In Benarsi Krishna v Karmayogi Shelters, the Supreme Court has decided that the word “party” in section 34 of the Arbitration and Conciliation Act, 1996, does not include a party’s agent. This, it is respectfully submitted, is incorrect or, at best, too widely stated. The important practical consequence of this proposition is that the period of limitation does not begin to run from the date of service on counsel. Since it is well-known that a Court has no power to condone a delay beyond the limit imposed by section 34 and its proviso, the exact date on which the period of limitation begins to run is of immense significance in arbitration law.
First, the facts: the claimant in the arbitration instituted proceedings for the breach of a collaboration agreement and obtained a successful award from a single arbitrator. This award was served on counsel for the respondent on 13 May, 2004. An application was filed to set aside this award on 3 February, 2005—plainly time-barred, if the date of receipt of the award was the 13th of May. Accordingly, a single judge of the Delhi High Court dismissed the petition. The Division Bench set aside this order, relying on the judgment of the Supreme Court in Union of India v Tecco Trichy Engineers, on the basis that service of the award had not been properly effected.
In considering this problem, it is important to carefully distinguish between two arguments: first, that the word “party” in section 34 excludes agents; and second, counsel has neither actual nor apparent authority to accept service. The first is a point of statutory construction, but the second calls for the application of well-known (if contentious) principles of the law of agency. The Supreme Court has, with respect unfortunately, accepted the first submission. In other words, it has held that the word “party” is defined as one who is party to an arbitration agreement and, as a matter of construction, does not include counsel. The following observations of the Court should be studied closely:
The expression "party" has been amply dealt with in Tecco Trechy Engineer's case (supra) and also in ARK Builders Pvt. Ltd.'s case (supra), referred to hereinabove. It is one thing for an Advocate to act and plead on behalf of a party in a proceeding and it is another for an Advocate to act as the party himself. The expression "party", as defined in Section 2(h) of the 1996 Act, clearly indicates a person who is a party to an arbitration agreement. The said definition is not qualified in any way so as to include the agent of the party to such agreement. Any reference, therefore, made in Section 31(5) and Section 34(2) of the 1996 Act can only mean the party himself and not his or her agent, or Advocate empowered to act on the basis of a Vakalatnama.
With respect, it is submitted that this conclusion contains two errors. The first is the contrast between an Advocate acting “for” the party and an Advocate acting “as” the party. This is a distinction without a difference unless one concludes that the word “party” in section 34 contemplates personal service—which was the issue before the Court. In other words, the first reason cannot be a reason for the conclusion that the word “party” excludes agents: it begs the question. The second reason given is that the word “party” is not qualified by Parliament to exclude agents. This, with respect, is a questionable proposition of law: the general rule of law is that a principal is bound by the acts of an agent and, with in relation to notice requirements, has been codified in the Companies Act and the Code of Civil Procedure. It is difficult to imagine that the Court intended every use of the word “party” in the Arbitration Act to refer to the party excluding its agents and yet the language in which its conclusion is expressed makes it difficult to resist this inference. Nor is this a surprising rule: when banks, for example, pay our electricity bill in accordance with a standing instruction, our debt to the Electricity Department is discharged because the law of agency treats the bank as our agent in relation to third parties, although it is our debtor with respect to the money it holds. For the same reason, a payment to our bank discharges a debt owed by any third party to us: the bank is our collecting agent. If a statute used the word “party” or “person” and defined certain legal consequences, it is therefore difficult to suppose that the word was intended to exclude agents.
In addition, one is bound to ask: what of legal entities? It is well-established law that a company acts through the deeds of human beings some of which are treated, by primary or secondary or other rules of attribution, as the acts of the company. In an outstanding judgment in Meridian Global, Lord Hoffmann explained that it is therefore misleading to talk of the company in anthropomorphic terms: the correct analysis is that the acts of certain persons are treated as the acts of the company by virtue of rules of law. These rules of law include rules of attribution and rules of agency. It is, in other words, impossible for a company to act (in the eyes of law) except through the acts of human beings whose acts are, by virtue of applicable rules of law, treated as its acts. It is difficult to reconcile this with the Supreme Court’s conclusion that service on the “party” excludes service on its agents. No doubt it will be suggested that there is a difference between “external agents” (like lawyers) and the company’s own agents (like the legal manager or CEO). That suggestion would be incorrect, because there is no difference at all in the eye of the law between external agents and internal agents: both are agents, albeit constituted differently and with different levels of authority.
The question, ultimately, is not whether the word “party” can include agents, for it plainly does, but whether the agent had authority to accept service. The authority of solicitors and counsel has always proved troublesome, generally in the context of settlement: in Waugh v HB Clifford and Sons  1 Ch 374, the defendant builder, who had instructed solicitors to settle a dispute with his customers by purchasing their houses, withdrew those instructions and told them not to settle. Unfortunately, this information did not reach the solicitor handling the case until after he had (subject to the question of authority) concluded a binding contract of settlement. Brightman LJ held that a solicitor have apparent authority to settle a dispute provided the terms of the settlement do not involve anything “collateral” to the dispute for which he was instructed. This, with one exception, echoes the analysis of a leading Indian decision: Surendra Nath v Tarubala Basi AIR 1930 PC 158, where the Privy Council held that a counsel has implied and apparent authority (arising from knowledge of implied authority) to compromise suits, but expressly declined to rule on whether this is the case where the agency is created by a written instrument, such as a vakalatnama.
The important question in this case—which was unfortunately not decided—was the scope of a counsel’s authority to accept service of an award. The question falls to be decided by asking the two usual questions: was there actual express/implied authority? If so, the matter ends there. If not, was there apparent authority? This would ordinarily arise from the existence of implied authority but would exist even if the implied authority did not exist in the particular case (for example, because of a prohibition not communicated to a third party). The added complication is the question of whether the claimant in the arbitration is entitled to ostensible authority with respect to an award sent to the defendant’s counsel by the arbitral tribunal, and whether authority, if any, exists after the arbitration is concluded. The Court records that one of these contentions was raised, but did not, in the result, have to rule on any because of the view it took on the meaning of the language of section 34(4).
The final point that should be made is about the reliance on Tecco Trichy: that was a case in which Lahoti CJ held that service on an unknown clerk in a large Government office does not constitute effective service. This conclusion can be ascribed to the traditional authority-based reasoning: such an employee is unlikely to have actual or ostensible authority to accept service. It should not be treated as authority for the general proposition that service excludes all agents, whatever their authority.