Chapter V: The Origin of Private Castles

We have now seen that history furnishes no instance of the existence of private castles among the Anglo-Saxons or the Danes (previous to the arrival of Edward the Confessor's Norman friends), and we have endeavoured to show that this negative evidence is of great significance. If, assuming that we are right in accepting it as conclusive, we ask why the Anglo-Saxons did not build private castles, the answer is ready to hand in the researches of the late Dr Stubbs, the late Professor Maitland, Dr J. H. Round, and Professor Vinogradoff, which have thrown so much fresh light on the constitutional history of England. These writers have made it clear that whatever tendencies towards feudalism there were in England before the Conquest, the system of military tenure, which is the backbone of feudalism, was introduced into England by William the Conqueror.[156] "Feudalism, in both tenure and government was, so far as it existed in England, brought full-grown from France," says Dr Stubbs; and this statement is not merely supported, but strengthened, by the work of the later writers named.[157] The institutions of the Anglo-Saxons, when they settled in England, were tribal; and though these institutions were in a state of decay in the 11th century, they were not completely superseded by feudal institutions till after the Norman Conquest.

We should naturally expect, then, that the fortifications erected by the Anglo-Saxons would be those adapted to their originally tribal state, that is, in the words which we have so often used already, they would be those of the community and not of the individual. And as far as we can discover the character of these fortifications, we find that this was actually the case. As we have seen, we find one of the earliest kings, Ida, building for the defence of himself and his followers what Bede calls a city; and we find Alfred and his children also building and repairing cities, at the time of the Danish invasions.

The same kind of thing was going on at about the same time in Germany and in France. Henry the Fowler (919-936), that great restorer of the Austrasian kingdom, planted on the frontiers which were exposed to the attacks of the Danes and Huns a number of walled strongholds, not only for the purpose of resisting invasion, but to afford a place of refuge to all the inhabitants of the country. He ordained that every ninth man of the peasants in the district must build for himself and his nine companions a dwelling in the "Burg," and provide barns and storehouses, and that the third part of all crops must be delivered and housed in these towns.[158] In this way, says the historian Giesebrecht, he sought to accustom the Saxons, who had hitherto dwelt in isolated farms, or open villages, to life in towns. He ordered that all assemblies of the people should be held in towns. Giesebrecht also remarks that it is not improbable that Henry the Fowler had the example of Edward the Elder of England before his eyes when he established these rows of frontier towns.[159]

The same causes led, on Neustrian soil, to the fortification of a number of cities, the walls of which had fallen into decay during the period of peace before the invasions of the Danes. Thus Charles the Bald commanded Le Mans and Tours to be fortified "as a defence _for the people_ against the Northmen."[160] The bishops were particularly active in thus defending the people of their dioceses. Archbishop Fulk rebuilt the walls of Rheims, between 884 and 900;[161] his successor, Hervey, fortified the town of Coucy[162] (about 900); the Bishop of Cambray built new walls to his city in 887-911;[163] and Bishop Erluin fortified Peronne in 1001, "as a defence against marauders, and a refuge for the husbandmen of the country."[164] But permission had probably to be asked in all these cases, as it certainly had in the last. The Carlovingian sovereigns represented a well-ordered state, modelled on the pattern of the Roman Empire; they were jealous of any attempts at self-defence which did not proceed from the State, and thus as long as they had the power they strove to put down all associations or buildings of a military character which did not emanate from their imperial authority.

The history of the 9th and 10th centuries is the history of the gradual break-up of the Carlovingian Empire, and the rise of feudalism on its ruins. In 877, the year of his death, Charles the Bald signed a decree making the counts of the provinces, who until then had been imperial officers, hereditary. He thus, as Sismondi says, annihilated the remains of royal authority in the provinces.[165] The removable officers now became local sovereigns. Gradually, as the Carlovingian Empire fell to pieces, the artificial organisation of the feudal system arose to take its place. By the end of the 10th century the victory of feudalism was complete; and the victory of feudalism was the victory of the private castle.

"The very word castle," says Guizot, "brings with it the idea of feudal society; we see it rising before us. It was feudalism that built these castles which once covered our soil, and whose ruins are still scattered upon it. They were the declaration of its triumph. Nothing like them had existed on Gallo-Roman soil. Before the Germanic invasion, the great landed proprietors dwelt either in the cities, or in beautiful houses agreeably situated near the cities."[166] These Gallo-Roman villas had no fortifications;[167] nor were the Roman villas in England fortified.[168] It was the business of the State to defend the community; this was the theory so long sustained by imperial Rome, and which broke down so completely under the later Carlovingians.

In the time of Charlemagne and Louis le Debonnaire, even the royal palaces do not appear to have been fortified. They were always spoken of as _palatia_, never as _castella_. The Danes, when they took possession of the palace of Nimeguen in 880, fortified it with ditches and banks.[169] Charles the Bald appears to have been the first to fortify the palace of Compiègne.[170]

Although there can be no doubt that private castles had become extremely common on the mainland of Western Europe before the end of the 10th century, it is more difficult than is generally supposed to trace their first appearance. Historians, even those of great repute, have been somewhat careless in translating the words _castrum_ or _castellum_ as _castle_ or _château_, and taking them in the sense of the feudal or private castle.[171] We have already pointed out that these words in our Anglo-Saxon charters mean a town or village.[172] The fact is that from Roman times until toward the end of the 9th century the words _castrum_ and _castellum_ are used indifferently for a fortified city or town, or a temporary camp. The expression _civitates et castella_ is not uncommon, and might lead one to think that a distinction was drawn between large and small towns, or forts. But it is far more likely that it is a mere pleonasm, a bit of that redundancy which was always dear to the mediæval scribe who was trying to write well. For as the instances cited in the Appendix will prove, we constantly find the words _castrum_ and _castellum_ used for the same town, sometimes even in the same paragraph. Later, from the last quarter of the 9th century to the middle of the 12th century, these same words are used indifferently for a town or a castle, and it is impossible to tell, except by the context, whether a town or a castle is meant; and often even the context throws no light upon it.

This makes it extremely difficult to say with any exactness when the private castle first arose. We seem indeed to have a fixed date in the Capitulary of Pistes, issued by Charles the Bald in 864,[173] in which he straightly ordered that all who had made castles, forts, or hedge-works without his permission should forthwith be compelled to destroy them, because through them the whole neighbourhood suffered depredation and annoyance. This edict shows, we might argue, that private castles were sufficiently numerous by the year 864 to have become a public nuisance, calling for special legislation. But the chronicles of the second half of the 9th century do not reveal any extensive prevalence of private castles. Indeed, after studying all the most important chronicles of Neustria and Austrasia during this period, the present writer has only been able to find four instances of fortifications which have any claim at all to be considered private castles; and even this claim is doubtful.[174]

When we come to the chroniclers of the middle of the 10th century we find a marked difference. It is true that the words _castrum_, _castellum_, _municipium_, _oppidum_, _munitio_, are still used quite indifferently by Flodoard and other writers for one and the same thing, and that in a great many cases they obviously mean a fortified town. But there are other cases where they evidently mean a castle. And if we compare these writers with the earlier ones in the same way as we have already compared the pre-Conquest portion of the _Anglo-Saxon Chronicle_ with the chroniclers of the 11th and 12th centuries, we find the same contrast between them. In the pages of Flodoard or Ademar the action constantly turns on the building, besieging, and burning of castles, which by whatever name they are called, have every appearance of being private castles. In fact before we get to the end of the century, the private castle is as much the leading feature of the drama as it is in the 11th or 12th centuries.

Why, then, had the chroniclers no fresh word for a thing which was in its essential nature so novel? The obvious and only answer is that the private castle in its earlier stages was nothing more than an embankment with a wooden stockade thrown round some _villa_ or farm belonging to a private owner, and was therefore indistinguishable in appearance, though radically different in idea, from the fortifications which had hitherto been thrown up for the protection of the community.[175] How easily we may be mistaken in the meaning of the word _castellum_, if we interpret it according to modern ideas, may be seen by comparing the account of the bridge built by Charlemagne over the Elbe, in the _Annales Laurissenses_, with Eginhards narrative of the same affair. The former states that Charlemagne built a _castellum_ of wood and earth at each end of the bridge, while the latter tells us that it was a _vallum_ to protect a garrison which he placed there. This, however, was a work of public utility, and not a private castle. But scanty as the evidence is, it all leads us to infer that the first private castles were fortifications of this simple nature.[176] Mazières-on-the-Meuse, which was besieged for four weeks by Archbishop Hervey, took its name from the _macerias_ or banks which Count Erlebald had constructed around it. It is impossible to say whether this enclosure should be called a castle or a town, but in idea it was certainly a castle, since it was an enclosure formed for private, not for public interests.

Whether these first private castles were provided with towers we have no evidence either to prove or to disprove. No instance occurs from which we can conclude that they possessed any kind of citadel, before the middle of the 10th century.[177] But before the century is far advanced, we hear of towers in connection with the great towns, which, whether they were originally mural towers or not, are evidently private strongholds, and may justly be called keeps. The earliest instance known to the writer is in 924, when the tower of the _presidium_ where Herbert Count of Vermandois had imprisoned Charles the Simple was burnt accidentally.[178] This tower must have been restored, as nine years later it withstood a six weeks' siege from King Raoul. A possibly earlier instance is that of Nantes, where Bishop Fulcher had made a castle in 889; for when this castle was restored by Count Alan Barbetorte (937-943), we are told that he _restored_ the principal tower and made it into his own house.[179] Count Herbert built a keep in Laon before 931; and this appears to have been a different tower to the one attached to the royal house which Louis d'Outremer had built at the gate of the city.[180] We hear also of towers at Amiens (950), Coucy (958), Chalons (963), and Rheims (988). All these towers, it will be observed, are connected with towns.[181] The first stone keep in the country for whose date we have positive evidence, is that of Langeais, built by Fulk Nerra, Count of Anjou, about the year 994; its ruins still exist.

But we are concerned more particularly here with the origin of the motte-and-bailey castle. The exact place or time of its first appearance is still a matter of conjecture. Certainly there is not a word in the chronicles which is descriptive of this kind of castle before the beginning of the 11th century.[182] The first historical mention of a castle which is clearly of the motte-and-bailey kind is in the Chronicle of St Florent le Vieil, where, at a date which the modern biographer of Fulk Nerra fixes at 1010, we learn that this same Count of Anjou built a castle on the western side of the hill Mont-Glonne, at St Florent le Vieil, on the Loire, and threw up an _agger_ on which he built a wooden tower.[183] In this case the word _agger_ evidently means a motte. But Fulk began to reign in 987; he was a great builder of castles, and was famed for his skill in military affairs.[184] One of his first castles, built between 991 and 994, was at Montbazon, not far from Tours. About 500 metres from the later castle of Montbazon is a motte and outworks, which De Salies not unreasonably supposes to be the original castle of Fulk.[185] Montrichard, Chateaufort, Chérament, Montboyau, and Baugé are all castles built by Fulk, and all have or had mottes. Montboyau is the clearest case of all, as it was demolished by Fulk a few years after he built it, and has never been restored, so that the immense motte and outworks which are still to be seen remain very much in their original state, except that a modern tower has been placed on the motte, which is now called Bellevue.[186]

It was a tempting theory at one time to the writer to see in Fulk Nerra the inventor of the motte type of castle, for independently of his fame in military architecture, he is the first mediæval chieftain who is known to have employed mercenary troops.[187] Now as we have already suggested in Chapter I., the plan of the motte-and-bailey castle strongly suggests that there may be a connection between its adoption and the use of mercenaries. For the plan of this kind of castle seems to hint that the owner does not only mistrust his enemies, he also does not completely trust his garrison. The keep in which he and his family live is placed on the top of the motte, which is ditched round so as to separate it from the bailey; the provisions on which all are dependent are stored in the cellar of the keep, so that they are under his own hand; and the keys of the outer ward are brought to him every night, and placed under his pillow.[188]

But unfortunately for this theory, there is some evidence of the raising of mottes at an earlier period in the 10th century than the accession of Fulk Nerra. Thibault-le-Tricheur, who was Count of Blois and Chartres from 932 to 962, was also a great builder, and it is recorded of him that he built the keeps of Chartres, Chateaudun,[189] Blois, and Chinon,[190] and the castle of Saumur; these must have been finished before 962. Now there was anciently a motte at Blois, for in the 12th century, Fulk V. of Anjou burnt the whole fortress, "_except the house on the motte_."[191] There was also a motte at Saumur;[192] and the plan of the castle of Chinon is not inconsistent with the existence of a former motte.[193] These instances seem to put back the existence of the motte castle to the middle of the 10th century.

We know of no earlier claim than this, unless we were to accept the statement of Lambert of Ardres that Sigfrid the Dane, who occupied the county of Guisnes about the year 928, fortified the town, and enclosed his own _dunio_ with a double ditch.[194] If this were true, we have a clear instance of a motte built in the first half of the 10th century. But Lambert's work was written at the end of the 12th century, with the object of glorifying the counts of Guisnes, and its editor regards the early part of it as fabulous. That Sigfrid fortified the _town_ of Guisnes we can easily believe, as we know the Danes commonly did the like (see Chapter IV.); but that he built himself a personal castle is unlikely.[195]

It is the more unlikely, because the Danes in Normandy do not appear to have built personal castles until the feudal system was introduced there by Richard Sans Peur. The settlement in Normandy was not on feudal lines. "Rollo divided out the lands among his powerful comrades, and there is scarcely any doubt that they received these lands as inheritable property, without any other pledge than to help Rollo in the defence of the country."[196] "The Norman constitution at Rollo's death can be described thus, that the duke ruled the country as an independent prince in relation to the Franks; but for its internal government he had a council at his side, whose individual members felt themselves almost as powerful as the duke himself."[197] Sir Francis Palgrave asserts that feudalism was introduced into Normandy by the Duke Richard Sans Peur, the grandson of Rollo, towards the middle of the 10th century. He "enforced a most extensive conversion of allodial lands into feudal tenure," and exacted from his baronage the same feudal submission which he himself had rendered to Hugh Capet.[198]

It is quite in accordance with this that in the narrative of Dudo, who is our only authority for the history of Normandy in the 10th century, there is no mention of a private castle anywhere. We are told that Rollo restored the walls and towers of the _cities_ of Normandy,[199] and it is clear from the context that the _castra_ of Rouen, Fécamp, and Evreux, which are mentioned, are fortified cities, not castles. Even the ducal residence at Rouen is spoken of as a _palatium_ or an _aula_, not as a castle; and it does not appear to have possessed a keep until (as we are told by a later writer) the same Duke Richard who introduced the feudal system into Normandy built one for his own residence.[200] It is possible that when the feudal oath was exacted from the more important barons, permission was given to them to build castles for themselves; thus we hear from Ordericus of the castle of Aquila, built in the days of Duke Richard; the castle of the lords of Grantmesnil at Norrei; the castle of Belesme; all of which appear to have been private castles.[201] But there seems to have been no general building of castles until the time of William the Conqueror's minority, when his rebellious subjects raised castles against him on all sides. "Plura per loca aggeres erexerunt, et tutissimas sibi munitiones construxerunt."[202] It is generally, and doubtless correctly, supposed that _aggeres_ in this passage means mottes, and taking this statement along with the great number of mottes which are still to be found in Normandy, it has been further assumed (and the present writer was disposed to share the idea) that this was the time of the first invention of mottes. But the facts which have been now adduced, tracing back the first known mottes to the time of Thibault-le-Tricheur, and the county of Blois, show that the Norman claim to the invention of this mode of fortification must be given up. If the Normans were late in adopting feudalism, they were probably equally late in adopting private castles, and the fortifications of William I.'s time were most likely copied from castles outside the Norman frontier.[203]

It might be thought that the general expectation of the end of the world in the year 1000, which prevailed towards the end of the 10th century, had something to do with the spread of these wooden castles, as it might have seemed scarcely worth while to build costly structures of stone. But it is not necessary to resort to this hypothesis, because there is quite sufficient evidence to show that long before this forecast of doom was accepted, wood was a very common, if not the commonest, material used in fortification. The reader has only to open his Cæsar to see how familiar wooden towers and wooden palisades were to the Romans; and he has only to study carefully the chronicles of the 9th, 10th, 11th, and 12th centuries to see how all-prevalent this mode of fortification continued to be. The general adoption of the feudal system must have brought about a demand for cheap castles, which was excellently met by the motte with its wooden keep and its stockaded bailey. M. Enlart has pointed out that wooden defences have one important advantage over stone ones, their greater cohesion, which enabled them to resist the blows of the battering-ram better than rubble masonry.[204] Their great disadvantage was their liability to fire; but this was obviated, as in the time of the Romans, by spreading wet hides over the outsides. Stone castles were still built, where money and means were available, as we see from Fulk Nerra's keep at Langeais; but the devastations of the Northmen had decimated the population of Gaul; labour must have been dear, and skilled masons hard to find. In these social and economic reasons we have sufficient cause for the rapid spread of wooden castles in France.

The sum of the evidence which we have been reviewing is this: the earliest mottes which we know of were _probably_ built by Thibault-le-Tricheur about the middle of the 10th century. But in the present state of our knowledge we must leave the question of the time and place of their first origin open. The only thing about which we can be certain is that they were the product of feudalism, and cannot have arisen till it had taken root; that is to say, not earlier than the 10th century.