Numerous theories have attempted to explain the function of the immense sloping defensive structures that enclosed many ancient cities in the Middle Bronze Age II (2000–1550 B.C.). Why so wide at the base? The rampart at Ashkelon, for example, is more than 50 feet high and more than 75 feet at the base.

At one time, most scholars attributed these formidable MB II fortifications to the “Hyksos,” the enigmatic and supposedly non-Semitic people from the north, who, with their superior weapons of war, invaded and conquered Canaan and Egypt in the 17th century B.C. The Hyksos had chariots, even at this early date. Ashkelon’s excavator, British archaeologist John Garstang, believed that Ashkelon’s earthworks, like those of the “Lower City” of Hazor, surrounded, not a city, but a huge chariot park. Storing chariots and their horses required a wide, protected area; hence the huge earthworks. The inhabited portion of Ashkelon, Garstang thought, was confined to a small mound inside the wide arc of the earthen embankments.

Other scholars proposed that the earthworks were related to chariotry, but for the opposite purpose: to keep them out rather than in.

Yigael Yadin dispelled both notions. His excavations at Hazor showed that the city extended throughout the lower portions of the site and was surrounded by impressive earthworks that formed the base of the fortification system. The earthworks, in other words, surrounded not a chariot park but an entire city. We now believe this to be true at Ashkelon as well. As for the other theory—that the earthworks were meant to repel chariots—Yadin pointed out that chariot warfare took place not around cities but in open plains away from cities.

Yadin believed the earthworks at Hazor were built to counter the battering ram, also once thought to have been introduced by the Hyksos. This seems highly unlikely, however, since we know that in later periods, the besiegers, such as the Assyrian king Sennacherib at Lachishi in 701 B.C. and the Romans at Masada in 73 A.D., actually built sloping siege ramps in order to move their battering rams into position for attacking weak points in the fortification line, such as the city gate. So I doubt that these huge earthworks were intended to counter the use of battering rams; they might have even aided in their use.

Archaeologists Peter Parr and G. R. H. Wright have proposed a more banal function for the sloping ramparts: to counter erosion of the tell, or artificial hill, formed by superimposed layers of settlement. The tell of some cities reached a considerable artificial height by the second millennium B.C. But surely, there must have been more energy-efficient and less costly ways of countering erosion than by installing tons and tons of earthen embankments around the site.

I would like to propose another solution. I agree with Yadin that the MB II ramparts were a defense against siege warfare. But these thick sloping embankments, often surrounded by a ditch or dry moat, were designed, not to counter battering rams; rather they were built in response to another very ancient city-conquering technique—tunneling, mining and sapping—in common use even in the medieval period, in fact right up to the invention of gunpowder.

While the city was under siege, a team of excavators from the attacking army would begin their tunnel at some distance from the fortification line they wished to undermine. Their object was to cause the fortifications to collapse or to sneak beneath them and then to surface inside the city, usually at night, to launch a surprise attack. It might take days, even weeks, for the “moles” to reach their objective. Once under the fortifications they might widen the tunnel in order to collapse the defenseworks above, or if that failed, to stoke the widened tunnel with combustibles, which would then be burned in order to precipitate collapse, while assault troops penetrated the breach above ground.

Obviously, the thick earthen ramparts of Ashkelon and many other Canaanite cities posed a serious obstacle to this siege technique. The amount of debris the tunnelers would have had to remove before reaching the wall line and towers was so great that this would give scouting parties, sent out by the besieged city, adequate time to spot the sappers and trap them or smoke them out. The ditch which surrounded many such ramparts was, whenever possible, dug to bedrock. This prevented the sappers from beginning their tunnel beyond the ditch and would give the scouts of the besieged city a better chance at spotting the entrance to a tunnel.

Tunneling through an MB II rampart was not only slow but also quite dangerous: The sand and soil fills, such as were used at Ashkelon, would have been extremely unstable. The tunnels would have been extremely susceptible to collapse (we know from experience just how unstable the balks or standing sections are at Ashkelon after more than one collapse).

When I presented this hypothesis to the premier military historian of the ancient Near East, Professor Israel Ephal of the Hebrew University, he was quick to accept the idea and then informed me that there was even an Akkadian word, pilsûu, that describes just such a siege technique, and it was already in common use by the early second millennium B.C. Thus the sapping or tunneling technique was known and used precisely at the time we find massive fortifications appearing in Syria and Canaan.

My conclusion is that the construction of immense sloping structures at the base of the city wall was not introduced by the Hyksos as foreign invaders. Indeed, the “Hyksos” were really Canaanites, anyway, as we now know from Manfred Bietak’s excavations at Tell-ed-Dab‘a, the Hyksos capital of Avaris in the Egyptian Delta. This fortification technique was an indigenous innovation of Canaanite cities to counter the besiegers’ tactic of tunneling to undermine the battlements or to enter the city clandestinely.