The beginnings of Scottish symbol stone art

Sculptured stones of the early historic period were classified by Joseph Anderson for his Rhind lectures of 1892 into three principal classes, a system that was adopted in the publication with Romilly Allen of Early Christian Monuments of Scotland (Allen and Anderson, 1903). The three classes have since been taken as broadly successive chronologically, if overlapping in their currency, from at least the seventh to tenth centuries ad. The basis of classification involved several criteria. Class I stones were undressed boulders, often of irregular shape, and their symbols, comprising animal or abstract motifs, were incised and relatively simple. Class II stones were cross-slabs, the central cross on one side generally imposing a degree of symmetry in the shape of the stone, with the cross and its surrounding designs or motifs carved in relief. The animal and abstract symbolism was evidently assimilated into the Christian iconography, though it never actually intrudes upon the outline of the cross. In the more elaborate cross-slabs, other biblical imagery such as Daniel in the lions' den or David in various roles fill the surrounding panels, while hunting or battle scenes may have had a secular or religious significance. On the Class III stones, the original symbolism disappears, and these are assumed to be among the latest, extending into the ninth century and beyond. That a system of classification should have survived in general use for more than a century may be surprising, even allowing for modifications like Henderson's (Henderson, I. 1987) addition of a Class IV to include early incised cross-slabs.

Along with the system of classification, it is arguable whether we should link the adjective 'Pictish' to symbol stone art, as has been the convention for more than a hundred years. Daniel Wilson (1851, 499) was among the first to attribute symbol stones to the Picts, but in doing so he was essentially arguing a native origin in preference to their introduction by Dalriadic Gaels or Norse invaders. Joseph Anderson in his Rhind Lectures (1881) had not made this equation, referring instead to symbol stone art as late Celtic, and only adopting a Pictish association in his later work with Romilly Allen. In the generalized sense of the Roman sources that implies that Picti were the totality of native communities north of the frontier, then symbol stones are indeed part of the Pictish cultural world. But attempts to identify a Pictish 'heartland' on the basis of archaeological distributions, whether of symbol stones, souterrains, long-cist cemeteries, settlements and fortifications, silver chains or other distinctive material types, have invariably foundered on the diverse geographical distributions and chronology of these classes of field monument or artefact types (Harding, 2004). Even the distribution of Class I symbol stones, generally acknowledged as the earliest, is not primarily concentrated in eastern Scotland so much as north of the Dee to Caithness and Orkney, with a few outliers to the west. So while symbol stones may be regarded as 'Pictish' in the sense of 'later Iron Age beyond the Roman frontier', any equation with an ethnic entity identified with later king-lists and their implied territories is probably best suspended as unproven.

Class I stones (Figure 11.5) do still appear to include the earliest, but the probability exists that this category includes examples from a considerable span of time. Unlike the cross-slabs, where the layout implies a degree of unitary composition, the placing of symbols on Class I stones gives no indication of necessary contemporaneity, and by analogy with earlier Iron Age metal-work we might expect that some of these carvings were composite achievements over time. The Inchyra, Perthshire, stone (Stevenson, 1959) is a clear demonstration of compound composition. There are three sets of symbols, two on one side and one on the other side of a truncated stone that in its final phase of use served as the capstone for an extended inhuman burial of uncertain date. The stone also bears ogham inscriptions, again evidently of more than one phase, and not necessarily in sequence with the succession of symbol carvings. Indeed, Stevenson's sequence, though plausible, is not definitive, and the span of time represented by this multiple re-use could extend backwards as well as forwards beyond the conservative time-span normally accorded to Class I stones.

The transition from Class I to Class II might also be expected to show intermediate variants. Though the cross-slab format would hardly have been adaptable to some Class I pillar stones, the two Glamis slabs could have been examples of re-used stones. In each case what becomes the back of the Class II cross-slab has irregularly disposed symbols that are simpler and less formally arranged that their successors on Class II slabs. In the case of the Glamis Manse cross-slab, Ritchie (Ritchie, A., 1989, 32) accepted that this was a case of re-use of an earlier stone, and demonstrated furthermore the experimental nature of the carvings on the cross-slab itself, a view that was echoed by the Hendersons (2004, 70—1, 83). If the cross-slab itself is early in the sequence, then its predecessor could easily have originated in the sixth rather than the seventh century. The Pabbay stone apparently represents the Christianizing of an earlier symbol stone, given the compressed field that the cross-carving occupies, and the fact that its arms are carved deeper than the adjacent symbols. In this instance it represents the combination of Class I symbols with what Thomas (1971) referred to as 'primary cross-marked stones', or Henderson's (Henderson, I., 1987) Class IV. Primary cross-marked stones in Ireland and Western Britain date from the sixth century, so that the Pabbay stone could again indicate the use of symbols by that time.

The dominant motifs of Class I stones are either abstract symbols or animal images. Human depiction or narrative scenes are a development of Class II. Classification imposes an artificial impression of standardization that is not the reality, and many of the abstract symbols defy concise description. Crescents, arcs, discs, double or multiple discs, discs with appendages, and various rectilinear shapes all feature on Class I stones. Some symbols have been viewed as representations of artefacts, either from contemporary material culture or inherited from earlier sources, whether Roman or native. The mirror has been claimed as based on actual artefacts that were current in the later pre-Roman Iron Age, while the comb, commonly linked with it, resembles later single-sided variants, both being passably realistic representations. Another symbol might represent a disc-shaped brooch, though matching its type exactly is not so easy.

Celtic Art Symbols

Figure 11.5 Symbol stone art. 1, Aberlemno, Angus; 2, Dunnichen, Angus; 3, Dyce, Aberdeen;

4, Newton House, Culsalmond, Aberdeenshire. Adapted from Allen and Anderson (1903).

Figure 11.5 Symbol stone art. 1, Aberlemno, Angus; 2, Dunnichen, Angus; 3, Dyce, Aberdeen;

4, Newton House, Culsalmond, Aberdeenshire. Adapted from Allen and Anderson (1903).

The conjoined double disc bears some resemblance to the central spine and terminal roundels of early Iron Age shields like Witham and Battersea, an observation in which I was anticipated by some seventy years by Sir Alfred Clapham (1934). In fact, the U-notched feature on the Inchyra and Newton House, Aberdeenshire (Figure 11.5, 4) stones echoes a much older (and likewise unexplained) association with shields of the later Bronze Age. The 'tuning-fork' and its more elaborate variants are not readily explained as representations of any known artefact. In any event, this is fundamentally irrelevant, since it has never been claimed that Celtic art derived its symbolic imagery from everyday artefacts, whether contemporary or reminiscent of a bygone era.

The V-rod and Z-rod present particular problems, since neither is a motif known from earlier times. Charles Thomas (1984) suggested that the V-rod was a broken arrow, and following the same line of reasoning the Z-rod could be thought of as a broken spear. Swords ritually bent in funerary contexts are widely known in Continental Europe, and in the Celtiberian world, where spears include the all-metal soliferreum type, as we have seen, these are commonly bent into three in graves (Lorrio, 1997, Figs 63, 64). The same practice would be largely undetectable where the spear shaft was made of wood, but bending double or treble, or breaking the shaft altogether, is simply the further compression of the token V-rod or Z-rod. Stevenson's (1993) parallel between the terminals of the V-rods with their fleur-de-lys and embellished terminals and representations in the Book of Kells makes a convincing case for regarding the eighth-century depictions as sceptres, but this need not mean that this was the function or meaning of the symbol in origin. Quite evidently 'pagan' symbolism was adopted and subsumed within the Christian iconography, so there is no need to assume constancy or continuity of meaning in the transition. On the other hand, the paired V-rods could have represented divination rods of the kind found in the so-called doctor's grave at Stanway, hence symbolizing magical or super-human powers.

It would be hard to sustain any 'Ultimate La Tène' influence in the abstract symbolism itself of the Class I stones, though a tentative case might be advanced on the basis of the filler elements of some symbols. Among abstract filler motifs, peltae and trumpet-spirals and S-scrolls most obviously recall an earlier tradition. Stevenson's chart (1955) showed a range of crescents with V-rod, the one motif that is regularly infilled rather than simply being depicted in outline. Some of the more complex examples, especially those with infilling, and indeed some of the more striking parallels with earlier La Tène styles, however, are from the later Class 2 cross-slabs. Reminiscent of earlier Celtic examples like the Besançon flagon or the Cerrig-y-Drudion bronzes is the fleshy vegetal style on the back of the Skinnet, Caithness, cross-slab, which has on its front side a fine dragon-pair that has acquired interlace manes. Likewise the Ulbster, Caithness, cross-slab has the same motif executed with the yin-yang technique of earlier La Tène art. Among Class 1 stones with scrolls the S-chains on the Dunnichen stone (Figure 11.5, 2) perhaps most closely might be paralleled in the early La Tène styles.

Spirals and interlocking spirals, too, are better represented in increasingly complex form together with interlaced ornament as background filling on the later cross-slabs. These motifs have much in common with background designs on illuminated manuscripts, as has been recognized since Romilly Allen and Joseph Anderson's pioneering catalogue, and must derive either from manuscript art or from a common pool of ornamental themes that provided the source for both. This is not to deny an ultimate debt to the pool of motifs and imagery of earlier Celtic art, but the derivation is thus at best very indirect and secondary. In its simplest form, the spiral is one of the motifs that has been regarded as one of several 'diagnostic endogenous shapes' (Dronfield, 1995), so that we would hardly claim, for example, any meaningful link between its occurrence in Celtic art and its use in megalithic art of late Neolithic Irish passage graves. But it may be worth observing that in application on the Class 2 stones, the complex spiral designs employ both of the principal forms of symmetry that can be identified in earlier Iron Age art, namely, rotational symmetry, as on one of the Nigg panels and bosses, and fold-over symmetry, as represented at Hilton of Cadboll, Shandwick and Meigle 4 and 9, for example. This may well result from the way in which the template for the design has been constructed based upon compass-work or a grid framework, as has been widely recognized in the illuminated manuscripts since Allen's pioneer study. While some survival of artistic conventions from an earlier period might be suggested, therefore, there really does seem to have been a renewed impulse into the repertory of the later seventh and eighth centuries.

Alcock (1993; 2003) stressed that, with the exception of the 'elephant' or 'beast', all the animal and bird representations from symbol stone art are remarkably naturalistic and evidently drawn from observation of the reality. It follows therefore that it is unnecessary to try to read into these symbols origins in earlier Celtic art, and equally unnecessary to imply any meaningful link between naturalistic animals in insular art and the Eurasiatic Iron Age art of the Altai (Thomas, 1961). It is worth noting, however, that the most frequently represented images, fish, snakes and eagles, are not the most characteristic of earlier La Tene art; earlier favourites, bull, boar and horse, are represented, but are in the minority in surviving examples. Dogs, sheep or goats do not register in the Class I repertory, so that animal symbolism is not simply a reflection of everyday life. The reason for the limited menagerie is not easy to explain. The combination of animals with abstract symbols certainly argues for a 'symbolic' meaning, perhaps linked to group identity or ritual associations, which is not to gainsay the view that 'a coherent symbolic language is hard to sustain' (Hicks, 1993, 199).

If animals were drawn from life, closely observed and naturalistic, then the 'beast' or 'elephant' is harder to explain as an aberrant dolphin or any other natural breed. Given the variety of fantastic beasts that occur on the developed Class II slabs, and the long antecedent history in Celtic art of fabulous animals, there seems to be no good reason why the fabulous should not appear alongside the naturalistic in symbol stone art. R. B. K. Stevenson (1955) and Charles Thomas (1961) each saw a quite different pedigree for the 'Pictish elephant', Stevenson citing Anglo-Saxon artefacts, illuminated manuscript art and Early Christian Irish artefacts, Thomas looking to early La Tene art and even its 'orientalizing' elements. Neither seem especially convincing, other than in the generalized sense of being comparative examples of fantastic beasts. More recent opinion (Hicks, 1993) has revived an older view (Clapham, 1934) that the 'elephant' derives from later Roman traditions. The 'orientalizing' theme was taken up by Thomas again in his pedigree of the Pictish S-dragon, which he relates to the dragon-pairs that characterize middle La Tene swords from Hungary westwards to the Thames. This comparison carries greater conviction, though the only Pictish opposed pairs, as opposed to solo sea-horses, are on Class II stones, and, assuming Class II stones are later than Class I, the route by which this particular theme entered symbol stone art is far from clear.

In fact, ducks, eagles, boars, deer, wolves and bulls tend to look broadly similar wherever they are depicted naturalistically, which is why dragon-pairs or fantastic beasts are perhaps more indicative of common traditions. Centaurs, hippocamps and voracious beasts are all absorbed into early Celtic art, and more particularly into the rather specialized and esoteric art of late pre-Roman Iron Age coins, but since they are obviously not exclusive to early Celtic art, we cannot assert that their appearance in the context of symbol stone art derives from continuity of that tradition. In fact, the appearance of these images on Class II symbol stones argues for the fusion of several different influences.

A major issue of contention has been the dating of symbol stones and symbol stone art. Robert Stevenson's enduring determination (1971; 1993) that symbol stone and cross-slab art was directly prompted by manuscript art has had a lasting and detrimental effect on the dating of insular sculpture. While there are some striking parallels, notably with Class II sculpture, it remains improbable that a widespread tradition of sculptural art should have been triggered by such a limited if influential set of gospel manuscripts, as the Hendersons have cogently argued. Certainly the Hilton of Cadboll stone gives the impression of a page from an illuminated manuscript (Stevenson, 1955, 116) and certainly the lion on the Class II Papil cross-slab (Figure 11.6A, 1) bears a marked resemblance to St John's lion from the Book of Durrow (Figure 11.6A, 2). Other parallels that have been asserted confidently as evidence for dating seem far less compelling. But these are both relatively late in the sculpture sequence, so the direction of influence is immaterial to the question of the origins of symbol stone art. Whether per contra the manuscripts were influenced by sculpture is equally arguable. Hicks (1993) re-asserted the basic principle that representation in art was more likely to progress from naturalistic to more stylized than vice versa, and that from this perspective the animals in the manuscripts were more stylized than those on symbol stones. On balance, it seems more likely that all — sculpture, manuscripts, metal-working and vernacular arts — were all drawing upon a common pool of styles and imagery that certainly must have been current before the later seventh and eighth centuries.

Plainly the problem stems from the lack of unequivocal archaeological associations for the Class I symbol stones. The earliest example from a reliable archaeological context is the much-cited slab with crude double-disc motif from Pool, Sanday, Orkney (Hunter, 1990), which was re-used as a paving slab in a structure of late fifth- or early sixth-century date. This secondary context suggests that the original use may have been from the fifth century. At Burghead, the context of the bull plaques (Figure 11.6B) is uncertain, but radiocarbon dates indicate the probable use of the fort from around the fourth to sixth centuries. In effect, it seems likely that, freed from the shackles of art-historical dating on the basis of perceived stylistic parallels, the archaeological arguments for fifth-century origins would gain credibility.

Antiquarians have for generations been absorbed by the meaning of symbol stone art. Various interpretations have been proposed, focusing on the stone as a statement of identity, authority, territorial rights or marriage alliances. All of these remain possibilities, though the lack of discrete distributional patterns, despite recurrences in combinations of motifs, remains an impediment to straightforward interpretation. Perhaps we would be better concerned with the function of symbol stones. Several, like Garbeg, Inverness and Golspie, Sutherland, were located in proximity to burials, and it

Art Animal Realistic

Figure 11.6 Animal imagery in later Iron Age insular art. A: animals in sculpture and manuscripts compared. 1, lion from Class 2 cross-slab, Papil, Shetland; 2, lion symbol, Book of Durrow, f. 191v; 3, eagle carving, Knowe of Burrian, Orkney; 4, eagle of St John, Corpus Christi MS 197B, f.1; 5, wolf carving, Ardross, Inverness; 6, wolf from Book of Kells, f. 76v. B: Burghead, Moray, bull carvings. 7, Burghead No. 3; 8, Burghead, No. 5. Adapted from various sources.

Figure 11.6 Animal imagery in later Iron Age insular art. A: animals in sculpture and manuscripts compared. 1, lion from Class 2 cross-slab, Papil, Shetland; 2, lion symbol, Book of Durrow, f. 191v; 3, eagle carving, Knowe of Burrian, Orkney; 4, eagle of St John, Corpus Christi MS 197B, f.1; 5, wolf carving, Ardross, Inverness; 6, wolf from Book of Kells, f. 76v. B: Burghead, Moray, bull carvings. 7, Burghead No. 3; 8, Burghead, No. 5. Adapted from various sources.

is likely that some at any rate served as memorials or funerary stele in a tradition known elsewhere in Britain from the late Roman and sub-Roman period (Thomas, 1984). Indeed, Alcock (2003) saw the pairing of symbols as a 'Pictish' version of a filiation formula cognate to the memorial formulae of Britain and Ireland. Though the symbols may be unique, there would be nothing exceptional in the practice of erecting stones as memorials.

0 0


  • ermias
    Is scottish art similar to celtic?
    7 years ago

Post a comment