SOURCE: “Arrian's Tactica,” in Aufstieg und Niedergang der röminschen Welt, edited by Wolfgang Hasse and Hildegard Temporini, Walter de Gruyter, 1993, pp. 312-37.

[In the following essay, Devine examines Arrian's treatise on the military tactics of the Roman army, Tactica, discussing the content, originality, and textual history of the work.]

I. L. FLAVIUS ARRIANUS: THE AUTHOR AND HIS CAREER

Of all the ancient tactical authors, the career and personal history of Arrian is by far the best known, even though much of his early military career has to be conjectured from geographical or historical allusions in his surviving works. Lucius Flavius Arrianus was born and educated at Nicomedia (Izmit), the capital of the province of Bithynia in north-western Asia Minor. Although the date of his birth is not attested, the fact that he held the consulship around 130 implies that he was born between A. D. 85 and 90, the normal age for the supreme magistracy in this period being around forty-two.¹ His family had apparently acquired Roman citizenship through the patronage of L. Flavius, an adherent of M. Antonius the Triumvir and himself suffect consul of 33 b. c. (Cass. Dio 49.44.3).² Around A. D. 108, Arrian attended the lectures of the celebrated Stoic philosopher, Epictetus, at Nicopolis in Epirus. Our author may have served in the Roman army as an equestrian praefectus cohortis or tribunus angusticlavus in legio VII Claudia during Trajan's Dacian Wars. Although he reported the emperor's Eastern Wars in his ‘Parthica’, there is no unimpeachable evidence that he saw active service in the Armenian and Mesopotamian campaigns of 114-17. However, he does seem to have made the acquaintance of the future emperor Hadrian before 112/13. In the Periplus, the literary account of his tour of inspection in the Black Sea (131/32), Arrian addresses the emperor in familiar, even intimate, terms (2.4). Arrian seems to have been adlected to the senate inter praetorios soon after Hadrian's accession, and then to have served as legatus legionis in a Danubian province during the early 120s. He was proconsul of the unarmed public province of Baetica in Spain around 125, before becoming suffect consul in 129 or 130, the first Bithynian to hold the consulship. From 130/31 to 137/38, he served as governor of the key frontier province of Cappadocia in eastern Asia Minor, repulsing an invasion by the nomadic Alani and driving them northward through the Caucasus in 135 (Cass. Dio 69.15.1). No official recognition for this campaign is attested, and it is possible that no recognition was merited, since the battle described in Arrian's Ectaxis contra Alanos (hereafter Ectaxis) is a tactical plan, rather than the account of an actual battle. And, in fact, no major battle between the Alani and Arrian's army can be proved.³ In 136 Arrian published his principal tactical treatise, … Tactica), his own revision and adaptation of a pre-existing Hellenistic tactical manual. Around 137/38 he retired to Athens, where he received honorary citizenship and was eponymous archon in 145/46, dying probably in the 160s.⁴

Arrian's best-known work, the Anabasis Alexandri (composed after 115 but before 125),⁵ is the fullest and most reliable surviving source for the reign and campaigns of Alexander the Great, beginning with those of 335 B. C. It is based on the Alexander-histories of Ptolemy, one of Alexander's most prominent generals and later king of Egypt, and Aristobulus, a Greek technical expert of some kind in Alexander's entourage. Arrian's account of Alexander is supplemented by the Indica, published at about the same time as the Anabasis and expanding on the details of Alexander's invasion of India as given in book V of the larger work.⁶

Although modern scholarship knows Arrian exclusively as a historian, in his own day his reputation was primarily that of a philosopher. This apparent paradox is due in part to the much broader conception, in the Hellenistic and Roman imperial period, of what a philosopher was and did. As also in the case of Aelian … writing on tactics and other technical topics counted as philosophy. Aristotle's writings on marine biology and Theophrastus' botanical treatise, to name only two instances, were certainly viewed in antiquity as philosophy. Indeed, two contemporary inscriptions are explicit in describing Arrian as a philosopher.⁷ Some of his reputation as such derived from his publication of his own summaries of Epictetus' lectures, but fragments of an Arrianic treatise on meteorology, a standard ‘philosophical’ topic, survive in Stobaeus. Arrian's strictly philosophical writings, if they ever existed, were lost by the time of Photius, the great Byzantine bibliographer who was twice Patriarch of Constantinople (858-867 and 878-886) and who excerpted many subsequently lost Arrianic works (cf. Photius, Bibl. cod. 58.5).

Linked with Arrian's philosophical ambitions was his conscious imitation of Xenophon, who in late antiquity was likewise more famous as a philosopher than a historian.⁸ Indeed, in two of his late works, the Cynegeticus and the Ectaxis (ca. 135/36), Arrian goes so far as to call himself Xenophon. The Byzantine biographical tradition, stemming from Heliconius of Byzantium, maintained that Arrian was actually called the New Xenophon in his lifetime. Works of the period allude to Xenophon the Elder, and a contrast with a younger aspirant to the name is implied. The device of usurped nomenclature is not, however, present in the Anabasis Alexandri, which title, evocative as it is of the Elder Xenophon's best-known work, is first attested by Stephanus of Byzantium and need not be Arrian's own: a second Xenophon would naturally be credited with a second ‘Anabasis’. Although Xenophon is undoubtedly a source of inspiration for the Alexander-history, there is no indication there that Arrian is casting himself in the role of the new Xenophon to the new Cyrus and his march up-country. The emulation is, however, explicit in the Periplus and the Cynegeticus, not to mention the Ectaxis, and probably only came quite late in Arrian's public life. Then, as the authoritative biographer of Epictetus, and himself a noted Stoic philosopher, prolific historian and geographer, successful military commander, and huntsman, Arrian convincingly resembled the real Xenophon. His close relationship to Hadrian could be seen to parallel that of Xenophon to Agesilaus III the Great of Sparta. The adoption of the Xenophontean name thus came late, and the nomenclature of Neos Xenophon may well have been conferred on Arrian as a title of honour, rather like those of the New Homer and the New Themistocles as conferred by the Athenians on C. Julius Nicanor.⁹

II. THE DATE AND OCCASION OF ARRIAN'S TACTICA

Unfortunately, the beginning of the Tactica was lost when a folio (181a) was torn out of the Codex Laurentianus gr. 55.4, the prototype of the entire manuscript tradition. Stadter¹⁰ argues convincingly that a total equivalent of 35 Teubner lines has thus disappeared, of which 26 would have constituted the formal preface. The size of the missing portion (at least 200 words) makes it almost certain that the lost preface would, by analogy with the extant one of Aelian, have contained a statement of Arrian's qualifications for writing such a treatise, a description of the intended audience, and a formal dedication of the work to the emperor.¹¹

The publication of the Tactica can, nevertheless, be precisely dated from the reference to Hadrian's vicennalia at Tact. 44.3, where the work's concluding lines and the verses of Terpander (frg. 6 Bergk = frg. 4 Diehl) quoted there indicate its ostensible purpose. It was meant as an occasional composition to commemorate the emperor's vicennalia, which, according to a calendar of the imperial cult (P. Oslo 3.77, lines 15-16), seems to have been celebrated on 13 December, A. D. 136. Hadrian's vicennalia, the first achieved since that of Tiberius in A. D. 34 (and, incidentally, the last until that of Diocletian in 304), was a highly apposite occasion for glorifying the emperor and his regime, and Arrian embellishes the latter part of his work with glowing references to Hadrianic policy and the themes of the reign.

While in the first part (1-32.2) of the Tactica Arrian gives a detailed but conventional account of the minor tactics and drill of the Hellenistic armies, in the second part (32.3-44.3) he produces, in overt compliment to Hadrian, a colourful, first-hand description of the parade-ground exercises of the Roman cavalry of his own day. This affords an opportunity, at Tact. 44.1-2, to list and praise Hadrian's cavalry reforms, while carefully noting their consistency with the mos maiorum. It may be noted, nonetheless, that the cavalry exercises described at Tact. 34-43 actually predate Hadrian.

At the time of composition Arrian was serving as legatus Augusti pro praetore in Cappadocia, and his frequent allusions to the Sarmatians and the Alani seem to be an attempt to draw attention to his own Transcaucasian campaign of 135. Moreover, at Tact. 42.1 our author portrays himself as “the good commander” conscientiously following the emperor's drill instructions. Arrian's governorship of Cappadocia appears to have ended in 137 or at least before Hadrian's death in the following year, and the underlying motive for the Tactica would thus have been one of inducing the emperor to continue the author in his command there or in some other armed province.¹²

III. THE FORM AND CONTENT OF ARRIAN'S TACTICA

1. THE HELLENISTIC ELEMENTS

Arrian's ‘Τεχνη τακτικη’ is, after Aelian's ‘Τακτικη θεωρία’ and Asclepiodotus' ‘Τεχνη τακτικη’, the fullest and most important of the surviving Hellenistic tactical manuals, the study of which is effecting a revolution in our understanding of the military techniques of both the Hellenistic states and the Romans, as well as even those of the Persians and the Carthaginians.

Comparison of Arrian's Tactica with the cognate tactical manuals of Asclepiodotus (first century B. C.) and Aelian (published ca. A. D. 106-113) is instructive.¹³ Of the three treatises, that of Aelian is the longest at 10,913 words (in the obsolete edition of the ‘Short Recension’ made by Köchly and Rüstow¹⁴), and that of Asclepiodotus, at 7,002 words (in the edition prepared by W. A. Oldfather for the Loeb Classical Library), is the shortest. Nonetheless, while the text of Arrian's ‘Tactica’ (in the Teubner edition of Roos and Wirth) runs to 9,460 words, of this approximately 2,600 words (327 lines in the Teubner edition) belong to the second, exclusively Roman, part of Arrian's treatise. This fact actually makes Arrian's Hellenistic section the shortest of the three Hellenistic tactical manuals. It should be noted, however, that the issue is complicated by the occurrence of a lacuna of indeterminate, though obviously not of very great, length (represented by a space of seven or eight letters in the prototypical Codex Laurentianus gr. 55.4) at Tact. 32.3 (in the bridge passage between the Roman and Hellenistic sections of the work). Moreover, part of section 1.1 (= 1.1K and part of 1.2K) is lost, and there are further lacunae at 10.1 (= 9.2K), 10.3 (i. e., at the beginning of 9.4K) and 29.2 (= 37.2K and 37.3K), and also possibly at 20.3 (= 24.3K) and 24.3 (= 28.3K).

A comparison of the beginning of Aelian's treatise with what remains of the opening of Arrian's reveals, despite some notable points of difference, a remarkable similarity of approach. In the Prooimion and the first section of his main text, Aelian explains his reasons for composing the ‘Τακτικη θεωρια’, freely admitting that he has no personal experience of warfare and no real knowledge of Roman military practice (praef. 1-2). Instead, he writes, like most of his predecessors, as a philosopher (and a perpetuator of the Stoic tradition, as we will see), expressing the hope that his efforts will find favour as “a Greek theoretical work and a polished dissertation” (praef. 6). In line with contemporary literary convention, Aelian justifies his undertaking by stressing his own erudition and ability, which transcends that of earlier writers on tactics (praef. 1 and 4-5). Arrian voices analogous sentiments in his “delayed introduction” to the Anabasis Alexandri (1.12.4-5), and their appearance in the lost introduction to the Tactica is a high probability. An authority on the Roman art of war through long personal experience (in contrast to Aelian), and a successful military commander like his model Xenophon, Arrian likewise saw himself writing primarily as a philosopher.¹⁵ And tactics, especially the tactics of a bygone age, was an appropriate subject for abstract discussion by a philosopher, which is the only interpretation that the unoriginal Hellenistic part of Arrian's treatise will bear. Both Aelian and Arrian complain that earlier tactical authors have written for a readership already familiar with tactical matters and have failed to explain their terminology (Tact. 1.2 = 1.3K; cf. Aelian 1.3). Arrian's express intention, like that of Aelian, is to combat the obscurantism and lack of clarity of these earlier tactical writers (Tact. 1.2-3 = 1.3-4K; cf. Aelian 1.4, 1.7). But Arrian (Tact. 1.4 = 1.3-4K) goes so far as to claim that he is the first to use his knowledge to correct the sheer obscureness of these antecedent authors. The same claim made implicitly by Aelian (1.4) suggests that both are in fact appropriating and adapting a claim from a common source—Poseidonius, probably, or perhaps even Polybius.

Further, Aelian considers the table of contents (the 113 ‘Headings of the Book’) which he prefixes to the ‘Τακτικη θεωρια’ a major contribution to the genre (praef. 7). Arrian's Tactica, like those of both Asclepiodotus and Aelian, to begin with probably had an index of section-headings prefixed to it. Arrian's failure to refer to his subsequently suggests that it was a somewhat more rudimentary affair than Aelian's, much more like the twelve headings with which the text of Asclepiodotus opens. Like his misguided and spurious claim of originality, it is certainly evidence that Arrian was unacquainted with Aelian's treatise and its superior model index.

Again, as autopsy of the Codex Laurentianus gr. 55.4 shows, Arrian also diverges from the mainstream tradition represented by Aelian and Asclepiodotus by failing to illustrate his treatise with diagrams—a deficiency which renders his text more difficult to follow—especially when he is describing such unusual formations as the cavalry εμβολοs (Tact. 16.6-8 = 18.4K) and the Thessalian ρομβοs (Tact. 16.3-5 [= 18.1-3K] and 17.1-3 [= 19.1-2K and 19.5K]). As, apparently, in the case of the prefixed index or table of contents, Arrian's failure to imitate the more comprehensive and lucid approach of Aelian and Asclepiodotus argues for his lack of familiarity with their rival productions.

Finally, it is in the addition of a major appendix, of a non-Hellenistic character, that Arrian diverges most radically from the rest of the tradition. But while the second part of the Tactica (32.2-44.3) is devoted to a tendentious account of the parade-ground exercises of contemporary Roman auxiliary cavalry (discussed below), the first part (1-32.2), as per convention, describes in detailed but generalized terms the minor tactics, drill, arms and equipment, tactical organization, and commands of the Hellenistic armies (rather than of the Macedonian army of Philip II and Alexander the Great).

And, indeed, as a section by section comparison of the three treatises reveals, the order of contents adopted by Arrian in the first part of the Tactica follows the pre-existing convention very closely. This order is as follows:¹⁶

I. Introduction (Tact. 1.1-3; cf. Aelian, praef. and 1.1-3).

II. Basic divisions of warfare (Tact. 2; cf. Aelian 2.1-5D [= 2.1-6K]; Asclep. 1.1).

III. Description of troops in terms of the equipment used (Tact. 3-4; cf. Aelian 2.6-11D [= 2.7-13K]; Asclep. 1.2-3).

1. Infantry (Tact. 3; cf. Aelian 2.6-8D [= 2.7-9K]; Asclep. 1.2).

2. Cavalry (Tact. 4; cf. Aelian 2.9-11D [= 2.11-13K]; Asclep. 1.3).

IV. Organization of the army (Tact. 5-10; cf. Aelian 3-10; Asclep. 1.4-3.4).

1. The need for organization (Tact.5.1-3; cf. Aelian 3; Asclep. 1.4).

2. The λοχοs (Tact.5.4-6.6; cf. Aelian 4-5; Asclep. 2.1-3).

3. Combinations of λοχοι (Tact. 7-8; cf. Aelian 6-7.3; Asclep. 2.4-5).

4. Light-armed troops and cavalry (Tact. 9.1-2; cf. Aelian 7.4-6).

5. The ideal number for the phalanx (Tact. 9.3-6; cf. Aelian 8; Asclep. 2.7).

6. The units of the army (Tact. 10; cf. Aelian 9; Asclep. 2.8-10).

V. Formations and the use of troops in battle (Tact. 11-19; cf. Aelian 11-23; Asclep. 3.5-9).

1. The phalanx (Tact. 11-12; cf. Aelian 11.1-4, 13-14; Asclep. 4.2, 5.1, 3.5-6).

a. Formations (Tact.11; cf. Aelian 11.1-4; Asclep. 4.2).

b. Arms and equipment (omitted; cf. Aelian 12; Asclep. 5.1).

c. The different ranks (Tact. 12.1-5, 12.10-11; cf. Aelian 13-14; Asclep. 3.5-6).

d. The spacing of the troops and the use of the sarissa (Tact. 12.6-10; cf. Aelian 14.2-6; Asclep. 5.1-2).

2. Light-armed troops (Tact. 13-14; cf. Aelian 15-16; Asclep. 6).

3. Archers and javelin-men (Tact. 15; cf. Aelian 17).

4. Cavalry (Tact. 16-18; cf. Aelian 18-20; Asclep. 7).

5. Chariots and elephants (Tact. 19; cf. Aelian 22-23; Asclep. 8-9).

VI. Movements (Tact. 20-27; cf. Aelian 24-35; Asclep. 10.1-21, 12.10).

1. Their names (Tact. 20; cf. Aelian 24; Asclep. 10.1).

2. Description of movements (Tact. 21-25; cf. Aelian 25-29; Asclep. 10.2-20).

3. Formations taken up by the phalanx (Tact. 26, cf. Aelian 30-31; Asclep. 10.21).

4. Means of relaying commands to the army (Tact. 27; cf. Aelian 35; Asclep. 12.10).

VII. Marches (Tact. 28-30; cf. Aelian 36-39; Asclep. 11.1-6, 11.8).

1. Order of the army (Tact. 28; cf. Aelian 36; Asclep. 11.1-2).

2. Formations on the march (Tact. 29; cf. Aelian 37-38; Asclep. 11.3-6).

3. The baggage-train (Tact. 30; cf. Aelian 39; Asclep. 11.8).

VIII. The manner of issuing commands (Tact. 31-32.2; cf. Aelian 40-42; Asclep. 12.11).

Detailed comparison with Aelian also reveals what Arrian omits: Aelian's more extended diatribe against previous tactical writers, and his declaration that he will illustrate his text with diagrams (1.4-6); his claim that tactics is the most useful of all the sciences, and his citation, in support of this contention, of Plato, Laws 626a to the effect that “all cities by their very nature wage undeclared war against all other cities” (1.7); Aelian's inclusion of fighting on rivers in the category of naval warfare, and his (unfulfilled) promise to discuss naval tactics later (2.1); the definitions of tactics advanced by Aeneas (“the science of military movements”) and Polybius (“whenever anyone takes an unorganized crowd, organizes it, divides it into files and, grouping these together, gives them a practical training for war”) (3.4); the information that the last man in a file is called the ονραγοs (5.1 ad fin.), and that the commander of a τελοs is called a τελαρχηs (9.7 ad fin.); the description of the arrangement within the phalanx of the phalangarchies, merarchies, tetrarchies, and syntagmata in terms of the proportionality of the fighting qualities of their commanders (10.1-4); the information that πεκνωsιs permits the individual soldiers to face about (11.3), while sυναοπιομοs permits neither withdrawals nor individual turns to right or left (11.4); the description of the standard-issue Macedonian shield and—with the present author's emendation¹⁷—the ten cubit (= 15 foot) sarissa (12; cf. Asclep. 5.1); the information that the men in the sixth and subsequent ranks of the phalanx do not extend their sarissai beyond the bodies of the men in the front rank—instead Arrian claims that the phalangites of the sixth rank extend their sarissai two feet beyond the front rank and modifies all the dependent statistics in the tradition accordingly (14.4-6, contradicting also Asclep 5.1-2); the information that some commanders have the troops posted in the rear equipped with longer spears than those in front, so that the weapon-heads of those stationed as far back as the third or fourth rank will project just as far (14.7); the observation that it is a great source of strength to a unit to have a senior commander posted not just in front but also in the rear (14.9); the file-closer … from the list of supernumerary men attached to each hekatontarchy (16.2, contradicting also Asclep. 6.3); the observations that whenever the number of the horses in the length (i. e. front) of a cavalry formation is equal to that of those in the depth, a square of number will be achieved but the shape of the formation will be an oblong rectangle with the depth greater than the length, and that whenever the shape of the squadron is square, the number of horsemen in rank will differ from that of those in file (18.9); the explanation of why the horses on the left and right sides and rear of the rhomboid-formation have to be stationed at a distance from each other (19.2); the observation that some authorities draw up the cavalry in a rhomboid-formation by both rank and file, others by neither rank nor file, some by file but not by rank, some by rank but not by file (19.3); the information that in drawing up a ρομβοs by rank and file, the middle rank in the formation is assigned an odd number of horsemen, and that this number is diminished by two per rank as the formation narrows down towards both the front and the rear (19.4); the manner of drawing up the ρομβοs by neither rank nor file, to make wheeling and riding through easier (19.6); the details of the positioning of the ilarch and the remaining horsemen in this way of drawing up the ρομβοs (19.7); the position of the zygarch or rank-commander (19.8); the formation of the rearward half of the ρομβοs (19.9); Polybius' ideal 64-man ile or squadron (19.10); the formation of the ρομβοs by file (19.11); the drill for facing about “spearward” and “reinward” by the horsemen in the ρομβοs (19.12); the drawing up of the squadron by rank (19.13); [Aelian's reference to one of his own illustrative diagrams (20.1)]; the exhortation to try out in the daily drills all the cavalry formations described, in order to determine which are the most useful and appropriate for actual fighting (21.1-3); the list of chariot-corps units—the zygarchy, syzygia, episyzygia, harmatarchy, keras, and phalanx (22.2); the division of chariots into “light-armed” and scythe-bearing (22.3; but cf. Tact. 2.5 = 2.6K); the nomenclature for the various elephant-corps unit-commanders and units—zoarch, therarch and therarchy, epitherarch and epitherarchy, ilarch and ilarchy, elephantarch and elephantarchy, keratarch or merarch and keratarchy, phalangarch and phalanx (23); the injunction that commands should be given in terms of fixed terminology, introducing the catalogue of terms for the various infantry manœuvres (24.1); explanation of the terms … “insertion”, … “supporting-position”, and … “forward position” (24.3); the remark that the tactical writers differ from one other as regards this terminology (24.4); the use of the movement called κλιοιs to repel flank-attacks and for making counterattacks (25.1); the definition of the kinds of … “about-face”, “spearward” or “away from the enemy” and “shieldward” or “towards the enemy” (25.2-4); the promise to explain how the … “quarter-turn” is used (25.6); the qualification that the οηνταγμα in making an επιοτροpη resumes close-order on its completion (25.7); the information that being drawn up by rank involves the individual soldier keeping himself equidistant from his comrades in both file and rank (26.1); the description of countermarches by row …, that is, the exchange of positions between the right and left within the row (28.4); the doubling of place occupied by a phalanx by depth, by means of half the men in each file countermarching to the rear (29.7D = 29.9K); the reversal of the countermarch, by means of the recall of those inserted as rear-rank men in the original files to their own separate files (29.8D = 29.10K); the account of the ορθια pαλαγξ (30.2); the statement that in νποταξιs the formation has the shape of a triple gate (31.4); the extended description of the various kinds of “wheel”—the “quarter-turn spearward”, “quarter-turn shieldward”, “half-turn spearward”, “half-turn shieldward”, and the reversal of each of these, the “three-quarter turn, spearward or shieldward” (32.1-9); the “compacting” … of the phalanx—on the right wing, on the left wing, and in the centre, and the reversal of each of these manœuvres (33.1-5); the injunctions that spears be held upright whenever “wheelings” are made in compact order, and that the light-armed infantry should be trained in the same drill (33.6); the observation that the commands for “facing-about”, “quarter-turns”, “half-turns”, “three-quarter turns”, and “resuming first position” are useful for confronting sudden approaches by the enemy, whether they appear on the right or the left, in front or in rear of the army's line of march, and that the same holds good for countermarches (34.1); the information that the Macedonians are said to have invented the Macedonian countermarch and the Spartans the Laconian countermarch (34.2); the claim that Philip II of Macedonia and Alexander the Great both preferred the Laconian countermarch, and only used the Macedonian variety when forced to do so by circumstance (34.3); the explanation that, in the Macedonian countermarch, sudden attacks on the rear tend to produce great disorder, thus increasing the danger of panic and enemy pursuit, since the troops are being marched around from the rear to a position behind the file-leaders—a manœuvre resembling flight, which emboldens the enemy (34.4); the claim that the Laconian countermarch has the opposite effect in the case of an attack on the rear, since the troops advance towards the enemy through their own formation (34.5); the comment that it is sometimes hard to find signals for all purposes, as critical moments present new circumstances to which the troops are unaccustomed, but since it is unlikely that all the various kinds of difficulties will arise simultaneously, distinct orders can still be given either by voice or visual signal (35.5D = 35.6K); the list of march-formations to be described (36.6D = 36.7K); the qualification that in the infantry … wedge-formation it is the file-leaders that end up actually engaging the enemy, and likewise in the … “hollow-wedge” (37.6-7); the definition of … “attenuation” as the reduction of the depth of a phalanx in terms of the number of ranks, the full phalanx-depth of sixteen being given as the example (38.3); the observation that those receiving commands given in haste must be on their guard against ambiguity, lest some do one thing, and others the opposite (40.1); the injunction that, in commands, the specific term precede the general term, to reduce the possibility of ambiguity (40.2, but cf. Tact. 31.3 = Aelian 40.3); the repetition of the injunction to put the specific term before the general (40.4); the quotation of Iliad 3.8-9 and the beginning and end of the preceding quotation of Iliad 4.428-431 (41.1); the last three lines of the quotation of Iliad 2.459-463 (41.2); the quotation of Iliad 3.1-2 (41.3); twenty-six of the listed commands (42.1); and the concluding comments on the principles utilized by the tactician (42.2; cf. Asclep. 12.11 ad fin.).

Some of the items omitted by Arrian, like the purely mathematical observation on the square of the number of horsemen in the front of an oblong cavalry formation, are of course of doubtful value for the understanding of Hellenistic minor tactics, though they are precisely the sort of thing that a ‘philosopher’ would be expected to reflect upon. But much of the omitted material, such as the progression of tactical units and the hierarchy of unit-commanders (notably in the elephant- and chariot-corps) and details regarding phalangite equipment (e. g. the standard-issue Macedonian shield), is of some importance for reconstructing the day-to-day practice and command-structure of the Hellenistic armies.

In view of the fact that Arrian had published a major work of military history, the Anabasis Alexandri (with a length of 80,714 words in the Roos-Wirth Teubner edition), some twelve to twenty-two years prior to the composition of his Tactica, the absence from the latter of any significant reference to, or example or illustrative material drawn from, the career of Alexander the Great or his opponents is remarkable. Were it not an established fact that both works derive from the pen of a single author, arguments based on the difference in content and the lack of any kind of intertextuality between them could be adduced in support of ascriptions to two separate writers.

A notable example of this apparent divergence is to be found in Arrian's failure, in discussing the term in the Tactica to recall his own use of εμβολοs or εμβολον to describe the formation taken up by the Macedonian phalanx at the battle of Pelion during Alexander's Illyrian campaign in 335 b. c. (Anab.1.6.3), a Persian cavalry formation at the battle of the Granicus (Anab. 1.16.7), and the grand tactical formation into which Alexander threw his Companion cavalry and part of his infantry phalanx at the crisis of the battle of Gaugamela (Anab. 3.14.2).¹⁸ 'νεμβολοs, instead, appears in the Tactica as an instantiation of πνκνωοιs: …

“(1) The phalanx is drawn up … depthwise …¹⁹ where a more compact order is required, as when it is necessary to dislodge the enemy by sheer compactness and force—(2) just as when Epaminondas at Leuctra drew up the Thebans themselves, and at Mantineia all the Boeotians, making, as it were, a wedge and leading it against the Spartan formation—or, again, where it is necessary to beat off those making a charge, just as when we have to take up a formation against the Sarmatians and the Scythians. (3) And ‘compacting’ … is the contracting from a more open to a more compact order as regards both rank and file, that is, in both length and depth”

(Tact. 11.1-3 = 11.1-3K).

The parallel passage in the authentic text or “Short Recension” of Aelian (11.3-5), preserved in the Cod. Laurentianus gr. 55.4, is more succinct, and devoid of such tactical examples: …

“(3) It is ‘compact order’ … whenever from a more open order the intervals are reduced so as to contract the formation as regards both rank and file, that is, as regards both length and depth, while still permitting the troops to face about … (5) ‘Compacting’ … is used whenever the general wishes to lead the phalanx against the enemy”.

Arrian' illustration is, however, closely paralleled in the interpolated or ‘Long Recension’ of Aelian (47.3-4 [= Dain L3-L4]):²⁰ …

“(3) Opposed to this [sc. the cavalry square-formation] is the infantry formation called the ‘wedge’ (εμβολοs), which has all of its sides made up of heavy infantry. This type of formation is derived from the cavalry wedge, but whereas in the case of the cavalry wedge one man is enough to lead the attack, three are required in that of the infantry wedge, one not being enough to engage the enemy alone. (4) By this device, Epaminondas the Theban, fighting the Spartans at Leuctra, defeated a very large force by compacting his army into a wedge”.

Arrian's reference to Epaminondas and the infantry εμβολοs indicates that this material, while not appearing in the authentic ‘Short Recension’ of Aelian, does nonetheless belong to the same, authentic Hellenistic tactical manual tradition, i.e. it is not merely an unjustified Byzantine interpolation, but derives ultimately from Poseidonius or even Polybius himself.²¹ Conversely, the interpolated ‘Long Recension’ of Aelian (39.4 [= Dain D4]) mentions the Sarmatians and Alani in a related context, which suggests derivation from Arrian.

Although Arrian's account of the cavalry wedge (Tact.16.6-8 = 18.4K)—from which, according to the interpolated passage of Aelian (47.3) quoted above, the infantry wedge is derived—is virtually identical with that of Aelian (18.4), our author does add a brief amplification: …

“… but the formation which is pointed, even if it advances drawn up in depth, yet, by wheeling with its leading point within a small arc, renders the entire formation easy to manœuvre”

(Tact. 16.8).

Arrian interpellates another, much more generalized, example at Tact. 11.4 (= 11.6K): …

“It is ‘locked-shield order’ … whenever from an existing ‘compact order’ … the formation is contracted still further, so that on account of the closeness of the troops it is not possible to turn the formation in either direction. And it is by means of such a locking of shields that the Romans form the ‘tortoise’ …”.

Although Arrian's account of the testudo is briefer and less informative than that of Cassius Dio (49.30.1-4), it is, after all, adduced only as an instance of ουναοπιομοs. The later reference to the testudo … at Tact. 36.1, as being the name of both the Roman infantry formation produced by synaspismos and an analogous Roman cavalry formation, is clearly intended to provide a link between the Hellenistic and Roman divisions of Arrian's treatise, and to indicate their interdependence.

But this stress on the interdependence of the two parts is somewhat artificial and the connections are a little contrived. This is evident from further instances where Arrian attempts to up-date or Romanize the content of the Hellenistic tactical manual he is adapting. Thus, while Aelian 2.10-11D (= 2.12-13K), for example, defines “spear-bearing cavalry” and “cavalry-skirmishers” as follows: …

“(10) Spear-bearing cavalry … are those which engage the enemy at close quarters and fight hand-to-hand on horseback with spears … (11) Of the cavalry those which fight from a distance with missiles are called skirmishers …”,

Arrian, who had had first-hand experience of the Alani and Armenians as governor of Cappadocia, and also probably of the Parthians earlier during Trajan's Parthian War, expands on Aelian's definitions, doubtless on the basis of autopsy: …

“Spear-bearing cavalry … are those which close with the enemy formation and fight with spears … or push with lances … in their onslaught, as do the Alani and the Sarmatians. Skirmishers … are those which fight from a distance with missiles, like the Armenians and those of the Parthians who are not lance-bearing”

(Tact. 4.3 = 2.12K).

Arrian, moreover, concludes this section on the various types of cavalry by adding information on Roman cavalry arms and equipment: …

“(7) Among the Romans, some cavalry carry lances … and charge in the manner of the Alani and the Sarmatians, while others carry javelins. … (8) A long, flat broadsword is suspended from their shoulders, and they bear flat oblong shields, an iron helmet, a woven corselet, and small greaves. (9) They bear javelins … for both purposes, both to throw from a distance, when required, and to hold in the hand when fighting at close quarters. And, if it is necessary to come together in hand-to-hand fighting, they fight with their broadswords. Some also carry small double-headed axes [or maces] with spikes set all around in a circle”

(Tact. 4.7-9 = 2.14K).

The subject of chariot-fighting, however, brings out the antiquarian in Arrian. In contrast to Aelian (22.2), Arrian omits the names of the various units of the chariot-corps, adding instead, at Tact. 2.5 (= 2.6K), a reference to many-poled … chariots, which he derives from the ‘Cyropaedia’ of his favourite model Xenophon (6.1.51-52, 6.4.2): …

“(4) Chariot-fighting is more diverse: (5) for there are light-armed chariots, such as those used in the Trojan War, or scythe-chariots like those used later in the Persian War, or those with the chariot horses either armoured or unarmoured, or chariots with one pole or two or even many-poled”

(Tact. 2.4-5 = 2.6K).

Roman originality in this regard lies in never having used chariots in battle: …

“(2) For the Romans have not at any time practiced fighting from chariots. Nor did the barbarians in Europe use chariots either, aside from those of the so-called British Isles, beyond the Great Sea. (3) These often used two-horse chariots, with small, poor-quality horses. Their two-man chariots (διpροι) are well adapted for running over all kinds of terrain and their ponies to enduring hardships. (4) Of the Asians, the Persians in ancient times practiced the use of scythe-bearing chariots and armoured horses, beginning in the time of Cyrus, (5) and before this the Greeks with Agamemnon and the Trojans with Priam used chariots with unarmoured horses. The Cyrenians also for some time fought from chariots. (6) But now all these practices have been abandoned, as has the use of elephants for warfare, except by the Indians and the Upper Ethiopians”

(Tact. 19.2-6 = 22.2-6K).

Arrian's excursus on the scythe-chariots of Asia again derives from Xenophon's ‘Cyropaedia’, which likewise credits Cyrus with the introduction of scythe-chariots (6.1.29-30) drawn by armoured houses (6.1.50), and claims that these superseded the unarmoured chariots utilized in the Trojan War (6.1.27) and in contemporary Cyrene (6.1.27). Xenophon recapitulates his account at 6.2.17 and 8.8.24.

Elephants, however, are patient of a more contemporaneous treatment. At 2.2 (= 2.3K), Arrian notes that elephants were used by Indian and Ethiopian armies, and later by the Macedonians and the Carthaginians, and occasionally even by the Romans, adding that some elephants are equipped with fighting-towers (2.4 = 2.5K).

Likewise, at Tact.18.4 (= 20.2K), Arrian, in describing the hierarchy of Hellenistic cavalry-units, adds that the Romans call the equivalent of a Hellenistic hipparchy (a unit of 512 horsemen) an … ala.

A similar attempt at up-dating and Romanizing Hellenistic material appears in Arrian's account of phalangite equipment: …

“For the full heavy infantry equipment, a helmet is also added, or Spartan or Arcadian conical caps …, and greaves, as with the ancient Greeks, or, as with the Romans, a single greave for the leg that is put forward in battle, and breastplates, some covered with plate-mail, some with fine iron chain-mail”

(Tact.3.5 = 2.10K).

The parallel passage in Aelian (2.6D = 2.7K) is more succinct, and omits the mention of Spartan and Arcadian πιλοι; and there is of course no reference to Roman practice.

Arrian's additions are often gratuitous and laboured, as at Tact. 27.4-5 (= 35.3-4K), where, to Aelian's mention of uneven ground as an obstacle to visible tactical signals, he adds a reference to hills. At Tact. 25.9 (= 29.7K), Arrian uses νπερpαλαγγειν instead of νπερκεραοειν (as in the parallel passages in Aelian 29.5D [= 29.7K] and Asclep. 10.18) merely for the sake of stylistic variation, thereby obscuring the original specific sense of the passage. Occasionally, indeed, Arrian's interpellations are totally irrelevant, as at 6.3 (= 5.3K), where he quotes Xenophon (Anab. 4.3.26) verbatim to show that the ενωμοτια contained less than half the number of troops in the λοχοs. In the passage cited, however, Xenophon is in fact speaking of the hundred-man Spartan λοχοs (3.4.21), not the basic sixteen-man unit of the Hellenistic period.

A major difference between Arrian and the other tactical writers of the same tradition—although in this case having no bearing upon our author's tendency towards the Romanization and up-dating of his source-material—concerns the length of the Macedonian sarissa. At Tact. 12.7 (= 14.2K), Arrian gives the length of the sarissa as 16 feet. According to Polybius 18.29.2 and Aelian 14.2, however, the sarissa was 16 cubits (24 ft.) long as regards the original design but in actuality only 14 cubits (21 ft.) long. Polyaenus (Strat. 2.29.2), indeed, credits the Eddessans at Cleonymus of Sparta's siege of their city (ca. 300 b.c.) with the use of sixteen-cubit (24 ft.) sarissai. But Asclepiodotus, on the other hand, says that the sarissa was not more than 12 cubits (18 ft.) and not less than 10 cubits (15 ft.) in length, a statement corroborated by (the emended text of) Aelian 12.²² The contemporary Aristotelian philosopher Theophrastus (370-288/85 b.c.) also claims that the “longest” sarissa was 12 cubits (Hist. Pl. 3.12.2). Since the entire Arrianic passage Tact. 12.7-9 (= 14.2-4K) is punctuated by measurements in feet …, it is probable that the variant is no mere scribal error, but goes back to the author's holograph. …

Similarly, Arrian contradicts Aelian (14.5) and Asclepiodotus (5.1) on the number of sarissai that extend beyond the front rank of the phalanx: …

“Therefore for each of the front-rank-man six sarissai project, in a semicircle, one behind another, so that each soldier is hedged about by six sarissai and pushes forward reinforced with the strength of six men, with which they are able to press heavily on”

(Tact. 12.10 = 14.5K).

To this may be compared the parallel passage in Aelian (14.5): …

“Therefore for each of those stationed in front rank five sarissai project, presenting a forbidding aspect to the enemy, and thus each man is hedged about by five sarissai and pushes forward reinforced with the strength of five, as can easily be seen”.

Since the only significant point of divergence here is a factual one, the remarkable verbal similarity suggests a common source, from which Arrian himself has chosen to differ for reasons which can no longer be reconstructed. Certainly no other surviving work of this tradition provides any support or justification for Arrian's modification.

Arrian's description of the arrangement of the sarissai in the Macedonian phalanx as overlapping “in a semi-circle, one behind another” is problematic. Arrian apparently visualizes the weapon-heads of the sarissai of the first six ranks as forming an arc in profile, with the shafts of the rearmost sarissai being held at 90° to the foremost, which would of course be an extremely difficult arrangement to sustain in a tactical advance. But the choice of the words “in a semi-circle” … is suspect, and the expression may in fact be gratuitously borrowed from the description of the cavalry wedge at Tact.16.6-8 (= 18.4K),²³ where it applies to the perimeter of the formation: …

“… this formation … seems useful because the leaders are posted in a semi-circle …, and the front narrowing down to a point makes it easy to break through every enemy formation, while permitting rapid wheeling and withdrawing movements”

(16.7).

A similar, less ambiguous, use of εν κνκλω to mean “the perimeter” of a formation occurs in Arrian's Romanizing description of the formation of the infantry variant of the testudo:

“(4) [The testudo] often takes the form of a square, or, where required, it advances as an oval … or an oblong. (5) Those troops on the perimeter … of the brick-formation … or drawn up in circular formation … hold their large oblong shields before them, while those deployed alongside them hold theirs over their heads, one shield overlapping the other”

(Tact. 11.4 - 5 = 11.6K).

One valuable contribution that Arrian does make to the Hellenistic tactical tradition is his definition, at Tact. 2.3 (= 2.4K), of the type of cavalry called αμpιπποι, which occurs, for example, in Antigonus Monophthalmus' order-of-battle at the battle of Paraitacene in 317 b.c. (Diodorus 19.29.2):²⁴

“… some cavalry are αμpιπποι. Some cavalrymen use only one horse, but αμpιπποι are those who tie together a pair of unsaddled mounts, so that they can leap from the one to the other”.

It is not surprising to find this obscure passage reproduced virtually verbatim in the ‘Suda’ (Suidas) and the Byzantine ‘Glossarium militare’ (‘Definitiones’) 25.

2. THE ROMAN ELEMENTS

It is only in the second, or Roman, section of the Tactica that we can discern any real originality on Arrian's part. Here, in general contrast to the Hellenistic section of his treatise, he is writing about subjects he knows from personal experience. Given the conditions that obtained on the eastern frontier of the Roman empire in the early second century, where the highly effective cavalry of the Parthians, Armenians, and Alani loomed large in the Roman military consciousness, it is not surprising that Arrian should decide to focus his attention on the drilling of auxiliary cavalry.

In marked contrast to his handling of Hellenistic tactical terminology, where Arrian is doing little more than merely plagiarizing the minute detail of the written tradition, technical terms are conspicuous by their absence from his account of the Roman cavalry exercises. Only four new terms are introduced—πετρινοs (37.4), ξννημα (42.4), τολοντεγον (43.2), and the attack called Κανταβρικν (40.1, 40.6)—the first three Celtic, the last Spanish in origin.

The πετρινοs, which Arrian claims is the most difficult of the basic cavalry exercises (37.4), surprisingly enough never reappears in the text of the Tactica after its original citation. The ξννημα, which seems to be an elaboration of the πετρινοs, may by contrast have originated as genuine battle-drill, since Arrian specifies that it is to be performed after the troopers have changed into their field uniforms, including breastplates (41.1). Both exercises are to be carried out while the rider is in the course of wheeling to the right (37.4, 42.2-3), i.e. the javelin is hurled across the front of the rider's body. In the case of the ξννημα, the final throw is made at an angle to the line of advance, and is preceded by two or three earlier throws while riding in a straight line. The τολοντεγον is a manœuvre that involves the horseman raising his shield over his head to protect his back from one opponent's strike and swinging his lance around to confront another opponent (43.2). The “Cantabrian attack” involves aiming a heavy spear, without an iron weapon-head, at an opponent's shield while making a full wheel around him (40.1-6).

In describing the Roman cavalry exercises, Arrian is at pains to stress that the emperor's specifications are being followed, actually noting this fact no less than three times (42.2, 42.4). And, indeed, the three Celtic drills are adduced primarily in support of an encomium of Roman and especially Hadrianic eclecticism. This approach is recapitulated in the treatise's concluding paean in praise of Hadrian for adapting the military techniques of the empire's barbarian enemies, in pursuance of the sound Roman tradition of appropriating excellence whatever the source (42.2, 42.4, 44.1). The cavalry drill of the Middle Empire is represented as a prime example of such commendable eclecticism, with, as we have seen, manœuvres borrowed from the Gauls and Spaniards, and further innovations from the West introduced by Hadrian himself. Indeed, Arrian's account of the cavalry drills seems to be intended to evoke Hadrian's imperatorial address at Lambaesis (ILS 9134, 2487). There is a striking agreement in content: both, for example, mention the “attack from concealment” (e tecto) and the “Cantabrian attack” (35.1, 40.1).

Despite Arrian's extensive personal acquaintance with Roman practice, his description of the auxiliary cavalry exercises is disappointing. Devoid of any didactic utility, his account is written from the standpoint of “top brass” judging a military display from the reviewing stand. Accordingly, he stresses the importance of good visibility and the need for the javelin-throwers to maintain proper intervals between their successive waves, so that good and bad performers can be distinguished and appropriate praise or blame can be awarded (38.2-6). The focus is on the outstanding performers, N.C.O.s like the decuriones, duplicarii, sesquiplicarii (42.1), and the draconarii, who carried the serpent standards and who, like all standard-bearers in all armies, had to be expert in the entire repertoire of drill (40.8-12). The average trooper, trained merely to follow the standards, is largely ignored (35.6). Arrian does, however, disparage display for its own sake (40.12), specifying twenty throws as the most any rider can make without cheating (40.11). Each trooper in each decuria is to perform individually, and the reviewing commander is to single out for particular commendation the decuria that contains the most troopers who excel in javelin-throwing, which Arrian extols as “practise for the actuality of combat” (42.5; cf. Aelian 21.2). But Arrian's object is clearly that of describing a mere festive display, not of instructing his readers in the mechanics of cavalry drill. His descriptions are occasionally obscure (as at 38.1 and 40.2), often degenerating into mere summaries of the proceedings, especially towards the end of the treatise (as at 43.1-4), where it is the inferior second-line squadrons that carry out the exercises, including the attack καντβρικη, albeit at a lower level of expertise and using slings and hand-thrown stones. Rather than stressing tactical utility, Arrian extols the beauty, elegance, astonishing qualities, and sheer splendour of these manœuvres (34.2, 34.4, 34.5, 35.1, 36.4, 38.2-4, 40.7, 40.12, 44.2).²⁵

Bosworth contends that the Tactica is intended to serve larger purposes than the essentially self-ingratiating plagiarism suggested above. According to him, “it provides a stylistically improved manual of Hellenistic tactics, emphasising the manœuvres which could be developed by contemporary commanders, and describes the training which inculcated the technical skills to carry them out”.²⁶ This is to see too much in it, since there is actually very little in the first part of the Tactica that can be divorced from Hellenistic military organization, arms, and equipment. Moreover, Arrian's main parallels—like that between Epaminondas' use of the grand tactical lambda-shaped infantry εμβολοs at Leuctra and Mantineia and the minor tactical Roman testudo—are hardly compelling. And, as the work and protestations of Aelian indicate, lightly reworking a tactical manual for dedication to someone like a militarily-successful or -interested emperor, though reprehensible in terms of the modern canons of originality, seems to have been a commonplace activity in the Middle Empire period. Arrian's superficially Romanized manual is thus a conventional, ‘philosopher's’ way of ingratiating himself with the reigning emperor, rather than an attempt at genuine originality or the production of a tactical guide of practical utility to contemporary commanders.

IV. THE SOURCES OF ARRIAN'S TACTICA

As is evident from the earlier comparison, the three tactical treatises of Arrian, Aelian, and Asclepiodotus are very closely related. H. Köchly²⁷ in fact believed that Arrian and Aelian were merely two recensions of the same treatise and printed them in parallel columns in his edition,²⁸ omitting altogether those sections of Arrian with a purely Roman content (i.e. 33.1-44.3 in the edition of Roos and Wirth). This view and the hypothesis subsequently advanced by R. Förster,²⁹ that Arrian drew on Aelian for his own Tactica, were decisively refuted by A. Dain.³⁰ Dain argued that Aelian and Arrian were independently derived from a common (lost) source, which, together with Asclepiodotus, in turn derived from a tactical treatise (also lost) by the Stoic philosopher Poseidonius of Rhodes (mentioned by both Arrian, Tact. 1.1 and Aelian 1.2). On this view, the differences which separate Asclepiodotus from Aelian and Arrian are the result of changes in the lost intermediate source. Stadter,³¹ however, argues that all three authors used Poseidonius directly, but Asclepiodotus introduced a number of modifications. The fact that Arrian's immediate source was either the Stoic Poseidonius or a known follower would explain his attraction for our author, the first part of whose Tactica should thus be interpreted as a fairly conventional retailing of an aspect of the Stoic philosophical tradition. Since Arrian (Tact.1.1) and Aelian (1.2) both imply that Poseidonius' treatise was, like those of Asclepiodotus and Arrian, entitled ‘Τεχνη τακτικη’, the title itself was clearly part of the tradition (a change of title possibly being another modest attempt at originality on the part of Aelian).

Poseidonius himself is known to have continued the ‘Histories’ of Polybius, and it is probable that Polybius' own (lost) tactical treatise (Arrian, Tact.1.1; Aelian 1.2) provided the basis for the Stoic's ‘Τεχνη τακτικη’. Certainly the striking resemblance between Polybius 18.29.2-5 and 18.29.1-30.4, on the one hand, and Aelian 14.2-6, Arrian, Tact. 12.6-10, and Asclepiodotus 5.1-2, on the other, argues very strongly for Polybius' position as the principal source for the entire tradition.

V. THE MANUSCRIPT TRADITION

The manuscript tradition of Arrian's Tactica is unitary and the number of manuscripts very limited.³² Only one, the Codex Laurentianus graecus 55.4, is of any independent value.

—Index siglorum [Dain] of known manuscripts:

F: Cod. Laurentianus gr. 55.4, saec. X.

B: Cod. Bernensis gr. 97, saec. XVI.

W: Cod. Parisinus gr. 2446, saec. XVII.

Cod. Parisinus suppl. gr. 270 (folios 169-181), saec. XVII.

Cod. Parisinus gr. 2539, saec. XVII.

C: Cod. Barberinianus gr. 59.

D: Cod. Barberinianus gr. 200.

e: Cod. Barberinianus gr. 245.

—Codex Laurentianus graecus 55.4.:

The “prototype”³³ of the entire manuscript tradition of Arrian's Tactica, like those of Aelian and Asclepiodotus, is to be found in the celebrated Cod. Laurentianus gr. 55.4, folios 182^r to 195^v, now in the Biblioteca Medicea-Laurenziana in Florence.³⁵ The Laurentianus is a parchment codex of 405 folios, each 32.5 cm. × 26 cm., dating from the time of the Byzantine emperor Constantine VII Porphyrogenitus (reigned a.d. 912-959). At the beginning of the fifteenth century it belonged to one Demetrius Lascaris Leontarios, who utilized the blank folios for family records (down to the death of his mother in 1450). In 1491 it passed into the possession of Lorenzo the Magnificent (died 1492) and thereafter (1508) into that of the Medici Pope Leo X, who finally deposited it in the Monastery of St. Laurence in Florence in 1521. It contains three separate collections of almost exclusively tactical works. They are as follows:

First Collection (Byzantine):

1^r: Constantine Porphyrogenitus, ‘Praecepta imperatori’

3^r: [Mauricius] Urbicius III, ‘Strategicon’

68^r: ‘De militari scientia’

76^v: ‘Hypotheseis (ex Polyaeno)’

103^v: ‘De Re strategica’

131^r: ‘Praecepta e Mauricio’

Second Collection (Hellenistic and Classical):

132^r: Asclepiodotus, ‘Techne tactica’

143^r: Aelian, ‘Tactica theoria’

159^v: Aeneas Tacticus, ‘Commentarius poliorceticus’

182^r: Arrian, ‘Techne tactica’

196^r: Arrian, ‘Ectaxis contra Alanos’

198^r: Onasander, ‘Strategicus’

Inserted between the Second and Third collections:

216^r: ‘Rhetorica militaris’

230^r: Sextus Julius Africanus, ‘Cestorum fragmenta’

Third Collection (Byzantine):

253^r: Leo VI, ‘Problemata’

281^r: Leo VI, ‘XVIII Tacticae Constitutiones’

379^r: Leo VI, ‘De Incursionibus necopinatis’

387^r: Leo VI, ‘De Obsidionibus’

394^r: Leo VI, ‘De navali pugna’

401^r: ‘Quomodo Saracenis debelletur’ (paraphrase of Leo VI)

403^r: Constantine Porphyrogenitus, ‘De moribus diversarum gentium’

The codex has suffered extensively from damp and mildew during the eleven centuries of its history, and the text is almost completely illegible in places, especially towards the left-hand margin of each page near the spine of the present sixteenth-century binding. This difficulty is compounded by the fact that the ink has, to varying degrees, come through the individual folios, as on folio 159^r where the last three lines of the text of Aelian are largely obscured by the ornamental band marking the beginning of the text of Aeneas Tacticus at the top of folio 159^v, which has also imprinted itself on the opposing page, folio 160^r. Here and there folios have been torn out, as in the case of 145a and, more significantly, 181a. The recto of 181a was either blank or carried a diagram like the series of five or six diagrams illustrating the text of Aelian, now largely obliterated (in five spaces on folios 151^v, 152^v, 153^r [three]), or the earlier sequence illustrating the text of Asclepiodotus (on the folios 136^v, 137^v, 138^v, 139^r, 139^v, 140^r, 140^v, 141^r [three]).

The text of the codex, 32 lines per each full page, is written throughout in an elegant miniscule hand current prior to the middle of the tenth century, and, where not faded or obscured, is relatively easy to read. It is, however, far from immune to minor scribal mistakes, although the overall standard of accuracy is fairly high. But comparison of Aelian 12 with Asclepiodotus 5.1 reveals an important instance of saut du même au même. The third sentence of Aelian 12 … corresponds to the second sentence of Asclepiodotus 5.1. …³⁵

VI. EDITIONS OF ARRIAN'S TACTICA

Although the editio princeps of Aelian was published by Francesco Robortello (1516-1567) in Venice as early as 1552 under the title ‘Aeliani de militaribus ordinibus instruendis more graecorum liber a Francisco Robortello Utinensi nunc primum graece editus multisque imaginibus, et picturis ab eodem illustratus’, the cognate treatise of Arrian had to wait almost a century for publication. The editio princeps of Arrian's ‘Tactica’ was published in Upsala in 1644 by Joannes Schefferus (Scheffer), together with Mauricius' (i. e. Urbicius III's) ‘Strategicon’, under the title ‘Arriani Tactica et Mauricii artis militaris libri duodecim omnia numquam ante publicata’, pp. 1-121. A second edition, based on Manuscripta ex Bibliotheca Claudii Salmasii (Cl. Salmasius, 1588-1653), was published by Nicolaus Blancardus (Blankaert) in Amsterdam in 1683 in ‘Arriani Ars Tactica, Acies contra Alanos. Periplus Ponti Euxini. Maris Erythraei. Liber de Venatione, etc. … Cum Interpretibus Latinis’, pp. 1-97. Subsequent editions were published in Paris in 1792-1811 and in 1845 (the Didot edition by Karl Müller).

The first critical edition of the Tactica was that of Köchly in H. Köchly and W. Rüstow, Griechische Kriegsschriftsteller 2.1. (Leipzig, 1855) 201-552. The Köchly edition was based primarily on two poor-quality copies of the Codex Laurentianus gr. 55.4, namely the sixteenth-century Codex Bernensis 97 and the seventeenth-century Codex Parisinus gr. 2446. Unfortunately, Köchly's work is vitiated by his failure to utilize or even consult directly the famous Codex Laurentianus. A. G. Roos' critical edition (1928) of Arrian was, however, based on the Florentine codex and, with G. Wirth's revision (1968), seems to approximate the author's holograph. No new critical edition is required.

VII. TRANSLATIONS OF THE TACTICA

Although a German translation of Aelian accompanies the Köchly edition, Köchly and Rüstow declined to translate those passages of Arrian that are not paralleled in the text of Aelian. Arrian's Tactica was, however, finally translated into German by Franz Kiechle, ‘Die Taktik’ des Flavius Arrianus, Bericht der Röm.-Germ. Kommission 45 (1964) 87-129. Charles Theophile Guischardt translated the first (i. e. Hellenistic) part of Arrian's Tactica in: Mémoires militaires sur les Grecs et les Romains 2 (Lyon, 1760). There is to date no published English translation.

Notes

1. Sir Ronald Syme, Tacitus, vol. 2 (Oxford, 1958) 653-656.

2. Sir Ronald Syme, The Career of Arrian, Harvard Studies in Classical Philology 86 (1982) 181-211, at 184 (= Idem, Roman Papers, vol. 4, ed. A. R. Birley [Oxford, 1988] 21-49, at 24).

3. E. L. Wheeler, The Occasion of Arrian's Tactica, Greek, Roman, and Byzantine Studies 19 (1978) 351-365, at 352.

4. A. B. Bosworth, From Arrian to Alexander: Studies in Historical Interpretation (Oxford, 1988) 17-24; Idem, Arrian and Rome: the Minor Works, above in this volume (ANRW II.34.1) 226-275, esp. 253-264.

5. A. B. Bosworth, A Historical Commentary on Arrian's History of Alexander (Oxford, 1980) 8-11, esp. 11: “a date of composition before 125”; Bosworth (above, note 4) 29-37, esp. 37: “we can date the work sometime towards the end of Trajan's reign”; and Bosworth, Arrian's Literary Development, Classical Quarterly N.S. 22 (1972) 163-185, esp. 185. Cf. Wheeler (above, note 3) p. 355, note 19.

6. Bosworth (above, note 4) 28.

7. AE 1968. 473 (from Corinth): on which see G. W. Bowersock, Greek, Roman and Byzantine Studies 8 (1967) 279-280; and AAA 3 (1970) 377-380 (= AE 1971.437) (from Athens).

8. The evidence is summarized in Bosworth (above, note 4) 25, Bosworth, A Historical Commentary on Arrian's History of Alexander (Oxford, 1980) 5-6, and Idem., Arrian and Rome etc., above in this same volume (ANRW II.34.1) 272-275 (V. ‘The New Xenophon’). Cf. A. Silberman, Arrien ‘Périple du Pont Euxin’: Essai d'interprétation et d'évaluation des données historiques et geographiques, above in this volume (ANRW II.34.1) 302 n. 147 and 303 n. 155.

9. Bosworth (above, note 4) 26-27; Bosworth (above, note 8) 6-7.

10. Classical Philology 73 (1978) 117-128, at p. 121, note 15.

11. Cf. Wheeler (above, note 3) p. 355, note 18.

12. Wheeler (above, note 3) 364-365.

13. For the date of Asclepiodotus, see W. A. Oldfather and C. H. Oldfather in: The Illinois Greek Club, Aeneas Tacticus, Asclepiodotus, Onasander (Loeb edition: Cambridge, Mass. and London, 1923) 230-237; and A. Dain, edited posthumously by P. Lemerle, Les Stratégistes Byzantins, Travaux et Mémoires 2 (1967) 317-392, at 326. For the date of Aelian, see A. Dain, Histoire du texte d'Élien le tacticien (Paris, 1946) 15-21, esp. 18-19; and A. M. Devine, Aelian's ‘Manual of Hellenistic Military Tactics’: A New Translation from the Greek with an Introduction, The Ancient World 19 (1989) 31-64, at 31. The Prooimion or preface to Aelian's ‘Τακτικη θεωρια’ is addressed to the reigning emperor, and although our surviving manuscripts all give the name of this emperor as “Hadrian” (praef. 1), it is obvious, from the reference (praef. 3) to the emperor's “deified father Nerva” (reigned a.d. 96-98), that the received text is corrupt and that Trajan (reigned 98-117) is the princeps addressed. The name of the distinguished consular Sextus Julius Frontinus, who encouraged Aelian's study of Hellenistic tactics in conversations during the author's stay in Formiae, also appears in this section (praef. 3). Frontinus was himself well-known, not only as a successful military commander (who had conquered the Silures in Britain in 78), but also as the author of two military treatises, the ‘Strategemata’ and the (lost) ‘De re militari’, and Aelian mentions him (1.2) in his catalogue of tactical writers (1.1-2). Frontinus died in 103, and the notice in the Prooimion does seem to suggest that the great consular was no longer writing at the time the ‘Tactica theoria’ was being composed. Aelian's allusions to the emperor's military prowess, though attended by hyperbole, provide us with a slightly later terminus post quem. At praef. 4 Aelian speaks of the emperor excelling “all the other generals who have ever been”, and at praef. 6 he alludes to his having “commanded in such great wars”. The plural in the latter case, together with the transcendent strategic status accorded the emperor, indicates the end of the Second Dacian War in 106 as the terminus ante quem non for Aelian's dedication. Trajan was in Rome from 106 or 107 until 113, when he departed the capital to begin the series of wars in Parthia and Armenia that occupied his attention until his death in 117. The publication of Aelian's ‘Tactica theoria’ can thus be plausibly assigned to these six or seven years of victorious peace, during which time Roman public interest in military affairs would have stood very high.

14. On the ‘Short’ and ‘Long’ (i. e. interpolated) Recensions, see A. Dain, Histoire du texte d'Elien le tacticien (Paris, 1946), passim, esp. 77-115; Devine (above, note 13), esp. 33-39 and 59-64.

15. See above, note 8.

16. The comparative method adopted here owes much, despite significant points of disagreement, to P. A. Stadter, The Ars Tactica of Arrian: Tradition and Originality, Classical Philology 73 (1978) 117-128, at 120.

17. Devine (above, note 13) 35 and 48, see also below, section V, ad fin.

18. On the various tactical senses of the term, see A. M. Devine, ΕΜΒΟLΟN: A Study in Tactical Terminology, Phoenix 37 (1983) 201-217, esp. 213-216 (for the Alexandrine instances).

19. Added by Roos on the basis of the parallel passage in Aelian.

20. Dain (above, note 14) 98.

21. On the historical authenticity of the passage and its content, see Devine (above, note 18) 203-210.

22. Anrw ii 34.1. See below, section V, ad fin.

23. On the cavalry wedge, see Devine (above, note 18) 201-202.

24. See further, A. M. Devine, Diodorus' Account of the Battle of Paraitacene, The Ancient World 12 (1985) 75-86, at 79-80; N. G. L. Hammond, A Cavalry Unit in the Army of Antigonus Monophthalmus: ASTHIPPOI, Classical Quarterly N.S. 28 (1978) 128-135, esp. 134.

25. Wheeler (above, note 3) 354-361, esp. 357. In general, A. B. Bosworth, Arrian and Rome etc., above in this volume (ANRW II.34.1) pp. 226-275, esp. 253-264, vastly overestimates Arrian's value and originality as a source for Roman and even Hellenistic tactics.

26. Bosworth, Arrian and Rome etc., above in this volume (ANRW II.34.1) p. 259.

27. H. Köchly, De libris tacticis, qui Arriani et Aeliani feruntur, dissertatio, Index lectionum (Zurich, 1851).

28. In H. Kochly and W. Rüstow, Griechische Kriegsschriftsteller 2.1 (Leipzig, 1855) 270-470.

29. Hermes 12 (1877) 426-449.

30. Histoire du texte d'Élien le tacticien (Paris, 1946) 26-40 (especially 39: stemma).

31. Classical Philology 73 (1978) 117-118.

32. Select bibliography: A. Dain, Les manuscrits des traités tactiques d'Arrien, Mélanges Bidez I, pp. 160-194 = Annuaire de l'Institut de philologie et d'histoire orientale 2 (Bruxelles, 1934) 157-184; ID., Le Parisinus gr. 2522, Revue de Philologie 67 (1941) 21-28.

33. To use the later terminology of Dain, not the “archetype”, as Roos-Wirth and Bosworth call it.

34. Select Bibliography: A. M. Blandini, Epistola de celeberrimo codice Tacticorum Bibliothecae Laurentianae (Florence, 1761) = id., Catal. cod. Manuscriptorum graecorum Bibl. Laurentianae 1 (Florence, 1768) col. 218-238; R. Förster, Studien zu den Griechischen Taktikern, Hermes 12 (1877) 426-471, esp. 427-430 and 459-461; The Illinois Greek Club, Aeneas Tacticus, Asclepiodotus, Onasander (Loeb edition: Cambridge, Mass. and London, 1923) esp. 18-19, 240-243, and 363-365; A. Dain, Histoire du texte d'Élien le tacticien (Paris, 1946) passim, esp. 183-202 and 375-377; id., Les manuscrits d'Énée le tacticien, REG 48 (1935) 1-32, esp. 6-10; id., edited posthumously by P. Lemerle, Les stratégistes Byzantins, Travaux et Mémoires 2 (1967) 317-392, esp. 382-385; A. G. Roos, revised by G. Wirth, Flavii Arriani Quae extant omnia, Teubner edition, vol. 2: Scripta minora et fragmenta (Leipzig, 1968) xx-xxiv; A. B. Bosworth, Arrian and the Alani, Harvard Studies in Classical Philology 81 (1977) 217-255, esp. 217 and 251; P. A. Stadter, The Ars Tactica of Arrian: Tradition and Originality, Classical Philology 73 (1978) 117-128, esp. 118-119.

35. Devine (above, note 13) 35.

All references here will be in terms of the Roos-Wirth (Teubner) edition of Arrian, unless indicated by the addition of K for the Köchly-Rüstow edition, the forthcoming Devine edition of Aelian (indicated by D as required), and the Oldfather edition of Asclepiodotus in the Loeb Classical Library.