Testing hoplite combat

[Erpingham]

So isn't the very first step to demonstrate the incompatibility of the standard model with carrying (forward facing) shields?

Didn't Connolley do this 20 years ago - I'm sure I've seen the pictures.  Or am I missing some damning reason why he did it wrong?

Overall, I'm a little concerned that all our experiments with broom handles and rakes might lead us to miss some ergonomic issues.  As mentioned above, different grips are better for different tasks.  Grips with the hand under the pike are, as Richard said, pretty good for bearing weight.  So in the renaissance charge (high) position, the left arm carrying the weight braces down through the body and steadies the weapon.    But that position is quite inflexible.
Polevaulter grip is a better dynamic (stabby) grip.  I think, however, it may be better stabbing down wards rather than level (its what polevaulters use it for if you think on it).  This is probably why it appears in the medieval repertoire - stabbing people over the front line or on the ground. 
I don't think the pool cue approach of sliding through the forehand is particularly good with an 18ft pike - it would pitch down I think - but experiments would be useful.  Also note any dynamic pike use needs more space than static use (especially behind), whichever hold you choose.

[Justin Swanton]

Useful pictures Justin. So as I understand it you have:

Front rank - 'standard low' (left under, right over, low)
Ranks 2-5 (or 6) - 'polevaulter high' (left over, right under, high)
Ranks 6+ (or 7+) - 'double over low' (left over, right over, low)

The front rank would be left and right over, low. I find it easier to manage the shield that way.

Why do you have a different grip for ranks 2-5 from 6+? Wouldn't it be easier (in terms of stepping forward as replacements) if they used the same grip? If a front ranker fell, would his second need to change grip or do you assume in the rough and tumble of combat such niceties wouldn't matter?

The only reason for the front ranker holding his sarissa low is as an anti-cavalry stance as shown in the Pergamon plaque. It's quite possible that if cavalry didn't feature then all the ranks could hold their sarissas high. Also quite possible that everyone used the same left and right hand over high grip with nobody using the polevaulter grip. The only advantage of the polevaulter grip is that one can better thrust the sarissa forward.

I'm pretty certain your final image (holding the sarissa by the end 30cm or so) would be impossible with a real sarissa, counterweighted or not.

That's just because all I had on hand was short stick. Visualise about 4 feet of sarissa projecting back past my right hand.

You say the telamon suports the shield and/or sarissa arm but given the angle it's at, the amount of support (weight bearing) looks fairly minimal. In order to provide support, would it not need to be so short that it was impossible to lower the left arm? How would this affect eg marching to combat? Both the counterweight and the telamon are speculative, with no hard evidence to support them (as we've discussed previously) so do you feel they are essential to your model?

I find that even when projecting the shield forwards as in the photos the telamon does give support. And if I rest the lower edge of the shield on my hip the telamon entirely supports the shield and my arm. Zero muscular strain.

The standard view of Hellenistic pike grips is:

Ranks 1-16 - 'standard low' (left under, right over, low)

Which has the benefit of simplicity (Occam would be pleased). I understand you feel this standard model is impossible due to the difficulty of the sarissa passing the shields. So isn't the very first step to demonstrate the incompatibility of the standard model with carrying (forward facing) shields? Then once you've shown that the standard model is impossible (or at least unlikely), your alternative three-grip model would become more attractive. As it is, your model is fine, and there's no reason I know of to declare it impossible, but it is also entirely speculative and logically unnecessary, and there doesn't seem any reason to favour it.

This will be definitively resolved by a reenactor phalanx trying out the two models and seeing which one works. For now we know three things: a) the height of a man in Antiquity, b) the size of a pike pelta, c) the width of a close order file. Putting these three together, I find it utterly impossible to make a close order phalanx work with the pikes presented low. From the second rank going back, the pikes will of necessity be horizontal about two feet or less above the ground. The phalanx will stop dead in its tracks if it tries to go up a slope inclined at more than about 9 degrees, or pass over an obstacle more than two feet in height. The second rankers going back will be unable to target the shields or any vulnerable part of their opponents' bodies, and those pikes will be useless against cavalry. Finally, the pikes, all jammed together in a space barely 10cm wide, will foul each other.

To overcome these difficulties one is obliged to deny a), b) or c). To date I haven't seen any convincing recreation of a close-order phalanx, shields overlapping, pikes low, anywhere.

[Justin Swanton]

So isn't the very first step to demonstrate the incompatibility of the standard model with carrying (forward facing) shields?

Didn't Connolley do this 20 years ago - I'm sure I've seen the pictures.  Or am I missing some damning reason why he did it wrong?

His experiments don't include pictures of a pike phalanx in close order with shields to the front and pikes held low. What I saw was an intermediate order setup with about a foot of space between the shields of each file.

Overall, I'm a little concerned that all our experiments with broom handles and rakes might lead us to miss some ergonomic issues.  As mentioned above, different grips are better for different tasks.  Grips with the hand under the pike are, as Richard said, pretty good for bearing weight.  So in the renaissance charge (high) position, the left arm carrying the weight braces down through the body and steadies the weapon.    But that position is quite inflexible.
Polevaulter grip is a better dynamic (stabby) grip.  I think, however, it may be better stabbing down wards rather than level (its what polevaulters use it for if you think on it).  This is probably why it appears in the medieval repertoire - stabbing people over the front line or on the ground. 
I don't think the pool cue approach of sliding through the forehand is particularly good with an 18ft pike - it would pitch down I think - but experiments would be useful.  Also note any dynamic pike use needs more space than static use (especially behind), whichever hold you choose.

All true. I need to make me a proper imitation pike and see how it handles.

[RichT]

Connolly did it (one cubit order, low pikes, forward shields) and said it worked fine (but Justin, as he says, wants pics).
Delbruck did it (one cubit order, doesn't specify the pikes or whether they had shields) and said it worked fine, and poured scorn on those who didn't accept his experiment.
Matthew so far as I know hasn't done it (one cubit order, low or high pikes, forward shields) but concludes it is impossible based, so far as I can tell, on drawing diagrams.
Justin hasn't done it but concludes it is impossible based on thinking about it and not being able to envisage it.

I have, broadly speaking, an exceptionally low opinion of the value of such experiments (practical or thought), though sometimes practical experiments can be better than nothing, so long as they are properly designed, properly conducted and (above all) properly documented, which they so rarely are.

But in this case we have a simple question: is one cubit order, low pikes, forward shields physically possible? Proving that it is or isn't should be quite straightforward given a modest outlay of time and effort (and yes, lone experiments with rakes and sticks in the back garden will not do the job). I think Justin and I were quite hopeful that Paul's experiment might be an opportunity to do this, but Paul reasonably enough had a different agenda (hoplites). My point is that until this point is proven, alternatives likes Justin's model are unnecessary and pointless, and recreating them is a waste of time - the first step has to be to prove that the standard model is not possible (or otherwise), and we can go from there.

[Patrick Waterson]

My point is that until this point is proven, alternatives likes Justin's model are unnecessary and pointless, and recreating them is a waste of time - the first step has to be to prove that the standard model is not possible (or otherwise), and we can go from there.

This may be an unduly restrictive approach: while I am essentially happy with the standard model, I would not seek to prohibit the offering of alternatives*, particularly testable alternatives.  Testing (or re-testing) the standard model is also a good thing to do, as it should inter alia help to determine whether or not slopes would actually be any sort of impediment and might provide further insights to question - or strengthen - the basic model itself.

*Cave canem in praesepio.

It may be that both the standard and alternative models work, the obvious objective in this eventuality being to find which works better.

[Justin Swanton]

I feel thought experiments are useful in that they point the way to further experimentation and hopefully an eventual certitude (that is the role of hypotheses in scientific investigation). Since a picture is worth a thousand words, here are my pictorial objections to the classical low hold system, derived from thought experimentation (and some measuring). Everything is to scale. The second picture shows the approximate maximum steepness of a slope that can be traversed by a phalanx using a low hold. It naturally applies to any obstacle two feet high or even less. Crossing the Pinarus at Issus and fighting up the opposing embankment isn't going to happen, and neither is the Macedonian attack across the ditch and palisade fieldwords at Sellasia.

[Erpingham]

As usual I admire your graphic skills but I am unclear at to why the phalangites don't raise their left arms, thus lifting their shields, in unison.  Does the suggested shield suspension prevent this? 

[Justin Swanton]

As usual I admire your graphic skills but I am unclear at to why the phalangites don't raise their left arms, thus lifting their shields, in unison.  Does the suggested shield suspension prevent this?

Yes. The shield is braced by the taut telamon, ochanon and arm working together. I find it impossible to raise the shield in the low hold position. Also if you did raise the shield too much you wouldn't be able to see where you're going.

[Erpingham]

As usual I admire your graphic skills but I am unclear at to why the phalangites don't raise their left arms, thus lifting their shields, in unison.  Does the suggested shield suspension prevent this?

Yes. The shield is braced by the taut telamon, ochanon and arm working together. I find it impossible to raise the shield in the low hold position. Also if you did raise the shield too much you wouldn't be able to see where you're going.

The obvious question would be is the shield suspension system right, or should it be adjusted?  I've no idea but raising questions is part of the thought experiment process.  On the latter point, from your scale diagram the phalangite could easily raise his shield 30cm without blocking his view forward, whch would make a difference. 

[Patrick Waterson]

As a follow up to Anthony's observation, I wonder whether a phalangite might, in lieu of raising the shield, simply depress the right shoulder/arm/hand a small distance, thus elevating the point the necessary few degrees to avoid ploughing it into terra firma.  Any thoughts on this?  Would bodily stance adjust slightly when moving uphill, when one's line of advance is no longer level?

[RichT]

I'm not (obviously) proposing the prohibition of speculative alternative models of Hellenistic pike use. I'm just saying that, given the lack of evidence supporting Justin's model, the only point in its favour is that it solves the problem of the standard model being physically impossible; but the physical impossibility of the standard model is (to say the least) undemonstrated. Connolly's practical experiments found that the standard model worked fine. Justin's thought experiments and diagrams find that it is impossible. In this case we just have to take our pick (and I pick Connolly).

If some future experiment proves the standard model impossible, then we can turn with delight to the solution offered by Justin's model. If future experiment proves both standard model and Justin's model possible, then we still have to chose. If that turns out to be the case, then I would argue 1) that Hellenistic pikes were in use for 300 years from Italy to India and we should no more expect every phalanx that used them to have done so in exactly the same way than that pikes from 1300 AD to 1600 AD were always used identically but also 2) that such evidence that we have (Pergamon plaque) and suggestions from literary descriptions (Polybius, tacticians) would tend to suggest that the standard model is probably correct (or I should say, much the most common).

Justin - as when we've discussed this before I think your problem is your insistence on the shields being held parallel to the front of the unit. If they are angled, then your objections would seem to go away. Now whether holding a shield at an angle would be practical in a combat formation is something we have no data on either so probably isn't a fruitful line of enquiry. As to shield hold, this is also entirely speculative. If your way of holding the shield is incompatible with the standard model, then most likely you are holding the shield wrong.

Edit to add re raising the pike - it's also possible to lean back slightly, as most people do naturally when carrying a pike (because it's heavy), which would also raise the point. I think (TBH) that the idea that pikes held low can't be carried up hills is not one we need to devote any time at all to refuting.

[PMBardunias]

One thing I can tell you from many experiences, is that you can never get caught with your spear under the line of overlapped shields. Your weapon is rendered useless and your foe will just step on it. Stabbing beneath a shield wall might seem like a good idea, but it does not work in practice once the shields overlap.

[Erpingham]

One thing I can tell you from many experiences, is that you can never get caught with your spear under the line of overlapped shields. Your weapon is rendered useless and your foe will just step on it. Stabbing beneath a shield wall might seem like a good idea, but it does not work in practice once the shields overlap.

With respect to your experience, may the experience of the pike phalanx be different?  In Justin's diagram, the pike head is 60cm off the ground.  That is a big step, perhaps difficult in a formed line.  Also, there are five more pikes behind the first which might target the exposed stomping leg.

We might note that the accepted anecdotal way in the Middle Ages (which may not in each circumstance be true but clearly circulated as an idea) was to bear down the pikes with the body of a hero, or hero on a horse.  This would take out a group of pikes at once, allowing following attackers into close contact.

[PMBardunias]

With respect to your experience, may the experience of the pike phalanx be different?  In Justin's diagram, the pike head is 60cm off the ground.  That is a big step, perhaps difficult in a formed line.  Also, there are five more pikes behind the first which might target the exposed stomping leg.

Ha, no sorry. I did not mean they just step on your spear.  When using these things they will move up and down while you strike and in general your foe will either knock it up and away, as is best with a hoplite dory, or down. Were we foyning and you were under the shield like that, I would pin your spear to the ground and you have very little ability to get over the top of my pike because you have very limited elevation. Then I would step on it if I advanced.

[PMBardunias]

Add to this what i have stated previously, that the deep pelta is not a good design to overlap. So much so that I think it is only because they stopped overlapping that the depth was possible in a rimless shield. I do not have a pair of deep peltae, but I do eat cereal.  See my experimental image below. The curve is similar to that of a roman, hemicylindrical scutum, and I have never seen these overlapped.  What they were probably doing is holding the front arm at 45 degrees or more as Connolly drew.