Did the Macedonian Phalanx practise othismos with its sarissas?

[Justin Swanton]

I think you are missing the fact that the ten spears bit refers to fighting Romans.  Romans don't have pikes - they throw things and close with swords.  To do this they must fight their way through layers of pikes.

OK. We are looking at two interpretations of the passage: the Romans have fight through 5 (or 6) rows of pikes successively or they have to fight through the whole lot at once. I opt for the latter as that makes better sense of Arrian: "Compactly they stand back successively so that each hoplite [phalangite in this context] in the front is covered by six sarissas and presses on with six forces whenever they bear down." Notice the "compactly" BTW - it implies less than 2 cubits per rank.

On pushing on the backs of the person in front with shields, it would be easier with a ported pike.  And before you ask, pretty impossible with a high charge position.  So, low charge might work but would be inefficient.  No, those pikes are held like that because the protruding pike points are more important than shoving people in the back.

Can you clarify for me? What is a ported pike and a high and low charge position?

Final thought.  Were Macedonian pike phalanxes designed to fight other pike phalanxes?  If they were for fighting hoplites or Persians, that row of pike heads would stop the enemy closing with their big shields and shorter weapons and give them the choice of being skewered or giving ground.  This sort of works too against Romans but they have a couple of tricks of their own - they throw things, they are more flexible in difficult terrain and they have fancy ways of using reserves.

They were designed for hoplites and Persians, their only serious enemies at that time (besides Thracians). They were probably designed more for hoplites as the default Macedonian hoplite would have been good enough for Persians. So they had to beat the hoplite phalanx at its own game. And we argue about what the hoplite phalanx's game was (no, I won't say the O-word). Just to say that the hoplite was well protected against spear thrusts.

[Erpingham]

Can you clarify for me? What is a ported pike and a high and low charge position?

How quickly the overarm sarissa thread fades into the past

Porte = holding at 45 degrees, front hand high, back hand low.  Not to be confused with high porte across the body, which is good for pushing but does not point pike points forward, which we know they did.

Low charge (Swiss style) = pike down at thigh level, arms straight or bent at elbow (there seem to be variations - I think the bent elbows one is more apt for comparison because of the need to hold shield). Front hand slightly ahead of body.

High charge - the classic "charge your pike" known to all re-enactor pikemen, with pike at neck level, front hand level with chest , back hand stretched backwards along shaft.  You advocated this as a possible sarissa hold.


[

They were designed for hoplites and Persians, their only serious enemies at that time (besides Thracians). They were probably designed more for hoplites as the default Macedonian hoplite would have been good enough for Persians. So they had to beat the hoplite phalanx at its own game. And we argue about what the hoplite phalanx's game was (no, I won't say the O-word). Just to say that the hoplite was well protected against spear thrusts.

Yes, but the hoplites game, as Paul B reminds us, is to close to spear fighting distance and ultimately to shield to shield.  He has to navigate his aspis between ranks of spears to do this, which may be possible individually but perhaps not as a body.  He can break formation and fight (abandoning his big advantage of a seried phalanx) or he can keep it and fall back (abandoning the forward momentum of his supporting ranks, however we envisage that).  This is a good balanced fight against good professional hoplites but will chew up ordinary city state militia types.  Persian Cardaces will find it very tough, provincial levies it doesn't look good at all.

[Patrick Waterson]

Arrian may actually be a better guide in this instance than Polybius, who was an Achaean cavalry officer and displays occasional misunderstandings of how a phalanx worked (e.g. in his critique of Callisthenes in Book XII).  Polybius has two cubits per man as file spacing but this is obviously incompatible with the rear ranks being able to add their weight.  Justin's one cubit or so per man for file spacing at contact (or just pursuant to contact, which may be the key here) makes eminent sense, because it brings the pike points forward and makes multiple application of points against an individual easier to achieve.  And it allows the rear ranks to 'add their weight' by actual contact with those in front.

[Justin Swanton]

Can you clarify for me? What is a ported pike and a high and low charge position?

How quickly the overarm sarissa thread fades into the past

Porte = holding at 45 degrees, front hand high, back hand low.  Not to be confused with high porte across the body, which is good for pushing but does not point pike points forward, which we know they did.

Low charge (Swiss style) = pike down at thigh level, arms straight or bent at elbow (there seem to be variations - I think the bent elbows one is more apt for comparison because of the need to hold shield). Front hand slightly ahead of body.

High charge - the classic "charge your pike" known to all re-enactor pikemen, with pike at neck level, front hand level with chest , back hand stretched backwards along shaft.  You advocated this as a possible sarissa hold.

I suspected that's what you meant (except for the porte - I'd completely forgotten that).

Y'know, there is actually an issue trying to wield a pike two-handed if you face directly forwards, and are in close order with a shield across your front. Here's a diagram with phalangites carrying the two standard shield sizes (64-66cm diameter for the smaller shield, 73-76cm diameter for the larger shield). Each phalangite occupies a close order interval of 48cm.



It's very difficult, especially for the larger shield, to get your left arm all the way across to the low position far enough so it can reach the pikeshaft where it would need to be to pass under the shields. OK with the high position though. In the low position your right hand can grip the pikeshaft well back from the left hand but in the high position it can't - I tried. To grip the pikeshaft comfortably with a decent gap between hands (Polybios' 2 cubits) you need to adopt a 3/4 stance, more so for the high position. For the porte the high position works fine even facing forwards.

In a serious pike-push crush your body is naturally forced to face ahead as the shield behind you pressing into your back flattens your torso into a forward position. In a high position your left hand can retain its grip on the pikeshaft but I've no idea how it would manage that in the low position whilst keeping the pike below the shields. The right hand in the high grip would have to move forwards, to just behind the left hand, though I suppose that if the pikehead is jammed in the shield of your opponent that isn't a problem - it would explain though Polybios' preoccupation with holding the pike at its point of balance, unlike the Renaissance pike that would just drop forwards in this position.

They were designed for hoplites and Persians, their only serious enemies at that time (besides Thracians). They were probably designed more for hoplites as the default Macedonian hoplite would have been good enough for Persians. So they had to beat the hoplite phalanx at its own game. And we argue about what the hoplite phalanx's game was (no, I won't say the O-word). Just to say that the hoplite was well protected against spear thrusts.

Yes, but the hoplites game, as Paul B reminds us, is to close to spear fighting distance and ultimately to shield to shield.  He has to navigate his aspis between ranks of spears to do this, which may be possible individually but perhaps not as a body.  He can break formation and fight (abandoning his big advantage of a seried phalanx) or he can keep it and fall back (abandoning the forward momentum of his supporting ranks, however we envisage that).  This is a good balanced fight against good professional hoplites but will chew up ordinary city state militia types.  Persian Cardaces will find it very tough, provincial levies it doesn't look good at all.

Here the phalangite wins. The hoplite can't fight him spear to spear, and he can't close to shield against shield. He has to let the phalangites stick their sarissas in his shield and then try to win the shoving contest, with the risk - not shared by the phalangites - of sarissa points pushing through his shield and armour and killing him. I know who I'd bet on.

[RichT]

[Dangun]

Nice graphic by the way...

In a serious pike-push crush your body is naturally forced to face ahead as the shield behind you pressing into your back flattens your torso into a forward position.

So is this an argument against scrum-O?

One for the proponents of scrum-O... how do we rationalise low casualties and scrum-O?

[PMBardunias]

My 2 drachmas:

A sarissaphoroi cannot push with a sarissa in the manner that hoplites pushed.  The reason has nothing to do with spacing, etc.  It is because the linkage between man and sarissa is no strong enough to support the force levels we see with hoplites.  You cannot grip a spear shaft strong enough to support a half ton of mass. 

But did sarissaphoroi engage in "othismos"?  Surely they did, the author's are quite clear.  But othismos just refers to a very crowded condition. This can exist with great force propagation forward, or with limited force propagation in an uncoordinated manner.  I have been in both kinds of crowds, and probably many of you have. Xenophon tells us of such pushing crowds trying to flee through a gate, ships at sea can crowd each other, even opposing arguments can jam to a stand-still. Roman's did it a bit at Zama, so the definition has to be quantitative, not qualitative.

Smythe is clear I think on what happens when two pike formations plow into each other: "after they have given their first thrust with their pikes and being come to join with their enemies front to front and face to face, and therefore the use and execution of the pikes of the foremost ranks being past, they must presently betake themselves to use of their swords and daggers"

He also tells us that the front ranks cannot foyne with their pikes unless they open their formation due to the closeness of the men behind them (and here I am reminded of Arrian's use of othismos in describing the crowding of the second rankers upon the first): "to the intent that they may have elbow roome enough without and impediment by the nearness of the ranks behind them, to pul backe their armes, and to thrust at their enemies approaching them at all the length they can of their armes and piques, and again with dexteritie to pull backe, and retire them to give new thrusts"

So, there would have been crowding within the sarissa phalanx pushing men forward, but not the kind of sustained pressure you can achieve with hoplites.  That said, Smythe also advocates the front ranks moving forward with swords, and if sarissaphoroi ever found themselves in dissarray and fighting shield on shield like hoplites, then it could occur.  But this is unlikely, because like the later pike men, it seem that the rear ranks would retain their pikes and still try to foyne around the front-like fighters rather than drop them and move up.

Now a caution.  It is not clear that a sarissa phalanx, formed closer can put out more force per square meter than a hoplite phalanx.  The problem is that the hoplite shields overlap much more solidly, which helps to synchronize the movement of men.  Also, I believe that the sarissa close order of 45cm per man required a sideways stance which allows no othismos of any kind.  In fact it probably arose specifically to counter any attempt at othismos by hoplites because you cannot push against a bunch of spear points and expect to live a long happy life.

Dangun mentioned the old argument that othismos would be too deadly to engage in.  This is not true, othismos is not comfortable, but unless your shield is crushed, as seems to have happened to some at Coronea, it is impossible to die. The pushing part of Othismos is actually too safe to account for battle casualties.  It is the vicious, short range sword or knife-fight, so close that you can smell the cheese and onions on your foe's breath, that kills you.

As of yet no one has seriously challenged the arguments I have put forward for othismos on grounds other than we cannot prove Greeks did it.  Fair enough, we were not there. But the possibility of elements of othismos as I present have not been refuted. Many of the old-school arguements against it can be shown to be frivolous. I think Mathew is the only author to try, but he clearly had no idea of the mechanics involved and spouted the old "men will be crushed", "12 ranks can't stand up to 50", "files can't generate such forces" (comical when the other argument is men will be crushed), and "no one would let you propel them into a hedge of spears".

[Justin Swanton]

My 2 drachmas:

A sarissaphoroi cannot push with a sarissa in the manner that hoplites pushed.  The reason has nothing to do with spacing, etc.  It is because the linkage between man and sarissa is no strong enough to support the force levels we see with hoplites.  You cannot grip a spear shaft strong enough to support a half ton of mass.

But it doesn't support half a ton. If five or six pikes are simultaneously involved in the sarissmos then each pike has to bear a maximum of 500kg/5 = 100kg or 500kg/6 = 83kg. A pole vaulter lifts his own weight with ease and the average weight of a polevaulter is 79kg.

But did sarissaphoroi engage in "othismos"?  Surely they did, the author's are quite clear.  But othismos just refers to a very crowded condition. This can exist with great force propagation forward, or with limited force propagation in an uncoordinated manner.  I have been in both kinds of crowds, and probably many of you have. Xenophon tells us of such pushing crowds trying to flee through a gate, ships at sea can crowd each other, even opposing arguments can jam to a stand-still. Roman's did it a bit at Zama, so the definition has to be quantitative, not qualitative.

I'm not quite clear on this. If you have force propagation forwards, and your men are in a straight file pushing shield against back, surely that is a pike version of othismos?

Smythe is clear I think on what happens when two pike formations plow into each other: "after they have given their first thrust with their pikes and being come to join with their enemies front to front and face to face, and therefore the use and execution of the pikes of the foremost ranks being past, they must presently betake themselves to use of their swords and daggers"

He also tells us that the front ranks cannot foyne with their pikes unless they open their formation due to the closeness of the men behind them (and here I am reminded of Arrian's use of othismos in describing the crowding of the second rankers upon the first): "to the intent that they may have elbow roome enough without and impediment by the nearness of the ranks behind them, to pul backe their armes, and to thrust at their enemies approaching them at all the length they can of their armes and piques, and again with dexteritie to pull backe, and retire them to give new thrusts"

So, there would have been crowding within the sarissa phalanx pushing men forward, but not the kind of sustained pressure you can achieve with hoplites.  That said, Smythe also advocates the front ranks moving forward with swords, and if sarissaphoroi ever found themselves in dissarray and fighting shield on shield like hoplites, then it could occur.  But this is unlikely, because like the later pike men, it seem that the rear ranks would retain their pikes and still try to foyne around the front-like fighters rather than drop them and move up.

This describes Renaissance pike formations for which I've seen no evidence of any kind of othismos. Lacking concave shields they couldn't have done it anyway.

Now a caution.  It is not clear that a sarissa phalanx, formed closer can put out more force per square meter than a hoplite phalanx.  The problem is that the hoplite shields overlap much more solidly, which helps to synchronize the movement of men.

Phalangite shields overlap just as solidly. They are smaller but the men are closer together so there is the same degree of overlap.

Also, I believe that the sarissa close order of 45cm per man required a sideways stance which allows no othismos of any kind.

Evidence for this?

In fact it probably arose specifically to counter any attempt at othismos by hoplites because you cannot push against a bunch of spear points and expect to live a long happy life.

And yet that seems to be exactly what happened at Sellasia.

[Dangun]

As of yet no one has seriously challenged the arguments I have put forward for othismos on grounds other than we cannot prove Greeks did it. 

Historicity, is a fairly significant exception.

But did sarissaphoroi engage in "othismos"?  Surely they did, the author's are quite clear...  As of yet no one has seriously challenged the arguments I have put forward...

To confidence with which a statement is made does not necessarily enhance its credibility.

[Patrick Waterson]

To confidence with which a statement is made does not necessarily enhance its credibility.

Although lack of refutability, or at least of refutation, does.

About the shields ...

If we visualise a phalanx from above, at one-cubit spacing each man probably has an angled body and definitely has an angled left arm and hence an angled shield.  I am assuming the centre of the shield more or less coincides with the elbow in order to let the left hand protrude sufficiently beyond the edge to grasp the sarissa shaft.

This gives us an angled shield, although one which would appear to be angled in the opposite direction to the body of the man in front.  Does this prevent the application of a push by the man behind?  I think it would still be possible, as the push would presumably be transmitted through chest, forearm and the part of the shield between elbow and hand as opposed to the whole shield.  And since the push would be delivered against what is essentially a cylinder (the torso of a man in a thorax) I do not think the angle would be an insuperable problem.

Exactly how much force a sarissa holder can bear and still hang onto his shaft is a matter beyond my physics, but to my mind the key point is that sarissa points would be going through targets and out the other side long before the limit was reached.  To get another phalanx moving back Sellasia-style would presumably require only as much pressure as makes the opposing front rank start to stagger back - and the side which reaches this threshold first will drive the other backwards until something else intervenes.

My current thinking is thus that an important part of phalanx fighting (as opposed to pike fighting, which I see as a rather broader spectrum) is the ability to close up the file spacing swiftly and with style, adding more - shall we call it 'thrust'? - more rapidly than the opponent can, and doing so without (this is important) any of one's own side cannoning into people ahead and knocking them off balance.  This kind of coordinated cohesive closure and concomitant thrust build-up would be what veterans such as the original Argyraspides could perform in the twinkle of an eye, with speeds and distances perfectly synchronised, while less expert opponents were still closing up and trying not to push each other over.

[Erpingham]

This describes Renaissance pike formations for which I've seen no evidence of any kind of othismos.

I will say again that renaissance pike fighting gives us an idea of what fighting in close formation with long pointy sticks might be like but we can't be sure the solutions they came up with are the same as those from a separate tradition nearly 2000 years before.

However, Smythe's description of pike fighting is about the most detailed we have from its period.  Obviously not everyone has picked up on the key points (though Paul clearly has).

Smythe derides pike foyning - it opens the formation to give people room to move weapons and reduces the supporting ranks to spectators.

What he prefers is to close up the formation from the sides and rear.  The first four ranks present their pikes at high charge, the remainder pikes upright.  The pike heads are in four even rows, 1 yd behind the previous rank (sound familiar?).  The ranks pack tight "as close as they can possiblie march pace with pace and step with step, as if they were one entire body" (this for some reason makes me think of a Genesis video  ).  On impact, the body all give a strong thrust at the faces of their enemy, which hopefully causes them to break.  If not, there is no room behind for the now extended pikes to be withdrawn for another thrust and the front men are best advised to draw sword and danger and stab their (still pike encumbered) foes.

So, by closing everyone up to virtually touching, the full weight of the formation is brought to bear.  Just like hoplites in a crowd-fight phase, the front men are using shorter weapons and actively trying to do the other guy harm, not just jostling and shoving.

What of this can apply a sarissa phalanx.  The closing tight on the back of the man in front can, and may be what the tacticians are talking about.  But did ranks of pikes slip past each other in a phalanx on phalanx encounter, to leave the file leaders using swords?  Or does that need a high hold on the pike , so that they slide over the opposition?  If we assume, as almost everyone does, that Hellenistic pikemen held pikes low, this couldn't happen.  So what did happen when five rows of points met five rows of points?

We have, I think plausibly explained the value of five rows of points against men with shields and spears.

[PMBardunias]

As of yet no one has seriously challenged the arguments I have put forward for othismos on grounds other than we cannot prove Greeks did it. 

Historicity, is a fairly significant exception.

But did sarissaphoroi engage in "othismos"?  Surely they did, the author's are quite clear...  As of yet no one has seriously challenged the arguments I have put forward...

To confidence with which a statement is made does not necessarily enhance its credibility.

It is not a matter of historicity. Half of those who read the relevant passages believe they are describing physical pushing, half don't. It is about differential interpretation and a lack of historical analogy. Without a time machine it cannot be resolved.

That is a neat trick on your second comment. You linked two comments which were unrelated in my original comments.  If you understood my position on what othismos was, then my surety in its existence in a Macedonian context is all but banal.  A state that the Greeks were describing as othismos is a feature of any close-in fighting in ranks. The difference is a matter of how much force the ranks generated.  In hoplites it could be huge and wholly survivable, in sarissaphoroi is was enough to keep their own front ranks moving forward and no allow backward movement, at Zama it was enough to at least herd the enemy ranks back if not push them in a coordinated manner.

[PMBardunias]

My 2 drachmas:

A sarissaphoroi cannot push with a sarissa in the manner that hoplites pushed.  The reason has nothing to do with spacing, etc.  It is because the linkage between man and sarissa is no strong enough to support the force levels we see with hoplites.  You cannot grip a spear shaft strong enough to support a half ton of mass.

But it doesn't support half a ton. If five or six pikes are simultaneously involved in the sarissmos then each pike has to bear a maximum of 500kg/5 = 100kg or 500kg/6 = 83kg. A pole vaulter lifts his own weight with ease and the average weight of a polevaulter is 79kg.

Sorry Rich, that is not how it works.  Each file is putting out force. if the force were being transferred from a spear 5 ranks back, unlikely because all of the spears would need to be getting shorter as they go forward in rank, then the force from the 6-8th rank would not be passed to the 4th rank in front. You have to choose one or the other.  But the whole unit is free to push with less force, which is a low energy type of othismos as I stated. The limit is really on how much force a shield can take from the point of a sarissa before being penetrated. Othismos requires that both sides resist, and I would not resist enough to get impaled.

Now a caution.  It is not clear that a sarissa phalanx, formed closer can put out more force per square meter than a hoplite phalanx.  The problem is that the hoplite shields overlap much more solidly, which helps to synchronize the movement of men.

Phalangite shields overlap just as solidly. They are smaller but the men are closer together so there is the same degree of overlap.

It is unlikely that sarissaphoroi stood any closer than hoplites with their peltae forward. It is not a matter of shield size, but human proportion. The average man's shoulders are 45 cm wide.  You cannot stand and fight in a 45cm space, you need room to be able to move your arms at least minimally- something closer to 60cm. This is also why I am sure they stood sideways at 45cm. I have yet to see anyone actually stand front on at 45cm and fight.  Mathew's measurements are bogus, the men are closer to 60cm than 45cm. The closest you will get is the shoulder to shoulder 18th-19th C marching formations.

In fact it probably arose specifically to counter any attempt at othismos by hoplites because you cannot push against a bunch of spear points and expect to live a long happy life.

And yet that seems to be exactly what happened at Sellasia.

It is not clear what happened at Sellasia.  One problem with the othismos debate is that people seem to want a level of consistency rarely met by ancient authors in their descriptions.  We have two things that occuss between masses of men: pushing and herding.  From the inside of a formation they feel the same.  Those men a few ranks back at Cannae or in the battle Procopius describes where the dead could not fall, felt all the sardine packing of othismos, yet it is unlikley that the enemy unit was physically pushing them in a consistent manner.  Instead they were herding the men into a tight space and the men's own outer ranks were pushing in- I just coined the term auto-othismos for this.  Cavalry could do this to infantry for example. Other times the opposing units are more likely physically pushing each other and transferring force from rear rank to rear rank, as in hoplite othismos.  Then we have what I think is going on in sarissa, where the rear ranks are pushing on their own front ranks, but the force is not being transferred above a minimal level through the pike shafts.

The difference between these is based on a) resistance. If the enemy, or your own men give more than a step, you are out of the packed condition. 2) The ability to transfer force forward. Here the sarissa fails. c) the ability to survive being crushed. Without an aspis or some other equivalent (I would love to see how renaissance breast and back plates held up to crushing) you have to either give ground, or push back on your own rear ranks, signalling them to let up. Unfortunately for the men at Cannae, stabby things were more inspiring that the pleas of their friends to spread out.

[RichT]

Just want to point out, if the 'Rich' was me, it wasn't me you are responding to but Justin. I'm with you on this one Paul, mostly!

As well as the difficulty of transmitting much push force through a held sarissa, there's the question of the sarissa itself. Modern reconstructions all have a certain degree of bend and if I understand anything at all about the mechanics of long slightly bendy poles, if you apply a very high degree of force to one end, they break. Matthew did force tests on his sarissa but just a two handed thrust test, not a sixteen man push.

Smythe and Montluc (or Monluc? Opinions seem to vary) seem to me to describe push of pike just as clearly, and in much the same words, as Polybius and the tacticians - they all talk about closing right up and pushing the man ahead while retaining one yard (two cubit) rank spacing. If that sounds strange then it is more likely to be due to us misunderstanding the realities of pike fighting than Polybius or Smythe or Montluc.

[Justin Swanton]

This describes Renaissance pike formations for which I've seen no evidence of any kind of othismos.

I will say again that renaissance pike fighting gives us an idea of what fighting in close formation with long pointy sticks might be like but we can't be sure the solutions they came up with are the same as those from a separate tradition nearly 2000 years before.

However, Smythe's description of pike fighting is about the most detailed we have from its period.  Obviously not everyone has picked up on the key points (though Paul clearly has).

Smythe derides pike foyning - it opens the formation to give people room to move weapons and reduces the supporting ranks to spectators.

What he prefers is to close up the formation from the sides and rear.  The first four ranks present their pikes at high charge, the remainder pikes upright.  The pike heads are in four even rows, 1 yd behind the previous rank (sound familiar?).  The ranks pack tight "as close as they can possiblie march pace with pace and step with step, as if they were one entire body" (this for some reason makes me think of a Genesis video  ).  On impact, the body all give a strong thrust at the faces of their enemy, which hopefully causes them to break.  If not, there is no room behind for the now extended pikes to be withdrawn for another thrust and the front men are best advised to draw sword and danger and stab their (still pike encumbered) foes.

So, by closing everyone up to virtually touching, the full weight of the formation is brought to bear.  Just like hoplites in a crowd-fight phase, the front men are using shorter weapons and actively trying to do the other guy harm, not just jostling and shoving.

What of this can apply a sarissa phalanx.  The closing tight on the back of the man in front can, and may be what the tacticians are talking about.  But did ranks of pikes slip past each other in a phalanx on phalanx encounter, to leave the file leaders using swords?  Or does that need a high hold on the pike , so that they slide over the opposition?  If we assume, as almost everyone does, that Hellenistic pikemen held pikes low, this couldn't happen.  So what did happen when five rows of points met five rows of points?

We have, I think plausibly explained the value of five rows of points against men with shields and spears.

OK, I think I get it. Using Renaissance pikefighting as a benchmark, we are to understand the 'weight' of a jhellenistic pike phalanx in a metaphorical sense, in that the men behind, by crowding forward, oblige the men in front to fight, even when they can no longer use their pikes, being obliged to resort to their secondary weapon. There is no serious physical pressure as such, just some persuasive nudging. Is that about it?

This however strains the text of Polybios and the tacticians, who speak of physical pressure of the bodies of the rear rankers making "charge very forcible": the men "press on with the weight of their bodies." It's a real physical pressure, not just a closing of the back door to prevent front rankers from taking early military retirement.