1989 INSC 0316 Union of India and Another Vs Raghubir Singh (Dead) By Lrs. Etc. Civil Appeals Nos. 2839-40 of 1989 Prithipal Singh and Others Vs Union of India Civil Appeal No. 4404 of 1985 Ram Mehar Raj Kumar and Others Etc. Vs Union of India and Others Slp (Civil) Nos. 8643 and 8661 of 1985 Delhi Cattle Breeding Farms Pvt. Ltd. Etc. Vs Union of India Slp (Civil) No. 8829 of 1985 (CJI R. S. Pathak, S. Natarajan, Rangath Misra, Sabyasachi Mukharji, E. S. Vankataramiah JJ) 16.05.1989 JUDGMENT PATHAK C.J. ­ 1. The question of law referred to us for decision in these cases is : "Whether, under the Land Acquisition Act, 1894, as amended by the Land Acquisition (Amendment) Act, 1984, the claimants are entitled to solatium at 30 per cent, of the market value irrespective of dates on which the acquisition proceedings were initiated or the dates on which the award had been passed ?" 2. It would suffice if we briefly refer to the facts in the Civil Appeals arising out of Special Leave Petitions Nos. 8194-8195 of 1985 - Union of India v. Raghubir Singh. 3. The land belonging to the respondents in village Dhaka was taken by compulsory acquisition initiated by a notification under section 4 of the Land Acquisition Act, 1894, issued on 13th November, 1959. The award with regard to compensation was made by the Collector on 30th March, 1963. A reference under section 18 of the Act was disposed of by the Additional District Judge on 10th June, 1968. He enhanced the compensation. The respondents preferred an appeal to the High Court claiming further compensation. During the pendency of the appeal, the Land Acquisition (Amendment) Bill, 1982, was introduced in Parliament on 30th April, 1982, and became law as the Land Acquisition (Amendment) Act, 1984, when it received the assent of the President on 24th September, 1984, when it received the assent of the President on 24th September, 1984. The High Court disposed of the appeal by its judgment and order dated 6th December, 1984. While it raised the rate of compensation, it also raised the fate of interest payable on the compensation, and taking into account the change in the law effected by the Land Acquisition (Amendment) Act, 1984 (referred to hereinafter as "the Amendment Act"), it awarded solatium at 30 per cent of the market value. The judgment and Order of the High Court is the subject of these appeals. 4. When these cases camp up before a Bench of two learned judges (E. S. Venkataramiah and R. B. Misra JJ.) on 23rd September, 1985, they referred to two earlier decisions of this court and expressed the view that the question set forth above required re-examinations by a larger Bench of five judges. It was further directed that the other questions involved in the petitions would be considered after the aforesaid had been resolved by the larger Bench. The two decisions referred to in the order of the learned judges are K. Kamalajammanniavaru v. Special Land Acquisition Officer ((1985) 1 SCC 582) decided by O. Chinnappa Reddy and Sabyasachi Mikharji JJ., on 14th February 1985, and Bhag Singh v. Union Territory of Chandigarh ((1985) 3 SCC 737) decided by P. N. Bhagwati C.J., A. N. Sen and D. P. Madon JJ. on August 14, 1985. 5. Solatium is awarded under sub-section (2) of section 23 of the Land Acquisition Act. Before the Amendment Act was enacted the sub-section provided for solatium at 15 per cent of the market value. By the change introduced by the Amendment Act, the amount has been raised to 30 per cent of the market value. Sub-section (2) of section 30 of the Amendment Act specifies the category of cases to which the amended rate of solatium is attracted. In K. Kamalajam-manniavaru ((1985 1 SCC 582), the two learned Judges held that sub-section (2) of section 30 referred to orders made by the High Court or the Supreme Court in appeals against an award made between April 30, 1982, and September 24, 1984, and that, therefore, solatium at 30 per cent alone pursuant to sub-section (2) of section 30 had to be awarded in such cases. In Bhag Singh ((1985 3 SCC 737), however, the three learned Judges held that sub-section (2) of section 30 referred to proceedings relating to compensation pending on April 30, 1982, or filed subsequent to that date, whether before the Collector or before the court or the High Court or the Supreme Court, even if they had finally terminated before the enactment of the Amending Act. In taking that view, they overruled K. Kamalajam-manniavaru ((1985) 1 SCC 582) and approved of the opinion expressed in another case, State of Punjab v. Mohinder Singh ((1986) 1 SCC 365), decided by S. Murtaza Fazal Ali, A. Varadarajan and Ranganath Misra JJ. on May 1, 1985. 6. At the outset, a preliminary objection has been raised by Shri B. R. L. Iyenger to the validity of the reference of these cases to a larger Bench. He contends that the mere circumstances that a Bench of two learned judges finds itself in doubt about the correctness of the view taken by a Bench of three learned judges should not provide reason for referring the matter to a larger Bench. The preliminary objection raised by Shri Iyenger has been vigorously resisted by the appellants. Having regard to the submissions made before us, we think it necessary to lay down the law on the point. 7. India is governed by judicial system identified by a hierarchy of courts, where the doctrine of binding precedent is a cardinal feature of its jurisdiction. It used to be disputed that judges make law. Today it is no longer a matter of doubt that a substantial volume of the law governing the lives of citizens and regulating the functions of the State flows from the decisions of superior courts. "There was a time," observed Lord Reid, "when it was though almost indecent to suggest that judges make law-they only declare it ... But we do not believe in fairly tales any more." (The Judge as Law Maker, p.22) In countries such as the United Kingdom, where Parliament, as the legislative organ, is supreme and stands at the apex of the constitutional structure of the State, the role played by judicial law-making is limited. In the first place, the function of the courts is restricted to the interpretation of laws made by Parliament, and the courts have no power to question the validity of parliamentary statutes, the Diceyan dictum holding true that the British Parliament is paramount and all powerful. In the second place, the law enunicated in every decision of the courts in England can be superseded by an Act of Parliament. As Cockburn CJ observed in Ex parte Canon Selwy ((1872) 36 JP 54) : "There is no judicial body in the country by which the validity of an Act of Parliament could be questioned. An act of the Legislature is superior in authority to any court of law." And Ungoed-Thomas J. in Cheney v. Conn ((1968) 1 All ER 779) referred to a parliamentary statute as "the highest form of law ... which prevails over every other form of law". The position is substantially different under a written Constitution such as the one which governs us. The Constitution of India, which represents the Supreme Law of the land, envisages three distinct organs of the State, each with its own distinctive functions, each a pillar of the State. Broadly, while Parliament and the State Legislature in India enact the law and the Executive Government implements it, the Judiciary sits in judgment not only on the implementation of law by the Executive but also on the validity of the legislation sought to be implemented. One of the functions of the superior judiciary in India is to examine the competence and validity of legislation, both in point of legislative competence as well as its consistency with Fundamental Rights. In this regard, the courts in India possess a power not given to the English courts. Where a statute is declared invalid in India, it cannot be reinstated unless constitutional sanction is obtained modified version of the statute is enacted which account with constitutional prescription. The range of judicial review recognised in the superior judicatory of India is perhaps the widest and the most extensive known to the world of law. The power extends even to examining the validity of an amendment to the Constitution, for, now, it has been repeatedly held that the no constitutional amendment can be sustained which violates the basic structure of the Constitution. (See His Holiness Kesavananda Bharathi Sripadagalvaru v. State of Kerala ((1975) 4 SCC 225 : 1973 Supp SCR 1) Smt. Indira Nehru Gandhi v. Shri Raj Narain ((1975) Supp SCC 1 : (1976) 2 SCR 347), Minerva Mills Ltd. v. Union of India, ((1980) 2 SCC 591) and recently in S. P. Sampath Kumar v. Union of India ((1987) 1 SCC 124 : (1987) 2 ATC 82 : (1987) 1 SCR 435). With this impressive expanse of judicial power, it is only right that the superior courts in India should be conscious of the enormous responsibility which rests on them. This is specially true of the Supreme Court, for, as the highest court in the entire judicially system, the law declared by it is, by article 141 of the Constitution, binding on all courts within the territory of India. 8. Taking note of the hierarahical character of the judicial system in India, it is of paramount importance that the law declared by this court should be certain, clear and consistent. It is commonly known that most decisions of the courts are of significance not merely because they constitute an adjudication on the rights of the parties and resolve the dispute between them, but also because on doing so, they embody a declaration of law operating as a binding principle in future cases. In this later aspect lies their particular value in developing the jurisdiction of the law. 9. The doctrine of binding precedent has the merit of promoting a certainly and consistency in judicial decisions, and enables an organic development of the law, besides providing assurance to the individual as to the consequence of transactions forming part of his daily affairs. And, therefore, the need for a clear and consistent enunciation of legal principles in the decisions of a court. 10. But, like all principles evolved by man for the regulation of the social order, the doctrine of binding precedent is circumscribed in its governance by perceptible limitations, limitations arising by reference to the need for readjustment in a changing society, a readjustment of legal normal demanded by a changed social context. This need for adapting the law to new urges in society brings hone the truth of the Holmesian aphorism that "the life of the law has not been logic, it has been experience" (Oliver Wendell Holmes : The Commin Law, p.5), and again when he declared in another study (Oliver Wendell Holmes : Common Carriers and the Common Law, (1943) 9 Curr LT 387, 388) that "the law is forever adopting new principles from life at one end", and "sloughing off" old ones at the other. Explaining the conceptual import of what Holmes had said, Julius Stone elaborated that it is by the introduction of new extra-legal propositions emerging from experience to serve as premises, or by experience-guided choice between competing legal propositions, rather than by the operations of logic upon existing legal propositions, that the growth of law tends to be determined (Julius Stone : Legal System & Lawyers Reasoning, pp. 58-59). 11. Legal compulsions cannot be limited by existing legal propositions, because there will always be, beyond the frontiers of the existing law, new areas inviting judicial scrutiny and judicial choicemaking which could well affect the validity of existing legal dogmas. The search for solutions responsive to a changing social era involves a search not only among competing propositions of law, or competing versions of a legal proposition, or the modalities of an indeterminacy such as "fairness" or "reasonableness", but also among propositions from outside the ruling law, corresponding to the empirical knowledge or accepted values of present time and place, relevant to the dispensation of justice within the new parameters. 12. The universe of problems presented for judicial choice-making at the growing points of law is an expanding universe. The areas brought under control by the accumulation of past judicial choice may be large. Yet the areas newly presented for still further choice because of changing social, economic and technological conditions are far from inconsiderable. It has also to be remembered that many occasions for new option arise by the mere fact that no generation looks out on the world from quite the same vantage-point as its predecessor, nor for that matter with the same perception. A different vantage point or a different quality of perception often reveals the need for choice-making where formerly no alternatives, and no problems at all, were perceived. The extensiveness of the areas for judicial choice at a particular time is a function not only of the accumulation of past decisions, not only of changes in the environment, but also new insights and perspectives both on old problems and on the new problems thrown up by changes entering the cultural and social heritage. 13. Not infrequently, in the nature of things, there is a gravity heavy inclination to follow the groove set by presidential law. Yet, a sensitive judicial conscience often persuades the mind to search for a different set of norms more responsive to the change social context. The dilemma before the judge poses the task of finding a new equilibrium, prompted not seldom by the desire to reconcile opposing mobilities. The competing goals, according to Dean Roscoe Pound, invest the judge with the responsibility "of proving to mankind that the law was something fixed and settled, whose authority was beyond question, while at the same time enabling it to make constant readjustments and occasional radical changes under the pressure of infinite and variable human desires. "(Roscoe Pound : An Introduction to the Philosophy of Law, p. 19). The reconciliation suggested by Lord Reid in The Judge as Law-Maker (pp. 25-26) lies in keeping both objectives in view, "that the law shall be certain, and that it shall be just and shall move with the times." An elaboration of this opinion is contained in Myres v. Director of Public Prosecutions ((1965) AC 1001 : (1964) 3 WLR 145 : (1964) 1 All ER 877), where he expressed the need for change in the law by the court and the limits within which such change could be brought about. He said : (Ibid at p. 1021) "I have never taken a narrow view of the functions of this House as an appellate tribunal. The common law must be developed to meet changing economic conditions and habits of thought, and I would not be deterred by expressions of opinion in this House in old cases. But there are limits to what we can or should do. If we are to extend the law, it must be by the development and application of fundamental principles. We cannot introduce arbitrary conditions or limitations : that must be left to legislation. And if we do in effect change the law, we ought, in my opinion, only to do that in cases where our decision will produce some finality or certainty." Whatever the degree of success in resolving the dilemma, the court would do well to ensure that although the new legal norm chosen in response to the changed social climate, represents a departure from the previously ruling norm, it must, nevertheless, carry within it the same principle of certainly, clarity and stability. 14. The profound responsibility which is borne by this court in its choice between earlier established standards and the formulation of a new code of norms is all the more sensitive and significant because the response lies in relation to a rapidly changing social and economic society. In a developing society such as in India, the law does not assume its true function when it follows a groove chased amidst a context which has long since crumbled. There will be found among some of the areas of the law, norms selected by a judicial choice educated in the experience and values of a world which had passed away 40 years ago. The social forces which demand attention to the cauldron of change from which a new society is emerging appear to call for new perceptions and new perspectives. The recognition that the times are changing the that there a occasion for a new jurisprudence to take birth is evidenced by what this court said in Bengal Immunity Co. Ltd., v. State of Bihar ((1955) 2 SCR 603 : AIR 1955 SC 661 : (1955) 6 STC 446), when it observed that it was not bound by its earlier judgments and possessed the freedom to overrule its judgment when it thought it fit to do so to keep pace with the needs of changing times. The acceptance of this principle ensured the preservation and legitimation provided to the doctrine of binding precedent, and therefore, certainly and finality in the law, while permitting necessary scope for judicial creativity and adaptability of the law to the changing demands of society. 15. The question then is not whether the Supreme Court is bound by its own previous decisions. It is not. The question is under what circumstance and within what limits and in what manner the highest court should overturn its own pronouncements. 16. In an examination of this question, it would perhaps be appropriate to refer to the response of other jurisdictions, specially those with which the judicial system in India has borne an historical relationship. The House of Lords in England provides the extreme example of a judicial body which until recently disclaimed the power to overrule itself. It used to be said that the House of Lords did never overrule itself but only distinguished its earlier decisions. An erroneous decision of the House of Lords could be set right only by an Act of Parliament. (See Street Tramways Co. Ltd. v. London County Council ((1898) AC 375) and Radcliff v. Ribble Motor Services Ltd. ((1939) AC 215, 245)). Apparently bowing to the pressure of a reality forced upon it by reason of a rapidly gathering change in the prevailing socio-economic structure, on July 26, 1966, Lord Gardiner L.C., made the following statement on behalf of himself and the Lords of Appeal in Ordinary : "Their lordships regard the use of precedent as an indispensable foundation upon which to decide what is the law and its application to individual cases. It provides at least some degree of certainty upon which individuals can rely in the conduct of their affairs, as well as a basis for orderly development of legal rules. Their Lordships nevertheless recognise that too rigid adherence to precedent may lead to injustice in a particular case and also unduly restrict the proper development of the law. They propose, therefore, to modify their present practice and, while treating former decisions of this House as normally binding, to depart from a previous decision when it appears right to do so. In this connection, they will bear in mind the danger of disturbing retrospectively the basis on which contracts, settlements of property and fiscal arrangements have been entered into and also the special need for certainty as to the criminal law." Since then, the House of Lords has framed guidelines in a series of cases decided up to 1975 and the guidelines have been summarised in Dr. Alan Paterson's "Law Lords" ((1982) pp. 156-157). He refers to several criteria articulated by Lord Reid in those cases. 1. The freedom granted by the 1966 Practice Statement ought to be exercised sparingly (the 'use sparingly' criterion) (Jones v. Secretary of State for Social Services (1972 AC 944, 966)). 2. A decision ought not to be overruled if to do so would upset the legitimate expectations of people who have entered into contracts or settlements or otherwise regulated their affairs in reliance on the validity of that decision (the "legitimate expectations" criterion) (Ross-Smith v. Ross-Smith and Indyka v. Indyka ((1969) AC 33, 69 : (1967) 2 All ER 689 : (1967) 3 WLR 510)). 3. A decision concerning questions of construction of statutes or other documents ought not to be overruled except in rare and exceptional cases (the 'construction' criterion (Ibid. supra note 22)). 4. (a) A decision ought not to be overruled if it would be impracticable for the Lords to foresee the consequences of departing from it (the 'Unforeseeable consequences' criterion) (Steadman v. Steadman (1976 AC 536, 542 : (1974) 2 All ER 977 : (1974) 3 WLR 56)). (b) A decision ought not to be overruled if to do so would involve a change that ought to be part of a comprehensive reform of the law. Such changes are best done 'by legislation following a wide survey of the whole filed' (the "need for comprehensive reform' criterion) (Myers v. DPP (1965 AC 1001, 1022 : (1964) 2 All ER 881 : (1964) 3 WLR 145); Cassell and Co. Ltd. Broome (1972 AC 1027, 1086); Haughton v. Smith (1975 AC 476, 500)). 5. In the interest of certainly, a decision ought not to be overruled merely because the Law Lords consider that it was wrongly decided. There must be some additional reasons to justify such a step (the "precedent merely wrong" criterion) (Knuller v. DPP (1973 AC 435, 455)). 6. A decision ought to be overruled if it causes such great uncertainly in practice that the parties' advisers are unable to give any clear indication as to what the courts will hold the law to be (the "rectification of uncertainly" criterion (Ibid., supra note 22; Oldendorff (E.L) & Co. Gam BH v. Tradax Export SA, 1974 AC 479, 533, 535 : (1972) 3 All ER 420)). 7. A decision ought to be overruled if, in relation to some board issue or principle, it is not considered just or in keeping with contemporary social conditions or modern conceptions of public policy (the 'unjust or outmoded' criterion Ibid., supra note 22; Conway v. Rimmer, (1968) AC 910, 938 : (1968) 2 WLR 998)). 17. Dr. Paterson noted that between the years 1966 and 1988, there were twenty-nine cases in which the House of Lords was invited to overrule one of its own precedents, that the House of Lords did so in eight of them, while in a further ten cases, at least one of the Law Lords was willing to overrule the previous House of Lords precedent. In a considerable number of other cases, however, the Law Lords seemed to prefer to distinguish the earlier decisions rather than overrule them. 18. The High Court of Australia, the highest court in the Common-wealth, has reserved to itself the power to reconsider its own decision, but has laid down that the power should not be exercised upon a mere suggestion that some or all the members of the later court would arrive at a different conclusion if the matter were res integra. In the Tramways' case ((1914) 18 CLR 54), Griffith C.J., while doing so, administered the following caution : "In my opinion, it is impossible to maintain as an abstract proposition that the court is either legally or technically bound by previous decisions. Indeed, it may, in a proper case, be its duty to disregard them. But the rule should be applied with great caution, and only when the previous decision is manifestly wrong, as, for instance, if it proceeded upon the mistaken assumption of the continuance of a repealed or expired statute, or is contrary to a decision of another court which this court is bound to follow; not, I think, upon a mere suggestion, that some or all of the members of the later court might arrive at a different conclusion if the matter was res integra. Otherwise, there would be grave danger of want of continuity in the interpretation of law." In the same case, Barton J. observed at p. 69 : "... I would say that I never though that it was not open to this court to review its previous decisions upon good cause. The question is not whether the court can do so, but whether it will, having due regard to the need for continuity and consistency of the judicial decisions. Changes in the number of appointed justices can, I take it, never of themselves furnish a reason for review ... But the court can always listen to argument as to whether it ought to review a particular decision, and the strongest reason for an overruling is that a decision is manifestly wrong and its maintenance is injurious to the public interest." 19. In the United States of America, the Supreme Court has explicitly overruled its prior decision in a number of cases and reference will be found to them in the judgment of Brandeis J in State of Washington v. Dawson and Co. (264 US 646 : 68 L Ed 219) where he said : (US p 238) "The doctrine of stare decisis should not deter us from overruling that case and those which follow it. The decisions are recent ones. They have not been acquiesced in. They have not created a rule of property around which vested interests have clustered. They affect solely matters of a transitory nature. On the other hand, they affect seriously the lives of men, women and children, and the general welfare. Stare decisis is ordinarily a wise rule of action. But it is not universal, inexorable command. The instances in which the courts have disregarded its admonition are many." Elaborating his point in his dissenting judgment in David Burnet v. Coronado Oil and Gas Company (285 US 393; 76 L Ed 815). Barandeis J. observe : (US pp. 406-408) "Stare decisis is usually the wise policy, because in most matters it is more important that the applicable rule of law be settled right. Compare-National Bank v. Whitney (103 US 99; 26 L Ed 443-444) This is commonly type even where the error is a matter or serious concern, provided correction can be had by legislation. But in cases involving the Federal Constitution, where correction through legislative action is practically impossible, this the lessons of experience its earlier decisions. The court bows to the lessons of experience and the force of better reasoning recognising that the process of trial and error, so fruitful in the physical sciences, is appropriate also in the judicial function." 20. The Judicial Committee of the Privy Council also took the view that it was not bound in law by its earlier decisions, but, in In re, Compensation to Civil Servants ((1929) AC 242; AIR 1929 PC 84, 87 : 56 MLJ 363), it declared that it "would hesitate long before disturbing a solemn decision by a previous Board, which raised an identical or even a similar issue for determination" and reiterated that reservation in Attorney-General of Ontario v. Canada Temperance Federation (1946 AC 193 : 62 TLR 199 (PC), and Phanindra Chandra Neogy v. The King (76 IA 10 : AIR 1949 PC 117). 21. These cases from England, Australia and the United States were considered by this court in Bengal Immunity Co. Ltd. v. State of Bihar, ((1955) 2 SCR 603 : AIR 1955 SC 661 : (1955) 6 STC 446), perhaps the first recorded instance of the Supreme Court in this country being called upon to consider whether it could overrule an earlier decision rendered by it. A Bench of seven judges assembled to consider whether the majority decision of a Constitution Bench of five judges in State of Bombay v. United Motors (India) Ltd. (1953 SCR 1069 AIR 1953 SC 252 : (1953) 4 STC 133), should be reconsidered. Four judges of the bench of seven said it should and voted to overrule the majority decision in United Motors' (1953 SCR 1069 AIR 1953 SC 252 : (1953) 4 STC 133). The remaining three voted to the contrary. Das, Acting C.J., speaking for himself and on behalf of Bose, Bhagwati and Jafar Imam JJ., preferred the approach adopted by the United States Supreme Court since, in the view of that learned judge, the position in India approximated more closely to that obtaining in the United States rather than to the position in England, where Parliament could rectify the situation by a simple majority, and to that in Australia, where the mistake could be corrected in appeal to the Privy Council. The learned judge observed : (SCR pp. 627-28) "There is nothing in our Constitution which prevents us from departing from a previous decision if we are convinced of its error and its baneful effect on the general interests of the public." And reference was made to the circumstance that article 141 of the Constitution made the law declared by this court binding on all courts in India. Speaking with reference to the specific case before the court, the learned judge referred to the far-reaching effect of the earlier decision in United Motors' (1953 SCR 1069 : AIR 1953 SC 252 : (1953) 4 STC 133), on the general body of the consuming public, and stated that the error committed in the earlier decision would result in perpetuating a tax burden erroneously imposed on the people, giving rise to a consequence "manifestly and wholly unauthorised." The learned judge observe : "It is not an ordinary pronouncement declaring the rights of two private individuals 'inter se'. It involves an adjudication on the taxing power of the States as against the consuming public generally. If the decision is erroneous, as indeed as conceive it to be, we owe it to the public to protect them against the illegal tax burdens which the States are seeking to impose on the strength of that erroneous recent decision." Cautioning that the court should not differ merely because a contrary view appeared preferable, the learned judge affirmed that "we should not lightly dissent from a previous pronouncement of this court." But, if the previous decision was plainly erroneous, he pointed out, there was a duty on the court to say so and not perpetuate the mistake. The appeal to the principle of stare decisis was rejected on the ground that (a) the decision intended to be overruled was a very recent decision and it did not involve overruling a series of decisions, and (b) the doctrine of stare decisis was not an inflexible rule, and must, in any event, yield where, following it would result in perpetuating an error to the detriment of the general welfare of the public or a considerable section thereof. 22. Since then, the question as to when the Supreme Court should overrule its own decision has been considered in several cases. Relying on the Bengal Immunity ((1955) 2 SCR 603 : AIR 1955 SC 661 : (1955) 6 STC 446), Khanna J. remarked that certainty in the law which was an essential ingredient of the rule of law would be considerably eroded if the highest court of the land lightly overruled the view expressed by it in earlier cases. One instance where such overruling could be permissible was a situation where contextual values gives birth to the earlier view had altered substantially since. In Maganlal Chagganlal P. Ltd. v. Municipal Corporation of Greater Bombay ((1974) 2 SCC 402 : (1975) 1 SCR 1), he explained : (SCC pp 425-26, para 22) "Some new aspects may come to light and it may become essential to cover fresh ground to meet the new situations or to overcome difficulties which did not manifest themselves or were not taken into account when the earlier view was propounded. Precedents have a value and the ratio decidendi of a case can no doubt be of assistance in the decision of future cases. At the same time, we have to, as observed by Cardozo, guard against the notion that because a principle has been formulated as the ratio decidendi of a given problem, it is, therefore, to be applied as a solvent of other problems, regardless of consequences, regardless of deflecting factors, inflexibly and automatically, in all its pristine generality (see Selected Writings, p. 31). As in life, so in law, things are not static." 23. In Lt. Col. Khajoor Singh v. Union of India ((1961) 2 SCR 828 : AIR 1961 SC 532), the majority of this court emphasised that the court should not depart from an interpretation given in an earlier judgment of the court unless there was a fair amount of unanimity that the earlier decision was manifestly wrong. In Keshav Mills Co. Ltd. v. CIT ((1965) 2 SCR 908, 921 : AIR 1965 SC 1636 : (1965) 56 ITR 365), this court observed that a revision of its earlier decision would be justified if there were compelling and substantial reasons to do so. In Sajjan Singh v. State of Rajasthan ((1965) 1 SCR 933, 947 AIR 1965 SC 845), the court laid down the test : '(I)s it absolutely necessary and essential that the question already decided should be reopened ?', and went on to observe : "The answer to this question would depend on the nature of the infirmity alleged in the earlier decision, its impact on public good and the validity an compelling character of the consideration urged in support of the contrary view." There can be no doubt, as was observed in Girdhari Lal Gupta v. D. N. Mehta ((1971) 3 SCC 189 : 1971 SCC (Cri) 279 : (1971) 3 SCR 748), that where an earlier relevant, statutory provision has not been brought to the notice of the court, the decision may be reviewed, or, as in Pillani Investment Corporation Ltd. v. ITO ((1972) 1 SCC 122 : (1972) 2 SCR 502), if a vital point was not considered. A more compendious examination of the problem was undertaking in Keshav Mills Co. v. CIT ((1961) 2 SCR 828 : AIR 1961 SC 532), where the court pointed out : (SCR p. 922) "It is not possible or desirable, and in any case, it would be inexpident to lay down any principles which should govern the approach of the court in dealing with the question of reviewing and revising its earlier decision. It would always depend upon several relevant considerations : What is the nature of the infirmity or error on which a plea for a review and revision of the earlier view is based ? On the earlier occasion, did some patent aspect of the question remain unnoticed, or was the attention of the court not drawn to any relevant and material statutory provision, or was any previous decision of this court bearing on the point not noticed ? Is the court hearing such plea fairly unanimous that there is such an error in the earlier view ? What would be the impact of the error on the general administration of law or on public good ? Has the earlier decision been followed on subsequent occasions either by this court or by the High Courts ? And would the reversal of the earlier decision lead to public inconvenience, hardship or mischief ? These and other relevant considerations must be carefully borne in mind whenever this court in called upon to exercise its jurisdiction to review and revise its earlier decisions. These considerations become still more significant when the earlier decision happens to be a unanimous decision of the Bench of five learned judges of this court." 24. Much importance has been laid on observing the finality of decisions rendered by the Constitution Bench of this court and in Ganga Sugar Corpn. Ltd. v. State of Utter Pradesh ((1980) 1 SCC 223 1980 SCC (Tax) 90 : (1980) 1 SCR 769, 782), the court held against the finality only where the subject was "of such fundamental importance to national life or the reasoning is so plainly erroneous in the light of later thought that it is wiser to be ultimately right rather than to be consistently wrong". 25. It is not necessary to refer to all the cases on the point. The broad guidelines are easily deducible from what has gone before. The possible of further defining these guiding principles can be envisaged with further juridical experience, and when common jurisprudential values linking different national systems of law may make a consensual pattern possible. But that lies in the future. 26. There was some debate on the question whether a Division Bench of two judges is obliged to follow the law laid down by a Division Bench of a larger number of judges. Doubt has arisen on the point because of certain observations made by Chinnappa Reddy J. in Javed Ahmed Abdul Hamid Pawala v. State of Maharashtra ((1985) 1 SCC 275 : 1984 SCC (Cri) 653 : AIR 1985 SC 231), Earlier, a Division Bench of two judges, of whom he was one had expressed the view in T. V. Vatheeswaran v. State of Tamil Nadu, ((1983) 2 SSC 68 : 1983 SCC (Cri) 342 : AIR 1983 SC 361), that delay exceeding two years in the execution of a sentence of death should be considered sufficient to entitle a person under sentence of death to invoke article 21 of the Constitution and demand the quashing of the sentence of death. This would be so, he observed, even if the delay in the execution was occasioned by the same necessary for filing an appeal or for considering the reprieve of the accused or some other cause for which the accused himself may be responsible. This view was found unacceptable by a Bench of three judges in Sher Singh v. State of Punjab, ((1983) 2 SCC 344 : 1983 SCC (Cri) 461 : AIR 1983 SC 465), where the learned Judges observed that no hard and fast rule could be laid down in the matter. In direct disagreement with the view in T. V. Vatheeswaran ((1983) 2 SCC 68 : 1983 SCC (Cri) 342 : AIR 1983 SC 361), the learned judges said that account had to be taken of the time occupied by proceedings in the High Court and in the Supreme Court and before the executive authorities, and it was relevant to consider whether the delay was attributable to the conduct of the accused. As a member of another Bench of two judges, in Javad Abdul Ahmed Hamid Pawala ((1985) 1 SCC 275 : 1984 SCC (Cri) 653 : AIR 1985 SC 231), Chinnappa Reddy J. questioned the validity of the observations made in Sher Singh ((1983) 2 SCC 344 : 1983 SCC (Cri) 461 : AIR 1983 SC 465), and went on to note, without expressing any concluded opinion on the point, that it was a serious question : "whether a Division Bench of three judges could purport to overrule the judgment of a Division Bench of two judged merely because three is larger than two. The court sits in Divisions of two and three judges for the sake of convenience and it may be inappropriate for a Division Bench of three judges to purport to overrule the decision of a Division Bench of two judges. Vide Young v. Bristol Aeroplane Co. Ltd. ((1944) 2 All ER 293 : (1944) 1 KB 718) It may be otherwise where a Full bench or a Constitution Bench does so". It is pertinent to record here that because of the doubt cast on the validity of the opinion in Sher Singh ((1983) 2 SCC 344 : 1983 SCC (Cri) 461 : AIR 1983 SC 465), the question of the effect of delay on the execution of a death sentence was referred to a Division Bench of five judges, and in Triveniben v. State of Gujarat, ((1988) 4 SCC 574 : 1989 SCC (Cri) 25 : AIR 1989 SC 142), the Constitution Bench overruled T. V. Vatheeswaran ((1983) 2 SCC 68 : 1983 SCC (Cri) 342 : AIR 1983 SC 361) 27. What then should be the position in regard to the effect of the law pronounced by a Division Bench in relation to a case raising the same point subsequently before a Division Bench of a smaller number of judges ? There is no constitutional or statutory prescription in the matter and the point is governed entirely by the practice in India of the courts sanctified by repeated affirmation over a century of time. It cannot be doubted that in order to promote consistency and certainty in the law laid down by a superior court, the ideal condition would be that the entire court should sit in all cases to decide questions of law, and for that reason the Supreme Court of the United States does so. But, having regard to the volume of work demanding the attention of the court, it has been found necessary in India as a general rule of practice and convenience that the court should sit in Divisions, each Division being constituted of judges whose number may be determined by the exigencies of judicial need, by the nature of the case including any statutory mandate relative thereto, and by such other considerations which the Chief Justice, in whom such authority devolves by convention, may find most appropriate. It is in order to guard against the possibility of inconsistent decisions on points of law by different Division Benches that the rule has been envolved in order to promote consistency and certainly in the development of the law and its contemporary status, that the statement of the law by a Division Bench is considered binding on a Division Bench of the same or lesser number of judges. This principles has been followed in India for several generations by judges. We may refer to a few of the recent cases on the point. In John Martin v. State of West Bengal ((1975) 3 SCC 836 : 1975 SCC (Cri) 255 : (1975) 3 SCR 211), a Division Bench of three judges found it right to follow the law declared in Haradhan Saha v. State of West Bengal ((1975) 3 SCC 198 : 1974 SCC (Cri) 816 : (1975) 1 SCR 778), decided by a Division Bench of five judges, in preference to Bhut Nath Mate v. State of West Bengal, ((1974) 1 SCC 645 : 1974 SCC (Cri) 300 : AIR 1974 SC 806), decided by a Division Bench of two judges. Again, in Smt. Indira Nehru Gandhi v. Shri Raj Narain (1975 Supp SCC 1 : (1976) 2 SCR 347), Beg J. held that the Constitution Bench of five judges was bound by the Constitution Bench of thirteen judges in His Holiness Kesavananda Bharati Sripadakalavaru v. State of Kerala ((1973) 4 SCC 225 : 1973 Supp SCR 1)4 SCC 225 : 1973 Supp SCR 1)4 SCC 225 : 1973 Supp SCR 1)4 SCC 225 : 1973 Supp SCR 1). In Ganapati Sitaram Balvalkar v. Waman Shripad Mage ((1981) 4 SCC 143) this court expressly stated that the view taken on a point of law by a Division Bench of four judges of this court was binding on a Division Bench of three judges of the court. And in Mattulal v. Radhe Lal ((1974) 2 SCC 365 : (1975) 1 SCR 127), this court specifically observed that where the view expressed by two different Division Benches of this court could not be reconciled, the pronouncement of a Division Bench of a larger number of judges had to be preferred to the decision of a Division Bench of a smaller number of judges. This court also laid down in Acharya Maharaj Shri Narendra Prasadji Anandprasadji Maharaj v. State of Gujarat ((1975) 1 SCC 11 : (1975) 2 SCR 317), that even where the strength of two differing Division Benches consisted of the same number of judges, it was not open to one Division Bench to decide the correctness or otherwise of the views of the other. The principle was reaffirmed in Union of India v. Godfrey Philips India Ltd. ((1985) 4 SCC 369 : 1986 SCC (Tax) 11). Which noted that a Division Bench of two judges of this court in Jit Ram Shiv Kumar v. State of Haryana, ((1981) 1 SCC 11 : (1980) 3 SCR 689) had differed from the view taken by an earlier Division Bench of two judges in Motilal Padampat Sugar Mills Co. Ltd. v. State of U. P. ((1979) 2 SCC 409 : 1979 SCC (Tax) 144 : (1979) 2 SCR 641), on the point whether the doctrine of promissory estoppel could be defeated by invoking the defence of executive necessity, and, holding that to do so was wholly unacceptable, reference was made to the well accepted and desirable practice of the later Bench referring the case to a larger Bench when the learned judges from that the situation called for such reference. 28. We are of the opinion that the pronouncement of law by a Division Bench of this court is binding on a Division Bench of the same or a smaller number of judges and in order that such decision be binding, it is not necessary that it should be a decision rendered by the Full Court or a Constitution Bench of the court. We would, however, like to think that for the purpose of imparting certainly and endowing due authority, decision of this court in future should be rendered by Division Benches of at least three judges unless, for compelling reasons, that is not conveniently possible. 29. Upon the aforesaid considerations and in view of the nature and potential of the questions raised in these cases, we are of the view that there was sufficient justification for the order dated September 23, 1985, made by the Bench of two learned judges referring these cases to a larger Bench for reconsideration of the question decided in K. Kamalajammanniavaru ((1985) 1 SCC 582) and Bhag Singh, ((1985) 3 SCC 737). The preliminary objection raised by learned counsel for the respondents to the validity of the reference is overruled. 30. We now come to the merits of the reference. The reference is limited to the interpretation of section 30(2) of the Land Acquisition (Amendment) Act, 1984. Before the enactment of the Amendment Act, solatium was provided under section 23(2) of the Land Acquisition Act (shortly "the parent Act") at 15% of the market value of the land computed in accordance with section 23(1) of the Act, the solatuim being provided in consideration of the compulsory nature of the acquisition. The Land Acquisition Amendment Bill, 1982, was introduced in the House of the People on April 30, 1982, and upon enactment, the Land Acquisition Amendment Act, 1984, commenced operation with effect from September 24, 1984. Section 15 of the Amendment Act amended section 23(2) of the parent Act and substituted the words "30 per centum" in place of the words "15 per centum". Parliament intended that the benefit of the enhanced solatium should be made available, albeit to a limited degree, even in respect of acquisition proceedings taken before that date. It sought to effectuate that intention by enacting section 30(2) in the Amendment Act. Section 30(2) of the Amendment Act provides : (2) The provides of sub-section (2) of section 23 ... of the principal Act, as amended by clause (b) of section 15 ... of this Act ... shall apply and shall be deemed to have applied, also to, and in relation to, any award made by the Collector or court or to any other passed by the High Court or Supreme Court in appeal against any such award under the provisions of the principal Act after the 30th day of April, 1982 [the date of introduction of the Land Acquisition (Amendment) Bill, 1982, in the House of the People] and before the commencement of this Act. 31. In construing section 30(2), it is just as well to be clear that the award made by the Collector referred to here is the award made by the Collector under section 11 of the parent Act, and the award made by the court is the award made by principal civil court of original jurisdiction under section 23 of the parent Act on a reference made to it by the Collector under section 19 of the parent Act. There can be no doubt that the benefit of the enhanced solatium is intended by section 30(2) in respect of an award made by the Collector between April 30, 1982, and September 24, 1984, Likewise, the benefit of the enhanced solatium is extended by section 30(2) to the case of an award made by the court between April 30, 1982, and September 24, 1984, even though it be upon reference from an award made before April 30, 1982. 32. The question is : what is the meaning of the words "or to any order passed by the High Court or Supreme Court on appeal against any such award ?" Are they limited, as contended by the appellants, to appeals against an award of the Collector or the court made between April 30, 1982, and September 24, 1984, or do they include also, as contended by the respondents, appeals disposed of between April 30, 1982, and September 24, 1984, even though arising out of awards of the Collector or the court made before April 30, 1982. We are of opinion that the interpretation placed by the appellants should be preferred over that suggested by the respondents. Parliament has identified the appeal before the High Court and the appeal before the Supreme Court by describing it as an appeal against "any such award". The submission on behalf of the respondents is that the words "any such award" mean the award made by the Collector or court and carry no greater limiting sense; and that in this context, upon the language of section 30(2), the order in appeal is an appellate order made between April 30, 1982, and September 24, 1984 - in which case the related award of the Collector or of the court may have been made before April 30, 1982. To our mind, the words "any such award" cannot bear the broad meaning suggested by learned counsel for the respondents. No such words of description by way of identifying the appellate order of the High Court or of the Supreme Court were necessary. Plainly, having regard to the existing hierarchical structure of fore contemplated in the parent Act, those appellate order could only be orders arising in appeal against the award of the Collector or of the court. The words "any such award" are intended to have deeper significance, and in the context in which those words appear in section 30(2), it is clear that they are intended to refer to awards made by the Collector or court between April 30, 1982, and September 24, 1984. In other words, section 30(2) of the Amendment Act extends the benefit of the enhanced solatium to cases where the award by the Collector or by the court is made between April 30, 1982, and September 24, 1984, or to appeals against such awards decided by the High Court and the Supreme Court whether the decision of the High Court or the Supreme Court are rendered before September 24, 1984, or after that date. All that is material is that the award by the Collector or by the court should have been made between April 30, 1982, and September 24, 1984. We find ourselves in agreement with the conclusion reached by this court in K. Kamalajammannivaru (Decd. by Lrs) v. Special Land Acquisition Officer ((1985) 1 SCC 582), and find ourselves unable to agree with the view taken in Bhag Singh v. Union Territory of Chandigarh ((1985) 3 SCC 737). The expanded meaning given to section 30(2) in the later case does not, in our opinion, flow reasonably from the language of that sub-section. It seems to us that the learned judges in that case missed the significance of the word "such" in the collocation of the words "any such award" in section 30(2). Due significance must be attached to that word, and, to our mind, it must necessarily intend that the appeal to the High Court or the Supreme Court, in which the benefit of the enhanced solatium is to be given, must be confined to an appeal against an award of the Collector or of the court rendered April 30, 1982, and September 24, 1984. 33. We find substance in the contention of the learned Attorney-General that if Parliament had intended that the benefit of enhanced solatium should be extended to all pending proceedings, it would have said so in clear language. On the contrary, as he says, the terms in which section 30(2) is couched indicate a limited extension of the benefit. The Amendment Act has not been made generally retrospective with effect from any particular date and such retrospectivity as appears is restricted to certain areas covered by the parent Act and must be discovered from the specific terms of the provision concerned. Since it is necessary to spell out the degree of retrospectivity from the language of the relevant provision itself, close attention must be paid to the provisions of section 30(2) for determining the scope of retrospective relief intended by Parliament in the matter of enhanced solatium. The learned Attorney-General is also right when he points out that it was never intended to define the scope of the enhanced solatium on the mere accident of the disposal of a case in appeal on a certain date. Delays in superior courts now extend to limits which were never anticipated when the right to approach them for relied was granted by the statute. If it was intended that section 30(2) should refer to appeals pending before the High Court or the Supreme Court between 30th April, 1982, and 24th September, 1984, they could well refer to proceedings in which an award had been made by the Collector from anything between 10 to 20 years before. It could never have been intended that rates of compensation and solatium applicable to acquisition proceedings initiated so long ago should now enjoy the benefit of statutory enhancement. It must be remembered that the value of the land is taken under section 11(1) and section 23(1) with reference to the date of publication of the notification under section 4(1), and it is that date which is usually material for the purpose of determining the quantum of compensation and solatium. Both section 11(1) and section 23(1) speak of compensation being determined on the basis, inter alia, of the market value of the land on that date and solatium by section 23(2) is computed as a percentage on such market value. 34. Our attention was drawn to the order made in State of Punjab v. Mohinder Singh ((1986) 1 SCC 365), but in the absence of a statement of the reason which persuaded the learned judges to take the view they did, we find it difficult to endorse that decision. It received the approval of the learned judges who decided Bhag Singh's case ((1985] 3 SCC 737), but the judgment in Bhag The well known rule is the property in all materials and Singh's case ((1985) 3 SCC 737), as we have said earlier, has omitted to give due significance to all the material provisions of section 30(2) and, consequently, we find ourselves at variance with it. The learned judges proceeded to apply the principle that an appeal is a continuation of the proceeding initiated before the court by way of reference under section 18, but, in our opinion, the application of general principle must yield to the limiting terms of the statutory provision itself. Learned counsel for the respondents has strenuously relied on the general principle that the appeal is a rehearing of the original matter, but we are not satisfied that he is on good ground in invoking that principle. Learned counsel for the respondents points out that the word "or" has been used in section 30(2) as a disjunctive between the reference to the award made by the Collector or the court and an order passed by the High Court or the Supreme Court in appeal and, he says, that, properly understood, it must mean that the period, 30th April, 1982, to 24th September, 1984, is as much applicable to the appellate order of the High Court or of the Supreme Court as it is to the award made by the Collector or the court. We think that what Parliament intends to say is that benefit of section 30(2) will be available to an award by the Collector or the court made between the aforesaid two dates or to an appellate order of the High Court or of the Supreme Court which arises out of an award of the Collector or the court made between the said two dates. The word "or" is used with reference to the stage at which the proceeding rests at the time when the benefit under section 30(2) is sought to he extended. If the proceeding has terminated with the award of the Collector or of the court made between the aforesaid two dates, the benefit of section 30(2) will be applied to such award made between the aforesaid two dates. If the proceeding has passed to the stage of appeal before the High Court or the Supreme Court, it is at that stage that the benefit of section 30(2) will be applied. But, in every case, the award of the Collector or of the court must have been made between April 30, 1982 and September 24, 1984. 35. In the result we overrule the statement of the law laid down in Mohinder Singh ((1986) 1 SCC 365) and in Bhag Singh ((1985) 3 SCC 737) and prefer instead the interpretation of Section 30(2) of the Amendment Act rendered in K. Kamalajammanniavaru ((1985) 1 SCC 582). 36. The case will now be listed before a Division Bench of three learned Judges for hearing on the merits of the other points raised in the cases.