Living on the Edge of Empire

Chapter 5

The Very Edge

Reappraising Romano-British Settlement in theCentral Pennines; the Littondale Experience

Keith Maude

ittondale is an unspoilt, remote, Yorkshire.Dale situated on the west side of Wharfedale

f ^ in the central Pennines. It is about 12.8 kmlong, with four small villages, two pubs and two postoffices. The remoteness of the Dale means thatprovisions must be purchased at the small towns ofGrassington or Settle each some twelve miles away,although in the past a greater degree of self sufficiencywould have been necessary, with the selling ofmanufactured articles and agricultural produce carriedout at market.

During the Roman period the North of England wason the 'edge of Empire' whilst the Dales area, far fromthe centres of political authority was on the 'very edge'.In comparison to the North West of England however,the Dales area is very rich in archaeological remains inthe form of earthwork sites (King 1986), with Littondalein particular having sites along the full length of thevalley.

These earthworks have been assumed to be theremains of early farmsteads with their associated fieldsystems. They can be seen from ground level but arebest viewed cither from the hillcrests at either side ofthe dale or if possible from alight aircraft or helicopter.

Reference to Figs 1 & 2 (the latter being a map of thearchaeological sites in upper Littondale betweenArncliffe and FoxUp) reveals the extensive nature of theremains and shows that the dale had been sub-dividedinto fields long before the enclosures of the eighteenthcentury.

These earlier fields divide the lower land in the dalebetween the river and the dale sides, although in severalplaces the dividing boundaries extend on to the presentmoorland. There once existed a series of settlementslocated on the spring line, presumably in order to ensurean adequate supply of water, with field systems andtrackways running down into the valley bottom. Theentire system has been interpreted, on typologicalgroups, as belonging to several periods, traditionallyseen as probably the Iron Age and early Roman eras.However, a lack of systematic field work in what hasbeen regarded as a well studied area (Hartley & Fitts1988, 118 n36), means that this assumption has not beenproved.

Prior to 1989 only three of these earthwork sites hadbeen excavated. Two sites in Pen y Ghent Gill, a smallwesterly tributary valley, were excavated in the 1930'sby Dr A Raistrick and Mr W Bennett (Raistrick 1939 &

Fig 5.7: Aerial viewlooking south-east atthe junction betweenLittondale andWharfdale. Note theearthwork fieldsystems that runbetween the uplands onthe right and the riverSkirfare on the left.

42 Archaeology North West Vol 3 (Issue 13 for 1998)

Living on the Edge of Empire

I km

Fig 5.2: Detailed plot of the ancient field systems within upper Littondale bet\veen Arncliffe and FoxUp, derivedfrom new aerial and earthworks surveys undertaken by the University of Manchester.

1960). These were at Dawson Close where earthworksettlement sites are located on the widest shelf of thelimestone. From the evidence of the recovered potteryand metal work an Iron Age date was assigned to thesesites.

Leeds University Department of Archaeology carriedout the third excavation at Thornber Barns in 1968 onthe well defined circular farmstead with adjacent fieldsand enclosures (Fig 5.1). This site is located below thelimestone scarp at Scoska Wood which is now thelargest remaining ash wood in the Yorkshire Dales.Within the wood to the north west of the site is ScoskaCave from which human skeletal remains (presumed tobe prehistoric) were recovered in 1905. A Neolithic flintscraper and a Roman coin circa AD 270 were recoveredfrom the scree outside the cave. The excavation atThornber Barns recovered rotary quern fragments,Romano-British pottery and a coin of Antonius Pius.

Further archaeological work since 1989 has beencarried out in the Dale by the Department ofArchaeology, University of Manchester with theintention of mapping this landscape through new aerialand earthwork surveys (Fig 5.2), and to date the putativefarmstead sites through sample excavation. Two sampleexcavations were carried out on the sites at Halton Gilland one at New Ing Barn to the north west of Littonvillage (Maude 1990).

The Halton Gill Site

The field systems on the hillside to the south of HaltonGill can easily be seen, especially when the sun is lowin the sky. What is not apparent is that the field systemssurround a farmstead. Air photographs however revealthe presence of the farmstead which is now bisected bythe Halton Gill to Stainforth road (Fig 5.2).

Tracks or drove ways lead to the site and the longestlinear field boundary which runs from the river to thefell is aligned on the site. Further field boundaries runparallel to this main boundary on either side of thefarmstead giving fields of almost constant width. Theentire system gives the impression of a planned layoutas opposed to one which has grown haphazardly.

Excavation was carried out on the site in 1991consisted of a series of test trenches which were placedin order to investigate the main farmstead and itsadjoining small enclosures. The largest trench wasexcavated across the main lynchet just to the north andadjoining the farmstead. Topsoil was very shallow andimmediately overlay the natural limestone surface sothat stratigraphy was virtually non existent. At the topof the upper bank of the lynchet however a small wallhad been built which had retained a relatively thick layerof soil and this area contained most of the artefactswhich were recovered, the main dating evidence beingprovided by the sherds of Romano-British pottery dated

Archaeology Norlh West Vol 3 (Issue 13 for 1998) 43

Living on the Edge of Empire

Field Systemsat

Halton GillLittondale

THORNBER BARNS

HALTON GILL

NEW ING BARN

Fig 3.3: Earthwork plans of the three farmsteads excavated within Littondale.

44 Archaeology North West Vol 3 (Issue 13 for 1998)

Living on the Edge of Empire

to the third century AD. The sloping face of the lynchetwas revetted with water worn cobble which hadpresumably been obtained from the River Skirfarewhich is approximately 200 metres distant. Thisrevetting like the small wall suggests that an attemptwas being made to minimise down slope soil erosion.

A trench in one of the adjoining enclosures containedthe bottom half of a flat quern which reinforces the thirdcentury date and a undatable whetstone was recoveredfrom a trench in the fields 100 metres to the north west.What appeared to be a saddle quern was recovered fromthe nearby field wall and one piece of beaker potterywith cord decoration was also recovered. This artefactis of a Bronze Age date, it was however stratified withinRoman period levels and is therefore thought to havebeen brought on to the site during that period.

The dating evidence suggests that the site was afunctioning farmstead in the late third century AD.

The New Ing Barn Site

This site lies at the foot of the hill slope to the northwest of Litton village below the old road from Litton toPen y Ghent and consists of a series of contiguousrectangular enclosures on top of a small ridge (Fig 5.3;Maude 1992).

Excavation was carried out in 1992 and a similarstrategy to the one employed at Halton Gill was used.As expected a similar stratigraphic sequence wasencountered with very shallow topsoil overlying thelimestone. At New Ing, however, structural componentsof the farmstead comprising a flagged area and apossible wall were still in place. The flags were madefrom gritstone, being readily available higher on the fell,in contrast to the limestone of the site (Fig 5.4).

Again a series of artefacts were recovered;Romano-British pottery which included part of amortaria, rim sherds of Huntcliffe ware, buff and greywares, a small blue bead and a, coin (Antonianius)which would appear to be of Tetricus II (AD 273) butwhich is thought to be a forgery? This period is thoughtto have been one with high inflation rates and perhapsas a consequence most counterfeit Roman coins are ofthis period. Also from a counterfeiters standpoint theDales area is an ideal location to operate from. It isremote and has lead mines which are known to havebeen operating during the Roman period and hence apossible supply of metal which can be used to debasesilver coinage was available. These artefacts againsuggest a date for the farmstead in the late third to earlyfourth centuries. As at Halton Gill, part of a rotary quernwas recovered. In this case it was a broken half of anupper stone made from millstone grit. The quern, unlikethe one from Halton Gill, did not have any apparentwear on its grinding face and there was a naturallyoccurring fault visible in the broken section of the quern.It would therefore appear as if the quern had beenbroken during manufacture, which would suggest thatquerns were being made on the site.

Discussion

Aerial and earthwork surveys by the University ofManchester have shown that seven early farmsteadsexist within the Littondale earthwork complex, andthese area all sited on the south-western side of the RiverSkirfare between Arncliffe and Halton Gill. (Maude1990). These are located along or close to the spring linethus ensuring a supply of water during the summermonths at which time the river is consistently dry. Thissouth-western valley location also provides shelter fromthe prevailing winds and the effects of the wind chillfactor in winter, and is a drier location than thenorth-eastern side of the valley.

So far three of these farmsteads have been excavatedand dated and artefactual evidence has been recoveredwhich suggests that they were functioning in the latterhalf of the third century AD. Although the sites aretypologically dissimilar, they are all closely dated to thesame period, which suggests that the whole valleysystem may belong to one period. In this case a relianceon a purely typological approach which suggesteddifferent periods appears to be wrong.

During the third century AD the area would have beenmore remote than today because of slowcommunication. Richmond, as long ago as the 1920's,suggested that some sort of social disturbance hadoccurred in the north of England during the latter halfof the third century, this being detectable fromexcavation evidence which showed that Roman fortsfrom Brough-on-Noe in the south to Hadrian's Wall inthe north had been reoccupied in the late third century.Although this view is now unfashionable, the rebuildingwork being interpreted as a period of restoration afteran era of peace and perhaps neglect (Hartley & Fitts1988, 31), what is undeniable is that within the Dalesonly the fort at Bainbridge was occupied continuouslyduring the Roman period, whereas the other-Pennineforts at Elslack, Ilkley and Burrow-in-Lonsdale wereabandoned and then rebuilt during the late third century.This places Littondale firmly within the highlandmilitary zone of Roman Britain.

This added military presence may account for theincreased farming activity during the period due to theexpanded population and the subsequent injection ofmoney into the local economy. Coins have been foundon two of the sites and also represent a common findfrom caves within the Dales area. This would appear tosuggest that a monetary market system was in use in theDales rather than just subsistence farming. No doubt themonetary economy ran in parallel with an exchange/barter system, although the presence of professionallymade pottery on the sites may imply a greater degree ofintegration with the wider Empire economy thanpreviously thought.

Finds of querns on all the excavated sites impliesfarmsteads with grain being ground on site for domesticuse. It does not prove, however, that cereals were beinggrown, for whilst it has been shown that it is possible to

Archaeology North West Vol 3 (Issue 13 for 1998) 45

Living on the Edge of Empire

Fig 5.4: Aerialview of theHalton GillRomano-Britishfarmstead andits associatedfield system. Ex-cavations haveshown that thissite appears tobe late thirdcentury AD inorigin.

grow cereals in the Dales, the area has always beenagriculturally marginal, the valley bottom ranging fromc 240m to c 300m above sea level. What the finds doshow is that querns were an essential part of everyhouseholds 'consumer durables' and a closeethnographic parallel can be made with the farmingcommunities of Orkney and Shetland where, at the startof this century, each household had its own rotary quem.On these islands flour was kept in the more durable grainform and ground into flour as required.

From an arable agriculture aspect Littondale is andwas a marginal area on 'the edge of Empire', but itwould appear that when the economic climate was rightit was developed agriculturally to supply the boomingmarket provided by the Roman military. Markets atnative settlements must have existed, but the mainmarket focal points will have been at the Roman forts;both Bainbridge and Burrow are known to have had vi'ci.

The main produce from Littondalc is likely to havebeen livestock, cattle and sheep (as in the Middle Agesand today), which would have been taken to market aswool, hides or on the hoof. Journeys to market wouldhave involved several days walk and all the forts arewithin one day's walk, although Ilkley al a distance of30 km is probably too far. Archaeological research inLittondale suggests that in the late third and early fourth

centuries AD, the Dales was an area of muchagricultural and possibly industrial activity, whilstmarginal from a geographical standpoint.

Returning to our concept of'edge of Empire' and 'thevery edge', both terms are subjective and to some extentof our own making. They are viable if viewed from thecentripetal aspect of Empire from a military, politicaland economic standpoint. Considered in the regionalcontext, however, they fail to adequately explain thelocal evidence. What is clear is that large parts of theLittondale field system would appear to be functioningduring the later Roman period, and that in this valley atleast the use of typological dating techniques on thecurvilinear and rectilinear earthworks interpreted asfarmsteads does not work. In other words there is nosubstitution for sample excavation in order to testhypothesises arrived at from aerial and earthworkevidence.

What is required now is the application of this sampleexcavation programme on the other four farmstead siteswithin Littondale, and throughout the Dales, in order toestablish the date of the surviving ancient agriculturallandscapes of these valleys, and in particular to test thehypothesis of an increase in agricultural activity duringthe late third century AD.

46 Archaeology North West Vol 3 (Issue 13 for 1998)