Today on the blog we welcome author MC Bishop, whose latest book The Pilum is now available to order. In today's post he puts theory into practice to see what the pilum was actually like to use as a weapon. 

Setting out to write The Pilum: the Roman Heavy Javelin, it was easy to work out the archaeological examples and literary sources that would help tell the story but what none of these did was tell me what it was like to use the weapon. To be sure, there were clues, once you knew where to find them, but hard data was rare while myths and speculation abounded. There had even been a few experiments, most of them fairly informal in nature, but numbers were, for the most part, lacking.

For this reason, I approached an old friend – one of the only people I know who had bothered to time the processes necessary to make ancient weapons – to help me test the pilum. Dr David Sim is a specialist on Roman metalworking and a practising blacksmith. Most importantly, he is always willing to throw a few weapons around in the name of the advancement of knowledge.

There was a series of questions to which we wanted answers, mostly focusing on those myths. Was the pilum designed to stick in an enemy shield? Was the tip hardened but the shank not? What about range? To guide us, we had those few obscure passages in the literary sources and some archaeological examples of damaged pila. Could we reproduce the various types of break or bend that we saw? In what follows, all of the damage was rectified in seconds by David nipping into his forge, next to where we were doing the tests: paradoxically, although the pilum could not be thrown back immediately, it was definitely a reusable weapon.

First we set up a series of drop tests to study penetration for different types of head and weight of weapon into a variety of targets, primarily simulated shields. This yielded the fascinating fact that steel and iron heads had exactly the same penetrative power against a simulated plank shield. Similarly, both were equally less effective against a section of ply shield. 

Pilum picture 1

This has a direct bearing on one of the old chestnuts of pilum mythology: the head was hardened but the shank was left unhardened so that it would bend. It would doubtless be a different story against any sort of metal armour, but shields are what matter for the time being. Why? As everybody knows, the pilum was designed to stick in the enemy’s shield. Textbooks have been repeating this second myth for years but the design of the pilum makes it quite plain that it was meant to penetrate the shield and stick in the person behind it. That at least makes sense for a weapon.

Let’s go back to that resilient little chunk of three-ply shield, simulating the typical legionary shield (as evidenced by actual finds from Fayum (Egypt), Masada (Israel), and Dura-Europos (Syria)). More energy was clearly needed to penetrate a ply shield and this could be achieved by adding weights to the pilum, which is precisely what happened in the 1st century AD. Since most barbarian foes used plank shields, it seems the legionary heavy javelin had been upgraded to be civil-war-ready!

But what about that ‘designed to stick in a shield’ myth: where did that come from? It is actually the product of one of the characteristics of written sources: they tend to favour the unusual over the mundane. In his Gallic War, Julius Caesar noted how, in his battle against the Helvetii, the enemy were so closely packed together that their shields overlapped and his legionaries’ pila pinned some of them together. He commented on it because it was unusual, rather than normal, for this to happen.

As for being ‘designed to bend’, our experiments showed that, when thrown, pila resolutely refused to bend whether they were stuck in a shield or the ground.

pilum picture 2

Shaking a shield to try and dislodge the javelin or grabbing it and trying to pull it out caused bending immediately below the head (and continuing with this would ultimately lead to failure of the shank).

pilum picture 4

Pilum heads would bend at the tip if they bounced off the shield board but one type of bend commonly seen amongst archaeological finds stubbornly refused to appear in our experiments – an angle of about 30º in the middle of the shank.

pilum picture 3

A possible solution occurred to us when we noted that this matched the angle of a pilum sticking out of the ground after being thrown. A forest of shafts protruding from the ground would present a formidable obstacle during a battle, particularly if it might be desirable to retreat through them, so deliberately treading them flat as men passed over them would solve their problem and ours – producing the bend that refused to occur naturally.

pilum picture 5PHOTO 5

Trying this first with iron rods [PHOTO 5] and then actual replica pila [PHOTO 6] reproduced the bend seen in the archaeological finds. Of course, as with all experimental archaeology, this just means that it could have happened in this way, not that it definitely did.

pilum picture 6PHOTO 6

One final detail has intrigued past experimenters but we chose to ignore: the maximum range of the pilum. Given that range was never raised as an issue by ancient commentators, then it probably never mattered to them. It was a shock weapon to be used immediately before two sides charged home; unlike light javelins such as the lancea, it was never about hurling the pilum long distances.

Whilst the summary results serve to illustrate my Osprey volume, our full report will be published in the Journal of Roman Military Equipment Studies in the near future.

The Pilum by MC Bishop is now availble to order from our website, which you can do by clicking here.