STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. ARRIAN'S ANABASIS AND BOOK XVII OF DIODORUS By R. B. Steele Frankel in Die Quellen der Alexanderhistoriker (pp. 460 ff.) con- cludes that Diodorus reproduces the work of Clitarchus with some modifications, and that he, as well as Aristobulus, the chief source of Arrian, made use of Callisthenes, Onesicritus, Chares, and Near- chus. The selection by Arrian of Aristobulus and Ptolemy is of itself an indication that their historical method differed from that of Clitarchus. He criticizes or omits many of the highly colored incidents found in Diodorus and has unnumbered details not re- corded by the latter. That they are widely different in statements of facts is admitted, and the object of this investigation is to determine whether they have retained a similar form of statement. The data will be presented under two heads: I, personal — historical, ethical, rhetorical; and II, individual words and syntax. I. PERSONAL Arrian, Anabasis (vii. 22. 5), 1 in an original discussion names Seleucus as the greatest of the followers of Alexander. In contrast with this Diodorus (xvii. 103. 6) and Curtius (ix. 8. 22-24), by implication, assign this position to Ptolemy. Diodorus refers to other writers only a few times, but Arrian, in addition to his main sources, frequently mentions others and also reports items from the logos. He speaks of heroes only incidentally (iv. 8. 3; iv. 11. 3), while Diodorus in his opening paragraph, section 4, declares that Alexander was equal to any of the ancients, and by an adjective or adverb presents the heroic deeds of Admetus (45. 6), of Erigyius (83. 6), and of Porus (88. 6; see 95. 2; 102. 4). The ethical environment set forth by the two writers is not the same. Diodorus (118. 1) mentions evaeffes as a spring of action; t6 Otiov (116. 1) and t6 ha.Ly.6vMv (116. 5). Arrian freely uses the last two expressions and has tv\6v a dozen times (see Boehner, De Arriani dicendi genere, 2 p. 39), with a few incidental references to 1 In succeeding references, with Arrian sc. Anabasis, and with Diodorus xvii. 2 Acta Seminarii Phil. Erlangensis, IV, 1 ff. [Classical Philology XV, July, 1920] 282 Arrian's "Anabasis" and Diodorus 283 riixv, whose importance in the eyes of Diodorus is shown by more than a score of occurrences. Various forms of rvyx&vu are found, the compound in iv- most freely in Arrian. In contrast with this are such expressions as avvej3ri yeviadai (Diodorus 63. 4; 66. 2; 100. 1, and 103. 7). Compare Arrian i. 1. 9; i. 27. 7; ii. 10.3; iv. 2. 5; v. 23. 5, usually more frequently, and with a negative (27.7; 46. 6); and with p-yi^s (38. 5; 99. 1; 100. 1). The latter is in Arrian (ii. 24. 6), but also \6yov (v. 7. 3; vi. 9. 5; vii. 7. 7). Some rhetorical differences are noticeable. There are a few examples of chiasmus in Arrian (e.g., i. 12. 9; ii. 7. 9), but, in con- trast with this, the etymological figure occurs frequently (vii. 17. 1; vii. 23. 8); dvaias .... dveadai (vii. 14. 1; vii. 17. 4). Boehner (p. 34) calls attention to the frequent repetition of the noun by Arrian, thus avoiding the use of a pronoun. However, this usage with the verb is just as noticeable, 'Appu/Jas yap voaui airidavtv. airWavtv 5e km 'Avt'ioxos (see i. 16. 6; i. 25. 8; ii. 4. 8; ii. 11. 5, 9; ii. 25. 2; ii. 27. 6; iii. 5. 4; iii. 5. 5). II. INDIVIDUAL WORDS AND SYNTAX A. INDIVIDUAL WORDS 1. Time. — The ways of fixing the center of reference for the actions of Alexander and his associates are very different. The fourscore occurrences of iv tovt($ in Arrian are not relatively as frequent as those of /xera ravra in Diodorus. One gives the actors in the midst of their activities; the other is retrospective. The plan of Arrian required a very free use of rbrt, for which he often substitutes kv re^rjs occurs (iv. 12. 2, 3; iv. 18. 7; vi. 11. 2) and ev 8e re? tyetfs (vii. 11. 8), while O-rjs is preferred by Diodorus, and except for one instance (Arrian v. 17. 4) he only has waXw, Arrian usually having €(U7raXtj' alone, but with es (iii. 12. 1) and with ds (ii. 11. 6). The general usage with ana is not the same, though both use it singly and repeated as an adverb. With participles it is limited in Diodorus to seven examples of irpaTTopAvois. But the chief difference is in the prepositional usage in Diodorus, a/xa 5k tovtw (33. 6) ; ap.a tovtois (1. 2). Arrian associates it not only with these pronouns but with others of which ol is the most noticeable. It occurs also with Arrian's "Anabasis" and Diodorus 285 common nouns and still more freely with proper names. Diodorus has dfwv with contrasted terms (13. 3), and Arrian has i\loi.s re bfxov koX iroXejuiois (iii. 11. 3; v. 17. 6) and it is repeated (iii. 22. 2; iv. 4. 7). Peculiar to Arrian is its association with r]b-q in nine pas- sages, and in the same number it does not differ from ap.a in em- phasizing the temporal phases of accompaniment (e.g., i. 8. 3; ii. 3. 5; ii. 27. 6). Arrian, giving the record by men who marched with Alexander, freely refers to the morrow, using wrepaia twenty-nine times, and with es ten. We find vvktos and vvnTwp in both works. Compare with these iv pvktL (v. 24. 1 [twice]; vi. 21. 4; vi. 26. 4), and iifia iv vvkt'i (iii. 10. 2). The far greater number of passages in Arrian fixing the time indicates that the ultimate source must have been a direct observer of the events recorded. 2. Place. — Some phases of local expression are as marked as are the temporal. Arrian usually has avrov instead of avTodi but uses the latter (iii. 2. 4; v. 5. 1; Diodorus 69. 6). The examples of ov are limited to ten in Arrian, with scores of occurrences of « while iravrrj and Tavrn are among the words he uses most freely. Diodorus has Kad&Kov nine times, and \ap.irp6s, opaios, Troj'reXijs, iroXvreXijs, and ra\v more freely than Arrian. 4. Nouns. — We shall present only nouns in -180X17, -opop.17, and -Xwpwi. 'E/Cj8oXi7 is found a score of times in Arrian (Diodorus 75. 2; 104. 3). The relative frequency for epfiokri, irpop.j8oXij is about the same, but irpofioXri seems limited to Arrian. On the other hand Diodorus has only six other compounds, not counting wrepfiokii, occurring twenty-four times. Peculiar to him also is irepi(3o\os (50. 3; 52. 3; 71. 5; 85. 3). "E.H.bpoy.i) occurs most freely of the nouns in -opop.17, fourteen times in Arrian. Other forms are hribpop.T] (Arrian i. 20. 5; vii. 21. 2), Karabpofxi] (iv. 1. 3; iv. 16. 6), awbpop.i] (vi. 3. 4; Diodorus 19. 1; 63. 1; 101. 4). Arrian has lyxupwi and «7rixwpioi, but irpo<7xwpioi more often than these two combined. Diodorus uses irX?7wv, and o^erepos the usage of the two is entirely distinct. Arrian has 01 in more than a score of passages, the larger part similar to ap.a ol ayav (i. 15. 3). The plural forms are noticeable (see Boehner, p. 31). These are sometimes reinforced as v re avr&v (iv. 13. 7); Ss avrovs (iv. 17. 6; Diodorus 22. 4). All the examples of on irep and the nine of o^eVepos are in Arrian, who also uses oirep much oftener than Diodorus. Notice Arrian oawvirep (iv. 21. 9) and similar forms (vi. 14. 4; vi. 29. 10; vii. 1. 6). Here also may be mentioned his thirteen examples of « roabvbe. He has tovs p.iv .... tovs 8e a few times, for which in corresponding passages Diodorus uses ovs p.ev .... ovs be. 6. Adjectives. — The different forms of irfij are not evenly dis- tributed between the two works, the most marked contrast being in the use of avtiiras, which is found at least seventy-nine times in Arrian; in Diodorus once (17. 4). However, the latter favors airas, 48 to 15, and has ttols relatively more freely, 148 to 190. An- other favorite of his is iSioj, fifty-two times, seven neuter and most of Arbian's "Anabasis" and Diodorus 287 them with ovvi{5ri. The plural is found eight times; also kclt' Wuxv (112. 3). Arrian uses the latter without a preposition (i. 26. 4); idiq., eight times, and is r& ISia (v. 25. 1). In contrast with Diodorus he has iwintiva a score of times, while naprepos, Kov6raTOs, and 8aos are favorites, the latter being especially noticeable with numerals. Diodorus is inclined to the use of oXos, ir\t)evyeoripois, with irapa. Arrian only uses ap^i, and Arrian's "Anabasis" and Diodorus 289 80 per cent of all the occurrences in personal connections, the larger part of the remainder in temporal expressions or with numerals. Place is indicated, aptcf>l rb itpbv (i. 23. 5; iii. 16. 5; vii. 20. 3). He has avr]Ke/3oX# (ii. 22. 2) ; and there are at least nine- teen other instances of the dative. 6) Associations : Both works have avev and m^XP'> for the most part, in entirely different associations. The Anabasis has kva in a few places with local connections, and avb. Kparos (iv. 17. 1), while Diodorus has avb. p&aov seven times. Arrian shows the greater freedom with 6.vrl (34 to 4) because it is used to give the details of changes in petty rulers, matters which are not usually mentioned by Diodorus, who has cbr6 to designate temporal relations (71. 1), biro Kvpov, other examples taking the form airb tovtuv yev6p.evos (73. 1 ; 81. 1; 95. 3; 106. 2). Time is expressed by e/c with a few nouns (Arrian vii. 14. 4, Ik 7rai56s); but generally with a pronoun, and iroXkov, a dozen times (Diodorus 99. 5), and Ik iraXeuoO, seven. With other adjectives and adverbs the usage is fairly free (see Boehner, p. 41). The larger part of the examples of iv are with tovtu (81), 290 R. B. Steele t<3 Tore (48), vvkt'l (11), and r\pipais (14). The first three are con- fined to the Anabasis, while the last is as freely used by Diodorus. The latter inclines to iv fiaxy (the plural in Arrian ii. 7. 7), the former to iv xtpa' 1 (Diodorus 56. 1). There are also occurrences with adverbs, and in this respect iv resembles e«, but the associations are for the most part different. In temporal expressions, Diodorus favors Kara, e.g., 101. 3, /card rov i^rjs irorov; and, excepting in a few passages with numerals and 38. 2, vwep r-qv r/'kudav, has virip only with the genitive — about twenty examples. In the Anabasis are the same number with numerals, and a small number with local associations. For virip = de, see Boehner (p. 46). The use of i-irl with the accusative is about the same; but Arrian has it more freely with the genitive and much more so with the dative. Of most interest are cos iiri, iiri fieya, and iiri iroXv. In Diodorus are more than a dozen examples with rinepas, and with this compare Arrian iv. 13. 5. Other temporal phrases are few, e.g., core iiri fiovkvrov (ii. 3. 3). Among the nouns in the dative we find fjfjiipq, (Arrian ii. 26. 4; iii. 4. 2; iii. 21. 6) and reXevrfi (i. 10. 5); while Diodorus has tovtois (8. 2; 28. 2; 84. 1). The free use of ap4>i in the Anabasis left little room for irepi in per- sonal connections, though with the accusative it gives both time and place, and also occurs with numerals. Most of the examples with the genitive in Diodorus express personal relations; in the Anabasis the nouns are mostly abstract, so that the proportion of the two classes differs widely. The dative seems limited to Arrian i. 22. 6; i. 27. 1; vi. 18. 2; vii. 24. 2, irepi t<2 dpovu. The relative frequency for irpos with all the cases is about the same, but there is not an even distribution. Diodorus avoids it with the genitive, but has it with the accusative more than twice as often as Arrian, and with the dative, usually tovtois, only half as frequently. A few passages (as Diodorus 40. 1 and 113. 1) give temporal relations, and local (25. 5). This usage is quite marked in the Anabasis: iii. 8. 7 and iv. 20. 1, irpos 'laaqi, for the usual iv 'laacS. Noticeable are nine examples of irpos eu (e.g., v. 3. 3; v. 5. 4; cf. 5. 6. 3). If we take into the count both prepositions and compound verbs there is no difference between Arrian and Diodorus, as the former has 29. 33 per page (Roos's ed.) and the latter 29. 31 (Fisher's Arrian's "Anabasis" and Diodorus 291 ed.). But the two elements are not equally distributed, the numbers for Arrian being 15. 24 and 14. 09, and for Diodorus 12. 69 and 16. 62. One emphasizes the preposition, the other the compound verb. Prepositions with the infinitive occur most freely in Diodorus, and Sid most often; in Arrian eir and iiri. Diodorus seems the more inclined to the use of inlocked preposi- tional phrases. This usage is not limited to a few prepositions for most of them occur in phrases expressing an attributive relation to some other phrase. A single instance from each will be enough for illustration: cbro rrjs ev 'IcnnS pax 7 )* (Diodorus 39. 1); cbro rod ewl dartpa rijs ToXews vovtikov (Arrian ii. 21. 8). A common rhetorical feature is the same preposition in suc- cessive phrases, e.g., Sid to SokcIv Sid tovtow (Diodorus 53. 1). See also with iiri (18. 4; 82. 1); with Kara (103. 7). The Anabasis has irapa and Trpos most freely (i. 19. 1; i. 25. 4, 9; ii. 2. 1; iii. 23. 8; iv. 1. 1; iv. 5. 1; v. 20. 6; i. 25. 4; ii. 17. 2; v. 26. 3). Arrian and Diodorus vary greatly in their use of participles, the average for one being 15. 8 per page; for the other, 10. 7. Examples of apas are noticeable in the Anabasis, and of xpt^ews in Diodorus, who has the genitive absolute about four and one-half times as often as Arrian. The usage with individual verbs also shows some clearly marked tendencies. One feature peculiar to Diodorus is virrjpxe instead of rji> (e.g., 7. 2; 34. 5; 71. 7; 90. 1). He is also more inclined to use elvai and its compounds with the infinitive (13. 6; 60. 4; 82. 8; 25. 4; 34. 1; 38. 6; Arrian v. 18. 5; vi. 27. 5). There are clear indications of selection in the case of a few verbs. BouXojuat, usually a participle, is the regular word for Diodorus, but e0e\w for Arrian, though he has some instances of the former (ii. 14. 4; v. 2. 3; v. 27. 1; v. 27. 7 [twice]; vi. 14. 3; vii. 10. 5), all in the presumed remarks of the actors. For the Anabasis Kakovpevos is regular (awewi-, vi. 3. 1). Diodorus has €7ri- in 20. 7, and iy- in 113. 1, though he prefers dvona^op.evos. Both have finite forms of the two verbs (see Boehner, Arriana). 1 There are clearly marked distinctions in the use of verbs of motion. T. Hultzsch 2 calls attention to the fact that forms of 1 Ada Seminarii Phil. Erlangensis, II. 501 ff. 2 De Eloculione Diodori Siculi. De Usu Aoristi et Imper/ecti, Pars I (1893), p. 22. 292 R. B. Steele ievai are rarely found in Diodorus. Arrian freely uses the aorist of the compounds in awo- and tin,-, and the imperfect of dra-, tin-, and juera-. There are scores of instances of ekavvu and of its compounds, but the occurrences of these and of a^Uero are few in Diodorus. The compounds of x w P* w ar e not equally stressed. That in irpo- is frequent in Arrian, while irpos- occurs less frequently; in Diodorus (tvv- is most freely used. He has wpocrayu oftener than irpo-, while Arrian prefers «ri-, though he has irpos- occasionally. A few verbs seem limited to Diodorus, as iqyyuje and ovviiyyure (41. 4; 49. 6), havvu and naravvo) (49. 6). He has most of the occurrences of aOpo'ifa and of ava^evyvvo); see in 31. 2, dpafei^as. Hultzsch (p. 84) states that forms of this verb in -jut do not occur in Diodorus (see Arrian i. 5. 1; and cf. i. 26. 5; ii. 5. 8; iii. 4. 5). There are similar differences in the use of /3dXXw, b'exop.at, 8l8ufu, Xa/i/3d»'w, irep.iru, oreXXco, as well as of a number of others occurring less frequently. A detailed statement will be given for verbs of dying. The regular one for Diodorus is reXeurdw; for Arrian, diroflj'jJo-Kco. Notice the occur- rences in corresponding passages in Diodorus (21. 6; 36. 6) and Arrian (1. 16. 6; ii. 11. 8). Though the same words are used the forms may differ, as in statements of burial rerekevT^KOTas (Diodorus 89. 3) ; TekevrqaavTas (Arrian v. 24. 6; Diodorus 14. 1; and 40, 1); but compare awodavovras (i. 23. 6). The same difference can be seen for eirecrov and aired avov in Diodorus (34. 5; 89. 1, 3) and in Arrian (i. 16. 4; v. 18. 2, 3). There is also a preference shown in the use of two other kindred verbs, for edXw is as noticeable in Arrian as is aveTXe in Diodorus. B. SYNTAX 1. Accusative of extent. — Diodorus has s with the infinitive, much less freely with finite forms of the verb, is the predominant particle, and that, numerically considered, -Iva and oVcos are not important. "I^a is commonly used to express local relations, but as a final particle it occurs in only four passages (see Diodorus 4. 8; 77. 7). Arrian has oVcos most freely with the indicative, as 6Vcos el%ov (i. 27. 7), and in the phrase awe/ify oVcos eluacre. It occurs with av (iv. 22. 6; vii. 15. 2). The ten final clauses are all affirmative, excepting in vi. 4. 3. Both affirmative and negative clauses occur oftener in Diodorus, the most noticeable example being oVcos jui) 5i>ecoercu kvkKovv (57. 5), which corresponds to similar statements in Arrian with wj (i. 4. 4; ii. 8. 4). The negative is used alone in ii. 3. 7; vi. 13. 1; vii. 13. 3. Arrian usually has were with finite forms of the verb (see Boehner, p. 56), while it was noticed in Diodorus seven times with the in- finitive, and with a negative 8. 4. 5. Temporal statements. — At the head of the list of temporal particles in the Anabasis is cbs with scores and scores of examples, occasionally with a correlative. Only twenty-seven were noticed in Diodorus, e. g., cos 5' riyyurav (11. 3). 'Eirav is limited to Arrian (iii. 3. 4; vi. 5. 7), and though he has more examples of eirei with causal force, Diodorus has the particle oftener. He uses lirabi] occasionally (Arrian 27), and tirtibav (106. 3), which occurs fifteen times in the Anabasis. Neither uses ore or orav freely, Arrian having Tore .... ore (ii. 1. 4; iv. 15. 1). Compare with these dirore .... rore (vii. 18. 6). He also has twenty-five other examples of dirore. There is a free use of core, twenty-three in- stances with finite forms of the verb, and of these iii. 20. 4 gives local relations. With the infinitive it occurs five times, as in iv. 7. 1. The meaning is "as long as" or "until," and with the latter meaning nine passages have av, and two (ii. 23. 3; iv. 30. 3) do not. 1 Dr. H. A. Short, AJP, VII, 167, n. 1. 294 R. B. Steele There are fifty-two examples of fare iiri with either temporal or local force (see Boehner, Arriana, p. 504). Neither re'ws nor ecos is freely used (the latter in ii. 8. 2; iv. 27. 2); without a verb, ews harbv koX tlKoai to fiaBos (i. 6. 1). Diodorus has it in 62. 1 and 92. 3; and with av in 7. 7; 26. 1; and 28. 2. Arrian has irplv as a preposition, irplv ipaovs (iii. 18. 6), in thirty passages with an infinitive, and a few times with finite forms of the verb, irplv Aapeiov re ireevy6ra yadovTO kclI irplv airoppayrjvcu (ii. 11. 2; ii. 11. 7; vi. 13. 2; vii. 14. 3; vii. 22. 1); and with the subjunctive /xi)-. . . . irplv .... kclt'i- Swctiv (v. 16. 3). The elements of expression in the works are widely different, and when these are wrought into connected narrative the results are not akin. There is a unity in the style of Arrian, and also, though a different one, in that of Diodorus, and the main features that we have considered are the same in the remaining parts of the works of both. Occasionally there is a similarity or equivalence of diction in the statement of some important point, while the context is entirely different. Arrian, TaOra 5e foairpal-aiJLevos iiravrjXOev els Maneb'ovlav (i. 11. 1) equals Diodorus, Mera 5£ raOra 6 piv ftacriXevs iiraveKduv . ... els rr/v Ma.Ktb'ovLav .... (16. 1), but the latter goes on to tell of a council of the leaders and of contests, at Dios to Zeus and the Muses, which Archelaus had established. Arrian says that they were at Aegae, and brings in the reference to the Muses under \eyovcrt.. Sia/Joum tov iropov (Arrian iii. 7. 5) is equivalent to rbv iropov .... paJdwv Sie0tj3eure ti\v bvvap.iv (Diodorus 55. 3), although the details in the entire chapter are not in harmony with those given by Arrian of the movements of Alexander from the Euphrates to the crossing of the Tigris. Interpretations of the main fact may widely differ. Arrian (i. 20. 1) says that Alexander had in mind to dismiss the fleet because of lack of money, and because the Persian fleet was not strong. Diodorus says that the design was to cut off hope of retreat, as Alexander had done at the Granicus, and as Agathocles did. Memnon advised, according to Arrian i. 12. 9, not to run any risk (p,r) Sid tavdvvov livat.) but to harry the land. The opposition expressed by Arsites pre- vailed, because the Persians were suspicious of him. Diodorus (18. 2) has the equivalent (/m? foa.Ki.vb'vvtbtiv), but Memnon further Arrian's "Anabasis" and Diodorus 295 advises to carry the war into Europe, and these suggestions were rejected because unworthy of the magnanimity of the Persians. As we have stated 77677 is freely used by Arrian, and occurs in i. 20. 3, indicating the previous appointment of Memnon, and this fact is stated again (ii. 1.1). Diodorus (23. 6; 29. 1) speaks of it as an act contemporaneous with the other events mentioned. Similar to this is Arrian (iii. 18. 10), where it is stated that Alexander found the Araxes already bridged, while Diodorus has Toe 'kpa^rjv Tora- jude feu£as bufiifiaxrt tovs crpaTt&Tas (69. 2). The data presented show two types of literary expression dis- tinct in many particulars. If derived from a common source, for successive stages of the history, they have been so transformed that the products are no longer similar. At the same time, there are some points of resemblance, the result of the statement of events common to every history of Alexander. These need no unity of source for their explanation while the diversity of expression is against rather than for such a theory. Vandeebilt University