Collections: Raising Carthaginian Armies, Part III: Generals, Warlords and Vassals

This is the third part (I, II) of our series looking at how Carthage’s complex, multipart armies were raised and constituted. Last time, we looked at the backbone of Carthage’s armies: North African troops levied out of Carthage’s subject communities in North Africa. These fellows seem to have been directly employed by the Carthaginian state, paid for their service, equipped by the state and mustered in Carthage itself.

But beyond these troops drawn from close to the core of Carthage’s empire, we also see a pattern of Carthage mobilizing troops from more loosely and indirectly controlled places. Carthage, by the third century, had extended a limited kind of control over the two kingdoms of Numidia (Massaesylii and Massylii) and Numidian cavalry – some of the best in the Mediterranean world – becomes an important component of Carthaginian armies. Meanwhile, Carthage also maintained a presence in Spain (dramatically expanded after 237) and Iberian warriors, men drawn from the communities of the Mediterranean coast of Spain, appear in Carthaginian armies, though as we’ll see their relationship to Carthage probably undergoes some pretty significant changes.

What unites these two groups is that we probably best understand their relationship to Carthage (at least, by 218) as something less than direct control but more than simply ‘mercenaries:’ instead as we’ll see these polities are more like vassals to Carthage, but with the notable quirk that their relationship with the Carthaginian state is often channeled through Carthage’s generals, who seem to have at times been at pains to develop deep personal ties with these communities as a means of gaining access to their military resources.

But first, as always, raising large armies of mercenaries, subject conscripts, vassal warlords and allies is expensive! If you too want to help me invade Italy with a multi-ethnic army of diverse origins in a doomed effort to stop the Roman Republic, you can help by supporting this project over at Patreon. If you want updates whenever a new post appears or want to hear my more bite-sized musings on history, security affairs and current events, you can follow me on Bluesky (@bretdevereaux.bsky.social). I am also active on Threads (bretdevereaux) and maintain a de minimis presence on Twitter (@bretdevereaux).

Generals and Magistrates

Before we look at how the Numidians and Iberians end up interacting with the Carthaginian state, we need to outline very briefly who exactly is the ‘face’ of the Carthaginian state that they see.

Carthage is somewhat striking in cleanly separating its chief domestic magistrates from its generals. Whereas in the Roman Republic, the consuls were both chief magistrates and chief generals (and in a Greek polis, while those two roles might be separate, they were often held by the same sort of elites), in Carthage, the chief magistracy and the office of general appear to have been mostly on different career paths. What follows, I should note, is essentially a summary of D. Hoyos, The Carthaginians (2010)’s description, with a few sprinkles from M.J. Taylor, “Generals and judges: command, constitution and the fate of Carthage” Libyan Studies 54 (2023).

The chief magistrates in Carthage were the sufetes (Latin: sufetes for Punic shophetim (špṭm)) two of which were elected each year. Early on it seems like these fellows probably had both domestic and military responsibilities, but by the time we can see them really clearly, they really do seem limited to civil affairs at home: handling legislation, managing courts, judging lawsuits, dealing with finances and son. But even though our Latin and especially Greek sources often call these guys kings (they weren’t, they were one-year elected magistrates), they came to have very little outward facing role.

Instead, armies were led by a general, the Punic term for which was rab mahanet (rb mḥnt), translating to something like ‘army leader.’ This office and the separation of civil and military leadership probably originates in the 500s, though even into the 400s and early 300s, sufetes sometimes show up also acting as generals, but this is clearly already very rare by Aristotle’s day (d. 322). By the third century the offices appear well and truly separate. The rabbim (that’s the plural) were elected, but did not serve fixed terms: instead they seem to have served for the length of the conflict or until they failed badly enough to be sacked from command (done by the ‘court of one-hundred and four’). The office was, as Hoyos notes, enough of a prize for seekers to “pay perfectly good bribes” for it, presumably because success provided for both fame and wealth, but failure was equally punished: unsuccessful generals tended to be put on trial and crucified for their failure. It is also striking that it seems like relatively few generals go on to be sufetes – some, but not many – suggesting that the position as rab mahanet was effectively the apogee of a separate career path for the Carthaginian aristocracy (so both sufetes and the rabbim are coming from the same background, the same aristocracy, but have different careers).

But what matters for us here most of all is that the office of general was a long-term office, generally held for years. Hamilcar Barca is a general from 247 to his death in 228, holding command for just short of two decades. Moreover, these generals tend to be assigned to specific theaters for extended periods (Hamilcar is in Spain from 247 to 241, then in North Africa from 241 to 237, then in Spain from 237 to 228), acting as the ‘face’ of Carthaginian rule in the region. If you are a local ruler who borders areas of Carthaginian control, the local Carthaginian rab mahanet is going to be a factor in your calculations for many years at a time, in stark contrast to the single-year terms of Roman magistrates and pro-magistrates.

Consequently, if you were a ruler who bordered Carthage, it would make a lot of sense to develop relationships with Carthaginian generals – or more specifically, with the Carthaginian general closest to you – as much as cultivating a relationship with the Carthaginian senate (the adirim, spelled ‘drm) or the sufetes. By contrast, a foreign power looking to interact with Rome on a long-term basis really wanted to interact with the Roman Senate, which represented the continuing component of Roman foreign policy.

All of which, as we’re going to see, matters for Carthage’s vassal states, beginning with Numidia.

Beginning With Numidia

We have only modest visibility into Carthage’s relationship with the Numidians (we are, ironically, better informed about the structure of Barcid control in Spain, despite its shorter duration), but Carthage’s relationship with Numidia was clearly an important component of the Carthaginian military system.

The Numidians were a Berber people, indigenous to North Africa. This was a primarily agricultural society (with some pastoralism) – the Numidians were not, for the most part, nomads and they had some urban centers like Cirta. Initially organized into tribes, by the time we see them clearly this tribal system of organization has coalesced into two larger proto-kingdoms, Massylii (to the east) and Masaesyli (to the west), which were rivals of each other. The rule of each of these kingdoms was held by a king, but the form of government here seems to be strongly personalistic, more akin to how a Gallic tribe with a king might function than the centralized, bureaucratic monarchies of the Near East.

The Carthaginian relationship with Numidia was complex: the Carthaginians never subdued this region, but at the same time until the Romans broke the back of Carthaginian power at the end of the Second Punic War, Carthage is clearly the dominant power compared to the two Numidian kingdoms (one thing the Romans do late in the Second Punic War is facilitate the unification of those kingdoms, creating a united Numidia hostile to Carthage). In particular, Carthage can play the two Numidian polities off of each other. That means that, somewhat paradoxically, Carthage is frequently fighting in Numidia and at the same time Numidians form a key part of Carthaginian armies.

Numidian armies seem to have generally been a mix of light-armed javelin infantry and light javelin cavalry.¹ The former was, so far as we can tell, of little use to Carthage (or anyone else) and I don’t believe we ever hear of Numidian infantry ever deployed abroad. But Numidian cavalry was some of the best there was. Numidian aristocrats fought on horseback as light javelin cavalry. They rode bareback and without bridles (Polyb. 3.65.6) or perhaps with only light bridles made of rushes (Strabo 17.3.7) on fairly small horses (Livy 35.11.6-11), wielding javelins and a light shield covered (or perhaps made of) hide (Strabo 17.3.7 again), but generally – it seems – without armor or helmets (though we do find imported armor in royal Numidian tombs).

Whereas Carthaginian, African or Liby-Phoenician cavalry fought as heavy ‘shock’ cavalry, the Numidians fought as expert cavalry skirmishers, moving rapidly, largely keeping out of close combat with heavier shock cavalry and skirmishing on the flanks. Numidian horsemen play key roles in Hannibal’s successes and the defection of Numidia to the Romans clearly weakens his army in the run-up to Zama, robbing him of one of his most potent tools.

What is perhaps most striking is the Carthaginian relationship with their Numidian vassals seems also to have been substantially personalistic. This is something large states often struggle with: it is hard for the institutional structures of a state to interface successfully with the highly personalistic governing structures of non-state peoples or peoples only beginning to develop state institutions.

The odd Carthaginian solution seems to have been the use of the generals – the rabbim mahanet – as a sort of personalistic interface with these communities. So – as Taylor (op. cit.) notes, we see Carthaginian generals developing close personal ties with Numidian kings and princes (that is, tribal leaders with large retinues) in order to get access to the military force those Numidian leaders can harness. So for instance Hamilcar maintained a close friendship with the Numidian aristocrat Naravas, through whom he got access to Numidian cavalry (Polyb. 1.78.1), while Hasdrubal Giscso cultivated a personal relationship – a guest-friend relationship – with the Numidian king Syphax (Livy 28.18, 29.23). Even closer ties were possible: during the Mercenary War, Hamilcar Barca promised his daughter in marriage to Naravas and nets 2,000 Numidian cavalry in his army in the bargain (Polyb. 1.78.1-9), while the Numidian, Mazaetullus – having seized power among the Massylii – aims to cement his position by marrying Hannibal Barca’s niece who had been, we are told, previously married to the former king Oezalces (Livy 29.29). The last Carthaginian-Numidian marriage we’re told about is that of Sophoniba, daughter of Hasdrubal Gisco, given in marriage to Syphax to secure that alliance (Livy 29.23).

What is striking about these sorts of ties is that these Carthaginian generals aren’t kings and aren’t even the chief magistrates in Carthage. However, the long-service structure of Carthaginian generalship lets Carthage ‘play’ at personalist politics, because a rab mahanet sticks around long enough to build those personal relationships.

In the meantime, Numidian cavalry provided a key military capability for Carthage, making up probably the largest single part of all of Carthage’s available cavalry in our 215 snapshot. It’s tricky to be precise because our sources repeatedly lump African and Numidian cavalry together, but Numidian and African cavalry together make up 11,000 of Carthage’s 21,000 cavalry in that 215 snapshot and that 11,000 in turn are probably at least 2/3rds Numidian, suggesting perhaps something like 7,500 Numidian cavalry, which is quite a large figure indeed. Carthaginian victories often rely substantially on their use of cavalry, so these Numidians, levied through important personal relationships with key Numidian aristocrats, form a crucial part of Carthage’s tactical approach.

It also helps us to understand what the Barcids are doing in Spain.

What the Barcids Are Doing in Spain

Carthage had a limited presence in Spain prior to the end of the First Punic War, particularly around Gades (modern Cádiz) although it isn’t always clear how direct Carthaginian control here was.

However after the end of the Mercenary War (in 237), the Carthaginians decide to send Hamilcar Barca to Spain with an army. We aren’t given a ton of reliable insight into this decision by our sources, but it seems like one factor here was a wariness of Hamilcar’s rising power. Hamilcar had commanded the land armies on Sicily during the First Punic War from 247 to the end in 241 – where he had held his ground against the Romans, only for the war to be lost at sea (leaving him undefeated) and then had led the successful Carthaginian effort during the Mercenary War (241-237). As noted above, he had strong relationships in Numidia, an experienced, veteran army and huge political cachet and so the decision to send him to Spain may have been a way to both ‘reward’ him (he is given basically carte blanche to do whatever he wants in Spain) and also get him out of Carthaginian politics where his influence would have been disruptive.

In any case, Hamilcar, followed by his son-in-law Hasdrubal the Fair, followed by his biological sons Hannibal, Hasdrubal and Mago, proceed to expand Carthaginian power in Spain dramatically, consolidating control over most of the Mediterranean coast of Spain south of the Ebro River. After Hamilcar dies in 229/8, Hasdrubal the Fair, his son-in-law, takes over and founds a new city to serve as a capital for this Barcid-controlled part of Spain, which he called Qart-hadasht (‘New City’) which is of course also the name of Carthage: a Carthage for Spain to go with the Carthage in Africa. The Romans will call this place ‘New Carthage’ (Nova Carthago) and of course today it is Cartagena.

Diodorus reports that, upon replacing Hamilcar, Hasdrubal the Fair was acclaimed strategos autokrator by the Iberians (Diod. Sic. 25.12). Of course strategos autokrator is a Greek title, something like ‘supreme general’ and we have no idea what the Iberians in their own language would have called the position, but as we’ll see it is going to function like a kind of ‘warlord of warlords.’ That position in turn is the one that Hannibal Barca will inherit on Hasdrubal the Fair’s death in 221 and he largely completes the conquest of the Iberian coastal zone.

There is sometimes a mistake here in understanding the where and who of this Carthaginian zone in Spain. The people who lived on the Mediterranean coast of Spain were the Iberians (iberes in Greek, hiberi in Latin), a people who spoke a non-Indo-European language and were culturally and ethnically distinct. To their south in Andalusia, you had another people, the Turdetani, who had their own different non-Indo-European language. By contrast, in the uplands of the Meseta lived the Celtiberians, who spoke a Celtic (and this Indo-European language). The Celtiberians will be the fellows who cause the Romans no end of trouble in the second century but they are not the fellows the Barcids are incorporating: Barcid influence is thin in Celtiberia as far as we can tell and direct control basically non-existent.

Instead, it is the Turdetani and Iberians who come under Carthaginian control in this period and who thus make up the ‘Iberians’ of Carthaginian armies. Our sources actually are fairly careful in this regard – in the rare cases that Carthaginian armies have meaningful numbers of Celtiberians in them, our sources go out of their way to comment on it.

That said the historical event of Carthage cementing control of the Iberians in Spain under the Barcids means that Carthage’s relationship to these fellows changes fundamentally at this point.

Recruiting Iberians

We’re also not super well-informed about the structures of government among the Iberians. Our sources treat the basic unit of Iberian organization as the oppidum – the fortified town – but these aren’t city-states. Instead, those towns are further grouped into ethic/tribal groupings (Oretani, Bastetani, Edetani, etc. etc.; these are all the Latin names), but these aren’t always political units.

Instead, when outside powers (Carthage, Rome) interact in political or military ways, they interact with figures that our Latin sources (esp. Livy) call reguli, ‘petty kings.’² It seems clear that what is implied here is that these fellows are ‘little’ not necessarily in the sense of the territory they rule, but in the limited nature of their power: they’re warlords, rather than true kings. Notably, they seem to lead coalitions of towns and those coalitions seem to have shifted regularly, with towns entering or exiting (note for instance Culchas who controls 28 towns in 206, but brings just 17 when he revolts against the Romans in 197; he’d be on the winning side of the wars until then, so he hasn’t lost these towns, they’ve just stopped following him).

What I think we need to understand here is that you have towns – that have their own town non-state/proto-state governments – which for collective warfare band together under charismatic war-leaders (the reguli) who then lead these coalitions. A system like that, if undisturbed from the outside, we would expect it to develop into true kingship as the reguli use their power and influence to make their position heritable and consolidate power. But that hasn’t happened yet, so the reguli are warlords or war-chiefs, leading ad hoc coalitions of towns and these coalitions are the largest political units in Iberian politics and warfare.

Beneath the reguli, our sources offer us almost no real visibility into Iberian politics, but archaeology reveals a consistent trend of aristocratic warrior burials around Iberian oppida, suggesting the same sort of non-state military aristocracy we’ve seen in Gaul and Celtiberia, where individual aristocrats wielded some small amount of military force on their own. In that case the reguli become immediately understandable as the leaders of coalitions of aristocrats who in turn dominate the local politics of their town and can mobilize their military forces.

In short then, this seems to be a non-state zone, where control of military force remains fragmented, with large scale and probably also small-scale warlords able to raise military force.

The Carthaginians seems to have recognized the potential of recruiting from this environment very early. Iberian mercenaries show up in Carthaginian armies as early as the fifth century, with Diodorus explicitly noting Carthage sending officers to Spain to xenologein, ‘foreigner-enroll’ Iberians into Carthaginian service (Diod. Sic. 13.44.6; 13.80.2); they’re explicitly sent with polla kremata, ‘a lot of wealth’ (which could mean money, but also trade goods or other stuff; I suspect if Diodorus had meant coined money, he’d have said arguros, ‘silver’ or a derivative) with which to do this, so service is being purchased. One wonders if these officers are hiring soldiers one-at-a-time or instead offering Iberian aristocrats access to some of the wealth that Carthage could provide if they would raise their retinue and lead it as part of Carthage’s army; I suspect the latter. Carthage was a large, developed city with access to a lot of wealth and access to a lot of stuff – metalwork, wine, import goods from the Eastern Mediterranean and so on – which could be important prestige goods in Spain, the sort of thing that an Iberian warlord could use as gifts himself to key retainers to cement his power or to demonstrate his own importance. We’ll come to this next time, but I don’t think Carthage ever had to pay these guys a lot of money, per se, even when the relationship was strictly transactional.

And that was probably the general state of affairs down to 237: Carthage, with lots of trade contacts into Iberia had plenty of opportunity whenever there was a war on to recruit Iberian infantry and even some cavalry as mercenaries and seems to have done so liberally.

Hannibal’s Iberians

This system has to change, of course, once Carthage is taking direct control over large parts of Spain.

The Barcids do not – probably could not – try to install some sort of bureaucratic system of direct rule in Spain, outside of a handful of new foundations (like New Carthage) designed as the hubs of Carthaginian control in the region. Instead, it seems like the Barcids brought with them the model of dealing with vassal polities they had from Numidia, getting access to military forces through personal relationships. That said, Barcid power was a lot more intense in Spain than in Numidia: the Carthaginians had never subdued the Numidians fully, but Hamilcar and his sons really do subdue the Iberians, demonstrating powerful military superiority.

What I think we should actually understand in Hasdrubal the Fair’s supposed title of strategos autokrator, is that the Barcids, in their conquest of Spain, have made themselves the warlords of warlords, essentially installing themselves at the top of a non-state pyramid of personalistic aristocratic-retainer relationships.

We see this most clearly in the replication of those marriage patterns we saw in Numidia above. Hasdrubal, just before he is proclaimed strategos autokrator, marries an unnamed Iberian princess (so the daughter of one of the reguli), presumably in so doing cementing his hold over a significant portion of Iberian military resources. Hannibal, likewise, marries an aristocratic woman from the Iberian town of Castulo that our sources name Imilce (Livy 24.41.7, Sil. Pun. 3.97-99). There’s also a ‘stick’ side to this: the Barcids take hostages from elite Iberian families to ensure loyalty. Presumably these hostages would be kept as honored guests in relative wealth and opulence in New Carthage (that is where they are when the Romans take the city), acting both as a kind of gift for loyalty and a potential punishment for disloyalty.

Crucially, this attachment is personal: Iberians are not loyal subjects of Carthage, they are vassals (sometimes of questionable loyalty) of the Barcids. This jumps out very clearly in Livy’s description of the defection of one of Hannibal’s reguli, Indibilis – who cites his services to the duces Carthaginienses, ‘to the Carthaginian leaders,’ which is to say the Barcids, not to Carthage itself. He cannot, after all, mean his services to the adirim or the sufetes – he’s never met either! Instead, Indibilis feels personally slighted by the Barcids and so is prepared to switch sides to the Romans if they will treat him better – an aristocrat defecting from being the vassal of one warlord to another. Notably, he clearly seems to understand himself as defecting not so much to Rome as to Scipio not-yet-Africanus (Livy 27.17).

Iberians at War

Now one of the things we discussed back when we covered non-state ‘tribal’ armies is that while non-state societies are generally very small – their personalistic systems of government do not scale well – they can often mobilize a very high proportion of their military resources. They may be poor and few, but they can get everything they have into the field, often allowing them to meet the armies of far larger and wealthier states in the field.

The Barcids, in making themselves warlords of warlords in Iberia have essentially taken their state-based military system, with Carthage’s large territory and revenues supporting an army of semi-professional conscript infantry levied from subject communities, and bolted on a non-state mobilization system. Indeed, they are able to bolt on several of them, because one Carthaginian general, with the wealth and power of the Carthaginian state behind him, can maintain several reguli as his vassals.

The result is that the Barcids, and thus Carthage, are able to pull tremendous amounts of troops out of Spain, despite the relative poverty of the societies in question. Prior to 237, Iberians made up a meaningful but by all accounts fairly small component of the Carthaginian army, minor compared to the role of North Africans and other groups. By contrast, in our 215 snapshot, Iberians make up the single largest portion of the Carthaginian army: 57,000 (counting cavalry) of the 162,000 man total, over a third of the entire force. And that in many ways understates the scale of Iberian participation, because the Iberian elements of Hannibal and Hasrubal(-his-brother)’s armies often take far heavier casualties than the African elements, which are then filled up – especially in the very heavy fighting in Spain – with ever more Iberians.

It is only through the tremendous mobilization of the Barcid’s Iberian vassals that Hannibal is able to rumble with Rome on an even footing in both Italy and Spain.

In terms of fighting style, the Iberians provided Carthage with a ‘medium’ infantry option. Metal armor of any kind (including helmets) were basically unknown in Spain in the third century, so Iberian warriors went to battle with at most textile defenses for the body and head. They did, however, use large shields – both a round shield of indigenous design (the caetra) and oval shields of the La Tène style were used – somewhat offsetting their lack of armor. The standard offensive loadout seems to have been a fairly typical thrusting spear, a single javelin and a sword as a backup weapon, the latter most commonly the devastating cutting falcata (our sources sometimes incorrectly want to give the Iberians the predecessor to the gladius Hispaniensis, but that sword derived from the Celtiberian variant of the La Tène sword, which was present in the Iberian coastal zone, but quite rare).³

In short they fought as spirited, but unarmored, ‘medium’ infantry, capable of skirmishing effectively with javelins but also able to hold as ‘line infantry’ when needed. Versatile and effective, but a lot lighter than the Roman heavy infantry they were often pitted against. Carthaginian generals are able to win pitched battles against the Romans with armies that rely heavily on these guys: Hannibal, of course, at Trebia (218), Trasimene (217) and Cannae (216), but also Hasdrubal(-brother-of-Hannibal), Mago(-also-brother-of-Hannibal) and Hasdrubal Gisco in twin battles on the Upper Baetis (211).

The weakness here is simply that Barcid control was personal, it was new and thus it was fragile. Once the Romans start really winning in Spain after 210, Carthaginian control begins collapsing as the reguli defect and bring over huge chunks of Carthaginian Spain with them. It is striking that the thing Hannibal wanted to do to Rome – trigger a mass revolt of Rome’s allies to undermine Roman military power – is precisely what P. Cornelius Scipio not-Africanus-yet does to Hannibal in Spain and then again in Numidia.

Importantly, we should probably not understand the Numidians at any time as ‘mercenaries‘ and equally after 237, we should not understand the Iberians as mercenaries either. Instead, Carthaginian generals – the rabbim – had developed a method of putting themselves into the reciprocal, personalistic systems of non-state governance and mobilization, enabling Carthage to access some of those non-state military resources in both Numidia and Spain.

Now, as an aside, I don’t necessarily think this structure was, in the long run, a good idea. The Carthaginians themselves, by 237, seem to have realized that Hamilcar Barca’s network of power was becoming dangerous (so they packed him off to Spain). My honest view is that had Carthage won the Second Punic War, they were in a position to effectively ‘speed run’ the Roman experience of the First Century, with powerful generals turning their loyal armies – loyalty secured through these sorts of personal ties – against the state. But of course Carthage didn’t win the Second Punic War, so that future remains only notional.

In the meantime, these systems enabled Carthage to pull an astounding amount of military power out of regions it only loosely controlled. That said, we’re not done with Carthage’s armies yet. Next time, we’ll look at Carthage’s allies and at long last at its actual mercenaries – there are some, I promise!

1. For more detail on Numidian warfare, the thing to read is W. Horsted, The Numidians (2021); Osprey volumes can be really hit or miss, but Horsted is largely a ‘hit.’
2. This is Latin rex, regis (m), ‘king’ with a dimunitive suffix, -ulus, creating regulus, reguli (m), ‘petty king, little king, kinglet.’
3. For more on how these guys fought, the absolute best thing in English is F. Quesada Sanz, Weapons, Warriors & Battles of Ancient Iberia (2023), trans. E. Clowes and P.S. Harding-Vera; if you want more, you need to read Quesada Sanz’ work in Spanish.