Ἰουδαίαν in Acts 2:9: a Diachronic Overview of its Conjectured Emendations

Abstract The appearance of Ἰουδαίαν in the table of nations (Acts 2:9–11) has troubled interpreters for centuries. Several scholars have proposed to emendate the text. The argumentations for such conjectures vary in elaboration and support. This article gives a diachronic overview of the conjectured emendations. It concludes with an evaluation of the discussion from a phenomenological perspective and a summary of the used argumentation, thereby providing input for a reversed engineering approach to the issue.


Introduction
In the context of the Pentecost story (Acts 2:1-13), the author mentions a list of nations, the inhabitants of which miraculously hear the apostles speak in their own language. Over the centuries, this list gave rise to a vast amount of discussion.¹ Especially, Ἰουδαίαν in 2:9 has been regarded as intrinsically difficult on the basis of three observations: (1) the reference to Judea and hence Jews hearing the apostles speak in their native tongue seems awkward,² (2) the reference to Judea (v. 9) does not fit very well in the geographical arrangement³ between Mesopotamia in the east and Cappadocia in the north⁴ (Figure 1), and (3) Ἰουδαίαν should be regarded as an adjective.
The difficulties are not equally weighed⁵ by interpreters, and diverse solutions have been offered. Literary connections with Old Testament table of nation traditions (esp. Gen 10),⁶ Old Testament prophecies like Isa 11:11, contemporary Jewish⁷ and astrological⁸ geographical lists have been suggested and debated. Furthermore, a background in contemporary classical geography (i.e. Strabo)⁹ has been discussed as well as the influence of the geographic viewpoint on Luke's programmatic perspective.¹⁰ Wendt's suggestion that the Pentecost miracle presupposes a "new language" solves the problem but is as ingenious as it is speculative.¹¹ Other interpreters tried to solve the difficulty by assuming a very early corruption in the transmission of the text. The next step to speculate about an alternative location in exchange for Judea was easily made, and a plethora of toponyms have been offered to emend the text. A partial overview of this discussion has been provided by Clemen,¹² Hatch,¹³ and Metzger,¹⁴ but the emendations proposed in the twentieth and twenty-first centuries have not been discussed systematically.
This article presents the discussion to date by providing an overview of the proposed conjectures, including the corresponding considerations, argumentation, and reception history in Section 2. Section 3 concludes this article with an evaluation of the discussion from a phenomenological perspective and a summary of the used argumentation.  7 von Dobschütz, "Zu der Völkerliste Act. 2, [9][10][11][407][408][409][410]Barrett,Acts I. 8 Weinstock,"The Geographical Catalogue in Acts II,[9][10][11][43][44][45][46]Bishop,"Burkitt." 9 Bauckham,"James and the Jerusalem Church,  2 A history of conjectures on Ἰουδαίαν in Acts 2:9 The way interpreters have tried to emend the text of Acts 2:9 can be distinguished in three categories: (1) a change in grammatical function, (2) a correction to an assumed corruption of the text, and (3) by conjecture of a different toponym.
The information provided here mainly follows what is incorporated in the Amsterdam Database of New Testament Conjectural Emendation (ADNTCE).¹⁵ Contrary to the ADNTCE, the data are classified according to the above-mentioned categories and subsequently presented in chronological order, thereby making the discussion and the interrelations between conjectures more explicit. For some cases, more information has been added to complement the data available in the ADNTCE.

Ἰουδαίαν as an adjective
It has been proposed to interpret the grammatical function of Ἰουδαίαν as an adjective. This proposal poses the question to which toponym it should be attached. In 1858, Heinrich Ewald, a German orientalist, Protestant theologian, and Biblical exegete, evaluates Ἰουδαίαν as "völlig unpassend" in the geographic arrangement, since he expects "das große Syrien" in the enumeration. He suggests Συρίαν might have been omitted during textual transmission.¹⁶ According to Ewald, the text should be restored to Ἰουδαίαν Συρίαν. He reaffirmed his position in 1872, now adding the error is Luke's who was unable to finish his work.¹⁷ Both Meyer¹⁸ and Wendt¹⁹ rejected this conjecture. However, a similar case is made by Martin Hengel who interprets Ἰουδαίαν as Greater Judea, i.e. Syria.²⁰ Adolf Hilgenfeld (1895) also took Ἰουδαίαν as an adjective.²¹ He attached it however to Mεσοποταμίαν Ἰουδαίαν. Sahlin supported this proposal but wrongly attributed it to Von Harnack.²² Metzger rebutted the idea since it is not clear to him "why Mesopotamia should deserve to be called 'Judean.'"²³

When in doubt, leave it out
Of the many solutions to the interpretive problem of Ἰουδαίαν in Acts 2:9, the remedy to regard it as a later inclusion²⁴ or a very early corruption²⁵ of the text for which we are not able to identify the original has been widely discussed. The idea to regard it as a corruption was introduced by the English theologian, historian, and mathematician William Whiston²⁶ (1746), although an early citation in Theophylact²⁷ (1097) might hint in its direction. This line of reasoning stems from the observation that a certain geographical clustering can be perceived in the enumeration of countries and peoples if Ἰουδαίαν is left out.²⁸ The view has been reinvented twice²⁹ and has been equally opposed³⁰ as advocated.³¹ Among its advocates, it finds Richard Pervo.³² Pervo's other option, to mark the spot with a blank space, indicating that the original cannot be identified with reasonable certainty, resonates with the opinion of Johannes Marinus Simon Baljon (1898), who was familiar with the readings Συριαν, Aρμενιαν, Bιθυνιαν, Iδουμαιαν, and Ποντον τε και Aσιαν and their originators. Ultimately, Baljon regarded Ἰουδαίαν as a corruption. He did not adopt any of the offered emendations.³³

Conjectured emendations
One of the earliest proposals to substitute Ἰουδαίαν with another toponym might be found in a writing of Aurelius Augustine (397).³⁴ He quoted Acts 2:9 with Ἀρμενίαν, but there is no accompanying remark. Although Tertullian (Adv. Jud. 7.4) cited the proposal, Augustine's solution did not convince many.³⁵ Some 15 years later, in 410, Jerome³⁶ cited Acts 2:7-11 in his commentary on Isaiah 4:11. In his citation, Ἰουδαίαν was substituted with Συρίαν. There is no discussion of the reading and therefore it is debatable whether it should be regarded as a proper conjecture. Baljon³⁷ and Blass³⁸ lend some support to it but do not seem very confident. Opponents simply advocate omission³⁹ or prefer different conjectures.⁴⁰ The German philologist and writer, Caspar Barthius, proposed Ἰδουμαίαν in 1624. Since the narrative is located in Judaea, it seems redundant to explicitly mention Jews in the enumeration of countries.
He therefore proposes to read Ἰδουμαίαν. Support for this conjecture can be found in Josephus and Pliny who distinguish Ἰδουμαία as a separate region from Palestine. Further support might be found in Stephanus who calls the Idumeans Eβραίων ἔθνος.⁴¹ This suggestion was reinvented in 1720 by Richard Bentley⁴² and once again by Otto Lagercrantz⁴³ in 1910. A few scholars⁴⁴ were in favour of this conjecture. Both Bloomfield⁴⁵ and Penn⁴⁶ argue for palaeographical confusion and they also provide manuscript support. Bloomfield claims support for this confounding from Josephus, and Penn refers to textual variants on Mk 3:7. Furthermore, understanding Ἰδουμαίαν as "that tract of country situated on the other side of Jordan, and south-east of Judaea, which was sometimes called Arabia Petraea,"⁴⁷ "exactly fits the geographical order."⁴⁸ Others⁴⁹ were at least familiar with the proposal. However, quite a number of opponents can be found for this conjecture.⁵⁰ The intrinsic difficulty of native Jews hearing their native language seems to have led another German, Erasmus Schmidius⁵¹, mathematician and philologist, to propose Ἰνδίαν in 1634.⁵² The logic behind this conjecture assumes a clustering according to the four cardinal directions on the compass. Exchanging Ἰουδαία for Ἰνδία would create a geographical cluster of Persia, Media, Parthia, Mesopotamia, and India in the East, before proceeding to the geographical clusters in the North, South, and West. Interestingly, there is a passage in John Chrysostom (403) which seems to offer support for this conjecture.⁵³ Although Schmidius' proposal was considered by some⁵⁴ in the early twentieth century, the overarching opinion was against it.⁵⁵ One of the reasons to discard the suggestions was the misfit of the geographical order.
In 1703, Joannes Georgius Graevius, a German-Dutch classical philologist and professor in Duisburg, Deventer, and Utrecht, proposed Γορδυαίαν or Γορδαίαν, some region of Armenia.⁵⁶ The conjecture was reinvented by Francis Crawford Burkitt, Norris Professor of Divinity at the University of Cambridge.⁵⁷ Burkitt discards Ἰουδαίαν based on the geographical arrangement and discusses Tertullian's Ἀρμενίαν. Although both Γορδυαία and Ἀρμενία appear to be ideal candidates from a geographical point of view, ultimately Burkitt prefers Γορδυαία on palaeographic grounds. His argumentation gained some support⁵⁸ but was mainly rejected.⁵⁹ In 1720, the English classical scholar, critic, and theologian, Richard Bentley, preferred Λυδίαν over Ἰδουμαίαν. His emendation did not receive much support. Although it was sometimes only mentioned,⁶⁰ it was already rejected by Heringa in 1793,⁶¹ followed by many others in subsequent years.⁶² A variation to this suggestion can be found in the proposal Kαππαδοκίαν τε καὶ Λυδίαν by Jacob Bryant⁶³ (1767) who substitutes and transposes the word order. Not much is known about its reception. It is mentioned by Van Manen⁶⁴ and criticized by Michelsen.⁶⁵ Gustav Georg Zeltner ( †1738), a Lutheran theologian from Germany, introduced Ἰδαιᾶν or Ἰδαίαν. His view is only known to us from Schulthess who acknowledges a phonetical resemblance but discards the toponym due to its geographical insignificance.⁶⁶  51 Knapp, "Sylloge," abbreviates Erasmus Schmidius to Erasm. He thereby creates an ambiguity which led later scholars to assume Desiderius Erasmus as the originator of the conjecture Ἰνδίαν, e.g. van Manen,231;Hatch,"Apostelgeschichte 2,9," 255;Metzger,Textual Commentary,254. 52 His formulation ("sunt … aliqui") might attribute the conjecture to others, although it could also imply his own authorship. Erasmus Schmidius, Versio Novi Testamenti nova, ad Graecam veritatem emendata, et notae ac animadversiones in idem: quibus partim mutatae alicubi Versionis redditur ratio, partim alia necessaria monentur. Accedit sacer contextus Graecus, cum Versione veteri: nec non Index Rerum et Verborum locupletissimus, 751. 53 Chrysostom, Homiliae in Acta apostolorum. See also the remark in the ADNTCE: "Chrysostom cannot be adduced as a witness to the reading Ἰνδίαν instead of Ἰουδαίαν, though it is remarkable that he mentions India as part of the climactic structure of his sentence (which contrasts Plato and Peter). Since he includes Ἰουδαίαν when he cites the text (in the same homily, Hom. Act In 1742, Thomas Mangey, an English clergyman and scholar, known for his edition of Philo, proposed to restore Kιλικίαν in the text.⁶⁷ Some support can be found in geographical lists in Philo as well as in Acts 6:9. However, Mangey himself already observed Jas 1:1 and 1 Pet 1:1 seem to contradict his proposal. Although his suggestion suits the geographical arrangement, it did not find acclaim.⁶⁸ The Dutch theologian and philologist Tiberius Hemsterhuis, Greek professor in Franeker and Delft, proposed Bιθυνίαν in 1766. A few scholars⁶⁹ followed. Van de Sande Bakhuyzen⁷⁰ and Valckenaer⁷¹ were most explicit in their support, and from these resources we can reconstruct the line of reasoning, which is based on geographic and palaeographic arguments and supported from the co-occurrence in enumerations of geographical areas in classical sources.⁷² The conjecture was widely discussed, though some scholars did not take a stance⁷³ but recognized a possible allusion to 1 Pet 1:1.⁷⁴ Others however rebutted this proposal, mainly because they favoured other emendations.⁷⁵ Several other proposals, although less widely and rigorously debated, have been offered: in 1818, Johannes Schulthess, a Swiss, reformed theologian, assumed that the original reading Ἱουναίαν is a halfcorrect rendering of a Semitic name near Ararat.⁷⁶ As an alternative to the option to omit Ἰουδαίαν, Jan Hendrik Adolf Michelsen, an Evangelical-Lutheran minister and modest adept of the Dutch radical critics, suggested Ἀραμαίαν (1879).⁷⁷ This position has also been put forward independently by Hatch in 1908.⁷⁸ Professor of New Testament and religious history at the University of Bonn, Carl Clemen (1895) ascribed the conjecture Jaudi to Gunkel.⁷⁹ This proposal was supported by Eissfeldt⁸⁰ but opposed by Hatch.⁸¹ The conjecture, however, appears to be based on erroneous transcription of Hebrew words.⁸² Thomas Kelly Cheyne, Oriel Professor of the Interpretation of Holy Scripture at Oxford, suggested Ἰωνίαν in 1901. Cheyne was inspired by a similar conjecture on 1 Macc 8:8 (which substitutes Ionia for India), and he argues that place names are easily confounded. He agrees with Blass that Judea is "intolerable" in the geographical arrangement in Acts 2:9-11.⁸³ However, Hatch preferred Ἀραμαίαν against it.⁸⁴ The eminent German biblical scholar, textual critic, orientalist, and editor of Novum Testamentum Graece, Eberhard Nestle, proposed Ἀδιαβαίαν in 1908.⁸⁵ His suggestion was implicitly contested by Samuel Krauss⁸⁶ who deduced from Rabbinic sources that "Erez Israel" could be used for (a part of) Mesopotamia, which in consequence might explain the occurrence of Ἰουδαίαν in Acts 2:9. Hoennicke simply preferred Ἰνδίαν or Ἰδουμαίαν over Ἀδιαβαίαν.⁸⁷ Although the German scholar Theodor Zahn previously expressed sympathy for Ἰνδίαν,⁸⁸ in 1916 he argued for Ἰουδαῖοι.⁸⁹ He appealed to an Old Latin translation to support his position.⁹⁰ Weinstock referred to this solution⁹¹ but suggests to either omit Ἰουδαίαν or, preferably, read Ἀρμενίαν.⁹² Ropes contested Zahn's claim to support from an ancient Latin manuscript.⁹³ Γαλατίαν or Γαλλίαν (both indicating the same area in Asia Minor)⁹⁴ was suggested in 1941 by Martin Dibelius, professor of New Testament in Heidelberg. He remarked "Judea may have been substituted by an unthinking copyist, especially since Judea is always close to the mind of a Bible reader."⁹⁵ Dibelius admitted there is no specific palaeographic reason, but thought his proposal fitted the geographical arrangement well. Metzger⁹⁶ referred in his rebuttal to Weinstock's argument about a geographic arrangement according to the zodiac circle,⁹⁷ but he seems neither convinced by that view.⁹⁸ After having evaluated several other conjectures, with special attention for Λυδίαν, Eberhard Güting (1975) suggested Λυκίαν, which he regarded "im hohem Maß als passend" due to its importance in Roman times.⁹⁹ Based on the geographical arrangement, John MacDonald Ross (1985) expected "a territory somewhere between Syria and the Caucasus Mountains". In his opinion, Ἰβερίαν (an ancient name for modern Georgia) fits this requirement.¹⁰⁰  81 Hatch, "Apostelgeschichte 2, 9," 255. 82 Instead of assuming some sort of corruption, the creative suggestion is to presuppose a Hebrew source from which the word ‫י‬ ‫א‬ ‫ד‬ ‫י‬ (yʾdy) could be rendered Judea equally well as Jaudi. However, since the Hebrew root for Judea is ‫י‬ ‫ה‬ ‫ו‬ ‫ד‬ ‫ה‬ and not ‫י‬ ‫א‬ ‫ד‬ ‫י‬ this suggestion can be safely rejected. 83 Cheyne, "India". 84 Hatch, "Apostelgeschichte 2, 9," 255.

Conclusion
The survey in the preceding section demonstrates the challenge posed by the text of Acts 2:9. Although interpreters detected serious internal difficulties with the reading Ἰουδαίαν, the supporting external manuscript evidence for this reading has been regarded as overwhelming.¹⁰¹ Simultaneously, the internal difficulties are not easily solved. Therefore, conjectural emendations abound: Cilicia, Armenia, Ida, Iounaia, Ionia, Jaudi, Iberia, Bithynia, Adiabene, Aramea, Idumea, Lydia, Gorduaia, Lycia, Galatia, Gallia, India, and Syria have all been suggested during the past centuries, cf. Figure 2.
When evaluating the historical overview from a phenomenological perspective, we observe that the issues with the originality of Ἰουδαίαν have been considered that serious, and each proposed conjectured emendation that unconvincing that numerous new attempts to solve the issue were attracted. Furthermore, confusion was created by imprecise formulation (see the case of Indian, note 51) or by attributing a conjecture erroneously ascribed to an honoured scholar (the case of Jaudi, see notes 79 and 82).
The overview is also illustrative in showing the diversity of considerations to opt for a certain candidate. In some cases, the fittingness in the geographical arrangement seems to have been the main motivation, while others sought to solve the issues in three different ways: by positing a scribal interpolation, by taking Ἰουδαίαν as an adverb (thus, interpreting a different grammatical function) to a different toponym, or by presuming palaeographical confusion.
In the end, "no one conjecture has proved generally acceptable."¹⁰² The unease remains and the discussion is undecided. In the second part of this article series, the issue will be revisited by addressing the  101 "Despite internal difficulties, the Committee was impressed by the overwhelming preponderance of external evidence supporting Ἰουδαίαν, and therefore retained it in the text." cf. Metzger,Textual Commentary,254. 102 Kilpatrick, "Conjectural Emendation in the New Testament," 351. In his evaluation of conjectural emendation to solve the issue in Acts 2:9, Kilpatrick arrives at a remarkable conclusion: "that emendation may destroy valuable evidence for the history of the list" (p. 353). In our opinion, this remark must be regarded as a non sequitur, since it is not clear to us how this question whether it might be possible to identify an acceptable alternative toponym assuming palaeographical confusion. It will use a computer algorithm to gauge the probability that Ἰουδαίαν could have been the result of a misreading of the original toponym due to letter confusion of majuscule script.