Aristides of Athens as Ecclesial Apologist: Syriac Translation, Greek Papyri, and Armenian Texts
The final installment of chapter 1 of my dissertation!
I am currently writing the dissertation for my second Ph.D., in Church History and Ecclesiology at Stellenbosch University. The tentative title of my dissertation is “Aristides of Athens as Ecclesial Apologist.” As I complete each portion, I will post the segment for anyone who wants to follow my progress. I will not be including the footnotes in these posts, so each chapter will typically be a couple thousand words longer than what’s posted here.
Chapter 1a: “Aristides of Athens as Ecclesial Apologist: Introduction”
Chapter 1b: “Aristides of Athens as Ecclesial Apologist: The Apologia in the Greek Barlaam and Ioasaph”
Chapter 1c is below.
1.7.2.2 The Syriac text of the Apologia
Scholars who give preference to the Syriac tradition have taken a very different position, contending that a translation undertaken with the goal of preserving a text in another language inevitably provides a superior witness to its source. Theodor Zahn summarizes the case for preferring the Syriac translation as the better witness:
Ein Uebersetzer, der als solcher die Pflicht, sein Original treu wiederzugeben empfindet, ist a priori ein besserer Zeuge, als ein Romanschreiber, der, statt selbst eine apologetische Rede vor einem heidnischen König zu erfinden, eine alte Apologie hierfür zustutzt, natürlich ohne damit irgendwelche Verpflichtung in Bezug auf Genauigkeit und Vollständigkeit zu übernehmen. (Zahn 1892:3)
J. Rendel Harris, editor of the editio princeps of the Apologia, preferred the Syriac text and argued that the second-century Roman philosopher Celsus must have been responding to a version of the Apologia of Aristides that closely agreed with the surviving Syriac text (Harris 1921:6–10). More recently, Pouderon and Pierre gave precedence to the structure of the Syriac text in the Sources Chrétiennes edition (Pouderon and Pierre 2003:133–156).
The Syriac text of the Apologia of Aristides is solely preserved in the sixth- or seventh-century Codex Sinaiticus Syr. S. Catherine 16 (Harris and Robinson 1891:4; Hennecke 1893:45; Pouderon and Pierre 2003:137; Rizk 2017:14). This vellum manuscript includes both theological and philosophical works, and the Apologia of Aristides has been copied between a theological text about Nilus the monk and a philosophical discourse from the Middle Platonist Plutarch. Whether intentionally or unintentionally, the copyist of this manuscript placed the apology of the philosopher Aristides squarely between ascetic theology (folios 1b—93a) and Greek philosophy (folios 105a—150b), such that this apologetic which is both theological and philosophical provides a verbal bridge between these two genres.
The Syriac translation represented in this sixth- or seventh-century manuscript seems to have been completed in the fourth century CE (Rizk 2017:14). This assessment is based in part on the frequency and Stoic function of “nature” in this version, which likely would have diminished if the translation had been completed later than the fourth century (Pouderon and Pierre 2003:138–139). A non-extant Greek text seems to have served as the source for the Syriac translation (Brock 1984:5.19; Harris and Robinson 1891:26; Lattke 2008:284; Pouderon and Pierre 2003:157–159).
In the absence of other Syriac translations of the Apologia and with no access to the Syriac translator’s Greek source, it is difficult to know how closely the Syriac text in Codex Sinaiticus Syr. S. Catherine 16 followed its source. Syriac translators of Greek texts in the fourth and fifth centuries tended to expand their sources, while Syriac texts in the sixth and seventh centuries were more likely to reproduce the Greek with great care (Brock 1984:2.3; Brock 1984:5.18–19). In some cases, Syriac translations that had been expanded in earlier centuries were reworked in the seventh century to represent their Greek sources more literally (Brock 1984:2.3–4). While it is conceivable that a fourth-century translator might have expanded the content of the Apologia, it is also possible that one or more translators or editors in the seventh-century could have produced a Syriac text that closely followed an earlier Greek version. Regardless of whether precedence is given to the Syriac or to the Greek, the Syriac text should not be dismissed as a potential witness to a Greek textual tradition earlier than the one represented in the speech of Nachor and elsewhere in the Greek Barlaam and Ioasaph.
New Testament scholar William Rutherford has argued for the priority, at the very least, of the Syriac description of Jewish worship and ethics. The Greek Apologia in Barlaam and Ioasaph preserves a harshly negative evaluation of Jewish knowledge of God (Apol. 14.3–4 Gr.), closely integrated with a narrative of past Jewish disobedience (Apol. 14:2b Gr.). The Syriac Apologia declares, in stark contrast, that the Jewish people’s ethics reveal their closeness to divine truth but that the Jews have erred by worshiping angels (Apol. 14.3–4 Syr.). Rutherford rightly recognizes that the textual changes which shifted the apology’s assessment of the Jewish religion did not take place over a long period of time. Instead, the changes were likely made all at once in the same manuscript, since the harsh appraisal of past Jewish disobedience (14.2b Gr.) and the positive appraisal of Jewish ethics (14.3–4 Syr.) are unlikely ever to have coexisted in the same text at the same time (Rutherford 2013:83). The reference in the Syriac text to Jewish worship of angels supports the supposition that the positive appraisal was written earlier than the negative one, because the accusation of angel worship in the Syriac mirrors a similar accusation in a second-century text quoted by Clement of Alexandria (“Ιουδαίους… λατρεύοντες ἀγγέλοις καὶ ἀρχαγγέλοις,” Strom. 6.5, from Kērygma Petrou).
The Syriac text also appears to preserve an earlier version of the classifications of humanity. In Syriac (as well as the Armenian), the apologist identifies four families of humanity: barbarians, Greeks, Jews, and Christians. The apology in the Greek Barlaam and Ioasaph, however, classifies humanity in three families—worshipers of false gods, Jews, and Christians—and then subcategorizes the “worshipers of so-called gods” (“οἱ τῶν παρʼ ὑμῖν λεγομένων θεῶν προσκυνηταὶ,” Apol. 2) into Chaldeans, Greeks, and Egyptians (see Table 1). William Simpson and later Markus Vinzent have argued that the fourfold structure articulated in the Syriac and Armenian texts is earlier; my research later in this dissertation reinforces their arguments for the priority of the Syriac and Armenian classification (Simpson 2015:123–125; Vinzent 2019:232–233).
1.7.2.3 The Greek text of the Apologia in the papyri
Portions of the Apologia also survive in three Greek papyrus fragments. All of the Greek fragments of the Apologia seem to have been copied no earlier than the third century and no later than the fourth century. Two fragments, P. Oxyrhynchus XV 1778 and P. Heidelberg inv. G 1013, originated in the same codex; these fragments are poorly preserved and include Apologia 4.3–6.2 with lacunae. These fragments of the Apologia have been dated to the fourth century; this estimation is based in part on paleographical similarities to P. Oxyrhynchus VI 847, a vellum fragment of John’s Gospel from the third or fourth century (Clarysse and Orsini 2012:457, 472; Comfort and Barrett 1999:683; Grenfell and Hunt 1908:4; Grenfell and Hunt 1922:1; Hagedorn 2000:40–44; Hunt 1922:1–6). A third fragment, P. London 223 olim inv. 2486, covers Apologia 15.4–16.2 and originates in a third- or fourth-century codex that also contained the Song of Songs in Greek. P. Montserrat Roca IV 44, a third-century fragment that preserves a portion of Song of Songs, may have been torn from this manuscript (Milne 1923:73-77; Roca-Puig 1975:89–91).
The stemma proposed by Bernard Pouderon and Marie-Joseph Pierre in their critical edition of the Apologia arranges the textual witnesses in three geographical branches: Egyptian, Greek, and Oriental. Pouderon and Pierre grouped all known Greek papyrus fragments in the same textual family and identified them as “Egyptian” (Pouderon and Pierre 2003:172). William Alexander Simpson’s dissertation research has demonstrated, however, that the classification of the witnesses by Pouderon and Pierre is in error (Simpson 2015:75–77). Careful comparisons of these fragments with the Syriac text and with the Greek Barlaam and Ioasaph reveal that the Greek papyri represent two distinct textual traditions. P. Oxyrhynchus XV 1778 and P. Heidelberg inv. G 1013 derive from a tradition (“Greek 1” in Simpson’s revised stemma) that is more reflective of the Greek source incorporated into Barlaam and Ioasaph (Simpson 2015:217–218, 222–223). P. London 223 seems to represent a differing textual tradition (“Greek 2,” in Simpson’s stemma) from which the Syriac translation was also derived (Simpson 2015:233; see especially the textual expansions illustrated in Simpson 2015:353–355). Some distinctive readings shared by P. London 223 olim inv. 2486 and the Syriac translation, such as references to burial practices and ascetic commitments, seem to reflect concerns that emerged later than the second century CE (Simpson 2015:223).
Although neither the Syriac translation nor the apology in the Greek Barlaam and Ioasaph can be identified precisely with the second-century words of Aristides, it seems that Barlaam and Ioasaph does, at the very least, preserve significant pre-fourth century portions of the Apologia. The sparing allusions to Scripture, the absence of clear references to God the Father, and the lack of explicit Trinitarian language all suggest a pre-fourth century basis for the apology (Vinzent 2019:256). Nevertheless, the surviving Greek texts can take the researcher only as far back as the fourth century with a high degree of confidence in the words of the text. Although the Greek Barlaam and Ioasaph maintains a text very similar to the text as it stood in the fourth century, “a broad expansion (and possible restructuring) of the Apology” might have taken place in the fourth century or earlier (Simpson 2015:77).
1.7.2.4 Armenian translations of the Apologia
The Apologia in the Armenian tradition is significantly shorter then the Greek or the Syriac, and it is at least possible that the Armenian rendering represents an earlier textual tradition than the Syriac or the Greek Barlaam and Ioasaph. If so, the length of the original apology might have been only a couple of chapters, in contrast to the sixteen chapters found in modern editions of the Syriac and Greek texts. The Apologia survives in Armenian both as an independent text and as a text embedded in Armenian translations of Barlaam and Ioasaph in the menologia.
1.7.2.4.1 The Armenian Apologia as an independent text
Three manuscripts preserve similar Armenian renderings of the Apologia as an independent apology: the tenth-century CE Codex Yerevan Matenadaran 2679 and two much later codices, Codex Yerevan Matenadaran 6228 and Codex Venice San Lazzaro 218 (Rizk 2017:18–20). The wording in Codex Yerevan Matenadaran 2679 and Codex Venice San Lazzaro 218 is virtually identical; it was the San Lazzaro manuscript that provided a basis for the Mekhitarist edition of the Apologia published in Armenian and Latin in 1878. The Armenian seems to have been translated from a Greek source (Pouderon 2000:173–193). The text incorporates grammatical constructions and theological language that characterized translations in the fifth century CE, when pre-Hellenizing translators frequently formulated rushed renderings of Greek texts (Rizk 2017:51–54, 236–237).
The independent Armenian Apologia is far shorter than either the Greek Barlaam and Ioasaph or the Syriac translation, concluding soon after summarizing the origins of the different families of humanity. Some later additions may be observed in the independent Armenian Apologia. The reference to Mary as Theotokos, for example, is unlikely to have been present in the text at any point prior to the hyperdyophysite controversies that culminated in the Council of Ephesus (431 CE). At the same time, the Armenian Apologia which survives as an independent text may at times preserve more primitive content than the Greek text. The Armenian agrees with the Syriac, for example, in identifying the families of humanity as “barbarians and Greeks, Jews and Christians,” (Apol. 2, Syr.) which seems to be the earlier reading.
It could be that the Armenian apology is shorter because the translator either possessed a partial text or intentionally produced a condensed version of a longer text. It is also conceivable that the Armenian translator was working from an earlier, now-lost Greek version of the Apologia that was shorter than the Syriac translation or the Greek text in Barlaam and Ioasaph (Simpson 2015:124–125; Vinzent 2019:232–233). If so, this short Armenian version might translate a very early stage in the development of the Apologia. Two sixteenth-century manuscripts (British Library Add MS 5118 and Patriarchal Library of Jerusalem MS 210) include the full catalog of Greek gods that appears in the Syriac version and in the Greek Barlaam and Ioasaph but not in the Armenian, except without the negative comments found in the Greek and Syriac versions. Markus Vinzent has suggested that such a list might suggest “an earlier stage of a non-Christian catalogue of Greek gods. Such a catalogue might have been integrated with… the Christian criticisms of the Greek gods, when the catalogue became integrated in the extended text” (Vinzent 2019:230). If the catalog of Greek deities was integrated into the text later, the initial second-century Apologia might have been more similar to the independent Armenian translation than to any extant Greek and Syriac versions.
1.7.2.4.2 The Armenian Apologia in Barlaam and Ioasaph in the menologia
Condensed versions of the Apologia of Aristides also appear at least three times in the Armenian menologium tradition, embedded as the speech of Nachor in Armenian renderings of Barlaam and Ioasaph. The stories of Barlaam and Ioasaph preserved in these versions have been abridged for liturgical usage, and the reworked Apologia differs in each one. Armenian menologia manuscripts in which Barlaam and Ioasaph has been appended include the fifteenth-century codices Bodleian Codex Marsh 438 and British Museum Orientalis 4580, as well as the seventeenth-century manuscript Bodleian Canonici Orientalis 131 (Baronian and Conybeare 1918:32–70, 80).
Although the speech of Nachor has been abridged in the Armenian menologia, sufficient remnants remain to recognize a text that is traceable to some version of the Apologia in the Greek Barlaam and Ioasaph. In Bodleian Codex Marsh 438, for example, the unwilling prophet Nachor mentions the superstitions of “the Greeks and the Egyptians and the Chaldeans,” a triad of ethnic classifications that appears in the Greek Barlaam and Ioasaph but never in the Syriac or independent Armenian versions of the Apologia. Whether translated from the Greek Barlaam and Ioasaph or from a Syriac rendering of the Greek (Conybeare 1896:135–139), the content in Apologia of Aristides as preserved in the menologia derives from a text traceable to the Greek Barlaam and Ioasaph. Since the content of the Apologia in the Armenian menologia seems to be dependent on the text preserved in the Greek Barlaam and Ioasaph, the Armenian menologium tradition provides little, if any, new information for the study of the second-century apology.
1.7.2.5 A Functional Approach to the Text of the Apologia
In the late nineteenth and early twentieth centuries, several scholars gave precedence to particular traditions based primarily on the language, Syriac or Greek (see, e.g., Zahn 1892:3; Harnack 1893:97). However, to prefer any version or tradition on the sheer basis of the language is to oversimplify the complex textual history of the Apologia of Aristides. It is not tenable, for example, simply to follow the Greek texts. Although the redactor who wove the Apologia into Barlaam and Ioasaph was faithful to his source text (Simpson 2015:231–232), this Greek textual tradition seems to have been altered at some previous point, resulting in a negative portrayal of the Jews (Rutherford 2013:83) and in a reclassification of the human race into three families instead of the earlier four (Vinzent 2019:232–233). Furthermore, as William Simpson’s dissertation research has demonstrated, the Greek fragments testify to two distinct Greek textual traditions. The elaborations in Apologia 15.4–16.2 that are apparent when P. London 223 olim inv. 2486 is compared with the Greek Barlaam and Ioasaph reveal that at least one of these traditions was already being enlarged no later than the fourth century CE (Simpson 2015:352–358). The Greek Apologia of Aristides in Barlaam and Ioasaph must be viewed, therefore, not simply as a second-century text but as a text as it was received no later than the eleventh century.
The Syriac and Armenian texts do appear to preserve some earlier concepts and structures. Yet solely following the Syriac text results in a different set of difficulties. Although the source text utilized by the Syriac translator preserved the original categorizations of humanity, the Syriac translator’s embellishments seem to have expanded the Greek source text (Simpson 2015:214, 233, 236).
Do these challenges place the original words of Aristides so far beyond the reach of contemporary scholars that little—if anything—can be said about the second-century Apologia? I contend that it is possible to engage with the apology as a second-century text; however, this possibility requires moving past any simplistic elevations of a single language or even a singular textual tradition. In 1894, a mere three years after the editio princeps of the Apologia was published, German church historian Friedrich Lauchert suggested that any quest for the second-century text would require interaction not only with the Greek text in Barlaam and Ioasaph but also with the Syriac and Armenian renderings (Lauchert 1894:291). After carefully engaging with all three textual traditions, I am convinced that Lauchert was right.
When the Greek, Syriac, and Armenian traditions are considered together, at least four observations may be sustained on the basis of historical and textual evidences:
The second-century apology included no less than a Greek version of the text represented in the independent Armenian text. Despite occasional additions in the Armenian version, most of the content in the independent Armenian text appears, in some form, in all of the textual traditions. Some scholars have suggested that these chapters might testify to a initial second-century Greek text that was shorter than any extant version. According to Markus Vinzent,
The lost longer Greek versions, which are attested by the papyri, the Syriac translation, and Eznik [of Kolb], might be fourth- and fifth-century elaborations and extensions of a shorter Greek version, attested by the Greek text in the [Barlaam and Ioasaph]. This Greek text was itself an elaboration of an even much shorter Apology that is witnessed by the extant Armenian version, which gives us only two initial chapters. (Vinzent 2019:230).
The Greek apology began to be altered no later than the fourth century CE. Close analysis of P. London 223 olim inv. 2486 reveals a third- or fourth-century text in which embellishments are already present (Simpson 2015:223–226, 352–358). References to fasting and to provisions for the burials of poorer members in this fragment are more congruent with fourth-century ecclesial concerns than with second-century apologetics. The redactor of the Greek Barlaam and Ioasaph and the translator of the Syriac inherited texts that included descriptions of “the Egyptians”; however, this too seems to reflect an addition to the second-century text (Simpson 2015:124–129). Snippets of later creedal and liturgical language have also been introduced into the apology in Barlaam and Ioasaph. Apologia 15 in the Greek text reads, “ἐν πνεύµατι ἁγίῳ ἀπ᾿οὐρανοῦ καταβὰς διὰ τὴν σωτηρίαν τῶν ἀνθρώπων, καὶ ἐκ παρθένου ἁγίας γεννηθεὶς,” which seems to reflect verbiage from the Creed of Nicaea (“γεννηθέντα… διὰ τὴν ἡμετέραν σωτηρίαν”) and the Niceno-Constantinopolitan Creed (“γεννηθέντα… διὰ τὴν ἡμετέραν σωτηρίαν κατελθόντα ἐκ τῶν οὐρανῶν… ἐκ πνεύματος άγίου καὶ Μαρίας τῆς παρθένου”).
The Apologia in Barlaam and Ioasaph preserves some second-century structures and arguments. Despite these later expansions, the Apologia in Barlaam and Ioasaph exhibits several features that seem to stem from the second century. The relative rarity of later theological concepts in the apology embedded in the Greek Barlaam and Ioasaph suggests that significant content may survive in a pre-fourth-century form. There is no Logos theology in the Apologia in Barlaam and Ioasaph, and neither the Trinity nor God the Father is mentioned explicitly (Picard 1892:56; Simpson 2015:145–161, 186). Additionally, this version of the Apologia never appeals to Old Testament prophecies or quotes New Testament texts. Unlike other prominent Christian apologies, the actual apologetic argument does not depend on biblical texts at all. If the text had been completely reworked in the fourth century or later, it seems likely that some of these gaps would have been filled. The Apologia has been embellished, but these expansions have not completely obliterated the structure and arguments of a more primitive text. Although certainty regarding the second-century form of the text is not possible, it seems likely that some second-century verbiage has survived in the Apologia embedded in Barlaam and Ioasaph.
The redactor of the Greek Barlaam and Ioasaph tended toward preservation of the text of the Apologia that he received. The redactor seems to have made some alterations in the Apologia. He may have changed “barbarians” to “Chaldeans,” for example, forging a connection to the text of the Georgian Balavariani that he inherited (Harris and Robinson 1891:70; see also Pouderon and Pierre 2003:150–154). Nevertheless, with relatively few exceptions, the redactor of Barlaam and Ioasaph preserved the text of the Apologia that he received. The most significant alterations which separate the non-extant second-century text of the Apologia from the apology embedded in Barlaam and Ioasaph seem likely to have occurred prior to the redaction that resulted in the Greek Barlaam and Ioasaph.
The prioritization of any version of the Apologia on the basis of language alone oversimplifies the textual issues presented by the various versions. This textual complexity justifies the decision of Pouderon and Pierre to include each text separately in their critical edition instead of assembling a hypothetical text from a puzzle of disparate elements in different languages. The result of such a hypothetical text would have been artificial at best, a monstrous patchwork at worst (Pouderon and Pierre 2003:8–10). Yet what other options are available for the consideration of this text as an example of second-century apologetics?
I propose a parsimonious approach that prioritizes portions of the text that are most likely traceable to the second century. Following the observations above, the following principles have guided my engagement with the texts:
Because the second-century apology included no less than a Greek version of the text represented in the independent Armenian text, I have prioritized the content found in the Greek text which also appears in the Armenian text. As a result, the primary arguments throughout this research rely on the components of the apology that are most likely to have been present in the second century.
Because the Greek apology began to be altered no later than the fourth century CE, I have followed the Syriac or Armenian texts when there are clear reasons to think that one of these texts has preserved earlier content. In the families of the human race and in the description of the Jews, for example, the Syriac text is more congruent with other second-century texts and provides the best explanation of the textual history.
Because the Apologia in Barlaam and Ioasaph preserves some second-century structures and because the redactor tended toward preservation of the text, I have cautiously utilized the Greek text when there are clear reasons to think that the content derives from the second century. If content in the Greek text parallels other second-century texts, for example, I have treated the content as secondary and supportive but ultimately uncertain.
I do not pretend that such an approach restores the Apologia to its second-century form. Until further manuscript evidence emerges, no such recovery is possible (Lattke 2008:392–393). Instead, this approach takes seriously the complex history of this text and prioritizes what is most certain while downplaying what is less certain. The result is a reasonable and functional textual basis for discussing the Apologia as an example of second-century apologetics.