From verba to voluntas

(a) Pre-dassical Roman law

The two maxims just mentioned reflect the struggle between verba and voluntas that for a long time dominated any discussion about the interpretation of contracts in Roman law.

¹ Examples in Schulz, RLS, p.

^E Festus, De verborum significatione. s.v. Ver sacrum, who continues: "[S]ed cum crudele vidcretur pucros ac puellas innocentes intcrficerc, perductos in adultam aetatem velabam atque ita extra fines suos exigebant." On this incident, see Honsell, Festschrift Coin%. vol. I, p. 139.

For recent criticism of the style of English legislative drafting, see, for instance, the remarks by two distinguished Hamlyn lecturers: Tony Honore, The Quest For Security: Employees, Tenants, Wives (1982). pp. 118 sqq. and P.S. Atiyah. Pragmatism and Theory in English Law (1987), p. 31 ("fM]ost legislation is not drafted in the form of a statement of true principle.... Much of it is drafted in the form of a set of specific rules, ad hoc solutions to particular problems. Nobody would read a modern English statute for its literary elegance as it was said that Flaubert used to read the French Code Civile.... Even when we do use legislation, an instrument well suited to the enactment of broad principles and generalisations, we find ourself so shackled by the traditional common law methodology, that we fail to use legislation in an effective and principled manner.... In particular, the detailed and crabbed style of legislative drafting means that it becomes almost impossible for the courts to draw principles from legislation, to treat legislation as a living graft on the common law, and to develop the law as an integral whole").

A considerable body of South African legislation is derived, either directly or indirectly, from English statutes, and so, too, is the South African style of legislative drafting. Statutes tend to be clumsily drafted, since the legislator anxiously tries to provide for every conceivable eventuality himself rather than to leave anything to the good sense of those interpreting the statute. Particularly odd, from a Continental perspective, are the long lists of (usually rather unhelpful) definitions with which many statutes commence (cf., for instance, Act 19/1893 where the term "banker" was defined as including "a body of persons... who carry on the business of banking"; on which, see National Housing Commission v. Cape of Good Hope Savings Bank Society 1963 (1) SA 230 (C) at 233). There is even an Interpretation Act (33/1957) which provides illuminating insights such as that "Christian name" means any name prefixed to the surname, whether received at Christian baptism or

(b) Post-classical jurisprudence

At the other end of the development of Roman legal science stood an exaggerated emphasis on subjective criteria.¹⁴ It was closely related to the final disappearance of the old word formalism and to a general trend to judge all human behaviour according to moral, more particularly Christian standards.'⁵ The content and character of every contract were determined by the intention of the contracting parties, and this intention to attain a specific legal consequence was usually referred to as "animus". According to Fritz Pringsheim, whose research has fundamentally shaped our views on the post-classical animus doctrine,¹⁶ the Byzantines were the first to attribute to the parties an intention directed towards procuring specific legal ends and to make the intention supreme, even where it was unexpressed and un- demonstrable. The Byzantine predilection to animus is obvious and unchallenged. Indisputably, it is the source of a considerable amount of interpolation of classical texts.

Classical Roman law did not normally concern itself with problems of evidence; the taking of evidence and its evaluation could be left to the

not, that "month" means a calendar month, or that "law" means any law, proclamation, ordinance, Act of Parliament or other enactment having the force of law. This approach to legislation neither presupposes nor engenders a particularly liberal approach to the interpretation of statutes. And, indeed, statutory interpretation in South Africa, as in England, is traditionally governed by the so-called "golden rule": "The rule by which we are to be guided in construing acts of Parliament is to look at the precise words, and to construe them in their ordinary sense, unless it would lead to any absurdity or manifest injustice" (Perry i>. Skinner (1837) 2 M & W 471 at 476); for all details, as far as South African law is concerned, cf. L.C. Stcyn, Die Uitleg van Wette (5th ed., 1981); Lourens M. du Plessis, The Interpretation of Statutes (1986). The equivalent of the "golden rule" for contract interpretation is the "plain meaning rule" (cf. supra, note 7; for details, see Liideritz, op. cit., note 2, pp. 65 sqq.) which applies to all those transactions that have been laid down in a document (no matter whether the formality is required by law or not). The plain meaning rule and the closely allied (procedural) "parol evidence rule" serve to protect reliance on the contractual declaration and are an expression of the objective (declaration-oriented) approach of the English courts to contract interpretation. According to the parol-evidence rule a written instrument may not be contradicted, added to or varied by oral evidence (cf.

ⁿ Kaser, RPr II, pp. 82 sqq.

^B Cf. generally Kaser, RPr II, pp. 60 sqq.; Biondi, DRC, vol. II, pp. 1 sqq.

^B Cf. particularly his article on "Animus in Roman Law", (1933) 49 LQR 43 sqq., 379 sqq.

¹⁷ (1933) 49 LQR 48.

^B Pringsheim, (1933) 49 LQR 48.

^B Giuseppe Gandolfi, Studi sull'ittterpretazione de%li atti negoziali in diritto romtmo (1966), pp. 243 sqq.; Franz Wieacker, (1966) 83 ZSS 437 sq., 444 sq. discretion of the judge (normally not a professional lawyer), who decided the issue not on the basis of fixed rules of proof but on that of his personal authority.²⁰ This was different in the so-called bureau­cratic, post-classical period of Roman jurisprudence.

"The innate tendency of every bureaucracy to convert the development of the law into the monopoly of a central office, to codify the law and to assure and supervise its strict application and enforcement, produced a complete change in the structure of Roman legal science."²¹

Hence the tendency to tie the judge to specific rules of evidence and to emphasize the problem of proof.²" "Nisi alia mens testatoris probetur",²³ "si voluntas testatoris... evidenter non ostenderetur":²⁴ these are the kinds of clauses which Justinian inserted into the classical texts. On the one hand, therefore, he instructed the judge to take account of the (real) intention of the parties; but at the same time he stressed the necessity that such an intention must be proved beyond reasonable doubt.

(Thus, what one has to guard against primarily is the temptation to put the past into the straitjacket of contemporary conceptions.^[3240] Classical Roman lawyers tended to look at legal transactions as a whole,^[3241] without isolating their individual compo­nents or dogmatizing the objective or subjective elements contained in them. The analysis of contract as being composed of two correspond­ing declarations of intention, both of which in turn are made up of an internal component (the intention—voluntas) and of an external one (the declaration—verba) is, as we have seen, of a much later date.^[3242] The Roman jurists did not think in terms of stereotyped categories; that obviously makes it difficult to generalize their solutions and to extract crisp and clear-cut formulae from our sources. Most remarkable and characteristic, probably, is the flexibility with which questions of interpretation were approached and decided.

Already by the time of the later Republic, the narrowness and rigidity of the old literal approach to interpretation had been left behind.^[3243] To a considerable degree, this was a consequence of the relaxation of the "external" formalism of Roman law. Trade links with non-Romans facilitated (sub voce "ins gentium")^[3244] the introduction and increasing popularity of new, informal types of transactions: transactions which no longer found their effective cause in compliance with a specific formality, but in the consensus, or the conventio,^[3245] of the parties. Thus it was with the consensual contracts that the wider and more liberal manner of interpretation gained ground. Here one did not necessarily have to stick to the meaning that was typically associated with the words contained in formal declarations, but one could freely take into consideration—and was, on account of the "ex bona fide" clause even encouraged to do so—the individual circumstances of each case. As in the case of error, testaments were the first formal type of transactions where the more modern, flexible approach established itself,^[3246] and by the time of high classical law, it prevailed with regard to formal contracts too.^[3247] "Nullum esse contractum, nullam obliga­tionem, quae non habeat in se conventionem": no matter whether a contract of sale, a mutuum^[3248] or a stipulation^[3249] had been concluded, the transaction was ultimately based on the consent of the individual parties concerned.

(c) The causa Curiana: the case before the court

The most prominent individual turning point within this development, incidentally, was the causa Curiana,^[3250] argued in 92 B.C. before the centumviral court.^[3251] It had all the ingredients of a sensational cause celebre, for it involved a clash between Lucius Licinius Crassus, the leading orator of his time,^[3252] and Quintus Mucius Scaevola, Pontifex Maximus, Consul in 95 B.C., author of an influential treatise on the ius civile^[3253] and probably the most outstanding jurist of the late Republic. The case was of particular significance, since it concerned the interpretation of a testament, i.e. a strictly formal legal act. It must have occurred quite frequently that a paterfamilias instituted as his heir a son who had not as yet attained the age of puberty. When the testator died, the pupillus (if such he still was) would, of course, become sui iuris, but he would not be able to make a testament himself: for persons under the age of puberty lacked the ius testamenti faciendi.^[3254] Thus, the danger existed that the testator's property might ultimately pass ab intestate Considering the specific importance attached by upper-class Roman citizens to testate succession (designation of an heir was regarded as the social duty of every bonus paterfamilias),^[3255] that was a result to be avoided. The paterfamilias was therefore allowed to appoint a substitute who was to become heir in case his son died as a pupillus. Such a clause, contained in the will of the paterfamilias, was known as a substitutio pupillaris, and it ran like this: "Titius filius meus mihi heres esto; si filius meus prius moriatur quam in suam tutelam venerit, tune Seius heres mihi esto."⁴*¹ A substitutio pupillaris thus covered the situation where the heir died before attaining the age of puberty; whether this occurred after he had accepted the inheritance or before the paterfamilias himself had died was immaterial. It did not cover the situation where the son did not in fact die at all for the simple reason that he had never been born.

This, however, is exactly what had happened in the causa Curiana: a testator by the name of Marius Coponius had instituted as heir a son whom he had only hoped to have;^{[3256] [3257] [3258] [3259] by way of a substitutio pupillaris, he had nominated a certain Curius as a substitute. But what Coponius had taken for granted did not materialize: for when he died, he neither had a son, nor was a postumus on its way. In such a case, a simple substitutio vulgaris would have helped. This was the appropriate device if the testator wanted to institute a third person as an alternative heir in the event of Titius (i.e. the person appointed in the first place) for some or other reason failing to inherit: be it because he refused to accept the inheritance, be it that he had predeceased the testator or that he had never been born. In order to be on the safe side, a testator who wanted to institute a future son was therefore well advised to combine a substitutio vulgaris (to cover the possibility that a son would not be born) and a substitutio pupillaris (to provide for the eventuality that by the time of the testator's death a son had been born but had died before having been in a position to make a will).[3260] A substitutio vulgaris was, however, not contained—expressis verbis—in the will before the court in the causa Curiana. The question therefore arose whether it could not possibly be read into the substitutio pupillaris. Scaevola, the jurist, rejected such a proposition; and, in fact, on a strict and literal reading of the substitutio pupillaris he was undoubtedly correct.[3261] The condition under which Seius had been instituted as alternative heir had not been met.}

(d) The causa Cunana: jurists and orators

But such a narrow construction no longer commended itself to the court. It was Crassus, the orator, who carried the day.^{[3262] [3263] He placed the emphasis on the overriding intention of the testator, rather than on the wording of the will. The testator had evidently wanted his son to inherit, but had regarded Seius as his second choice. This is why he had instituted Seius to become heir after his son. The possibility that he might not have a son had evidently not occurred to him. It is quite obvious, however, that he would have preferred Seius to inherit his estate in this event too, rather than to die intestate. Undoubtedly, therefore, it was in accordance with the testator's (unexpressed and hypothetical) intention, and generally in his interests to construe the substitutio pupillaris as containing a substitutio vulgaris.[3264]}

The causa Curiana, no doubt, was only one milestone within a long development, and Crassus' victory did not mean that henceforth a will- oriented, subjective or individualizing approach had definitely superseded or replaced the old and rigid literal interpretation.^[3265] In fact, throughout classical Roman law we find cases where the verba prevailed against voluntas.⁵³ But it became an oft-quoted⁵⁴ precedent, a cause celebre, where at a crucial juncture a free exegesis had triumphed over rigidity. For this is probably the most important feature of the causa Curiana: it demonstrates the flexibility that had been gained by the time of the late Republic. A case no longer necessarily had to be decided on the basis of the verba; other factors were taken into account and, where appropriate, they could even determine the result of the decision.

The causa Curiana is also characteristic of the influence of Hellenistic theories of forensic rhetoric on contemporary Roman jurisprudence.⁵⁵ Roman court orators adopted the Greek "status" doctrine,⁵⁶ particu­larly the antithesis of verba fscriptum) and voluntas (aequitas). Rhetoric, of course, was a theory of advocacy, and the orator had to adopt whichever view suited the interests of his clients.⁵⁷ Thus, he was not necessarily on the side of voluntas or aequitas; it could just as well be his duty to plead a case on the basis of a strictly literal interpretation. ⁵⁸ The function of the Roman jurist was a different one. ^[3266]

decision too much. For a rejection of Stroux's view cf. also Feliciano Serrao, Ctassi, Partiti e Legge neila Repubblica Romana (1974), pp. 142 sqq.; Burge, op. cit., note 11, pp. 46 sqq.

For the interpretation of wills cf. e.g. Wieling, Testamentsauslegung, op. cit., note 36, pp. 59 sqq., 107 sqq. and passim.; cf. also Felix Wubbe, "Der Wille des Erblassers bei lav. D. 32, 100, 1", in: Iuris Professio, Festgabe fur Max Kaser (1986), pp. 371 sqq.

³¹ Cf. the references to Cicero and Quintilian in Schulz, RLS, p. 79, Gandolfi, op. cit., note 19, pp. 291 sqq. and Wieacker, (1967) 2 The Irish Jurist 157 sqq.

^s Cf. generally Stroux, op. cit., note 40, passim (with too extreme and far-reaching conclusions); contra (equally extreme) e.g. Gerhard von Beselcr, "Recuperationes iuris antiqui", (1938) 45 BIDR 169 sqq.; cf. also Schulz, RLS, pp. 76 sqq.; Behrends, Fraus legis, pp. 73 sqq.; for a more balanced evaluation, see Wunner, Contractus, pp. 182 sqq.; Kaser, RPr I, jp. 236; Honsell, Festschrift Coing, vol. I, pp. 143 sq.; and, in particular, Franz Wieacker, "Ober das Verhältnis der romischen Fachjurisprudenz zur griechisch-hellenistischen Theorie", (1969) 20 Iura 469 (on the whole, possibly still underrating the influence of rhetoric); Frier, Roman Jurists, pp. 95 sqq., 127 sqq. ("The Ciceronian court, with its shameless tattoo of loci communes, formed... an indispensable laboratory where Rome's fledgling legal scientists could create and test their abstract rules in relation both to specific cases and community values.... It seems reasonable to discover, within the intensely competitive arena of forensic discussion, many of the impulses towards breadth and equity which gave Roman private law its vitality as a living system and its vast influence as a dead one" (pp. 137 sq.); cf. also p. 267: "This indirect and selective flow of legal ideas from judicial oratory into law is perhaps... the real basis Roman law's growth into maturity"). Cf. now also the magisterial summary of the discussion by Wieacker, RR, pp. 662 sqq.

³⁵ For details cf. Uwe Wesel, Rhetoristhe Statuslehre und Gesetzesauslegung der romischen Juristen (1967), pp. 22 sqq.; Franz Horak, "Rhetorische Statuslehre und der moderne Aufbau des Verbrechensbeg riffs", in: Festgabe für Arnold Herdlitczka (1972), pp. 121 sqq.; Richard A. Bauman, "The 'leges iudicorum publicorum' and their interpretation in the Republic, Prindpate and later Empire", ANRW, vol. II, 13 (1980), pp. 112 sqq.; Wieacker, RR, pp. 669 sqq.

⁵⁷ Schulz, RLS, pp. 76 sq.; cf. also Gandolfi, op. cit., note 19, pp. 257 sqq.; Wieacker, (1969) 20 lura 475; Burge, op. cit., note 11, pp. 58 sqq.

^s On the "relativistic framework of rhetorical argument" see Frier, Roman Jurists, pp. 127 sqq. ("All that rhetoric offered was a loose framework of alternatives for interpretation"); Frier provides a brilliant and fascinating analysis of Cicero's speech pro Caecina which, according to him, reflects the crucial transformation that the Roman judicial system He was concerned with questions of law, not with the selection and skilful presentation of an argument that could possibly strengthen the case of a particular client. "The question of law, namely which of the two interpretations, the literal or equitable, ought to prevail", states Schulz,^{[3267] 59 [3268]} "was simply outside the province of rhetoric"; and he concludes that the Roman jurists cannot have found anything worth learning in rhetoric. But this view is much too negative. The elements of rhetoric were taught at school,^[3269] and every Roman jurist must therefore, from an early age, have been familiar with the relevant "status" and techniques of argument. The speeches of the orators in court merely reminded them of what they already knew themselves: namely, that, as for many other problems, more than one point of view can be adopted on the question of the interpretation of contracts, wills or statutes. The jurists were constantly made aware of the arguments for and against a literal interpretation, and it is hard to believe that this should not have affected them at all. Of course, they still had to decide which argument was to prevail. But the important progress that had been made lay in the fact that both rigid and wide, both objective and subjective interpretation could now prevail.^[3270]

(e) The "individualizing" approach

The causa Curiana has taken us into the field of the law of succession. So do most of the cases contained in the Digest which deal with problems of interpretation. They would all have to be considered for a more thorough appraisal of the approach of the classical Roman lawyers on this matter; for strict dogmatic borderlines between the interpretation of contracts, of testaments, and even of statutes did not exist.^{[3271] [3272] [3273] [3274] [3275] [3276] Of course, the Roman lawyers were aware of the differences between the individual types of transactions: the conflict of interest between declarant and recipient of the declaration in the case of contract; strict observance of the prescribed formalities (but also: no reasonable reliance on the part of any addressee that needed to be protected!) in the case of wills. We do not find any express statement in the sources analysing or clarifying these policy issues, but they are reflected in the way the Roman lawyers dealt with the individual cases brought before them. The dominant impression is the highly "individualizing" approach adopted throughout,64 and as far as the construction of wills was concerned, that entailed specific significance of "mens testatoris" or "quid sensit testator".65 In the case of contract, it was a slightly different criterion by which the process of interpretation was guided. "Id quod actum est" is the phrase that we find emphasized again and again. The felling of timber was sold for five years. Whose is the mast which might fall from the trees? "[P]rimum sequendum [est] quod appareret actum esse.1,66 A piece of land had been sold, and the parties had provided that the water rights should go with it. Does that include the right of way to the water? "[R]espondit sibi videri id actum esse."67 Some water pipes were supposed to pass into the property of the purchaser of an estate. What about the reservoir from which the water is drawn through the pipes? "[R]espondi apparere id actum esse, ut id quoque accederet, licet scriptura non continetur."68 Or, to take the case of a stipulation: somebody had promised to give or to perform something "kalendis lanuariis". Which date did he have in mind? " [S]i autem non addat quibus lanuariis, facti quaestionem inducere... quid inter eos acti sit [utique enim hoc sequimur quod actum est]."69}

if) Id quod actum est

Id quod actum est⁷⁰ referred to the common intentions of the individual parties to a contract, as they became apparent from the specific context within which the negotiations had taken place and the declarations had been made. This "context" included the individual, personal circum­stances of the contracting parties as well as those of their expectations that were based on other than purely internal motives. The literal meaning of the words used by the parties and other objective standards were relevant wherever they could help to determine id quod actum est; otherwise they were only fallen back upon where the individual circumstances remained in the dark ("si non appparet quid actum sit").⁷¹ In this connection, certain standard arguments, based on general ^{[3277] [3278] experience or policy, could also be of assistance:72 "commodissimum est id accipi, quo res de qua agitur magis valeat quam pereat";[3279] 73 "[i]n obscuris inspici solere, quod verisimilius est aut quod plerumque fieri solet"[3280] or the well-known interpretation contra proferentem.[3281] They were all of a fairly subordinate significance in Roman law; it was not until the time of the glossators that they began to be used as general maxims of interpretation.}

All in all, I think that the position under classical Roman law can with some justification be described as an (unstable but) happy equilibrium.^[3282] Neither verba nor voluntas reigned supreme: the original archaic formalism in interpretation had been left behind, but it had not (yet) been replaced by an equally radical subjectivism. The advance of voluntas not only started in post-classical times; it in fact contributed to the refinement of classical law. The Byzantine animus theory was therefore neither novel nor intrinsically false: "The idea is an old and a great one; it is only its exaggeration that creates the novelty or the danger" and that turns what was once progress into decline.^[3283]

(g) Excursus: the animus novandi

Thus, not even every reference to animus is spurious. It was often only by isolating and dogmatizing it that Justinian interfered with the sources of classical law. Merely by way of example,^[3284] we may here refer to the problem of the animus novandi. Novation was defined by Ulpianus as "prioris debiti in aliam obligationem... tnmsfusio atque translatio".^{[3285] [3286] An existing obligation was extinguished and substituted by an new one. A novation was effected by way of a stipulation that was causally framed; it referred to what was owed under the previous obligation. 0 It was often used, for instance, in order to achieve a change in the person of the debtor or of the creditor. "Quod mihi Seius debet, mihi dari spondes?" "Spondeo":[3287] [3288] [3289] on account of this stipulation the promisor replaced Seius as the debtor of "ego". Seius' obligation was terminated ipso iure. In order to have this "novatory" effect, the (new) stipulation basically had to comprise the same debt (idem debitum), but it also had to contain a new element (aliquid novi).82 In the above example, the change of creditor is the new element. Often this pair of objective criteria (idem debitum—aliquid novi) was both satisfactory and sufficient to determine when a particular stipulation had to be viewed as a novation. But sometimes it was not. In our discussion of suretyship stipulations we have seen that a sponsio could, according to Proculian practice, be concluded both ex intervallo and in the absence of the main debtor.83 That entailed a change of the standard formula, which could now no longer refer to idem, but had to read, for instance, like this: "Quod Seius mihi debet, mihi dari spondes?" It was exactly the same as in the case of a novation. And yet, in the one instance (sponsio) the promisee was to become debtor alongside Seius, in the other (novatio) he was to replace him. It was in order to deal with these kinds of problems that the classical lawyers brought in a new, subjective requirement: they made the decision dependent upon the intention of the parties, and this was the origin of animus novandi.[3290] [3291] [3292] [3293] [3294] In some of our sources even the very term appears to be of classical origin, although, as a rule, a formulation such as "hoc agere, ut (novetur)" was used.83 Thus, for instance, Ulpian amplifies his definition of novation with the words: "... hoc est cum ex praecedenti causa ita nova constituatur, ut prior perematur. 1,86 As a result, therefore, the construction of a stipulation as a novation depended on a characteristic mixture of both objective and subjective criteria. It was only in post- classical times that animus novandi became the dominant feature in the dogmatic make-up of this institution and conclusively determined the novatory effect of a transaction.[3295]}

III.