INTRODUCTION
One of the enduring enigmas of early medieval northern Britain is the identification of a place which Bede records as urbs Giudi. Urbs Giudi features in Bede’s account of ‘How the Britons were ravaged by the Irish and the Picts’ (Vt Brettones a Scottis uastati Pictisque).Footnote 1 Bede clarifies Gildas’ description of ‘the Scots from the north-west and the Picts from the north’ as gentibus transmarinis ‘overseas peoples’,Footnote 2 that is, peoples who invaded Britain from across the sea. He explains that:Footnote 3
Transmarinas autem dicimus has gentes non quod extra Brittaniam essent positae, sed quia a parte Brettonum erant remotae, duobus sinibus maris interiacentibus, quorum unus ab orientali mari, alter ab occidentali Brittaniae terras longe lateque inrumpit, quamuis ad se inuicem pertingere non possint. Orientalis habet in medio sui urbem Giudi, occidentalis supra se, hoc est ad dexteram sui, habet urbem Alcluith, quod lingua eorum significat Petram Cluit; est enim iuxta fluuium nominis illius.
This follows an earlier passage in the Historia ecclesiastica:Footnote 4
Est autem sinus maris permaximus, qui antiquitus gentem Brettonum a Pictis secernebat, qui ab occidente in terras longo spatio erumpit, ubi est ciuitas Brettonum munitissima usque hodie, quae uocatur Alcluith; ad cuius uidelicet sinus partem septentrionalem Scotti, quos diximus, aduenientes sibi locum patriae fecerunt.
Although Bede does not mention them by name, these ‘arms of the sea’ are clearly identifiable as the firths of Forth and Clyde, which cleave what is now central Scotland. This is evident from Bede’s references to their great size (longe lateque, permaximus) and the Antonine Wall that runs between them, while Alcluith/Petram Cluit, near the river of the same name, is readily identifiable as the distinctive landmark of Dumbarton Rock. In marked contrast to Alcluith, the location of urbs Giudi remains controversial. With some understatement, Kenneth Jackson described Orientalis habet in medio sui urbem Giudi as ‘a rather obscure sentence … which has caused some perplexity’.Footnote 5 The conundrum surrounding the identification of urbs Giudi is given greater irony by Bede’s stated intention ‘to remove all occasions of doubt about those things I have written’ and, reflecting his status as a ‘truthful historian’ (uerax historicus), his commitment to recording information ‘in a straightforward manner’ (simpliciter).Footnote 6
Despite describing urbs Giudi as being literally ‘in the middle of ’ the Firth of Forth, Bede’s reference is both terse and ambiguous, fuelling speculation about its location. Over twenty different sites have been proposed for urbs Giudi since the sixteenth century.Footnote 7 Taking a literal interpretation of Bede’s description, urbs Giudi was traditionally sought among the islands of the Firth of Forth, usually Inchkeith.Footnote 8 From the mid-twentieth century, urbs Giudi was tentatively identified with the Roman forts at Cramond, four miles (6.4 km) north-west of the centre of Edinburgh, or Inveresk,Footnote 9 six miles (9.7 km) east-south-east of Edinburgh city centre, both on the southern shore of the Firth of Forth. Stirling is the location of urbs Giudi favoured by most scholars,Footnote 10 although it has been rejected by others.Footnote 11 As a result, support for Stirling has waned since the 1980s.Footnote 12 This uncertainty is reflected in the subsequent proposed identifications of urbs Giudi with Stirling, Inveresk or even Edinburgh,Footnote 13 despite a strong argument against the latter,Footnote 14 Stirling or Cramond,Footnote 15 and ‘possibly Inveresk but also possibly Stirling’.Footnote 16
The location of urbs Giudi remains unresolved. The most recent study adds further uncertainty by suggesting additional identifications. James Fraser not only revives cases for Cramond Island (Midlothian) and Blackness (West Lothian) as the site of urbs Giudi, originally made in the nineteenth century,Footnote 17 but also proposes a new location, Carlingnose Battery, North Queensferry (Fife).Footnote 18 As the first candidate site for urbs Giudi on the northern shore of the Firth of Forth, within the territory of historical Pictland, the latter marks a departure from previous studies.Footnote 19 Fraser suggests that Carlingnose Battery was the site of urbs Giudi on the grounds that the strait between Queensferry and North Queensferry lies at ‘the most obvious “mid-point” of the estuary’,Footnote 20 consistent with Bede’s in medio sui ‘in the middle of itself’, and that an elevated coastal promontory was the type of location favoured by builders of early medieval fortresses.Footnote 21 However, there is no evidence for the presence of any fortifications on Carlingnose Point until the 1650s, when it may have been the site of a battery intended to prevent Cromwell’s army from crossing the Forth,Footnote 22 and it was used to defend the entrance to the Forth Estuary from the late eighteenth century.Footnote 23 The key factor in the attractiveness of Carlingnose Point as a fortification was the development of modern artillery.
Fraser’s proposed alternative locations for urbs Giudi have received a mixed reception, being described as both ‘convincing’ and ‘contentious’.Footnote 24 His identifications have also been criticised on the grounds that they ‘involve disregarding Bede’s distinction between “Orientalis (sinus) habet in medio sui” and “occidentalis supra se, hoc est ad dexteram sui”’.Footnote 25 Fraser deepens, rather than resolves, the enigma of urbs Giudi. Instead of recent research narrowing the field of candidate sites, opinions on urbs Giudi have become increasingly divergent as additional possible locations are proposed. The prospect of identifying urbs Giudi appears more remote than ever.
This paper does not rehearse well-established arguments or indulge in further speculation concerning the identification of urbs Giudi. Instead, it reassesses one strand of evidence in an attempt to break the scholarly stalemate. Bede records four points about urbs Giudi: its place-name, location, and its status as both a counterpart to Alcluith and as an urbs. Although usually translated ‘city’,Footnote 26 Bede uses urbs to refer to a fortress or stronghold.Footnote 27 This paper focusses on Bede’s description of the location of urbs Giudi. This has been attempted on several occasions previously.Footnote 28 Fraser, for example, suggests that ‘careful consideration of his [Bede’s] geographical references … may help scholars to locate … urbs Giudi’.Footnote 29 However, analyses of the relevant passage tend not to be objective but are influenced by previous studies and preconceptions about the likely location of urbs Giudi. The continued absence of any consensus on this reveals that a different approach is required, beginning with a detailed examination of the terms used by Bede and their possible implications. Surprisingly, given the level of interest in urbs Giudi, previous analyses of the vocabulary used by Bede are superficial. This study, therefore, takes a literary and lexical approach, focussing on the range of possible meanings expressed by Bede’s reference to urbs Giudi and following Bede’s words to see where, if anywhere, they lead on the ground. Identifying the most plausible interpretations within the context of the topography of the Firth of Forth may provide new insights and possibly even assist the identification of urbs Giudi. This paper does not begin with the intention of making a case in favour of, or against, any particular identification of urbs Giudi. Instead, it aims to be objective about the possible location of urbs Giudi and unencumbered by the baggage of five centuries of inconclusive debate and speculation.
After reviewing Bede’s interest in geography and his abilities as geographer, this paper assesses the plausibility of the enduring theory that urbs Giudi was an island. It then seeks fresh insights into the identity of urbs Giudi by comparing the key details from Bede’s description of it as being situated in medio sui, referring to the eastern sinus, with other instances within the Historia ecclesiastica and across Bede’s other writings in which Bede uses the phrase in medio or the word sinus. This opens a new dimension to the debate, because Bede’s wider use of these terms presents the possibility of determining their most probable meaning in his description of urbs Giudi.
Urbs Giudi justifies renewed study for several reasons. It merits attention by virtue of its appearance in the Historia ecclesiastica, one of the outstanding literary accomplishments of the early medieval period, alone. The Historia ecclesiastica is a profoundly important work of, and source for, early medieval Insular history. The enigma of urbs Giudi features in successive editions and translations of, and commentaries on, the Historia ecclesiastica and has been the subject of several studies.Footnote 30 This alone makes urbs Giudi of intrinsic interest. In addition, urbs Giudi is one of only four places referred to by Bede in what is now central Scotland, while its status as an urbs indicates that urbs Giudi was a major, probably royal, power centre. This makes urbs Giudi of archaeological and historical importance as well as being of wider significance to the study of early medieval northern Britain, particularly the political organisation of a region about which very little is known.
Making it of added interest, urbs Giudi is usually identified with ‘the city called Iudeu’ (urbem quae vocatur Iudeu) recorded in the Historia Brittonum. Footnote 31 This was the scene of atbret Iudeu, the ‘distribution of Iudeu’, in which Oswiu, king of Northumbria, was forced to surrender his treasure to an alliance of Penda, king of Mercia, and British kings in 654.Footnote 32 A reading of one manuscript of the Historia Brittonum locates Iudeu in the region of Manau, a British territory straddling the Forth Estuary.Footnote 33 However, it is impossible to determine whether this represents anything more than a scribal error.Footnote 34 Although they are usually thought to be Old English and Brittonic reflexes of the same name,Footnote 35 any resemblance between the place-names Giudi and Iudeu may be more perceived and/or superficial rather than actual.Footnote 36
Moreover, the location of Iudeu is disputed. Not surprisingly, given the prevailing association of urbs Giudi with Stirling, most scholars also identify Iudeu with Stirling,Footnote 37 specifically its Castle Rock.Footnote 38 Alternatively, Iudeu has been identified with Carriden, Cramond, Inveresk, and as a possible nickname for Din Eidyn, Edinburgh.Footnote 39 By contrast, Tim Clarkson challenges the association of Iudeu with urbs Giudi and the Forth of Forth, suggesting that Iudeu was located ‘in southern Deira or north-east Mercia rather than at Stirling’ — or, presumably, elsewhere in, on, or near the Firth of Forth — and that it may be equated with the unidentified *Iudanburg (Iudanbyrig),Footnote 40 ‘somewhere in eastern Mercia or southern Northumbria’.Footnote 41 Consistent with this, the identities of the warring parties, who were from Mercia, Northumbria, and North Wales, imply a location much further south than the Firth of Forth. This may be compared with, and is possibly supported by, the location of Oswiu’s revenge the following year. Oswiu defeated and killed Penda in battle near the river Winwaed (prope fluuium Uinued), in the district of Leeds (in regione Loidis).Footnote 42
In the absence of any reliable evidence for the location of Iudeu, its identification with urbs Giudi is doubtful and, therefore, Iudeu is not discussed further here. As a result, there appears to be no reliable source for urbs Giudi other than Bede’s Historia ecclesiastica, while the meagre nature of the evidence available emphasises the importance of its analysis and interpretation.
BEDE AS A GEOGRAPHER
The extensive corpus of Bede’s surviving writings amply demonstrates his richly-descriptive language skills.Footnote 43 Indeed, Bede has been praised as ‘without question the most accomplished Latinist produced in these islands in the Anglo-Saxon period’.Footnote 44 Traditionally, Bede’s prose style was thought to be characterised by its clarity, simplicity and debt to Classical Latin.Footnote 45 By contrast, recent studies demonstrate that Bede’s prose varies considerably in style and that his Latin can be more complex.Footnote 46 Although this conclusion is drawn primarily from the study of Bede’s exegeses, the prose style of the Historia ecclesiastica can also be very ornate, as the sparrow simile demonstrates,Footnote 47 while elaborate wordplay features prominently throughout his writings.Footnote 48 Might this be relevant to Bede’s reference to urbs Giudi? Certainly, the difficulties presented by this passage — and, therefore, of locating urbs Giudi — may be contrasted with Charles Plummer’s frequently-quoted claim that ‘it is very seldom that we have to pause to think of the meaning of a sentence’ in the Historia ecclesiastica. Footnote 49
Bede was a polymath. The anonymous Life of Bede, copied in the late eleventh and early twelfth centuries,Footnote 50 emphasises the importance of geography to Bede by placing it first in a list of his many and varied interests.Footnote 51 This is apparent from the start of the Historia ecclesiastica, which begins with a geographical study, a description of Britain, following Gildas.Footnote 52 His abilities as a geographer are also evident in his exegeses, where ‘Bede takes great care to account for geographical features’ and ‘demonstrated the greatest attention to geographical questions, creating examples for later scholars to follow’.Footnote 53 As ‘a scholar quite exceptionally conscious of geography’,Footnote 54 Bede displays a ‘fascination with the landscape of history’.Footnote 55
Although his comments are limited, Bede includes enough detail to demonstrate his knowledge of the Forth-Clyde isthmus. He notes that the ‘two wide and long arms of the sea’ do not meet and locates Alcluith on the right (north) bank of the western sinus, near the river Alcluith shares its name with.Footnote 56 This accurately locates Alcluith near the point where the River Clyde widens into the Clyde Estuary. Bede is not only aware that the Antonine Wall crosses the Forth-Clyde isthmus but that it ‘starts almost two miles west of the monastery at Aebbercurnig (Abercorn) in the place which the Picts call Peanfahel’ (Kinneil) and ‘stretches westward as far as Alcluith’.Footnote 57 Although the eastern terminus of the Antonine Wall is actually at Carriden,Footnote 58 just under four miles (6.5 km) from Abercorn, Peanfahel means ‘end of the wall’.Footnote 59 Moreover, Bede records that the Anglo-Saxon name for Peanfahel was Penneltun and Old English tun probably denotes a royal estate.Footnote 60 This suggests that Peanfahel referred to a wider area, which may have included the end of the wall. Bede also demonstrates his familiarity with the toponymy of some of these places. As well as recording both the ‘Pictish’ and Old English names for Kinneil,Footnote 61 Bede also includes a gloss on Alcluith, ‘a name which in their language means Petram Cluit’. This is a convincing account and its accuracy implies that the information Bede provides on urbs Giudi is also reliable. It was probably based on sources originating in the Anglian monastery at Abercorn,Footnote 62 on the southern shore of the Firth of Forth, although Bede may have been drawing — directly or indirectly — on multiple informants.
As an accomplished geographer and Latin scholar, Bede would have selected the words he uses to describe urbs Giudi with great care, skill and attention to detail. Although he refers to urbs Giudi only in passing, this is not a casual reference or throwaway remark but part of a larger, carefully constructed and complex passage. We know what Bede wrote, but what, exactly, did he mean? And what implications do his lexical choices have for the identification of urbs Giudi? Although Bede’s reference to urbs Giudi is terse, it need not be interpreted in isolation. A reassessment of Bede’s description of urbs Giudi, in which the words and expressions Bede uses are compared with other examples elsewhere within the Historia ecclesiastica or in his other writings, may help to clarify his intended meaning and, perhaps, assist the identification of urbs Giudi.
URBS OR INSULA?
Bede records that urbs Giudi was located in medio sui, referring to the eastern ‘arm of the sea’, that is, the Firth of Forth. Until the mid-twentieth century, the literal interpretation of this phrase led most scholars to conclude that urbs Giudi was located on an island, ‘the insular town of Giudi’.Footnote 63 Of the several islands in the Firth of Forth proposed, Inchkeith was an early and enduring favourite for the site of urbs Giudi. Footnote 64 Many editors and translators of Bede’s Historia ecclesiastica also identified urbs Giudi with Inchkeith, with varying degrees of certainty.Footnote 65 Not only is Inchkeith one of the largest islands in the Firth of Forth but, although closer to the Fife coast, it is also more realistically described as being located ‘in the middle’ of the Firth of Forth than most of the other islands.
Claims that urbs Giudi was located on an island persist. Urbs Giudi is still identified as Inchkeith in the Penguin Classics translation of the Historia ecclesiastica,Footnote 66 while Fraser resurrects earlier suggestions that Cramond Island may be the site of urbs Giudi. Footnote 67 Indeed, Fraser claims that the Middle Irish and Brittonic names for the Firth of Forth, Muir nGiudan and Merin Iodeo respectively, strengthen the case ‘that Giudi might lie in an island’.Footnote 68 Moreover, Fraser not only interprets the urbs of Iudeu, with which urbs Giudi is often identified, as an island fortress,Footnote 69 but also asserts that it was located ‘in the midst of the Firth of Forth, probably in or opposite West Lothian’.Footnote 70
Despite the popularity of the insular hypothesis, there are archaeological, topographical and textual arguments against an island location for urbs Giudi. Most obviously, no physical traces of an urbs have been identified on any of the islands in the Firth of Forth. Some antiquarians attempted to explain this by claiming that urbs Giudi was ‘probably built of wood, as no vestige of it has been seen for many ages’.Footnote 71 This is unconvincing, as many early medieval royal centres recorded in northern Britain were strongly fortified with earthwork, stone or timber-laced defences.Footnote 72 There are, however, exceptions. A timber palisade provides the simplest form of defence and elaborate timber enclosures occur on some early medieval royal sites in northern Britain.Footnote 73 Large double-palisaded enclosures are prominent features of the Northumbrian royal centres at Milfield, Sprouston and Yeavering,Footnote 74 while an enclosure at the Pictish possible royal centre at Rhynie (Aberdeenshire) comprised a complex wall of posts and planks, perhaps with a wall-walk.Footnote 75 However, the landscape contexts of these sites contrast markedly with the small islands in the Firth of Forth. In addition to the absence of structural evidence, in the form of fortifications, there are no small finds that might indicate the presence of an undiscovered early medieval urbs on any of these islands.
The absence of archaeological evidence may be assessed in conjunction with the unsuitability, both logistically and defensively, of an insular urbs. Most of the islands in the Firth of Forth are small and rocky, while all are exposed to sea-borne assault, as repeated English — and one French — attacks on, and occupations of, Inchcolm, Inchkeith and the Isle of May between the fourteenth and mid-sixteenth centuries attest.Footnote 76 Although several islands in the Firth of Forth were fortified,Footnote 77 this did not begin until the sixteenth century, when the development of artillery gave them a new role in defending the ports of, and shipping within, the firth. The challenges associated with constructing, resupplying and defending an early medieval royal centre make these islands unlikely locations for the site of urbs Giudi. While Peter Hunter Blair entertained the possible identification of urbs Giudi with the tidal Cramond Island, he rejected Inchkeith, ‘an island which lay some four miles offshore in the middle of a tidal estuary’, as a ‘scarcely conceivable’ location for an early medieval royal centre.Footnote 78 Jackson concurred, noting that ‘Dark Age’ urbes were not sited on ‘little islands’ because it would have required fleets to make them viable.Footnote 79 Supporting this, there are no clear or convincing parallels for the insular location of an early medieval urbs.
Bede does not attach the word insula to urbs Giudi. However, this could be explained either as indicating that Bede did not think that urbs Giudi was an island or, alternatively, that he was under the impression that describing it as being ‘in the middle’ of a body of water adequately conveyed that this place was an island. Some scholars proposed ingenious solutions to Bede’s omission of a reference to an island. Sir John Rhys speculated that Bede confused two separate places sharing the same name, Insula Giudi and Urbs Giudi, the former in the Firth of Forth, the latter on its southern shore, at Carriden or possibly Edinburgh.Footnote 80 If correct, this would require the identification of two separate places named Giudi although they were, presumably, associated by more than just their name. However, Rhys did not provide any evidence to support his theory, for example, in the form of attestations of an *insula Giudi or an equivalent Celtic place-name, and none is apparent.
Another theory is that the location of urbs Giudi may be resolved by inferring the existence of a scribal error. According to John Edward Lloyd, Bede mistakenly refers to Giudi as an urbs when it was actually an insula. Footnote 81 However, a copying error seems most unlikely because of the reliability of the text of the Historia ecclesiastica and its preservation in four early manuscripts.Footnote 82 The earliest of these, the Moore manuscript, dates to 734 × 737,Footnote 83 no more than two years after Bede’s death (735) and around six years after the date traditionally attributed to the completion of the Historia ecclesiastica, in c. 731. Unusually for an early medieval source, these manuscripts reveal that the text has been transmitted with great accuracy, with the result that major emendations are required or acceptable only in exceptional cases.Footnote 84
The only way of reconciling Bede’s description of urbs Giudi with an insular setting might be if Bede was either unaware that urbs Giudi was located on an island or if his account is somehow confused. However, it seems unlikely that Bede would have referred erroneously to an insula as an urbs for several reasons. Firstly, Bede’s competency as both a geographer and a Latin scholar makes such a mistake improbable. Bede’s reliability as a commentator in general and as a geographer in particular is apparent from his extensive writings, where no parallels to such an inaccuracy can be identified. Moreover, Bede may have been less likely to make such a mistake when referring to an urbs, as this term denotes the most important settlements. Bede refers to only eight urbes in Britain, four of which are in northern Britain: Bebbanburgh (Bamburgh, Northumberland), Coludi urbem (Coldingham, Berwickshire), Alcluith, and urbs Giudi. Footnote 85 Secondly, the importance Bede attaches to islands makes it implausible that he would have either confused or not mentioned its insular status if urbs Giudi was (on) an island. Insularity and remoteness were fundamental to Bede’s vision of Britain and its place in the wider world from the opening words of the Historia ecclesiastica, Brittania Oceani insula ‘Britain is an island of the ocean’,Footnote 86 a theme which dominates the first chapter. Indeed, the island of Britain defines the extent of Bede’s history, its insular status providing his geographical terms of reference, as well as a powerful ideological role, underpinning ‘the ideal of unified church in Britain’ and a driver for ‘the internal colonialism of religion’.Footnote 87
Above all, Bede regards islands as places of great symbolic and spiritual significance. They not only mirror the island of Britain but, by evoking the Garden of Eden, are also associated with the terrestrial paradise.Footnote 88 Bede was particularly interested in islands as monastic retreats and these feature prominently throughout the Historia ecclesiastica: Farne Island, Inishboffin, Iona, Lindisfarne and an unnamed island (St Herbert’s Isle) on Derwentwater.Footnote 89 Bede also refers to Anglesey, the Isle of Man, the Orkney Isles, the Isle of Thanet and the Isle of Wight as islands.Footnote 90 To Bede, islands were places of great sanctity, separated from the world and ideally suited to an ascetic life of devotion, contemplation and spiritual salvation.
By contrast, there is no unequivocal evidence for the occupation of islands by early medieval power centres in northern Britain. Although it is argued that both Dumbarton Rock and Dunadd, a fortified royal centre of the Scots of Dál Riata,Footnote 91 were tidal islands during the mid-first millennium AD,Footnote 92 this seems unlikely in the case of Dunadd, at least. Moreover, this need not have any implications for the location of urbs Giudi. The status of Alcluith and urbs Giudi as counterparts does not imply that, even if Dumbarton Rock was a tidal island, it follows that urbs Giudi was too. And, despite the defensive potential of a tidal island being indicated by the siege of Theodoric, king of Bernicia, on Lindisfarne by Urien, king of Rheged, in the 570s,Footnote 93 this event may have been the exception rather than the rule. In addition, it is unclear whether Lindisfarne was a royal power centre at the time, or simply an expedient defensive location to which Theodoric had retreated.
Bede’s fascination with both the geography and sanctity of islands is also evident in his Life of Cuthbert. For example, he records the location of Coquet Island, explains how the island acquired its name and states that ‘It was renowned for its monasteries’.Footnote 94 Bede makes a distinction between Cuthbert’s choice of a ‘true island’ as his place of retreat and the tidal island of Lindisfarne. To Bede, Farne Island is ‘an island … in the middle of the sea’.Footnote 95 Emphasising its insular status, Bede adds that Farne Island is ‘cut off on the landward side by very deep water and facing, on the other side, out towards limitless ocean’.Footnote 96 By contrast, Lindisfarne is a semiinsula, literally ‘half-island’, because it ‘is an island in the strict sense of the word only twice a day, when cut off by the tide. When the tide is out it is joined to the mainland’.Footnote 97 This distinction is of potential relevance to one of the proposed sites of urbs Giudi. Cramond Island is also a tidal island, although William Skene suggested that this is inconsistent with the location of urbs Giudi: ‘By the expression “in medio”, Bede seems to imply … an island at all times surrounded by the sea’.Footnote 98
Ultimately, the evidence of Bede’s own words is inconclusive in attempting to determine whether urbs Giudi was located on an island or not. More conclusively, no archaeological evidence of an early medieval fortress has been discovered on the islands of the Firth of Forth. Moreover, these islands make unlikely locations for a stronghold of this period because they could not have been defended effectively from sea-borne attacks. Indeed, the islands of the Firth of Forth contrast markedly with the hill-top and coastal promontory locations of known early medieval fortifications in northern Britain.Footnote 99 As a result, it seems implausible that urbs Giudi was either an island or was located on an island. How, then, can this be reconciled with Bede’s description of urbs Giudi as being situated ‘in the middle’ of the Firth of Forth? Although there appears to be an inherent and fundamental contradiction in Bede’s account, this rests entirely on the interpretation of in medio sui.
IN MEDIO SUI
The Latin adverbial phrase in medio sui is fundamental to locating urbs Giudi. In medio sui comprises a preposition, in ‘in’, plus medio, the ablative of medius ‘central, middle’,Footnote 100 used as a noun/substantive in this context, and sui, the genitive of the reflexive pronoun sē, here meaning ‘itself’.Footnote 101 As a result, the interpretation of in medio sui, literally ‘in the middle of itself’, referring to the eastern sinus, initially appears to be straightforward. Most translations of Bede’s Historia ecclesiastica translate in medio sui fairly literally, as ‘in the middle of’,Footnote 102 ‘in the midst of’,Footnote 103 ‘in its midst’,Footnote 104 [a]u milieu de,Footnote 105 and [a] la moitié du. Footnote 106 Not surprisingly, this prompted the (fruitless) antiquarian quest for urbs Giudi among the islands of the Firth of Forth.
The ‘Halfway Along’ Hypothesis
By contrast, Bertram Colgrave’s translation of the Historia ecclesiastica interprets in medio as ‘half way along’.Footnote 107 This divergence from the literal translation of in medio is not explained. However, it reflects contemporary thinking about the location of urbs Giudi, given the absence of plausible locations for a fortress among the islands in the Firth of Forth and, therefore, the perceived inadequacy of the literal translation ‘in the middle’. As early as 1885, it was noted that in medio could mean either ‘in the middle’, in the sense of between two opposite shores, or refer to ‘the middle’ laterally, between the two ends of the Firth of Forth.Footnote 108 And, in 1908, Alan Anderson argued that in medio meant that urbs Giudi was ‘not necessarily on an island’,Footnote 109 although he did not propose any alternative interpretations.
By the mid-twentieth century, the search for urbs Giudi had shifted from the islands of the Firth of Forth to its coast. Colgrave and Mynors do not refer to Hunter Blair’s suggestion that ‘the site of urbs Giudi should be sought in some suitable position, such as Cramond … or perhaps Inveresk, on the southern shore of the Forth’,Footnote 110 but note that ‘[t]here is some possibility that it may be Inveresk’.Footnote 111 It could be argued that Cramond or Inveresk are situated approximately in the middle of the southern shore of the Firth of Forth, particularly in the absence of maps or surveys and given the inability to see from one end of the firth to the other because of its curving course.Footnote 112 This led Hunter Blair to suggest that in medio refers to a location ‘not out in the middle of the Forth, but half way along it’,Footnote 113 then to state as fact that ‘urbs Giudi … according to Bede lay halfway along the Firth of Forth’.Footnote 114 Hunter Blair subsequently noted that ‘the words in medio …. are commonly taken to mean “in the middle of”, but “half-way along” seems to be a better translation’.Footnote 115 ‘Halfway along’ was already in use before its adoption by Colgrave and Mynors,Footnote 116 was declared to be a ‘perfectly acceptable’ translation of in medio,Footnote 117 and is now accepted widely.Footnote 118 One French translation of the Historia ecclesiastica follows Colgrave’s interpretation: À mi-chemin du bras oriental ‘Halfway up the eastern arm’ but, inconsistently with this, notes Stirling as the possible location of urbs Giudi. Footnote 119
The only dissent was from Kenneth Jackson,Footnote 120 who disputed as ‘incorrect’ the assessment that ‘halfway along’ is a ‘perfectly acceptable’ interpretation, describing it as ‘not nearly so natural’ a translation as ‘in the middle’ and even ‘forced’. Despite Jackson’s misgivings, a possible parallel may be identified in Classical Latin in colle medio, literally ‘in mid-hill’, but translated as ‘half-way up the hill’.Footnote 121 However, there are other objections. Firstly, and most fundamentally, in medio does not mean literally ‘halfway along’. Secondly, Bede uses a different expression to express this meaning. This is apparent from his description of the position of a ship in the English Channel as ad medium itineris, literally ‘to the middle of the journey’, which Colgrave translates as ‘half-way across’.Footnote 122 Thirdly, Bede was skilled in both Latin and geography and appears to have had access to sources originating in Abercorn. If urbs Giudi was located ‘halfway along’ the Firth of Forth, then Bede might be expected to have described it in such terms and to have specified which shore it stood on, just as he does with Alcluith. But he does not. Moreover, the manner in which Bede contrasts the location of Alcluith, on the right hand bank of the Clyde Estuary, with that of urbs Giudi implies that urbs Giudi was situated on neither the northern nor southern shores of the Firth of Forth but in medio sui, wherever that may be. Nowhere else does Bede use the phrase in medio in the sense ‘halfway along’.
It seems unlikely to be coincidental that ‘halfway along’ emerged at the same time as Hunter Blair’s proposed location of urbs Giudi at Cramond or Inveresk.Footnote 123 The translation of in medio appears to have changed from ‘in the middle’ to ‘halfway along’ in order to reflect what was then considered to be a plausible theory concerning the location of urbs Giudi, rather than being based on an objective analysis of Bede’s text. The translation ‘halfway along’, therefore, rests on a circular argument in which a proposed location for urbs Giudi has influenced the interpretation of the text and vice versa, providing false corroboration of both. As a result, ‘halfway along’ cannot be accepted as a reliable translation of in medio. This has fundamental implications for the location of urbs Giudi. If urbs Giudi was neither ‘in the middle’ of, nor ‘halfway along’, the Firth of Forth then where was it?
The ‘Apex of a Triangle’ Hypothesis
Graham proposed a third interpretation, suggesting that Bede ‘may well have imagined the Firth [of Forth] to be a more or less triangular opening with a wide base to seaward’.Footnote 124 As a result, Bede may ‘have used the words in medio sui to mean “at its apex”, regarding the apex as a kind of mid-point between evenly converging shores’.Footnote 125 Interpreting this as referring to the angle between the northern and southern shores of a triangular-shaped firth, Graham suggested a location for urbs Giudi at the tidal reach of the Firth of Forth: ‘Stirling, placed as it is at the highest point that a seaman would be likely to reach, would exactly fit this meaning of in medio sui’.Footnote 126 Interpreting in medio sui as ‘at its apex’, Graham concluded that urbs Giudi, if located at Stirling, may legitimately be described as ‘in the middle of a bay’. Believing that Bede’s claimed perception of a ‘triangular’ firth may explain his description of it as longe lateque, Jackson accepted Graham’s interpretation of in medio as referring to the apex of a triangle and, therefore, found the identification of urbs Giudi with Stirling ‘entirely convincing’.Footnote 127
Graham’s interpretation is followed in the most recent translation of the Historia ecclesiastica, although his paper is not referred to there. In a marked departure from previous translations, Paolo Chiesa renders the passage as L’insenatura orientale ha al vertice la città di Giudi ‘The eastern inlet has the city of Giudi at its apex’.Footnote 128 Consistent with this, Jackson’s identification of urbs Giudi with Stirling is referred to, although it is noted that this is disputed.Footnote 129
Fraser suggests that Graham’s ‘translation of in medio sui as “at its highest point” surely strains credulity’,Footnote 130 although this could be said to be equally true of more widely-accepted translations of the phrase. Most obviously, the Firth of Forth does not resemble a triangle because it does not have ‘evenly converging shores’ but irregularly curving coasts instead. If Bede had perceived the Firth of Forth as a triangle, he might have been expected to describe it as such. Instead, Bede refers to this body of water as a sinus. Footnote 131 In addition, the Firth of Forth could only form a triangle if it is treated in isolation. By contrast, Bede perceives the eastern sinus not as a separate body of water but as an ‘arm of the sea’ that penetrates deep inland. Given ‘[t]he usual medieval belief about the flow of rivers … that it was supplied from the sea’,Footnote 132 this perception must have seemed self-evident in the case of firths and estuaries. Emphasising the weakness of his argument, Graham contradicts his own hypothesis by suggesting that his interpretation of in medio ‘would be even more natural if the sinus was thought of as a kind of rounded pocket’,Footnote 133 rather than a ‘triangle’.
Jackson makes two more points in support of Graham’s ‘triangle’ hypothesis. Firstly, he compares the Firth of Forth with the Moray Firth and Dublin Bay,Footnote 134 but this is irrelevant to Bede’s description because all three have unique topographies. Secondly, in an attempt to reconcile Bede’s use of in medio sui with the proposed location of urbs Giudi at the head of the Forth Estuary, Jackson claims that Bede ‘must have known this region by hearsay only’, thereby introducing ‘a certain vagueness and inaccuracy about the true character of these two waters’.Footnote 135 Jackson follows Graham’s presumption that Bede ‘was not … familiar with the Firth of Forth’ but ‘depended on the reports of sailors or other casual informants’.Footnote 136 Only by finding fault with Bede and/or his sources does the ‘triangle’ hypothesis stand up. However, this contrasts with Bede’s geographical skills and access to reliable sources originating at Abercorn. The ‘triangle’ hypothesis is a weak argument,Footnote 137 a contrived attempt to rationalise Bede’s description and make it compatible with the identification of urbs Giudi with Stirling.
Despite this, it is widely agreed that Castle Rock, Stirling, is the probable site of an early medieval fortress.Footnote 138 A massive volcanic crag-and-tail formation towering over the surrounding alluvial plain of the Carse of Stirling, this was occupied by a royal castle by the early twelfth century, although its earlier history remains obscure. Its strategic location, commanding the lowest crossing of the Forth and the ‘Stirling Gap’ between surrounding hills, makes Stirling the ‘gateway to the Highlands’. This ensured that its castle was besieged repeatedly during the Scottish Wars of Independence and later conflicts while its intended relief led directly to the Battle of Bannockburn in 1314.Footnote 139 No evidence of early medieval activity on Castle Rock has been detected, presumably because it has been destroyed by, or lies beneath, successive phases of medieval and later fortifications. Despite rejecting its identification with urbs Giudi, Fraser concedes that Stirling ‘suggests itself’ as the site of an early medieval royal power centre.Footnote 140 Could Graham and Jackson have arrived at the correct location for urbs Giudi but for the wrong reasons?
Bede’s references to urbs Giudi and Alcluith within the same sentence appear to imply some form of equivalent status. Inferring from Alcluith, this is presumably as impressive and important fortifications of the northern Britons, on opposite sides of the Forth-Clyde isthmus. Bede seems to indicate that urbs Giudi and Alcluith may have been similar types of places but that their relationship with the estuaries on which they were situated was dissimilar. Bede describes the locations of Alcluith, on the right bank of the western sinus, and urbs Giudi, in medio sui, referring to the eastern sinus, in very different and precise terms. In doing so, he appears to make a fundamental distinction between their situations relative to the estuaries with which they are associated. In particular, the absence of a reference to which bank urbs Giudi stood on is surely significant. Bede implies that urbs Giudi was not located on either bank of the eastern sinus. This impression is reinforced by the precision with which Bede records that urbs Giudi was situated in medio sui, although the significance of in medio sui in this context is obscure. Whatever in medio sui means, it appears to undermine the arguments in favour of a location for urbs Giudi on either shore of the Firth of Forth and, therefore, most of the sites proposed previously, including Cramond, Inveresk, and Edinburgh, as well as Fraser’s proposed sites of Blackness, Cramond Island, and Carlingnose.Footnote 141 Given the doubts that in medio sui refers to an island in the Firth of Forth or a location on either of its shores, this leaves Stirling as the only remaining plausible, but disputed, identification from suggested locations for urbs Giudi. Is Stirling compatible with a place which is described as being located in medio sui of the eastern sinus?
Other Instances
Bede’s language appears to be precise but his intended meaning of in medio sui and, therefore, its implications for the location of urbs Giudi, are unclear. Although its translation as ‘halfway along’ is rejected above,Footnote 142 in medio is capable of more than one reading, as its use in other contexts indicates.
The phrases in medio, in medium occur seventeen times in the Historia ecclesiastica. Footnote 143 Bede uses medius in two different ways, as a noun, in ‘the middle’ of something, and as an adjective, in ‘mid-’ something’. Although the reference to urbs Giudi as being situated in medio sui employs medius as a noun, adjectival examples are also revealing and are therefore included here. For clarity, adjectival phrases are distinguished from nominal phrases where they are discussed. Published translations of in medio, in medium are often too free to convey nuances and do not always distinguish between medius as an adjective and as a noun or in plus accusative ‘into’ and in plus ablative ‘in’. As a result, literal translations are sometimes included here in an attempt to capture the exact meaning of Bede’s phrases.
Seven examples of Bede’s use of in medio refer to a centrally-placed location within a building. These include porticus … in medio pene sui altare ‘in the middle of the chapel is an altar’,Footnote 144 literally ‘the chapel … has an altar almost in the middle of it’, and ecclesia … in cuius medio ‘in the centre of the church’,Footnote 145 literally ‘the church … in the midst of which …’. Bede also describes how Edwin constructed a church at York, ‘in the midst of which (in cuius medio) the chapel which he had first built was to be enclosed’.Footnote 146 In addition, Bede uses in medio of sites associated with Christ.Footnote 147 The expression carries the same meaning in domestic and secular contexts: accenso … foco in medio, literally ‘with the hearth in the middle [of the hall]’,Footnote 148 and accenso grandi igne in medio … domus ‘a great fire burned in the midst of the dwelling’.Footnote 149 Elsewhere, Bede refers to in medio tuguriunculi mei ‘in the midst of my little dwelling’.Footnote 150 Bede’s intended meaning is also clear in his description of Earth as orbis … in medio totius mundi positus ‘a sphere set in the middle of the whole universe’.Footnote 151 Bede was familiar with the Biblical concept of the world in which Jerusalem was located in medio gentium, at the centre of the earth with the peoples or nations arranged around it, even if he himself remained unconvinced: Hierusalem … putant ibi mediam esse terram ‘some suppose Jerusalem to be at the centre of the earth’,Footnote 152 where mediam esse terram means literally ‘to be mid-earth’. This example differs from the others discussed above because here medius is the adjective not the noun. Nevertheless, all these instances of the literal sense of in medio (and one of mediam) appear to be straightforward and incontrovertible.
Bede also employs in medio when referring to groups of people or animals. He mentions a tribune who procedit in medium ‘came into the midst’ of a crowd.Footnote 153 Bede records that Æthelthryth (Etheldreda), abbess of Ely, died in medio suorum ‘in the presence of her nuns’,Footnote 154 literally ‘in the midst of her people’,Footnote 155 and was buried in medio eorum ‘among them’,Footnote 156 literally ‘in the midst of them’. In another context, Bede uses the phrase in medio luporum ‘among the wolves’,Footnote 157 literally ‘in the midst of wolves’.Footnote 158 In these examples, Colgrave usually translates in medio literally and Leo Sherley-Price more freely. In some cases, Bede’s use of in medio may be better expressed in English as meaning something being ‘surrounded by’ another thing or things, rather than by the strict sense of being precisely ‘in the middle’ of something.
In medio occurs in a wide range of other contexts in the Historia ecclesiastica. For example, using medius as an adjective, rather than as a noun, Bede relates how monks were caught in a storm in medio mari ‘in the midst of the sea’,Footnote 159 literally ‘in mid-sea’, when returning from Farne Island. According to Bede, Farne Island was insula medio in mari posita ‘an island in the middle of the sea’,Footnote 160 literally ‘an island situated in mid-sea’. This does not reflect geographical reality, because Inner Farne, the location of Cuthbert’s cell, lies only 2.4 km (1.5 miles) off the Northumbrian coast. Bede’s descriptions of monks at sea and the location of Farne Island may be paralleled by his use of in medio to express examples of something being ‘surrounded by’ other things, as both the monks and the island were ‘surrounded by’ the sea, that is, cut off from the mainland, regardless of how close to the shore they were. Here, Bede’s use of in medio conveys the isolation of Farne Island, emphasising Cuthbert’s ascetic credentials, spiritual devotion and life of austerity. This sense of in medio mari may be better expressed as ‘far out to sea’.Footnote 161 In this context, Bede’s use of (in) medio may parallel the sense of Classical Latin medius ‘(in contexts emphasising remoteness from the edge, boundary, etc.) [t]he inside of, the middle of, the heart of’.Footnote 162
Another example where Bede uses the adjectival phrase in medio is in medio sermone, literally ‘in mid-talk’.Footnote 163 This is translated ‘in the midst of their talk’ and ‘in the course of conversation’.Footnote 164 The intended sense is apparent from the context, a meeting in which Cuthbert foresees the death of King Ecgfrith and informs Ecgfrith’s sister Ælfflæd, abbess of Coquet Island. The emotional and spiritual intensity of this discussion suggest a more elegant, but flexible, translation, that Cuthbert and Ælfflæd were ‘deep in conversation’. However, as Bede relates that Cuthbert and Ælfflæd continued their discussion after Ælfflæd asked about Ecgfrith, in medio sermone may be taken literally to mean that she posed her question ‘in the middle’ of a much longer conversation about a range of issues.
Bede’s commentary on the Canticle of Habakkuk contains his gloss on the biblical image of being in medio duorum animalium ‘[b]etween two living beings’,Footnote 165 literally ‘in the middle of two animals’. Bede’s objective here is to establish what these ‘two living beings’ might signify. His suggestion that they may denote Moses and Elijah refers to Matthew 17:3, where these prophets talk to Christ, who might then be described as being ‘in their midst’. Bede also suggests that the ‘two living beings’ may represent the two thieves crucified on either side of Christ. In both cases, Bede’s suggestions stem from taking in medio fairly literally as an indication that the words from Habakkuk must refer to another biblical event in which something was located either alongside two other things (Christ, Moses and Elijah) or in between them (Christ and the two thieves).
Returning to the Historia ecclesiastica, Bede’s description of Hell from Dryhthelm’s vision indicates his non-literal application of in medium. In it, Bede uses in medium as a nominal phrase three times when describing tormented souls in Hell: prosiliebant miserae in medium rigoris infesti ‘they leapt into the midst of the deadly cold’ and resiliebant rursus urendae in medium flammarum inextinguibilium ‘they jumped back only to burn once again in the midst of the unquenchable flames’, before Trahentes autem eos maligni spiritus descenderunt in medium baratri illius ardentis ‘The evil spirits dragged them down into the midst of the burning pit’.Footnote 166 The last passage may be translated literally as ‘dragging them [the souls], the evil spirits descended into the midst of that burning pit/Hell’. This may express a sense of something being ‘surrounded by’ something else, like some examples of in medio discussed above. However, Bede’s description of Hell presents a more complex use of in medium because it refers to motion occurring in different dimensions simultaneously, with the demons not only pulling souls into a fiery pit but also downwards inside it. This is emphasised by Colgrave’s translation of cum longius subeuntibus ‘as they descended deeper’ later in the same sentence.Footnote 167 In these cases, in medium refers to both ‘midst’ and ‘depth’. Bede’s account of Dryhthelm’s vision of Hell provides the strongest evidence that he uses in medium in a non-literal sense, at least in some instances.
Assessing in medio
It is evident from the examples discussed above that Bede’s usual practice was to employ the nominal phrase in medio ‘in the middle’ literally. This, by itself, does not exclude the possibility of a non-literal meaning for in medio sui as it applies to the location of urbs Giudi in relation to the eastern sinus. This is because Bede employs in medio in different ways in different contexts. In medio clearly carries a literal meaning when applied to a physical structure, whether a chapel, church, hall or dwelling, or even to the world itself. Beyond the confines of a building, in medio appears to carry a more flexible sense. When applied to a group of living beings, whether nuns or wolves, in medio may still mean ‘in the middle’, but in the sense ‘surrounded by’. These (predominantly) small-scale, local contexts, both structural and social, account for most instances of in medio in the Historia ecclesiastica.
By contrast, Bede appears to use in medio/in medium in a different sense when referring to wider physical contexts, specifically the location of Farne Island and perceptions of Hell. In these open settings, Bede employs in medio to express depth and/or distance, with the possible meanings ‘far out to’ and ‘deep into’. In this manner, Bede uses in medio to convey the remoteness of Farne Island, with its associated perils for the monks caught in a storm on the open sea, and to evoke the desperate fate of those souls dragged into the centre and depths of Hell, as perceived in Dryhthelm’s vision. Although these instances are in the minority, they indicate a broader range of meanings for in medio that are dependent on context. Bede’s application of the nominal phrase in medio sui in his reference to urbs Giudi falls into this select category, concerning as it does the location of a fortress within the wider setting of the Firth of Forth. Indeed, in both its geographical scale and estuarine setting, Bede’s reference to urbs Giudi may be compared with his use of the adjectival phrase in medio, in a marine or coastal context, to describe the location of Farne Island. In the cases of both Farne Island and urbs Giudi, Bede is pinpointing specific locations within their more extensive geographical settings, a tiny island in the vastness of the North Sea and an urbs within the great eastern sinus respectively.
These distinctions in Bede’s use of in medio have potential implications for the interpretation of the passage concerning urbs Giudi and the location of this stronghold. A combination of text and topography suggests that Bede’s description of urbs Giudi being situated in medio sui may carry a secondary, more obscure, meaning rather than literally ‘in the middle of itself’. Although they are not identical — in medio is an adjectival phrase when referring to Farne Island, whereas in medio sui, used to locate urbs Giudi, is a nominal phrase — Bede’s description of Farne Island as in medio mari may imply that urbs Giudi was not ‘in the middle of’ the Firth of Forth at all. This is consistent with the absence of archaeological evidence for an urbs on the islands in the Firth of Forth and their poor suitability for the site of an early medieval fortress. Similarly, there is no evidence to support more imaginative interpretations of in medio as meaning ‘at the highest point’,Footnote 168 or ‘half way along’.Footnote 169
That in medio sui may carry a more opaque secondary meaning has been overlooked previously in the debate on urbs Giudi. Bede’s description of Farne Island and his accounts of Dryhthelm’s vision indicate that Bede sometimes employs in medio to convey depth, distance and/or remoteness. These meanings may be inferred from their contexts. In medio sui may also carry a subsidiary and figurative, rather than literal, meaning when applied to urbs Giudi because this is another example of Bede’s use of this phrase within a more extensive geographical setting. Moreover, this is the only reading that makes sense. In this context, in medio sui reflects the great size, both length and breadth, of the Firth of Forth. By describing its location as in medio sui, Bede did not necessarily mean that urbs Giudi was literally ‘in the middle’ of the Firth of Forth but may use this phrase to express that it was ‘far along’, ‘deep inside’, ‘at the heart of’, ‘in the depth, at the deepest or furthest point of’, or something similar. Rather than a site on an island in, or on the shore of, the Firth of Forth, in medio sui expresses remoteness, implying a location that was further upstream and deeper inland. This is consistent with a location at or near the head or tidal limit of the Forth Estuary. This parallels Graham’s theory that urbs Giudi was sited ‘at the highest point’ or ‘apex’ of the Firth of Forth,Footnote 170 where the Forth Estuary is intended.Footnote 171 The only previously-proposed location for urbs Giudi that is consistent with this re-reading of in medio is Stirling. This reinterpretation of in medio sui counters the objection ‘that Stirling is “too far inland”’ to be the location of urbs Giudi. Footnote 172 On the contrary, Stirling’s location at the head of the Forth Estuary and a short distance below its normal tidal limit fits this re-reading of Bede’s description of urbs Giudi in a way that no other proposed location in, on or around the Firth of Forth does.
The evidence for this reinterpretation of in medio sui as it relates to urbs Giudi is suggestive but far from conclusive. With this in mind, the significance of the topographical terms Bede uses to locate urbs Giudi, fretum and sinus, may also be reassessed.
SINUS
Previous attempts to locate urbs Giudi have focussed on the phrase in medio sui at the expense of the analysis of other terms which may be significant topographically. According to Bede, the Picts and the Scots were separated from the Britons by duobus sinibus maris … longe lateque ‘two wide and long arms of the sea’, which he also describes as duo freta uel sinus … maris, ‘two channels or arms of the sea’.Footnote 173 The western of these is introduced as sinus maris permaximus ‘a very wide arm of the sea’,Footnote 174 subsequently referred to simply as a sinus ‘arm’.Footnote 175 Bede uses fretum and sinus — apparently synonymously, despite their different meanings — when referring to the firths of Forth and Clyde.
Fraser attaches importance to Bede’s reference to the Firth of Forth as a fretum ‘channel, strait’.Footnote 176 However, the Firth of Forth is no more accurately described as a ‘channel’ or ‘strait’ than a ‘bay of the sea’. Fretum may carry various associations in this context from its Classical Latin sense ‘[a] place where the sea boils up’.Footnote 177 This may refer literally to turbulent waters. Exposed to the North Sea, the Firth of Forth can be blasted by violent storms and sudden squalls, such as that which forced Alexander I (1107–1124) to seek shelter on the island of Inchcolm.Footnote 178 Alternatively, or in addition, fretum may hold a metaphorical significance, perhaps alluding to the violence of Pictish raids across the Firth of Forth. Although of intrinsic interest, these meanings are not relevant to the location and identification of urbs Giudi and, therefore, fretum is not discussed further here. By contrast, sinus occurs more frequently than fretum in Bede’s references to these firths, indicating that it is the more significant of the two terms, while its ablative plural sinibus appears in the passage concerning urbs Giudi. Sinus is fundamental to Bede’s perception or ‘mental map’ of the Firth of Forth and, therefore, the identification of urbs Giudi.
The Forth: River, Estuary, and Firth
Despite noting that ‘negative evidence is dubious grounds for presuming that Bede did not know about Din Etin [Edinburgh]’, Fraser claims that Bede displays no awareness of either the River Forth or that it flowed into the eastern sinus. Footnote 179 Bede does not refer to the firths of Forth and Clyde by what would appear to be the most obvious term, Latin aestuarium ‘[a]n inlet, etc., covered by the sea at high tide, tidal opening’, ‘river estuary’,Footnote 180 ‘tidal inlet or channel, estuary’.Footnote 181 Nevertheless, this is unlikely to reflect the limitations of Bede’s knowledge. The Firth of Forth is the major geographical feature on the east coast of what is now Scotland and was known from at least the second century AD, when Ptolemy recorded it as Boderiae aest[uarium]. Footnote 182 Bede was almost certainly aware that this was an estuary from his sources originating at Abercorn, skills as a geographer, and familiarity with such environments. Indeed, Bede ‘spent his entire life near a tidal estuary’,Footnote 183 probably the Wear,Footnote 184 and refers to ‘we who live at various places along the coastline of the British Sea’.Footnote 185 As a result, Bede’s writing is ‘the product of a littoral, as much as [a] literary, environment’.Footnote 186 This is reflected in Bede’s interests in coastal fauna and tides.Footnote 187
Although Bede does not refer to the eastern sinus as a river, its estuary, and firth explicitly, this may be because he regarded their relationship as self-evident. Bede was certainly aware that the western sinus was associated with a river: ‘above the western branch … is the town of Alcluith … which … stands near the river (fluuium) of that name’.Footnote 188 Thomas Charles-Edwards assesses that Bede ‘undoubtedly combined the Firth of Clyde with the estuary of the River Clyde as the western sinus, so he may also have combined the Firth of Forth with the lower course of the River Forth as the eastern sinus’.Footnote 189 Charles-Edwards’ ‘lower course of the River Forth’ conforms with local usage but refers to what is more accurately described as the Forth Estuary.Footnote 190 Just as Bede refers to the Firth of Clyde and the Clyde Estuary jointly as the western sinus, it may reasonably be inferred that Bede’s eastern sinus comprises both the Firth of Forth and the Forth Estuary. This may be supported by Bede’s comment that both the eastern and western sinūs extend for a considerable distance inland: Brittaniae terras longe lateque inrumpit, literally ‘penetrates the lands of Britain far and wide’.Footnote 191
This point, if accepted, has major implications for the location of urbs Giudi. If the eastern sinus comprises both the Firth of Forth and the Forth Estuary, then it follows that a location in medio sui may not have been located in or on the Firth of Forth, but further upstream, on the Forth Estuary. This is paralleled by Dumbarton Rock, which is located on the Clyde Estuary. Just how far up the Forth Estuary urbs Giudi was situated is unclear and is dependent on the interpretation of in medio sui and Bede’s awareness and perceptions of the extent of this watercourse. However, if the revised reading of in medio sui as ‘far along’, ‘deep inside’, ‘at the heart of’ or ‘in the depth, at the deepest or furthest point of’ is accepted,Footnote 192 then this may suggest that urbs Giudi was located at or near the head of the estuary, placing it at, or in the vicinity of, Stirling.
The Forth as a Sinus
Sinus is not an obvious term for the Firth of Forth because it cannot be described accurately as a curving bay.Footnote 193 Indeed, the Firth of Forth is rarely referred to in print as a bay and, on the few occasions it has been, only by authors who appear to have limited familiarity with it.Footnote 194 Most translators of the Historia ecclesiastica have sought to resolve this apparent inconsistency between text and topography by adapting Classical Latin sinus, ‘a curved indentation in the coastline, bay, gulf’,Footnote 195 to suit Bede’s application of it, interpreting it as an ‘arm’,Footnote 196 ‘branch’,Footnote 197 ‘creek’,Footnote 198 or ‘inlet’ of the sea,Footnote 199 ‘estuary’,Footnote 200 ‘firth’,Footnote 201 archaic ‘frith’,Footnote 202 or simply ‘sea’.Footnote 203 More conventionally, other translators retain ‘bay’,Footnote 204 or ‘gulf’,Footnote 205 of the sea.
Medieval Latin sinus ‘curved or narrow area of water, enclosed by coastline, bay, strait, arm of the sea’ may be used to describe a wide variety of bodies of water.Footnote 206 Bede himself uses sinus flexibly, even inconsistently. He explains how ‘the Red Sea is divided into two gulfs’ (Mare Rubrum … scinditur … in duos sinus), the Persian and the Arabian,Footnote 207 refers to the English Channel as ‘the sinus between Britain and Gaul’ (gallico sinu brittanias),Footnote 208 and describes Ambleteuse Bay, at the mouth of the Slack Estuary, as a sinus maris ‘bay of the sea’.Footnote 209 And, although Bede refers to the firths of Forth and Clyde as sinūs, the Humber estuary is the Humbrae fluminis Footnote 210 and Humbrae fluminis maximi Footnote 211 ‘the (great) river Humber’. The contrast between the bodies of water referred to as sinūs by Bede suggests that this is not, either primarily or exclusively, a precise topographical term but perhaps has another significance.
Could successive scholars have misinterpreted Bede’s intended meaning of sinus as it relates to the firths of Forth and Clyde? Some nineteenth-century dictionaries include a more specific topographical sense of Latin sinus ‘anything bent or winding’Footnote 212 and ‘a bending, winding, a bay, gulf, frith, creek’.Footnote 213 Peter Heylyn (1599–1662), the English ecclesiastic and geographer, defines a sinus as ‘a Creek or Bay … a Sea contained within a crooked or circling shore, wherewith it is almost environed … and this is sometimes called a Gulf’.Footnote 214 Although Heylyn’s definition applies to the Firth of Forth, it may have been influenced by Bede’s Historia ecclesiastica, which was first published in England in 1643.Footnote 215 An earlier topographical sense is recorded as sinus ‘the turning or hollownes of water bankes’, giving sinuosus ‘that hath many turnings or windings’,Footnote 216 and Sinus, ripa, vel litus incurvum sive sinuosum, fluminis sinus ‘the turning, winding, or hollownesse of waterbanks, the curving reach of a river’.Footnote 217 These definitions make a direct link between Latin sinus as a topographical term and its adjectival form sinuosus ‘[c]haracterized by curves, sinuous, winding … (of rivers, coasts),Footnote 218 ‘characterized by bending, curving, twisting, winding, or sim[ilar]’,Footnote 219 from which English sinuous ‘[c]haracterized by or abounding in turns, curves, or sinuosities; sinuate, curving’ is derived.Footnote 220 Could these meanings express the sense in which Bede employs sinus in the context of urbs Giudi? Charles-Edwards suggests that Bede describes the location of urbs Giudi as in medio because he uses a term for the Firth of Forth, sinus, which suggests ‘a curving indentation’, although he does not elaborate on this.Footnote 221 However, Charles-Edwards also attempts to explain Bede’s use of sinus in this context by suggesting that ‘Bede may well have had only a very general knowledge of the shape of the Firths of Forth and Clyde’.Footnote 222
‘Crooked, turning, bending, winding, curving’ describes the course of the Firth of Forth more accurately than Graham’s ‘triangle’ or ‘rounded pocket’.Footnote 223 However, it is even more relevant to, indeed defines, the lowest reaches of the River Forth and most of the Forth Estuary. Between Craigforth, 1.2 miles (2 km) west-north-west of Stirling Castle, and Alloa, these sections of the River Forth and Forth Estuary follow such a serpentine course that an 18 mile (29 km) length covers a distance of only 7 miles (11.6 km) as the crow flies. This is known as the Links of Forth.
Most previous attempts to reconcile terminology and topography assume that Bede uses sinus to refer to the Firth of Forth, rather than to the Forth Estuary and the Firth of Forth. As a result, relying on the dictionary definition of medieval Latin sinus when attempting to locate urbs Giudi may introduce a circular argument, giving false confirmation that Bede’s sinūs refers only to the firths of Forth and Clyde, rather than to their more extensive watercourses. The Links of Forth suggests an alternative identification and interpretation of sinus which has been overlooked previously in the debate on urbs Giudi. It seems unlikely to be coincidental that Bede applies the term sinus to one of the most sinuous estuaries in Britain, renowned for its meandering course. This reinforces the argument that sinus applies not only to the Firth of Forth but also includes the Forth Estuary. Bede may have been aware of the Links of Forth from sources originating in Abercorn and/or clerics who travelled along the Forth Valley on their way between Iona and Lindisfarne. Sinus may be a synecdoche, referring to both the estuary and firth as a whole by their most distinctive feature. Stirling is located on this meandering section of the Forth Estuary, just below its normal tidal limit. Indeed, although the ‘halfway along’ hypothesis is rejected above,Footnote 224 it may be significant that Stirling lies at approximately the mid-point of the combined river, estuary and firth of Forth, depending on how the source of the former and mouth of the latter are defined.
Lexical Range and Metaphorical Meanings
Bede eschews the more obvious term aestuarium in favour of sinus when referring to the Forth, both estuary and firth. Given his skills as both a geographer and Latin scholar, this indicates that Bede selected sinus carefully to convey his intended sense(s). Bede’s use of sinus may reflect his command and imaginative use of Latin to express a more complex geographical or wider reality. Although sinus may refer to the sinuous course of the Forth Estuary, its rich lexical range suggests that it may hold additional meaning(s).
Perhaps significantly, sinus ‘a bay of the sea’ is very much a subsidiary sense in both Classical and Medieval Latin, comprising only the eleventh of twelve definitions listed in the Oxford Latin Dictionary and the tenth of eleven meanings in the Dictionary of Medieval Latin from British Sources. Its primary sense is ‘[t]he cavity or fold produced by the looping of a garment; … clothes draped in folds’ and, by association, ‘the part of the body covered by the sinus, the breast’.Footnote 225 From its meaning ‘[t]he fold of a garment over the breast or the breast itself as the part where a person, etc, is held as a demonstration of affection, for protection, etc’, sinus also carries the figurative senses ‘the place where secrecy is preserved’ and ‘[b]osom, refuge, shelter’.Footnote 226 Older Latin dictionaries also define these meanings as ‘[a] hiding-place, place of concealment … in secret’,Footnote 227 ‘a refuge, place of safety’.Footnote 228 These senses are also shared with medieval Latin sinus: 1 ‘fold (of garment, in quot[ations] fig[uratively])’, b ‘fold of garment over the breast’, c ([in a] transf[erred sense]) ‘pocket, purse’; 2 ‘breast, bosom’; 3 ‘embrace’; 4 (transf[erred] & fig[uratively]) b (with implication of secrecy, refuge, comfort or sim[ilar]) … ; 5 ‘the breast as the seat of the thought or emotion, heart, mind’; 6 ‘innermost part, heart’ (of a place); 7 ‘swelling, curvature’ (of sail); 8 ‘curved recess, hollow interior, cavity’; 9 ‘curved form or part of object’; 10 ‘curved or narrow area of water, enclosed by coastline, bay, strait, arm of the sea’; 11 ‘sine’ (math[ematical]).Footnote 229
Bede was ‘well aware that its [sinus’] primary meaning referred to a fold in a garment and hence (though not in a strictly physical sense) to the bosom’.Footnote 230 For example, Bede describes how Bothelm placed moss from a holy cross in sinum sibi ‘in his bosom’,Footnote 231 and refers to in sinum Abrahae ‘in the bosom of Abraham’,Footnote 232 symbolising heaven and the resurrected life. Despite this, these senses and their potential significance have been overlooked in previous discussions of urbs Giudi. More surprisingly still, another topographical application of sinus in the Historia ecclesiastica has not been noted in this context before.
Sinus also occurs in Bede’s Latin gloss on the Anglo-Saxon place-name Streanæshalch, Whitby (North Yorkshire).Footnote 233 Whitby’s coastal location initially appears to be consistent with the sense sinus ‘a bay of the sea’. However, sinus fari, ‘the bay of the light house’ or ‘bay of light’,Footnote 234 is not a Latin translation of Streanæshalch. Instead, sinus fari, possibly a pun on sinus maris ‘bay of (the) sea’, can also be read as ‘bosom of light’.Footnote 235 Sinus fari refers both literally to the bay on which Whitby is located and metaphorically to the radiant jewel worn in the bosom of Breguswith, or the folds of her garment, in her prophetic dream,Footnote 236 the saintliness of her daughter Hild, abbess of Whitby, and the monastery at Whitby as the beacon shining the traditions of the Roman Church over the island of Britain.Footnote 237
This gloss provides an insight into the complexity and richness of Bede’s imagery and his metaphorical use of a term in what initially appears to be a straightforward topographical reference. This calls for a reassessment of sinus as it relates to the Forth and urbs Giudi.
Reassessing Sinus
Bede was familiar with, and played on, the lexical range of sinus. Sinus fari demonstrates that Bede invests sinus with a deliberate ambiguity, using it to mean not only ‘a bay of the sea’, but also ‘a fold of garment over the breast’ and, by association, ‘the bosom’. As sinus fari echoes the duobus sinibus maris in the passage on urbs Giudi, is it plausible that sinus carries only a literal and topographic sense when applied to the Forth? This seems unlikely for several reasons. Firstly, Bede may have used the more obvious aestuarium if he was referring only to the Forth Estuary and Firth of Forth. Instead, Bede employs sinus because of its wider associations and greater literary potential. Supporting this, sinus has an extensive lexical range and carries a rich symbolism. Secondly, Bede uses sinus to express complex metaphorical meanings elsewhere in the Historia ecclesiastica, as sinus fari attests. Thirdly, the Firth of Forth and Forth Estuary do not conform to the standard topographical sense of sinus ‘a bay of the sea’, indicating that this term has been selected for another reason. Fourthly, several of those senses are directly relevant to the Forth Estuary and/or to urbs Giudi itself. Fifthly, sinus ‘a fold of garment over the breast’ parallels a wider pattern in Bede’s wordplay. Bede uses two homonyms involving clothing elsewhere,Footnote 238 including a recurring pun on habitus ‘garment’ and ‘character, nature’.Footnote 239 Sixthly, exclusively topographical readings of sinus have proved unproductive in previous attempts to locate urbs Giudi, indicating that a metaphorical sense may be more relevant.
Bede’s words are carefully selected and his sentences are skilfully constructed, while his rhetoric displays ‘a style grounded on the premise of a real correspondence between language and physical reality’.Footnote 240 As with sinus fari, Bede employs sinus as a monolingual pun in his reference to urbs Giudi. Playing on its ambiguity, Bede appears to use sinus in several different senses concerning the Forth and urbs Giudi. Firstly, Bede stretches the topographical meaning of sinus ‘a bay of the sea’ to apply to the Firth of Forth. Secondly, its primary sense, sinus ‘[t]he hanging fold of the upper part of the toga, about the breast, the bosom of a garment; also the bosom of a person; sometimes also the lap’,Footnote 241 may refer to both the winding course of the Firth of Forth and, in particular, the Forth Estuary. Here, the undulating folds of a loose garment are a metaphor for the meandering course of the Links of Forth. Thirdly, sinus ‘curved, winding’ expresses the sinuous course of the Links of Forth. Fourthly, sinus ‘bosom’ conveys the topography of the Firth of Forth, enclosed between curving coastlines. This sense is echoed in a 1578 reference to ‘a gret Nauie … sett in that bosum and arme of the Sey’,Footnote 242 where Scots bosum is a translation of the original Latin sinus. Footnote 243 In this case, the presence of a hostile English fleet in the very heart of Scotland expresses a sense of violation. Sinus ‘bosom’ also applies to the Links of Forth, where ‘[m]any peninsulæ are embosomed in the watery foldings’.Footnote 244 The dominant feature here is Castle Rock, Stirling, which sits in the bosom not only of the surrounding hills but also of the winding Links of Forth at its foot, as if caught metaphorically in the folds of a garment.
A possible objection to the interpretation of sinus ‘bosum; winding’ as a metaphorical reference to the sinuous course of the Links of Forth is that Bede also describes the Clyde as a sinus. However, the Firth of Clyde and Clyde Estuary also follow a crooked course, although not as winding as the Forth. Moreover, Bede may have applied sinus to the Clyde by association, because he was referring to both the Forth and Clyde but had descriptions of the Forth — reports of which he may have been more familiar with — in mind, either for literary effect or simply for convenience. More prosaically, Bede’s reference to both the Forth and Clyde as sinūs may reflect his wider and flexible use of the term and its lexical range. So rich are the range of meanings, literal and metaphorical, of sinus that not all them can have applied to every body of water Bede refers to as a sinus.
Bede may also use the transferred and figurative senses of sinus ‘the interior, innermost part, heart, centre (of a place)’ and, by association or implication, ‘a hiding-place, place of concealment, in secret’ to express the character, location and perhaps even function of urbs Giudi. Sinus ‘heart, centre (of a place)’ echoes in medio but may express the status of this location as a perceptual and/or socio-political centre, perhaps consistent with an urbs as a seat of power, rather than in a strictly geographical sense. Sinus ‘the interior, innermost part, heart, centre (of a place)’ implies a location which is not in or on the Firth of Forth but further upriver and, therefore, less accessible, more remote, ‘hidden’ or even ‘secret’. This is more consistent with the Forth Estuary and River Forth, which penetrate deep into the interior.
The terminology of Bede’s reference to urbs Giudi implies an unfamiliar, perhaps even remote, location. This may reflect the combination of geography and navigation. Although Stirling is only 20 miles (33 km) from Abercorn as the crow flies, the sailing distance is doubled by ‘the endless windings’ of the Links of Forth,Footnote 245 which also create a strangely disorientating experience for those travelling up the estuary.Footnote 246 Nestled deep within the eastern bosom, at the head of a sinuous estuary, Bede implies that urbs Giudi was concealed within the interior. The obscurity of Bede’s reference to urbs Giudi may be a playful reflection of the perceived remoteness or hidden nature of this place, at least when viewed from a Northumbrian perspective.
Another transferred and figurative sense of sinus is also of potential relevance to urbs Giudi. Urbs Giudi presumably represents a latinised form of a native place-name. As Giudi is a Celtic place-name,Footnote 247 urbs may represent a latinised form of a Northern Brittonic place-name element.Footnote 248 A possible candidate here may be dīn ‘“a fort”, often — but not necessarily — a hill-fort’.Footnote 249 If so, this may imply that Bede employs sinus in a fifth sense, as ‘a refuge, place of safety’, ‘protection, asylum’. Paralleling this, dīn also means ‘fortress, place of refuge’,Footnote 250 while it is suggested that ‘[p]erhaps “a place of refuge” comes closest to the core sense’ of dīn. Footnote 251 In addition, the Forth Estuary, particularly the Links of Forth, provided sheltered berths for shipping.Footnote 252 The natural defensive strength of Castle Rock, Stirling, the fortified character of an urbs and its strategic domination of the surrounding area reinforce the status of this location as ‘a refuge, place of safety’.
The multiple senses in which Bede appears to have used sinus invite analysis of its other meanings to see if any of these are also applicable to urbs Giudi. In what may be another play on words, sinus ‘in the power or possession of ’ may imply that urbs Giudi was a socio-political, perhaps royal, centre. This is consistent with its status as an urbs and fits the context of urbs Giudi as the eastern counterpart of Alcluith, the fortified power centre of the kingdom of Dumbarton and, later, Strathclyde. However, the socio-political context of urbs Giudi lies beyond the scope of this paper.
CONCLUSIONS
Generations of scholars have struggled to reconcile the apparent inconsistencies between the location Bede gives for urbs Giudi, the topography of the Firth of Forth, and the absence of any evidence for an early medieval fortress there. Previous attempts at rationalising this dilemma have resorted to the claimed limitations of Bede’s knowledge of the Forth.Footnote 253 This not only underestimates Bede’s skills as a geographer and his access to first-hand sources on this region, acquired from members of the community of Abercorn, but also overlooks his command and playful use of Latin. An alternative approach suggests that the Historia ecclesiastica contains subtle clues to the location of urbs Giudi. Analysis indicates that in medio and sinus are used by Bede in elliptical, presumably deliberately ambiguous, senses that have eluded interpretation previously. In particular, sinus carries a wide range of literal, topographical and metaphorical meanings. Like sinus fari, the use of sinus to describe the location of urbs Giudi may be identified as another example of Bede’s wit and wordplay. Bede gives a playfully-imaginative location for urbs Giudi, rich in metaphor and allusion. This may be interpreted as a display of Bede’s learned knowledge, of both geography and Latin, and was perhaps intended to impress and/or raise a knowing eyebrow from those who knew urbs Giudi by experience or report.
Acknowledging the complexity of Bede’s text and the prominence of wordplay in this passage yields new insights into his reference to urbs Giudi, enabling it to be viewed from fresh literary, metaphorical and topographical perspectives. The ambiguity, lexical range and subsidiary and figurative senses of in medio and sinus have fundamental implications for the location of urbs Giudi. In medio, in this context meaning ‘deep inside, far along, at the deepest or furthest point, in the heart of’, or similar, indicates that urbs Giudi was located not ‘in the middle of’ or ‘halfway along’ the Firth of Forth but further upstream and, therefore, deeper inland. This is supported by the lexical range of sinus, which is more applicable to the Forth Estuary, particularly the sinuous Links of Forth, than the Firth of Forth. The effect of this is not simply to extend the area within which urbs Giudi was located but to shift the focus of the search for it away from the Firth of Forth to the upper reaches of the Forth Estuary. Although the range of meanings held by both in medio and sinus is difficult to capture in translation in a single phrase, Bede evokes a place nestled ‘deep inside the eastern bosom’ but also ‘far along/at the deepest point of the eastern meander’. The sense is perhaps phrased more naturally, but ambiguously, in translation as ‘at the head of the eastern estuary’ or ‘at the top of the eastern inlet’, implying a place at or near the tidal limit of the Forth.
Only one location for urbs Giudi on the Forth Estuary has been proposed previously: Stirling. Stirling is sited in a sinus ‘bosom’ by virtue of being nestled within the folds of the meandering Links of Forth and located on the valley floor, framed by surrounding hills. Bede’s terse phrase evokes the distinctive landscape setting of Stirling. Castle Rock, Stirling, is located only 2 km south-south-east and downstream from the normal tidal limit of the Forth Estuary. The presence of a harbour at Stirling is first recorded in 1147,Footnote 254 although it may have had much earlier origins, and it thrived until the First World War, with ships trading as far afield as Archangel and the Baltic Sea.Footnote 255 Commonly regarded as the head of navigation on the Forth, Stirling Harbour, at Forthside, is 62 miles from the mouth of the Firth of Forth at the Isle of May. As one might expect, given his background and the probable origin of his source(s), Bede’s perception of urbs Giudi is essentially a littoral one, approached from the Forth estuary and through the Links of Forth. This perspective is now unfamiliar, as Stirling’s status as a port declined after 1914,Footnote 256 the last merchant vessels docked there during the 1950s,Footnote 257 and this stretch of the Forth is rarely used for boating now.
This paper does not prove the identification of urbs Giudi with Stirling. However, if its analysis is accepted, it excludes all previously-proposed sites for urbs Giudi with the exception of Stirling. It does this without requiring any criticism of Bede’s sources, his understanding of them, or imaginative massaging of Bede’s meaning to explain how Stirling may be described as in medio sui, referring to the eastern sinus. Stirling not only presents a very plausible site for urbs Giudi but also one which may now be recognised as fitting the location Bede gives for it. This reassessment has reached the same conclusion for the location of urbs Giudi as Graham and Jackson, but as a result of a very different reading of the evidence and using an alternative approach, based on a comparative analysis of Bede’s use of in medio sui and sinus, and taking their lexical range and elliptical senses into account.
The fortified place par excellence at Stirling is, without doubt, the imposing Castle Rock, with formidable natural defences on all but its narrow eastern side. This provides an ideal site for an early medieval urbs. However, no evidence of an early medieval presence on Castle Rock has been discovered, at least so far, presumably because of its continued occupation since at least the twelfth century. On the basis of the text of the Historia ecclesiastica, the terminology Bede uses, and on topographic grounds, the best candidate for the location of urbs Giudi is Castle Rock, Stirling. Although two other sites for urbs Giudi at Stirling have been proposed,Footnote 258 neither of them are as attractive a site for an urbs because they do not offer the strong natural defences that Castle Rock possesses.
This paper is unlikely to be the final word on the identification of the controversial urbs Giudi but it is hoped that it places the debate on a new footing by identifying a wider range of meanings in, and clarifying the intended senses of, Bede’s terse and ambiguous phrasing. When taken together, these point to the identification of Bede’s urbs Giudi with Stirling, where the most plausible location for an early medieval fortification is Castle Rock.Footnote 259
