cavalry wheeling

[Patrick Waterson]

Thank you for laying out your thoughts, Richard, as this enables us to pinpoint where the questions arise.

1) The problem of moving bodies of cavalry in close order across terrain, as efficiently as possible and with minimum disorder (which we can take as read as militarily desirable) hasn't changed over the centuries, so while different solutions might have been found at different times (such as the wedge, which AFAIK isn't common in other periods), the fundamental problem and therefore the fundamental solution are unlikely to be very different.

But specifically how was it done?  Generalised philosophy unfortunately tells us nothing about technique, and can be seriously misleading when battlefield conditions have changed considerably over the centuries.

2) There is no evidence that ancient cavalry manoeuvred in some fundamentally different way from cavalry of other periods; while there is little evidence (but see below) of how they did manoeuvre, in the absence of evidence to the contrary our starting position should be that similar problems were met with similar solutions. While different theories could be devised, in the absence of any evidence to support them, or any reason to prefer them to the default position, they remain just interesting theories, and cannot supplant the default position.

Although the organisation and command structure of 'ancient cavalry' seems to have differed somewhat, e.g. officering by file, and form tends to follow function.

3) Such evidence as there is for wheeling in the Hellenistic tactical manuals may apply only to infantry, but it is reasonable to assume, in the absence of contradictory evidence, that it would apply to cavalry also, and as formations are described as turning like a single body, which seems to mean wheeling in the way we are familiar with from other periods, it is reasonable to suppose that this is how cavalry wheeled also.

Although this tells us nothing about the manner in which they turned, whether by rank or by file.  Given that they were organised by file, one would expect them to manoeuvre as a series of files as opposed to a series of ranks.

4) Byzantine tactical manuals (so far as I know them) also don't give precise details but seem to describe sub units (bandon size, around 300 men) wheeling as above, as single bodies, and larger formations wheeling by bandons aligning on each other, which is what we would expect from other periods also.

This is what I would have thought: the subunits wheel individually and the unit as a whole sorts itself out at the conclusion of the procedure.  This is pretty much what I have been attempting to convey, or thought I was.  That the individual files within a subunit would align on each other I would have thought elementary.

5) The wedge and rhombus are a way to ease the wheeling of a formation as a whole by following the leaders/edges; there is no suggestion that this involved wheeling by files (for example) nor that it replaced wheeling by files in square/oblong formations - the whole point of wedge/rhombus seems to be that it is an improved way to wheel the unit as a body.

No disagreement there, although I wonder at the 'no suggestion that this involved wheeling by files ... nor that it replaced wheeling by files in square/oblong formations' as there is equally no suggestion that it did not involve such a process.

6) People with practical experience (such as reenactors) of moving formations on horseback come up with similar solutions.

In other words, people who begin with a 20th/21st century equestrian tradition come up with what they perceive to be a 19th century solution to impose upon a 2nd century problem.

Some may consider the question of manoeuvre by rank or by file trivial - and perhaps it is.  Where it seems to me to be important is that there is a widespread tendency among authors to underestimate historical depths, both for cavalry and infantry, and we could maybe benefit from keeping in mind classical depths and classical organisational principles when attempting to second-guess tactical procedures from the classical era.

[Dangun]

Is it Drungus, Δροῦγγος and Δρουγγιστί – a Gallicism and Continuity in Roman Cavalry Tactics?

Going through the bibliography, I found that Cavalry Operations in the Ancient Greek World (Gaebel, 2002) is available online as a PDF... http://www.kavehfarrokh.com/wp-content/uploads/2010/02/Gaebel_202_Cavalry_Operations_in_the_Ancient_Greek_World.pdf

[Jim Webster]

Is it Drungus, Δροῦγγος and Δρουγγιστί – a Gallicism and Continuity in Roman Cavalry Tactics?

Going through the bibliography, I found that Cavalry Operations in the Ancient Greek World (Gaebel, 2002) is available online as a PDF... http://www.kavehfarrokh.com/wp-content/uploads/2010/02/Gaebel_202_Cavalry_Operations_in_the_Ancient_Greek_World.pdf

grabbed that one thanks , now to find time to read it 

[Mark G]

You really are overplaying the idea the organisation by files determined all movement, Pat.

18th, 19th, 20th century cavalry was also organised by files.  It had no bearing on movement whatever.

And in all probability, so did all earlier periods too .

It proves nothing to this discussion.

[Patrick Waterson]

You really are overplaying the idea the organisation by files determined all movement, Pat.

It did not 'determine all movement'; it simply meant that cavalry did not manoeuvre by ranks and this has a bearing on the way the Hadrian's Cavalry re-enactment group are trying to do things.  There was also obvious suitability for a tactical repertoire that included javelin-shooting by files.

It did of course have a bearing on how cavalry units were handled, and one good feature of Cavalry Operations in the Ancient Greek World (Gaebel, 2002) is that the author brings out the difference in pace and timing between a classical battle and a gunpowder era or modern one. 

18th, 19th, 20th century cavalry was also organised by files.  It had no bearing on movement whatever.

Actually on the battlefield they usually organised by ranks.  Napoleonic practice was to form in two single-rank lines about 200 yards apart, the second line acting as a support for the first and being able, if the first line was demolished by musketry during a charge, to sweep into the infantry while they were reloading.  Compared to the classical period, the gunpowder battlefield was a completely different tactical environment and cavalry techniques and handling were optimised for that battlefield (and remained so even when battlefield conditions again changed during 1861-1914.  New techniques then enabled cavalry to be effective in a more limited way from 1918 to 1945 and further developments in tactics and handling enabled Grey's Scouts in Rhodesia to be very useful up to 1979).

[Mark G]

Ffs, it's back to square one again.

Look.  The basic organisational unit in the early modern period was the file.  They did not manoeuvre by files, but that was the basic organisational unit.
File, then three's, etc, up to half squadrons, squadrons, and regiments'.

But movement was by squadron (and half sqdn for deploying).

They did not move by ranks, they moved by sqdn but MAINTAIN ranks while moving.

You cite ancient cavalry as being organised by files.  Ok, so what?

That has absolutely no effect on how they move, and you cannot cite any evidence that it does.

[Patrick Waterson]

They did not move by ranks, they moved by sqdn but MAINTAIN ranks while moving.

You cite ancient cavalry as being organised by files.  Ok, so what?

That has absolutely no effect on how they move, and you cannot cite any evidence that it does.

"The formation that would add most to the beauty of the exercises at the inspections has already been explained. Provided his horse is strong enough, the leader should ride round with the file that is on the outside every time. He will be galloping all the time himself, and the file whose turn it is to be on the outside with him will also be galloping. Thus the eyes of the Council will always be on the galloping file, and the horses will get a breathing space, resting by turns." - Xenophon, Cavalry Commander 3.9

We may note that the unit of cohesion during the wheel is the file, not the rank.

The mistake made by the Hadrian's Cavalry re-enactors has been to organise their turma by ranks instead of by files, which means their 30-man formation is incorrectly deployed at 10 wide and 3 deep when it should be the other way around.  As a result, it takes at least three times as long as it should to complete a wheel and gives a misleading impression of how a turma appeared on the battlefield or exercise ground.  This was the original point at issue and everything I have seen in this discussion convinces me it is upheld.

[Mark G]

I see you tit-bit of Xenohpon, and raise you a much more specific bit.

"Now the state has divided the cavalry into ten separate regiments. I hold that within these you should, to begin with, appoint file-leaders1 after consulting each of the colonels, choosing sturdy men, who are bent on winning fame by some brilliant deed. These should form the front rank. [3] Next you should choose an equal number of the oldest and most sensible to form the rear rank. To use an illustration, steel has most power to cut through steel when its edge is keen and its back reliable. [4]
To fill the ranks between the front and rear, the file-leaders should choose the men to form the second line, and these in turn the men to form the third, and so on throughout. In this way every man will naturally have complete confidence in the man behind him. [5]

You must be very careful to appoint a competent man as leader in the rear.2 For if he is a good man, his cheers will always hearten the ranks in front of him in case it becomes necessary to charge; or, should the moment come to retreat, his prudent leadership will, in all probability, do much for the safety of his regiment. [6]
An even number of file-leaders has this advantage over an odd, that it is possible to divide the regiment into a larger number of equal parts.
The reasons why I like this formation are these. In the first place, all the men in the front rank are officers; and the obligation to distinguish themselves appeals more strongly to men when they are officers than when they are privates. Secondly, when anything has to be done, the word of command is much more effective if it is passed to officers rather than to privates. [7]
Let us assume that this formation has been adopted: every file-leader must know his position in the line of march by word passed along by the colonel, just as every colonel is informed by the commander of his proper place in the charge. For when these instructions are given there will be much better order than if the men hamper one another like a crowd leaving the theatre. [8] And in the event of a frontal attack, the men in the van are far more willing to fight when they know that this is their station; so is the rear-rank in the event of a surprise attack in the rear, when the men there understand that it is disgraceful to leave their post. [9] But if no order is kept there is confusion whenever the roads are narrow or rivers are being crossed; and when an action is fought no one voluntarily takes his post in the fighting line. "

Xenophon, Cavalry Commander II

Look Patrick,
Ranks!!!
Regiments.
movement orders issued to regiments not files.
Emphasis on order.
Files only referenced in relation to knowing relative position to follow. (which is EXACTLY the same as early-modern usage of cavalry files)
Officers in the FRONT LINE, required to know their position in the FRONT LINE.
etc.


Chapter III, on the other hand, from which you selectively quote your supposed proof, is a lot more specific than you would let on.

for Chapter III directly relates to processions and festivals, not to military organisation and movement.  In fact, the emphasis of the chapter itself is on delighting spectators.  it is a bit like taking the red arrows displays as indicative of sound air combat techniques.

One wonders how you managed to not read chapter two on your way to finding your selectively quoted proof from chapter three.

probably by using the same method that led you to not notice this bit from chapter five

[6] Another way of exaggerating the apparent strength of your force is to arm the grooms with lances or even imitation lances, and put them between the cavalrymen, whether you display the cavalry at the halt or wheel it into line. Thus the bulk of the company is bound to look denser and more massive [7]
On the other hand, if your object is to make a large number look small, then, assuming that your ground affords cover, you can obviously conceal your cavalry by having part in the open and part hidden. If, however, the whole of the ground is exposed, you must form the files into rows and wheel, leaving a gap between each two rows1.

form your files into rows and wheel.  hardly the manoeuvre by file that you would have us believe.

or this from chapter 8

Should it happen at any time that the cavalry forces engaged are about equal, I think it would be a good plan to split each regiment into two divisions, putting one under the command of the colonel, and the other under the best man available. [18] The latter would follow in the rear of the colonel's division for a time; but presently, when the adversary is near, he would wheel on receiving the order and charge. This plan, I think, would make the blow delivered by the regiment more stunning and more difficult to parry

yet again, regiments and half regiments are the manoeuvre unit, wheeling by such is described.  ditto, [23] and [24] - more regiments wheeling.

I also note, Xenophon, Greek, circa 400 BC.
Hadrian, Roman circa 120 AD
some 500 years difference. 
we see front ranks emphasised in medieval, in renaissance, in the horse and musket era, and in the modern era.  where we have such evidence, it also specifies organisation by file, but movement by squadron or troop with an emphasis on maintaining ranks.  clearly, changing eras is not relevant to the very basics.


now we must let you have the last word, because no thread can possibly end without you having the last word, and it is inconceivable that you could be grown up enough to admit you are simply wrong, so go to it.  but I think there is only one person remaining who thinks as you do, and you can spot him in the mirror.

[Patrick Waterson]

Mark, the authoritative text is the Greek.  'Ranks' is a translator's interpolation.  'Colonel' is a mis-rendering of phylarch and 'regiments' mis-conveys phylai.  Athens raised ten separate cavalry contingents, presumably one from, or associated with, each of the ten traditional phylai, 'tribes'.

'Ranks' is a complete red herring here.  Take "These should form the front rank." (Xen CC 2.2)  In the original Greek, this is kai toutous men prōtostatas dei einai, which translates as: "And these (toutous) the sum of (einai]) the file leaders (prōtostatas) [should form being understood]."

There is no word in the Greek which can be taken as 'rank' here or anywhere else in Cavalry Commander.  (The nearest word to 'rank' in Greek appears to be metapon, which normally means 'front' or 'frontage'.)

Now look at the way the unit is organised and how control is exerted: "Let us assume that this formation has been adopted: every file-leader must know his position in the line of march by word passed along by the phylarch, just as every phylarch is informed by the hipparch of his proper place in the charge."

The building-block and unit of organisation, and for that matter movement, is the file, controlled by its leader.  Obviously a unit manoeuvres as a unit, but not by ranks.

The translator is also misleadingly free in 5.6:
"whether you display the cavalry at the halt or wheel it into line"

The Greek is: ēn te hestēkos epideiknuēs to hippikon ēn te paragēs

En = if; te = you; hestēkos = make to stand; epideiknuēs = display, show off, exhibit; to hippikon = the cavalry; en te = if you; paragēs = march the men up from the side, bring them from column into line.

For some reason, the translator likes to render 'paragego' as 'wheel', even though this is not supported by the lexicon and does not seem appropriate to the context, e.g. Hellenica VII.5.22 where Epaminondas, having drawn up his line, proceeds to deepen (not 'wheel') his left.

Again, in V.6, "If, however, the whole of the ground is exposed, you must form the files into rows and wheel, leaving a gap between each two rows."

The Greek is: ēn de pan kataphanes ē to khōrion, dekadas khrē stoikhousas poiēsanta dialeipousas paragein

ēn de pan kataphanes ē to khōrion = If the whole of the locality is exposed (devoid of concealment); dekadas = the ten-man files; khrē = proclaim, order; stoikhousas =  draw up in a line or row; poiēsanta = make happen; dialeipousas = leaving an interval between; paragein = march the men up from the side, bring them from column into line.

Curiously, enough, this does actually involve a wheel, as the end file is being deployed parallel with the enemy, raising its spears to the vertical, while the remaining files lower theirs and thus are partly concealed, make it impossible for the enemy to estimate their numbers.  The unit is presenting its flank to the enemy in order to conceal its strength.  Note again how the file is the key component unit and the files from in 'rows' but not ranks; each 'row' is simply a file facing the enemy along its length.  This is a deception technique, not a combat formation.

So, Mark, I can see where you get excited over this, but do be aware that there is no basis in the Greek for what you think you have discovered.  Greek cavalry, like Roman, organised and operated by files.  A unit was made up of files.  When it manoeuvred, the files kept station on each other.  This has nothing to do with individual maturity, only source statements.

[RichT]

Please consider this an addendum and not denying Patrick his last word, or Mark his conclusion - but there are just a couple of points of fact to clarify, and one of opinion, before the thread dies a natural death.

Take "These should form the front rank." (Xen CC 2.2)  In the original Greek, this is kai toutous men protostatas dei einai, which translates as: "And these (toutous) the sum of (einai]) the file leaders (protostatas) [should form being understood]."

Not quite - einai is a verb (eimi). "These should be the protostatas".

There is no word in the Greek which can be taken as 'rank' here or anywhere else in Cavalry Commander.  (The nearest word to 'rank' in Greek appears to be metapon, which normally means 'front' or 'frontage'.)

Not true - such a word is protostatas - the 'standers in front' or 'front rankers' - also the 'file leaders' of course, since the totality of the file leaders in a formation is, naturally, the front rank, just as the totality of file closers is the rear rank, the totality of second-in-file is the second rank, and so forth. This is clearly shown in (for example) the very familiar Polybius 18.29-30 (and I trust nobody would argue that the phalanx constituted individual files with no ranks, and where you also find the, also familiar, expression 'kat epistaten kai kata parastaten', 'by standers-behind and standers-beside' or 'by file and rank'). Greeks and Romans aren't as precise in their language as we might like - look at Latin 'ordinum', which serves for rank, file and formation generally. Greek tactical language is based around the file but it is just two ways of describing the same thing.

And to go beyond the simply factual:

The building-block and unit of organisation, and for that matter movement, is the file, controlled by its leader.  Obviously a unit manoeuvres as a unit, but not by ranks.

This discussion is dogged by the fact that there is no practical difference I can see between 'manoeuvring by file' and 'maoneuvring by rank', unless there is some insistence that files manoeuvred individually (each file waiting until the one beside had completed its manoeuvre), which is self-evidently absurd. Otherwise the practical result is the same. In the case of the Hadrianic reenactors, as a unit of 30 they might well have been better off forming column of threes to perform their manoeuvres, which would be easier - but which isn't manoeuvring by files, it is manoeuvring in column of threes by 30-man subunit. The same applies to Patrick's exposition of his theory on the first page of this thread. But in the case of actual cavalry and infantry it is clearly stated that turning maneuvres were done by the syntagma (Asclepiodotus) or bandon (Maurice) - units of 256-300 men with a front of 16-30, depending on depth. These manoeuvred as single units, and there is no evidence that they split into column of threes (or other subunits of 30 or so) to manoeuvre, still less that they (somehow) broke up into individual files. The practical result would therefore be a turning manoeuvre by ranks 16-30 men long (in which we can all happily grant that the rest of the files would follow the file leader - what else could they do?) The files of the formation would retain the same relative positioning throughout, thus maintaining ranks, as in all other known periods.

Hopefully we can leave it at that, unless to consider applications in wargaming terms again (such as - do wargaming rules take (enough) account of different formations - lines v. drungi, etc).

[Erpingham]

Hopefully we can leave it at that, unless to consider applications in wargaming terms again (such as - do wargaming rules take (enough) account of different formations - lines v. drungi, etc).

We might really need to explore that question separately from wheeling.  It seems to me that we might get entangled in the nature of wedge/rhomboid/flight of cranes formations.  Some, as we know from previous discussions, believe these formations were used to provide almost magical combat advantages, as opposed to being primarily manoeuver strategies.

One thing which seems uncontested about ancient cavalry formations is they tended to be deeper than later Western style (17th century on) ones.  I suspect it would be almost impossible to turn these on a fixed pivot, so they would use a moveable pivot.  The swinging gate pivoting move to swing a unit into the flank of another, so satisfying on a wargames table, would seem very doubtful in practice.

[Patrick Waterson]

One thing which seems uncontested about ancient cavalry formations is they tended to be deeper than later Western style (17th century on) ones.  I suspect it would be almost impossible to turn these on a fixed pivot, so they would use a moveable pivot.  The swinging gate pivoting move to swing a unit into the flank of another, so satisfying on a wargames table, would seem very doubtful in practice.

A very good point, although here we meet the divide between simulation and reality.  In reality, if you as a classical cavalry commander want to position your formation to hit an opponent in flank, you will anticipate and be ready rather than waiting for the opponent to finish his bound and then begin yours.  Hence the 'swinging gate', although geometrically and vectorially dubious, is probably a closer approximation to reality in overall effect than insisting upon the unit concerned having an increment of forward movement in conjunction with the wheel.

This matter of intersecting timings is responsible for many wargames abstractions, e.g. the DBA base frontage matchup and ability to pivot on the spot to face a charge, which seem inexplicable in terms of actual battlefield manoeuvring.

There is no word in the Greek which can be taken as 'rank' here or anywhere else in Cavalry Commander.  (The nearest word to 'rank' in Greek appears to be metapon, which normally means 'front' or 'frontage'.)

Not true - such a word is protostatas - the 'standers in front' or 'front rankers' - also the 'file leaders' of course, since the totality of the file leaders in a formation is, naturally, the front rank, just as the totality of file closers is the rear rank, the totality of second-in-file is the second rank, and so forth.

My turn for a nit-pick.

One may see in this the concept of a rank, but not the organisational or operational use of one.  Polybius himself is reduced to using zugou, from zugon, literally a 'yoke-fellow', to delineate soldiers standing side by side (XVIII.29.5).  While such placement obviously existed, the concept of the rank as an entity evidently did not.  (This may have changed by the time Arrian wrote his Order of Battle Against the Alans, if anyone wishes to see how the subject was handled there.)

And to go beyond the simply factual:

The building-block and unit of organisation, and for that matter movement, is the file, controlled by its leader.  Obviously a unit manoeuvres as a unit, but not by ranks.

This discussion is dogged by the fact that there is no practical difference I can see between 'manoeuvring by file' and 'maoneuvring by rank', unless there is some insistence that files manoeuvred individually (each file waiting until the one beside had completed its manoeuvre), which is self-evidently absurd. Otherwise the practical result is the same.

It is the same in that the unit gets from point A to point B in the same order it started, but confusing the two leads to mistakes in re-enactment such as putting one's decurions at the end of a rank rather than the beginning of a file, with concomitant misrepresentation of formation (10x3 instead of 3x10) and loss of historicity.  The distinction in theory is minor, as the men are going to be keeping the same relationship to each other whether organised by rank or file, but significant in that various classical tactics, (e.g. the 'caracoling' file delivering a constant output of javelins) become impossible if one tries to operate by rank.  When attempting re-enactment, such apparently trivial distinctions do matter.

[Patrick Waterson]

One thing which seems uncontested about ancient cavalry formations is they tended to be deeper than later Western style (17th century on) ones.  I suspect it would be almost impossible to turn these on a fixed pivot, so they would use a moveable pivot.  The swinging gate pivoting move to swing a unit into the flank of another, so satisfying on a wargames table, would seem very doubtful in practice.

Thinking further about this, it does bring up the question of what would happen if we had the classic tabletop situation of infantry units engaged with infantry units, but one side has a cavalry unit standing next door to the end friendly infantry unit but unengaged.

1) Would this actually happen in real life?
2) If it did, what sort of manoeuvre would the cavalry perform in order to close with and grievously harm the enemy?

The 'revolving gate' pivoting on the man closest to the enemy seems out (or is it?).  My own particular conjecture would be a 'J-shaped charge', i.e. moving forward beyond the enemy line, turning round and then having a jolly good charge into the enemy rear.  A file or two could be left to cap the enemy flank for neatness if desired.

Any thoughts?

[Jim Webster]

The 'revolving gate' pivoting on the man closest to the enemy seems out (or is it?).  My own particular conjecture would be a 'J-shaped charge', i.e. moving forward beyond the enemy line, turning round and then having a jolly good charge into the enemy rear.  A file or two could be left to cap the enemy flank for neatness if desired.

Any thoughts?

the problem with the J shaped charge is that if both units are moving there's a risk that they'd both miss.

[Erpingham]

Thinking further about this, it does bring up the question of what would happen if we had the classic tabletop situation of infantry units engaged with infantry units, but one side has a cavalry unit standing next door to the end friendly infantry unit but unengaged.

1) Would this actually happen in real life?
2) If it did, what sort of manoeuvre would the cavalry perform in order to close with and grievously harm the enemy?

Any thoughts?

Good questions and ones I'd also thought about after the initial exchange above.

Whether it would happen in real life is surely a matter of how we got to be in the situation.  I can't see a cavalry unit moving up to formate on the corner of an enemy infantry unit.  But if it had originally been engaged to the front and it's enemy had disengaged or fled, something akin to this could happen.

What it would do would again depend on the battlefield situation.  If moving up with no opposition to interfere, I think it would seek to swing out wide of the fight and deliver a charge into the enemy flank or rear.  If opposition is about, it might attack it or hang back to cover the flank of its own infantry.

If it came to be in the situation by departure of its enemy to the front, it would come down to discipline.  Would it break ranks to pursue, in which case parts of it would doubtless clash with the flanks of the infantry unit.  If it held discipline, it's leaders may still feel that pursuit of the enemy to the front is the best option.  It might push on past the flank then regroup behind the infantry, which would return us to whether it would be free to deliver an attack into the enemy rear.

In all of these options, there is the question of how much would our cavalry be working to strict parade ground formation in the midst of a melee?  We know that impact cavalry down the ages knew the idea that it should try to hit the enemy together, rather than in dribs and drabs.  What happens after that, though, seems to become more fluid.