The Frankish Period

Add a subpage to this text

Read the help page for usage.

 

Engels’ manuscript The Frankish Period was not published during his lifetime.

The manuscript has two parts; the first part includes two chapters: “The Radical Transformation of the Relations of Landownership under the Merovingians and Carolingians” and “The District and Army Structure”. The second part of the manuscript bears the title “Note: The Franconian Dialect”. Each part is a complete whole. All headings are given in conformity with Engels’ manuscript.

The Frankish Period was first published in full (“The Franconian Dialect” was published earlier, see Note 47) in: Marx and Engels, Works, First Russian Edition, Vol. XVI, Part I, Moscow, 1937, and in the language of the original in: Friedrich Engels, Zur Geschichte und Sprache der deutschen Frühzeit. Ein Sammelband. Berlin, 1952, S. 97-152.

The Frankish Period (without “Note: The Franconian Dialect”) was published in English for the first time in: Marx, Engels, Pre-Capitalist Socio-Economic Formations. A Collection, Progress Publishers, Moscow, 1979.

THE RADICAL TRANSFORMATION OF THE RELATIONS

OF LANDOWNERSHIP UNDER THE MEROVINGIANS

AND CAROLINGIANS

The mark system remained the basis of almost the entire life of the German nation till the end of the Middle Ages. Eventually, after an existence of one and a half millennia, it gradually disintegrated for purely economic reasons. It succumbed to economic advances with which it was unable to keep pace. We shall later examine its decline and ultimate destruction and we shall see that remnants of the mark system continue to exist even today.

It was only at the expense of its political importance that it was able to survive for so long. For centuries it had been the form embodying the freedom of the Germanic tribes. Then it became the basis of the people’s bondage for a thousand years. How was this possible?

The earliest community, as we have seen, comprised the whole people. Originally the people owned all the appropriated land. Later the whole body of inhabitants of a district [Gau], who were closely interrelated, became the owners of the territory settled by them, and the people as such retained only the right to dispose of the tracts which had not yet been claimed. The populace of the district in their turn handed over their field and forest marks to individual village communities, which likewise consisted of closely kindred people, and in this case too the land that was left over was retained by the district. The same procedure was followed when the original villages set up new village colonies—they were provided with land from the old mark by the parent village.

With the growth of the population and the further development of the people the blood-ties, on which here as everywhere the entire national structure was based, increasingly fell into oblivion. This was first the case with regard to the people as a whole. The common descent was less and less seen as real consanguinity, the memory of it became fainter and fainter and what remained was merely the common history and the dialect. On the other hand, the inhabitants of a district naturally retained an awareness of their .consanguinity for a longer time. The people thus came to mean merely a more or less stable confederation of districts. This seems to have been the state of affairs among the Germans at the time of the great migrations. Ammianus Marcellinus reports this definitely about the Alamanni, and in the local law it is still everywhere apparent. The Saxons were still at this stage of development during Charlemagne’s time and the Frisians until they lost their independence.

But the migration on to Roman soil broke the blood-ties, as it was bound to. Although the intention was to settle by tribes and kindreds, it was impossible to carry this through. The long marches had mixed together not only tribes and kindreds but also entire peoples. Only with difficulty could the blood-ties of the individual village communities still be held together, and these became thus the real political units of which the people consisted. The new districts on Roman territory were from the start, or soon became, judicial divisions set up more or less arbitrarily—or occasioned by conditions found already in existence.

The people thus disintegrated into an association of small village communities, between which there existed no or virtually no economic connection, for every mark was self-sufficient, producing enough to satisfy its own needs, and moreover the products of the various neighbouring marks being almost exactly the same. Hardly any exchange could therefore take place between them. And since the people consisted entirely of small communities, which had identical economic interests, but for that very reason no common ones, the continued existence of the nation depended on a state power which did not derive from these communities but confronted them as something alien and exploited them to an ever increasing extent.

The form of this state power depends in its turn on the form of the communities at the time in question. Where, as among the Aryan peoples of Asia and the Russians, it arises at a time when the fields are still cultivated by the community for the common account, or when at any rate the fields are only temporarily allotted by it to individual families, i.e. when there is as yet no private property in land, the state power appears as despotism. On the other hand, in the Roman lands conquered by the Germans, the individual shares in arable land and meadows already take, as we have seen, the form of the allodium, the owners’ free property subject only to the ordinary mark obligations. We must now examine how on the basis of this allodium a social and political structure arose, which—with the usual irony of history—in the end dissolved the state and completely abolished the allodium in its classical form.

The allodium made the transformation of the original equality of landed property into its opposite not only possible but inevitable. From the moment it was established on formerly Roman soil, the German allodium became what the Roman landed property adjacent to it had long been—a commodity. It is an inexorable law of all societies based on commodity production and commodity exchange that the distribution of property within them becomes increasingly unequal, the opposition of wealth and poverty constantly grows and property is more and more concentrated in a few hands. It is true that this law reaches its full development in modern capitalist production, but it is by no means only in it that this law operates. From the moment therefore that allodium, freely disposable landed property, landed property as commodity, arose, from that moment the emergence of large-scale landed property was merely a matter of time.

But in the period we are concerned with, farming and stock-breeding were the principal branches of production. Landed property and its products constituted the by far largest part of wealth at that time. Other types of movable wealth that existed then followed landed property as a matter of course, and gradually accumulated in the same hands as landed property. Industry and trade had already deteriorated during the decline of the Roman empire; the German invasion ruined them almost completely. The little that was left was for the most part carried on by unfree men and aliens and remained a despised occupation. The ruling class which, with the emerging inequality in property, gradually arose could only be a class of big landowners, its form of political rule that of an aristocracy. Though, as we shall see, political levers, violence and deceit contribute frequently, and as it seems even predominantly, to the formation and development of this class, we must not forget that these political levers only advance and accelerate an inevitable economic process. We shall indeed see just as often that these political levers impede economic development; this happens quite frequently, and invariably when the different parties concerned apply them in opposite or intersecting directions.

How did this class of big landowners come into being?

First of all we know that even after the Frankish conquest a large number of big Roman landowners remained in Gaul, whose estates were for the most part cultivated by free or bound copyholders against payment of rent (canon).

Furthermore we have seen that as a result of the wars of conquest the monarchy had become a permanent institution and real power among all Germans who had moved out, and that it had turned the land which had formerly belonged to the people into royal domains and had likewise appropriated the Roman state lands. These crown lands were constantly augmented by the wholesale seizure of the estates of so-called rebels during the many civil wars resulting from the partitions of the empire. But rapidly as these lands increased, they were just as rapidly squandered in donations to the Church and to private individuals, Franks and Romans, retainers (antrustions) and other favourites of the king. Once the rudiments of a ruling class comprising the big and the powerful, landlords, officials and generals had formed, during and because of the civil wars, local rulers tried to purchase their support by grants of land. Roth has conclusively proved that in most cases these were real grants, transfers of land which became free, inheritable and alienable property, until this was changed by Charles Martel.

When Charles took over the helm of state, the power of the kings was completely broken but, as yet, by no means replaced by that of the major-domos.

The class of grandees, created under the Merovingians at the expense of the Crown, furthered the ruin of royal power in every way, but certainly not in order to submit to the major-domos, their compeers. On the contrary, the whole of Gaul was, as Einhard says, in the hands of these

“tyrants, who were arrogating power to themselves everywhere” (tyrannos per totam Galliam dominatum sibi vindicantes).

This was done not only by secular grandees but also by bishops, who appropriated adjacent counties and duchies in many areas, and were protected by their immunity and the strong organisation of the Church. The internal disintegration of the empire was followed by incursions of external enemies. The Saxons invaded Rhenish Franconia, the Avars Bavaria, and the Arabs moved across the Pyrenees into Aquitania. In such a situation, mere subjection of the internal enemies and expulsion of the external ones could provide no long-term solution. A method had to be found of binding the humbled grandees, or their successors appointed by Charles, more firmly to the Crown. And since their power was up to then based on large-scale landed property, the first prerequisite for this was a total transformation of the relations of landownership. This transformation was the principal achievement of the Carolingian dynasty. The distinctive feature of this transformation is that the means chosen to unite the empire, to tie the grandees permanently to the Crown and thus to make the latter more powerful, in the end led to the complete impotence of the Crown, the independence of the grandees and the dissolution of the empire.

To understand how Charles came to choose this means, we must first examine the property relations of the Church at the time, which anyway cannot be passed over here, being an essential element of contemporary agrarian relations.

Even during the Roman era, the Church in Gaul owned considerable landed property, the revenue from which was further increased by its great privileges with regard to taxes and other obligations. But it was only after the conversion of the Franks to Christianity that the golden age began for the Gallic Church. The kings vied with one another in making donations of land, money, jewels, church utensils, etc., to the Church. Already Chilperic used to say (according to Gregory of Tours):

“See how poor our treasury has become, see, all our wealth has been transferred to the Church.”

Under Guntram, the darling and lackey of the priests, the donations exceeded all bounds. Thus the confiscated lands of free Franks accused of rebellion mostly became the property of the Church.

The people followed the lead of the kings. Small man and big could not give enough to the Church.

“A miraculous cure of a real or imagined ailment, the fulfilment of an ardent wish, e.g. the birth of a son or deliverance from danger, brought the Church whose saint had proved helpful a gift. It was deemed the more necessary to be always open-handed as both among high and low the view was widespread that gifts to the Church led to the remission of sins” (Roth, p. 250).

Added to this was the immunity protecting the property of the Church from violation at a time of incessant civil wars, looting and confiscation. Many a small man thought it wise to cede his property to the Church provided he retained its usufruct at a moderate rent.

Yet all this was not sufficient for the pious priests. With threats of eternal punishment in hell they virtually extorted more and more donations, so that as late as 811 Charlemagne reproaches them with this in the Aachen Capitulary, adding that they induce people

“to commit perjury and bear false witness, so as to increase your” (the bishops’ and abbots’) “wealth”.

Unlawful donations were obtained by hook or by crook in the hope that, apart from its privileged judicial status, the Church had sufficient means to cock a snook at the judiciary. There was hardly any Gallic Church Council in the sixth and seventh centuries that did not threaten to excommunicate anyone trying to contest donations to the Church. In this way even formally invalid donations were to be made valid, and the private debts of individual clerics protected against collection.

“We see that truly contemptible means were employed to arouse, again and again, the desire for making donations. When descriptions of heavenly bliss and infernal torment were no longer effective, relics were brought from distant parts, translations were arranged and new churches built; this was a veritable business in the ninth century” (Roth, p. 254). “When the emissaries of the St. Medard monastery in Soissons by much assiduous begging obtained the body of Saint Sebastian in Rome and in addition stole that of Gregory, and both bodies were deposited in the monastery, so many people flocked to see the new saints that the whole area seemed to be swarming with locusts, and those seeking relief were cured not individually but in whole herds. The result was that the monks measured the money by the bushel, counting as many as 85, and their stock of gold amounted to 900 pounds” (p. 255).

Deceit, legerdemain, the appearance of the dead, especially of saints, and finally also, and even predominantly, the forging of documents, were used to obtain riches for the Church. The forging of documents was—to let Roth speak again— “practised by many clerics on a vast scale ... this business began very early.... The extent of this practice can be seen from the large number of forged documents contained in our collections. Of Bréquigny’s 360 Merovingian certificates nearly 130 are definitely forgeries.... The forged testament of Remigius was used by Hincmar of Reims to procure his church a number of properties, which were not mentioned in the genuine testament, although the latter had never been lost and Hincmar knew very well that the former was spurious.”

Even Pope John VIII tried to obtain the possessions of the St. Denis monastery near Paris by means of a document which he knew to be a forgery. (Roth, pp. 256 ff.)

No wonder then that the landed property the Church amassed through donations, extortion, guile, fraud, forgery and other criminal activities assumed enormous proportions within a few centuries. The monastery of Saint-Germain-des-Prés, now within the perimeter of Paris, at the beginning of the ninth century owned landed property of 8,000 mansi or hides, an area which Guérard estimates at 429,987 hectares with an annual yield of one million francs=800,000 marks. If we use the same average, i.e. an area of 54 hectares with a yield of 125 francs=100 marks per hide of land, then the monasteries St. Denis, Luxeuil, St. Martin de Tours, each owning 15,000 mansi at that time, held landed property of 810,000 hectares with an income of 1^1/2 million marks. And this was the position after the confiscation of Church property by Pepin the Short! Roth estimates (p. 249) that the entire property of the Church in Gaul at the end of the seventh century was probably above, rather than below, one-third of the total area.

These enormous estates were cultivated partly by unfree and in part also by free copyholders of the Church. Of the unfree, the slaves (servi) were originally subject to unmeasured service to their lords, since they were not persons in law. But it seems that for the resident slaves too a customary amount of duties and services was soon established. On the other hand, the services of the other two unfree classes, the colons and lites (we have no information about the difference in their legal position at that time) were fixed and consisted in certain personal services and corvée as well as a definite part of the produce of their plot. These were long established relations of dependence. But for the Germans it was something quite new that free men were cultivating not their own or common land. It is true that the Germans met quite frequently free Roman tenants in Gaul and in general in territories where Roman law prevailed; however during the settlement of the country care was taken to ensure that they themselves did not have to become tenants but could settle on their own land. Hence before free Franks could become somebody’s copyholders they must have in some way or other lost the allodium they received when the country was being occupied, a distinct class of landless free Franks must have come into existence.

This class arose as. a result of the beginning concentration of landed property, owing to the same causes as led to this concentration, i.e., on the one hand civil wars and confiscations and on the other the transfer of land to the Church mainly due to the pressure of circumstances and the desire for security. The Church soon discovered a specific means to encourage such transfers, it allowed the donor not only to enjoy the usufruct of his land for a rent, but also to rent a piece of Church land as well. For such donations were made in two forms. Either the donor retained the usufruct of his farm during his lifetime, so that it became the property of the Church only after his death (donatio post obitum). In this case it was usual, and was later expressly laid down in the kings’ Capitularies, that the donor should be able to rent twice as much land from the Church as he had donated. Or the donation took effect immediately (cessio a die praesente) and in this case the donor could rent three times as much Church land as well as his own farm, by means of a document known as precaria, issued by the Church—which transferred the land to him, usually for the duration of his life, but sometimes for a longer or shorter period. Once a class of landless free men had come into being, some of them likewise entered into such a relationship. The precaria they were granted seem at first to have been mostly issued for five years, but in their case too they were soon made out for life.

There is scarcely any doubt that even under the Merovingians relations very similar to those obtaining on Church estates developed also on the estates of the secular magnates, and that here too free and unfree rent-paying tenants were living side by side. They must have been very numerous as early as Charles Martel’s rule for otherwise at least one aspect of the transformation of landownership relations initiated by him and completed by his son and grandson would be inexplicable.

This transformation depended basically on two new institutions. First, in order to keep the barons of the empire tied to the Crown, the Crown lands they received were now as a rule no longer a gift, but only a “beneficium”, granted for life, and moreover on certain conditions nonfulfilment of which entailed the forfeiture of the land. Thus they became themselves tenants of the Crown. And secondly, in order to ensure that the free tenants of the barons turned up for military service, the latter were granted some of the district count’s official powers over the free men living on their estates and appointed their “seniores”. For the present we need only consider the first of these two changes.

When subduing the rebellious small “tyrants” Charles probably—we have no information regarding this—confiscated their landed property according to old custom, but in so far as he reinstated them in their offices and dignities he will have granted it to them entirely or in part as a benefice. He did not yet dare to treat the Church land of recalcitrant bishops in the same way. He deposed them and gave their positions to people devoted to him, though the only clerical trait of many of them was their tonsure (sola tonsura clericus). These new bishops and abbots then began at his bidding to transfer large tracts of Church land to laymen as precaria. Such instances had occurred earlier too, but it was now done on a mass scale. His son Pepin went considerably further. The Church was in decay, the clergy despised, the Pope, hard pressed by the Langobardi, depended exclusively on Pepin’s support. He helped the Pope, favoured the extension of his ecclesiastical rule and held the Pope’s stirrup. But as a remuneration he incorporated the by far largest part of the Church land into the Crown estates and left the bishops and monasteries an amount just sufficient for their maintenance. The Church acquiesced passively in this first large-scale secularisation, the synod of Lestines confirmed it, albeit with a restrictive clause, which was, however, never observed. This huge mass of land placed the exhausted Crown estate once more on a secure footing and was to a large extent used for further grants, which in fact soon assumed the form of ordinary benefices.

Let us add here that the Church was soon able to recover from this blow. Directly after the conflict with Pepin the worthy men of God resumed their old practices. Donations came once more thick and fast from all directions, the small free peasants were still in the same sorry plight between hammer and anvil as they had been for the past 200 years. Under Charlemagne and his successors they fared far worse still and many entrusted themselves and all their possessions to the protection of the crosier. The kings returned some of their booty to favoured monasteries, and donated vast stretches of Crown land to others, especially in Germany. The blessed times of Guntram seemed to have returned for the Church during the reign of Louis the Pious. The monastery archives contain especially numerous records of donations made in the ninth century.

The benefice, this new institution, which we must now examine closer, was not yet the future fief, but certainly its embryo. It was from the outset granted for the common span of life of both the conferrer and the recipient. If one or the other died, it reverted to the owner or his heirs. To renew the former relationship, a new transfer of property to the recipient or his heirs had to be made. Hence the benefice was subject to both “throne-fall” and “home-fall”, to use a later terminology. Throne-fall soon fell into desuetude; the great beneficiaries became more powerful than the king. Home-fall, even at an early stage, not infrequently entailed the re-transfer of the estate to the heir of the former beneficiary. Patriciacum (Percy), an estate near Autun, which Charles Martel granted as a benefice to Hildebrannus, remained in the family passing from father to son for four generations, until in 839 the king presented it to the brother of the fourth beneficiary as full property. Similar cases occur quite frequently since the mid-eighth century.

The benefice could be withdrawn by the conferrer in all cases in which confiscation of property was applicable. And there was no shortage of such cases under the Carolingians. The risings in Alamannia under Pepin the Short, the conspiracy of the Thuringians and the repeated risings of the Saxons invariably led to new confiscations, either of free peasant land or of magnates’ estates and benefices. This occurred also, despite all treaty stipulations to the contrary, during the internal wars under Louis the Pious and his sons. Certain non-political crimes were also punished by confiscation.

The Crown could moreover withdraw benefices if the beneficiary neglected his general obligations as a subject, e.g., did not hand over a robber who had sought asylum, did not turn up armed for a campaign, did not pay heed to royal letters, etc.

Furthermore benefices were conferred on special terms, the infringement of which entailed their withdrawal, which of course did not extend to the rest of the property of the beneficiary. This was the case, for example, when former Church estates were granted and the beneficiary failed to pay the Church the dues that went with them (nonae et decimae). Or if he let the estate deteriorate, in which case a year’s notice was usually first given as a warning so that the beneficiary could improve matters to avert confiscation which would otherwise follow, etc. The transfer of an estate could also be tied to definite services and this was indeed done more and more frequently as the benefice gradually developed into the fief proper. But initially this was by no means necessary, especially with regard to military service, for many benefices were conferred on lower clerics, monks, and women both spiritual and lay.

Finally it is by no means impossible that in the beginning the Crown also conferred land subject to recall or for a definite period, i.e. as precaria. Some of the information and the procedure of the Church make this probable. But at any rate this ceased soon for the granting of land as a benefice became prevalent in the ninth century.

For the Church—and we must assume that this applied to the big landowners and beneficiaries as well—the Church, which previously granted estates to its free tenants mostly only as precaria for a definite period of time, had to follow the stimulus given by the Crown. The Church not only began to grant benefices as well, but this kind of grant became so predominant that already existing precaria were turned into lifelong ones and imperceptibly became benefices, until the former merged almost completely into the latter in the ninth century. Beneficiaries of the Church and also of secular magnates must have played an important part in the state as early as the second half of the ninth century, some of them must have been men of substantial property, the founders of the future lower nobility. Otherwise Charles the Bald would not have so vigorously helped those who had been without reason deprived of their benefices by Hincmar of Laon.

The benefice, as we see, has many aspects which recur in the developed fief. Throne-fall and home-fall are common to both. The benefice, like the fief, can only be revoked under certain conditions. The social hierarchy created by the benefices, which descends from the Crown through the big beneficiaries—the predecessors of the imperial princes—to the medium beneficiaries—the future nobility—and from them to the free and unfree peasants, the bulk of whom lived in mark communities, formed the foundation for the future compact feudal hierarchy. Whereas the subsequent fief is, in all circumstances, held in return for services and entails military service for the feudal lord, the benefice does not yet require military service and other services are by no means inevitable. But the tendency of the benefice to become an estate held in return for services is already obvious, and spreads steadily during the ninth century; and in the same measure as it unfolds, the benefice develops into the fief.

Another factor contributed to this development, i.e., the changes which took place in the district and army structure first under the influence of big landed property and later under that of the big benefices, into which big landed property was increasingly transformed as a result of the incessant internal wars and the confiscations and retransfers associated with them.

It is evident that only the pure, classical form of the benefice has been examined in this chapter, which was certainly only a transitory form and did not even appear everywhere simultaneously. But such historical manifestations of economic relations can only be understood if they are considered in their pure state, and it is one of the chief merits of Roth that he has laid bare this classical form of the benefice, stripping it of all its confusing appendages.

THE DISTRICT AND ARMY STRUCTURE

The transformation in the position of landed property just described was bound to influence the old structure. It caused just as significant changes in the latter, and these in their turn had repercussions on the relations of landed property. For the present we shall leave aside the remodelling of the political structure as a whole and confine ourselves to an examination of the influence the new economic position exerted on the still existing remnants of the old popular structure in the districts and the army.

As early as the Merovingian period we frequently encounter counts and dukes as administrators of Crown estates. But it was not until the ninth century that certain Crown estates were definitely linked to the countship in such a way that the count of the day received their revenue. The formerly honorary office had been transformed into a paid one. In addition to this we find the counts holding royal benefices granted to them personally, which is something self-evident under the conditions of that time. The count thus became a powerful landowner within his county.

First of all it is obvious that the authority of the count was bound to suffer when big landed proprietors arose under him and side by side with him. People who had often enough scorned the commands of the kings under the Merovingians and early Carolingians could be expected to show even less respect for the orders of the count. Their free tenants, confident ‘of the protection of powerful landlords, just as frequently disregarded the count’s summons to appear in court or turn up for his levy to the army. This was one of the reasons that led to grants being made in the form of benefices instead of allodial grants and later to the gradual transformation of most of the formerly free big estates into benefices.

This alone was not enough to ensure that the free men living on the estates of the magnates did in fact perform their services to the State. A further change had to be introduced. The king saw himself compelled to make the big landlords responsible for the appearance of their free tenants at court and for their performance of military and other traditional services to the State in the same way as hitherto the count was held accountable for all free inhabitants of his county. And this could only be accomplished if the king gave the magnates some of the count’s official powers over their tenants. It was the landlord or beneficiary who had to make sure that his people appeared before the court, they therefore had to be summoned through him. He had to bring them to the army, therefore the levy had to be effected by him, and so that he might always be held accountable for them he had to lead them and have the right to impose military discipline on them. But it was and continued to be the king’s service that the tenants performed, and the recalcitrant was punished not by the landlord but by the royal count, and the fine went to the royal fisc.

This innovation too goes back to Charles Martel. At any rate only since his time do we find the custom of high ecclesiastical dignitaries taking the field themselves, a custom which, according to Roth, was due to the fact that Charles made his bishops join the army at the head of their tenants in order to ensure that the latter turned up. Undoubtedly this also applied to the secular magnates and their tenants. Under Charlemagne the new arrangement is already firmly established and universally enforced.

But this caused a substantial change also in the political position of the free tenants. They who had formerly been on an equal footing with their landlord before the law, however much they depended on him economically, now became his subordinates also in the legal sphere. Their economic subjection was politically sanctioned. The landlord becomes Senior, Seigneur, the tenants become his homines, the “lord” becomes the master of his “man”. The legal equality of the free men has disappeared; the man on the lowest rung of the ladder, his full freedom already greatly impaired by the loss of his ancestral land, moves down another step nearer the unfree. The new “lord” rises that much higher above the level of the old communal freedom. The basis of the new aristocracy, already established economically, is recognised by the State and becomes one of the fully operative driving wheels of the State machinery.

But alongside these homines made up of free tenants there existed yet another kind. These were impoverished free men who had voluntarily entered into the service or become retainers of a magnate. The retinue of the Merovingians were the antrustions, the magnates of that time will likewise have had their retainers. The retainers of the king were, under the Carolingians, called vassi, vasalli or gasindi, terms which had been used for unfree men in the oldest codes of common law, but had now come to mean usually free retainers. The same expressions were applied to the grandee’s retainers, who now occur quite commonly and become an increasingly numerous and important element of society and State.

Old treaty formulas show how the grandees came to have such retainers. One of them (Formulae Sirmondicae 44) for instance says:

“Since it is known to one and all that I have not the wherewithal to feed and clothe myself, I ask of your” (the lord’s) “piety that I may betake and commend myself into your protection” (mundoburdum—guardianship, as it were) “so that ... you will be obliged to aid me with food and clothing, according as I shall serve you and merit the same; in return, may I be obliged to render you service and obedience in the manner of a freeman (ingenuili ordine); nor shall it be in my power to withdraw from your authority and patronage during my lifetime but I shall spend my days under your authority and protection.”

This formula provides full information about the origin and nature of the ordinary relations of allegiance stripped of all alien admixtures, and it is especially revealing because it presents the extreme case of a poor devil who has been reduced to absolute penury. The entry into the seignior’s retinue was effected by the two parties reaching a free agreement—free in the sense of Roman and modern law—often rather similar to the entry of a present-day worker into the service of a manufacturer. The “man” commended himself to the lord, and the latter accepted his commendation. It was confirmed by a handshake and an oath of allegiance. The agreement was lifelong and was only dissolved by the death of one of the two contractors. The liege man was obliged to carry out all services consistent with the position of a free man which his lord might impose on him. In return the lord provided for his keep and rewarded him as he thought fit. A grant of land was by no means necessarily involved and in fact it certainly did not take place in all cases.

Under the Carolingians, especially since Charlemagne, this relationship was not only tolerated but directly encouraged and eventually, it seems, made compulsory for all ordinary free men—by a Capitulary of 847—and regulated by the State. For example, the liege man could unilaterally annul the relationship with his lord only if the latter attempted to kill him, hit him with a stick, dishonour his wife or daughter or deprive him of his hereditary property (Capitulary of 813). The liege man moreover was bound to his lord as soon as he had received a value equivalent to one solidus from him. This again clearly shows how little at that time the vassal relationship was linked with the granting of land. The same stipulations are repeated in a Capitulary of 816, with the addition that the liege man was released from his obligations if his lord wrongfully attempted to reduce him to the status of an unfree man or failed to afford him the promised protection although he was able to do so.

With regard to his retainers the liege lord now had the same rights and duties towards the State as the landlord or beneficiary had with regard to his tenants. As before they were liable to serve the king, but here too the liege lord was interposed between the king and his counts. The liege lord brought the vassals to court, he called them up, led them in war and maintained discipline among them, he was responsible for them and their regulation equipment. This gave him a certain degree of penal authority over his subordinates, and was the starting point of the feudal lord’s jurisdiction over his vassals, which developed later.

In these two additional institutions, the formation of the retainer system and the transfer of the official powers of the counts, that is the State, to the landlord, the holder of a Crown benefice, and the liege lord over his subordinates—both tenants and landless retainers, who were soon all to be called vassi, vasalli or homines—in this political confirmation and strengthening of the actual power of the lord over his vassals we see an important further development of the germ of the fief system contained in the benefices. The hierarchy of social estates, from the king downwards through the big beneficiaries to their free tenants and finally to the unfree men, has in its official capacity become a recognised element of the political organisation. The State recognises that it cannot exist without its help. We shall see later how in actual fact this help was given.

The differentiation between retainers and tenants is only important in the beginning, in order to show that the dependence of free men came about in two ways. The two types of vassals very soon merged inseparably, in name as well as in fact. It became more and more customary for the big beneficiaries to commend themselves to the king, so that they were not only his beneficiaries but also his vassals. It was in the interest of the kings to make the magnates, bishops, abbots, counts and vassals swear the oath of allegiance to them personally (Annales Bertiniani 837 and other documents of the ninth century); consequently the distinction between the general oath of the subject and the specific oath of the vassal was bound to disappear soon. Thus all the great men gradually became vassals of the king. The slow transformation of the big landowners into a special estate, an aristocracy, was herewith recognised by the State, incorporated into the State structure and became one of its officially functioning elements.

Similarly the retainers of the individual big landowners gradually became tenants. Apart from providing board at the manorhouse, which after all could only be done for a small number of people, there was but one way of assuring oneself of retainers, that is by settling them on the ground, by granting them land as a benefice. A numerous militant retinue, the main prerequisite for the existence of the magnates in those times of perpetual fighting, could therefore only be obtained by granting land to the vassals. Consequently landless retainers gradually disappear from the manor while the mass of those settled on the lord’s land grows.

But the more this new element penetrated the old structure, the more it was bound to weaken the latter. The old direct exercise of State power by the king and the counts was more and more replaced by an indirect method; the seignior, to whom the common free men were increasingly tied by personal allegiance, now stood between them and the State. The count, the mainspring of the mechanism of State, was bound to recede into the background more and more, and so he did. In this situation Charlemagne acted as he generally used to do. First he encouraged the spread of the vassal relationship, as we have seen, until the independent small free men had almost disappeared, and when the weakening of his power to which this led became obvious, he tried to help it on its feet again by State intervention. Under such an energetic and formidable ruler this could be successful in some cases, but the force of circumstances created with his help asserted itself inexorably under his weak successors.

Charlemagne’s favourite method was to send out royal emissaries (missi dominici) with plenipotentiary powers. Where the ordinary royal official, the count, was unable to stem the spread of disorder, a special envoy was expected to do so. (This has to be historically substantiated and amplified.)

There was, however, another method, and this was to put the count in such a position that he had at his disposal material means to enforce his authority which were at least equal to those of the magnates in his county. This was only possible if the count too became a big landowner, which again could be brought about in two ways. Certain estates could be attached to the office of the count in the various districts as a sort of endowment, so that the count of the day administered them ex officio and received the revenue they yielded. Many examples of this kind can be found, especially in documents, from as early as the end of the eighth century, and this arrangement is quite usual from the ninth century onwards. It is self-evident that such endowments come for the most part from the king’s fiscal estates, and as early as the time of the Merovingians we often find counts and dukes administering the king’s fiscal estates situated in their territory.

Strangely enough there are also a good many examples (and even a formula for this purpose) of bishops using Church property to endow the office of the count, of course in the form of some sort of benefice since Church property was inalienable. The munificence of the Church is too well known to allow of any other reason for this but dire need. Under the growing pressure of neighbouring secular magnates no other resort was left to the Church but to ally itself with the remnants of the state authority.

These appurtenances associated with the count’s post (res comitatus, pertinentiae comitatus) were originally quite distinct from the benefices which were granted personally to the count of the day. These too were usually distributed generously, so that, endowment and benefices taken together, countships, originally honorary positions, had by then become very lucrative posts, and since Louis the Pious they were, like other royal favours, bestowed on people whom the king wanted to win over to his side or of whom he wanted to be sure. Thus it is said of Louis the Stammerer that he “quos potuit conciliavit [sibi], dans eis comitatus et abbatias ac villas” {Annales Bertiniani 877). The term honor, formely used to designate the office with reference to the honorary rights connected with it, acquired the same meaning as benefice in the course of the ninth century. And this necessarily caused a substantial change in the character of the count’s office, as Roth rightly emphasises (p. 408). Originally the seigniory, in so far as it was of a public character, was modelled upon the office of the count and invested with some of the count’s powers. Then, in the second half of the ninth century, the seigniory had become so widespread that it threatened to outweigh the count’s office and the latter could only maintain its authority by more and more assuming the characteristics of seigniory. The counts increasingly sought, and not without success, to usurp the position of a seignior vis-à-vis the inhabitants of their districts (pagenses) with regard to both their private and their public relations. Just as the other “lords” sought to subordinate the small people in their neighbourhood, so the counts tried, in an amicable way or by force, to induce the less well-off free inhabitants of their district to become their vassals. They succeeded the more easily as the mere fact that the counts could thus abuse their official power was the best proof that the remaining common free men could expect very little protection from the royal authority and its organs. Exposed to oppression from all quarters, the smaller free men had to be glad to find a patron, even at the cost of relinquishing their allodium and receiving it back as a mere benefice. Already in the Capitulary of 811 Charlemagne complained that bishops, abbots, counts, judges and centenarii by continuous legal chicanery and repeated summonses to the army reduced the small people to such a state that they agreed to transfer or sell their allodium to them, and that the poor bitterly lamented that they were being robbed of their property, etc. The greater part of free property in Gaul had in this way already passed into the hands of the Church, the counts and other magnates by the end of the ninth century (Hincmar, Annales Remenses 869). And somewhat later no free landed property belonging to small free men existed any longer in some provinces (Maurer, Einleitung, p. 212). When the increasing power of the beneficiaries and the declining power of the Crown had gradually caused benefices to become hereditary, the count’s office as a rule became hereditary too. If we saw the beginnings of the subsequent nobility in the large number of royal beneficiaries, here we see the seed of the territorial sovereignty of the future princes that evolved from the district counts.

While thus the social and political system changed completely, the old constitution of the army, based on the military service of all free men—a service which was both their right and their duty—remained outwardly unchanged, except that where the new relations of dependence existed, the seignior interposed himself between his vassals and the count. However, year by year the common free men were less able to carry the burden of military service. This consisted not only of personal service; the conscript had also to equip himself and to live at his own expense during the first six months. This continued until Charlemagne’s incessant wars knocked the bottom out of the barrel. The burden became so unbearable that in order to rid themselves of it the small free men began en masse to transfer not only their remaining property but also their own person and their descendants to the magnates, and especially to the Church. Charlemagne had reduced the free warlike Franks to such a state that they preferred to become bondsmen or serfs to avoid going to war. That was the consequence of Charlemagne’s insistence on maintaining, and even carrying to the extreme, a military system based on universal and equal landownership by all free men, at a time when the bulk of the free men had lost all or most of their landed property.

The facts, however, were stronger than Charlemagne’s obstinacy and ambition. The old army system was no longer tenable. To equip and provision the army at the expense of the State was even less feasible in that age of a subsistence economy run practically without money or commerce. Charlemagne was therefore obliged to restrict the liability to service in such a way that equipment and food could still remain the responsibility of the men themselves. This was done in the Aachen Capitulary of 807, at a time when the wars were reduced to mere border fights, and the continued existence of the empire seemed, on the whole, ensured. Firstly all the king’s beneficiaries without exception had to turn up, then those owning twelve hides (mansi) of land were to appear clad in armour, and therefore presumably also on horseback (the word caballarius—knight is used in the same Capitulary). Owners of three to five hides of land were also obliged to serve. Two owners having two hides of land each, three owners having one hide of land each, or six owners each possessing half a hide of land, had to send one man equipped by the others. As to free men who had no land at all but personal property worth five solidi, every sixth of them was to take the field and receive one solidus as pecuniary aid from each of the other five men. Moreover the obligation of the various parts of the country to take part in the fighting, an obligation which applied fully when the war was waged in the neighbourhood, was in the case of more distant wars reduced to between one-half and one-sixth of the total manpower, depending on the distance from the theatre of war.

Charlemagne evidently attempted to adapt the old system to the changed economic position of the men liable to military service, to rescue what he could still rescue. But even these concessions were of no avail, and he was soon compelled to grant further exemptions in the Capitulare de exercitu promovendo. The whole contents of this Capitulary, which is usually regarded as antecedent to that of Aachen, shows that it was undoubtedly drawn up several years later. According to it, one man has to do military service from every four hides of land, instead of three as previously. The owners of half a hide of land and those without land appear to be exempt from military service, and as regards beneficiaries their obligation is also restricted to the provision of one man for every four hides of land. Under Charlemagne’s successors the minimum number of hides of land obliged to provide one man seems even to have been raised to five.

It is strange that the mobilisation of the armoured owners of twelve hides of land seems to have encountered the greatest difficulties. At any rate, the order that they must turn up clad in armour is repeated innumerable times in the Capitularies.

Thus the common free men disappeared to an increasing extent. Just as the gradual separation from the land had driven part of them to become vassals of the new big landlords, so the fear of being completely ruined by military service actually drove the other part into serfdom. How rapidly this submission to servitude proceeded can be seen from the polyptychon (land register) of the Saint-Germain-des-Prés monastery, which then still lay outside Paris. It was compiled by abbot Irminon early in the ninth century, and among the tenants of the monastery it lists 2,080 families of colons, 35 of lites, 220 of slaves (servi), but only eight free families. In the Gaul of those days, however, the word colonus definitely denoted a serf. The marriage of a free woman to a colonus or slave subjected her Jo the lord as defiled (deturpatam) (Capitulary of 817). Louis the Pious commanded that “colonus vel servus” (of a monastery at Poitiers) “ad naturale servitium velit nolit redeat”. They received blows (Capitularies of 853, 861, 864 and 873) and were sometimes set free (see Guérard, Irmino). And these enthralled peasants were by no means of Romance stock, but according to the testimony of Jacob Grimm (Geschichte der deutschen Sprache, I, p. [537]), who examined their names, “almost exclusively Frankish, far outweighing the small number of Romance ones”.

This huge rise in the unfree population in its turn changed the class relations of the Frankish society. Alongside the big landlords, who at that time rapidly emerged as a social estate in its own right, and alongside their free vassals there appeared now a class of unfree men which gradually absorbed the remnants of the common free men. But these unfree men had either themselves been free or were children of free men; those who had lived for three or more generations in hereditary bondage formed a small minority. Moreover, for the most part they were not Saxon, Wendish, or other prisoners of war brought in from outside, but natives of Frankish or Romance origin. Such people, especially when they began to constitute the bulk of the population, were not as easy to deal with as inherited or foreign serfs. They were not yet used to servitude, the blows which even the colonus received (Capitularies of 853, 861, 873) were still seen as a humiliation and not as something natural. Hence the many plots and risings of unfree men and even peasant vassals. Charlemagne himself brutally crushed an uprising of the tenants of the bishopric of Reims. In a Capitulary of 821 Louis the Pious mentions slaves (servorum) plotting in Flanders and Menapiscus (on the upper Lys). Risings of the liege men (homines) of the Mainz bishopric had to be put down in 848 and 866. Orders to stamp out such plots are reiterated in capitularies from 779 onwards. The rising of the Stellinga in Saxony must likewise be included here. The fact that from the end of the eighth century and the beginning of the ninth gradually a definite limit was fixed for the obligations of the unfree men, and even of the settled slaves, and that this limit, which was not to be exceeded, was laid down by Charlemagne in his Capitularies, was obviously a consequence of the threatening attitude of the enthralled masses.

The price therefore which Charlemagne had to pay for his new Roman Empire was the annihilation of the social estate of common free men, who had constituted the entire Frankish people at the time of the conquest of Gaul, and the division of the people into big landlords, vassals and serfs. But with the common free men the old military system collapsed, and with these two the monarchy went down. Charlemagne had destroyed the foundation of his own power. It could still sustain him, but under his successors it became evident what the work of his hands had been in reality.

NOTE: THE FRANCONIAN DIALECT

This dialect has received curious treatment from philologists. Whereas Grimm let it disappear into French and High German, more recent ones grant it a spread extending from Dunkirk and Amsterdam to the Unstrut, Saale and Rezat, and in some cases even as far as the Danube and, through colonisation, to the Riesengebirge. While even a philologist like Moritz Heyne constructs an Old Low Franconian language from a manuscript of the Heliand prepared in Werden, a language that is almost pure Old Saxon with a very faint tinge of Franconian, Braune lumps together all the truly Low Franconian dialects without further comment as Saxon here and Dutch there. And finally Arnold limits the territory conquered by the Ripuarians to the area north of the watershed of the Ahr and the Mosel, letting everything situated to the south and south-west be occupied, first by Alamanni, later exclusively by the Chatti (whom he also lumps together with the Franks), thus letting them speak Alamannic Chattish.

First let us reduce the Franconian language area to its real limits. Thuringia, Hesse and Main Franconia have no other claim whatever to be included in it except that in the Carolingian period they were part of what was called Francia. The language spoken east of the Spessart and Vogelsberg and the Kahler Asten is anything but Franconian. Hesse and Thuringia have their own independent dialects, being inhabited by independent peoples; in Main Franconia a mixed Slav, Thuringian and Hessian population was permeated with Bavarian and Frankish elements and evolved its own peculiar dialect. Only if one employs as the main criterion the extent to which the High German sound shift penetrated into these dialects can these three linguistic branches be assigned to Franconian. Yet as we shall see, it is precisely this procedure which creates all the confusion when the Franconian language is assessed by non-Franks.

Let us commence with the oldest records and first view Moritz Heyne’s so-called Old Low Franconian in the correct light. The so-called Cotton Manuscript of the Heliand, prepared in Werden and now preserved in Oxford, is supposed to be Old Low Franconian because it was produced in the monastery of Werden, still on Frankish soil though close to the Saxon frontier. Here the old tribal boundary is, to this day, the boundary between Berg and Mark; of the abbeys situated in between, Werden belongs to Franconia, Essen to Saxony. Werden is bounded in the immediate vicinity, to the east and north, by indisputably Saxon communities; in the plain between the Ruhr and the Lippe the Saxon language pushes forward in places almost to the Rhine. The fact that a Saxon work is copied in Werden, obviously by a Frank, and that here and there this Frank has let slip from his pen Franconian word forms, is far from being sufficient reason to declare the language of the copy to be Franconian. Apart from the Cotton Heliand Heyne considers as Low Franconian a number of fragments from Werden that show the same character, and the remains of a psalm translation, which according to him originated in the area of Aachen, whereas Kern (Glossen in der Lex Salica) states quite simply that it is Dutch. In fact it does contain purely Dutch forms on the one hand, but also genuine Rhenish Franconian forms and even traces of the High German sound shift. It obviously originated on the frontier between Dutch and Rhenish Franconian, say between Aachen and Maastricht. Its language is much later than that of the two Heliand manuscripts.

The Cotton Heliand alone is enough, however, for us to establish beyond doubt from the few Franconian forms that occur in it some of the main differences between Franconian and Saxon.

I. In all Ingaevonian dialects the three persons of the present indicative plural all have the same ending, namely a dental preceded by a vowel: Old Saxon -à, Anglo-Saxon -dh, Old Frisian -th (which probably also stands for -dh). Thus Old Saxon hebbiad means “we have, you have, they have”; similarly, all three persons of fallan, gewinnan are the same: fallad, winnad. It is the third person that has taken over all three, but, mark well, with the specifically Ingaevonian loss of n before -d or -dh, the loss affecting all the three dialects mentioned. Of all living dialects, only Westphalian has preserved this peculiarity; to this very day Westphalian has vox, ji, se hebbed, etc. The other Saxon dialects no longer retain this feature, nor does West Frisian; they differentiate the three persons.

The West Rhenish psalms have, like Middle High German, -n in the first person plural, -t in the second, -nt in the third. However, at times the Cotton Heliand has, besides the Saxon forms, quite different forms: tholônd—they suffer, gornônd—you complain, and as the imperative, marient—announce, seggient— say, where Saxon requires tholôd, gornôd, mâriad, seggiad. These forms are not merely Franconian, they are in fact genuine local Werden, Berg dialect to this day. In Bergish we also find that all three persons of the present plural are the same, but end not as in Saxon in -à, but as in Franconian in -nt. As opposed to Märkish wi hebbed, there right on the border they say wi hant, and as in the above imperative seggient they say seient ens—[German] sagt einmal (tell me). On the basis of this simple observation, that here in Bergish the three persons have been levelled, Braune and others have quite simply declared the entire Bergish highlands to be Saxon. The rule certainly advanced into the area from Saxony; unfortunately, however, it is put into effect in the Franconian manner, thus proving the reverse of what it is intended to prove.

The loss of n before dentals in the Ingaevonian dialects is not restricted to this case; it is less common in Old Frisian, but fairly widespread in Old Saxon and Anglo-Saxon: mudh—Mund [mouth], kudh—kund [known], us—uns [us], odhar—ein anderer [other]. The Frankish copyist of the Heliand in Werden twice has the Franconian form andar for odhar. The Werden tax registers alternate between the Franconian form of the names Reinswind, Meginswind and the Saxon Reinswid and Meginswid. The psalms of the left bank of the Rhine, on the other hand, regularly have munt, kunt, uns; only once have the so-called Lipsius Glosses (excerpted from the lost manuscript of these psalms) farkutha abominabiles instead of farkuntha. Similarly, the Old Salic records have consistently preserved the n in the names Gund, Segenand, Chlodosindis, Ansbertus, etc., which is irrelevant. The modern Franconian dialects regularly have the n (sole exception in Bergish is the form 05—uns [us]).

II. The linguistic records from which the so-called Old Saxon grammar is usually constructed all belong to south-western Westphalia, Münster, Freckenhorst, Essen. The language of these records shows a few essential deviations not only from the general Ingaevonian forms, but also from such forms as have been preserved for us in proper names from Engern and Eastphalia as genuine Old Saxon; however, they are in curious agreement with Franconian and Old High German. The latest grammarian of the dialect, Cosijn, therefore even terms it Old West Saxon.

Since in this investigation we must almost totally rely on proper names in Latin documents, the demonstrable differences in form between West and East Saxon can only be few in number; they are restricted to two cases, but these are very important.

1. Anglo-Saxon and Old Frisian have -a in the genitive plural of all declensions. Old West Saxon, Old Franconian and Old High German, on the other hand, have -6. So what is the correct Old Saxon form? Should this dialect in fact deviate from the Ingaevonian rule on this point?

The documents from Engern and Eastphalia provide the answer. In Stedieraburg, Horsadal, Winethahûsen, Edingahûsun, Magathahurg and many other names, the first part of the compound is in the genitive plural and has -a. Even in Westphalia the -a has still not entirely disappeared: the Freckenhorst Roll once has Aningera lô and Wernera-Holthuson and the -o in Osnabrück is likewise an old genitive plural.

2. Similarly, the weak masculine in Franconian, as in Old High German, ends in -o, as opposed to Gothic-Ingaevonian -a. In Old West Saxon -o is likewise established as the rule; thus another deviation from Ingaevonian usage. But this by no means applies to Old Saxon as a whole. Not even in Westphalia did -o apply without exception; alongside -o the Freckenhorst Roll already has a whole succession of names in -a {Sîhoda, Uffa, Asica, Hassa, Wenda, etc.,); the Paderborn records in Wigand nearly always show -a, only exceptionally -o; in documents from Eastphalia -a dominates almost exclusively; so that Jakob Grimm {Geschichte der deutschen Sprache) already comes to the conclusion that there can be no mistaking the fact that -a and -an (in oblique cases) was the original Saxon form common to all parts of the nation. The advance of -o instead of -a was not restricted to Westphalia either. At the beginning of the 15th century the East Frisian men’s names of the chronicles, etc., almost regularly have -o; Fokko, Occo, Enno, Smelo, etc., as opposed to the earlier -a still preserved in odd cases in West Frisian.

It may therefore be taken for established that both deviations of West Saxon from the Ingaevonian rule are not originally Saxon but caused by foreign influence. This influence is easily explained by the fact that West Saxony was formerly Frankish territory. Only after the departure of the main mass of the Franks did the Saxons move across the Osning and Egge gradually up to the line that even today divides Mark and Sauerland from Berg and Siegerland. The influence of the Franks who remained behind and have now merged with the Saxons shows in those two cases of -o instead of -a; it is still unmistakable in the present-day dialects.

III. A peculiarity of the Rhenish Franconian language which extends from the Ruhr to the Mosel is the ending of the 1st [person] present indicative in -n, which is best preserved in cases where it is followed by a vowel: dat don ek—das tue ich [I do that], ek han—ich habe [I have] (Bergish). This verb form applies to the whole lower Rhine and the Mosel, at least as far as the Lotharingian border: don, han. The same peculiarity is already found in the left-bank Rhenish psalms: biddon—ich bitte [I ask], wirthon—ich werde [I become], though not consistently. This-n is lacking in the Salic dialect; there even the oldest record has ec forsacho, gelobo. It is also missing in Dutch. Old West Saxon is here distinct from Franconian in so far as it knows this -n in one conjugation only (the so-called second weak): skawôn—ich schaue [I look], thionôn—ich diene [I serve], etc. It is quite alien to Anglo-Saxon and Old Frisian. We may therefore assume that this -n is also a Franconian remnant in Old West Saxon.

Apart from the numerous proper names preserved in documents, etc., and the glosses of the Lex Salica, which are often distorted past recognition, we have almost no remains of the Salic dialect at all. Nevertheless, Kern (Die Glossen in der Lex Salica) has removed a considerable number of these distortions and established the text, in many cases with certainty, in others with great likelihood, demonstrating that it is written in a language that is the immediate precursor of Middle and Modern Dutch. But the material reconstructed in this way is naturally not directly applicable for the grammar. Apart from this, all we possess is the brief abjuration charm added to the Capitulary of Carloman of 743 and probably drawn up at the synod of Lestines, thus in Belgium. And here we come across two characteristic Franconian words right at the outset: ec forsacho—ich entsage [I renounce]. Ec for ich [I] is widespread among the Franks even today. In Trier and Luxemburg eich, in Cologne and Aachen êch, in Bergish ek. Though written Dutch has ik, ek is often heard in the vernacular, particularly in Flanders. The Old Salic names Segenandus, Segemundus, Segefredus are unanimous in showing e for i.

In forsacho, ch stands for g between vowels: this occurs elsewhere in the records (rachineburgius) and is even today a sign of all the Franconian dialects from the Palatinate to the North Sea. We shall return to these two chief characteristics of Franconian—e often for i, and ch between vowels for g—in the individual dialects.

As the result of the above investigation, which may be compared with Grimm’s statements about Old Franconian in the Geschichte der deutschen Sprache at the end of the first volume [p. 547], we may posit this thesis, which anyway is hardly disputed now: that in the 6th and 7th centuries Franconian was already a dialect of its own, forming the transition between High German, in particular Alamannic, and Ingaevonian, in particular Saxon and Frisian, and at that time still completely at the Gothic-Low German stage of shifting. But once this has been conceded it has also been acknowledged that the Franks were not a mish-mash of different peoples allied by external circumstances, but a main German people in their own right, the Iscaevonians, who probably absorbed foreign constituents at different times but also had the strength to assimilate them. Similarly we may regard it as proven that each of the main branches of the Franconian people already spoke a peculiar dialect at an early stage, that the language divided into Salic and Ripuarian and that many distinguishing peculiarities of the old dialects still live on in the present-day vernacular.----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Let us now move on to these still living dialects.

1. There is no longer any doubt that Salic lives on in the two Netherlands dialects, Flemish and Dutch, and at its purest in the areas that have been Frankish ever since the 6th century. For after the great tidal waves of the 12th, 13th and 14th centuries had wiped out almost all Zeeland and formed the Zuider Zee, the Dollart and the Jade, thus breaking the geographical, and also the political, cohesion of the Frisians, the remains of old Frisian liberty succumbed to the pressure of the surrounding landed gentry, and with it, almost everywhere, the Frisian language, too. To the west it was hemmed in or wholly suppressed by Dutch, to the east and north by Saxon and Danish, in all cases leaving behind strong traces in the invading language. In the 16th and 17th centuries the old Frisian area of Zeeland and Holland became the centre and mainstay of the struggle for independence in the Netherlands, just as they were already the seat of the main trading towns of the country. Thus it was chiefly here that the modern written language of the Netherlands came into being, absorbing Frisian elements, words and word forms, which can be clearly distinguished from the Franconian foundation. On the other hand, the Saxon language advanced from the east on to formerly Frisian and Frankish territory. It must be left to detailed research to draw up the exact boundaries; purely Salic are only the Flemish-speaking parts of Belgium, North Brabant, Utrecht, along with Gelderland and Overijssel with the exception of the easterly, Saxon areas.

Between the French linguistic boundary on the Maas and the Saxon boundary north of the Rhine, the Salians and the Ripuarians clashed. We shall discuss later the matter of the demarcation line, which here too has yet to be ascertained by detailed study. But first let us consider the grammatical peculiarities of Dutch.

As for the vowels, we see at once that i is replaced by e in the true Franconian manner: brengen—bringen [bring], kreb—Krippe [crib], hemel—Himmel [sky], geweten—Gewissen [conscience], ben— bin [am], stem—Stimme [voice]. This is even more frequently the case in Middle Dutch: gewes—gewiss [certain], es—ist [is], selver— Silber [silver], blent—blind [blind], where Modern Dutch has gewis, is, zilver, blind. Similarly in the vicinity of Ghent I find two places, Destelbergen and Desteldonck, according to which Distel [thistle] is to this day Destel. Middle Dutch, raised on pure Franconian soil, is here in exact agreement with Ripuarian, while the Modern Dutch written language, having been exposed to Frisian influence, is less so.

Further, again in agreement with Ripuarian, o replaces u before m or n plus following consonant, though not so consistently as in Middle Dutch and Ripuarian. Beside konst, gonst, kond, Modern Dutch has kunst, gunst, kund [art, favour, known]; yet both agree in having mond—Mund [mouth], hond—Hund [dog], jong—jung [young], ons—uns [us].

In contrast to Ripuarian, the long i (ij) has become ei as far as pronunciation is concerned, which does not yet seem to have been the case in Middle Dutch. However, this ei is not pronounced as High German ei = ai, but really as e + i, though not quite as thin as, e.g., the ej of the Danes and Slavs. Scarcely divergent from this sound is the diphthong written not ij but ei. Corresponding to High German au we find ou, ouw.

The umlaut has disappeared from the inflexion. In the declension singular and plural have the same stem vowel, as do indicative and subjunctive in the conjugation. On the other hand, umlaut does occur in word formation in two forms: 1. in the [mutation] of a to e by i common to all post-Gothic dialects; 2. in a form peculiar to Dutch that did not develop until later. Middle Dutch and Ripuarian still both have hus—Haus [house], brun— braun [brown], rum—geräumig [roomy], tun—Zaun [fence], plural huse, brune. Modern Dutch has only the forms huis, bruin, ruim, tuin (ui =High German eu), which are alien to Middle Dutch and Ripuarian. On the other hand, eu is already displacing short o (High German u) in Middle Dutch: jeughet, beside joghet, Modern Dutch jeugd—Jugend [youth]; doghet—Tugend [virtue], dor—Tür [door], kor—Wahl [choice], alongside the forms with eu; Modern Dutch permits the forms deugd, keur, deur only. This is in perfect agreement with the eu that developed from the 12th century in Northern French for Latin stressed o. Kern draws attention to a third case: the mutated form ei from ê (ee) in Modern Dutch. All these three forms of umlaut are unknown in Ripuarian, as in the other dialects, and are a special characteristic of Dutch.

Ald, alt, old, olt, uld, ult turn into oud, out. This transition is already present in Middle Dutch, in which, however, guldin, hulde, sculde still occur alongside goudin, houde, scoude (sollte) [should], so that it is possible to establish roughly the time when it was introduced. It is also peculiar to Dutch, at least as opposed to all the other continental Germanic dialects; it does, however, exist in the Lancashire dialect of English: gowd, howd, owd for gold, hold, old.

As far as the consonants are concerned, Dutch has no pure g (the guttural Italian, French or English g). This consonant is pronounced as a strongly aspirated gh, which in certain sound combinations does not differ from the deeply guttural (Swiss, Modern Greek or Russian) ch. We have seen that this transition of g into ch was already known in Old Salic. It is also found in a part of Ripuarian and the Saxon dialects that developed on formerly Frankish soil, e.g. in Münsterland, where, as in Bergish, even initial j , especially in foreign words, on occasion sounds like ch, and it is possible to hear Choseph and even Chahr (Jahr) [year]. If M. Heyne had taken this into account, he might have spared himself his difficulty with the frequent confusion and mutual alliteration of j , g and ch in the Heliand.

In some cases Dutch retains the initial wr: wringen—ringen [ring], wreed—cruel, harsh, wreken—rächen [avenge]. There is also a remnant of this in Ripuarian.

The softening of the diminutive -ken to -tje, -je is derived from Frisian: mannet je—Männchen [little man], biet je — Bienchen [little bee], halsje — Hälschen [little throat], etc. But k is also retained: vrouken — Frauchen [little woman], hoedeken — Hütchen [little hat]. Flemish better preserves the k, at least in the vernacular; the famous little man in Brussels is called Manneken-Pis. The French have thus borrowed their mannequin, and the English their manikin, from Flemish. The plural of both endings is -5: vroukens, mannetjes. We shall come across this -s again in Ripuarian.

In common with Saxon and even Scandinavian dialects, Dutch shows the loss of d between vowels, especially betwen two e’s: leder and leer, weder and weer, neder and neer, vader and vaer, moeder and moer—Mutter [mother].

The Dutch declension shows a complete mixture of strong and weak forms, so that, as the plural umlaut is also lacking, the Dutch plural forms only in the rarest cases agree with even the Ripuarian or Saxon ones, and this, too, is a very tangible characteristic of the language.

Common to Salic and Ripuarian and all the Ingaevonian dialects is the loss of the nominative indicator in er, der, wer [he, the, who]: Dutch hij, de (article) and die (demonstrative pronoun), wie.

To go into the conjugation would take us too far. What has been said here will suffice to distinguish the present-day Salic language everywhere from the neighbouring dialects. Closer examination of the Dutch dialects is bound to bring to light much of importance.

II. Rhenish Franconian. With this term I denote all the remaining Franconian dialects. I do not place Salic in opposition to Ripuarian in the old manner, and there is a very good reason for this.

Even Arnold has drawn attention to the fact that the Ripuarians in the proper sense occupied a relatively limited area, the southern boundary of which is more or less marked by the two places Reifferscheid near Adenau and near Schleiden. This is correct in so far as in this way the purely Ripuarian territory is demarcated linguistically too from the territories occupied by genuine Ripuarians after, or at the same time as, other German tribes. Since the name Low Franconian has already acquired another meaning which also includes Salic, I am left only with the term Ripuarian—in the narrower sense—to denote the group of closely related dialects which extend from the Salic linguistic boundary up to this line.

1. Ripuarian. The dividing line between this group of dialects and the Salic by no means coincides with the Dutch-German border. On the contrary, the major part of the district Rees, where in the area of Wesel Salic, Ripuarian and Saxon meet, still belongs to Salic on the right bank of the Rhine. On the left bank the areas of Kleve and Geldern are Salic, roughly as far as a line drawn from the Rhine between Xanten and Wesel, south of the village of Vluyn (west of Mors) and from there south-west towards Venlo. A more exact definition of the boundary is only possible on the spot since many Ripuarian names have been preserved on the maps in Salic-Dutch form as the result of many years of Dutch administration not only in Geldern but also in the county of Mors.

From the area of Venlo upwards the greater part of the right bank of the Maas seems to be Ripuarian, so that here the political border nowhere crosses Salic territory but only Ripuarian and this extends almost as far as Maastricht. Names in -heim (not -hem) and the specifically Ripuarian ending -ich occur here in great numbers on Dutch territory, further south already names in -broich (Dutch -broek), e.g. Dallenbroich near Roermond; likewise in -rade (Bingelrade near Sittard, plus Amstenrade, Hobbelrade and 6 or 7 others); the little piece of German territory that has fallen to Belgium to the right of the Maas, is entirely Ripuarian (cf. Kriitzenberg, 9 kilometres from the Maas, with Kruisberg, north of Venlo). Indeed, left of the Maas, in the Belgian so-called Limburg I find Kessenich near Maaseyk, Stockheim and Reekheim on the Maas, Gellik near Maastricht as proof that no purely Salic population lives here.

The Ripuarian border with Saxony starts from the area of Wesel, running south-east at an increasing distance from the Rhine, between Mülheim on the Ruhr and Werden on the Franconian side and Essen on the Saxon side, to the border between Berg and Mark, here even now the border between the Rhine Province and Westphalia. It does not leave this border until south of Olpe, where it proceeds eastwards, dividing the Siegerland as Franconian from the Saxon Sauerland. Further east, the Hessian dialect soon takes over.

The above-mentioned southern border with the dialect which I term Middle Franconian is in rough agreement with the southern borders of the old districts of Avalgau, Bonngau and Eiflia, and from there runs westwards to Wallonia, keeping rather to the south. This area thus circumscribed includes the big old district of Ripuaria as well as parts of the districts adjoining it to the north and west.

As already stated, Ripuarian agrees in many respects with Dutch, but in such a way that Middle Dutch is closer to it than Modern Dutch. Ripuarian agrees with Modern Dutch in its pronunciation of ei = e+i and ou for au, the transition of i to e, which goes much further in Ripuarian and Middle Dutch than in Modern Dutch: the Middle Dutch gewes, es, blend, selver (silver) are still good Ripuarian to this day. Similarly, and consistently so, u changes into o before m or n with a following consonant: jong, lomp, domm, konst. If this following consonant is a d or a t, this changes to g or k in some dialects; e.g. honk—Hund [dog], plural höng, where the softening to g is an aftereffect of the loss of the final vowel, e.

However, the situation as regards umlaut in Ripuarian is very different from that of Dutch; it is in general agreement with High German, and in odd exceptions with Saxon (e.g. hanen for Hähne [cocks]).

Initial wr has become hardened to fr, retained in fringen—to wring water out of a cloth, etc., and frêd (Dutch wreed) with the meaning hardy, weather-beaten.

For er, der, wer it has hê, de, wê.

The declension is midway between High German and Saxon. Plural forms in -5 are common, but are hardly ever in agreement with the Dutch; this -5 becomes -r in local High German in correct memory of the linguistic development.

The diminutive -ken, -chen is changed to -sehen after n: männschen; the plural has -5 as in Dutch (männsches). Both forms extend all the way into Lorraine.

r is lost before s, st, d, t, z, the preceding vowel remaining short in some dialects, being lengthened in others. Thus hart [hard] becomes halt (Bergish), haad (Cologne). In the process st becomes seht through Upper German influence: Durst [thirst] — doascht (Bergish), dôscht (Cologne).

Similarly, initial si, sw, st, sp have become schl, etc., through High German influence.

As in Dutch, pure g is unknown in Ripuarian. Some of the dialects on the Salic border, as well as Bergish, have aspirated gh for initial and medial g, though softer than in Dutch. Thé rest have j . Final g is everywhere pronounced as ch, though not like the hard Dutch sound, but like the soft Rhenish Franconian ch, which sounds like a hardened j. The essentially Low German character of Ripuarian is attested by terms such as boven for oben [above].

The majority of the voiceless consonants are everywhere at the first stage of the sound shift. Only t and medial and final k, occasionally p, show the High German sound shift in the southern dialects: they have lôsze for löten—lassen [let], holz for holt [wood], rich for rik—reich [rich], êch for ek—ich [I], pief for pipe—Pfeife [pipe]. But et, dat, wat and a few others are retained.

It is this not even consistently carried out intrusion of the High German sound shift in three cases on which the usual demarcation of Middle and Low Franconian is based. But in this way a group of dialects that belong together on. account of definite features in the sound system, as demonstrated, which are still recognised in the popular mind as belonging together, are torn apart arbitrarily and on the basis of a characteristic that is here quite fortuitous.

Quite fortuitous, I say. Each of the other Central German dialects, Hessian, Thuringian, Upper Saxon, etc., is generally speaking at a specific stage of the High German sound shift. They may show rather less shifting on the Low Saxon border and rather more on the Upper German border, but that is at most only enough to justify local differences. Franconian, on the other hand, shows no shifting at all on the North Sea, Maas and Lower Rhine, on the Alamannic border almost entirely Alamannic shifting; in between there are at least three intermediate stages. The shift thus penetrated into Rhenish Franconian when it had already developed independently, splitting it up into several pieces. The last trace of this shift need not by any means vanish on the border of a particular group of dialects that was already in existence; it may die out in the midst of such a group, as it in fact does. On the other hand, the truly dialect-forming influence of the shift, as we shall see, does indeed cease on the border of two dialect groups that were already different beforehand. And did not the schl, schw, etc., and the final seht come to us from High German in a similar way and at an even later date? These however—at least the first—even go deep into Westphalia.

The Ripuarian dialects formed a fixed group long before some of them learnt to shift t and medial and final k and p. How far this change was able to advance within the group was and remains for the group a matter of chance. The dialect of Neuss is identical with that of Krefeld and München-Gladbach—apart from minor differences that a stranger cannot hear at all. Nevertheless, one is supposed to be Middle Franconian, the other Low Franconian. The dialect of the Berg industrial country merges into that of the south-west Rhine plain in imperceptible stages. And yet they are supposed to belong to two totally different groups. For anyone who is at home in the region it is obvious that book-learning is here forcing the living dialects, with which it is scarcely acquainted if at all, into the Procrustean bed of characteristics constructed a priori.

As a result of this purely superficial distinction the southern Ripuarian dialects are lumped together into a so-called Middle Franconian with other dialects from which they diverge, as we shall see, far more than they do from the so-called Low Franconian. Owing to the same superficial distinction, a narrow strip is held back because you are at a loss what to do with it and are finally obliged to declare one part Saxon and another Dutch, which is in glaring contradiction to the state of affairs in these dialects.

Let us take, for instance, the Bergish dialect, which Braune without much ado calls Saxon. It forms, as we have seen, all three persons plural of the present indicative in the same way, but as in Franconian, with the ancient form -nt. It regularly has o instead of u before m and n followed by a consonant, which according to the same Braune is definitely un-Saxon and specifically Low Franconian. It agrees with the other Ripuarian dialects in all the characteristics set out above. While it imperceptibly merges into the dialect of the Rhine plain from village to village, from farm to farm, it is most sharply separated from the Saxon dialect on the Westphalian border. Perhaps nowhere else in all Germany is there such an abruptly drawn linguistic border as here. And what a distance between the languages! The whole vowel system seems to be turned upside down; the sharp Low Franconian ei contrasts abruptly with the broadest ai, just as ou contrasts with au; not one of the many diphthongs and vocalic glides is in agreement; here sch as in the rest of Germany, there s-ch as in Holland; here wi hant, there wi hebbed; here the dual forms get and enk used as the plural (German ihr and euch), there only i, ji, and ü, jü; here the sparrow is called common Ripuarian Masche, there common Westphalian Lüning. Not to mention other peculiarities specific to the Bergish dialect which also suddenly vanish here on the border.

The individuality of a dialect is most apparent to the stranger if the person in question is not speaking dialect but High German, which is more intelligible to the stranger, and which in the case of most Germans is, of course, strongly coloured by their respective dialect. But then the allegedly Saxon inhabitant of the Berg industrial district is for the non-native quite indistinguishable from the inhabitant of the Rhine plain, who is supposed to be Middle Franconian, except for the somewhat more harshly aspirated gh, where the other says j. A man from Heckinghaus in Berg (from Oberbarmen, left of the Wupper), however, and a man from Langerfeld in Mark, who lives scarcely a kilometre further east, are further apart in the local High German of everyday life than the man from Heckinghaus and one from Coblenz, let alone anyone from Aachen or Bonn.

The advance of the shift of t and final k makes such a small impression on the Rhenish Frank himself as a linguistic boundary that even in an area well known to him he will first have to reflect where the border runs between t and z, k and ch, and that, when crossing this border, he finds that one comes almost as naturally to him as the other. This is made even easier by the many High German words with shifted 5z, z, ch and / that have entered the dialects. A striking example is afforded by the old Bergish penal code from the 14th century (Lacomblet, Archiv, I, p. 79 ff.). There we find zo, uiss (aus), zween, bezahlen; alongside them in the same sentence: selten, dat nutteste (nutzeste); likewise Dache, redelich beside reichet (reicht); Upiaden, upheven, hulper (Helfer) beside verkouffen. In another paragraph p. 85 it has alternately zo and tho—zu. In short, the dialects of the mountain and the plain are continually getting mixed up without this disturbing the scribe in the slightest. As usual, this final wave with which the High German sound shift washed over Frankish territory was also the weakest and shallowest. It is surely of interest to mark out the line showing how far this wave extends. But this line cannot be a dialect boundary; it is not able to tear apart an independent group of anciently and closely related dialects and provide the pretext for allocating the fragments thus violently divided to more distant groups in contradiction with all linguistic facts.

2. Middle Franconian. From the above it is quite obvious that I place the northern border of Middle Franconian much further to the south than is customary.

From the fact that the Middle Franconian region on the left bank of the Rhine seems to have been in the possession of the Alamanni at the time of Clovis, Arnold finds reason to investigate the place-names there for traces of Alamannic settlement, and comes to the result that it is possible to prove the existence of a pre-Frankish, Alamannic population as far as the line CologneAachen; the traces, most numerous in the south, naturally becoming rarer and rarer to the north. The place-names, so he says, point to

“a temporary advance by the Alamanni as far as and beyond the area around Coblenz and Aachen, and also a longer occupation of the Wetterau and the southern areas of Nassau. For the names with the genuine Alamannic endings -ach, -brunn, -felden, -hofen, -ingen, -schwand, -stetten, -wangen, and -weiler, which nowhere occur in purely Frankish territory, are found scattered from Alsace onwards over the entire Palatinate, Rhenish Hesse and Rhenish Prussia, only they become rarer to the north, giving way more and more to the Franconian names par excellence in -bach, -berg, -dorf, -born, -feld, -hausen, -heim, and -scheid” (Deutsche Urzeit, p. [140]).

Let us first examine the allegedly Alamannic names of the Middle Franconian country. I have not found the endings -brunn, -stetten, -felden, -wangen anywhere on the Reymann map (which I am using here, let it be said once and for all). The ending -schwand occurs once: Metzelschwander Hof near Winn weiler, and then again Schwanden north of Landstuhl. Thus both in the Upper Franconian Palatinate, with which we are not concerned here. In -ach we have along the Rhine Kreuznach, Bacharach, Hirzenach near St. Goar, Rübenach near Coblenz (Ribiniacus of the SprunerMenke District Map), Andernach (Antunnacum of the Romans), as well as Wassenach. Now, as the Romanised Celtic ending -acum occurs generally the whole length of the left bank of the Rhine in Roman times—Tolbiacum (Zülpich), Juliacum (Jülich), Tiberiacum (Ziewerich) near Bergheim, Mederiacum—in the majority of these cases the choice of the form -ach for -ich, at most, might betray Alamannic influence. Only one, Hirzenach (=Hirschenbach), is definitely German, and this was formerly called Hirzenowe, Hirschenau, not Hirschenbach, according to the district map. But how then do we explain Wallach, between Büderich and Rheinberg, close by the Salic border? At any rate it is certainly not Alamannic.

In the Mosel region there are also a few -ach: Irmenach east of Bernkastel, Waltrach, Crettenach near Trier, Mettlach on the Saar. In Luxemburg Echternach, Medernach, Kanach; in Lorraine on the right of the Mosel only: Montenach, Rodelach, Brettnach. Even if we wished to concede that these names indicate an Alamannic settlement, then it is only a thinly scattered one, which, moreover, does not extend beyond the southernmost part of the Middle Franconian territory.

There remain -weiler, -hofen, and -ingen which require closer examination.

Firstly, the ending -weiler is not properly speaking Alamannic but the provincial Latin villarium, villare, and is found only very exceptionally outside the old frontiers of the Roman Empire. The Germanisation of villare to weiler was not the privilege of the Alamanni, but they had a predilection for using this ending also for new settlements in large numbers. In so far as Roman villaria occurred, the Franks too were obliged to take over the ending, Germanising it as wilare, later weiler, or drop it altogether. Probably they did now one, now the other, just as they certainly gave new settlements names in -weiler here and there, but far more rarely than the Alamanni. Arnold cannot find any important places in -weiler north of Eschweiler near Aachen and Ahrweiler. But the present importance of the place has nothing to do with it; the fact of the matter is that on the left bank of the Rhine the names in -weiler extend almost as far as the Salic border to the north (Garzweiler and Holzweiler are less than five miles from the nearest Dutch-speaking place of the Geldern area) and north of the line Eschweiler-Ahrweiler there are at least twenty of them. They are, understandably, commonest in the vicinity of the old Roman road from Maastricht via Jülich to Cologne, two of them, Walwiller and Nyswiller, even being on Dutch territory; are these Alamannic settlements too?

Further south they hardly occur in the Eifel at all; the Malmedy section (Reymann, No. 159) has not one single case. In Luxemburg, too, they are rare, as on the lower Mosel and as far as the crest of the Hunsrück. Yet they frequently occur on the upper Mosel on both sides of the river, becoming increasingly common towards the east, becoming more and more the dominating ending to the east of Saarlouis. But this is where the Upper Franconian language begins, and here it is not disputed by anyone that the Alamanni had occupied the country before the Franks.

Thus for the Middle Franconian and Ripuarian area the -weiler do not indicate Alamannic settlement any more than do the many -villers in France.

Let us move on to -hofen. This ending is still less exclusively Alamannic. It occurs throughout the Franconian area, including present-day Westphalia, which was later occupied by the Saxons. On the right bank of the Rhine just a few examples: Wehofen near Ruhrort, Mellinghofen and Eppinghofen near Duisburg, Benninghofen near Mettmann, another Eppinghofen near Dinslaken, in Westphalia Kellinghofen near Dorsten, Westhofen near Castrop, Wellinghofen, Wichlinghofen, Niederhofen, two Benninghofens, Berghofen, Westhofen, Wandhofen, all on the Hellweg, etc. Ereshofen on the Agger, Martis villa, reaches back into pagan times, and the very designation of the god of war as Eru proves that no Alamanni are conceivable here: they called themselves Tiuwâri, thus calling the god not Eru but Tin, later shifted to Ziu.

On the left bank of the Rhine it is even more difficult to demonstrate the Alamannic derivation of -hofen. There is another Eppinghofen south-east of Xanten, hence possibly Salic already, and from there on to the south the whole Ripuarian area is teeming with -hofen, alongside -hof for single farms. But if we proceed to Salic country, it gets even worse. The Maas is accompanied by -hofen on either side, from the French linguistic boundary onwards. For the sake of brevity let us pass to the west bank straight away. In Holland and Belgium we find at least seven Ophovens, in Holland Kinckhoven, etc; for Belgium let us first turn to the section for Löwen (Reymann, No. 139). Here we find Ruykhoven, Schalkhoven, Bommershoven, Wintershoven, Mettecoven, Helshoven, Engelmannshoven near Tongern; Zonhoven, Reekhoven, Konings-Hoven near Hasselt, further west Bogenhoven, Schuerhoven, Nieuwenhoven, Gippershoven, Baulershoven near St. Truyen; most westerlv Gussenhoven and Droenhoven east and north-east Tirlemont (Thienen). The section for Turnhout (No. 120) has at least 33 -hoven, most of them on Belgian territory. Further to the south-west the -hove (the dative -n is regularly dropped here) skirt the entire French linguistic border: from Heerlinkhove and Nieuwenhove near Ninove, which is itself a Romanised -hove,— omitting the intermediate ones, about ten in number—to Ghyverinckhove and Pollinchove near Dixmuyden and Volckerinckhove near St. Omer in French Flanders. Nieuwenhove occurs three times, which proves that the ending is still living among the people. In addition a great number of single farms in -hof On this basis the supposedly exclusively Alamannic character of -hofen may be judged.

Finally to -ingen. The designation of common descent with -ing, -ung, is common to all the Germanic peoples. Since settlement took place by kin, the ending plays an important part in place-names everywhere. Sometimes it is linked, in the genitive plural, with a local ending: Wolvarad-inga-husun near Minden, Snotingaham (Nottingham) in England. Sometimes the plural alone stands for the designation of place: Flissingha (Vlissingen), Phladirtinga (Vlaardingen), Crastlingi in Dutch Frisia; Grupilinga, Britlinga, Otlinga in Old Saxony. These names have mostly been reduced to the dative nowadays, ending in -ingen, rarely in -ing. Most peoples know and employ both forms; the Alamanni, so it seems, chiefly the latter, at any rate now. Since, however, this also occurs among the Franks, Saxons and Frisians, it is very audacious to immediately deduce Alamannic settlement from the occurrence of placenames in -ingen.

The above mentioned names prove that names in -ingas (nominative plural) and -ingum, -ingon (dative plural) were nothing unusual either among the Frisians or among the Saxons, from the Scheide to the Elbe. Even today the -ingen are no rarity throughout Lower Saxony. In Westphalia on either side of the Ruhr, south of the line Unna-Soest, there are at least twelve -ingen, alongside -ingsen and -inghausen. And as far as Franconian territory extends, we find names in -ingen.

On the right bank of the Rhine we first find in Holland Wageningen on the Rhine and Genderingen on the Ijssel (and we exclude all possibly Frisian names), in the Berg country Huckingen, Ratingen, Ehingen (close behind them on Saxon territory Hattingen, Sodingen, Ummingen), Heisingen near Werden (which Grimm derives from the Silva Caesia of Tacitus and which would thus be very ancient), Solingen, Hnsingen, Leichlingen (on the district map Leigelingon, thus almost a thousand years old), Quettingen and on the Sieg Büdingen and Röcklingen, not counting two names in -ing. Henningen near Rheinbrohl and Ellingen in the Wied area provide the link with the area between Rhine, Lahn and Dill, which at a low estimate counts 12 -ingen. It is pointless to go any further south, since here begins the country that indisputably passed through a period of Alamannic settlement.

Left of the Rhine we have Millingen in Holland above Nimwegen, Lüttingen below Xanten, another Millingen below Rheinberg, then Kippingen, Rödingen, Höningen, Worringen, Fühlingen, all further north than Cologne, Wesselingen and Köttingen near Brühl. From here the names in -ingen follow two directions. In the High Eifel they are rare; we find near Malmedy on the French linguistic border: Büllingen, Hünningen, Mürringen, Iveldingen, Eibertingen as a transition to the very numerous -ingen in Luxemburg and on the Prussian and Lotharingian upper Mosel. Another connecting line follows the Rhine and the side valleys (in the Ahr area 7 or 8) and finally the Mosel valley, likewise after the area above Trier, where the -ingen predominate, but cut off from the great mass of Alamannic-Swabian -ingen first by the -weiler and then by the -heim. So if we, according to Arnold’s demand, “consider all the facts in their context”, we shall come to the conclusion that the -ingen of the upper German Mosel area are Franconian and not Alamannic.

How little we need Alamannic help here becomes even clearer as soon as we trace the -ingen from the French-Ripuarian linguistic border near Aachen on to Salic territory. Near Maaseyk west of the Maas lies Geystingen, further west near Brée Gerdingen. Then we find, turning back to section No. 139, Löwen: Mopertingen, Vlytingen, Rixingen, Aerdelingen, Grimmersingen, Gravelingen, Ordange (for Ordingen), Bevingen, Hatingen, Buvingen, Hundelingen, Bovelingen, Curange, Raepertingen, Boswinningen, Wimmertingen, and others, in the area of Tongern, St. Truyen and Hasselt. The most westerly, not far from Löwen, are Willebringen, Redingen, Grinningen. Here the connection seems to break off. But if we move on to territory that is now French-speaking but from the 6th to the 9th century was in dispute between the two languages, we find from the Maas onwards an entire belt of French -ange, a form which corresponds to -ingen in Lorraine and Luxemburg too, stretching from east to west: Ballenge, Roclenge, Ortrange, Lantremange, Roclange, Libertange, Noderange, Herdange, Oderinge, Odange, Gobertang, Wahenges; slightly further west Louvrenge near Wavre and Revelinge near Waterloo form the link with Huysinghen and Buisinghen, the outpost of a group of over 20 -inghen, stretching south-west of Brussels from Hal to Grammont along the linguistic boundary. And finally in French Flanders: Gravelingen, Wulverdinghe (thus exactly the Old Saxon Wolvaradinges-hûsun), Leubringhen, Leulinghen, Bonninghen, Peuplingue, Hardinghen, Her melinghen, near St. Omer and as far as behind Boulogne Herbinghen, Hocquinghen, Velinghen, Lottinghen, Ardinghen, all sharply distinguished from the even more numerous names in -inghem (= -ingheim) in the same area.

Thus the three endings which Arnold regards as typically Alamannic turn out to be every bit as much Franconian, and the attempt to prove an Alamannic settlement on Middle Franconian territory before the Franconian one on the basis of these names must be considered to have failed. While the possibility of a not very strong Alamannic element in the south-eastern part of this territory can still be conceded.

From the Alamanni, Arnold leads us to the Chatti. With the exception of the Ripuarians proper, they are supposed to have occupied the area south of the Ripuaria district, the same one, in other words, as we call Middle and Upper Franconia, after and alongside the Alamanni. This too is substantiated by references to the Hessian place-names found in the area beside the Alamannic ones.

“The agreement in the place-names on this and the other side of the Rhine as far as the Alamannic border is so peculiar and so striking that it would be a miracle indeed if it were coincidental; on the other hand, it seems quite natural as soon as we assume that the immigrants gave their native place-names to their new domiciles, as still occurs in America all the time.”

There is little to object to in this sentence. But all the more to object to in the conclusion that the Ripuarians proper had nothing to do with the settlement of the whole Middle and Upper Franconian country, that we only find Alamanni and Chatti here. Most of the Chatti who left their home for the west seem to have joined the Iscaevones from time immemorial (as did the Batavi, Canninefates and Chattuari); and where else should they turn? In the first two centuries A. D. the Chatti were only linked with the other Herminones in the rear through the Thuringians; on the one side they had the Ingaevonian Cherusci, on the other the Iscaevones, and before them the Romans. The Herminonian tribes, which later appear united as Alamanni, came from the heart of Germania, having been separated from the Chatti for centuries by Thuringians and other peoples and having become more alien to them than the Iscaevonian Franks, with whom they were allied by a centuries-old brotherhood in arms. The Chatti’s participation in the occupation of the area in question is thus not doubted. But the exclusion of the Ripuarians is. This can only be proved if no specifically Ripuarian names occur there. The situation is quite the reverse.

Of the endings stated by Arnold to be specifically Franconian, -hausen is common to Franks, Saxons, Hessians and Thuringians; -heim is Salic -ham; -bach Salic and Lower Ripuarian -beek; of the others, only -scheid is really characteristic. It is specifically Ripuarian, just like -ich, -rath or -rade and -siepen. Further, common to both Franconian dialects are -loo (-loh), -donk and -bruch or -broich (Salic broek).

-scheid occurs only in the mountains and, as a rule, in places on the watershed. The Franks left this ending behind throughout Westphalian Sauerland as far as the Hessian border, where it occurs, only as mountain names, as far as eastern Korbach. On the Ruhr Old Franconian -scheid encounters the ending in its Saxon form, -Schede: Melschede, Selschede, Meschede; in the near vicinity, Langscheid, Ramscheid, Bremscheid. Frequent in the Berg area, it is found as far as the Westerwald and into it, but not further south, on the right side of the Rhine. Left of the Rhine, however, the -scheid understandably do not commence until the Eifel; in Luxemburg there are at least 21 of them, in the Hochwald and Hunsrück they are common. But as south of the Lahn, here too, on the eastern and southern sides of the Hunsrück and Soonwald, they are joined by the form -schied, which seems to be a Hessian adaption. Both forms together move southwards across the Nahe as far as the Vosges, where we find: Bisterscheid west of Donnersberg, Langenscheid near Kaiserslautern, a plateau called Breitscheid south of Hochspeyer, Haspelscheid near Bitsch, the Scheidwald north of Lützelstein, and finally as the southernmost outpost Walscheid on the north slope of the Donon, even further south than the village of Hessen near Saarburg, the most advanced Chattic outpost in Arnold.

Also specifically Ripuarian is -ich, from the same root, Gothic -ahva (water), as -ach; both are also German forms of the Belgian-Roman -acum, as proved by Tiberiacum, on the district map Civiraha, today Ziewerich. It is not very frequent on the right side of the Rhine; Meiderich and Lirich near Ruhrort are the most northerly, from where they skirt the Rhine as far as Biebrich. The plain on the left of the Rhine, from Büderich opposite Wesel onwards, is full of them; they cross the Eifel as far as the Hochwald and Hunsrück, but vanish in the Soonwald and the region of the Nahe, even before -scheid and -roth stop. In the western part of our territory, however, they continue to the French linguistic border and beyond. The Trier area, which has a lot of them, we shall pass over; in Dutch Luxemburg I count twelve, on the other side, in the Belgian part, Törnich and Merzig (Messancy—the spelling -ig makes no difference, etymology and pronunciation are the same), in Lorraine, Soetrich, Sentzich, Marspich, Daspich west of the Mosel; east of it Kuntzich, Penserich, Cemplich, Destrich, twice Kerprich, Hibrich, Hilsprich.

The ending -rade, -rad, on the left bank of the Rhine -rath, also considerably exceeds the bounds of its old Ripuarian homeland. It fills the whole Eifel and the middle and lower Mosel valley, as well as its side valleys. In the same area where -scheid mixes with -schied, -rod, -roth occurs alongside -rad and -rath on both banks of the Rhine, also of Hessian origin, except that on the right bank, in the Westerwald, the -rod extend further north. In the Hochwald the northern slope has -rath, the southern slope -roth, as a rule.

The least advanced is -siepen, shifted -seifen. The word means a small stream-valley with a steep fall and is still in general use with this meaning. Left of the Rhine it does not extend far beyond the old Ripuarian border; on the right it is found in the Westerwald on the Nister and even near Langenschwalbach {Langenseifen).

To examine the other endings would take us too far. But at any rate we may assert that the countless -heim, which accompany the Rhine upstream from Bingen deep into Alamannic territory and are found everywhere where the Franks settled, are not Chattic but Ripuarian. Their home is not in Hesse, where they rarely occur and seem to have entered later, but in the Salic country and the Rhine plain around Cologne, where they occur alongside the other specifically Ripuarian names in almost equal numbers.

Thus the result of this investigation is that the Ripuarians, far from being held back by the stream of Hessian immigration at the Westerwald and Eifel, on the contrary overran the entire Middle Franconian area themselves. And more strongly in a southwesterly direction, towards the upper Mosel area, than to the south-east towards the Taunus and the area of the Nahe. This is also corroborated by the language. The south-western dialects, right into Luxemburg and western Lorraine, are much closer to Ripuarian than the eastern ones, particularly those on the right bank of the Rhine. The former might be regarded as a more High German shifted extension of Ripuarian.

The characteristic thing about the Middle Franconian dialects is firstly the penetration of the High German sound shift. Not the mere shift of a few tenues to aspirates, applying to relatively few words and not affecting the character of the dialect, but the beginning shift of the voiced-stopped consonants, which brings about the peculiarly Middle and Upper German confusion of b and p, g and k, d and t. Only where the impossibility of making a sharp distinction between initial b and p, d and t, g and k appears, in other words what the French particularly mean by accent allemand—only then does the Low German feel the great cleft which the second sound shift has torn through the German language. And this cleft runs in between the Sieg and the Lahn, the Ahr and the Mosel. Accordingly, Middle Franconian has an initial g which is lacking in more northern dialects, whereas medially and finally it still pronounces a soft ch for g. Furthermore, the ei and ou of the northern dialects turn into ai and au. A few genuinely Franconian peculiarities: in all the Salic and Ripuarian dialects Bach, unshifted Beek, is feminine. This is also true at least of the largest, western part of Middle Franconian. Like the numerous other -backs with the same name in the Netherlands and on the lower Rhine, the Luxemburg Glabach {Gladbach, Dutch Glabeek) is also feminine. On the other hand, girls’ names are treated as neuter: it is not only das Mädchen, das Mariechen, das Lisbethchen, but also das Marie, das Lisbeth, from Barmen to Trier and beyond. Near Forbach in Lorraine the map, originally made by the French, shows a “ Karninschesberg” (Kaninchenberg). Thus the same diminutive -sehen, plural -sches, which we found above to be Ripuarian.

With the watershed between Mosel and Nahe and on the right bank of the Rhine with the hill-country south of the Lahn, a new group of dialects begins:

3. Upper Franconian. Here we are in a region which was indisputably first Alamannic territory by conquest (disregarding the earlier occupation by Vangiones, etc., of whose tribal affinities and language we know nothing) and where a fairly strong Chattic admixture can be readily conceded. But here too the place-names, as we need not repeat, indicate the presence of not insignificant Ripuarian elements, especially in the Rhine plain. And the language even more so. Let us take the southernmost definable dialect which at the same time has a literature, that of the Palatinate. Here we again encounter the general Franconian inability to pronounce medial and final g in any other way but as a soft eh. They say there: Vbchel, Flechel, geleche (gelegen) [lain], gsacht—gesagt, licht—liegt, etc. Similarly the general Franconian w instead of b in the medial position: Bûwe—Buben, glâwe—glauben (but i glâb), bleiwe, selwer—selbst, halwe—halbe. The shift is far from being as complete as it looks; there is even reverse shifting, particularly in foreign words, i.e. the initial voiceless consonant is shifted not one stage forwards, but backwards: t becomes d, p becomes b, as will be seen; initial d and p remain at the Low German stage: dun—tun, dag, dame, dür, dodt; but before r: trinke, trage; paff—Pfaff, peife, pah—Pfalz, parre—Pfarrer. Now as d and p stand for High German t and pf, initial t is shifted back to d, and initial p to b, even in foreign words: derke—Türke, dafel—Tafel, babeer—Papier, borzlan—Porzellan, bulwer—Pulver. Then the Palatinate dialect, agreeing only with Danish on this score, cannot tolerate any tenues between vowels: ebbes—etwas, labbe—Lappen, schlubbe—schlüpfen, schobbe—Schoppen, Peder— Peter, dridde—dritte, rodhe—raten. The only exception is k: brocke, backe. But in foreign words g: musigande—Musikanten. This is also a relic of the Low German stage of the sound system which has spread out further by means of reverse shifting; only because dridde, hadde remained unshifted could Peter become Peder and the corresponding High German t receive the same impartial treatment. Similarly, the d in halde—halten, aide—alte, etc., remains at the Low German stage.

Despite the decidedly High German impression it makes on Low Germans, the dialect of the Palatinate is far from having adopted the High German sound shift even to the extent that our written language has preserved it. On the contrary, by means of its reverse shift the Palatinate dialect is protesting against the High German stage, which, having entered from without, proves to be a foreign element in the dialect to this day.

This is the place to look at a feature that is usually misunderstood: the confusion between d and t, b and p and even g and k among those Germans in whose dialects the voicedstopped consonants have undergone the High German sound shift. This confusion does not arise as long as everyone speaks his own dialect. On the contrary. We have just seen that the native of the Palatinate, for example, makes a very nice distinction here, so much so that he even shifts back foreign words in order to adapt them to the requirements of his dialect. The foreign initial t only becomes d for him because written German t corresponds to his d, foreign p only becomes b because his p corresponds to written German pf. Nor do the voiceless consonants get mixed up in the other Upper German dialects as long as people speak dialect. Each of these dialects has its own, precisely applied sound-shift law. But the position is different as soon as the written language or a foreign language is spoken. The attempt to apply to it the shifting law of the dialect concerned—and this attempt is made involuntarily—collides with the attempt to speak the new language correctly. In the process the written b and p, d and t lose all fixed meaning, and thus it is that Börne, for instance, in his letters from Paris complains that the French were unable to distinguish between b and p, because they obstinately insisted that his name, which he pronounced Feme, commenced with a p.

But back to the Palatinate dialect. The evidence that the High German sound shift was foisted on it from without, so to speak, and has remained a foreign element to this day, not even reaching the sound-system stage of the written language either (far exceeding which the Alamanni and the Bavarians on the whole preserve one Old High German stage or another)—this proof alone suffices to establish the predominantly Franconian character of the Palatinate dialect. For even in Hesse, which is much further north, the shift has, on the whole, been carried further, thus reducing the allegedly chiefly Hessian character of the Palatinate dialect to modest proportions. In order to offer such resistance to the High German sound shift hard by the Alamannic border among the Alamanni that remained behind, there must have been at least as many Ripuarians alongside the Hessians, who were themselves essentially High Germans. And their presence is further proved—apart from the place-names—by two generally Franconian peculiarities: the preservation of the Franconian w instead of b medially, and the pronunciation of g as ch in medial and final positions. To this may be added a lot of individual cases of agreement. With the Palatinate Gundach—”guten Tag”—you will get by as far as to Dunkirk and Amsterdam. Just as “a certain man” is ein sichrer Mann in the Palatinate, in the entire Netherlands it is een zekeren man. Handsching for Handschuh [glove] corresponds to the Ripuarian Händschen. Even g for j in Ghannisnacht (Johannisnacht [midsummer night]) is Ripuarian and extends, as we have seen, into the Münster area. And baten (to improve, be of use, from bat—-better), common to all the Franks, and the Netherlanders too, is in current use in the Palatinate: '5 badd alles nix—it's all no use—where the t is not even shifted to High German tz but is softened to d between vowels in the Palatinate manner.