Sarthak Sahoo*

The Constitution of India is routinely amended. Occasionally, a constitutional court will render such amendments unconstitutional. In that case, if the amendment purported to change the text of the Constitution, does its invalidation ‘revive’ the provision as it originally stood? This article argues in the affirmative. It argues that the case law as it stands does not unequivocally rule on this point, and to the extent it does, should be overruled. It collapses the distinction between statutory revival and constitutional revival, as well as judicial invalidation and Parliamentary repeal, to support the constitutional revival thesis. Finally, it employs an abductive analysis of the Constitution’s amendment method to establish its revivalist reality.

The Supreme Court of India has currently reserved judgment in Property Owners Association v State of Maharashtra (2024), a case that asks whether a constitutional provision’s original form, when amended, gets revived if the amendment is invalidated.

I will not be dealing with the particular facts of the case. Interested readers may refer to Neha Vinod’s piece at the Supreme Court Observer, which traces an illuminating history of case law on this point, as well as addresses the relevant factual detail. She concludes that the question largely depends on a practical consideration: whether the non-revival of the original provision would create constitutional chaos? I refer to this as the ‘consequence’ test.

In this blog, I argue that this should not be the test to apply. Instead, the Court should affirm the doctrine of revival for constitutional provisions wholesale. The structure of this post is as follows. Firstly, I provide a precis of Vinod’s argument, and the case law it relies on. Secondly, I seek to make certain clarifications with reference to revival as it applies to differently to constitutions and statutes; and by Parliament or judicial review. Thirdly, I argue that the doctrine of revival must apply to constitutional provisions for constitutional, statutory, and conceptual reasons. Lastly, I apply Professor Richard Albert’s framework of constitutional change to support the argument that India’s amending style entails a doctrine of revival.

As It Lies

Vinod illustrates that the doctrine of revival has been affirmed by the Court in Ammer-un-Nissa Begum v Mahboob Begum (1952) and State of Maharashtra v Central Provinces Manganese Ore Co (1976). Interestingly, Central Provinces differs from Firm ATB Mehtab Majid & Co v State of Madras (1962) and BN Tiwari v Union of India (1962), both cases of coordinate strength. The Firm ATB negated the doctrine by stating that amendments entailed a two-step substitution process:

1. removing the original provision; and
2. inserting the new provision.

This would mean that invalidation would only remove the latter step, thereby not reviving the original provision.  It is this theory of substitution that Central Provinces argues against, holding it to be a unified process, thereby implying that invalidation restores status quo ante.  This is also supported by Waman Rao v Union of India (1980), which held that Article 31C was revived in its original form when an amendment to it was struck down in Minerva Mills v Union of India (1980).

Observing this doctrinal confusion, Vinod instead points to Indian Express Newspapers (Bombay) v Union of India (1984) which held, albeit loosely, that this question depends on the ‘totality of circumstances and the context in which they are used.’ Khehar J has since developed this reasoning further in the NJAC (2015) case referring to three factors to consider – Parliament’s intention, the amendment’s affect in toto, and the effect of striking down the amendment. Kurien J, agrees not with the factors, but the conclusion on grounds of necessity. Interestingly, Waman Rao, by preserving Article 31C as it stood post-Kesavananda but pre-Minerva, validates revival doctrine applies. Being a larger bench than NJAC, it may raise issues with the latter’s validity.

Nonetheless, the view of NJAC can be called governing as it is not, strictly speaking, incompatible with either theories of revival or non-revival. One can easily retrofit Article 31C as one of those cases which would passes the revival test under Khehar J’s framework.

I concede this position for the sake of argument. However, given that Property Owners Association is a nine-judge bench, I argue that the Court should overrule this particular view by a smaller bench, especially given that this is nowhere near stare decisis consensus.

The Two Dualities

At the outset, two conceptual dichotomies must be dealt with. Firstly, except for Waman Rao and NJAC, other cases often relate to amendments to statutes and not constitutional provisions. Therefore, their application to constitutional provisions is suspect. Secondly, there exists a distinction between Parliament repealing a law and a court striking it down. Constitutional provisions retain the amended text even after invalidation, whereas statutes cease to do so. Therefore, Parliament must make the effort of revival in the case of the constitution. In my view, both of these objections fail.

The distinction between cases governing statute revival and constitutional revival is a fair one. Even if all statute cases supported the theory of revival (albeit they don’t), it would be perfectly legal for a smaller bench to hold in relation to a constitutional provision that they do not apply as the ratio in those cases concern a different type legal instrument. However, the argument does not cease there – as this is merely a descriptive question of form. The next question is whether there is anything that justifies having different rules of revival between constitutional provisions and statutes.

The answer, emphatically, is no. Firstly, although both types of amendments are sourced from different powers – the constituent and constituted power, and have different voting requirements, they are put into force using the same mechanism – an Act of Parliament. Therefore, no distinction can be drawn between them as a matter of mechanics. If ‘substitution’ is a unitary or a two-step process – it applies similarly to constitutions and statutes in principle.

Understandably, some might view this as a redux of Golaknath, where the Court infamously equivocates and equalises the legislative character of Parliament with its constituent character – mostly on tenuous procedural similarities between ordinary law and constitutional amendments. See Krishnaswamy at pp 7 to 9.  However, my argument echoes Kesavananda in recognising that a constitutional amendment is sourced from a superior constituent power, one that cannot be subject to the constituted one. Instead, my argument pertains to the Court’s function of judicial reviewing these different types of legislative outputs. In this behalf, revival in a purely procedural question, and the fact that substantively, whether a change is arising from the constituent or constituted power does not affect the Court’s act of judicially reviewing its constitutionality. This function of the remains entirely the same on procedural grounds.

On the second question, the distinction between judicial invalidation and Parliamentary repeal is similarly placed. India’s constitutional structure is such that judicial invalidation operates against the Act of Parliament that amends the Constitution, much like how it functionally would to an ordinary statute (as proved above).

The Court’s decisions do not change (or delete) the phrasing of constitutional provisions because the Court does not ‘amend’. See e.g., para 7 of Schedule X, which although declared unconstitutional, remains in the constitution effectless. The Court deals not with the field of the Constitution, but rather of the underlying statute. Therefore, when the Court declares a constitutional amendment invalid, it is doing the same thing that would happen to a statute on repeal – it turns void.

(The only different between these two actions may be their retrospective validity. Repealed statutes are usually invalid thereafter; whereas judicially invalid statutes are void ab initio save for prospective overruling.)

The conclusion to be drawn, here, is that the doctrine of revival should, as a matter of legal policy and constitutional coherence, apply identically to cases of statutory amendments and constitutional amendments; and irrespective of whether the source of this Parliamentary repeal or judicial invalidation is. Of course, any constitutional amendment to Article 368 or 13 may change this position, but that is not the case at present.

To note, even if this position is not accepted, the rest of my argument remains agnostic to this equivalence – and deals with revival’s argument independently, mostly for constitutional provisions, but can also sometimes be extended to statutory ones.

Walking the Dead

The strongest argument against revival lies in believing that amendments operate in two steps: (1) removing the original provision and then (2) inserting the new one. This is wrong on three levels – constitutionally, statutorily, and conceptually. However, their reasoning overlaps considerably.

Article 368, which enables constitutional amendments, provides this ‘by way of addition, variation or repeal’. On the constitutional level, this two-step model of amendment vitiates the distinction between a ‘variation’ type amendment and ‘repeal followed by addition’ amendments. It would render the constitutional power of variation as a superfluous feature, as anything it could do, could instead be done by repeal and addition. This view is attested to by the case law (para 100) of the Court, which provides that the Constitution should be interpreted in such a manner that does not render its terms redundant. Contrarily, the plain meaning of ‘variation’ appeals to the fact that the text is changed and not removed and then something else put into its place.

On the statutory level, constitutional amendments, when varying provisions, employ the following phrasing (For instance, see Section 3 of the 101st CAA):

In the [provision] of the Constitution, for the words “ABC”, the words “XYZ” shall be substituted.

Right away, the use of the word ‘substitute’ cannot possible be read as infracting the meaning of ‘variation’, as the latter being the parent power of the former. As for use of ‘substitute’ in the first place, it may be noted that such amendments use ‘inserted’ and ‘omitted’ to operationalise ‘addition’ and ‘repeal’ in their constitutional form as well. See Sections 7 and 8 ibid.

At this point even if the two-step process was true, given that this process of ‘substitution’ occurs in the same provision of the Act with no phrasal distinction between removal and addition, no severability can be had between the two phenomena.

On the conceptual level, the two-step process implies that there can exist a purgatory between a law’s amendment. It implies that the subliminal space between a phrase being removed and another being added, there can exist a state of limbo. This, however, is simply untrue, as operating upon Presidential assent, variation amendments do not have a temporal dimension, as the passage and assent of repealing and adding amendment acts put together might.

(To illustrate, if statute A removes a provision and statute B inserts something in its place; and the President assented (and notified) to statute B before A, such amendment would have no legal effect per se, for it would be based on an impossibility. This shows a temporal dimension between statues that does not exist when dealing with these two processes being put into one provision. The temporal dimension between two provisions in the same statute is unclear to me.)

From Form to Function

Whereas the above arguments inevitably conclude that revival occurs, I now point to a theoretical constitutional practice that supports this view. Prof. Richard Albert, in The Architecture of Constitutional Amendments (2023) distinguishes amendments to constitutions in four forms. See Albert at pp. 5 to 11.

1. Appendative: Where amendments, no matter how they change the existing constitution, get added at the end of the constitution. For example, the United States Constitution.
2. Integrative: Where amendments, when they make a change, are accompanied by information about the character of the change (say, in a footnote). For example, the Indian Constitution.
3. Invisible: Where amendments, when they make a change, simply do so, with the post facto constitution leaving no trace of the previous one. For example, the Portuguese Constitution.
4. Disaggregate: Where amendments, when they make a change, are not placed in a single legal instrument, and hence don’t affect their presentation in any way. For example, the United Kingdom’s political constitution.

Here, it is important to caveat that the fact that India uses the integrative method has no normative value. This is because the aforementioned taxonomy is descriptive. See Cunha in Albert at p. 76.  However, the use of the integrative method is not wholly irrelevant to the question of how constitutional amendments occur.

There are two inferences that can be drawn from the use of this method. First, that the Constitution continues to account for a theoretical position ex ante where the amendment in question had not occurred. This is unlike the invisible model, which fails to do so; and unlike the disaggregated and appendative models, which sidestep the act of variation altogether. Second, that the Constitution of India is not sui generis in the way it deals with amendments to it, as compared to statutes.

The second inference is useful for obvious reasons. It lends credit to my prior thesis that if statutory provisions do revive upon being rendered unconstitutional, this continues to apply to the Constitution. 

The first inference deserves some consideration. It is crucial to note that legal propositions may not just be true by deductive reasoning. They may also be achieved by abductive reasoning, which involves the examination of certain facts, and thereafter coming to conclusions about what legal conditions give rise to that fact. As Prof. Bjarte Askeland notes, abductive legal reasoning ‘aims at inferring which rule a given fact can be subordinated to within the realm of legal norms.’ [emphasis original] A strong example of this is purposive interpretation of statutes, where a given factual posture is categorised compatible or incompatible with legal norms only the degree to which the severity of their harm was supposed to be covered by norm itself.

In this case, is trivially true that while the integrative method could be used in a Constitution that does not expect revival, the use of such method would be abductively arbitrary and useless, as it would not be subordinate to any legal norm that gives rise to its condition. On the other hand, the integrative method would make perfect sense in a Constitution that can turn back the clock on its provisions. Therefore, on practical grounds, this serves as evidence that Constitutional practice supports revival.

Conclusion

There are two views on the removal of an amendment. The first, that an amendment that varies an existing provision or statute that replaces another is a unified process. Hence, when that amendment is invalidated, the old provision or statute revives. The second, that it consists of two parts: removal of the old, insertion of the old.

This blog has firstly established that the consequence test’s governing value notwithstanding, the Property Owners Association Court can make an original determination on the question of revival. It has secondly established that revival should operate (or not operate) in the same way in case of either constitution or statute; be it via Parliamentary repeal or judicial invalidation. It has thirdly established that for the constitutional amendment, a two-step process is incoherent due to constitutional, statutory, and conceptual reasons. Lastly, it has demonstrated that the unified theory has support from India’s constitutional amendment practices in light of Prof. Albert’s framework of constitutional change. 

*The author is a student of law at the Rajiv Gandhi National University of Law. He holds interests in constitutional law, public international law, and legal theory. The author would like to thank Pranav Aggarwal for his comments on the original draft.