War FantasyHistorical Armies – Army of King Tomislav
Baptized Croatia musters as many as 60 thousand horse and 100 thousand foot, and galleys up to 80 and cutters up to 100. The galleys carry 40 men each, the cutters 20 each and smaller cutters 10 each. This great power and multitude of men Croatia possessed until the time of Prince Krasimer. But when he was dead and his son Miroslav, after ruling for four years, was made away with by the ban Pribonias, and quarrels and numerous dissentions broke out in the country, the horse and foot and galleys and cutters of the Croat dominion were diminished. And now it has 30 galleys and *** cutters, large and small, and *** horse and *** foot.
Above is the passage from De Administrando Imperio, by Roman Emperor Constantine VII. It is probably the only direct source for military strength of Tomislav’s Croatia we have, and had been long a subject of debate. Generally speaking, the 19th century Romantic Croatian historians had uncritically accepted it, using it to prove the military strength of Tomislav’s Croatia – even though it really requires no proving, considering his achievments could only have been possible with a strong army. But massive military strength apparently shown by Constantine was used to prove that Croatia was a Balkan superpower, something that was followed by 1990s and some of the later Croatian historians. Authors such as Šanjek, Milošević and Ivan Vujčić presented the above numbers as factual.
First historian to comment Constantine’s numbers was Franjo Rački. He did not bring numbers into question, but at first commented that they likely correspond to time of Duke Trpimir before changing his opinion and concluding that they are more likely to time of King Tomislav. Ivan Kukuljević Sakcinski merely used the numbers as-is to estimate population of Tomislav’s Croatia as being anywhere from 1 600 000 and 2 000 000 people. Vjekoslav Klaić considered Constantine’s numbers realistic in the context of the territorial extent of Croatian kingdom of the time as well as Tomislav’s military successes. Bulgarian historian M.S.Dironov concluded that Porphyrogenetus’ data referred to period of Tomislav’s rule when he had integrated Bosnia, or at least its larger, western part, into Kingdom of Croatia. Slovene historian Lj. Hauptmann does not bring the numbers into question at all, but merely concludes that they refer to the time after year 925. G.Manojlović spent a lot of time and space trying to justify Porphyrogenetos’ unusually large numbers. He concluded that Porphyrogenetos’ data was based on an older document due to usage of the present time, and also that he never included data for numbers of Croatian army after the half of 10th century, despite intending to. He also concluded that the numbers related to king Tomislav. Large number of infantry and cavalry Manojlović attempted to explain by noting that this was not a professional army similar to that of the Byzantine Empire, but rather a gathering of all military-capable males under banners of their chiefs. He also believed that numbers Constantine had provided included not only troops of actual Kingdom of Croatia, but also those of the Duchy of Neretva, as well as the Serbs whom Tomislav had sheltered from extermination in Croatia following Serb defeat against Bulgaria. Thus he concluded that the numbers presented included all areas under Tomislav’s leadership and not just Croatia itself. F Šišić uncritically accepts Porphyrogenetos’ numbers, merely commenting that they had made Croatia a major military power of the time, as does G Novak.
Overall, numbers as presented in the opening paragraph tended to be uncritically accepted until the half of the 20th century.
Conversely, modern-day, predominantly Communist, historians however tended to equally uncritically reject it as worthless propaganda. Bogo Graffenauer first expressed doubt in validity of Constantine’s report in 1950s based on the demographic impossibility of such numbers, and this soon spread to school textbooks. Serbian historian B Ferjančić noted that “Constantine’s numbers are undoubtably exaggerated, as they would assume that Croatia had military force greater than the Byzantine Empire. Yugoslav textbooks such as those of Olga Salzer in 1960 as well as modern (post-2000) Croatian textbooks unequivocally reject Constantine’s data as exaggerations. Petrić-Ravančić textbook decisively refutes accuracy of Constantine’s numbers stating that “… these numbers can in no way be correct, as they would mean that Croatia of the time had military larger than that of the Byzantine Empire itself”. Bekac-Štambak textbook likewise states that numbers are clearly overstated. In modern Croatian historiography, Nada Klaić rejected these numbers as exaggerated as they would require population of over 2 million people, while she accepts Graffenaueur’s assumption that Croatia had had only just over half a million inhabitants. J Lučić and V Košćak however uncritically accept Constantine’s numbers, while Ivo Goldstein and A Nazor spend a lot of time refuting them. Goldstein stated that Constantine’s numbers may be ten times as large as actual numbers.
Goldstein’s argument rests on few main arguments:
But the last note is actually misrepresentation of historical source by Goldstein. Goldstein presents ARF as stating that Ljudevit had lost 3 000 infantry and 300 cavalry in the fighting in Dalmatia. Even if that were the case, this would in fact imply greater than 1:10 ratio of cavalry to infantry as cavalry is generally better able to escape. But that is not what the source says at all: actual statement is that Ljudevit had lost 3 000 soldiers killed and 300 horses captured. Further, since Ljudevit was clearly still capable of resistance after losing 3 000 soldiers, this implies that both his field army he had led into Dalmatia and especially his overall military were significantly stronger than 3 000 men.
At any rate, Goldstein’s remarks would produce a military of 16 000 men, of whom some 1 600 would be cavalry and 15 000 infantry.
Ante Nazor notes that analysis of sources supports the view that early medieval armies in the area of Croatia could not have been very numerous. He bases this on data on demography of medieval Croatia, required expenses of equipping the army described and comparative analysis of multiple sources. He concludes that Tomislav’s army may have numbered some 2 500 “standing” soldiers and overall force of up to 10 000.
In general, “romantics” uncritically accept various impossibilities while “deconstructivists” negate even the most solidly established truths (with few exceptions such as Ante Nazor). In both cases, historiography is led by politics and propaganda, with nationalist romantic historians seeking to maximize Tomislav’s achievments and the military power of the Croatian kingdom, while Yugoslav Communist historians tend to try and minimize them. Neither group has shown any capacity for actual critical thought and consideration of the historical data nor shown any inclination to try and fit information given by the Emperor into the historical, sociopolitical and geopolitical context which Tomislav’s Croatia existed in.
So what was the real strength of Tomislav’s Croatia?
Historical Context
While Tomislav’s kinghood is contested in some circles (e.g. so-called-historian Ivo Goldstein), letter of Pope John X from 925 clearly states “To dearest Tomislav, King of Croats…”. Thus Tomislav was clearly the first king of Croatia. His kinghood also makes sense in the light of his achievments, as Tomislav had united – for the first time ever – Pannonian and Maritime Croatia into one kingdom (for geographical reference, Pannonian Croatia will have included modern-day Zagreb area and Slavonia, while Maritime Croatia will have included modern-day Lika, Dalmatia, Bosnia and parts of Herzegovina), although Slavonia in reality may have remained a sort of buffer zone between Croatia and Magyaria / Hungary. Tomislav did so after defeating Magyar invasion, having earlier defeated Bulgarian army under Algobotur somewhere near river Vrbas. He was also the first Croatian ruler that was unquestionably independent of Frankish dominion, as until 888. Croatia was still under Frankish overlordship.
So it is quite clear that a) Tomislav was in fact the king of Croats and b) Tomislav’s Croatia was a strong military power. Unfortunately, this does absolutely nothing to help us answer the question “How strong?”. Magyar invasion was likely just a plundering raid, while victory over Bulgarians was a consequence of an ambush in the mountains of Herzegovina. In neither case will Tomislav have required a massive military force, especially since he will have had advantage of terrain in both cases. Even if – which is doubtful – victory over Magyars was in Slavonia, swampy and forested terrain of the 10th century Slavonia will have given Croatian, predominantly infantry, army a major advantage over mounted Magyars. Fact that Croatian army was active deep in Slavonia can be seen from a warrior burial in Susedgrad near Zagreb. In this grave was found spata K which is typical for 10th century Croatia, while swords of these type were not found at all in the areas surrounding the kingdom. Late type of this sword means that it could only have come from the early 10th century, that is, time of expansion of Tomislav’s Croatia. So while we cannot be certain Tomislav had actually conquered Slavonia, at the very least it was heavily contested between Croatia and Hungary.
There are multiple arguments for why Constantine’s data may be inaccurate. Two main arguments are as follows:
Thing is, none of these “pass the smell test”.
The opening paragraph was, as noted, written by the Roman Emperor Constantine VII. Constantine VII Porphyrogenetos was born in 905 AD, crowned in 913 AD and shared a throne with multiple co-emperors until 945 AD, most notably Romanos I Lakapenos (ruled 920 – 944 AD). In 945 AD he finally achieved practically sole rule. Legally, he never ruled alone – immediately upon becoming a sole emperor, he crowned his then-7-years-old son Romanos II as a co-emperor. This makes Constantine a very close contemporary of King Tomislav, as Tomislav was crowned a Duke of Maritime Croatia in 910 AD and was crowned a King of Croatia in 925 AD, reigning until 928 AD at least. Thus Constantine VII will have been 20 years old at the time of Tomislav’s coronation as a king, a situation very different from typical medieval chroniclers who are oftentimes decades or centuries removed from the events they are writing about. And while Constantine’s work was finalized and published in 945 AD, he began writing it in 926 AD, during Tomislav’s reign.
Work De Administrando Imperio actually never had a title. Name itself was only made up in 1611 AD by the Dutch diplomat and scientist Meursius, but the work was in reality something of a letter and a guide written by Constantine VII for his son and heir Romanos (future Roman Emperor Romanos II). And since Constantine VII was as noted Tomislav’s contemporary, and an Emperor of the state that had at the time the most sophisticated diplomatic and espionage complex in Europe if not wider, he can be expected to have access to factually correct data. Purpose of the work as a guide for a future emperor also precludes the propagandist purpose historians had offered as a reason for the high numbers of Croatian troops described: a work written as an internal guide for one’s own son is not going to have propagandized numbers, unless they are considered factually accurate by the writer.
Yet number of 160 000 soldiers for Croatia of the time is basically impossible, for several reasons.
Numbers in Medieval Sources
When looking at numbers in medieval sources, one has to be careful of several things:
When it comes to context, most important factor is that numbers tend to be one of several types:
Of the three, last number is of least use here, simply because medieval Croatia – as typical of the medieval states with the exception of the Byzantine Empire – did not have a an official establishment with prescribed number of troops. In fact, the ninth and tenth century Byzantine Empire was the only European state that did. Treadgold places Byzantine army in 899 AD at 124 000 soldiers and marines and 34 200 oarsmen. All of these were professional soldiers, a situation unlikely in Croatia.
Census of troops is also useless, as it is not known that medieval Croatia ever carried out such a census before its union with Hungary. In fact, carrying out a census will have been impossible because as stated above, most medieval states did not have a prescribed number of troops and thus military would be in a constant if low-level flux. And even if census had existed, it will have been necessary to first establish whether it lists the entire military or only a portion of it.
Observational guesses are typically of limited use. These are arrived at by observing an army in the field and assigning a round number – say, 100 000 men. But for large armies, such numbers are usually greatly exaggerated due to normal human tendencies (just as police and the press often exaggerate size of large crowds). Moreover, even if number of any given field army is correct, it still gives absolutely no indication of actual number of troops available to the state. A country with underdeveloped logistics may be able to field only a small field army despite having technically more troops available than a country with better developed logistics.
In case of Constantine’s text, what he is doing is discussing total size of Tomislav’s army based on his own estimate. So this is not a field army, and it is not a reliable number.
Demography and Logistics
Early medieval demographic estimates provide the clearest refutation of the idea that Croatia could have raised so many men. Tomislav’s Croatia was certainly a very powerful state, and after his victories over Magyars / Hungarians and Bulgars almost certainly extended to natural river borders of Drava – Danube – Sava – Drina – Neretva. This means that it had included modern-day Croatia (without Istra), Syrmia and Bosnia, but likely excluded Herzegovina as Duchy of Neretva was likely not under Tomislav’s direct authority.
- Croatia: 56 594 km2
- remove Istria County: – 2 813 km2
- add Serbian Syrmia: + 3 942 km2 (?)
- BiH: 51 209 km2
- remove Herzegovina: – 12 188 km2 (?)
- TOTAL: 96 744 km2
All and all, a total area of 95 000 – 105 000 km2 is likely. Whether Tomislav however had practical control over Slavonia is uncertain, but in any case majority of population of Croatia of the time was likely in the maritime Croatia (a.k.a. Dalmatia).
But what about population density?
We know that medieval Croatia was not a densely populated country. While exact estimates are difficult due to lack of data (especially hard data such as censi and tax records), it is possible to arrive at approximate estimates by using contextual clues. Climate, vegetation and soil conditions are also not well known. Yet through study of sources and comparisons with other areas of Europe it is nevertheless possible to arrive at some estimates.
Notable here are descriptions of Crusaders who had passed through Croatia in the 11th century. Crusades had passed through Maritime Croatia – today’s Dalmatia and western Bosnia. Specifically, they passed from Senj eastwards (Lika, western Bosnia, Dalmatian highlands). Chronicler describes these lands as being full of hills, forests, large rivers, streams and pastures, and also as being “wet and foggy and almost entirely swampy”. 10th century sources mention swamps in Istra as well, and much if not most of Pannonia was also swampy.
One also needs to consider that amount of land necessary to support a peasant family was much greater than today. Today, one needs 200 square feet of garden, 200 square feet of fruit trees, 1/4 acre or 10 890 square feet of pasture for a cow and chickens, 500 square feet for pigs (they and chickens can use the same pasture as the cow), 1/2 acre or 21 780 square feet for lamb and goats, 2 acres or 87 120 square feet for meat cows, for a total of 120 690 square feet (2,77 acres / 11 212 m2) of land to feed a family. During medieval times, a peasant household required at least 90 000 m2 to live (Byzantine modios), but by 15th century in Western Europe it could be up to 120 000 m2 (German hufe – 121 406 m2; Hungary – 30 acres = 121 406 m2). Thus, amount of land necessary was some ten times greater. But Croatian population of the time – as with much of early medieval Europe – had only small amounts of actually cultivated land. Most of the food was provided by animal husbandry, that is, livestock farming.
Going from this, estimates posit that average population density must have been low. How low can be seen from estimates for population density of similarly pastoral areas. V Rogić estimated average density of former Yugoslavia in early Middle Ages as being around 5 people per km2. B Grafenauer estimated population density of early medieval Slovenia at 6 to 8 people per km2. Ancient Germania had density of some 4 to 7 people per km2, which is relevant as we know it had similar living conditions to early medieval Croatia: primarily animal husbandry with very few farms and large areas of land covered by forests and swamps. Density may however have been increased by refugees from Moravia (conquered by Magyars in the early 10th century), but not significantly so.
So what about numbers?
During the Carolingian age, eastern Francia (roughly today’s Germany) had some 2 500 000 people at cca 400 000 km2, for density of cca 6,25 people per km2. V Rogić estimated that total early medieval population of the area of former SFRY (255 804 km2) was around 1 000 000 people, for density of cca 3,9 people per km2.
According to this, in 1 000 AD, Kingdom of France had 7 200 000 people, Hungary and England 1 250 000 people each, Ireland 630 000 and Croatia with Bosnia may have had a total of 698 000 people at a slightly larger surface area of 107 000 – 114 000 km2. This will have given Croatia a population density of cca 6,32 people per km2, compared to 9,59 per km2 for England and some 14 per km2 for France.
In 900 AD, population will have been some 70% – 90% of population in 1000 AD. This means that Tomislav’s Croatia will have had population density of 4,42 – 5,67 people per km2 at area of 95 000 – 105 000 km2, for total population of anywhere between 420 000 and 600 000 inhabitants.
To stay as favorable as possible I will herefore use the 90% estimate as well as the maximum (105 000 km2) surface area, leading to a total population of 600 000 people.
But how does this translate into military?
Quite logically, in fact. Soldiers need to be trained, equipped and fed – even the local militia. And this requires smiths, cooks, tailors and other specialists. Further, neither women nor children will be used as soldiers by any sane society, nor will be the people too old to fight. What this means is that even for the absolute balls-to-the-wall general mobilization, typically done only by barbarian nomadic societies, maximum proportion of fighting men is somewhere in the 25% range, as in, all of their military-aged males. Extreme military mobilization by settled societies with part-time militia armies (e.g. Roman Republic) could mobilize some 6 – 8% of population, and less militarized states could count on being able to mobilize some 4 – 5% of the population. But these armies were cca 90% infantry. Roman manipular legion had 4 200 infantry and 300 cavalry (93% infantry), and in barbarian armies, proportion of infantry was even greater. Mobilization was typically “take a guy, give him a spear and a shield, and send him to fight”. And in the early middle ages, even the “basic” equipment – shield and spear – was already expensive.
Cavalry requires much more support than infantry. Horse is expensive, horseman training is also expensive. In order to ensure survival of this already massive investment, ruler will also want to make sure that the cavalryman has adequate protection – metal helmet, hopefully metal armor – which drives up the cost even further. During Charlemagne’s time, equipment of a cavalryman cost some 44 solidi – or some 15 cows, with horse alone costing 3 to 6 cows. Other sources give breakdowns which are even higher, up to 45 cows:
- metal helmet = 6 cows
- scale armor = 12 cows
- horse = 12 cows
- sword = 7 cows
- spear and shield = 2 cows
Thus a minimum was some 40 cows, and even more if one wanted additional equipment such as dagger or greaves. It is consequently obvious that a cavalryman had to be a rich person indeed. Frankish cavalryman had to own at least 12 mansi (plots) of land to be capable of service. As will be seen below, this ratio of 12 plots per cavalryman was nearly identical in the Byzantine Empire, and thus can be assumed to hold true for Croatia as well.
And when on campaign, horse’s massive food requirements mean that the cavalryman has to have several pack animals as well, as well as the servant or servants. And to provide for all of this, cavalryman will have far larger holdings than a simple infantryman – which is the entire reason why knightly class appeared. Luckily, it is possible to see roughly how much land will soldiers require.
The closest approximation here is the Byzantine Empire. It was geographically proximate to Croatia, and while its military was professional, basic support requirements for troops do not change that much. And for Byzantine soldier to be supported, one required a lot of land. In the Byzantine Empire of the time, one small farmer owned typically 30 modioi of land (some 90 000 – 100 000 m2). As price was typically 3/4 of nomisma for one modius of first-class arable land, 30 modioi meant 22,5 nomismata.
Soldiers’ land holdings as prescribed by Constantine VII were as follows:
- cavalryman = 4 lbs of gold / 288 nomismata (=12,8 farmer plots)
- infantryman = 50 modii = 1,7 farmer plots
Nikeforos II prescribed holdings of 12 pounds of gold for a heavy cavalryman (cataphract), though that is less relevant as there are no suggestions Tomislav ever had this type of soldiers in his army.
I will assume that marines of ships had same holdings as infantry, despite the fact that in the Byzantine Empire marines held four pounds of gold worth of land, same as the cavalry. Since purpose here is to arrive at the maximum number of soldiers Croatia could have possibly raised, and Croatian warships were basically coastal fishing ships, this is not too much of an issue.
Further, since Croatia of the time was under both the Frankish and Byzantine influence, and equipment will have been similar, this means that values seen above for Byzantine cavalry (which are not cataphracts) and infantry will have been roughly similar for Croatia. It should be noted that in both Francia and Byzantium of the time, armor was worn only by cavalry – infantry had either no armor or else textile armor only. Heavy cavalry, equipped with metal helmet, metal armor for the rider and textile for the horse, lance, sword, dagger and bow, will have been exceptionally rare. Even in the Byzantine Empire, cataphracts numbered some 5% of the army – remainder were 25% light cavalry and 70% infantry.
So if we take the numbers provided by the translation, the results are as follows:
- 60 000 cavalry = 240 000 lbs of gold = 768 000 plots
- 100 000 infantry = 170 000 plots
- 80 galleys with 40 men each = 3 200 men = 5 440 plots
- 100 boats with 20 men each = 2 000 men = 3 400 plots
In total, absolute minimum for the above force would be the equivalent of 946 840 plots of land. Even assuming that the infantry and the marines are peasant militia with only one plot of land each, it is impossible to go below 873 200 plots of land.
This means that minimum population Croatia will have required to field such numbers is anywhere between 4 366 000 and 4 734 200 people, requiring very nearly modern-day population densities (41,6 to 49,8 people per km2). And the real dealbreaker here is cavalry: anywhere between 81% and 88% of land goes to supporting the cavalry. In fact, while not directly comparable, 15th century Hungary could field some 80 000 mostly light cavalry from population of cca 4 million people – but no infantry. Of these, 16 000 came from the Kingdom of Croatia. And Tomislav’s Croatia was heavily forrested, which means that raising cavalry and especially feeding it on campaign a very difficult if not impossible task. Further, while general military obligation providing infantry was definitely not impossible for Tomislav’s Croatia, fact that it was a monarchy and with strong cavalry indicates that the process of class stratification was already underway, which may well have limited the number of available infantry as well.
It is undeniable that Tomislav’s Croatia did have strong cavalry. This is proven by textual records as well as the spurs found in burial sites and pictorial evidence. But “strong” is a relative word, and as can be seen from the discussion of environmental factors, conditions in Croatia of the time were hardly favorable for large-scale horse breeding. This is especially true for the mountainous littoral Croatia – the most densely settled area of early medieval Croatia. For comparison,Croatia in 1890 had 2 186 410 people and 248 053 horses (1 horse on 9 inhabitants) while in Bosnia in 1879 ratio was approximately one horse for every 7 inhabitants and in Dalmatia in 1869 only one horse for every 27. Average for all three areas was one horse for 9 inhabitants, but Bosnian average may well be assumed for Bosnia and Dalmatia. Thus cavalrymen will have been relatively rare, and likely limited to vassals and escorts of ruler and more notable individuals. In Anglo-Saxon terminology, cavalrymen would be equivalent to huscarls while infantry will have been equivalent to fyrd. Because Tomislav’s cavalry was very well equipped however – stirrups had been found in Croatia as early as 7th century, compared to 8th century in France and 10th century in England. Most likely cause of this was Avar influence.
10th century Croatian cavalrymanMilitary organization of Tomislav’s Croatia also makes a cavalry-heavy army unlikely. Specifically, Croatia of the time was organized into županije (counties) ruled by župan (count), who was a royal appointee. There were six counties during time of Trpimir, and 11 at least during Tomislav’s time – Constantine Porphyrogenetus actually lists the counties which he considers part of Croatia: Hlijevana, Cetina, Imotski, Plivanj, Pesent, Primorje / Coastal, Bribir, Non, Knin, Sidraš, Nin, Krbava, Lika, Gacka (note tha Krbava, Lika and Gacka were technically not counties but were ruled by ban of Croatia). He also specifies that Croatia begins at Cetina and stretches until Labin in Istra.
Counties were centered on strong fortifications such as Klis, built on raised land which dominated their surrounding, controlling roads and mountain passes. Each county was ruled by a count who also led county’s armed forces, and there was a separate area ruled by ban (viceroy) who had his own army. This setup, while based around militia as opposed to professional troops, is curiously similar to Byzantine army of the time, which was organized into provincial territorial armies (themata) and the professional standing army commanded by the Emperor (tagmata). Pop Dukljanin (Presbyter Diocleas) states that each county had a count and a captain, with captain leading 100 men, while viceroy’s area had 7 captains. With 11 counties and viceroy’s area this makes for 18 captaincies and thus a maximum of around 1 800 somewhat professional soldiers (likely cavalry). Even if we assume the king had a separate royal army equal in size to viceroy’s army, number of professional / career soldiers / cavalry still only reaches 2 500 men.
There are issues with these numbers. Captains (satnici, stotnici) only first appear in Croatian sources in the 11th century. While they may have existed earlier, that is necessarily only speculation. Nor is it clear that the Emperor had listed all of Croatian counties. Porphyrogenetos does not mention the organization of area between Velebit and Raša, which can be assumed to have some military force at least.
Yet cavalry numbers are definitely not small for the 10th century “standing” army, as later documents show that cavalry was only very few in number. Pacta Conventa, a likely 13th century document that may or may not reflect the actual state in 11th century, states that each Croatian tribe was to equip at least 10 armored cavalry at its own expense. Duke of Krk, Bartul, received Modruša county with the obligation of raising 10 men-at-arms (armored cavalrymen) for campaigns within the borders of the Kingdom of Croatia, and 4 cavalrymen outside it. In 1224, Klokok tribe in Gorska county received royal freedoms alongside the obligation of equipping 15 armored cavalry and 100 infantry. Golden Bule to Gradec in 1242 obliged the city to provide 10 fully armed soldiers (meaning heavy cavalry). In 1293, viceroy Pavle Šubić was, due to his honor of viceroy, obliged to provide 500 soldiers. This was the primary expeditionary army, as infantry only really served in the local defense.
Thus, traditionally accepted number of troops for Tomislav’s Croatia is impossible. At the very least, infantry ought to outnumber cavalry far more than it does in the sources. Further, while reliance on the peasant militia for infantry may have given Croatia relatively large numbers of recruitable army, only a small portion of it would be actually useful for deployment on campaign as peasant infantry could not be marched for significant distances. Essentially, any “field” army would consist of a cavalry core plus however many infantry could be collected from the immediate area.
What this means is that numbers of Tomislav’s army, and especially those of its cavalry component, have to be significantly reduced.
Since population of Tomislav’s Croatia cannot plausibly go above 600 000 people and 120 000 plots at most, simple arithmetic reduction while keeping proportion will lead to a military of at most 8 000 cavalry, 16 000 infantry and 1 600 sailors. Taking the above data for number of horses in 1890 however, conclusion could be drawn that there may have been some 70 000 – 85 000 horses in total in Tomislav’s Croatia. Since cavalryman required more than one horse, and war horses were in any case a minority of all horses, this again confirms that number of cavalrymen in Tomislav’s army must have been in single-digit thousands.
Thankfully, reduction does not need to be, and cannot be, proportional. Few if any tenth century European armies had anywhere near close to a third of cavalry as proportion of the army (sole exception being the Byzantine Empire, at 20 – 25%).
Cutting cavalry to 1 800 – 2 500 will allow the number of infantry to be raised to some 90 000 men, which is a far more plausible relation. Basic overview would be thus:
- 120 000 total plots
- 21 600 – 32 000 plots cavalry
- 95 680 – 85 280 plots infantry
- 2 720 plots sailors
From the above, Tomislav’s Croatia may have had:
- 1 800 – 2 500 cavalry
- 95 680 – 85 280 infantry
- 1 600 sailors
- 20 galleys with 40 men each
- 40 boats with 20 men each
or a total of 90 000 – 100 000 troops at its maximum population. At minimum population of 420 000 men, total number of soldiers may have been as low as 60 000. Compared to the numbers usually cited for Constantine’s work (162 500 total), these values are between 37% and 62% of the total.
Situation is even worse since Constantine does not include Slavonia as part of Tomislav’s Croatia. This makes sense since while Tomislav had definitely beaten the Magyar / Hungarian invasion, it is not clear whether he could have expanded his authority until Drava. Considering the likely very lightly populated nature and difficult swampy terrain of the area between Sava and Drava, it is likely that neither Magyars nor Tomislav actually fully incorporated Slavonia into their states, but instead left it as a sort of “no mans land” or a buffer zone between the two.
Field army, of course, will have been smaller. Considering the factors seen above we can assume that the maximum deployable army will have been entire cavalry + infantry of one to two counties. This makes the field army total as some 1 800 – 2 500 cavalry and maybe 7 000 – 6 000 infantry. Cavalry could be brought down further, making maximum field army some 900 – 1 200 cavalry and 7 000 – 6 000 infantry, far closer to actual early medieval field armies. For comparison, Frankish army that had defeated the Magyars at Lechfeld in 955 AD likely numbered only 7 000 – 8 000 men.
But above are just principles. While I am confident that the framework established above is accurate, it can only ever set principles to how many soldiers Croatia could have had and thus produce a context and framework for discussing the actual sources. Closer look at actual source is required to see what more likely the matter.
Constantine Porphyrogenetos – De Administrandi Imperio
As noted earlier, Constantine Porphyrogenetos’ De Administrondo Imperio (DAI) is the sole real source for numbers of military of Tomislav’s Croatia, but contextual clues show that the reading shown in the first paragraph cannot be considered reliable. But where other historians had predominantly ended up rejecting the numbers outright, Živković believes that Constantine’s numbers are completely reliable and precise… when read properly. Specifically, he believes that numbers had been fundamentally misread.
While contextual clues are a key tool for historians, one should not rely on them too much. As Warren Treadgold noted in his discussion with John Haldon, “faith in superiority of assumptions over evidence could be used to reject any source or sources”. Živković’s reinterpretation solves this problem by offering interpretation which brings DAI text to a point where it actually agrees with the various contextual considerations discussed previously.
Problem with De Administrando Imperio is that it is preserved in only one medieval manuscript, and that one is not the original either. Copyist errors are common in the hand-written manuscripts, and especially so in manuscripts which a) utilized Byzantine relatively clumsy numerical system and b) had been copied by people outside the culture they had been originally written in. In fact, Živković notes that not one author had actually bothered to take a look at the original document before making his argumentation.
The passage in question is as follows:
Ὅτι ἡ βαπτισμένη Χρωβατία ἐκβάλλει καβαλλαρικὸν ἕως τῶν ξ’ χιλιάδων, πεζικὸν δὲ ἕως χιλιάδων ρ’ καὶ σαγήνας μέχρι τῶν π’ καὶ κονδούρας μέχρι τῶν ρ’ καὶ αἱ μὲν σαγῆναι ἔχουσιν ἀνὰ ἀνδρῶν μ’, αἱ δὲ κονδοῦραι ἀνὰ ἀνδρῶν κ`, αἱ δὲ μικρότεραι κονδοῦραι ἀνὰ ἀνδρῶν ι’
Typical Croatian translation reads thus:
Krštena Hrvatska postavlja konjaništva do 60 tisuća a pješadije do 100 000 i sagena do 80 i kondura do 100. Na sagenama imaju po 40, na kondurama po 20 a na manjim kondurama po 10 ljudi.
Christian Croatia sets cavalry up to 60 thousand and infantry up to 100 000 and sagenas up to 80 and conduras up to 100. On sagenas they have 40 men each, on conduras 20 men each and on smaller conduras 10 men each.
According to Živković, however, transcription and its translation are an issue. First and greatest problem is the word χιλιάδων after the letter ksi (ξ’), that is the number 60. He believes that the word does not denote thousands at all. And he is correct because the Paris manuscript actually has two alphas (ξ’ ᾷᾳ) in the place of thousands. Moravcsik transcribed the shorthand as χιλιάδων, but older editions of DAI show that editors had transcribed it in many various ways. Dubrovnik researcher A Banduri, who did have access to the original, simply left the shorthand out. Bekker and Rački did not leave it out, but translated it as “thousands”.
Yet if ᾷᾳ really does stand for “thousands”, that would be the unique case in this manuscript. At no other place is the number “thousand” shortened as “ᾷᾳ”. In fact, when writing “thousand” the writer is using exclusively χιλιάς , and thus ᾷᾳ is never and cannot be used for the word thousand. Several notes at the margins also imply that the transcriber had issues understanding the original he was transcribing. Živković thus interprets the ᾷᾳ as meaning “ἀλλάγια” (allagia, allagion). This is a Byzantine military term which in general can mean a shift or rotation, but by the 10th century was being increasingly used as an alternate name for cavalry bandon. But what size is the bandon?
As with any military unit, size of bandon had changed over time. Yet the sources temporally closest to Tomislav’s time are remarkably consistent in the size of the allagion. The Praecepta Militaria of Emperor Nikephoros II Phokas (963 – 969) has the basic cavalry unit being bandon of 50 men, and same size of the bandon is repeated in the Taktika of Nikephoros Ouranos (980 – 1010). Campaign Organization and Tactics states that cavalry units should number up to 300 men, however, and Taktika of Leo VI (895 – 908) states the size of 250 to 350 men for bandon. The “imperial” allagion could have been far larger, even up to the stated 300 man size.
From the above, it can be concluded that 60 allagia numbered most likely a total of 3 000 cavalry – but they could have also numbered 9 000, 15 000, 18 000 or even 21 000 men.
So what about the infantry?
Expression “πεζικὸν δὲ ἕως χιλιάδας ρ`” can only be read as “infantrymen up to / around 100 000”. Yet as shown in the previous demographic discussions, such number of infantry is possible only as an extreme upper end, and only if cavalry numbers significantly less than 3 000. Using a maximum of 120 000 plots of land with 3 000 cavalry (36 000 – 38 400 plots) and 1 600 sailors (1 600 plots?) would leave infantry with a maximum of 82 400 plots of land, and this is ignoring the possibility that some soldiers may have had more land than strictly necessary. Having 100 000 infantry would leave only 18 600 plots for cavalry, which would limit their numbers to no more than 1 500 or so. And if infantry had truly numbered around 80 000, would not Constantine had written so (“infantrymen up to / around 80 000”).
Živković concluded that the problem was again in the transcription. Transcriber already mixed up ρ and κ in the transcript, which means that the most likely solution is that instead of ρ᾿ (100) the text had originally read κ᾿ (20). F Rački made nearly the exact same mistake some eight centuries later, mixing up the number of sagenas (80) with the number of sailors (40):
“takova (Hrvatska) mogaše diognuti na noge 100.000 pješakov, 60.000 konjanikov, a na moru 5.200 brodarov, koji oboružahu 40 velikih a sto manjih brodov.”9
such (Croatia) could raise 100 000 infantrymen, 60 000 cavalrymen, and at sea 5 200 sailors who armed 40 large and 100 small ships.
Živković also believes that the “reading” of Croatian naval capacity is accurate. And it really has no reason not to be. Eighty galleys at 40 men each and 100 cutters at 20 men each represents a total of 5 200 men. For comparison, Byzantine dromond carried either 108, 120 – 160 or 300 men depending on the size. At this size, Croatian galleys would be closest to smaller of the Viking longships (karvi – 6 to 16 benches = 12 to 32 rowers; snekkja – 20 benches = 40 rowers) and much smaller than the larger longships (skeid or dreki, at 30 benches / 60 rowers and above). By comparison, 959 Byzantine navy had 34 200 oarsmen and 8 000 marines, or 42 200 men in total. This incidentally made it have 0,35% of Empire’s population, where 1 600 men of Croatian navy would be either 0,38% of 420 000 population or 0,27% of the 600 000 men figure. Even 5 200 men of the navy would not be impossible however, remembering that these were mostly not professional soldiers.
Conclusions
Number of 165 000 soldiers for Tomislav’s Croatia is plainly impossible. Kingdom did not have either the population density nor organization capability required to raise such an army, even if we assume that the entire army consisted of lightly equipped peasant levies – which in itself is very different from what Constantine states. How large exactly the number of soldiers available was is far more difficult to estimate, but a combination of multiple sources (De Administrando Imperio and various other sources and contextual clues) can lead to a reasonably accurate conclusion.
Overall, most likely strength of Tomislav’s Croatia is:
- 3 000 cavalry (60 allagia)
- 20 000 infantry
- 5 200 sailors (80 galleys and 100 cutters)
Sources
Proba_93-146str_2crop.pdf (srce.hr)
#armies #army #Croatia #croatianArmy #croatianKingdom #europe #kingTomislav #kingdomOfCroatia #militaryHistory
