A Powerful Place of Pictland: Interdisciplinary Perspectives on a Power Centre of the 4th to 6th Centuries ad

Abstract

OUR UNDERSTANDING of the nature of late and post-Roman central places of northern Britain has been hindered by the lack of historical sources and the limited scale of archaeological investigation. New work at Rhynie, Aberdeenshire, Scotland (NJ 49749 26345), has begun to redress this through extensive excavation and landscape survey. This has revealed a Pictish central place of the 4th to 6th centuries ad that has European connections through material culture, iconography and site character. In addition to reviewing the place-name and historical context, this article outlines preliminary reflections on five seasons of excavation and survey in the Rhynie landscape. The article also provides a detailed consideration of chronology, including radiocarbon dating and Bayesian statistical analysis. The results reveal the multi-faceted nature of a major, non-hillfort elite complex of Pictland that comprised a high-status residence with cult dimensions, a major centre for production and exchange, and a contemporary cemetery. A series of sculptured stones stood in association with the settlement and cemetery and the iconography of the stones, along with the wider archaeological evidence, provides a rich dataset for a renewed consideration of the central places of early medieval northern Britain with broader implications for the nature of power and rulership in late and post-Roman Europe.

Acknowledgements

Many thanks to the students, staff and volunteers who have made Rhynie 2011–2017 possible and to ‘Rhynie Woman’, who have brought so much to the community element of the project. Fieldwork at Rhynie has been funded by the University of Aberdeen Development Trust, the British Academy, Historic Environment Scotland, the Society of Antiquaries of Scotland, the Carnegie Trust for the Universities of Scotland and Aberdeenshire Council Archaeology Service. The writing of this article was also supported by a Leverhulme Trust Research Leadership Award (RL-2016-069).

Notes

8 Ibid, 312.

9 Eg Hope-Taylor 1977; Hamerow 2012, 23, 102–9; Thomas 2013, 126; Brennan and Hamerow 2015.

10 Eg Thomas 2013, 112.

11 Esmonde Cleary 1989, 138–61; Ulmschneider 2011, 157–60; Hamerow 2012, 23.

12 Eg Aitchison 1994; Newman 1997; Lynn 2003; Bhreathnach 2014.

13 Stout 1997, 24–9; O’Sullivan et al 2008, 59–61. Crannogs were also an important form of high-status residence and nexus of power in early medieval Ireland, eg Hencken 1950; Gleeson and Carty 2013; O’Sullivan et al 2014, 58–62.

14 Eg Críth Gablach dating to around 700 ad: Kelly 1991, 30; Stout 1997, 18; O’Sullivan et al 2014, 82–3.

15 Alcock 2003, 179.

16 Eg Bannerman 1974, 15–16; Alcock 2003, 179–200; Woolf 2007; Fraser 2009; Evans 2014.

17 Lane and Campbell 2000.

18 Ibid, 37; Fraser 2009, 213–24, 301–2.

19 Mac Airt and Mac Niocaill 1983, 130, 154, 176, 234, 326 (AU 658.2, 694.6, 722.3, 780.1, 870.6); Stokes 1896, 253 (AT [752].2); Alcock and Alcock 1990, 98; Adomnán, Life of St Columba, 123 (II 15), in Sharpe.

20 Ritchie 1994, 3–5; Woolf 2007, 87–130, 312–42; Fraser 2009.

21 Driscoll 1998, 169–70.

22 Alcock et al 1989, 192; Woolf 2007, 105.

23 Adomnán, Life of St Columba, 184 (II 35), in Sharpe.

24 Ibid, 181–2, (II 33).

25 Bannerman 1974, 15–16; Anderson 2011, 250–2; Taylor 2011, 73.

26 Carver 2011.

27 Conceived in 1973. Alcock et al 1986; Alcock and Alcock 1990, 216.

28 Alcock and Alcock 1987; 1990, 216; 1992; Alcock et al 1989.

29 Ibid, 195.

30 Alcock 1988, 31; 2003, 179–200.

31 The initial aims of the project in 2011 were modest: to detemine the character, extent and survival of the cropmark features; to obtain scientific dating for the enclosures and to develop a deeper understanding of the Rhynie landscape. With the identification of the Barflat enclosure complex as early medieval in date, the aims shifted toward recording the major constituents of the monument complex through strip and map excavation. The work at the Barflat site and cemetery was carried out as part of the Rhynie Environs Archaeological Project (REAP), directed by the Universites of Aberdeen and Chester. The work in the wider landscape is being conducted as part of the Northern Picts project based at the University of Aberdeen.

32 See Roy’s 18th-century map <http://maps.nls.uk/geo/roy/#zoom=13&lat=57.3445&lon=-2.8818&layers = roy-highlands> [accessed 15^th November 2018]

33 Simpson 1930, 48–52; RCAHMS 2007, 121; Brander 2014, 29. The Mounth is an eastwards projection of the Cairngorms which results in a narrow lowland pass between central and north-eastern Scotland, a strategic routeway in itself.

34 RCAHMS 2007, 121.

35 Ibid, 103–5.

36 Ibid, 103.

37 See Taylor 2011.

38 1226 Moray Reg no 69 (Innes 1837, 73–76); note also Ryni 1226 x 1242 Moray Reg no 81 (Innes 1837, 90–5).

39 Grigg 2015, 85–6.

40 Watson 1926, 34.

41 Rivet and Smith 1979, 447; Isaac 2005, 202.

42 There remains the problem of the extra syllable in Ryan, as opposed to Rhyn-ie. This may have come about through assimilation to the well-known Irish surname Ryan, although a mid-17th-century form (Loch Rian c 1636 x 1652 Gordon 61 [manuscript map]) suggests it may be part of the original development of the name.

43 Anderson 2011, 245; Broun 2007, 75–9. For the tale lists, see Dobbs 1949, 137–8; Mac Cana 1980, 46, 47, 61, 63; Evans 2014, 66–7.

44 Dobbs 1949, 137–8; Atkinson 2006.

45 Watson 1926, 115; Anderson 2011, 245.

46 Broun 2007, 79.

47 Alexander, 1952, 186, 244, 325, 330, 333; Broun (forthcoming), drafts of which the author has kindly made available before publication. On March, see Taylor with Márkus (2012, 439). Our thanks go to an anonymous referee for drawing our attention to Marchmar.

48 Mac Airt and Mac Niocaill 1983, 448; Broun forthcoming.

49 Ibid.

50 Duncan 1975, 188; Barrow 1988, 4; Ross 2015, 100.

51 Anderson 2011, 268, 276, 284, 288–9. The same details are also found in the Verse Chronicle inserted into the Chronicle of Melrose (Anderson 1922, vol I, 604; Broun 1999, 153–60).

52 Moray Reg no 30 (Innes 1837, 22–23); Watson 1926, 440–1; Alexander 1952, 382–3; Taylor 2011, 105, 108; Taylor and Márkus 2012, 301–2, 504; Beam et al 2012, <http://db.poms.ac.uk/record/factoid/46205/> [accessed 15^th November 2018]

53 Simpson 1922, 147–50, 152. The parishes in the lordship were Kinnoir, Essie, Rhynie, Dunbennan, Ruthven, Glass, Drumdelgie, Botarie and Gartly (Ross 2015, 100).

54 GD44/51/747/1, Rental of the Lordship of Huntly, 1600 held at the NAS (now NRS) recorded the market at Rhynie as being worth £80, four times that of Huntly. Subsequent rentals show the market at Huntly expanding exponentially, though 18th-century records suggest that Rhynie was still the site of a ‘great fair’. Many thanks to Colin Shepherd for pointing these records out.

55 Cf Simpson 1922, 147–50, 152.

56 The lack of certainty really highlights the lack of a historical framework for this period in northern Scotland and the importance of archaeological evidence for revealing more about the character of 1st millennium ad sites and landscapes in this region.

57 We can perhaps loosely draw on the Anglo-Saxon evidence here. Chris Wickham suggests that the origins of Anglo-Saxon kingdoms lay with relatively small territories — equivalent to the size of a modern English county or less. By ad 600, kingdoms such as East Anglia, Wessex, Deira, and perhaps Mercia and Hwicce, were the size of around two modern English counties, but the Tribal Hidage records much smaller polities still in existence even in the 7th century (Wickham 2015, 313; see also Arnold 1997, 208–9).

58 For Pictish art see Henderson and Henderson (2004) and references therein.

59 In 1958 Isobel Henderson identified Rhynie as a possible origin centre for the Class I tradition based on the five stones then identified. However, she thought that Class I stones may have originated further north and that Rhynie was an important place through which the tradition disseminated (Henderson 1958, 54–5). Few scholars would attempt to identify an origin centre today, but the number of stones from certain sites is likely to be significant in terms of addressing the importance of sites without historical sources. Gondek (2006) outlines how the relative investment in the number of stones and complexity of carving and symbology is likely to be a good indication of centres of power in early medieval Scotland that can both support and add nuance to our historical understandings of the period.

60 RCAHMS 2008, 38–41; 2007, 119–22.

61 There are other concentrations of Class I stones known, such as the four at Inverurie and two others at Broomend of Crichie and Brandsbutt (all in central Aberdeenshire), but in this case the distribution is over a larger area and there is little archaeological information available to assess the significance of this particular concentration (RCAHMS 2007, 26).

62 Shepherd and Shepherd 1978, 215.

63 Turner 1994. See discussion below for the potential significance of the Cunningsburgh place name. See Shepherd and Shepherd (1978, 215) for further parallels for Rhynie Man on Class II Pictish stones.

64 RCAHMS 2008, 98; Henderson and Henderson (2004, 70) suggest that the Golspie sculptor was using a style that expressed a ‘conscious archaism’.

65 RCAHMS 2008, 70, 80. And now a new stone from Tulloch, Perth (Mark Hall pers comm).

66 Shepherd and Shepherd 1978.

67 RCAHMS 2007, 121.

68 A small excavation also occurred in 2006 near the complex, focusing on a Bronze-Age ring-ditch house, following geophysical survey, see Gondek and Noble (2015).

69 RCAHMS 2007, 122.

70 Ibid. See fig 6.20 in RCAHMS (2007) for enclosures of comparable size in Aberdeenshire.

71 Wainwright 1955, 88; Ralston 1997, 34; Carver et al 2016, 228.

72 Noble et al 2013, 1142.

73 S1 at Portmahomack was around 14.5 m x 10 m in maximum dimensions (Carver et al 2016, 240).

74 Ibid.

75 Shepherd and Shepherd 1978, 211.

76 RCAHMS 2007, 122.

77 Ibid 2007, 121.

78 Henderson 1907, 163.

79 Logan 1929, 56.

80 Greig 1994, 26.

81 Campbell and Maldonado forthcoming; Mitchell and Noble 2017, 17.

82 RCAHMS 2007, 103–5.

83 Cook 2010, 79.

84 Small and Cottam 1972; Peteranna and Birch forthcoming. Current Historic Environment Scotland funded work at Tap o’ Noth aims to date the major enclosing elements of the fort.

85 RCAHMS 2007, 101; Cook 2011, 216.

86 New excavations by the Northern Picts project (2018–2019) aims to more firmly establish the chronology and character of the Cairnmore site.

87 Gondek and Noble 2011; 2015.

88 RCAHMS 2007, 73; Gondek and Noble 2015, 128; George Currie pers comm.

89 RCAHMS 2008, 40. There are a number of recumbent stone circles in the area from which this stone could have come. See Clarke (2007) for a list of other possible prehistoric standing stones reused as Class I symbol stones. Clarke argues that prehistoric stones were reused as a reaction against Christianity missionary activity, but this does not seem wholly convincing for the Rhynie case or for north-eastern Scotland more generally in the 5th or early 6th centuries. See Clancy (2008) for Christianity in north-eastern Scotland.

90 See also RCAHMS 2007, 82–3; Gondek and Noble 2015.

91 The number of artefacts from the site is particularly notable given that the site is plough truncated, with the majority of contemporary land surfaces removed by cultivation and later land-use practices. The excavation strategy was also evaluative in nature, with the majority of deposits left in situ for future investigation, suggesting that the number of artefacts outlined here is only a small fraction of the total assemblage that survives on site and, given truncation, is likely to be a smaller fraction again of the original total.

92 Campbell 2007, 18–24.

93 Ibid, 64–9.

94 Ibid, fig 13.

95 See Alcock (1972), Warner (2000) for general site overviews and Campbell (2007, 110) for locational aspects of these two sites.

96 See Fowler (1963) for brooch typology.

97 See discussion in Youngs (2005).

98 Campbell and Heald 2007, 177.

99 Ashmore 1999. See Tab 1 for all the dates included in the modelling.

100 The samples were pre-treated, combusted, graphitised and measured by accelerator mass spectrometry at the Scottish Universities Environmental Research Centre. See Dunbar et al (2016).

101 Calibrations have been calculated using the probability method (Stuiver and Reimer 1993).

102 Buck et al 1996. Bayesian statistics provide a method of allowing different types of information (eg radiocarbon dates, phasing and stratigraphy) to be combined to produce realistic estimates of calendrical dates. The technique used is a form of Markov chain Monte Carlo sampling and has been applied using the program OxCal v4.2. Details of the algorithms employed by this program are available from the online manual (https://c14.arch.ox.ac.uk/oxcal.html) or in Bronk Ramsey 1995; 1998; 2001; 2009. In the model, the calibrated radiocarbon dates are shown in outline and the posterior density estimates produced by the chronological modelling are shown in solid black. Highest posterior density intervals used to summarise these distributions in the text are given in italics to reflect the fact that they are modelled and emphasise that they are not absolute and would change given a different set of parameters or ‘prior’ beliefs.

103 Eg palisade posthole samples (SUERC-35639/40/43, -66398, and -66425). The two results from palisade posthole 15333 (SUERC-66398 and -66245) are statistically consistent (T’ = 2.3; ν = 1; T’ (5%) = 3.8). All tests for statistical consistency follow the method of Ward and Wilson (1978).

104 (SUERC-35638, -66394/9, -66400, -66424, and -66431/2). The two results from context 15334 are not statistically consistent (T’ = 4.6; ν = 1; T’ (5%) = 3.8; Ward and Wilson 1978), nor are the two results from 15345 (T’=6.6; ν = 1; T’ (5%) = 3.8). This would suggest the fill from the palisade trench contains material deposited over a protracted period of time, and this is consistent with an interpretation of much of the material coming from redeposited backfill after the palisade had been dismantled.

105 Of the four results from samples submitted from the inner ditch, SUERC-45548 can be placed later in the stratigraphy than SUERC-45554. SUERC-45553 is from the basal fill in the outer ditch and earlier than SUERC-45547, while SUERC-35637 comes from a lower fill than SUERC-35649. There are two results (SUERC-66388/9) from context 15191 that are statistically consistent (T ’= 0.3; ν = 1; T’ (5%) = 3.8) and could be the same actual age.

106 Of the five results (SUERC-45538/9 and -45544/6/9) from structure S1, there are two in stratigraphic order. SUERC-45538 is earlier than SUERC-45546. There are two results (SUERC-66390/1) from context 15112 associated with structure S4. The two results are statistically consistent (T’=0.0; ν = 1; T’ (5%) = 3.8) and could be the same actual age. There are three results (SUERC-35641/2 and -35648) from structure S3, with SUERC-35648 coming from a discrete charcoal spread from within the structure. Structure S3 is definitely later than the inner ditch and probably later than the outer ditch. The overall modelling explores both possible readings of the stratigraphy.

107 (SUERC-66426/30). These measurements are statistically consistent (T’=1.7; ν = 1; T’ (5%) = 3.8) and could be the same age.

108 Woolf 2006, 192, 195–6, 199–201.

109 Fraser 2009, 50–1, 54–8, 212–16, 224.

110 Adomnán, Life of St Columba, 196 (II 42), n 294 and 324, in Sharpe; Woolf 2006, 201; Fraser 2009, 99–103.

111 Fraser 2009, 68–9. AU 536.3 (AT kl. 47.1) (perditio panis); AU 539.1 (perditio panis); AU 545.1 (mortality); AU 549.3 (mortality) (CS 551); AU 554.2 (AT kl. 61.2, CS 554.2) (pestis).

112 Fraser (2009, 94–5) has argued that this could have been a later addition c 700 when Iona was politically close to Fortriu, in order to show that Fortriu was involved in Argyll affairs at an early date, but the obscurity of the item’s wording would make it very poor propaganda.

113 Fraser 2009, 68–9. AU 536.3 (AT kl. 47.1) (perditio panis); AU 539.1 (perditio panis); AU 545.1 (mortality); AU 549.3 (mortality) (CS 551); AU 554.2 (AT kl. 61.2, CS 554.2) (pestis).

114 For the precariousness of most of Irish society, and the dire social consequences of natural disasters, see Corráin (2005, 574–83); Fraser (2009, 68) has speculated that the plagues may have reached Scotland. For rulers as decisive for a kindgom’s natural as well as human prosperity, see Oakley (2010, esp. 20–1, 46, 155–7), and for Ireland specifically, see Jaski (2000, 58, 73–81).

115 Clancy 2008.

116 Ralston 2004; Noble et al 2013.

117 Noble et al 2013, online appendix.

118 Alcock et al 1989, 200.

119 Ibid, 201.

120 Peteranna and Birch forthcoming.

121 Alcock et al 1989, 202.

122 Lane and Campbell 2000, 251.

123 Close-Brooks 1986, 145.

124 MacIver and Noble 2017.

125 A building of around 9 m x 5 m was found during 2017 excavations at Burghead (Sveinbjarnarson and Noble 2017).

126 Eg Fenton and Walker 1981, 73–76; Hunter 2007, 48–50; see RCAHMS (2007, chapter 3) for a detailed consideration of the destruction of archaeological remains in Aberdeenshire in cultivated landscapes and chapters 7–8 for a consideration of the lack of rural settlement in a post-Iron-Age context.

127 As argued for early Pictish structures at Portmahomack (Carver et al 2016, 86). Similar issues surrounding the survival (or lack of survival) of buildings at early medieval sites exist in western Britain, at Cadbury Castle, Somerset, for example. At Cadbury, Alcock found that there were few earthfast components to the rampart or to a putative timber hall found within (Alcock 1972, 176–7; 1995)

128 Hope-Taylor 1977, 6.

129 While the Rhynie Barflat buildings were much more modest than those found at Anglo-Saxon hall complexes, it is important to stress the early date of the Craw Stane complex, with the structures at Rhynie more comparable to Anglo-Saxon buildings of the 5th–6th centuries. Anglo-Saxon buildings of this period rarely exceed 12 m in length (Hamerow 2012, 22). The post-built structures 2, 3 and 4 at Mucking (Essex), for example, were just over 10 m in length though, unlike in many Pictish buildings, earthfast elements were evident marking the line of the exterior walls (Hamerow 1993, 8, figs 54–56). The evidence from Rhynie shows that northern traditions may have been important models for emulation during the development of the Anglo-Saxon elite complexes of the 6th and 7th centuries ad.

130 Stout 1997; O’Sullivan et al 2008, 58.

131 Warner 1988, 58; 2000.

132 Warner 1988, 57.

133 Ibid.

134 Campbell 2007, 123–4.

135 Ibid, 123.

136 See Campbell (2007, tabs 17–19). Table 19 shows comparable sites in Scotland — to date only silver-working, not gold, has been found at sites in Pictland. The sword/dagger pommel from Rhynie is the first from a defended site in Pictland. See also Campbell (1996).

137 Curle 1982; Lane and Campbell 2000; Laing and Longley 2006.

138 Nieke 1993, 128; Campbell 2007, 116.

139 RCAHMS 2008, 64, 66; Gondek 2015; Toolis and Bowles 2017.

140 Turner 1994.

141 Stewart 1987, 48–9. The earliest forms he gives are: Konesbrocht, Konosbrocht, Konnesbrocht 1507, Cunnisburgh 1565.

142 For a thoughtful and usefully cautious discussion of this name in relation to the place, see Johnston (1999). For the argument that the eponymous ‘king’s fort’ is Mousa Broch, so named by the Norse because of its relative magnificence and great size, see Smith (2016, 16). Thanks to Brian Smith for these references.

143 Noble et al 2013, 1147.

144 Hope-Taylor (1977, 97–102) argued that the hall complex was aligned on prehistoric barrows and the hall complex included a cult structure (D2) as a central component of its layout.

145 See, for example, Newman (2007), papers in Schot et al (2011), Gleeson (2012) and Gleeson and Carty (2013).

146 Eg Brink 1996, 237; Ringtved 1999, 50; Watt 1999, 173; Hedeager 2011, 152.

147 Eg Sundqvist 2002, 166; Ringtved 1999, 49; Hedeager 2011, 152.

148 I 37, II 33. In Ireland the status of druids had diminished by the 7th–8th centuries, but they were still present and influential enough to be included in lists of privileged people in law tracts (Kelly 1991, 59–60).

149 Alcock 1972, 80–81; Hencken 1950, fig 40a. The Cadbury example was found in the entranceway of the fort, perhaps ritually deposited. It was associated with Anglo-Saxon metalwork of probable 6th-century date (Alcock 1972, 104, pl 79). Parallels can also be drawn with the axe hammer in the Sutton Hoo ship burial (Suffolk), although again this Anglo-Saxon example is likely to be later in date than the Rhynie Man depiction.

150 Blair 1995; Campbell and Maldonado forthcoming.