THE INTERPRETATION OF DOCUMENTS:

SAYING WHAT THEY MEAN OR MEANING WHAT THEY SAY

By Mr. SUNDARESH MENON* ( Current Chief Justice of Singapore)

An excellent Article by one of the most brilliant mind in Law on an Article that is a Conundrum even to this day and age. The best Article that I have come across in my early morning sojourn every day.

The Speech

I. INTRODUCTION

I have chosen to speak on the topic “The Interpretation of Documents: Saying what they mean or meaning what they say?”. The subject of interpretation has attracted considerable debate through the years, with judges, practitioners, academics and jurists each offering different perspectives. This ought not to be

Surprising, given that according to some English estimates, as much as 90 per cent of a Judge’s work involves an element of interpretation. As students, many of you would have learnt and applied various maxims of interpretation, such as the “mischief” rule or the “contra proferentum” rule. Even just from this, you would have discovered that interpretation involves so much more than the formulaic or mechanical application of maxims. See

*In E. Heward, Lord Denning, A Biography, 2nd ed. (London: Barry Rose Law Publishers Ltd, 1997) at p 62, it was estimated that over 90% of the reported cases of the UK Court of Appeal in 1989 related to the interpretation of statutes. Separately, Lord Steyn observed that interpretation “amounts to the preponderant part of the legal work of English judges, perhaps as high as 90 per cent” (J. Steyn, “The Intractable Problem of the Interpretation of Legal Texts” [2003] 25(5) Sydney Law Review 5).

It is a fascinating subject that undergirds much of what you will ever do as lawyers. I propose to begin this tour of some selected highlights by defining interpretation. Next, I consider why so many disputes before the courts involve questions of interpretation. Against this background, I trace the way in which the law on interpretation has developed through the years, focusing specifically on the interpretation of statutes and contracts. I will then survey some of the difficult issues that arise in statutory and contractual interpretation. Finally, I touch on what the boundaries might be for a court faced with a question of interpretation.

II. WHAT IS THE PROCESS OF “INTERPRETATION”?

Interpretation refers to the process by which meaning is ascribed to the expressions found within a legal text. Such meaning can be ascertained from two possible perspectives.

The Objective Test

The first is to apply an objective test of interpretation to ascertain what a reasonable person would understand the legal text to mean. This, at least in theory, is the approach adopted under the common law. The Subjective Approach

The subjective approach, on the other hand, requires the court to ascribe meaning to the text by having regard to what it perceives or discerns were the actual intentions of the drafters. This is prevalent in civil law systems and is also reflected in a number of international conventions on commercial transactions, for example, in the Vienna Convention and the UNIDROIT Principles of International Commercial Contracts 2010.

In contract law, the subjective test rests in part on the notion that the premise of the law of obligations can be traced “to the will of the individual as [a] self-governing agent”. For that reason, the actual intentions of the parties are treated as relevant, or even paramount.

In contrast, the objective test is premised on the notion of certainty. It recognises that the courts have “no direct access to [the parties’] subjective mental states, no window into their minds”. The primary way to ascertain the parties’ subjective intentions would be to consider their oral evidence, but such a process is sub-optimal in an adversarial system of litigation. Disputes might only arise years after the contract had been entered into, by which time such evidence might not be available or reliable due to the passage of time. These uncertainties also incentivise opportunistic behaviour and in turn reduce the likelihood of the disputants reaching a settlement. There are other knock-on effects arising from such uncertainty, not least, its impact on commercial trade. As Lord Devlin observed more than half a century ago:

… in the service of commerce the letter is in many ways more significant [than the spirit of the contract]. This is because in most commercial contracts many more than the original parties are concerned. The contract is embodied in a document which may pass from hand to hand when the goods it represents are sold over and over again to a string of buyers … The spirit of the contract gets lost on these travels and the outward form is all that matters.

However, it would be overly simplistic and incorrect to conclude that the objective approach disregards the actual intentions of the parties. Rather, what the objective approach does is to shift the burden. Instead of requiring the courts to ascertain the parties’ subjective intentions after the fact, with all its concomitant difficulties, the burden is placed on the contracting parties to ensure at the outset that their respective subjective intentions are accurately encapsulated within the four corners of the legal text. The objective test facilitates this process by providing a somewhat well-developed and more or less predictable framework for interpreting the text. Although the international conventions or instruments which I have mentioned are rooted in the subjective approach, it has been observed that these are “often likely to achieve objective results in practice”.

It is not uncommon to see the terms “interpretation” and “construction” being used interchangeably. However, there is a difference between the two. Whereas the process of interpretation is constrained by the specific words used in the legal text, the process of construction is broader and permits the court to ascertain the parties’ intention, by reference to the contract as a whole. It consists of both interpretation (which is dependent on the specific words used) and implication (which concerns the “presumed intention” of the parties that has not actually been expressed in the text). I will return to this a little later, but in this light, let me turn to consider some of the reasons why interpretive disputes arise in the first place.

III WHY DO DISPUTES RELATING TO THE INTERPRETATION OF LEGAL TEXTS ARISE?

I think there are at least four such causes: — the imprecision of language, — imperfect information, — the cost-benefit analysis and — cognitive limitations.

A. The imprecision of language

Dictionaries supply the possible meanings of a particular word, but provide little or no assistance in determining the meaning that was actually intended by the drafter. Language is inherently imprecise, especially at the “penumbra” as Professor Hart described it. Words and phrases can bear multiple meanings. Indeed, the very same sentence could bear contradictory meanings, depending on whether it was expressed sincerely or sarcastically. As a result, the process of ascertaining meaning by relying on the dictionary alone can be iterative in nature and is liable to frustrate more than it might illuminate.

Let me illustrate the point with an example which Professor Hart put forward: a law that says “No vehicles shall be allowed in the park”. According to the Oxford dictionary, a vehicle is “a thing used for transporting people or goods, especially on land”. On this definition, cars, trains and Segways would all equally qualify as “vehicles”. Adopting a literal approach, one might conclude that prams were banned from the Botanic Gardens. But few would understand the rationale for such a rule. The purpose and context in which the words are found are therefore of paramount importance in assessing what meaning is to be ascribed to them. If, for example, the law was found in a statute dealing with motorised vehicles or pollution, the meaning of “vehicle” would be clearer. The simple point is that the inherent imprecision of language renders a written text liable to be interpreted in many different ways.

B. Imperfect Information

The second point ought not to be controversial. Drafters work without the benefit of hindsight and with imperfect foreknowledge. It is possible to classify information relating to the future into two categories made famous by the former US Secretary of Defence, Donald Rumsfeld, namely “known-unknowns” and “unknown-unknowns”. “Known-unknowns” refer to events for which the exact outcome is not known with certainty, but the likelihood and consequences of its occurrence are known in advance. For example, we could today determine the likelihood of it raining tomorrow even though we would not know for a fact whether it will rain. On the other hand, “unknown-unknowns” refer to events which are not at all known to the drafter at the time of drafting. For example, the fact that information could be transmitted by “tweeting” would have been an “unknown-unknown” a decade ago, as Twitter was only founded in 2006. It follows that we can confidently expect legal texts, taken on their own, to be incomplete, especially where “unknown-unknowns” are concerned.

C. Cost-Benefit Analysis

Thirdly, even assuming an impressive degree of foresight, it would seldom make economic sense to draft a complete legal text to deal with every possible eventuality. Negotiating and drafting a legal text can frequently involve considerable cost. Economists often employ the concept of marginal cost, which suggests that a party should invest an additional dollar if the value of the benefits from that investment is likely to exceed its cost. In this regard, it is useful to compare the cost and benefit at two distinct periods of time, namely, at the time of drafting (“front-end cost”) and at the time of enforcing the contract (“back-end cost”). Where the back-end cost is prohibitive, for example, if protracted and expensive proceedings are anticipated, it might be worthwhile investing at the front-end by drafting a more complete legal text. Insofar as front-end cost is concerned, a further distinction can be drawn between drafting rules and standards, with rules being generally more costly to draft but cheaper to enforce than standards are. Given the variable ways in which these considerations might be balanced in a cost-benefit analysis, it is entirely rational to conclude that legal texts will often be imprecise or incomplete.

Cognitive Limitations

The last point relates to the limits of the human brain in handling information. I can illustrate this point with a familiar example. Because there is a finite limit to the storage and processing abilities of the human brain, no matter how diligently you might have prepared for the examinations, there remains a chance that you might not recall all that you thought you had memorised. In other words, try as we might, we are always liable to miss out on certain points that we would have included had we thought about it or remembered it at the time the text was being concluded.

Conclusion: Why the need for interpretation arises

When all four factors are considered together, it is easy to see why disputes over interpretation are bound to arise. The imperfect mind, bedevilled with imperfect foresight and knowledge, and subject to economic constraints, directs the drafting of a legal text using language that is inherently imprecise. At the time of the dispute, however, it is interpreted by judges who are blessed with the benefit of perfect hindsight. There is thus an inevitable mismatch between the circumstances in which the drafter says what he thinks he means, and those in which the court says what it thinks the drafter meant by what he had said.

Of course, the foregoing factors do not completely explain why disputes arise. Rather, this is often down to the fact that where the parties perceive a gap or ambiguous language that is not ideally suited to deal with the actual situation at hand, they have the incentive to behave opportunistically. This problem will be exacerbated where both parties are relatively optimistic about the merits of their respective cases. While it is possible to constrain a party’s incentive to act opportunistically, such methods are imperfect. Seen in this context, the volume of cases before the courts which involve the interpretation of legal texts is unsurprising.

IV. THE PRINCIPLES OF INTERPRETATION

In a perfect world, the court would have no difficulty saying what the legal text means because there would be equivalence between what the text said, and what its drafters meant. But this is the stuff of fantasy. In the imperfect world in which we live, should the court limit itself to saying what the meaning is of the text before it, or should it go further and say what it considers the drafters meant even if, occasionally, that does some violence to what they had, in fact, said? There are parallels in the way the law has developed its response to this question in statutory and contractual interpretation. Broadly speaking, the law has evolved from a literal approach, to what is commonly known as the purposive approach (in statutory interpretation) and the commercial or contextual approach (in contractual interpretation).

Principles of Statutory Interpretation

In the very early days in the UK, statutes were drafted by judges, who would rely on their prior knowledge in aid of interpretation. Subsequently, as the roles and powers of the executive, legislature and judiciary became more defined, the literal approach towards statutory interpretation came to dominate. The prevailing attitude in the 19th century is perhaps aptly summarised by Pollock, when he observed that “Parliament generally changes law for the worse, and … the business of the judges is to keep the mischief of its interference within the narrowest bounds”. In particular, an exclusionary rule (which eschewed the use of Parliamentary materials to construe the language of an Act) developed out of a blend of constitutional and practical objections against the use of such extrinsic materials. However, the exclusionary rule was slowly but surely abrogated with the passage of time, and this can be illustrated by a series of Lord Denning’s judgments. In 1958, Lord Denning upheld the exclusionary rule, stating that “we do not refer to the legislative history of an enactment”. However, by 1971, Lord Denning took to Hansard to ascertain the mischief of the Act he was interpreting. And in 1983, Lord Denning took things a step further by relying on his own research of Hansard even though counsel had not been allowed to present their arguments on such material given the exclusionary rule. In his judgment, Lord Denning noted:

In most of the cases in the courts, it is undesirable for the Bar to cite Hansard or for the judges to read it. But in cases of extreme difficulty, I have often dared to do my own research. I have read Hansard just as if I had been present in the House during a debate on the Bill. And I am not the only one to do so.

Not infrequently, one begins by swimming against the tide and then finds that the tide has turned. The exclusionary rule was eventually relaxed in authoritative terms by the majority decision of the House of Lords in Pepper v Hart . Lord Browne-Wilkinson, delivering the majority judgment, thought it would be artificial for the courts to apply “highly technical rules of construction” when faced with more than one possible meaning of a statutory provision, when Parliament could have considered the very question as it passed the legislation. Lord Browne-Wilkinson was putting, more formally, what Lord Denning had earlier said when he observed that judges should not “grope about in the dark for the meaning of an Act without switching on the light[s]”.

The position in Singapore is largely similar to that which obtains in the UK. Before s 9A of the Interpretation Act was introduced, there had been no principled way to determine whether the literal or purposive approach should be applied in a given case. Further, the courts were generally reluctant to rely on Parliamentary materials to aid the interpretation of statutes until s 9A was enacted in 1993, shortly after Pepper v Hart was decided. This now mandates the use of the purposive approach in statutory interpretation, which the courts have confirmed and applied in numerous subsequent cases.

Principles of contractual interpretation

In contractual interpretation, the law has likewise shifted from a literal approach to a broader, contextual (or commercial) approach. In tracing the developments in this field, Lord Steyn aptly observed as follows:

[T]here has been a shift from strict construction of commercial instruments to what is sometimes called purposive construction of such documents … About the fact of the change in approach to construction there is no doubt … [As] Lord Diplock … observed:

… if detailed semantic and syntactical analysis of a word in a commercial contract is going to lead to a conclusion that flouts business common sense, it must be made to yield to business common sense.

21. As in the area of statutory interpretation, an analogous exclusionary principle, commonly referred to as the parol evidence rule, developed in the area of contractual interpretation to exclude extrinsic evidence in aid of interpretation, subject to various exceptions.33 In Singapore, the admissibility of extrinsic evidence is generally governed by the rules of evidence, which can be found first in the Evidence Act34, and then in the common law.35

22. The genesis of the modern contextual approach towards contractual interpretation can be traced to Prenn v Simmonds, where Lord Wilberforce observed that “[t]he time has long passed when agreements … were isolated from the matrix of facts in which they were set and interpreted purely on internal linguistic considerations”.36 Since Prenn v Simmonds, a large (and still growing) body of case law has developed to decipher what this “contextual”, “commercial”, “common sense” or “modern” approach involves.37 The leading decision in this regard is Investors Compensation Scheme38, where Lord Hoffmann (in his oft-cited judgment) laid down five principles of contractual interpretation. In his view, interpretation is the

33 See V K Rajah JA, “Redrawing the Boundaries of Contractual Interpretation: From Text to Context to Pre-text and Beyond” (2010) 22 Sing. Ac. L.J. 513 at 514.

34See, especially, ss 94–99 of the Evidence Act (Cap. 97, 1997 Rev. Ed. Sing.).

35 Sembcorp Marine Ltd v PPL Holdings Pte Ltd and another and another appeal [2013] SGCA 43 [Sembcorp Marine] at [65(b)].

36Prenn v Simmonds [1971] 1 W.L.R. 1381 at 1383–1384.

37McMeel, supra note 10 at para. 1.39.

38 Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 W.L.R. 896 [Investors Compensation].

12

“ascertainment of the meaning which the document would convey to a reasonable person having all the background knowledge which would have been available to the parties in the situation in which they were at the time of the contract”. 39 The Singapore courts have likewise embraced the applicability of the “contextual” approach in contractual interpretation, although I should caution that there are some important differences in the way the law has developed in the UK and Singapore.40

Against this background, I turn to survey some of the challenging issues inherent in the “purposive” and the “contextual” approaches.

V. WHAT ARE SOME OF THE CHALLENGES OF THE “PURPOSIVE” APPROACH AND THE “CONTEXTUAL” APPROACH?

The purposive approach considered

Section 9A of the Interpretation Act mandates the courts to prefer “an interpretation that would promote the purpose or object underlying the written law…”. In so doing, the courts are entitled to consider the extrinsic materials listed under s 9A(3), including any explanatory statements to the Bill or any relevant material in the Parliamentary debates. The Court of Appeal has held that reference to such materials is permitted even “if the words of the statutory provision are unambiguous or do not produce unreasonable or absurd results”. However, in deciding whether

For reasons explained in Sembcorp Marine, supra note 35 at [33], Lord Hoffmann’s comments are perhaps more appropriately restated in terms that interpretation is “the ascertainment of the meaning which the expressions in a document would convey to a reasonable person” instead [emphasis added] or not to consider such extrinsic materials, or the weight to be accorded to them, the courts should have regard to “the desirability of persons being able to rely on the ordinary meaning … taking into account its context … and the purpose or object underlying the written law”, and also, “the need to avoid prolonging legal or other proceedings without compensating advantage”.

The reasons for preferring the purposive approach are simple. Given the inherent inadequacies of the draftsman’s foresight and the imprecision of the linguistic tools at his disposal, a rule that allows the court in some way and to some degree to bridge the gulf in time and to have an insight into what the draftsman was hoping to achieve seems entirely reasonable. Lord Steyn put it thus:

No explanation for resorting to purposive interpretation of a statute is necessary. One can confidently assume that Parliament intends its legislation to be interpreted not in the way of a black letter lawyer, but in a meaningful and purposive way giving effect to the basic objectives of the legislation.

But the purposive approach is not without its critics. Lord Steyn himself had been fairly robust in voicing some objections in his extra-judicial analysis of Pepper v Hart, resting these on both constitutional and practical grounds. The principal objections can be summarised thus:

(a) The constitutional objections: Under the constitutional framework, it is Parliament’s role to legislate and the courts’ role to interpret the words enacted. The intention of Parliament is in truth nothing more than a “chimera” or a “mythical beast” and can be manipulated by judges to enable them to reach their desired ends. Furthermore, care should be taken not to cast what are in fact the views of members of the Executive, as Parliament’s intention. In any event, those subject to the law should be able to ascertain the scope and effect of the legislation easily, and by reference to sources that are publicly available, authoritative and final.

(b) The practical objections: Allowing Parliamentary debates to be cited before the courts would “add greatly to the time and expense involved in preparing cases involving the construction of a statute”. In most cases, such a search “would throw no light on the question before the court”. Further, it might not always be practicable for counsel to obtain access to older reports of Parliamentary debates.

It is noteworthy that Lord Mackey’s dissent in Pepper v Hart was grounded on practical rather than on constitutional grounds. His Lordship was concerned about the possibility of an “immense increase in the cost of litigation” and felt it prudent to study whether the relaxation of the exclusionary rule would result in a substantial increase in cost before acting. Experience has shown that Lord Mackay’s concerns have not been borne out. The purpose of a statute is often easily ascertained, especially when compared to contracts, and this is facilitated by the availability of a reliable and official record of Parliamentary proceedings in the form of Hansard. Moreover, the technological advancements post-Pepper v Hart have mitigated most practical objections as our Parliamentary reports are now available and searchable online. The difficulties relating to some of the older reports might remain but these are likely to become less acute with time as statutes are amended to maintain their relevance. Insofar as the problem involves the introduction of irrelevant material before the courts, this is a practical problem that infects all proceedings, not just those involving statutory interpretation. Its remedy lies elsewhere, perhaps through the use of appropriate costs orders, rather than in pre-emptively ordering the exclusion of such material entirely.

Turning to the constitutional objections, it may be noted that Lord Browne-Wilkinson, in Pepper v Hart, did take these into account. See Pepper v Hart, supra note 26 at 634C-D, where

“Lord Browne-Wilkinson noted “I have come to the conclusion that, as a matter of law, there are sound reasons for making a limited modification to the existing rule (subject to strict safeguards) unless there are constitutional or practical reasons which outweigh them.” [emphasis added]”.

These are certainly not to be viewed lightly, as they call into question the constitutional role of the courts. However, as I have already noted, one has also to consider the constraints that drafters face when drafting a statute, in particular, the imprecision of language and imperfect knowledge. As statutes are durable pieces of legal text that can only be amended by Parliament, it is unsurprising that they are typically drafted at a high level of generality, with greater reliance on the use of standards rather than rules to enhance their prospect of passing the test of time. In this light, it is evident that in some cases at least, confining oneself to the four corners of the text to discern the legislative aim would be a futile exercise because the words, in and of themselves, would not suffice to answer the question. Forcing the courts to work with blinkers could actually result in decisions that are contrary to what Parliament had intended and this could give rise to even greater constitutional concerns.

Another strand of the constitutional objections relates to the way that proceedings in Parliament are conducted, leading to questions as to whether the Legislature’s “intentions” can truly be isolated or defined. Singapore has a unicameral legislature, unlike the UK: to that extent, some of the objections raised there would not apply here. In substance, the constitutional objections reflect the tension between the risk of judicial legislation on the one hand and the need to ensure that justice is actually done on the facts of each case, where the standards have to be applied to the actual situations that arise before the courts.

I can illustrate how we have resolved this tension by reference to two cases. First, in Dorsey James Michael, the respondent commenced an action to administer pre-action interrogatories on the appellant. The High Court Judge granted the order in part and the appellant appealed. Under the Supreme Court of Judicature Act, any judgment or order of the High Court is ordinarily appealable as of right. This is subject to certain exceptions including where a Judge makes an order specified in the Fourth Schedule to the SCJA,* one provision of which states “No appeal shall be brought … where a Judge makes an order giving or refusing interrogatories”. The respondent argued that the plain and ordinary meaning of the term “interrogatories” encompassed pre-action interrogatories and therefore applied to strike out the appeal.

The Court of Appeal rejected the respondent’s contention that the “plain meaning” was correct or applicable in this instance. It examined the purpose of the amendments introducing the relevant provision into the Fourth Schedule. From the Minister’s Second Reading speech, it was clear that the amendments restricting appeals to the Court of Appeal were intended to apply only to orders that were made at the hearing of interlocutory applications. An application for pre-action interrogatories is not an interlocutory application as leave to administer the pre-action interrogatories constitutes the only relief sought in the entire action. Once the application has been ruled upon by the court, the proceedings conclude and would not be followed by any other steps leading to any trial or further disposal of that matter. This case brings to life the dangers of trying to construe the purpose(s) of a statute by relying only on the words used, without regard to any extrinsic material.