sign in sign up
<<
Home , Machicolation, Murder hole, Tower house, Yett

The defensibility of Irish Houses - A study.

60 The Studies Group Journal No 4: 010-11 The defensibility of Irish Tower Houses - A study.

The defensibility of Irish Tower Houses - A study. uncan Berryman ower-houses are often considered to be small , with similar defensive fea- tures and functions. hey are small, single , often four or five storeys high and have a simple plan. hey were most likely to have been accommodation for the small- er land-owning lordship, both Gaelic and Anglo-Norman. ower-houses became more numerous from the late-fifteenth and early-sixteenth century; they mainly fell out of use after a few hundred years, but some remain occupied today. ower-hous- es are found across Ireland, with concentra- tions in the southern Counties, the Pale - the area around Dublin - and southern County Down [see CSGJ 20, 7-9]. Similar buildings can be found in Scotland, mainly around the Borders, where they are called Peel owers. he tower-houses of Scot- Fig. 1. Castle Ward - Co own. Example of battle- land are similar in appearance, but differ in ment-level box over the entrance. design.

Many scholars, such as Leask,om McNeill, of Queen's University Bel- Sweetman, homson and McNeill, have fast, rejects the idea that tower-houses were placed tower-houses alongside other cas- primarily for defence [1997, 217-221], in- tles in their respective studies. his exem- stead his studies have stressed the social plifies the position that tower-houses hold factors and the architectural design of the in the field of castle studies, being seen as towers. Research by Rory Sherlock [2007, a relatively minor area of study. It is true 59] and Gillian Eadie [forthcoming] has that they share many features with their attempted to investigate how the domestic larger counterparts, but they have a very functions of a tower-house would have dissimilar position in the social scale and operated. Recent research carried out at must serve slightly different functions. he Queen's University, Belfast has taken a earliest work on tower-houses was carried slightly different approach to the study of out by Leask [1941, 75-91], this formed tower-houses [Berryman, 2008]. his re- two chapters in his book of Irish castles. search has taken a sample of tower-houses Similar work was carried out by Sweetman from across three counties of Ireland, Co. [2005, 137-174] in his book on Irish medi- Down, Louth and Meath, rather than study eval castles. However, neither of these every tower-house in one County. Instead evaluated the effectiveness of the defence of looking at the tower as a whole, this or living facilities. erry Barry, of rinity study focused on one important feature of College, Dublin, considers tower-houses to the tower-house - the door - crucial to the be primarily defensive and has based his defence of the tower. Being the only entry, studies on an attempt to date them and to it was central to the tower's social function search for their origins [1987, 180-190]. and its every day life.

The Castle Studies Group Journal No 4: 010-11 61 The defensibility of Irish Tower Houses - A study.

Fig. 2. erryhivenny Castle c.1643. Conjectural re- Fig. 3. Audley’s Castle, early C15. The south-east construction, after Leask, showing extent of the bawn. façade. Arch machicolation above the entrance in the south (left) , at ground floor level. lost.

The Bawn either as a result of decay or of removal. Research by Aideen Burke at Derryhiven- he largest defensive feature that the tow- ny (Fig. 2) has shown an extensive com- er-house could enjoy was a bawn wall. plex that has left no surface remains [see his was usually a three metre high wall CSGJ 20, 50, & The erryhivenny Castle that surrounded the tower to form an enclo- Project]. A bawn wall would have been a sure. his feature could have prevented very expensive construction project and it attackers from gaining direct access to the is possible that most lords could not afford tower-house, they would have to breach the to construct one to protect their property. wall first. A bawn wall serves the same Another explanation for the lack of bawns purpose as a perimeter wall of a great cas- is that they were not needed; there may not tle; protecting the from enemy assault. have been enough violence to justify the However, extant bawns are quite rare; ap- expense of a bawn wall. proximately only twenty-five percent of tower-houses still have evidence of their Arch Machicolation bawn’s existence; this is similar to the num- A tower-house was usually designed with ber of mottes which have baileys attached machicolation to protect the doorway. to them [Mallory & McNeill, 1991, 262]. hese are much like the machicolation he lack of baileys in Ireland contrasts with found in great castles; they have similar their relative abundance in , where functions and differ only in design. he eighty percent of mottes have a arch machicolation is closest in design to [Mallory & McNeill, 1991, 262]; this may those of great castles. It is formed from an indicate a trend in Ireland for not creating arch between two projecting on the an enclosure to protect subsidiary build- front face of the tower. his design is the ings. It is possible that more towers had most effective at defending the doorway, bawns, but they have not been preserved; but it leaves the defenders vulnerable to

6 The Castle Studies Group Journal No 4: 010-11 The defensibility of Irish Tower Houses - A study.

Fig. 4. Audley’s Castle c. C15. The south-east façade. Arch machicolation at level above the en- trance. attack [Berryman, 2008, 36]. Arch machi- colation is only found in three tower-houses in south-east County Down; namely Aud- ley’s (Figs. 3, 4), Kilclief (page 242) and Jordan’s (Fig. 14) [McNeill, 1997, 213]. Box Machicolation & he most common design of machicolation Fig. 5. Aughnanure Castle c. C16. From the east. The rooftop machicolation is reconstructed, but the corner is box machicolation. his is formed by bartizans are original. projecting part of the exterior wall out- wards on corbels [Berryman, 2008, 36]. The or iron grate he box machicolation provides a small Probably the most effective defensive fea- opening above the doorway, giving the de- ture of a tower-house was a yett; an iron fender a smaller field of vision, but provid- grill or grate that could have been closed ing them with a good defence against across the main door to protect it from attackers. attack and was secured with a chain Machicolation is relatively com-[Sweetman, 2005, 140]. A yett performed mon and is found across Ireland; examples the same function for a tower-house as a include Castle Ward (County Down) (Fig. did for a great castle. It is very 1) and Clara (County Kilkenny). Some rare to find a yett still in place, such as at tower-houses, such as Aughnanure (Fig. 5) Cregg Castle (Fermore, Co. Cork) (Fig. 6); (County Galway) and Ballymalis (County but a number of towers have evidence that Kerry) feature corner bartizans; these are one was once fitted to the main doorway similar in design and function to box mach- (e.g. Athclare, Fig. 7). his evidence in- icolations, but are found at the corners of cludes holes for the hinges in the dressed the tower-house at first or second floor stone of the door jambs and a hole that level [Sweetman, 2005, 150 & 171]. Mach- passes through the wall for the chain to icolation appears to be quite an effective secure the yett closed. A yett could have form of defence, but possibly designed protected the door from a battering ram more for their image than their functional- attack by dissipating the force of the attack ity, as they either constrain the defender or through the walls and preventing the ram leave them open to attack. from hitting the door. If the attackers tried

The Castle Studies Group Journal No 4: 010-11 63 The defensibility of Irish Tower Houses - A study.

Fig. 6. Cregg Castle (Fermore, Co. Cork). Original Fig. 7. Athclare Castle, Co. Louth. Slot for the chain yett. Image © avid Newham Johnson. to close the yett. to burn the door down, a yett could have es in social conditions and a reduction in prevented them from getting into the tower- the need for high security in homes house. However, research shows that ap- [Simms, 1975]. hus, a possible explana- proximately twenty percent of tower-hous- tion for the lack of is that not all es were fitted with a yett [Berryman, 2008, tower-houses needed to be protected from 39]. his may suggest that tower-houses serious attack. were not designed to be primarily defen- Murder holes sive as it contrasts with evidence from great castles. Many of the great castles, Inside a tower-house, there were a number such as Colchester, Rochester [oy, 1985, of ‘defensive’ features that could have been 70 & 76], or of enclosure cas- present. One feature that was taken directly tles, such as Bodiam or Caernarfon [oy, from the design of great castles was the 1985, 196 & 216], had to pre- ‘’. he majority of tower- vent an attacker breaking through the door- houses have a 'murder hole' over the lobby, way. If so many great castles are which is the equivalent of siting them over constructed with a portcullis, why are more the entrance passage of the . As tower-houses not provided with yetts? in great castles, the 'murder hole' allowed ower-houses were constructed a number the defenders to drop rocks or other projec- of centuries after the great castles, and one tiles on to their attackers who had broken would assume that they would have copied through the main door [McNeill, 1992, 98]. the designs of the great castles. Contrary to However, the 'murder hole' is often very previous research [Barry, 1987, 181], it is small compared to the area of the lobby; likely that this period of history saw chang- this means that it would have been difficult

64 The Castle Studies Group Journal No 4: 010-11 The defensibility of Irish Tower Houses - A study.

Fig. 8. Audley’s Castle, Co. own. Murder hole exit. for the defenders to hit their attackers, un- less they stood directly under the 'murder hole'. One tower-house, Audley's Castle (County Down), may suggest a different function for the 'murder hole' (Figs. 8, 9). he 'murder hole' of Audley's Castle is situated above the door to the ground floor Fig. 9. Audley’s Castle. Murder hole entrance along the spiral staircase. room, not near the main doorway; it is very small with a restricted opening. he im- plausibility of using this 'murder hole' for sively security [Simpson, 1992, 319]. he defence leads to the consideration of other control of the doorway and access to the functions, such as communication between castle was the main role of the porter, but the first floor and the ground floor, possibly he would also have looked after the castle between the lord and his porter concerning when the lord was not present [Simpson, admittance to the tower-house. All murder 1992, 321]. he porter may not have been holes provide only a restricted view of the so important in a tower-house, but someone lobby and thus other possible uses should needed to open the door and control access be considered for them as well. to the lord's accommodation. The Lobby & Draw-bars. The Defensibility Test Almost all tower-houses have a Very few academics have ques- number of doors opening off the lobby; tioned the physical strength of the tower- usually these lead to a ground-floor room house door itself. It is either assumed to be and the staircase; in some cases there is a too weak to withstand assault or strong third door leading to a guard room. Often enough to hold out during a . Recent- there is evidence that these doors were ly, an experiment was carried out by re- securely closed with a draw bar attached to searchers at Queen's University Belfast to the inside of the door. hese draw-bar di- find out whether a tower-house door could mensions are usually very similar to the have survived an attack by battering ram or main doorway. he ability to lock these by fire. It was possible that all the force of doors indicates that there may have been the battering ram would have been dissipat- someone inside the ground floor room, pos- ed through the walls of the tower-house and sibly a porter. he porter of great castles all the heat from the fire would escape up- served a number of functions, not exclu- wards.

The Castle Studies Group Journal No 4: 010-11 65 The defensibility of Irish Tower Houses - A study.

minutes for the door to burn down completely. Despite their lack of ex- perience in using a batter- ing ram, the team quickly learnt how to use the ram. hey succeeded in break- ing the draw bar within the first ten hits. Howev- er, the broken draw bar prevented the door from opening, as it was still in position and it took a total of fifty-four hits to knock the door out of the door- way. he whole ramming exercise lasted about five Fig. 10. Queen’s University battering ram team starting work. Below: Fig. 11. minutes, this included Fire just lit. two changes of the team; an experienced team o do this, two doorways werecould probably have managed it in approx- built, one to test the battering ram and the imately three minutes. other to test fire . he doors were made of 30mm thick oak planks, which were at- his experiment shows that tower- tached to bracing planks with bolts. he houses could not have withstood an or- door was hinged by projecting the ends of ganised attack from a team of experienced the first plank into holes in the ground and raiders and a tower-house's defensibility lintel. he doorways were constructed of was not as effective as its apparent defen- concrete blocks and mortar and were built sive features might suggest. he best de- to measurements taken during survey fensive features that a tower could have work. One door was securely closed by a draw bar, this was 80mm square made of solid oak and a battering ram was formed from an ash tree. Wood was stacked in front of the other door to test the fire. he wood in front of the door that was to test the burning was ignited with newspaper and matches and more wood was added as the fire grew. In medieval times, the fire was probably lit away from the tower and then brought to the door. he door quickly became very blackened and the lintel reflected the heat back onto the door. After about twenty minutes, the fire had begun to penetrate through weak- nesses in the door; eventually the door then began to burn on its own. It took forty

66 The Castle Studies Group Journal No 4: 010-11 The defensibility of Irish Tower Houses - A study.

be viewed as essentially a device for closing the door to keep the elements out; it was not a significant defen- sive feature. An ‘evolu- tion’ of defensive features can be seen in tower-hous- es. he most basic design is a door and a draw bar; this combination would have been ineffective against an attack. he next step would be to defend the door with a yett or portcul- lis. his option would have reduced the effect of a bat- tering ram and allowed a Fig. 12. Above: The battering ram team breaking through after 5 minutes. Fig. 13. Below: The oak door burnt down completely after about 45 minutes. door to hold out for much longer. he most effective means of inhibiting an at- been built with were a yett or a first floor tacker from gaining access was to place the entrance, both are found in Irish tower- doorway on the first floor. A flight of stairs houses. Yetts were rare, but are found in would have prevented the use of a battering approximately one quarter of all tower- ram against the door and thus stopped an houses [Berryman, 2008, 39]. First floor attack. entry is mainly restricted to tower-houses in the south-west corner of Ireland [Samu- ower-houses have many features el, 1998, 111-7]. Examples include Leam- that appear defensive in nature, but they con and Kilcoe castles built between 1427 and 1470 and Dunalong, built between 1460 and 1500; Bunratty is also an example of a very large tower-house with a first- floor entrance built after 1450 [see CSGJ 23, 89-98]. A first floor entry would have made it impossible for attackers to use a battering-ram against the door; it could also have made lighting a fire very difficult. If the stairs were made of wood, they could have been destroyed if the tower was going to be attacked, thus preventing the attackers reaching the doorway. If the attackers could break the door down in such a short time, then these are the most important defences, all the others are of little use and only for decoration and display. Conclusions hese results have implications for castle studies in general. he draw-bar can only

The Castle Studies Group Journal No 4: 010-11 67 The defensibility of Irish Tower Houses - A study.

Video footage of this experiment can be viewed at: www.qub.ac.uk/schools/gap/Education/Archaeologya ndPalaeoecology/StudentAchievements Bibliography Barry, ., 1987, The Archaeology of Medieval Ireland, London: Methuen & Co. Berryman, D., 2008, oorways to Irish Tower Houses: a study, BA Dissertation, Queen's Uni- versity Belfast. Berryman, D., 2009, ‘Home security: how strong was a ower House door’, Archaeology Ireland 23 [Summer 2009], 8-10 Eadie, G., forthcoming, The functions and clas- sification of tower houses, PhD thesis, Queen's University Belfast. Leask, H.G., 1941, Irish Castles and Castellated houses, Dundalk: Dundalgan Press. Mallory, J.P. & McNeill, .E., 1991, The Ar- chaeology of Ulster, Belfast: Institute of Irish Studies, QUB. McNeill, . E., 1992, Castles, London: English Fig 14. Jordan’s Castle, near Ardglass, Co. own. Heritage Third example of the ‘arch machicolation’. McNeill, . E., 1997, Castles in Ireland, Lon- could only have been used to defend the don: Routledge tower as a last resort. his experiment has hompson, M.W., 1987, The ecline of the shown that a simple, and typical, tower- Castle, Cambridge: Cambridge University Press. house was not designed to be exclusively or primarily defensive; this is supported by Samuel, M., 1998, 'A entative Chronology for the lack of effective defensive features, ower Houses in West Cork', Journal of the such as a yett. herefore, although tower- Cork Historical and Archaeological Society 103, 105-24 houses appear similar to great castles, they are not the same and should be studied in Sherlock, R., 2007, 'he Social Environment of isolation, with different research objectives the Irish ower House', in The Castle Studies and methods of study. Group Journal 20, 59 [email protected] Somerset Fry, P., 1996, Castles of Britain and Ireland, Newton Abbot: David & Charles. cknowledgements Simms, K., 1975, 'Warfare in the Medieval I would like to thank Patton Construction Gaelic Lordships', in The Irish Sword 12, 98-108. for supplying the materials and Ardnavally Scout Centre for hosting the experiment. I Simpson, G.G., 1992, 'Claves Castri: the role of would also like to thank Dr om McNeill the gatekeeper in Scottish Medieval Castles', and Dr Gillian Eadie for their guidance and Château Gaillard 15, 319-324 assistance throughout this experiment. Sweetman, P.D., 2005, Medieval Castles of Ire- land, Cork: he Collins Press. oy, S., 1985, Castles, New York: Dover.

68 The Castle Studies Group Journal No 4: 010-11