The Word He Left Out

RE: “Two-Thirds of Whom, Revisited: A Reply to Dr. I.S. Koroma of Kornel Law.”

On 10th of August this year, I wrote a piece on air while traveling to the United States. It was on the constitutional amendment process that dominated the debates at the time in Freetown which was my place of departure, with some venturing to describe the process as a possible treasonable conduct by the Speaker and if the President were to assent the Constitution of Sierra Leone (Amendment) Bill 2025 which was passed in Parliament on that day, himself inclusive as would any other participant. Only as a matter of principle and in acknowledgment of the work of a dear colleague have I decided to do this response after my attention was curiously drawn to the rejoinder to that my earlier piece by Dr. Emmanuel Saffa Abdulai. Yes, for over two decades, he has been my friend since our student days at Fourah Bay College where we both teach at the postgraduate school.¹

Dr. Abdulai has done what a good colleague across the corridor is supposed to do: he has taken my argument seriously enough to write sixteen thousand words against it, and he has done so with real skill. I am grateful for the engagement, and I agree with him on more than his piece might suggest; that this is a genuine constitutional puzzle, that section 108(2)(b) is a badly drafted provision, and that the missing word "all" deserves an explanation rather than a shrug. In fact, I emphasized the horrible drafting in my earlier piece. Where I part company with him is what follows from all of that. His reply builds an elaborate argument around one missing word while passing over a second missing phrase that sits far closer to home, inside the very same chapter of the very same Constitution, and that phrase (not the comparative tour through Harare, Nairobi, Pretoria and Kampala), is what actually decides this case. I will do a short, if you like the old phrase, surrejoinder.

The provision Dr. Abdulai does not mention

Section 91(1) of the 1991 Constitution of Sierra Leone is the general rule for how Parliament decides anything. It provides, in substance, that except where the Constitution otherwise provides, questions before Parliament are determined by a majority of the Members present and voting.² That opening qualifier; "except as otherwise provided in this Constitution," is doing real work. It tells us that Section 91(1) is the default, and that wherever the Constitution wants something different, it says so specifically. It also tells us, in the plainest possible terms, exactly what phrase our own drafters used when they meant a denominator confined to the Members who turn up and vote: "present and voting." That is the model. That is the sentence our drafters reached for whenever presence, not total membership, was what they intended.

Section 108(2)(b) does not use that phrase. It does not say, "two thirds of the Members present and voting." It says, "not less than two thirds of the Members of Parliament," full stop, the same bare formula Section 91(1) exists precisely to displace when the Constitution "otherwise provides." Dr. Abdulai's entire argument is built on the significance of one missing word, "all." I did not just write it down, I watched him say it many times. He does not once engage with the significance of a missing phrase, "present and voting," sitting one section away, in a provision whose only function is to supply exactly that qualifier when the drafters wanted it and to yield the floor when they did not.

This is not a minor oversight in an otherwise complete piece. It is the single most relevant textual comparator in the entire Constitution, and it points the opposite way from where Dr. Abdulai's piece takes the reader. If "the Members of Parliament" in section 108(2)(b) already meant, by some silent default, the Members present and voting, then section 108(2)(b) would simply be restating section 91(1) at a different fraction (I raised this point in my earlier piece and Dr. Abdulai did not seem to grasp it). But section 91(1) already tells us what "otherwise provided" looks like in this Constitution's own drafting convention, and section 108(2)(b) does not use that convention. On the ordinary canon that a specific provision governs over a general one; generalia specialibus non derogant,³ section 108(2)(b) is the "otherwise provided" that section 91(1) itself anticipates, and it displaces the presence-based denominator altogether, not merely the fraction attached to it. Dr. Abdulai either ignored this fact or was never in tune with it.

What the Supreme Court of Sierra Leone has already told us

Dr. Abdulai reaches for foreign case law i.e., the American National Prohibition Cases of 1920 as, in his words, the single most relevant piece of comparative authority on this question. He does not mention that Sierra Leone's own Supreme Court has already spoken to the correct method for reading contested constitutional text. In Alhaji Sam Sumana v. Attorney-General and Minister of Justice and Victor Foh, the Court held that where constitutional language is clear and unambiguous, effect must be given to the ordinary meaning of the words employed.⁴ That was an application, in our own jurisdiction, of the long-settled literal rule associated with R. v Judge of the City of London Court.⁵ The ordinary meaning of "the Members of Parliament" is the body corporate, the whole 149-seat institution the Constitution elsewhere calls Parliament, not a subset defined by whoever happens to be in the chamber on a given afternoon. If the framers wished to shrink that denominator to those "present and voting," Section 91(1) shows they knew exactly how to say so and did not say so here.

Dr. Abdulai is right that Commonwealth authority which was discussed in that my earlier piece; Bribery Commissioner v. Ranasinghe⁶ and Attorney-General for New South Wales v. Trethowan⁷ establishes that entrenchment clauses bind Parliament and that courts will look behind a purported amendment to verify it was validly passed. He treats this as background that "answers a question nobody is asking." I would put it differently: that doctrine is precisely why the denominator question matters so much, and why it should be resolved in favour of the reading that keeps the entrenchment meaningful. A manner-and-form requirement that can be satisfied by manufacturing a quorum's worth of "present and voting" supporters, after a minority has been squeezed out of the chamber, is not much of an entrenchment at all.⁸ The Ranasinghe and Trethowan line of authority exists to stop governments from amending constitutions through procedural shortcuts. Reading Section 108(2)(b) down to a "present and voting" threshold does not honour that doctrine; it hollows it out.

The comparative argument, reconsidered

Dr. Abdulai makes much of the fact that Zimbabwe, Kenya, South Africa and Uganda all spell out "all members" or "the entire membership" for their equivalent thresholds, while Sierra Leone's text does not. He treats the absence as proof that Sierra Leone must have meant something narrower. There is a more obvious reading of the same evidence. Every one of those constitutions is significantly younger than ours; Uganda's dates to 1995, South Africa's to 1996, Kenya's to 2010, Zimbabwe's to 2013, and each was drafted after decades of exactly the kind of dispute Sierra Leone is now having. Constitutional drafters across the Commonwealth learned, from lived experience with ambiguous "two-thirds" clauses, that the phrase invites litigation unless nailed down explicitly. The later constitutions did not invent a new rule; they closed a loophole that older, looser drafting like ours had already exposed. That later drafters felt the need to spell out "all members" is evidence that unadorned "two-thirds of the members" language was already understood, well before 2013, to run the real risk of being read as reaching the full membership, which is exactly why careful drafters started saying so in terms, to remove any doubt either way. It is not evidence that the older, unadorned formula must mean the opposite. Whether Dr. Abdulai tries to edify it as a probable mistaken interpretation or a misapplication of procedure by the Speaker as he did on his Truth Media interview recently, or just blatantly as refusing to take a clear position on the law, as I have demonstrated, that argument bears no edible fruit.

What the drafting history actually shows

Dr. E.S. Abdulai's history is accurate as far as it goes: the 1961 Independence Constitution said, "all the members of the House,"⁹ and the 1978 Constitution, drafted by a government that had just abolished multi-party politics, dropped the word "all."¹⁰ He asks us to draw the inference that the 1978 drafters wanted a weaker, more easily satisfied threshold, and that the 1991 framers, by re-adopting the 1978 wording, inherited that weaker meaning. But a one-party government stripping away the word "all" is at least as plausible a story about removing the explicit two-election, two-session safeguard that accompanied it in 1961 as it is a story about the denominator. The 1961 clause was not simply "all the members"; it required passage in two separate parliamentary sessions separated by a general election. That is what a government building a one-party state actually needed to remove; the mechanism that gave the electorate a chance to intervene between readings. Removing that safeguard does not tell us anything reliable about whether "the Members of Parliament," standing alone, was still understood by the drafters and the public to mean the whole House. Given that every other African constitution examined on this subject matter treats the whole-House meaning as the default that then needs to be made explicit, the more natural reading of the 1978 and 1991 drafters' silence is that they assumed the ordinary meaning still applied, not that they silently narrowed it.

The Policy argument turned around

I consider Dr. Abdulai's strongest rhetorical point to also be his weakest legal one: that a fixed threshold lets a minority "freeze" the amending power by staying away, and that this rewards walkouts over votes. But that is precisely, and by design, what every serious entrenchment clause is meant to do to at least some degree. The whole point of requiring a supermajority for constitutional change, as opposed to an ordinary majority, is to make the amending power harder to use than ordinary legislative power deliberately, and for the electorate's protection, not the government's convenience. Comparative scholarship on constitutional amendment design treats amendment difficulty as a feature to be calibrated, not an accident to be minimized. Richard Albert's work on amendment rules observes that where designers want an amendment process insulated from transient majorities, they set the threshold high enough that it cannot be met by momentum alone.¹¹ A threshold that a governing party can satisfy merely by holding 67% of whoever shows up on the day, regardless of how many seats the opposition actually holds, is not a serious check on anything. It converts a supposedly extraordinary power into an ordinary one, contingent only on the government's ability to keep enough of its own side in the room. Dr. Abdul would not agree with this proposition but I trust the Supreme Court knows that the guiding principle is not in favour of who is available. Sierra Leone's post-war constitutional settlement was built, after a civil conflict fuelled in no small part by one-party domination, on the premise that the most consequential decisions; removing a President, declaring or extending emergency powers, amending the Constitution itself, should require support that cannot be manufactured by attendance-taking. Section 89's lower quorum for ordinary business proves the opposite of what Dr. Abdulai wants it to prove: it shows that where the drafters were content with a low, presence-based bar, they built one for ordinary business. That they did not carry that same logic into Section 108 is the point, not an embarrassment to be explained away.¹²

National Prohibition Cases does not rescue the present-and-voting reading

Further, Dr. Abdulai calls the American National Prohibition Cases¹³ the single most relevant comparative authority in this debate. It is relevant, but not in the way he suggests, and it is readily distinguishable. The U.S. Supreme Court held that the two-thirds required to propose a constitutional amendment under Article V means two-thirds of the members present, assuming a quorum, relying on its earlier decision in Missouri Pacific Railway Co v Kansas.¹⁴ But here is the catch, the American Constitution has nothing resembling Sierra Leone's section 91(1). Article I already establishes that a majority of each House of the US Congress constitutes a quorum "to do business,"¹⁵ and Article V's two-thirds requirement was read against that quorum backdrop because Congress has no explicit constitutional formula anywhere in its text for "members present and voting" against which Article V's silence could be measured by negative implication. The American courts had to guess at Article V's meaning in isolation, because their Constitution gave them no internal comparator.¹⁶ Sierra Leone's courts do not have to guess. Section 91(1) supplies the missing comparator directly, in the same Constitution, in the same chapter dealing with the same institution. Where the U.S. Supreme Court had only silence to work with, our Supreme Court has an explicit textual signpost telling it exactly what "present and voting" language looks like when the drafters intend it, and section 108(2)(b) does not carry that signpost. An American holding driven by the absence of any internal comparator is not strong authority for how to read a provision that has one. Here again, Dr. Abdulai completely missed the point.

What Uganda's 2017 vote actually confirms

On this point, Dr. Abdulai is right that Uganda's Constitution spells out "all Members of Parliament," and right that the 2017 age-limit vote passed 317 to 97 even after the opposition, led by Winnie Kiiza, walked out.¹⁷ But, he draws only half the lesson from his own example. Uganda's drafters wrote "all" because they wanted a threshold impervious to exactly the tactic Sierra Leone's opposition later attempted, and it worked as designed: the walkout changed nothing, because the fixed denominator did not move with attendance. That is not an argument that unadorned "two-thirds" clauses mean something narrower elsewhere; if anything, it is a demonstration of what happens once a fixed-denominator clause meets a boycott; the boycott fails, exactly as an entrenchment clause should ensure. Sierra Leone's Supreme Court, asked to choose between a reading that makes a walkout irrelevant and one that makes a walkout decisive, should not need much persuading as to which result serves the entrenchment section 108 was designed to provide.

Where this actually leaves us as a nation

None of this means Dr. Abdulai's reading is not available on words alone, taken in isolation and without their surrounding provisions. Reasonable lawyers, including the sitting Speaker, have adopted it. But section 108(2)(b) does not sit in isolation.¹⁸ It sits one section away from section 91(1), which supplies the very phrase "present and voting" that the present-and-voting reading needs and does not have; it sits in a Constitution whose own Supreme Court has directed that clear words be given their ordinary meaning; and it governs the one class of legislative business that the framers, in the shadow of a war fuelled by the collapse of political pluralism, had every reason to make genuinely difficult to change. The missing word Dr. Abdulai spent sixteen thousand words on is real, but it is not the missing word that matters most. That word is "present," and it is missing from section 108(2)(b) precisely because the framers did not intend the threshold to move with attendance. If the Supreme Court of Sierra Leone is looking for the reading that respects both the words actually used and the reasons a supermajority clause exists at all, it is not the one Dr Abdulai defends.¹⁹