Δια as a Polysemous Preposition in Early Byzantine Greek: “Dead Ends” and Other Uses in the Qurrah Archive (VIII AD)

Abstract

In this article, I offer a systematic description of the various uses of the preposition διά in the Early Byzantine archive of Qurrah ibn Sharik (VIII AD), an archive in which the preposition is attested remarkably frequently. Functionally, the use of διά is reminiscent of the Classical period, in that various older uses are attested that no longer occur in Modern Greek (such as PATH, INTERMEDIARY, and INSTRUMENT). However, there are also various innovative uses that are attested neither in the Classical nor in the Modern period (such as AGENT, SOURCE, and OPPONENT). The occurrence of these “dead ends” shows that the functional development of prepositions from Classical to Modern Greek should not be thought of as linear.

Keywords:

• διά
• prepositional system
• functional development
• Early Byzantine Greek
• documentary papyri
• Qurrah ibn Sharik

Acknowledgements

An earlier version of this paper was presented at the conference The Language(s) of the Papyrus Archives (Ghent, 18 September 2015). I would like to thank the participants of that conference, as well as Mark Janse and an anonymous referee, for helpful comments.

Notes

1 Cf. Bortone (2010, 155).

2 See Bortone (2010, 193) for this term.

3 Cf. Bortone (2010, 215).

4 There is some discussion about the status of the “semantic” roles proposed in Table 1: positing them entails polysemy of the preposition, but in some cases, the context seems to make a rather important contribution. From a Cognitive Linguistic point of view, we can say that the more a preposition is used in a certain context, the more that context will become part of its semantics (see, e.g. Langacker’s [1987] notion of “entrenchment”). In other words, some roles will be more “semantic” than others (PRICE, for example, depends entirely on the prepositional complement).

5 For further background on this archive, see Abbott (1938), Bell (1910, xi–xlv), Richter (2010) and Papaconstantinou (2015a, 2015b). Alternatively, the archive is called “the archive of Basil the Pagarch”.

6 Archives can be defined as groups of texts that have been collected in antiquity by persons or institutions, for example, because they were useful and needed to be kept, or because they had sentimental value. See further Vandorpe (2009).

7 Due to the introduction of new writing materials such as paper, and the rise of Arabic in Egypt.

8 The Trismegistos database records 182 texts in the archive, 178 of which certain, and 4 of which uncertain. For an overview, see http://www.trismegistos.org/archive/124.

9 Richter (2010, 215) relates the use of Arabic to the government at Fustat, that of Greek to the provincial and urban elite, and that of Coptic to the villages and local elite.

10 For an overview, see Richter (2010, 197–198).

11 On the notion of “pagarch”, see Liebeschuetz (1974).

12 Cf. Papaconstantinou (2015a, 44).

13

To receive a letter from the governor actually meant to receive two letters, one of them written in Greek and therefore rather easily comprehensible to the pagarch and his staff, the other one written in Arabic, thus not comprehensible (at least not to everybody), but nevertheless to be archived in the office. (Richter 2010, 200)

14 In a later article, Bell offers a less positive evaluation of the style of the archive:

The wordy, rambling style, with constant repetition of the same ideas, and with clause tacked on to clause, participle piled on participle, and a want of any real unity of construction – this, as well as the curious grammatical usages, is thoroughly typical of the series. (Bell 1926, 276–277)

15 Cf. also Mayser (1934, 426), who notes that there are no examples in the Ptolemaic papyri.

16 Translations are my own unless otherwise indicated. Relevant instances of διά have been marked in bold.

17 Note that the preposition is a restoration by the editor, and not part of the extant text.

18 Luraghi (2000, 276).

19 The terms “trajector” and “landmark” stem from Cognitive Linguistics, and very generally refer to a participant that is in primary focus (the trajector) versus a participant that is in secondary focus (the landmark).

20 Cf. Luraghi (2014, 68).

21 Luraghi (1989, 294) calls it “the most common means for expressing cause”.

22 See SB.10.10454, l. 4 (709 AD); P.Lond.4.1344, l. 5 (710 AD); P.Lond.4.1350, l. 10 (710 AD).

23 See P.Ross.Georg.4.16, l. 5 (710 AD).

24 Contrast, e.g. ὡς ἴδον Ἥϕαιστον διὰ δώματα ποιπνύοντα (Hom. Il. 1.600) “as they saw Hephaestus puffing through the palace” with βὰν δ᾽ ἰέναι προτέρω διὰ δώματος, ἧος ἵκοντο Τηλέμαχον (Hom. Od. 15.109–10) “they walked through the hall, until they reached Telemachus”.

25 For similar examples, see P.Lond.4.1346, l. 18 (710 AD); SB.10.10459, l. 13 (VIII AD).

26 E.g. πλησίον τῆς στράτας (Geo. Syc. V. Theod. Syc. 67.1) “near the street”; ῥογεύσωμεν αὐτὸ εἰς τὰς στράτας τοῖς πτωχοῖς (Call. V. Hyp. 20.1) “let’s give this in the streets to the poor”.

27 For further discussion of this example, see Section 2.2.5.

28 Cf. Luraghi (2014, 70).

29 Cf. Luraghi (1989, 300).

30 Compare our observations on CAUSE in Section 2.1.2.

31 For similar examples, see, e.g. P.Lond.4.1358, l. 7 (709 AD); P.Lond.4.1363, l. 7 (710 AD); P.Lond.4.1375, l. 12 (710 AD); P.Lond.4.1371, ll. 10–11 (710–711 AD).

32 For similar examples, see, e.g. P.Berl.Frisk.6, l. 14 (710 AD), SB.3.7241, l. 22 (710 AD), and perhaps P.Lond.4.1361, l. 3 (710 AD).

33 Cf. Bell (1910, 29).

34 For similar examples, see, e.g. P.Lond.4.1337, l. 22 (709 AD); P.Berl.Frisk.6, ll. 10, 19 (710 AD); P.Ross. Georg.4.8, l. 5 (711 AD); SB.10.10459, l. 8 (VIII AD).

35 For similar examples, see, e.g. P.Lond.4.1339, l. 8 (709 AD); P.Lond.4.1337, ll. 20–21 (709 AD); P.Lond.4.1356, l. 10 (710 AD); P.Ross.Georg. 4.14, l. 5 (710 AD).

36 See, e.g. P.Lond.4.1394, l. 10 (708–709 AD); P.Lond.4.1346, l. 9 (710 AD); P.Lond.4.1350, l. 15 (710 AD); P.Lond.4.1353, ll. 9, 26 (710 AD); P.Lond.4.1359, l. 11 (710 AD); P.Lond.4.1369, l. 15 (710 AD); SB.3.7241, l. 45 (710 AD).

37 See Crespo (1988, 48–49) for an overview.

38 For some examples, see P.Lond.4.1332 (708 AD), l. 15; P.Lond.4.1333, l. 17 (708 AD); P.Lond.4.1339, l. 2 (709 AD). Also, note the expression πᾶσαν σπουδὴν ποίησον in P.Lond.4.1335, l. 16 (709 AD).

39 Cf. Luraghi (2014, 67).

40 For similar examples, see, e.g. P.Lond.4.1337, ll. 6, 14–15 (709 AD); P.Lond.4.1365, l. 5 (710 AD); P.Lond.4.1370, l. 10 (710 AD).

41 For similar examples, see, e.g. P.Lond.4.1398, l. 3 (709 AD); P.Lond.4.1362, l. 9 (710 AD); P.Lond.4.1380, l. 17 (710 AD).

42 For another possible example, see P.Lond.4.1349, l. 5 (710 AD).

43 See, e.g. Jannaris (1897, 375) and Bortone (2010, 235).

44 Cf. Section 2.2.1.

45 Cf. similarly Cadell’s (1983, 113) rendering of SB.10.10454, l. 3 (709 AD).

46 For similar examples, see, e.g. P.Lond.4.1381, l. 4 (708–710 AD); P.Lond.4.1337, l. 19 (709 AD); P.Lond.4.1346, l. 13 (710 AD); P.Lond.4.1349, l. 14 (710 AD); P.Lond.4.1366, l. 12 (710 AD); P.Lond.4.1380, l. 26 (710 AD).

47 Luraghi (2003, 31) notes, however, that in the Classical period the use of such source expressions to denote the agent remains rather marginal. See further George (2005).

48 As one of the reviewers notes, this provides a counter-example to the principle of “unidirectionality” which is maintained in grammaticalization theory, according to which linguistic change is unidirectional from lexical to grammatical meaning (see, e.g. Haspelmath [2004]).

49 Cf. Section 2.2.4.

50 See Luraghi (2000, 288).

51 Cf. Mayser (1934, 509), who notes competition between ὑπό, ἀπό, ἐκ, διά, and παρά.

52 Cf. Ljungvik (1932, 29–30).

53 Cf. Wenger (1906, 10): “es sei darum hier besonders hervorgehoben, daß διά zuweilen auch zur Bezeichnung einer Zahlung ab aliquo angewendet wird, wo wir eher παρά erwarteten.”

54 For further examples, see Wenger (1906, 10–11) and Ljungvik (1932, 29–30).

55 For some late examples, see, e.g. P.Harrauer.60 (VII AD); P.Lond.5.1745, l. 1 (VII/VIII AD); P.Lond.5.1749, l. 1 (VII/VIII AD); P.Pintaudi.27, ll. 1–2 (VII/VIII AD).

56 Another example could, perhaps, be found in P.Strasb.7.680, l. 3 (VII AD), but it is unclear whether διά really indicates SOURCE.

57 See Bell (1926, 266–273) for further details.

58 Richter (2010, 209–211) discusses some examples of linguistic interference (Greek–Arabic) in the Qurrah dossier, but these all belong to the lexical level.

59 Cf. Richter (2010, 216): “there is hardly any reason to believe that language contact between Arabic and Greek speakers and individual Arabic–Greek bilingualism in early Islamic Egypt reached a considerable extent.”

60 For some more examples (not mentioned by Bell), see CPR.30.21, l. 5 (ca. 640–700 AD); P.Amh.2.154, ll. 4–5 (ca. 643–644 AD?); SB.26.16350, l. 2 (ca. 643–644 AD); CPR.25.33, l. 5 (VII AD); P.Apoll.5, l. 1 (VII AD). There are no examples of this use in literary texts.

61 Jernstedt’s suggestion is also rejected by Ljungvik (1932, 25).

62 Wolf (1912, 50) mentions the use of παρέρχομαι διά (with the genitive) in John Malalas’ Chronography (VI AD), and notes that it is used there with the meaning of “vorbeikommen”.

63 Perhaps we should not entirely exclude interference between the two prepositions (compare our observations in Section 2.2.5). On the general semantic similarity between the two prepositions, see Luraghi (2003, 202).

64 See, e.g. Bortone (2010, 122).

65 On documentary papyri, see, e.g. Halla-aho (2010).

Additional information

Funding

My work was funded by the Fund for Scientific Research – Flanders [grant number FWO13/PDO/008].