an irish view of scottish mottes journal 2020-21rev4... · 2020. 12. 17. · an irish view of...

131THE CASTLE STUDIES GROUP JOURNAL NO. 34

An Irish view of Scottish mottes

Tom McNeill

Abstract:

In both countries castles, many of whichfeature mottes, are introduced in the 12thcentury, later than the 11th century in Franceand England. They were part of a majorre-organisation of the societies, linked to theintroduction of Anglo-Norman lords and lord-ship, principally to emphasise the importanceof manorial centres. As such they cast light onthe pattern of settlements of the time, afterthe reforms. Detailed study of their distribu-tions, however, shows that, instead of a sys-tematic introduction of castles with nucleatedvillages and towns, the situation was muchmore varied. They might represent smaller,dispersed settlements in the 13th century, inline with the earlier, Gaelic pattern continuedunder the new administration. The chronologyof the mottes in Scotland and Ireland hasimplications for mottes in other areas.

Mottes are towers. The basic purpose of bothwas to provide an imposing structure foranyone who wished to raise their house upto dominate the landscape and to add anelement that would make it resemble a castlefrom the outside. Although they should beregarded simply as components of castles,they have been classified as a monument intheir own right, defining a type of castle. Assuch, they have been assumed to have asingle purpose and have been written intoone or more general roles. Wood and earthconstructions are often described negatively;as temporary, even makeshift, structures tobe replaced as soon as possible by morerespectable, because more permanent, stonebuildings. Mottes were assigned to a short-lived role, pre-eminently as being erected inthe course of a military campaign. They wereseen as primitive and in most general books

on castle produced before 1990 mottes aretreated as a type of castle in the first chapterbefore the author moves on to discuss thedevelopment of the “real” stone castles.Significantly, only one book has ever beenproduced on them in English (Higham &Barker, 1992). Many mottes have surviveddown to modern times, partly because theyare difficult to remove, while the amount ofland cleared by their removal is small, so thatit is often not worth the effort. Mottes havebeen studied in two ways. The first isindividually, through excavation of the topwhich has disappointed, because theyproduced such varied results, although themodel of Abinger, excavated in 1950, wasideal: a simple timber tower on top of amotte. Unfortunately for those who sawmottes as a type, later results varied wildlyfrom this, from completely blank sites to thecomplexities of Hen Domen. The secondapproach is to see them as additions to lordlysites, and focus on the sites which they markout for. This means looking at them enmasse, mapping the patterns which theydisplay: geographical, topographical andsocial, the line followed in this article.

In both Scotland and Ireland, becausethey are viewed as early and intended to bereplaced by proper castles, they have beenassociated with the first stages of an influxof English lords in the 12th century. Thisinflux was very important and led to a greatchange in the economy and settlement ofboth countries – feudal order – and mottestherefore have been studied very much inthis light. In Ireland, this change is portrayedas a solely military conquest, partly becauseGiraldus Cambrensis (a man with an agendaif ever there was one) tells us that it was one,but he is backed up by many otherdocumentary sources, although counteredby many others. Mottes have been built into



this story ever since Orpen’s magisterial,four-volume account of the Normans inIreland, published in 1911-1920. Becausethey were the remains of castles, which weremilitary structures, they provided support forthe idea of a purely military conquest.Conversely, because they were linked to amilitary conquest, this showed that they weremilitary structures. A perfect feedback loopwas established, largely without examiningthe pattern of mottes themselves. Over thelast fifty years, with the development of thestudy of the archaeology of later medievalIreland, mottes have been listed anddescribed to allow us to have a picture oftheir occurrence and siting, not seen before;this evidence and some conclusions from itwere gathered together in an articlepublished in the Archaeological Journal nineyears ago (McNeill, 2011) and may be brieflysummarised here.

The first observation is that the distributionof mottes in Ireland is very varied. The overalldistribution of mottes as a whole does notcompare, even slightly, with the boundariesof the English lordships established in the12th century. Their distribution becomessparser south and west of Dublin; in particularthere is only a handful identified from thecounties of Waterford, Cork and Limerick. Ifthey were an essential part of a war for theconquest of military lordship in Ireland, it isvery hard to explain why the lords of thesouth could do without them. Either motteswere not essential as military structures orthe “conquest” was much more peaceful thanit is portrayed; or both. This argument isreinforced by the way that the “less military”moated sites appear to replace mottes in thesouth as demarcators of lordly capita. Somemottes can be identified as being put upeither as part of a military campaign or

Fig.1. Ireland: map of mottes (left) compared with a map of moated sites (McNeill 2011)



intended for a primarily military role. Two ofKing John’s Justiciars erected mottes atimportant sites: John de Gray, Bishop ofNorwich at Clones and Belturbet in 1212 andHenry of London, Archbishop of Dublin, atRoscrea in 1213. Analysis of the concentrationof mottes along the western boundary of thelordships of Meath and Louth shows that theywere put up as a group and that their purposewas to provide strong-points to counteractraids on these lordships from the west afterthe 1220s. Being able to identify a (small)minority of mottes as being built primarily forwar, emphasises that the majority were not,simply additions to demarcate the caput of alordly holding

Among these “civilian” mottes, proliferatingalong the eastern side of the island, thedistribution is varied between lordships. InLeinster lordship, mottes seem closelyassociated with knights’ fees, but in Meath,Louth and Ulster they spread further down thetenurial hierarchy. This applies particularly to

the Lordship (later the Earldom) of Ulster, thepart closest to Scotland. In the administrationof the land, it has long been recognised thatthe secular holdings, the manor were closelyrelated to the parishes. Both were establishedat about the same time and it would havebeen an obvious plan to make the two newunits coterminous: lords of the manor couldhave the prestige (and advowson) of a parish,while the parish priest had a clear patron andthe two could support each other. From this,it has been an easy step to assume that thisunit would translate from administration ofthe land and its revenues into actualsettlement. From the 1960s, when the studyof deserted villages in England seemed tomake the idea of a village the norm ofmedieval settlement, writers have expectedthat a nucleated pattern of manor, parish andvillage would have been what the new lordswould have established in Ireland. However,since then we have come to realise that thenucleated village is not necessarily the only

Fig.2. The motte at Kilwaughter, Co. Antrim. The church site lies to the right of the buildings ofthe recent castle.



pattern of English medieval settlement andthat a more dispersed pattern of freetenancies was common, especially in landswhich needed to be cleared for arable, or inareas where pasture was stronger. Thisbrings us to two questions. The first is howfar did the establishing (by whatever means)of these new lordships actually initiate theeconomic and social revolution that they arecredited with; i.e. how real was this “feudalagenda”? The second is what does thedistribution of mottes tell us of theorganisation of the land in practice asopposed to its administration; in particular,can we see it as a totally new construct or didit contain significant elements of continuityfrom the old landscape ?

In Meath and Louth we see both the presenceof many mottes at manorial capita, as wewould expect, but also others away fromthem, presumably marking lesser holdings.There is a warning, however: the coincidenceof parish and manor with church and motteis not as frequent as we might expect if thelandscape was being re-ordered andnucleated villages based on arable farmingintroduced. For Scotland, the case of Ulsteris clearly the most useful. From early in the12th century, forces from the west ofScotland and the Isle of Man are to be foundactive in Ulster. The first English lord, Johnde Courcy, married the daughter of the Kingof Man. King John made an effort to bolsterhis power in the North Irish Sea area byexpanding the power of the Lords ofGalloway. After 1212, he granted lands inwhat is now north Co. Antrim to Alan ofGalloway and his cousin Duncan of Carrick;the first may not have resulted in muchsettlement but Duncan does seem to haveorganised his lands, from Glenarm to Larne.Hugh de Lacy expelled Alan and Duncan whenhe re-gained the Earldom of Ulster in 1226,

and he divided north Antrim into three majorblocks: his own lands in the north-west, fromColeraine to Ballycastle and Ballymoney; thesmall Irish client kingdom of Ui Tuirtrearound Ballymena; and the coast fromRathlin island and Ballycastle to just south ofBallygalley, which he granted to the Scottishfamily of the Bissets (see McNeill, 2019-20for details). Not only is Antrim closegeographically to Ayrshire and Galloway, butit shared lordships as well.

We may analyse the pattern of mottedistribution and land holdings in Ulster forevidence on the view that they are of apackage of changes brought in by the newlords, in a break with the earlier organisation.In favour of the idea of change, we have thesame relationship of manor and parish aselsewhere and we can see evidence of a newemphasis in agriculture. The Earl’s capita, aremarked by castles and borough towns;Carrickfergus and Dundrum with stonecastles, Antrim and Dundonald by largemottes. In the core areas of the Earldom wesee evidence of a new agriculture. In northDown the mottes are focused on good land,but the evidence is clearer in mid Antrim,between Antrim town and Carrickfergus. Thedistribution of lesser mottes may becompared with that of the centres of earlierestates, the raised raths, which tend to bewidely scattered and sited at the boundarybetween the better, lower land and the uppergrazing. Mottes occupy lower ground, thearea suitable for arable, and are more denselydistributed, reflecting, perhaps, smaller butmore intensively farmed estates. By contrast,we have the have a good proportion of themottes sited at places which are eitherundocumented or which we may identify, notas manorial centres, but the holdings of freetenants of manors. The manorial structure inthese cases appears to be simply one of rent



collection, not part of a systematicagricultural pattern. In two parts of Ulster,perhaps with different regimes the Churchestates of southeast Co. Down the sites ofBallynarry (Davison 1961-2) and Lismahon(Waterman 1959) appear to be developed ontop of earlier raised raths; Lismahon inparticular, seems to be the house of a bailiffon the small estate of Christchurch, Dublin,not a full manor. We find the same innortheast Co. Antrim, along the boundariesbetween the Earl’s lands and those of theBissets: the site of Drumadoon (McSparron &Williams 2009) and the similar site atDoonmore (Childe 1938) while the raised rathat Knockaholet appears to have had a mottebuilt on it, as does the one at Broomore. Afurther twist is provided by the Irish kingdomof Ui Tuirtre, where there are a number ofmottes, including the fine motte and bailey ofHarryville in Ballymena. Mottes are notnecessarily confined to English lords.

When we try to turn from the mottes, and thecapita that they mark, to the actualsettlement patterns, we see the same varietyin the evidence. A good introduction comesat the site of Killyglen (McNeill, 2019-20).Here what appeared to be a classic site of anucleated centre, with an Earl’s manor(possibly), parish church and site of anucleated village, proved to be more complexwhen subjected to detailed survey andexcavation. The space between the motteand the church which appeared to be likely tocontain the village houses proved to have noremains of occupation. Instead, the parishchurchyard seems to have been imposed ona small, irregular settlement of perhaps halfa dozen houses. This fits well with the sortof irregular settlement excavated at theBishop of Connor’s manor of Portmuck inIslandmagee (Anderson & Rees, 2004 ). Themottes in Carncastle parish seem to representsuch small settlements within the

administrative framework, as recorded in asurvey of Bisset lands in 1278. In Ulster,therefore, mottes articulate a land wherenucleated organisation with arable farmingco-existed with less formal, small groups ofhouses; change in the landscape along withcontinuity of central places; English mottesbeing used by Irish lords to embellish theirpower centres. The mottes lead us into amuch more complex world than the simplepicture of an English occupation imposing awholly new, regular pattern of organisationand settlement on the land.

From the lands of the Scottish Bissets in Ulsterto the lands of South-West Scotland is only ashort sail. A general map of mottes inScotland was produced (Stell, 1975) ratherbefore a reliable one was produced forIreland. The studies of mottes in the twocountries share a number of general issues.In both, castles, typified by mottes, feature asa key element in the imposition of a “feudal”English lordship culture, along with charteredboroughs and Church reform. In Ireland thiswas attributed totally to military conquest butin Scotland it was seen as having beenintroduced by the native Kings, notably DavidI, which meant that it was more acceptable.The resulting merger of the two lordships,Gaelic and Anglo-Norman, in Scotland wasseen as coming from the “native” lordsaccepting the incoming system, rather thanof compromise by the Anglo-Normans toaccommodate local conditions. One problemof explanation that they shared was in thegaps in the overall distribution. In Ireland, thelack of mottes in the south challenged theassertion that they were a necessary part ofthe military conquest of the land. In Scotland,the absence of mottes in Lothian in particularbut also in other regions posed a similarproblem (Simpson and Webster 1972). Thiswas where royal power was strong andshould, therefore, have seen David’s scheme



Fig.3. Map of mottes in Scotland (Stell 1975)



imposed most clearly, leaving a goodscattering of mottes for us to see. The ideaof a systematic imposition of Anglo-Normansettlement scheme has been recentlychallenged in favour of a narrative which isless sudden and confrontational and morevaried (Oram, 2006).

The main concentration of mottes is in thesouth-west of the country, facing the coast ofUlster. Two detailed regional studies carriedout within the area: in the upper Clyde valley(Tabraham 1977-8) and in Galloway (Tabra-ham 1984). In both there was the same closerelationship between parish and manor thatwe saw in Ireland. The settlement along theupper Clyde involved the establishing of elevenmanors and parishes, all granted to Flemingsin what is described as a simultaneous act oforganising the landscape in the new order ofthe early 12th century. Although the unifiedadministrative organisation of motte, manorand parish is symmetrical and visible in nearlyall the manors, the actual settlement is morediverse. Only at two manors, Biggar andCarnwath, are the present village settlementsand the mottes demonstrably close to eachother, while at five they are clearly separated.The churches at these five are found with thevillages. The mottes, therefore, appear tooccupy a separate site; whether because it wasan older centre, or because the lords wantedto distance themselves from the peasantry orbecause the village has migrated away, wecannot tell. The point is that the conjoinedpackage of village, church and motte is not thenorm. In Galloway, likewise, the majority ofmanors have a motte at the site of the mano-rial caput. Oram (2000, 228-9), however, haspointed out that they do not correspond to asingle phase of simultaneous construction ofmanorial mottes, but stretch over a longerperiod; at Borgue a second site may have beenadded well into the 13th century, and that byan English lord. However, five other parishes

have more than one site: a second motte atAnwoth and Dalry; rock-cut sites at Parton andTwynholm. The latter two are particularlyreminiscent of older, less important centres,up-graded in the new style, within the manorswhich also had mottes at their capita. Oram’sdescription of the settlement of Galloway in the12-14th centuries (2000, 253-60) gives a picturefamiliar from the Irish evidence: “the normalpattern of rural society in Galloway was one ofdispersed settlement in a number small touns”.More nucleated settlements at manorial capitamight be either imposed on these small tounsor grow out from them but their scale remainedmodest, a handful of houses (perhaps six tonine) informally laid out at the site.

In terms of a legacy in the two countries, didthe mottes that had marked many of themanorial sites in Ireland and Scotland continueto be focal points even if the mottes might notbe actively used after the crisis of the 14thcentury ? The obvious way that this could bedemonstrated would be if the tower houses,which became the equivalent manorialstructures of the 15th century, were builtbeside the old mottes. To investigate thisquestion, we need to look where mottes andtowers are found in the same areas: in Irelandwe can pick out Cos. Meath, Louth and easternDown. In Meath and Louth the towers arerarely sited near mottes. From thedocumentary evidence, we can see that themottes occupy sites which were ancient seatsof power and they continue to be so, usuallywithout towers. By contrast, the towersoccupy sites which are new centres, often ofnew families who have gained land in theperiod 1350-1450. As the mottes celebratedthe winners of the 12th century, so the towerscelebrate those of the 15th. East Downpresents a different narrative. The motteswere sited at the manorial centres on the goodarable land. By contrast, the towers arefocused on landing places along the coastline,



principally of Strangford Lough, reflecting achange in the source of wealth. In the 13thcentury arable farming provided the profit; inthe 15th century it was trade in hidesespecially which was the main earner,organised through the newly prosperous portof Ardglass. The power centres moved to theshore to take advantage of the newconditions. The overall pattern of towerbuilding in Scotland certainly contrasts withthat of mottes. The absence of mottes inLothian disappears and the distribution oftowers reflects the distribution of thecountry’s wealth much better than did themottes (Dunbar, 1975).

In both Ireland and Scotland the study of theMiddle Ages is very strongly divided by theadvent of Anglo-Norman lords, which is seen asa major threshold in the history of the countries,between Early and Late Medieval periods. Itmay be seen as a disaster by Irish nationalistsor as a positive initiative leading to the creationof a strong kingdom in Scotland but the impactis seen in both as massive, involving theintroduction of a new social and economicorder, and a new political regime. This isfounded on a feudal attitude to land, withmanors and castles and borough townsarticulating trade, and on a new organisation ofthe Church, with new monasteries and anetwork of parishes connected to the manors.Mottes are seen as the archetypical early castlesand a crucial part of this revolution. When welook at them in detail, however, what we findis that mottes in 12-13th century Ireland andScotland, even when attached to domesticrather than military sites, and in the lesserestates, are not a single thing. They are not partof an indissoluble package of manor andnucleated settlement combined into a system.Along with the establishment of parishes andthe founding of towns, they are elements whichmight be deployed in combination with theothers or independently of them. Whether he

would have a motte or not was a choice for thelord concerned; along with a dependentmoated enclosure (a bailey) or an enclosurewithout a motte (a ringwork or a moated site)it might be up to the whim of the lord or thecustom of the barony in which he held hismanor. Like a tower, the presence of a mottehas no prescribed purpose, except display; itwas flexible and adaptable to individualcircumstances. The package did not need toarrive as a single event; castle, town ormonastery could come at different times, asindividual items, or not at all.Two different schemes, based on two differ-ent chronologies, emerge from the study ofmottes in Ireland and Scotland: a short and along chronology. The short chronology iscomes from the idea of a single package,associated with the idea of a single introduc-tion. In Scotland this is brought about throughthe policy of David I; in Ireland through mili-tary conquest. The belief in a sudden intro-duction reinforces the idea of a coherentprogramme for change; the two ideas reflectand reinforce each other. In Ireland this hasmade the castles into military structureswhich were vital to the conquest. Opposedto this is the idea of a long chronology whichemphasises the amount of co-operation andcompromise that went on. The new lords inIreland and Scotland built their estates out ofthe existing settlement patterns and adaptedthem to local conditions. At the same time,local lords adopted Anglo-Norman ways. Thecastles, as identified by the mottes, show thistwo-way process clearly. Understanding thisis helped by realising that a motte is not acastle in itself but is the part of one that hassurvived the best, structures put up to pointout particular houses in the landscape. Mottestudies become settlement studies.How do the issues raised in Ireland and Scotland,in particular substituting a long chronology, affectthe study of mottes in England and Wales, or



even France ? Their studies also deploy theexamination, usually through excavation, ofindividual mottes and castles. It started from thebelief that mottes were a type of site only tofounder on results which were so varied as todeny that idea. Alongside it went the mappingof mottes and the study of their distribution,notably by Ivens (1984), Pounds (1990, especiallychapter 3) and Butler (1994). As we have seen,it is in the gaps in the distributions that much ofthe interest of these studies lies. The earlier workwas based very strongly on a short chronology,like the one deployed in Ireland and Scotland. Itwas one of the fatal flaws in R. A. Brown’s viewin his dispute with Davison over “the origins ofthe castle in England” (note that castle is in thesingular): “to argue for castles in pre-ConquestEngland....is to argue for pre-Conquest Englishfeudalism” (1967, 137).

Both he and Davison identified mottes as thefirst castles in England. Beyond this contro-versy, a short chronology for mottes in Englandstresses the same ideas as in Ireland and Scot-land: a sudden deployment of mottes and anessentially military role for their use. The intro-duction occurs after 1066 and the military roleis seen in the use of mottes in the war betweenStephen and Matilda; the construction ofmottes is often attributed to either of theseevents, even confined to them. However, thiswill not stand up to the evidence from Scotlandor Ireland. Mottes were deployed in thesecountries through the 12th and into the 13thcentury by magnates who came from England.We may presume that they did so because theywere used to seeing them in use where theycame from and saw value in them, which mustextend their effective life in England as well, intothe 13th century. The same idea of a longchronology should apply to England as well.Along with this goes a greater stress on themas being not purely military features, but morelike towers, structures which added primarilyprestige to their sites but which might have

been used for different purposes. Motte studieshave tended to tail off in the last twenty yearsoutside Ireland, where they have been kept goingby the advances in archaeological survey andmapping of sites in general, but also, of course,because of the controversy over the idea of theNorman Conquest. In part this relative neglectresults from the Lewis Warren’s maxim thatnothing kills off a topic as effectively as a goodbook on it, in this case Higham and Barker’s. Italso comes more, however, from the corralling ofmottes into a short chronology and a militarynarrative, which runs counter to the whole trendof castle studies in recent decades. They need tobe re-absorbed into castle studies and theirgeneral narrative, as the features they were, justas towers have been liberated from their role asmilitary keeps. After that, the message needs tobe communicated to the world beyond the stu-dents of castles, in schools and in descriptions bypublic authorities, as seen in the example ofDroagh motte in the appendix.

BibliographyAnderson, S. and Rees, A. R. 2004 ‘Theexcavation of a medieval rural settlement atPortmuck, Island Magee, County Antrim’, UlsterJournal of Archaeology, 63, 76-113Brown, R. A. 1967 ‘An historian’s approach tothe origins of the castle in England’,Archaeological Journal, 126, 131-48.Butler, L. 1994 The origins of the honour of Rich-mond and its castles, Château Gaillard, 16, 69-80.Childe, V.G. 1938 Doonmore, a castle moundnear Fair Head, Co. Antrim, Ulster Journal ofArchaeology, 1, 122-35Davison, B. K. 1961-2 Excavations at Ballynarry, Co.Down, Ulster Journal of Archaeology, 24-5, 39-87.Dunbar, J. G. 1975 Tower houses, in McNeill &Nicholson, 1975, 71-2, map 82.Higham, R. & Barker, P. 1992 Timber Castles, Batsford.Ivens, R. J. 1984 Deddington castle, Oxfordshire, andthe English Honour of Bayeux, Oxoniensia, 49, 101-19.McNeill, P. and Nicholson, R. (eds) 1975 AnHistorical Atlas of Scotland c.400-c.1600, University



Appendix: The motte at Droagh, Co. Antrim

of St Andrews Press. Expanded and reissued asMcNeill, P.G.B. and MacQueen, H.L. (eds) 1996Atlas of Scottish History to 1707, Edinburgh: TheScottish Medievalists and Department ofGeography, University of Edinburgh, laterpublished online at:https://scotlandsplaces.gov.uk/digital-volumes/published-gazetteers-and-atlases/atlas-scottish-history-1707/atlas-scottish-history-1707McNeill, T. E. 2011 ‘Mountains or molehills? Differ-ent uses for mottes in the Lordships of EasternIreland’, Archaeological Journal, 168, 227-271.McNeill, T. E. 2019-20 ‘The medieval settlementsite at Killyglen, County Antrim’, Ulster Journalof Archaeology, 75, 70-93McSparron, C. and Williams, B. 2009 ‘Theexcavation of an Early Christian rath with latermedieval occupation at Drumadoon, Co, Antrim’,Proceedings of the Royal Irish Academy, 109C,105-64Oram, R. 2000 The Lordship of Galloway, John Donald.Oram, R. 2006 ‘Castles and colonists in Twelfth-and Thirteenth-Century Scotland: the case ofMoray’, Château Gaillard, 22, 289-98.Pounds, N. J. G. 1990 The medieval castle inEngland and Wales, Cambridge University Press.Simpson, G. G. and Webster, B. 1972 ‘Charterevidence and the distribution of mottes inScotland’,Château Gaillard, 5, 175-92. Reprintedwith minimal changes in Stringer, K.J. (ed.) 1985Essays on the Nobility of Medieval Scotland(Edinburgh: John Donald), 1-24.Stell, G. Mottes in McNeill & Nicholson, 1975,28-9, map 22.Tabraham, C. J. 1997-8 ‘Norman Settlement inUpper Clydesdale’, Transactions of theDumfriesshire and Galloway Natural History andAntiquarian Society, 53, 114-128.Tabraham, C. J. 1984 ‘Norman settlement inGalloway: recent fieldwork in the Stewartry’,Breeze, D. (ed.) Studies in Scottish Antiquity, JohnDonald, pp. 87-124.Waterman, D. M. 1959 ‘Excavations at Lismahon,Co. Down’, Medieval Archaeology, 3, 139-76.


Figs. 4-5. This notice, erected at the site by someone,gives a good picture of the public knowledge aboutmottes and earthwork castles in general. Of thestatements in this notice, including thereconstructed drawing (note the angle of theearthwork) there is only one which might bedefended as probably true. That is that “mottessometimes had a bailey extending from the base ofthe mound”, in Ulster only about 20% had. All therest are either definitely untrue or dubious.

an irish view of scottish mottes journal 2020-21rev4... · 2020. 12. 17. · an irish view of...

Documents