We may turn from the excavation of a few mottes to consideration of the distribution (by class as well as geographically) of all those identified in the field within the various lordships. Leinster serves well as an entry into this discussion because in 1247 the lordship was surveyed in detail so that it could be divided among the five heiresses of the last Marshal lord (Brooks, 1950). The survey listed all the demesne lands farmed directly by the lord, and also each parcel of land held by military tenure, for the service of one knight, of more than one, or else for fractions of a knight. We have therefore a list of all the main estate centres in the lordship within three generations of its first organisation; it is possible to be quite confident of the identification of most of these place names on the modern map. If we combine this with field survey of the surviving monuments, we can begin to examine the social and geographical context of the mottes (McNeill, 1990b). We can then use Leinster as a yardstick by which to measure other lordships.
The first conclusion which emerges from such a study is that the social pattern of building mottes is not random. Of the 27 holdings, held in return for the service of half a knight, only four are now marked by mottes: none of the holdings held for less than half a knight have mottes now. On the other hand, of the 46 holdings held for a full knight’s service or more, 19 are now marked by mottes. Put proportionally, 40 per cent of the higher holdings have mottes against 15 per cent of the half-fees. It is difficult to believe that this is a difference which reflects the resources available to the two groups. Full knight’s fees were obviously larger than half-fees, but hardly so much so that the lesser tenants could not have erected mottes if they had wanted, or been encouraged, to do so. For example, if the new lords had been afraid of a rising by their Irish peasantry, is it likely that tenants of half-fees would have felt so much safer than those who held full fees? It seems odd that they would have thought the rebels would be so discriminating as to attack only one sort of tenant rather than the other. The result of this formal basis for the erection of mottes is that the lordship has a fairly even but rather sparse distribution of mottes over its area, a density which we may take as a norm for comparison for other lordships.
In the hinterland of the lordships of Meath and, less clearly, English Oriel, the same situation prevails. In the good lands of the plain of Meath (the bulk of the present county), the centres of the holdings of the major tenants are where the great majority of mottes are sited. The position is less clear partly because we have not got the sort of systematic contemporary survey that we have for Leinster. In Ulster, also, we have to try to make up such a list of major tenants’ holdings from scattered sources, which are even less complete than for Meath. However, something of the same pattern emerges for the core area of the lordship in south-east Co. Antrim and in coastal Co. Down.
In the three last-named lordships, Meath, Oriel and Ulster, there is a contrast between the settled cores and their borders to the north-west and west. Mottes proliferate along the borders, at the most extreme in the north-western borders of Meath and Oriel, where some mottes are only one or two miles from their nearest neighbour (Fig. 41). They make a dense line on the map along the strip of land where the plain gives way to the low hills and broken country of the present counties of Cavan and Monaghan. Interestingly, the line continues into the lordships of Oriel, in spite of the difference not only of lord but also of date: Meath was seized by Hugh de Lacy from the 1170s, but Oriel was not occupied by English lords until after the death of Murchad Ua Carbaill in 1189. In Ulster, the mottes are most densely distributed at the north-east corner of Lough Neagh and west of Dundrum Bay, the two points of easy entry into the lordship from the west.
The mottes in the border zones of these lordships are sufficiently densely distributed for the zones to be defined on a map. The borders concerned were not defined by lines of mottes spaced at intervals of ten or more miles, as has been proposed (Waterman, 1963, 77; Jope, 1966, 105; Graham, 1980, 52). To have an effective barrier of fortification for an Irish lordship it would have had to counter raids by small groups of mobile forces who might well enter the lordship at night. Ireland is a country with many small hills and valleys; in the twelfth century it was also well wooded. Intercepting a raiding force between bases ten miles apart would have needed constant patrolling by groups of men large enough to take on the raiding force themselves: sending five miles for help, without radios or helicopters, would be futile.
No medieval lordship could afford to keep permanent forces on the alert for long. The defence of a lordship depended essentially on rallying the tenants to fight on their own behalf: to gather them needs time and well-known meeting points. The best strategy is to intercept raiders on their way out, when they are burdened with loot and tired from the effects of rape and pillage, or have
Figure 41 Map of mottes in the lordships of Meath and Oriel
Been mauled by a successful defence. Again, it is no use having bases ten miles apart. We see the system in action in the 1211-12 Pipe Roll, where there are accounts for groups of 20-50 men paid for guarding districts. They presumably acted as an immediate reaction force, but mainly to hold a base and act as a stiffening to the local tenants when they arrived. The bases these men occupied would have to be selected according to an assessment of the route that the raiders were likely to follow, and to be astride it. A successful defence plan would need many potential bases to cover all eventualities. It is likely that only one would be chosen on each occasion and many sites might be left empty at any one time: indeed, if there is no reason to expect a raid, all the frontier bases might be abandoned.
The second thing that we can note about these border mottes is that many more of them have baileys attached to them than do the mottes of the hinterland (Fig. 42). In Ulster, one bailey (or two, counting the possible site at Downpatrick) still exists in the core area of the lordship, out of some 49 mottes. Nine baileys are found in the 19 mottes in the two key border areas noted above, or 47 per cent. In Meath and Oriel 55 per cent of the baileys are found attached to mottes along the border: only 31 per cent of the mottes in the hinterland have baileys. The latter figure drops to 25 per cent for the present Co. Meath alone. In the three Leinster counties of Kildare, Carlow and Wexford, nine out of 49 mottes (18 per cent) have baileys. Consistently in these lordships, some 50 per cent of the border mottes have baileys, twice as many as the mottes of the hinterland.
This difference between some areas and others in the frequency of baileys contrasts with the position elsewhere in the British Isles, where very few mottes lack baileys. Not only are they not almost automatic in Ireland, but the baileys are also smaller than elsewhere. It is not uncommon for a motte to have been added within an earlier rath, taking up a fair proportion of the enclosure, which may have been only 30 m (100 ft) in diameter to begin with. Why this should be is one of the real puzzles about these early castles in Ireland. Along the borders, it may be linked to the sort of use outlined above. The baileys could mark off those mottes where it was expected that forces might be stationed in times of tension and raids, protected within the bailey enclosure. This would certainly fit the results of the excavation of Duneight, itself one of the outer Co. Down sites, where the poor quality of the structures found would seem to be more what might be provided for troops. If the role of baileys along the borders may be explained thus, their small size and relative rarity in the core areas of the lordships cannot.
If the lordships north of Leinster present a contrast in their militarised borders, as evidenced by the density of their mottes, the lordships of Munster display a different contrast. Here we are faced with an extraordinary lack of mottes in the present counties of Waterford, Tipperary, Cork and Limerick. Between them, there are only about 22 mottes in these four counties. In part this may be due to a lack of field survey, but the counties of Tipperary and Cork have both been covered, and Limerick to some extent as well. It was this absence of mottes which prompted Twohig and others to postulate the existence of ringworks to fill up the gap, an argument criticised earlier in this chapter. The absence is almost certainly a real one, not produced by differential destruction or the selective use of a different form of earthwork castle in these lordships. We are faced with a situation, not of a lack of castles as such in southern and eastern Munster, for there are stone castles and some earthwork ones, such as the royal mottes of Lismore or Tibberaghny and Philip of Worcester’s Knockgraffon, but of a lack of the proliferation of castles, such as we can see in the lordships of the east and north.
World History
