download in pdf


Iván Siklósi*


Private Autonomy and Its Restrictions in Roman Law: An Overview Regarding the Law of Contracts and Succession


I Introduction

According to the famous statement of Heinrich Heine, Corpus iuris is the ‘Bible of egoism’. Although Heine’s conclusion is somewhat excessive, it is out of question that private autonomy had great importance in Roman law, and not only in private law,[1] in which autonomy had a fundamental importance (especially in the law of contracts, which primarily contains ‘dispositive’ rules, from which the parties may differ by mutual consent, and in the law of succession, considering the principle of testamentary freedom)[2], but in public law[3] as well.

‘Autonomy’ is a modern concept (the idea of autonomy and individualism was only developed in the legal science of the 19th century[4])[5] but its application to Roman law cannot be considered as anachronistic or unhistorical. In Roman society, privacy was largely respected. The private law legislation intruded into the private sphere relatively rarely but, if the legislator did so, the relating legal norms often became unpopular (in this regard, we can refer e.g. to the Augustan laws on family relations[6]).

As an example of Roman aversion to the law’s intrusion into the private sphere, one can refer, inter alia, to the problem of expropriation. Expropriation did not win a wide range of applications in classical Roman law.[7] It tended to be applied instead in the age of the later Roman Empire, which was not so ‘sensitive’ to private sphere.[8]

It is beyond question that the autonomy of the will shows its clearest expression in the law of contracts and in the law of succession. In this paper, we only want to deal with a few problems of autonomy in the context of the law of contracts and law of succession, with somewhat generalised references to some famous topics of Roman law in the context of private autonomy and its restrictions.


II Some Questions of Private Autonomy in the Roman Law of Contracts

The question of autonomy in the context of the law of contracts is a very complex issue.

On the one hand, the law of contracts is based, in theory, on the freedom of parties, for whom it is allowed to decide to enter or not to enter into a contractual relationship; and, if they want to, they are permitted to decide, in what type, form, and content to conclude the contract. The contracting parties may differ from the provisions of positive law if it is not prohibited by a mandatory rule (ius cogens – see e.g. the provisions limiting interest rates).[9] This is the essence of the modern principle of contractual freedom.

On the other hand, formalism (ritual forms) had a great importance in the archaic age of Roman law. In this period, the essence of contract was not the individual will of the parties but the rite. The fundament of the binding force of a contract was the form containing ritual elements (see, for example, the institution of sponsio which was a contract of a ritual nature). The principle ‘voluntas mater contractuum est’ [‘(contractual) will is the mother of conventions’] is only valid in the developed Roman law, in which the contractual will of the parties (voluntas) has great importance, and in which many contracts are formless. Nevertheless, the ‘requirement of standardisation’ (Typenzwang) remained in the whole history of Roman law, although it had been significantly dissolved by means of (praetorian) actiones in factum conceptae; by means of expanding the scope of actionable pacts; or by means of giving an actio praescriptis verbis, etc. The principle of ‘pacta sunt servanda’ was not the outcome of Roman law but this fundamental principle was developed by the scholars of canon law and, later, by the representatives of natural law in the modern age.[10] The distinction between ‘pactum’ and ‘contractus’ has an important role in Justinianic law, too: only parties to agreements with a so-called civilis causa (the legal basis for the contract to be sued) were legally compellable, even in this period of Roman law.

Apart from contractual formalities, the private autonomy of the contracting parties was respected, and it was only rarely restricted.

According to the famous sentence of Cervidius Scaevola, ‘ius civile vigilantibus scriptum est’: civil law was written for vigilant people (Scaev. D. 42, 8, 24), i.e. it was made for those who are diligent in protecting their own rights. According to Reinhard Zimmermann, there was very little in the Roman law of contracts to limit this core feature of economic liberalism. The law merely provides the framework within which the individuals may operate.[11]

Except in, for example, the system of causes of invalidity or the problems of novatio, this paper only brings up the following examples of private autonomy.

a) It was permitted in Roman law to have a special agreement regarding the liability of the parties or the risks.

As for the contractual liability, it was permitted in classical and in Justinianic Roman law as well, to soften or to intensify the liability of the parties (compared to the provisions of positive law). For example, on the one hand, liability only for dolus (‘deceit’ or ‘fraud’) instead of liability for culpa (‘fault’ or ‘negligence’), or a responsibility for culpa instead of custodia-liability was allowed to be specified, and, on the other hand, the establishment of a more rigorous culpa– or an objective custodia-liability (‘safekeeping’) instead of dolus– or culpa-liability was also permitted. Nevertheless, the contractual freedom of the parties had its limits. The liability for dolus – since dolus is contrary to the objective principle of bona fides – was not allowed to be excluded; such an agreement was null and void in Roman law,[12] too. It should generally be noted that bona fides (good faith and fair dealing) can itself be considered as a limit of contractual autonomy,[13] too.[14]

As for the allocation of risks, it was based decisively on the private autonomy of the parties. An agreement regarding the risks can have a special importance, for example in a contract of sale. According to the principle ‘periculum est emptoris’, the purchaser, when the sale has been completed, must assume the risk.[15] This rule was only concerned, however, with ‘acts of God’ (vis maior, i.e. events which human weakness cannot prevent[16]), and the allocation of other contract-specific risks was the object of the agreement of the parties. Other examples of the special clauses concerning the allocation of risks can be mentioned from the sphere of locatio conductio [cf. e.g. Flor. D. 19, 22, 36 on the topic of locatio conductio operis (contract of enterprise)].

b) A good example of the restriction of contractual autonomy is the rule of laesio enormis in postclassical Roman law.

According to classical law, the determination of the purchase price was fully an object of the parties’ agreement. In this regard, the ‘economic liberalism’ and private autonomy were fully respected. Taking advantage of one another was ‘naturally’ permitted for the contracting parties (cf. Ulp. D. 4, 4, 16, 4: ‘…in pretio emptionis et venditionis naturaliter licere contrahentibus se circumvenire.’). In a transaction of purchase and sale it is naturally conceded that the parties can either purchase or sell something for more or for less, and hence mutually circumvent one another (cf. Paul. D. 19, 2, 22, 3). The law of contracts primarily provides a framework within which individuals may operate, and it does not usually have a protective function. A notable exception was the Roman legislation against usury, but it is out of the topic of the contract of sale. No attempts were made in classical Roman law to interfere with the freedom of the parties to a contract of sale to fix the price.[17]

The rescission of the sale – due to the lack of equivalence – was originally and even later excluded. However, in case of laesio enormis,[18] which was a product of the economic crisis in the end of the 3rd century, the freedom to determine the purchase price was already restricted, but only in the sale of real estate. By means of the disputed legal institution of laesio enormis, the legislator restricted the private autonomy of the parties with regard to the sale of real estate, which could be rescinded by the seller if he did not receive even half of the real value of the estate.[19] The legal construction of laesio enormis (which rule shows an entirely different approach compared to classical law) can be regarded, in our opinion,[20] as one of the cases of annulment according to ius civile in Roman law.


III Private Autonomy in the Roman Law of Succession

Private autonomy is of great importance, by nature, in the law of succession, too.

Testamentary freedom is undoubtedly one of the most important expressions of private autonomy. The testator’s freedom to make a last will was already provided by the Twelve Tables (cf. XII tab. 5, 3[21]). However, in Roman law, testamentary freedom was restricted from several aspects. Two examples are highlighted as follows.

a) The testator’s freedom was significantly restricted through the statutes on the limits of legacies. Such were the lex Furia testamentaria, which fixed the maximum amount of a legacy at one thousand assēs (this was the earliest statute setting limits for legacies[22]); the lex Voconia, according to which nobody could receive by legacy more than the heir;[23] and the lex Falcidia, which provided that legacies should not exceed three quarters of the testator’s estate.[24] These laws can be considered as sharp restrictions of testamentary freedom and, therefore, can be regarded as relatively rare signs of Roman law interventions into the private sphere.

b) Another significant limitation of testamentary freedom was the regulation considering debita portio, from the classical era of Roman law. According to this, the descendants, or, in their absence, the ascendants, or, in their absence, the siblings of the testator shall have at least one fourth of the legitimate portion of inheritance (cf. Ulp. D. 5, 2, 1; Inst. 2, 18, 1). In modern jurisdictions, reserved portion can be regarded as the most significant limit of testamentary freedom. In civil law jurisdictions, reserved portion is an essential part of the inheritance law, contrary to modern English law, in which all property may be disposed of by will, and in which a reasonable part or a reserved portion is currently not institutionalised (as opposed to early English Common law in which a writ de rationabili parte bonorum was available for the wife and the children of the deceased).

When it comes to the private autonomy in the law of succession, it is particularly important to refer to the Roman principle of favor testamenti, too. From the classical period of Roman law, this principle became a widely applied rule.[25] According to favor testamenti, in conditions mentioned in wills, the intention, rather than the words of the testator, should be considered. (In connection with the principle of favor testamenti, the partial invalidity e.g. had of greater importance in the Roman law of succession,[26] than that in the law of contracts.)

* Iván Siklósi (PhD) is assistant professor at Eötvös Loránd University, Faculty of Law, Department of Roman Law and Comparative Legal History (e-mail:

[1] The division of law into ‘private’ and ‘public’ law branches itself (a distinction that, as is well-known, has no fundamental importance in the legal thinking of common law jurisdictions) was based on the dichotomy of utilitas privata (private interest) and utilitas publica (public interest); cf. Ulp. D. 1, 1, 1, 2. On the problem of the distinction between private and public law, see e.g. Gábor Hamza, ‘Reflections on the Classification (divisio) into “Branches” of Modern Legal Systems and Roman Law Traditions’ in Cosimo Cascione and others (eds), Fides Humanitas Ius. Studii in onore di Luigi Labruna, vol 4 (Editoriale Scientifica 2007, Napoli, 2449–2476) 2449ff.

[2] As for the law of things, we can refer, inter alia, to Hadrian’s and Justinian’s regime on treasure trove (cf. Vita Hadr. 18, 6 and Inst. 2, 1, 39). According to Inst. 2, 1, 39, if anyone found treasure in his own land, the Emperor Hadrian, following natural equity, adjudged to him the ownership of it. Hadrian established the same rule when the treasure was found by accident in a sacred or religious place. If the treasure was found in the land of another by accident, and without specially searching for it, Hadrian gave half to the finder, half to the owner of the land; and upon this principle, if the treasure was found in a land belonging to the Emperor, he decided that half should belong to the latter, and half to the finder. Consistently with this, if anyone finds treasure in a land belonging to the imperial treasury, or in a public place, half belongs to the finder, and half to the treasury (fiscus), or the civitas. The individualist and liberal approach of Roman law is reflected in this regime on treasure trove, as a kind of expression of private autonomy. (The original concept by Hadrian related to treasure trove is currently amended with numerous ‘public law elements’, even in those legal systems that are based on the Roman law tradition, since it is obvious that nowadays treasures of great archaeological and cultural importance would not to be awarded exclusively to the finder or for the landowner. An exclusively ‘private law approach’ seems to be unsustainable today, as the ruling of treasure trove deserves a complex approach, according to which any treasure could be regarded as a national heritage or even a kind of ‘common heritage of mankind’.)

[3] Without a detailed discussion we can refer e.g. to the Italian and provincial administrative units (civitas; municipium; colonia; res publica; vicus; pagus) having autonomy. In this regard, we note that civitates were considered as private individuals in Roman law (see Gai. D. 50, 16, 16: ‘civitates… privatorum loco habentur’), having their own legal capacity (cf. Flor. D. 46, 1, 22).

[4] See Alfons Bürge, Das französische Privatrecht im 19. Jahrhundert (2nd edn, Klostermann 1995, Frankfurt am Main) 65.

[5] One can refer in this regard, for example, to the concept of legal relation (Rechtsverhältnis), which was elaborated by Friedrich Carl von Savigny, on the basis of the Kantian concept of autonomy of will. In Savigny’s famous ‘System des heutigen römischen Rechts’ (the theory of private law of which can be regarded as the philosophy of positive law based on Kant’s works), the ‘great Lord’ of legal science emphasised that the essence of legal relation is the independent reign of individual will. On this topic, see from the modern literature e.g. Alejandro Guzmán Brito, ‘Los orígenes del concepto de “Relación Jurídica”’ (2006) 28 Revista de estudios histórico-jurídicos 187–226, 187ff. The modern concept of contract developed by the Pandectist legal science in the 19th century was also based on the Kantian concept of autonomy of will. As for the juridical act (legal transaction), the Pandectist concept of juridical act matches the principle of private autonomy perfectly [cf. Werner Flume, Allgemeiner Teil des bürgerlichen Rechts, vol 2, Das Rechtsgeschäft, (4th edn, Springer 1992, Berlin – Heidelberg – New York) 23]: juridical act is an act regarding the manifestation of private autonomy; in other words, a juridical act is a private declaration of will (i.e. declaration of will made by a private individual). [However, as for the autonomy of contractual will, it already appeared e.g. in the works of Pothier, who strongly accentuated the importance of the contractual will of the parties. Cf. e.g. Robert-Joseph Pothier, Traité des obligations, vol 1 (Debure l’aîné 1764, Paris) 9. It also deserves mentioning that the modern concept of private autonomy has its roots in canon law, too. On this problem see e.g. Peter Landau, ‘Pacta sunt servanda. Zu den kanonistischen Grundlagen der Privatautonomie’ in Mario Ascheri and others (eds), ‘Ins Wasser geworfen und Ozeane durchquert.’ Festschrift für Knut Wolfgang Nörr (Böhlau 2003, Köln 457–474) 457ff.]

[6] See the lex Iulia de maritandis ordinibus and lex Papia Poppaea, which granted various privileges to married people and parents of children, but imposed several economic and social disadvantages on unmarried persons and childless married persons. On this topic see e.g. Pál Csillag, The Augustan laws on family relations (Akadémiai Kiadó 1976, Budapest); Riccardo Astolfi, La ‘lex Iulia et Papia’ (4th edn, CEDAM 1996, Padova); Dieter Nörr, ‘Planung in der Antike. Über die Ehegesetze des Augustus’ in Tiziana J. Chiusi, Wolfgang Kaiser, Hans-Dieter Spengler (eds), Historiae iuris antiqui. Gesammelte Schriften, vol 2 (Keip, 2003, Goldbach bei Aschaffenburg) 1093ff; Hans Hermann Seiler, ‘Römische Ehe und augusteische Ehegesetzgebung’ in Hans Hermann Seiler, Geschichte und Gegenwart im Zivilrecht. Ausgewählte Schriften (Elke Hermann ed, Heymanns 2004, Köln–Berlin–München) 117ff.

[7] Instead of expropriation, which is an institution of ‘administrative law’, purchases were concluded instead – sometimes under political pressure – with the persons concerned.

[8] By nature, Roman ownership (dominium, proprietas) had some other limits, too. It was by no means an unrestricted right. Among the restrictions on ownership, one can also refer to the iura vicinitatis, or e.g. to the prohibitions of alienation. Confiscation was often applied in Roman law as punishment. Moreover, the origin of modern prohibition of abuse of rights is rooted in Roman law, too [cf. e.g. XII tab. 4, 2b (regarding the sanction of three times sale of a filius familias); Ulp. D. 8, 5, 8, 5 (the famous case of taberna casiaria)], although – as an expression of private autonomy – ‘no one who does what he has a right to do is considered to commit fraud’ (‘Nullus videtur dolo facere, qui suo iure utitur’; see Gai. D. 50, 17, 55).

[9] In general, it should be noted that violation of a mandatory rule may lead to invalidity in Roman law, too. The system of causes of invalidity is a strong restriction of the private autonomy of the contracting parties. A contract may be invalid, for instance, due to a mistake, deception, or because it is against the law or good morals, or because the contract was made in circumvention of the law. The relevant rules are constraints of private autonomy.

[10] For the roots of the pacta sunt servanda principle, see from canon law e.g. the relevant discussions of Bernardus Papiensis, Vincentius Bellovacensis, and Hostiensis. Cf. especially the statement ‘pax servetur, pacta custodiantur’ in Liber Extra 1, 35, 1 de pactis: ‘Pacta quantumcunque nuda servanda sunt’. From the later (rationalist) natural law literature see primarily Grotius, De iure belli ac pacis, 2, 11 De promissis.

[11] Reinhard Zimmermann, The law of obligations. Roman foundations of the civilian tradition (3rd edn, Oxford University Press 1996, Oxford) 258.

[12] Cf. Paul. D. 13, 6, 17 pr. regarding commodatum (loan for use); Ulp. D. 16, 3, 1, 7 regarding depositum (deposit); Ulp. D. 50, 17, 23 regarding, in general, the bonae fidei iudicia; see, in addition, Paul. D. 2, 14, 27, 3.

[13] Cf. Giovanni Meruzzi, L’exceptio doli dal diritto civile al diritto commerciale (CEDAM 2005, Padova), 203ff (‘come limite all’autonomia contrattuale’).

[14] On the problems of bona fides see from the Hungarian literature the works of András Földi, e.g.: András Földi, A jóhiszeműség és tisztesség elve [(Principle of good faith and fair dealing) ELTE ÁJK 2001, Budapest].

[15] Cf. Paul. D. 18, 6, 8 pr.: ‘[…] perfecta emptione periculum ad emptorem respiciet.’ On the principle ‘periculum est emptoris’ see e.g. Wolfgang Ernst, ‘Periculum est emptoris’ (1982) 99 (1) Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung 216–248, 216ff; Martin Bauer, ‘Periculum emptoris.’ Eine dogmengeschichtliche Untersuchung zur Gefahrtragung beim Kauf (Duncker & Humblot 1998, Berlin); Éva Jakab, Risikomanagement beim Weinkauf (C. H. Beck 2009, München).

[16] See Gai. D. 44, 7, 1, 4. Furthermore, see Gai. D. 13, 6, 18 pr.; Ulp. D. 19, 2, 15, 2; C. 4, 65, 28. As for the literature on vis maior see, for example, Adolf Exner, Der Begriff der höheren Gewalt (vis maior’) im römischen und heutigen Verkehrsrecht (Hölder 1883, Wien); Alfredo De Medio, ‘Caso fortuito e forza maggiore in diritto romano’ (1908) 20 Bullettino dell’Istituto di Diritto Romano ‘Vittorio Scialoja’ 157–209, 157ff; Theo Mayer-Maly, ‘Höhere Gewalt: Falltypen und Begriffsbildung’ in Hermann Baltl (ed), Festschrift für A. Steinwenter (Hermann Böhlaus Nachfolger 1958, Graz–Köln, 58–77) 58ff; Wolfgang Ernst, ‘Wandlungen des “vis maiorˮ-Begriffes in der Entwicklung der römischen Rechtswissenschaft’ (1994) 22 Index. Studi camerti di studi romanistici 293–321, 293ff; Felix Wubbe, ‘Vi tempestatis’ in Jean-François Gerkens (ed), Mélanges F. Sturm, vol 1 (Editions juridiques de l’Université de Liège 1999, Liège, 579–593) 579ff.

[17] See Zimmermann (n 11) 258.

[18] See C. 4, 44, 2 and 8.

[19] From the virtually boundless literature of laesio enormis, see e.g. René Dekkers, La lésion énorme (Librairie du Recueil Sirey 1937, Paris); Boudewijn Sirks, ‘La “laesio enormis” en droit romain et byzantin’ (1985) 53 Tijdschrift voor rechtsgeschiedenis 291–307, 291ff; Martin Pennitz, ‘Zur Anfechtung wegen “laesio enormisˮ im römischen Recht’ in Martin J. Schermaier, J. Michael Rainer und Laurens C. Winkel (eds), Iurisprudentia universalis. Festschrift für Theo Mayer-Maly (Böhlau 2002, Köln–Weimar–Wien, 575–589) 575ff; Matthias Armgardt, ‘Zur Dogmengeschichte der „laesio enormis’ – eine historische und rechtsvergleichende Betrachtung’ in Karl Riesenhuber, Ioannis K. Karakostas (eds), Inhaltskontrolle im nationalen und Europäischen Privatrecht (De Gruyter 2009, Berlin, 3–18) 3ff; Daniil Tuzov, ‘La “rescissio” della compravendita nel diritto romano tardoclassico e postclassico’ in Luigi Garofalo (ed) ‘Actio in rem’ e ‘actio in personam’. In ricordo di Mario Talamanca (CEDAM 2011, Padova, 835–891) 835ff; Aleksander Grebieniow, ‘La “laesio enormis” e la stabilità contrattuale’ (2014) 61 Revue internationale des droits de l’Antiquité 195–216, 195ff; Paola Lambrini, ‘Ipotesi in tema di rescissione per “lesione enorme”’ in Zuzanna Benincasa, Jakub Urbanik (eds) Mater familias. Scritti romanistici per Maria Zabłocka (2016, Warszawa, 453–464) 453ff. – From Diocletian’s time, the other famous example of the direct intervention into the price was the edictum de pretiis rerum venalium (301) [cf. Siegfried Lauffer (ed), Diokletians Preisedikt (De Gruyter 1971, Berlin)].

[20] See Iván Siklósi, A nemlétező, érvénytelen és hatálytalan jogügyletek elméleti és dogmatikai kérdései a római jogban és a modern jogokban [(Theoretical and dogmatic questions of the inexistence, invalidity, and ineffectiveness of juridical acts in Roman law and in modern legal systems), ELTE Eötvös Kiadó 2014, Budapest] 234ff.

[21] XII tab. 5, 3 (of which several versions are known, cf. Cic. De invent. 2, 50, 148; Pomp. D. 50, 16, 120; Paul. D. 50, 16, 53 pr.) provides the opportunity to bequeath [by will (probably by mancipatio familiaetestamentum per aes et libram)] the property (familia pecuniaque) of a Roman citizen sui iuris.

[22] Cf. Gai. 2, 225; L. s. reg. 1, 2.

[23] Cf. Gai. 2, 226.

[24] Cf. e.g. Gai. 2, 227.

[25] Cf., in general, Pap. D. 35, 1, 101 pr.; Paul. D. 50, 17, 12; Marc. D. 34, 5, 24.

[26] Cf. e.g. Pap. D. 5, 2, 15, 2; Pap. D. 5, 2, 28; Paul. D. 5, 2, 19; Paul. D. 28, 5, 20, 2; Ulp. D. 5, 2, 24; Marci. D. 28, 7, 14; Inst. 2, 14, 10.