (1960), and Twenty-Fifth Amendment, 111 Cong. Rec. 15,216 (1965),
and the Senate vote on the Twenty-Third Amendment, 106 Cong. Rec. 12,858 (1960), were
all taken without recording yeas and nays, so it is entirely possible that these amendments
received only two-thirds of a quorum, not two-thirds of the whole membership of each cham-
ber. In the case of the Twenty-Third Amendment, for example, the Congressional Record
reports that, “[i]n the opinion of the Chair, two-thirds of the Senators present and voting . . .
voted in the armative.” Id. at 12,858. The Congressional Record does not indicate how
many senators were “present and voting.” Senator Everett Dirksen had “suggest[ed] the
absence of a quorum” immediately before the vote, but the quorum call was rescinded by
unanimous consent in response to a request from Senator Lyndon Johnson. Id. In the
House, the Twenty-Third Amendment passed by “two-thirds,” but an earlier roll call had
revealed that only 325 members were present. Id. at 12,562, 12,571. Two-thirds of 325 is well
under the 290 members that would constitute two-thirds of the full membership. The lack
of clarity on the final votes for these amendments reflects that the quorum rules were taken
for granted.
327. See generally David A. Strauss, The Living Constitution (2010) [hereinafter
Strauss, The Living Constitution] (describing the large role played throughout U.S. history
2021] PUZZLES AND POSSIBILITIES 2373
in detail, this has not happened with Article V. Recurring contestation over
amendment validity has not yielded anything like a robust body of
congressional, executive, or judicial precedent on the key questions raised
by Article V. A large proportion of the interpretive puzzles that section II.A
identifies—many of which go to the core of the amendment project—have
never been resolved.
The Supreme Court has directly addressed the meaning of Article V
only a handful of times. To recapitulate: In Hollingsworth v. Virginia, Justice
Chase asserted during oral argument that the President “has nothing to
do with the proposition, or adoption, of amendments to the
Constitution.”
328
Over 120 years later,
329
in Hawke v. Smith (No. 1), the
by a “common law approach to constitutional interpretation”); William Baude,
Constitutional Liquidation, 71 Stan. L. Rev. 1 (2019) (describing liquidation); Caleb
Nelson, Originalism and Interpretive Conventions, 70 U. Chi. L. Rev. 519, 547 (2003)
(arguing that the Founding generation “expected subsequent practice to liquidate [consti-
tutional] indeterminacy” and thereby “produce a fixed meaning for the future”).
328. 3 U.S. (3 Dall.) 378, 381 n.* (1798); see also Hawke v. Smith (No. 1), 253 U.S. 221,
229 (1920) (citing only Hollingsworth for the statement that “[a]t an early day this court
settled that the submission of a constitutional amendment did not require the action of the
President”); supra notes 117–130 and accompanying text.
329. We are not aware of any Supreme Court decisions about the amendment process
in the long interval between Hollingworth and Hawke. In Luther v. Borden, 48 U.S. (7 How.)
1 (1849), which concerned the validity of a new purported state constitution in Rhode
Island, the Court suggested in dicta that questions regarding the amendment process may
be nonjusticiable. Id. at 53. In White v. Hart, 80 U.S. (13 Wall.) 646 (1872), the Court
likewise suggested that the validity of the Reconstruction Amendments was nonjusticiable.
Id. at 649. In Myers v. Anderson, 238 U.S. 368 (1915), which concerned the constitutionality
of Maryland’s “Grandfather Clause,” several parties raised constitutional objections to the
Fifteenth Amendment. The attorney defending the state’s law—named, oddly enough,
William Marbury—argued that the Fifteenth Amendment was unconstitutional as applied
to Maryland under the Equal Surage Clause, because Maryland had never ratified the
amendment. In openly racist terms, Marbury submitted that “compelling” states to expand
the franchise to non-whites “would be in substance and eect depriving the original
State . . . of all representation in the Senate.” Vose, Constitutional Change, supra note 229,
at 39 (quoting Brief for Plainti in Error, Myers v. Anderson, 238 U.S. 368 (1915)). An
amicus brief also rehearsed procedural objections to the passage of the Fourteenth and
Fifteenth Amendments. Id. at 40–41. The Court did not address these arguments in its
unanimous opinion in Myers, nor in another case decided the same day about Oklahoma’s
Grandfather Clause, Guinn v. United States, 238 U.S. 347 (1915). There is an intriguing
hint, however, that the arguments may have gotten some internal traction at the Court.
Marbury’s son later recounted in a letter to Justice Felix Frankfurter that Justice James Clark
McReynolds—an “intimate” of Marbury who joined the Court while Myers was pending—
told Marbury that “the Fifteenth Amendment election cases had been the subject of a
terrific controversy.” Benno C. Schmidt, Jr., Principle and Prejudice: The Supreme Court
and Race in the Progressive Era, Part 3: Black Disfranchisement From the KKK to the
Grandfather Clause, 82 Colum. L. Rev. 835, 867 (1982) (quoting Letter from William L.
Marbury, Jr. to Felix Frankfurter (Sept. 25, 1958)). Marbury’s son went on: “Apparently,
Justice Lurton had prepared a dissenting opinion in which he followed my father’s
argument in the Myers case. This so scandalized the Chief Justice that he suggested that
Lurton resign. When Lurton refused to do this the majority of the court held up a decision
until his death.” Id. Justice Frankfurter passed the letter on to Alexander Bickel, though
2374 COLUMBIA LAW REVIEW [Vol. 121:2317
Court held in a “shocking[]”
330
decision that a state could not require a
proposed amendment to be submitted to a popular referendum after
being ratified by the state legislature.
331
In the National Prohibition Cases,
the Court concluded, in an unreasoned summary opinion, that Congress
does not need to declare that an amendment is “necessary” when it pro-
poses an amendment; that the two-thirds vote in each chamber “is a vote
of two-thirds of the members present—assuming the presence of a
quorum”; and that the Prohibition Amendment was “within the power to
amend reserved by Article V of the Constitution.”
332
In Dillon v. Gloss, the
Court stated in dicta that ratification of an amendment “must be within
some reasonable time after its proposal” and that Congress could fix “a
definite period for ratification . . . as an incident of its power to designate
the mode of ratification.”
333
In Leser v. Garnett, the Court held that a state
could not place limits on its legislature’s power to ratify a federal amend-
ment in its state constitution and that “ocial notice” of ratification from
a state to the U.S. Secretary of State “was conclusive upon him” and “upon
the courts.”
334
In United States v. Sprague, the Court held that, as between
state legislatures and state conventions, the “choice . . . of the mode of rat-
ification, lies in the sole discretion of Congress.”
335
Finally, in Coleman v.
Miller,
336
which the next section discusses in more detail, a splintered
Court determined that “the eect both of previous rejection and of
noting his “heavy doubts about its accuracy.” Id. (quoting Letter from Felix Frankfurter to
Alexander Bickel (Sept. 23, 1959)).
330. The Amendment Stands, N.Y. Times, June 3, 1920, at 10.
331. 253 U.S. 221, 231 (1920); see supra notes 202–208 and accompanying text. In
Arizona State Legislature v. Arizona Independent Redistricting Commission, 576 U.S. 787
(2015), the Court suggested in dicta that Hawke extended to the governor’s role in the
amendment process. Id. at 808 (“In the context of ratifying constitutional amendments, in
contrast, ‘the Legislature’ has a dierent identity, one that excludes the referendum and
the Governor’s veto.”). On the other hand, Justice William Rehnquist, in a solo opinion
denying a stay, wrote that the question whether the word “Legislatures” in Article V
“encompasses the voters of a State who have power to enact laws by initiative” is “by no
means settled.” Uhler v. AFL-CIO, 468 U.S. 1310, 1311 (1984) (Rehnquist, J., in chambers)
(discussing the meaning of the word “Legislatures” in the Application Clause).
332. 253 U.S. 350, 386 (1920); see supra notes 210–211 and accompanying text.
333. 256 U.S. 368, 375–76 (1921); see supra notes 196–198 and accompanying text. On
why this statement is dicta, see Cong. Pay Amendment, 16 Op. O.L.C. 85, 92–93 (1992).
334. 258 U.S. 130, 136–37 (1922); see supra notes 214–225 and accompanying text. The
Court also rejected the argument that the Nineteenth Amendment was invalid because it
made “so great an addition to the electorate . . . without the State’s consent.” Leser, 258 U.S.
at 136. The Court did not explain its reasoning, other than to say that the argument would
also render the Fifteenth Amendment invalid. Id.
335. 282 U.S. 716, 730 (1931); see supra notes 199–201 and accompanying text.
336. 307 U.S. 433 (1939). Coleman concerned the ratification of an amendment pro-
posed by Congress in 1924 that would have empowered Congress to regulate “the labor of
persons under eighteen years of age.” Id. at 435 n.1 (quoting Child Labor Amendment, 43
Stat. 670 (1924)). The proposed amendment fell short of ratification. See John R. Vile,
Child Labor Amendment, in Encyclopedia of Constitutional Amendments, Proposed
Amendments, and Amending Issues, 1789–1995, at 47, 48–49 (1996).
2021] PUZZLES AND POSSIBILITIES 2375
attempted withdrawal” and “whether [an] amendment had been adopted
within a reasonable time” were political questions, “not subject to review
by courts.”
337
These opinions cover relevant ground, but not all that much
ground—only a fraction of the landscape of interpretive puzzles canvassed
in section II.A. And the extent to which the Court has settled even the
questions it has purported to answer is far from clear. Nearly half of the
above-listed statements on Article V are dicta, unreasoned ipse dixits, or
both. Few if any of the opinions are widely respected by constitutional
lawyers, and questions such as the President’s role in the amendment
process and the meaning of Hollingsworth remain the subject of debate, at
least in law reviews.
338
Several opinions have been blatantly contradicted
by subsequent practice: The Twenty-Second Amendment was approved by
the House without “a vote of two-thirds of the members present,” as was
said to be necessary in the National Prohibition Cases;
339
the Twenty-Seventh
Amendment mocks the Dillon Court’s contention that it was “quite
untenable” to think an amendment proposed in 1789 was still pending in
1921;
340
and states have continued to experiment with direct democracy in
the ratification process, Leser notwithstanding.
341
Among the
uncontradicted opinions, most just label certain issues as political or
nonjusticiable, casting no light on their merits.
342
337. 307 U.S. at 449, 454 (opinion of Hughes, C.J.). The quotations are taken from
Chief Justice Hughes’s “Opinion of the Court,” which was joined by Justices Harlan Fiske
Stone and Stanley Forman Reed. Id. at 435. Justice Hugo Black, writing for himself and
three other Justices, would have ruled that “Congress has sole and complete control over
the amending process, subject to no judicial review.” Id. at 459 (Black, J., concurring). The
Court subsequently confirmed that in Coleman it had “held that the questions of how long
a proposed amendment to the Federal Constitution remained open to ratification, and what
eect a prior rejection had on a subsequent ratification, were committed to congressional
resolution and involved criteria of decision that necessarily escaped the judicial grasp.”
Baker v. Carr, 369 U.S. 186, 214 (1962). OLC has questioned the authoritativeness of both
Coleman opinions. Cong. Pay Amendment, 16 Op. O.L.C. 85, 93 n.11, 99–102 (1992).
338. See generally, e.g., Black, On Article I, supra note 21; Tillman, A Textualist
Defense, supra note 119; Joshi, supra note 130; Mike Rappaport, Does the Constitution
Require Constitutional Amendments to Be Presented to the President?, Law & Liberty
(Jan. 17, 2013), https://lawliberty.org/more-on-lincoln-signing-the-13th-amendment
[https://perma.cc/V3BQ-LAVY].
339. 253 U.S. 350, 386 (1920) (citing Mo. Pac. Ry. Co. v. Kansas, 248 U.S. 276 (1919));
see supra section III.J.
340. Dillon v. Gloss, 256 U.S. 368, 375 (1921); see supra section III.K.
341. See, e.g., Uhler v. AFL-CIO, 468 U.S. 1310, 1311–12 (1984) (Rehnquist, J., in cham-
bers); Kimble v. Swackhamer, 439 U.S. 1385, 1385 (1978) (Rehnquist, J., in chambers).
342. The case law in the lower federal courts is similarly sparse. And given the de facto
repudiation of Article V principles announced by the Supreme Court, one would expect any
lower court precedent to be even less stable. The two most significant lower court cases both
involve the ERA. The first is Dyer v. Blair, 390 F. Supp. 1291 (N.D. Ill. 1975) (three-judge
court), an opinion authored by then-Judge John Paul Stevens about ratification of the ERA
in Illinois. The court held that state legislatures were free to set their own voting and
quorum rules for ratifying federal constitutional amendments, but that a state constitution
2376 COLUMBIA LAW REVIEW [Vol. 121:2317
It would be fair, of course, to say that some questions have been
resolved by a combination of Article V’s text, judicial doctrine, and histor-
ical tradition. No amendment would be rejected at this point for lack of
presidential presentment (although the rise of textualism and originalism
as interpretive methods may make the practice of nonpresentment
increasingly awkward).
343
Nor would an amendment be rejected for receiv-
ing only two-thirds of a quorum of both houses, rather than two-thirds of
the entirety of both houses (although it is conceivable that a member of
Congress could persuade colleagues with such an objection).
344
But few
and far between are the Article V practices that have been “open, wide-
spread, and unchallenged since the early days of the Republic,”
345
and it
seems to us that the unsettled questions substantially exceed the settled
ones in both number and significance.
Perhaps the simplest response to anyone who nonetheless maintains
that history has resolved most of the important puzzles in Article V is to
point to the two most prominent recent amendment eorts and the legal
tumult each has occasioned. While the controversies over the Twenty-
Seventh Amendment and the ERA have idiosyncratic elements, the proce-
dural issues they raise are strikingly basic.
346
They highlight just how little
the law of Article V has developed since the Founding. If controversy over
the Twenty-Seventh Amendment has quieted a bit in recent years, this
probably has more to do with the amendment’s relative unimportance and
the unlikelihood that Congress would ever contravene it than with consid-
ered acceptance of its validity.
347
* * *
In sum, the Supreme Court’s unanimous pronouncement that a
“mere reading demonstrates” that “Article V is clear in statement and in
could not specify a supermajority requirement that would bind a state legislature acting on
such an amendment. Id. at 1306–08. The second is a district court decision from Idaho
holding that states could rescind prior ratifications and that Congress’s attempt to extend
the ERA ratification deadline was invalid. Idaho v. Freeman, 529 F. Supp. 1107, 1150–54 (D.
Idaho 1981), vacated as moot sub nom. Carmen v. Idaho, 459 U.S. 809 (1982). The Supreme
Court granted certiorari before judgment, but the new ratification deadline expired while
the case was pending, so the Court vacated the district court’s decision as moot.
343. See supra section III.B; supra note 21 and accompanying text (discussing the view
that nonpresentment is incompatible with clear constitutional text).
344. See supra section III.C (discussing the plausibility of the view that a two-thirds vote
of the full chamber is required).
345. NLRB v. Noel Canning, 573 U.S. 513, 572 (2014) (Scalia, J., concurring in the
judgment) (describing the sorts of practices that “should guide [the Court’s] interpretation
of an ambiguous constitutional provision”).
346. See supra sections III.K, III.L.3.
347. See Strauss, Irrelevance of Amendments, supra note 42, at 1486–87 (describing the
Twenty-Seventh Amendment as having “no significant eect” and “remarkable for [its] rel-
ative lack of importance”).
2021] PUZZLES AND POSSIBILITIES 2377
meaning, contains no ambiguity, and calls for no resort to rules of con-
struction”
348
seems to us exactly backward. As Part II illustrates, a mere
reading demonstrates that Article V is shot through with ambiguities and
calls for constant resort to rules of construction. As Part III shows, this
ambiguity is borne out by, and persists through, 230 years of history.
Americans tend to take the text of the Constitution as a given and to regard
constitutional interpretation as something that operates on that text. But
if our descriptive arguments thus far have been sound, then there is no
constitutional text that precedes interpretation. To identify what is or is
not “in” the constitutional text is itself a complex act of constitutional
interpretation. And given the amount of legal uncertainty that continues
to enshroud the Article V process, one can expect all but the most uncon-
troversial new amendments to elicit credible challenges to their validity in
the years ahead.
B. Revisiting Coleman
If we are right that controversy over the validity of amendments is
likely to persist for the foreseeable future, then it behooves us to reflect on
our system for resolving such disputes. When some groups in the polity
insist that an attempted amendment has satisfied the rules of Article V and
others disagree, with both sides advancing credible legal arguments, who
should decide whether the amendment has become part of the
Constitution? We submit that the degree of ongoing Article V ambiguity
documented in this paper lends new support to the Coleman Court’s
inclination to leave such matters to Congress, while also suggesting possi-
ble institutional innovations.
In Coleman v. Miller, a splintered Supreme Court determined in 1939
that Congress has the ultimate authority to promulgate or proclaim an
amendment after ratification by the states.
349
Chief Justice Hughes, writing
for himself and Justice Harlan Fiske Stone in what was styled the “Opinion
of the Court,” found that Congress has “control over the promulgation of
the adoption of [an] amendment.”
350
He further found that the two ques-
tions before the Court—the eect of a state’s prior rejection of an amend-
ment on a subsequent ratification, and whether too long a period had
elapsed between proposal and ratification—were political questions not fit
for judicial review.
351
Justice Hugo Black penned a concurrence, joined by
Justices Felix Frankfurter, Owen Roberts, and William O. Douglas. He
similarly suggested that it was Congress’s role to “proclaim[]” an
amendment part of the Constitution, and he would have held that all
348. United States v. Sprague, 282 U.S. 716, 730 (1931).
349. 307 U.S. 433, 450 (1939).
350. Id.
351. Id. at 450–54.
2378 COLUMBIA LAW REVIEW [Vol. 121:2317
questions related to the amendment process were “political” and beyond
the purview of the courts.
352
Coleman’s reference to a congressional “promulgation” or “proclama-
tion” power has been criticized on the ground that Article V makes no
reference to any role for Congress after the initial proposal of an amend-
ment. The text instructs that an amendment is “valid . . . when ratified by
the Legislatures of three fourths of the several States,” not when promul-
gated by Congress following the final state ratification.
353
And the Supreme
Court has in fact adjudicated a number of disputes about the amendment
process, albeit infrequently, going all the way back to Hollingsworth,
354
which predates Marbury.
355
These criticisms of Coleman have some force, but they slight a pruden-
tial rationale for the Court’s ruling that the Court itself did not articulate
but that this paper’s historical account helps to reinforce.
356
On account
of the underdeterminacy of Article V and the momentousness of formal
constitutional change, there is a strong possibility that significant conflicts
over the validity of amendments will arise and yield no clear answers, only
“political questions.”
357
Partly for this reason, the Court has avoided reach-
ing the merits of an Article V dispute for over ninety years.
358
There is a
consequent interest in enabling some nonjudicial institution to resolve
authoritatively whether the Constitution has been amended.
The ongoing saga of the ERA exemplifies this concern. At least three
pending lawsuits allege, not at all frivolously, that the ERA is already part
of the Constitution.
359
Meanwhile, proponents of the ERA are pressing
Congress to retroactively extend the ratification deadline, which, if done,
352. Id. at 457–58 (Black, J., concurring).
353. U.S. Const. amend. V.
354. Hollingsworth v. Virginia, 3 U.S. (3 Dall.) 378 (1798); see also supra notes 328–337
and accompanying text (cataloging cases).
355. Marbury v. Madison, 5 U.S. (1 Cranch) 137 (1803). For prominent criticisms of
Coleman along these lines, see Cong. Pay Amendment, 16 Op. O.L.C. 85, 99–105 (1992);
Dellinger, Legitimacy of Constitutional Change, supra note 16, at 397–405; Paulsen, A
General Theory, supra note 50, at 709–18.
356. Cf. Pozen & Samaha, supra note 69, at 732–33, 775–76 (explaining that “pruden-
tial” arguments in constitutional law tend to emphasize system-level considerations of
administrability, workability, and the like).
357. See Baker v. Carr, 369 U.S. 186, 217 (1962) (listing factors that indicate the pres-
ence of a “political question”). In some Article V disputes, not just one or two but all six of
the Baker factors may arguably be triggered. See Thomas E. Baker, Towards a “More Perfect
Union”: Some Thoughts on Amending the Constitution, 10 Widener J. Pub. L. 1, 7–8 (2000)
(“The Article V amendment power bears all the constitutional hallmarks of a nonjusticiable
or political question . . . .”); Thomas Millet, The Supreme Court, Political Questions, and
Article V—A Case for Judicial Restraint, 23 Santa Clara L. Rev. 745, 763–66 (1983)
(analyzing the applicability of the Baker factors to Article V cases).
358. The last such ruling was United States v. Sprague, 282 U.S. 716 (1931), issued in
February 1931. See supra note 335 and accompanying text.
359. See supra note 308 and accompanying text.
2021] PUZZLES AND POSSIBILITIES 2379
would immediately raise a dicult new Article V question.
360
No one can
predict with any confidence whether or how the constitutional fate of the
ERA will be determined, as all three branches of government continue to
vie for interpretive supremacy with regard to Article V
361
and many lawyers
now flatly deny the precedential value of Coleman.
362
The status quo of second-order uncertainty about how to resolve first-
order Article V uncertainty does not necessarily reflect any sort of crisis in
self-government or the rule of law. Policentric decisionmaking procedures
can generate deliberative and participatory benefits as well as stable polit-
ical equilibria under certain conditions, including on questions of consti-
tutional interpretation.
363
In contrast to the “juricentrism” that
characterizes much of contemporary constitutional culture,
364
a more fluid
form of departmentalism has prevailed in the Article V context through-
out U.S. history. We could keep muddling through.
Yet even if the status quo is defensible, it carries significant costs in
terms of predictability, eciency, and popular responsiveness. The ERA
example is, again, instructive, as countless hours of legal and political
mobilization have been devoted not only to the question whether we
should have an ERA but also to the question whether we do have an ERA—
with no clear end to the legal wrangling in sight and the very real possibil-
360. See supra note 307 and accompanying text. On whether Article V allows Congress
to retroactively extend the ERA’s deadline, compare Held et al., supra note 256, at 128–31
(yes), and Gerard N. Magliocca, Buried Alive: The Reboot of the Equal Rights Amendment,
71 Rutgers U. L. Rev. 633, 644–50 (2019) (yes), with Idaho v. Freeman, 529 F. Supp. 1107,
1150–54 (D. Idaho 1981) (no), vacated sub nom. Carmen v. Idaho, 459 U.S. 809 (1982), and
Ratification of the Equal Rights Amendment, 44 Op. O.L.C., slip op. at 24–37 (Jan. 6, 2020)
(no).
361. See supra notes 265–272 and accompanying text.
362. See Kalfus, supra note 313, at 445 (“Many scholars have concluded that the Court’s
reformulation of political question analysis, coupled with Coleman’s conflict with Court
precedent, render Coleman ‘dead.’”).
363. See, e.g., Neal Devins & Louis Fisher, Judicial Exclusivity and Political Instability,
84 Va. L. Rev. 83, 103 (1998) (arguing that “judicial exclusivity” over constitutional deci-
sionmaking “encourages acrimony, not cooperation”); Robert C. Post & Reva B. Siegel,
Legislative Constitutionalism and Section Five Power: Policentric Interpretation of the
Family and Medical Leave Act, 112 Yale L.J. 1943, 2023–37 (2003) [hereinafter Post & Siegel,
Legislative Constitutionalism] (defending a model of “policentric constitutional interpreta-
tion” for Section 5 of the Fourteenth Amendment). The term “policentric” refers to “the
distribution of constitutional interpretation . . . across multiple institutions, many of which
are political in character,” and is distinct from the term “polycentric,” which describes
“problems that contain multiple and interdependent solutions.” Post & Siegel, Legislative
Constitutionalism, supra, at 2022–23.
364. See David E. Pozen, Judicial Elections as Popular Constitutionalism, 110 Colum. L.
Rev. 2047, 2055–56 (2010); see also Nikolas Bowie & Daphna Renan, The Separation-of-
Powers Counterrevolution, 131 Yale L.J. (forthcoming 2022) (manuscript at 22–46) (on file
with the Columbia Law Review) (discussing the “juristocratic” understanding of separation
of powers disputes that has prevailed since Reconstruction).
2380 COLUMBIA LAW REVIEW [Vol. 121:2317
ity that the ERA will fail without creating any precedent on Article V out-
side the executive branch.
365
Most worrisome, in a nation with one of the
most demanding amendment criteria and lowest formal amendment rates
in the world,
366
the absence of any clear Article V dispute-resolution mech-
anism makes it harder to revise the constitutional text by perpetuating
legal uncertainty and proliferating institutional veto points. Even after
supermajorities of both chambers of Congress and the state legislatures
have approved a change to this text, legal objections from any one of the
federal legislative, executive, or judicial branches may be enough to derail
an attempted amendment. It is for these sorts of reasons that many theo-
rists of constitutional design, from Madison in 1787 to the Venice
Commission in 2009, have advised that confusion regarding the amend-
ment rules themselves “ought to be as much as possible avoided.”
367
If, in line with this view, one of the existing organs of government
ought to be assigned primary responsibility for resolving disputes over
amendment validity, Congress seems best suited to the task.
368
The basic
reasons are straightforward. Unlike the judiciary, Congress is not bound
by a case or controversy limitation. It can take years for a case to wend its
way up to the Supreme Court, and many amendments do not give rise to
365. Within the executive branch, OLC recently issued a substantial opinion concluding
that the ERA is dead. See Ratification of the Equal Rights Amendment, 44 Op. O.L.C. (Jan.
6, 2020) (slip opinion). The opinion immediately drew fierce criticism and pledges of defi-
ance. See, e.g., Garrett Epps, Bill Barr Doesn’t Get to Decide What’s in the Constitution,
Atlantic (Jan. 16, 2020), https://www.theatlantic.com/ideas/archive/2020/01/trump-doj-
era-constitution/605047/ (on file with the Columbia Law Review) (noting that ERA advocates
“reject [OLC’s] arguments” and quoting Reva Siegel for the view that “[t]he Justice
Department does not have authority to decide the legal validity of the ERA”); Press Release,
Rep. Carolyn B. Maloney, Maloney Statement on OLC Response to Archivist on ERA
Ratification (Jan. 8, 2020), https://maloney.house.gov/media-center/press-releases/
maloney-statement-on-olc-response-to-archivist-on-era-ratification
[https://perma.cc/72WA-UQ96] (“I do not believe that the OLC has the final word to
dictate how Congress or the states proceed in amending the Constitution.”).
366. See Richard Albert, American Exceptionalism in Constitutional Amendment, 69
Ark. L. Rev. 217, 225–31 (2016) (reviewing comparative evidence and explaining that,
among democracies, “the United States ranks in the three lowest average annual revision
rates”); Donald S. Lutz, Toward a Theory of Constitutional Amendment, in Responding to
Imperfection, supra note 1, at 237, 260–61 (finding that the U.S. Constitution has the
world’s “second most dicult amendment process,” behind only the now-defunct Yugoslav
Constitution).
367. 2 The Records of the Federal Convention of 1787, supra note 89, at 630 (James
Madison); see also supra notes 89–90 and accompanying text.
368. To say that the validity of contested amendments is a “political question” entrusted
primarily to Congress is not necessarily to deny any role for the judiciary. It may be that
judicial review is appropriate for “plainly ultra vires action”—for instance, if Congress were
to assert that thirty-five was three-fourths of fifty. Richard H. Fallon Jr., Political Questions
and the Ultra Vires Conundrum, 87 U. Chi. L. Rev. 1481, 1490 n.36 (2020) (citing Laurence
H. Tribe, A Constitution We Are Amending: In Defense of a Restrained Judicial Role, 97
Harv. L. Rev. 433, 433 (1983)). In addition, the prudential argument for congressional pri-
macy may be weaker in the case of an amendment proposed by an Article V convention or
an amendment that reforms Congress itself.
2021] PUZZLES AND POSSIBILITIES 2381
justiciable controversies at all. Should an Article V controversy be enter-
tained, the Court’s precedents are sparse and not well ordered in this area.
Congress, on the other hand, can act promptly to assess the validity of an
amendment as soon as the final state has purportedly ratified. More
important, questions about the Article V process will often involve highly
charged, legally underdetermined judgments, which risk taxing the
Court’s institutional competence and compromising its institutional
clout.
369
These separation of powers concerns become all the more acute
when Article V has been activated in order to overturn the Court’s own
constitutional rulings.
370
The sociological legitimacy of amendments, this
paper shows, has never had a tight relationship with their procedural
propriety.
371
As the most geographically representative, deliberatively
transparent, and electorally accountable branch, Congress will in general
be best positioned to determine whether an amendment has gained broad
social acceptance and to generate additional political support once such a
determination has been made.
372
Relative to the Court, the executive branch also has stronger demo-
cratic credentials and the ability to act with greater dispatch. But the text
of Article V makes no mention of the executive. And one of the very few
propositions of Article V law that historical practice has settled is that the
President has no legal role in proposing an amendment to the states.
373
In
light of this practice, which dates back to the Bill of Rights, it is odd to
369. Cf. Tara Leigh Grove, The Supreme Court’s Legitimacy Dilemma, 132 Harv. L.
Rev. 2240, 2250–54 (2019) (book review) (explaining how politically divisive rulings can
undermine the Court’s sociological legitimacy). Of course, the Court could bring more
determinacy to the law of Article V by weighing in on more Article V questions. But the
history of widespread disapproval and even defiance of the Court’s pronouncements on
Article V, see supra notes 338–341 and accompanying text, gives reason to doubt that the
Court could ever definitively clear up this body of law when motivated majorities object to
its conclusions.
370. See Laurence H. Tribe, A Constitution We Are Amending: In Defense of a
Restrained Judicial Role, 97 Harv. L. Rev. 433, 444–45 (1983).
371. See supra Part III.
372. See Neal Kumar Katyal, Legislative Constitutional Interpretation, 50 Duke L.J.
1335, 1363 (2001) (“Society needs a democratic mandate rather than a judicial one for some
decisions . . . . Impeachment, war powers, and the decision whether a constitutional amend-
ment has been ratified are some examples.”); Post & Siegel, Legislative Constitutionalism,
supra note 363, at 2030 (“If the Court has particular strengths in explicating the
Constitution as a rule of law, Congress is especially well-situated to respond to changes in
constitutional culture.”); see also Larry Alexander & Frederick Schauer, On Extrajudicial
Constitutional Interpretation, 110 Harv. L. Rev. 1359, 1377 n.80 (1997) (explaining that “a
single national legislature” like the U.S. Congress may be best positioned to fulfill an
“authoritative settlement function” for certain constitutional issues, including political
questions). Insofar as Congress, in resolving specific Article V disputes, is more likely than
the Court to leave some play in the joints of Article V law, we believe this is a virtue rather
than a vice. See infra section V.C.
373. See supra sections III.A–.B.
2382 COLUMBIA LAW REVIEW [Vol. 121:2317
think that the President would play any sort of significant role at the con-
clusion of the Article V process.
374
It is odder still to think that such a role
would be played by subordinate executive ocers such as the Archivist of
the United States, who currently has the statutory duty to “publish[]” an
amendment upon receiving “ocial notice” that it has been adopted.
375
That said, Congress is less than ideal as an arbiter of Article V validity
in a number of respects. Rising levels of polarization within each chamber
increase the risk of raw partisan conflict, gamesmanship, and perceptions
of bias. Congress’s popular approval rating currently sits near a historic
low.
376
And in some instances, a disputed amendment might aect
Congress itself in ways that call into question its members’ ability to assess
the amendment’s constitutional status in good faith, as when the
Seventeenth Amendment fundamentally changed the Senate by introduc-
ing direct election of senators. This section has suggested that the Coleman
Court was correct to conclude that, for settling whether an amendment
has satisfied Article V, Congress will typically be the least bad option
among the branches of government. Might there be any other options?
There are no silver bullets, but potentially useful subconstitutional
moves can be made within the Coleman framework. One possibility is a
national referendum. Beginning with the Eighteenth Amendment’s time
limit,
377
Congress has on several occasions built a ratification condition
into the text of an amendment. In a similar spirit, Congress could include
a provision in future proposed amendments stating that they will be inop-
erative unless validated in a certain sort of post-ratification referendum,
thereby eectively precommitting Congress to “proclaim” validated
amendments despite any procedural objections that might arise along the
374. One could perhaps argue, on some sort of division of power grounds, that the
exclusion of the President from the “proposal” stage makes it all the more appropriate that
the President be given a decisive role at the “promulgation” stage. Cf. Black, On Article I,
supra note 21, at 899 (maintaining that Hollingsworth should not be extended “one inch”).
But it would create a textually as well as functionally bizarre asymmetry to read Articles I and
V to exclude presidential presentment on the front end of the amendment process while
requiring it on the back end. And given the extraordinary amount of national consensus
already required by the amendment process, there are powerful democratic reasons not to
introduce yet another hard veto point.
375. See supra note 265 and accompanying text. This duty used to be performed by the
Secretary of State. See Act of Apr. 20, 1818, ch. 80, § 2, 3 Stat. 439; Bernstein, supra note 25,
at 540 n.218 (explaining that Congress transferred responsibility for certifying amendments
from the Secretary of State to the Administrator of General Services in 1951 and then to the
Archivist in 1984). We have not found any evidence suggesting that Congress intended to
delegate to the Archivist the authority to resolve the status of contested amendments. It is
perfectly consistent with the statutory text to regard a congressional proclamation of an
amendment’s ratification as the relevant “ocial notice” that triggers the Archivist’s duty
of publication.
376. See Harry Enten, Congress’ Approval Rating Hasn’t Hit 30% in 10 Years. That’s a
Record., CNN (June 1, 2019), https://www.cnn.com/2019/06/01/politics/poll-of-the-
week-congress-approval-rating [https://perma.cc/QP8P-6BR7].
377. See supra notes 196–198 and accompanying text.
2021] PUZZLES AND POSSIBILITIES 2383
way.
378
Alternatively, Congress could reserve for itself or for each chamber
separately the authority, through a simple majority vote, to call for a
national referendum on an amendment’s validity after the amendment
has—in the view of the members calling for the referendum—attained a
sucient number of ratifications.
379
Another possibility is for Congress to
convene, either through the ordinary legislative process or through the
Article V mechanism just described, a special commission to issue a non-
binding opinion on the validity of a disputed amendment and the steps
that would be required, if any, to cure legal defects. Congress could form
such a commission today to help it assess the status of the ERA and build
bipartisan buy-in for whatever position it ultimately adopts.
380
The above discussion provides only a skeletal sketch of these options,
which would take a whole other paper to elaborate in full. The key point,
for present purposes, is that ongoing confusion over how to decide Article
V disputes creates an opportunity to innovate in limited, pragmatic ways
that honor Coleman’s prudential wisdom while moderating some of the
risks of relying on Congress. By enlisting the assistance of a referendum or
commission, Congress can remain in the driver’s seat when it comes to
judging Article V amendments without necessarily serving as the exclusive
or even the final decisionmaking body.
378. See Amar, America’s Constitution, supra note 49, at 418. Amar, recall, has argued
that a national referendum may be used to bypass Article V altogether. See supra notes 4, 7,
58 and accompanying text. The proposal here could be seen as a hybrid between Amar’s
much bolder proposal and the existing Article V process. Any such referendum would add
another step, and another eective veto point, to what is already an arduous amendment
process. But we expect that a referendum would be very unlikely to fail if the amendment
in question had already plausibly navigated the hurdles of Article V—and that a successful
referendum would put immense, productive pressure on Congress to accept the result forth-
with. Consider, by way of analogy, the pressure that the Brexit referendum exerted on the
U.K. Parliament. See Meg Russell, Brexit and Parliament: The Anatomy of a Perfect Storm,
74 Parliamentary As. 443, 448 (2021) (“[T]he clear reservations of many parliamentarians
about the Brexit decision were overshadowed by the referendum result, with most MPs on
both the government and opposition side accepting that it must be respected. Parliament
hence actively, if reluctantly, ceded its sovereignty to the public on the principle of Brexit.”).
In moving to submit a potential amendment directly to the American people for an advisory
referendum, one U.S. Senator argued to his colleagues in 1861 that the people’s “senti-
ments and their opinions will be our safest guide upon this question, . . . disabled as we are
by our own distractions and divisions in Congress from acting upon it.” Cong. Globe, 36th
Cong., 2d Sess. 264 (1861) (statement of Sen. Crittenden).
379. This proposal might be challenged on Chadha grounds, see generally INS v.
Chadha, 462 U.S. 919 (1983), but such a challenge would be undercut by building the
option to call a referendum into the text of the amendment itself. By doing this, the
referendum would already plausibly have navigated the Article V process by the time it is
initiated. See Amar, America’s Constitution, supra note 49, at 418.
380. Such a commission, in our view, would ideally be instructed by Congress to apply
some version of the Article V Thayerianism proposed infra section V.C.
2384 COLUMBIA LAW REVIEW [Vol. 121:2317
V.
EMBRACING ARTICLE V AMBIGUITY: LESSONS FOR INTERPRETERS AND
REFORMERS
The dispelling of any notion that Article V supplies a straightforward
guide to amendment holds lessons for academics and advocates as well as
government ocials. That descriptive undertaking is, again, the heart of
this paper, and space constraints require us to be brief in this final Part.
Here, we first argue that our account of Article V ambiguity complicates
two prominent debates in constitutional theory.
381
We then close by
considering how it could open up new possibilities for constitutional
change.
A. Originalism and Textualism
One of the most sweeping developments in constitutional law over the
past several decades, as countless commentators have observed, has been
the rise of originalism and its close cousin textualism as a preferred mode
of interpretation.
382
With the confirmation of Justice Amy Coney Barrett,
there are now multiple “self-avowed originalists” on the Court.
383
Scholarly
interest in originalism shows no signs of abating.
Originalism is a complex phenomenon, more a cluster of related
methodologies than a single, well-defined one.
384
But an important nor-
mative justification for all its variants has been the idea that it is the inter-
pretive approach most consistent with a commitment to popular
381. Our account also has potential implications for more general debates in legal the-
ory. For instance, scholars “typically assume that bright-line rules are more constraining on
judicial and administrative decisionmakers than context-saturated standards.” Connor N.
Raso & William N. Eskridge, Jr., Chevron as a Canon, Not a Precedent: An Empirical Study
of What Motivates Justices in Agency Deference Cases, 110 Colum. L. Rev. 1727, 1812
(2010). The Article V ambiguity and flexibility documented in these pages challenge that
assumption. As a formal matter, the key terms in Article V are much more rule-like than
standard-like: There are no references in the text to reasonableness, equity, all-things-
considered balancing, or anything of the sort. It is partly for this reason, we suspect, that
Article V is believed by many to be more constraining than it really is.
382. See generally Jeremy K. Kessler & David E. Pozen, Working Themselves Impure: A
Life Cycle Theory of Legal Theories, 83 U. Chi. L. Rev. 1819, 1844–47 & n.63 (2016) (sum-
marizing originalism’s academic and political development since the 1970s and collecting
citations to intellectual histories).
383. Ilan Wurman, What Is Originalism? Debunking the Myths, Conversation (Oct. 24,
2020), https://theconversation.com/what-is-originalism-debunking-the-myths-148488
[https://perma.cc/R7CD-W679] (listing Justices Amy Coney Barrett, Neil Gorsuch,
Clarence Thomas, and—more controversially—Brett Kavanaugh in this category).
384. See, e.g., Lawrence B. Solum, Cooley’s Constitutional Limitations and
Constitutional Originalism, 18 Geo. J.L. & Pub. Pol’y 49, 54 (2020) (describing originalism
as “a family of constitutional theories” united by certain ideas about the fixity and constrain-
ing force of the constitutional text).
2021] PUZZLES AND POSSIBILITIES 2385
sovereignty.
385
The original meaning of the constitutional text, on this
view, “is both binding and uniquely legitimate” because of the way the text
was ratified through the special supermajoritarian procedures laid out in
the Constitution.
386
The rules of Article V play a crucial role in this picture.
They not only determine which amendments have become part of the
canonical document but also, once satisfied, endow the communicative
content of these texts with a legal authority that never fades over time,
unless and until a new Article V amendment comes along and overrides
them.
387
“The Article V amendment process and originalism,” in John
McGinnis’s words, “march under a single banner.”
388
This paper’s descriptive account helps reveal a new sense in which this
picture of Article V as arbiter and embodiment of the sovereign will may
be too simple. The vast majority of recognized amendments have not been
adopted in clear compliance with Article V.
389
All but one has been
adopted in clear defiance of a requirement of presidential presentment
that, as a textual matter, plausibly applies to the Article V process.
390
The
single most transformative set of amendments, many believed at the time
of their adoption and still believe today, amounted to a “naked violation[]
of Article V.”
391
And ever since the Founding, the meaning of Article V has
been continually contested without, on many significant questions, reach-
ing any resolution. Nothing in the language or law of Article V tells us
definitively, for instance, that the Twenty-Seventh Amendment is part of
the Constitution while the ERA is not. In U.S. constitutional culture, the
degree to which an attempted amendment has complied with a discrete
set of legal rules has always had an attenuated connection to its ultimate
constitutional fate. What is and is not “in” the Constitution is ultimately
385. See Kurt T. Lash, Originalism, Popular Sovereignty, and Reverse Stare Decisis, 93
Va. L. Rev. 1437, 1440 (2007) (stating that “popular sovereignty and the judicially enforced
will of the people” is “the most common and most influential justification for originalism”).
386. Lawrence B. Solum, Communicative Content and Legal Content, 89 Notre Dame
L. Rev. 479, 514 (2013); see also Colby, supra note 150, at 1631–38 (summarizing the stand-
ard “popular sovereignty” argument for originalism).
387. See Steven G. Calabresi & Saikrishna B. Prakash, The President’s Power to Execute
the Laws, 104 Yale L.J. 541, 551 (1994) (“The central premise of originalism . . . is that the
text of the Constitution is law that binds each and every one of us until and unless it is
changed through the procedures set out in Article V.”); cf. Stephen E. Sachs, Originalism
as a Theory of Legal Change, 38 Harv. J.L. & Pub. Pol’y 817, 820–21 (2015) (“What original-
ism requires of legal change is that it be, well, legal; that it be lawful, that it be done according
to law. This is a requirement of procedure, not substance.”).
388. John O. McGinnis, Protecting the Originalist Constitution, 42 Harv. J.L. & Pub.
Pol’y 81, 85 (2019). “And what does that banner read? It says, ‘We the People,’ and not ‘We
the Elite Judges.’” Id.
389. See supra Part III.
390. See supra section III.B; supra note 21 and accompanying text.
391. Ackerman, We the People, supra note 7, at 111; cf. Colby, supra note 150, at 1630–
31, 1662–66 (suggesting “the possibility that the shortcomings in the framing of the
Fourteenth Amendment could seriously undermine the normative appeal of originalism
more generally”).
2386 COLUMBIA LAW REVIEW [Vol. 121:2317
determined by a complex, ongoing, and not especially predictable dialec-
tic between formal government actions and popular and political beliefs
and perceptions. Provocatively put, the Constitution is not so much a col-
lection of texts that have passed through a unique legitimating process as
a collection of texts that Americans agree to regard as the Constitution,
despite wide variations in and serious questions about the manner in
which they were added.
392
If this is right, it calls into question the legal and empirical basis for
the popular sovereignty justification of originalism. There are no clear
rules to indicate, in many cases, whether an attempted act of sovereign
constitutional authorship meets the criteria laid out in the Constitution.
Nor do the details of how any given piece of purported constitutional text
was created—including both the forms that were followed and the degree
to which the overall process could be characterized as supermajoritarian—
seem to determine its sociological legitimacy in any mechanical manner.
The process of formal constitutional change has always been less legalistic
and more chaotic than standard originalist narratives about Article V seem
to presume.
None of this necessarily upends “the great debate” between original-
ism and living constitutionalism or indicates that originalism is less demo-
cratic than alternative approaches.
393
Critics have pointed out numerous
other diculties with the popular sovereignty case for originalism.
394
And
proponents of originalism have oered numerous other justifications that
do not depend on the character of the process that led to the text’s enact-
ment. Prakash, for instance, defends originalism as a logical entailment of
the hermeneutic enterprise while expressly “contest[ing] the interpretive
assertion that . . . originalism is a legitimate means of making sense of the
Constitution merely or primarily because of the manner in which the
Constitution was ratified and amended.”
395
William Baude and Stephen
Sachs have launched a defense of originalism as “our law” on openly pre-
sentist, positivist grounds, with no direct connection to Article V or VII.
396
Our intervention in the debate is simply to suggest that the popular sover-
392. There is an interesting parallel, which we do not have space to pursue, to the his-
torical processes through which certain texts are canonized as part of sacred scripture while
others are denied that status in a given theological tradition. See Frank Kermode, The
Canon, in The Literary Guide to the Bible 600 (Robert Alter & Frank Kermode eds., 1987).
The Constitution itself can be thought of as a kind of canon in this sense.
393. See generally Lawrence B. Solum, Originalism Versus Living Constitutionalism:
The Conceptual Structure of the Great Debate, 113 Nw. U. L. Rev. 1243 (2019).
394. See Colby, supra note 150, at 1662–63 (summarizing diculties).
395. Saikrishna B. Prakash, The Misunderstood Relationship Between Originalism and
Popular Sovereignty, 31 Harv. J.L. & Pub. Pol’y 485, 486 (2008).
396. See, e.g., William Baude & Stephen E. Sachs, Grounding Originalism, 113 Nw. U.
L. Rev. 1455 (2019); William Baude, Is Originalism Our Law?, 115 Colum. L. Rev. 2349
(2015). For an elegant rebuttal, see Richard Primus, Is Theocracy Our Politics?, 116 Colum.
L. Rev. Sidebar 44 (2016).
2021] PUZZLES AND POSSIBILITIES 2387
eignty case for originalism is subject not only to the standard counterargu-
ments about the dubious democratic bona fides of an ancient text but also
to a new counterargument about the dubious legal bona fides of nearly
every provision of that text.
B. Amendment Inside and Outside Article V
A foundational debate in constitutional theory concerns whether and
how the written or big-C Constitution may legitimately be updated outside
the procedures specified in Article V.
397
The “outsider” position is most
closely associated with Amar and Ackerman.
398
According to Amar, the
Constitution is best read to preserve for the people an unenumerated right
to amend its terms through something akin to a national referendum.
399
According to Ackerman, the Reconstruction Amendments were adopted
in violation of Article V but were nonetheless legally legitimate because
their adoption conformed to the true, unwritten criteria for higher
lawmaking.
400
Conversely, the Twenty-Seventh Amendment was adopted
in conformity with Article V but is nonetheless illegitimate because it
violated those unwritten criteria.
401
The “insider” position is the
conventional view—the view that Article V supplies the exclusive route to
formal constitutional change, that “[n]othing new can be put into the
Constitution except through the amendatory process.”
402
Our account collapses some of the space between these two positions.
Given the long history of procedural creativity and the pervasive legal
uncertainty that we document, there is no clear line demarcating what is
“inside” or “outside” Article V. Ackerman labors heroically to show that
the Reconstruction Amendments were valid additions to the Constitution
even though brought into existence in a manner that is very plausibly
inconsistent with Article V. We do not disagree with Ackerman’s
conclusion; we disagree with the premise that the Reconstruction
Amendments were quite so extraordinary in this regard. Within broad
boundaries, the degree to which an attempted amendment stays inside the
four corners of Article V has not been decisive in determining whether it
397. The unwritten or small-c constitution has been updated many times over by judicial
decisions, framework statutes, and other developments that have little to do with Article V.
See supra notes 42–43 and accompanying text. The focus here is on the more radical prop-
osition that the constitutional text itself may be changed outside Article V.
398. Cf. James E. Fleming, We the Unconventional American People, 65 U. Chi. L. Rev.
1513, 1540 (1998) (book review) (“In recent years, some constitutional scholars have noted
the emergence of a ‘Yale school’ of constitutional theory, by which they refer to Ackerman’s
and Amar’s theories of amending the Constitution outside Article V.”).
399. See Amar, Consent of the Governed, supra note 3, at 462–94; Amar, Philadelphia
Revisited, supra note 4, at 1044–76.
400. See Ackerman, We the People, supra note 7, at 99–252.
401. Id. at 490 n.1.
402. Ullmann v. United States, 350 U.S. 422, 428 (1956); see also supra note 3 and
accompanying text.
2388 COLUMBIA LAW REVIEW [Vol. 121:2317
becomes accepted by Americans as part of the written Constitution. All
constitutional amendment, in this sociological sense, takes place “outside”
Article V.
That being the case, it is notable how little headway Amar’s relatively
straightforward proposal has made in convincing Americans that they can
amend the Constitution through a national referendum established by
and employing a simple majority vote. Especially in light of the constitu-
tional order’s steady shift toward greater nationalism and majoritarian
democracy ever since the Civil War, his proposal strikes us as no less plau-
sible as a matter of constitutional text, structure, and “spirit”
403
than the
notion that an Article V amendment could (like the Twenty-Seventh
Amendment) remain pending for more than two centuries before being
ratified. Yet, even though the law of Article V is so unsettled and so many
amendments have questionable Article V credentials, the assumption that
all revisions to the written Constitution must be pursued through the
Article V process continues to hold a powerful sway on the U.S. legal and
political community, at least among elites.
404
This persistent combination
of Article V obscurity and Article V exclusivity suggests that the two may
reinforce one another: If determined majorities had not found Article V
to have such play in the joints, its status as the exclusive gateway to the
constitutional text may well have proved unsustainable long ago.
405
These observations may also hold a clue as to how Article V exclusivity
could unravel in the future. If Amar’s proposal is just as constitutionally
coherent as numerous amendments that are accepted as part of the docu-
ment, then the same sort of political mobilization that potentiated those
amendments may be sucient to potentiate amendment-by-referendum
403. Amar, Consent of the Governed, supra note 3, at 460–61.
404. See Sanford Levinson, The Political Implications of Amending Clauses, 13 Const.
Comment. 107, 114 (1996) (remarking that “Amar’s method” of amending the Constitution
by popular referendum “is not only untried but also, for most Americans, I suspect
unthinkable”).
405. Cf. Philippe Nonet & Philip Selznick, Law and Society in Transition: Toward
Responsive Law 76–79 (Routledge 2d ed. 2017) (1978) (discussing the tension that all legal
institutions face between the need to limit discretion and discipline decisionmaking, on the
one hand, and the need to remain responsive to new pressures and contingencies, on the
other). Article V’s status as the exclusive gateway to the constitutional text may also have
proved unsustainable if the Court had not, over time, been so accommodating of legal and
social change outside the Article V process. See Strauss, The Living Constitution, supra note
327, at 115–16 (arguing that “[s]ome form of living constitutionalism is inevitable” in light
of the diculty of formal amendment).
2021] PUZZLES AND POSSIBILITIES 2389
as well.
406
At least for those willing to look past its novelty,
407
Amar’s pro-
posal is not necessarily more legally outlandish than any number of things
that have been done in the name of Article V. The constitutional text,
accordingly, is not the principal problem for Amar; constitutional culture
is.
408
Perhaps calling attention to just how fast and loose Americans histor-
ically have played with Article V, as this paper has done, will conduce to
greater cultural openness to experimenting with other legally plausible (if
unavoidably problematic) modes of updating the constitutional text in the
service of deepening democracy. But in case Amar’s argument is destined
to remain o-the-wall in our lifetimes, we close with a more modest reform
proposal of our own.
C. Loosening the Constitutional Cage Through Article V Thayerianism
For all of the ambiguities we have identified, the hard core of Article
V remains. Unless a first-ever Article V convention is called, those who
would revise the written Constitution need to convince others that two-
thirds of both chambers of Congress have approved, and three-quarters of
the states have ratified, an amendment. There are many dierent ways to
count to two-thirds and three-fourths, as we emphasize throughout Parts
II and III, but a plausible double-supermajoritarian showing of some sort
must be made. In comparison with the approaches taken by other
democracies, this is an exceptionally dicult amendment process. And
the actual rate of amendment in the United States has been exceptionally
low.
409
As many have argued, the practical diculty of revising the written
Constitution invites judges to update supreme law through creative
406. See Schauer, Amending the Presuppositions, supra note 3, at 160–61
(“[C]onstitutions are always subject to amendment by changes—amendments—in the prac-
tices of a citizenry, in the practices of its ocials, and in the practices of its judges.”). The
precise mechanisms and pathways through which constitutional orthodoxies change over
time remain “enigmatic,” Pozen & Samaha, supra note 69, at 792, but potential contributors
presumably include social movements, partisan politics, and judicial appointments as well
as academic argument.
407. Cf. Leah M. Litman, Debunking Antinovelty, 66 Duke L.J. 1407, 1427–65 (2017)
(criticizing the view that “novelty” is evidence of a constitutional problem in the legislative
context).
408. See 1 Laurence H. Tribe, American Constitutional Law 109–10 (3d ed. 2000)
[hereinafter Tribe, American Constitutional Law] (observing that the sociological legiti-
macy of a non-Article V amendment would ultimately depend upon “social and cultural
practices,” but expressing doubt that such an amendment could succeed anytime soon); see
also Richard Albert, The Case for Presidential Illegality in Constitutional Amendment, 67
Drake L. Rev. 857, 873–75 (2019) (discussing, with reference to examples from other coun-
tries, how “the sociological and moral force” of a successful yet “illegal” constitutional ref-
erendum could compel “the legal and political elite to recognize the validity of this
unconventional change to the U.S. Constitution”). See generally Peter Suber, The Paradox
of Self-Amendment: A Study of Law, Logic, Omnipotence, and Change (1990) (describing
various modes of amending an amendment process).
409. See supra note 366 and accompanying text.
2390 COLUMBIA LAW REVIEW [Vol. 121:2317
readings and undermines even the possibility of genuine collective self-
rule
410
—leaving our democracy “trapped inside the Article V cage.”
411
Yet without exiting the cage altogether, as an Amarian referendum or
a true revolution would have us do, perhaps we might loosen its bars.
When one combines the classic democratic case against Article V’s “obdu-
racy”
412
with this paper’s new account of Article V’s ambiguity, the promise
of what could be called Article V Thayerianism comes into focus. The label
refers to James Bradley Thayer’s famous proposal that judges should defer
to a coordinate legislative branch on constitutional questions except when
the latter has made a “very clear” mistake.
413
Thayerianism is typically
advocated today as a means of respecting the constitutional judgment and
authority of the legislature and thereby limiting the countermajoritarian
character of constitutional law
414
—virtues that become all the more signif-
icant under conditions of high-level interpretive uncertainty.
415
This argu-
ment transposes readily to the Article V context, where
countermajoritarian concerns are especially acute and legal uncertainty is
especially rife. If it were to take hold, Article V Thayerianism would neu-
tralize the potential chilling eects of this uncertainty and promote insti-
tutional innovation in the amendment process. Simply put, Article V
Thayerianism would make it easier to amend the Constitution, at least at
the margins.
Article V Thayerianism could be operationalized in more or less
ambitious ways. A narrow version would have judges refrain from blocking
amendment proposals by Congress or amendment ratifications by state
legislatures on Article V grounds, unless the actions are seen as “clearly”
inconsistent with Article V. Given Coleman’s nonjusticiability ruling, this
version would not necessarily mark any observable advance on the status
quo—though Coleman’s precedential status is far from secure and could
410. See, e.g., John Ferejohn & Lawrence Sager, Commitment and Constitutionalism,
81 Tex. L. Rev. 1929, 1954 (2003) (discussing “the standard critique” that “the obduracy of
Article V acts to suppress the people’s voice in our constitutional aairs, and thus is either
flatly undemocratic, or at least more antidemocratic than we would like”); David Singh
Grewal & Jedediah Purdy, The Original Theory of Constitutionalism, 127 Yale L.J. 664, 668,
679 (2018) (book review) (arguing that “the Article V amendment procedure [has come]
perilously close to choking o further sovereign action by the people,” producing “a politi-
cal community that is at once committed to ruling itself and unable to do so”).
411. Sanford Levinson, Our Undemocratic Constitution: Where the Constitution Goes
Wrong (and How We the People Can Correct It) 20–21 (2006) [hereinafter Levinson, Our
Undemocratic Constitution] (capitalization omitted).
412. Ferejohn & Sager, supra note 410, at 1954.
413. James B. Thayer, The Origin and Scope of the American Doctrine of Constitutional
Law, 7 Harv. L. Rev. 129, 144, 154–55 (1893).
414. Cf. G. Edward White, Historicizing Judicial Scrutiny, 57 S.C. L. Rev. 1, 48–50 (2005)
(explaining that Thayer himself did not defend the clear-mistake rule, as “[m]odern com-
mentators tend to,” in terms of “countermajoritarian constraints”).
415. See Vermeule, Judging Under Uncertainty, supra note 14, at 230–88.
2021] PUZZLES AND POSSIBILITIES 2391
use buttressing.
416
A more expansive version, which we favor but which
raises additional complications, would be internalized by legislative and
executive ocials as well as by judges. Members of Congress, on this
approach, would likewise apply a rule of clear mistake when called on to
resolve the validity of contested steps forward that are taken in the amend-
ing process. The hard core of Article V would not be aected. Aspiring
amenders would still need to persuade two-thirds of Congress and three-
fourths of the states in a constitutionally credible manner. But the fuzzy
edges around that core would become more permeable, zones of permis-
sion and experimentation rather than additional vetogates on the path to
constitutional change.
As with all versions of Thayerianism, just how permissive Article V
Thayerianism would prove may vary depending on who is applying it.
Those who see more clarity in the law of Article V will tend to find more
clear mistakes in need of correction. In principle, however, the Thayerian
proposal is potentially compatible with any interpretive approach, as it says
nothing about the method of constitutional interpretation that is to be
used—only about what is to be done when that method generates an
uncertain legal conclusion. Originalism could lead to an especially permis-
sive version of Article V Thayerianism, given how little light was shed on
the workings of Article V during its drafting and ratification
417
and the
growing recognition among originalists that constitutional decisionmak-
ers must rely on normative judgment in situations where the communica-
tive content of the constitutional text is too vague or ambiguous to fully
determine a legal result.
418
What would Article V Thayerianism look like in practice? Most imme-
diately, the ERA would be recognized as part of the Constitution as soon
as Congress so declares.
419
As Part III discusses, the Article V objections to
the ERA are substantial but not “clearly” fatal, especially if Congress takes
new action to extend the ratification deadline.
420
Accordingly, if Congress
416. See supra section IV.B.
417. See supra note 17 and accompanying text.
418. See Pozen & Samaha, supra note 69, at 777–78; see also Lawrence B. Solum,
Originalist Methodology, 84 U. Chi. L. Rev. 269, 295 (2017) (noting that one approach
originalists might take in situations of textual underdeterminacy is to adopt “a Thayerian
default rule of deference to democratic institutions”).
419. As Julie Suk details in a forthcoming article, Ruth Bader Ginsburg’s writings and
testimony before becoming a judge support a primary role for Congress in resolving the
current status of the ERA. See Julie C. Suk, The Procedural Path of Constitutional Inclusion:
Justice Ginsburg’s Cautious Legacy for the Equal Rights Amendment, 110 Geo. L.J. (forth-
coming 2022) (on file with the Columbia Law Review).
420. See supra section III.L. A recent district court opinion called the question whether
Congress “may revive the ERA” a “dicult issue” and did not resolve it. Virginia v. Ferriero,
525 F. Supp. 3d 36, 61 (D.D.C. 2021). In its 2020 opinion finding the ERA to be dead, OLC
acknowledged that Congress’s authority to “modify” a ratification deadline presents a
“dicult question.” Ratification of the Equal Rights Amendment, 44 Op. O.L.C., slip op. at
3 (Jan. 6, 2020).
2392 COLUMBIA LAW REVIEW [Vol. 121:2317
were to pass a joint resolution retroactively waiving the deadline and
directing the Archivist of the United States to publish the ERA as the
Twenty-Eighth Amendment, the Archivist should promptly do so.
421
More
generally, Article V Thayerianism might empower aspiring amenders to
take advantage of quorum rules, to consider amendments passed by one
chamber in the other chamber during subsequent legislative sessions,
422
to
demand a supermajority vote for rescissions of amendment proposals and
ratifications, to hold binding or advisory state referenda on ratification,
423
and so on and so forth.
424
Perhaps the most promising avenue of Thayerian experimentation
would involve ratification through conventions. Congress could, for
instance, propose an amendment and provide for at-large elections of state
convention delegates on a single day. Doing so would eectively create an
amendment “Election Day” and convert each state vote into a referendum
421. A middle option, which we find appealing, would be for Congress to retroactively
extend the deadline while at the same time providing that it will treat rescissions as eective.
See Magliocca, supra note 360, at 635; Hemel, supra note 317. This would be a departure
from strict Thayerianism, in the sense that it would not read Article V in the most permissive
fashion possible. But such a departure may be prudentially warranted in light of the unique
circumstances of the ERA (specifically, the multiple rescissions and twofold extension of the
ratification deadline), and it would return the ratification process to the state legislative
arena for a final push. Cf. Goldwater v. Carter, 444 U.S. 996, 1003 (1979) (Rehnquist, J.,
concurring in the judgment) (suggesting that the ecacy of rescission “‘might be answered
in dierent ways for dierent amendments’” (quoting Dyer v. Blair, 390 F. Supp. 1291, 1302
(N.D. Ill. 1975) (three-judge court))).
422. For an arguable limit case, the House in 1969 passed a constitutional amendment
abolishing the Electoral College and providing for direct election of the President, but the
Senate did not follow suit. See 115 Cong. Rec. 26,007–08 (1969). Is it clear, as a matter of
Article V law, that the Senate could not revive this proposed amendment? See Seth Barrett
Tillman, Noncontemporaneous Lawmaking: Can the 110th Senate Enact a Bill Passed by
the 109th House?, 16 Cornell J.L. & Pub. Pol’y 331, 345 n.31 (2007) (arguing that the House
and Senate need not act contemporaneously in the amendment process).
423. Unless it is overturned or defied, Hawke v. Smith (No. 1), 253 U.S. 221 (1920),
would prevent a state from holding a referendum undercutting ratification after legislative
approval. But a state could still, consistent with Hawke, hold a referendum prior to legislative
consideration and choose to bind itself to the result.
424. Article V Thayerianism is a heuristic to adjudicate present and future disputes over
the validity of amendment eorts, such as the ERA. For disputes about past amendment
eorts, we believe it makes sense for systemic stare decisis reasons to defer to longstanding
social and ocial consensus about the textual content of the Constitution. As William Baude
has suggested to us, one might call this the “pocket constitution method”: Copies of the
Constitution in general circulation and carried around by people are presumptively correct,
absent a clear inconsistency with Article V. Hence, while there may be a colorable argument
that Article the First, on congressional apportionment, satisfied the requirements of Article
V, see supra section III.L.1, that argument is not so clearly correct as to upset the overwhelm-
ing consensus spanning more than two centuries that the amendment was never adopted.
Conversely, while the Titles of Nobility Amendment appeared for a time in (some) pocket
constitutions, see supra section III.L.2, its noncompliance with Article V was clear. The
Twenty-Seventh Amendment is an intermediate case. Despite the controversy surrounding
its initial promulgation, it has appeared in pocket constitutions for three decades now and
is not clearly in violation of Article V. See supra section III.K.
2021] PUZZLES AND POSSIBILITIES 2393
on the proposed amendment. Congress could even include a provision in
the proposed amendment stipulating that it will be inoperative unless
approved by a majority of all voters on Election Day. In addition to increas-
ing the democratic legitimacy of the resulting amendment, this maneuver
would streamline the Article V process in two ways: It would require only
a single electoral victory in each state, rather than a victory in both houses
of bicameral state legislatures,
425
and it would circumvent partisan
gerrymandering in those legislatures.
More radical possibilities are also imaginable. Consider the final
clause in Article V providing “that no State, without its Consent, shall be
deprived of its equal Surage in the Senate.”
426
For those who believe the
two-senators-per-state rule to be inconsistent with the value of political
equality, this clause reflects “a truly extraordinary” and grossly undemo-
cratic “instance of dead-hand control.”
427
As Robert Dahl once lamented,
“those fifteen words end all possibility of amending the constitution in
order to reduce the unequal representation of citizens in the Senate.”
428
But is this so clear? Many workarounds to the Equal Surage Clause have
been proposed over the years, from an amendment repealing the clause
paired with another amendment changing the Senate apportionment for-
mula,
429
to an amendment preserving the Senate in name but “relocating”
its powers to a new body,
430
to the proposition that the one-person-one-
vote equal protection principle announced in Reynolds v. Sims should be
understood to apply to the Senate,
431
to the centuries-old argument that
the clause has never been legally operative.
432
These proposals vary in their
degree of legal boldness; depending on one’s general approach to consti-
tutional interpretation and one’s specific views on the Equal Surage
425. Cf. Donald S. Lutz, Toward a Theory of Constitutional Amendment, 88 Am. Pol.
Sci. Rev. 355, 361 (1994) (assuming “that a unicameral legislative process is one half as
dicult as a bicameral one”).
426. U.S. Const. art. V.
427. Michael J. Klarman, The Framers’ Coup: The Making of the United States
Constitution 627 (2016).
428. Robert A. Dahl, How Democratic Is the American Constitution? 145 (2d ed. 2003).
429. See, e.g., Tribe, American Constitutional Law, supra note 408, at 111–12; Lynn A.
Baker & Samuel H. Dinkin, The Senate: An Institution Whose Time Has Gone, 13 J.L. &
Pol. 21, 67–70 (1997).
430. Amar, America’s Constitution, supra note 49, at 293.
431. See, e.g., Eric W. Orts, Senate Democracy: Our Lockean Paradox, 68 Am. U. L.
Rev. 1981, 1999–2009, 2071–75 (2019) (oering a detailed version of this proposal, imple-
mentable by statute). The Reynolds Court expressly declined to apply the one-person-one-
vote principle to the Senate, citing the Senate apportionment formula’s “ingrained [status]
in the Constitution” and the “unique historical circumstances” attending its creation. 377
U.S. 533, 574 (1964).
432. See supra note 61 and accompanying text.
2394 COLUMBIA LAW REVIEW [Vol. 121:2317
Clause, some may seem clearly impermissible. But most are “at least think-
able” as a matter of orthodox constitutional interpretation.
433
For those
who find such proposals not just legally thinkable but legally credible,
Article V Thayerianism and the principle of popular sovereignty that
underwrites it counsel openness to reform.
Even more important than its direct eects on the legal environment
for attempted amendments, however, may be Article V Thayerianism’s
indirect eects on the cultural environment. A striking finding from the
comparative constitutional literature is that the procedural diculty of a
country’s formal amendment rule is not strongly correlated with its rate of
amendment.
434
Some countries with stringent amendment rules rewrite
their constitutions frequently; some countries with lax amendment rules
rewrite their constitutions only rarely. More consequential than the
amendment rule itself, it seems, is the prevailing “amendment culture.”
435
In the United States, Jackson suggests that perceptions of the diculty of
satisfying Article V not only tend to be “overstated” but also have become
“self-fulfilling” by deterring political actors from trying to pursue
amendments.
436
The determinants of amendment culture are enig-
matic,
437
so we cannot say with any confidence what the precise eects of
interpretive reform would be, in the United States or any other system.
There is also an endogeneity complication because, just as the embrace of
Article V Thayerianism might change U.S. amendment culture, a change
in U.S. amendment culture might be needed for Article V Thayerianism
to gain traction. Cultural change has to start somewhere, however, and it
does not seem far-fetched to think that growing appreciation for just how
many questions the law of Article V does not clearly resolve, combined with
growing levels of scholarly support for Article V Thayerianism, could
433. Sanford Levinson, There Is a Way, but Will There Ever Be a Will?: Comments on
Eric Orts’s Senate Democracy, 69 Am. U. L. Rev. Forum 159, 164 (2020); see also David B.
Froomkin & A. Michael Froomkin, Fixing the Senate: A User’s Guide 67 (Univ. of Mia.,
Legal Stud. Rsch. Paper No. 3797782, 2021), https://ssrn.com/abstract=3797782
[https://perma.cc/YG6R-B4ZC] (“[T]he Entrenchment Clause presents a less substantial
obstacle to reform than is sometimes imagined . . . .”).
434. See Tom Ginsburg & James Melton, Does the Constitutional Amendment Rule
Matter at All? Amendment Cultures and the Challenges of Measuring Amendment
Diculty, 13 Int’l J. Const. L. 686, 711 (2015) (“[A]mendment culture is more important
than institutional constraints in explaining amendment practices.”); Bjørn Erik Rasch,
Rigidity in Constitutional Amendment Procedures, in The Constitution as an Instrument of
Change 111, 121 (Eivind Smith ed., 2003) (noting that the “empirical relationship between
rigidity and amendment is . . . not very robust”).
435. Ginsburg & Melton, supra note 434, at 699–701, 708–13.
436. Jackson, The (Myth of Un)Amendability, supra note 30, at 584–90, 602–05; see also
Vicki C. Jackson, Democracy and Judicial Review, Will and Reason, Amendment and
Interpretation: A Review of Barry Friedman’s The Will of the People, 13 U. Pa. J. Const. L. 413,
433–52 (2010) (book review) (discussing and critiquing “amendophobia” in contemporary
U.S. constitutional culture).
437. See Ginsburg & Melton, supra note 434, at 687, 701.
2021] PUZZLES AND POSSIBILITIES 2395
nudge the U.S. “legal complex”
438
in a more amendment-friendly direc-
tion. Above and beyond any discrete arguments we have advanced, we
hope this paper contributes to such a shift.
C
ONCLUSION
A written constitution’s rules for its own amendment help “define its
very essence.”
439
Our ambition in this paper has been to explore the
“essence” of the U.S. Constitution through a critical study of how the rules
of Article V have been developed and applied over time. The standard
account of Article V depicts it as all but freezing our constitutional text, if
not our democracy as well.
440
Yet for all of Article’s V ostensible “clarity”
441
and “rigidity,”
442
the picture that emerges from this study is one of
persistent legal contestation, confusion, and innovation.
Descriptively, we have tried to oer a robust empirical showing of the
underdeterminacy of the American rule of recognition for constitutional
enactments. Prescriptively, we have suggested that this showing holds
untapped promise for those who wish to revise the Constitution. Ever since
the Founding, the bars of “the Article V cage”
443
have been significantly
looser than the conventional wisdom appreciates. And still to this day, the
law of Article V remains remarkably unsettled not just on minor
technicalities but on fundamental questions of substance and procedure.
Recovering the full story of Article V adventurism, and recognizing the full
scope of legal discretion left in its wake, are important tasks for enriching
our understanding of constitutional history and theory. They may also be
necessary first steps toward unfreezing the constitutional text today.
438. See Lucien Karpik & Terence C. Halliday, The Legal Complex, 7 Ann. Rev. L. &
Soc. Sci. 217, 220, 233–34 (2011) (explaining that “constitutional politics invariably are
embedded within a politics of the legal complex,” defined as “a cluster of legal actors,”
including academics, “related to each other in dynamic structures and constituted and
reconstituted through a variety of processes”).
439. Amar, Philadelphia Revisited, supra note 4, at 1102 n.208.
440. For the freezing metaphor, see Daniel Lazare, The Frozen Republic: How the
Constitution Is Paralyzing Democracy (1997); see also Albert, Constitutional Amendments,
supra note 113, at 96 (collecting similar statements by leading scholars).
441. E.g., Albert, Constitutional Disuse or Desuetude, supra note 14, at 1035; Harrison,
supra note 14, at 459, 461; Paulsen, A General Theory, supra note 50, at 761.
442. E.g., Heather K. Gerken, The Hydraulics of Constitutional Reform: A Skeptical
Response to Our Undemocratic Constitution, 55 Drake L. Rev. 925, 938 (2007); Aziz Z. Huq,
The Function of Article V, 162 U. Pa. L. Rev. 1165, 1168–69 (2014).
443. Levinson, Our Undemocratic Constitution, supra note 411, at 20.
2396 COLUMBIA LAW REVIEW [Vol. 121:2317
