Introduction

Parties to contracts are commonly interested in committing the terms to writing. The texts that they produce are meant to record the act that constitutes the contract as well as the terms that regulate the parties’ future conduct. In all these respects, the Greek language has already provided clear terminology in the Classical period. The verb δανείζω was used to denote the assignment of a quantity of genus chattels on the understanding that the assignee would in the future repay the assignor commodities of equal quantity and quality.1 The verb μισθόω was used for the assignment of a specific person or property for a designated time frame and purpose, with return to the assignor effected once the term of the contract ends.2 Finally, the medial voice of ἀποδίδωμι was used to denote the permanent conveyance of title to an object by sale.3

In all three cases, the contract also recorded remuneration. In sales, the document records the non-recurrent delivery of the consideration. It is commonly labelled τιμή. Documents recording loans stipulate the assignee’s duty to remit the remuneration (in this case the interest) periodically, routinely termed in the Ptolemaic and Roman periods τόκος or τόκοι. The same is the case with leases. Before the Byzantine period the routine terms are (in Egypt)4 φόρος, ἐκφόριον, or ἐνοίκιον. Terminological unity is evident in additional contexts as well. The verb ἀποδίδωμι, already encountered in the medial voice to denote the act of sale, is used in the active voice to designate the delivery of genus chattels in the context of payment of rent, repayment of a loan, restitution of dowries or the delivery of a commodity in the context of a sale on delivery.5

Beyond these examples of terminological unity, the documents exhibit unity of format. Loan and lease contracts, in particular, feature much of the same structure from the Classical through the Byzantine period: an account of the act that constitutes the contract, the amount of the remuneration, the modalities of its payment, and the termination of the contract. They exhibit additional constant elements as well: an account of the date and place of its composition, the kyria clause—professing the future applicability of the document as a piece of evidence, a praxis clause, setting forth the measures that the obligee can take to recover the debt from the obligor, and in some contexts, clauses introducing penalties and compensations.

The available source material, however, also exhibits much change. Recent collections of legal documents on stone show much regional variation.6 The source material from Egypt, the focus of the present study, is uniquely apt to exhibit diversity: documents originating in many parts of Egypt give evidence of regional peculiarities.7 In addition, the land of the Nile yields continuous evidence from an extremely long period. The earliest Greek document from Egypt, P.Eleph. 1, dates to 310 BCE; the latest postdates the Arab occupation by almost two centuries.8 In between, the number of documents of Egyptian provenance is incomparably greater than that from any other region in the Greco-Roman world. By the time I composed this Introduction (November 2022), I had analysed 1,629 lease contracts from Egypt, 1,590 acts of sale, 1,167 contracts related to credit, 354 labour contracts, 247 testamentary dispositions, 161 marital dispositions, and 142 diaireseis. Most types of documentation are attested for all three periods studied—the Ptolemaic, the Roman, and the Byzantine—thus allowing us to trace their development, as well as the development of the structure and terminology evidenced in legal documents through time and place, to an extent unparalleled in any other ancient land, province, or region.

The present study aims at presenting the reader with an analysis of the clauses of Greek legal documents preserved on papyrus and other perishable material, in Greek and Latin, from Egypt and other parts of the Eastern Roman Empire, in the Hellenistic, Roman, and Byzantine periods.9 It is based on the databank synallagma, now placed in the servers of The University of Münster (Fachbereich Papyrologie) (https://synallagma.uni-muenster.de/), cordially hosted by Professor Patrick Sänger. The initial funds for the creation of the databank were provided by the Israel Science Foundation in 2005. At a later stage, the databank was developed in collaboration with Professor Michele Faraguna, at that time of the university of Trieste. The team led by Professor Faraguna undertook the analysis of clauses on legal documents on stone from Greece and the Hellenistic world outside Egypt. I hope that the result of that survey will be published as well. In addition, I leave out of the present study all accounts of legal documents in Demotic, Coptic and Arabic, though I hope that the present book will stimulate the creation of similar accounts by specialists in those areas. Documents composed, mostly in Latin, in the Roman and Byzantine periods outside Egypt are also not systematically treated in the present monograph.

Legal documents, as defined here, are composed to record the creation of rights to an object or a person by way, in particular, of sale, loan, lease, labour contract, marital arrangement, will, and diairesis. Documents recording the termination of these rights—the vast corpus of receipts—have been registered in the Synallagma databank, but are left out of the present analysis. A ‘clause’ is defined in this book as a routinely and repeatedly applied phrase that conveys unique information about the terms of the activity recorded in the document of which it is a part. A clause need not be syntactically independent; especially in the Roman and Byzantine periods, many clauses are syntactically subordinate to others: relative clauses, absolute genitives, and nominal infinitives embedded in a relative clause that gives, for the most part, the consequences or the background of the act recorded in the preceding or following clause.10 All these clauses, however, convey new information consisting of a subject, a predicate, and, in most cases, a direct and an indirect object.

The number of clauses studied here is 281. Some occur in just a few texts and are used within a limited region and timespan. Others, however, are extremely common and indispensable in any legal document. Such is the case with two clauses not studied here: the clause that reports the date and place of composition and the address clause, which records the author of the document and its addressee in the context of the epistolary format (cheirographon, cheir, grammation). The user can consult the raw data recorded in synallagma.11 The book does aim, however, at offering an analysis and an updated account of other hitherto attested clauses. For each clause addressed, the book records key syntactical traits and terminology, places and times of attestation, and, perhaps most importantly, particular formulations that are invoked in its context. For example, scribes frequently refer back to earlier contracts that the parties concluded; this is the case in 207 documents. Yet a study of the wording used in each type of contract yields eight or even nine different formulations by which this act is recorded.

The purpose of this book is to present these formulations, but not necessarily to explain each exhaustively—a task occasionally undertaken by scholars studying individual clauses. The account of the clause is preceded by a link to the synallagma databank, which provides relevant metadata, as well as access to the document itself. That account is then followed by a list of attestations. In the lists I register available texts and offer a typological analysis of each. In my descriptions of individual attestations, I provide no information about individual documents only if the formulation is entirely identical everywhere. This, however, is rarely the case. In most cases, if an overarching typology is possible, I add for each individual document its type-number; otherwise, I add key elements that I consider conducive to the future typologization of the clause.

This monograph is devoted to the analysis of the language and sphere of application of individual clauses. For this reason, while I realize that potential readers will be interested in studying just one specific clause, an attentive reading of the entire monograph may allow them to assess underlying trends, similarities, and connections, both those treated here and those they may be reach by studying additional sources. For as the reader of the following pages will observe, clauses are rarely developed in isolation; rather, those that deal with similar subject matter tend to use similar vocabulary. Such is the case, for example, in the clause that records the acts of leasing, selling, bequeathing, dividing, or exchanging landed property; in all instances, the same terminology is used to describe the object of conveyance. Similar use is attested of clauses that report the capacities of a conveyee following the act of conveyance, anticipate the conveyance of genus chattels, prescribe penalties, stipulate maintenance, and so on. Accordingly, the clauses are organized into chapters and sections that underscore their terminological affinity. By assembling the vocabulary used in each cluster (see especially the index verborum), I also give the reader a glimpse at the way Greek scribes in Egypt, as well as those in other regions that have yielded papyri, conceptualize key institutions pertaining to property, family, and contract law.

Before I turn to the clauses, I wish to underline key issues that have been addressed in multiple studies of Greek legal documents since the end of the nineteenth century. I will also indicate where I believe additional scrutiny is required.

1. The act of contracting: A legal document commonly records the past physical act that brought about the creation of the contractual relationship, which is for the most part an act of surrender.12 In the earliest stage (Classical and Hellenistic periods), the act is reported through the indicative of the aorist tense. The verb varies in accordance with the nature of the contract: ἐδάνεισεν in loans, ἀπέδοτο in sales, ἐμίσθωσεν in leases, ἐξέδοτο in marriage documents and contracts of apprenticeship, διέθετο in wills. In the course of time, the protocol style gives way to a text introduced by the verb ὁμολογέω, that will be termed in the course of this book homologia.13 Even then, however, the recipient of the object, the obligor, acknowledges the same past act of surrender. But must a document record the past performance of an act of surrender to become effective? In one substantial group of documents, the lease hypomnemata, the document is framed as an application by the prospective lessee to the owner to take possession of the object in lease (βούλομαι μισθώσασθαι). If the misthosis is indeed a physical act of surrender, it must postdate the composition of the document and becomes valid only after the performance of an additional act. If such an act has ever been performed, however, it has left no trace in the available documentation. In addition, there are cases, in particular in the context of contracts of labour, in which the document records the declarant’s consent to perform the required task—in the future, of course.14 Is the contract in these cases consensual rather than real?15
2. The clauses: After recording the act of contracting itself, the document spells out the terms of the surrender. Since documents that record an established type of surrender (say, a loan) always record the clauses that regulate the payment of interest, the terms of repayment, and the consequences of default, it can be assumed that these key duties automatically arise from the act of contracting as recorded at the beginning of the document and set out in the verb of contracting. Namely, the use of δανείζω entails, in itself, the obligations of paying interest and repaying the loan.16 If so, the clauses are inserted not to introduce the obligations themselves but to establish the modalities of their performance: the rate of the interest and the time, place, and manner of repayment. It may be asked, however, to what extent this ‘automatism’ obtains in other clauses of the document, such as the clauses that penalize overdue payment and praxis. It may also be asked whether the use of an alternative vocabulary to record the loan, viz., through ἔχω, has the same consequences. Would, in other words, the simple acknowledgment of the ‘having’ of fungibles—without indication of legal category, the causa, interests, or return clause—in itself entail the performance of these additional terms?
3. Penalty and enforcement: All contracts establish obligations of commission or omission. One of the most detailed accounts of such obligations appears in lease contracts that require the lessee to cultivate the leased land. In that particular case, the document does not prescribe a sanction for violation. Other obligations are generally enforced by sanctions: payments in cash and in kind in the context of loans, leases, sales on delivery, in marriage documents, in contracts of labour that mandate a given duration of stay (paramone) as well as in sales contracts that enforce the conveyor’s duty to protect the conveyee’s possession. In all these cases, violation is liable to a fine that also serves as compensation for nonperformance.       In other respects, however, the nature of the sanction varies.17 In cases of payments on account of loans, leases, and marriage arrangements, if the original payment is due in cash, the clause stipulates an increase, usually by fifty per cent. If it is due in kind, the clause introduces a payment proportional to the purported value of the commodity.18 To obtain compensation, a creditor may invoke the right of execution (praxis) against the obligor and his property, in accordance with a procedure set forth in the document.19 In (mostly land) conveyances, the conveyor must protect the conveyee’s right to the object of conveyance.20 If he fails to do so, he shall be subject to a severe penalty: repayment of the consideration increased by fifty or one hundred per cent, compensation for damages and expenses, and a lump-sum fine payable both to the conveyee and to the fisc.21 In contrast to the former group, however, there is no enforcement of the act of conveyance; there is no praxis.

4. The clauses in the homologia: As soon as the document adopts the structure of an homologia, the individual clauses are absorbed into its cohesive textual flow. The obligor, who acknowledges his past performance of the act of contracting, may also acknowledge, in the same context, his duty to satisfy derivative requirements in the future. Anticipating a future act of the obligee is problematic, however, and it is impossible to record through the homologia injunctions, especially those directed against the obligee, just as is the documentation of rights of the obligee that are constituted by the act of contracting.22 To my knowledge, the phenomenon has never been studied, at least not in the context of Greek papyri from Egypt.
5. ‘Unnatural clauses’: Some clauses discussed below do not derive generically from the act of contracting. Such is the case with the prodoma, which in the early Ptolemaic period was a loan given by the lessee to the lessor on the occasion of the act of leasing. In formulaic terms, the documentation of the prodoma brings about the creation of a hybrid document that integrates elements of loan and lease contracts alike.23 The prodoma, however, is just one of a larger group of clauses that abound in lease and labour contracts and stipulate provisions in cash and or kind.24 These clauses, which exhibit distinct vocabulary and syntax, require further scrutiny. In the context of the present book, I will point out the use, in this context, of the genitive absolute, especially in what I term ‘the sphere of liability formulation’ (ὄντος/οὔσης περί [contracting party] τοῦ/τῆς [name of the liability]), and single out the specific array of verbs and nouns that are employed in these clauses.
6. The quality of the document: The document is meant to serve as evidence of the contract that it records. But would it serve as an exclusive piece of evidence? And what if its contents conflict with the contents of law or convention? In addition, could a document that names a certain person as the obligee be used by others to enforce the performance of the duties recorded? And would the place of composition of the document—by a public or private scribe—impact the applicability of the document and, if so, in what sense? These exhaustively studied questions were addressed, within the text of the legal document, in the kyria clause. Already in the Ptolemaic period, it became established that the document could be presented as reliable evidence of the contract ‘everywhere’. It has also been frequently stated that the document can be presented as evidence by anyone, but the communis opinio is that it can be served as evidence by a third party only as the representative of the obligee; there is no bearer instrument in Ptolemaic and Roman Egypt.25 While I share this view, the data discussed in the relevant entry may allow the reader to investigate further this important question. In the Roman period, the focus shifted somewhat as the kyria clause came to focus also on the ‘physicality’ of the document—particularly the number of texts issued, the identity of the scribe, and the ‘public’ quality of the document. When a text was privately composed, it was said to be as applicable as if composed by a public scribe or deposited in a state archive.26 But what exactly does this clause mean, and could a privately composed document be invested with the quality of a public document simply by inserting that clause? Since the most recent discussion of the subject is more than half a century old,27 further investigation may be in place.
7. Authentication: A legal document serves as proof of its contents. To verify its authenticity, the act of contracting and composition may be attended by witnesses or composed by a public scribe or a notary. Composition by a public scribe accords the document additional authority to that provided by a privately composed instrument and is obligatory for some types of acts.28 For this reason, documents composed privately may be replaced with a new public instrument (a procedure attested in both the Ptolemaic and the Roman periods) and then, in the Roman period, may be accorded the quality of a public document ex post. From the mid-Ptolemaic period onward, documents could be composed privately and not be attended by witnesses. In this case, the document is validated through its autographic drafting by the obligor. Autographic drafting then also becomes obligatory in publicly composed documents, as the obligor autographically confirms the terms of the contract beneath the body of the text.29 The analysis of the hypographe presented below lays the foundation for further research. It enables us to evaluate how the development of the hypographe influenced the language of legal documents in the Ptolemaic and Byzantine periods, how the Greek hypographe interacted with documents written in Demotic or Latin, and what factors affected its length and the choice between subjective formulations (e.g., beginning with ὁμολογέω) and objective ones.
8. Vocabulary: the present study underscores the use of the same terms in different contractual contexts and clauses. Terms that record objects have been extensively studied; the taxonomy of agricultural land, facilities, houses, slaves, animals, and various kinds of fungibles has been exhaustively analysed in the past. In some areas, however, renewed analysis is required and may now be more easily undertaken in light of the observations made in the present book. Terms denoting capacity, for example, appear in all types of clauses and are used both positively—to denote an acquired capacity—and negatively—to indicate restrictions placed on the obligor (particularly in the case of mortgage) for the duration of the contract. For this reason, I intend to give an exhaustive account of the verbs used in this context in each particular document. An equally central group consists of the records of duties in the variety of contracts of lease and labour. The same vocabulary is then used in clauses that anticipate the termination of the contract; there they appear in the perfect participle and modify the object of the return.
9. Tenses: a final observation relates to tenses. Originally, in the early Ptolemaic period, the past act of contracting is reported in the indicative mood of the aorist tense. In clauses that stipulate duties, the scribe uses the present imperative if the obligation is repetitive or continuous. The aorist is used for a single, nonrecurrent activity, as in, for example, the repayment of a loan or the restitution of an object of rent. In the regime of the homologia, by contrast, the verb ὁμολογέω routinely takes the perfect infinitive to denote the past act of contracting. The clauses that stipulate the obligor’s duties use not the aorist but the future tense in either the indicative mood or in the infinitive. The aorist and the present retain their position in the capacity clauses. Introduced by the impersonal ἔξεστι, the aorist infinitive denotes the capacity to perform a nonrecurrent act (e.g., registration of title) while the present infinitive expresses continuous disposal. In a sense, the most innovative contracts are labour contracts, particularly those that record the assignment of a person to a designated task; here a mixture of the present and future participle and the infinitive of the aorist tense is attested. The most substantial transformation takes place in the Byzantine period, as the aorist infinitive gains dominance even where aspectually, one would expect the present tense to appear. This change occurs between the fourth and sixth centuries, a period that remains seriously under-investigated in the context of legal documents.

How to Use This Book

Here are some guidelines for potential users:

1. I discuss 281 distinct types of clauses in this book. Each entry contains, beyond the discussion itself, a link to the relevant documents in the synallagma databank. Since synallagma continues to evolve while this book presents a conclusive list, occasional incongruences may occur, but these do not affect the overall proportions or significance of the clauses. Each discussion is followed by a list of documents in which the clause is exhibited; the list aims to be exhaustive. Here too, since the essays on some clauses were composed before the finalization of the lists, the reader may again find incongruences. As before, the addition of texts does not alter the overall picture.
2. As for the designation of texts, I employ the following method: A number followed by a letter without a dot, (e.g., 126a) indicates the original publication number. By contrast, ‘126.a’ indicates the fragment number, meaning that the clause appears in the first fragment of Papyrus 126. Column numbers are expressed in Arabic numerals, e.g., 126.2. Hence, if a papyrus published as no. 126a contains two fragments and the clause appears in lines 20–22 of the third column of the second fragment, it will be marked as 126a.b.3.20–22.
3. To specify places and dates of composition, I rely first and foremost on those given by the Heidelberger Gesamtverzeichnis but I depart from this rule here and there, especially when I identify considerations that may impel revision.
4. Most of the Greek texts are followed by translations. Whenever I give a verbatim translation produced elsewhere, I indicate its source at its end. When no such indication appears, the translation is mine even if partly based on earlier translations.
5. Transliterated Greek and Latin texts are given in italics if used in the main text. In the typological analysis, in contrast, italics are limited to Latin editorial terms only (e.g., vel sim.).
6. In the lists, I created a system of numbers, acronyms, and abbreviations, most of which are explained at the beginning of the book and also, occasionally, in the relevant entries. Since some clauses use the same terminology, I have aimed to maintain uniformity.
7. In the lists, an asterisk indicates that the word marked has been entirely restored. Its absence, however, is not a sure indication that restoration is lacking. The reader should always consult the original before drawing further conclusions.
8. Typologies are always reported in square brackets after the indication of date and place. In determining the typology of each item, I frequently use multiple elements set apart by punctuation marks. A semicolon establishes the strongest break by separating two unrelated terms that are considered in the typology. In reference to the tabelliones clause, for example, one document is described as [1; egr.] since its identification as Type1 has no apparent connection to the use of ἐγράφη in the text. The semicolon is also used to indicate the end of a direct quotation from the papyrus, always opened by ‘ed.:’: ‘ ed. αβγδ;’. If the quotation is the only or the last element of the typology, the reader encounters the marking ;];. If the end of the quoted text is restored, the reader even encounters ];]; indicating the end of a restored text, the end of the quotation, the end of the typological analysis and the end of the record of the text. Colons are used mostly when the two elements have a strong syntactical connection, as in a verb and a direct object. Finally, a comma is used whenever terms are connected by a conjunction. The use of all three elements is demonstrated in the following analysis of a clause that records the duties of a leasing tenant: [2; pur1: spor., kop.: chort.]. The pur1 mark indicates that the purpose of the lease is introduced by εἰς (‘for the purpose of’), followed by σπορά (‘sowing’), κοπή (‘cutting’), and χόρτος (‘fodder’).
9. For each clause, I create a link to the relevant data in the synallagma databank. The link has several merits. First, it gives the reader access to the text of the document and other metadata in the papyrological navigator, the Heidelberger Gesamtverzeichnis, and the trismegistos portal (trismegistos.org)—and, via all three, to a digitized picture of the papyrus at the websites of the different collections. The synallagma databank, however, may also offer opportunities for independent study by enabling the reader to plot the documents farther along in terms of time, region, transaction value, identity of the parties, format of the text, and identity of the scribal office—all of which are left out of the discussion in this book.
10. To facilitate document structure analysis, I use an index locorum. Instead of page numbers, I indicate the relevant clause within the document. For this purpose, I found it particularly important to avoid in the first chapter the continuous numbering adopted elsewhere in the book. Accordingly, in the combination ‘subscr. 2; val. 1, 3; crea. 20; remun. 23; sec. 15; exp. 6’, ‘subscr. 2’ indicates that the text is followed by an hypographe, ‘val. 1, 3’, that it ends with the kyria, and with the stipulation clause, ‘crea. 20’ record the payment of money in anticipation of future delivery of a chattel commodity and ‘remun. 23’ the terms of the delivery itself. A trained reader may recognize patterns and documents even without consulting synallagma.
11. The present volume is meant to be follow by a second one, an extensive lexicon of the vocabulary of the legal document. In the present volume, since the text will be first published as a searchable .pdf file, the reader may conduct a lexical search independently. However, I find the addition of a preliminary selective index of words useful because it, too, may indicate a shared cross-genre vocabulary.
12. Since the text is first published online, it may still be updated and corrected occasionally. Readers who note inaccuracies or wish to comment in any other way are very welcome to contact me at uiftach@tauex.tau.ac.il.