The Origins of the Feudal System: Chapter One

That Military Benefices Did Not Exist Under the Roman Empire

This is Chapter 1 from the book The Origins of the Feudal System: Benefices and Patronage During the Merovingian Period by Numa Denis Fustel de Coulanges. That book is itself Volume 5 of an epic series on the history of France written by Coulanges, titled History of the Political Institutions of Ancient France.

One of Coulanges’ concerns in this book, which was compiled posthumously by Camille Jullian, is to show that the feudal system which developed in France was uniquely French— and given his nationalist concerns in 19th century France—especially that it was not a mere German imposition. In Chapter 1 (below), he shows that feudalism was not established by the military land grants under the Roman Empire, contrary to the widely accepted position of French historians and legal theorists in the early modern period, the continuing relevance of which Coulanges initially downplays, but then shows is still somewhat popular even into the 19th century.

Please note that while I have dutifully transcribed all of Coulanges’ footnotes in case someone wants to chase down one of his citations, I have not translated them (they are in a mix of French and Latin). If there is a particular note of his you would like translated, please let me know and I’d be happy to help. I have also included a few notes of my own (marked) where I thought the text or my translation could bear some explanation.

My translation is based on the second edition published by Librairie Hachette in 1900. You can access the digital version via BNF Gallica. Please do not miss the previous entry in this project, the introduction.

---

No one today believes anymore in the Roman origin of the benefice¹ and the fief. It must be said, however, that this view was held for a long time, and by exceedingly erudite men. The view bears examining, just as much as the contrary view.

Not that anyone ever held that feudalism existed under the Roman Empire; it is obvious that it did not take off until long after the disappearance of that Empire. But it was said that under the Roman Empire there already were military benefices and a type of fief for soldiers’ use. And then it was added that the Germans, after their invasions, hastened to adopt this Roman custom, developed and spread it, and made of it a widespread institution.

Loyseau, in his Traité des Seigneuries² that he published in 1608, explains the origin of the fief thus: “The Franks, having seized the lands of Gaul, established upon them fiefs, that is to say, charged them with always attending the prince in war: an invention that had been begun by the Roman emperors, who themselves, in order to protect their borders, decided to give these very lands to their most distinguished officers and soldiers, a form of largesse which they termed ‘benefice,’ to be held by them only as long as they should be soldiers.”³

Sixty years later, the great scholar of the Romans, Jacques Godefroi,⁴ published the Theodosian Code with his own admirable commentary. Arriving at Book 7, Title 15, we find a law of Honorius conceived thus: “The lands which by the farseeing goodness of our ancient predecessors were given to soldiers called gentiles for the protection of the borders of the empire, we have learned that they are sometimes held by men who are not soldiers; nevertheless be it known that these landholders are constrained to the service of protecting the border, and if they default in this duty, they must leave these lands and hand them over to gentiles or veterans.”⁵ To this law of Honorius, Godefroi adds an edict of Theodosius II in which it is also seen that the agri limitanei⁶ could belong only to soldiers and could not be sold to men who were not soldiers.⁷ He brings together with these two legislative texts two passages from the Historia Augusta.⁸ The first is from Lampridius, who says that Alexander Severus “gave the territories taken from the enemy to the leaders and soldiers of the frontier lands, on the condition that their heirs would be soldiers and that the territories would never belong to men who were not soldiers.” The second is from Vopiscus, who relates that Probus “gave to the veterans certain territories in Isauria, adding that their male offspring must be soldiers starting from the age of 18.” Upon these four texts, Godefroi makes this reflection: “Here is a type of fief; after all, here is found the necessary condition, which was that the receiver owed to the bestower his loyalty and military service.”⁹ And a bit further on: “It was with this sort of fief and benefice that the emperors compensated the labors of their soldiers.”¹⁰ Comparing again these soldiers to the Laeti of the laetus lands¹¹ and even to vassals, he adds: “They were a type of vassal vowed to war.”¹² He enumerates the conditions attached to these imperial concessions and counts three: the first is military service; the second is the prohibition on selling the land, at least to men who are not soldiers; the third is heredity, constrained to male succession. And he concludes: “All this very much recalls the nature of fiefs.”¹³

This same stance was taken in the following century by the Abbé Dubos. The Roman Emperors, he writes, parceled out these borderlands between their soldiers “on the condition that the State would always remain the true owners of these estates,” and that these lands would not pass “to heirs from the original awardee” unless these heirs bore arms. “It is widely regarded” that the distribution of lands under these conditions “as the very origin of possessions known in history under the name of fiefs.”¹⁴ The author cites as support for his theory the same texts Godefroi had already pointed out; but he adds a passage from Saint Augustine conceived thus: “It is well known that the soldiers of that century, since they wished to receive from their temporal lords a temporal benefice, began by linking themselves towards them by an oath and undertook to keep faith with their lords.”¹⁵ So behold, in a single sentence of Saint Augustine’s, the mention of “lords,” “benefices,” of “oath,” and of “fidelity.” It appears that the entire feudal contract is contained in this sentence.¹⁶

This theory did not end with the Abbé Dubos. It can be found, in truth much softened, among many erudite men of our own century.¹⁷ Among them there is just this contradiction, that after having recognized a sort of military fief among the Romans, they then claim that feudalism comes from Germanic customs.

Now the existence of military benefices and near-fiefs among the Romans is not a fact that should be concealed or downplayed. If their existence is proved, then that is the source, or at least one of the sources, of the feudal regime. As soon as the military benefice or the fief existed in the Empire, even if it held a very small place, it could have been the germ from which feudalism sprang. It would be possible that, from a purely local and exceptional institution, the Germans could have made a general and ruling institution. It would then be of no little importance to state that their origin point was in the Empire, and it would have to be admitted that the Roman Emperors instituted military benefices, from which came the Merovingian benefices and fiefs. We must see whether this fact is true, and in order to do that, examine the texts upon which it is believed to rely.

First: Lampridius, after having mentioned some wars in Mauritania, in Illyria, and in Armenia, says that the lands taken from the enemy were given by the Emperor to the generals and soldiers of the corps which protected the borders. He speaks of giving, donavit.¹⁸ The Emperor, in truth, imposed as a condition upon the recipients of the land that their heirs would be constrained to service. But we must remark that this condition did not change the nature of the act, which was truly a gift.

Second: Vopiscus relates that, the mountains of Isauria being infested by brigandage, Probus chased them out, and then searching for a means of preventing the return of this scourge, came upon the idea of setting up veterans there. “All places difficult to access, he gave as private property to the veterans, in adding the measure that from the age of 18, their children, at least their male children, would be sent into the corps of troops, from fear that, if they remained in their mountains, they would take to the habit of brigandage.”¹⁹ There are three things to note in this passage. First, the historian speaks of a real gift, properly speaking, privata donavit. Thus he does not speak of a condition placed upon this type of property: he says merely that the Emperor added that the sons of these men would be sent to the army; this was an administrative measure that he took and which conformed to the military legislation of the time. In the end, the historian says nothing of the sort that these lands were heritable only from male to male; he indicates no principle of succession, and leaves us to suppose that these sorts of estates were subject to the ordinary rules which ruled the succession of the privata.

It is necessary to read these two passages very superficially in order to see in them a resemblance to fiefs or benefices. This is a question of plain and total gift; donavit, says Lampridius; privata donavit, says Vopsicus. This is the opposite of the benefice and the fief. These concessions are hereditary, something the benefice never was in fact, and which the fief never was by right. Nothing shows that masculine succession was the rule, as later for the fief. In the end, if the condition of military service was attached to it, this itself, far from bearing a resemblance to the Merovingian benefice, is what distances it from it the most; for we will show shortly that the benefice, the origin of the fief, was never subject to a condition of military service.

Let us subject to the test the two legislative texts which were invoked, adding as well another law of Honorius’ in the Justinian Code and a fragment from the jurist Paul in the Digest.²⁰

All these texts are concerned with lands situated on the borders, agri limitanei. We know that the Emperors, in order to protect their immense borders without having need of too many soldiers, garnished them with a line of little forts, castella, and a wall, limes.²¹ The soldiers who guarded them were called castellani or limitanei.²² Around each castellum and along the length of the limes, by rule the land was allowed to the use of the soldiers. Each little corps of troops had a few pastures for feeding their livestock,²³ and also a few fields under cultivation. This was for the profit of the soldiers, and something of a complement to their pay. It was above all a resource to help their families live; for the soldiers of the Empire, whether they were barbarian or Roman, normally had their wives and children with them.²⁴

It is these types of land which are at question in these law codes. The emperors saw that abuses had been committed, that individuals had occupied or bought many of these lands. They made laws in order to restore these lands to the soldiers, annulling any sale as illicit.²⁵

But these lands were not fiefs. Just a moment’s attention is required to perceive a fundamental difference between them and fiefs. These lands were held in common by each corps of soldiers. They belonged indivisibly to this legion,²⁶ to that cohort, to this troop of gentiles, to the men of this castellum. But they did not belong individually and privately to each soldier or each officer. If one soldier were to be transferred from one of these corps to another, he did not keep any part of the ager limitaneus. This collective character of possession is precisely the opposite of what we find in the fief. Never did a fief belong collectively to a corps of soldiers. On the contrary, it will be of the very essence of the fief to be individual, as will be the services and obligations attached to it. It is the same with the Merovingian benefice. The agri limitanei of the emperors thus have nothing in common with the benefice or the fief.

Now, as for the passage of Saint Augustine’s, where the Abbé Dubos finds the mention of lords, of benefices, and of the oath of fidelity to the lord. The homily in which we find this sentence is not Saint Augustine’s. It is an apocryphal homily²⁷: it was composed in the Middle Ages, and the passage cited was borrowed from Yves de Chartres, who lived at the beginning of the 12^th century.²⁸ This passage on the feudal oath thus was not written in the time of the Roman emperors; it was written in the thick of feudalism.²⁹

It must be added that the word beneficium, a term used quite often in the language of the Empire, was never applied to land. One reads often that a man obtained some land “by the beneficence of the prince,” but never does one read that the land was “a benefice conceded by the prince.”³⁰ Every goodwill concession concerns a gift, free and clear in perpetuity, not a temporary concession—conditional, revocable—as benefices would be later on. The word beneficiarius comes up frequently as well in the language of the Empire; but it is said of officers, named by the choice of the general³¹ or the soldiers, who obtained some sort of favor, such as higher pay or exemption from levy-duty or were attached to some particular service.³² There is not a single instance to be found in which the word’s meaning approaches that which it had in the Middle Ages.

Thus there is not a single text from the Roman Empire which shows us military benefices or fiefs. We see neither lands conceded for the term of the recipient’s life on the condition of his feudal service, nor anything that resembles relief,³³ resumption,³⁴ or the rule of male succession. The position which would derive the fief from a creation of the Roman Empire for the sake of her soldiers must, thus, be totally discarded. The fief and the military benefice were never a Roman institution.³⁵

1. [Translator’s Note:] The word used here by Coulanges, bénéfice, can have a variety of meanings, but he uses it here in its general, historical sense: land conceded by a chieftain or king in exchange for military service. The more familiar term “fief” eventually replaced “bénéfice,” but since Coulanges uses both words, I’ve decided to use the otherwise obscure English “benefice.” Cf. below, fn. 30, for a brief discussion of the philological history of beneficium, the Latin term which bénéfice is derived from.

2. [Translator’s Note:] Charles Loyseau (1564-1627) was an immensely influential French jurist whose writings are of foundational importance for understanding the social structure of the French ancien régime. Coulanges is here discussing his Traité des seigneuries, or Treatise on Manors, though his Traité des ordres et simples dignitez (in English, commonly referred to as A Treatise on Orders) is his most well-known work, which describes the three Orders, or Estates, of traditional France.

3. Loyseau, Traité des Seigneuries, c. 1, édit. de 1620, p. 12. – Cf. Ét. Pasquier, Recherches de la France, livre II, c. 16, édit. de 1723, t. I, p. 128 : « Auguste commença de donner aux soldats quelques assiettes de terres, laquelle coutume fut estroitement observée par ses successeurs. De ces départements et distributions de terres nous voyons assez fréquente mention ès anciens juriconsultes. Ces terres ne se distribuaient qu’a des soldats : elles ne leur étaient d’abord octroyées qu’à vie, et ne devinrent héréditaires qu’au temps de l’empereur Alexandre Sévère… Ainsi firent les Francs.” – La même opinion est exprimée par Charondas, dans ses notes sur Bouteiller, édit. de 1603 ; p. 480 – D’autre part, Dumoulin avait soutenu avec une très grande énergie que les fiefs étaient une création des Francs et n’avaient rien de commun avec le droit romain, voy. édit. de 1681, t : I, p. 3–5.

4. [Translator’s Note:] Jacques Godefroi was a jurist and Genevan politician. His edition of the Theodosian Code became the standard edition of the work.

5. Code Théodosien, cum commentariis Gothofredi, VII, 15, édit. Ritter, t. II, p. 398.

6. [Translator’s note:] Latin, for “frontier farms”

7. Novelles de Théodose II, édit. Haenel, XXIV, § 4, p. 105-106, ou au Code Justinien, XI, 60 (59), 3.

8. [Translator’s note:] A collection of Roman biographies, supposedly the work of six different authors, among whom are two discussed below, viz. Lampridius and Vopiscus.

9. Godefroi, édit. Ritter, t. II, p. 398: Est haec species quaedam feudi. Sane similis lex erat fundorum constituendorum, ut qui praedium acciperet, danti fidem et militiae ferme munus exhiberet.

10. Ibidem, p. 599: Veteranos hoc quasi feudi beneficiique genere pensari, post exsudatos militaie labores…

11. [Translator’s Note:] The laeti were barbarians admitted into the Roman Empire, given land (i.e. laetus land), and in return, expected to supply troops for the Roman army.

12. Ibidem: Vasallorum et hominum genus militiae adstrictum.

13. Godefroi, édit. Ritter, t. II, p. 400: Quae pleraque ad feudorum naturam proxime accedunt. – Déjà Casaubon, dans ses notes sur Vopiscus, 1603, avait dit : Hanc esse quamdam speciem feudi, vel potius initia quaedam ejus juris quod feudorum appellatione est designatum.

14. Dubos, Établissement de la monachie française, 2^e édit., 1743, t. I, p. 82. – Plus loin, t. II. p. 518, il revient sur le même sujet, et par une confusion à peine croyable il assimile ces « bénéfices militaires » des empereurs romains aux « terres saliques » de l’époque mérovingienne.

15. Saint Augustin, Sermo in vigilia Pentecostes: Notum est quod milites saeculi BENEFICIA temporalia a temporalibus DOMINIS accepturi prius militaribus SACRAMENTIS obligantur, et DOMINIS SUIS FIDEM se servaturos profitentur.

16. Cf. encore Abbé Garnier, Traité de l’origine du gouvernement français, 1765. p. 104. (Il prononce nettement le mot de « bénéfices militaires » à propos du texte de saint Augustin ; mais il suit Dubos de très près ; cf. p. 49.)

17. Serrigny, Droit public et administrative romain, t. I, p. 365-372 ; C. Dareste de la Chavannem, Histoire des classes agricoles, p. 68-71 ; Révillout, Étude sur le colonat, dans la Revue historique du Droit, t. III. 1857, p. 213 ; M. Garsonnet compare aussi ces concessions à des fiefs, Histoire des locations perpetuelles, p. 165, sans dire pour cela que les fiefs en viennent.

18. Lampride, Vie d’Alexandre Sévère, 53: Sola quae de hostibus capta sunt, limitaneis ducibus et militibus donavit, ut eorum essent si heredes eorum militarent, nec unquam ad privatos pertinerent. – Les mots ad privatos s’opposent ici à milites et désignent des hommes qui ne sont pas soldats. C’est le sens du mot privatus ou de l’expression privatae conditionis au Code Théodosien, VII, 15, 2. Cf. Godefroi, Glossarium, au mot privatus.

19. Vopiscus, Vie de Probus, 16: Potentissimo quod latrone Palfuerio capto Isauriam liberavit… Hoc dixit : Facilis est ab istis locis latrones arceri quam telli. Veteranis omnia illa quae anguste adeundum loca PRIVATA DONAVIT, addens ut eorum filii ab anno octavo decimo, mares duntaxat, ad militem mitterentur, ne latrocinare unquam discerent.

20. Code Justinien, XI, 60 (59), 2 (Code Théodosien, VII, 15, 2). – Paul, au Digeste, XXI, 2, 11 : Possessiones (trans Rhenum) ex praecepto principali veteranis in praemia adsignatas.

21. Spartien, Vie d’Hadrien, 12: Per ea tempora et alias frequenter in plurimis locis, in quibus barbari non fluminibus sed limitibus dividuntur, stipitibus magnis in modum muratis saepis funditus factis atque annexis barbaros separavit. – Ammien Marcellin, XXVIII, 2 : Valentinianus magna animo concipiens et utilia, Rhenum omnem a Raetiarum exordio ad usque fretalem Oceanum magnis molibus communiebat, castra extollens altius et castella turresque assiduas per habiles locos et opportunos qua Galliarum extenditur longitudo, nonnunquam etiam ultra flumen aedificiis positis subradens barbaros fines. – Code Théodosien, VII, 15, 1 : Munitionem limitis atque fossati. – Sur ce fossé du côté de la Germanie, voir Cohausen, der romische Grenzwall, 1884, etc.

22. Castellanus miles, Code Théodosien, VII, 15, 2. – Duces et milites limitanei, Lampride, Vie d’Alexandre Sévère, c. 58. – Sur les limitanei milites opposés aux milites comitatenses, voy. une loi de 389 au Code Théodosien, VIII, 4, 17 ; une novelle de Théodose, XXIV, édit. Raenel, p. 102, et une loi du Code Justinien, I, 27, 2, § 8 (Voir la Notitia Dignitatum, commentaire de Boecking, p. 515 et suiv. Un diplôme militaire du IIIe siècle mentionne les milites castellani, Ephemeris epigraphica, t. IV, p. 508).

23. Tacite parle déjà de cet usage, Annales, XIII, 55 : Agros vacuos et militum usui sepositos… Partem campi jacere in quam pecora et armenia militum transmitterentur.

24. Novelle de Théodose II, Haenel, p. 105, 106 (Code Justinien, XI, 60 (59), 5) : Agros limitaneos universos cum paludibus et omni jure quos, ex prisca dispositione, limitanei milites ab omni munere vacuos ipsi curare pro suo compendia atque arare consueverunt.

25. Ibidein: Si ab aliis possidentur, cujuslibet spatii temporis praescriptione cessante, ab universis detentoribus vindicatos iisdem militibus sine ullo collectionis onere, sicut antiquitus statutum est, volumus assignari. Si quis forte, quo minim : audere debuerat, emptionis titulo memorati juris possidet praedia, competens ei actio adversus venditorem intacta servabitur. – Code Justinien, XI, 60 (59), 2 : Quicumque castellorum loca quocumque titulo possident, cedant ac deserant, quia ab his tantum fas est possideri quibus adscripta sunt et de quibus judicavit antiquitas. Quod si quispiam in his locis non castellamus miles fuerit detentator inventus, capitali sentantia… plectatur.

26. Corpus inscriptionum latinarum, t. II, n 2916-2920 : Terminus Augustalis dividit prata legionis quartae et agrum Juliobrigensem.—Henzen, n 6825 : Pequarius [legionis]. Cf. L. Renier, Inscriptions de l’Afrique, n 129 et 425 [Corpus inscriptionum latinarum, t. VIII, n 2553, 2827].

27. On trouvera ce sermon in vigilia Pentecostes dans l’edition de Louvain, 1635, et dans l’édition des Bénédictins, 1685. Dans la première, il est au t. p. 687, parmi les sermons apocryphes ou douteux. De même dans l’édition des Bénédictins, au. t. IV, p. 278 ; et les savants éditeurs le font précéder de cette note, qui aurait dû frapper l’abbé Dubos : ex Ruffino, Caesario, Gregorio, Yvone Carnotensi collectus. Ce sermon n’est en effet qu’une sorte de centon.

28. [Translator’s Note:] Yves de Chartres, also known as Ivo Carnutensis, or Ivo of Chartres, was a medieval French churchman, scholar of canon law, prolific letter-writer, and bishop of Chartres. He is venerated as a saint in the Roman Catholic Church.

29. Cette phrase citée comme étant de saint Augustin, on la trouve dans les œuvres d’Yves de Chartres, édit. de l’abbé Migne, Patrologie, t. CLXII, col. 604. Il n’est pas possible de soutenir que la phrase ait été empruntée par Yves à saint Augustin ; la lecture et la comparaison des deux sermons ne permettent pas cette supposition. C’est un anonyme qui l’a empruntée à Yves et l’a insérée maladroitement dans un sermon qu’il a mis sous le nom de saint Augustin. Yves de Chartres est mort en 1116. [Translator’s Note : In fact, he died in 1115.]

30. Beneficium se dit de toute sorte de bienfait ou faveur, qu’elle soit accordée par l’Etat (ex : Cicéron, Pro Archia, V ; Ad familiares, V, 20, 7 ; Philippiques, II, 36, 91) ou qu’elle le soit par le prince. Le Liber beneficiorum, dont il est parlé plusieurs fois chez les Agrimensores, édit. Lachmann, p. 203, p. 295, p. 400, était un registre où l’on tenait note de tous les dons du prince, soit en terres, soit en autres objets. (Cf. Lampride, Vie d’Alexandre, 46. Il est fait mention du primicerius beneficiorum dans la Notitia Dignitatum, d’un a commentariis beneficiorum dans une inscription, Gruter, 578, 1.)

31. Sur l’expression ordinem consequi beneficio (ducis) non virtute, voir Hirtius, De bello Africano, 54 ; Tacite, Histoires, I, 25 ; Suétone, Tibère, 12. (Cf. Handbuch de Marquardt, Roem. Staatsverwaltung, t. II, 2^nd édit., p. 549)

32. Voir les inscriptions, Corpus inscriptionum latinarum, t. III, n 1781, 1906, 1909, 1910, 2023, 3161, 3270, 3955, 4820, 5953 ; t. VIII, n 2080, 2401, 2226, 2564, 2798, 2829, 10717. – Cf. Spartien, Vie d’Hadrien, 2 ; Végèce, II, 7. Le mot beneficiarius se dit aussi de soldats détachés de l’armée pour le service des magistrats ; Pline, Lettres, X, 21 et 27, édit. Keil, etc. (Dans son livre sur les Institutions politiques et administrative de la France, t. I, 1890, p. 431, M. Viollet semble confondre les expressions et les faits, lorsqu’il dit que les soldats appelés beneficiarii recevaient des terres et qu’il parle « des droits d’un soldat sur son bénéfice ». Il n’y a, je crois, rien de pareil dans les textes.)

33. [Translator’s Note:] Relief (le relief in French) was the tax that the heir of a fief had to pay to an overlord before taking possession of the fief he was receiving from the prior fief-holder. This French practice was introduced in England after the Norman Conquest.

34. [Translator’s Note:] Feudal resumption (la commise in French) was the confiscation of a fief by the overlord, if the fief-holder did not properly render submission or homage.

35. Nous ne faisons que répéter ici ce que nous écrivions déjà en 1875 (Revue des deux mondes, 15 mai, p. 452-453). Nous sommes surpris que M. Garsonnet, dans son Histoire des locations perpétuelles, p. 244, nous attribue d’avoir soutenu l’opinion que nous avons au contraire combattue très nettement.