We may pick out several themes for discussion among these castles. The first is their actual existence in stone at all, at this date. The lords who caused them to be built had come to a country where, as far as we know, there was no significant tradition of secular building in mortared stone. In the Church there had been round towers (which involved scaffolding and lifting gear) since the tenth century; complex mouldings and architectural carving in Irish Romanesque dated only from the early twelfth century; the elaborate planning of monastic complexes for Savignac or Cistercian houses dated from the same time.
Figure 30 Limerick castle: general view from the Shannon bridge
Figure 31 Limerick castle: north-west tower and gate house
Organising the building of their castles must have needed a special effort, with at least the master masons brought into Ireland. Combined with the church building they patronised, they must have given a major boost to the building industry.
This effort they clearly found essential. John de Courcy had a castle to retreat to in Co. Antrim, surely Carrickfergus, in 1178, the year after his seizure of the lordship. Unless the great tower of Nenagh was built during the minority of Theobald Walter II, or he used remarkably old-fashioned masons after 1221, it must have been built by his father before 1206. Nenagh was granted to Theobald and Adare to Geoffrey de Marisco in about 1199. The window style Geoffrey’s masons used would have been outdated a decade or two later, so he must have started work soon after the grant. Here we see some of the organisation at work, for he employed men who had been working at the Cistercian abbey nearby to build his hall. Building in stone, rather than earth and timber, was neither cheap nor quick: these lords were sending a clear message of their intentions to commit themselves to Ireland.
A second theme is that they liked great towers in their castles. Of the baronial castles we have considered here, only Adare (which has a detached tower of uncertain date and status) and Carlingford (which has a dominant gate tower) do not have a clear great tower. At Carrickfergus, Trim, Dunamase and Maynooth, they are rectangular; at the others round. For the majority, it is clear that they were not built to accommodate halls. Trim and Carrickfergus have fine rooms on
Figure 32 Plans of the gate houses of Nesles-en-Tardenois and Limerick castles
Their third floors, but not only are they hopelessly remote for access by the general public for whom halls were intended, but there are halls in the courtyard. Like the round towers, these were towers for the accommodation of the lord and his household alone. Conversely, at Maynooth and Dunamase (in so far as we can say anything about the latter) the great towers were almost entirely taken up by the hall. The great towers were either for the lord’s chamber or for the hall, but not for both. Nor can they be distinguished by size, but only by analysis of their internal plan in the context of the other buildings of the castle.
The traditional name for such towers, ‘keeps’, implies a military role for them: to be the strong point and place of ultimate resort for the whole castle. This role the great towers of castles in Ireland were ill suited to perform. The greatest, Trim, was weak enough to need to be reinforced later in the thirteenth century. Dividing the lord’s accommodation off from the great hall meant that, if either was used as a place of strength, inevitably the other element was abandoned in the case of a serious attack. The price of giving up the great hall, the most
Figure 33 Carrickfergus castle: east tower
Expensive room in the castle and seat of the lord’s prestige, to take the brunt of the damage seems a remarkable one, and the same applies to the lord’s chamber. As ‘keeps’ they are inadequate. Nor were the curtain walls very much more consistently military in their design. The great hall windows of Adare, Trim and Carrickfergus were all obvious sources of weakness. Carlingford and Dunamase are strongly defended by nature and by construction. The two royal castles of Dublin and Limerick are strong, but Dungarvan was not. The second period of work significantly strengthened Carrickfergus, which illustrates the point: the first period placed comfort as high as defence in priority. The main purpose of these castles for the lords was to give a suitably impressive setting for their exercise of the new lordships they had acquired. This is seen at its most extreme at Carrickfergus, where the dominant tower with John de Courcy’s lodging took up more than a quarter of the castle’s area.
The deliberate nature of this choice is reinforced when we remember that the castles were all paid for and designed by men from outside Ireland. The lords had wide connections with the Angevin court and there was no native tradition of castle building to cause this pattern. These connections and a knowledge of ideas outside Ireland show freely in the designs. The most obvious is the popularity of the round great tower, which we must surely attribute to the prestige of the great tower of Pembroke castle (Renn, 1968; King, 1978). Its builder was William Marshal, lord of Leinster, and it was set at the main point of embarkation for Ireland. It is interesting to compare it with Nenagh. The latter takes the idea of the round great chamber tower from Pembroke, but its walls (themselves the thickest of the round towers of Ireland’s castles) are some 70 cm (28 in) thinner than those of Pembroke, so that the windows give more light; Nenagh also has fine carving, which Pembroke lacks. There is nothing of Pembroke’s elaborate fighting top at Nenagh: Theobald Walter asked for a more comfortable chamber tower than his model. A second general parallel is the use of the arrow loops with plunging loops which became popular in the early thirteenth century to control the base of walls, at Trim, Carlingford, Nenagh and Dunamase.
Two specific parallels are interesting. The first is the use of towers on the Lacy castles at Trim and at Carlingford, which go from square to semi-octagonal. This is surely a personal thing, recalling the north-east and north-west towers at the Lacy caput of Ludlow (Renn, 1987, Fig. 8); we might almost talk of homesickness. The second is the round west gate tower at Trim, taken from the Poitevin castle of Coudray-Salbart (Curnow, 1980), but with an added barbican. This was clearly a question of de Lacy providing a suitably impressive parallel as the main element in the show front of his new castle. The town gate, with its square to polygonal plan, the parallel with his home castle, is almost a private matter, a personal reference; the west gate (the approach from Dublin) a design inviting comment and explanation to impress a guest.
Constructionally, round towers present the builder with one problem, that of flooring the space. In a rectangular building, it is easy to run joists between the two nearer sides, although there may be differing ways of securing them in to the walls. At three of the round towers of Ireland, the holes for the joists are still preserved enough to allow them to be measured and their pattern planned (Fig. 34). Inchiquin and Clogh Oughter both have a very similar scheme, with a single central beam providing support to the beams running across it. The first floor at Dundrum is different, with three lower beams providing greater support; given the relative irregularity of the beams, however, this is as likely to be the result of a nervous carpenter giving more support in case of trouble as a careful assessment of the risks. The second floor is different: it uses half the number of beams, and the crossbeams are not set lower than the others, but must have been elaborately jointed where they met. This might reinforce the view (Jope, 1966, fig. 135) that the second floor is later than the rest of the tower. This view is based solely on the nature of the mural chamber vaults, and ignores the continuity of the stair, chimney flue and the absence of a change in masonry. What this information means is unclear: the ways used to floor the round towers of Wales would be the first point of comparison to see if there are different craft traditions at work here.
The royal castles compare interestingly with the baronial ones. Dublin and Limerick share the idea of the roughly polygonal plan with corner towers and a pair of smaller towers at the gate, but with no great tower. Again, as with round towers, this is a theme which links the castles of south Wales, the work of William Marshal or Hubert de Burgh (brother of William, lord of lands in Tipperary and Limerick), and then France (Knight, 1987; Curnow, 1980; Heliot 1965). The French ideas show most strongly in the design of the gates at Dublin and Limerick (Fig. 31). They do not have true gate houses in the sense of single buildings with towers on either side of the passage, bound together at first-floor level; instead there are two small towers set close together, with the gate between them. This is unlike the work of William Marshal at Chepstow or Usk, where there is no gate tower at all, although here we have the two similar gate houses of Dunamase and Carlingford. Nesles-en-Tardenois has the same design as Limerick, with the towers circular within (see Fig. 32); Dourdan has similarly small towers, but projecting back behind the curtain to provide a second pair of rooms behind the towers, which leads, through Boulogne, Bolingbroke or Beeston, to the full gate houses of the later thirteenth century (Chatellain, 1981; Curnow, 1980).
The royal work of 1215-23 at Carrickfergus was much less ambitious, although also putting general principles into practice. The aim was to strengthen the defences and the method was to deploy archery in the same way as in a number of late twelfth-century castles built by the Angevins in England, from the Avranches tower at Dover or the Bell tower at the Tower of London around 1190, to the Buttevant tower at Corfe of 1202-4 (Renn, 1969; RCHM, 1970, 60), as well as in France. At Carrickfergus, and in the north-east tower of Limerick castle, two or three arrow loops were gathered into the one embrasure, giving the archer a choice of fields to control (Fig. 35). At Carrickfergus, a more specific idea was employed: to place a strong tower, to be defended by archers firing
Figure 34 Plans of the floor joists in the great towers at Dundrum (Co. Down), Clogh Oughter and Inchiquin castles
Through multiple loops, which would control a weak point, typically a salient angle. Clearly the masons on the royal works in Ireland were at least aware of the latest ideas of castle defence design in England or France.
Defence was not the whole purpose of a castle, however: the administration of the lordship depended on it, and the lord lived there in as much comfort as he could. Below the major castles, which were the heads of lordships, stood the lord’s tenants. These take us to questions of the provision of lesser fortifications, not meant to be the set pieces of design and major centres of power in the land: the second line of castles. This matter of the proliferation of castles affected all the lords, with their tenants and their lesser manors, and to this we must now turn.
Figure 35 Two examples of the royal works in Ireland: multiple arrow loops in the east tower of Carrickfergus castle and the north-east tower at Limerick castle
World History
