By Shadrack Muyesu

Under the title “Why a constitution cannot be unconstitutional”, the October 2016 issue carried an antithesis of my “controversial” work, “Unconstitutionality of the Constitution”, that ran in the September 2016 edition. The author, Newton Arori, questioned the idea that there exist superior norms within and without the Constitution, against which constitutional amendments ought to be gauged for constitutionality.

Citing an array of sources, Arori provided an alternative in the rationale that constitutional clauses were all complementary, with none being superior to the other. He also noted the absence of a constitutional forum within which the constitutionality of constitutional amendments could be challenged. In doing so, he referred the reader to Article 2(3) of the Constitution (of Kenya, 2010), which protects the Constitution from any challenge on its validity or legality before any court or State organ.

Finally, while accepting that unchecked, numerous amendments could result in a starkly self contradicting document, he stressed that the drafters considered a vaccine to this affliction in the requirement that, where required, proposed changes must be subjected to a referendum which should determine their adoption. Even then, he emphasised that contradictions were a natural feature of the Constitution, which courts retained the duty to reconcile and resolve. To aid his course, he cited the locus classicus on separation of powers in Marbury vs. Madison

I shall endeavour to address Arori’s worries in due earnest but before I do so, it’s only prudent I clarify a misrepresentation others like Arori, might have adopted. My work wasn’t concentrated on the Constitution of Kenya, 2010; certainly, it did not question its constitutionality. Its sole purpose was to debunk the popular yet incorrect thinking that constitutions cannot be unconstitutional. My reference to the Constitution of Kenya was merely for demonstration purposes – that indeed there existed superior norms. That said, whether a constitution is unconstitutional or not is purely a jurisprudential discourse. While positivists tend to favour the cue taken by Article 2(3), naturalists are agreeable that indeed constitutions can be unconstitutional. I shall explain in due earnest.

Legitimacy and superior norms
Can a constitution be illegitimate?
In answering this question, I shall retain the information of the widely accepted definition of a constitution as a social contract between governors and the governed. As creatures of society, constitutions betray the supervening public morality. Public morality shifts with the growing complexities of society, which phenomenon Rousseau classes, “the sentiment of their existence – the ability of human beings to change their material and psychological relations to one another” (Jean Jacques Rousseau, 1762). As I highlighted in my initial article, a constitution must be adaptable to societal evolution: an un-amendable constitution adopted by a generation long since dead can hardly be viewed as a manifestation of the consent of the governed. A constitution that isn’t progressive at the instance of the citizens is therefore illegitimate (Edmund Burke, 1984; James Madison, 1987).

According to Rousseau, a constitution is a custodian of the general will of the governed. By studying this, will we understand their general (public) morality. But do constitutions always demonstrate this public morality? Of course not. Ideally, the general will should be what the citizens of the State have decided together in a general assembly – a phenomenon common with the direct democracies of the old.

The convenience of indirect democracy, however, results in a “general will that is a transcendent incarnation of the citizens’ common interests, which exist in abstraction from what any of them really wants.” Simply put, the modern constitution cannot accommodate all interests; it is a compromise that dilutes public morality. That courts are only limitedly influenced by popular sentiment means that the constitutional interpretation of the courts – which positivists like Arori so believe in – is, too, a compromise that often dilutes and, in some cases, even erodes public morality. My conclusion, that occasionally citizens end up being governed by laws they never chose, that such laws though legally valid are illegitimate, is therefore anything but far-fetched.

I will cite a few examples. While studies have shown that the Kenyan population is largely hostile to homosexuality, the High Court has interpreted the Constitution as protecting of minority (including gay) rights (Eric Gitari vs. Non-Governmental Organisations Co-ordination Board & 4 Others [2015] eKLR). In addition, for a population traditionally hostile towards promiscuity, the law steers clear of entirely banning promiscuous acts – choosing instead to only criminalise “carnal knowledge against the order of nature”, as well as “living off the gains of prostitution”.

In Gitari, the court opined that though the Penal Code criminalises carnal knowledge against the order of nature, it does not define what carnal knowledge against the order of nature is, all but legalising gay intercourse. And here is a personal favourite: while public sentiment against atheism (and devil worship) remains loud, the Constitution emphasises the secularity of the State. This may also be observed of the two- thirds gender rule in a society that is generally patriarchal, and the clamour for decisive (punitive) presidential action against runaway corruption when the Constitution sacrificed a strong presidency at the altar of rule of law and separation of powers.

According to Rousseau, a State that defies the general will (indeed, public morality) is equally illegitimate. Similarly, a constitution cannot be legitimate if it misrepresents the intentions of the citizens. For a general will to be truly general, it must come from and apply to all, substantively and formally. For Rousseau, this can only happen in societies that are sufficiently homogenous. He asserts that, in well ordered societies (sufficiently enlightened, virtuous and agreeable to restrictions), the necessary compromise can be achieved – which really is an agreeable middle ground, a plastic general will and therefore, a simulated public morality. Unfortunately, highly socio-economic and culturally stratified and rural societies, such as Kenya, remain incapable of such a compromise.

Yet, Rousseau isn’t entirely pessimistic. As a way forward, he recommends (or notes?) a reliance on a virtuous legislator, who is to convince citizens of the need for compromise and, in the process, inculcate in them a variant public morality in the acceptance of the plastic general will (the constitution as a compromise). He is, however, clear that the legislator should not create a new morality, but tailor the existing public morality to come up with an agreeable compromise.

To formalise this new general will (the constitution), Rousseau believes that citizens must be virtuous and knowledgeable, otherwise they won’t notice it when the legislator is lying to them. For him, “good will can only be willed by good citizens”. As I adopt this position, I invite the reader, indeed, Arori, to consider the number of times legislators have misguided citizens. As rising disaffection with some of the novelties of the Constitution in an expanded Legislature and a limp Senate demonstrate, the legislator, for his own selfishness, constantly fails in his function of inspiring the required public morality – a sense of collective identity.

‘Substitution’ doctrine

To drive home the idea that constitutional clauses are equal, Arori cites the dictum of Justice Kato in “Rwanyarare and Haj Badru Wegulo vs. Attorney General (Constitutional Petition No. 5 of 1999): “…the court had no power to declare one article of the Constitution inconsistent with another, but could deal with the question as to whether or not the correct procedure was followed when the act was passed…”

Curiously, Kato seems to accept that constitutional amendments are illegitimate if they fail on procedure. Thus the question begs, what creates this procedure? Is it the Constitution? And does the Judiciary reserve the right of constitutional interpretation in cases of a conflict? If yes, was it correct for him to rule that the court had no power whatsoever to declare one part of the Constitution inconsistent with another? Of course not. And does this collapse Arori’s argument that constitutional amendments cannot be unconstitutional? In my opinion, it does!

Commendably, Arori cites quite a number of authorities to assert his claims. As well, my initial work cited authoritative rulings of its own, key among them being the dissenting opinion of Kennedy J. in State (Lemmon) vs. Ryan J., and the landmark ruling of the Supreme Court of India in Kesavanada Bharati vs. State of Kerala and Anr. I will also refer Arori to Colombia where the Constitutional Court issued two decisions striking down two bodies created by Legislative Act 1 of 2015, a constitutional reform act aimed at an extensive constitutional overhaul of the separation of powers. The court relied on the substitution of the constitution doctrine, which enabled it to strike down constitutional amendments that, in its view, went beyond merely amending the existing constitutional text to “replace” core constitutional principles. I will also refer him to Peru where the Supreme Court rejected the idea that constitutional reform is merely a political act that could not be subjected to judicial scrutiny.

In civil applications Nos. 0050-2004-AI/TC, No. 004-2005-PI/TC, No. 007-2005-PI/TC, and No. 009-2005-PI/ TC (Reform of the Pensionary System of Law” cases), the Supreme Court observed that all creatures of the Constitution, including Congress, bore a duty to maintain a fidelity to the Constitution and its values – which fidelity, alongside constitutional principles and basic democratic values, the Judiciary bore a duty to protect.


Islamic constitutionalism, which is generally inferior to modern secular constitutionalism, offers us the best avenue of appreciating unconstitutional constitutions. Sharia Law retains supremacy even where the legal system is seemingly secular – which inclines judges to favour a naturalist approach to constitutional interpretation. Where, as in the Islamic Republics of Iran and Kuwait, this supremacy is codified, the interpretation though positivistic, will still favour natural (Sharia) law.

In the latter instance, Islamic legal science, fiqh, which is generally flexible in nature, can be classed as the principle determinant of the public morality of these societies. Fiqh is itself founded on an unchanging Sharia Law, which Secondary Law (the Constitution) must be agreeable to, it hence its status as a superior constitutional norm without the Constitution. Therefore, a constitutional provision, amendment or interpretation that flies off the handle of fiqh is unconstitutional – similarly so to a constitution.

Granted, where the law allows, the public may demand a secular constitution without the burdens of fiqh, but this should be interpreted as nothing beyond an acceptable shift in public morality (see The Relationship between International Law, Islamic Law and Constitutional Law of the Islamic Republic of Iran – A Multi-layer System of Conflict, by Ramin Mostchhagi. As well, the Preamble of the Iranian Constitution, 1979, identifies the Constitution as a social construct evidencing the public morality of the Iranian people). It is curious to note that, unlike its Kenyan equivalent, the constitution of Iran lacks a constitutional supremacy clause. Article 61 is particularly instructive that the courts, in carrying out their functions, ought to be instructed by Islam (beyond the Constitution).

By way of summarising this discourse, constitutions can be illegitimate if they misrepresent the nature of the society they are meant to govern. I have also demonstrated the possibility of superior clauses within the constitution and superior norms without the constitution – the latter classed as “constitutions without constitutions”.

Finally, I also demonstrated that constitutional amendments can be unconstitutional if they are arrived at unprocedurally, or if they defy superior norms within and without the constitution, and that constitutions can be unconstitutional, where they laden with inconsistencies or stand in defiance of superior norms.

But is this unconstitutionality justiciable? Well, this depends on the jurisdiction: in parliamentary supremacies and natural law states, this is very much possible. ^