Volume II | Chapter 11 | Viking-Period FormsNext Back to catalogue index
by R.N.Bailey

HOGBACKS (G.I., p. xxi, figs. 5–7)

The fifteen hogbacks from the Cumbrian peninsula share decorative details and variations in form with monuments to the east of the Pennines and in Scotland (Lang 1972–4; idem 1984). Thus the shape of Kirkby Stephen 8 is best paralleled at Ingleby Arncliffe and Lythe in Yorkshire; the pilaster arrangement of ornament on Lowther 6 and Aspatria 6 has analogues at Brompton and Wycliffe in Yorkshire and Inchcolm, Fife; and the running scroll which decorates the walls of Penrith 6 and 9, and Brigham 10, is a type of hogback ornament which is distributed as widely as Repton in Derbyshire, Crathorne in Yorkshire and York. There are, however, certain variations which are specifically Cumbrian.

Most distinctive is the tall thin section (Bailey 1980, 98). This is a feature shared by Appleby, Bromfield 4, Aspatria 6, all four hogbacks from Penrith, both hogbacks from Gosforth, and two of those (4 and 5) from Lowther as well as Bolton le Sands in Lancashire; Cross Canonby 5, in its restored form, betrays the same feature. Outside the area only Govan (Renfrewshire) in the Clyde valley approaches these proportions, and its 'stopped-plait' decoration significantly echoes a distinctively Cumbrian form of ornament (Lang 1972–4, pl. 14).

A second Cumbrian predilection is the preference for figural decoration on the sides of the stone. Both hogbacks from Gosforth, Lowther 4 and 5, Penrith 7, another from Heysham in Lancashire and (in a slightly different manner) Cross Canonby 5 all share this taste. It is otherwise only found at Sockburn in the Tees valley, co. Durham (Cramp 1984, pls. 138, 741; 139, 745; 146).

Within Cumbria there are smaller links between specific hogbacks. Cross Canonby 5 and Gosforth 5 are uniquely linked in the hogback series by their setting of the end-beast head on the ridge-line some distance from the gable end. An undulating snake running along the bottom of the wall underlies the ornament on Lowther 4 and 5 as well as others at Penrith and Cross Canonby, whilst the wall ornament is not clearly panelled off from its surround on three of the Penrith stones, on either of the Lowther hogbacks 4 and 5, or on Gosforth 5.

This whole class of monuments has been described as 'non-Christian in character' and 'thoroughly pagan in conception' (Smyth 1979, 273). These descriptions are unjustified. There is nothing 'pagan' about the form, which represents a modification of the Anglian shrine-tomb (Bailey 1980, 92–7). There is nothing 'pagan' either about their production, since they are carved by the same masons as carved the crosses; this is clear from the technical and ornamental evidence available at such sites as Aspatria and Gosforth. Nor need their end-beasts have any more pagan significance than the animals which perch in similar positions on the metalwork shrines of Christian Italy, Germany, or Ireland (ibid., 97–8).

What then of pagandom in decoration? Of the fifteen Cumbrian hogbacks three are ornamented with a vine scroll which, if it had a perceived significance, carried a Christian symbolism. The gable-end of Gosforth 5 is decorated with a Crucifixion, Lowther 6 has a cross on both gables, and the Cross Canonby stone has an orans. The snake combats of Gosforth 5 and Penrith 7, like the similar theme on the cross-shaft from Great Clifton, could carry a Christian meaning (Bailey 1980, 138–42). Most of the other members of this group are decorated with animals or interlace and this kind of ornament had long found a place in the most Christian of contexts in pre-Viking England: witness the Lindisfarne Gospels.

We are left with three stones whose decoration might more plausibly be interpreted as pagan: Gosforth 4, and Lowther 4 and 5. It must be admitted that these stones carry no overt Christian symbolism. Nor is there any doubt that their battle scenes can be paralleled in the pre-Christian art of Gotland – the themes no doubt inspired by models circulating in perishable media like tapestries, metalwork, wood carvings and shield paintings. There is, however, no reason to think that they are specifically 'pagan' as distinct from 'secular' statements. Like the warrior portraits of Yorkshire they can be interpreted as celebrating the secular, heroic ideals of the new patrons of sculpture. But these ideals had long existed in pre-Viking Christian Britain, for an interest in the deeds of a real or imagined past is well evidenced in the monasteries of Anglian England and on the continent (Wormald 1978, 48–52). It was reflected in the literature emanating from those monasteries and, in the Franks Casket, even achieved sculptured form. At the very core of Anglo-Saxon Christianity there was a fascination with the noble deeds of heroes. Within the Viking period, and under changed non-monastic patronage, this interest is now given sculptural expression. There is, however, no call to interpret this taste as one lying beyond the pale of medieval Christianity.


Among the more ambitious crosses of pre-Viking England were those whose shafts had the usual rectangular section at the top but whose lower parts were cylindrical. The type is sometimes difficult to distinguish, in its fragmentary survivals, from contemporary decorated columns but there is sufficient evidence, nevertheless, to suggest that the form was not limited to any particular part of the country. There are pre-Viking examples in Yorkshire at Dewsbury and Collingham (Collingwood 1907a, 296; idem 1915a, 156) and Cumbria provides both Dacre 1 and Beckermet St Bridget 1 (see above, p. 12).

In the Viking period this form of shaft continued in use. The geographical centre of this continuity lay in the Peak District (Kendrick 1941b; Pape 1945–6; Bu'lock 1960) where both decorated and derivative non-decorated versions are found, forming what Radford rightly considered as the '. . . impoverished continuity of a native tradition' (Radford 1961, 210).

The six Cumbrian examples need not be viewed as reflexes of this Peak District group, even though the broad collars of Penrith 4–5 and Beckermet St Bridget 2 can be closely paralleled at sites like Leek, and Chebsey (Staffordshire). The ornament at Penrith and Beckermet, for example, is of a kind unfamiliar in the Peak District but well evidenced around the Solway. Similarly, the organization of the decoration and the general proportions of Gosforth 1 (and the related Gosforth 7) make this a unique monument which owes nothing to the carvings of Derbyshire and Staffordshire. Workington 2, if it is indeed a round-shaft derivative, is equally linked to local west-coast groups and, in its use of a scalloped frame without collar, may well have been influenced by Gosforth 1 which is the only other carving in England to have this particular form. The Cumbrian and Peak District groups are therefore probably separate, if parallel, developments from an established Anglian type of monument. The fact that a Viking-age round-shaft derivative, Beckermet St Bridget 2, stands alongside an Anglian example of the same type, further strengthens this argument for a separate local development.

An associated group of carvings, again found in variant forms all over the north, is made up of crosses which have an orthodox rectangular-sectioned shaft but combine this with ornament proper to round-shaft derivatives. In Cumbria this influence of the one type on the other takes various forms. The raised collars of the circle-heads Rockcliffe 1 and Bromfield 2 probably derive from this source. More certainly, the shape of Penrith 2 so exactly echoes in its profile the adjacent Penrith 5 that its source of inspiration cannot be doubted. The remains of an angular moulding on Kirkby Stephen 1 also seem to represent a north-western adaptation of a round-shaft derivative type of decoration; it can be seen in more elaborate form further south at Whalley and Bolton le Moors, Lancashire, (Brown 1937, pl. ciii) and had already been used earlier at Sandbach, Cheshire (Radford 1957).


The most significant feature of the forty or so cross-heads which belong to the tenth or eleventh centuries is that over half are free-armed and thus in an Anglian tradition. Among these the dominant type is that employed by the spiral-scroll school which has markedly dumpy proportions with short arms and a large boss occupying the whole of the centre of the cross-head. Two other forms, characteristic of the north-west, require more extended treatment: the hammerhead and the circle-head.

HAMMERHEADS (G.I., fig. 2, type A5; Fig. 6a)

The form which Collingwood christened 'hammerhead' is one in which the end of the upper arm is broadened to the width of the transverse arms. In Cumbria the best examples are Carlisle 4 and its ring-headed equivalent, Addingham 1. Closely associated, though with less exaggeration of the upper arm, are Brigham 6, Bromfield 1, Dearham 3 and Walton. There can be no doubt that this is a type whose distribution is concentrated in the north-west; the other known examples are at Gargrave and Middlesmoor in west Yorkshire and at Kilmorie in Wigtownshire (Collingwood 1927a, figs. 111–13). The origin of the form is not clear, though it is tempting to follow Collingwood's suggestion that it represents an exaggerated rendering of those cross-heads whose arm-ends were of oblong block shape (Collingwood 1915a, 279; idem 1927a, 86–7, 90–2). What makes this explanation particularly attractive is that two of the best examples of such oblong-block types come from the Anglian north-west at Carlisle and Lancaster (ibid., figs. 104, 137). What is more, the Kilmorie slab carries both a hammerhead and a form which is intermediate between that fully-developed class and heads of the Carlisle/Lancaster type. Whilst this remains the most likely background for this regional development it should be recorded that the oblong block type of cross-head was not restricted to the north-west; there is a fine example at Upleatham in north Yorkshire (Collingwood 1911a, 301).

CIRCLE-HEADS (G.I., fig. 3, 3)

A second distinctive sub-group of crosses is the circle-head type of which nine (perhaps twelve) examples survive in Cumbria from nine (or eight) sites (Aspatria 1; Bromfield 2–3; Dearham 1; Gilcrux; Irton 2; Muncaster 1; Rockcliffe; Workington 3; Aspatria 7(?); Beckermet St John 6(?); Workington 6(?); in general, see Bailey 1980, 177–82). These carvings are linked by the fact that the connecting ring forms a continuous circle which appears to overlie the arms. The ends of the arms consequently project like ears beyond the circle. The distribution of this type of cross-head is centred on the western seaboard between Cumbria and north Wales, with further outliers in Cornwall (Langdon 1896, 170–4; Collingwood 1928a; Bu'lock 1958; Bailey 1974a, i, 106–37; idem 1980, 177–82).

If we leave aside for a moment the Cornish examples and the solitary Yorkshire representative from Gargrave, then it is possible to distinguish between Cheshire and Cumbrian treatments of this type of head. In Cumbria the circle is always decorated with a continuous plait and set in a different plane to the arms. By contrast, in Cheshire the circle is ornamented with key patterns or pellets and is not so distinctively separated from the arms. With one exception (Rockcliffe) the Cumbrian spandrels are always pierced; the Cheshire group – again with one exception at Bromborough – lack this piercing and fill the space with a boss set in the arm-pit of the head. Rockcliffe, whilst using a spandrel boss, employs it in a manner imitative of Whithorn fashions and is thus quite distinct from the Cheshire group. In its turn Bromborough, though exhibiting pierced spandrels, preserves other Cheshire features.

The chronological relationship between the two groups is as yet uncertain but, since the outliers at Gargrave and in north Wales seem to draw upon both Cheshire and Chester characteristics, and since there are some links between individual stones in each set, it seems reasonable to assume that they flourished contemporaneously with each other.

The Viking-age date of the Cumbrian group is not only clear from the use of a variant of the ringed form of head but from the decoration of this closely interlinked set of crosses. Ring-chain forms are used on Dearham 1 and Bromfield 2, Jellinge-influenced beasts appear on Gilcrux and Rockcliffe (and perhaps on Aspatria 1), whilst Workington 3 seems to carry one of the few examples of Mammen art in England. On this basis the likelihood is that the group flourished in the tenth rather than the eleventh century.

The shape is a north-western invention. The well known 'Cross of the Scriptures' (Henry 1967, pl. 90) at Clonmacnoise (co. Offaly) has a similar type of continuous circle (partly masked by disk ornament) and there is certainly an Irish taste for spandrel bossing before, during, and after the Viking period which might be seen as relevant to the Cheshire details. Yet, whilst the original idea may have been influenced by such Irish crosses, the fully developed circle-head is a phenomenon which is restricted to the north-west coastal strip.

Within the Cumbrian group certain crosses are closely linked together. The shafts at Bromfield and Rockcliffe are uniquely connected in their collars and relatively square sections as well as in their use of a ring-chain derivative (G.I., fig. 26, type cv), idiosyncratic interlace patterns and the disposition of their zoomorphic ornament. Aspatria 1, Dearham 1 and Muncaster 1 share a similar panel layout of the shaft. Aspatria 1 shares with Melling and Lancaster the use of a ring-knot with multiple concentric circles, one of which is formed by a line of pellets. Since the Lancaster example also employs a curious knot-pattern otherwise only known at Aspatria and Bromfield (and there are similarities between Aspatria and Melling in the handling of the interlace) it seems reasonable to associate this north Lancashire pair with the Cumbrian set.

Next Back to catalogue index