The Syriac language is best known from the large corpus of Christian literature that began to be produced by the 2nd cent. and continues to be composed until the present day. Syriac is, however, also attested in a much smaller corpus of non-Christian documentary texts that date from the 1st to the 3rd cent. These texts are traditionally termed ‘Old Syriac’. The largest group of Old Syriac texts are inscriptions, which number over 100 (ed. Drijvers and Healey). In addition to the Old Syriac inscriptions, there are also the Old Syriac documents, which include three Syriac parchments as well as several Syriac texts found on (primarily) Greek documents from the Middle Euphrates region.
The first of the Old Syriac parchments was discovered in 1933 during excavations of Dura-Europos , an important military outpost for the Parthian Empire and then the Roman Empire on the Euphrates. This document, which is now known as P. Dura 28, is a bill of sale for a female slave dated 9 May 243 (the most accessible version of the text is in Drijvers and Healey, 232–6 [s.v. P1]; but see also Bellinger and Welles; Goldstein; Healey 2009, 264–75 [with plate 9]; Torrey; Welles, Fink, and Gilliam, 142–9 [with pl. LXIX, LXXI]; plates are also found in Moller, 185–6). In addition to this Syriac parchment, more than 150 other documents were found at Dura-Europos, most of which are in Greek or Latin though there are also a few in Iranian or Aramaic (ed. Welles, Fink, and Gilliam). For more than fifty years, P. Dura 28 was without parallel, but in 1988 two similar Old Syriac parchments, now known as P. Euphrates 19 and 20, came to light. Though not found in controlled excavations, it is likely that both of these documents originate from Appadana (Neapolis), just north of Dura-Europos on the Euphrates. P. Euphrates 19 is a transfer of debt dated 28 Dec. 240 (the most accessible version of the text is in Drijvers and Healey, 237–42 [s.v. P2], but see also Teixidor 1989, 220; 1990, 144–54 [includes two plates]; Brock; Aggoula; Healey 2008; 2009, 252–64 [with plate 8]). P. Euphrates 20 is a property lease dated 1 Sept. 242 (the most accessible version of the text is in Drijvers and Healey, 243–8 [s.v. P3]; see also Teixidor 1989; 1990, 154–57; 1991–1992 [includes two plates]; Brock; Aggoula). These two Syriac parchments belong to a cache of texts that also includes 19 Greek papyri and parchments (ed. Feissel and Gascou [and Teixidor]). On several of these Greek texts, there is additional writing in Syriac. P. Euphrates 6 (with its duplicate P. Euphrates 7), for instance, contains a bill of sale for a slave in Greek, which is followed by seven lines of Syriac summarizing the sale as well as a list of witnesses and guarantors (verso), both in Syriac. In addition, a Syriac subscription occurs on two of the Greek papyri, viz. P. Euphrates 3 and its duplicate P. Euphrates 4. This subscription is the earliest witness to the use of Syriac on papyrus (see Papyri, Syriac).
Since their discovery, the Old Syriac documents have proven important on a number of fronts. In both their content and form they have added to the growing body of data for the Aramaic legal tradition. The three Old Syriac parchments are of the Doppelurkunde-type, in which the main text is presented on the lower portion of the parchment and an abbreviated summary occurs on the top. This top section would have originally been rolled up and sealed in order to prevent tampering. This practice is well attested both in Greek and Aramaic papyrology, and it may have antecedents in the cuneiform tradition where tablets were enclosed in clay envelopes. The legal terminology in the Old Syriac parchments also points to continuity with other Aramaic legal texts, such as the Elephantine papyri, the Samaria Papyri, the Nabataean papyri, and the Jewish legal texts from the Judaean desert and the ‘Cave of Letters’ in Naḥal Ḥever (see Healey 2005). In P. Dura 28, for instance, it is stated: ‘From today until forever, you, Tiro, the buyer, and your heirs will have authority (šlyṭ) over this female slave whom I have sold to you, to possess or to sell or to do with her whatever you wish (kl d-tṣbʾ)’ (ln. 11–12). This type of statement — commonly called a Kyrieia-clause — is found throughout the Aramaic legal tradition, as well as in Greek papyri.
The dating formulae in the Old Syriac documents have provided valuable information for the chronology of the last kings of Edessa (see Abgarids). P. Dura 28 shows that the Abgarid dynasty must have come to an end in 212/213 when the city became a Roman colonia. On the basis of P. Euphrates 19, which states that 28 Dec. 240 is the 2nd year of king Abgar, it can be established that the dynasty was restored in 239 (or late 238) under Abgar X, son of Maʿnu. Finally, it seems that Edessa must have reverted to a colonia by 242 on the basis of P. Euphrates 20, which gives 1 Sept. 242 as year 30 of the colonia.
The Old Syriac documents are also an important witness to an early stage of the Syriac language. The script of the Old Syriac documents is not the monumental Esṭrangela that is found in most of the Old Syriac inscriptions and the oldest Syriac literary mss., but rather it is of a more cursive type (especially in P. Euphrates 20) that also occurs in some mosaic inscriptions. These documents, thus, corroborate the existence of a cursive script, which was used alongside Esṭrangela already in the earliest period of Syriac and which would later develop into Serṭo (see Healey 2000). In orthography, the Old Syriac documents have features in common with the Old Syriac inscriptions, but different from Classical Syriac, e.g. the defective writing of historical short *u (e.g. ḥšbn ‘reckoning’) versus the plene writing with a mater lectionis in Classical Syriac (e.g. ḥwšbn) and the representation of the etymological lateral *ɬ (= traditional ś) with š (e.g. ʿšryn ‘twenty’) versus s in Classical Syriac (e.g. ʿsryn). In one important diagnostic feature, however, the Old Syriac documents differ from most of the Old Syriac inscriptions in agreement with Classical Syriac: the prefix marker of the 3rd-person masculine prefix-conjugation is y in most of the Old Syriac inscriptions, but n in both the Old Syriac documents and Classical Syriac. This shows that there was already linguistic variation in Old Syriac (see recently Healey 2008). Finally, the Old Syriac documents provide evidence for Syriac-Greek language contact already in the middle of the 3rd cent. as there are a relatively large number of Greek lexical borrowings (20 types that occur more than 50 times) in the documents, several of which show a significant degree of incorporation into the Syriac language, e.g. ʾsṭrṭgwtʾ ‘strategos-ship’ (< Greek stratēgós) and ʾrkwnwtʾ ‘archon-ship’ (< Greek árchōn).