previous sub-section
One The Study of Attic Fortifications
next sub-section

Closing the Gates of Attica?

A salient characteristic of the fortresses of Attica is the fact that they came into existence at various times. Several already existed before the Peloponnesian War, a few were built during it, and a few were built after


that war. It is possible to explain this proliferation of fortresses as the evolution of a homogeneous system, the objectives and functions of which were essentially the same at the end of the classical era as they had been at its beginning. If this was the case, then the functions of these fortifications must be equally intelligible early on, when there were few fortresses in the Attic countryside, and later, when they were more numerous. Modern investigators, however, have often explained the origins of the fortifications of Attica in terms of their ultimate dispositions, when, to all appearances, these fortresses guarded all of the important routes and passes crossing the land frontier of Attica. Such an explanation involves the awkward assumption that the earliest garrison forts of Attica were built as elements of a system of frontier defense that did not become comprehensive until the final generation of fortresses was built more than a century later.

The introduction to Lilian Chandler's 1926 article on the northwest frontier of Attica embraces such an assumption:

Of all states in ancient Greece, Attica seems to have had the most interesting and complete system of land defences. A chain of important fortresses, of most of which there are still considerable remains, follows the line of the Kithairon-Parnes range: Eleutherai, Oinoe, Panakton, Phyle, Dekeleia, Aphidna and Rhamnous. It may appear at first that this series of strongholds was designed expressly to mark off Athenian territory, but whilst incidentally and in large measure they served this end, in origin they were intended rather to defend the various roads from Attica into Boeotia.[33]

The assertion that forts were intended to defend roads is a modern deduction, supported by no ancient authority. Yet the seemingly systematic arrangement of forts along the roads leading into Attica is, to many observers, evidence enough that defense of roads and passes was not only the ultimate, but, as Chandler stresses, the original , purpose of these fortifications. Chandler was not the first to reach this conclusion. The system evident in the fortifications of Attica that Chandler goes on to describe had been outlined more than thirty years earlier by a Prussian military cartographer, Captain Winterberger, who had surveyed portions of the northwestern frontier of Attica for Curtius and Kaupert's Karten yon Attika . Winterberger's summary description of a "planmässiges System der Grenzvertheidigung" was reflected in Arthur Milch-hoefer's extensive commentary on the Karten yon Attika , which in turn provided the basis for Chandler's survey of the subject in English.[34]


Winterberger's description was itself the logical outgrowth of observations previously made at the beginning of the nineteenth century by Colonel (then Captain) William Leake in his influential work on Attic topography. Leake, whose classical studies were pursued while gathering military intelligence in Turkish Greece, was sufficiently impressed by the disposition of Attic border forts along the passes to Boiotia that he was ready to suppose the existence of remains of further forts completing the system where none actually existed. So, for example, he associates the classical place-name Melainai with medieval remains around the Byzantine monastery of Hosios Meletios on the southern slopes of Kithairon-Pastra, and he speaks of the place "as a castle on the frontier, for this situation would exactly serve to complete a chain of fortresses defending the passes of the Attic mountains towards Boeotia, of which the other links were Oenoë, Harma, Phyle, Deceleia and Sphendale."[35] George Finlay, a British philhellene, historian, and contemporary of Leake, turned classical topography to modern strategy in his history of the Greek War of Independence when he wrote that the Greeks might have cut off the Turks besieging Athens in 1826-27 "by a line of posts, extending from Megara to Eleutherae, Phyle, Deceleia, and Rhamnus."[36] It seems beyond a doubt that considerations of contemporary military strategy had a profound influence on the interpretation of Attic fortifications by nineteenth-century classicists.

Although the thesis that the Attic forts were intended to guard, or in some sense, to control, the roads leading into Attica has not been universally accepted, it has remained the most influential explanation of their purpose.[37] So, for instance, in his study of Greek fortifications published in 1971, F. E. Winter states:

In Attica the fortresses of Phyle and Gyphtokastro [Eleutherai] are both in a position to exercise complete control over their respective passes. Only a large army would have any chance of capturing them by direct assault, and then only at the cost of heavy casualties. Yet even the largest army


could ill afford to pass them by, leaving them free to harry its rear and sever its lines of communication.[38]

If complete control of passes were in fact the object of Attic garrison forts, this goal could not have been achieved until the fourth century, since they manifestly failed to control passes during the Peloponnesian War. A superficial view of the evidence makes this interpretation seem plausible, since it was only after the Peloponnesian War that the Attic border fort system achieved its most complete form, with the addition of the two fortresses named above; and between the Peloponnesian War and the beginning of Macedonian domination in 322 Attica was invaded only once, by the Spartans under Sphodrias at the beginning of the Boiotian War in 378. Whether or not the Attic border forts were from the first intended to guard roads, as Chandler and others have claimed, a circumstantial case can thus be made that this was their function in the fourth century.

A closer look at the sources, however, raises suspicions about the cogency of such circumstantial evidence. For instance, according to Plato's priorities for defensive works,

(Laws 778e), garrison forts would deserve special mention alongside, or even before, barrier walls if they had served this purpose in the fourth century. Had Plato's prejudices against circuit walls so completely blinded him to their usefulness? Or was he merely ignorant of recent innovations in the arts of defense? Before these questions can be answered, we must consider whether such an explanation of the Attic garrison forts is inherently plausible.

We can readily take the first step in this process by reviewing the most thorough exposition of this approach to Attic fortifications, presented by Josiah Ober in his thesis, published as Fortress Attica in 1985. According to Ober, in the fourth century the border forts of Attica were employed, for the first time, as elements in a system of preclusive frontier defense based on the control of all major routes into Attica. The forts enabled the Athenians to maintain troops on the frontiers of Attica year-round, so that, in the event of an invasion, they could harass and detain the enemy until the arrival of the main force from Athens, promptly summoned by signals via appropriately placed signal towers.[39] Ober schematizes the operation of the system in the following terms:

Even if the enemy forces had succeeded in forcing the pass, until they had taken the fortress that guarded it they would not be able to advance into Attica, since they could hardly afford to leave a significant garrison intact


which could attack their baggage train as it marched past. Furthermore, the fortress threatened the enemy line of retreat; if the invasion should fail, the possibility arose of being trapped between the main Athenian army and the garrison. The invading army would have to turn aside and attempt to reduce the fort before proceeding. The relief forces from Athens were therefore granted as much time as it would take the enemy to capture the fortress . . . . [The relief forces] would then proceed to attack the invaders at or near the fortress  . . . in the borderlands. Attica and its vital economic resources would therefore be protected.[40]

Here we must note again the absence of any textual support for such an interpretation of garrison forts. No fortified circuit held by a regular garrison is ever said to have been an obstacle to an invading army. There were many occasions when bodies of troops were posted at passes to prevent the passage of enemy forces, but in all cases these were extraordinary forces assembled ad hoc. Never is the presence or absence of a garrison fort circuit said to affect the defensibility of a route, and never was a perennial garrison, a regular element of the

, in Attica or elsewhere expected to prevent or delay the general invasion of a region.[41]

The archaeological evidence likewise provides no support for this interpretation, for garrison forts were situated according to criteria other than the defense of passes. The first criterion was the natural defensibility of the location of the fort itself; it ought to be a strong place (

, or ), well suited by nature to be difficult to seize by assault. The second criterion, which often actually had priority over the first, was inhabitability, as determined chiefly by the availability of water.[42] A location that met these criteria was often attractive for civilian habitation as well, and so garrison forts were frequently situated on or immediately adjacent to civilian settlements (as at Eleusis, Rhamnous, Aphidna, Oinoe, and Eleutherai). Garrison forts and their associated settlements were of course served by roads, but the placement of forts with reference to roads was secondary to the above criteria. As to actual passes or strategic narrows along roads, only one fortress stood in close proximity to a pass on the Attic land frontier (pace Winter above), and that was Eleutherai. Yet not even the fortress at Eleutherai (nor any garrison fort elsewhere in Greece, to my knowledge) physically obstructed passage along a major route.


No one denies, therefore, that an invading army could walk right past a fort, whether its garrison was confined within it or was partly out skirmishing on the mountainsides. In either event, a numerically insignificant garrison force could do little more than momentarily annoy a passing army.[43] Therefore, in order for forts to have had their preclusive effect, they must, it is argued, have compelled the invader to stop and attack them.

This is the heart of the thesis on the fourth-century approach to the defense of Attica as recognized by Winter and elaborated by Ober. It depended absolutely upon an invading commander's decision to stop and attack a fort before proceeding. The defenders of Attica, once they had built a fort and allotted a garrison to it, had no further control over that decision. Could that decision have been as inevitable as Ober must needs argue it was?

He, like Winter, appeals to the vulnerability of baggage trains and lines of retreat if forts were left intact along the way. As to baggage trains, only in the case of Eleutherai in the Kithairon pass does a road come so close to a fortress that a train could be struck by missiles from it. Since the Eleutherai pass is easily circumvented by other routes, including the nearby Dryos Kephalai pass, there is no reason to believe that an invading commander would have had to delay his advance in order to attack Eleutherai or any other Attic fort.[44] As to lines of retreat and how they might have affected a decision to invade, an invading commander would proceed only if he had confidence in his ability to overcome the enemy wherever they might appear in strength. He would therefore regard garrison forts as no more a threat to his eventual withdrawal than a hindrance to his advance.

This conclusion is amply supported by the testimony of Xenophon, who, writing shortly before the middle of the fourth century, on two occasions discusses the hypothetical consequences of an invasion of Attica. In his treatise on the Athenian cavalry commander, Xenophon considers the following scenario:

If the enemy invades Athenian territory, in the first place, he will certainly not fail to bring with him other cavalry besides his own and infantry in


addition, whose numbers he reckons to be more than a match for all the Athenians put together. Now provided that the whole of the city's levies turn out against such a host in defence of their country, the prospects are good. For our cavalrymen, God helping, will be better, if proper care is taken of them, and our heavy infantry will not be inferior in numbers, and I may add, they will be in as good condition and will show the keener spirit, if only, with God's help, they are trained on the right lines. And, remember, the Athenians are quite as proud of their ancestry as the Boeotians. But if the city falls back on her navy, and is content to keep her walls intact, as in the days when the Lacedaemonians invaded us with all the Greeks to help them, and if she expects her cavalry to protect all that lies outside the walls, and to take its chance unaided against the foes,—why then, I suppose, we need first the strong arm of the gods to aid us, and in the second place it is essential that our cavalry commander should be masterly. For much sagacity is called for in coping with a greatly superior force, and an abundance of courage when the call comes.[45]

While the thesis espoused by Ober would have us expect to find new strategies reflected in a fourth-century source, Xenophon's scenario is surprisingly consonant with the Periklean approach to the defense of Attica. The options for the Athenians, in the event of an invasion from Boiotia, are either to attempt to match forces with the enemy in open battle or to avoid battle with a powerful enemy in Attica and to withdraw within walls, while relying on the navy to strike against the enemy's homes and calling upon the cavalry to harry the invader and limit his depredations by preventing him from dispersing his forces to devastate or plunder. There is no hint that forts along the borders prevented, or even delayed, the arrival of the invader outside the walls of Athens.[46]


Xenophon addresses different concerns in his treatise on the revenues of Athens, but his reflections on the strategic situation of Attica under invasion are entirely consonant with the previous passage. In speaking of enhancing the revenues of the state from the silver mines of southern Attica, Xenophon discusses how minimal the effects of an invasion would be on this resource:

I reckon that, even in the event of war, it would not be necessary to abandon the silver mines. There are, of course, two fortresses in the mining district, one at Anaphlystos on the south side, the other at Thorikos on the north. The distance between them is about sixty stades [just under twelve kilometers]. Now if there were to be a third stronghold between them on the highest point of Besa, the works would then be linked to one or another of the fortresses, and at the first sign of a hostile movement, every man would have just a short distance to go in order to reach safety. If the enemy came in force, they would certainly carry away any grain, wine, or livestock that they found outside; but the silver ore, if they were to seize it, would be of no more use to them than so many stones. And how could an enemy ever invade the mining district? The distance between Megara, the nearest city, and the silver mines, is of course much more than five hundred stades [about 100 kilometers]; and Thebes, which is the next nearest, lies at a distance of much more than six hundred stades [over 120 kilometers]. If, then, the enemy is marching on the mines from some such point, they are bound to pass Athens. And if their numbers are small, they are likely to be destroyed by our cavalry and our patrols. On the other hand, it would be hard for the enemy to march with a large force, leaving their own property unprotected. For when they arrived in the mining district, the city of Athens would be much nearer to their own states than they themselves would be. But even supposing that they should come, how could they stay without supplies? To send part of their forces in search of food would endanger both the foraging party and their overall objectives, while if the whole force is continually foraging it will sooner find itself besieged than besieging.[47]


In this passage, the function of rural forts during an invasion is discussed, though they are the forts along the northern edge of the mining district, not on the borders of Attica. Xenophon points out that there would be little to fear in the mining district from a raiding party entering Attica, for such a small force would likely be destroyed before its arrival in this distant corner of Attica by the combination of the quickly deployed cavalry and the already deployed patrols, the peripoloi . A major invasion force, however, would certainly be able to range and plunder at will throughout Attica. What would protect the mining district in that event would be the fact that everything vital (especially the miners themselves) could be readily withdrawn to the safety of convenient forts, while the silver ore left behind would be of no immediate value to the enemy and would be too cumbersome to move. As a consequence, there would be no point to an invasion of the mining district, especially since such a course would leave the enemy's own territory undefended; and being at the end of a long march, it would force the enemy to disperse and possibly lose manpower in foraging for supplies while achieving no useful offensive purpose. Here Xenophon assumes that an invading army would outman the Athenians and would have no difficulty moving anywhere in Attica. There is a conspicuous lack of any reference to border forts in a context that would, according to Ober's thesis, be most appropriate for their discussion.

These two passages from Xenophon, the first written perhaps before 362 and the second in the later 350s, are the only explicit discussions in fourth-century literature of the potential Athenian responses to an invasion of Attica.[48] Xenophon made these observations during the very period in which Ober claims that the system of border forts and towers was being perfected, having been under the guidance of a "coherent and ongoing program of defensive preparations" for some time.[49] Yet there is not the slightest trace of the system described by Ober in the writing of Xenophon.

Indeed, although Ober must assume that the policy of preclusive border defense was implemented by the Athenians through a process of


ongoing public debate, he is embarrassed by the absolute silence of even the orators on the subject. He attempts to explain it away by asserting that "discussions of border defenses probably tended to make for rather dull orations,  . . . and speeches concerned with the technicalities of the fortification system were not chosen for copying and preservation."[50] This is special pleading. For in laying the foundations for his interpretation of the border defenses of Attica, Ober argues that the collective Athenian psyche was in the grip of a "defensive mentality" to such an extent that "the fourth-century Athenian lived in terror of enemy invasion and wanted desperately to be allowed to go about his business in peace and safety."[51] There is no doubt, as has been pointed out at the beginning of this chapter, that the Athenians valued highly the security of Attica. But if there were any radically new departure embodied in the fourth-century approach to

, and if, as Ober argues, that approach was sometimes at odds with such factors as financial constraints on the one hand and opposing policies, such as the imperialism of 395-387 or the aggressive foreign policy of Demosthenes on the other hand, then the subject of border defenses would have been ripe for declamatory pyrotechnics. In fact, forts are mentioned in speeches, as are various other preparations for war, and the ideal of defending the homeland is brought up often enough in fourth-century Athenian rhetoric. Yet nowhere is there any hint that the Athenians had created, or thought that they had created, an impermeable barrier of fortresses and watchposts around Attica.[52]

Arguments from silence are never, by themselves, fully convincing. There is a passage from Demosthenes, however, which appeals to the common knowledge of his audience about the nature of defensive preparations and is more telling than most about the full range of Athenian measures. Not only is this passage silent about a preclusive barrier of frontier fortresses, but it absolutely excludes the possibility of such a defensive policy. In his speech On the Naval Boards , delivered in 354, Demosthenes advocates practical measures the Athenians could take to strengthen their military preparedness. His specific advice is introduced after the following prefatory remarks:

If indeed there were one kind of force suitable for defence against Persians and another for defence against Greeks, then we might reasonably


be suspected of marshalling ourselves against the King; but when all preparation for war is on the same lines [italics added] and the main objects of an armed force are the same—to be strong enough to repel the enemy, to assist one's allies, and to preserve one's own possessions—why, having open enemies enough [in Greece], must we be looking out for another? Let us make our preparations against them [i.e., Greeks], and then we shall defend ourselves against him too, if he ventures to molest us.[53]

By defense against Greeks, Demosthenes particularly has in mind defense against the Thebans (as he makes explicit in On the Naval Boards 33-34), who dwell on the very borders of Attica. The policy that Demosthenes goes on to advocate is a revision of the procedures for financing naval operations, designed to make it easier to man the fleet. The premise of these prefatory remarks, epitomized in the italicized portion above,

, would be manifestly false if the Athenians, as Ober argues, had labored to create a unique frontier defense system.

Given the inherent implausibility of the hypothetical system together with the silence of the orators, the silence of Xenophon, the silence of Plato, and of all other sources, we must conclude that Ober and his predecessors have created e silentio a fabulous structure. Ober's "preclusive defense system" never existed except as a modern figment.

previous sub-section
One The Study of Attic Fortifications
next sub-section