Whose Vote is it Really? The Article 121 Conundrum in the Tata Boardroom

I. Introduction

Once again, Bombay House¹ found itself in the eye of a storm on 17 September 2026. The six-member board of Tata Sons voted by a majority to (re)appoint N Chandrasekharan as chairman for a third 5-year term. The lone dissenter was Noel Tata, one of two nominee directors representing the Tata Trusts, which control 66 per cent of Tata Sons. Further, N Chandrasekaran recused himself after disclosing his conflict of interest, and Harish Manwani took his place as acting chairman. While the Secretarial Standards (SS-1) describe an ‘acting’ chairman as a ‘chairman of the meeting’, we will use the terms interchangeably. It was in this capacity that Manwani cast the fateful vote that appointed Chandrasekaran for a third term.

The primary issue revolves around the interpretation of Article 121 of the Articles of Association (AoA) of Tata Sons, which requires both a majority of the board and a majority of the nominee directors to pass resolutions, subject to the chairman’s casting vote to resolve an ‘equality of votes’. Nominee directors are appointed specifically by the two Tata Trusts under Article 104B as long as they hold up to 40 per cent of Tata Sons.²

The key question is whether a deadlock amongst the nominee directors can be broken by the chairman’s casting vote? In brief, we argue that the most persuasive interpretation of Article 121 permits the chairman to break ties between nominee directors, and that ‘chairman’ includes an ‘acting chairman’. Part II of this piece explores the applicability of the casting vote, Part III examines Manwani’s capacity to cast it, and Part IV concludes.

II. Casting Vote in Affirmative Voting Rights

The preliminary question is whether Article 121 is the AoA’s sole decision-making clause, or a special clause setting out affirmative voting rights for nominee directors? Umakanth Varottil notes that the Article either proposes a ‘double-barrel majority requirement’—prescribing both general and special decision-making rules—or its formulation reflects an ‘if/then’ structure, i.e., where a threshold condition is met, affirmative voting rights apply. He argues that the semantic structure of the clause favours the if/then reading. To him, the use of ‘which’ in Article 121 at least minimally indicates that matters requiring a majority is a threshold condition for activating affirmative voting rights of nominee directors. An implication of the if/then reading, as Varottil notes, is that the ‘sole’ purpose of the Article pertains to setting up the affirmative voting rights mechanism. This implies that the casting vote mechanism applies squarely to this sole purpose.

However, we can potentially infer a contradictory reading from Tata Consultancy Services Limited v Cyrus Investments Private Limited (2021). Describing affirmative voting rights, the Supreme Court of India (“SCI”) observed that Article 121 ‘speaks only about the manner in which matters before any meeting of the Board shall be decided.’³ The court’s use of ‘any’ (mirroring the language in Article 121) supports the inference that Article 121 is the primary provision for decision-making procedure.⁴  This is bolstered by the Article’s heading ‘Matters How Decided’, and its reading with Article 121A, which says, ‘the following matters shall be resolved upon by the Board…’ Crucially, Article 121 is not titled ‘Matters which require majority, How Decided’, indicating that the ‘double-barrel majority’ reading is the stronger reading of the AoA more broadly.

A second question is whether a disagreement between the nominee directors constitutes an ‘equality of votes’ in the first place? The Concise Law Dictionary defines ‘equality’ as ‘[a] word which conveys the idea of identity in size and form’.⁵ While it doesn’t define ‘equality of votes’, this sufficiently indicates two threshold conditions: one, both sides must record an ‘identity’ of votes in size; two, the requirement for passing the resolution must be a ‘majority’, i.e., one side should have a greater number of votes. Under Article 121, both conditions are fulfilled for both ‘votes of the board’ and ‘votes of the nominee directors’.⁶  A deadlock between an even number of nominee directors is indistinguishable from a deadlock between an even number of directors generally. Therefore, an ‘equality of votes’ exists in this case.

Varottil argues that under a ‘double barrel’ reading, Noel Tata’s position that the casting vote rule applies to the first prong (full board majority) but not the second (special majority amongst nominee directors) is plausible. We disagree because it is plausible to interpret the casting rule as applicable to both prongs for a few reasons. Article 121 is silent on whether it applies to one or the other. It merely stipulates that the chairman is empowered to act ‘in the case of an equality of votes’ (this argument was first made by Sudipto Sarkar in his pre-meeting opinion to Tata Sons, found here). On a plain grammatical reading of the text, if we accept that an equality of votes can arise in both prongs, then the singular device of the casting vote resolves both.⁷

Furthermore, the Trusts’ position would require reading in certain implied terms (specifically, the words ‘an equality of votes of the Board’). In Tata Consultancy, the SCI deemed the AoA to have the status of a contract, which meant that the courts lacked the power to rewrite it by judicial fiat.⁸ However, Indian contract law does recognise judicial capacity for reading in implied terms. In Nabha Power Ltd v Punjab State Power Corporation Limited (2017), the SCI noted that ordinarily courts must read contracts as per their ‘express terms’.⁹ Implied terms therefore are an exception, for when a five-condition test is satisfied. Condition (2) is that of necessity, i.e., implied terms must not be read in if ‘the contract is effective without it’. While Condition (5) holds that the implied term cannot ‘contradict any express term…’¹⁰

How do we interpret effectiveness in Condition 2? It is possible to argue that the express term is effective, to the extent that there are no operational impediments to executing it.  Commentators counter-argue that the express term inverts the protection granted to nominee directors under Article 121 read with 104B, destroying the purpose of the clause and allowing an errant board to appoint a chairman despite the owners’ objections through a single defecting nominee. This would violate the object of the AoA itself. However, the lack of operational impediments by itself does not satisfy the business efficacy standard. The SCI, in Nabha, held that courts should look beyond operability to whether it frustrates the business purpose or not.¹¹

In response, we would like to make three arguments. First, the purpose of Article 121 remains intact under the broad interpretation. In Tata Consultancy, the court noted that with a 66 per cent stake in Tata Sons, the Trusts are well-positioned to pack the board. The explicit choice to require the minimum number of nominee directors to be at least one-third of the board strength as opposed to packing the board with nominee directors (through their 66 per cent stake in Tata Sons), along with non-nominee directors being allowed to be chairman, indicates that the Articles never sought the board to be subservient to the Trusts.¹² Furthermore, the court also held that one must balance the dual-responsibility of the director, both to the company (under Section 166(2) of the Companies Act 2013) and to the nominating trust.¹³ The manner in which the nominator’s interest and company’s interest are to be weighed and balanced is unclear in Tata Consultancy and prior judgments of the SCI and Bombay High Court have taken differing stances.

The broad interpretation therefore does not affect the protection from operating per se; it merely refuses to grant a veto arbitrarily (i.e., individual nominee directors have a veto when their number is two and don’t otherwise). Further, the characterisation that a single nominee director must defect to override the owner’s decision overlooks the requirement that the board would need at least 50 per cent or more of the nominees (depending on whether the number of nominee directors are odd or even) to override the shareholders directive.¹⁴

Second, the AoA and the Companies Act 2013 have alternate remedies, which protect the Trusts’ ownership interest. The Trusts may replace any nominee director under Article 104B if they are displeased with their balancing of company and trust interests. A prerequisite for being a chairman is to be a director. Therefore, the majority may—even on occasion where a chairman is selected through an ‘errant board’—prevent such person from being reappointed as director or even remove them from the position of director entirely, thus vitiating their chairmanship. Third, the business efficacy test applies to the broad and narrow readings of Article 121. Applying the test to the narrow reading subverts the objective of the 2014 amendment to the Article, which swapped ‘unanimity’ amongst nominee directors for a ‘majority’.

Regarding condition (5), if we accept the Trusts’ reading, which requires adding three words to Article 121, we narrow the broader interpretation. This implies that the power to break ties between nominee directors must be stripped from the Tata Sons board. Additionally, if we take the broad interpretation as the grammatical baseline (or plain reading), the implied terms will clearly contradict the express terms. Therefore, the justification for reading in implied terms is quite limited.

Both factors point to strong reasons for disagreeing with Varottil’s conclusion that accepting the double-barrel view supports the Trusts’ view. But what is a harmonious interpretation of Article 121? Considering the arguments made above, we propose a single-ballot model for Article 121. This model envisages a single ballot that must satisfy two majority requirements for success, crucially without requiring two separate votes. This implies that the chairman’s role is to ascertain the bona fides of the votes and determine if the dual majority has been met; if not, they get to exercise a second casting vote to break an ‘equality of votes’. Viewed in this light, one can see that Article 121 is modelled on Regulation 68 (Table F under Schedule I of the Companies Act), which likewise does not distinguish between a ‘majority of votes’ in the board generally from a special majority in a protected sub-class. It ensures that where an equality of votes arises and AoAs do not account for it, the chairman breaks it.

III. Chairman v Acting Chairman? – The Capacity Question

To determine whether the casting vote would lie with Manwani in his position as acting chairman, one must examine the position of the chairman. The Companies Act envisages certain circumstances where the office of the chairman is not filled, and by extension, no casting vote exists. A combined reading of Regulations 45 and 68 of the Model Articles of Association (Table F of Schedule I of the Companies Act 2013)  and the definition of chairman under SS-1 raises two issues. First, SS-1 seeks to distinguish between chairman of the board and chairman elected for a meeting. Since the Model Articles are silent on the definition of a chairman, we extend this distinction into the Model Articles, which grant the casting vote only to the chairperson of the board.¹⁵ We argue that if Regulation 68 instead used only the phrase ‘chairperson’ without the qualifying words ‘of the Board’ subsequently, then it would include the chairperson of a meeting being granted a casting vote. The Model Articles are of importance here because paragraph 5 of SS-1 makes casting vote powers subject to the Articles.¹⁶ Therefore, where the Model Articles do not use the standalone word ‘chairperson’ but instead specify ‘Chairperson of the Board’, it must be interpreted as the Model Articles providing for otherwise and taking priority per SS-1.

Second, the use of the term ‘if any’ in Regulations 45 and 68 indicates that a chairperson’s office need not always be filled. We adopt a reading that treats the chairman as an “office” independent of the person. Regulation 70(ii) provides that where the chairperson’s office is vacant, or the chairperson is absent, the directors may choose one of the present directors to be the chairperson of the meeting. The use of the word ‘may’ rather than ‘shall’ in this clause further bolsters the argument that the Articles do not compel the appointment of a chairperson. While paragraph 5.1.2 of SS-1 requires that directors shall elect one of themselves to the chair, it makes it subject to the Articles. As the Model Articles do not mandate the appointment of a chairperson of the meeting, the same would take priority. Although it is unheard of in practice that a board meeting is conducted without either the chairperson of the board or of the meeting, a technical reading of the Model Articles suggests such a possibility.

However, given that Article 120 of the AoA of Tata Sons uses the term ‘shall’ rather than ‘may’, once Chandrasekaran recused himself, they were under an obligation to select a chairman for the meeting to vote on his re-appointment.¹⁷  Once Manwani was selected as the chairman for the meeting, the casting vote extended to him since Article 121 uses the word ‘chairman’ without any qualifier. SS-1 applies here as the Tata Sons AoA does not define ‘chairman’ within the scope of board meetings.¹⁸

Taking note of the distinction between ‘chairman of the board’ and ‘chairman of the meeting’, Tata Trusts could potentially argue that even if the casting vote applies to nominee directors, the ‘chairman of the board’ is uniquely empowered to exercise that vote per the AoA. This stems from the special appointment procedure outlined in Article 118, which empowers the Trusts to play an outsized role in the appointment of the chairman. Article 120 of the AoA does not provide for such involvement in the selection of an acting chairman, simply requiring the present directors to choose one among themselves without guidance on how that choice is to be made.¹⁹ In essence, if the ‘chairman of the board’ is conflicted out, deadlocks should not be resolved. This adds a second layer of protection for the Trusts’ interests.

We disagree with this inference, while conceding that Article 118 is the special provision governing chairman appointments (which is the current case).²⁰ As K Satish Kumar argues, the principle of generalia specialibus non derogant, subjects the ‘general’ casting vote mechanism in Article 121 to the special appointments provision in Article 118. We argue that Kumar’s opinion is weakened by the text of Article 118, which envisages a two-stage appointment process: a selection committee (whose composition the article determines) must recommend a name, and, more importantly, that individual must be formally appointed in accordance with the process prescribed by Article 121. Therefore, to apply the generalia specialibus principle is improper, because Article 118 specifically invokes compliance with Article 121 as a prerequisite. It also means that while the procedure might be generally applicable to all board decisions (including matters listed in Article 121A), it is the specific procedure relevant to carry out the object of Article 118.²¹

There’s a deeper principle that the Trusts would rely on, i.e., that the law contemplates a material difference between the two categories of ‘chairman(s)’ described above. While our argument is based on Tata Sons AoA on board meetings lacking a definition of ‘chairman’, we offer a second argument to tackle this deeper principle. Consider two hypotheticals under the Trusts’ position: one, where the chairman is conflicted out under the Act or the AoA, and two, where the chairman is unable to attend the meeting because of an illness or other prior commitments. In the first scenario, while the question of whether Chandrasekaran was legally required to recuse himself is contestable (see here), we will assume that he acted in good faith.

If the Trusts’ interpretation is accepted, then the acting chairman in both scenarios lacks the power to break ties with a casting vote. However, it is crucial to consider the temporal difference between the two hypotheticals. In the former, the inability to break the deadlock is permanent, i.e., the chairman’s conflict does not disappear with time. Rather, a material change is required (like the chairman deciding not to seek another term) for the conflict to disappear. In the latter, the inability is temporary, i.e., the chairman’s conflict will disappear with time. Their health improves, or they return from travel and so on. This implies that, superficially, though the acting chairman is similarly restrained in both cases, in one of them deadlock is the outcome, by design.

In the absence of a meaningful distinction between these types of resolutions, one is hard-pressed to justify the difference between the outcomes. Consider a third hypothetical: if the resolution concerns listing Tata Sons in pursuance of the Reserve Bank of India’s (RBI) mandate, then Chandrasekaran would not be conflicted out and could resolve any equality of votes. How is a resolution to list Tata Sons distinct from, say, one to acquire a battery manufacturer where the chairman is an investor, or one that pertains to the appointment of a chairman? They differ, but only trivially, because their difference does not require different voting outcomes, and if it does, it is reasonable to presume that the AoA would provide for it expressly. If anything, Article 121A read with Article 118 shows that all three types of resolutions are likely to be ‘resolved upon’ subject to the procedure outlined in Article 121.

IV. Conclusion

Commentators and retained counsel alike have framed the dispute in grandiose terms. Harish Salve characterised it as an attempt to modernise an ‘antiquated’, ‘feudal trusteeship’. In other words, this was a profound and concerning oddity at the heart of India’s modern corporate landscape. In response, Anup Salve invoked the ethos of ‘corporate constitutionalism’, emphasising the need to focus on the founding documents of Tata Sons. In this piece, we have sought to provide a persuasive answer to the central question of the scope and extent of Article 121 by carefully studying the ‘austere’ question (as Anup describes it). We believe that while Tata Sons might have a plausible case, it might be more appropriate to amend Article 121 for absolute clarity.

1. The head office of the Tata Group.[↩]
2. See Tata Sons Articles of Association, p 26.[↩]
3. [2021] 9 SCC 449 [19.25].[↩]
4. It is important to note that the court does not seek to answer this question, rather this is an inference we make, building upon the court’s acceptance of Article 121’s language.[↩]
5. See P Ramanatha Aiyar, Concise Law Dictionary (7th Edn, Lexis Nexis 2022) 510.[↩]
6. The 2014 Amendment to the AoA specifically requires a ‘majority’ of nominee directors to vote in favour. See (n 2) p 31.[↩]
7. If anything, applying a canon of interpretation like noscitur a sociis can assist us in concluding that the casting vote requirement applies more clearly to the special majority requirement—which immediately precedes it—as opposed to the general majority requirement, which is framed as conditional. This is perhaps why Varottil supports the if/then reading of the clause as more natural when considering only its semantic structure.[↩]
8. See Tata Consultancy (n 3) [18.5].[↩]
9. [2018] 11 SCC 508 [72].[↩]
10. Nabha (n 9) [43], quoting BP Refinery (Westernport) PTY Limited v President, Councillors and Ratepayers of the Shire of Hastings [1977] 180 CLR 266.[↩]
11. Nabha (n 9) [45].[↩]
12. See Tata Consultancy (n 3) [19.25].[↩]
13. ibid.[↩]
14. While Tata Trusts would be accurate in noting that the 50 per cent is only one nominee in this case, that’s particular to the current situation. The court must examine the issue at a structural level, otherwise the 2014 amendment, which reduced a ‘unanimity’ requirement to a ‘majority’ one would be nullified because of a specific number of current nominee directors. If anything, that’s the argument we highlight when we discuss the problem of an arbitrary veto.[↩]
15. We use the term ‘chairperson’ while analysing the model articles under the Companies Act, 2013 as that is the terminology used. SS-1 and the Tata Sons AoA both use the term ‘chairman’ which we have used throughout the article.[↩]
16. See Institute of Company Secretaries of India, Secretarial Standards on Meetings of the Board of Directors (SS–1) (2024) 16.[↩]
17. See (n 2), p 30.[↩]
18. Even if one were to consider the description of chairman under Article 88 of the Tata Sons AoA, it includes the chairman of the meeting within its scope.[↩]
19. Or any other procedure contemplated in the AoA or not at all as noted above in accordance with the Model Articles.[↩]
20. It must be noted that other commentators (found here and here) have discussed the distinction between an appointment and a re-appointment, and whether current resolution can really be a case of re-appointment or not. While we acknowledge the question, we think it is beyond the scope of the current essay to unpack the nature of appointment under consideration. Instead, when we refer to the principle of principle of generalia specialibus non derogant, vis-a-vis Articles 121 and 118, we are assuming that it pertains to a question of ‘appointment’ for the purposes of examining the powers and functions of an acting chairman under Article 121.[↩]
21. In what can best be described as a drafting quirk, Article 118, as quoted by the SCI in Tata Consultancy, actually mis-described Article 121. The court says (paragraph 16.33), ‘For the purpose of selecting a new Chairman of the Board of Directors…a Selection Committee shall be constituted…to recommend the appointment of a person…and the Board may appoint the person so recommended as the Chairman of the Board of Directors, subject to Article 121 which requires the affirmative vote of all Directors appointed pursuant to Article 104B’ (emphasis supplied). Clearly this does not capture the 2014 amendment to Article 121. While interesting to note, it does not affect the interpretation in this case because of the phrase ‘subject to’ being added by the draftspersons. Even if Tata Trusts were to argue that as per principle of generalia specialibus non derogant, Article 118 required unanimity, and for all other matters, a majority of nominee directors sufficed as per Article 121, it would not be a viable position. This is because Article 118’s appointment is made ‘subject to’ the procedure outlined in Article 121. The remainder of the clause—‘which requires…’—only (re)describes the content of Article 121 (as indicated by the use of ‘which’, denoting that the ensuing text is merely Article 121 repeated) and does not intend to modify it in any meaningful manner. Therefore, simply applying Occam’s Razor, it’s far more likely to be a drafting oversight than a special carve-out for certain matters.[↩]