This paper argues that the town-by-town narratives of the Catalogue of Ships are organized according to the realities of local geography. The poet's methodical organization becomes apparent when lists of Catalogue sites are interpreted in light of two previously unnoticed units of oral compositional thought: syntactical and line-by-line groups, which frequently map onto local topographic features. In addition, we argue that the exceptional way that syntactical groups are used in the Boiotian contingent may suggest its origin in an oral-traditional Thebaid.
the catalogue of ships in book two of the iliad is an enormous feat of memory, listing the twenty-nine geographical regions that supported the Greek expedition to Troy and enumerating toponyms within those contingents, [End Page 1] for two hundred and three place names, landmarks, and ethnics in all.1 It is no wonder, then, that before embarking upon his virtuoso two hundred and fifty line recreation of the Greek fleet, in which he will also tell us about its leaders, their histories, and the number of their ships, Homer asks the Muses to help him remember the details (2.488–92).
πληθὺν δ᾽ οὐκ ἂν ἐγὼ μυθήσομαι οὐδ᾽ ὀνομήνω,οὐδ᾽ εἴ μοι δέκα μὲν γλῶσσαι, δέκα δὲ στόματ᾽ εἶεν,φωνὴ δ᾽ ἄρρηκτος, χάλκεον δέ μοι ἦτορ ἐνείη,εἰ μὴ Ὀλυμπιάδες Μοῦσαι, Διὸς αἰγιόχοιοθυγατέρες, μνησαίαθ᾽ ὅσοι ὑπὸ Ἴλιον ἦλθον…2
I could not recount their numbers nor name them,not if I had ten tongues and ten mouths,and an unbreakable voice and a brazen chest within,unless the Olympian Muses, daughters of aegis-bearingZeus, would remind me how many came under Ilium.
Homer's invocation raises a major question: this inspiration of the Muses aside, how does a poet working in an oral tradition, long before the development of maps, compose and organize such a detailed geographical tour of the Greek world?3 This paper reevaluates the geospatial organization of Homer's lists of place names, thereby revealing that many contingents are much better organized than previously understood. We suggest that this high level of geospatial organization, in turn, serves the poet as a mnemonic device.
introduction i: new geospatial paradigms
Scholars have pointed out that the Catalogue displays an overall geospatial organization.4 The poet proceeds in three more or less contiguous, contingent-by-contingent itineraries around the Greek world: itinerary one [End Page 2] begins in Boiotia, winds west toward Delphi, north through Lokris, back east toward Euboia, south through Attika towards a thorough tour that covers the Peloponnese, then west from Elis towards the Ionian islands, ending at last on the mainland in Aitolia; itinerary two breaks off from central Greece, beginning in Krete, then moving west and north through Rhodes and the other Dodekanese; itinerary three begins back on the mainland, with the Myrmidons in northern Greece, and winds its way through eastern Thessaly (with a detour to Dodona), bringing the Catalogue to a close with the Magnetes in the far north east.5 It has been argued that this generally contiguous plan serves the oral poet as a "spatial mnemonic," or mental roadmap, facilitating his cataloguing of place names, a theory we shall discuss in more detail.6
These three contingent-by-contingent itineraries have led scholars to conclude that Homer's town-by-town narration of sites is disorganized by comparison. Kirk, for example, notes the clear geographical sequence of the three contingent-by-contingent itineraries, but finds little evidence of an underlying order on a town-by-town basis: [End Page 3]
Although the places named for each contingent are usually in no particular logical order, the contingents themselves are presented, for the most part, in a conspicuous geographical sequence.7
Richard Seaford makes a similar claim:
Another feature [of the Catalogue of Ships] that is inconsistent with a unified conception of space is that—although the contingents within each sub-list are in geographic order—the towns and other localities … within each contingent are not listed in geographic order; and moreover there is only one statement of any spatial relation between them, and only one instance of them defining the boundaries of an area.8
Underlying these arguments is the assumption that the best measure of geospatial organization is the travel itinerary, that is, a narrative that recounts toponyms in a sequence suitable for travel. So long as the travel-itinerary is the implicit standard of geographic organization, these scholars are correct that the town-by-town narratives within the contingents do not betray much in the way of order.
But there are other ways of organizing space. This article departs somewhat from the itinerary paradigm of geospatial organization. By interpreting portions of the Catalogue according to a new geospatial paradigm, we show that many of the twenty-nine contingents do, in fact, demonstrate a considerable organizational sophistication, even at the local level. This high level of organization comes to light more clearly when the contingents are analyzed according to two new interpretive tools, the first of which we call syntactical groups, and the second, line-by-line groups. A syntactical group consists of a single town or a group of towns governed by one verb; a line-by-line group consists of a single town or a group of towns contained within a single verse line.9 These units of poetic-geographical organization frequently map onto local topographic features of the landscape, such as travel routes, waterways, mountains, valleys, peninsulas, and islands. When the toponym lists of various [End Page 4] contingents are divided according to their syntactical and/or line-by-line groups, it becomes clear that the place names within a given syntactical or line-by-line group frequently cluster around such landmarks. From the perspective of these geographical sub-regions, a number of the contingents reveal themselves to be remarkably well ordered.
Analysis of Homer's narrative of towns according to syntactical and line-byline groups has also allowed us to formulate a separate and novel interpretation of the Boiotian contingent, an exceptional region in a number of ways, but especially geospatially. Whereas the syntactical and line-by-line groups of many other contingents very clearly reflect geographical sub-regions, the same is not true of Boiotia. Instead of distributing towns according to shared geographical sub-regions, the Boiotian catalogue is enumerated as a rough circuit around the unnamed city of Thebes. Therefore, instead of moving through an overview of local geographical landmarks or sub-regions, as in many other contingents, the Boiotian contingent has a geospatial narrative focus on a central point: Thebes itself. We will suggest that this perspective, akin to a teichoskopia, is perhaps borrowed from an oral Thebaid.
It must be stated at the outset that the study of the Iliad poet's knowledge and organization of geographical information is by nature speculative. This is so, in large part, due to the difficulty of securely identifying Catalogue sites with actual archaeological remains. As Dickinson has pointed out, there is always a risk of circularity: if the presence of Mykenaian archaeological remains is taken as evidence for the identification of a particular Catalogue site, this assumes what requires proof, namely, that Catalogue sites are Mykenaian.10 The evidence is further confused by the efforts of various ancient towns, in an attempt to provide themselves with Homeric prehistories, to associate themselves with Catalogue sites, regardless of the historicity of such claims.11 In addition, the testimonies of ancient geographers like Strabo, upon whose claims many identifications rely, do not always withstand scrutiny.12 Nevertheless, we hope that the present analysis can usefully illuminate Homer's oral-poetic mastery of the landscape, despite the admitted difficulty of identifying Catalogue sites with actual archaeological remains.
introduction ii: cognitive maps and memory
The analysis of the Catalogue of Ships according to syntactical and line-by-line groups, besides illustrating Homer's command of geospatial knowledge, also [End Page 5] demonstrates one method by which an oral poet was able to remember and deploy such a trove of highly detailed information. The division of contingents into syntactical and line-by-line groups results in a methodical arrangement of clusters of place names. Such arrangements differ notably from itineraries, in that they do not distribute names in a sequence suitable for travel. We suggest that the organization of a given contingent into an orderly succession of clusters of place names is precisely what allows the poet to remember the place names associated with that contingent: organizing information into spatially contiguous or spatially related segments facilitates the memory. According to this theory, therefore, poets use different geospatial mnemonic devices to aid recollection of different layers of geographical organization. At the contingent-by-contingent level, the poet's memory is facilitated by the itinerary paradigm. Within the contingent, however, the poet uses geospatially related sub-groups of place names: syntactical and line-by-line groups. Despite clear differences between the itinerary method and the organizational mode presented in this paper, they both facilitate recall by organizing toponyms into spatially related sequences.
Our argument that the use of syntactical and line-by-line groups has the potential to serve as a mnemonic device therefore builds on the findings of a number of scholars who have successfully used data from the field of cognitive psychology about visuospatial thinking to illuminate the ways that oral-traditional poets employ spatial imagery and itineraries as memory aid. Once again, we are not proposing that the toponym lists of the contingents are equivalent to itineraries. When analyzed according to syntactical and line-by-line groups, however, a number of contingents are revealed as orderly sequences of roughly contiguous geographical sub-regions.
Most recently, Jenny Strauss Clay has illustrated how spatial imagery and cognitive mapping function in the poet's visualization of the battlefield before Troy, allowing Homer to organize and remember complex battle sequences and enabling his audience to visualize them.13 Minchin, in her work on Homer and memory, has argued that the poet's mnemonic strategy for the Catalogue of Ships is fundamentally visuospatial: he recites his list in three spatially contiguous itineraries that help the poet remember it [fig. 1].14 In elaborating the visuospatial memory tools used by the poet, both Clay and Minchin build on [End Page 6] the work of psychologists like Rubin, who has demonstrated a fundamental connection between visuospatial thinking and oral tradition, arguing forcefully that, "oral traditions are spatial."15 Although Rubin does not note Homer's use of a geographical visuospatial mnemonic in his brief discussion of the Catalogue of Ships, he does emphasize the value of visualizing sequences in space as a memory tool for oral performers.16 Rubin also notes the value of concrete, easy-to-visualize imagery for oral poets, and holds up the Catalogue of Ships as an example. Whereas Rubin dwells primarily on the fact that the Catalogue mainly consists of simple, concrete imagery, thus making it easier to visualize and therefore to remember, we add that the poet also visualizes and remembers his Catalogue as a geographical schema.17
All three of these scholars have noted that Classical authors were well aware of the power of visuospatial imagery as a mnemonic device, as exemplified by the method of loci described by Cicero and attributed to Simonides, while both Clay and Minchin have related this method to various spatial mnemonics employed by Homer.18 The method of loci, whereby one distributes items to be remembered throughout a "memory palace" or along a route and then "visits" those places in the mind's eye to retrieve the information, continues to be recognized as a powerful mnemonic device.19 As Clay has pointed out, Homer's use of loci demonstrates the mnemonic power of linking items to be remembered with a visualized space.20 Because Homer's contingent-by-contingent narratives resemble three itineraries, Clay and Minchin have drawn a connection to the loci-method. We argue here that the poet uses a similar geospatial mnemonic within many of the contingents, at the level of the town-by-town narrative: rather than organizing the towns into travel routes, the poet organizes them into syntactical and line-by-line groups that are associated with spatially related landmarks or sub-regions. This method results in a highly organized cognitive map for a number of contingents. We believe that such a map, in addition to demonstrating Homer's knowledge of local Greek geography, also facilitated the poet's memory of geospatial information. The recall of a contingent would also trigger recall of an associated hierarchy of geospatial information: within each contingent are particular landmarks [End Page 7] or sub-regions; clustered around each landmark or sub-region are associated groups of individual toponyms.
When Homer's toponym lists are interpreted according to syntactical and line-by-line groups, it therefore becomes possible to visualize many of the more organized contingents as a series of such sub-regions or landmarks, with a group of toponyms attaching to each one. Such an organizational schema reveals a four-level hierarchy of geospatial knowledge relating to the Catalogue that is organized by and accessible to the poet: at the most general level (Level 1) are the three contingent-by-contingent itineraries; next (Level 2), there are the individual contingents; moving to the following layer of detail (Level 3), there are the subgroups of toponyms within each contingent, clustered around local land features or sub-regions; finally and most specifically (Level 4), there are the individual place names within each syntactical or line-by-line group.21 As such, the itinerary model, wherein locales are placed in an order that could plausibly serve as a travel route, does not hold at the most local level. As the poet deploys more specific geospatial information, the basis of organization is no longer dependent on the itinerary model, but upon the linking of place names with contiguous or spatially related sub-regions or landmarks.22
To summarize the goals of this paper: we aim to establish that Homer frequently possesses knowledge of regional Greek geography; this knowledge is demonstrated by the Catalogue's high level of geospatial organization even at the town-by-town level; this organization comes to light especially by analysis of toponym lists by means of two related interpretive tools: syntactical and line-by-line groups, units of compositional thought that frequently map onto local topography; given this high level of geospatial organization, it is likely that Homer employs his geographical knowledge as a spatial mnemonic. A series of eight case studies will demonstrate the way in which syntactical and line-by-line groups serve to organize local geographical units on a micro level and facilitate the composition of the Catalogue. After the case studies, we offer an analysis of the exceptional geospatial perspective of the Boiotian catalogue. [End Page 8]
case study one: syntactical groups in the mykenaian contingent (fig. 2)
Agamemnon's contingent contains four syntactical groups, all of which fall along travel routes.23 The narration of the syntactical groups starts in the east and proceeds to the west in an orderly manner. The first syntactical [End Page 9] group is composed of Mykenai (1A), Korinth (1B), and Kleonai (1C).24 We identify this cluster of three sites as a syntactical group because they share a common verb (569–70):
Οἳ δὲ Μυκήνας εἶχον, ἐϋκτίμενον πτολίεθρον,ἀφνειόν τε Κόρινθον ἐϋκτιμένας τε Κλεωνάς …
They who held the well-built citadel of Mykenaiand wealthy Korinth and well-built Kleonai …
Not only does the verb εἶχον govern each of the three toponyms in the first syntactical group, but each of the three sites also falls along a distinct travel corridor between Mykenai and Korinth by way of Kleonai.25
The second syntactical group, composed of Orneai (2A), Araithyrea (2B), and Sikyon (2C), is located directly to the west of the first group, distributed along a parallel travel corridor. Once again, Orneai, Araithyrea, and Sikyon belong to the same syntactical group because they are all governed by the same verb (571–72):
Ὀρνειάς τ' ἐνέμοντο Ἀραιθυρέην τ' ἐρατεινὴνκαὶ Σικυῶν', ὅθ' ἄρ' Ἄδρηστος πρῶτ' ἐμβασίλευεν …
…and who inhabited Orneai and lovely Araithyreaand Sikyon, where Adrastos first was king …
As was true of the first syntactical group, all three of these locations fall along a clear north-south travel corridor, composed of two routes: Orneai-Araithyrea (Phlious) and Araithyrea (Phlious)-Sikyon. The Orneai-Araithyrea route was a means of travel to Argos, "passing from the southwestern part of the Phleiasian [Araithyrean] plain into the valley of Leondion to Sterna in the upper valley of the Inachos river."26 The route from Araithyrea north to Sikyon was a direct route to the Gulf of Korinth.27 The toponyms comprising the first two syntactical groups of the Mykenaian contingent, then, are arranged along two parallel north-south travel corridors, and Homer's narrative of those syntactical groups has moved from east to west. [End Page 10]
The third and fourth syntactical groups continue the westerly movement, both units falling west-east along the northern coast of Achaia instead of the interior, north-south locations of the first two groups. Each of these routes falls along what must have been a well-defined travel corridor: the northeastern coast of Achaia, certainly accessible by water and probably by coastal road. Each of the toponyms in the third and fourth syntactical groups falls within the territory of the Achaian League as formed in the seventh century b.c.e., so it is to be expected that a network for trade and communication would have existed between these locations from at least that period.28 The Sikyonians, who are the last entry in the second syntactical group, are supposed (Paus. 7.26.2–3) to have launched an attack on Hyperesia on the north coast of Achaia, which may indicate some measure of connectivity at an earlier date. More concretely, it is notable that Pausanias himself followed a land route that encompassed the sites in our third and fourth syntactical groups. Although Pausanias's second century c.e. date is very late from a Homeric standpoint, Valderrama argues that the road he followed existed throughout Classical antiquity. This roadway, stretching from Aigion to Korinth, would have provided important connections between Achaia and the Korinthia, as well as providing access to interior hinterlands by means of more minor routes. Valderrama further proposes that the coast road would have been supplemented by coastal navigation.29 There is other evidence to suggest that Homer would have conceived of the northeastern coast of Achaia as a travel corridor. The stretch of coast between Korinth and Patras was a region of good to moderate connectivity in the ancient Peloponnese, as demonstrated by network analysis of Peloponnesian sites on the Peutinger Table, which shows way stations on the northern coast at Korinth, Sikyon, Aegeira (Hyperesia), Aigion, and Patras.30 Although this latter evidence stems from the Roman period, it supports Valderrama's conclusion that a coast road would have permitted land travel at earlier dates, although perhaps not at the interregional level during the Mykenaian period.31 As for water routes, Sherratt has argued that sea traffic along the northwest coast of Achaia increased circa the twelfth century b.c.e. as a result of a northward shift in east-west trading routes.32 There is limited [End Page 11] evidence of an influx of Korinthian pottery to coastal Achaian sites during the late eighth century, indicating east-west coastal trade in the Korinthian gulf.33 Much later, Pausanias (7.26.14) speaks of a nautical route between Hyperesia, (which he there calls Aigeira) and Aristonautai, the port of Pellene at the far eastern border of Achaia. Although Pausanias is much later than Homer, his description of this nautical route nonetheless provides evidence of connectivity between the eastern and more central portions of Achaia. Still late in relation to the Catalogue of Ships, but earlier than Pausanias, Agathemeros (second or first century b.c.e. according to Rizakis) lists Patras as a waypoint on the nautical route between Korinth and Brindisi.34 This again suggests that there were active sea routes along the coast of northern Achaia.
Strikingly, Homer lists the towns of both syntactical groups three and four from west to east. The third syntactical group, governed by the verb εἶχον (574), is composed of Hyperesia (3A), Gonoessa, and Pellene (3C). Homer begins in the west with Hyperesia, which Kirk, following Pausanias (7.26.2), identifies with later Aegeira.35 The location of Gonoessa is unknown, but, given the west to east movement of syntactical groups three and four, we posit that it probably lies between Hyperesia and Pellene. Finally, Pellene is the furthest point east in this syntactical group, uncontroversially located just to the west of Sikyon.
The fourth and final syntactical group also falls on the northern coast of Achaia, to the west of Hyperesia. Once again, Homer narrates west to east, beginning with Aigion (4A) and moving east to conclude the contingent with Helike (4B). The fourth syntactical group also includes a noun that refers to the extent of the coastline: Aigialos (575; not pictured).36 These place names are governed by the verb ἀμφενέμοντο (574).
This test case, then, reveals that, when sites are considered according to syntactical groups, Homer's Mykenaian narrative proceeds in an orderly fashion, from east to west. Each of the four syntactical groups is distributed across an easily recognizable geographical region. In the east, there are two inland travel corridors, the first stretching between Mykenai and Korinth, the second between Orneai and Sikyon. In the western half of the narrative, Homer moves west to east along the coast in two blocks: Hyperesia to Pellene and Aigion to Helike. [End Page 12]
A number of things are notable about Homer's orderly narrative of the Mykenaian contingent. Clearly, the poet's arrangement of sites into syntactical-geographical clusters demonstrates knowledge about the Korinthia and northern Achaia. Moreover, the organization of sites into successive topographical units suggests that the poet uses geography as a spatial mnemonic, visualizing the landscape of the Mykenaian contingent in his mind's eye during composition.37 Minchin has discussed how sequential spatial order is a powerful mnemonic tool. Although Minchin's focus is on the contingent-by-contingent map in particular, her findings are also relevant for the study of the towns within each contingent:
The sequential order of the 'map' directs search in memory. It assists in the recall of the principal regions and the peoples who dwelt there … This is the first level of organization within the structure of the catalogue. Within the broad categories at this first level further information has been nested: individual entries at each level hold further memories at a lower level. Thus higher-order spatial cues prompt other associations … The major geographical or demographical headings of the catalogue (such as the Phokians, the Lokrians, Euboia, or Athens), therefore, cue further lower-order place-names.38
Understanding the Mykenaian contingent as a sequence of syntactical groups allows us to build upon Minchin's discussion of the "nesting" of geospatial cues, providing greater specificity about the poet's process of recall. At the broadest, most general level, the poet understands the Mykenaian contingent as part of a contingent-by-contingent narrative (Level, 1 as articulated in the introduction, above). However, the Mykenaian contingent also exists as its own entity (Level 2), containing its own, subordinate groups of geospatial units (syntactical groups, in this case; Level 3). The Mykenaian contingent may be visualized as a spatially related sequence of four such units. Each unit, in turn, contains even more specific information: the names of each town associated with a given geographical sub-region (Level 4). Thus, the principle that, according to Minchin, allows the poet to remember the Catalogue at the contingent-by-contingent level—the sequential order of the cognitive map—also holds at the local level, but in a more sophisticated way. When visualized merely in their narrative order, the individual towns seem to lack geographical organization because their sequence could not serve as a plausible travel itinerary. As we have seen, however, the narration of toponyms does not lack in geospatial order, because each of the individual towns may [End Page 13] be visualized as part of a syntactical group associated with a particular subregion. These sub-regions are spatially contiguous, providing the sequential order that the poet needs to facilitate recall. Whatever the nature or source of that visualization—i.e., wherever the poet got this information and wherever it falls on a continuum between highly schematic abstraction and realistic approximation based on autopsy—the end result is an orderly narrative of sites that can be mapped onto the realities of local geography.39
case study two: syntactical groups in argos (fig. 3)
[End Page 14]
Homer uses two verbs in narrating the Argos plain and the Akte Peninsula, dividing the contingent into a first syntactical group of seven sites (Argos [1A], Tiryns [1B], Hermione [1C], Asine [1D], Troizen [1E], Eione [1F], Epidauros [1G]) and a second of two sites (Aigina [2A], Mases [2B]).40 With the first verb, εἶχον (559), Homer begins at Argos and then circumnavigates the Akte peninsula. First come Argos and Tiryns, in the Argive plain itself. Continuing to the east, the narrator travels down the southern coast of the peninsula to Hermione and Asine, which share a formulaic phrase: "controlling the deep gulf" (βαθὺν κατὰ κόλπον ἐχούσας, 560). We are then guided back up to the northern side of the Akte penninsula, to Troizen, Eione, and Epidauros.
Having circled around Akte from the west to the east and back again, Homer now uses a second verb (ἔχον 562) to mention the island of Aigina, off the northern coast, and the town of Mases, on the southern coast of Akte, before proceeding to the adjacent kingdom of Agamemnon and the Mykenaians. The locations of all sites in the Argive contingent are relatively uncontroversial, save for Eione. BK does not venture a particular location for Eione, but suggests that it should be located on the Saronic side of the Akte peninsula, rather than the Argolic.41 The site of Eione, notes Kirk, is "unknown, although an ancient conjecture identified it with Methana."42 We therefore place Eione at Methana on our map, although that location is obviously up for debate.
It is immediately clear that the distribution of syntactical groups in the Argive contingent maps neatly onto local geography: the first syntactical group describes a more or less continuous circuit around the Akte peninsula, while the compositional break marked by the introduction of the second syntactical group reflects a geographical break away from the mainland to Aigina, [End Page 15] before the poet returns to Mases, back on the southern coast of the peninsula. It is notable, in addition, that the narration of Argos proceeds very nearly in geographical order. The Argive line-by-line groups accentuate this order: Argos and Tiryns appear together in line 559, in the west-southwest portion of the contingent; line 560, moving southeast, contains Hermione and Asine; in line 561 appear Troizen, Eione, and Epidauros, on the northeastern side of the peninsula; the final line (562) contains Aigina and Mases, corresponding to the most prominent geographic break in the contingent.
Two places where the narration diverges from geographical order are 1) that Mases follows Aigina, after a circuit of the peninsula has already been completed, thus passing over the town's natural place in the order, and 2) that Hermione is named before Asine. While the exceptions to the orderly town-by-town progression in no way negate the overall well-organized nature of the Argive contingent—with syntactical groups reflecting two important geographical divisions of the region, and line-by-line groups marking even more specific sectors—they are nonetheless worth discussing in the context of the Homeric tradition.
As it turns out, Homer may be depending on poetic tradition in both instances of diversion from orderly town-by-town progression in the Argive contingent. An explanation for the positions of Mases and Hermione is suggested by a passage from Hesiod's Catalogue of Women, where Aigina and Mases occur in tandem, as do Hermione and Asine (155. 44–51 Most; 204. 44–51 MW; the text is that of Most).
Αἴας δ' ἐκ Σαλαμῖνος ἀμώ. ητος πολεμ.ι.σ.τὴςμνᾶτο· δίδου δ' ἄρα ἕδνα ἐ. [ο]ι.κότα, θαυματὰ ἔργα·οἳ γὰρ ἔχον Τροιζῆνα καὶ ἀγ[χ]ίαλον Ἐπίδαυροννῆσόν τ' Αἴγιναν Μάση. τά τε κοῦρ.ο.[ι] Ἀχαιῶνκαὶ Μέγαρα σκιόεντα καὶ ὀφρυόεντα Κό.ρ.ινθον,Ἑρμιόνην Ἀσίνην τε παρὲξ ἅλ. α. ν.αιετα.ώσας,τῶν ἔφατ' εἰλίποδάς τε βόας κ[α]ὶ. [ἴ]φ.ι..α. μ.ῆ.λασυνελάσας δώσειν· ἐκέκαστο γὰρ ἔγ.χεϊ μ.α.κρῶι..
Ajax from Salamis, the excellent warrior, wooed; he offered seemly wedding-gifts, marvelous works: for those who possessed Troizen, and Epidaurus by the sea, and the island Aegina and Mases, the Achaean youths, and shadowy Megara and beetling Corinth, Hermione and Asine dwelling beside the sea—he said that their rolling-footed oxen and plump sheep he would drive together and give: for he excelled with the long spear (trans. Most).
Although Hesiod depicts the above locations as subject to Ajax (not Diomedes), he groups the sites as Homer does. Aigina is paired with Mases, [End Page 16] which may help to explain why Mases is not named along with the other sites on the peninsula: it was traditionally associated with Aigina, as the formulaic line of the Hesiodic catalogue suggests.43 The Hesiodic fragment also pairs Hermione and Asine, in that order, followed by a shared phrase: παρὲξ ἅλα ναιεταώσας (dwelling beside the sea). The towns likewise share a formulaic phrase in the Catalogue of Ships, metrically equivalent but different in wording: βαθὺν κατὰ κόλπον ἐχούσας (possessing a deep gulf). The Iliadic formula has provoked various interpretations.44 Most significant, however, is that the Hesiodic fragment, like the Catalogue of Ships, also treats Hermione and Asine as a pair, conjoining them. Moreover, Hesiod names the towns in the same order as does Homer (Hermione-Asine); this traditional order may result in part from metrical ease, since Asine (˘˘–) would be more difficult to place before Hermione (–˘˘–).45 In any case, Argos comes close to an accurate town-by-town progression; the only discrepancies appear traditional.
The toponyms of the Argos contingent, like those of Mykenai, may be visualized as a sequence of nested geospatial cues; this high level of geospatial organization may well have served the poet as a mnemonic device. At the most general level (Level 1 as articulated above), Argos may be visualized as part of the contingent-by-contingent narrative. However, the Argive contingent also exists as its own entity (Level 2), containing its own, subordinate groups of sequential, geospatial units (syntactical groups and line-by-line groups in this case; Level 3). Each unit, in turn, contains even more specific information: the names of each town associated with a given geographical sub-region (Level 4). In this case, the town-by-town narrative proceeds almost in geographical sequence, which may have further facilitated recall.
case study three: syntactical groups in phokis (fig. 4)
Of the nine places mentioned in Phokis, five towns are securely located, Pytho (Delphi), Daulis, Panopeos, Hyampolis, and Lilaia, as is the Kephisos River.46 It is possible to offer plausible suggestions for the locations of the remaining [End Page 17]
sites, Kyparissos, Krisa, and Anemoreia. For Kyparissos and Krisa, we follow McInerney and the Barrington Atlas.47 Anemoreia will be discussed in detail.
At first glance, the disposition of the four syntactical groups in Phokis is both very intuitive and very confusing. It is immediately clear that the narrative of Phokian sites centers in some way around Mount Parnassos, as is especially evident when the place names are considered in their syntactical [End Page 18] groups. Group one, governed by ἔχον (519), is composed of Kyparissos (1A), Pytho (Delphi; 1B), Krisa (1C), Daulis (1D), and Panopeos (1E). These sites all lie to the south of Parnassos, in the Pleistos and eastern Kephisos river valleys. Group two, governed by ἀμφενέμοντο (521) and composed of Anemoreia (2A) and Hyampolis (2B), is problematic. Hyampolis is securely located across the Kephisos on the eastern frontier of Phokian territory, near Eastern Lokris; Anemoreia, on the other hand, though most likely south of Parnassos, is very controversial. The final two syntactical groups have only one place name each. Group three, governed by ἔναιον (522), names the Kephisos River (3A), thereby extending the reach of the narrative further to the north of Parnassos and back towards the west. Group four completes the circuit with Lilaia (ἔχον, 523; 4A), which is said to be at the headwaters of the Kephisos.
So it is clear that the four syntactical groups of the Phokian contingent are to be understood as describing the Phokian territory in a rough circuit around Parnassos: first the sites located along the Sacred Way, then moving into the Kephisos River valley and up the river towards Lilaia at the headwaters. The main difficulty of the Phokian contingent is the location of Anemoreia, which has long been and continues to be contested. This matter is of chief importance for our argument, because its position has the potential to define the nature of the second syntactical group.
Scholars have traditionally assigned Anemoreia to the modern town of Arachova, west of the modern Zimeno Pass in the Pleistos River valley, on the south slope of Parnassos.48 As Rousset has recently argued, however, there is very little evidence to support this identification. Rousset suggests instead that the position of Anemoreia cannot be determined with more specificity than to place it in the Pleistos River valley, somewhere between Delphi and Phlygonion.49 While we do not aim to settle the matter of Anemoreia's location, [End Page 19] a rereading of the evidence in light of Homeric syntactical groups may enable a slightly more specific, although tentative, location.
The matter hangs on the location of an ancient landmark mentioned in a treaty of circa 140 b.c.e. (FD III 2:136.18–33), demarcating the boundary between Delphi, on the one hand, and the Ambryssians and Phlygonians, on the other.50 Here are the remains of that treaty, along with McInerney's translation (with minor modifications underlined).51
τὸ [κα]λ. ούμε[ν]ον ΔΕ. […]ΣΣ[………] […]ΕΩΝ, τὰ δ[ὲ]ε̣[ὐώ]νυμα ὡς ὕδωρ ῥεῖ εἶναι Δ[ελφῶν] ἕω. [ς Α]ἰγωνείας.[Ἀπὸ] δὲ Αἰγωνείας ὡς ὕδωρ ῥεῖ δι[ὰ τ]ῆς [χαρ]άδρας <ἐπὶ> τὸν[λ]όφον τὸν καλούμενον Κέρδωνα ἕ.[ως τῆ]ς ὁδοῦ[τ]ῆς ἐπὶ τὸν πρῖνον φερούσης· τὰ δέξ. [ια εἶναι] Φλυγο-[ν]έων καὶ Ἀμβρυσσέων, τὰ δὲ εὐώνυμα. [εἶναι] Δελφῶν.[Ἀ]πὸ δὲ τῆς ὁδοῦ ἐπὶ τὸμ πρῖνον τὸν ἐπ[ὶ τοῦ] Κατο-πτηρίου· ἀπὸ δὲ τοῦ Κατοπτηρίου εἰς ὀ[ρθὸν κ]ατὰῥάχιν ἐπὶ τὸν Ὀξὺν λίθον· ἀπὸ δὲ τ[οῦ Λίθ]ου [εἰ]-ς ὀρθὸν ἐπὶ τὸν Πέτραχον· ἀπὸ δὲ τοῦ [Πετράχου]εἰς ὀρθὸν ἐπὶ τὸν Παρνασσόν· τούτων. [τὰ μέν]πρὸς ἡλίου ἀνατολὴν εἶναι Φλυγωνέω. [ν καὶ Ἀμ-]βρυσσέων, τὰ δὲ πρὸς ἡλίου δύσιν εἶναι Δε[λφῶν].Τὸ δὲ ὕδωρ τὸ παρὰ τὴν Αἰγώνειαν ῥέον εἶνα[ι]κοινὸν πάντων. Εἰ δέ τινά ἐστιν ἱερὰ ἐν τού-τοις τοῖς τόποις, ὑπάρχειν αὐτὰ κατὰ τὰ ἐξ ἀρχῆ[ς].
… the so-called DE … SS … is to belong to the … while the / territory on the left, as the water flows, is to be the property of Delphi as far as Aigoneia. / From Aigoneia, as the water flows, [the border shall run] through the ravine, towards the / hill called Kerdon as far as the road / bearing towards the holm oak. The territory on the right-hand side belongs to the Phlygonians and Ambryssians, while the territory to the left is the property of Delphi. / [The border shall run] from the road to the holm oak which stands on Lookout / Point. From Lookout Point [it continues] straight on along / the ridge to the sharp rock. From the rock straight / on to Petrachos. From Petrachos / straight on to Parnassos. That which / is towards the east is the property of the Phlygonians and the Am / bryssians, while everything to the west belongs to Delphi. / The water which flows past Aigoneia is / to be common to all. If there are any sanctuaries in / these areas, let them continue to function according to customary usage. [End Page 20]
Of major importance here is the so-called Lookout (Katopterios), which Strabo (9.3.15) notes is in the vicinity of Anemoreia. McInerney contends that the Katopterios should be identified as a peak of Parnassos rising above Arachova, thus allowing the identification of Anemoreia with that modern town.52 Yet the description of the border running up to the Lookout point (Katopterios) does not point exclusively to a position above Arachova, as McInerney interprets it.53 If the torrent that sets the course of the border flows northward from the vicinity of Kirphis, as McInerney himself argues, it is at least as plausible to locate the Katopterios further towards the eastern edge of the Pleistos River valley.54
Indeed, it may be that the unnamed torrent and accompanying ravine here described should be identified with the upper reaches of the modern Platanias River, a tributary of the Kephisos that flows approximately north-west in the vicinity of Ambryssos and Phlygonion on its way from Mount Kirphis.55 If the torrent in question is not the Platanias, it may well have been another roughly north-flowing stream coming off the north face of Mt. Kirphis. In either case, [End Page 21] it would make sense to look for the Katopterios, and therefore Anemoreia, somewhere on the southeastern slopes of Parnassos, by the opening of the Kephisos River valley, perhaps to the west or even north-west of Daulis.56
This tentative location for Anemoreia also has the advantage of helping to make sense of the syntactical groups for the Phokian contingent. According to the tentative positions of the towns followed here, all the sites in syntactical group one fall along the main travel routes to Delphi from the south-west and south-east. That is, all five of the towns in the syntactical group fall along a continuous travel corridor composed of the Sacred Plain and the Sacred Way. The Sacred Way approached Delphi from the east, passing first through Daulis and Panopeos, and then the probable site of Kyparissos near the modern Zimeno Pass, before finally reaching Delphi; but many travelers also approached Delphi from the south-west, via the harbor and plain below Krisa (the modern village of Chrisso), known as the Sacred Plain.57 Thus, the five towns in the first syntactical group are part of a continuous travel corridor running east-west past Delphi, a combination of two common routes providing access to the holy site. The second syntactical group contains two sites, neither of them on the Sacred Way or Sacred Plain: Anemoreia and Hyampolis. With Anemoreia tentatively located on the southeastern slopes on Parnassos near Daulis, these two sites both stand near gaps opening out onto the Kephisos River valley. Group three moves the narrative north and west along the Kephisos river, upstream; with Lilaia, the single member of group four, we reach the headwaters of the Kephisos, on the northern slopes of Parnassos, in the western frontier of Phokis with Doris.58 [End Page 22]
This narrative of Phokis is plausible and well-organized. The poet begins along the Sacred Plain and Sacred Way in syntactical group one, leaves that travel corridor towards the north and east with group two, Anemoreia and Hyampolis, proceeds north and west with group three, the Kephisos, and concludes across Parnassos to the north from Delphi, at Lilaia. As with the contingents in the previous case studies, the toponym list of Phokis nests the individual towns within geographically determined subgroups, four syntactical groups in this case. The poet proceeds through each of these subgroups in succession, suggesting that he visualized this contingent as a series of geographically related units. This cognitive map probably served the poet as a mnemonic device.
case study four: line-by-line groups in krete (fig. 5)
[End Page 23]
The sites in the Kretan contingent are all in the center of the island and may be generally located with a fair degree of certainty.59 Most problematic are Lyktos, Lykastos, and Rhytion.60 None of the Kretan sites is deemed "unlocated" by the directory of the Barrington Atlas [Map 60]. Our placement of the Kretan sites accords with the locations cited there. The seven toponyms of the Kretan contingent all comprise a single syntactical group, governed by εἶχον (646). Within this single syntactical group, however, there are three line-by-line groups (i.e., groups of place names that share a single verse): 1) Knossos (1A) and Gortyn (1B; 646); 2) Lyktos (2A), Miletos (2B), and Lykastos (2C; 647); and 3) Phaistos (3A) and Rhytion (3B; 648). It has been suggested in the past that the Kretan catalogue is jumbled, and that "some of the places are taken out of any logical order."61 By considering the disposition of sites according to line-by-line groups, however, it becomes clear that the poet's narrative is carefully crafted to reflect the different regions of central Krete.
The first Kretan line-by-line group runs down the north-south axis of the island, naming one town north of the Ida massif (Knossos) and one town south of it (Gortyn). The remaining narrative of Kretan sites will accord with this north-south division of the island. Line-by-line group two consists of sites in the northern half of central Krete, above Mounts Ida and Aigaion: Lyktos, Miletos, and Lykastos. The third and final line-by-line group names two towns in the south, Phaistos and Rhytion, both of which share the Mesara Plain with Gortyn. It may be concluded, therefore, that the line-by-line grouping of Kretan place names reflects the realities of local geography, just as syntactical groups have been shown to do. Homer carefully divides central Krete into northern and southern portions, grouping the sites accordingly.
Analysis of the Kretan contingent according to line-by-line groups therefore brings to light an underlying, geospatial order that may have served the poet as a visuospatial mnemonic device. As articulated, this geospatial order consists of four levels of nested geospatial cues. At the most general level (1), Krete is understood as part of the contiguous, contingent-by-contingent sequence. However, the Kretan contingent also exists as its own entity (2), containing its own, subordinate groups of spatially related units (line-by-line groups in this case; 3). Each unit, in turn, contains even more specific information: the names of each town associated with a given geographical sub-region (4). [End Page 24]
case study five: line-by-line groups in aitolia (fig. 6)
The poet lists five sites in the Aitolian contingent, all in a single syntactical group, governed by ἐνέμοντο (639). These place names fall into two line-by-line groups: 1) Pleuron (1A), Olenos (1B), and Pylene (1C; 639), and 2) Chalkis (2A) and Kalydon (2B; 640). Of these, there is controversy about the locations of Olenos and Pylene. The latter is tentatively placed by the Barrington at modern Magoula, also cautiously accepted by Hope Simpson and Lazenby (HSL).62 This is more or less in agreement with Kirk's assessment that Pylene was "probably in the same general region" as Pleuron and Olenos.63 Strabo (10.2.6) says that Olenos was near New Pleuron, but was destroyed by the Aitolians. Since New Pleuron was north and slightly west of Old Pleuron, that would suggest that Strabo's location for Olenos is also north-west of Old Pleuron. HSL, on the other hand, doubt that Olenos and Pleuron would be so close together, and propose instead that Olenos may be [End Page 25] found at Agios Ilias, on the opposite side of the Aitolic Lagoon from Pleuron.64 We follow Strabo in our placement of Olenos, because that location seems to fit the narrative model of the contingent better than the other option. This reasoning is, admittedly, quite speculative, but the logic is as follows. With Olenos and Pylene located as detailed above, then the two line-by-line groups follow a clear pattern: group one (639) begins at Pleuron on the south-west side of Mt. Arakynthos and proceeds straightforwardly up the coast, to the north-west, through Olenos, followed by Pylene. Line-by-line group two (640) begins to the south-east of Mt. Arakynthos with Chalkis, and then, like group one, moves up the coast to the north-west, ending at Kalydon in a similar narrative structure as seen in the fourth and fifth syntactical groups of the Mykenaian contingent.
Most important for our purposes, however, is the following conclusion: wherever Olenos is tentatively located—whether HSL are correct in placing it at Agios Ilias on the opposite side of the Aitolic Lagoon, or if we are correct in putting it between Pleuron and Pylene—the Aitolian contingent breaks down into two line-by-line groups that appear to respond to a local topographical feature. Line-by-line group one (639) contains the three coastal sites to the south-west of Mt. Arakynthos, and its narrative tends to the north-west; line-by-line group two (640) contains the two coastal sites to the south-east of Mt. Arakynthos, and its narrative also tends to the north-west. As was the case with the previous case studies, therefore, the poet subdivides the Aitolian contingent into organized, geospatial sub-regions (comprised of two line-by-line groups in this case) that distribute toponyms according to local land features. The poet shows more general geospatial knowledge (contingent-by-contingent level), while triggering recollection of increasingly specific knowledge (line-by-line and town-by-town levels).
case study six: line-by-line groups in lokris (fig. 7)
The poet lists eight place names in the territories of the Eastern Lokrians, although most of these are problematic. Despite the difficulty of finding precise locations for some towns, however, enough information regarding their general position remains to establish that the poet is clustering the Lokrian sites into line-by-line groups, constituted according to shared geographical features. The Lokrian contingent, like the Kretan and the Aitolian, contains only a single syntactical group, governed by ἐνέμοντο (531). Within this single syntactical group, however, there are three line-by-line groups: 1) [End Page 26]
Kynos (1A), Opous (1B), and Kalliaros (1C; 531); 2) Bessa (2A), Skarphe (2B), and Augeiai (2C; 532); and 3) Tarphe (3A) and Thronion (3B; 533). These three line-by-line groups, we suggest, correspond to three geographical sub-regions within Lokris.
The first two locations in line-by-line group one are the most straightforward of the contingent. Scholars agree that Kynos is to be placed on the hill called Pyrgos, on the coast northeast of modern Livanates.65 Opous is most probably located at modern Atalanti, at the western extreme of the Atalanti plain.66 If Opous was not at Atalanti, it was at modern Kiparissi, at the southern extreme of the Atalanti plain.67 The precise location of Kalliaros is unknown, but modern scholars put it somewhere in the plain of Atalanti, for reasons well summed up by Kramer-Hajos:68
Kalliaros and Bessa were both unknown already to Strabo, who places Kalliaros in a plain (probably based on no more than the name of the town, taking Kalliaros [End Page 27] as from καλός, beautiful, and ἀρόω, to till) … (Strabo 9.4.5). It is likely that Kalliaros is to be sought in or near the plain of Atalanti: this would provide a logical itinerary after Kynos and Opous, and the D-Scholia (on Hom. Il. 2.531) make Kalliaros the son of Opous, which suggests that Kalliaros was located in the same general area as Opous. All evidence is however circumstantial.
The evidence is incomplete, but we follow Barrington and other commentators in placing Kalliaros somewhere in the plain of Atalanti, perhaps at Skala Atalantis or Kiparissi.69 After all, as Dakoronia notes, Kalliaros probably lay on a plain, and "the only plain in this part of Greece deserving of this description is the plain of Atalante."70 All three of the towns in the first line-by-line group, therefore, fall somewhere on or around the plain of Atalanti, in the eastern portion of Homer's narrative of Lokris. The cluster of sites in line-by-line group one, in other words, appears to reflect a feature of the local landscape.
Of the names in the second line-by-line group (Bessa, Skarphe, and Augeiai ), Bessa poses the most difficulty. Its location is unknown to Strabo (9.4.5), who argues only that, based on etymology, it must have been in a wooded glen (βῆσσα = wooded combe, glen [LSJ]).71 There is, then, very little to go on. Dakoronia proposed the modern site of Roustiana, in a mountainous, wooded region, west-northwest of Livanates, where there is a previously undiscovered acropolis containing Mykenaian pottery.72 Skarphe, the second term in the second Lokrian line-by-line group, "was presumably at the site of the later Skarpheia somewhere near the southern shore of the Malian gulf."73 Pascual and Papakonstantinou (PP) put the town one kilometer southeast of modern Molos, at Trochala.74 The location of Skarphe, in turn, may provide us with an approximate location for Augeiai, which Strabo claimed was in Skarpheian territory.75 This scenario, if trusted, would have important implications for the second Lokrian line-by-line group. If Bessa is [End Page 28] at or near Roustiana and Augeiai is somewhere near Skarphe, then line-byline group two proceeds according to a clear geographic rationale: whereas line-by-line group one was contained by the Atalanti plain in the far eastern portion of the contingent, line-by-line group two constitutes the northern coast of the contingent, tending east-west along the southern coasts of the Northern Euboian and Malian Gulfs.76
The precise locations of the sites in the third line-by-line group (Tarphe, Thronion ) are likewise difficult. PP argue convincingly that Tarphe should probably be identified with modern Trikorfo, about 2.5 km west of Thronion, on the opposite side of the Boagrios.77 Kirk follows HSL in proposing modern Pikraki for ancient Thronion; Barrington and PP agree.78 For our purposes, however, the exact positions of Tarphe and Thronion are not of foremost importance, because Homer provides us with a concrete bit of detail about their locations: both are "beside the flows of the Boagrios."79 The ancient Boagrios River is also known as the modern Platanias or Xerias.80 Since both Tarphe and Thronion are located along the Boagrios, then the third line-by-line group yet again clearly reflects a local geographical reality: the sites in line-by-line group three are selected for their shared proximity to the Boagrios river.
The Lokrian contingent, although full of difficulties, provides enough information to strongly suggest that Homer narrates the contingent in clusters of toponyms grouped according to shared geographical traits. Line-by-line group one features sites at the far east of the contingent, in or about the Atalanti plain; line-by-line group two stretches east-west along the northern coast of the contingent; line-by-line group three contains sites along the Boagrios River, in the contingent's west. The Lokrian contingent, like those in the other case studies, is probably organized into a series of nested geospatial cues, with more general knowledge triggering recollection at the line-by-line and town-by-town levels. [End Page 29]
case study seven: line-by-line groups in the koan contingent (fig. 8)
The final stop on Homer's tour of the Aegean is the Koan contingent, a group of islands that all fall along a north-south axis. The contingent contains five place names, all listed, once again, within a single syntactical group, governed by εἶχον (676). The poet enumerates the places in two line-by-line groups: 1) Nisyros (1A), Krapathos (1B), and Kasos (1C; 676); 2) Kos (2A) and Kalydnai (2B; 677). Only Kalydnai is at all controversial, the question being whether by that name the poet intended only Kalymnos, or if he meant to include one or more of the nearby islands. For our purposes, that controversy is not of the utmost importance. Most relevant, instead, is the poet's division of the islands into two groups, south and north, according to line-by-line groups. [End Page 30]
The first site mentioned is the relatively insignificant island of Nisyros, which has rightly led commentators to wonder at its placement in the contingent: "The placing of the small and obscure Nisuros at the head of this island group is perhaps fortuitous, or perhaps determined by metre or even alliteration (to keep the four places with initial kappa together)."81 Certainly, these suggestions are all possible, but there is a good geographical explanation as well. Nisyros is directly to the west of Syme, the only toponym in the preceding contingent. The poet's contiguous movement directly to the west towards Nisyros effectively divides the sites into two sections, north and south (cf. Krete and Euboia, two islands whose contingents appear to be divided between northern and southern sections). These geographical divisions are mirrored in the poet's line divisions: in verse 676 come the islands to the south, Nisyros, Krapathos, and Kasos; in verse 677 come the islands to the north, Kos and the Kalydnai.82 As in the other case studies, the poet's tidy organization of the contingent into clusters of toponyms corresponding to geographical sub-regions probably served as a mnemonic device.
case study eight: syntactical and line-by-line groups in euboia (fig. 9)
[End Page 31]
The Euboian contingent is composed of one syntactical group and one line-by-line group. The syntactical group contains five sites, Chalkis (1A), Eretria (1B), Histiaia (1C), Kerinthos (1D), and Dion (1E), all of which are governed by the verb ἔχον (536). The line-by-line group includes two sites, Karystos (2A) and Styra (2B), both in line 539.83 The poet begins in the center of the island with Chalkis and Eretria on the west coast, the two most important sites on Euboia, then moves to the north with Histiaia, Kerinthos, and Dion (syntactical group); he then proceeds to the southern portion of the island, ending his narrative of Euboia with Karystos and Styra (line-by-line group). The southern endpoint is fitting, because it points the way towards Homer's next stop, Athens.84
Most importantly for our purposes, the two Euboian subgroups correspond to a north-south division of the island. The five sites in the syntactical group are all on the coasts of the northern and central portions of the island, while the line-by-line group falls on the southern part, which is separated from the rest by the narrow neck of land around modern Limni Dystos. The south's relative lack of settlement and barrenness also distinguish it from the rest of Euboia.85 Once again, therefore, the distribution of toponyms according to syntactical and line-by-line groups reflects the realities of the local landscape that serve poetic memory.
initial conclusions and speculations
These case studies demonstrate that the poet incorporates knowledge of local geography into a town-by-town description of a number of contingents. The poet's geographical knowledge especially comes to light when the distribution of place names is considered from the point of view of syntactical and line-by-line groups. The disposition of these units of toponyms frequently reflects local topographic features, such as travel routes, mountains, valleys, peninsulas, coasts, and waterways. We suggest, in addition, that an oral poet would have used such topographic knowledge as a spatial mnemonic device, allowing the poet to visualize (even in a highly abstract manner) and thereby remember the large number of sites to be recited. [End Page 32]
The potential source of the poet's knowledge—whether it originated in certain cases from autopsy, whether it was part of a Pan-Hellenic poetic tradition, or whether poets stitched it together from a series of local traditions—is impossible to determine with any precision. Nonetheless, the following general remarks may be offered. If Visser and Kirk are correct, then the language of the Catalogue bears the same hallmarks of oral composition as the rest of the Iliad and the Catalogue was therefore likely composed as an integral part of the whole.86 From this it follows that, as McInerney has concluded, "the Catalogue, judged solely in terms of composition, is a product of the eighth century."87 While the poem may be an eighth century composition, it includes historical and linguistic material that may be traced back to the Mykenaian age. As such, it would be reasonable to expect that the Catalogue, like the rest of the Iliad, contains an amalgam of elements from different historical periods in the hands of an eighth century poet. Indeed, as scholars like Kirk and Dickinson have discussed, the Catalogue cannot be understood simply as a Mykenaian document, since the poet gets much wrong about the Mykenaian world.88 At the same time, the Catalogue does clearly include material that is probably Mykenaian in origin. Nor is the Catalogue, as has been argued, for example, by Giovannini, a reflection of iron or even archaic age Greece, contemporary with the poet, as the inclusion of many Mykenaian sites and the omission of important geometric and archaic sites suggests.89 In other words, there is no one simple answer to the question of the origins or date of the geographical knowledge displayed by the Catalogue of Ships.
Yet our study seeks to advance the discussion of the poet's geographical knowledge. First of all, the poet displays a strikingly detailed and concrete knowledge of local geography in the above case studies, wherever and however that knowledge was obtained, and regardless of the time to which that knowledge best corresponds. Moreover, if, as we suggest, syntactical and line-by-line groups were tools used by the poet to remember troves of geographical information, this appears to be a means by which the poetic tradition could preserve detailed knowledge of a regional landscape over many generations. It seems plausible, therefore, that such detailed geographical information [End Page 33] could have been part of a living oral tradition, especially since the language of the Catalogue suggests it was composed as part of the Iliad as we know it.90
This article does not intend to be a complete appraisal of Homer's knowledge and use of geography in the Catalogue of Ships. Yet it is our hope that the methods detailed here may be useful to future research on the places named in the Catalogue. We have already suggested several instances in which the organizational techniques pointed out here may potentially be of help in locating controversial sites. The patterns modeled in our analysis may point the way forward in other regions as well, and we will look to extend our findings in the near future, especially in the central, southern, and western portions of the Peloponnese, the Ionian islands, and Thessaly, regions not discussed here. These patterns may prove helpful not only in the analysis of other regions [End Page 34] based on their similarities to the findings here, but also based on potential differences. One region where thinking about divergences from Homer's neat organization of contingents into geographical clusters has been particularly fruitful is Boiotia, to which we now turn in the final portion of this paper.
boiotia and an exceptional use of syntactical groups (fig. 10)
Even a cursory look at the diagram for the Boiotian contingent reveals that geographical-syntactical clusters do not operate in that region as in the eight case studies presented above. The poet uses neither syntactical nor line-by-line groups in any easily perceptible way to organize the landscape narrative. Instead of proceeding in units grouping place names according to travel corridors or topographical features (like Mykenai or Krete) or in an order that approximates an itinerary (like the Argive contingent, which also clearly corresponds to the topography of the Akte peninsula), the Boiotian sites appear sprinkled all over the map. The Boiotian contingent serves, then, as an exception that proves the rule: in light of the eight previous examples of the poet's geospatial orderliness, the apparent disorder of the Catalogue's first contingent demands explanation. We argue here that, in fact, syntactical groups do play an important role in the Boiotian contingent, but that their role is not the same. In the other eight contingents, syntactical and line-by-line [End Page 35] groups are clusters of sites grouped according to local topography. In Boiotia, however, syntactical groups function differently: they reflect the poet's mental topography of the unnamed city of Thebes and its seven famous gates.
Previous commentators have sought to impose a rough order on the list of place names. Visser, for example, has suggested that the poet's narrative accords with the realities of Boiotian geography in at least one way: he proposes that the poet divides the realm into east, west, and central portions reflecting the division between the region's eastern and western plains, as well as the historical, mythical, and religious importance of Thebes and Onchestos toward the center. The narrative order of places would therefore seem to group towns according to their positions near these two plains, indicating that the poet does indeed have some knowledge about the relative placement of sites.91 This geographical division may reflect the political realities of Bronze Age Boiotia in one way or another.92 But the extent to which the poetic-geographical organization of Boiotia differs from the methods outlined in our eight case studies is quite striking.
This difference in geographical organization is certainly pointed, but hardly the only way in which Boiotia stands out. There is, first and foremost, the well-known question of priority.93 Why should the Boiotians be first, if they are featured relatively little in the remainder of the Iliad, and what might that apparent discrepancy in treatment tell us about the relationship between the Catalogue of Ships and the rest of the poem? Boiotia is also exceptional with regard to the sheer number of place-names it contains: twenty-nine, more than twice as many sites as any other contingent. Why such an array of sites for a contingent that is, again, relatively unimportant elsewhere in the Iliad? Both of these questions may potentially be resolved by reference to the broader legend of the Trojan War: it was at Aulis that, with the allies assembled, Agamemnon was delayed in his expedition, and where Iphigenia met her fate.94
Yet it may be, too, that the Boiotian catalogue, which feels out of place in the Iliad despite the importance of Aulis, belonged originally to an oral tradition that centered on Boiotia in particular. The existence of a Boiotian tradition of catalogue poetry that was subsequently borrowed or absorbed by the poet of the Iliad might explain why that contingent finds itself so [End Page 36] abundantly treated in the Catalogue of Ships. Scholars have proposed such an explanation before, encouraged especially by a special connection, real or imagined, between Boiotia and catalogue poetry.95
There is yet another peculiar feature about the enumeration of place names in the Boiotian contingent: it makes no mention of Thebes, clearly the most important town in the region in history and legend. Homer does name Hypothebes, a settlement variously located directly below the Kadmeia or as far away as Potnia, but never the city of Thebes itself.96 Why should a contingent with so many place names make no mention of the most important one?97 The absence of Thebes itself from the enumeration is even more confounding when one notes that the Iliad poet clearly knows of Thebes and names the town in other contexts (e.g., Il. 4.378; 14.323; 19.99). One possibility is that Homer excludes Thebes from his enumeration of sites because he knows that, by the time of the expedition to Troy, Thebes had been sacked by the Epigonoi and therefore ceased to exist from the standpoint of legend, a knowledge he does indeed demonstrate in Book 4 (4.376–410).98 In that case, the poet would be adjusting his catalogue to fit with the legendary timeline. Although this explanation may account for the absence of Thebes from the Boiotian contingent, it does not account for the unusual copiousness of place names for the region. In any case, if Homer is indeed deliberately excluding Thebes from the Catalogue on the basis of its previous destruction in accordance with the mythological timeline, we might have expected the poet to acknowledge that omission in some way.99
Another potential solution suggests itself in light of the geographical organization of the Boiotian contingent. Instead of using syntactical or line-by-line groups to create an orderly narrative of sites grouped according to geographical sub-regions, the poet may base his enumeration of Boiotian towns on an alternate method of geo-poetic organization. The Boiotian place names are [End Page 37] all disposed around an unnamed central point, which is the unacknowledged focus of the narrative: Thebes (fig. 10). In other words, the poet's treatment of the Boiotian catalogue is akin to a teichoskopia. It is chiefly for this reason that Thebes goes unnamed: the poet is working in a tradition that visualizes itself on the Theban walls, listing, or even pointing towards, sites around or at a distance from the city's gates.
Although the Catalogue of Ships does not contain any other examples of such narratives around a central point, this geo-poetic model is found with relative frequency elsewhere in the tradition. Most obviously, there is the famous teichoskopia in Book 3 of the Iliad, where Helen looks down from the Trojan wall at the Greek forces arrayed on the battlefield and describes a number of the generals to Priam. Homer does not, in the teichoskopia of Book 3, list the generals together with their places of origin, because the immediate organizational scheme is their position on the field. More importantly the Trojan Catalogue, which follows directly upon the Catalogue of Ships in Book 2, is yet another example of listing sites from the perspective of a central point. "There," as we have said elsewhere,
Homer enumerates the Trojan allies, prefacing his list with a description of a hill on the battlefield (Il. 2.811–815). This high vantage point offers the poet a viewing place akin to a teichoskopia, from which the Trojans and their allies could be distinguished. In this case, Homer does list the towns from which the battalions came, as he had done earlier for the Greeks. When plotted on a map, as in Boiotia, the listed towns radiate out from the central unnamed city of Troy.100
Also reminiscent of this narrative technique is the Homeric Hymn to Apollo, where the sites Leto visits during her wanderings are enumerated as a wellorganized circuit around the central point of Delos.101 The teichoskopic perspective, then, constitutes a geospatial narrative with a focus on the center and is a fairly common narrative technique in the Homeric corpus.102
Given the several unusual characteristics of the Boiotian contingent, not least its large number of place names and striking teichoskopic narrative style, it is tempting to search for a predecessor in the oral tradition that could have influenced the poet of the Iliad. We do not mean to suggest that the Boiotian catalogue was not composed by our poet as an integral piece of the Catalogue [End Page 38] of Ships, but that its unusual features, which set it apart from the other contingents, indicate that it may have developed in a tradition of its own. Homer thus would have been aware of the tradition that gave rise to the Boiotian catalogue, which he then appropriated for use in the Iliad.
In fact, a plausible, if speculative, context for the development of a teichoskopic catalogue of Boiotia is not difficult to imagine: the pre-Iliadic oral Thebaid. Tsagalis has recently provided an overview of the wide range of connections between the Theban and Trojan epic traditions as well as the main issues relevant to their study.103 Torres-Guerra has argued that Homer knew of Theban sagas, as can be demonstrated from the text of the Iliad.104 It is clear from Book Four that Homer is aware of traditions about the Seven Against Thebes and the Epigonoi (4.376–410); the same passage mentions the famous seven gates (406). Pache has discussed how the teichoskopia as well as the construction and destruction of the Achaian wall are points at which the poet of the Iliad sets himself in competition with wall-motifs from the Theban tradition.105 Singor has argued, moreover, that the Achaian wall of the Iliad, which he claims has seven gates, is probably based upon material from the oral Thebaid.106 It is possible, therefore, that Homer was aware of a Theban teichoskopia from a traditional Thebaid.
The large number of place names in the Iliadic Boiotian catalogue may be based upon just such a set piece, where the Theban allies from across the region would have been enumerated by a narrator on the wall of Thebes, looking out at the disposition of troops at the seven gates of the city.107 It is striking that several later treatments of the Seven Against Thebes share a teichoskopic [End Page 39] perspective. A scout reports the position of the enemy at each of the gates in the Seven of Aeschylus, in a dramatic teichoskopic narrative lasting several hundred lines (375ff.). Book 7 of Statius's Thebaid, moreover, offers a teichoskopic catalogue of Boiotian allies enumerated in seven contingents.108 Perhaps these subsequent Theban teichoskopiai evoke a no longer extant Theban version.
It may be that the enumeration of sites in the Theban contingent of the Catalogue of Ships was influenced by an oral Thebaid not only in its narrative focus on Thebes, unnamed in the Iliad, but in its syntax, as well. The syntactical groups of the Boiotian contingent do not reflect local topography, as they do in other contingents. But they may serve another function here. The Boiotian contingent, as it turns out, contains seven syntactical groups, one for every gate of seven-gated Thebes.109 Each of these syntactical groups, therefore, may correspond to a contingent placed at each of the seven gates in the teichoskopia of the oral Thebaid. The teichoskopic vantage of the Boiotian catalogue, borrowed from the Thebaid, accounts for its radically different geospatial framework and exceptional use of syntactical groups. This Theban perspective may have served Homer as a spatial mnemonic, allowing him to visualize the towns of Boiotia in seven groups corresponding to the seven gates. In the case of Boiotia, too, then, traditional, geospatial material may be influencing the poet of the Iliad and facilitating the poetic performance.
Homer's town-by-town narratives of the eight case studies above are organized according to the realities of local geography. As we show, the orderly arrangement of these individual contingents becomes apparent when their place names are interpreted in light of two previously unnoticed units of compositional thought: syntactical and line-by-line groups, which frequently map onto local topographic features. In addition, the line-by-line groups tend to be used in contingents that contain only a single syntactical group, at least as far as the admittedly limited sample size of this study indicates (Euboia is an exception). We contend, moreover, that the high degree of geospatial order apparent in these contingent narratives makes it very likely that the poet [End Page 40] employed geography, even at a local level, as a spatial mnemonic device. The relative ease with which this well-organized material could be remembered, in turn, suggests its potential to exist as part of a body of oral-traditional geographical knowledge, perhaps inherited as part of a pre-Homeric tradition available to the Iliad poet. In the case of the Boiotian catalogue, the poet could likewise have employed syntactical groups, but in a different way. Instead of mapping onto features of local topography, the syntactical groups of the Boiotian contingent may correspond to seven groups of allies, each one located at one of the famous seven gates of Thebes. If the teichoskopic vision of the Boiotian catalogue is correct, it may point to origins in an oral-traditional Thebaid. Even this exceptional case, therefore, is part of a well-wrought geospatial plan behind Homer's masterpiece of memory.
* The authors thank the editor of TAPA, Andromache Karanika, as well as the anonymous reviewers, for their many helpful comments. We thank Edith Nally for creating the images used here and for many useful conversations, especially about Boiotia. We also wish to thank the present and former faculty of the Scholars' Lab at the University of Virginia for allowing us to visualize and refine the theories laid out here. In particular, we are grateful to Bethany Nowviskie, director emerita of the Scholars' Lab, as well as to Jeremy Boggs, David McClure, Wayne Graham, Chris Gist, and Kelly Johnston. Our work was funded by two Buckner W. Clay Grants from the Institute of the Humanities and Global Cultures (UVA), a grant from the Center for International Studies (UVA), a grant from the Vice President for Research (UVA), a travel grant from the Scholars' Lab, and a travel grant from Topoi at the Freie Universität Berlin. All images were initially created using the mapping tool Neatline (neatline.org). Figures 2-10 utilize the Pelagios Digital Map of the Roman Empire as their base-layer (http://pelagios.org/maps/greco-roman/about.html).
1. Latacz et al. 2010: 146. Latacz et al. is hereafter referred to as BK. The Catalogue, of course, is not only focused on the geography of Greece, but also its people and myth. On the former, see Heiden 2008; on the latter, see Visser 1997 passim and 746–50.
2. We use Munro and Allen's OCT edition for the text of Homer. Translations are our own unless marked otherwise.
7. Kirk 1985: 183. See also Irby 2012: 86: "Within each region, Homer proceeds only roughly in a circumnavigational fashion, ordering place-names according to the demands of his poetic rhythm rather than the practical considerations of a journey." While Irby is certainly correct to point out that contingent narratives are not designed as practical guides for travel, many nonetheless prove to be far more organized than Irby implies, despite moving "roughly in a circumnavigational fashion" at times.
14. Minchin 2001: 84–87. See also 79–84 on the catalogues and memory more generally. Minchin 2008 is a valuable resource on the topic of the Iliad and spatial memory; but her treatment of the Catalogue of Ships there summarizes the findings of Minchin 2001.
22. If Homer had been a prose writer, he might well have opted to transmit his geographical knowledge in more straightforward itineraries. Metrical constraints may be one reason that the poet does not do so. A possible example is mentioned in the discussion of the Argive contingent. There may be others.
23. Some of the arguments in this case study are briefly summarized in Evans and Jasnow 2014: 318. Likewise, a brief summary of arguments pertaining to the Mykenaian and Boiotian catalogues may be found at our website, http://ships.lib.virginia.edu/.
24. Unless otherwise noted, the placement of all toponyms follows the locations in the Barrington Atlas of the Greek and Roman World and may be confirmed there or in the online gazetteer Pleiades (http://pleiades.stoa.org/). Only difficult or controversial locations will be discussed in detail, as well as those not listed by Barrington.
28. See the helpful map in the article of Y. Lafond, Brill's New Pauly, s.v. Achaeans, Achaea.
39. For an overview of sources on memory function, see Minchin 2001: 86n31–33 and Minchin 2008: 9–13. Minchin 2008 evaluates the use of spatial memory in the composition of the Iliad and draws comparisons with living oral-poetic traditions, including examples that map out sequences of geographical landmarks. For sources and an overview of mnemonics as they relate to locations in space and oral tradition, see Clay 2011: 111–19, with notes 40–46. On the visuospatial aspects of oral tradition, see Rubin 1995: 46–47, 50–51, 60–63 and Zeman 2016: 168–69. For a broad overview of space and epic, see Skempis and Ziogas 2014: 1–8. Lateiner 2014 offers a wide-ranging survey of space and place in Homer. On the way that nesting spatial information within a hierarchy assists recall, see Neisser 1988: 369–70 and Paivio 2014: 436. On the hierarchical, nested nature of spatial memory and memory more generally, see Neisser 1989: 75–76 and Paivio 2014: 39–40. On the hierarchical nature of memory, and on the way that organization facilitates recall, see Mandler 1967: 366, 370–71 and Paivio 1979: 213–14. For the "cognitive map" as a schema that serves to stimulate recall, see Neisser 1989: 76–77; see also Ryan 2003. On imagery and concreteness assisting memory, see Paivio 2014: 22–24, 66–69, 436. For an overview of historical mnemonic practices, including ancient traditions and their empirical grounding, see Paivio 2014: 18–24. See also Yates 1966: 17–26. For a very useful overview of issues relating to mental imagery, see Thomas 2017.
41. BK: 181.
43. See Finkelberg 1988 for a comparison of Ajax's geographical realm in the Catalogue of Women with the Argive realm in the Catalogue of Ships. Finkelberg 1988: 35 demonstrates that Ajax's catalogue in Hesiod "is a genuine traditional version independent" of the Catalogue of Ships.
47. McInerney 1999: Appendix 1; Barrington Atlas (Map 55). Rousset 2002: 157 does not accept the position of Anemoreia at Arachova, arguing that its position cannot be determined with more specificity than to place it in the Pleistos River valley, somewhere between Delphi and Phlygonion.
48. For discussion and summary of evidence, see McInerney 1999: Appendix 1.18 and Rousset 2002: 29–30, 63n73, 155–57. See also Oulhen's article in Hansen and Nielsen 2004: 404 for an overview of the issues relating to the toponym.
49. Rousset 2002: 157 takes issue with McInerney and others who place Anemoreia at Arachova. Note, as well, that the location of Phlygonion is itself up for debate. Rousset 2002: 45 suggests a position not far to the east of modern Arachova, at modern Bania, while McInerney (Map 7 and Appendix 1.21) locates it to the south and west of Ambryssos, near modern Stiri. Barrington (Map 55) tentatively suggests a location just to the south west of the Phokikon, at modern Ano Tseresi. The location of Phlygonion is, therefore, contested, to say the least. It may well have been located further east than Rousset allows.
53. The stream that serves as the border in the beginning of the document is probably flowing from the vicinity of Mount Kirphis, along a roughly north-south axis, since the land to the left (west) is to belong to the Delphians, while the land to the right (east) is to belong to the Ambryssians and Phlygonians (the inscription makes it clear that left and right are according to the perspective of one facing downstream). We may add that the stream is probably toward the eastern portion of Kirphis, since it will be serving as a border between Delphi and towns to its east. The border continues on its roughly northward bearing from the eastern portion of Kirphis, until an unknown location by the name of Aigoneia. From there, still following the course of the stream, it proceeds to an unknown hill by the name of Kerdon, and then to a road that leads towards a Holm Oak, maintaining all the while the west-east division of the territories. It then continues from the road to the Lookout, which Strabo claims was a cliff extending from Parnassos. From the Lookout, the border continues up a ridge, ascending further up Parnassos.
54. McInerney 1999: 76–77. McInerney contends that the treaty begins by identifying the border with a north flowing torrent (line 19 of the treaty, "as the water flows"), but then, after Aigoneia, switches course and turns westward down a different torrential ravine (line 20 of the treaty, "as the water flows"), which he identifies as the Pleistos river valley. However, there is no need to imagine that the border jags west after reaching Aigoneia, instead, we may suppose that it continues its general northward tendency, more or less following in the direction of the torrent it began with. Rousset 2002: 155 and fig. 5 also understands the border to run north-south.
55. At least if one follows the placement of Phlygonion as construed by Barrington's or McInerney 1999. This would be a convenient choice, since the treaty at hand is precisely between Delphi and the Ambryssians and Phlygonians.
56. The possibility that the Katopterios may be a more easterly projection of Parnassos, and that Anemoreia is therefore further to the east than previously allowed, seems to gain support from Strabo's description of eastern Phokis. The geographer clearly conceives of Anemoreia as one of the towns in and about the Kephisos valley, where he groups it among the following sites and in the following order: Daulis, Panopeus, Anemoreia, Hyampolis, Parapotamia.
58. The above suggestion about Anemoreia's location is one possible way to account for Homer's second syntactical group in the Phokian contingent. Although we contend that the above grouping of sites constitutes a well-organized narrative of Phokis that demonstrates Homer's knowledge of the region, it may nonetheless be noted that Hyampolis is strikingly more to the east than the rest of the sites named in the contingent. However, Hyampolis's similarity in name to another Phokian toponym leads us to speculate—and speculate is the operative word here—that another place name may originally have been grouped with Anemoreia at an earlier point in the tradition: Hyampeia. See Oulhen's article in Hansen and Nielsen 2004: 405 for basic information about Hyampeia. This name designates the more easterly of the two peaks rising over the Kastalian Spring at Delphi (RE s.v. Hyampeia [Felix Bölte]). Paired with Anemoreia, this would make for a tidy second syntactical group indeed. Could it be that, at an earlier layer of the tradition, Hyampeia was named together with Anemoreia, and it was only at a later point Hyampolis took its place? In that case, the second group would recount two locales high above the sacred road. This admittedly speculative suggestion gains some plausibility based not only on the similarity of the two toponyms, but also from the fact that Hyampeia could replace Hyampolis in line 521 without any other changes to the verse. It is interesting to note, as well, that Pausanias followed a south-north path over Parnassos to Lilaia (10.33.3–5), not too different from the imaginary path of the second syntactical group if Hyampeia did replace Hyampolis.
59. BK: 209.
62. HSL: 108.
64. HSL 107–8.
69. Our position is that noted as the approximate location in Barrington (Map 55). See also Fossey 1990: 75, who puts Kalliaros at Skala Atalanis, and Dakoronia 1993: 120–24, who puts it at Kiparissi.
74. PP: 109–10. This location is just southwest of the approximate location for Skarphe as listed by Barrington (Map 55).
75. Strabo 9.4.5 as emended by Meineke in his 1852 Teubner edition; HSL: 49.
76. PP: 128–30 suggest a location near Augeiai and Skarphe for Bessa, because Strabo mentions it in his description just before he mentions Augeiai (9.4.5). Just before Strabo mentions Bessa, however, he mentions Kalliaros, which was on the Atalanti plain. By the logic of PP, therefore, the placement of Bessa at modern Roustiana works just as well. Even if PP are correct that Bessa was near Skarphe, however, it is worth noting that the second line-by-line group would still make a well-defined geographical unit of towns on the north shore of Epiknemidian Lokris, by the mouth of the Boagrios River.
77. PP: 147–48.
79. For this reading of 2.533, see also PP: 139, 148 and Homer Encyclopedia s.v. Boagrios.
80. PP: 133–37.
83. Barrington (Map 55) notes uncertainty about the identity of the ancient settlement at Dion. Despite the uncertainty, BK: 172–73, Kirk 1985: 204, and HSL: 51–55 all place Dion in the far northwest of Euboia, close to Cape Kenaion.
90. Knowledge of travel routes or regional geography has influenced Homeric poetry in numerous locations. Hera's route from Olympos to Ida by way of Pieria, Emathia, Thrace, Mt. Athos, Lemnos, Imbros, and Lekton is described at Il. 14.225–30 and 280–85. See Janko in Kirk et al., vol. 4, 186–87. Similarly the poet describes Telemachos's route from Ithaka to Sparta by way of Pherai in Od. 3–4 (Od. 3.1–4: Ithaka-Pylos in one night; 3.485–4.1: Pylos-Pherai, Pherai-Sparta in two days). See Heubeck, West, and Hainsworth 1988 Vol. 1: 64–65, 159–60; Mark 2005: 170. Telemachos's return itinerary is more fully elaborated, coasting from Pylos past the streams of Krounoi and Chalkis toward Pheai, past Elis, and back toward Ithaka by way of the Sharp Islands (Od. 15.295–300).
Various Homeric Hymns also recount geographical details more or less in the form of itineraries. In the Hymn to Demeter, the goddess, in disguise, recounts a fictional voyage on a ship from Krete to Thorikos, in southern Attika, and then to Eleusis after a period of wandering (123–34; Seaford 2012: 27–28; Richardson 1974: ad loc.). The Homeric Hymn to Apollo includes three itineraries featuring well-organized geographical progressions. Leto's wanderings (itinerary one) constitute an orderly tour of the Aegean, centered around the important cultic site of Delos (30–49). In his journey to find a site for his oracle (itinerary two), Apollo travels an orderly route from Olympos to Delphi (216–86). In the third Apolline itinerary, the god presses a boatful of Kretan sailors into service at his Delphic temple. The itinerary that they travel, both before and after Apollo commandeers their vessel and sails to Delphi, is recounted in rich detail, though the voyage up the west coast of the Peloponnese is not without organizational difficulties (391–523). See Richardson 2010: ad loc. and Maps 1–2 for details of Leto and Apollo's travels. See also Baltes 1981 for a thorough overview and stylistic discussion of the catalogues. The Homeric Hymn to Hermes also contains a circuit of mainland Greece and overlaps with parts of the second itinerary of Apollo. First, Hermes travels from Mt. Kyllene to Pieria (70–74), from Pieria to Onchestos (87–93), from Onchestos to the River Alpheios (95–141), and finally back to Mt. Kyllene (142–320). Then Hermes and Apollo travel together from Mt. Kyllene to Olympos, from Olympos to Pylos and the Alpheios, and at last back to Mt. Olympos (320–578). For the details of this journey, see Vergados 2013: ad loc. and Maps 1a-1b.
94. See, e.g., BK: 154.
99. As, for example, he does with the dead Protesilaos, erstwhile leader of the Phylake contingent (Il. 2.698–702). Another instance may be seen at Il. 3.236–42, where Helen searches in vain during the teichoskopia for her brothers, the Dioskouri. Homer then acknowledges that Kastor and Polydeukes were already dead and buried in Lakedaimon (3.243–44).
102. At an earlier stage in the tradition, the Boiotian catalogue may have narrated the array of Boiotian towns not according to their actual locations, but according to their stations on a field of battle, as in Iliad 3.
104. Torres-Guerra 1995: 75. See also Torres-Guerra 2015 for a discussion of the fragments, background, and dating of the Thebaid. Though he posits a date of 573 b.c.e., for the text being fixed in writing, he concludes on 243, "[I]t seems possible that an oral version of the Thebaid existed prior to the Iliad, influencing its composition …"
107. Since the Boiotian catalogue from Book 2 of the Iliad sketches a rough circuit around the town of Thebes, moving in a somewhat east-west direction, it is possible to imagine two organizational modes for the 'original' teichoskopia from the Theban sagas. 1) It may simply have been a matter of describing the towns participating in the conflict, without reference to their disposition on the battlefield. 2) The ordering of the towns may indeed have reference to their disposition on the battlefield or at the seven gates. We tend to favor this second option, for reasons to be discussed momentarily.
109. 1) ἐνέμοντο (496): Hyria, Aulis, Schoinos, Skolos, Eteonos, Thespia, Graia, Mykalessos; 2) ἐνέμοντο (499): Harma, Eilesion, Erythrai; 3) εἶχον (500): Eleon, Hyle, Peteon, Okalea, Medeon, Kopai, Eutresis, Thisbe, Koroneia, Haliartos; 4) ἔχον (504) Plataia; 5) ἐνέμοντο (504): Glisas; 6) εἶχον (505): Hypothebai, Onchestos; 7) ἔχον (507): Arne, Mideia, Nisa, Anthedon.