Egyptian Law, Saite to Roman Periods

In their efforts to reunite Egypt after a period of political division during the Third Intermediate Period (ca. 1050–664 BCE), the kings of the Twenty-sixth Dynasty (664–526 BCE) imposed a number of administrative and legal reforms, which included the propagation of a new cursive script (demotic) for documentary and legal texts.¹ Juridical terminology changes, new types of legal documents appear, and the phrasing and makeup of already existing document types is modified, which implies new legislation and the distribution of new official templates to the notaries by the central administration.

However, the legal measures of the Twenty-sixth Dynasty seem not to have been put into the shape of a single legal corpus until Darius I ordered a collection of Egyptian law. According to a report on the verso of pBib.Nat. 215, col. c.6–16,² Darius I in his third regnal year sent a letter to his satrap at Memphis commanding him to unite “the wise men … from among the soldiers, priests, and scribes of Egypt” and to write down “the earlier law of Egypt (pȝ hp (n) Km(.t) n (tȝ) ḥȝ.t), until regnal year 44 of Amasis,” that is, 527 BCE. In spring 526 BCE, after a reign of only a few months, Amasis’s son Psammetichus III had been defeated by Darius’s predecessor Cambyses II.³ It is not stated how far back this collection was supposed to go, but one may assume that the legal reforms of the early Twenty-sixth Dynasty were fully included, and perhaps even the legislation of preceding kings. “Temple matters” were added “in the manner of the law of Egypt,” and a translation of the Egyptian text was drawn up in Aramaic, the lingua franca of the Persian empire.

The redaction of the collection is said to have taken sixteen years, which shows its enormous scope, and it was divided into eight⁴ subsections called ỉpt: literally, “beaker,” possibly designating a cylindrical metal sheath as protection for papyrus scrolls.⁵ The old translation, “tablet,”⁶ clearly influenced by the biblical “tables of the Law” or the Roman “law of the Twelve Tables,” was recently defended by Quack,⁷ but his argument that Egyptians occasionally did write important documents on metal tablets is not conclusive. The tablets he refers to contain short religious texts and were held in the hand by the priests as an aide-memoire during their service, and it is very unlikely that the complete legal collection would have fitted on just eight tablets, even if inscribed recto-verso. Besides, in all the cases that Quack evokes, the Egyptian word used for “tablet” is not ỉpt, but ʿn. What is more, Diodorus I.75.6 likewise relates that the Egyptian law was redacted in eight volumes, which were posed before the judges during the trial, and the word he uses is βιβλίοι “scrolls” and not πίνακες “tablets.”

Despite this report, some scholars still do not believe in a “codification” of Egyptian law under Darius. Following Frei,⁸ the report in pBib.nat. 215 verso c.6–16 had been used by biblical scholars to bolster the theory of the imperial authorization of the Torah; as a consequence, those who reject this theory tend to reject the historicity of pBib.nat. 215 verso c.6–16 in the same breath, usually without any background knowledge of Egyptian legal history.⁹ Others admit that there was some kind of redaction activity under Darius, but object to the term “codification.” Redford, for example, argues that the Saite legal and administrative system was sufficiently elaborate to have brought forth a law collection (or even several) before the Persian period, so that there would have been no need of a codification, pretends that there is no evidence for Egyptian courts using a “codification” or similar systematic law corpus in their judgment; and doubts that Darius could have stopped Egyptian judges using Egyptian law if he had wanted, so that one should not speak of an “imperial authorization.” According to Redford, Darius’s order concerned only the translation of Egyptian laws into Aramaic, with a focus on those laws that concern economic aspects, especially within the temple sphere.¹⁰ However, all these arguments can be refuted: seeing that the Saite period was a formative era, it is more plausible that new legislation was first sent out from the chancellery piecemeal, as it was issued. Even if it is likely that compendia of these laws were redacted from time to time, for example at the ascension of a new king, this does not contradict the claim that exactly such an update was ordered by Darius, especially as the probably quite significant number of laws enacted during the forty-four-year rule of Amasis might never have been included in a compendium. Thus, even if partial collections probably existed already, Darius might still have perceived a necessity to reunite them in a single comprehensive and better-organized corpus. Moreover, when looking at the post-Persian period, as is necessary if we want to evaluate whether a law collection that might be the one ordered by Darius was used by Egyptian judges, we find that there actually are strong indications for such a use. Greek and Demotic texts of the Ptolemaic periods refer to a corpus known as “the law(s) of the country” (ὁ νόμος/οἱ νόμοι τῆς χώρας),¹¹ “the law of Egypt” (pȝ hp n Kmy(.t)),¹² or just “the law” (pȝ hp)¹³ as being the juridical basis of Egyptian legal documents and of judgments by the indigenous courts (wpṱ.w / λαοκρίται; see below). This corpus was still referred to in the Roman period¹⁴ when Egyptians tried, not always successfully, to convince Roman officials acting as judges to apply “the law(s) of the Egyptians” (ὁ νόμος/οἱ νόμοι τῶν Αἰγυπτίων).¹⁵ The use of the singular—occasional in Greek, but regular in Egyptian—is significant, as this shows that at least for the Egyptians, this was a single, unified composition, representing not just a more or less haphazard collection of laws, but “the law.”

What is more, there are fragments of several demotic manuscripts of a law collection: the two best-preserved manuscripts¹⁶ are known under the names Zivilprozeßordnung¹⁷ and Hermopolis Legal Code.¹⁸ In addition, fragments of a second-century CE copy of a third-century BCE Greek translation have been discovered.¹⁹ Fragments of two different demotic copies of a didactic legal manual that was obviously used in the training of judges are equally known²⁰—some of the laws cited there are also preserved in the law collection fragments, proving its dependence on the same legal corpus.

The collection contains legal precepts in three different forms:

Even though by chance none of the laws cited in the preserved process protocols survives in one of the fragmentary law collection manuscripts as well, their terminology and phrasing are identical to those of the laws in the law collection and didactic manual fragments, and their citation method shows that the same corpus is meant.²⁵

No copies of the Aramaic translation of the corpus have hitherto been discovered, but indirect evidence for its existence comes from Aramaic legal documents of the fifth century BCE found at Elephantine, which contain numerous expressions (e.g., šmh “his/her name” for Egyptian (n-)rn=f/=s “(of) his/her name,” i.e., “the aforementioned”²⁶; “You have made my heart content with … ,” “I am far from you concerning … ,” “I will clean it for you”; cf. also the standard description of real estate) that are direct translations from the Egyptian²⁷ and thus most likely were inspired by the document templates in the Aramaic version of the Egyptian legal corpus.²⁸

The earliest preserved copies of the demotic version of the corpus of Egyptian law and of its didactic commentary were written down in the early Ptolemaic period, but internal evidence shows that at least some of the laws recapitulated there date to the Saite period, if not even earlier times.²⁹ Moreover, in one of the manuscripts a subheading containing an Old Persian loan word survives, which indicates that these Ptolemaic period copies do not go back directly to Redford’s hypothetical Saite period law collections, but rather to a text redacted in the Persian period.³⁰ Despite their very fragmentary character, the manuscripts testify to the wide variety of legal domains addressed in the law collection (see the next section), which contradicts Redford’s claim that Darius’s purpose was basically to assure Persian economic and fiscal interests. That Darius would not have been able to stop Egyptians from using their laws is irrelevant, as he obviously did not want to anyway, and in fact quite the opposite: often taken as evidence for the Persian administration’s intrusion upon the internal affairs of an Egyptian temple,³¹ the satrap’s exhortatory answer to the priests of Elephantine concerning the election of unsuitable candidates for the post of temple administrator³² more likely demonstrates the Persian official’s displeasure with the priests’ not keeping to their own rules and regulations (i.e., the “temple matters” mentioned above), recently included as laws in the royal collection and thus accessible to him via the Aramaic translation.³³

Thus, it seems difficult to argue that the corpus of Egyptian law copied and referred to in the Ptolemaic and Roman periods is not the corpus of laws redacted at Darius’s behest, and even more so, to deny this legal collection, which obviously united the claims of exhaustiveness and future validity, the designation “codification.”

Only a small part of the original law corpus is known so far, due to the manuscripts’ fragmentary conservation. The subjects treated in those fragments nevertheless provide a glimpse of the enormous breadth of the original work and thus support the claim that the collection of laws might have been considered comprehensive. They cover (although not necessarily in full): leases and their use in connection with contestations of property titles³⁴ (see “Leases,” below); sʿnḫ documents³⁵ (a special type of maintenance document; see “Matrimonial Property Settlements and Maintenance Obligations,” below); contestations of property titles for real estate and other neighborhood related disputes³⁶; inheritance³⁷ (see also “Transfer and Division of Inheritance,” below); sale of tombs (see “Sales,” below)³⁸; handling of claims based on different types of documents, oral depositions, and so forth, in court³⁹; rules and regulations pertaining to priests and priestly offices⁴⁰; loans and interest (see “Loans and Other Forms of Indebtedness,” below)⁴¹; and servitude (see also “Sales” below).⁴²

Perhaps by pure chance, no laws about how to deal with assault, manslaughter, murder, rape, abduction, and other major crimes survive in the law collection. A selection of such criminal laws is mentioned by Diodorus I.77–78, and evidence from Egyptian sources may one day come forth to confirm or challenge the exactness of his report. It is still uncertain which legal institution was competent to deal with crimes and how this may have changed over time. As in the New Kingdom, it is possible that local judges only dealt with comparatively minor offenses: a case of double murder was brought before the king in the Twenty-sixth Dynasty, and in the Ptolemaic period cases of theft and assault are on the one hand attested in petitions to the police and lower civil administrators, and on the other hand in temple oaths (see the section “Legal Procedure”), and thus seem at least sometimes to have been judged by Egyptian local courts.⁴³ Due to the character of these sources, little is known about punishment. Convicted thieves had to make restitution for the stolen objects or at least their value; an additional fine was not always demanded.⁴⁴ Victims of assault usually demanded a monetary compensation in their petitions, but it is unclear how the amount was determined; interior regulations of religious associations often contained a catalog of fines payable in cases of offenses committed by members against each other (e.g., injuries, defamation, or adultery with a colleague’s wife), but these cases were obviously dealt with before the assembly (see also the section “Courts and Judges,” below).⁴⁵

In the New Kingdom the jurisdiction had been divided in two levels. Courts (qnb.wt) existed on a regional level, probably in all towns and even villages, and were constituted from local dignitaries: according to a decree of Horemheb (Eighteenth Dynasty, 1319–1292 BCE), of high priests, mayors, and priests. The number of judges seems to have been flexible, with three being a minimum, and they were probably convoked anew (by drawing lots from the number of eligible persons?) for each single session. On a higher level, there were two or three “great courts” (qnb.wt ʿȝ.wt or qnb.wt wr.wt) based in the royal capitals (Thebes, Memphis, and/or Heliopolis), presided over by the vizier(s) or even the king himself; only these great courts had jurisdiction over crimes that were punishable by mutilation or the death penalty, and they also had jurisdiction over cases about land property and those involving officials.⁴⁶

With the shift from abnormal hieratic to demotic during the Twenty-sixth Dynasty, the local qnb.t courts changed their name: demotic texts now call the local judges nȝ wpṱ.w and the court house pȝ ʿ.wy wpy(.t) “the house/place of judgment” or pȝ ʿ.wy nty ỉw nȝ wpṱ.w n-ỉm=f “the house/place in which the judges are”; the word qnb.t does not survive in demotic except for idiomatic expressions like ḏd qnb.t ỉrm NN “to litigate with NN” or qnb(y).t “legal document to be used as proof in a lawsuit.” Due to a dearth of published process protocols from the Late Period, the earliest known sources for the exact composition of these local wpṱ.w courts date to the Ptolemaic period; at least by this time, they consist of three priests of the local temple.⁴⁷ It should be recalled here that the Hermopolis Legal Code, the only manuscript of the law collection of which the original archaeological context is known, comes from a temple archive.⁴⁸ With the didactic manual mentioned above, we have some indications about the modalities of the formation of judges: the text is a sequence of questions and answers covering a wide range of legal knowledge and using citations of laws and legal clauses from the document templates contained in the law collection. Its structure implies that it was used by those studying for the bar to prepare for some sort of exam.⁴⁹

Muhs claims that the high courts (qnb.wt ʿȝ.wt) no longer existed in the Saite period⁵⁰; this is certainly true in the sense that the term no longer appears. However, there are indications that the institution itself continued to exist until at least the mid-Twenty-sixth Dynasty. In pRylands 9 XI.18–19, a court session held by Psammetichus I at his residence in a case of double murder is related,⁵¹ but this might also refer to direct royal justice without the assistance of other officials as judges. However, in XV.8–11 mention is made of a second plea before the king at his residence, this time Psammetichus II, concerning a priestly position and the corresponding income. The plaintiff was sent off because the king was ill, then decided to file his suit before “the vizier and the judges” by having his deposition taken in “the house of judging.”⁵² That the vizier is mentioned as well might indicate that the plaintiff is not simply pleading before the local court of the residence town, but before the successor institution of the New Kingdom qnb.t ʿȝ.t.

During the First Persian period (Twenty-seventh Dynasty), an Aramaic document from fifth-century Aswan mentions dyny mdntʾ “judges of the province,”⁵³ who might correspond to the local courts of three judges (wpṱ.w)⁵⁴: in the same document three persons with Iranian and Babylonian names are explicitly labeled “the three judges.” As the context is extremely broken, it remains unclear whether this means that the Saite court system was extended in the Persian period to non-Egyptian communities, in this case, the garrison town of Aswan, which accordingly appointed the statutory three judges from their own elite, or that the Egyptian local courts of priestly judges were completely replaced by officials of the Persian administration, as maintained by Fried, who points out that all judges known from Aramaic documents bear Iranian (or Babylonian) names.⁵⁵ Strictly speaking, this is no argument against purely Egyptian communities still having Egyptian judges, as there are, of course, no Aramaic documents preserved there. The information from pBib.Nat. 215 verso c.6–16 about the Aramaic translation of the Egyptian law collection (discussed above) also cuts both ways: the collection in itself implies, on the one hand, that Darius wanted Egyptian law to be applied under his rule (at least, presumably, by and to Egyptians), and on the other hand, that this law was also supposed to be accessible to Persian officials who did not read demotic, and thus that they might at least occasionally have been called upon to judge indigenous Egyptians. Since no demotic process protocols of the Twenty-seventh Dynasty are known (or at least edited), Fried’s claim that the Persians completely abolished Egyptian courts is an argumentum e silentio.

Another problem is posed by the dyny mlkʾ “judges of the king,” who might correspond to the earlier high court(s) in the capital(s).⁵⁶ Given the evidence of pBerlin P 13489,⁵⁷ they might (also?) have been active locally, as their court session with the garrison commander of Aswan probably took place there. Either one has to follow Fried’s opinion that in the Persian period there were two kinds of local courts—one of judges appointed by the province (however that was supposed to work), the other of judges appointed directly by the king⁵⁸—or one might assume that the central high courts became mobile courts of royally appointed magistrates in the Persian period, which would make them forerunners of (or even the inspiration for?) the Ptolemaic χρηματισταί (see below). In any case—and even if one remains skeptical about the complete replacement of the Egyptian local judges by Persian officials—it is likely that these “royal judges” were non-Egyptian functionaries, and it remains questionable whether they were bound to apply Egyptian law when judging Egyptians.⁵⁹ An allusion to a central high court of the Late Period might be found in the report of Diodorus about pre-Ptolemaic juridical practice (I.75.3–6), describing a court of thirty men from Thebes, Memphis, and Heliopolis presided over by a high judge wearing a precious stone pendant of “Truth” (i.e., a figure of the goddess Maat, often shown to be worn by viziers), but it is not clear at all to which period exactly he (or rather his source, Hekataios of Abdera) refers.⁶⁰

No primary sources about either local or central courts are available for the period from the Twenty-eighth Dynasty to the Second Persian period. As the wpṱ.w reappear in the better documented Ptolemaic period, it is probable that they simply had continued to be active.

After the conquest of Egypt by Alexander the Great, more and more Greek settlers appeared in the country, and with them the juridical systems of their hometowns. The three Greek poleis Alexandria, Ptolemais, and Naucratis undoubtedly had their own full set of Greek juridical institutions, those of Alexandria being calqued on the Athenian legal constitution.⁶¹ But larger Egyptian towns with significant Greek populations also possessed Greek courts (δικαστήρια) of usually ten judges (δικασταί), who were chosen from among the local elites by drawing lots. The Ptolemaic king was the highest judge; to initiate a lawsuit, Greeks and Egyptians alike had to address a petition (ἔντευξις) to him.⁶² Of course the king only rarely became active himself, and generally these petitions did not even reach him personally, but were handed to a district level functionary, usually the nome governor (στρατηγός), who either decided on the spot or ordered one of his subordinates to attempt an arbitration to reconcile the parties. If this did not succeed, the functionary decided whether to hand the case to a Greek magistrate to be judged or to assign it to the appropriate Greek or Egyptian court.⁶³ Greek documents call these last λαοκρίται “judges of the people” (with λαός in the slightly deprecative sense of “common people, indigenous population” as opposed to Ἕλληνες), which has led some scholars to doubt that their adjudications were legally binding, because they pretend that there was a difference in meaning between κρίνειν and δικάζειν, with the first supposedly meaning “to arbitrate” and the second “to judge.”⁶⁴ But not only can no such distinction be discerned in the Greek material from Ptolemaic Egypt—κρίνειν and δικάζειν are used interchangeably for “to judge,” while “to arbitrate” is expressed by διάλυσειν⁶⁵—the choice of κρίνειν over δικάζειν is easily explained by the fact that κρίνειν, like wpy (and its pre-demotic equivalent wḏʿ), has the basic meaning “to separate, divide, split up,” and κρίται is thus a more literal translation of wpṱ.w. The claim that the judgments of wpṱ.w/λαοκρίται had no real authority because they supposedly could not be implemented immediately and depended on the losing party’s agreement to draw up a “document of being far” (see “Legal Procedure” and “Sales,” below)⁶⁶ misconstrues the basic workings of Egyptian jurisdiction: it was exactly by forcing the losing party to draw up this document that the Egyptian judges implemented their judgment.⁶⁷

In order to fully integrate these Egyptian tribunals into the Ptolemaic legal system, they, like the Greek courts, were each assigned a Greek official (εἰσαγογεύς), who was charged with preparing the cases and summoning the parties and also made sure that the documents adduced as evidence were valid.⁶⁸ At some point before 258 BCE, Ptolemy II⁶⁹ introduced mobile courts (χρηματισταί),⁷⁰ whose three members were chosen for a comparatively short term of office from among the polis elites. They were considered direct delegates of the king; as such, no appeal was possible against their judgment.⁷¹ In the second and first centuries BCE, the χρηματισταί became permanently installed in different districts.⁷² Cases in which Egyptians pled against Egyptians before the χρηματισταί are attested,⁷³ and one wonders why these were not brought before an Egyptian λαοκρίται court, as a trial before the χρηματισταί meant that all Egyptian documents had to be translated into Greek first. The problem of which court was competent—hitherto seemingly decided on an ad hoc basis by the officials who dealt with the ἔντευξεις—was finally decided by a decree of 118 BCE: the language of the document on which the claim was based now determined whether the χρηματισταί or the λαοκρίται courts were to judge the case.⁷⁴ Curiously enough, the decree in question is actually the last securely dated attestation of the Greek word λαοκρίται; while this has been taken by some scholars as a sign that, despite this decree, the λαοκρίται were soon superseded by the χρηματισταί,⁷⁵ this could equally well be explained by the immediate effect of the decree: if the decision to pass a case on to the Egyptian courts was no longer made on a case-by-case basis by a Greek official, but was regulated by law, this would of course significantly diminish their mention in the Greek documentation. However, it has to be noted that in 51 BCE a case concerning an Egyptian marriage settlement document was indeed decided before the χρηματισταί and not the λαοκρίται.⁷⁶

It is interesting to note that many Egyptian religious associations, from at least 223 BCE onward, stipulate in their rules and regulations that members had to sue each other before a court formed by “those of the ‘house’” (i.e., the priests of the temple with which they were associated or even their comembers within the association?) instead of before the predominantly Greek administrative authorities (ṯs sḥn rmṯ ỉ.ỉr sḫy “commander, administrator [or other] man in power”), and that they were not to appeal against a judgment from the temple authorities before these administrative authorities, or they would incur sometimes quite heavy fines.⁷⁷ In one case, there was also a fine for anyone who appeals against a judgment from “those of the house” before another religious association—but only if he loses his case again.⁷⁸ This shows that the underlying motivation was probably not simply to keep one’s problems within the community, but perhaps also to bind the members of the associations to Egyptian law, which was not necessarily applied by Ptolemaic officials acting as judges.⁷⁹ It might be from these internal courts of religious associations that the last temple oaths came (see the section “Legal Procedure”),⁸⁰ as it remains doubtful whether Egyptian courts officially survived into the Roman period. All juridical activities seem to have been taken over by Roman officials, the foremost being the prefect as representative of the emperor, who either gave judgment on his yearly visits to selected towns in the countryside (χώρα) or delegated decisions to subordinates. The iuridicus Alexandreae (Greek δικαιοδότης) and the responsible (curator) for the imperial exchequer (ἴδιος λόγος) also held judicial responsibilities.⁸¹ None of these Roman officials were in any way bound to apply Egyptian law, although they occasionally did so; several Greek process protocols of the Roman period make reference to the corpus of Egyptian law, presumably using the Greek version, copies of which still circulated in the mid-second century CE (see the section “From Laws to Law”).

It remains unclear how one approached the Egyptian court in the pre-Ptolemaic period. As mentioned above, a petition to the king was necessary in the Ptolemaic period, at least in the third and second centuries BCE. The procedure proper started with the deposition of a written statement of one’s claim that was handed to the defendant. The latter answered with a writ of his own, to which the plaintiff responded again; the plaintiff then wrote a final statement, and all four documents were handed to the judges.⁸² The Egyptian legal system knew no lawyers⁸³; the parties were expected to simply state their case, and the judges were supposed to identify the applicable laws—for this, they had been formed to find their way around the law corpus (see the sections “From Laws to Law” and “Courts and Judges”). As many Egyptians, and especially women, were illiterate, most parties probably needed (and could get) some kind of unofficial legal help through the scribes who wrote down their statements for them, even if just by putting their reports in coherent form.

The parties were then summoned to the “house/place of judgment,” which could be a building or an open air space in the shade of a temple gateway or similar public place.⁸⁴ Their writs were read before them, and they were asked to confirm their statements. Even though women, according to Egyptian law, had full legal capacity and did not need a guardian to act for them, it seems to have been common that they chose to be represented by a man who spoke for them in court; compare the judges’ question to the female claimant at the beginning of a process in 170/171 BCE: “Is there a man who makes your plea?”⁸⁵ The parties were questioned by the judges, and the documents on which their claims were based were brought forward, authenticated, and read in court; claims that could not be corroborated by a document had to be sworn to. The judges had comparatively little leeway in conducting their instruction, as the procedure to be followed for a given scenario, down to the phrasing of oaths, was established in detail by the regulations of the law corpus.⁸⁶ If the outcome of a case depended on one of the parties’ agreeing or refusing to swear the oath preformulated by the judges, the session was interrupted, and the party in question was sent to the place in the temple precinct where oaths usually were taken (hence the appellation “temple oath”), together with witnesses and an ostracon (inscribed pottery sherd or stone flake) or piece of papyrus on which the oath was written.⁸⁷

In their final decision, the judges cited the laws they were basing their verdict on; a protocol of the proceedings, including copies of all documents read in court, was drawn up and presumably filed in an archive. In cases of disputed property, the judges obliged the losing party to have a “document of being far” (see “Sales” below) drawn up, through which the winning party could later prove its title.

It is possible that notaries’ offices were already attached to temples in the Twenty-sixth Dynasty, as it is attested in the Ptolemaic period. It is, however, difficult to know whether the notary scribes (sẖ or sẖ qnb(y).t) had to be priests, since they rarely mention other titles in their signatures; if they do, these are in fact priestly.⁸⁸ The Greek expression μονογράφος implies that there was only one single notary scribe per notary office, but as Arlt has shown, in practice two notary scribes could be employed at the same time in the larger offices, and quite often notary scribes were surrounded by family members or apprentices as helpmates or even official representatives.⁸⁹

In the Ptolemaic period, notary scribes in Upper Egypt and some Fayum villages tended to have long office terms and often passed the office on to a son, who before this had usually been their representative, resulting in veritable dynasties of notary scribes. In Lower Egypt and parts of the Fayum, shorter office terms, sometimes just one year, were the norm, and there are cases of notary scribes getting back into office after an interval during which other, unrelated scribes had held it.⁹⁰ These differences can be explained in two ways: either regional traditions prevailed in the absence of legislation on the matter, or the strong southern tradition of hereditary offices simply overruled a legal obligation to appoint notary scribes for a year only, so that outsiders never had a chance to be chosen as long as there were candidates from the established families.

In the Roman period, Egyptian notary scribes, now called νομογράφος, “law scribe” or sẖ qnb(y).t (sẖ mtn) “document scribe (and registration scribe”), often no longer signed.⁹¹ At least in the Fayum (and after the reign of Augustus, no demotic legal documents are known from other parts of Egypt)⁹², the notary’s office (γραφεῖον) was still assigned on a yearly basis,⁹³ and the same incumbent (ὁ πρὸς τῷ γραφείῳ “the one of the notary’s office”) could be reappointed several years in a row or again after several years of absence from the position.⁹⁴ The new obligation to write not only demotic legal documents, but also the lengthy Greek subscriptions (see “Forms of Legal Documents,” below), initially called for joint ventures between a Greek ὁ πρὸς τῷ γραφείῳ and an Egyptian sẖ qnb.t, but it is likely that in the first century CE Egyptian notary scribes emerged who knew enough Greek to draw up both parts of a document.⁹⁵

Little is known about the notaries’ training. As model documents were included in the corpus of Egyptian law and questions concerning the typology of legal documents and the correct understanding of document clauses are part of the didactic manual (see the section “From Laws to Law”), one might expect notary scribes to have been schooled together with the judges, that is, at the temple. But while at least in some smaller towns the uniformity of the handwriting used for notary documents and religious texts shows that basic scribal training for candidates of different professions took place in a single school, most likely that of the local temple,⁹⁶ there are indications that in the Ptolemaic and Roman periods, at any rate, the apprentices’ further formation consisted mainly, if not entirely, of hands-on training in a notary office. By the Ptolemaic period, distinct local variations in the phrasing of the most common document types had emerged.⁹⁷ Nevertheless, a good notary’s apprentice did not just copy his seniors’ documents, but also learned how to adapt the basic document forms to the parties’ individual requirements.

Two forms of demotic legal documents need to be distinguished: the sẖ-document, which has to be drawn up by a notary scribe and authenticated by witness signatures and/or an official registration (see below), and the “letter style” šʿ.t-document, which is drawn up by the party himself (or any other literate person in his name) and also may bear witness signatures, but is not usually registered.⁹⁸

The form of the document influenced the phrasing of the document frame (introduction and conclusion), but not necessarily the phrasing of the document body. Some kinds of transactions—for example sales, donations, or matrimonial property settlements—needed to be documented by sẖ-document; others, like leases, agreements, work obligations or loans, could be either one or the other. A special subtype of the šʿ.t-documents, known as “double document,” appeared in the Ptolemaic period. It was modeled on the Greek “six-witness-document” (ἑξαμάρτυρος συγγραφή), with a sealed inner part and an open outer part, and was usually used for obligations toward the state authorities, notably warrantor’s guarantees, leases of state monopolies, and loans, often including a promissory oath on the king.⁹⁹

A specific feature of demotic legal documents is that they are generally unilateral, with the document body phrased as direct speech of the obligated party (issuer of the document) to the beneficiary (recipient of the document). This unilateral statement of obligation became a “contract” between the two parties through the acceptance of the document by the second party. Thus, a demotic legal document, despite bearing a date of issue to which the text referred back as the date from which it was valid (“from today onward, until eternity”), was not in force until the beneficiary actually received it. As a consequence, sales documents could also be used conditionally, that is, as security for loans or maintenance obligations or even as testamentary property transfers. In these cases, the issuer simply handed the document to a trusted third party (ʿrbṱ “document keeper”) instead of the beneficiary or left it with the notary; the beneficiary only received it if—or when—the condition (nonfulfillment of the loan repayment/maintenance obligations or death of the issuer) came to pass.¹⁰⁰ It has to be underlined that theses conditions were usually not made explicit in the document text; thus, conditional sales, when found out of context, are often indistinguishable from normal sales. However, certain Ptolemaic period mortgage documents conflate the loan and the conditional sale as security into a single text.¹⁰¹ On the other hand, Roman period mortgage documents sometimes combined the two separate documents—a Greek loan and a demotic sale of real estate (consisting of a sẖ ḏbȝ ḥḏ and a sẖ wy)—on a single sheet of papyrus; if the loan was not repaid in time, the document keeper or notary could simply cut off the loan document and hand what was now a normal sales document for the mortgaged real estate to the creditor¹⁰².

The usual, and originally the only, way to guarantee the authenticity of a sẖ-document was through witness signatures—at least four, sometimes also multiples thereof: eight, sixteen or exceptionally even thirty-six, either on the recto underneath the text or on the verso of the papyrus. Seidl thought that the term qnb(y).t designated specifically a sẖ-document with sixteen witness signatures,¹⁰³ but evidence from the didactic legal manual shows that documents with only four witness signatures could also be labeled as qnb(y).t,¹⁰⁴ and possibly even šʿ.t-documents with witness signatures fell under this heading¹⁰⁵; the meaning of qnb(y).t seems simply to be “a document that can be used as evidence in court,” as the word is etymologically connected to the pre-demotic word qnb.t “law court” (see above). Until the end of the third century BCE, sẖ-documents could also contain witness copies; that is, some of the witnesses recopied the complete document text in their own handwriting on the recto of the papyrus, which accordingly had to be quite large.¹⁰⁶ When the authenticity of the document was contested in court, the witnesses, or if they were deceased, their children, were called upon to testify under oath—in the case of the children, that they recognized their fathers’ handwriting.¹⁰⁷

In its efforts to harmonize the coexisting Egyptian and Greek legal systems, the Ptolemaic administration introduced new ways of authentication for the demotic documents. From about 264 BCE onward, copies of legal documents had to be deposited in an archive (κιβωτός, literally “box”); in 146 BCE a decree stipulated that short Greek abstracts had to be added to the demotic legal documents and the key data added in Greek to a register in the local notary office. Only documents that contained the Greek registration note (ἀναγραφή) underneath the demotic text and on which, accordingly, taxes had been paid, were henceforward accepted as evidence in court. In the early Roman period, the registration process became even more sophisticated, as copies now not only had to be kept in the notary office, but also had to be sent to a central archive at Alexandria (Ναναῖον); from 72 CE onward an additional archive in the district capital (βιβλιοθήκη ἐγκτήσιον) watched over property rights that had to be entered in a register.¹⁰⁸ During the first years of Roman rule, the Greek subscriptions under the demotic documents were extended: first the issuer, then also the beneficiary now had to add a substantial first-person statement in Greek (the second party, under certain circumstances, could also sign in demotic), either personally or through a writing assistant, presumably because in Roman law unilateral documents were not recognized as valid contracts. These parties’ statements preceded the notary’s Greek abstract of the contract and the registration notice.¹⁰⁹ Thus, the demotic legal documents became bilingual documents, although not in the sense that the demotic and Greek texts were completely congruent.¹¹⁰ With these innovations, witness signatures on demotic documents became useless and disappeared completely by the end of the first century BCE, although the phrase “and I have instructed the witnesses to write and the notary official to hand over the document after registration” curiously lingered in the first party’s Greek statement for at least another fifty years.¹¹¹

Since only a comparatively small part of the once quite extensive collection of laws has survived (or been identified and edited), we do not possess the legally prescribed document templates for all types of transactions. Information about the legal minimum requirements and the development of a given document type can also be gleaned through a comprehensive diachronic study of all extant examples, but this has not been done for all types, and new editions continue to add material. Only the most common transaction types are treated here.

The document usually used for sales¹¹² was called sẖ ḏbȝ ḥḏ “document of paying money.” The seller declares that he is content with the money he has received from the buyer (although the amount is not mentioned); describes the sold object; states the transfer of the title to the buyer; and guarantees to act against third-party claims, notably by personally furnishing any kind of proof by document or oath that could be demanded of the buyer to clear his title. Sales documents may contain third-party agreements from persons who have an interest in the sold object, usually the seller’s wife or children, but also of creditors or co-heirs.¹¹³

The fact that demotic “documents of paying money” did not state the purchase price allowed the use of this type of document for property transfers for which it may seem doubtful that more than a nominal sum or any money at all was paid, such as between parents and children. If one adds to this the fact that demotic documents became valid in practice only when the beneficiary actually received them, it becomes clear that many of these “sales” between family members might actually have been property transfers mortis causa (see “Transfer and Division of Inheritance” below), although due to a strong aversion to alluding to one’s own death, the issuer did not state this expressly.¹¹⁴

The most common demotic legal documents are those concerning the sale of real estate (e.g., houses, fields, gardens, building plots); the earliest known example is pBM EA 10117 (542 BCE),¹¹⁵ the latest pVienna D 10086 (85 CE).¹¹⁶ From the early Ptolemaic period onward¹¹⁷ a “document of being far” (sẖ wy) was added to the “document of paying money” for sales of real estate.¹¹⁸ This type of document was originally conceived (and up to the Ptolemaic period exclusively used) as an acknowledgment of lack of title resulting from an amiable agreement, a compromise, or a lost process.¹¹⁹ Instead of documenting the transfer of property as a necessary second step in the sales transaction after the agreement on the price in the “document of paying money,” as was formerly believed, the “document of being far” joined with a “document for money payment” is in fact a document anticipating a judgment in favor of the buyer.¹²⁰ The selling party thus could no longer contest the sale by filing a lawsuit, unless by proving that both documents were forgeries. Occasionally the texts of the “document of paying money” and the “document of being far” were executed on the same papyrus (e.g., pBerlin P 3146, Thebes, 194 BCE; pBerlin P 3101, Thebes, 118 BCE¹²¹), and this might even have become the standard practice in the Roman period.¹²²

Sales documents for animals are comparatively frequent in the Twenty-sixth and Twenty-seventh Dynasties¹²³ and become rarer during the Ptolemaic period¹²⁴; the last known examples date to 17 BCE.¹²⁵ Demotic sales documents for slaves are only attested in the pre-Ptolemaic period (earliest example: 517 BCE¹²⁶; last example: 342–338 BCE¹²⁷), whereas during the Ptolemaic and Roman periods, sales of slaves are exclusively documented in Greek, which might be the effect of an otherwise unattested decree. A special category is the sale of one’s property and of oneself as a slave, presumably to work off debts¹²⁸—a practice that might be a legal circumvention of the forceful seizure of the debtor’s person by the creditor, which according to Diodorus I.79.3, had been forbidden by Bokchoris (Twenty-fourth Dynasty, ca. 720–716 BCE). This document type was adapted once for an adoption,¹²⁹ and in the Ptolemaic period quite frequently for self-dedications¹³⁰—by paying (!) a monthly fee to become the slave of a divinity, the issuer expected to be protected against demons and diseases.¹³¹ Demotic sales documents can also concern charges and the income therefrom, for example “temple days” (i.e., income from priestly offices), one “temple day” corresponding to 1/360 of the temple’s yearly income. Occasionally the documents somewhat misleadingly phrase sales of “temple days” as sales of a part of a sanctuary, that is, as a real estate sale.¹³² Equally, sales concerning offices and income of funerary priests are usually phrased as sales of the tombs for which the service was performed; however, the actual sale of already occupied tombs as real estate was forbidden by law¹³³: The tomb sales documented on two Memphite stela of the Twenty-sixth Dynasty, which bear hieratic (Louvre C 101) and hieroglyphic (Florence 1639)COMP “Florence 1639” is not an (author date) citation, but, like “Louvre C101”, an inventory number.?> transcriptions of demotic originals, concern newly built, unoccupied tombs¹³⁴; pBM 10388 is the sale of a plot on which a tomb was going to be built.¹³⁵ The sale of “water days,” that is, the right of access to irrigation for a day, was a specialty of the Western Oases.¹³⁶ It is to be presumed that the sale of moveable objects generally did not have to be recorded, as this would have rendered daily commerce too cumbersome. The few examples of demotic sales documents for movables concern comparatively large and valuable objects (e.g., a loom)¹³⁷ or involve prepayment (e.g., for wine shipments).¹³⁸

No document was necessary for a marriage as such to be valid¹³⁹; the common accord of the spouses to live henceforth as man and wife was sufficient. However, to guarantee the woman maintenance and to secure the children’s inheritance right to their father’s property, it was possible to draw up a document either on the day of marriage or later. Following Pestman, three different types can be distinguished.

In documents of type A,¹⁴⁰ attested between 535¹⁴¹ and 60 BCE,¹⁴² the man states that he has made the woman his wife, that he has given her money as a “woman’s gift,” and that he will give her maintenance in money and kind; in case of a divorce on his initiative, he promises her a further payment of money and/or a third of their joint property, but in case of a divorce on her initiative, she has sometimes to pay back part of her “woman’s gift.” There are two known exceptions to this: pBerlin 3078¹⁴³ and pLibbey¹⁴⁴ are both made by the wife for the husband, and in them the woman promises to pay back part of the “woman’s gift” in case she divorces him and acknowledges that she will lose all rights to her share of their common property in this case; there is no stipulation that the husband has to pay something in case he divorces her. Usually, the eldest son or all the children of their marriage (even if not yet born) are appointed as sole heir(s). Occasionally, the document includes a list of personal objects that the woman has brought into the new household, which the man promises to hand back to her (or to replace if lost or broken) if she leaves him.

Type B¹⁴⁵ is the type that is attested the longest (earliest example¹⁴⁶: 517 BCE; last known example¹⁴⁷: 43 or 53 CE). Here, the man acknowledges that the woman has paid him a sum of money, usually called “money to become wife,” and that he has to pay this money back in case of a divorce, usually within a thirty-day period; the earliest example adds, like some of the documents of type A, that the woman in this case also receives a third of their joint property. Ptolemaic and Roman period examples add to this the obligation to pay the woman maintenance during the marriage and/or in case the man does not pay back the “money to become wife” within the stipulated period; some documents state explicitly that the woman can ask for the money back on her own initiative. The entire possessions of the man are given as security for the fulfillment of these obligations.¹⁴⁸

Pestman’s type C¹⁴⁹ is called sẖ n sʿnḫ in demotic—a term that often is misleadingly translated as “annuity document.” The earliest surviving example is quite late (365 BCE),¹⁵⁰ but documents of this type are attested for the Twenty-sixth Dynasty,¹⁵¹ and a template of a sẖ n sʿnḫ is actually preserved in one of the fragmentary copies of the law collection.¹⁵²

As in type B, the issuer acknowledges receiving a sum of money, this time called sʿnḫ; he promises to give this money back if he is asked for it, but until then, or if he is unable to pay it back within the stipulated time, he has to provide maintenance (ʿq-ḥbs, literally, “ration (and) clothing”) to the woman. Thus, sʿnḫ does not mean “annuity,” as it is often translated, but refers to the basis on which an annuity can be obtained, as can be seen in O.Manawir 6997, a sales document for one “water day” per month to a woman, possibly the seller’s daughter: the “water day” is described as “your private sʿnḫ,” meaning that by leasing her irrigation rights, she can obtain a permanent income.¹⁵³ As in type B, the man acknowledges that his entire property serves as guarantee, but at least from the third century BCE onward, it became customary to go one step further by actually mortgaging this property through a conditional sales document (see “Sales,” above), drawn up at the same time as the sʿnḫ document; occasionally, this sale was even written on the same papyrus as the sʿnḫ document.¹⁵⁴

While the first two types are explicitly limited to marital situations between issuer and beneficiary, type C could also be (but in practice only rarely was) used to settle maintenance on a daughter-in-law¹⁵⁵ or even one’s mother.¹⁵⁶ That the broad context of this document type was nevertheless that of a marriage is argued by Johnson, and indeed all “anomalous” cases can be explained by the husband being replaced by either his father (which might have been necessary when the maintenance engaged also his son’s future inheritance as guarantee) or his son, probably after the husband had died.

According to the template and the Saite period attestations, sʿnḫ documents were originally made by a man to another man, with the woman only being mentioned as beneficiary of the maintenance, and the last example for this practice dates to the early Ptolemaic period.¹⁵⁷ However, from at least the Thirtieth Dynasty onward, it was also possible to address the woman directly,¹⁵⁸ and with the exception just mentioned, this was the usual model in the Ptolemaic period. As women in Egypt had full legal capacity, one might wonder why sʿnḫ documents were originally not addressed directly to them. Johnson argues that this is a relict of a pre-demotic practice, as in abnormal hieratic marriage documents the groom always addressed the father of the bride and never the bride herself,¹⁵⁹ but while this is certainly true, the question remains why the father (or another male parent—the template leaves this open) was addressed in the earlier periods and why this practice changed during the fourth century BCE. A possible answer could be that this was caused by a slow rise in the average age at which girls were married: if most of them had been minors and thus unable to act for themselves in the sixth century BCE, by the third century BCE most women might have been adults (according to antique standards) at their marriage. Even though there is little direct evidence for this theory, it might be significant that, while the age of majority in Egypt was probably thirteen or fourteen years in the Ptolemaic and Roman periods, evidence from census lists of the Roman period shows that comparatively few women married under age eighteen.¹⁶⁰

Whatever the underlying reasons, the adaptation of the document form to the new custom of addressing the woman directly was unproblematic from a legal point of view, as the law already foresaw that, if the male relative died without having made over the document to someone else, the woman herself could sue her (ex-)husband for eventual arrears of her maintenance.¹⁶¹ Other changes from the original template are the inclusion of a clause about the inheritance rights of the children of the two parties (taken over from document type A), which of course only appears when the woman herself is being addressed by her husband, and of a clause that, without saying so directly, pertains to an eventual divorce: whereas the law seems to have allowed the issuer of a sʿnḫ document to give back the sʿnḫ money of his own accord if he wanted to stop the maintenance payments,¹⁶² this option was explicitly excluded in those documents drawn up directly for a woman by adding a clause that gave her alone the right to end the maintenance by asking for her sʿnḫ money. Thus, even if the man wanted to divorce the woman or the woman herself divorced, he still had to pay her maintenance until she herself chose to receive the sʿnḫ money back instead.

Most surviving demotic lease documents are for land (fields, gardens, or orchards).¹⁶³ Templates for leases of various objects (a field, weaver’s workshop, brewery, poultry farm, shop, garden, etc.) are conserved in the Hermopolis Legal Code,¹⁶⁴ although as a sort of parenthesis in the broader context of contestations to a property title: if an object has been leased out for three consecutive years by its possessor (who thus renders his claims to the title public) without anyone having contested him, his title becomes clear.

The demotic documents’ one-sided phrasing, from indebted issuer to beneficiary recipient, poses some problems when the transaction concerns obligations from both parties, which is usually the case in leases: on the one hand the lessor had to guarantee the lessee unmolested use of the object during the time of the lease, and on the other hand the lessee had to promise the payment of the rent and sometimes also of taxes. Nevertheless, most demotic leases followed the traditional unilateral phrasing, but while Upper Egyptian leases were normally issued by the lessee, in the Fayum, the issuer could also be the lessor.¹⁶⁵ One option for binding the second party more strongly was the inclusion of a first-person statement of the second party, adapted from the formula originally intended for third-party agreements (see above), and/or an autograph signature of the second party under the document body.

Leases always mention the length of the occupation: in land leases, this was often not a year, but just the “growing period” from sowing to reaping, with the possibility to tacitly renew the contract sometimes being excluded.¹⁶⁶ Demotic land leases for 99 years are also attested.¹⁶⁷ Land lease documents are quite flexible: they can stipulate that the taxes have to be paid directly by the lessee or be added to the rent payment (if the lessor engages himself to pay them),¹⁶⁸ that the rent is to be paid as a fixed sum in advance or as a percentage of the harvest,¹⁶⁹ and so forth. Penalties in case one of the parties did not fulfill his obligations were usually symmetrical, but the amount varied from lease to lease, and they could either profit the royal cult (“for the burnt offerings before pharaoh”) or the damaged party, or even both.¹⁷⁰

When documenting a loan,¹⁷¹ one could either phrase this as a receipt followed by the commitment to pay back, or as a promissory note, that is, an acknowledgment of debt that did not necessarily state how this debt was contracted. In the second case, it is sometimes difficult to know the interest rate, as only the overall debt is ciphered (“capital and interest” or “with its interest included”). Loans in kind were always submitted to 50 percent interest per year, a rate that was clearly meant for grain loans used for sowing, as is also shown by a problem in the didactic manual, where it is explained that the interest of a long-term grain loan is to be calculated not on overall length, but by counting how many times the due day corresponding to the end of the harvest has passed.¹⁷² However, the same interest rate of 50 percent is also applied to wine loans, which made buying wine on credit a costly undertaking. Interest rates for money loans were initially very divers, but the compound interest was capped at 100 percent of the capital;¹⁷³ this law is also mentioned by Diodorus, who ascribes it to Bokchoris.¹⁷⁴ Under Ptolemy II, the maximum interest for money loans was limited to 24 percent per year,¹⁷⁵ and at least in the first century CE, to 12 percent per year (1 percent per month).¹⁷⁶ Failure to repay the loan at the stipulated time resulted in a fee of 50 percent of the amount owed (demotic pȝ 1 r 1 ½ “the one to one-and-a-half,” Greek ἡμιόλιον).

At the end of the second century BCE, demotic notarial documents for money loans disappear¹⁷⁷; demotic money loans of the first century BCE are all šʿ.t documents.¹⁷⁸ In the Roman period there may even have been a prohibition against using demotic documents for money loans, which led to bilingual loans on mortgage combining a demotic sales document as security with a Greek loan document on the same papyrus.¹⁷⁹ However, demotic šʿ.t documents for grain loans were still in use in a temple context until at least the end of the first century BCE.¹⁸⁰

Recent research shows that only larger sums were actually lent against mortgages, while smaller sums often were secured by pledges (moveable objects left with the creditor) or through personal warranty of a third party.¹⁸¹

Drawing up a document for the transfer of property mortis causa was not obligatory, but if a person died without having made a will, the legal succession was regulated by law¹⁸²: the property was immediately taken into possession by the deceased person’s eldest son as main heir, or if the deceased had only daughters, the eldest daughter. However, as soon as the younger children demanded their share (if not immediately, probably at the latest when the second eldest sibling reached majority and could act independently), the eldest son had to divide the inheritance or, if he had already sold part or all of the property, the resulting profit. The law corpus presents two possibilities for this division: if the inheritance consisted of various mobile and immobile objects, it could be split into lots that were, necessarily, not of perfectly equal worth; the eldest son then chose his lot first and the other children after him by order of birth, with sons preceding daughters. The second possibility was to divide the property virtually, that is, to split the usufruct among the co-heirs, with the eldest son receiving a double share as well as the shares of all siblings who had died childless between their parent’s death and the division. The law collection prescribed this method explicitly in cases where a single house was the only inheritance.¹⁸³ The division among the co-heirs would have had to be documented in order to give them full title, and indeed such documents are attested: they often took the form of either donations or fictitious sales from the main heir to each of the co-heir(s), with the co-heir(s) in turn setting up a “document of being far” (see “Sales,” above) for the others’ shares.¹⁸⁴ If the division was among several co-heirs, it was also possible to draw up a shareholders’ agreement for each co-heir by the others as a group. These are often referred to as “division documents” because of their characteristic introductory formula: “We have divided with you, you have divided with us .” This type of document is still attested in the first century CE.¹⁸⁵

If a person did not want his property to be transferred according to the legal order of succession, he could actively attribute his inheritance or parts of it through notarial documents. By inserting the appropriate clause in a marriage settlement document (see “Matrimonial Property Settlements and Maintenance Obligations,” above), one could designate the eldest son from this marriage as sole heir even before he was born; in this case, the younger siblings later were not able to claim shares from him. However, usually all children from the marriage were designated as heirs in the settlement, thus excluding children from a later marriage unless the first wife and/or the eldest son from the first marriage agreed to their inclusion. This is the moot point of the process reported in pBM 10591 recto, where the judges ruled that, as such an agreement had been made, the younger half-brother of the claimants’ husband did have the right to inherit.¹⁸⁶ In any case, without further specifications, the simple mention of the children being heirs in a marriage settlement did not change the division pattern from the one defined by law. If for whatever reason a man did not have this clause included in his marriage settlement, he could have another document drawn up later, for example upon the birth of his first child.¹⁸⁷ A person other than the eldest son could also be designated sole heir through a notarial act,¹⁸⁸ but nonrelatives probably had to be adopted first: indirect evidence is the adoption document mentioned above, which, however, only states that the adopted is to be considered the adopter’s son and has to act accordingly.¹⁸⁹

If a person wanted to divide his property among several heirs, but according to his own designs, he could either assign specific objects to each one of them through fictitious sales documents (see “Sales,” above) or virtual shares of his property through either fictitious sales or share allocation documents; these last are often also referred to as “division documents.”¹⁹⁰ In each case, a separate document was made for each beneficiary. As mentioned above, the effectiveness at the death of the issuer was practically never spelled out in these documents.¹⁹¹ A share allocation document is basically a donation that, instead of a specific object, allots a share of one or several objects (e.g., “I have given to you my 1/x part of …”); its phrasing can, however, also follow that of promissory notes instead (“To you belongs a 1/x part of my …”). Often the legator specified the other shareholders and the size of their shares—this also happened sometimes in fictitious sales documents about shares.¹⁹²

Virtual division was often practiced even if the object was not just a single house, at least as a first step. However, co-heirs sometimes decided later to split up the joint property into distinct lots, such as by separating a field or building ground into plots or dividing a house into apartments.¹⁹³ If, however, the number of heirs was too large or the property too small, this would have resulted in lots that were of little use to anyone; one of the co-heirs could instead decide to buy out the others, or a third person could successively buy up the co-heirs’ shares.¹⁹⁴