Grass enclosure, beside unclassified road from Minions to Doublebois, about 1.5 km west of St Cleer, about 1.3 metres (4 feet) from St Cleer 3

First recorded 1600 in present location (Camden 1600, 155). Despite various vicissitudes (see Okasha 1993, 213), monument has remained there.

Monument broken but stable; ornament clear; text clear; situation good

One section of the shaft of a composite monument. The shaft tapers upwards and is markedly stepped out near the base, presumably where the lower, slightly rougher, part was intended to be buried in the ground. The top of the stone contains a rectangular socket, designed to take the tenon of a second section of cross-shaft. On each face of the shaft, within a broad incised edge-moulding, are single panels with interlace decoration and an inscription, the interlace being well laid out and executed in broad flat bands.

A (broad): This face contains only the inscribed text. The text is incised in five horizontal lines inside a panel measuring 66 cm high by 59 cm. It is complete and legible and uses a predominantly insular script, the letters measuring between 5 and 14 cm in height. The text reads:

B (narrow): The single panel on this side carries closed circuit four-strand plait.

C (broad): This face has a panel of four symmetrically-placed knots, each consisting of two rings interlaced diagonally.

D (narrow): This incomplete panel carries the remains of closed circuit turned pattern B.

The monument is the bottom part of a composite cross, containing a socket into which further (missing) sections would have been added. As such, it is comparable to the composite monuments found in Wales, for example Llantwit Major 2 or 4 or Llandough in Glamorganshire (Redknap and Lewis 2007, 373–6, 382–9, 329–37). Like St Cleer 3 (the second stone on the site) it is a member of the Panelled Interlace group of sculpture (Chapter IX, p. 85).

The text reads: DONIERT : ROGAVIT PRO A[N]IMA, that is, 'Doniert requested [this cross or prayers] for [his] soul' (Ill. 55). Camden first suggested that Doniert was the ninth-century king Dumnarth or Dungarth (Camden 1607, 139; see Okasha 1993, 215). This identification is fully discussed below. From this clearly legible personal name, the stone is popularly known as the 'Doniert Stone'.

Dating this cross is difficult. In the inscription we have, on the face of it, a clear guide. Were the cross indeed commemorative of the documented Doniert who died in 875, then this memorial is likely to have been set up fairly soon after. However, other broadly similar monuments in other parts of the country tend to be assigned slightly later dates (see discussion of groups, Chapter IX, p. 86). This is because of the fact that the simple knots employed in the decorative scheme are generally assigned a 'Viking Age' date and because, although they can be found in pre-Viking Age work, they tend to occur more frequently in the later sculpture (see Bailey 1980, 71–2). Particularly clear and fine examples of two of the knots on the Doniert Stone can be seen, for example, on the characteristically Viking period hogbacks at Brompton in northern Yorkshire, dated to the first half of the tenth century (Lang 2001, 74–6, 78, ills. 82–5, 104, 106). In Wales, the cross-shaft at Llantwit Major (4) which contains rows of interlinked oval loops like the Doniert Stone, is assigned an early tenth-century date, even though this is somewhat at odds with the earlier dating suggested by the archaic letter forms in the monument's inscription (Redknap and Lewis 2007, 382–9, especially p. 387).

In favour of accepting the Doniert Stone as a late ninth-century monument is the fact that such an identity with a historical figure would be readily accepted elsewhere and even used to guide the local chronology for sculpture (see, for example, Edwards 2007, 115). Llantwit Major 4 and its simple knots, for which the inscriptions on their own might suggest a slightly earlier date than that which is actually assigned, may also help the case. Like a number of the Llantwit Major stones, the Doniert Stone has interlace carefully constructed on a grid by a sculptor with a good understanding of how this type of decoration works (unlike the 'free' interlace sometimes seen on later monuments). Finally, as this may be one of the earliest sculptured monuments in Cornwall, it may perhaps be regarded as experimental: the choice of simple knots may in part have been made in response to the toughness of the granite of which it is carved. However, one of the authors (E.O.) considers that the identification of Doniert on the stone with the documented Doniarth cannot be accepted as certain (see the further discussion below).

Although, to modern eyes, the Doniert Stone and the other stone on the site (St Cleer 3) appear to sit in the middle of nowhere, the location is in fact remarkably well chosen. The stones are close to Liskeard, a lys and probable seat of Doniert (see below). At St Cleer, the monument is actually far more visible than it would have been in Liskeard, since it lies on major east–west and north–south routeways linking Launceston and Callington with St Neot, Cardinham and Bodmin on the one hand and Liskeard to Camelford and the Tintagel area on the other. The east–west route is testified on various old maps, for example Moll's of 1724 (Quixley 1966, 35), while the northern part of the north–south route was described as a magnum iter plaustrorum 'great way of waggons' in the mid-thirteenth century, and as 'the roadway between Liskeard and Cambleford' in about 1696 (Henderson, C. and Coates 1928, 71; Holden et al. 2010, 313, 315). It is also above the ford over the River Fowey which both these routes would have used, and moreover is close to the boundary of St Cleer parish. Being a stopping point throughout the ages for almost all travellers into Cornwall may also help to explain why it is one of the earliest and most fully recorded of Cornish early medieval monuments.

Further Discussion of the Name (O.J.P.)

The form Doniert (on the stone) is a variant spelling of the name Dungarth which occurs under the year 875 (for 876) in the Welsh Annals (so MS. A, c. 1100; MS. B, thirteenth century, has another spelling, Dumnarth: Morris 1980, 48 and 90; Dumville 2002, 12–13). The original (British) form of the name was *Dubno-gartos (alternatively *Dumno-), etymologically 'deep-protection' or 'world-protection'. The variation between o and u in the spelling of the first element, Dubno-, is attested as early as the first century bc in Gaulish personal names: Domne-, Domni-, Domno-, Dubno-, Dumno- (Schmidt 1957, 195–6, 198, 199–200). There was also later variation between these vowels in Cornish and Welsh (though not rigidly distributed), making Doniert a typically Old Cornish form contrasting with Welsh Dungarth and Dumnarth (Jackson 1953, 272 and 274–5).

The spellings Don- and Dun- (but not the later variant Dumnarth, meaning 'Duvnarth') show loss of [v] (from earlier b/m) before n; this sporadic loss is paralleled elsewhere, and is discussed, in this and other names, by Jackson (1953, 421–2).

After n (and sometimes after other consonants too) the following g, standing at the beginning of the second element of a compound name, was vocalised to j], as shown in the form Doniert; thus the spelling Dungarth in the Annals actually meant 'Duniarth'. Spellings indicating this vocalisation occur in Breton already in the ninth century, so the Cornish form Doniert is unexceptional at that date (Jackson 1953, 439; Jackson 1967, 720–1); but archaic spellings with g also continued to be used, as shown in the Annals.

The least regular variation between the two forms of the name is the difference of spelling between er and ar. Original er often became ar in Brittonic under certain circumstances, for instance Germanus > Welsh Garmon (Jackson 1953, 280–1); but original ar (as in *Dubnogartos) would not normally be expected to be spelt er, as it is in the lapidary form Doniert. Two possible explanations for that spelling can be suggested. Jackson has pointed out that the vocalisation of -g-to -j- is attested in Breton chiefly before front vowels (i or e), in names such as Festgen or Festien, Moetgen and Moetien (all of the second half of the ninth century); he has compared the Old Cornish name spelt both as Gurient (Bodmin no. 27, second half of the tenth century) and more archaically as Wurgent (Bodmin no. 22, early eleventh century), possibly the same man but there is no particular reason to think so (Jackson 1967, 720–1; Förster 1930, 90 and 88). So perhaps the g, in becoming vocalised before a back vowel in the name 'Doniarth', caused the following a to be irregularly fronted to e so as to provide the more usual context for the preceding j] sound. Alternatively it might be that the spelling er in Doniert was a hypercorrection: someone who habitually wrote 'correct' er in words pronounced with ar, as in 'Garmon' written as German(us), might occasionally have substituted er even in a word where it was not etymological.

Finally, in the lapidary form Doniert the final -t stands for th, since -t always became [θ] in this position in all the Brittonic languages, from the sixth century on, although it continued often to be written t, as here (Jackson 1953, 148–9 and 570–2).

Thus the three forms Doniert, Dungarth and Dumnarth represent the same name, pronouned 'Doniarth' (or 'Donierth') in Old Cornish. Since the -t in the lapidary form and the g in the Annals form are both misleading to modern readers, the reconstructed Old Cornish form 'Doniarth' is used here to show the pronunciation. The name was in common use in the Gaelic world, but very rare in the Brittonic one. Examples of early Irish Domongart appear, for example, in the collections of genealogies, both saintly and secular (O'Brien 1962, 593, five men; Ó Riain 1985, 241 and 282, three saints, two other men). But in the Brittonic world the only other historical example noted of the name is a lay witness called Dofngarth occurring in a south Welsh charter dated approximately to the 860s, a date interestingly close to that of the Cornish king (Evans and Rhys 1893, 169, twelfth-century MS; Davies 1979, 106). There is also a legendary Dyfnarth or Dyfnfarch ap Prydain ('son of Britain') occurring in Welsh genealogies (Bartrum 1966, 184b, references); and one literary instance of the name in Welsh, Gwrgwst Letlwm a Dyfnarth y uab 'Gwrgwst half-naked and Dyfnarth his son' (Bromwich and Evans 1988, line 993), but this character was envisaged as a northerner, so his name is likely to be a rendering of the well-known Irish Domongart; compare Dunart brenhin y Gogled 'Dunart king of the North' in the same text (Jackson 1982, 20–1; Bromwich and Evans 1988, line 254 and note to line 993).

In fact the second element of the name, garth 'enclosure, protection', is altogether rare in Brittonic personal names. Apart from the name Doniarth, the only other Brittonic personal name noted as containing the element is Guengarth, the name of both a cleric and a layman in the charters of the Book of Llandaf (Evans and Rhys 1893, index, 401a), and of a saint so named in the twelfth-century Life of St Cadog (Wade-Evans 1944, 130), who gave his name to the parishes of Llanwenarth (Monmouthshire, SO 2714) and St Weonards (Herefordshire, SO 4924); possibly also the personal name Gistlerth, if from gwystl 'pledge' + garth (Evans and Rhys 1893, 260, 264: does the spelling -er- here provide a parallel for that seen in Doniert?). Apart from Doniert, there are no other known instances of this element in personal names from Cornwall, and none at all from Brittany (Cane 2003, 166, 176 for the element in Welsh names, and 240, 264, 282 and 288 for its absence in Cornish and Breton names). This element garth has been interpreted as Welsh garth 'height', thus 'leader' (Cane 2003, 166, 176), but the spelling -gort in the Irish cognate shows it is the word meaning 'enclosure' instead, thus here 'protection'. The name 'Doniarth' is thus rare and distinctive in the Brittonic world.

The rarity of the personal name has important implications for the sculpted stone. The monument is large and impressive and was erected for someone highly significant, incorporating a style of decoration newly-developed in Cornwall. It stands three miles from Liskeard, which was the largest comital manor in the county in 1066–86, and which was already in use as an administrative centre by the Anglo-Saxon government in the early eleventh century. The name of Liskeard (containing Cornish-language *lys 'court') shows that it had previously been a ruler's court under the native administration which preceded that of the Anglo-Saxons. It is natural to conclude that Liskeard was a native court which was taken over by the incoming Saxon administration, and therefore that it was in use as a royal court by the last native rulers in the late ninth century, immediately before the Saxon administration became fully established. Thus Doniarth, the last native king of Cornwall, is implicitly linked with Liskeard. In this context the hunting visit of Alfred to the Liskeard area in 865–70 gains increased significance (Stevenson 1959, 55; Keynes and Lapidge 1983, 89, 254–5): at that date Doniarth (died 876) was presumably already on the throne, or his immediate predecessor. It is most unlikely that Alfred, as crown prince of Wessex, would not have visited the sub-king of Cornwall when hunting in the Liskeard area, and in fact his choice of area may have been due to the significance of its court at the period.

Given Doniarth's implicit association with Liskeard, there is a further deduction to be drawn regarding the nearby stone. It is most unlikely that a second person of sufficient importance to be commemorated on so impressive a monument would also have borne this rare Brittonic name in the same area at the same period. The cumulative evidence therefore indicates that this monument was indeed erected to the last king of Cornwall who died in 876, and hence that the stone provides valuable dating-evidence for this style of decoration. Professor Julia Crick kindly informs me that the forms of the letters in the inscription, and particularly the Insular r and a, if occurring in a manuscript (the forms of which this inscription seems to be imitating quite closely) would be entirely compatible with dating the stone to this period.

The conclusion can be taken a little further, though more tentatively. Of the four monuments in Cornwall using this style of decoration, two stand at a single spot (St Cleer 2 and 3, the 'Other Half Stone'), erected beside one of the major east–west routeways (see Ills. 58–9). The third monument of this type stands two miles away, at St Neot's church (Ills. 151–4), where King Alfred, as a youth, sought the shrine of the local saint to obtain a cure for his troubling disease. The style of that monument too can thus be comparatively dated to King Alfred's time. It then becomes a reasonable conjecture to attribute the St Neot 1 cross-shaft possibly to the king himself, perhaps given in gratitude for the cure which he obtained there. It need not have been produced at the time of his visit (c. 865–70), but could have been given later during the king's lifetime, or at his death in 899 (see p. 172).

The fourth monument using this style of decoration is that at St Just-in-Penwith, at the opposite end of the county (Ills. 100–2); its existence slightly reduces the cohesiveness of the east Cornish group, but it is smaller, and the geographical closeness of the three large cross-shafts to one another, outside Liskeard, remains notable. The two monuments in this style surviving in Devon are at Copplestone, four miles outside Crediton, at the fork of the two main roads to north Devon, and at Exeter (Cramp 2006, 82–3, 86–7, ills. 10–14, 26–9); Crediton and Exeter were both major centres in the ninth century and early tenth, so the location of the three Cornish stones near the administrative centre of Liskeard is consistent with the pattern seen in Devon.

Late ninth century, if the identification of Doniert with the historical king is accepted, or tenth century

Camden 1600, 155; Camden 1607, 139 and fig.; Speed 1676, 21 and fig.; Gibson 1695, cols. 9, 23 and fig.; Lhwyd 1700, no page numbering; Moll 1724, 'Cornwal' figs., no page numbering; Moyle 1726, 185, 187–8, 237–8 and fig.; Norden 1728, 58–9 and fig.; Hals 1750, 46–8 and fig.; Borlase, W. 1754, 360–3 and fig.; Britton and Brayley 1801, 388; Polwhele 1803, ii, 195–7; (—) 1807, 717–18; Lysons and Lysons 1814, ccxxii–ccxxiii, ccxlv; (—) 1817, 39; Gilbert, C. 1820, 947 [1]; Bond 1823, 195–201 and fig. opp. p. iv; Hitchins 1824, i, 439–43, ii, 153; Gilbert, D. 1838, i, 178–84, fig. 195; Penaluna 1838, i, 108; Redding 1842, 57 and fig.; Spence 1849, 211–15, pl. 32, fig. 2; Hingston 1850, no. 47, fig. 47; (—) 1851, 193; Collins 1851, 35; Coppard 1851, 205; Coppard 1853, 73–4; Blight 1858, 128–9 and fig.; Wilkinson 1862, 52; Paull 1866–7, xi; Polsue 1867, 205–6; (—) 1869, xxii; Haddan and Stubbs 1869, 699; Iago 1871–3, 67; Rhys 1875, 369–70; Huebner 1876, 8, no. 22, and fig.; Iago 1878–81, 399; Hare 1883–5, 20; Langdon, Arthur and Allen, J. R. 1888, 303, 311–14, 318, 324; Allen, J. R. 1889, 128–9, 220–1; Langdon, Arthur 1889a, 319, 325–7, 334, 337, 343 and figs.; Langdon, Arthur 1889c, 239; Langdon, Arthur 1890–1, 36, 43, 46–7, 53, 87 and figs.; Langdon, Arthur 1892, 37n; Borlase, W. C. 1893, 184–5; Langdon, Arthur 1894, 309; Langdon, Arthur and Allen, J. R. 1895, 52, 57, 60, no. 33 and fig.; Langdon, Arthur 1896, 377–9, passim and figs.; Norway 1897, 211–12; Rhys 1905, 74; Daniell 1906, 244, 290; Langdon, Arthur 1906, 408–9, 411–12, 419, 438, 441, pl. V, fig. 36; (—) 1910–11b, 281–4 and figs.; Rhys 1918, 193; Jenner 1919, 137; Henderson, C. 1925, 73; Macalister 1929, 188; Andrew 1932, 25–7; Hencken 1932, 268, 271, 295; Nance 1933a, 36; Nance 1933b, 10; Andrew 1933–6, 112–39; Malim 1936, 73, 79–82, 84 and fig.; Nance 1936, 15–16; (—) 1943–50, 242; Macalister 1949, 184, no. 1054, and fig.; Ellis, G. 1952–3f, 177–9 and figs.; Jackson 1953, 421; Radford 1953, 23; Thomas, A. C. 1957–8, 69; Allen, J. 1967, 13; Thomas, A. C. 1967a, 104; Pevsner 1970, 15–7, 163–4; Thomas, A. C. 1971b, 65 and fig.; Ellis, P. 1974, 19; Axford 1975, 87–8, 180; Radford 1975, 14; Pearce 1978, 168 and fig.; Thomas, A. C. 1978, 76–9, fig. 4.1; Pearce 1981, 270; Weatherhill 1985, 26, 35 and fig.; Higgitt 1986b, 141; Preston-Jones and Rose 1986, 159; Thomas, A. C. 1986, 67–8 and fig.; Todd 1987, 295; Okasha 1993, 213–17 and passim, no. 43, fig. II.43; Hingley et al. 1995, 66; Langdon, Andrew 1996a, 25; Pearce 2004, 314; Langdon, Andrew 2005, 35, no. 41, and figs.; Turner 2006a, xvii, 163, pl. 11b; Turner 2006c, 33, 37–9, fig. 13; Preston-Jones 2011, 274

[1] This reference may instead be to St Cleer 3.