Claudius and his elephants

[Duncan Head]

According to the sequence in Appian, Hasdrubal is sent to hunt elephants once news arrives of Scipio's preparations in Sicily, but before Scipio sails for Africa. Scipio crosses to Sicily in 205, and thence to Africa in 204, but I'm not sure (without checking sources for Scipio's timetable) if we can date it closer than that. Does Scipio sail in the spring of 204 as one might expect? Anyway, it looks as if Hasdrubal's safari starts at some time in 205.  The battle of Zama is in 202, normally dated to October. So to catch and train (at least the first batch of) these elephants, the Carthaginians have about three years.

What sources do we have on how long it takes to train a war-elephant? Arthasastra II.31 suggests perhaps five years (if elephants are to be captured at the age of 20 but are "of the lowest class" but presumably fit for service at 25). I suspect that in 2-3 years the elephants could be adequately trained, though by no means veterans.

[Mark G]

There is a lot of detail in the ancient battles thread on Zama from last year which is worth looking through.

I read it that some panicked early - trumpets etc - and ran to the edges of the battle line, where they were a significant contributor to the numidians scarpering far sooner than was expected.

The rest did stay and fight, were given a hard time by the javelins, and when they charged (or more accurately, re charged, since it was clearly a combat flurry sort of affair), often found they ran down a blind alley, where they were shot by javelins on all sides.

I don't see a simple elephant attack, loud noise, elephants run.
nor do I see a simple elephants all charge down the lanes and die

I think there was a good bit of fighting, but after the worst trained ones fled at the trumpet blasts (taking the numidians with them), it was just a matter of not-too-great time .

But my reading does concur with yours on Appian about training.

And Pat and I fundamentally disagree on most things on this battle.

[Patrick Waterson]

The Roman point of view (which may not be an expert opinion) is given in African War 27.

"Scipio meanwhile began to train the elephants in the following way.  He drew up two battle lines.  One was of slingers, to face the elephants in the role of the enemy and fire small stones against the front line formed by the beasts.  Then he drew up the elephants in a line and posted his own battle-line behind them, so that when the fire of stones began from the enemy side and the elephants wheeled round in terror towards their own side, his men should throw stones at them and drive them back again towards the enemy.  This is a slow and painful process; for elephants are unruly beasts who can be trained only with difficulty, by dint of many years' teaching and prolonged practice; and even then, when they are brought out into the battle line they are a menace to both sides."

Aulus Hirtius' gloomy assessment seems a little pessimistic, but he seems to think that 'years' (plural) are needed, which would make one wonder why Scipio would even bother with Caesar expected at any time.  It looks as if Scipio may have been attempting a crash course, a departure from normal training methods, in the hope of achieving in weeks what would normally take rather longer.

As it happened, when committed to battle the elephants followed their training with absolute fidelity.  The Battle of Thapsus opened with Caesar's slingers and archers sending their missiles at the dense mass of elephants on Scipio's left, which accordingly caused them to turn in the direction of their own troops.  These, however, did not emit the shower of stones customarily used in training - and so the elephants simply kept on going and caused havoc to their own side, which perhaps prompted Hirtius' comment.

Scipio Africanus did indeed set sail in the spring of 204 (Appian Punica III/13).  Hasdrubal's elephant hunt is mentioned in Punica II.9 and hence occurred in 205 BC.  Hence there would have been something like three years to train the new elephants.

[Citizen6]

I should probably change my wording a tad. I will say insufficiently trained rather than poorly trained as the latter implies something of the quality of the mahouts and we have no information either way in that regard.

I would suggest though that three years is the absolute maximum but it's more likely two to two and a half years. First the elephants have to be found and caught. This is a slow process in itself. There is reasonable evidence to suggest that the elephant population in North Africa was already in decline (keeping in mind that war was a very secondary industry to ivory). It would take a couple of months for the elephant to recover to a point where it could start working with a mahout due to the neck wounds and starvation that were part of the standard method of capture. Even today elephants require at least a year of conditioning to get to a point where they can be taken out and start learning work skills.  Most of that year is used getting the elephant accustomed to the mahout and the routine noise of living around cars and people. Now the more the animal is afraid of something the more time will be required. So I think it safe to assume that battlefield and combat sounds and sights would take longer. Intensification or quicker training does not work as it makes the animals more likely to rebel and thus ruin the bond with the mahout, who effectively has to start all over to rebuild trust.

Elephants are young when captured and like humans are not as emotionally stable until in their twenties. Duncan refers to a source stating five years to train but there are Indian sources (which I can't put my hand on at the moment) that say ten or even fifteen to do it properly. This puts the elephants in their early twenties (assuming the average capture age of five to ten years which apparently is ideal). So while the elephants may have been trained to carry infantry and be comfortable around masses of people they were likely quite young on average and well short of what is considered ideal in terms of training duration. As such, I am far from convinced that they were anything more than the equivalent of recruits in the Pacific in WW2 with four weeks of basic training under their belts. That is, casualties waiting to happen. Especially when you factor n the superior preparations of Scipio.

[Duncan Head]

Elephants are young when captured and like humans are not as emotionally stable until in their twenties. Duncan refers to a source stating five years to train but there are Indian sources (which I can't put my hand on at the moment) that say ten or even fifteen to do it properly. This puts the elephants in their early twenties (assuming the average capture age of five to ten years which apparently is ideal).

Arthashastra suggests that they should be captured at the age of twenty; so we may have differences within the Indian sources. Not thatthat is likely to have had very much bearing on Hasdurbal, who more likely just grabbed what he could get.

It is also true that, even if Hasdrubal had three years, this only applies to the first batch of elephants he acquired. Did he capture 80 beasts all at once, or are we to assume several captures over the period 205-04? If so, then even if there was time to train the first batch of elephants acquired, later "recruits" won't have had as long.

[Patrick Waterson]

... I am far from convinced that they were anything more than the equivalent of recruits in the Pacific in WW2 with four weeks of basic training under their belts. That is, casualties waiting to happen. Especially when you factor n the superior preparations of Scipio.

I would respectfully disagree: the very fact of Scipio's preparations indicates the seriousness with which he treated the elephants' combat potential.  The problem at Zama seems not to have been that the elephants or their training were deficient, but that Hannibal's battle plan went awry, as a result of which the elephants were not closely supported by the Celtic and other troops in the first line, who as far as I can establish were meant to follow them closely and take advantage of the gaps they created.

The question of Hannibal's battle pan has been discussed elsewhere on this site, but in a nutshell my understanding is that Scipio fooled Hannibal into believing that the latter was superior in cavalry, keeping Masinissa's arrival secret, and so Hannibal planned what was essentially a replay of Bagradas with refinements and variations plus a double envelopment with his line of veterans.  The elephants and Celts were supposed to go in together, followed and closely supported by the citizen spearmen, who would catch 'leakers' as their counterparts did at Bagradas.  Meanwhile Hannibal's cavalry would deal with Laelius' outnumbered and perhaps even outclassed troops and clear the way for a decisive envelopment on both flanks by the veterans.  All this went out of the wiindow when Hannibal suddenly realised as the armies finished drawing up that not only was he not superior in cavalry, he was actually outnumbered.

His cavalry were therefore unable to perform their part, and he seemed unable to improvise a new plan.  With the Carthaginian cavalry hesitating the Celts seemed to advance hesitantly, too, leaving the elephants unsupported, and when the Carthaginian cavalry withdrew the Carthaginian spearmen seem not to have moved at all, leading eventually to ructions with the Celts in the first line.

Had Hannibal been joined at the last minute by Vermina, who was en route with several thousand cavalry reinforcements, then Hannibal could have operated what I see as his original plan and the Roman and Numidian cavalry would have had their hands too full to bother with concentrated volleys at elephants.  Looking at the description of Hasdrubal's elephants at the Metaurus (Polybius XI.1):

"The elephants too had proved to be of no more service to one side than to the other, for as they were hemmed in between the two armies and exposed to a hail of missiles, they threw both the Roman and the Spanish ranks into disorder."

These, we may note, are elephants whose length of service and degree of training has not been questioned.  Their behaviour seems, if anything, less controlled than that of Hannibal's elephants at Zama, most of which actually stayed on course despite all the distractions.

[Mark G]

You do seem to be assuming one draft of elephants.

I would expect a small cadre retained in Carthage at all times.  a call comes in to get more, quickly, and a large hunt is sent out.  but I would also expect more (smaller) drafts of hunts coming in on a fairly regular basis.

given the depth of the emergency with scipio landing in Africa, all of these assets would have been sent out, and hence, varying degrees of training and hence also, varying degrees of success - some (less trained) running at the noise, others making a decent fist of it (there is no indication this was over that quickly), and some having a bit of local success.

[Patrick Waterson]

Probably a fair point, though there were still elephants in Carthaginian hands to be surrendered following Zama, indicating that even when making a 'maximum effort' they did not cut corners to include everything they had.  Hannibal of all people would appreciate that a poorly-trained elephant is a disaster waiting to happen.

We may note that Bagradas saw about 100 elephants fielded (just under) - this may have been Hannibal's own target figure, but it seems he had to be satisfied with about 80.

[Citizen6]

I cannot, despite a significant search, find the Indian sources I referred to in an earlier post. As such I'm not sure if they were an actual quote of an Indian source or a modern second hand interpretation of a source. Therefore the information can be ignored as having no evidentiary value for this discussion.

I would respectfully disagree: the very fact of Scipio's preparations indicates the seriousness with which he treated the elephants' combat potential.

I'm not sure this really indicates anything about the elephants. It does indicate that Scipio, as a good commander, prepared well. He would have known he would face elephants, or at least had a high expectation of such, given that the Carthaginians had deployed them in every other theatre. A good commander does not train to the level of the enemy. Training is carried out to improve strengths and exploit enemy weaknesses while maximizing enemy losses and minimizing your own. So I doubt Scipio would have prepared any differently, irrespective of what he thought of the elephant quality.

I totally agree with Mark's comment. There would have been some experienced elephants and I have generalized in my posts for the sake of brevity. But were they a significant number? I was reading Appian on Zama (Book VIII.43) and he gives a slightly different version to Polybius. Stating that the elephant wings folded quickly to the cavalry but the centre penetrated the Roman lines and caused some havoc and deaths amongst the armoured troops until Scipio brought up more cavalry and velites to assail them with javelins. This deployment is what I would expect, in that Hannibal would place his strongest elephants in the centre given he (if my understanding of his intended strategy is correct) was aiming to break through there.

Looking at the description of Hasdrubal's elephants at the Metaurus (Polybius XI.1):

"The elephants too had proved to be of no more service to one side than to the other, for as they were hemmed in between the two armies and exposed to a hail of missiles, they threw both the Roman and the Spanish ranks into disorder."

These, we may note, are elephants whose length of service and degree of training has not been questioned.  Their behaviour seems, if anything, less controlled than that of Hannibal's elephants at Zama, most of which actually stayed on course despite all the distractions.

To me, and I realize this is easy to say in hindsight from the comfort of my back verandah in the 21stC, Hasdrubal's deployment was a prime example of what not to do with elephants. He deployed a small number of elephants in an enclosed valley with narrowly deployed troops in front and behind. That, in and of itself, would be stressful to an elephant, as they would not be able to ascertain a clear means of retreat. Once the javelins started flying, it would only take two or three of such a small herd to panic for that to become a rout. Personally, I think he'd have been far better off using them to deny a flank rather than offensively.

Most of the sources (if not all) we have on training elephants are from the east and apply to a different species completely. We know both through the ancient sources and modern experience that the same techniques can be used for both species but that doesn't mean that both species are equally bold in combat. Scipio at Magnesia kept his sixteen elephants in reserve because he knew they would not compete with the Seleucid Indians. Due, in part, to the size differential but also to North African elephants supposedly being more timid. If this is true it might explain some of the elephant outcomes in the Punic wars. But, a consequence of such timidity, if it exists, is that North African elephants would take longer to train to war quality than Indians, all other things being equal. I have been unable to find any good information in this regard. It may be true, but it may a falsehood that has been repeated so many times that it becomes "truth". For example, the assertion that African elephants are harder to teach than Indians...it just isn't correct.

While there may have been some well trained / veteran elephants at Zama (probably in the centre), the Carthaginian's repeated deployment of small numbers of elephants (eg. 10 at Metaurus, 14 at Utica, 16 on each flank at Ilipa) as compared to the 70, 80 or even 140 that they regularly fielded a generation before in the First Punic War suggests they were struggling to keep up supply. It just seems a bit odd to me that two authors would specifically mention the Carthaginian effort to obtain more elephants for Scipio's invasion if this new batch didn't contribute significantly to their present pool of Punic pachyderms.   

[Patrick Waterson]

Funnily enough, Appian (Punic Wars III/13) has Hasdrubal with 140 elephants in 204 BC.

"Hasdrubal, Syphax, and Masinissa encamped not far from each other near the city of Utica, to which Scipio had been driven by the winds, and he also was camped hard by. Not far from him was Hasdrubal with an army of 20,000 foot, 7000 horse, and 140 elephants."

This is before Scipio's night attack on Hasdrubal's and Syphax's camps during a truce.

The small numbers of elephants in armies outside Carthage seem to have been a reflection of the difficulty of shipping them rather than the difficulty of training them.

The fact that it was the elephants on the ends of the line which broke away towards the cavalry I see as being more down to their position at the ends of the line, which left them with a quick way out of the 'noise zone', than any difference in training.  If anything, the ones at the end of the line would presumably be the best-trained, because they are acting as 'markers' for the rest and keeping them in line by the example of their presence.

Out of interest, how would you have expected Hannibal's elephants to have behaved at Zama if trained to the limit possible?  How would such elephants have reacted differently?

[Mark G]

The same as they did at bagradad

[Citizen6]

Funnily enough, Appian (Punic Wars III/13) has Hasdrubal with 140 elephants in 204 BC.

There's a couple of possibilities to this. Firstly, they are the newly captured elephants from his hunt that go on the be the 80 at Zama. Not all elephants are suitable for training so maybe they rejected 60 for battle. I don't buy this as a solution but I have seen it suggested (in Kistler I think). Secondly, and I don't know if anyone has even stated this previously, though to me it is more the far more likely, is an error in the text. Given Appian is relatively highly regarded as an author, I'm inclined to give him the benefit of the doubt. Personally, I suspect that there may have been a copyist's error somewhere along the line.  I find it a little too coincidental that the number of elephants captured and/or killed in Scipio's night attack against Hasdrubal and Syphax adds up to 14 (8 killed and 6 captured - Livy XXX 6.7-9).

"Both of the generals, however, made their escape, and out of so many thousand armed men two thousand infantry and five hundred horsemen escaped half-armed, many of the men wounded and scorched by the flames. [8] Slain or burned to death were some forty thousand men, more than five thousand captured, many Carthaginian nobles, eleven senators. [9] Of military standards a hundred and seventy-four were taken, of Numidian horses over two thousand seven hundred. Six elephants were captured, eight destroyed by sword or by fire."

Also, I cannot see how pretty much the entire army could have been slaughtered in this attack and yet 126 elephants somehow managed to escape (and then aren't subsequently noted as doing so by either Appian or Livy). 

The small numbers of elephants in armies outside Carthage seem to have been a reflection of the difficulty of shipping them rather than the difficulty of training them.

Not at all...Hanno has 50 at Heraclea (Polybius I.19); Hasdrubal has 140 (new) elephants at Lilybaeum (I.38); and supposedly Hasdrubal (son-in-law) had ~200 in Spain (Diodorus XXV.12). So shipping doesn't seem to have been too much of an issue during the First Punic War and the subsequent interbellum.

The fact that it was the elephants on the ends of the line which broke away towards the cavalry I see as being more down to their position at the ends of the line, which left them with a quick way out of the 'noise zone', than any difference in training.  If anything, the ones at the end of the line would presumably be the best-trained, because they are acting as 'markers' for the rest and keeping them in line by the example of their presence.

But to argue this would surely suggest that the best elephants (on the flanks) did the worst (fled without contact) while the worst elephants (centre) performed the best (penetrating and heavily disrupting the principes). (As per Appian)

Out of interest, how would you have expected Hannibal's elephants to have behaved at Zama if trained to the limit possible?  How would such elephants have reacted differently?

There's no way of knowing how that might have turned out. I suspect that Scipio's preparations were such that even veteran elephants would have had difficulty but the fact that a significant number got through in the centre and caused havoc suggests that the Romans may have had a worse time for it...probably not enough to lose but certainly not the rout that it was. I don't base my opinion on the outcome of Zama so much as the peripheral information that I've outlined. One thing that I haven't brought up yet but also contributes to my opinion, albeit as corroborating rather than direct evidence, is the difficulty that Ammianus Marcellinus has with the Persian elephants (History XXV.3.11).

"11. On the other hand the Persians, fighting with increased spirit, shot forth such clouds of arrows, that we could hardly see the shooters through them; while the elephants, slowly marching in front, by the vast size of their bodies, and the formidable appearance of their crests, terrified alike our horses and our men."

Admittedly these are Indians and some are armoured but both he and Julian (and presumably other senior officers) would have been very well versed in military history as part of their classical education and as such, in known anti-elephant tactics. Other parts of the History shows they employed many of these tactics and yet still had much trouble with the elephants.

[Patrick Waterson]

Funnily enough, Appian (Punic Wars III/13) has Hasdrubal with 140 elephants in 204 BC.

Personally, I suspect that there may have been a copyist's error somewhere along the line.  I find it a little too coincidental that the number of elephants captured and/or killed in Scipio's night attack against Hasdrubal and Syphax adds up to 14 (8 killed and 6 captured - Livy XXX 6.7-9).

Step one in these circumstances is to look at the Greek, which gives elephantas hekaton epi tessarakonta (one hundred and forty elephants), and I cannot see how this could possibly be a copyist's error for elephantas dekatessera (fourteen elephants).

Also, I cannot see how pretty much the entire army could have been slaughtered in this attack and yet 126 elephants somehow managed to escape (and then aren't subsequently noted as doing so by either Appian or Livy). 

The structures that would have trapped humans would just be so much matchwood for panicked elephants.  Considering that an assumed 126 elephants with various degrees of a) scorching and b) induced paranoia about military camps presumably got away, I think the Carthaginians would have done well if they did retrieve 80 which proved sufficiently tractable to once again accompany an army.  Granted the lack of explicit mention means we have no direct confirmation that 100+ elephants were extant following the conflagration, but neither Livy nor Appian seemed particularly interested in their welfare.  Conversely, had Scipio bagged over a hundred roasted elephants, I would have expected some mention.

The small numbers of elephants in armies outside Carthage seem to have been a reflection of the difficulty of shipping them rather than the difficulty of training them.

Not at all...Hanno has 50 at Heraclea (Polybius I.19); Hasdrubal has 140 (new) elephants at Lilybaeum (I.38); and supposedly Hasdrubal (son-in-law) had ~200 in Spain (Diodorus XXV.12). So shipping doesn't seem to have been too much of an issue during the First Punic War and the subsequent interbellum.

All of which is very true, but we were considering the Second Punic War, in which fewer elephants seem to have been transported for whatever reasons were the case.  My suspicion (and this is just a guess) is that following the First Punic War the cash-strapped Carthaginians cut back heavily on elephant transports, so that Hasdrubal in Spain could have received a succession of small shipments during peacetime but large-scale shipments over any distance during wartime were out of the question whether the Romans were contesting the routes or no.

The fact that it was the elephants on the ends of the line which broke away towards the cavalry I see as being more down to their position at the ends of the line, which left them with a quick way out of the 'noise zone', than any difference in training.  If anything, the ones at the end of the line would presumably be the best-trained, because they are acting as 'markers' for the rest and keeping them in line by the example of their presence.

But to argue this would surely suggest that the best elephants (on the flanks) did the worst (fled without contact) while the worst elephants (centre) performed the best (penetrating and heavily disrupting the principes). (As per Appian)

It is not (as I see it) a matter of better or worse performance as of some elephants having a discernible way out to the flank and the rest having none.

Out of interest, how would you have expected Hannibal's elephants to have behaved at Zama if trained to the limit possible?  How would such elephants have reacted differently?

There's no way of knowing how that might have turned out. I suspect that Scipio's preparations were such that even veteran elephants would have had difficulty but the fact that a significant number got through in the centre and caused havoc suggests that the Romans may have had a worse time for it...probably not enough to lose but certainly not the rout that it was. I don't base my opinion on the outcome of Zama so much as the peripheral information that I've outlined. One thing that I haven't brought up yet but also contributes to my opinion, albeit as corroborating rather than direct evidence, is the difficulty that Ammianus Marcellinus has with the Persian elephants (History XXV.3.11).

"11. On the other hand the Persians, fighting with increased spirit, shot forth such clouds of arrows, that we could hardly see the shooters through them; while the elephants, slowly marching in front, by the vast size of their bodies, and the formidable appearance of their crests, terrified alike our horses and our men."

Admittedly these are Indians and some are armoured but both he and Julian (and presumably other senior officers) would have been very well versed in military history as part of their classical education and as such, in known anti-elephant tactics. Other parts of the History shows they employed many of these tactics and yet still had much trouble with the elephants.

True: Alexander's men at the Hydaspes also handled elephants quite effectively, but did they like it?  Assuredly not! 

[Citizen6]

Step one in these circumstances is to look at the Greek, which gives elephantas hekaton epi tessarakonta (one hundred and forty elephants), and I cannot see how this could possibly be a copyist's error for elephantas dekatessera (fourteen elephants).

The structures that would have trapped humans would just be so much matchwood for panicked elephants.  Considering that an assumed 126 elephants with various degrees of a) scorching and b) induced paranoia about military camps presumably got away, I think the Carthaginians would have done well if they did retrieve 80 which proved sufficiently tractable to once again accompany an army.  Granted the lack of explicit mention means we have no direct confirmation that 100+ elephants were extant following the conflagration, but neither Livy nor Appian seemed particularly interested in their welfare.  Conversely, had Scipio bagged over a hundred roasted elephants, I would have expected some mention.

In light of that I'm more than happy to concede that it's not a transcription error, like I stated earlier, my Latin is poor and my Greek non-existent. Sadly when I went to school classics were considered pointless and weren't taught. I'm not attached to a university in any way and so rely heavily on my own library of English translations and to a lesser degree on the Lacius Curtius website. Appian could still be mistaken, but I have no evidence at all to back that up. It was only a theory, now dead, like so many elephants ....c'est la vis.   

I must admit I hadn't thought of 60 being wounded sufficiently to preclude them from Zama and that is a very real possibility. So maybe Kistler's source is correct after all. Though it really just takes us back to where we began with a bunch of elephants captured by Hasdrubal of debatable quality. I still find it very odd that they aren't mentioned subsequently especially as the cavalry and infantry escapes are. And there is no chance that the fire consumed them totally. It's just not possible.

All of which is very true, but we were considering the Second Punic War, in which fewer elephants seem to have been transported for whatever reasons were the case.  My suspicion (and this is just a guess) is that following the First Punic War the cash-strapped Carthaginians cut back heavily on elephant transports, so that Hasdrubal in Spain could have received a succession of small shipments during peacetime but large-scale shipments over any distance during wartime were out of the question whether the Romans were contesting the routes or no.

I had thought of the Roman domination of the sea and the difficulty of moving a large fleet of elephants about. But given how long Hannibal was in Italy and the fact that the Romans didn't dominate the sea near Spain, even if the Carthaginians did have to ferry elephants in small loads they had more than enough time to build up a very large reserve force in Spain (and to my understanding there just isn't evidence that they did compared to a generation before). The First Punic War ends in 241, Hamilcar dies in 228 and Hasdrubal in 221 so the 200 elephants must have been there somewhere in those seven years (but certainly well after the first war - though where they all went is a mystery as subsequent engagements only account for about 100 of them).

It is not (as I see it) a matter of better or worse performance as of some elephants having a discernible way out to the flank and the rest having none.

But there isn't anything to suggest that they turned around en masse and fled away from the Romans which would surely be the most likely thing for a fleeing elephant (such as the left flank ones did). Once they were amongst the Roman lines and heavily pressured by javelin fire, Scipio's lanes provided a clear path out. But the flanks don't even make it this far. Ultimately though, here we are conjecturing about the mindset of another species in an unusual situation (and of the inherent nature of that situation, we really know quite little beyond some limited accounts). Maybe if we're really lucky that new discovered Macedonian city off the coast of Egypt will have a complete history of the Hannibalic wars from a third party perspective carved in stone.....but I'm not too expectant. 

True: Alexander's men at the Hydaspes also handled elephants quite effectively, but did they like it?  Assuredly not!

But they did have sarrisas, and from a personal perspective I think I'd much rather face elephants in close combat as part of a pike phalanx then with short swords. 

At the end of the day I suspect we're just going to have to agree to disagree on this one, though it has been an interesting and engaging debate.

[Jim Webster]

Conversely, had Scipio bagged over a hundred roasted elephants, I would have expected some mention.

A vote of thanks to the caterers perhaps?

;-)

Jim