The Journal of Scottish Name Studies Vol. 4

The Journal of Scottish Name Studies Vol. 4

edited by Richard A.V. Cox and Simon Taylor

Clann Tuirc 2010 The Journal of Scottish Name Studies4 (2010) edited by Richard A.V. Cox and Simon Taylor

First published in Scotland in 2010 by Clann Tuirc, Tigh a’ Mhaide, Ceann Drochaid, Perthshire FK17 8HT Printed in Wales by Gwasg Gomer, Llandysul

ISSN 1747-7387

© text: the authors 2010 © book and cover design: Clann Tuirc 2010

All rights reserved. No part of this book can reproduced in any form, or by any means, known or otherwise, without the prior consent of the pubisher.

The Journal of Scottish Name Studies JSNS is a peer-reviewed journal that exists to publish articles and reviews on place and personal names relating to Scotland, her history and languages. JSNS is published with the endorsement of The Scottish Place-Name Society, for whose members a discount is available – the Society web site is .

Editors Professor Richard A.V. Cox and Dr Simon Taylor

Reviews Editor Gilbert Márkus

Editorial Advisory Board Professor Dauvit Broun Dr Rachel Butter Professor Thomas Clancy Mr Ian Fraser Dr Jacob King Mr Gilbert Márkus Professor W. F. H. Nicolaisen Professor Colm Ó Baoill Dr Maggie Scott Mr David Sellar Dr Doreen Waugh

Subscriptions Visit , or contact the publisher by e-mail at fios@ clanntuirc.co.uk, or by post at the above address.

Contributions Prospective contributors to the Journal should refer to the Notes for Contributors, available from the publisher and at . Contents

A Grammar of Manx Place-names George Broderick 1

Memories, Meids and Maps: the Shetland Place Names Project Eileen Brooke-Freeman 43

Scottish Gaelic Sannda and Its Aliases Richard A.V. Cox 61

Scotland’s -ham and -ingham Names: a reconsideration Alan G. James 103

The Shadow of ‘Onomastic Graffiti’ Denis Rixson 131

Varia Jacob King Aberkarf 159

Review Henry Gough-Cooper Terry Kinder, ed., Life on the Edge. The Cistercian Abbey of Balmerino, Fife (Scotland) 169

Bibliography for 2006–2009 173 Simon Taylor

Notes on Contributors 187

A Grammar of Manx Place-names1

George Broderick University of Mannheim

1 Introduction From August 1989 to July 2005 a systematic collection of place-names from both oral and documentary sources was undertaken in the Isle of Man under the auspices of the Manx Place-Name Survey. This resulted in the publication by Max Niemeyer Verlag, Tübingen, of Place-Names of the Isle of Man in seven volumes (Broderick 1994–2005). The material netted was considerable, both in the Scandinavian as well as in the Goidelic (Gaelic) component, the latter being much greater than the former. Gillian Fellows- Jensen (Fellows-Jensen 2005) has provided an extensive linguistic outline of the Scandinavian component. The remaining task is to provide the same for the Goidelic component. The presence of some five ogam stone inscriptions in Man (MacManus 1997, 44 §4.2(c)) testifies to Goidelic speech there from c. 500 AD until the end of the seventh century.2 Before then British speech was apparently spoken there (cf. Jackson 1953, 173) until the sub-Roman period and perhaps thereafter. The archaeological record tells us of the presence of a significant number of keeill-sites relating to the early Christian church in Man (sixth to seventh centuries; cf. Kermode 1909–18), as well as some 15 ring-forts of the type found in Ireland during the sixth to eighth centuries AD,3 suggesting that Goidelic speech was also present in Man at that time.4 The language of the population of Man at the time of the Scandinavian arrival (ninth century) was likely Goidelic. As we have noted, British was evidently spoken, but with the incoming of Goidelic settlers from Ireland around 500 AD the population would likely have become predominantly Goidelic-speaking by the onset of the Scandinavian period (c. 900 / 925–

1 This is an expanded and revised version of the same as appears in the Introduction to Vols 1–6 of PNIM. 2 Two further ogam stones from Man belong to the ‘scholastic’ type (McManus 1997, 170, note 2 and 130, §7.5(ii)), dated by R. I. Page (1983) to not before the mid-12th century. 3 Cf. Davey (2005, 335). His map shows 31 ring-forts of which 10 have so far been identified by aerial photography and five are ‘possible’. No comment is made about the others. See also summary reports in Medieval Archaeology 19 (1975), 230–31; 20 (1976), 174; 21 (1977), 216. 4 The Goidelic period in Man traditionally dates from c. 500 AD, on the basis of the five ogam inscriptions noted above, to c. 900 AD, the date of the first known Scandinavian settlements in Man (Wilson 2008, 25).

The Journal of Scottish Name Studies 4, 2010, 1–42 2 George Broderick 1266).5 Though there seem to have been some pre-Scandinavian Goidelic (and British?) place-names in Man (Broderick 2008), it seems certain that Goidelic continued in use throughout the Norse period (Thomson 1983) and thereafter became the main language of the native population until c. 1880–1920, when it was replaced by English. In this regard a number of names show developments deriving from English pronunciation. The extent of the material here gives us a clearer insight into earlier stages of Manx Gaelic, showing regular developments as found in other branches of Goidelic in Ireland and Scotland. The material admits of the following aspects of Manx grammar relevant to place-names, which, as we shall see, spans the entire gamut of Goidelic speech in Man, from its arrival c. 500 AD to its demise and obsolescence during the 19th and 20th centuries.

2. The structure of Manx place-names 2.1 Simplex place-names consisting of a noun only Simplex place-names of Goidelic origin also occur in Man and mostly seem to have originated as the names of natural and topographical features, surviving as treen, quarterland, field, mountain names etc. They stand on their own. Apart from the name Man itself 6 (cf. PNIM VII, 337–38, Broderick 2008, 166–67), the following two names in Man are, in Manx terms, of some antiquity:

Appin Appyn in Glenfaban 1376–77 PA ‘abbeyland’ (ScG apainn)7 (PNIM I, 19). See also Nappin (an apainn) JU (PNIM II, 257ff.), LE (PNIM III, 441). Rushen russin 1257 RU (diminutive of ros (Ir. ros, roisean), OIr. ross < PCelt. *φrostos > PBrit. *rossos (cf. Schrijver 1995, 407; PNIM VI, 165–66) ‘any kind of elevated area of land, e.g. hill, plateau’, cf. with rhosfa (< *rhos- ma) ‘mountain pasture’, Skt prº-stha ‘plateau’. However, judging from its

5 The year 1266 marks the formal ending of Norwegian conrol in Man, after Man and the Hebrides were sold in that year to the kingdom of Scotland. For details of the Scandinavian period in Man see Wilson 2008, McDonald 1997, Sellar 2000, Broderick 1995 and 2003. For an assessment of the Norse on the 26 known rune-stones in Man, see Page 1983. For an appraisal of the continuity of Goidelic speech in Man throughout the Scandinavian period, see Megaw 1976 (1978) and Thomson 1983. 6 (< IE *men, *mon ‘protrude, rise, cf. W. mynydd ‘mountain’ < *monijo- (cf. LEIA, s.v. muin), with suffix *-aua, the sense being ‘mountain island’ or ‘high island’, i.e. protruding out of the sea, as seen from the sea or from adjacent coasts (cf. also PNRB 411; Sims-Williams 2000, 7; Schrijver 1995, 96). 7 Manx examples are given their equivalent in either Irish (Ir.) or Scottish Gaelic (ScG), mostly the latter. A Grammar of Manx Place-names 3 distribution pattern, ros in place-names frequently means ‘promontory’. Meanings such as ‘wood, copse, level tract of arable land’ etc. are secondary.8

There is no indefinite article in Manx, so baatey can mean ‘boat’ etc. or ‘a boat’ (ScG bàta), etc. The (definite) article could also be implicit inan element of this type, e.g. Mx Rushen ‘the hill, plateau’, ScG beannan ‘the peaked place’. Nouns standing alone also function as place-names in English and Scandinavian place-names, with the element of definiteness implicit and not stated, even though many languages normally employ a definite article, cf. Brooke (Leicester, Norfolk; OE brōc), Breck (Lancashire; Scand. brekka), Dal (Denmark; Scand. dalr).9 Sometimes the simplex names in Man are latterly accompanied by the English definite article. The recorded examples include the following:

Barra Bara 1920 PA ‘channel through which boats reach the Niarbyl mooring’, cf. Ir. barra ‘a sand-bar running into the sea’ (PNIM I, 52). Bollagh the Bollough 1704, Bollaugh 1709 GE ‘pass, track between curragh and mountain land’ (ScG bealach) (PNIM I, 210). Carraghan (mountain) Carraughan 1707 BN ‘(place of) rocks, rocky place’ (Ir. carrachan) (PNIM V, 76). Carran (mountain) fo’n Charron (‘under the Carron’) 1868, Carran 1898 RU ‘crown of the hill’ (ScG caran) (PNIM VI, 371). Carrick (mountain) (the) Carrick 1704 LE, AbQL ‘rock (on land)’ (Ir. carraig) (PNIM III, 317). Carrick The Carrick1957 RU ‘rock (in bay)’ (Ir. carraig) (PNIM VI, 372), Yn Carrick [n kark] 1929 MA (PNIM IV, 72). Chibberan the Chibberyn 1923 PA ‘place of wells’ (OIr. tiprán) (PNIM I, 67). Croggane Balla Crokane 170410 Crogane 1739 AN ‘little hill (farm)’ (Ir. cnocán) (PNIM III, 99), with voicing of intervocalic /k/ in the Manx reflexes. Cronk Cronk 1688–89 BA ‘hill’ (Ir. cnoc) (PNIM II, 163). Driney Dryny 1704 GE ‘place of blackthorns’ (Ir. draigheanaigh) loc./dat. (PNIM I, 245). Druin Drinn 1704 LE ‘humpback, hill, ridge’ (ScG dronn, gen. druinne, dat. druinn) (PNIM III, 377). Gilcagh TR Gilcagh 1515, Guilcagh QL (the) Guilckaugh 1704 AN ‘(place of) broom’, Mx giucklagh (C., 80j) (Ir. giolcach), with metathesis of -c- and

8 Cf. Mac Giolla Easbuig (1981,154–155). 9 Robert L. Thomson and Gillian Fellows-Jensen, pers. comm. August 1993. 4 George Broderick -l- in the Manx dictionary form (PNIM III, 122, 124). Glannan (the) The Glannan 1735 GE ‘the glen area’ (ScG gleannan) (PNIM I, 261), cf. ScG beannan, above. Rheast Rheast [ri:st] 1990 GE ‘wasteland’ (ScG riasg) (PNIM I, 261 et passim). Rhennie Renny 1704 GE ‘ferny place’, Mx rhennee (ScG rainich) loc./dat. (PNIM I, 285). Scigheaghin Scigheaghin 1744 PA ‘(place of) thornbushes’ (ScG sgitheachan) (PNIM I, 158). Sloc slock 1898 RU ‘pit, depression’ (Ir. sloc) (PNIM VI, 492) etc.

2.2 Close-compound names A common place-name formation found in Continental Celtic as well as in British and Goidelic is the nominal compound of the structure noun + noun (determinans + determinatum), in which the first noun qualifies or defines the second, e.g. Gaulish: Lindumagos ‘water-field, field of water’ (River Limmat, Germany), Rigomagos ‘kings’ plain, plain of king(s)’ (Remagen) (Zimmer 2004, 207), British: Camulodunon ‘the fortress of (the war-god) Camulos’ (Colchester) (PNRB 294–95), Durovernon ‘alder-swamp of / by the fortress (on a plain), fortress alder-swamp’ (Canterbury) (PNRB 353). Also a feature of this structure is adjective + noun, whereby the qualifying adjective is pre-posed, i.e. the first element defines the second. Again in Gaulish: Noviomagos ‘new field’ (Nijmegen), Vindobona ‘white, bright ground’ (Vienna) (Zimmer 2004, 207); British: Camboglanna ‘crooked (river)bank’ (Castlesteads, Hadrian’s Wall) (PNRB 293–94), Vindolanda ‘white, bright area’ (Chesterholm, Hadrian’s Wall) (PNRB 502). Compound names are also found in Irish place-names, though seemingly not as frequently as in Gaulish or British, e.g. Darmhagh (< *daromagos ‘oak-plain, plain of oaks), though bearing Goidelic initial stress (cf. Mac Giolla Easbuig 1981). A small number of Goidelic close-compound place-names of both types and bearing initial stress has survived in Man.11 Recorded examples include the following:

2.2.1 Noun + noun Here the first noun occasions lenition, where appropriate, in the following noun:

10 Balla has likely been prefixed here specifically to refer to the farm, rather than the hill. 11 For details of close-compound place-names in Ireland, see Mac Giolla Easbuig 1981 and Tempan 2009. For Islay and Jura, see Drummond 2009; for Lewis, see Cox 2002, 16–19, 33. A Grammar of Manx Place-names 5 Mooragh Mooraugh 1685 MA [murx], [murk] 1989–92 ‘a level strip of land along the sea-coast’, OIr. murbach, muirbech12 (DIL M, s.v.), Ir. muirmhagh (Di. 770), cf. with morfa < PCelt. *mōri-magos ‘sea-plain’ (cf. PNIM IV, 196; also LE PNIM III, 438).

2.2.2 Adjective + noun Here the adjective occasions lenition, where appropriate, in the following noun:

Doolough Dufloch1280 13, the Dock 1808, Dock 1841 LE [dulx], [dulk] 1989–92 ‘black lake, marsh’ (ScG dubh loch) (PNIM II, 239, III, 378). Douglas dufglas 1257 ON [dgls] 1989–92 ‘black / dark water’, Ir. dubhghlais < PCelt. *dubo-glassio (PNIM VII, 29).14

A perhaps later development involving close-compounds concerns names bearing forward stress on the second syllable of the compound (as marked, below). However, there is no evidence that these structures were created with a forward stress pattern: as all examples consist of a qualifier + generic, the stress would originally have lain on the first syllable (as above), while late analogical levelling lies behind the current stress pattern (as in §2.2.2.3, below). The recorded examples include the following:

2.2.2.1 Adjective + noun with attested forward stress Arderry Ardarry 1704 [adri] 1990–91 BN ‘high wood, copse’, Mx ard- derry (ScG àrd doire) (PNIM V, 26). Ardwoailley Ard-willy 1838 [adwli] 1991 ML ‘high fold’, Mx ard-woailley (ScG àrd bhuaile) (PNIM VI, 183), Ardwillagh [dwljax] 1991 LO (PNIM IV, 219). Breckwoaillee Breckyoaly 1709 PA ‘speckled fold’, Mx breck-voaillee (Ir. breac- bhuailidh) (PNIM I, 59), [brekwli] 1991 ML (PNIM VI, 183). Camlorge, Camlork TR Camlorge 1507, QL Camlorge 1704, Camlorke 1719 [kam lk] 1991 BN ‘crooked ridge, path’ (ScG cam lorg) (PNIM V, 75). Corvalley Corvalley 1643 [kvalj] 1991 RU ‘remote farm, farm on / by a round hill’ (Ir. corr-b(h)aile) (PNIM VI, 385).

12 The second element in -bach, -bech may represent OIr. fótbach, gen. sing. often fótmaige, -maig, perhaps through association with mag ‘plain’, lit. ‘sod-breaking’ (DIL F, s.v. fótbach), then referring to raised ground. 13 The -f- would represent the medial spirant /v/ before a liquid (see foregoing footnote). Vocalisation of such spirants was generally complete by the 17th / 18th century. 14 Not found in Man, but used there, is Mx Divlyn, Ir. Duibhlinn ‘black / dark pool’ (> Dublin), with survival of the medial spirant before a liquid (see previous footnote). 6 George Broderick Corvonagh Corvonny 1737, Corvonnagh 1740 [kvon] 1991 MR ‘remote turbary’, Mx corvonagh (Ir. corr + m(h)óna(i)dh) (PNIM V, 185–86). Lheakerroo Leykerrow 1692, the lhea Cherroo 1779 [likro] 1991 PA ‘half quarterland’, Mx lieh-cherroo / kerroo (ScG leth-cheathramh) (PNIM I, 139), with or without lenition. Shennvalley shennvalley 1704 [s′nvalj], [s′nval] 1991 ML ‘old farm’, Mx shenn valley (ScG seann bhaile) (PNIM VI, 171).

2.2.2.2 Adjective + noun with probable forward stress (starred) Cam-*woillee Cam-woillee 1796 MA ‘crooked fold’, Mx cam-woaillee (Ir. cam-bhuailidh) (PNIM IV, 93). Clagh-*Wooilley Clagh-Wooilley 1765 MA ‘stone fold’, Mx clagh-woailley (ScG clach bhuaile) (PNIM IV, 126). Doo*halloo Doo-Halloo 1761 BA ‘black land’, Mx doo-halloo (ScG dubh thalamh) (PNIM II, 117), BR (PNIM III, 192), RU (PNIM VI, 413). Gar*woailley Garwooley 1841 MA ‘short fold’, Mx giare-woailley (ScG geàrr bhuaile) (PNIM IV, 32). Glass*thalloo Glas-Thallow 1778 MA ‘grey / green land’ (ScG glas talamh) (PNIM IV, 128). Glass*voailley Glass-volley 1722 RU ‘grey / green fold’, Mx glas-voailley (ScG glas bhuaile) (PNIM VI, 258). Min*gary Mingary 1830 BN ‘small garey (sourland)’, Mx myn-g(h)arey (ScG mion gheàrraidh) (PNIM V, 27), wingerry 1783 BN (ibid.), with lenition and suppressed Manx article. Shenn*awin Shinown 1946 AN ‘old river’ (ScG seann abhainn) (PNIM III, 155). Shenn*lhargee Shenn Largy 1788 LE ‘old hillslope’ (ScG seann leargaidh) (PNIM III, 299). Shenn*thalloo Shen thalloo 1770 GE ‘old land’ (ScG seann talamh) (PNIM I, 189).

2.2.2.3 With adjective following its noun Ballacoar Ballacoare 1703 LO ‘remote farm’, Mx balley coar (ScG baile còrr) (PNIM IV, 230). Bwoaill’ Ard Ballerd 1920 PA ‘high fold’, Mx (ScG buaile àrd) (PNIM I, 43). Crockan Minna Crockan Minna 1693 AN ‘little hills’, Mx cronkyn mynney (ScG cnocan miona) (PNIM III, 98), with absence of nasalisation in Crockan. A Grammar of Manx Place-names 7 Slieau Corr Sliew Corr 1725 MA ‘rounded, remote hill’, Mx slieau corr (ScG sliabh còrr) (PNIM IV, 155).

In Manx, as in Insular Celtic in general, there was a tendency to abandon the close-compound noun / adjective + noun in favour of the parallel structure of noun + defining genitive. In Irish, internal evidence suggests that the shift away from compounds had begun at an early stage, since very few such compounds occur in the earliest literature (leaving aside the obvious learned calques on Latin words). What is of interest is the number of the Irish noun + noun type that have cognates in Welsh, indicating perhaps that many such compounds are survivals from the Common Celtic period (Mac Giolla Easbaig 1981, 152). The available evidence (Mac Giolla Easbuig 1981, Tempan 2009) suggests that the nominal compound had more or less ceased to be productive by the early Old Irish period (c. 700–725 AD) and offers a date of around 400 AD for the introduction of non-compound names.15 Anders Ahlqvist (1985, 4) links this change with the establishment of VSO as the unmarked order of Irish sentences, which is dated as being not later than 700 AD. This, if correct, would likely explain the few attested survivals of compound names in Man, i.e. that this type of name was becoming obsolescent during the period of initial Goidelic settlement in Man (c. 500 AD) but survived long enough to leave behind some examples.

3 The morphology of nouns in Manx place-names Nouns in Manx deriving from original Old Irish masculine, feminine and neuter nouns come down either as masculine or feminine. From the Classical period of Manx (18th century, essentially that of the Manx Bible translation) down through to Late Manx (19th and 20th centuries) nouns came more and more to be treated as masculine only, unless obviously otherwise, e.g. ben ‘woman’, or marked as feminine, for example in nominal phrases such as fud ny h-oie ‘throughout the night’ (ScG air fad na h-oidhche). Manx nouns used in Manx place-names often reflect original Goidelic acc. / dat. singular forms expressing ‘motion towards’ or ‘rest’ – in place- names there may, of course, be loss of a locative preposition. The list includes the following:

15 Oliver Padel (1985, xv–xvi) warns against too rigid a deliniation between the type noun + noun and noun-phrasal names. Examples from British demonstrate quite some overlap in time. 8 George Broderick Mx m. dat. thie ‘house’, OIr. nt. nom. / acc. tech gen. tige / taige, dat. tig / taig (neuter s-stem) (DIL T, s.v. tech; GOI §338) (ScG taigh). Mx f. nom. bwoailley, gen. bwoaillagh, dat. bwoaillee ‘cattle fold, pen’, OIr. nom. buaile, gen. buailed, dat. buailidh (also ModIr.) (dental stem) (DIL B, s.v. búaile). Mx f. dat. (later nom.) keeill ‘cell, church, chapel’, n. *kial, gen. killey, OIr. f. nom. cell, gen. cille, dat. (later nom.) cill (fem. ā-stem) (DIL C, s.v. cell ). The original nominative form is to be seen in the plural kaltin, kaltiin (Phillips; GEM 299), kialteenyn (C. 108). Mx f. nom. lhargey / liargey, gen. lhargagh / liargagh, dat. lhargee / liargee ‘hillside, hillslope’, OIr. f. nom. lerg, lerga, gen. lerge, dat. leirg, leargaidh (ā, iā-stem) (DIL L, s.v. lerg). Mx f. nom. moain, gen. moaney, dat. moanee ‘bog, peat-moss, turf, turbary’, OIr. f. nom. móin, gen. móna, -adh, acc. / dat. mónai (i-stem). In later MidIr. also a dental stem: gen. mónad, dat. mónaid (DIL M, s.v. móin). Mx f. dat. (later nom.) traie ‘shore’, OIr. m. and f. dat. (later nom.) tráig, also trág (o-stem m.), gen. trágha, dat. tráig (i-stem) (DIL T, s.v. tráig) (ScG tràigh).

With locative preposition: Niarbyl PA ‘to / at the tail’ (Ir. i n-earball ) (PNIM I, 145–46).

3.1 Sing. article with original masc. noun

3.1.1 With Manx article: The singular article in Manx is y / yn which occasions no change in the initial consonant of the following masculine noun. The masculine genitive of the article, viz. y / yn, occasions lenition (spirantisation) in the initial consonant of the following noun, except dentals. In original masculine nouns in Manx with initial vowel, the original -t of the article does not appear in the Manx reflexes, e.g. Mx yn ean ‘the bird’, ScG an t-eun (cf., also, Yn Ellan, below), as though Manx had generalised the feminine inflection.

Inclose Inclose 1709 BA ‘the enclosure’, Mx yn close (Ir. an clós) (PNIM II, 185). See also §3.1.2. Yn Carrick 1929 MA ‘the rock’ (ScG a’ charraig) (PNIM VI, 72). See also §2.3. Yn Ellan Yn Ellan 1814 AN ‘the island’, Mx yn ellan (ScG an t-eilean) (PNIM III, 117). A Grammar of Manx Place-names 9 3.1.2 With English definite article However, in many modern place-name forms of this type the English definite article the has replaced Mx y / yn and this has now come to be fairly standard in Manx place-names. Occasionally, the accompanies y/yn where the latter has become inseparable from its noun (see The Inclosebelow; see also §3.1.1. above):

The Carrick The Carrick 1957 RU ‘the rock (in bay)’ (ScG carraig) (PNIM VI, 372). See also §2.3. The Garey the garey ground 1642, gary 1709 LE ‘the garden or enclosure’ (ScG gàrradh) (< ON garðr) (PNIM III, 384–85). The Inclosethe Inclose 1704 BA ‘the enclosure’ (with the Manx article attached to its noun) – unless this name is to be taken as Eng. ‘The In-Close’ as opposed to the ‘Out-Close’ (PNIM II, 185). See also §3.1.1.

3.2 Sing. article with original fem. noun

3.2.1 With Manx article In the above examples the misdivision of the article arises in Manx, but the sense of the article, being present or required, is continued on in English.

Neary Nary 1715 LE ‘the shieling, summer-pasture’, Mx yn eary (ScG an àirigh) (PNIM III, 445). Naaie (passim) ‘the homefield, flatt’, Mxyn aaie (ScG an fhaiche).

3.2.2 With English definite article The CregThe Creg1957 GE ‘the rock’ (ScG a’ chreag) (PNIM I, 228). The Neary the Nary 1737, the Arie 1764 PA ‘the shieling’, Mx yn eary (ScG an àirigh) (PNIM I, 100–01). The Lhargan The Liaggin 1750 GE ‘the hillslope’, Mx lieckan (ScG leacann) (PNIM I, 208). The NaaieThe Naaie 1990 MI (also with Mx def. art.) (PNIM II, 22). The Nyarn The Nyarn 1740 GE ‘the share, portion’, Mx yn ayrn (ScG an earrann) (PNIM I, 190), also with Mx def. art. and with Ny- indicating a palatal n- before a palatal vowel.

3.2.3 Lenition of fem. noun after Manx article Originally the Manx article lenited the initial consonant of a following original feminine noun, e.g. yn chreg ‘the rock’. However, during the course of the latter 10 George Broderick part of the 18th and into the 19th and 20th centuries lenition in original /k/, /g/, /d/, viz. /x/, //, // (and their palatalised variants, viz. /x′/, /′/, /′/) in any position began to lose their spirantisation (and palatalisation) and return to their radical forms, viz. /k/, /g/, /d/. Original feminine nouns in initial /s/- show t-lenition after the nominative / accusative of the article. Nouns in palatal /s′/ + vowel are sometimes prone to this whether they are originally feminine or not. The article itself (as here) may sometimes be suppressed:

(Yn) Chillagagh /t′s′ilgrx/ (sic) 1990 BA ‘the place of the willows’ (Ir. an t-seileagach) (PNIM II, 139), cf. Shellagagh (PNIM II, 199).

This also applies to the preposed adjective shenn ‘old’, ScG sean(n), when qualifying an original feminine noun (see also §2.2.2.2).

(Yn) Chenn Reinnagh Chen-rennagh 1764 MI ‘the old bracken area’ (ScG an t-seann raineach) (PNIM II, 30). The Chennn Chroit(the) chan chroit 1784 BA ‘the old croft’ (ScG an t-seann chroit) (PNIM II, 141), with Eng. def. art. reinforcing the suppressed Mx def. art.

3.3 Original masc. noun with adjective In Manx the qualifying adjective normally follows the noun (but see §2.2). After original masculine nouns no lenition occurs in the following adjective. So we find:

Bayr Dowin Bare Dowin 1794 LE ‘steep road’ (Ir. bóthar domhain) (PNIM III, 307). Cronk Beg Knock begg 1687 AN ‘big hill’ (ScG cnoc beag) (PNIM III, 106). Cronk Breck Knocke brecke 1636 AN ‘speckled hill’ (ScG cnoc breac) (PNIM III, 108). Ellan Bane Ellanbane 1662 LE ‘white island’ (ScG eilean bàn) (PNIM III, 381).

However, in the latter stages of the language ‘abuse’ of lenition can take place in attributive adjectives. So we find:

Cronk Vane Cronk Vane 1929 PA ‘white hill’ (ScG *cnoc bhàn) (PNIM I, 89). Cronk Vooar Cronk Vooar 1822 PA ‘big hill’ (also Cronk Mooar) (ScG *cnoc mhòr) (PNIM I, 89). A Grammar of Manx Place-names 11 Mullagh Vedn molagh bane road 1834, Mollagh vedn 1920 PA ‘white summit’, Mx mullagh bane (ScG mullach bàn) (PNIM I, 143). The latter is the name of a ‘rough road with white boulders at the top leading toward Slieu-whallin from Glenmaye’. The entry for 1834 makes clear the expected form. Cronk Vane Beg Cronkvane beg 1704 GE ‘little white hill’ or ‘little Cronk Bane’ (ScG *cnoc bhàn beag) (PNIM I, 237). Here Cronk Vane is a unit treated as masculine (and vane presumably had no gender) qualified by the adjective beg in contrast, real or implied, to *Cronk Vane Mooar.16 Cf. Traie Vrish Beg, below. Traie Vrish Beg / Mooar Traie Vrish beg / mooar 1930 PA ‘little / big broken shore / Traie Vrish’ (ScG tràigh bhriste + beag / mòr) (PNIM I, 169). Though the verbal adjective vrisht makes clear that traie is here treated as feminine, the unit traie vrisht is treated as if masculine, as beg and mooar indicate. Lhargey Vane Meanagh Liargy Vane Meanagh 1775 GE ‘middle Lhargee Vane (white hillslope)’ (ScG leargaidh bhàn + meadhanach) (PNIM I, 189).

3.4 Original fem. noun with adjective Attributive adjectives to original feminine nouns follow their noun and are lenited. Failure or absence of lenition begins to be found during the course of the nineteenth century.

3.4.1 With regular lenition Awin Veg (Awin) Veg 1929 MR ‘little river’ (ScG abhainn bheag) (PNIM V, 148). Bwoailley Veen Boalley veen 1755 MI ‘level fold’ (ScG buaile mhìn) (PNIM II, 97). Clagh Vedn Clagh vedn 1963 BR ‘white stone’, Mx clagh vane (ScG clach bhàn) (PNIM III, 216), with preocclusion (see §3.27 below). Creg Veg Craig Veg 1834 GE ‘little rock’ (ScG creag bheag) (PNIM I, 228). Creglhea, the the Creglhea 1800 PA [krgle] 1991 ‘at the grey rock’, cf. Ir. creag liath, loc. / dat. without preposition creig léith, as suggested by the pronunciation (PNIM I, 84). Croit Cheyll Croit Kheik 1787 GE ‘narrow croft’ (ScG croit chaol ) (PNIM I, 211). Note that Kh- would indicate lenition in the adjective. Croit Vane Crott Vaane 1768 BN ‘white croft’ (ScG croit bhàn) (PNIM V, 89). Faaie Ghressagh Fhai-Chressagh, Fhai-yhressagh 1757 JU ‘briary flatt’ (ScG

16 Contrastive usage is also found in Carloway, Lewis (cf. Cox 2002, 52–53), though different from that found in Man. 12 George Broderick faiche dhreasach) (PNIM II, 258). Immyre Ghoo Immyr Ghoo 1754 MI ‘black ridge’ (ScG imir *dhubh) (PNIM II, 97). This word is treated as masc. in ScG. Lagwoailley Veg / Vooar Lagwilley veg / woar 1837 GE ‘little / big Lag Woailley (fold hollow)’ (ScG lag bhuaile bheag / mhòr) (PNIM I, 261). Lheeannee Hirrym Leni herim 1704 GE ‘dry meadow’ (cf. Ir. léana, ScG lèana + tioram) (PNIM I, 273). Léana / lèana is treated as masc. in Ir. and ScG, but usually as fem. in Mx. See also next. Lheeanney B / Vroghe Leane brough or Dirty Meadow 1711, Leaney Vroagh 1752 BA ‘dirty meadow’ (Ir. léana bhróghach) (PNIM II, 191). Lhargee Vollee Lhargy-Volly 1830 GE ‘pass slope’ (ScG leargaidh bhealaich) (PNIM I, 272). Ooig Vooar Ough wooare 1920 PA ‘big cave’ (ScG ùig mhòr) (PNIM I, 148). The formough could also be (Mx) oghe ‘oven’ (cf. ScG aghann), not unsuitable as a name for a cave.

3.4.2 With failure or absence of lenition Creg Bane Creg Bane 1955–57 PA ‘white rock’ (ScG creag + bàn) (PNIM I, 76). Eary Mooar Eary Moar 1830, 1861 and thereafter GE ‘big shieling’ Eary Voar 1851 (ScG àirigh + mòr) (PNIM I, 249). Faaie Dressagh Faiy Dressagh 1708 JU ‘briary flatt’ (ScG faiche + dreasach) (PNIM II, 241). Ooig Dorraghey Ooig Dorraghey 1930 PA ‘dark cave’ (ScG ùig + dorcha) (PNIM I, 147).

Failure of lenition in initial /f/, /d/, /g/, /k/, /s/ begins to be seen in Classical Manx (18th century) and is complete by the period of Late Manx (mid-19th century). As this came to be the rule we should speak of ‘absence’ rather than ‘failure’ of lenition, since its occurrence was no longer expected. Note that in Standard Manx Orthography dh, th represent a form of interdental articulation, viz. [dh], [th] and not [] or [h] (or palatalised variants thereof), as in Gaelic orthography. However, they are very irregular in occurrence.

3.4.3 Nouns of irregular gender Some nouns are treated irregularly, for a while masculine, for a while feminine (leaving out the uncertainty as to gender during the 19th and 20th A Grammar of Manx Place-names 13 centuries), their gender being determined by the treatment of their qualifying adjective. It may be that, at times, a given form represents an oblique case form surviving the loss of a head-word. So, for instance, Lhargee Vooar (ScG leargaidh mhòr) ‘big hillslope’ may represent a feminine simplex form with adjective, or a masculine genitive singular form with loss of head-word, e.g. (balley yn) lhargee vooar (with lhargagh) ((baile an) leargaidh mhóir) (leargach) ‘farm of / by the big hillslope’, or a dative form, viz. lhargee vooar (ScG leargaidh mhòir).

Lheeanney B / Vroghe Leane brough or Dirty Meadow 1711, Leaney Vroagh 1752 BA ‘dirty meadow’ (Ir. léana bhróghach) (PNIM II, 191). Lhergyvreck Largybrack 1693, Lhergy-Breck 1777 MI ‘speckled hillslope’ (ScG leargaidh bhreac). However, all entries from 1693 to 1777 treat Lhargee as masculine, thereafter feminine: Liargy Vrick 1803, Lhergyvreck 1957.

3.5 Nouns with adverbs of place and direction and with prepositional phrases

3.5.1 Nouns with adverbs of place and direction Most examples show the ‘non-motion’ as opposed to the ‘motion’ form of the adverb. The recorded examples include the following:

Bwoaill’ ny Garnane Heese / Heose Boal-ne-garnane heese / heiose 1837 GE ‘lower / upper Bwoaill’ ny Garnane (fold of the little cairns)’ (Ir. buaile na gcárnán shíos / shuas) (PNIM I, 261). Breck-woailley Heear Brack woaley Heare 1766 GE ‘west Breck-Woailley (speckled fold)’ (Ir. breac bhuaile thiar) (PNIM I, 201). Bwoailley ny Goar Heear Bwoaley-ne-Goare heare 1769 GE ‘west Bwoailley ny Goar (fold of the goats)’ (Ir. buaile nan gabhar thiar) (PNIM I, 233). Croit Veg Sthie / Mooie Croit veg sthie / mooie 1827 PA ‘inner / outer little croft / Croit Veg’ (ScG croit bheag a-staigh / a-muigh) (PNIM I, 88). Curragh Hiar Curragh Hiar 1783 PA ‘east curragh’ (Ir. currach thear) (PNIM I, 93). Magher y Vullee Sthie / Mooie Magher Eh vullie stie / mooie 1832 GE ‘inner / outer Magher y Vullee (the top field)’ (ScG machair a’ mhullaich a-staigh / a-muigh) (PNIM I, 270). Magheryn Shiar Maugheryn Shar 1759 PA ‘east fields’ (ScG machairean sear) (PNIM I, 142). 14 George Broderick Neigh Haule 1920 GE ‘the yonder flatt’, Mx yn aaie hoal (ScG an fhaiche thall ) (PNIM I, 201). Y Woaill’ Ard Hiar / Heear Wooil Ard har / heer 1824 GE ‘the east / west Bwoaill’ Ard (high fold)’ (Ir. a’ bhuaile ard thear / thiar) (PNIM I, 260).

3.5.2 Nouns (specifics) governed by a preposition Not many examples of this type are recorded, but they include the following:

Coan fo Hraie Coan-vo-raie 1828 MI ‘hollow under / by the shore’ (ScG cabhan fo thràigh) (PNIM II, 42), with lenition after (Mx) fo ‘under’ in this instance. Folieu ffolewe 1637–38 MA ‘below, at the foot of the mountain’ (ScG fo shliabh) (PNIM IV, 106). In later Manx, except in fossilised phrases, e.g. fo niaghtey (Ir. fo shneachta, ScG fo shneachda) ‘under snow’, the prep. fo does not lenite a following noun.

3.6 Sing. article with original masc. noun + adjective The few surviving examples contain the definite article in English, but which originally would likely have had the article in Manx:

Bayr Keyl, the The Baar Keall 1798 LE ‘the narrow road’, Mx bayr keyl (Ir. an bóthar caol ) (PNIM III, 307). Reconstructed this would be *yn bayr keyl. Cronk Mooar, the The Cronk Mooar 1821 PA ‘the big hill’ (ScG an cnoc mòr) (PNIM I, 87). Reconstructed this would be *yn cronk mooar.

3.7 Sing. article with original fem. noun + adjective Examples of the Manx article y (yn before a vowel or lenited f-) with feminine nouns are more prolific, since the nouns affected have either initial vowel or f- (which disappears under lenition), therefore taking yn:

Yn Eary Vooar Yn Nary vore 1709 GE ‘the big shieling’, Mx (ScG an àirigh mhòr) (PNIM I, 249). Yn Aaie Ghorrym Y Nhaai ghorrym 1738 BA ‘the blue flatt’ (ScG an fhaiche ghorm) (PNIM II, 188). Yn Aaie Vooar E Nay Voare 1795 JU ‘the big flatt’ (ScG an fhaiche mhòr) (PNIM II, 266). Yn Aaie Gheinnee Neigh Ghenny 1923 GE ‘the sandy flatt’ (ScG an fhaiche ghainmhich) (PNIM I, 201). With failure or absence of lenition in the adjective: A Grammar of Manx Place-names 15 Yn Eary Glass Nary Glass 1745 GE ‘the green shieling’ (ScG an àirigh + glas) (PNIM I, 246). Here absence of lenition (from // to /g/) took place early (see §3.2.3, above).

3.8 Article + adjective + noun A small number of this combination is attested. The Manx article survives better in instances of vocalic anlaut.

3.8.1 With Manx article: Nerlough Nardlougher 1634 MI ‘the high rushy place’, Mx yn ard leayghyr (ScG an àrd luachair) (PNIM II, 90). The earlier forms suggest Mx ard ‘high’, with the Manx article yn, as the first element. The second element may suggest ScG luachair ‘rushes’, the whole having initial stress, reduction of the second syllable, and (eventual) loss of the third.

3.8.2 With English definite article: Shennvalley, the the Shenvally 1704 GE ‘the old farm’, Mx shenn valley (ScG seann bhaile) (PNIM I, 290), Shenvala 1704 (with no def. art.). Reconstructed with def. art.: *yn shennvalley (ScG an seann bhaile).

3.9 Noun + original gen. sing. of noun

3.9.1 Original masc. noun with original gen. sing. of noun This type of name is common in Man, i.e. the specific is in the genitive case, whether masculine or feminine, and is formally attached to its head-noun functioning as an adjective.

Ballabrooie Ballabrooy 1712 PA ‘river-bank farm’ (ScG baile bruaich) (PNIM I, 179). Ballachrink BallaCroink 1704, Ballacrinck 1736 GE ‘hill farm’ (ScG baile cnuic) (PNIM I, 180) et passim. Ballaglonna Ballaglanna, Ballaglonna 1704 ML ‘glen farm’, Mx balley glionney (ScG baile gleanna) (PNIM VI, 40–41). Ballakerka Balla-Cerkey 1709, Ballakerka 1891 PA ‘rock farm’, Mx balley carrickey (Ir. baile cairrge) (PNIM I, 38–39), Ballacarghee 1662 LE (PNIM III, 289). Cass Strooan Cass strooan 1921 GE ‘stream foot’ (ScG cas sruthain) (PNIM I, 218), Casstruan (‘or Castrogan popularly’) 1898 RU (PNIM VI, 373–74). For intervocalic -g- see PNIM VI, 374. 16 George Broderick Close Shilly Close shillee 1992 GE ‘willow enclosure’ (Ir. clós seilich) (PNIM I, 296). Conrhennie (the) Conrenney 1759 LO ‘ferny hollow’, Mx coan rhennee (ScG cabhan rainich) (PNIM IV, 275). Cronk Aittin Knock aggin 1698 MI ‘the hill of gorse’ (ScG cnoc aitinn) (PNIM II, 62). Cronk Juckley Cronk Juckley 1810 MI ‘broom hill’, Mx cronk giucklee (giucklagh) (Ir. cnoc giolcaigh) (PNIM II, 63). Cronk Sharree Cronk Sharrey 1963 PA ‘foal hill’ (ScG cnoc searraich) (PNIM I, 89). Cronk Urley Cronk urley 1774 MI ‘the hill of advice, counsel’, Mx cronk coyrley (ScG cnoc comhairle) (PNIM II, 63–64, DMPN 95). Curragh Kella Corrough Keli 1704 GE ‘wood / grove curragh’ (ScG currach coille) (PNIM I, 242). Glen Maye glanmoy 1666 PA ‘open-plain glen’ (Ir. gleann maighe) (PNIM I, 108–09), OIr. mag nt., Ir. maige, dat. maig (GOI §338), with apocope of maighe in the Manx reflex. Kenaa Kenai 1699 GE ‘ford end’, Mx kione aah (ScG ceann àtha) (PNIM I, 257). Kionslieu Kan slew, Keoun-slew 1704 PA ‘mountain end’, Mx kionslieau (ScG ceann slèibhe) (PNIM I, 130). Mullagh Crink mullaugh Crink 1764 LE ‘hilltop’ (ScG mullach cnuic) (PNIM III, 317).

3.9.2 Original fem. noun with original gen. sing. of noun This is an extension of the foregoing, but with lenition in some cases.

Bwoaillee lieau Bolylew 1704 PA ‘mountain fold’ (ScG buaile shlèibhe) (PNIM I, 62–63). Bwoaillee Shoggyl Boaly Shoagell 1704 GE ‘rye fold’ (ScG buaile seagail ) (PNIM I, 216), Booill-shoggill 1787 BN (ScG buail’ seagail ) (PNIM V, 71). Chibber Lansh Chibber Lansh 1929 LE ‘the well of health’, Mx chibbyr hlaynt (Ir. tioba(i)r shláinte) (PNIM III, 320). Cooildarree Cool-darry 1835 MI ‘oaken nook’, Mx cooil darree (ScG cùil daraich) (PNIM II, 32, 58). Earylieau Earylheau 1956 GE ‘mountain shieling’ (ScG àirigh shlèibhe) (PNIM I, 249). Lheeannee Creggey Leinee Cregie 1733 JU ‘rock meadow’ (ScG lèana creige) (PNIM II, 252). A Grammar of Manx Place-names 17 Thalloo Chooilley Thalloo Chooilley 1829 BA ‘corner land’ (ScG talamh chùil) (PNIM II, 202).

3.9.3 With gen. sing. of noun marked only by position Faaie Clagh Fayee Clough 1741/42 GE ‘stone flatt’ (ScG faiche + clach) (PNIM I, 261), with absence of lenition and inflection, viz. chloie (ScG chloiche), in the qualifying noun; or this is a genitive plural (cf. §3.18).

3.10 Noun + genitival unit (noun + noun) This type of name is sparsely recorded in the Manx place-name corpus.

Strooan Bayrlieau Struan Bouilloue 1816 PA ‘mountain road stream’, Mx strooan bayr lieau (Ir. sruthan bóthair shléibhe) (PNIM I, 164). Here the constituent bayr lieau qualifies strooan. Curragh Ghaaue Hulby Curragh Ghew-hulby 1832 BA ‘the curragh / marshland of the smith of Sulby, the Sulby smith’s curragh’ (Ir. currach ghabha Shulbaigh) (PNIM II, 169). Here Manx Gaaue Hulby (with lenition in the genitive dependent place-name) is treated as a single unit, the first part of which is also subject to lenition in genitive relationship with preceding curragh. Croit (e) Ghaaue Hulby Croit Ghaue Hulby, Croit e Ghaue Hulby 1840 LE ‘the Sulby smith’s croft’ (Ir. croit (a’) ghabha Shulbaigh) (PNIM III, 356). A similar name and situation to the foregoing. The presence of the article in the second example could suggest that Ghaue Hulby is seen as a single unit. Or that the name has developed as follows: Croit e Ghaae  Croit e Ghaaue Hulby ‘the Sulby Croit e Ghaae’.

3.11 Noun + y / yn + original masc. gen. sing. noun

3.11.1 Gen. sing. of o-stems In o-stems the genitive singular masculine is formed by lenition of the initial consonant, where applicable, after the genitive of the masculine article followed by palatalisation of the final consonant.

Ballig Ballaluig 1704 MI (et passim) ‘the farm of / by / at the hollow’, Mx balley y luig (lag) (ScG baile an luig) (lag) (PNIM II, 44–45). Balynemade TR Balynemade 1515 MI ‘the farm of the (wild) dogs’, Mx baile ny moddey (ScG baile nam madadh) (PNIM II, 46). Claddagh yn Ollee Cloddaugh e Nolly 1728, Claddagh E Niolley 1767 PA 18 George Broderick ‘the claddagh (grass bank) of the cattle, the cattle claddagh’, Mx ollagh (ScG cladach an eallaich)(eallach) (PNIM I, 68). The entry for 1767 shows the n of the article palatalised by the initial syllable of eallach, as expected. Gob y Charry Gob-a-charry 1920 PA ‘the point of the foal (rock)’, Mx gob y charree (sharragh) (ScG gob an t-searraich) (searrach) (PNIM I, 116). Close y Toalt Close Toalt 1704 GE ‘the barn enclosure’, Mx close y toalt (soalt) (Ir. clós an t-sabhailt) (sabhal(t)) (PNIM I, 222), with excrescent -t. Cronk y Ching Cronk e King 1752, Cron[k] a Khing 1871 PA ‘the hill of the head / mound’, Mx cronk y ching / king (kione) (ScG cnoc a’ chinn) (ceann) (PNIM I, 90–91). The genitive in the 1752 example is marked by inflection but not by lenition, in the 1871 example by both – Kh- would repr. /x′/-). As already noted (§3.10), lenited /k/, /g/, /d/, giving /x/, //, // respectively (and their palatalised forms giving /x′/, /′/, /′/), began losing their spirantisation (and palatalisation) during the course of the late 18th and early 19th centuries. Cronk y Voddy Cronk y Voddy 1845 GE ‘the hill of the churl’, Mx cronk y voiddee (boiddagh) (ScG cnoc a’ bhodaich) (PNIM I, 238), or ‘hill of the (wild) dog’ (ScG cnoc a’ mhadaidh). Cooil y Churree Cool a Chuire 1641 JU ‘the marsh / curragh nook’, Mx curragh (ScG cùil a’ churraich) (PNIM II, 230). Cooyl y Chrink 1813 GE ‘behind the hill’, Mx cooyl y chrink (cronk) (ScG cùl a’ chnuic) (cnoc) (PNIM I, 203). Glen Fessan Glen Fessan 1956 MI ‘the glen of the parson, the parson’s glen’, Mx glion y phesson (< Eng. ‘person’), or MacPherson’s glen (ScG gleann ’ic a’ Phearsain) (PNIM II, 72). Glen Wyllin Glan vullen 1700, Glan willin 1704 MI ‘the mill glen’, Mx glion y wyllin (mwyllin) (ScG gleann a’ mhuilinn) (muileann) (PNIM II, 73). Magher y Vullee Magher Eh vullie 1832 GE ‘the top field’, Mxmullagh (ScG machair a’ mhullaich) (mullach) (PNIM I, 270). Magher y Vayr Magher-e-Vair 1760 GE ‘the field of / by the road, the road field’, Mxbayr (ScG machair a' bhàir (bhòthair)) (PNIM I, 202).

3.11.2 Of other stems Ballalheih Ballaley 1704 MI ‘the physician’s homestead / farm’, Mx balley y lhee (Ir. baile an leagha, nom. liaigh) (PNIM II, 31), Ballaleigh LE (PNIM III, 302) (i-stem). Gob yn Ushtey Gob-yn-ushtia 1920 PA ‘the point of the water(fall)’, Mx ushtey (ScG gob an uisge, nom. uisge) (PNIM I, 117) (io-stem). Gob y Glionna Gob-E-Glionney 1775 GE ‘the mouth of the glen’, Mx gob y ­g(h)lionney (glion) (ScG gob a’ ghleanna, nom. gleann) (PNIM I, 253), A Grammar of Manx Place-names 19 with original /′/ delenited and depalatalised to /g/ in the Manx form (s- / n-stem).

3.11.3 With complete loss of original /′/ (original //) in sandhi Thoar y (gh)lionnaThirlyonna 1927 PA ‘the dungfold / bleaching green of / by the glen’, Mx thoar y (gh)lionney (ScG todhar a’ (gh)leanna) (PNIM I, 39). The spelling -ly- in the 1927 example would indicate loss of the preceding palatalised gh-, viz. /′/.

3.11.4 With lenition / non-lenition of f- in a noun after the Manx article

3.11.4.1 With lenition Croe ny Ree Cro ny ree [krnri] 1990 MI ‘the pen of / by the heather’, Mx croe yn reoaie (ScG crò an fhraoich) (fraoch) (PNIM II, 21), with the feminine genitive singular of the article ny for yn to assist the rhythm of the phrase (see also §3.16).

3.11.4.2 With non-lenition Cronk y Freoaie Cronk y Frey 1819 PA ‘the hill of the heather, the heather hill’, Mx freoagh (ScG cnoc an fhraoich) (fraoch) (PNIM I, 87), with suppressed lenition of /f/ under which it would have disappeared.

3.11.4.3 Manx cl- for Gaelic tsl- In Irish and Scottish Gaelic original masculine nouns in initial sl-, sr- take initial t- (originally part of the masculine genitive singular of the article in Old Irish, viz. in(d), int (before sh) (GOI 294)) in the genitive. In Manx tsl- develops into cl-:

Giat y Clieau Giat y Clieau [gjatkl′u] HLSM II, 502 LE ‘the mountain gate’ (ScG geata an t-slèibhe) (PNIM III, 389, HLSM II, 502).

3.12 Noun + y / yn + original masc. gen. sing. + qualifying adjective Examples are not so great in number but some show the expected inflection in the adjective:

Cronk y Close Mooar Cronk-E-Close moar 1838 PA ‘the hill of the big enclosure, the Close Mooar hill’, Mx Cronk y C(h)lose Mooar (Conamara Ir. cnoc an c(h)lós mór) (PNIM I, 37). 20 George Broderick Croit y Voddee *Vrick Crot y Voddy Vrick 1760 AN ‘the croft of the speckled mastiff’ (ScGcroit a’ mhadaidh bhric) (PNIM III, 109), with inflection and lenition also in the qualifying adjective. This combination is rare, even in place-names. Eary y Phoill Vooar Eary-foal-woare 1920 PA ‘the shieling of the big pool’ (ScG àirigh a’ phuill mhòir) (PNIM I, 81). The genitive is marked by lenition after the article in both the noun and adjective and inflection in the noun. The Manx orthography here does not reveal any inflection in the adjective. Glion Cosshie Vooar Glen Cosshie Vore 1720–21 PA ‘the glen of the big rye- grass’ (ScG gleann a’ chuisich mhòir) (PNIM I, 107), with lenition and palatalisation in the noun and lenition (and probably palatalisation also) in the adjective, though this is not shown.

3.13 Noun + dependent noun + y / yn + original masc. gen. sing. Ballakilleyclieau Ballakilley 1704, Ballakilley Clieau 1759 MI ‘the mountain of Ballakilley (church farm)’, Mx balley killey yn Clieau (ScG baile cille an t-slèibhe) (sliabh) (PNIM II, 29–30).

3.14 Noun + ny + original fem. gen. sing. This type of name is well represented in Manx toponomy:

Baie ny hOoig Baie ny Ooig 1929 PA ‘the bay of the cave, the cave bay’ (ScG bàgh na h-ùige) (PNIM I, 23), with apocope of final / / in polysyllables. Ballaharra the ballaharrey 1686, Ballnaharra 1704 GE ‘the boundary farm’ (ScG baile na h- + OIr. airbe, Ir. airbhe) (PNIM I, 184). Ballakerka BallneKarkey 1704 PA ‘the rock farm’, Mx balley ny carrickey (Ir. baile na cairrge) (PNIM I, 38), Ballnacarga 1662 LE (PNIM III, 289). Bayr Gaish Bayr Gaish 1925 MI ‘the road of the causeways’, Mx bayr ny *gaish (Ir. bóthar na gceis) (PNIM II, 47). Bregnakell Bregnakell 1991 MI ‘the partly ploughed field of / by the wood’, Mx brackan na keyll(ey) (Ir. *breacan na coille) (PNIM II, 53), with apocope in the dependent genitive of the Mx reflex. Bulnahughtee Boalne hoaghty 1704 PA ‘the fold / field of / on the slope’, Mx bwoaill’ ny h-ughtee (ughtagh) (ScG buail’ na h-uchdaich) (PNIM I, 60–61). Close ny Haah Close Ne Hah 1780 GE ‘the enclosure of / by the ford’, Mx close ny h-aah (Ir. clós na h-áth) (áth) (PNIM I, 221). Cosh ny Lheeanney Coash na leannay 1712 PA ‘(the area at) the foot of the meadow’, Mx lheeanney (ScG cois na léana) (PNIM I, 76). Croit ny Greiney Crot Ne greaney 1817 PA ‘the sunny croft’, Mx grian (ScG A Grammar of Manx Place-names 21 croit na gréine) (grian) (PNIM I, 85–86). Croit ny Harrey Crot-ne harey 1784 PA ‘the croft of / by the boundary fence’, Mx croit ny h-*Arra (ScG croit na h- + OIr. airbe, Ir. airbhe) (PNIM I, 86). Drimnahairy Drim ne hary 1728 MI ‘the ridge of the shieling, the shieling ridge’, Mx dreeym ny h-eary (ScG druim na h-àirigh) (PNIM II, 66). Faaie ny Hawin Fien ny Howin 1970 PA ‘the river flatt, homefield’ (ScG faiche na h-abhainn) (PNIM I, 44). Gob ny Beinney Gob-ny-benya 1920 PA ‘the point of the mountain’, Mx bing (ScG gob na beinne) (beinn) (PNIM I, 115). Gob ny h-*Ink Gob ne ink 1845 GE ‘the point of the track / point of land etc.’ (cf. ScG gob na h-eanga) (PNIM I, 269), with apocope and devoicing of final -/g/ in the Mx reflex. Kione daa voaney Kione E daa Voaney 1793, Kion Davoaney 1795 MI ‘the end of two turbaries’, Mx moain (Ir. ceann dá mhóna) (móin) (PNIM II, 97). Magher ny Kirka Magher ny Kirka 1970 PA ‘the (Carrick) rock field’, Mx carrick (Ir. machair(e) na cairrge) (carraig) (PNIM I, 39). Strooan ny Kirkey Struanakirky 1704 GE ‘the hen stream’, Mx kiark (ScG sruthan na circe) (cearc) (PNIM I, 293).

3.14.1 The ‘alternative’ Manx gen. in -agh (ScG -ach) The following examples display what the late Robert L. Thomson called the ‘alternative’ Manx genitive, i.e. the genitive in -agh (ScG -ach), treated as though a lenited velar stem. This treatment is usually but not exclusively applied to nouns in -r, though mainly in but not exclusive to Phillips (c. 1610), e.g. Mx pooar ‘power’, gen. (Phillips) poyragh, cf. also the 18th-century Biblical magher ‘field’ machair( ), gen. magheragh, originally feminine, as in Scottish Gaelic.

Close ny Chibberagh Close ne chibbraugh 1704 BA ‘the enclosure of the well, the well enclosure’, Mx chibbyr (OIr. tipra; Ir. clós na *tiobarach / *tiobaireach) (tiobar) (PNIM II, 147), with syncope in the Manx example. Cronk ny Killagh Cronk ne Killagh 1861 GE ‘the hill of the chapel, the chapel hill’, Mx keeill (ScG cnoc na *cilleach) (PNIM I, 236). Strooan ny Poyllagh Streeuan ne pollaugh 1740 PA ‘the pool stream’, Mx poyll (ScG sruthan na *pollach) (poll ) (PNIM I, 164). Bayr ny Quaiylagh Bare ne Quailagh 1810 GE ‘the road of / to the court, the court road’, Mx quaiyl (Ir. bóthar na cómhdhálach) (PNIM I, 205). Cronk ny Bingagh Cronk ny bingagh 1760 GE ‘the hill of the jury, the jury hill’, Mx bing (ScG cnoc na *binneach) (PNIM I, 237). 22 George Broderick This ending can also apply to masculine nouns in -r – unless this is a case of a double suffix:

Croit yn Idderagh Crohtt Negeragh 1776 GE ‘the weaver’s croft’, Mx croit yn idderagh (fidder) (ScG croit an *fhigheadarach) (figheadair) (PNIM I, 193).

3.14.2 Absence of inflection in the fem. gen. noun Feminine nouns with absence of inflection in the genitive also occur, i.e. they are genitive by position, not by inflection. The form taken is that of the nominative with lenition after the article.

Creg yn Annag Creg yn nhannag 1812 GE ‘the rock of the crow, the crow’s rock’, Mx fannag (ScG creag + an fheannag) (PNIM I, 229). Thenh- implies palatalisation of n of the article, as one would expect, before a front vowel (here after loss of f- through lenition). Thie y Lhargee Veg thie e Lirgy-veg 1842 PA ‘the house of / by the Lhargee Veg (little hillslope)’ (ScG taigh + an leargaidh bheag) (PNIM I, 39), with the nom. / acc. y Lhargee Veg genitive by position only. The in the 1842 example probably represents the centralising of /a/ before /r/ (see §3.30).

3.15 Noun + ny + original fem. gen. sing. + qualifying adjective / noun In the following examples inflection but, as expected, no lenition in the dependent noun and its qualifying adjective are to be found:

Brackan Mooar ny Bwoaillee Mullee Breggan Moar ne Boillee-Mollee BA ‘the big breggan17 of / at the top fold’, Mx brackan mooar ny bwoaillee mullee (Ir. *breacan mór na buailidh mullaigh) (PNIM II, 166). Garey ny Hary Biggy garey ny hary Biggy 1746 LE ‘the enclosure of / by the little shieling’, Mx garey ny h-eary *biggey (ScG gàradh na h-àirigh bige) (beag) (PNIM III, 386). Killabreaga Killabrickey 1704 LE ‘(the farm of / by) the speckled church’, Mx (balley ny) killey *brickey (ScG (baile na) cille brice) (PNIM III, 414), cf. Ecclesbrek or Falkirk (ScG An Eaglais Bhreac). Knock / Cronk na h-Arrey Laa Knock-na-Heary-La 1784 PA ‘the hill of the day watch’ (Ir. cnoc na h-aire láe) (PNIM I, 132).

With absence of inflection in a polysyllabic qualifying adjective in the Manx reflex:

17 Cultivated strip of land in a field. A Grammar of Manx Place-names 23 Glion ny Moaney Mollagh Glen-ny-monee mollagh 1805 MI ‘the glen of the rough turbary’ (ScG gleann na mònadh molaich) (PNIM II, 74). See also Moaney Mollagh, §3.20.

3.16 Gen. fem. sing. of the article ny for gen. masc. sing. y / yn There is a tendency in Manx place-names to replaceyn with ny, as it was felt to ‘improve’ the flow of the phrasal name and therefore be more ‘authentic’. As a genitive feminine singular marker, ny occasions no lenition in the following noun, so that a form like Cronk ny Fasney (ScG *cnoc na fasgnaidh), below, would be felt to be in order. Levelling of the form of the article and of restoration of radical initials in noun + art. + noun structures is well-attested in anglicised Gaelic place-names in Ireland and Scotland.

Creg ny Vraane Creg ny vraane (sic) 1925 PA ‘the rock of the quern, the quern rock’, Mx creg yn vraain (ScG creag na brathan) (brà) (Ir. creag na brón) (brón) (PNIM I, 77). Croit Bal ny Trouan Croit-balny-trouan 1785 MI ‘the croft of the farm of / by the stream, Baltruan croft’, Mx croit balley’n trooan (strooan) (ScG croit baile an t-sruthain) (sruthan) (PNIM II, 61). Cronk ny Fasney Cronk ny Fasney 1882 GE ‘the hill of the winnowing’, Mx cronk yn asney (ScG cnoc an fhasgnaidh) (fasgnadh) (PNIM I, 237). Mx fasney, ScG fasgnadh ‘act of winnowing’ (originally a u-stem) reflects o-stem declension and, in the genitive singular after the article, undergoes lenition of the initial and palatalisation of the final consonant, viz. ScG an fhasgnaidh, Mx *yn asnee. On the other hand Cronk-fasnee (like Eng. Millinghill ), without article and with the specific as quasi-adjective, might have been admissible. However, Manx verbal nouns in -ey (ScG -adh) tend to use a genitive singular in -ee (ScG -(a)idh) in direct association with the head-noun they modify in nominal constructions, e.g. Mx dooinney-moyllee ‘best man’ (lit. ‘praising man’) (duine molaidh), otherwise no inflection takes place. Gob y Clieau Gob ny Clieu 1923 PA ‘the point of the mountain’ for Mx gob y clieau (slieau) (ScG gob an t-slèibhe) (sliabh) (PNIM I, 115).

3.17 Plural nouns This structure is rare in Manx place-names.

Chibberaghyn Chibberaghyn 1793 BN ‘wells’ (Ir. *tiobrachan) (PNIM V, 79). 24 George Broderick 3.17.1 Plur. noun with plur. adjective This structure is rare in Manx place-names.

Cronkyn Mynney Cronck an minney 1732 GE ‘little hills’ (ScG cnocan miona) (PNIM I, 226).

3.17.2 Plur. noun with qualifying noun in genitive This structure is not common in Manx place-names.

Bwoailtyn Killey Bwoilthyn Killey 1825 BR ‘church folds’ (ScG buailtean cille) (PNIM III, 213).

3.17.3 Ny + plur. noun This structure shows small representation in Manx place-names.

Ny Brooinyn Ne Brewnyn 1823 BA ‘the banks’, Mx ny brooinyn (Ir. *brua(ch)-anan) (PNIM II, 181). For the double plural here, see §3.17.5, below. Ny Crink ny Chruink 1738 LE ‘the hills’, Mx ny crink (ScG na cnuic) (PNIM III, 448).

3.17.4 Ny + plur. noun + qualifying adjective So far as is known, there are no extant examples of this feature with the plural form of the Manx article.

Brooinyn Glassey the Brownen Glassey 1735 BA ‘green banks’ (Ir. *brua(ch)- anan glasa) (PNIM II, 134); here with the Eng. plural definite article.

3.17.5 Double plural This structure is rare in Manx place-names.

Brooinyn Glassey Brownen Glassey 1728 BA ‘green banks’ (Ir. *brua(ch) anan glasa) (PNIM II, 134). Here Mx broogh ‘bank’, Ir. bruach, has attracted the plural suffix-yn twice, evidently as a form of reinforcement, viz. broogh + yn + yn, probably due to loss of medial /x/. Alternatively, this may contain a plural morpheme *-ynnyn, cf. Glionteeynyn, below. Glionteeynyn Mynney Gloan-tee-nyn-mynney 1829 BN ‘little glens’ (Ir. *gleanntaidh-annan miona) (cf. also PNIM V, 103). Here the older plural A Grammar of Manx Place-names 25 form Mx *gliontee (cf. OIr. glennta, ModIr. gleannta, ScG gleanntaidh) has attracted to it the plural ending *-ynnyn (cf. ScG -annan).

3.18 Noun + gen. plur. noun Some examples of this structure are noted, though they may represent forms containing the genitive plural of the Manx article, often lost in transmission, e.g. by speakers of English. They include:

Cooil Shellagh Quooleshellagh 1637–38 MI ‘willow nook’ (Ir. cúil ­s(h)‌eil- each / *s(h)ealach) (PNIM II, 59), with loss of lenition in the gen. pl. of the following noun; or the name may represent cúil nan seileach / *sealach. See also §3.23.1. Cronk Sharrey Cronk Sharrey 1963 PA ‘the hill of foals / inland boulders’, Mx cronk sharree (sharragh), i.e. Ir. cnoc + nom./acc. plur. of searraigh), but genitive by position; or sharree is gen. sing. and functions as an adjective qualifying cronk, viz. ‘foal / boulder hill’ (Ir. cnoc searraigh) (PNIM I, 89). Curragh Feeagh Curragh ffeaugh 1704 GE ‘the marshland / curragh of ravens, raven curragh’ (ScG currach fhitheach) (PNIM I, 241), with non-lenition of f- in the Mx genitive plural; or the name represents ScG currach nam fitheach. See also §3.23.1. Faaie Clagh Fayee Clough 1741/42 GE ‘the flatt of stones, sone flatt’ (ScG faiche + clach) (PNIM I, 261), with absence of lenition. See also §3.9.3. Glion Shellan Glion Shellan 1957 GE ‘the glen of (the) bees, (the) bees’ glen’ (ScG gleann sheillean) (PNIM I, 253), with non-lenition of sh- in the Mx genitive plural; or the name represents gleann nan seillean. See also §3.23.1.

3.19 Noun + ny(y) + gen. plur. noun In Manx the genitive plural of the article is ny + eclipsis (voicing of /p/, /t/, /k/, /f/ and nasalisation of /b/, /d/, /g/) of the initial consonant of the following noun, but nyn before vowels. Eclipsis and non-eclipsis can take place in the same phrasal name, though often (but not exclusively) in variants with nominative / accusative forms in the genitive plural. Genitive plurals of o-stems identical in form with the nominative singular, widespread in Irish and Scottish Gaelic, are only vestigial in Manx, almost entirely in place- names.

3.19.1 With eclipsis Akyr ny Moghtyn Acre ny moughtyn 1776 BA ‘the acre of the poor’ (Ir. acair 26 George Broderick na mbochtan) (PNIM II, 105). Here we have eclipsis in the gen. plur., as expected, but with the gen. plur. having the same form as the non./acc. plur., instead of the same form as the nom. sing., viz. (Mx) akyr ny moght, as in Close-ny-moght (below). See also Aker ne Boghtyn and Acre ne Boughtin, §3.19.2, below. Broogh nyn Ean Broogh-nyn-Ean 1776 GE ‘the river-bank of the birds’ (ScG bruach nan eun) (PNIM I, 189). Close Tunnag Close ne dunnag 1684 BA ‘the enclosure of the ducks, the duck enclosure’, Mx close ny dhunnag (thunnag) (Donegal Ir. clós na dtunnag) (PNIM II, 148–49). Close ny Moght Close ne Moght 1815 LE ‘the enclosure of the poor, the poor enclosure’, Mx close ny moght (Ir. clós na mbocht) (PNIM III, 334). Close ny Muck Close-ny-Muck 1785 LE ‘the enclosure of the pigs, the pigs’ enclosure’ (Ir. clós na muc) (PNIM III, 336). Ellan ny Braase Ellan-ny-brace 1920 PA ‘the island of the potatoes, the potato island’, Mx praase (PNIM I, 102). Ellan ny maghyl Ellan ny Maughol 1957 PA ‘the island of the crook-sticks, staffs, croziers’, Mx bachyl (Ir. oileán na mbachall ) (PNIM I, 102). Glion ny Mreck Glan ne mreck 1728 PA ‘the glen of the trout, the trout glen’, Mx breck (Ir. gleann na mbreac) (PNIM I, 113–14). Gob ny Greg Gob-ney-Greg 1817 GE ‘the point of the rocks’, Mx creg (Ir. gob na gcreag) (PNIM I, 214). Kerroo ny Glough Kerrow na Glaugh 1705 GE ‘the quarterland of the stones, stoney quarterland’, Mx kerroo ny glach (clagh) (Ir. ceathramh na gcloch) (PNIM I, 258). Knock ny Veoghaig Knock ne vachaage 1769 PA ‘the hill of the periwinkles, crusted shells’, Mx knock ny veoghaig (feoghaig) (Ir. cnoc na bhfaochóg) (PNIM I, 50).18 Slieau ny Garnane Slieau ny Garnane 1871 RU ‘the mountain of the little cairns’, Mx slieau ny garnane (carnane) (Ir. sliabh na gcarnán) (PNIM VI, 491). See also §3.19.2, below. Talloo ny Doltan Talloo ny doltan 1795 BA ‘the land of the ruined houses’, Mx thalloo ny dholtan (tholtan) (Ir. talamh na *dtolltan) (PNIM II, 202). Tollow ne Garnane Tollow ne Garnane BA ‘the land of the little cairns’, Mx thalloo ny garnane (Ir. talamh na gcarnán) (PNIM II, 203).

18 For the development of G. ó and á to [] in Manx, cf. Jackson 1955; Broderick 1986. If faochóg ‘periwinkle’ is the correct interpretation here, the development of G. /o/ to Mx [] (as expected) is shown here, in contradistinction to its retention in the name Traie Fogog [fgog] ‘the shore of periwinkles, periwinkle shore’ GE (PNIM I, 294). A Grammar of Manx Place-names 27 3.19.2 With absence of eclipsis Aker ne Boghtyn Aker-ne-Boghtyn 1773 BA ‘the acre of the poor’, Mx akyr ny boghtyn (ScG acair nam bochdan) (PNIM II, 106), acre ne boughtin 1737 JU (PNIM II, 206). See also §3.19.1, above. Chibber ny Fainnaghyn Chibber ny fainnaghyn 1882 GE ‘the well of the rings’ (ScG tiobar nam fàinneachan) (PNIM I, 218). Close Tunnag Close na Tunnagg 1704 BA ‘the enclosure of the ducks’ (ScG clós nan tunnag) (PNIM II, 148–49). See also §3.19.1, above. Kerroo ny Glough Kerrow ne Claugh 1735 GE ‘the stony quarter’, Mx kerroo ny glagh (clagh) (ScG ceathramh nan clach) (PNIM I, 258–59). Ooig ny gollan-geayee Ough-na-gallon-gaie 1920 PA ‘the swallows’ cave’ (ScG ùig nan gobhlan-gaoithe) (PNIM I, 149). Slieau ny Garnane Shleaw-ny-Carnayn 1762 RU ‘the mountain of the little cairns’ (ScG sliabh nan càrnan) (PNIM VI, 491). See also §3.19.1, above.

3.20 Survival of gen. or genitival phrase with ellipsis of noun of location This structure is not uncommon in Manx place-names.

Kella the Kealley 1676 LE ‘(the farm of / by the) wood’, Mx (bally ny) keylley (keyll ) (ScG baile na coille) (PNIM III, 410). Killabrega Killabrickey 1704 LE ‘(the farm of / by the) speckled church’, Mx (bally ny) killey *brickey (keeill vreck) (ScG baile na cille brice) (cill bhreac) (PNIM III, 414). Moaney Mollagh MI ‘(the glen) of the rough turbary’, Mx (glion ny) moaney mollagh (ScG gleann na mòna molaich) (PNIM II, 88). See also Glion ny Moaney Mollagh, §3.16. Nayre Yarg Nayre Yarg 1728 GE ‘(? the fold of the) red grass’, Mx (bwoailley) yn aiyr yiarg (Ir. (buaile) an fhéar dhearg) (PNIM I, 248), with historical lenition in the adjective. Vaaish Vooar Veaish Voar 1831 GE ‘(the farm) of / on the steep-sided hill’, Mx (balley) y vaaish vooar (Maase Mooar) (ScG (baile) a’ mhàis mhòir) (màs mòr), with lenition and inflection in the noun and lenition in the adjective; or Mx baaish ‘forehead, temples’, Ir. baitheas, gen. baithis ‘crown of head, pate’ (PNIM I, 295).

3.21 Noun + ny + gen. plur. noun + adjective When plural nouns in genitive by position have a qualifying adjective no eclipsis takes place at all, either in the noun or the qualifying adjective, i.e. the genitival relationship is syntactic, not inflectional. The structure is nom. / acc. 28 George Broderick form plus nom. / acc. form, the latter in genitival position following the genitive plural of the article ny(n). Note the plural form of the qualifying adjectives after their plural nouns; in this respect concord is maintained.

Creg ny Bollan Beggey Creg ny Bollan beggey 1923 PA ‘the rock of the little wrasse / rockfish’ (ScG creag nam ballan beaga) (PNIM I, 77). Gob ny Creggan Glassey Gob ny Creggan Glassey 1957 MI ‘the point of the grey rocks’, Mx gob ny creggyn glassey (ScG gob nan creagan glasa) (PNIM II, 75). Magher ny claghyn baney Magher ny claghyn baney 1923 PA ‘the field of the white stones’ (ScG machair nan clachan bàna) (PNIM I, 142).

Without the qualifying adjective we would expect something like *magher ny glagh, as in Kerroo ny Glough GE, or *gob ny greg, as in Gob ny Greg GE, §3.19.1 above; unless the examples are simply later forms containing erstwhile names, as in §3.23.

3.22 Noun + noun phrase in genitival position This structure is not uncommon in Manx place-names.

Chibbyr Slieau Maggle Chibber slew ne Magarell 1704 MI ‘the well of / at / on the mountain of the testicles,19 the well of Slieau Maggle, Slieau Maggle well’, Mx chibbyr slieau ny maggle (ScG tiobar sliabh nam magairle) (PNIM II, 55). Coan Slieau Meayl Coan Sliew-meil 1748 ON ‘the hollow of / by the barren mountain, Slieau Meayl hollow’ (ScG cabhan sliabh maol ) (PNIM IV, 385). Creg Cosh ny Lheeanney Creg coash na leannay 1712 PA ‘the rock of Cosh ny Lheeannee (the foot of the meadow)’ (ScG creag cois na lèana) (PNIM I, 76). See also §3.15. Garey Thie ny Puddase Garey tye-ny-Puddaise 1811 MI ‘the enclosure of Thie ny Puddase (the house of the potatoes)’ (Ir.gàradh taigh nan *púdás) (PNIM II, 102). Faih Yhean Vallieyre faih Yhean Vallieyre 1797 MI ‘the flatt of Jane of Balleira, Jane of Balleira’s flatt’, Mx faaie Yane Valleira (PNIM II, 43), with lenition in both dependent genitives, with Jane dependent on faaie and Balleira on Jane. Faaie Gob ny h-*Ink Faye Gob ne ink 1845 GE ‘the flat of / at Gob ny

19 According to oral tradition, sheep were formerly gathered here and castrated. A Grammar of Manx Place-names 29 h-*Ink (the point of the river-meadow)’ (ScG faiche gob na h-eanga) (PNIM I, 269). See also §3.14. Lheeanney Gob y Lionna Lheney Gob e Lhenney 1837 GE ‘the meadow by / at Gob y Lionna (the mouth of the glen)’, Mx lheeannee gob y (gh) lionna (ScG lèana gob a’ (gh)leanna) (PNIM I, 274). Lheeannee Vill Jane Leany Veljeayn 1787 MI ‘Bill Jane’s meadow’ (PNIM II, 43), with lenition in the first dependent genitive, as Bill Jane may be regarded as a single unit. Rheast Croit na Moaney Reaist Croit ne moany 1772 SA ‘the wasteland of Croit na Moaney (the turbary croft)’ (ScG riasg croit na mòna) (PNIM V, 277).

3.23 Reduction of gen. plur. ny / nyn to y / yn or complete loss of article In addition to the use of the Manx genitive plural of the article, viz. ny (nyn before vocalic anlaut) there also occurs a reduced form in y (yn before vowels).

Creg a Ruddyn Creg-a-ruddyn 1920 PA ‘the rock of the seal(s), the seal rock’, Mx creg y raun, creg ny raun (ScG creag an ròin, creag nan ròn), here with preocclusion in the Manx form, see §3.27, below (PNIM I, 76). Creg yn eean Creg yn Eayn 1929 PA ‘the rock of the birds, the bird rock’ for Mx creg nyn eean (ScG creag nan eun) (PNIM I, 79)

The two foregoing examples could conceivably be genitive singular, but this might suggest (a rock of) a particular seal or bird, rather than seals or birds in general, as would be implied by a genitive plural.

3.23.1 Loss of ny(n) or restoration of the radical in English environment Cooil Shellagh Quooleshellagh 1637–38 MI ‘the nook of the willows, the willow nook’ (ScG cùil nan seileach / *sealach) (PNIM II, 59), see also §3.18. Curragh Feeagh Curragh ffeaugh 1704 GE ‘the marshland / curragh of ravens, raven curragh’ (ScG currach fhitheach) (PNIM I, 241), with non- lenition of f- in the Manx genitive plural; or the name represents ScG currach nam fitheach, see also §3.18. Glion Shellan Glion Shellan 1957 GE ‘the glen of (the) bees, (the) bees’ glen’ (gleann sheillean) (PNIM I, 253), with non-lenition of sh- in the Mx genitive plural; or the name represents ScG gleann nan seillean, see also §3.18. 30 George Broderick

3.23.2 With complete loss of ny(n) in English environment Creg Voillan Creg Vollan 1923 PA ‘the rock of the seagulls, the seagull rock’, Mx creg ny voillan (Ir. creag na bhfaoileann) (PNIM I, 78), with eclipsis. Lhargey Gaish Larga-gash 1686 MI ‘the hillslope of / by the bog-roads’, Mx lhargey ny *gesh (*kesh) (Ir. learga na gceis) (PNIM II, 83), with eclipsis.

3.24 Noun + gen. of a personal name In Old Irish the genitive of a personal name was treated like any other indefinite dependent genitive, i.e. as an adjective in the same position, and so was lenited after a feminine noun in the nominative singular and after any singular noun in the dative (GOI §361). Lenition of genitive masculine personal names may originate by analogy with personal names following the genitive singular of mac ‘son’, viz. mhic, ’ic.

3.24.1 With original masc. noun Clyeen Clyfinn 1704, Clyeen 1810 MI ‘Finn’s rampart’ (Ir. claidhe Fhinn) (PNIM II, 57). Earlier entries (1704–1802) suggest no lenition of F-. Close Viggal Close Viggel 1799 GE ‘Michael’s enclosure’ (Ir. clós Mhíchíl ) (PNIM I, 222), with lenition of M-. Cronk Yhonnil Cronk yhonnil 1782 MI ‘Donald’s hill’, Mx Cronk Ghonnell (ScG cnoc Dhòmhnaill ) (PNIM II, 50), with lenition of D-. Initial yh- would represent the voiced velar fricative //.

3.24.2 With original fem. noun Aryhorkell Aryhorkell 1515 TR MI ‘Thorkell’s shieling’ (ScG àirigh Thorcaill ) (PNIM II, 18), with a loan from the ON personal name Þorkell, and with lenition of T. Croit yoan E Corris Croit yoan E Corris 1830 PA ‘John Corris’s croft’, Mx croit Yuan y Corris (’ic Mhuireasa) (PNIM I, 86), with lenition of /d′/. The Manx personal name (with a radical initial /d′/) is likely a development from Eng. John and so would not be all that old; unless it is Ewan, as in Ballakewn ML Mx balley y Kewn (ScG baile ’ic Eòghainn) (PNIM VI, 46–47). Eary Himmin Ary himmin 1704 GE ‘Shimmin’s / Simon’s shieling’ (ScG àirigh Shiomain) (PNIM I, 247), with lenition of S-. Keeill Pharlane Keeill Pharlane 1957 MI ‘(St) Bartholemew’s Church’ (ScG Cill Phàrlain) (cf. also PNIM II, 77), with lenition of P-. Keeill Pherick GE ‘(St) Patrick’s Church’ (ScG Cill Phàdraig) (PNIM I, 264), with lenition of P-. A Grammar of Manx Place-names 31 Keeill Vael20 Keeill Vael 1957 MI ‘(St) Michael’s Church’ (ScG Cill Mhìcheil ) (PNIM I, 77), with lenition of M-. Rengorrey rengorrey 1731 MI ‘(Mac)Gorree’s ridge’, Mx rheynn G(h)orree / y Gorree (rinn Ghoraidh / ’ic Goraidh)21 (cf. also PNIM II, 44); lenition of G- is not shown here, see §3.2.3, above.

3.24.3 With absence of lenition Lenition may be absent. (Although feminine names in literary sources in Manx are too rare for any definite conclusions to be drawn, they can remain unlenited in genitive position).22

Close Michael Close Michael 1797 GE ‘Michael’s enclosure’ (PNIM I, 220), with English personal name. Creg Tim Creg Tim 1920 PA ‘Tim’s rock’ (PNIM I, 78), with English personal name. Croit Sandy Croit Sandy 1824 GE ‘Sandy’s / Alexander’s croft’ (croit Alasdair) (PNIM I, 231), with Scottish personal name. Keeill Moirrey Keeill Moirrey 1957 GE ‘(St) Mary’s Church’, Mx Keeill Voirrey (Ir. Cill Mhuire) (PNIM II, 257). Skeeylley Vreeshey [skil vris′] (with lenition), also Skeeylley Breeshey [skil bris′] (without lenition) ‘Kirk Bride’ (ScG sgìre cille Bhrìghde ‘parish of the church of Bridget’) (HLSM II, 510).

Where it is felt that a personal name would in some way lose its ‘completeness’ under lenition, i.e. initial /s/-, /t/- > /h/-, /f/- > ø, then absence of lenition can take place. Foreign names are usually exempt from lenition, as in Welsh and Irish. But cf.

Close Hom Close hom 1687 LE ‘Tom’s enclosure’ (with lenition) (PNIM III, 330). Lenition may have taken place here, as Tom was felt not to be a ‘foreign’ name, whereas Tim may have been.

20 For a discussion on the name Mael, see PNIM II, 15. 21 According to Richard Cox (2007b, 69), ScG Goraidh < Gofraid (AU 942–1075) shows loss of f without compensatory lengthening, as in Cliosgro < ON Klifsgróf ‘[the] stream of the steep ascent’ (Cox 2002, s.v. Cliosgro). 22 In Scottish Gaelic (Carloway, Lewis) female personal names remained unlenited in genitive position, until very recently, when some sporadic lenition of feminine names crept in (cf. Cox 2002, 52). 32 George Broderick 3.25 Noun + gen. sing. of an original surname in ó Derivation: OIr. úa, óa, ó (earlier aue), io-stem, gen. haui, uí, masc. ‘grandson; a male descendant, a spiritual descendant’; fem. ‘granddaughter’ úa. In the formation of sept names the plural of úa is found followed by the genitive of the eponym, e.g. Uí Néill; seemingly belonging to a later fashion of nomenclature than collective names (DIL U, s.v. úa, óa, ó). Ó-surnames form a very small number of Manx Gaelic surnames and seemingly reflect the earliest strata of Goidelic settlement in Man (c. 500–900 AD).23

Arernan TR Aryeuzryn 1280, Arernan 1511 ML ‘Ó Uidhrin’s shieling’ (Ir. áirghe uí Uidhrin) (PNIM VI, 29–30). Arnycarnygan TR Arnycarnygan 1515 PA ‘O’Kennaghan’s / O Carnahan’s portion’, Mx ayrn y Charnaghan (Ir. earrann uí Chearnacháin) (PNIM I, 21). Arragon TR Are rogan 1507, QL Rogane 1748 SA ‘O’Rogan(e)’s shieling (Ir. àirghe uí Ruadhagáin) (PNIM V, 225). Aryrody TR Aryrody 1515 GE ‘O’Rody’s shieling’ (Ir. áirghe uí Rodaighe) (PNIM I, 176). Baldhoon QL Balladoyne 1654 LO ‘O’Doyne’s farm’ (Ir. baile uí Dhubháin) (PNIM IV, 221–22), with radical /d/- for lenited //. Ballafarrant QL Ballafarrant 1666 ML ‘Farrant’s farm’ (Ir. baile uí Fhearáin) (PNIM VI, 36–37), with loss of lenition of F- and with excrescent -t. Ballagarraghan TR Baly Dorghan 1524, QL Ball gorrahan 1704 GE ‘O’Dorgh-an’s farm’ (Ir. baile uí Dhorcháin) (PNIM I, 182, 204), with original dh- // falsely delenited to /g/ in the Manx form. Balladoyne TR Balydoyne 1524, QL Balladoyne 1704 GE ‘O’Doyne’s farm’, Mx O’Duinn (Ir. uí Dhuinn, uí Dhoinn), with radical /d/- for lenited //. Orthographic oy = [ui] in Early Northern English. Later, [ui] generally fell in with [ai], which is preserved in dialect, while [i] is normally restored from the spelling.24 Ballavarran QL Ballavarran 1704 JU ‘O’Barron’s farm’ (cf. Ir. baile uí Bhearáin) (PNIM II, 219). Ballavarvane AbQL Ballavarvane 1666 ML ‘Ó Borbhain’s farm’ (Ir. baile uí Bhorbháin) (PNIM VI, 58). Ballayolgane QL Joalgane 1703, Ballagolgane 1704, Ballayolgane 1825 LO

23 For Irish versions of Manx surnames, see MacLysaght 1985; for Scottish Gaelic versions, see Black 1946. 24 Robert L. Thomson, pers. comm. August 1993. A Grammar of Manx Place-names 33 ‘O’Golgan’s farm’ (Ir. baile uí Ghealgáin) (PNIM IV, 251). Ballahain QL Ballaeyne 1636, Ballahane 1643 RU ‘Ó Heidhin’s farm (Ir. baile uí h-Éidhin) (PNIM VI, 341).

3.26 Noun + gen. sing. of an original surname in mac Derivation: OIr. mac (macc), o-stem, gen. maicc, meicc, masc. ‘son; male descendant, boy, lad (from childhood to the age of bearing arms)’. After the introduction of surnames (10th–11th centuries) mac followed by gen. masc. personal names becomes a common form of surname and often changes to Mág (gen. Méig) in Ireland; the usage is also extended to foreign names (DIL M, s.v. mac (macc)). Mac-surnames form by far the majority of Manx Gaelic surnames. Manx reflexes showmhic + original lenited initial consonant of the surname, except in initial C or G due to homorganic inhibition.

Ballacain PA ‘Cain’s farm’ (Ir. baile ’ic Catháin) (PNIM I, 25). Ballacallin PA ‘Callin’s / MacAleyn’s farm’ (ScG baile ’ic Ailein) (PNIM I, 25). Ballacallow MI ‘Callow’s farm’ (ScG baile ’ic Amhlaigh)25 (PNIM II,19). Ballacarnane MI ‘Crinán’s farm’ balecrinane 1608 (Ir. baile ’ic Críonáin) (PNIM II, 20). Ballacojeen LO ‘MacFadyn’s farm’ (Ir. baile ’ic Pháidín) (PNIM IV, 230). Ballacorlett MI ‘Corlett’s farm’ (ScG baile ’ic *Thòirleoid ), with a loan from the ON personal name Þorljótr (PNIM II, 25). Ballacorteen MA ‘MacMartin’s farm’ (Ir. baile ’ic Mháirtín) (PNIM IV, 33). Ballacrye BA Crye’s farm’ (Ir. baile ’ic Craith) (PNIM II, 112). Ballafajeen MI ‘Páidín’s farm’ (Ir. baile Pháidín) (PNIM II, 26). Ballagawne MI ‘Gawne’s / the smith’s farm’ (ScG baile mhic a’ Ghobhainn) (PNIM II, 27). Ballahimmin GE ‘Shimmin’s / Simon’s farm’ (ScG baile Shiomain) (PNIM I, 185). Ballakaighin GE ‘Kaighin’s farm’ (ScG baile ’ic Eachainn) (PNIM I, 187). Ballakelly AN ‘Kelly’ farm’ (Ir. baile ’ic Ceallaigh) (PNIM III, 43). Ballaleece GE ‘Leece’s farm’ (ScG baile ’ic Gille Ìosa) (PNIM I, 191). Ballaquayle ON ‘Quayle’s farm’ (ScG baile ’ic Phàil ) (PNIM IV, 371). Ballaqueeney RU ‘MacSweeney’s farm’ (ScG baile ’ic Shuibhne) (PNIM VI, 349).

25 For a discussion of the ScG personal name Amhlaigh from a possible ON Ánlaif, see Cox 2007a, 141). 34 George Broderick Ballaquiggin BN ‘Quiggin’s farm’ (Ir. baile ’ic Uiginn) (PNIM V, 59). Bwoaillee Carney PA ‘Carney’s fold’ (Ir. buailidh ’ic Cearnaigh) (PNIM I, 27). Coan y Vurraghey Coan vourghey 1731 MI ‘Curphey’s hollow’, Mx coan y Vurraghee (ScG cabhan ’ic Mhurchaidh) (PNIM II, 42). This form gives Mx y Curghey, later anglicised to Curphey [kø:fi].

3.27 Preocclusion The articulation of Manx final /l/, /m/, /n/, // in original long stressed syllables (in mono- and disyllablic words) can be accompanied by their corresponding preocclusive, viz. [dl], [bm], [dn], [g]. Preocclusion can lead to the shortening of original long syllables (cf. HLSM III, 28–34):

shooyll ‘walk’ [s′ul], [s′udl] ScG siubhail. trome ‘heavy’ [trom], [tro()bm] ScG trom. bane ‘white, fair’ [bn], [b()dn] ScG bàn. lhong ‘ship’ [l], [lg] ScG long.

3.28 Final -/g′/, -/′g′/ in Old Irish become devoiced to -/k/, -/k/, -/k′/ in Manx ScG Pàdraig ‘Patrick’ : Mx Perick /prik′/ (HLSM II, 491). ScG easbaig ‘bishop’ : Mx aspick /aspk′/ (HLSM II, 13). ScG fang ‘sheep-pen’ : Mx *fank, as in Ballanank /balnak/ (baile an fhaing < Scand. fang ‘catch’) (PNIM VI, 51), here with absence of palatalisation in the Manx reflex. ScG eang ‘track, land’ : Mx *ink, as in Gob ny h-*Ink Gob ne ink 1845 GE ‘point of the track / point of land etc.’ (ScG gob na h-eanga) (PNIM I, 269), with apocope and devoicing of final -/g/ in the Manx reflex.

3.29 Medial and final Ir. and ScG /sk/, /s′k′/ > /st/, /s′t′/ in Manx

3.29.1 Palatalised variants Ir. uisce, SCG uisge ‘water’  Mx ushtey /us′t′/ (HLSM II, 470). ScG rèisg gen. ‘moor, marsh’ (nom. / acc. riasg)  Mx reisht */res′t′/ (PNIM VI, 30).

3.29.2 Non-palatalised variants Ir. iasc, ScG iasg ‘fish’  Mx eeast /jist/ (HLSM II, 142). ScG iasgadh ‘act of fishing’  Mx eeastagh /jistax/ (HLSM II, 142–43). ScG riasg ‘moor, marsh’  Mx reeast /rist/ (PNIM III, 457). A Grammar of Manx Place-names 35 3.30 Centralising stressed vowels in the environment of /r/ or r-clusters In Late Manx stressed long or short vowels, or stressed short vowels secondarily lengthened, can be realised as [] or [] in the environment of /r/ or r-clusters (but also to a lesser extent in the environment of /t/, /b/, /d/, /g/, /m/, /n/, /l/, /s/, /s′/, /x/; cf. HLSM III, 44–48), with or without loss of /r/.

Ballargey PA [blg] (Ir. baile leirge) (PNIM I, 45). Doarlish Ard PA [d()lis′d] (cf. ScG dòirling + àrd ) (PNIM I, 96–97). Faaie oarn MR [fin] (sic) (ScG faiche eòrna) (PNIM V, 197).

3.31 Back formations from Old Norse originals When Old Norse loan-words or loan-names have radical initials which correspond to lenited consonants in Manx, such consonants may be ‘restored’ to their appropriate radical form in Manx. The samples are scarce:

ON [v] (or earlier [w]) vágr, dat. vági  Mx [b] baie ‘bay’.26

3.32 Incident names A small number of places are named from incidents of one sort or another, e.g. murder. One or two names are recorded both from oral tradition and documentary sources that reflect an incident – usually a murder – that took place at a particular spot:

Clagh y Dooinney Marroo Clagh y dooinney marroo 1925 LE ‘the stone of the dead man’ (ScG clach + an duine marbh) (PNIM III, 323 for details); late. Cley varr E-Dooiney Cley varr E-Dooiney 1791 PA ‘(the) hedge which killed the man’, Mx cleigh varr y dooinney (Ir. *claidhe a mharaigh an duine); no further details are known (PNIM I, 70).

3.33 Names reflecting older Manx traditions One or two names represent older Manx traditions long since disappeared. This feature is rare in Manx place-names.

Cleigh Feeiney Cliy [F ]ennia 1818, Cleih Fennah 1930 PA ‘the fence of the Fianna’, Mx cleigh ny *Feeiney (Ir. claidhe na Féinne) (PNIM I, 70).

26 See also Cox 2002, 52 for examples from Lewis. 36 George Broderick Clyeen Clyfinn 1704, Clyeen 1810 MI ‘Finn’s rampart’ (Ir. claidhe Fhinn), with non-lenition in the name Finn in Manx, as suggested here by the earlier entry (PNIM II, 57).

The foregoing would refer to the tales surrounding Fionn MacCumhail, the central hero in the Fenian Cycle of Gaelic folk tradition (cf. DCM 230– 233). In Manx tradition a number of Fenian stories were written down by Ned Beg Hom Ruy (Edward Faragher of Cregneash, 1831–1908) (cf. Broderick 1974–76; 1982; 1983). A version of the Manx traditional song Fin as Oshin (‘Fin and Ossian’) was collected in Man c. 1761 / 62 or c. 1769 (cf. Broderick 1990).

4 Conclusion Traditional Goidelic, or Gaelic, speech arrived in Man c. 500 AD and died with the last reputed native Manx Gaelic speaker, Ned Maddrell, on the 27th of December 1974, extending over a period of some 1500 years. Part of the cultural baggage of Goidelic in Man are the place-names which were bestowed over a similar period, some earlier than others, depending on structure. The development and range of place-name structures in Man can be briefly sketched as follows:

(i) Simplex nouns and close-compounds (noun + noun; adjective + noun) The earliest type consists of simplex names containing a noun only, e.g.Appin, Rushen (§2.1), followed by the close-compound type made up of noun + noun (Mooragh) (§2.2.1) and adjective + noun (Doolough, Douglas) (§2.2.2), all bearing initial stress. This structure was already becoming obsolescent c. 400–500 AD, but nevertheless managed to set foot in Man, albeit on a modest scale, before its apparent demise around 700 AD.

(2) Nouns with the sing. article Developing on from the foregoing we have single nouns but accompanied by the article, either masculine (Yn Ellan) (§3.1) or feminine (Neary) (§3.2). Then come the structures noun + adjective, again either masculine (bayr dowin) (§3.3) or feminine (awin veg) (§3.4), or noun + adverb of place (naaie hoal ) or preposition (coan fo hraie) (§3.5); noun with article and adjective (*yn cronk mooar), (yn eary vooar) (§3.6–7); article + adjective + noun (Nerlough) §3.8). A Grammar of Manx Place-names 37 (3) Nouns with a dependent gen. sing. noun Head-nouns are either masculine (ballabrooie) or feminine (cooil darree) (§3.9); they may form part of a genitival unit (strooan bayrlieau) (§3.10).

(4) Nouns with the article and gen. sing. noun with or without qualifying adjective Without qualifying adjective, masculine (claddagh yn ollee) (§3.11) or feminine (baie ny h-ooig) (§3.14); with inflected qualifying adjective, masculine (croit y voddee vrick) (§3.12), feminine (garey ny h-eary biggey) (§3.15).

(5) Plur. nouns alone, with plur. adjective or dependent gen. noun Plurals consisting of a noun only (chibberaghyn) (§3.17); with plur. adjective (cronkyn mynney) (§3.17.1); with dependent gen. noun (bwoailtyn killey) (§3.17.2).

(6) Plur. article with plur. noun alone, or with adjective Alone (ny brooinyny) (§3.17.3); with adjective (the brooinyn glassey), with Eng. def. art. (§3.17.4)

(7) Double plurals Glionteeynyn Mynney (§3.3.17.5).

(8) Noun with gen. plur. noun, with or without plur. article With gen. plur. noun (Cooil Shellagh) (§3.18), with plur. article and gen. plur. noun (close ny moght) (§3.19).

(9) Survival of gen. or gen. phrase with loss of headword With genitive (kella); with genitival phrase (killey brickey) (§3.20).

(10) Noun with plur. article + gen. plur. noun + plur. adjective, or with gen. noun phrase With plur. article + gen. plur. noun and plur. adjective, Gob ny Creggyn Glassey (§3.21); with a gen. noun phrase (Creg Cosh ny Lheeanney) (§3.22).

(11) Noun with gen. of a personal name With masculine headword (Close Yhonnil ) (§3.24.1); with feminine headword (eary Himmin) (§3.24.2). 38 George Broderick (12) Noun with surname in ó and mac In ó (Aragon – Ir. áirighe uí Ruadhagáin) (§3.25); in mac (Ballacain – Ir. baile ’ic Catháin) (§3.26).

(13) Preocclusion With [dl] shooyl [s′udl], with [bm] trome [trobm], with [dn] bane [be()dn], with [g] lhong [lg] (§3.28). There then follow a small number of minor phonetic developments.

(14) Incident names and names reflecting older Manx traditions Incident name (clagh y dooinney marroo) (§3.31); name reflecting older Manx traditions (cleigh feeiney) (§3.32).

During the period when Manx was the everyday language of the ordinary people of Man, from the 14th to the 19th centuries, we see, for instance, that concord was maintained between article and noun (§3.2.3, §3.7), and noun and qualifying adjective(s), whether as generic or specific (§§3.3–4, §§3.6– 7); lenition (§3.4.1, §3.7) and eclipsis (§3.19.1) were applied whenever expected; inflection where appropriate (§3.9–14) etc.; thus indicating that Manx Gaelic functioned in vibrant fashion like any other living dialect of Goidelic. Towards the end we begin to see simplification, then reduction, taking place, e.g. lack of concord between noun and adjective (§3.3, §3.4.2, §3.7, §3.9.3, §3.11.4.2), absence of inflection (§3.14.2), non-application of lenition (§3.4.2, §3.7, §3.24.3) and / or eclipsis (§3.19.2) etc.,27 thus mirroring the demise and loss of vitality of Manx as a living language. In short, the Manx place-name evidence spans the period of Goidelic presence in Man, from its initial presence c. 500 AD down through the 1500 years of its life there until its replacement by English during the 19th and 20th centuries.

Phonetic note A superscript tick indicates palatalisation, e.g. [k′, l′, s′ t′].

Abbreviations Parish: AN – Andreas, AR – Arbory, BA – Ballaugh, BN – Braddan, BR – Bride, GE – German, JU – Jurby, LE – Lezayre, LO – Lonan, MA – Maughold, MI

27 For details of these and other features of language obsolescence in Man, see Broderick 1999. A Grammar of Manx Place-names 39 – Michael, ML – Malew, MR – Marown, ON – Onchan (Conchan), PA – Patrick, RU – Rushen, SA – Santan.

Other: AbQL – abbeyland quarterland (land division) AU – Annals of Ulster Eng. – English G. – Gaelic IE – Indo-European Ir. – Irish MidIr. – Middle Irish ModIr. – Modern Irish OE – Old English OIr. – Old Irish ON – Old Norse PBrit. – Proto-British PCelt. Proto-Celtic Phillips – (Early Manx; see Moore and Rhŷs 1893–94) QL – quarterland (land division) Scand. – Scandinavian ScG – Scottish Gaelic TR – treen (land division) W. – Welsh Skt – Sanskrit

References Ahlqvist, Anders, 1985, ‘The ordering of nominal compounds in Irish’, Historical semantics. Historical Word-Formation, Trends in Linguistics: Studies and Monographs 29 (Berlin: Mouton), 1–9. Black, George, 1946, The Surnames of Scotland: their origin, meaning and history (New York: Birlinn; repr. 1999, 2004). Broderick, George, 1974–76, ‘Four Manx Folktales’, Béaloideas 42–44, pp. 41–61. Broderick, George, 1982, ‘Manx Stories and Reminiscences of Ned Beg Hom Ruy’, Zeitschrift für celtische Philologie 38, 113–78. Foreword and Manx Gaelic texts. Broderick, George, 1983, ‘Manx Stories and Reminiscences of Ned Beg Hom Ruy’, Zeitschrift für celtische Philologie 39, 117–84. Translation and linguistic notes. 40 George Broderick Broderick, George, 1984–86, A Handbook of Late Spoken Manx, Vol. I Grammar and Texts, Vol. II Dictionary, Vol. III Phonology (Tübingen: Niemeyer). Broderick, George, 1990, ‘Fin as Oshin’, Celtica 21, 51–60. Manx-Gaelic folksong with linguistic notes. Broderick, George, 1994–2005, Place-Names of the Isle of Man, 7 vols (Tübingen: Niemeyer). Broderick, George, 1995, Cronica Regum Mannie & Insularum. Chronicles of the Kings of Man and the Isles (Douglas: Manx National Heritage; reprinted 2004; 1st edn 1979). Broderick, George, 1999, Language death in the Isle of Man. An investigation into the decline and extinction of Manx Gaelic as a community language in the Isle of Man, in Linguistische Arbeiten 395 (Tübingen: Niemeyer). Broderick, George, 2003, ‘Tynwald: a Manx cult-site and institution of pre- Scandinavian origin?’, Cambrian Medieval Celtic Studies 46, Winter 2003, 55–94. Broderick, George, 2006, A Dictionary of Manx Place-Names (Nottingham: English Place-Name Society). Broderick, George, 2008, ‘Pre-Scandinavian place-names in the Isle of Man’, in Padel and Parsons 2008, 165–84. C. – A Dictionary of the Manks Language (Cregeen 1835). Cavill, Paul and George Broderick, eds, 2007, Language contact in the place- names of Britain and Ireland (Nottingham: English Place-Name Society). Cowan, Edward J., and R. Andre, eds., 2000, Alba: Celtic Scotland in the medieval era (East Linton: Tuckwell Press). Cox, Richard A. V., 2002, The Gaelic Place-Names of Carloway, Isle of Lewis. Their structure and significance (Dublin: Dublin Institute for Advanced Studies). Cox, Richard A. V., 2007a, ‘Notes on the Norse impact on Hebridean place- names’, Journal of Scottish Name Studies I, 139–44. Cox, Richard A. V., 2007b, ‘The development of Old Norse -rð- in (Scottish) Gaelic’, in Cavill and Broderick 2007, 57–96. Cregeen, Archibald, 1835, A Dictionary of the Manks Language (Douglas: Quiggin). Davey, Peter J., ed., 1978, Man and Environment in the Isle of Man, BAR British Series (Oxford: University Press). Davey, Peter J., 2005, ‘Placenames and the physical and human geography of the Isle of Man: an overview’, in PNIM VII, 325–36. DCM – A Dictionary of Celtic Mythology (MacKillop 1998). Di. – Foclóir Gaedhilge-Béarla – Irish-English Dictionary (Dinneen 1927). A Grammar of Manx Place-names 41 DIL – Contributions to a Dictionary of the Irish Language (Royal Irish Academy 1912–76). Dinneen, Patrick S., 1927, Foclóir Gaedhilge agus Béarla. An Irish-English Dictionary (Dublin: The Irish Texts Society). DMPN – A Dictionary of Manx Place-Names (Broderick 2006). Drummond, Peter, 2009, ‘Close compound mountain toponyms in Islay and Jura’, in McClure, Kirk and Storrie 2009, 50–61. Fell, Christine, et al., eds., 1983, The Viking Age in the Isle of Man (London: The Viking Society for Northern Research, University of London). Fellows-Jensen, Gillian, 2005, ‘The Scandinavian element in the place-names of the Isle of Man’, in PNIM VII, 357–370. GEM – Glossary of Early Manx (Thomson 1954-57). GOI – A Grammar of Old Irish (Thurneysen 1946). HLSM – A Handbook of Late Spoken Manx (Broderick 1984–86). Jackson, Kenneth H., 1953, Language and History in Early Britain (Edinburgh: University Press). Kermode, P. M.C. (1909–18), The Manx Archaeological Survey ... Keeills and Burial Grounds in the Sheadings of Glenfaba, Michael, Ayre, Garff and Middle (Douglas: Johnston). LEIA – Lexique étymologique de l’irlandais ancien (Vendryes 1959-). Mac Giolla Easpaig, Dónall, 1981, ‘Noun + noun compounds in Irish place- names’, Études celtiques 18, 151–63. MacKillop, James, 1998, A Dictionary of Celtic Mythology (Oxford: University Press; paperback edn 2000). MacLysaght, Edward, 1985, The Surnames of Ireland (Dublin: Irish Academic Press; repr. 1991–2001). McClure, J. Derrick, John M. Kirk and Margaret Storrie, eds, 2009, A Land that lies Westward. Language and culture in Islay and Argyll (Edinburgh: John Donald). McDonald, R. Andrew, 1997, The Kingdom of the Isles: Scotland’s western seaboard c. 1100–c. 1336 (East Linton: Tuckwell Press). McManus, Damian, 1997, A Guide to Ogam (Maynooth: An Sagart). Megaw, Basil, 1976, ‘Norseman and native in the Kingdom of the Isles: a reassessment of the Manx evidence’, Scottish Studies 20, 1–44. Reprinted with minor revision in Davey 1978, 265–314. Moore, A.W., with John Rhŷs, 1893–94, The Book of Common Prayer in Manx Gaelic. Being translations made by Bishop [John] Phillips in 1610, and by the Manx clergy in 1765, Manx Society Vols. XXXII–III (Oxford: Oxford University Press). 42 George Broderick Padel, O. J., and David N. Parsons, eds, 2008, A Commodity of Good Names. Essays in Honour of Margaret Gelling (Donington: Shaun Tyas). Page, R. I., 1983, ‘The Manx rune-stones’, in Fell 1983, 133–46. PNIM – Place-Names of the Isle of Man (Broderick 1994–2005). PRNB – Place-Names of Roman Britain (Rivet and Smith 1979). Rivet, A. L. F., and Colin Smith, 1979, The Place-Names of Roman Britain (London: Batsford). Schrijver, Peter, 1995, Studies in British Celtic Historical Phonology (Amsterdam: Rodopi). Sellar, W. D. H., 2000, ‘Hebridean sea kings: the successors of Somerled 1164–1316’, in Cowan and McDonald 2000, 187–218. Sims-Williams, Patrick, 2000, ‘Degrees of Celticity in Ptolemy’s names: examples from Wales’, in Sims-Williams and Parsons 2000, 1–15. Sims-Williams, Patrick, and David N. Parsons, eds, 2000, Ptolemy: towards a linguistic atlas of the earliest Celtic place-names of Europe (Aberystwyth: Cambrian Medieval Celtic Studies). Tempan, Paul, 2009, ‘Close Compounds in Irish place-names’, in McClure, Kirk and Storrie 2009, 62–78. Thomson, Robert L., 1954–57, Glossary of Early Manx, Sonderabdruck aus Zeitschrift für celtische Philologie 24/4, 272–307; 25/1–4, 100–40; 27/1–2, 264–308, 79–160. Thomson, Robert L., 1983, ‘The continuity of Manx’, in Fell 1983, 169–74. Thurneysen, Rudolf, 1946, A Grammar of Old Irish, revised and enlarged edition, translated from the German by D. A. Binchy and Osborn Bergin (Dublin: Dublin Institute for Advanced Studies; repr. 1970). Vendryes, J., et al., 1959–1978, Lexique étymologique de l’irlandais ancien (Paris: Centre national de la Recherche scientifique ; repr. 1981–1996 by the Dublin Institute for Advanced Studies, Dublin). Wilson, David M., 2008, The Vikings in the Isle of Man (Aarhus: Aarhus University Press). Zimmer, Stefan, 2004, Die Kelten – Mythos und Wirklichkeit (Stuttgart:Theiss). Memories, Meids and Maps: the Shetland Place Names Project1

Eileen Brooke-Freeman Shetland Amenity Trust

Shetland Amenity Trust is a charitable organisation dedicated to preserving, enhancing and promoting Shetland’s cultural and natural heritage, with responsibility for Shetland’s archaeology, biological records, architectural and environmental improvement, woodlands, Shetland Museum and Archives and the tourism organisation, Promote Shetland. In 1998 it was recognised that place-names are an important component of our cultural heritage, fitting logically within the trust’s remit.2 A project plan was developed and funding secured3 to set up a pilot project to record all available information on Shetland’s place-names and establish a comprehensive database linked to digital maps in order to enable users to relate the names precisely to their locations. I took up the post of Shetland Place Names Officer in August 2001. The project planned to build on the achievements of place-name scholars Jakob Jakobsen and John Stewart who dedicated years to recording and interpreting Shetland place-names. Jakobsen in the 1890s and Stewart in the 1950s amassed vast collections of Shetland place-names which they scrutinised and interpreted in detail, culminating in Jakobsen’s The Place- Names of Shetland and Stewart’s Shetland Place-names. Their approach was to arrange the names by linguistic element rather than geographically, resulting in some difficulties when relating the names precisely to their locations. Subsequently one of our key aims was to locate all recorded names on maps, augmenting Jakobsen’s and Stewart’s work, not only by adding more names, but by putting the names they collected and explored linguistically into their geographical context.4 Great public interest and support has been stimulated throughout Shetland and beyond, resulting in considerable amounts of previously unrecorded information being brought to our attention. From a three-year Heritage 1 This paper is based on a presentation given to the SPNS Conference in Glasgow in November 2009. It builds on a presentation of the first two years of the project, published in Gammeltoft, Hough and Waugh 2005. 2 Dr Doreen Waugh was pivotal in the establishment of the project, having originally brought forward the idea. 3 Funders of the three-year pilot were Heritage Lottery Fund, Shetland Islands Council, Shetland Enterprise Company and Shetland Amenity Trust. 4 Methods explored by Jakobsen, Stewart and the Shetland Place Names Project are explored in more detail in Sigurðardóttir and Smith 2010.

The Journal of Scottish Name Studies 4, 2010, 43–60 44 Eileen Brooke-Freeman Lottery Fund pilot, the Shetland Place Names Project has now become one of Shetland Amenity Trust’s core activities.

Sources An early task involved conducting an audit of all available information. This has proved to be an ongoing process because, in addition to collections deposited in the more obvious locations like the Shetland Archives, the National Archives of Scotland, the School of Scottish Studies and local museums,5 new lists and documents in private hands have been unearthed and continue to come to light. Individuals have compiled lists or created maps of local place-names, many of which have never been incorporated in official maps. We are trying to collate all these resources to make the information more widely accessible. Oral records are particularly important. Old place-names are a particularly fragile resource in the current era of digital mapping and satellite navigation. Our modern lifestyle means that increasingly fewer people use coastal landmarks or waypoints through the hills. A vast number of names relate to fishing and crofting, with individual rigs, dykes, geos and rocks all carrying names passed down between generations of people living in the same area. Additionally, names have frequently been written down but not recorded on maps, again highlighting an urgency to talk to older residents who may remember them and therefore help locate them before the information vanishes forever. The project focuses on recording these oral names first, prior to checking documentary sources and undertaking linguistic analysis. The urgent need to preserve the oral record – recognised by both Jakobsen and Stewart – is still very apparent but, sadly, some of our key informants have now died, while other, potential informants have either died or become incapacitated before being interviewed.

Recording An excellent starting point has been working with the network of local history groups. There are currently 20 groups who meet mainly through the winter months, although some continue their activities throughout the year. They have been instrumental in identifying potential informants for each local community. To date we have worked directly with 14 history groups, staging recording sessions within their communities. Other key groups

5 One such example is a map and list of almost 700 Fetlar coastal names compiled by Roger Smith and the late Charlie Thomason of Fetlar, which is deposited in the Fetlar Interpretive Centre. Memories, Meids and Maps: the Shetland Place Names Project 45 include the Women’s Royal Voluntary Service and Care Centres, where staff have acted as recorders, helping to note names on maps and recording sheets from day-care residents. Work with primary schools has included supplying local information, maps and documents for local studies and Viking topics in order to supplement national curriculum resources.6 Recording place- names has also formed part of EcoSchools and other school assignments. A range of recording techniques have been employed, including using copy maps and recording sheets, taping and digitally recording conversations, studying old maps and aerial photographs, walking the ground and photographing sites and extracting information from documentary sources. Fig. 1 shows names from Hamnavoe, Eshaness, annotated on an old photograph.7 Each place-name is recorded on a copy map or aerial photograph with a running number. Details about the feature are added to the recording sheet, including an English ‘phonetic’ spelling, along with the precise geographic location and a record of any alternative names. A description of the type of feature and information about origins of the name or suggested interpretations are also noted. All known names are recorded, including those for features such as rigs, tracks, outbuildings, wells, mills, noosts (places where boats are drawn up) and rocks, whether named on the map or not. Names that are mis-named or mis-positioned on the maps are also corrected. Recording conversations helps determine pronunciation, and therefore spelling, and yields further background information about the locality. Some locations in the field are recorded with Global Positioning Systems (GPS) whereby satellites provide a grid reference which can later link to digital maps. Various tools act as memory triggers, including Ordnance Survey (OS) maps (especially the c.1880, 1901 and 1973 editions), photographs and aerial photographs. Particularly valuable have been copies of the RCAHMS Royal Air Force aerial photographs of 1944–1951. We have teamed these up with transcripts of the names collected by John Stewart in 1951 to help pinpoint individual rig, hill and coastal names. We are also extracting information from documentary and printed sources deposited in the local museums, library and archives. It is also very important to cross-check written evidence with current local knowledge to avoid mistakes in suggested origins of names. This is demonstrated in the following example: Stewart (1987, 317) lists Fersills originating from

6 Prior to my input, school children were introduced to Viking place-names through a workbook page of Lincolnshire examples. 7 Photograph annotated by the late Ronnie Johnson, Eshaness. 46 Eileen Brooke-Freeman Courtesy Ronnie Johnson Courtesy Ronnie Fig. 1: An annotated photograph: Hamnavoe place-names 1: An annotated photograph: Hamnavoe Fig.

The Journal of Scottish Name Studies 4, 2010, 43–60 Memories, Meids and Maps: the Shetland Place Names Project 47 Fig. 2: John Stewart of Whalsay’s 1951 questionnaire (SA D.27/1/1) 1951 questionnaire Whalsay’s of Stewart 2: John Fig.

The Journal of Scottish Name Studies 4, 2010, 43–60 48 Eileen Brooke-Freeman ON forsoela f. ‘a shady place’. Firsills is listed on an original record sheet and current local knowledge confirms the house name as coming from the occupant Taamie Fraser (nicknamed Taamie Firsil) who died there in 1902.8

John Stewart Archive The papers of John Stewart of Whalsay, deposited in the Shetland Archives, are one of our most useful sources.9 In 1951 Stewart, a Shetlander living and working in Aberdeen, followed the model of similar projects in Western Norway and issued almost 5,000 numbered questionnaires to schoolchildren to take home to their parents and grandparents and to other houses in the district (Fig. 2). The questionnaire asked respondents to record every name they knew, spelling the names as they said them, grouping them by names on their croft, names in the hill, names at the shore and names in the sea. The response was overwhelming. Almost 1,200 original sheets are now deposited in the Shetland Archives, some listing up to 250 names. From the record sheets, Stewart extracted lists of names for his own field notes which he added to and amended on his annual field trips. He traced every 6 inch : 1 mile OS map for the whole of Shetland, copying all the names as they appeared on the map. These were stuck into books and gradually annotated and amended during his annual field trips. He then proceeded to compile alphabetical lists of names covering each parish; these lists are estimated to contain over 30,000 different place-names. John Stewart’s detailed study of Shetland island- and farm-names, was published posthumously as Shetland Place-Names in 1987, 10 years after his death. It comprises almost 4,000 names but, to his disappointment, the immense volume of information amassed meant that he was unable to subject the entire collection to the detailed scrutiny he desired.10 In order to help make the entire collection more accessible, we are striving to pinpoint names recorded in this way by Stewart on maps, adding information about these and other known place-names. Transcripts of the original sheets are taken out to people in each community in order to try

8 Helen Jamieson, born in 1907, recited the following poem which also verified the link with the name Grindypoil (also recorded by Stewart 1987, 220): Thomas R Fraser is my name and England is my nation Grindypoil is my domain and it is my possession And when I’m dead and in my grave and all my bones are rotten Look at this and think of me and say he’s quite forgotten. 9 SA D.27/1/1–100. 10 In 1963 Stewart wrote: ‘My place-name stuff is an immense accumulation now. The problem is to get anything published … I would certainly be willing to publish it as a book or books. The difficulty is finding time to polish it off as I would like it.’ (Stewart 1987, viii). Memories, Meids and Maps: the Shetland Place Names Project 49 and verify names and, more importantly, locate them on the map. It is a slow process as many of the sheets are difficult to read due to Stewart’s work method: as he copied names onto his subsequent lists, he crossed them off the record sheets, often almost completely obliterating them. The lists work in the same way as OS maps, jolting memories and resulting in sheet names being corrected and added to. They are particularly effective when used in conjunction with enlarged copies of the RCAHMS aerial photographs and, as a result, we are pinpointing the exact location of many of these, in many cases, almost obsolete place-names.

The Shetland Place Names Database We established the Shetland Place Names Database, which is linked to digital maps. It is based on the Scottish Place-Name Database11 and includes fields for linguistic, historical and geographical analysis of place-names within their environment. Possible searches include all the names for a particular area, all occurrences of a particular feature (e.g. all noosts or skerries), all names that include a specific linguistic element (e.g. all borg- or kví-names) and all the names which include a particular word (e.g. baa). Unlike the Scottish Place- Name Database, the Shetland Place Names Database precisely locates each place-name with a 10-figure grid reference showing the location on maps. This is achieved through our Geographic Information System (GIS).12

GIS digital map system The key to recording and presenting data is the use of digital mapping. We have developed the interface between the database and GIS so that we can enter place-names by first plotting them on digital maps and then transferring them into the database for the addition of further information such as the source, classification, informant details and pronunciation, before the data is exported back into GIS. This enables quick and accurate data entry. Digital mapping also provides us with a powerful means of presenting data. We can show layers of information gathered from different sources and maps can be created based on the specific searches previously mentioned, e.g. all the names for one village, all names containing the element borg and all rigs or baas (Fig. 3). Results appear on the Shetland map and in a table, which can be enlarged in order to study a specific area or individual place-names. When new names are incorporated, a

11 The Scottish Place-Name Database is based at the University of Edinburgh’s Department of Celtic and Scottish Studies. 12 ESRI ArcGIS 9. 50 Eileen Brooke-Freeman Fig. 3: Digital mapping output 3: Digital Fig. © and database right Crown Copyright and Landmark Information Group Ltd. All rights reserved. Licence no. NG00228.

The Journal of Scottish Name Studies 4, 2010, 43–60 Memories, Meids and Maps: the Shetland Place Names Project 51 Fig. 4: Da Rentils wi da Muckle Knowe wi da Muckle Rentils 4: Da Fig.

The Journal of Scottish Name Studies 4, 2010, 43–60 52 Eileen Brooke-Freeman map is generated and taken to informants for checking – both names and location; any necessary amendments and additions are made with updates subsequently made to the database.

Fishing meids A parallel project in conjunction with the Shetland Biological Records Centre has recorded Shetland’s fishing meids. Before the advent of navigation with radar and GPS, fishing grounds around Shetland were located by taking the transect of two meids – each meid involving lining up two landmarks. Meids were used traditionally to locate species of fish, but many meid names refer to features which can only be seen from the sea, while some are alternative names, either descriptive names – reflecting the view from the sea – or noa names, which arose due to taboos or superstitions over not using certain words while in a boat. Many meid names, however, are now at risk of being lost. We interviewed 16 retired fishermen to collect information on meids and the fishing grounds they were used to locate. Information was gathered on place-names relating to meids and the wider coastline and we amassed an oral history relating to the contemporary environment, fishing and culture. This yielded a detailed history of the halibut-line, haddock-line and seine-net fisheries with descriptions of the grounds, details of place-names and notable words and phrases. The place-names used in meids vary depending on the fishermen’s knowledge of the land and their distance from it; sometimes the sea names are distinctively different or more descriptive than the land names. A common tradition was that the use of correct names could bring bad luck while fishing, with the result that noa names were created, often describing the appearance of landmarks seen from fishing grounds. In Unst, Da Noup was Da Dongi ‘the heap’ and Housafield was Da Jokkel ‘the projecting knowe’. The heavy surf at Strandibrough in Fetlar led to its sea name Da Fustra. Scottie’s Wart has two aliases: Da Dai to Unst fishermen and Da Dwarg to Yellmen. Northmavine features with alternative sea names include Ness of Olnesfirth (Da Hill of Hamar), Hill of Haggrister (Da Felsen), Holms of Burravoe (Da Flooery Holms) and Ronas Hill (Bloberg, which describes its appearance from a distance as a blue rock). Papa Stour’s Virdiefield and the Hill o Feallie are described in terms of their proximity to the mainland: Oot Feallie and In Fiellie, respectively. The sea name for Da Noup in Foula is Jockie’s Cap, from its appearance when aligned with the South Ness. Da Clifts, Cunningsburgh, are called Da Brunes (from ON brún f. ‘brow, slope’). The valley of Quarff – clearly visible both east and west of Shetland – is known as Quarff Scord Memories, Meids and Maps: the Shetland Place Names Project 53 and features in several meids. Groups of prominent hills often had collective names, for example Da Pobies (Unst), Da Scally hills (Lang Kames), the three Anderhills (Bressay) and the three Hallilees in the south mainland. Some place-names have been forgotten on land but preserved in meids. Houlastongas was the fishermen’s name for the sound between the Holm o Skaw and Whalsay (from ON holmr m. ‘holm’ and ON stong f. ‘pole’).13 Burgidale was an old name for the valley between Sumburgh Head and Mid Head, named after the fort once situated at Sumburgh Head (from ON borg f. ‘fort’ and dalr m. ‘dale, valley’).14 Today there is evidence of ditches and banks but any structures seem to have been obliterated when the building of Sumburgh lighthouse commenced in 1819. The name Burgidale is now only preserved in the meid Water in Burgidale : seen from 13 miles to the east, the valley dips below the horizon giving the appearance of water in the dale. From the recordings, transcripts were made and reports compiled covering each type of fishery and ground.15 Additionally, a database and GIS system allows us to plot the landmarks and grounds, linking it to the Shetland Place Names Database (Fig. 4). The transcripts of the fishermen’s stories have been used to produce a book; Water in Burgidale by Charlie Simpson was published in August 2010 and gives a detailed insight into in the days of fishing before the use of navigational aids.

Place-names on pavers The Meids Project is just one example where the Shetland Place Names Project has worked with other areas of Shetland Amenity Trust. Another initiative involved recycling place-names in a new and exciting way. The steps and terrace outside the new Shetland Museum and Archives Boat Hall were laid with recycled glass pavers, 46 of which incorporated place-names. These paving slabs are inlaid with crushed serpentine mixed with resin, an initiative developed through working with Shetland Amenity Trust’s own recycling company, Enviroglass, and local artist Alan Hart. The steps feature coastal names, moving from sea at the bottom to land at the top, using green serpentine for the sea names and dark red for the land ones (Fig. 5). On the terrace, the place-names relate to the land and take one on a tour of Shetland from Finniquoy (Fair Isle) through to Hamar (Unst) at the top.

13 Jakobsen (1993, 104) notes that ‘In Norway “Stang” (Eng. pole) occurs as a name of isles and peninsulas, and commonly as the name of a headland, in the compound “Stangnes”. – Shetla. Stonger-holm [ståņgrom], the name of a holm off Nesting.’ 14 The Ancient Fort of Swenbrugh was recorded by Timothy Pont and published by Johann Blaeu in 1654. 15 Shetland Amenity Trust Meids reports. 54 Eileen Brooke-Freeman

Fig. 5: Pavers at the Shetland Museum and Archives Boat Hall

Photo Eileen Brooke-Freeman

The Journal of Scottish Name Studies 4, 2010, 43–60 Memories, Meids and Maps: the Shetland Place Names Project 55

Fig. 6: The distribution ofkleber , klamel or klemer-names

After Birmingham Archaeology

The Journal of Scottish Name Studies 4, 2010, 43–60 56 Eileen Brooke-Freeman My challenge was to select a range of interesting place-names no more than twelve characters in length, which linked to the themes in the museum displays. I wanted to incorporate both particularly unusual place-names, such as Brimfooster, Winyadepla, Houlastongas, Cockstool and Veltamatas, and names common throughout Shetland, such as Swarta Skerry ‘black skerry’, Loomishun ‘small loch of the diver’, Gorsendi Geo ‘dyke-end geo’, Da Vadill ‘the shallow place in water which can be forded’, Kirkabister ‘church farm’ and Linga ‘heather isle’. It was also important to ensure good geographic coverage, have examples of common geographic features and consider the look and sound of each name. Some of the names used were only known orally and had been newly recorded as part of the Shetland Place Names Project. All the selected place-names provide clues about Shetland’s history, the landscape or the habitats of birds, animals and plants. Names chosen to represent natural features include geos, taings, ayres, scords, stacks and skerries. Burgataing in North Roe and Brindister denotes the location of a broch (ON borg f. ‘fort’) and Gulga in Sandwick was once a gallows site (ON gálgi m. ‘gallows’). Bragaster in Papa Stour features in Shetland’s earliest document, dated 1299, as Breka sætr ‘homestead on the slope’, whilst Pettawater in the Lang Kames is linked to stories of picts (ON péttr m. ‘pict’) or trows. Klingrahool, Catfirth has wild roses (ON klungr m. ‘brier’) growing alongside the nearby burn and several Ern Stacks were the nesting place for sea eagles (ON örn m. ‘sea eagle’). Other interesting names include Hegary’s Böd on Gunnister Voe (named after the German trader Simon Harratstay or Hagarskilde, trading there in 1582), Diggers Rest in Yell (built by Andrew Anderson of Cunnister on his return from the Australian gold-mines in the late 19th century) and Harley Street in Scalloway, where the Scalloway Health Centre is found today.

Kleber Another cross-Trust activity saw me working closely with archaeologists and geologists to research kleber-names. This formed part of a book considering the history of soapstone in Shetland (Foster and Turner 2009). Many of the places mentioned carry kleber, klamel or klemer-names: the Shetland dialect word for soapstone is kleber (from ON klé m. ‘loomweight’ and ON berg n. ‘stone’, one of the original uses of steatite). This was a fascinating assignment and involved layering the place-names on geological maps and conducting site visits with a geologist to verify the presence of the soft stone. I was delighted to add a further 10 names to the archaeologists’ maps – including one where the geologists had missed a small outcrop on the shoreline, although it was clearly evident from the place-name, Clibberswick (Fig. Memories, Meids and Maps: the Shetland Place Names Project 57 6). In Fetlar one of the earliest forms of the name, Claeberg, was plotted on one of the few detailed John Stewart maps.16 Today this appears to be a ridge where the softer rock has been cut away leaving only the harder rock behind – there are distinct hollows at the north end of the ridge suggesting the cutting away of the steatite. Other linked trust activities have involved providing place-name information for the Shetland Geopark website and presentations and providing material to be incorporated in the Museum and Archives displays on German trading, the St Ninian’s Isle treasure, the Gunnister Man and emigrants. The latter project was part of the 2010 Shetland Hamefarin (‘homecoming’) and considered how Shetlanders took local names with them when they emigrated. We identified and plotted Shetland place-names throughout the world and collated photographs of house and road signs. Shetland Amenity Trust has for many years participated in transnational projects working with colleagues throughout what can be described as the Viking world. The latest of these is the Northern Periphery Programme THING Project which considers North-Atlantic assembly sites.17 We are examining and comparing thing-sites, developing site management and tourism strategies, interpreting sites and telling the common thing story, including considering the wider use of ting, thing, ding and fing-names. I am leading Shetland’s work package, which forms 45% of the entire project, and we have subsequently been able to appoint a Place Names Assistant to ensure the Shetland Place Names Project keeps up the momentum during the period to June 2012. This is an exciting development and we are now tackling the backlog of names to incorporate into the database, transcribing names from Stewart and other archive sources, continuing working with local groups and establishing new contacts. The database will soon reach 10,000 entries, over half of Stewart’s 30,000 place-names have been transcribed and we have a busy year of activities planned with local groups.

The Shetland Place Names Project and other organisations The project has collaborated with other agencies such as Shetland Arts for their annual writing prize in 2006 which featured place-names in poetry,18 Shetland Islands Council Ferries providing guided trips around the northern isles of Yell, Unst and Fetlar, Hjaltland Housing Association supplying information to help them choose new housing scheme and street names,

16 SA D.27/1/14. 17 Northern Periphery Programme 2007–2013 Thing sites International Networking Group (THING Project). 18 The New Shetlander 238. 58 Eileen Brooke-Freeman and the Shetland Cattle Breeders Group researching place-names associated with cattle for their centenary book (2009). I further worked with Shetland Folklore Development Group and the Northern Periphery Programme Sagalands project identifying the location of sites relating to trows and places that feature in the sagas.

Outreach We also undertake a range of outreach work, including a series of talks to a variety of groups and conferences, as well as exhibitions staged at various events. We have participated in tour guide training and an introduction to place-names sources in the Shetland Archives. We are currently developing an introduction to a place-names course. Attending local shows has been particularly worthwhile, often yielding information on further potential contacts throughout the isles. A monthly column appears in the Shetland Life magazine, each issue considering a particular place-name or place- name element, giving local examples and seeking further information from readers. This often results in new informants coming forward, for example an article on burial sites in the hill yielded several new names and locations not previously mapped. We also produce an occasional newsletter and have submitted articles for other local publications.

Conclusion The future of the project looks really healthy. Our next major challenge is to make the maps and database more universally accessible. This will, of course, take time and money, but it will be pleasing to feel that we are using the latest technology to present a complete picture of what is happening in the landscape. There is immense scope for presenting and using the data. Base maps can be linked to detailed databases with potential to link in old maps, aerial and site photographs, archive and printed source material and sound recordings. Modern technology dictates our goal of making the information available to all interested parties worldwide and to broaden the scope for the future researcher, be it the local resident, the tourist, the school child, the person in search of a house name, the historian, the official charged with producing maps or road signs, the historical geographer, the geologist, the archaeologist or the linguist.

References Blaeu, Johann, 1654 (2006), Atlas Novus, based on maps surveyed by Timothy Pont and reworked by Robert Gordon of Straloch and James Gordon of Memories, Meids and Maps: the Shetland Place Names Project 59 Rothiemay, reprinted as The Blaeu Atlas of Scotland, with translation by Ian Cunningham, by Birlinn in association with the National Library of Scotland, Edinburgh 2006. Also available at . Gammeltoft, Peder, Carole Hough and Doreen Waugh, eds, 2005, Cultural Contacts in the North Atlantic Region: the evidence of names (Lerwick: NORNA, Scottish Place-Name Society and Society for Name Studies in Britain and Ireland). Foster, Amanda K. and Val E. Turner, eds, 2009, Kleber: Shetland’s Oldest Industry . Shetland Soapstone Since Prehistory (Lerwick: Shetland Amenity Trust). Jakobsen, J., 1993, The Place-Names of Shetland (Lerwick: Shetland Library; 1st pub. 1936). Northern Periphery Programme 2007–2013 Thing sites International Networking Group (THING Project). See and . RCAHMS: Royal Commission on the Ancient and Historical Monuments of Scotland, Royal Air Force Post-war Survey 1944–1951 (Shetland Amenity Trust duplicate copy gifted by RCAHMS). SA: Shetland Archives, D.27/1/1–100: Papers of John Stewart. Shetland Amenity Trust Meids Project: reports on the fishing grounds of Shetland, 2001–2004. Shetland Cattle Breeders Group, 2009, A Pictorial ‘Daander Trowe’ Shetland’s Crofting Culture. Centenary Celebration. Illustrating the role of the Shetland Coo – A Breed from the Past with a Place in the Future (Lerwick: Shetland Cattle from Shetland Breeders Group). Sigurðardóttir, Turið and Brian Smith, eds, 2010, Jakob Jakobsen in Shetland and the Faroes (Lerwick, Shetland Amenity Trust and the University of the Faroe Islands). Stewart, J., 1987, Shetland Place-names (Lerwick, Shetland Library and Museum).

Further Reading Brooke-Freeman, E. L., 2003, ‘John Stewart’s Shetland Place Name Survey’, The New Shetlander 223. Brooke-Freeman, E. L., 2005, ‘Shetland Place Names Project’, in Gammeltoft, Hough and Waugh, 42–57. Brooke-Freeman, E. L., 2006, ‘Place names and poetry: the 2006 Shetland Arts and New Shetlander Writing Prize’, The New Shetlander 238. Brooke-Freeman, E. L., 2009, ‘Locating Shetland Steatite using Place Name 60 Eileen Brooke-Freeman Evidence’ in Forster and Turner, 18–26. Brooke-Freeman, E. L., 2009, ‘Place names relating to kye’, in Shetland Cattle Breeders Group, 67–69. Brooke-Freeman, E. L., 2010, ‘Putting Jakobsen on the map: collecting Shetland place names in the twenty-first century’, in Sigurðardóttir and Smith, 99–109. Jakobsen, J., 1928 and 1932, An Etymological Dictionary of the Norn Language in Shetland (London and Copenhagen). Simpson, C.H., 2003, ‘On meids and fish stocks’, The New Shetlander 225. Simpson, C.H., 2010, Water in Burgidale: Shetland Fisheries in a Pre-electronic Age (Lerwick, Shetland Amenity Trust). Stewart, J., 1951, ‘Shetland Place-Name Collection’, The New Shetlander 28. Stewart, J., 1953, ‘The Place-Names Survey’, The New Shetlander 36. Tait, I., 2010, From Old Rock to New Life. Shetland Museum and Archives exhibition for Shetland Hamefarin 2010 (Lerwick, Shetland Museum and Archives). 0 10 20 40 60 M

0 10 20 40 60 80 100Km

Scottish Gaelic Sannda and Its Aliases

Richard A.V. Cox Sabhal Mòr Ostaig

§1 Introduction The Scottish Gaelic names Àbhainn, Sannda and An Spàin all denote the small island that lies off the southern tip of Kintyre, within the parish of Southend. The following article looks at the history and etymology of these forms.1

§2 Avona Porticosa Francis Groome’s Ordnance Gazetteer of Scotland (1884, 93) describes Aven as ‘a modern provincial abbreviation of “Avona Porticosa”, the ancient name of the island Sanda in Southend parish, Argyllshire.’ Of the island’s natural harbour, his 1896 edition states that:

[It] was a common station of the Scandinavian fleets during the contests for the possession of Kintyre and The Hebrides. The island, in this con- nection, was then called Avona Porticosa – a name which it still retains, in the abbreviated form of Avon, among the Highlanders; but it figures, under its more proper name of Sanda, in the more ancient record of Adamnan’s Life of Columba. Francis Groome, Ordnance Gazetteer of Scotland, 1896, 318

Similar descriptions go back to the late 18th century:

SANDA, an island, in the parish of Southend, district of Cantyre, county of Argyll; containing 11 inhabitants. This is a small island, lying near the outer extremity of the peninsula of Cantyre, and measuring about a mile and a half in length and half a mile in breadth; its name is of Scandinavian origin, and signifies ‘Sand Island.’ It possesses a good natural harbour, although between the island and the main land the sea is extremely turbulent and dangerous, and for two or three months in the year the place cannot be approached by a small boat. Sanda was a common station for the Scandinavian fleets during the contests so long carried on for the possession of Cantyre and the neighbouring islands. Samuel Lewis, A Topographical Dictionary of Scotland, 1846: s.n. Sanda

1 A version of this article was presented at the Scottish Place-Name Society’s one-day conference at the University of Glasgow on the 7th November 2009.

The Journal of Scottish Name Studies 4, 2010, 61–102 62 Richard A.V. Cox The island of Sanda, separated from the main land by a channel three miles across, is of irregular form, about four miles in circumference, and being covered with good pasture, serves the purpose of a large sheep- farm. It has passed, at different times, under different names, though its present appellation is considered the most ancient, on the authority of Adomnan, Abbot of Iona, who wrote the life of St. Columba in the year 680. During the visits of the Scandinavians to these coasts, and their attacks upon the district for the possession of Cantyre and the adjacent islands, Sanda, according to the historian Buchanan, was an important station for their fleets; when the Danish fleet assembled here the isle was called Avona Porticosa, and by the natives it is still termed Aven. The sound is much frequented for its anchorage by small vessels sailing up the Frith of Clyde, which has about twelve fathoms of water at three miles from the shore. Ibid.: s.n. Southend

At the time [Sanda] was the rendezvous of the Danish fleet, it was called Avona Porticosa; and is still called Aven by the Highlanders. Sanda, how- ever, is the more ancient name, as appears from the life of St Columba, written by Adomnan, Abbot of Iona, in the year 680. New Statistical Account, 1834–45, Vol. 7: Argyll, Southend, 415

[Sanda’s] ancient importance, as the station of the Scandinavian fleets during the contests for the possession of Cantyre and the neighbour- ing islands, is well known; and the anchorage is still frequented by the smaller classes of vessels which navigate the Clyde. John Macculloch, A Description of the Western Islands of Scotland, 1819 II, 440

In former days, this anchorage was of more importance than it is now; Sanda having been a common station for the Scandinavian fleets dur- ing the contests so long carried on for the possession of Cantyre and the neighbouring islands. The name Avona and Avon, by which it was known, is a corruption of the Danish Hafn, a haven. John Macculloch, The Highlands and Western Isles of Scotland, 1824 II, 68–69

Sunda island, above a mile and a half in length, and half a mile in breadth, is noted as the place of rendezvous for the Danish fleet, in their descents Scottish Gaelic Sannda and Its Aliases 63 on those coasts: hence it was called Avona Porticosa ... James Playfair, A Geographical and Statistical Description of Scotland, 1819 II, 7

Not far from this rock is the island of Sanda, above a mile and a half in length, and half a mile in breadth, famed as the place of rendezvous for the Danish fleet, in their excursions to these coasts. Hence it went under the name of Avona Porticosa, and is still sometimes called Aven. Sanda, however, is the more ancient, as well as the more common name as ap- pears from St Columba’s life by Adomnan. Statistical Account of Scotland, 1791–99, Vol. 3: Argyll, Southend, 363

The salient elements within these accounts are that the primary nameof the island is Sanda §3, meaning ‘sand’; the island offers good anchorage or harbourage, which was formerly used by the Scandinavians; the island was previously known as Avona Porticosa, or Avon (Aven) for short. The formSanda , however, is seen as the older name, ‘as appears from St Columba’s life’ (SAS ); and Macculloch takes the name Avon to be from the Danish word ‘Hafn, a haven’. A considerable degree of interdependence is apparent between these descriptions, but all would appear to go back ultimately, although not necessarily directly, to George Buchanan’s description of 1582:

A promontorio Cantiera paullo plus mille passus abest Avona, id est, portuosa; id cognomen adepta ab statione navium, quod, cum Dani in- sulas tenebant, ad eam classes eorum cursus dirigebant. Robert Fribarn’s edition: George Buchanan, Rerum Scoticarum Historia, 1727, 22

‘From the Promontory of Kintyre a little over a mile lies Avona, i.e. “well supplied with a harbour”; having got that name from the anchoring of ships, because, when the Danes held the islands, their fleets would set course for it.’2

2 Sutton 2009, §32: A promontorio Cantera paulo plus mille passus abest Avona, id est portuosa. Id cognomen adepta ab statione navium, quod cum Dani insulas tenebant ad eam classes eorum cursus dirigebant (); ‘From the promontory of Cantyre, a little more than a mile, lies Avona, now Sanda, called Portuousa, i.e., fit for a port. It got that name from being a road for ships, for when the Danes possessed those islands their fleets directed their course thither for shelter’ (

Avoyn. Befor the south poynt of the promontory of Kyntyre, lyes be ane myle of sea, ane iyle neire ane myle lange, callit the iyle Avoyn, quhilk iyle is obtained that name fra the armies of Denmark, quhilkis armies callit it in their leid Havin, it is inhabit and manurit, and guid for shipps to lay one ankers. Donald Monro, Description of the western isles of Scotland, called Hybrides ... 1549, 1774, 6

‘Avoyn. To the extent of one sea mile off the southern promontory of Kintyre lies an island nearly a mile long called the Isle Avoyn, which island obtained that name from Denmark’s armies, which armies called it Haven in their language; it is inhabited and fertile and good for ships to lie at anchor.’5

scothist/1eng.html#I.32>). Aikman (1827, 43) translates: ‘Little more than a mile from the promontory of Cantyre lies Avona, [now Sanda,] that is Portuosa, full of havens, a name affixed on account of its being a naval station; for, when the Danes had possession of these islands, it was the general rendezvous for their fleets.’ 3 A particularly debased form of Buchanan’s Avona Portuosa, whose development (presumably based upon both a mistranscription and a supposed etymology) defies reconstruction, occurs in an anonymous description of the first half of the 17th century: ‘And eastward from Dunawardie two mylls off the land there is ane litle Illand of ane Myll length and half ane myll breadth called Awin, which the Romans did call in the tyme of Julius Cæsar, Porta Eosa Avona. Upon the Landsyde of it, next to Cantyre is a verie good harborie. On the east end of it is the Sheep Illand where there is verie manie Coneys and arrettis. The streame runns so swiftlie that no shipps can remaine near it, except they be within the harborie. In this Illand of Awin there is ane litle Chappell ...’ (from ‘Ane Description of Certaine Pairts of the Highlands of Scotland’ in Mitchell 1907–08 II, 144–92: 187; Mitchell (ibid., xxiii) notes that ‘Sibbald says in his Repertory of Manuscripts, p. 22, that this was a communication to Robert Gordon, and Bishop Nicholson says that it was “by a Native”’). 4 This is the view of R.W. Munro (2002, 291). 5 An almost identical version is printed in MacFarlane’s Geographical Collections: Scottish Gaelic Sannda and Its Aliases 65 Our oldest source, however, is John of Fordun (1384 x 1387, Skene 1871– 72 I, xiv).6 His list, de insulis Scociæ, includes:

Insula Awyn, ubi cella Sancti Adamnani, ibique pro transgressoribus re- fugium. Walter Goodall’s 1759 edition [Edinburgh College Library MS]: Joannis de Fordun Scotichronicon I, lib. ii, cap. x, 45

‘The Isle of Awyn, where the chapel of St Adomnán is, and where there is a refuge for transgressors.’

In other MSS, this reads:

Insula Awyne, ubi capella Sancti Anniani, ibique pro transgressoribus refugium. Thomas Hearne’s 1722 edition [Trinity College, Cambridge MS]: Johannis de Fordun Scotichronicon, lib. ii, cap. x, 81

Insula Aweryne, ubi capella sancti Sannani, ibique pro transgressoribus refugium.7 William F. Skene’s 1871–72 edition [The Wolfenbüttel MS]: Johannis de Fordun: Chronica Gentis Scotorum I, lib. ii, cap. x, 43

‘Avoyn. Befor the south poynt of the promontory of Kyntyre lyes be ane myle of sea, ane Iyle neire ane myle lange callit the Iyle of Avoyn quhilk Iyle is obteinit that name fram the armies of Denmark, quhilkis armies callit it in their leid Havin. It is inhabit and manurit and guid for shipps to lay one ankers.’ (Donald Monro, A Description of the westerne isles of Scotland by Mr. Donald Monro, quho travelled through maney of them in Anno 1549 in Mitchell 1907–08 III, 262–302: 265). From a different MS: ‘Before the south poynt of the Promonterie of Kintyre lyis be ane lang myle of sea ane Ile neirest ane myle lang callit the Ile of Avoin, quhilk Ile has obteinit that name fra the Armes of Denmark, quhilk Armes are callit in their leid Havoin, inhabite and manurit, gude for schipps to ly on ankeris’ (Monro [1549] 2002, 301). 6 While Buchanan is known to have used John of Fordun as a source for his own history (Innes [1729] 1885, 201–15), this was not the case in his preliminary description of Scotland. Thomas Pennant (1772, 218 and note) summarises both Buchanan and Monro: ‘Sanda, or Avoyn, or island of the harbours’. 7 ‘Aweryne (Sanday), where is the chapel of Saint Sannian, and a sanctuary for transgressors’ (Skene 1871–72 II, 39; Skene writes Averyne in his notes (p. 386)). Skene gives a footnote to two other MS readings (including Trinity College’s) of the saint’s name, but not of the place-name (ibid I, 43). In his (later) edition of Vita Columbae, however, Skene (1874, cxlviii) quotes Goodall’s 1759 edition: ‘Insula 66 Richard A.V. Cox Of equal importance is the title of a c.1644–51 description of the island (see Appendix), which contains our earliest genuinely Gaelic form of the name:

Insulae Sandae seu Avoniae, Hibernice Abhuinn, Brevis Descriptio. R. P. fratris Edmundi mac Cana ‘A short description of the Island of Sanda, or Avon, in Irish Abhuinn, by the Reverend Father, Brother Edmund MacCana’

The latest of our early forms are Robert Gordon c.1636–52 (map of Kintyre: Cantyre) Yl. Avon or Sanda and Blaeu 1654 (surveyed by Timothy Pont between 1583–96) Avon or Sanda.8 In addition to his map’s Avon, Blaeu’s texts give latinised Avona and Avena.9 For the modern form and etymology of this name, see §5.

Awyn ...’, as above. The editors of Origines Parochiales Scotiae (II, 9) translate ‘Isle Awyn’, also following Goodall’s edition. 8 The editors of Origines Parochiales Scotiae (I, 9) state that Timothy Pont gives ‘Yl Avon or Sanda’; however, although Yl. occurs frequently in island names given by Pont, this is not the case in this instance on the map that I have had access to: Stone 1991, Plate 19, and NLS . Martin Martin (1703, 228) gives ‘The Isle Avon’. KAS (1938, 5) gives forms cited in Origines Parochialis Scotiae: ‘Aven, Avon, Avona Porticosa’, adding, ‘Old Name for Sanda’. 9 Blaeu’s description (f. 119) of Sanda is taken directly from George Buchanan: ‘A promontorio Cantera paulo plus mille passus abest Avona, id est, portuosa: id cognomen adepta ab statione navium: quod cum Dani insulas tenebant, ad eam classes eorum cursus dirigebant’, although he adds in italics, ‘Avonam vide in tabula Kanteræ’ (NLS: style=maps.xsl&wid=700&hei=500&browser=win_ie&plugin=false>), which Ian Cunningham translates as, ‘A little more than a mile from the promontory of Kintyre is Sanda Island, that is with a harbour; the name is taken from an anchorage for ships; because when the Danes held the islands, their fleets directed their course to it. See Avon on the map of Kintyre.’ (NLS: ). Also in Blaeu’s altas, the form Avena (f. 9v; not Avenna as in Gammeltoft 2006, 77) occurs in Andreae Melvini Scotiae Topographia, Andrew Melville’s ‘Topography of Scotland’, a long poem to Prince Henry Frederick, at this point evidently based upon Buchanan’s own text: ‘A promontorio Canteira abscessit Avena / Passus mille, sua naves statione receptans; // Dirigeret cum cursum ad eam olim Danica classis.’ (NLS: ), which Cunningham translates as ‘A mile distant from the promontory of Kintyre is Sanda, receiving ships in its anchorage, since once the Danish fleet directed its course to it.’ (NLS: ). Avona and its variants are omitted from later maps. Scottish Gaelic Sannda and Its Aliases 67 §3 Sannda The island is known asSandey in Hákon Hákonsson’s Saga, written shortly after King Hákon’s death in 1263:10

Eftir þat sigldi Hakon konungr vndan Melansey ok laa vm nottina vndir Hersey ok þadan vndir Sandey ok sua til Satirismula ok kom vm naatti- na nordr vndir Gudey. þadan sigldi hann vt i Jlarsund ok laa þar .ij. nætr. Vigfusson and Unger’s edition: ‘Hákonar Saga hins gamla’, Flateyjarbok III, 227, §28111

‘After that King Hakon sailed away from Holy Island, Lamlash, and lay during the night off Arran and then off Sanda and so to the Mull of Kintyre and came during the night northwards off Gigha; then he sailed out into the Sound of Islay and lay there two nights.’12

Similarly, in Magnus Barefoot’s Saga, in Heimskringla, written c.1230 of events about the turn of the 11th century (1093–1103):

Vítt bar snjallr á slétta Sandey konungr randir Linder and Haggson’s edition: ‘Saga Magnús konungs berfœtts’, Heimskringla, §1013

Vitt bar sniallr a sletta Sandey konvngr rander Finnur Jónsson’s edition: [Magnus Barefoot’s Saga], Eirspennill, Chap. 7, 11814

10 In 1264–65 by Sturla Þórðarson (Pulsiano and Wolf 1993, 259). 11 The reference HSH.326 Sandey in Gammeltoft 2006, 67, 77 is expanded in error as in the Eirspennill redaction. 12 Bremner 1923, 251 (after Dasent [1894] 1997 II, chap. 326, p. 362): ‘After that king Hacon sailed away from Malas-isle (Lamlash), and lay for a night under Arran, and thence under Sandisle, and so to the Mull of Cantire, and came in the night north under Gudey. And thence he sailed out into the Islay-sound, and lay there two nights.’ Munch (1859, 446) translates into Norwegian: ‘Derefter seilede Kong Haakon fra Melasø og laa om Natten under Herø; derfra seilede han under Sandø og forbi Santíresmulen og kom om Natten nord til Gudø. Derfra seilede han til Ilarsund, hvor han laa i to Nætter ...’ 13 An older version occurs in Morkinskinna, 317 §132: ‘Vitt be{ sna{ aslętta / Sandey konvngr randir’. 14 The reference MSB.9Sandey is given in error in Gammeltoft 2006, 77. 68 Richard A.V. Cox Map (a) The Northern and (b) the Southern Hebrides Cape Wrath N

Lewis The Point of Stoer

St Kilda Clachan Shannda

N. Uist

Skye

Barra Canna Sanndaigh

0 10 20 40 60 M

0 10 20 40 60 80 100Km

Sanndraigh (a)

(b)

Jura Tarbert

Firth of Clyde Sound of Islay Islay Gigha

Muasdale Carradale Kintyre Arran Holy Isle

Pladda

Rathlin Mull of Kintyre

Sruth na Maoile Sannda © Maps in Minutes 2004 TM East Ulster Larne TM 2004 © Maps in Minutes © Maps

The Journal of Scottish Name Studies 4, 2010, 61–102 Scottish Gaelic Sannda and Its Aliases 69 ‘The valiant king bore far the shields upon the plain of Sandey’15

Campbell (1924, 11) notes – without expansion – that ‘[t]he island is on occasion written of as Sanderey’,16 taking it to be an ‘elaboration of the shorter name’ and echoing Paterson and Renwick’s (1900, 203) ‘Another name for Sanda is Sanderey, which may mean Sand eyrr ey, Sand beach island ...’ However, these assertions appear to be the result of confusion with a superficially similar name or names in the Outer Hebrides: Captain Thomas (1880, 365) states that ‘[i]n 1202 Olaf was in Sanday (Sandarey, Barra; or Sand (Sandar), North Uist).’17 Thomas’s reference is to Hrafn’s Saga:18

Þessu næst kómu þeir skipi sínu í góða hôfn við ey þá, er Sandey heitir, ok þar reistu kaupmenn hafnarmark. Svá segir Grímr:

Hér hefir beitt at brattri Bótólfr skipi fljótu, áðr fell sær um súðir, Sandeyju, skæ branda.

Reisti sjálfr, ok sýsti, snarr félagi harra hafnarmark, fyrir hrefnis happsverk gota sterkan.

Þeir lágu við Suðreyjar í góðu lægi nôkkurar nætr. Þá réð fyrir Suðreyjum Óláfr konungr. ‘Hrafns Saga’ (Helgadóttir 1987, 21–22)

15 Collectanea de Rebus Albanicis, 1847, 348: ‘The valiant king bore far and wide / The shields upon the plain / Isle of Sandey.’ 16 So KAS 1938, 29. 17 Thomas’s view is repeated in the form of a letter published posthumously by Vigfussen in his Icelandic Sagas (1887, xxxvii note 1) and cited verbatim in Tjomsland 1951, 59–60. (Vigfussen (ibid., xxxviii) refers to Thomas’s 1880 article as ‘Proc. Antiq. Scot. Vol. XI’, in error for ‘... XIV’.) 18 Specifically to Vigfusson’s edition of Hrafn’s Saga (1878 II, 292, Chap. 11): ‘Þessu næst kómu þeir í góða höfn við ey þá er Sandey heitir, ok þar reistu kaupmenn hafnar-mark. Svá sagði Grímr: ¶ Hér hefir beitt at brattri Bótólfr skipi fljótu / (áðr fell sær um súðir) Sandeyju skæ branda: / Reisti sjálfr, ok sýsti, snarr félagi harra, hafnar-mark fyrir hrefnis, happs-verk, gota sterkan. ¶ Þeir lágu við Suðreyjar í akkeris lægi nökkurar nætr. Þá réð fyrir Suðreyjum Óláfr konungr.’ 70 Richard A.V. Cox ‘After that they came into a good harbor on an island called Sandey, and there the merchants raised a harbor mark.

‘Grím said:

Here Bótólf has brought the swift ship to the steep Sandey After the strong waves powerfully washed the sides of the ship. The brave companion of kings himself raised the harbor mark And performed a good deed for the strong ship.

‘They lay at anchor off the Sudreys for some nights. King Ólaf ruled the Sudreys at this time.’ (Tjomsland 1951, 28–29)19

Apart, however, from Olaf’s connection with Lewis – which, in Thomas’s view (loc. cit.) ‘probably included the whole Long Island or Outer Hebrides’ – there is no reason to associate Hrafn’s Saga’s Sandey with either the Barra or Uist name. Thomas’s Sandarey is ScG Sanndraigh, an island lying between Vatersay and Pabbay, south of Barra. His North Uist Sandar is a speculative plural (Norse) form, i.e. ‘[the] sands’: the name survives in Eng. Clachan Sands (NF873765), OS 1882 Clachan Sanda, and ScG Clachan Shannda, which Iain Mac an Tàilleir (Mac an Tàilleir, s.n. Clachan Sands) translates as ‘ “The churchyard of Sanda”, which is “sand river”, from Norse’. However, besides there being no river at the location, ScG [sA)u)d]20 is genitive of *Sannd, from ON Sand acc. ‘[the] sand’. Certainly, ScG Sanndraigh is not derived from ON Sand-ey ‘sand-isle’, which consists of the stem form of sandr m. ‘sand’ and ey f. ‘island’, and which yields ScG Sanndaigh off Canna in the Small Isles and Sannda off Kintyre (see below).21

19 A translation of the prose element is also found in Anderson 1922 II, 359: ‘Next, they came into good harbourage, beside an island that is called Sand-ey. And there the merchants raised a harbour mark ... They lay beside the Hebrides, at anchor, for some nights. King Olaf ruled then over the Hebrides.’ 20 Lenited in the context of ScG Clachan Shannda [kH;Axan ha)u)d] (Eairdsidh MacGilleathain, Solas); as opposed to a ScG *[sa)u)da], for ON Sand-á ‘sand-river’. Cleasby and Vigfusson (1874 s.v.) note the frequent use of Icelandic sandr (sg) and sandar (pl.), as well as compound forms, in local place-names in Iceland. On the west of North Uist, unrelated OS 1881 Sandary (NF7367) is a shieling name in ON ‑ærgi nt.; so also Horisary and Loch Dusary, to the west of Sandary. 21 The ScG reflex of ONey - ‘island’ in island names in the southern Inner Hebrides is usually ‑a -[a]. ScG Sanndraigh (Blaeu 1654 Sandrera, Monro [1549] 1774, 30 Scottish Gaelic Sannda and Its Aliases 71 Hrafn’s Saga’s Sandey, then, is likely to refer to one of these two islands: Sanndaigh Canna or Sannda Kintyre. Alan Anderson tentatively suggests Sanndaigh (1922 II, 787).22 Guðrún Helgadóttir (1987, 76) goes further, supporting the identification by noting the suitability of the island as a place of residence for a Highland chieftain (whom Hrafn and his company call upon).23 However, a similar, perhaps better, case could be made on this basis in conjunction with Kintyre’s Sannda. Nor is the saga evidence incontrovertible. At the start of their summer journey from Iceland to Norway, Hrafn and his companions spend a long time at sea, before getting a favourable wind. Eventually, they are driven southwards until they come across birds from Ireland and are finally carried to Scotland, where they lie fyrir Stauri ‘off [the] Staurr’.24 A southerly gale drives them in a sea so strong it is like nothing experienced even undan Hvarfinu á Skotlandi ‘off Cape Wrath in Scotland’. They avoid shipwreck via miraculous intervention and proceed through the Hebrides, Hrafn acting as pilot, until they arrive at Sandey. The same voyage is recounted, although in far less detail, in Bishop Gudmund’s Saga (Vigfussen

Sanderay, [1549] 2002, 324 idem) is probably from ON Sand-eyrarey ‘[the] island of the sand-(gravel)bank’, with gen. sg. of eyrr f. ‘(gravel)bank’, which is suitable topographically. 22 Anne Tjomsland (1951, 59–60) supports Captain Thomas in placing Sandey in the Outer Hebrides (cf. note 17, above). 23 Power 2005, 12, and Jesch, forthcoming, also identity Sandey with Sanndaigh Canna. 24 Which is taken to be The Point of Stoer in Assynt (Anderson loc. cit.; Tjomsland 1951, 26), although other features may have borne such a Norse name. John MacKay (1890, 121) derives Stoer from either Gaelic or Old Norse. Scots/Eng. Stoer is in fact from ScG Stòr, which in turn is a loan-name from ON Staur(in) acc. m. ‘(the) stake’ (cf. Watson 1906, 367–68; Henderson 1910, 351), in reference to the rock stack, Scots/Eng. The Old Man of Stoer; cf. a similar rock formation in Skye, ScG Bod Stòirr ‘the penis of Stòrr’, which has been sanitised as ScG Bodach Stòirr, now Bodach an Stòirr under the influence of Scots/Eng. The Old Man of Storr (Dwelly, s.n. Storr, Old Man of: ‘Bod Stòrr [source: W.J. Watson] (euphemistically Bodach Stòrr)’; note also Marwick 1923, 261: ‘Professor Watson informs me that the Old Man of Storr in Skye is a mere euphemism; the real name is Bod Storr—the phallus of Storr’). The Assynt ScG name has now acquired the article – An Stòr (Mac an Tàilleir, s.n. Stoer) – presumably under the influence of Scots/Eng. The Old Man of Stoer, which is rendered Bodach an Stòir in Gaelic, so also Rubha an Stòir for Rubha Stòir (cf. : Bodach an Stòrr, Rubha an Stòrr). The difference between ScG stòrr (Skye) and ScG stòr (Assynt), if orthographic forms represent an (historical) difference in pronunciation, arises as follows: ON staurr nom. yields ScG stòrr; ON staur acc. yields ScG stòr (cf. ScG sgòd < ON skaut nt. (Cox 2002, 297)). Formally, gen. sg. of ScG stòrr is stòirr, although this falls together in sound with stòrr through assimilation of non-palatalised and palatalised unlenited rr, hence written forms in (gen. sg.) stòrr; the gen. sg. of ScG stòr is stòir; furthermore, ScG written forms may be confused with Scot/Eng. forms in storr. 72 Richard A.V. Cox 1905, 613): a north-westerly wind drives the company southwards to Hirtir (St Kilda), after which they continue south by Ireland, sailing south of Ireland until they come up against stormy weather. Having made appropriate vows, the storm abates and they are able to sail directly to Norway. From the detail of these two accounts, it would appear that neither Sanndaigh Canna nor Sannda Kintyre can be ruled out entirely. Kintyre’s Sannda appears in Latin in texts at the beginning and end of the 17th century: in Retours, (Argyll 21) 1619 20 solidatis terrarum insulæ de Sanda and (Argyll 93) 1695 20 solidatas terrarum insulæ de Sanda; and in a number of documents relating to the Franciscan mission to Scotland in the early part of the century: in Edmund MacCana’s description of the island (see Appendix) c.1644–51 (gen.) Insulæ Sandæ, (nom.) Insula Sanda, Sanda; in a letter from Patrick Hegarty to Hugh de Burgo (Giblin 1964, 33) 1624 in Sandam insulam; in a report by Cornelius Ward (ibid., 50) 1625, or soon after, in Sandam insulam; and in a report by Cornelius Ward (ibid., 149) post 1631 Sanda insula. In addition to Blaeu 1654 and Gordon c.1636–52 (see above), map forms include Moll 1745 Sanda I., Roy 1747–55 Sanda Id, van Keulen c.1780 Sana I., Thomson 1820 Island of Sanda and OS 1869 Sanda Island. Scots / Eng. Sanda25 and Sanda Island 26 are, of course, from ScG Sannda, Kintyre Gaelic /sada/ (Holmer 1962: 17, recorded 1937/38),27 a loan-name from ON Sand-ey (see above). The Norse name presumably refers to the nature of the north-facing bay, opposite Kintyre, which is sandy.28 Hamish Haswell- Smith’s (2004, 5) suggested derivation from ‘ON sandtange [leg. sand-tangi] “sandspit”’ is unlikely from a topographical point of view and is formally impossible from a phonological one. The notion that the nameSannda goes back to Adomnán §2 begins with the firstStatistical Account. It is unclear, however, since the name in any form is not mentioned in Vita Columbae, how the idea developed.29 Campbell (1924, 10–

25 E.g. Fraser-MacIntosh 1895, 36: ‘The island of Sanda, of old Avon ...’. 26 E.g. Holmer 1957, 49. 27 (Lenited) genitive in Murchison 1960, 64: ‘chaidh a’ cheud “Chlansman” air tìr air eilean Shannda’ (‘the first Clansman went aground on the island of Sannda’). Less appropriately, Skene (1871–72 II, 39) writes Sanday and the website gives Gaelic Sandaigh (for the final syllable, see note 21, above). ON -nd is treated in Scottish Gaelic as an (originally geminate) unlenited non-palatalised // – a velarised dental nasal – plus stop. The preceding stressed vowel is lengthened (diphthongised) in this environment in most Scottish Gaelic dialects, e.g. Sanndaig [sa)u)tIkJ] (Lewis; Cox 2002 s.n.) (O’Rahilly 1976, 49–52; Ó Baoill 1990, 131; Cox 2000, 213–16), the exceptions being the dialects of Argyllshire, Arran, East Perthshire and parts of Aberdeenshire (Grannd 2000, 53–54, 138; SGDS, e.g. items 163, 196, 329). Scottish Gaelic Sannda and Its Aliases 73 11) surmises that the connection was made through a misidentification with Adomnán’s Sainea;30 this seems plausible. Additionally, a conflation of John of Fordun’s note on the dedication on the island – cella Sancti Adamnani (Goodall 1759 I, 45; Edinburgh College Library MS (which Reeves 1857, 87 quotes)); capella Sancti Anniani (Hearne 1722, 81; Trinity College, Cambridge MS)31 – with the island’s name (although it is not (Lat.) Sanda but Awyn, Awyne, respectively, that John mentions) may have aided the misidentification.32

§4 An Spàin Campbell (1924, 11) comments that ScG An Spain [leg. An Spàin, with ScG spàin ‘spoon’] or Scots / Eng. The Spoon ‘is known to the people on the south of Arran’ – from where the eastern end of Sannda resembles the bowl and the western end the handle of an upturned spoon – while Duncan Colville (KAS 1938, 29 and 31) notes that An Spain [sic] or Spoon Island is ‘[s]aid to be used by mariners for Sanda Island’, for ‘Sanda resembles a spoon when seen from Pladda or Ireland’ (p. 31).33 Whether the Gaelic or English form here came first is perhaps open to question.

§5 Àbhainn While Avona Porticosa §2 is a ghost name,34 Buchanan’s Avona is a latinised form of MacCana’s c.1644–51 Abhuinn (MS Abhuin, see Appendix), modern 28 Martin 2009, 31: ‘It ought to be Old Scandinavian Sandey, sandy island, but it is hard to see why such a name would be bestowed on a rocky island, unless it could relate to the sandy seabed at the approaches to the harbour on the north side of the island.’ For an aerial view, see Google Earth (): Sanda, UK. 29 At any rate, the island is probably unlikely to have borne a Norse name during the seventh century. 30 Which has otherwise not been identified with certainty. Anderson and Anderson (1991, lxxiv) suggest possibly Colonsay; Dr Reeves and Wentworth Huyshe suggest Shuna (cited in Campbell loc. cit.). 31 capella sancti Sanniani (Skene 1871–72 I, 43; Wolfenbüttel MS). 32 Skene (1871–72 II, 386) concludes the dedication is to Senchan; the editors of Origines Parochiales Scotiae (II, 9 and note 12) to Ninian. Edmund MacCana (c.1644–51) is clear that the chapel was dedicated to St Ninnian and that the sons of a most holy man, Senchan, were buried there (see Appendix). 33 Pladda lies just off the south-east coast of Arran. (There is no connection between the name An Spàin and Uri Geller, famous for his spoon bending, who early in 2009 bought The Lamb, an uninhabited lump of volcanic rock in the Firth of Forth ( – accessed 28/10/09).) 34 Parallel to the term ghost word – a word that has entered the language through the perpetuation, in dictionaries etc., of an error (Collins) – the term ghost name refers to a name that has entered the nomenclature through the perpetuation, in written or oral sources, of an error. 74 Richard A.V. Cox ScG Àbhainn (e.g. Dwelly, s.n. Sanda Id.); cf. Kintyre Gaelic dol do dh’Ābhuin ‘going to Sanda’ (Henderson 1910, 180)35 and Arran Gaelic tha tigh solus ann an Àbhuinn /ha tej sols a  Evin′/36 ‘there is a lighthouse on Sanda Island’ (Holmer 1957, 49), the pronunciation of which is reflected in W. J. Watson’s orthographic representations (1926, 91) of Arran Gaelic, viz Ēibhinn and Eubhainn.37 Without offering any grounds, Watson (1926, 91) states that ‘[t]he pre- Norse name of Sanda is preserved in Gaelic still as Àbhainn’. As for Àbhainn being pre-Norse, nothing places it for certain earlier than John of Fordun (1384 × 1387, §2), although there are no internal (i.e. linguistic) dating criteria that prevent it being older. Watson does not state explicitly whether he thought Àbhainn was Gaelic in origin or not.

§6 Adomnán’s Ommon Skene’s (1874, 328) tentative suggestion38 that John of Fordun’s Awyn (supposing an original Old Gaelic lenited -b- ‑[B]‑) is connected with Adomnán’s Ommon (with geminate -mm-) (Reeves 1857, lib. i, cap. 36, p. 70; Skene 1874, lib .i, cap. 29, p. 136) is disposed of correctly by Watson (1926, 91) on phonological grounds.39

§7 EG abhuinn, ScG abhainn ‘river’ It has been suggested that Àbhainn is simply ScG abhainn f. ‘river’.40 Campbell (1924, 11: Abhuinn [sic]) implies that this was a local folk etymology for the name: ‘the explanation rendered is the geographic fact that, as it is expressed by Dean Monro’ [rather by the anonymous author of ‘Ane Description of Certaine Pairts of the Highlands of Scotland’ (note 3, above)], ‘[t]he streame runns so swiftlie that no shipps can remaine near it, except they be within the harborie’ (Mitchell 1907–08 II, 187). The derivation is ruled out topographically,

35 KAS (1938, 16 and 29) gives Eilean Abhainn, noting (p. 2) that ‘grave and acute accents in a number of Gaelic words are omitted in printing’ – in fact, lengthmarks appear to have been omitted from the printing process entirely in this publication. 36 In which solus (solas) fails to show genitive inflexion. 37 Mac an Tàilleir (2003, 103) writes ‘Àbhainn or Eabhainn [leg. Èabhainn]’. Martin (2009, 8) quotes Watson’s forms without lengthmarks. The latter name form is similarly misspelt Eabhainn on the Tìr Chaluim Chille (2003) map. 38 A suggestion quoted by Fowler (1894, 47 note 4) in his edition of Columba’s life. 39 Ommon remains unidentified (Sharpe 1995, 298 note 160; Anderson and Anderson 1991, lxxv). 40 E.g. Haswell-Smith (2004, 5); further, ‘[the name] could also be a reference to the Scottish Gaelic “abhainn” which means a river’ (). Scottish Gaelic Sannda and Its Aliases 75 however, because (a) there are no rivers on the island and (b) because the Gaelic term for the stream (i.e. race or current) of the description is sruth41 – off the Mull of Kintyre, specifically Sruth na Maoile.42 Formally, the derivation would also be ruled out phonologically because of the mismatch between the long stressed vowel of the place-name, on the one hand, and the short vowel of the appellative, on the other, but see the discussion on vowel length below §10.

§8 ScG àmhainn ‘oven’ Haswell-Smith (2004, 8) notes that ‘[t]he central valley [in the island] traps the heat as it is sheltered from all the prevailing winds.’ The valley is now

41 Although this has not prevented the supposed lexical sense of the name of the island being extended to ‘current’ in the case of ‘Abhuinn (le courant)’ (). 42 Literally ‘the stream of A’ Mhaol (‘the mull’)’ (cf. The Sea of Moyle). Specifically between Kintyre and Sanda, this is ‘Tiompanach, The. The name given to the swift current in the Sound of Sanda’ (KAS 1938, 32), ScG An Tiompanach (Dòmhnall Iain MacAonghais, Scarp: ‘an sruth aig ceann an iar an eilein’ (‘the current at the western end of the island’, pers. comm.). Tiompan is used occasionally of hills, e.g. Tiompan [tJH)u)m8pan] of a headland or bluff (Lewis; Cox 2002, 381), Maoil an Tiompain (Loch Broom; Watson 1976, 247), Knocktimpen (Dumfries and Galloway; Watson 1924, 144), Màm an tiompain (Gillies 1906, 218) and possibly Trumpan (Skye; Forbes 1923, 435), and derives, via OG timpán ‘timbrel, drum; some kind of stringed instrument’, from Lat. tympanum (DIL). John Purser (pers. comm.) assures me that tiompan refers historically only to a stringed instrument in Scottish Gaelic tradition (see also Purser 2007, 35). In place-names, however, the sense of hill may have arisen via biblical usage: while the Old Testament was translated into Gaelic – published between 1783–1801 – from Hebrew (which uses Hebrew toph ‘timbrel, tambourine’), Gaelic tiompan acc. sg. (Job 21:12; Psalms 81:2), tiompan dat. sg. (Psalms 149:3, 150:4) and tiompanaibh dat. pl. (Exodus 15:20, Judges 11.34, 2 Samuel 6:5, 1 Chron 13:8) are used consistently for the Latin Vulgate’s tympanum acc. sg. (Job 21:12; Psalms 81:2), tympano dat. sg. (Psalms 149:3, 150:4) and tympanis dat. pl. (Exodus 15:20, Judges 11.34, 2 Samuel 6:5, 1 Chron 13:8), all of which are rendered timbrel in the King James Version. Cormac’s 10th-century definition, or

rather etymology, of timpán reads, ‘Timpān .i. tim (.i. bocc [added from another MS]) .i. sail 7 bān .i. umæ bīs inti. Uel quasi simpān a simphonia .i. ōn bindius.’ (Meyer 1994, 109 §1258) (‘Timpán, i.e. pliant (i.e. soft), i.e. [?like] willow + [?]bright, i.e. it is made of copper. Or like simpán from simphonia, i.e. from the sweetness of sound.’). Whether *An Tiompanach (< tiompan + suffix of place (Cox 2002, 60)) is a transferred name and once referred to a headland or nearby hill is unknown; alternatively, it may have referred to eddies or whirlpools in the current, in the same way that ScG coire ‘cauldron’ has been used in a similar sense elsewhere, e.g. Coire Bhreacain (Cox 1998, 26–28). For the semantic extension of tiompan, cf. Eolas an Spealaidh / Cum bogha air do chorp / ’S cum a mach do thiumpan. / Tiumpan = the posterior = Deireadh. [‘Scythe lore: keep your body bowed (arched) and stick your bottom out. Tiumpan = posterior = rear.’] (Alexander Carmichael’s fieldwork notes (Stiùbhart 2009, 142)). 76 Richard A.V. Cox called Lag nan Gàidheal ‘the hollow of the Gaels’ (OS 1869 Lag nan Gael; KAS 24, idem), but we might speculate that it was once called ‘the oven’. It is not known what the Southend Gaelic form of the Gaelic word for ‘oven’ was, but were it anything like Jura Gaelic òbhan [ovn] (with a non-nasal stressed vowel (George Jones, pers. comm.)), as opposed to, for example, Lewis Gaelic àmhainn [a)vIJ]), it might conceivably have developed into Àbhainn. However, such a name would be relatively recent and would be more likely to occur along with its article. For these reasons, ScG àmhainn, or similar, may be excluded from further consideration.

§9 ON hófr ‘hoof’ or ON háfr ‘bag-net’ Àbhainn might derive from an Old Norse form with a suffixed article, e.g. ON Hófinn acc. ‘the hoof’, with hófr m.,43 or ON Háfinn acc. ‘the bag-net’, with háfr m.,44 either of which might have been applied, on account of its shape, to the central valley, or perhaps even the bay itself. With loss of initial h-, and with or without o ~ a alternation, either form might yield Àbhainn.

§10 ON hôfn Monro implicitly and Macculloch explicitly, however, take the name to be from ‘Danish havin /hafn’ (i.e. ON hôfn) ‘haven, harbour’.45 Indeed, one of the most significant physical features concerning the island is the bay, which offers protection from both south-westerlies and strong tidal currents (Sruth na Maoile §7, and An Tiompanach note 42), affording safe anchorage and a beach where boats can be dragged clear of the water. ON hôfn f. ‘haven’ is, therefore, a natural candidate for a derivation of Àbhainn.46 ON hôfn occurs in several place-names in the Hebrides, e.g. Tamnabhagh (Lewis), Tamhnaraigh (Lewis and the Summer Isles), Na Hamhn (Eriskay and Mull) and Port na h-Abhainn(e)) (earlier Port na h‑Abhann; Islay) (Cox 2008). As in Port na h‑Abhainn(e), though, we should expect a short stressed vowel here, which would nominally rule out hôfn’s candidacy.47

43 Cf. Mod. Norwegian Hoven and Hestehoven (Møre og Romsdal; id. 269358 and 261897, respectively (). 44 A word borrowed into Scottish Gaelic (Oftedal 1956, 106). 45 So KAS (1938, 19): ‘Havin. Name given to Sanda by the Scandinavians.’ 46 In spite of the evidence from John of Fordun to Groome §2, there is no justification for saying that ‘The Norse sometimes referred to Sanda as ‘Havn’ because it provided a reasonable offshore haven or harbour for boats’ (Haswell-Smith 2003, 3). 47 The diphthongisation inTamhnaraigh and Na Hamhn is due to the vocalisation of the Early Gaelic nasal fricative mh [B)] < ON f [B] (ibid., 54–55). Scottish Gaelic Sannda and Its Aliases 77 Lengthening of short stressed vowels, however, has been recorded in the Gaelic dialect of Kintyre. Neil MacDougall of Carradale was recorded by Derick Thomson in 1951 for the Linguistic Survey of Scotland (for whose Gaelic results, see the Survey of the Gaelic Dialects of Scotland (SGDS)). The resulting tapes were transcribed by Anthony Dilworth – point 36 throughout – who marks a degree of lengthening in some originally (and, elsewhere, normally) short stressed vowels: half-long, with several examples before -r, in arbhar (SGDS item 54), bradan (120), carbad (153), farsuing (399) and marbh (602); long in fasdaidh (400); but short in aran (53) and short before the nasal labio-dental fricative in amharc (3648) and sgamhan (758). Due to a misunderstanding – as Cathair Ó Dochartaigh, SGDS editor, explains (ibid. I, 85) – MacDougall was also interviewed using the Survey’s questionnaire by Fred MacAulay in 1954. Ó Dochartaigh remarks that there are a number of ‘striking differences’ between Dilworth’s and MacAulay’s transcriptions. Some of these, he continues, ‘may be due to the quality of the tape recording used for point 36 [which was rather poor], others to the fact that [Dilworth] had been working on dialects of the north-centre mainland and was not familiar with Kintyre Gaelic.’ For all of the above items (although fasdaidh was not recorded), MacAulay – point 37 throughout – gives short vowels. Anthony Dilworth, however, also transcribed tapes for the Survey – point 38 throughout – of John Taylor of Muasdale, recorded by Derick Thomson that same year. Of the nine items above, four show greater lengthening in Muasdale, one greater lengthening in Carradale (point 36) (Fig. 1).

Stressed-vowel lengthening in Carradale and Muasdale, Kintyre, in SGDS data Fig. 1 SGDS item Stressed-vowel quantity in Carradale (pt 36) and Muasdale (pt 38) short half-long long 2 abhainn 36 38 36 amharc 36 38 38 53 aran 36 38 54 arbhar 36 38 120 bradan 38 36 153 carbad 36 38 399 farsuing 36 38 400 fasdaidh 36 38 602 marbh 36 38 758 sgamhan 36 38

48 Point 36 has a bilabial rathar than a labio-dental fricative. 78 Richard A.V. Cox The same phenomenon is recorded in the Linguistic Atlas of the Survey of Irish Dialects (LASID), for which Colm Ó Baoill undertook fieldwork in 1961; indeed, the following examples of lengthening of short stressed vowels (Fig. 2) derive from his interview with the same Neil MacDougall that had been interviewed for SGDS in 1951 and 1954. The list is not intended to be exhaustive but illustrative of the fact that there is a tendency to lengthening, at least with this speaker.

Stressed-vowel lengthening in Carradale, Kintyre, in LASID data Fig. 2 LASID item Stressed-vowel quantity in LASID IV, 212–17 short half-long long 1 tarbh Carradale Carradale 2 reic Carradale 2, 3 crodh Carradale Carradale 22 bainne Carradale 28 tabhann Carradale 30 mart Carradale Carradale 30 agad Carradale 39 feasgar Carradale 42 cartadh Carradale 44 uisge Carradale Carradale 125 gairm Carradale 138 coileach Carradale 139 briste Carradale 178 marcachd Carradale 205, 206, 208 cat Carradale Carradale 209 craiceann Carradale 212 losgadh Carradale 220 a-nis Carradale 239 boireannach Carradale 242 cleachdte Carradale 246, 248 teanga Carradale 265 deasachadh Carradale 339 litir Carradale 340 goirid Carradale

The same phenomenon occurs with Ó Baoill’s informant from Muasdale, Neil Thomson (Fig. 3).49

49 Some similar traits are found in the responses of Neil Thomson’s cousin, George Thomson, also interviewed by Colm Ó Baoill for the Linguistic Survey, e.g. short and half-long 558, 559, 560 seachad; half-long 544 cidsin; half-long and long 576, 577, 578 losgadh; long 576 ochd. Scottish Gaelic Sannda and Its Aliases 79 Stressed-vowel lengthening in Muasdale, Kintyre, in LASID data Fig. 3 LASID item Stressed-vowel quantity in LASID IV, 212–17 short half-long long 608, 609, 629 achadh Muasdale Muasdale 645 coisich Muasdale 620, 655, 663 coirc Muasdale Muasdale 656 seagal Muasdale 701 marbh Muasdale 710, 934 math Muasdale Muasdale 710 obair Muasdale 718 tarann Muasdale 732, 733, 734, 1013 dhachaigh Muasdale Muasdale Muasdale 742 agam Muasdale 750 botal Muasdale 751, 752 deoch Muasdale 757, 773 bochd Muasdale Muasdale 791 seirbheis Muasdale 847 uisge Muasdale 858 fasgadh Muasdale 866, 867 fliuch Muasdale Muasdale 880 marbhadh Muasdale 872, 976, 1043 goirid Muasdale Muasdale 977 loch Muasdale

Yet, apparently, no trace of such a phenomenon appears in Nils Holmer’s study of Kintyre Gaelic, for which he undertook fieldwork in 1937 (Holmer 1957, vii), which included interviews with both Neil MacDougall (Ca 1) (Holmer 1962, 4), and Neil Thomson (La 12) (ibid., 3) – unless it be in his note 2 on p. 36 (ibid.): having discussed the glottal stop between vowels in Kintyre Gaelic, Holmer turns to the question of syllabification and syllable limits, adding, ‘Some people even make a short stop before the syllable-ending consonant, as teine /tSe>n′/ ‘fire’ ..., coileach /ke>lax/ ‘rooster’ ..., gealach /g′a>ax/ ‘moon’ ..., sileadh /Si>l/ ‘drizzling’ ...’ (The symbol >/ / in Homer’s transcription here signifies a slight narrowing, but not occlusion, of the vocal chords, resulting in ‘a reduction in intensity of a preceding vowel’ (ibid., 35), rather than a half-long vowel, as it does in this article, in SGDS (I, 113) and LASID (I, xxiii).) We may speculate that what was a sporadic, weak glottal stop after stressed vowels in the more southerly dialect of Kintyre developed during the middle of the 20th century into (sporadic) lengthening of the stressed vowel. From the evidence above, it appears that the development could occur in open syllables (crodh, math), before voiceless (cat, litir) and voiced stops (obair, 80 Richard A.V. Cox bradan, seagal), before -s and s-clusters (deasachadh, uisge), at least some l- and n-phonemes (coileach, bainne), and particularly before -r and r-clusters (tarann, tarbh); there is also evidence of lengthening before the voiceless fricative -ch (deoch), as well as voiced fricatives (abhainn, amharc).50 According to the SGDS data (Fig. 4), lengthening of short stressed vowels is mainly restricted to southern Kintyre; outwith southern Kintyre, lengthening occurs only in one instance: amharc in Arran. Glottal stops, which occur only in short reflexes of stressed vowels, not in half-long or long varieties, are found mostly outwith southern Kintyre: 25% in northern Kintyre (nine out of 36 examples) and 10.77% in Arran (seven out of 65). Of the four examples in southern Kintyre (which represent 8% of the 50 examples), two are from point 37, Carradale, and two from point 39, Killean and Kilmory (about two and a half miles north of Muasdale), all recorded by MacAulay.

Glottalisation & lengthening of stressed vowels in Kintyre & Arran in SGDS data Fig. 4 SGDS item pts 31–35 (Arran), 36–39 (s. Kintyre), 40–42 (Gigha + n. Kintyre) half- short long long 2 abhainn 31 32 33 34 35 37 39 40 41 42 36, 38 36 amharc 32 34 36 37 39 40 41 42 38 31, 33, 35 53 aran 31 32 33 34 35 36 39 41 38 54 arbhar 31 32 33 34 35 37 39 40 41 42 36, 38 120 bradan 31 32 33 34 35 37 38 39 40 41 42 36 153 carbad 31 32 33 34 35 37 39 40 41 42 36 38 399 farsuing 31 32 33 34 35 37 39 40 41 42 36, 38 400 fasdaidh 31 32 33 34 35 39 40 41 36, 38 601 marbh 31 32 33 34 35 37 39 40 41 42 36 38 690 radan 31 32 33 34 35 37 39 40 41 42 36, 38 726 sabhal† 31 32 33 34 35 37 39 40 41 42 36, 38 727 sagart 31 32 33 34 35 36 37 38 39 40 41 42 38‡ 758 sgamhan 31 32 33 34 35 36 37 39 40 41 42 38 † (-bh-: 36 [/], 38 [w], 42 zero, elsewhere [v]) ‡ examples occur with both short and half-long stressed vowels Key: Bold = stressed vowel with glottal stop. Points where absent were not recorded. 50 For more on the incidence of glottal stops in Argyllshire, particularly Jura, see Jones 2000. Grannd’s (2000) comparative study of Islay Gaelic with Argyllshire and Arran Gaelic is restricted largely to lexical variation and word geography and provides no evidence here. Scottish Gaelic Sannda and Its Aliases 81 The evidence, then, suggests that the incidence of glottal stops before consonants and pausa and lengthening of stressed vowels are in more or less complementary distribution within the area as a whole. Lengthening may initially have been a development from a stressed vowel + glottal stop; in general, however, it may have been a parallel, but alternative development more or less restricted to southern Kintyre. (Indeed, a stressed vowel following a weak glottal stop may have occasionally been perceived as a half-long or long vowel by fieldworkers.51) While this feature of lengthening may have had its focus within southern Kintyre, it would be wrong to restrict the area of lengthening entirely to southern Kintyre. Ó Baoill’s informant from Tarbert in northern Kintyre, Mary MacKinnon, also shows signs of the same development, e.g. half-long 48 achadh, 63, 66 goirt, 87 olann; short and long 89 lomamh (for lomadh); and short, half-long and long 61, 66, 68, 73 bainne (LASID IV, 212). On the other hand, she appears to use glottals stops more often than Carradale or Muasdale informants, e.g. 54, 55, 56 bleoghann, 72 soitheach, although it is difficult to make comparisons between one informant and another due to differences in the substance and volume of data-sets. Similarly, LASID data

51 Yet the question of the differences in vowel length in Neil MacDougall’s speech as recorded by Anthony Dilworth and Fred MacAulay remains unresolved: Dilworth’s lengthened vowels are not matched by glottal stops in MacAulay’s transcriptions. Dilworth’s examples of lengthening are supported independently by Colm Ó Baoill’s transcriptions, which begs the question whether MacAulay may have unconsciously normalised apparently irregular vowel lengths in his own material. Fred MacAulay (1925–2003), from Solas, North Uist, took a diploma in phonetics at Edinburgh before working on the Linguistic Survey between 1951–54, taking up a full-time post as fieldworker in 1952. Anthony Dilworth had already heard Kintyre Gaelic, among many other dialects, as a student, before joining the Linguistic Survey as a fieldworker in 1954, with which he continued for three years. Of the transcription process, he recalls the excellent sound-editing facility, with its capacity to isolate and/or compare words or sequences of words, provided by Tony Anthony in the Department of Phonetics at Edinburgh: ‘Bha an t-inneal-èisteachd a rinn Tony Anthony ann an Roinn nam Foneataigs fìor mhath dha-rìreadh. Dh’fhaodadh tu an aon fhacal no sreath a chluinntinn uair is uair agus bha sin a’ toirt cothrom dhut na fuaimean a sgrìobhadh ann am foneataigs air do shocair gu math ceart. Cha bhiodh tu an urrainn a bhith ag iarraidh air seann sluagh faclan a’ cheisteachain a ràdh ach a dhà no trì tursan air neo dh’fhàsadh iad searbh den chùis oir bha tòrr cheistean ann. Bha e comasach cuideachd an aon fhacal o dheifir àiteachan a chur air lùib teip agus coimeas a dhèanamh eatarra.’ (pers. comm.). Although Heinrich Wagner comments in the introduction to LASID IV (p. viii) that Ó Baoill, ‘having gained some further experience in the field of Scottish Gaelic phonetics, ... is now not pleased with some of his transcriptions which were partly made from tape ...’, the general picture of lengthening in stressed vowels remains. 82 Richard A.V. Cox for Arran (also recorded by Colm Ó Baoill) shows a tendency, albeit a less pronounced one, for lengthening short stressed vowels (Fig. 5).

Sporadic stressed-vowel lengthening in Arran in LASID data Fig. 5 LASID item Stressed vowel quantity in LASID IV, 212–17 short half-long long 1 tarbh Arran Arran 2 reic Arran 22 bainne Arran 28 tabhann Arran 30 agad Arran 44 uisge Arran 45 thàinig Arran 125 gairm Arran 138 coileach Arran 139 briste Arran 178 marcachd Arran 205, 206, 208 cat Arran 209 craiceann Arran 242 cleachdte Arran 339 litir Arran

However, unless Holmer simply failed to record examples of lengthened short vowels (cf. his Rathlin abhainn [o>in] [o>n] (with hiatus, Holmer 1942, 156), as opposed to LASID’s (I, 280) lengthened form, [on] (point 67)), the negative evidence of his description of Kintyre Gaelic would suggest that the development of sporadic lengthening of short-stressed vowels in southern Kintyre may post-date George Henderson’s Ābhuin (1910) and Watson’s Ēibhinn and Eubhainn (1926). If so, we should look for an alternative cause for the lengthening of the stressed vowel in Àbhainn. There may be an alternative. Like most folk etymologies, ours may contain an element of truth, and the conflicting claims of a connection with both ON hôfn and ScG abhainn may indeed have some basis in fact. On the one hand, the relative importance of Sannda’s harbour within the area is highlighted consistently in early descriptions by reference to the Norse use of the harbour and their naming it Hôfn. This scenario is supported by the topography of the island’s anchorage and the nature of Sruth na Maoile; by the fact that the island is a stepping-stone between Ireland and Scotland, Scottish Gaelic Sannda and Its Aliases 83 and between Arran and Gigha and Islay;52 and by the frequency with which the Norse element hôfn occurs in Norse loan-names throughout the Suðreyjar. From this perspective, ON hôfn is the perfect candidate for the derivation of Àbhainn. On the other hand, the place-name is traditionally said to be synonymous with the Scottish Gaelic term abhainn ‘river’. We know little of the Gaelic dialect of Southend, i.e. of the Mull of Kintyre and Sannda. Holmer remarks that in his time – 1937 – the dialect ‘hardly survives’; ‘it was different from the north Kintyre dialect, and more like Irish’ (Holmer 1962, 1–2), but the little we do know appears to add nothing to the discussion on vowel length (ibid., 2 and 106–07). Yet, we can be certain that, as part of a linguistic continuum,53 the dialect of Southend would have held affinities with both Ireland and Scotland. In an East Ulster Irish context, lengthening of the stressed vowel in abhainn ‘river’ occurs through vocalisation of the intervocalic fricative.54 Thus LASID (I, 280) records Rathlin [on] (point 67), Omeath [oJ] (point 65) and Inishowen [oJ] (point 68). Although this is in contrast to Holmer’s Rathlin [o>in] [o>n] (1942, 156, in which the superscript dot (>) indicates hiatus55), it is possible that Southend Gaelic abhainn was pronounced something like *[a>I] or *[aI] with a half-long or long stressed vowel after vocalisation of the fricative,56 and this is supported by the Ulster form of the name, [iniS o>in],57 with a half-long stressed vowel. On this count, it is conceivable that ScG Àbhainn derives from ON Hôfn ­ /hBn/, via Early Gaelic *Abhan */aBn/, dat. *Abhain */aBin′/, with a short 52 The island is described as ‘within dangerously easy reach of Clann Dòmhnaill [Islay] and their Mac Aoidh allies [Kintyre and Bute]’ (Forte 2008, 211). 53 See, for example, Ó Baoill 1978 and 2000. 54 This phenomenon is unconnected to the lengthening of stressed vowels in Kintyre and its environs discussed above. 55 ‘In Rathlin the hiatus is not so clearly marked as in Scotland: cumhang “narrow” may rime with uan “lamb” ’ (loc. cit.). 56 In Mid-Argyllshire, the fricative has generally been replaced by hiatus, in Islay generally by a glottal stop (see SGDS II, 3), and it is curious that Holmer (1938, 116) records [nj] in Islay. 57 Brian McLaughlin, whose family was from Carnlough (in sight of Sannda), eight miles north of Larne, as told to him by his father, who reported that his own grandmother (Margaret McLaughlin, c.1840–1933, from Slanesallagh, four miles from Carnlough) had told him that this was Sannda’s proper name. Margaret’s mother, Catherine’s, family seems to have been engaged in transporting salt from Carrickfergus to Scotland. Catherine may have been a Gaelic speaker: she was born close to the Glens of Antrim and her father, Para Bán ‘fair Para’, was nicknamed Brochan Scallta (‘paltry porridge’), apparently in reference to his employment during the Great Famine (1845–49) in the distribution of what was considered to be poor quality Indian meal. (Brian McLaughlin, pers. comm.) 84 Richard A.V. Cox stressed vowel (cf. John of Fordun’s late 14th-century Scots form, Awyn). Later the bilabial fricative develops into a labio-dental fricative in Scottish Gaelic, viz *Abhan */avn/, dat. *Abhain */avin′/ (cf. Monro 1549 Avoyn, Mac Cana 1621–22 Abhuinn, SAS 1791–99 Aven) but is vocalised in Southend Gaelic, as in the case of intervocalic -bh- in abhainn in East Ulster Irish, with compensatory lengthening of the stressed vowel yielding *Abhan */an/, dat. *Abhain */ain′/. We may assume that this dative form was homophonous, or nearly so, with the local pronunciation of abhainn ‘river’, which thus gave rise to the folk etymology that the name meant ‘river’ – although the sense had to be altered to ‘stream’ and the currents of Sruth na Maoile, in order for the etymology to be credible. If this is so, neighbouring communities (e.g. Arran /Evin′/ (Holmer 1957, 49; §5, above)58) appear to have adopted the long stressed vowel of */ain′/, while preserving the labio-dental fricative of earlier */avin′/. Meanwhile, radical *Abhan may survive in George Campbell Hay’s Tarbert form,59 Àbhann (rad. an latha thogas Àbhann ‘when we set course on Sanda’, Byrne 2000 I, 424 (also 277, 316), dat. ’s i ’seòladh mach o Àbhann ‘as out she sails from Sanda’, 425), whose development was presumably on the analogy of the development of *Abhain > *Àbhainn.60

§11 Conclusion Of the various possible derivations of Àbhainn discussed above, ON Hôfn seems contextually to be the most likely. The historical phonology for the development of Àbhainn from ON Hôfn, however, is not entirely clear. Our knowledge of the Gaelic dialect of Southend in very limited. There is evidence that some lengthening of short stressed vowels took place in southern Kintyre and there is evidence to suggest that lengthening was in more or less complementary distribution with the development of a glottal stop in similar environments in neighbouring areas, although the sporadic nature of the lengthening may suggest that the development was one of free variation. On the face of it, the lack of comment on lengthening of short stressed vowels in Holmer’s description of Kintyre Gaelic suggests that the phenomenon was a developing one during the 20th century, one which may be too late to account

58 The reflexE / / may occur for /A/ in Arran and, to a lesser extent, in Kintyre; see Holmer 1957, 49, and 1962, 41. 59 During his formative years, Hay learnt Gaelic from his maternal great-aunts in Tarbert, Kintyre, and from local fishermen (Martin 1984, 48–71; Byrne 2000 II, 3–6). 60 EG abann, a form derived from ab as a fem. n-stem (DIL, s.v.; Vendryes, s.v.), dat. abainn, so modern nominativised abhainn. Scottish Gaelic Sannda and Its Aliases 85 for lengthening in Àbhainn. Alternatively, lengthening of *Abhainn to Àbhainn may have paralleled the development of the appellative abhainn ‘river’ > *à’ainn in the local Gaelic dialect, assuming that this was the same as that which took place in Ulster, cf. Rathlin [on] ‘river’. If so, while neighbouring dialects, by virtue of the presence of a long stressed vowel, have in effect re-borrowed the name, they have preserved earlier intervocalic -bh-, perhaps on account of the long-standing folk etymological connection with abhainn ‘river’. It remains to be seen, however, whether further evidence will come to light which can support one or other of these hypotheses. It is possible, then, although not certain, that we here have two different Norse names: one for the island itself, ON Sandey (yielding ScG Sannda), and one for the harbour, ON Hôfn (yielding ScG Àbhainn).61 One can suppose that in general the latter might have held greater significance for shipping as, in times of need, boats would have set course for the haven, Àbhainn, rather than the island, Sannda, and this may help explain why Àbhainn, as opposed to Sannda, survives in Arran Gaelic. While the onomastic meaning of Àbhainn transfers to the island in Gaelic usage – ousting use of Sannda in some communities, acquiring equivalent sense in others – it is the original island name that is borrowed into Scots / English, hence Sanda, a fact which presumably reflects administrative usage. The shape of the island as seen from the sea gives rise to a third form, An Spàin. If ScG Àbhainn ever had a hold in modern Scots / English usage, it has been eclipsed by the English form of An Spàin, Spoon Island.62 In the early part of the 20th century, it is probable that Sannda, Àbhainn and An Spàin belonged to different user groups (Cox 1990, 46–48): Sannda to the islanders themselves and to the people of Kintyre – Eng. Sanda to officialdom; Àbhainn to the fishing communities of Kintyre and Arran; and An Spàin or Spoon Island to the people of south Arran and to a principally non-fishing, boating community.

Phonetic note [ ] enclose phonetic transcriptions, whose symbols represent actual pro- nunciation. / / enclose phonemic transcriptions, whose symbols represent contrastive units of sound within the dialect or language concerned. ScG [d] is devoiced; [d t ;  J] are dentals; [; ] are velarised; both [J] and / ′/ indicate palatal consonants. 61 There are parallels, albeit that the islands concerned are larger than Sannda: Na Hamhn on Eriskay (ON Eiríksey ‘Eirík’s isle’), Na Hamhn and Na Hamhnan on Mull (ON Mýl ), as well as Port na h‑Abhann (< *Abhann) on Islay (ON Íl ), all from ON Hôfn (Cox 2008, 51–55). 62 E.g. . 86 Richard A.V. Cox Acknowledgements My sincere thanks to Thomas E. V. Pearce for checking my translations of George Buchanan and Edmund MacCana and for his helpful comments on them; Angus Martin and Tom Schmidt for references; George Jones, Dòmhnall Iain MacAonghais, Eairdsidh MacGilleathain and Brian McLaughlin for forms; Cairistìona Cain, Librarian at Sabhal Mòr Ostaig, for help in tracking down sources; Bibliothèque royale de Belgique for permission to reproduce Edmund MacCana’s MS (Plates 1–2); John Purser for his discussion of tiompan; Anthony Dilworth for his reminiscence on his work at the Linguistic Survey of Scotland; and to Colm Ó Baoill, Mícheál Ó Mainnín and Seòsamh Watson for reading an earlier draft of this article and for their helpful suggestions.

Appendix

A Description of Sanda by the Rev. Father Edmund MacCana

Introduction On the 4th of January 1619, as missionaries for the revival of Catholicism in the Western Highlands and Islands of Scotland, two Irish-speaking priests from Ulster, Patrick Brady and Edmund MacCana, along with laybrother John Stuart, travelled from the Irish Franciscan College of St Anthony at Louvain, arriving in Scotland about the middle of March. Patrick Brady chose the Highlands as his territory; Father MacCana concentrated on the Isles (Giblin 1964, ix–xii). ‘Towards the end of 1620, [MacCana] was arrested and kept in prison for two years, after which he was banished and ordered not to enter England or Scotland again under pain of death’ (ibid., x). Appointed by Pope Urban VIII in 1623 (Stevenson 1980, 53), MacCana was one of four missionaries to go from Louvain to Scotland in 1624: ‘[h]e definitely laboured on the mission, but it is not possible to say how long he actually stayed there, and of the four missionaries least of all is known about his activities’ (Giblin 1964, xi). It is uncertain when MacCana wrote his description of Sannda. The editors of Origines Parochiales Scotiae (II, 820 note 5) date the MS to c.1600, so also Campbell (1924, 3: ‘presumed to be dated Circa 1600’), while Reeves (1864, 133) suggests ‘the early part of the seventeenth century’. These dates may be too early. On the 15 July 1624, Patrick Hegarty ‘explained the meaning of the sacred vestments to the islanders, and preached to them; they had been instructed in the elements of the faith four years before by another Irish Franciscan, but had seen Scottish Gaelic Sannda and Its Aliases 87 no priest since then’ (Giblin 1964, 32–33). It is likely that this other Franciscan was Edmund MacCana. His description, then, which we may infer from its contents was not written on Sannda, may date at the earliest from c.1621–22 and the period of his imprisonment in Scotland. Nevertheless, as indicated above, MacCana returned to Scotland in 1624, and, although it is not clear when he left the country again, the Franciscan mission to Scotland lasted until c.1637 (ibid., xv). However, reference within the description to MacCana’s poor recollection (quorum mihi ... memoria non suppetit ...) and to ‘this war’ (ante hoc bellum) may indicate an even later date, perhaps some time between 1644–1651 and the Wars of the Three Kingdoms.63 This accords with William Reeves’s (1854, 44) dating of MacCana’s Itinerarium in Hibernia ex relatione R. P. Fratris Edmundi MacCana, which ‘appears from internal evidence to have been written shortly after 1643, and to have been intended as a topographical contribution to the antiquarian store which the Irish Franciscans of Louvain had, with such laudable zeal, been for years accumulating in the service of their beloved country.’

The manuscript (Plates 1–2) The MS is housed in Bibliothèque royale de Belgique, 5301–20 fos 276–77, and is cited in the following:

Bindon (1847, 491), in his catalogue of MSS relating to Ireland in the Burgundian Library: ‘[Vol. XVIII] No. 5308. “Descriptio Insulae Sandae,” ... commences, “Insula Sanda est.”’

Smith (1854, 217): ‘Insulæ Sandæ, seu Auoniae, Hibernicé ABHUIN, brevis descriptio, R. P. fris Edmundi Mac Cana.’

Origines Parochiales Scotiae II, 1855, 820, note 5: ‘MS. in the Burgundian Library at Brussels, circa 1600, entitled Insulæ Sandae seu Auoniae Hibernice Abhuinn Brevis Descriptio, by Friar Edmund McCana.’

Skene (1874, clxviii): ‘Father Mac Cana’s MS account of the island states that in Irish it is called Abhuinn, Latinized Avonia.’

van Den Gheyn (1907, 49), in his catalogue of MSS in Bibliothèque royale de Belgique: ‘[4641 (5301–20)] 12. (F. 140–41 [leg. 276–277]) Insulae

63 cogadh sa na tri Rioghachtuibh ‘this war of the three Kingdoms’ (Gille-Críost MacBheatha – Christopher Beaton – writing in what is now known as the Book of Clanranald during the 1690s or early 1700s (Cameron 1892, 176; Bannerman 1998, 16–17)). 88 Richard A.V. Cox Sandae seu Avoniae Hibernie [sic] Abhyn brevis descriptio R. P. fratris Edmundi Mac Cana.’

Campbell (1924, 3–4): ‘A short description of the Island of Sanda, or Avon – in Irish Abhuinn by the Reverend Father, Brother Edmund McCana.’

Transcription While the original has been edited for punctuation and capitalisation, and abbreviations expanded silently, deletions have not been reproduced. On the few occasions where characters against the right margin, having been obscured by the binding process, are no longer visible, recourse has been made to Charles Mac Donnell’s transcription in Reeves 1864, 133–34. Diacritics have been ignored and allographs (e.g. < i, j>) rationalised.

[fo. 276] Insulæ Sandæ seu Avoniæ, Hibernice Abhuinn,64 Brevis Descriptio. R. P. fratris Edmundi mac Cana

Insula Sanda est in oceano Scotico ad occasum, uno milliari a Kinntiriæ continenti sejuncta; complectitur in circuitu unum magnum milliare. Solum jucundum, fructuum ac frugum, si coleretur, ferax. In ea est ædicula S. Ninniano sacra, ad cujus cænobium in Galvidia tota insula spectat. Conjunctum huic ædiculæ est ossarium sive sepulchretum quat- uordecim filiorum sanctissimi viri Senchanii Hiberni, sanctitate illus- trium. Saxeo murulo septum, in quo sunt septem grandia et polita saxa, quibus sanctissima corpora teguntur, in quorum medio stat obeliscus, altior hominis statura (ut mihi jam suggerit memoria). Nemo mortalium impune ingreditur illum murulum. Lepidum est quod mihi retulerunt insulani: gallinam id loci ingressam ova peperisse et exclussisse; pullos, cum jam præ ætate egredi poterant, omnes intortis collis – insigni spec- taculo – processisse. Retulit mihi etiam grandior natu insulanorum, et ferme omnium pater, hoc prodigium quod subscribo. Ængussius Mac- Donellus, Kenntiriæ ac Insulæ Ilæ dinasta (quem ipse jam olim vidi), in- gressus est aliquando insulam, multa comitante caterva, inter quos etiam fuit præcipua Kinntiriæ juventus. Cum forte dinasta ac cæteri nobiles de rebus seriis tractarent, juventus, ut solet, se pilæ ac clavarum ludo exercebat; pila vi clavæ impulsa, priusquam ab adversa manu juvenum excipi posset, altius in sacrum sepulchretum volavit. Juvenis, memor loci

64 MS Abhuin. Scottish Gaelic Sannda and Its Aliases 89 religionis, injecit tantum alterum pedem et manum, ad extrahendam pilam. Ab incolis reprehenditur, quod sacri loci majestatem violaverit idque criminis eum impune minime laturum denunciant. Ille lusum ni- hilominus cum sociis persequitur. Exacto lusu, ac appetente nocte, in hospitium se recipit, ad focum sedet; cooriuntur statim ingentes dolores in toto pede quem in sepulchretum intulit. Insulani significant divinam esse ultionem læsæ religionis. Intumuit mirum in modum pes, adeo in- flatus divina ultione ut equi magnitudinem exequaret.

[fo. 277] Sub mediam noctem juvenis expirat. Omnes Deum laudant; sancta cor- pora deinceps religiosius venerantur. Hinc discendum quantam habeat rationem et curam sanctorum suorum Deus optimus maximus, quorum sacrilegam irrisionem et contemptum impius Calvinus, novus evange- lista, orbi intulit, aut potius intrusit. Magnum hoc miraculum excitavit in animis spectatorum et ex ipsis audientium, etiam a nostra religione aversorum, sanctorum hominum reverentiam.

In illa insula fuit repertum brachium Sancti Ultani, quod, thecæ argen- teæ inclusum, ante hoc bellum religiose servabatur a viro generoso ex inclyta Mac Donellorum familia. Fons est ibi non procul a sacello per- ennis aquæ, miraculis, ut insulani et multi ex continenti mihi dixere, nobilis. Frequentabatur quidem meo tempore ab accolis circumquaque, maxime ab iis in quorum animis aliquæ reliquiæ priscæ religionis reside- bant. Sunt multa alia mira et jucunda quæ homines mihi fide dignissimi de hoc loco retulerunt, quorum mihi et memoria non suppetit et tem- pore excludor.

Illis sacris cineribus hoc quod sequitur rude epitaphium cum ibi essem posui; atque ad illud sacrum sepulchretum tertio sacris misteriis cum magna animi mei recreatione sum operatus.

Corpora bis septem, totum veneranda per orbem, Senchanii65 natum Sanda beata tenet. Doctorum divumque parens, Hibernia quondam Quos genuit sanctos Scotia terra tegit. Scotia dicta minor, multis celebrata trophæis,66 Matris in amplexu, pignora cara tenet. 65 MS Senchanii. 66 Inserted interlineally above multis celebrata trophæis are the words genuit qu< > Scotia major. 90 Richard A.V. Cox

Plate 1 © Bibliothèque royale de Belgique

‘Insulae Sandae seu Avoniae Hibernie Abhuin brevis descriptio R. P. fratris Edmundi Mac Cana’, MS 5301–20, fo. 276, reproduced here with the kind permission of Bibliothèque royale de Belgique.

The Journal of Scottish Name Studies 4, 2010, 61–102 Scottish Gaelic Sannda and Its Aliases 91

Plate 2 © Bibliothèque royale de Belgique

‘Insulae Sandae seu Avoniae Hibernie Abhuin brevis descriptio R. P. fratris Edmundi Mac Cana’, MS 5301–20, fo. 277, reproduced here with the kind permission of Bibliothèque royale de Belgique.

The Journal of Scottish Name Studies 4, 2010, 61–102 92 Richard A.V. Cox Sanda tibi cedit, veterum celebrata Camœnis, Bettiginum gazæ, ripa beata Tagi. Hos igitur sacros cineres devotus adora, Quisquis in Hebrigenum littora tuta venis.

[The following lines have been added as a footnote] Corpora bis septem, septem conduntur in urnis, Ut natu gemini, sic videantur humo.

Translation67 A short description of the Island of Sanda, or Avon, in Irish Abhuinn, by the Reverend Father, Brother Edmund MacCana.

[fo. 276] ‘Sanda is an island to the west in the Scottish sea, separated by a mile from the mainland of Kintyre; in circumference it is just over a mile. The soil is agreeable and, if cultivated, would be prolific in crops and fruits. In it is a chapel consecrated to St Ninian, to whose monastery in Galloway the whole island belongs. Adjoining this shrine is the ossuary or burial place of the fourteen sons of a most holy man, Senchan, an Irishman, who were renowned for their sanctity. It is surrounded by a low stone wall, in which are seven large polished stones, by which the most sacred bodies are covered, in the middle of which stands an obelisk higher than a man’s stature (as it now occurs to my memory). No mortal enters that enclosure with impunity. What the islanders have told me is charming: a hen, having entered the place, laid its eggs and hatched them; the chicks, when they were old enough to come out, all appeared with twisted necks – a remarkable sight! One of the older inhabitants of the islands, and father of nearly all the rest, also related to me this amazing story which I append. Angus MacDonell, Lord of Kintyre and of the Island of Islay (whom I myself saw once), entered the island one time, a large crowd in attendance, among whom were the principal youth of Kintyre. When, as it happened, the lord and other nobles were discussing serious mat- ters, the youths, as is their wont, exercised themselves at a game of shinty [‘ball and sticks’]; a ball hit by a caman stroke flew over into the sacred cemetery, before it could be caught by a hand of the opposing youths. The young man, mindful of the place’s religious significance, entered it

67 A sometimes inflated translation is given by the Rev. Father Butler in Campbell 1924, 3–4, a heavily truncated one in Origines Parochiales Scotiae II, 820. Scottish Gaelic Sannda and Its Aliases 93 using only one foot and only one hand to extract the ball. He is taken to task by the natives because he had violated the dignity of the holy place; and they declare that his crime will not go unpunished. Nevertheless, he continues that game with his companions. The game being finished, and the night approaching, he goes to the guests’ lodging, and seats him- self at the fireplace; great pains suddenly break out throughout the foot which he had put into the cemetery. The islanders indicate the divine vengeance is for lack of piety. The foot swelled to an astonishing extent, thus inflated by divine wrath until it equalled a horse’s in size.

[fo. 277] ‘During the middle of the night the youth expires. All praise God, and subsequently they venerate the sacred bodies with greater devotion. Hence it is learnt to what extent the most good and most high God keeps the reckoning and care of his saints, whose sacreligious mock- ery and contempt unholy Calvin, the new Evangelist, has introduced, or rather intruded, into the world. This great spectacle kindled in the minds of the spectators, and even among those who just heard about it, even those who had turned away from our creed, a reverence for holy men.

‘In that island the forearm of Saint Ultan was found, which, enclosed in a silver case, was carefully preserved before this war by a nobleman of the renowned family of the MacDonnells. There is a perpetual water spring there not far from the chapel, known for miracles, as the islanders and many from the mainland have told me. Indeed it was frequented in my own time by neighbours on all sides, especially by those in whose minds any vestige of the old religion remained. There were many other wonderful and delightful things that men most worthy of confidence reported to me about this place, which for me both memory fails and time excludes.

‘When I was there I placed over those sacred ashes the rough epitaph which follows, and at that sacred cemetery I thrice performed the sacred mysteries with great refreshment of mind.

‘Fourteen bodies, throughout the world revered, Of Senchanius born blessed Sanda holds. Ireland, the mother of divine teachers, once Begat the saints whom Scotland’s soil covers. 94 Richard A.V. Cox Scotland the minor, famed for its many memorials, Holds these beloved pledges in a mother’s embrace. Sanda, renowned for its ancient poetry, yields to you, Treasures of Mount Bettigo, blessed shores of the River Tagus.68 O faithful one, entreat these sacred ashes, then, you who comes to the safe shores of the Hebrigenae.69

[Footnote] ‘Fourteen bodies, preserved in seven urns, as they were twins at birth, so are they seen in the earth.’

References Adomnán: see Reeves 1857; Fowler 1894; Skene 1874; Anderson and Anderson 1991; Sharpe 1995. Aikman, James, 1827, The History of Scotland, Translated from the Latin of George Buchanan, Vol. 1 (Glasgow: Blackie, Fullarton & Co., and Edinburgh: Archibald Fullarton & Co.). Anderson, Alan Orr, 1922, Early Sources of Scottish History AD 500 to 1286, 2 vols (Edinburgh and London: Oliver and Boyd). Anderson, Alan Orr, and Marjorie Ogilvie Anderson, revised by Marjorie Ogilvie Anderson, 1991, Adomnán’s Life of Columba (Oxford: Clarendon Press). Bannerman, John, 1998, The Beatons: a medical kindred in the classical Gaelic tradition (Edinburgh: John Donald). Beeching, H. C., 1900, The Poetical Works of John Milton (Oxford: The Clarendon Press). Bindon, S. H., 1847, ‘On the MSS relating to Ireland in the Burgundian Library at Brusselles’, Proceedings of the Royal Irish Academy III, 477–502. Blaeu, Ioannes, ‘Cantyre Chersonesus, Cantyr a Demie-yland’, and ‘Vistus

68 Bettigo in India (Ptolemy, e.g. Lib. VII, cap. 1 §33, ¢pÕ toà Bhttigë Ôrouj (Nobbe 1843– 45 II, 148) ‘from the mountain of Bhttigë’) and the Tagus in the Iberian Peninsula (Ptolemy, Lib. II, cap. 5 §4, T£gou potamoà ˜kbola… (Nobbe, ibid. I, 80) ‘the mouth of the river T£goj’) may be being used here to represent natural wonders of the world. 69 Nom. pl. of Hebrigenum, which is taken to be a short form of Hebrigenarum ‘of the Hebrideans’, gen. pl. of a compound, Hebrigena ‘Hebridean’ (as in Romigena ‘Roman’ and the ninth-century theologian Johannes Scotus Eriugena’s name, whose pseudonym contains Ériu ‘Ireland’), cf. John Milton’s penè totis finibus Angligenûm (‘Ad Joannem Rousium’, Beeching 1900, 170.32) ‘almost all the lands of the English’. Scottish Gaelic Sannda and Its Aliases 95 Insula, vulgo Viist, cum aliis minoribus ex Aebudarum numero ei ad meridiem adjacentibus’, Theatrum orbis terrarum sive Atlas novus (Amsterdam 1654), surveyed by Timothy Pont between 1583–1596, in Jeffrey Stone, Illustrated Maps of Scotland from Blaeu’s Atlas Novus of the 17th Century (London: Studio Editions Ltd, 1991): Plates 19, 44. Bremner, Robert Locke, 1923, The Norsemen in Alban (Glasgow: Maclehose, Jackson & Co.). Buchanan, George, [1582] 1727, Rerum Scoticarum Historia, ed. Robert Fribarn (Edinburgh: Jo. Paton). Buchanan, George, [1582] 2009, Rerum Scoticarum Historia (1582), ed. Dana F. Sutton (Irvine: University of California), . Buchanan, George, 1827: see under James Aikman. Byrne, Michel, 2000, Collected Poems and Songs of George Campbell Hay (Deòrsa Mac Iain Dheòrsa), 2 vols (Edinburgh University Press for the Lorimer Trust). Cameron, Alexander, 1894, Reliquiae Celticae II, ed. Alexander Macbain and John Kennedy (Inverness: The Northern Counties Newspaper and Printing and Publishing Company). Campbell, Sheriff J[ohn] MacMaster, [1924], ‘The Island and House of Sanda, with Related Passages from the Annals of Kintyre’ (reprinted from The Campbeltown Courier), [entitled, p. 1] ‘ “The Island and House of Sanda.” A lecture delivered by Sheriff J. MacMaster Campbell ... to Kintyre Antiquarian Society, on 13th Novermber, 1924’. Cleasby, Richard, and Gudbrand Vigfusson, 1874, An Icelandic-English Dictionary (Oxford: The Clarendon Press). Collectanea de Rebus Albanicis, ed. The Iona Club (Edinburgh: Thomas G. Stevenson 1847). Collins: Collins English Dictionary, Complete and Unabridged (6th edn; Glasgow: Collins, 2003). Colville, Duncan: see under KAS. Cox, Richard A.V., 1990, ‘Place-nomenclature in the Context of the Bilingual Community of Lewis: Status, Origin and Interaction’, Proceedings of the 2nd International Conference on the Languages of Scotland (Glasgow 1988), in Gaelic and Scots in Harmony, ed. Derick S. Thomson, 1990 (Glasgow: Department of Celtic, University of Glasgow), 43–52. Cox, Richard A.V., 1998 ‘Onomastic Luggage: Variability in the onomastic landscape’, Nomina 21, 15–28. Cox, Richard A.V., 2000, ‘The Phonological Development of Scottish 96 Richard A.V. Cox Gaelic uinneag “window” and Related Questions’, Scottish Gaelic Studies XX, 212–21. Cox, Richard A.V., 2002, The Gaelic Place-names of Carloway, Isle of Lewis: Their structure and significance (Dublin: Dublin Institute for Advanced Studies). Cox, Richard A.V., 2008, ‘The Development of Old Norse -fn in (Scottish) Gaelic’, The Journal of Scottish Name Studies 2, 51–68. Dasent, George W., transl., [1894] 1997, The Saga of Hacon and a Fragment of the Saga of Magnus with Appendices 2 vols (facsimile repr., Felin Fach: Llanerch Publishers; 1st pub. 1894). DIL: Dictionary of the Irish Language and Contributions to a Dictionary of the Irish Language, ed. E.G. Quin et alii (Dublin: Royal Irish Academy 1913–1976). Dwelly, Edward, 1988, The Illustrated Gaelic–English Dictionary (10th ed., Glasgow: Gairm Publications; 1st pub. 1901–11). Eirspennill: Eirspennill – Am 47 fol – Nóregs konunga sôgur: Magnús góði– Hákon gamli, ed. Finnur Jónsson (Kristiania: Julius Themtes Boktrykkeri, 1916). Flateyjarbok: Flateyjarbok: En samling af Norske Konge-sagaer, 3 vols, ed. Guðbrandr Vigfusson and C. R. Unger (Christiania: P. T. Mallings Forlagsboghandel, 1860–1868). Forbes, Alexander R., 1923, Place-names of Skye and Adjacent Islands: with lore, mythical, traditional and historical (Paisley: Alexander Gardner). Fowler, J. T., 1894, Adamnani Vita S. Columbae, edited from Dr. Reeves’s text ... (Oxford: Clarendon Press). Fraser-MacKintosh, Charles, 1895, The Last MacDonalds of Isla (Glasgow: Celtic Monthly Office). Fribarn 1727: see under George Buchanan. Forte, A. D. M., ‘A Preliminary Account of an Early Sixteenth-Century Episode of Highland Piracy’, Scottish Gaelic Studies 24, 207–14. Gammeltoft, Peder, 2006, ‘Scandinavian Influence on Hebridean Island Names’, Names through the Looking-glass: Festchrift in Honour of Gillian Fellows-Jensen, ed. Peder Gammeltoft and Bent Jørgensen (Reitzels: Copenhagen), 53–84. Giblin, Cathaldus, 1964, Irish Franciscan Mission to Scotland 1619–1646: Documents from Roman Archives (Dublin: Assisi Press). Gillies, H. Cameron, 1906, The Place-Names of Argyll (London: David Nutt). Goodall, Walter, 1759, Joannis de Fordun Scotichronicon ... Vol. I (Edinburgh: Robert Fleming). Scottish Gaelic Sannda and Its Aliases 97 Gordon, Robert, c.1636–52: Cantyre, . Grant, Seumas, 2000, The Gaelic of Islay: a comparative study (Aberdeen: Department of Celtic, University of Aberdeen). Groome, Francis H., 1884, Ordnance Gazetteer of Scotland: a survey of Scottish Topography, statistical, biographical, and historical Vol. I (Edinburgh, London, Glasgow, Aberdeen: Thomas C. Jack, Grange Publishing Works; first published 1882). Groome, Francis H., 1896, Ordnance Gazetteer of Scotland, facsimile copy on-line at , ed. Mike Spathaky (2001, 2006) – 13/8/09. Haswell-Smith, Hamish, 2003, An Island Odyssey: among the Scottish Isles in the wake of Martin Martin (Edinburgh: Canongate). Haswell-Smith, Hamish, 2004, The Scottish Islands: a comprehensive guide to every Scottish island (Edinburgh: Canongate; 1st pub. 1996). Hearne, Thomas, 1722, Johannis de Fordun Scotichronicon ... (Oxford: The Sheldonian Theatre). Heimskringla: Heimskringla: eða Sögur Noregs konunga Snorra Sturlusonar, ed. N. Linder and H. A. Haggson (Uppsala: W. Schultz, 1869–72). Helgadóttir, Guðrún P., ed., 1987, Hrafns saga Sveinbjarnarsonar (Oxford: Clarendon Press). Henderson, George, 1910, The Norse Influence on Celtic Scotland (Glasgow: James Maclehose). Holmer, Nils M., 1938, Studies on Argyllshire Gaelic, K. Humanistiske Vetenskaps-Samfundet i Uppsala 31: 1. Holmer, Nils M., 1942, The Irish Language in Rathlin Island, Co. Antrim, Royal Irish Academy Todd Lecture Series XVIII (Dublin: Royal Irish Academy). Holmer, Nils M., 1957, The Gaelic of Arran (Dublin: Dublin Institute for Advanced Studies). Holmer, Nils M., 1962, The Gaelic of Kintyre (Dublin: Dublin Institute for Advanced Studies). Innes, Thomas, [1729] 1885, A Critical Essay on the Ancient Inhabitants of the Northern Parts of Britain, The Historians of Scotland VIII (Edinburgh: William Paterson). Jesch, Judith, forthcoming, ‘The Threatening Wave: Norse poetry and the Scottish Isles’, in the proceedings of the Maritime Societies of the Viking and Medieval World conference (Kirkwall, 2008), ed James Barrett. John of Fordun: see under Walter Goodall, Thomas Hearne and William Skene. 98 Richard A.V. Cox Jones, George, 2000, ‘Beagan mu’n Stad Ghlotasach ann an Gàidhlig Ceann a Deas Earraghaidheil’, Scottish Gaelic Studies XX, 201–11. Jónsson, Finnur: see Eirspennill. KAS: Kintyre Antiquarian Society, 1938, The Place Names of the Parish of Southend [ed. Duncan Colville] (Kintyre Antiquarian Society: Campbeltown Courier Office). See also under Martin 2009. LASID: Linguistic Atlas and Survey of Irish Dialects, 4 Vols, Vol. I ed. Heinrich Wagner (1981; 1st pub. 1958), Vol. IV ed. Heinrich Wagner and Colm Ó Baoill (1969) (Dublin: Dublin Institute for Advanced Studies). Lewis, Samuel, 1846, A Topographical Dictionary of Scotland: 445–61 ‘Saddell – Selkirkshire’: Sanda ; 467-489 ‘Skye – Stewarton’: Southend – 13/8/09. Linder, N., and H. A. Haggson: see Heimskringla. Mac an Tàilleir, Iain, 2003, Ainmean-Àite: . Mac Cana, Edmund: see Appendix. Macculloch, John, 1819, A Description of the Western Islands of Scotland including the Isle of Man ... Vol. II (London: Printed for Archibald Constable and Co. Edinburgh; and Hurst, Robinson, and Co. Cheapside, London). Macculloch, John, 1824, The Highlands and Western Isles of Scotland ... Vol. II (London: Longman, Hurst, Rees, Orme, Brown, and Green, Paternoster Row). MacFarlane, Walter: see Arthur Mitchell. MacKay, John, 1890, ‘Sutherland Place Names’, Transactions of the Gaelic Society of Inverness XV, 1888–89, 107–22. MacMaster Campbell: see Campbell. Martin, Angus, 1984, Kintyre: The Hidden Past (Edinburgh: John Donald Publishers Ltd). Martin, Angus, 2009: The Place Names of the Parish of Southend, compiled by Duncan Colville, revised and supplemented by Angus Martin (1st pub. in 1938 by KAS, q.v.; Campbeltown: Kintyre Antiquarian & Natural History Society). Martin, Martin, 1703, A Description of the Western Islands of Scotland [1695] (1st pub. 1703; 2nd ed. 1716; repr. 1884, 1934; facsimile copy in Michael Robson, Curiosities of Art and Nature: The new annotated and illustrated edition of Martin Martin’s classic A Description of the Western Islands of Scotland, Port of Ness: The Islands Book Trust, 2003). Scottish Gaelic Sannda and Its Aliases 99 Marwick, Hugh, 1923, ‘Celtic Place-names in Orkney’, Proceedings of the Society of Antiquaries of Scotland 57, 251–65. Meyer, Kuno, [1912] 1994, Sanas Cormaic: An Old Irish Glossary, compiled by Cormac Úa Cuileannáin, King-bishop of Cashel in the tenth-century, facsimile reprint with the editor’s corrections (Felinfach: Llanerch Publishers; 1st pub. in 1912 in Anecdota from Irish Manuscripts IV (Dublin: Halle A.S.)). Mitchell, Arthur, 1907–08, Geographical Collections Relating to Scotland Made by Walter MacFarlane, 3 vols (Edinburgh: printed at the University Press by T. and A. Constable for the Scottish History Society; II 1907, III 1908). Moll, Herman, 1745, The Shire of Bute: contains Bute and Arran with Cantire which is Part of Argyle Shire, . Monro, Donald, [1549] 1774, Description of the Western Isles of Scotland, called Hybrides; by Mr Donald Monro High Dean of the Isles, who travelled through the most of them in the year 1549 [printed in the year 1773] (Edinburgh: William Auld). Monro, Donald, [1549] 2002, A Description of the Occidental i.e. Western Isles of Scotland, in Martin Martin, A Description of the Western Isles of Scotland Circa 1695 and A Voyage to St Kilda with Donald Monro, A Description of the Occidental i.e. Western Isles of Scotland, with Introductions by Charles W. J. Withers and R.W. Munro (Edinburgh: Birlinn; 1st pub. 1999), 289– 339. [This is from a different MS to the 1774 edition.] Morkinskinna: Morkinskinna, ed. Finnur Jónsson (København: Háskóli Íslands, 1932). Munch, P. A., 1859, Norges konge-sagaer fra de ældste tider indtil anden halvdeel af det 13de aarhundrede efter Christi fødsel I (Christiania: Feilberg Landmart), – accessed 25/10/09. Munro, R.W., 2002: Introduction to Monro [1549] 2002. Murchison, Thomas M., ed., 1960, Prose Writings of Donald Lamont (Edinburgh, Oliver and Boyd for the Scottish Gaelic Texts Society). New Statistical Account Vol. VII: Argyle: Southend (Edinburgh, 1843–45). NLS: National Library of Scotland. Nobbe, Carolus Fridericus Augustus, 1843–45, Claudii Ptolemaei Geographia 3 vols (Lipsiae sumptibus et typis Caroli Tauchnitii). NSA: New Statistical Account. Ó Baoill, Colm, 1978, Contributions to a Comparative Study of Ulster Irish and Scottish Gaelic (Belfast: Institute of Irish Studies, The Queen’s University of Belfast). 100 Richard A.V. Cox Ó Baoill (1990), ‘Rhyming Vowels before Long Liquids in Scottish Gaelic’, Éigse XXIV, 1990, 131–46. Ó Baoill, Colm, 2000, ‘The Gaelic Continuum’, Éigse XXXII, 121–34. O’Rahilly, Thomas F., 1976, Irish Dialects Past and Present (Dublin: Dublin Institute for Advanced Studies; 1st pub. 1932). Oftedal, Magne, 1956, The Gaelic of Leurbost, Isle of Lewis, A Linguistic Survey of the Gaelic Dialects of Scotland III, Norsk tidsskrift for sprogvidenskap Supplementary Vol. IV (Oslo). Origines Parochiales Scotiae: Origines Parochiales Scotiae: the antiquities ecclesiastical and territorial of the parishes of Scotland Vol. 2 (Edinburgh: W. H. Lizars, Glasgow: J. Smith & Son, London: S. Highley, Jun.,1855). OS: Board of Ordnance 6" : 1 mile map, 1869, 1881, 1882. Paterson, John, and John Renwick, 1900, ‘Report of a Visit to Sanda and Glunimore’, Transactions of the Natural History Society of Glasgow 5 (New Series), 197–204. Pennant, Thomas, 1772, A Tour in Scotland and Voyage to the Hebrides, 1772 (Chester: John Monk, 1774; 2nd ed., London: Benj. White, 1776; London: Benj. White, 1790). Playfair, James, 1819, A Geographical and Statistical Description of Scotland ... Vol. II (Edinburgh: Archibald Constable and Co.; London: Longman, Hurst, Rees, Orme and Brown, and Hurst, Robinson and Co.). Pulsiano, Phillip, ed., and Kirsten Wolf, co-ed., 1993, Medieval Scandinavia: An Encyclopedia (New York: Garland). Purser, John, 2007, Scotland’s Music: a history of the traditional and classical music of Scotland from early times to the present day (Edinburgh and London: Mainstream Publishing in conjunction with BBC Scotland). Reeves, William, ed., 1857, Vita Sancti Columbae (Dublin: The Bannatyne Club). Reeves, William, 1854, ‘Irish Itinerary of Father Edmund MacCana translated from the Original Latin and Illustrated by Notes’, The Ulster Journal of Archaeology 2, 44–59. Reeves, W[illiam], 1864, ‘On the Island of Sanda’, Proceedings of the Royal Irish Academy 8, 132–35. Retours: Retours of Services of Heirs: Inquisitionum ad Capellam Domini Regis Retornatarum Abbreviatio 1544–1699 Vols I–III, ed. Thomas Thomson (CD-Rom: Edinburgh: The Scottish Genealogy Society, 1999; Book: 1st pub. 1811–16). Roy, William, 1747–55, Military Survey of Scotland, . Scottish Gaelic Sannda and Its Aliases 101 SAS: Statistical Account of Scotland. SGDS: Survey of the Gaelic Dialects of Scotland 5 vols, ed. Cathair Ó Dochartaigh (Dublin: Dublin Institute for Advanced Studies, 1994). Sharpe, Richard, translator, 1995, Adomnán of Iona: Life of St Columba (Penguin Books). Skene, William F., 1871–72, Johannis de Fordun: Chronica Gentis Scotorum, 2 vols (Edinburgh: Edmonston and Douglas). Skene, William F., ed., 1874, Life of Saint Columba, Founder of Hy, written by Adamnan, ninth abbot of that monastery, a re-edition of Reeves 1857, with a translation by Bishop Forbes, The Historians of Scotland VI (Edinburgh: Edmonston and Douglas). Smith, J. Huband, 1854, ‘The Shrine of Saint Patrick’s Hand, with Notices of Some Similar Reliquiaries’, Ulster Journal of Archaeology II, 207–23. Statistical Account of Scotland Vol. 3: Argyll, Southend (Edinburgh, 1791– 99). Stevenson, David, 1980, Alasdair MacColla and the Highland Problem in the Seventeenth Century (Edinburgh: John Donald). Stiùbhart, Domhnall Uilleam, 2009, ‘Alasdair MacGilleMhìcheil agus Cultar Dùthchasach’, in Richard A.V. Cox, ed., Dualchas agus an Àrainneachd (Slèite: Clò Ostaig), 135–60. Stone, Jeffrey: see under Ioannes Blaeu. Sutton 2009: see under George Buchanan. Thomas, [Captain] F. W. L., 1880, ‘Traditions of the MacAulays of Lewis’, Proceedings of the Society of Antiquaries of Scotland XIV, 363–431. Thomson, John, 1820, Northern Part of Argyll Shire. Southern Part, . Tìr Chaluim Chille [map], Iomairt Cholm Cille, 2003. Tjomsland, Anne, 1951, ‘The Saga of Hrafn Sveinbjarnarson: the life of an Icelandic physician of the thirteenth century’, Islandica XXXV. van Den Gheyn, J., 1907, Catalogue des manuscrits de la Bibliothèque Royale de Belgique (Bruxelles: Henri Lamertin). van Keulen, Johannes, c.1780, Nieuwe Caart van de Rivier van Glasgow in ‘t Groot, . Vigfusson, Gudbrand, 1878, ed., Sturlunga Saga including the Islendinga Saga of Lawman Sturla Thordsson and Other Works, 2 vols (Oxford: Clarendon Press). Vigfusson, Gudbrand, 1887, Icelandic Sagas and Other Documents Relating to the Settlements and Descents of the Northmen on the British Isles Vol. 1 (London: HMSO). 102 Richard A.V. Cox Vigfusson, Gudbrand, and F. York Powell, 1905, Origines Islandicae I (Oxford: Clarendon Press). Vigfusson, Guðbrandr, and C. R. Unger: see Flateyjarbok. Watson, W. J., 1906, ‘Some Sutherland Names of Places’, Celtic Review II, No. 8, 1906, 232–42 and 360–68; repr. in Watson, William J., 2002, Scottish Place-Name Papers (London and Edinburgh), 56–75. Watson, W. J., 1924, ‘The Celts (British and Gael) in Dumfriesshire and Galloway’, Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society, 3rd Series, XI, 1923–24, 119–48. Watson, William J., 1926, The History of the Celtic Place-names of Scotland (Edinburgh and London: William Blackwood & Sons). Watson, W. J., 1976, Place-names of Ross and Cromarty (Inverness: Ross and Cromarty Heritage Society; 1st pub. 1904). Scotland’s -ham and -ingham Names: a reconsideration

Alan G. James Scottish Place-Name Society

Place-names in southern Scotland ending in -ham and, especially, -ingham, have received a fair amount of attention from scholars in the past 30 years, being rightly recognised as traces of an early phase – some would say the earliest phase – of English-speaking colonisation north of Hadrian’s Wall. W. F. H. Nicolaisen’s chapter on ‘Early English Names’ in Scottish Place-Names (1976 [2001], ch. 5) was the starting-point and remains the main reference-point for their study. At the heart of that chapter is a careful consideration of a range of candidates that may be considered to be among the earliest Anglian place- names in Scotland. Out of this discussion, Nicolaisen drew three – Coldingham BWK, Tyninghame ELO and Whittingehame ELO – as being of a type that, in southern and eastern England, correlate closely with archaeological evidence for a relatively early phase of Anglo-Saxon colonisation, dating from the period before conversion to Christianity. In later printings of that work he added a fourth, Penninghame WIG.1 These he declared to be the earliest. Although his judgements have been widely accepted by place-name scholars and are frequently cited by archaeologists and historians, they have always seemed to me to be surprising, and I take this opportunity to explain my doubts and to suggest some possible alternative approaches to the interpretation – both philological and historical – of these and other -ham and -ingham names in the early Bernician heartland (the lower Tweed basin and that of the East Lothian Tyne, and along the Berwickshire and East Lothian coasts), in Galloway and in lands overlooking the Clyde coast.2

-hām in settlement names Firstly, we need to consider the generic, -hām. This is, of course, the English word ‘home’, Scots ‘hame’, a habitative term referring to a settlement.3 However it is important to note that present-day English place-name scholars recognise that it referred as much to the landholding associated with the

1 See under -ing³ and subsequent sections below. 2 An earlier version of this paper was presented at the spring 2009 meeting of the Scottish Place-Name Society at the Catstrand in New Galloway and reported in Scottish Place-Name News 27, Autumn 2009. I am grateful to several who heard the paper or read the report and offered helpful suggestions, to anonymous referees and, as ever, to Dr Simon Taylor for unfailing help and encouragement. 3 The fullest account of its OE senses relevant to toponymy is A. H. Smith (1956a), 226–29.

The Journal of Scottish Name Studies 4, 2010, 103–30 104 Alan G. James settlement as to the homestead at its heart, just as a ‘farm’ is both a landholding and a set of buildings. So, in place-names, -hām may have referred to quite a substantial area, comparable to a later parish, and its main settlement may not always have been at or near the place that preserves -ham in its present- day name.4 Barrie Cox has demonstrated that this element was the most favoured habitative term among the earliest Germanic-speaking settlers and that it remained in use up to around the time of Bede, but radical changes in settlement patterns, land-holdings and fiscal systems caused it to fall into disuse as a place-naming term from about the mid-eighth century, being superseded by a range of other nouns, most commonly -tūn.5 This observation is vividly illustrated by the only concentration of place-names in ‑ham to be found in (or just outwith) Scotland, which is in the Tweed basin below Melrose, down to the Merse. Norham NTB and (probably) Yetholm ROX were heads of sċīras;6 both of these became mediaeval parishes, as did Edrom BWK, Ednam ROX, Oxnam ROX and Smailholm ROX. Midlem (Bowden ROX), Birgham and Leitholm (both Eccles BWK)7 and Kimmerghame (Edrom BWK), though not parishes, are all in favourable locations showing continuity of settlement. The significance of this group of -ham names has been overlooked by those archaeologists and historians who insist that Northumbrian rule was maintained by a tiny minority of warrior aristocrats dominating a linguistically and ethnically Celtic population.8 Such an ‘élite dominance’ hypothesis might be valid for the earliest period of rule from Bamburgh and even during the heyday of the palace complex at Yeavering, but these -ham names around the lower Tweed point to vigorous and determined English-speaking colonisation by the early eighth century. The fact that none of them (except, doubtfully, Kimmerghame) have personal names as specifiers suggests that they are unlikely to have been royal land- grants to favoured followers. Opportunistic or sponsored immigration from

4 Indeed, Gillian Fellows-Jensen (1990) argues that hām referred primarily to extensive estates and only came to be attached to particular settlements when these estates were disintegrating in the later Anglo-Saxon period. 5 Cox 1973, 15–73, and 1976, 61–63; see also Gelling 1984, 65. 6 I.e. major territorial units within the early Northumbrian kingdom; see Barrow 1973 [2003], 28–32 with map 4. Norham may not have acquired that secular status until it became an important ecclesiastical site; see further under ‘-hām in monastic names’, below. 7 On Birgham, see further under ‘-hām in monastic names’, below. 8 Hope-Taylor proposed such a context for the palace complex at Yeavering NTB in the 6th– early 7th centuries (1977, 276–300) but was not responsible for the widespread assumption that things remained so thereafter. Scotland’s -ham and -ingham Names: a reconsideration 105 the expanding areas of English-speaking population further south are more likely to have been factors in the formation of this toponymic geography.9

-hām in monastic names However, there is another consideration to be borne in mind when examining -hām. As Victor Watts reminds us in his discussion of the change of Hagustaldes ea to Hagustaldes hām, Hexham NTB, this element continued to be used in Old English, at least up to the early ninth century, as an appellative referring specifically to a religious house (and, presumably, its landholding) (1994, 135–36).10 Returning to Norham NTB, this was, or became, an important ecclesiastical site, housing for a time St Aidan’s wooden church and the bodies of St Cuthbert and King Ceolwulf. Their translation occurred, according to the Historia de Sancto Cuthberto, some time in the second or third quarters of ninth century.11 We cannot be sure whether Norham was then already a mynster 12 or was only established as such at that date; its status as head of a sċīr may only date from its temporary elevation to locus of the chief relics of St Cuthbert’s community.13 Still, the Northumbrian chronicle incorporated by Symeon of Durham in Historia Regum Anglorum adds the information that it was ‘anciently called Ubbanford’.14 This strongly suggests another ‘monastic’ re-naming like that at Hexham. Whether it originated in the eighth century or the ninth, its new name may very well be understood as ‘the religious house to the north (of Hexham?)’. It may also be worth noting that Roger of Howden added Birgham BWK to his version of the list of possessions of Lindisfarne (Anderson 1908, 60 note 8). It is conceivable that this was an Anglian mynster, successor to a British church or religious house implied by the parish name Eccles.15 If so, Birgham may well be another ‘monastic’ -hām in the Tweed valley.16

9 On the early history of the Bernician heartland, see J. E. Fraser’s cautious review of the evidence (2009, 149–54). 10 See also Smith 1956a, 227. 11 In South 2002, §9, 48–50; Symeon of Durham Historia Dunelmensis Ecclesiae in Arnold 1882, §5, 52–53, refers only to the building of a church there and the translation of Ceolwulf. On the date and historical context see Cambridge 1989, Crumplin 2004, 12 and note 42, and Woolf 2007, 79–86. 12 The house and church of a religious community, monastic at least in a broad sense, having some pastoral and civil-administrative responsibilities for a fair-sized area round about; see Blair 2005, 2–5, 80–83 et passim, and Foot 2006, 34–42, 283–91 et passim. 13 A group of fragments of stone crosses found there ‘represent the most up-to-date ninth- century taste of a strongly Deiran type’ (Cramp 1989, 227), though two possibly architectural fragments may be from the eighth century, according to Cambridge 1989, 371. 14 In Arnold 1882, II.101, s.a. 854; see Woolf 2007, 82. 15 Barrow 1973 [2003], 28–32, Hough 2009 and James 2009a, especially 130–32. 16 For a suggestion that Edrom may have been another mynster, see note 65 below. 106 Alan G. James It is interesting that, when we look north of the Lammermuirs, we find fewer place-names in -ham, but among those we do find are Morham ELO, where a fragment of Anglian sculpture in the church bears witness to an important ecclesiastical site in the Northumbrian period, probably a mynster – a mynster at Morham might well have succeeded a Celtic church and monastery associated with the royal centre on Traprain Law, nearby to the east; Auldhame ELO, where recent excavations have revealed a significant early Christian site;17 lost Pefferhame, mentioned in Historia Regum as a possession of Lindisfarne, doubtless on one of the Peffer Burns and presumably another ‘monastic’ hām;18 and, of course, Tyninghame, the monastery founded by followers of St Baldred.19 This suggests a rather different situation in East Lothian, one where large-scale, English-speaking colonisation may not have taken place but where Northumbrian hegemony was exercised to a significant degree through land- grants to religious communities able to provide civil and ideological leadership favourable to the house of Bamburgh.20

-ing3 Mention of Tyninghame brings me to the -ingham names and the need to clarify the central element, -ing-. Hugh Smith, in his work on place-name elements for the English Place-Name Survey, distinguished four usages of -ing in Old English name-formation:

(i) -ing1 a common nominal suffix, having a range of origins, forming terms for (male) persons (e.g. æðeling ‘a nobleman’), diminutives (e.g. dēorling

17 See www.aocarchaeology.com/field-archaeology/auldhame.htm – accessed 01/08/09. According to the Aberdeen Breviary, this was one of the three churches granted to St Baldred by St Kentigern. The others were ‘Cunninghame’ (probably a mistranscription of Tyninghame) and Preston ELO. The latter is likely to be a later name, reflecting the allocation of landholdings to priests, which John Blair and Thomas Pickles both see as a development of the ninth century onward (Blair 2005, 383–85; Pickles 2009, especially 42–49, 71–82, 84–86, 91–97). There are other Prestons: in BWK and (with Preston Merse) KCB, as well as Prestonpans (with Preston Tower) ELO and Preston Law PEB; Prestonfield MLO was Prestisfelde 1375 × 6 (Dixon 1947, 116) but clearly belongs in the same group. Pickles’s painstaking study overlooks these northernmost Prestons. 18 In Arnold 1882, II.101, s.a. 854: see Anderson 1908, 60 note 8. 19 See discussion of this monastic place-name under ‘Tyninghame’, below. 20 For other possible ‘monastic’ hām names, see under ‘Tigbrethingham’, ‘Edingham and Twynholm’ and ‘Eaglesham’, below. The only other, presumably ‘secular’, hām- names in ELO are Lyneringham and Whittingehame, discussed below, and Oldhamstocks near the coast and the border with BWK – another *ald-hām. Letham is obscure but unlikely to incorporate -hām. Scotland’s -ham and -ingham Names: a reconsideration 107 ‘darling’), concrete nouns (e.g standing ‘a standing, especially for a hunter to shoot game’) and feminine abstract nouns (e.g. rydding ‘a clearing’). (ii) -ing² a suffix (perhaps a further usage of (i)) that forms place-names (usually masculine or neuter, always singular) from topographical terms (including pre-existing Brittonic names, see below), as well as verbal roots, descriptive adjectives, words for animals and vegetation and personal and ethnic names. (iii) -ing3 a patronymic forming individual personal names but also, in the plural, ethnic names, see below. (iv) -ing4 an associative or connective particle, linking a personal name or other specific term to a generic such as -tūn in a semantically flexible way.21

The one labelled -ing1 need not detain us but we shall need to attend to -ing² and -ing³ and, at least in passing, -ing4.22 We begin with -ing3, the patronymic ‑ing:23 thus, X-ing means ‘the son or descendant of, or a person closely associated with, X’. The plural X-ing³as meant more broadly ‘descendants, kin, dependants of X’. In place-names in the south, we find -ing3 in the plural, so Hastings SSX is Hæstingas− ‘[place belonging to] the “clan” (as we might say) of one Hæst− ’.24 It used to be thought, reasonably enough, that names of this kind were very early sites of the original Germanic-speaking tribal settlements and Nicolaisen found no names of this sort in Scotland formed simply with this suffix, like Hastings, nor any formed from its dative plural, like Reading BRK, æt Rædingum.25 Still we must note that, as John Dodgson demonstrated in 1966, there is actually no correlation between the distribution of these names and that of pagan Anglo-Saxon burials: names of the Hastings and Reading types seem to reflect a later, post-Conversion stage in the expansion of English- speaking settlement (Dodgson 1966). We shall also need to bear in mind again a possible ecclesiastical association: the plural -ing³as can also be used of religious

21 Summarising A. H. Smith 1956a, 284–86, 290–91, 298–303. 22 For -ing², see under ‘Lyneringham and Tyninghame’, below, and for -ing4, see notes 53, 78 and 95, below. Smith’s account should be regarded as a pragmatic taxonomy of usages in OE place-naming; the Germanic origins and inter-relationships of the four types are complex matters that need not concern us here. 23 Discussed in detail by Smith 1956a, 290–91, but see observations in the following note. 24 Smith, rather confusingly, divorces the plural forms from -ing3, dealing with them separately in ibid., 298–303. His detailed discussion remains very important but has to be read in the light of the subsequent revisions by Dodgson and Cox, for which see below. 25 Nicolaisen 2001, 89–93, 95–98. There are none in CMB or WML either and probably none in NTB or DRH; see Watts 2002, xiii–xvii, and ibid., 26, for the complicated case of Cleatlam DRH. 108 Alan G. James communities: the Berclingas, for example, were in 804 the monks of Berkeley GLO. As is the case with the formation of ‘monastic’ names with -hām, this usage may be later than that of the true patronymic (Ekwall 1960, 39). Formations with the genitive plural of -ing3, -inga-, are like those with -ingas or ‑ingum, likewise generally associated with the expansion and development of Anglo-Saxon landholding in the early Christian period rather than the earliest, pagan settlements. Apart from the possible ‑inga-hām formations about to be considered, it is unlikely that there are any with ‑inga- plus any other generic in Scotland. Names like Edington BWK and Haddington ELO show no evidence in their early recorded forms for a vowel between -g- and -t-, so they are unlikely to be *-inga-tūn; they are probably *-ing²- or *-ing4-tūn, which could date from any time from the rise of tūn as the main habitative element (about the mid- eighth century) to the end of the Northumbrian period (Nicolaisen 2001, 89–93, 95–96, 98). However, Barrie Cox found that there is a correlation, to an impressive degree, between the archaeological evidence for pre-Christian Anglo-Saxon settlement, along Roman roads or navigable waterways, and place-names formed with -inga-hām, where -hām is combined with the genitive plural of -ing3 (Cox 1973, 15–73). Nottingham, for example, is Snotingahām ‘hām of the clan of Snot’ and is associated with evidence for the furthest penetration of pagan Anglo-Saxon settlement up the river Trent. Nicolaisen saw this -ingahām formation in Whittingehame, Coldingham, Tyninghame and Penninghame,26 and archaeologists and historians – naturally accepting his authority – have been eager to see them as traces of the earliest, pre-Christian Anglian colonisation of the north. Place-name scholars, too, have followed suit: Ian Fraser, for example, stated that ‘the earliest example of Anglian nomenclature in the border is probably Coldingham’ (1995, 184) and Carole Hough, having reviewed the implications of Dodgson’s and Cox’s findings regarding the‑ hām and (apparent) -ingahām names, declared that ‘the earliest strata of A-S settlement names from England are not after all absent from the Border counties’ (1997, 150). Yet it is odd that ‘the earliest strata’ should be present when, as we have seen, the later strata (names in -ingas, ‑ingum, or -inga- with a generic other than ‑hām) are conspicuous by their absence. Still, Coldingham and Tyninghame are, as Nicolaisen showed, each recorded from the late 11th century onward with a vowel between the g and the h and, in at least one case for either name, the vowel is a.27 Given this evidence it may seem

26 Nicolaisen 1976, 71–73 for the first three; for his views on Penninghame, see under that name, below. 27 See forms listed under ‘Coldingham’ and ‘Tyninghame’ below. Scotland’s -ham and -ingham Names: a reconsideration 109 perverse to question whether they are genuine examples of the pre-Christian ‑ing³ahām. Yet it needs to be emphasised that such names are concentrated in England in the areas of large-scale, primary Anglo-Saxon settlement in the east and south-east. Cox made this clear and Watts has expressed the view that there are probably no -ingahām names in County Durham.28 There are no such names in Northumberland and, indeed, the northernmost certain examples are probably in the region of early Deiran settlement, like Goodmanham YOE, Collingham YOW and Walkingham YOW.29 Cumberland has Addingham in the upper Eden Valley, Hensingham on the coast and Whicham in Copeland. Back in 1950 (before the studies of Smith, Dodgson and Cox on -ing formations) the English Place-Name Survey editors treated all these as -ingahām, but only in the case of Whicham do the early forms show evidence for a vowel between ‑ing- and -hām. Here, the development to Whicham implies a palatalised -ġ-, *Hwīt-inġī-hām, named after one Hwīta, identical to Whittingehame, and, as I shall argue below, that is certainly not an ‑ing³ahām.30 Addingham and Hensingham are much more likely to be -hām formations with -ing2- or -ing4-. There are no ‘‑ingham’ names of any kind in Westmorland. Whittingham in Lancashire, again named after a Hwīta, does seem to be an ‑ing³ahām: if so, it is the only reasonably certain example west of the Pennines.31 There are none in Cheshire (Dodgson 1997, 278–88) and it is indeed doubtful whether there are any in the English west midlands. So Nicolaisen’s view of these names does need to be questioned.

Coldingham Coldingham is, as Nicolaisen says,32 well documented, though there are lacunae in the historical and archaeological records. St Æbbe’s double house was known to Bede as Coludi Urbs, presumably Latinising Colodesburh as found in the Anglo-Saxon Chronicle (MS E).33 Colud was apparently a personal name, or

28 Watts 2002, xiii–xiv. He rejects Skerningham and Wolsingham – see his discussions of these, ibid., 113, 141 – and regards Billingham as doubtful; as his discussion (ibid., 8) shows, this was probably an ‑inġī-hām formation like Whittingehame, for which see below. 29 Though Collingham may be from a lost stream or hill name,*Colling, itself an -ing2 formation; compare A. H. Smith 1963, 174, with Watts 2004, 151, but note the archaeological evidence reported in Copley 1988, 47. 30 See under ‘Lyneringham and Whittingehame’, below. 31 Misplaced in Lincolnshire by Hough 2001, 100. It was only just in Lancashire, being on the west bank of the River Hodder, the old boundary with the West Riding of Yorkshire. The modern pronunciation with [-im] favours -ingahām, though early forms do not rule out yet another *Hwīt-inġī-hām. 32 Nicolaisen 1976, 20; see also ibid., 72–73. 33 Earl and Plummer 1892, vol. 1, 39, s.a. 679. 110 Alan G. James at least it was taken for such by Northumbrian English speakers.34 Urbs / burh probably refers to the pallisaded site on Kirk Hill, the summit of St Abbs Head. This may have been a fort – maybe more of a status symbol thana serious military installation – built before the time of the monastery, perhaps by the eponymous Colud, or it may have been a prestigious royal religious house from the start. Leslie Alcock’s excavations revealed probable evidence of a monastic site here, though a lot has tumbled into the sea in the past 1300 years,35 and, indeed, a number of religious houses in southern England had names with -burh.36 It is a magnificent location, in clear view of Lindisfarne and Bamburgh to the south and of Fife Ness to the north, but very exposed. The chief buildings in the monastic landscape may have been at the more sheltered location of Coldingham from the start or else they may have been re-located there after they were burnt down, soon after Æbbe’s death c. 686.37 Bede does not indicate that the house had been re-established in his time: it may not have been restored until well into the eighth century. According to the Aberdeen Breviary, when St Æbbe’s relics were found in 1188 by the Benedictines of the then newly-established Coldingham Priory, they were taken to the oratory (presumably of the earlier nunnery), which was then in ruins but which became the priory church.38 This implies that the nunnery was believed by the Benedictines to have been, at least by later Northumbrian times, on the site subsequently occupied by the priory.39

34 It is presumably Celtic, perhaps incorporating the honorative suffixuđ - seen in Welsh names like Maredudd. However, it is an oddity in the corpus of early Celtic personal names. A hypothetical Celtic hill-naming word, *colǖd (gender uncertain) might be derived from British

*colo:t-. < early Celtic *colout- < IE(NW) *kolh1- (o-grade of *kelh1- ‘rise, stand up’) + -t-, cf. Latin collis, Greek kolōnós and Germanic *ul-ni- > OE hyll > ‘hill’. 35 Alcock 1986, 255–79; see also Blair 1992, 228 (Fig. 10.1) and 259–61, and Alcock 2003, 183. 36 On the significance of burh in OE monastic names, see Blair 2005, 249–51, 269–70, Parsons and Styles 2000, 77, and Draper 2009, especially 104–07. It is worth noting that several known or possible ‘monastic’ -burh names appear to have OE personal names as their specifiers, e.g. Amesbury WLT, Avebury WLT, Charlbury OXF, Congresbury SOM, Ramsbury WLT, Shaftesbury DOR, Tilbury ESX, Tisbury WLT. Are these unknown monastic founders, or (like Colud, if that was a personal name) earlier laymen, even pagans, associated with these sites (several of which are, like Coldodesburh, at or close to iron-age forts)? Several double houses or nunneries have feminine names + -burh: these certainly or probably do commemorate foundresses, e.g Alberbury SHR, Alderbury WLT, Bibury GLO, Fladbury WOR, Heytesbury WLT, Tetbury GLO, but we do not have *Æbbesburh. 37 Historia Ecclesiastica IV 23[25] and Epistola ad Ecbertum 5, 9 and 10, in Plummer 1896, I, 262, 408–09, 412–14. 38 For a detailed discussion of the cult in the 12th century, see Bartlett 2003, xii–xxix. 39 A strap-end that might have been from an early mediaeval book-binding has been found there (see Alcock 1986). The only, somewhat doubtful, evidence for the history of the house Scotland’s -ham and -ingham Names: a reconsideration 111 Be all that as it may, the earliest records of the name Coldingham from the central middle ages do show -ingaham: ‘mansio de’ Collingaham 1095 × 1100 Durham Miscellaneous Charters 973, Coldingaham c. 1125 CDS.40 It would certainly seem to be *Colud-inga-hām ‘hām of the clan of Colud ’. So Nicolaisen assigns the name, as well as the foundation of the religious house, to ‘the second or third quarter of the seventh century’ (2001, 27). Nevertheless, there are grounds for doubt. Nicolaisen is disregarding both Cox’s emphatic association of -ingahām names with pre-Christian Anglo-Saxon settlement and the fact that Bede does not mention Coldingham, only Coludi Urbs. There is no evidence for the name Coldingham until around 1100. Can there be any other explanation? As has been mentioned, the plural of -ing³, ‑ingas, was used in naming religious communities.41 Watts in his discussion of -hām points out that, both at Hexham and at Lastingham YNR, it was used to re-name monasteries, perhaps as late as the eighth century, on the basis of an earlier name for the location (1994, 134–36). St Cedd’s monastery at Lastingham was known to Bede as *Laestinga-eu42 ‘dry land in a wet or marshy area43 of the [people known as] *Læstingas’. It was evidently re-named, presumably after Bede’s time, incorporating the early ‘clan’ name *Laest -ing³as but using it to name the monastic community. So *Laest-ing³-ahām was the ‘house and landholding of [the religious community known as] *Læstingas’. The underlying perception here and at Berkeley seems comparable to that seen in early Christian Ireland, where, when a monastery was established, the peasant- farmers and other men on its estate were enrolled, willy-nilly, as ‘members’ of the monastic community, being referred to as manaich ‘monks’ (Sharpe 1992, 102). Something similar could have happened at Coldingham. If we apply the

in the later Northumbrian period is Matthew Paris’s account in his Chronica Majora (trans. Anderson 1908, 61–62) of a Viking raid there in about 870, when St Æbbe the Younger and her companions took drastic measures to preserve their chastity. Matthew may have had access to a lost source from one of the northern monasteries. 40 Subsequent records mostly have -ingham: Coldingham c. 1130 ESC no. 90, Coldinghamsire 1173 × 1178 RRS ii no. 181, ecclesia de Koldingham c. 1250 St A. Lib. 31, but Goldingeham in the 13th-century Scalachronica cited by Nicolaisen does show a vowel between g and h. I am grateful to Dr Simon Taylor for information on these early forms, which add to those given by Nicolaisen 1976, 20–21 and 72–73. In the case of Collingaham, scribal confusion with Collingham YOW (mentioned above) seems unlikely, though not impossible. For a full discussion of the MS context of the late 11th-century charters containing this name, including Durham Miscellaneous Charters 973 (printed ESC no. 16), see Duncan 1999, with place- name forms from various related MSS usefully listed on p. 18. 41 See ‘-hām in monastic names’ above. 42 Historia Ecclesiastica III.23, in Plummer 1896, I, 174–75. 43 See Gelling and Cole 2002, 37–44. 112 Alan G. James analogy of Lastingham here and suppose that *Colud-inga-hām was not an early ‘ethnic’ -ing³a-hām but a name given to the later religious house, it may have been based in a similar way on *Coludingas as the name for the local populace and/or district.44 That would still have been an-ing 3 formation, though such an underlying form need not necessarily be earlier than the mid-seventh century, and the later form with -hām would not be directly comparable to the English ‘pagan’ -ing³a-hām names, being given quite possibly no earlier than the mid- eighth century.

Tyninghame45 At Tyninghame the monastery was founded some time after the death of St Baldred (recte probably Baldhere) in 756 and cropmarks near the village indicate a series of buildings comparable to the early Anglian ‘halls’ at Yeavering etc. beside a series of enclosures.46 Early recorded forms of the name include:47

in Tininghami Annals of Lindisfarne, s.a. 756, in the first scribal hand [Nicolaisen 1976, 72] Tinningaham c. 1050 Historia Sancti Cuthberti 4, p. 4 [MS c. 1150] Tiningehame 1094 ESC no. xii Tinningham c.1140 Taylor 2009, 606 (Augustinian Account) [18th c. copy] Tiningaham 1140 × 1148 Historia Regum s.a. 757 Tyningham 1153 × 1165 RRS i no. 219 Tiningeh’ 1178 × 1188 St A. Lib. 45 Tiningeham 1240 St A. Lib. 165 Tiningham 1240 St A. Lib. 166 Tiningeham 1240 St A. Lib. 168 Tynigham 1274 Bag. p. 33 Timingham 1275 Bag. p. 57 Timingeham 1275 Bag. p. 58 [repeat of preceding; the MS fo. 60v seems to read Tinungeham] Tinnyghame 1562 Assumptions 2

44 It could presuppose a former *Coludinga- plus an unknown generic, as in Læstinga-eu, which was replaced by -hām, but as we have seen, such formations seem otherwise absent from Scotland. 45 Nicolaisen 2001, 93–94, uses the spelling Tynninghame, which reflects the modern pronunciation; however Tyninghame is the current OS form. 46 See plan in Lowe 1999, 32. 47 I am very grateful to Dr Simon Taylor for these, which amend and expand the list given by Nicolaisen 2001, 93–94. Scotland’s -ham and -ingham Names: a reconsideration 113 Although what may be the earliest recorded form is in Tininghami,48 both Historia Sancti Cuthberti and Historia Regum show forms with -ingaham. In the latter two forms, the genitive plural ‑inga- does seem to be, as Nicolaisen puts it, ‘beautifully preserved’. He is right, too, to point out that, while -ingahām formations are normally based on personal names, a few in England are based on topographic names, as Tyninghame is based on that of the East Lothian Tyne. However, Avening in Gloucestershire, cited by Nicolaisen, seems to be the only other example of an -ing³ name apparently based on a river name.49 Perched on a steep hillside below the Cotswold scarp, Avening overlooks a minimal stream today lacking any name, Avon or otherwise, and is hardly likely to have been the basis of an ethnonym – so Avening is itself a problematic case. And, again, the remoteness of Tyninghame from any certain ‑ingahām names must give us pause.50 We may well have a similar case here to that at Coldingham. On the analogy of Lastingham, *Tīningas ‘dwellers on the Tyne’ may have been current as a local folk- and / or district-name, though, being anomalously based on the river name, the formation might have been an analogous coinage, invented only when the monastery was founded around the mid-eighth century.51 Either way, as at Coldingham, Tīn-inga-hām ‘the house and landholding of [the religious community known as] *Tīningas’ would be an -ing³a-hām name but need not be anything like so early as those of southern and eastern England.

Lyneringham and Whittingehame In considering Nicolaisen’s third supposed -ingahām, Whittingehame, we need to turn our attention to Smith’s -ing2.52 It is what I call the ‘naming after’ -ing: ‘A-ing- B’ is ‘B named after (a pre-existing) A’. A likely example of -ing2- with -hām is Lyneringham,53 a name recorded in a 15th-century document in the

48 The i- here is somewhat mysterious. It appears to be a Latin 3rd declension i-stem ablative. It can hardly be the OE locative. 49 Smith 1964, 86. Skerningham DRH, on the River Skerne, might seem a closer parallel, but Watts rejects it (2002, xiv and 113). 50 Beside the considerations that follow, it is possible that 12th-century scribes were influenced by the frequency of -ingVham place-names in Northumbria (from -inġīhām in the north, -ingahām in the south) in standardising their spellings of this name. 51 Cf. note 44, above. Again, *Tīninga- may have preceded some other generic (-eġ could have been appropriate here, as at Lastingham), but that too would have been a name of a type apparently absent from Scotland. 52 Discussed in detail by Smith 1956a, 285–90. 53 Alternatively, it might be an example of Smith’s -ing4, see above; what follows in this paragraph would remain equally valid. 114 Alan G. James parish of East Linton ELO:54 it would be the ‘hām named after Lyner’, Lyner being probably a lost Celtic stream name (cf. the river Lynher in Cornwall).55 Such a name could have been formed at any time during the currency of hām in these parts from the mid-sixth to early eighth centuries. Whittingehame lies a couple of miles to the south of East Linton, beyond Traprain. It is like those -ingham names in Northumberland which trip up the unwary by being pronounced -[indm].56 John Dodgson, in his important series of articles on Old English -ing,57 argued that such -[indm] names belong to the class that Smith had labelled -ing2 but that they preserve an archaic Germanic (indeed, Indo-European) locative inflexion, -ī, so that ‘X-inġī-hām would be a ‘hām at the place named after X’. Whittingehame is the ‘hām (here, the estate) at the place named after Hwīta’. Hwīta was a fairly common Old English personal name: there is a Whittingham pronounced [wit-indm] on the River Aln in Northumberland,58 and we have already encountered Whicham in Cumberland with the same origin, as well as Whittingham in Lancashire, a possible -ing³a- hām. How many different Angles named Hwīta were involved we have no way of knowing.59 That such names preserve an archaic feature of Old English suggests that they were formed pretty early. Dodgson reckoned that they probably originated in the pre-Christian period, though not necessarily in the very earliest phases of settlement. His attempts to square this with a distribution largely to the west and north of those parts where Anglo-Saxon pagan burials have been found have a whiff of special pleading. It is of interest in the Northumbrian context that the runic inscriptions on the Ruthwell Cross and the Franks Casket both contain examples of this locative form, but in both cases the noun is preceded by a preposition, implying that the locative sense of the inflection had become moribund. Dodgson regarded these as late seventh century but current scholarship places them in the first half of the eighth, again favouring a rather later date for the -inġī-hām names than he proposed, not necessarily pre-Christian.60

54 See Bain 1887 and W. Patterson, ‘East Lothian’, at – accessed 13/07/09. I am grateful to Mr Patterson for additional information about the location. 55 See my discussion of this in James 2010. 56 Bellingham, Ealingham, Edlingham, Eglingham, Ellingham, Eltringham, Ovingham and Whittingham. 57 Dodgson 1967a, 1967b and 1968; also of importance is Section I of Dodgson 1967c [1997]. 58 Whence, according to Historia de Sancto Cuthberto, the Northumbrian prince Guðfriþ was rescued from slavery (South 2002 §13). 59 On Whicham CMB, and also Whittingham LNC, see under ‘-ing³ ’ above. 60 See Page 1973, 66–68, 174–82, 188–89; Cassidy 1992, 19–20, 22–25; MacLean 1992; and Meyvaert 1992, 148–50, 164–65. Scotland’s -ham and -ingham Names: a reconsideration 115 Nicolaisen recognised the relevance of Dodgson’s findings to Whittingehame, but he nevertheless attempted to force that name into the -ing³a-hām class. After declaring that Whittingehame, Tyninghame and Coldingham are all ‘genuine -ingahām names’, he went on to discuss the [-indʒm] pronunciation of Whittingehame, declaring ‘it is therefore OE *Hwītingahām “settlement of Hwīta’s people”, or more likely *Hwītingiahām or *Hwītindahām “the settlement at Hwīting (= at the place named after Hwīta)”’ (2001, 93). This is most confusing. *Hwītindahām would be impossible in OE, as palatalized [ġ] cannot occur before a back vowel like [a], while *Hwītingiahām is a monstrosity, grafting the genitive plural of -ing3 on to the locative singular of -ing2 . The formation is *Hwīt-inġ2ī-hām, which is indeed ‘at the place named after Hwīta’, but is formed with the locative singular of -ing2. ‑ing³-ahām has nothing to do with Whittingehame, which is an early Anglian name, but not necessarily any earlier than other -hām names in Bernicia.

Tigbrethingham An ecclesiastical ‘-ingham’ presumably in south-east Scotland that has so far escaped the attention of place-name scholars is the enigmatic Tigbrethingham recorded in the list of possessions of Lindisfarne included in the Historia Regum.61 Its location is unknown, though, as the list appears to be geographically ordered, it may lie somewhere between Melrose and Abercorn.62 This solitary record gives no encouragement for another -inga-hām here; a formation with ‑ing²- or -ing4- is more likely. It is hard to make much of Tigbreth- as an Old English word, unless it is a miscopying (with -t- for -c-) of an Old English personal name ending -brecht, i.e. -ber(c)ht with metathesis. More possible might be -ing-hām (say -ing²-) added to a neo-Brittonic name-phrase, *Tī-vrech ‘a multi-coloured house’, adopted at a time when the final [ɣ] was sufficiently audible for Northumbrian English speakers to have treated it as palatalised [ġ], and when the lenited [] was still heard by them as [b]: the former would have been possible up to the early ninth century (Jackson 1953, 456–57 and 470), but the latter would have been extremely early in the settlement period, very unlikely in this region (ibid., 558 and 560–61).63 However, an alternative, very speculative, suggestion would be to read the core of the name as a version of *Bret-ing4-hām ‘an estate or, here, probably

61 In Arnold 1882, II.101, s.a. 854: see Anderson 1908, 60 note 8. 62 Alex Woolf’s suggested identification with Stow MLO has much to commend it (2007, 235). The only other published suggestion I have come across is C. Ralegh Radford’s tentative identification with Hoddom DMF (1953, 153), which seems very unlikely in a list of names otherwise all in the Bernician heartland. See also note 65, below. 63 But note that Peebles, if it is from *peïl, presents a similar problem; see Jackson 1953, 553. 116 Alan G. James a monastery, associated with Britons’.64 Tig- would then be a secondary, prefixed element. It looks like Middle Irish tigh ‘house’ but the problem with an early Gaelic modification is that, while this name appears in the listin Historia Regum, it is omitted (along with Abercorn and Edinburgh) from the corresponding list in Historia de Sancto Cuthberto. The list inHistoria Regum is taken to reflect conditions when the whole of Lothian and the Tweed basin was still under Northumbrian rule, the one in Historia de Sancto Cuthberto a stage (probably in the second quarter of the 11th century) when that was no longer the case (see Woolf 2007, 235). Therefore it is unlikely that the name is a post- Northumbrian, Gaelic adaptation, but it could be the neo-Brittonic cognate tī added by the early ninth century. So we might have a British monastery named *Bret-ing4-hām after the Britons by the Northumbrian Angles, then renamed by the Britons themselves *Tī-bretingham ‘house called Bretingham’, which name the English-speakers subsequently re-adopted as *Tībretinghām! A complicated tale but, in a context where Brittonic remained current, probably as the majority language, and Northumbrian English was a colonial intruder, such linguistic give-and-take would have been entirely possible.65 In any case, whatever its origin, Tigbrethingham should be noted as another ‘monastic’ -hām.

Edingham and Twynholm En route to Nicolaisen’s fourth ‘-ingham’, Penninghame WIG, we need to pause briefly at a couple of hām- s in the Stewartry of Kirkcudbright. Edingham, by Dalbeattie KCB, Nicolaisen wisely excluded from his group of possible -ing³a-hām names on the grounds that there are no reliable early forms for it (2001, 93, 95). Nevertheless, Daphne Brooke treated it as one and argued for the presence of an Anglian mynster (with perhaps a British predecessor) here (1987; also Brooke 1991, 301). We are yet further away from any genuine, early -ing³a-hām names or evidence for pagan Anglian settlement. If Brooke’s case for a mynster here can be sustained on historical or archaeological evidence, we might possibly have another ‘monastic’ -ing³a-hām of the Lastingham type, but the toponymic evidence is insufficient. As to the first element, it is probably an elided form of an Anglian name – or hypocorism – beginning Ead-, but the only thing we can be reasonably sure about is that Edingham is a likely

64 On OE Bret[t]- in place-names in the north (admittedly uncommon), see James 2008, 191–92. 65 Given that Abercorn appears as Eoriercorn in Historia Regum, the degree to which this name has been garbled can hardly be overestimated. I am grateful to Dr James Marple for the further suggestion that *Edringhām ‘religious house on the Adder’ could be involved and that this could have been an earlier name for Edrom (*Eder-hām) BWK, where Anglian sculpture again suggests a possible mynster. Scotland’s -ham and -ingham Names: a reconsideration 117 -hām name, witness to Anglian colonisation west of the Nith by the mid-eighth century. Nicolaisen 2001, 98–99, and Brooke 1991, 297, both consider Twynholm KCB to be another -hām, Brooke suggesting an Old English personal name Twicga as a possible specifier. This is perhaps favoured by one early form which she cites, Twignam 1154 × 1165.66 Other records include Twenham 1200 × 1206 and Tuinham 1287.67 The specifier might be Northumbrian Old English[be] * twēon- ‘between’68 or [ġe]twinn ‘double, two-fold’, but neither has any apparent application to the location of Twynholm. Nevertheless, the settlement occupies a central position in the West Stewartry; it is probably another -hām and possibly a mynster, a ‘monastic’ -hām.69

Penninghame Farther to the west, in Wigtownshire, Penninghame was declared by Nicolaisen in the first edition of Scottish Place-Names to be among the ‘non-genuine examples’ of ‑ingahām (1976, 72) and, indeed, that it only ‘may or may not contain the element hām’ (ibid., 76). However, in an additional note to the 1979 reprint (facing p. 1), he wrote:

‘P. 76 ... The spellings Tuinham (for Twynholm) and Peningham (for Penninghame) in Bagimond’s Roll of 1287 not only predate the evidence so far available by more than three centuries, they also strongly suggest that these two names are genuine examples of compound names in OE -hām and ‑ingahām, respectively. They therefore significantly strengthen our toponymic evidence for an early Anglian presence in Galloway.’ and in a footnote to the Introduction to the second edition (xx), he declared:

‘It is, however, necessary to revise the statement concerning Penning- hame WIG (p. 93) as the discovery of the spelling Peningham in 1287 in Bagimond’s Roll shows it to be a genuine -ingahām name and a witness to an early Anglian presence in Galloway. This is supported by other fifteenth-century spellings.’

66 Loc. cit., although this is absent from her database (as digitised by Simon Brooke, no longer available on-line, as at 16/04/2010); the earliest record there is Twenham. 67 Bagimond’s Roll, cited by Nicolaisen 1979, opposite p. 1; see below under Penninghame. 68 Cf. (æt) Tweoxn-eam (A-SC(A), s.a. 901) 10th c., (æt)Twynham (934) 17th c. (Gover 1961, 54), the former name of Christchurch HMP, being an elided form of West Saxon *[be]twēoxn- ēam ‘between rivers’. A Northumbrian form would have been *[be]twēon-ēam, but Twynholm is not ‘between rivers’. 69 But see James, ‘A note on the two Barloccos KCB’, forthcoming. 118 Alan G. James With all respect to a great place-name scholar, it shows nothing of the sort. There is no evidence in this 1287 form70 or any other early recorded form of the name for any vowel between -g- and -h-71 and, of course, Penninghame is about 170 miles from the nearest reasonably certain ing³a-hām name72 and even further from the nearest archaeological evidence for pre-Christian Anglian settlement. It is most unlikely to be -ing³a-hām. These early forms do, however, confirm that it is certainly a hām, probably a relatively early farming settlement and estate, though the possibility of an ecclesiastical connection here should be noted. Penninghame became a large mediaeval parish, with a moated grange (and possible residence) of the bishops of Whithorn at the central settlement.73 It may well have had an earlier close association with Northumbrian Whithorn: a monastic territory extending from the Isle of Whithorn up to the head of the Cree would have been a typical major landholding of the early mediaeval period. With this in mind, Carole Hough’s proposal, that it is not an ‑ingham of any kind, but rather *Pening-hām ‘a hām assessed at a penny’, seems an attractive one (2001). Early modern forms of the name quite often treat it as Penyhame (or similar) and the pennyland was, of course, an important unit of fiscal assessment in 12th–13th-century Galloway.74 Was Penninghame a hām assessed at a penny? We have seen that hām is an early Old English toponymic term. Even if it referred to a mynster, the formation would hardly be later than around 800, probably a good deal earlier. Coins were being minted under the Northumbrian kings in the eighth century and the word pening was used for the silver coin otherwise called sċeatta.75 However, the fact that such Northumbrian peningas went out of production by 790, along with the point that hām was obsolete as a habitative term by that date, implies a limited window of opportunity for such a formation in the early to mid-eighth century. But monetary valuation of land entails more than the presence of coins, it implies a fully-functioning monetary economy. How likely was that in north-west Northumbria in the mid-eighth century? The coins found at Whithorn, as at other high status sites, could well have been used only in transactions with visiting merchants,

70 Nicolaisen 2001, 99, adds Peningham 1287 and Penynghame 1425 to Penygham 1644, which was the only form given in Nicolaisen 1976. McQueen 2008, 44–45, adds Pennynghame 1532. 71 As was noted by Hough 2001. 72 Taking this to be Whittingham LNC (see note 31, above). 73 NX406608. See Hall 2006, 175–77, idem 2009, 59 and 64, and R. Smith 2009, 78. 74 See Oram 1987. 75 First minted in York in the late seventh century by Aldfrith; the main issues in the 8th century were under Eadberht and Athelwald Moll. Eight such coins, from c. 737–c. 780, were found during the excavations at Whithorn, see Pirie 1997. Scotland’s -ham and -ingham Names: a reconsideration 119 or as prestige gifts to local grandees and important visitors. There is really no evidence for the circulation of coins away from high-status sites in northern Northumbria / Cumbria, or indeed anywhere in ‘Scotland’, until around 1100.76 Whatever the origins of the concept, the name ‘pennyland’ seems more likely to be an early Scots formation of the 12th century.77 So we are left with the remote possibility that there may have been a short-lived, local monetary economy in the regio of Whithorn in the mid-eighth century. But if Penninghame were named then as *Pening-hām, it would have been revolutionary at the time and it would have remained wholly exceptional in north-west Northumbria for three or four centuries. An alternative, perhaps more likely, interpretation is to see Penninghame as an -ing2-hām formation like Lyneringham, a ‘hām named after (a pre-existing) Penn’.78 What was Penn? Hough is rightly doubtful about it (or Penna) as a personal name (2001, 101),79 but it could well be topographic. Penn is a very common and significant Brittonic place-name element, generally referring (in inland place-names) to the ‘head’ or ‘end’ of a high ridge or hill-spur (see Gelling and Cole 2000, 210–13). Is it appropriate to Penninghame? Such a feature might have been anywhere in the extensive parish and the name in -hām may have referred originally to the whole of the district.80 Nevertheless, the site of Penninghame old church is likely to have been a central place. It is on some of the best farmland in the parish, with the earthworks of the episcopal grange nearby, at the southern end of Barr Hill, a ridge ending in a sharp point. That feature might well have been named, or at least referred to, by Brittonic speakers as *Pen[n].81 Either way, English speakers would have taken that as its name and formed *Penn-ing2-hām ‘estate named after (the headland called) Penn’. If Penninghame was *Penn-ing2-hām, as a ‘secular’ hām name, it is unlikely to have been formed later than the mid-eighth century. The terminus post quem is more difficult to determine: the date at which The Machars peninsula was brought under Northumbrian rule is unknown, though archaeological evidence suggests that the monastic site at Whithorn may have come under

76 Scottish pennies were only introduced under David I in the 1130s. 77 See Oram 1987, also idem 2000, 237–38. 78 Alternatively, -ing4, ‘hām associated with Penn’. 79 It could perhaps be a hypocorism, even an English hypocorism of a Celtic name, but that is a rather desperate solution. 80 See note 5, above. 81 Viewed across Wigtown Bay from the hills to the east, Barr Hill has the classic ‘pen’ profile seen at Penyghent YWR and Pendle LNC. The old church site is at NX 412612. I am grateful to Mr B. Ditchburn for information about this location. 120 Alan G. James Northumbrian control near the end of the seventh century (Hill 1997,17, 37).82 The establishment of the Anglian episcopal see at Whithorn shortly before 731 was probably the culmination of a process of ‘anglicisation’ that had been under way for some time. A tentative but reasonable date for the name would be in the late seventh or early eighth century.

Cunninghame and Incuneningum Penninghame is not the only apparent ‘-ingham’ name in the Machars. On the River Bladnoch, about 3½ miles / 5km upstream from Wigtown, is a farm named Cunninghame.83 Little can be said about this name as early forms are lacking. It could be a transferred name from the better-known Cunninghame AYR or it could even have been named after a farmer with the surname Cunninghame, but its presence here raises a tantalising possibility which will become apparent as we examine the Ayrshire settlement and district name.84 Consideration of the latter has to begin with Bede’s Incuneningum, the regio ... Northanhymbrorum where the pious paterfamilias named Dryhthelm experienced an alarming prevision of the life to come.85 Since 1896, this has been identified with Cunninghame AYR.86 That the eponymous *Cunen- could have borne the Brittonic name *Cönan, Welsh Cynan, Irish Conán, has often been noticed,87 and the prominence of one Cynon of Aeron in the awdlau attributed to Aneirin has encouraged speculation on the assumption that Aeron was the Ayr basin. However, Nicolaisen was unhappy with this identification on historical grounds (2001, 91–92): there is indeed no documentary record of Northumbrian activity in these parts until 750, when Eadberht ‘added to his kingdom the plain of Cyil (Kyle), with other regiones’,88 and 756 when

82 If (and it is a big if) this was part of the British kingdom of Rheged (ruled from Dunragit WIG?), a date early in Oswy’s reign is plausible, i.e. 643 or soon after. See Smyth 1984, 24–26, and Fraser 2009, 176–79. Both Smyth and Fraser locate Rheged south of the Solway, west of the Pennines, on toponymic grounds that most present-day place-name scholars would find questionable at best. The case for Dunragit is at least as good, but this is not the place to pursue the debate. 83 Brooke 1991, 297, MacQueen 2008, 45–48. 84 See note 104, below. 85 HE V.12, in Plummer 1896, I, 304. 86 Plummer 1896, II, 295, citing a Mr Moberly, pers. comm. 87 Most recently by Woolf 2007, 4 note 4. Cynan has been derived from *Cunagnos, but note Sims-Williams’s view (2003, 160) that ‘there may never have been a British *Cunagnos ... Cynan may be either wholly Irish in origin or a hypocoristic form of a Welsh Cyn- name using a productive Irish-derived suffixán - ’. 88 Bædæ Continuatio, in Plummer 1896, I, 362, as translated in Anderson 1908, 56. See Campbell 1979, 48, on the significance of regio; regiones might be interpreted as ‘sub- kingdoms’. Scotland’s -ham and -ingham Names: a reconsideration 121 he led an ill-fated expedition to Alclud (Dumbarton) and Ovania (Govan).89 Nevertheless, it is quite conceivable that the rulers of Bamburgh at the height of their power in the seventh century, from the time of Æthelfrith’s triumph at Degsastan (c. 603) to Ecgfrith’s disaster at Nechtansmere / Dunnichen (685), held sway over the whole of Clydesdale south of the Antonine Wall as well as Inverclyde and the Clyde coast. Eadberht’s efforts in the eighth century may have been intended to recover what he saw as rightly his own. An early Anglian settlement here, perhaps a royal vill,90 would have made a great deal of sense, especially when Oswy (642–70) was at the height of his power. Guarding the entrance to the Clyde, facing the Cenél Comgaill lands of Cowal across the estuary and Bute down the water, and within sight of Alclud, citadel of the Britons of the Clyde, the strategic importance of this region cannot be overstated. However, there are problems for the toponymist. Incuneningum is, as Nicolasien points out, the dative plural of an -ing³ name, a name of the Reading type, and, as we have seen, there seem to be no definite cases of such a name in Scotland or in northernmost England.91 They are, as Dodgson’s studies (1966 and 1967b) show, characteristic of areas near to (though typically not actually in) the regions of primary Anglo-Saxon settlement. We might expect such a name in or close to Deira but Cunninghame is at the opposite extreme of Northumbrian territory.92 The certain, early recorded forms for Cunninghame could be derived from Incuneningum but they are open to other interpretations that may be more appropriate in the historical and geographical contexts:93

Cunegan 1131 × 1141 David I Chrs. no. 57 [‘probably 1136’; = Glas. Reg. no. 9]94 Conigham 1165 Glas. Reg. no. 24 [rubric; second n suspension mark. Superscript a]

89 The Northumbrian chronicle used by Symeon of Durham,Historia Regum, in Arnold 1882, I, 49. On the identification ofOvania as Govan, see Breeze 1999 and Forsyth 2000. 90 See Campbell 1979, 48, for the suggestion that -ingās and -ingahām names may have been royal vills. 91 See note 25, above. 92 There are serious difficulties with Nicolaisen’s alternative suggestion (2001, 92)ofan association with Cuncaceastre, Chester-le-Street DRH, which need not detain us here. Geographically, it is at least in the right direction. 93 I am most grateful to Dr Simon Taylor for these, which amend and greatly extend the list given by Nicolaisen 2001, 91. 94 Taylor notes (pers. comm.), ‘G.W. S. Barrow (ed.) is clear that both ‘Registrum Vetus’ and ‘Liber Ruber’ have Cunegan (David I Chrs. p. 81); the king grants to the church of St Kentigern of Glasgow ‘totam decimam meam de meo chan, in animalibus et porcis, de Stratgriva et Cunegan et de Chul et de Karric’ (in each year save when the king sojourns in those parts and eats his cain there).’ 122 Alan G. James decimam chan de Charri et de Chil, de Cunigham, de Stragrif, de Lerghes 1165 Glas. Reg. no. 24 Kunigham 1179 Glas. Reg. no. 51 [papal confirmation] Chunigham 1182 Glas. Reg. no. 57 [papal confirmation] in Cunygha1245 Kelso Lib. no. 281 [rubric; ecclesiam de Kilmaur’ in Cunygha] in Cunygha1245 Kelso Lib. no. 281 [ecclesiam de Kilmaur’ in Cunygha] tenura Willelmi de Cunygha1245 Kelso Lib. no. 2 in tenemento suo de Cunygham 1277 Glas. Reg. no. 230

If we leave Incuneningum aside and focus on the 12th- and 13th-century forms recorded above, we see a consistent final -ham, strongly suggesting that Cunninghame was in origin a -hām formation, not -ingum. It might be an -ing²-hām, but the frequent -ig- rather than -ing- needs to be noted: it can hardly be blamed on scribal omission of an abbreviation mark on every occasion. And it is difficult, without resorting to implausible hypocoristic forms, to identify any Old English common noun or personal name that could explain Cun-.95 A very interesting possibility, however, is cyning ‘king’. Formations with uninflected cyning qualifying an Old English habitative term are unusual but are to be found, for example Kennington KNT.96 The rounded vowel [ü] in the first syllable would have been unrounded to [i] in Northumbrian Old English after about 800.97 If the name had been adopted by speakers of early Cumbric, they may have treated it as [u] or [i] (there was no short [] in neo-Brittonic), reducing it to [] in low- stressed position.98 But if it had been adopted at an early date by Goidelic speakers (which is, of course, entirely possible here), they would probably . . have treated [ni] as [ui], with [u] followed by a ‘slender’ [n]. Gaelic influence might also account for the reduction of [g] to [g], the -ig forms reflecting a Gaelicised form like *Cuiníg-ham.99 When Gaelic superseded both Northumbrian Old English and Cumbric as the language of these

95 We might compare Kingham OXF, Gelling 1954, 360, an -ing4- formation *Cæ-ġ-inga-hām, but, for reasons already presented, any such formation seems very unlikely here. 96 Ekwall 1960, 272 s.n.; the two Kinetons, GLO and WAR, have early forms varying between cyng ‘king’ and cyne ‘royal’, see Smith 1964b,14 and Gover et al. 1936, 282. However, it has to be admitted that there seem to be no definite cases in England of cyning (inflected or not) with the generic -hām. 97 Campbell 1959, 15 and 132. 98 Jackson 1953, 274–78, 302–03 and 664–66. 99 Compare Custantín < Constantin- (Simon Taylor, pers. comm.). Scotland’s -ham and -ingham Names: a reconsideration 123 parts, the Gaelic-influenced form of the name would have prevailed, being then taken into Scots with ‑inghame eventually restored on the analogy of other names with that ending.

Eaglesham In view of its strategic location, already mentioned, a Northumbrian ‘king’s estate’ here would itself be of considerable historical significance. But Cunninghame cannot be considered without attention to its neighbour across the hills, Eaglesham RNF. This ‑hām,100 overlooking what must have been major centres of royal and church power in the kingdom of Alclut – Paisley, Renfrew and Govan – is very likely to be *eclēs-hām, incorporating the anglicised form of Brittonic *eglẹ:s, from British Latin *eclesia ‘the Church as an organised community, either locally or universally’.101 The importance of this element as evidence of British ecclesiastical presence at the time of Anglo-Saxon settlement or colonisation has been recognised since the seminal works of Kenneth Cameron and Geoffrey Barrow,102 but Carole Hough has recently drawn attention to the important point that, as *eclēs occurs in English place-names only in simplex forms or as a specifier, we cannot assume that it was understood or regarded by English-speakers as anything but a place-name: *eclēs-hām would be ‘*hām named “Eclēs” ’ (2009, 109–24). In the same publication, I have argued that, in naming or speaking of places, Brittonic *eglẹ:s meant primarily a ‘church landholding’ and that, for the incoming English-speakers, whether or not they understood (or cared about) its ecclesiastical associations, an *eclēs was first and foremost a desirable piece of real estate (James 2009, 126–30). Still, it is very unlikely that Eaglesham was colonised by pagan Angles under Æthelfrith and, in this location, there is a reasonable possibility that a British church estate was transferred to English ecclesiastical control. So it may be another ‘monastic’ -hām ‘estate of [the monastic community at] Eclēs’.103 *Eclēs-hām and *Cyning-hām, ‘church estate’ and ‘king’s estate’? They

100 As is confirmed by the early forms Egglesham 1161 RRS i no. 184 [16–17 c. copy; part of the grant made to Walter FitzAlan (Steward) by David I, confirmed by Malcolm IV], and ‘ecclesiam parochialem de Eglishame’ 1427 × 1433 Glas. Reg. ii no. 340 [also Eglisham]. Again, I am most grateful to Dr Taylor for these details. 101 Thomas 1981, 148: only doubtfully ‘a church building’; see also James 2009a, 126–27. 102 Cameron 1968, 87–92, Barrow 1973 [2003], 7–39, and idem 1983. 103 See James 2009, 137–39, where I discuss Eaglesham in greater detail; see also Hough 2009, 121 note 16. Although we take different standpoints on the historical background of *eglẹ:s / *eclēs, either Hough’s position or mine could be consistent with the view expressed above. 124 Alan G. James make a tempting pair in this crucial location!104 Still, we must acknowledge that this interpretation of Cunninghame is very speculative, and even that of Eaglesham is not beyond challenge. What is pretty certain, and itself very significant, is that they are both -hām names, evidence of control and colonisation of this area at some stage during the Northumbrian period. Two historical contexts are most plausible. One would be, as already suggested, during the climax of Bernician imperium in the mid-seventh century. At this time, hām would certainly have been current as the appropriate term for newly- acquired landholdings. The other would be in the period following Eadberht’s (re-)annexation of Kyle and other regiones (750). By this time, hām was going out of use in the south, as the kind of territories to which it referred were beginning to give way to increasingly specialised, self-contained and directly- taxed holdings. However, such developments were yet to reach the north and, for lands lately conquered or re-conquered, hām would probably still have been the preferred label. We do not know how permanent Eadberht’s (re-)conquests proved, though the scatter of -wīc names in AYR and RNF, especially Prestwick AYR, suggests that English-speaking rule was still in force when large estates were being reorganised, probably well into the ninth century.105

Conclusion All the -hām names in southern Scotland are of great interest as traces of the expansion of English-speaking colonisation under Northumbrian rule, but Nicolaisen’s attempt to associate the -ingham names among them with the pre-Christian ‑ing³a-hām names in southern and eastern England has been an unfortunate distraction. The unlikelihood of there being any such names in Scotland needs to be emphasised. The ingham- names north of the Border are not necessarily Scotland’s earliest English place-names. Coldingham and Tyninghame in particular are anomalous: -ing³a-hām in form, but probably of

104 And, returning to Cunninghame WIG, the possibility of a royal estate on the Bladnoch, close – but not too close – to Whithorn is interesting: as at Cunninghame and Eaglesham, it would be consistent with the frequent juxtapositon of royal and ecclesiastical centres of power in the early middle ages. 105 See fuller discussion in James 2009, 138–39, and, for the dating of Prestwick compare note 17 on Preston ELO, above. Mr Alex Maxwell Findlater (pers. comm.) has suggested that Martnaham AYR (Dalrymple) might be an -ingham name. Early forms (Longmertenock 1302 × 1304 CDS ii, 1608, Mertnem 1465 RMS ii no. 841 etc.) do not favour this, though it may be another -hām. OE mearþ ‘a marten’ might be involved but note that ‘marten’ is of 15th-century origin in English, replacing earlier ME martre (from OF, cf. Scots mertrick), which itself had replaced mearþ (OED). Alternatively, the very common OE compound mere- tūn, ‘lake-farm’, would be topographically appropriate, perhaps in the dative plural *[æt] mere- tūnum (see Smith 1956b, 38–39). Scotland’s -ham and -ingham Names: a reconsideration 125 a different and later origin than the pre-Christian ‘ethnic’ names of the south. And the historical salience of these two monastic names should not overshadow the significance of the otherham - and -ingham names in the Bernician heartland. Some of these might date from the latest pre-Christian period, the reign of Æthelfrith (592–616), but the era of Northumbrian state-formation under Edwin (616–33), Oswald (633–42), Oswy (642–70) and Ecgfrith (670–85) seems the likeliest context for the important cluster of -hams in the lower Tweed basin and the Merse – the historical significance of these deserves greater recognition. The more dispersed ‘secular’ -hāms around the North Sea coast and north of the Lammermuirs can be seen as evidence of less intensive Anglian colonisation in the same period. This would apply to Oldhamstocks, Lyneringham and Whittingehame – the last, being formed with the archaic locative -inġī-hām, is of particular linguistic interest, but it need be no earlier than the mid-seventh century. The second main theme of my argument has been to draw attention to the use of hām in (re-)naming religious houses. I suggest that this, rather than the ‘pagan’ -ing³ahām formations of southern and eastern England, may better explain Coldingham and Tyninghame. In these two cases, (re-)naming with -ingahām may be associated with their (re-)foundation in the mid-eighth century – such ‘monastic’ hām names need not be as early as the ‘secular’ ones. Norham, likewise, may represent ‘monastic’ re-naming, even as late as the mid- ninth century. On the other hand, Morham, Auldhame and lost Pefferhame (and possibly Brigham and Tigbrethingham) may also be ‘monastic’ hām names, though in these cases we cannot know whether they were later in origin than their ‘secular’ neighbours. The context in which all the ‘-ham’ and ‘‑ingham’ place-names of south-east Scotland are best understood is one in which the development of the Anglian Church was closely interlinked with the processes of secular state-formation. To the west, apparent -hām and -ingahām place-names are few and far between and their formations, especially their specifiers, are in most cases highly debatable. Nevertheless, the very fact that there are -hām names in Galloway and even overlooking the mouth of the Clyde is of considerable historical interest. Edingham, Twynholm, Penninghame, the two Cunninghames and Eaglesham are all witnesses to the extension of Northumbrian imperium up to the mid- eighth century. While none of these can be definitely identified as ‘monastic’ hāms, it is a very tentative possibility at Edingham and Twynholm, while ecclesiastical involvement in the naming process is possible at Penninghame and quite likely at Eaglesham. The Cunninghames, especially the northern one (being a district name), offer the tantalising possibility of royalhāms . So, turning our attention away from the probably futile quest for pre- 126 Alan G. James Christian English place-names to a consideration of the light that elements known to have been current in pre-viking northern England may cast on the territorial politics of the growing Northumbrian state, and on the closely- associated development of the Northumbrian Church, would seem a fruitful change of direction in the study of the Old English place-names of southern Scotland.106 There is, of course, a pressing need for full place-name surveys of the counties of the region comparable to the almost completed volumes on the place-names of Fife.107 But, even at the present time, increased accessibility of reliable transcriptions of early forms, the steadily growing body of archaeological data and the recent eruption of interest among historians in the post-Roman era offer an exciting prospect to those with a knowledge of the topography and an enthusiasm for the history of the region and who are in touch with current English place-name scholarship but willing to question its automatic applicability to names in ‘Anglo-Saxon Scotland’.

References Alcock, L., et alii, 1986, ‘Reconnaissance excavations on early historic fortifications and other royal sites in Scotland, 1974–84: 1, Excavations near St Abb’s Head Berwickshire, 1980’, PSAS 116, 255–79. Alcock, L., 2003, Kings, Warriors, Craftsmen and Priests in Northern Britain AD 550– 850, Society of Antiquaries Monograph Series (London). Anderson, A. O., 1908, Scottish Annals from English Chroniclers (London). Arnold, T., ed., 1882, Symeonis Dunelmensis Opera Omnia, Rolls Series (London). Bain, J., 1887, ‘Notes on the Trinitarian or Red Friars in Scotland, and on a recently discovered charter of Alexander III confirming the foundation of Houston by Cristiana Fraser, widow of Sir Roger de Moubray’, PSAS 12, 28–32. Barrow, G.W. S., 1973, The Kingdom of the Scots (London; 2nd edn, Edinburgh 2003). Barrow, G.W. S., 1983, ‘The childhood of Scottish Christianity: a note on some place- name evidence’, Scottish Studies 27, 1–15. Barrow, G.W. S., 1998, ‘The Uses of Place-names and Scottish History – Pointers and Pitfalls’, in S. Taylor, ed., The Uses of Place-Names (Edinburgh), 54–74. Bartlett, R., ed. and trans., 2003, The Miracles of St Æbbe of Coldingham and St

106 Elements that merit similar reconsideration include the habitative terms burh, boðl, wīċ and worþ (the first two discussed in Barrow 1998, 67–69, with distribution maps, the last three in Nicolaisen 2001, 100–8), as well as Old English topographical and land-use terms, especially in locations where there is historical or archaeological evidence for Northumbrian- period settlement. 107 See, for example, Taylor 2009. The only county surveys for southern Scotland comparable in scope to those of the earlier volumes of the English Place-Name Survey – i.e. substantial, but far from comprehensive – are Macdonald 1941 and Dixon 1947. Scotland’s -ham and -ingham Names: a reconsideration 127 Margaret of Scotland (Oxford). Blair, J., 1992, ‘Anglo-Saxon minsters: a topographical review’, in Blair and Sharpe 1992, 226–66. Blair, J., and R. Sharpe, eds, 1992, Pastoral Care before the Parish (Leicester). Blair, J., 2005, The Church in Anglo-Saxon Society (Oxford). Bonner, G., D. Rollason and C. Stancliffe, eds, 1989,St Cuthbert (Woodbridge). Breeze, A., 1999, ‘Simeon of Durham’s Annal for 756 and Govan, Scotland’, Nomina 22, 133–36. Brooke, D., 1987, ‘The deanery of Desnes Cro and the church of Edingham’, TDGNHAS 62, 48–65. Brooke, D., 1991, ‘The Northumbrian settlements in Galloway and Carrick: an historical assessment’, PSAS 121, 295–327. Cambridge, E., 1989, ‘Why did the Community of St Cuthbert settle at Chester-le- Street?’ in Bonner et al. 1989, 367–86. Cameron K., 1968, ‘Eccles in English Place-Names’, in M.W. Barley and R. P. C. Hanson, eds, Christianity in Britain 300–700 (Leicester), 87–92, repr. in Cameron 1977, 1–7. Cameron, K., ed., 1977, Place-Name Evidence for the Anglo-Saxon Invasion and Scandinavian Settlements (Nottingham). Campbell, A., 1959, Old English Grammar (Oxford). Cassidy, B., 1992, ed., The Ruthwell Cross (Princeton NJ). CDS: Calendar of Documents relating to Scotland, ed. J. Bain et al., 5 vols (Edinburgh 1881–1986). Copley G., 1988, Early Place-Names of the Anglian Regions of England BAR British Series 185 (Oxford). Cox, B., 1973, ‘The Significance of the Distribution of English Place-Names inHām in the Midlands and East Anglia’, Journal of the English Place-Name Society 5, 15– 73, repr. in Cameron 1977, 55–98. Cox, B., 1976, ‘The Place-Names of the Earliest English Records’, Journal of the English Place-Name Society 8, 12–66. Cramp, R., 1989, ‘The Artistic Influence of Lindisfarne within Northumbria’, in Bonner et al. 1989, 213–28. Crumplin, S., 2004, Rewriting History in the Cult of St Cuthbert from the Ninth to the Twelfth Centuries, PhD thesis, University of St Andrews, on-line at . David I Chrs.: The Charters of David I: the Written Acts of David I King of Scots, 1124– 53, and of His Son Henry, Earl of Northumberland, 1139–52, ed. G.W. S. Barrow (Woodbridge 1999). Dixon, N., 1947, Place-Names of Midlothian, PhD thesis, University of Edinburgh, on-line at . Dodgson, J. M., 1966, ‘The Significance of Place-Names iningas - , -inga- in South- 128 Alan G. James East England’, Mediaeval Archaeology 10, 1–29, repr. in Cameron 1977, 27–54. Dodgson, J. M., 1967a, ‘The ing- - in English place-names like Birmingham and Altrincham’, Beiträge zur Namenforschung NF2, 221–45. Dodgson, J. M., 1967b, ‘Various forms of OE -ing in English place-names’, Beiträge zur Namenforschung NF2, 325–96. Dodgson, J. M., 1967c, ‘The English Arrival in Cheshire’,Transactions of the Historical Society of Lancashire and Cheshire 119, 1–37, repr. with addenda and corrigenda in Dodgson 1997, 271–88 and notes at p. 306. Dodgson, J. M., 1968, ‘Various English place-name formations containing Old English -ing’, Beiträge zur Namenforschung NF3, 141–89. Dodgson, J. M., completed and ed. by A. R. Rumble, 1997, The Place-names of Cheshire: 5.2 EPNS LXXIV (Nottingham). Draper, S., 2009, ‘Burh names in Anglo-Saxon England’, Journal of the English Place- name Society 41, 103–18. Duncan, A. A. M., 1999, ‘Yes, The Earliest Scottish Charters’, Scottish Historical Review 78 (1), 1–38. Earl, J., and C. Plummer, eds., 1892, Two of the Anglo-Saxon Chronicles Parallel (Oxford). Ekwall, E., 1960, Concise Oxford Dictionary of English Place-Names (4th edn, Oxford). ESC: Early Scottish Charters prior to 1153, ed. A. C. Lawrie (Glasgow 1905). Fellows-Jensen, G., 1990, ‘Place-Names as a Reflection of Cultural Interaction’, Anglo-Saxon England 19, 13–21. Foot, S., 2006, Monastic Life in Anglo-Saxon England c. 600–900 (Cambridge). Forsyth, K., 2000, ‘Evidence of a lost Pictish source in the Historia Regum Anglorum of Symeon of Durham’, with appendix by J. T. Koch, in Taylor 2000, 19–34. Fraser, I., 1995, ‘Border Place-Names’, in Oman 1995, 179–92. Fraser, J. E., 2009, From Caledonia to Pictland: Scotland to 795 (Edinburgh). Gelling, M., 1954, The Place-names of Oxfordshire: 2 EPNS XXIV (Cambridge). Gelling, M., 1984, Place-Names in the Landscape (London). Gelling, M., and A. Cole, 2000, The Landscape of Place-Names(Stamford). Glas. Reg.: Registrum Episcopatus Glasguensis (Bannatyne and Maitland Clubs 1843). Gover, J. E. B., A. Mawer and F. M. Stenton, in collaboration with F. S. Houghton, 1936, The Place-Names of Warwickshire EPNS XIII (Cambridge). Gover, J. E. B., 1961, The Place-Names of Hampshire, typescript in the care of the English Place-Name Society. Hall, D., 2006, Scottish Monastic Landscapes (Stroud). Hall, D., 2009, ‘“A loop in the Forth is worth an Earldom in the North”: the rediscovery of Scotland’s monastic landscapes’, TDGNHAS 83, 55–65. Hill, P., 1997, Whithorn and St Ninian: excavation of a monastic town 1984–91 (Stroud). Hope-Taylor, B., 1977, Yeavering: an Anglo-British Centre of Early Northumbria (London). Hough, C., 1997, ‘The Earliest Old English Place-Names in Scotland’, Notes and Scotland’s -ham and -ingham Names: a reconsideration 129 Queries 44(2), 148–50. Hough, C., 2001, ‘The Place-Name Penninghame (Wigtownshire)’, Notes and Queries 48(2), 99–102. Hough, C., 2009, ‘Eccles in English and Scottish Place-Names’, in Quinton 2009, 109–124. Jackson, K. H., 1953, Language and History in Early Britain: a chronological survey of the Brittonic languages 1st to 12th c. AD (Edinburgh). James, A. G., 2008, ‘A Cumbric Diaspora?’, in Padel and Parsons 2008, 187–203. James, A. G., 2009, ‘*Eglēs /Eclēs and the formation of Northumbria’, in Quinton 2009, 125–50. ․ James, A. G., 2010, ‘A Slippery Customer: PrIE *slei- and its descendants in Celtic Place-names in Britain’, Nomina 33, forthcoming. Kelso Lib.: Liber S. Marie de Calchou (Bannatyne Club 1846). Lowe, C., 1999, Angels, Fools and Tyrants (Edinburgh). Macdonald, A., 1941, The Place-Names of West Lothian (Edinburgh and London). Maclean, D., 1992, ‘The date of the Ruthwell Cross’, in Cassidy 1992, 71–94. Macgregor, L. J., and B. E. Crawford, eds., 1987, Ouncelands and Pennylands (St Andrews). MacQueen, J., 2008, Place-Names of the Wigtownshire Moors and Machars (Stranraer). Metcalfe, D. M., ed., 1987, Coinage in Ninth Century Northumbria, BAR British Series 180 (Oxford). Meyvaert, P., 1992, ‘A new pespective on the Ruthwell Cross: Ecclesia and Vita Monastica’, in Cassidy 1992, 95–166. Nicolaisen, W. F. H., 1976, Scottish Place-Names: their study and significance (London; 2nd impr. with add. information 1979; new edn with minor changes, Edinburgh 2001). Oman, D., ed., 1995, The Borders Book (Edinburgh). Oram, R., 1987, ‘Davachs and Pennylands in South-West Scotland: a review of the evidence’, in Macgregor and Crawford 1987 46–59. Oram, R., 2000, The Lordship of Galloway (Edinburgh). Padel, O. J., and D. N. Parsons, eds, 2008, A Commodity of Good Names: essays in honour of Margaret Gelling (Donnington). Page, R. I., 1973, An Introduction to English Runes (London). Parsons, D. N., and T. Styles, 2000, The Vocabulary of English Place-Names: Brace to Cæster (Nottingham). Pickles, T., 2009, ‘Biscopes-tūn, muneca-tūn and prēosta-tūn: dating, significance and distribution’, in Quinton 2009, 39–107. Pirie, E. J. E., 1996, Coins of the Kingdom of Northumbria, c. 700–867 (Llanfyllin). Pirie, E. J. E., 1997, ‘The early mediaeval coins’, in Hill 1997, 332–45. Plummer, C., ed., 1896, Venerabilis Baedae Opera Historica (Oxford). PSAS: Proceedings of the Society of Antiquaries of Scotland. Quinton, E., ed., 2009, The Church in English Place-Names (Nottingham). 130 Alan G. James Ralegh Radford, C., 1953, ‘Hoddom’, Antiquity 27: 107, 153–60. Rollason, D., ed., 2000, Libellus de exordio atque procursu istius, hoc est Dunhelmensis, ecclesie = Tract on the origins and progress of this the Church of Durham / Symeon of Durham (Oxford). RRS i: Regesta Regum Scottorum vol. i (Acts of Malcolm IV ), ed. G.W. S. Barrow (Edin- burgh 1960). RRS ii: Regesta Regum Scottorum vol. ii (Acts of William I ), ed. G.W. S. Barrow (Edin- burgh 1971). Sharpe, R., 1992, ‘Churches and communities in early mediaeval Ireland’, in Blair and Sharpe 1992, 81–109. Sims-Williams, P., 2003, The Celtic Inscriptions of Britain: phonology and chronology, c. 400–1200 (Oxford and Boston MA). Smith, A. H., 1956a, English Place-Name Elements part one, EPNS XXV (Cambridge). Smith, A. H., 1956b, English Place-Name Elements part two, EPNS XXVI (Cam- bridge). Smith, A. H., 1963, The Place-Names of the West Riding of Yorkshire: 4 EPNS XXXIII (Cambridge). Smith, A. H., 1964a, The Place-Names of Gloucestershire: 1 EPNS XXXVIII (Cam- bridge). Smith, A. H., 1964b, The Place-Names of Gloucestershire: 2EPNS XXXIX (Cambridge). Smith, R., 2009, ‘Evidences for “Lost” Thirteenth Century Enclosure/Courtyard Castles in South-West Scotland: and overview and assessment’, TDGNHAS 83, 66–79. Smyth, A. P., 1984, Warlords and Holy Men (London). St A. Lib.: Liber Cartarum Prioratus Sancti Andree in Scotia (Bannatyne Club 1841). South, T. J., ed., 2002, Historia de Sancto Cuthberto: A History of Saint Cuthbert and a Record of His Patrimony, Anglo-Saxon Texts No. 3 (Cambridge). Taylor, S., ed., 2000, Kings, Clerics and Chronicles in Scotland, 500–1297 (Dublin). Taylor, S., with G. Márkus, 2009, Place-Names of Fife Vol. 3 (St Andrews and the East Neuk) (Donington). Thomas, [A.]C., 1981,Christianity in Roman Britain to AD 500 (London). TDGNHAS: Transactions of the Dumfries and Galloway Natural History and Antiquarian Society. Watts, V., 1994, ‘The Place-Name Hexham: a mainly philological approach’,Nomina 17, 119–36. Watts, V., 2002, A Dictionary of County Durham Place-Names (Nottingham). Watts, V., 2004, The Cambridge Dictionary of English Place-Names (Cambridge). Woolf, A., 2007, From Pictland to Alba 789–1070 (Edinburgh). 0 10 20 40 60 M

0 10 20 40 60 80 100Km

The Shadow of ‘Onomastic Graffiti’

Denis Rixson Mallaig

‘I am afraid the shadow of “onomastic graffiti” still hangs over Scottish toponymy’. On several occasions I have heard this rueful comment, or one similar, fall from the lips of Simon Taylor (see also Taylor 2001, 112–13). He was referring to a famous phrase used by Professor W. F. H. Nicolaisen in describing Scandinavian place-names on the west coast. This article is an attempt to dispel that shadow. It is based on research into land-assessment in the West Highlands and Islands. Although Professor Nicolaisen was referring particularly to Arran, his remarks extended to mainland Argyll. Accordingly I will confine my examples to the area between Cowal and Glenelg, forsaking all the Hebrides. I will not attempt to include the mainland north of Glenelg, although I will refer to research in that area. The phrase ‘onomastic graffiti’ appeared in an article by Nicolaisen called ‘Arran Place Names – A Fresh Look’, in Northern Studies 28, 1992. He writes (pp. 7–8):

‘Although many of the names of Scandinavian origin are now names of farms and villages, none of them started out as such. They are all names of coastal features or of features easily seen or reached from the coast, almost within sight of their boats. These include two bay names, ... eight names of small river valleys or ravines ... and three river names. ... The only ex- ception is Goat Fell, the most prominent mountain on the island, easily seen from the sea and the dominating feature in the island landscape even for non-resident visitors, for this is what the inventory of Scandinavian place names represents – not the nomenclature of a settled people but of occasional, albeit fairly regular but not always very welcome, visitors. It is a nomenclature that experiences the island from the sea, not only visually but also while exploring and utilising it. It is a sailor’s toponymic vocabu- lary and that of the fisherman and the hunter and the herdsman involved in transhumance. It is very similar to the Scandinavian nomenclature which one can distil out of the toponymy of the west coast of the Scottish mainland, especially Argyll. For the island dwellers of the Hebrides, if the names they left behind are anything to go by, Arran was experienced more as part of the mainland to the east than of their own insular kingdom. Not one Norse name indicates that its giver ever stayed on during winter or had any permanent dwelling in the island ...

The Journal of Scottish Name Studies 4, 2010, 131–58 132 Denis Rixson ‘Thus the Norse place-names of Arran are a nomenclature of outsiders who, like the water, greedily encroached upon the island, snatching at it at every high tide and being beaten back by every ebb. ... they are the names of seasonal intruders depleting the rivers and grazing their heif- ers and their yearlings on shielings on the best grassland easily acces- sible from the shore. These names are more like onomastic graffiti ...’

Taking this as our basis we can characterise minimal and maximal positions for Norse place-names on the west coast. The minimalist position is as set out by Professor Nicolaisen above (cf. also 1969, 16). He has been supported by Fraser (1978, 25 and 1985, 176). It has an unassailable logical simplicity; circularity even. Certain place-names can be associated, by their very nature, with settlement. These will include habitative elements like ON ból, bólstaðr, staðr, býr (all names for ‘farm’). If we find such, then we know there was settlement. Thus far, incontrovertible. The deduction then drawn is that if we don’t find them we cannot prove settlement – in fact, the minimalist would argue, there is a presumption against settlement. This deduction is at least questionable. Absence of habitative place-names does not, of itself, prove absence of settlement but it is a strong argument for doubting it. Ranged against this self-contained and self-fulfilling hypothesis are a whole host of contrary opinions arguing, to a greater or lesser degree, for what we might call a maximalist position. By contrast with the minimalists it is not possible to define a single maximal position since various types of supporting evidence have been summoned by individual proponents. However, we can set out some of the lines of argument employed:

(a) Arguments by analogy: topographical names are used in different parts of the Scandinavian world to indicate primary settlement units. In other colonies they are demonstrably the names of some of the earliest farms. (b) Arguments by density: the sheer quantity of Norse names in parts of the mainland like Kintyre strongly suggests settlement. Such names are mutually supportive. (c) Arguments of opposition: it is nonsensical to suggest that names like Lussa (rivers in Kintyre and Knapdale) or Smirisary, Assary and Ulgary (shielings in Moidart) would ever embed amongst a Gaelic-speaking population unless there was significant Norse settlement.

There are two aspects to this controversy: one logical, the other onomastic. Let us deal with the logical issue first. Unfortunately, the way the minimalist The Shadow of ‘Onomastic Graffiti’ 133 argument is constructed, no onomastic evidence is allowed as relevant unless it has to do with habitative names. Anything topographical can immediately be dismissed as peripheral. So whatever arguments the maximalists may muster; however many dale-names they collect; however rich the analogies with Iceland or the Faeroes; whatever the attractions of user-group theory; whatever the common-sense appeal of their position; they cannot dent the rock-hard simplicity of the minimalist position. In logical terms it is a pre-emptive strike: no habitative names, no settlement. This has not stopped the maximalists from trying; they still fire off the occasional empirical missile in the direction of the minimalists, only to see it fall short of persuasive force or be deflected by logical inviolability. As Simon Taylor observes, the shadow still falls. What can be done about this? There are really only two options. One is to try and accumulate such a weight of onomastic evidence for the maximalist position that the minimalists are somehow borne down or discredited. Historically this seems to have been the line favoured by many researchers. In this case, however, the argument is never won, in logical terms, and the shadow will always threaten. The other is to tackle our opponents on their own ground. Can we find sufficient habitative names to undermine the minimalist position from within? However, in order to do this we must forego, completely and consistently, any of the usual maximalist arguments. Out must go at least 23 dale-names in Kintyre; out go the Norse personal names in mainland farms; out go the butter-shielings (e.g. Smirisary), the inland waterfalls (e.g. Glenforslan) and salmon rivers (e.g. Lussa); out go the islands of Kerrera, Bernera, Eriska, Ramsay and Pladda in Loch Linnhe. We must abandon every analogy; abjure all user-group theory; we must only use terms that the minimalists would recognise as their own; in effect we must turn their own weapons against them. In doing this, there are two separate sets of evidence we can deploy. Firstly we have the habitative place-name elements which the minimalists have already conceded. These includeból, bólstaðr, staðr and býr where what is being described is basically a farming unit. There are some names on the mainland which already fit into this category. I have mapped these (see Map) and detailed them individually in the Appendix, along with some more doubtful possibilities. Taken together these make the beginnings of a case for Norse settlement in mainland Argyll. Secondly I would argue that we can fairly extend this list. There are other words the Norse used which imply habitation by their context or syntax, even if they are not words which indicate habitation by their meaning or semantics. (For instance, it might be hard to imagine them occurring without habitation. They might appear in clusters or in association with a habitative name,as around Loch Morar where we find Scamadale, Swordland and Meoble.) As a 134 Denis Rixson Map: Norse settlement names on the mainland west coast (Cowal to Glenelg) Norse settlement names on the mainland west coast. (Cowal to Glenelg)

20

21 2 1 22 3

16 12

15 23 4

8

9

19 ??

5

13 18 = ‘Habitative’ names habitative names 6 10 proposed= Proposed additionsadditions to 17 set of ‘habitative’ names 7 to set of habitative names 11 14 24

The Journal of Scottish Name Studies 4, 2010, 131–58 The Shadow of ‘Onomastic Graffiti’ 135 preliminary set I would suggest ON vôllr ‘field’, land ‘land’, garðr ‘enclosure’ and borg ‘fort’. (Another would be ON skiki ‘strip’ but because it does not seem to occur in the area of survey I shall not discuss it in detail.) Since there is a danger that these terms may be disallowed before my arguments even get off the ground, I think it is necessary to defend them individually:

ON vôllr ‘field’ Summers on the west and north-west coasts of Scotland are short. Proverbially, Scotland only has two seasons: June and winter. In the north-west, even that can seem generous. For the sake of argument let us presume that the Norse had only occupied the Hebrides and were seasonal visitors to the mainland. Even if the climate was kinder we cannot envisage Norse farmers in the islands bringing their livestock to the west coast (or Arran) before May and it is unlikely any would have lingered for the equinoctial gales of September. On this basis it is difficult to believe they would have established fields or built enclosures. A field suggests permanence. It is a description of part of a farm of a different order to a shieling. It presupposes an area of ground that is physically recognisable – even if not enclosed – and near to farm-buildings. It might have been used for cattle or crops, either to keep stock in or keep stock out. It is a permanent, not a seasonal, feature of the farming landscape. (Nicolaisen (1989, 80) includes vôllr amongst the ‘man-made aspects of a farmstead’. However, Fraser (1978, 21) lists vôllr as a topographical name.) In Scandinavian Scotland descriptions associated with Norse fields include ‘long’ (Langwell, Langal), ‘broad’ (Braal), ‘narrow’ (Mial); for horses (Rossal) or sheep (Syal); or with a ‘thing’ (Dingwall). (Nicolaisen 1982, 83; Fraser 1995, 100; Fraser 1986, 31; Macbain 1894, 241; Fellows-Jensen 1996, 22–24.) They occur throughout northern Scotland and in the Hebrides as far south as Bute. (It would be particularly difficult to dismiss names including ‘thing’ as merely topographical: the word suggests judicial assemblies and permanent habitation.) Before leaving vôllr-names I would like to draw attention to the incidence of the word ‘summer’ (< ON sumar) in place-names on the west coast. Seasonal occupation, I would argue, was by local residents, not annual migrants from the islands. In 18th-century estate maps we sometimes come across fields described as ‘Wintertown’, i.e. that part of the farm to which the animals were brought down in winter (see RHP 12324/1 Balligrundale & Tirone, Lismore; RHP 12324/2 Killean, Lismore; Robinson 1988, plan of Achar (Duror), facing p. 97). In summer the process was reversed. Many, if not most, farms in the Highlands and Islands practised transhumance. There was a migration of livestock to the hill pastures as they flourished in the warmer 136 Denis Rixson weather. Thistransfer lasted anything from a few weeks to several months and in some cases the grazings might be many miles away. In Skye there is a Sumardale (summer-dale), at NG 3736 in Bracadale, which has a value of 3½ d. in 1706. On 16th December 1539 some 19 Highlanders were accused of besieging ‘Alexander Makcloid of Dunvegane in his House of Summerdaill, in the moneth of Maii last bipast’ (Pitcairn 1833 I, Pt 1, 225). Macleod would obviously have been more vulnerable in Sumardale than in his winter quarters of Dunvegan Castle. This is reminiscent of one of the more notorious episodes in Clanranald history. There is a tradition that the chief’s family summered in Arisaig, although Castle Tioram lies in Moidart. The bay in Arisaig is called Loch nan Ceall and, c. 1588, Allan Og, the heir, was murdered here by his brothers whilst shooting seals off the rocks (MacDonald and MacDonald 1900, 295–96). From two other parts of the mainland we have enigmatic place-names which seem to suggest an association between the great families and summer residences. There is a Somerville Bank by Loup in Kintyre which appears in Langlands’ map of 1793. (Loup was the estate centre of the MacAlisters, one of the oldest Macdonald branches.) Unfortunately I know of no early reference. There is also a Summerhill (NR 7221), situated behind Askomill, which was one of the principal Macdonald residences in Kintyre. (Askomill was where Sir James Macdonald of Dunnyveg burned the house of his father, Angus, in January 1597 (Pitcairn 1833 III, Pt 1, 5–6).) Again I know of no reference before 1809 (Martin 2009a, 43). Fellows-Jensen (1984, 156) discusses the possibility that ON sumar-hliðar ‘summer slopes’ lies behind place-names in Iceland and England. Now I cannot construct a grand theory from little snippets like the above, each and all of which may explicable in other ways. However it does make me wonder if there were summer-fields in the same way as there were wintertowns, and that these summer-fields may have dated back to Norse times.

ON land ‘land’ Marwick (1952, 231–32) writes of land-names in Orkney:

‘In general, they are substantial fertile farms (or tunships) lying in the body of the parish, so to speak. They are without any doubt very early and venerable settlements’.

Thomson (1995, 59) sounds a note of caution by pointing out thatland :

‘was a very general term which could be applied in a wide variety of The Shadow of ‘Onomastic Graffiti’ 137 circumstances varying from district-names to quite minor non-habitative features’.

Fraser (1978, 24 and 1986, 30) gives a Sutherland example where Merkland may be for ‘march-land’ since it lay on an important watershed. Watersheds and drainage basins certainly lie behind some of our earliest district definitions. In light of these comments we should scrutinise our examples to see whether they apply to farms and whether they are associated with other elements which are also Norse. If they pass both tests then we may admit them as habitative.

ON garðr ‘enclosure’ Nicolaisen (1969, 8) decided to look at some of Marwick’s place-name elements, ‘in the wider context of Norse settlement in Scotland’. He eliminated garðr from his reckoning:

‘because it, or a closely related cognate, frequently occurs as a loan-word in Gaelic place-names in the Hebrides, as well as in bilingual Norse- Gaelic names’.

Nevertheless he includes garðr (1989, 80) amongst the ‘man-made aspects of a farmstead’. Fraser (1978, 19) includes garðr in his list of ‘Habitative Names’ but comments, ‘It is difficult to say whether this is a true habitative term.’ Macbain (1894, 232) was aware of the complications with garðr but states that ‘when Garry is prefixed the name is Gaelic’. He then gives examples (Croigarry, Asmigarry) where garry is the last element and comments that these ‘appear to be Scandinavian’. Whilst acknowledging the dangers, I think we can include garðr-names on the mainland where they are combined with another Norse element and follow Norse word-order. If they satisfy both conditions it is difficult to discount their significance. We can also take into account the prevailing onomastic context. It is extremely unlikely that the Norse would have made enclosures if they only ever visited the mainland for seasonal grazing. Enclosures were usually built to fold cattle or sheep, to prevent livestock from straying into corn-land or to protect some grass for winter. Why would seasonal migrants ever want to enclose fields? By definition they had no crops to protect.

ON borg ‘fort’ It is difficult to see why the Norse would name a fortified site in their own language unless they recognised it and maintained it as such. Borg implies 138 Denis Rixson military possession. It was now a Norse fortified site. There are many hundreds of Iron Age forts and fortlets along the west coast and throughout the Hebrides. Most of these are not called Borg. Many are now just nameless duns. Those which are named Borg were occupied by their new Norse masters; they were adopted, adapted and maintained by the new elites. This supposition is supported by the fact that the names are sometimes of the form Borrodale (i.e. ON borg + dalr ‘fort-dale’) where both elements are Norse. When we find an associated name such as Camus nan Gall ‘strangers’ bay’ – which is the name of the beach at Borrodale, Arisaig – then a Norse connection becomes even more likely.

ON skiki ‘strip’ Fraser (1978, 21) includes skiki as a topographical element and gives examples. In 1995, 98, Fig. 21, he gives a map of skiki and vôllr names in the far north- west. Clark (1993, 3–7) also gives instances. A strip of land surely indicates habitation. Farmers do not graze animals on strips but they do grow crops thereon. However, I will not deal with this term directly since it does not seem to occur in mainland Argyll. There are a fair number of occurrences in the far north and north-west. There are also some possible skiki-names in Mull (see Appendix).

This list is not exhaustive. There are other terms whose meanings warrant inclusion. Nicolaisen (1989, 81) mentions ON velta ‘ploughed soil’. Fraser (The settlement names of Gairloch parish, 6) discusses Taagan (< ScG Na Tathagan) as a possible loan from an Old Norse word for ‘infield’ or ‘homestead’ (taða). Part of Fernoch (NR 7285) in North Knapdale was Fernach beg antagan in NAS GD437 no. 32 of 1459 or Farnagyntagan in AS i no. 268 of 1643. Nicolaisen (1989, 80) includes topt ‘house-site, toft’ amongst the ‘man- made aspects of a farmstead’. Mindful of his comments about garðr I have excluded topt from my survey because it certainly passed into Gaelic and we find it several times in island place-names. Macbain (1894, 238–39) gives examples from Lewis and Skye. There are also Totamore and Totronald in Coll. The only time it seems to occur as a farm-name in the area of study is as Daltote (NR 7583) in Knapdale. There are also terms which do not imply habitation by their meaning but which strongly suggest it in their cultural context. Nicolaisen (1979–80, 111– 112), whilst discussing Lewis, describes a ‘second category of simple name, that referring to the man-made landscape’. His first example is Cross (< ON Kross ‘cross’); his second is borg. In fact, within Lewis the word Kross appears not just in its simple form as Cross (NB 5062, 4 d. in 1754, Johnson’s map of 1821) but also as a compound in Crossbost (NB 3924, Blaeu, Johnson’s map of 1821). The Shadow of ‘Onomastic Graffiti’ 139 Nicolaisen gives further examples of Kross appearing as a farm-name in Orkney and Norway. Reverting to the west coast we have Cross (NM 6691) as a 3 d. farm in South Morar as well as Ardnacross (NR 7626), Crossibeg (NR 7422) and Crossaig (NR 8351) in Kintyre. There are also Crosshill (NR 7119) in Kintyre and Crossal (NG 4531) in Skye which may both be from kross + vôllr. It is difficult to see how the Norse word for a cross could become a farm-name in Morar unless the Norse settled there and erected (or preserved) a noteworthy cross. (The site is not at a crossroads.) The likeliest scenario for the survival of such names is that they were first coined by Norse occupants and these forms adopted by current and later Gaelic-speaking inhabitants. A ‘Langal’ or ‘long field’ would be inconceivable as the coinage of a wholly Gaelic-speaking community. It is equally unlikely to be the legacy of a seasonal visitor. It only makes sense if we think of it as minted by a Norse occupier and as having then passed into the vocabulary of the local speech-community for long enough for it to survive after its parent language had died. Of course one could make precisely the same arguments for topographical terms. The difference is thatv ôllr, land and garðr indicate, by their very nature, fixed and settled farming communities, whilst borg suggests military strongholds to protect the new name-giving elite in their recently- acquired possessions.

Dale-names In the controversies over the relative significance of habitative as opposed to topographical place-names it is probably the case that dale (< ON dalr) is the key point of division. There are a great many dale-names scattered down the west coast of Scotland. If we accept them as indicating settlement then plainly the whole of the west coast falls within the orbit of the Norse. If we deny them then we can construe a division between (Gaelic) Innse Gall (the islands of strangers) and Earra-Gàidheal or Argyll (the coastland of Gaels). I do not want to reopen this aspect of the debate but would like to point to a distinction within the class of dale-names which is relevant to the wider issue. Dale-names come in many shapes and sizes. They can be simple: (?)Daill, NM 7906 in Craignish, and (?)Daal, NM 6269 Ardnamurchan (cf. Nicolaisen 1979–80, 111), or compound: Torrisdale, NR 7936, and Ugadale, NR 7829, in Kintyre. In either form they may have glen tagged on the front (Glendale, NG 1749, Duirinish, Skye; Glenaladale, NM 8275, Moidart). Within the group of wholly Norse compounds there are some mainland dale-names which would be difficult to dismiss as merely topographical. One example is the name Scamadale (< ON skamr ‘short’ + dalr ‘dale’). In the area under study there are: 140 Denis Rixson Coire Sgamadail, NG 8815, Glenelg (Coire) Sgamadail, 1 d., NG 8007, Knoydart Scamadale, 1 d., NM 7090, South Morar Scamodale, 1 d., NM 8373, Sunart, on the south-east side of Loch Shiel Scammadale, 4 mk, NM 8820, Kilninver parish, Lorn

In all cases these are, or lie at the mouth of, short glens – in the case of Scamadale, South Morar, spectacularly so. The examples in Glenelg, Sunart and Lorn are some distance inland and four of the five became farms. It is difficult to see how a Gaelic speech-community would adopt such names unless they were coined by Norse settlers. We could make similar claims for other mainland dale-names where both elements are Norse. Dibadale (or similar, in Lewis, Skye, Rum, Mull and Kincardine) is from Norse Djúpadalr ‘deep dale’ (Nicolaisen 1979–80, 113; Watson 1904b, 7–8). Swordale (or similar, in Lewis, Skye, Ross and Cromarty and Sutherland) is from Svarðdalr ‘sward dale’ (Nicolaisen 1979–80, 113; Macbain 1894, 224 and 1902, 67; Watson 1904b, 87; Fraser 1986, 30). Torrisdale (Sutherland, Kintyre) is the personal name Þórir + dalr (Macbain 1894, 225; Watson 1904a, 27; Fraser 1978, 21). So, whilst it must be admitted that all dale-names are topographical, it could be argued that some dale-names are packed with more habitative potential than others. If they are well inland; if they become farm names; if they contain personal names; if they comprise more than one Norse element; then they are less likely to indicate seasonal migration.

Lost names It is important to recognise that we do not yet have the full complement of Norse names on the west coast. Many of the humbler documents such as rentals and most of the surviving estate maps remain unpublished. There are a number of early names still to be recovered. It is by no means the case that all the names recorded in early documents or on estate maps are utilised in academic research. Sidrigill appears as a farm in North Kintyre several times in the earliest documents but has been ‘lost’ for centuries. Several dale-names in Kintyre and Knapdale now only attach to stretches of water: Loch Dirigadale at NR 7245, Loch Garasdale at NR 7651 and Loch Racadal at NR 7665. At least these are recorded. But what is now Black Loch at NR 7117 in South Kintyre was Loch Grunidel in Roy(PC 53) of the mid-18th century. In ABA DR 4/9/7, which is a farm plan of Dalriach by John Waterston in 1836, it is Loch Grunidale. That name has now disappeared. Equally tradition, rather than the map, gives us Stangadill in Morvern. The Shadow of ‘Onomastic Graffiti’ 141 What can the evidence from land-assessment contribute? One of the dangers of looking at place-names in the abstract is that, devoid of context, we weigh them all equally. We concentrate simply on the names, the words that compose them, their meanings and derivations. In some situations that is desirable, but not if we wish to weigh them for their habitative content or grade them as indicators of settlement. Settlement is not abstract, it is concrete. Settlers take over or establish farms. They inherit or assign extent and boundaries; they impose spatial definitions. Some units contained others – a farm might include several fields and more than one shieling. Farms were dynamic; some expanded and absorbed their neighbours; others contracted by subdivision. At various stages all farms were given valuations for fiscal purposes. The following paragraphs are based on my own research into land-assessment within the area of study. This is now complete but not yet published. In our area we have to deal with at least three fiscal systems. At the top, or youngest level, we have merklands which were imposed by the Lowland Scottish government from the 12th century. This system was comprehensive and gives us complete coverage of the west coast. Below this we find the ouncelands and pennylands of the Norse. Ouncelands probably date from c. 850, pennylands from c. 995 and the introduction of the Hiberno-Norse coinage. (They were integrated at the rate of 1 ounceland to 20 pennylands.) In some areas, e.g. from Morvern northwards, we can map this Norse fiscal system fairly completely. In others, e.g. around Loch Awe, the picture is blurred. Beneath this we have the davachs, quarterlands and eighthlands of the Picts or Britons – possibly modified by an intermediate layer of Dalriadic houses. The lower (earlier) networks are successively less complete but we can use each land-assessment layer, and the likely exchange rates between them, to reconstitute the one beneath. We can then plot the values of individual farms to help us gauge how important they were in earlier centuries. This is not an exact science but there have been many attempts to differentiate between primary and secondary settlement and there is no reason why we cannot use the data from land-assessment towards this. Our documentary sources for land-assessment on the west coast begin in the 13th century, but for some areas (Skye, Lewis etc.) they may be as late as the 18th century. Our cartographic evidence is also generally late. We have the Pont and Blaeu maps from the late 16th and 17th centuries. We have a map of the farms of Tiree in the 17th century but in most cases our estate maps date from after 1750. Our understanding of land-assessment is partial and patchy. But the advantage of having three fiscal networks which covered the landscape like layers of cloth is that we can establish an extent from our knowledge of the pattern or from a different layer, even where parts of the fabric are missing or worn. If, for instance, we know the merkland valuations and we also know 142 Denis Rixson the exchange rate locally between merklands and pennylands, then we can fill in the gaps in our knowledge of the pennyland layer. Likewise, if we know the ratios between merklands and davachs or between ouncelands and davachs, then we can establish the pattern of davachs, quarterlands and eighthlands. The rules of arithmetic validate this process. This allows us to assume that the patchwork of farms formed a cohesive, systematic overlay, that very often the oldest things in the human landscape are the farms, which, although conceptual from one point of view, had precise physical boundaries: this burn, that cairn or lochan, the hill ridge from here to there. The perennial competition for resources, added to the failings of human nature, meant that farm boundaries were often in dispute. In Argyllshire such arguments fill the pages of 18th-century documents and estate maps. But in a way these controversies reinforce our notion of a systematic network. It is only within such a comprehensive system that all these march disputes take place. The existence of the farms is not disputed but the ownership of a particular grazing slope or an island formed when a river changed course. As a basis for logical argument the land-assessment system has weaknesses. It is not everywhere complete and our evidence is not always early, but it is cohesive and it is comprehensive. How can we use it to detect settlement patterns? First of all the word ‘settlement’ can be misleading. The first Norse incomers to the area were not just settlers, they were farmers. Yes, they included warriors, skalds, merchants, craftsmen, fishermen and shipwrights, but overwhelmingly they were farmers. We have not a scrap of archaeological or other evidence for any urban centres or commercial entrepôts. Some Norse may have earned their living from fishing or crafts or trade, but the predominant form of economic activity on the west coast has always been farming. Like their successors in later Highland history they may have performed other functions as well, but fundamentally they were farmers or farm-owners. To secure the basics for life, they had to grow crops and keep livestock. How did they see this landscape? Did they see it as the first colonists saw the forests and plains of North America? Did they carve it out on a mathematical basis in terms of a grid of straight lines with carefully defined areas? No; nor did they inherit from hunter-gatherer societies. They inherited from a culture that was already old, numerate and literate; that already had a system of land assessment; that measured land in terms of productivity, not area, and where borders were defined in easily expressed, strictly physical terms. Thenew farmers often just took over, adopted or adapted the old farms. We can even see the areas where they left the previous system largely alone, e.g. the Ross of Mull or the parishes of eastern Cowal where there is barely a Norse farm-name. The Shadow of ‘Onomastic Graffiti’ 143 Conclusions There are 10 habitative names plotted on the map, with a further 14 which I think may be regarded as habitative. These represent only a tiny fraction of the total number of Norse place-names on the mainland west coast. We can now accept that there was significant Norse settlement within the area of study. There are some districts which are relatively blank, but each of these should be judged in its own context. On the basis of other place-names, South Knapdale does not seem particularly Norse, but Morvern probably was. Similarly, there appears to have been only a very limited degree of Norse settlement in Western Cowal, virtually none in Eastern Cowal. In the case of Kintyre, which is a long thin peninsula, Norse settlement was probably comprehensive. Further north in Argyll, it may have been confined to the western coastal strip. Despite achieving fiscal control it is unlikely there was much Norse penetration around Loch Awe or on the east side of Loch Linnhe in Duror, Glencoe or Mamore. Having said that, I have no doubt they established a trading station at the foot of the Great Glen. How else would there be an ‘Invershippinish, two miles south of Inverlochy’ (‘History of the Macdonalds’, Highland Papers I, 40)? We can also make a general criticism of the minimalist argument in that it is highly abstract. It simply looks at linguistic definitions. It takes the meanings of certain words and implies that those meanings can be used to define the historical process of settlement. The true picture is more complex and we have to be more discriminating when looking at Norse occupation of the west coast. This is a large area and settlement may have taken place over centuries. Different areas may have experienced different densities of settlement at different times. Some districts may have suffered a sweeping military invasion, others may have seen more stability and a certain amount of dynastic intermarriage. It appears that the area around Loch Morar was much more Norse than Knapdale, simply on the basis of the frequency of Norse place-names and the fact that they include some of habitative type. By admitting only certain categories of place- name element we are being precise, but possibly too narrow. Alongside the place-names there are other factors we should take into account. There may be more mileage in studying individual farms rather than settlement in the abstract. Each occurrence of the name-types suggested above should be judged in its context. Where an element such as ON garðr or vôllr is only part of a farm name then any other elements should be taken into account. If they too are Norse the overall place-name is more likely to have been coined by Norse-speaking settlers. Equally we should look at the geographic environment. If there are other Norse names in the immediate vicinity then this makes settlement more 144 Denis Rixson concrete. These names should also be studied in the context of land-assessment data. To do so helps us to establish relative weightings. Because some names are more valuable than others in fiscal terms this may have a bearing on their relative importance in onomastic terms. The minimalists have already done something similar by according special status to key habitative elements. But language is not the only register of value. In historical terms we have fiscal values established by arithmetic. This aspect of the human landscape was already embedded in the toponymy when the Norse arrived. All place-names are equally place-names but the minimalists have established an abstract gradient of value according to their habitative content. There is no reason why we cannot use the concrete historical evidence from land-assessment to challenge, support or subvert this. Land-assessment evidence concentrates on farms, those names which became farms. Despite the distance in time and the caution we should exercise in utilising such data, land valuations do allow us to make comparisons between farms and the names they bore. If we can assess relative importance in fiscal terms and if we are dealing with farm units that have been much the same since pennylands were established c. 995 AD, then we can assess their relative importance as indicators of settlement or habitation. (If a farm has preserved its valuation intact then this presupposes unchanged boundaries.) In a sense such units are, or become, proxy habitative names because they indicate units of habitation. So Meoble (< ON ból , NM 7987, South Morar), which was 1 d., is a more important name than (Allt) Maodail (NM 7783, Arisaig) which does not seem to have carried an assessment. Equally Garrabost (< ON bólstaðr, ?11 d.) and Pabail (< ON býli ‘farm’, 7½ d.), both on the Eye peninsula, Lewis, were more important farms than Shader (2 d.), Sulaisiadar (1 d.) and Seisiadar (½ d.) (< ON setr ‘shieling’) towards the eastern end of the same peninsula. These examples would gratify a minimalist heart since they imply ból, bólstaðr and býli were more important than setr. However, at 12 mk the undivided Langal (< ON vôllr) was one of the most important farms in medieval Bute; likewise the 6 d. of Mingary (< ON garðr) in Ardnamurchan. Land-assessment data is not without its problems. Firstly, our documentary sources are often late. Secondly, farms were living entities: some expanded, others contracted. A good example is Feorlig (NG 2943) in Duirinish, Skye. Nominally this farm appears to have been a farthingland (< ScG feòirlig or feòirling ‘farthing’). In fact it enters the records at 3 d., presumably because the name of one of the subsidiary units replaced that of the parent farm. Nevertheless, used cautiously, land assessment allows us to build a picture of earlier settlement and allows us to categorise. The agricultural economy of the Highlands was incredibly conservative. Time and again we can show that farm The Shadow of ‘Onomastic Graffiti’ 145 units, their names, their valuations and their boundaries remained unchanged for centuries. It is also the case that the minimalist argument rests too heavily on the assumption that somehow the coastline was a boundary. It was not. If we are looking for geographical features that were decisive political and cultural boundaries on the west coast in early times then we should be looking at watersheds. Many of the earliest district definitions seem to be in terms of watersheds and drainage basins. The western seaboard was not the limit for Norse settlers, rather the mountain ranges to the east. I have already indicated (2002, 21–24) where settlement may have extended to within the area known as Na Garbh Chrìochan ‘the rough bounds’. The north-south line of Coire nan Gall (‘strangers’ corrie’) place-names is a more reliable boundary than the coast. The idea of seasonal migration from islands to mainland is unrealistic and impractical. The only context in which this theory is credible is with regard to timber extraction, just as described by Adomnan in earlier times.

Coda In some areas the Norse settled densely, whether drawn by better land or secure in their own numbers or because the area could be more easily defended. Other districts may have been left largely unaffected as bastions of native population. Kintyre and Knapdale are physically adjacent. The frequency of Norse place-names and the land-assessment values of farms with Norse names both suggest that Kintyre was more thickly settled than Knapdale, West Cowal than East Cowal, and Bute than either part of Cowal. For example, North Kintyre was probably 400 mk or 5 ouncelands, South Knapdale was 120 mk or 1½ ouncelands. I have added up the merkland valuations of all those farms which have wholly Norse names (e.g. Ugadale, Kintyre). I have omitted anything hybrid or doubtful. In South Knapdale there is only one farm with an unequivocally Norse name and that is Ormsary at 4 mk – although I do not think Ormsary is a true -ary ‘shieling’ name. In North Kintyre there are some 21 farms which have wholly Norse names, with a total valuation of c. 70¼ mk. In percentage terms this makes the Norse content of South Knapdale farms about 3⅓%, but 17½% in North Kintyre. Accordingly we can claim more Norse settlement for Kintyre. We must also take into account the political process of settlement. Sometimes settlement was so heavy as to transform local demography; at other times it was probably confined to political overlordship. In some locations the Norse may just have swamped the natives by force of arms and then sheer numbers. In other areas their numbers may have been more modest but they achieved control either by conquest or intermarriage with local dynasties. In a 146 Denis Rixson few districts local families may have been left in charge, as long as they respected the new political arrangements and paid the appropriate taxes. How can we gauge the relative densities of Norse settlement? How can we claim that one area was more Norse than another? The minimalist concentrates on the quality of the names, measured by their habitative content. The maximalist approach has other strengths. By looking at quantity and context – onomastic, geographic, fiscal – we can apply relative weightings to the various Norse elements and naming conventions. This better reflects the murky realities of Norse settlement. It allows us to build a composite picture of where the Norse had more impact, where less. Settlement was not uniform, any more than the landscape, natural or human, in which the settlers arrived. It was always modified by circumstance. These circumstances included political, military, logistical, dynastic and demographic factors. Norse occupation and settlement was always going to be more problematic in Cowal than in Bute or Arran. The latter are islands and access has to be by sea, which the Norse controlled.

Appendix

I have listed here (in capitals) all names utilised in the text or map, along with a selection of earlier forms. Where there is little doubt about a name I have not listed all the early sources; where there is ambiguity I have listed more. Each name is numbered to ease identification on the map. To these I have added some names (underlined) which I have not used but which might be relevant. They include names for which I have no early forms or have seen no convincing derivations. I am not suggesting they are all proven, but that they deserve further attention. For brevity I have not given every source but have included a representative selection with dates and valuations where I have them.

1. habitative names (a) < ON ból or bólstaðr 1 MEOBLE, NM 7987, South Morar Mayboll, NAS E764/11/1/5 1694 1 d. Meoble, C. Fraser-Mackintosh, Excerpta e Sasinarum Registris Vice- Comitatum de Inverness, Vol. II No. 950, 1829 2 ARNIPOL & ARNABOL BURN, NM 7484, Arisaig Ardnopoill, 1 d., NRAS 2950/1/38 1616 Arnaboll, 1 d., AS ii no. 459, 1633 The Shadow of ‘Onomastic Graffiti’ 147 3 POLNISH, NM 7582, Arisaig Boeness, 1 d., GD 201/1/362/3 1699 Bullneish, 1 d., NAS E744/1/1 1748 4 RESIPOLE, NM 7264, Sunart Reischepoll, 3 mk, ER xvii, 624, 1541 Ryissabull, RMS iii no. 3085, 1545 Reseboll, Fraser-Mackintosh, The Last Macdonalds of Isla, 39–40, 1595 5 GLENSKIBLE, NR 8860, North Kintyre Glenskipinche, 1 mk, RMS ii no. 2261, 1495 Glenskippaill, 2 mk, OPS II, I, 30, 1511 Glenskippel, AS ii no. 883, 1655 6 KILLEPOLL (now Calliburn), NR 7125/7225, North Kintyre Keyllpoll, Robertson’s Index, p. 26 No. 27 (Robert I) Killepoll, 3 mk, ER xii, 352–66, 1502–05 7 KILLYPOLE, NR 6417, South Kintyre Calybole, RMS ii no. 1485, 1481 Kellipull, Munro: ALI, 202, 1449 ×1490(?) Kellable, RMS vii no. 760, 1612 Callibill, ‘twa groat land’ (i.e. 8 s. 4 d. or ⅝ mk), Kintyre Magazine No. 37 p. 25, 1674 Cox (1994, 50–57) suggests on phonological grounds that Meoble, Resipole and perhaps Arnabol actually derive from ON pollr ‘pool’. Gammeltoft (2001, 300–01, 310), although he does not address Cox’s grounds, thinks Arnabol and Meoble probably derive from ON ból. (On Gammeltoft’s p. 311, Resipole and Ryissabull should be equated.) Daill, NR 8290, North Knapdale was: Daltinabill, AS i no. 192, 1622 Daltinable, AS ii no. 561, 1636 Daltinable, 2 mk, AS ii no. 830, 1654 Daltnabill, 2 mk, GD1/426/1/26/1 1681 Clamboyle, c. NG 7712(?), Glenelg Clamboyle, 10 d., OPS II, II, 829–30, 1583 Clamboill, Inverness Retours 19, 1608 Clambuil, NRAS 2950/2/488/11/1 & 2 of 1735 Clamboll, NRAS 2950/2/488/18 of 1773 There are two places in Kintyre, with very similar names which might 148 Denis Rixson derive from ON ból. Puball, c. NR 712262, lay in North Kintyre and is now represented in the name of a burn. OS 6" 1st Series Sheet CCLI (1866–67) states that urns and coins were found at a ‘Circle’ at Puball. RCAHMS Argyll Vol. I No. 255 describes an earthwork here. There is also reference to Pubill which I think was a separate settlement in South Kintyre, probably near Innean Gaothach (which was at NR 5913). (See Martin 2009b, 29–30, for a possible location on North side of Glenadale.) This was Bugill in HP iii, 75, 1596; Pubill in HP iii, 82, 1605; Pubill in 1678 rental (NLS MS 3367). I doubt either derive from Gaelic puball ‘tent’.

(b) < ON býr 8 SOROBA, NM 8627 / 8628, Kilmore & Kilbride parish, Lorn Sorapa, 12 mk, GD 112/2/107 1470 –12 mk made it a big farm Soroba, GD 112/2/40/1/5 1510 Soirbe, Blaeu (Lorn) 9 SOROBA, NM 8004, Craignish Soroba, 5 d. (with Ilan M’nevin), Miscellany IV, 292 No. 4, quoting Argyll Inventory Vol. 1, 308, 1412 The two farms called Soroba are from ON saurr ‘mud’ + býr ‘farm’. (See Macbain 1894, 224; Nicolaisen 1960, 56; 1979–80, 118.) For discussion of bý-names and distribution map, see Grant 2005, especially 129–130, 137). 10 SMERBY, NR 7522/7523, Kintyre Smerbe et Clachyn, 8 mk, ER xii, 698–703, 1505 Smerby, 7 mk, Argyll Valuation Roll 1751 Smerby is from ON smjôr ‘butter’ + býr ‘farm’; cf. Smirisary (Moidart, Wester Ross) where the first element is smjôr ‘butter’ and the ending -ary from ON ærgi ‘shieling’ (Macbain 1902, 77; Watson 1904b, 233; Fraser 1995, 101).

Nicolaisen (1969, 10–11) points out that ON staðir (plural of staðr) develops into the ending -ston in Orkney and gives the examples of Grimeston, Hourston, Berston, Herston and Tormiston. (See also Marwick 1952, 234–35; Fellows-Jensen 1984, 157–59; Waugh 1989.) There are some similar names on the west coast which may repay further study. These include Auliston in Morvern, Coustonn and Troustan in Cowal, Caddleton in Kilbrandon, Ardtun (NM 3923) in the Ross of Mull. I am doubtful that any of these are due to influence from English except, possibly, the two examples from Cowal. It has been suggested to me that some may contain ON -tún. It is also possible that The Shadow of ‘Onomastic Graffiti’ 149 more than one final element is involved. The Ardtun example becomes more interesting because it included Eorabus (NM 3823 – a name I cannot trace before 1878) and lies close to Assapol (NM 4021) which was Assaboll in 1587– 88. (See also Gammeltoft 2001, 301.) I give variant spellings of the mainland examples below: Auliston, NM 5457/5557, Morvern 1 ounceland (=20 d.) Hawlaste in RMS i no. 520, 1346 × 1372‒73 2½ mk Alastill, RMS ii no. 2216, 1494 2½ mk Haulestin, ER xiii, 217, 1509 2½ mk Aulastill, ER xv, 675, 1528‒29 2½ mk Hawlestin, ER xvii, 646, 1541 Aulastine, Argyll Valuation Roll 1751 Coustonn, NS 0774, Inverchaolain parish, Cowal Colstane, RMS ii no. 1059, 1472 5 mk Colstane & Stroune, An Inventory of Lamont Papers No. 91, 1519 5 mk Colfstane et Strone, RMS iii no. 1882, 1538–39 Cowston, Pont(14) 5 mk Cowlstoun & Strone, AS i no. 284, 1646 For the ending in -stane, cf. Marwick 1952, 234. Troustan, NS 0776, Inverchaolain parish, Cowal Trolstir, RMS ii no. 1059, 1472 5 mk Troistir, An Inventory of Lamont Papers No. 91, 1519 5 mk Troster, ER xiv, 626, 1519–20 5 mk Troister, RMS iii no. 1882, 1538–39 Trouster, Pont(14) Trowstoun, Argyll Valuation Roll 1751 Caddleton, NM 7816, Kilbrandon parish, Lorn Caddiltoune, RRS v no. 27, 1313 Caddiltoun, 4 mk, AS ii no. 237, 1628, no. 434, 1632, no. 522, 1634 Kattildan, Blaeu (Lorn) Coldtoun, APS vii, 340, c. 1631–34

2. Proposed additions to the established set of habitative names (a) < ON vôllr 11 CROSSHILL, NR 7119, South Kintyre Crossall, 3 mk, RMS iii no. 3085, 1545 Crossell, 3 mk, RMS iv no. 1272, 1558 150 Denis Rixson Crossall, 3 mk, AS i no. 371, 1658 Crosshill, Argyll Valuation Roll 1751 This is now Crosshill and is a farm by a small hill (83 metres) just south of Campbeltown. The earlier spellings suggest a Norse word rationalised to English ‘hill’. Whether the Norse original was ON vôllr ‘field’ or fjall ‘hill’ is debatable. Crosshill was important historically because it was the source for Campbeltown’s water supply, even before the current reservoir. Perhaps the siting of a cross here reflects that. I prefer ‘field’ to ‘hill’ as an element in a farm-name and would point to Crossal (NG 4531, < ON kross + vôllr ‘cross- field’) in Minginish, Skye, as a parallel. 12 LANGAL, NM 7169, Moidart Langoll, GD 201/1/170 1700 Langal, GD 201/5/1257/1 1718 Langol, 1 mk, NAS E744/1/1 1748 Langal is from ON langr ‘long’ + vôllr ‘field’. There is also a Longoll (with a Borrodale) in Applecross (see OPS II, II, 404). 13 LAGALGARVE, NR 6529/6629, North Kintyre Lagolgarreiff, 20 s., RMS viii no. 545, 1623 This looks to be from ONlágr ‘low / low-lying’ + vôllr ‘field’ with Gaelicgarbh ‘rough’ tagged on. 14 DALSMIRREN, NR 6413, South Kintyre Dalsmerill, RMS ii no. 1485, 1481 Dalsmerill, ER xii, 356–57, 1502–05 The first element here is a word which passed into Gaelic asdail (CPNS, 414) with the meaning of ‘meadow’. The next two elements suggest Norse smjôr ‘butter’ + vôllr ‘field’ which would imply a shieling.

Other names which may be worth further investigation include: Lioul, NG 8420, Glenelg Lyolbegg, 3 d., NRAS 2950/2/488/11/1 & 2 of 1735 Leolbegg, 3 d., NRAS 2950/2/488/12 & 13 of 1754 & 1755 Legoillmoir, Inverness Retours no. 19, 1608 Lyolmore, NRAS 2950/2/488/11/1 & 2 of 1735 (Beinn) Mhialairigh, NG 8012, Glenelg Meillarie, 5 d., OPS II, II, 829 of 1583 Miolary, NRAS 2950/1/380/28 of 1804 The Shadow of ‘Onomastic Graffiti’ 151 (Allt) Miol-airighe, NM 7899 / 7999, Knoydart Millarie, 2 d., RMS ix no. 677, 1637 Mialary, Plan of Knoydart 1812 Stoul, North Morar, NM 7594 Stull, 1 d., NLS Delvine Papers MS 1313 fo. 8 1633 Stouil, GD 44/25/29/2/5 1760 The loch above is known as Lochan Stole. A place called Stole (in Norway) is mentioned in Orkneyinga Saga (Pálsson and Edwards 1981, 113–115). Gemmil, NM 7805, Craignish Gemmill, 1 mk, Argyll Valuation Roll 1751 Old Statistical Account Vol. 7, 442 (1793) claims this is a Danish name. Askol, Glassary One of our earliest documents for the West Highlands gives 51 d. in Glassary in 1240 (HP ii, 121–24, and facsimile facing p. 227). These include 7½ d. Rudol and 5 d. Askol. Rudol is now Rhudle but Askol may not be the present Asknish. In the case of Rudol the second element appears to be from ON dalr ‘dale’ but possibly the second element in Askol is from ON vôllr?

(b) < ON garðr 15 MINGARY, NM 5063, Ardnamurchan Meare, 3½ mk, ER xvii, 623, 1541 Mingarie, 6 d., NAS CS 46/1913/May No. 6, 346–48, 1696 16 MINGARRY, NM 6870/6970, Moidart Mingary, 1 d., NAS E744/1/1 1748 17 MINGARY, NR 6419, Kintyre Mingary (Ardnamurchan) and Mingarry (Moidart) are beside two of the most important castles on the western seaboard (Castles Mingary and Tioram, respectively). The castle connection is particularly striking. Douglas Simpson (1968, 133–37) pays tribute to these early fortifications on the west coast. Were they perhaps successors to Norse military structures? Mingary in Kintyre is not obviously near a castle, although it is close to several important farms (e.g. Tirfergus and Machrihanish) as well as the parish church of Kilkivan. There is also a Mingary (NM 4155 / 4156) in Quinish, Mull which was worth the equivalent of 2 d. in 1528. In all cases this name is probably from ON mikill ‘great’ + garðr ‘enclosure’ (Macbain 1902, 73, 76). All three mainland Mingarys have other farms with Norse names in the neighbourhood. 152 Denis Rixson 18 SKEROBLINGARRY, NR 7026, North Kintyre Skerblinggorie, 3 mk 5 s. 8 d., RMS viii no. 545, 1623 I know of no other name quite like this except its pair Skeroblinraid, NR 7025 (also known as Park). The first element is reminiscent of the ‘skeir’ in Skeirchenzie, which was the Iona estate in this part of Kintyre. Both Skeroblins were part of the Skeirchenzie estate but appear in different parishes in 1678. Skeroblin may be for Skeirblaan which is given as a farm name in a partial list of the lands of Skeirchenzie in a document from the Argyle Inventory dated 1576 (OPS II, I, 21). The ‘skeir’ in Skeirblaan may, like Gaelic sgìre ‘parish’, derive from Anglo-Saxon scir which became ‘shire’ in English (Macbain 1911, 315). The second element is the personal name Blaan or Blane, presumably the 6th-century saint who is associated with the church, farm and parish of Kilblaan in South Kintyre. The final element is ONgarðr . If the first element is indeed ‘skeir’ then I do not know whether it is used here in the sense of parish or lay administrative district (Barrow 1973, ch. 1). 19 KELDOUENEGARTH, 4 d. in Glassary in 1240 (HP ii, 121–24, and facsimile facing p. 227); since lost. Niag-ard, NG 7406, Knoydart The ending in -rd is modern. All early forms show an ending in -rt. Nisgart, 1 d., RMS ix no. 677, 1637 Neugart, 2½ d., NLS MS 1313 p. 83, 1730 Neiguart, 2½ d., NLS MS 1313 Printed Sale Sheet 1768 Nieugart, Plan of Knoydart 1812 Barnayarry, NM 7817, Kilbrandon parish, Lorn There is a Barnacarry in Kilmore and Kilbride parish, a Barrnacarry in Kilninver parish and a Barnayarry in Kilbrandon parish – all in Lorn. Given the vagaries of spelling, it is sometimes difficult to differentiate these and I am extremely uncertain whether Barnayarry is a -garðr place-name. However early spellings probably include: Barrangerre, OPS II, I, 103 of 1470 Barrangearre, 6 mk, GD 112/2/107 of 1470 Barrangaire, 6 mk, GD 112/3/5 of 1470

(c) < ON land 20 SWORDLAND, NG 8817, Glenelg (now Suardalan) Swordilan, 3 d., NRAS 2950/2/488/11/1 & 2 of 1735 Swordland, 3 d., NAS RH9/3/142 of 1738 The Shadow of ‘Onomastic Graffiti’ 153 21 SWORDLAND, NM 7891, North Morar Swordland GD 44/25/29/2/5 1760 Swordland, ½ d., NLS Delvine Papers MS 1313, Glengarry Estate Rental 1762 Swordland is likely to derive from ON svôrðr ‘sward’ + land ‘land’. Samadalan, NG 7306, Knoydart, is spelled Sandland in RMS ix no. 677, 1637, but other spellings suggest the second element may have been from ON holmr ‘islet’ not land. (Santilman or Santdman in NLS Delvine Papers MS 1313 fo. 8, 1633; Sandlemain in NLS Delvine Papers MS 1313 p. 83, 1730 and Printed Sale Sheet 1768; Sandliman in Fraser-Mackintosh 1897, 123, c. 1761.) Exactly the same thing seems to happen with what is now Samalan Island (NM 4536) off Inch Kenneth, Mull. Samuel Johnson calls it Sandiland in 1773. OS 6" 1st Series Sheet LXXXII of 1878 calls it Sandland Island. I think something similar occurs with Samalaman (NM 6677 / 6678) in Moidart. Accordingly I have not included Sandland on the map. Glenforslan, NM 7573, Moidart Glenforsalan, 1 mk, NAS E764/1/1 1748 Glenforslan, 1 mk, NAS E764/24/2 1770 Fors is Old Norse for ‘waterfall’ – the glen forms an impressive hanging valley – but I am not sure what the last element represents. Retland, South Morar, is from Gaelic (cf. Camas Réidhlean). Greenland (NR 7323, Kintyre), along with Norway, is recent (Martin 2009a, 26, 37). According to ABA DR4/9/136 (undated) Greenland was formerly High Smerby. For further discussion of -land names, see Saerheim 2001 and 2005.

(d) < ON borg 22 BORRODALE, NM 6984, Arisaig Borwadall, 4 mk, RMS ii no. 2438, 1498 Borridill, 4 d., NRAS 2950/1/38 & 39 1616 23 GLENBORRODALE, NM 6161, Ardnamurchan Glenborvicdall, 1 mk, ER xvii, 622, 1541 Glenboradill, 5 d., Argyll Valuation Roll p. 74, 1751 24 BORGADALE, NR 6206/6306, South Kintyre Bargadullmoir, 2 mk, RMS iii no. 3085, 1545 Burgadillbeg, 1 mk, Rental 1678 (NLS MS 3367) 154 Denis Rixson To this list we should very probably add Coire Bhorradail (NM 6250), in Morvern, which lies at the head of the glen behind Dun Fhionnàiridh. In 1843 Rev. John M’Leod, minister, wrote the following in his account of one of Somerled’s victories over the Norse: ‘Two of the leaders, Borradill and Lundy, were slain in adjoining cor- ries, which still bear their names, and another, Stangadill, was so closely pursued, that to escape the sword he leaped into a boiling linn, which, in commemoration of the event, is still known as Eass Stangadill’. Parish of Morvern p. 179, New Statistical Account Vol. 7 (1845) Coire Lunndaidh (NM 6450) is less than a mile east of Coire Bhorradail and there are waterfalls in the near vicinity, although the name Stangadill has disappeared from the map. Folklore has helped preserve two Norse place- names, albeit in a new guise. In view of Nicolaisen’s comments about Arran it is worth noting that there are possibly two ‘Borg-dales’ in Bute. There is Birgidale (NS 0759), which was 7½ mk in 1437, though subsequently divided into Birgidale Crieff and Birgidale Knock ‒ 7½ mk would have made it a large and significant farm. There is also a Dumburgadale at NS 063660 in OS 6" 1st Series Sheet CCIV of 1863. Similarly there is a Birgidile at NM 829402 on Lismore and a Borrodale (by Longoll) in Applecross (see OPS II, II, 404).

(e) < ON skiki This does not seem to occur on the mainland within the area of study but I give the following possible examples from Mull: Girgiscaig (no longer on the map but by Tobermory), in NAS E645/1/1 of 1718. It is Gurasgag in AS ii no. 1434 of 1668 and Girkiskay in Argyll Valuation Roll p. 68 of 1751 Gleann Alasgaig, NM 5020/5120, Brolos

More doubtfully I mention: Allt Ghreagasgaig, NM 5837, Forsay Abhainn Diosgaig, NM 6223, Moloros. This was Desgaig in 1493‒94 (RMS ii no. 2200), ⅝ mk Deestage in 1509 (ER xiii, 212), Daskaig in Blaeu Uisge Fealasgaig, NM 5324, Brolos These last three are all (now) connected with water. The first two elements in Finiskaig (NM 8794), North Morar, may be for fionn ‘white’ + uisge ‘water’. The Shadow of ‘Onomastic Graffiti’ 155 Abbreviations ABA Argyll and Bute Archives, Lochgilphead ALI Acts of the Lords of the Isles (see under J. and R.W. Munro) APS Acts of Parliament of Scotland (see under T. Thomson) AS Argyll Sasines (see under H. Campbell) CPNS The History of the Celtic Place-Names of Scotland (see under Watson) ER Exchequer Rolls (see under J. Stuart) GD Gifts and Deposits, National Archives of Scotland HP Highland Papers NAS National Archives of Scotland NLS National Library of Scotland NRAS National Register of Archives Scotland (via NAS) OPS Origines Parochiales Scotiae OS Ordnance Survey RHP Register House Plan, National Archives of Scotland RMS Registrum Magni Sigilli Regum Scottorum (Register of the Great Seal) RRS Regesta Regum Scottorum (for Vol. 5, see under Duncan)

References An Inventory of Lamont Papers, 1914 (Scottish Record Society: Edinburgh). Argyll Valuation Roll, 1751, manuscript copies in NAS, Argyll & Bute Archives (Lochgilphead) and Mitchell Library, Glasgow. Barrow, G.W.S., 1973, The Kingdom of the Scots (London). Blaeu: maps available through NLS website. Campbell, H., ed., 1933–34, Argyll Sasines 2 vols (Edinburgh). Clark, A., 1993, The Vikings in Ross and Cromarty (Ross-shire). Cox, R.A.V., 1994, ‘Descendants of Norse Bólstaðr?: A Re-Examination of the Lineage of Bost & Co.’, in Peoples & Settlement in North-West Ross, ed. J.R. Baldwin (Edinburgh), 43–67. Cox, R.A.V., 2002, The Gaelic Place-Names of Carloway, Isle of Lewis (Dublin). Douglas Simpson, W., 1968, The Ancient Stones of Scotland (London). Duncan, A.A.M., ed., 1988, The Acts of Robert I 1306–1329, Regesta Regum Scottorum V (Edinburgh). Fellows-Jensen, G., 1984, ‘Viking Settlement in the Northern and Western Isles – the Place-Name evidence as seen from Denmark and the Danelaw’, in A. Fenton and H. Palsson, eds, The Northern and Western Isles in the Viking World (Edinburgh), 148–168. Fellows-Jensen, G., 1996, ‘Tingwall: the significance of the name’, in D.J. Waugh, ed., Shetland’s Northern Links (Edinburgh), 16–29. 156 Denis Rixson Fraser, I.A., no date, The Settlement Names of Gairloch Parish (Ross and Cromarty Heritage Society). Fraser, I.A., 1978, ‘The Norse Element in Sutherland Place names’, Scottish Literary Journal – Language Supplement No. 9, 17–27; given at ASLS Conference in May 1978. Fraser, I.A., 1985, ‘The Place-names of Argyll – An historical perspective’, in Transactions of the Gaelic Society of Inverness 54, 1984–86. Fraser, I.A., 1986, ‘Norse and Celtic Place-Names around the Dornoch Firth’, in J.R. Baldwin, ed., Firthlands of Ross and Sutherland (Edinburgh), 23–32. Fraser, I.A., 1995, ‘Norse settlement on the north-west seaboard’, in B. E. Crawford, ed., Scandinavian Settlement in Northern Britain (London), 92–105. Fraser-Mackintosh, C., 1895, The Last Macdonalds of Isla (Glasgow). Fraser-Mackintosh, C., 1897, Antiquarian Notes (Inverness). Fraser-Mackintosh, C., Excerpta e Sasinarum Registris Vice-Comitatum de Inverness, MS volumes in the Fraser-Mackintosh Collection, Inverness Reference Library. Gammeltoft, P., 2000, ‘Why the Difference? An Attempt to Account for the Variations in the Phonetic Development of Place-Names in Old Norse bólstaðr in the Hebrides’, Nomina 23, 107–19. Gammeltoft, P., 2001, ‘The place-name element bólstaðr in the North Atlantic area’ (Copenhagen). Grant, A., 2005, ‘The Origin of the Ayrshire Bý Names’, in Cultural Contacts in the North Atlantic Region: The Evidence of Names, eds. P. Gammeltoft, C. Hough and D. Waugh (Lerwick), 127–140. Highland Papers I (Scottish History Society: Edinburgh, 1914). Highland Papers II (Scottish History Society: Edinburgh, 1916). Highland Papers III (Scottish History Society: Edinburgh, 1920). Kintyre Magazine (Kintyre Antiquarian and Natural History Society: Campbeltown). Jennings, A., 2004, ‘The Norse Place-Names of Kintyre’, in Scandinavia and Europe 800–1350, eds. J. Adams and K. Holman (Brepols). Jennings, A., and Kruse, A., 2009, ‘One Coast – Three Peoples: Names and Ethnicity in the Scottish West during the Early Viking period’, in A. Woolf, ed., Scandinavian Scotland – Twenty Years After (St Andrews). Kruse, A., 2004, ‘Norse Topographical Settlement Names on the Western Littoral of Scotland’, in J. Adams and K. Holman, eds, Scandinavia and Europe 800–1350 (Brepols). Macbain, A., 1894, ‘The Norse element in the Topography of the Highlands and Isles’, Transactions of the Gaelic Society of Inverness 19 (1893–94), 217– 245. The Shadow of ‘Onomastic Graffiti’ 157 Macbain, A., 1902, ‘Place-Names of Inverness-shire’, Transactions of the Gaelic Society of Inverness 25, 55–84. Macbain, A., 1911, An Etymological Dictionary of the Gaelic Language (Stirling). MacDonald A., and A. MacDonald, 1900, The Clan Donald Vol. II (Inverness). Martin, A., ed., 2009a, The Place-Names of the Parish of Campbeltown (Campbeltown). Martin, A., ed., 2009b, The Place-Names of the Parish of Southend (Campbel- town). Marwick, H., 1952, Orkney Farm-Names (Kirkwall). Miscellany: Miscellany of the Scottish History Society IV, 1926 (Scottish History Society: Edinburgh). Munro, J. and R.W., eds, 1986, Acts of the Lords of the Isles (Scottish History Society: Edinburgh). Nicolaisen, W.F.H., 1960, ‘Norse Place-Names in South-West Scotland’, Scottish Studies 4, Pt 1, 49–70. Nicolaisen, W.F.H., 1969, ‘Norse settlement in the Northern and Western Isles’, in Scottish Historical Review XLVIII, 6–17. Nicolaisen, W.F.H., 1979-80, ‘Early Scandinavian naming in the Western and Northern Isles’, in Northern Scotland 3, No. 2, 105–121. Nicolaisen, W.F.H., 1982, ‘Scandinavians and Celts in Caithness: The Place- Name Evidence’, in J.R. Baldwin, ed., Caithness, A Cultural Crossroads (Edinburgh), 75–85. Nicolaisen, W.F.H., 1989, ‘Imitation and Innovation in the Scandinavian Place-Names of the Northern Isles of Scotland’, Nomina 11, 75–85. Nicolaisen, W.F.H., 1992, ‘Arran Place Names – A Fresh Look’, Northern Studies 28, 1–13. Origines Parochiales Scotiae II, Pt I (Edinburgh, 1854). Origines Parochiales Scotiae II, Pt II (Edinburgh, 1855). Pálsson, H., and P. Edwards, 1981, Orkneyinga Saga (Harmondsworth). Pitcairn, R., 1833, Ancient Criminal Trials in Scotland (Edinburgh). Plan of Knoydart, copied from plan of 1812 (NLS EMS S.461.B). Pont – maps available through NLS website. Registrum Magni Sigilli Regum Scottorum, ed. J.M. Thomson (Edinburgh, repr. 1984). Retours: see under T. Thomson. Rixson, D., 2002, Arisaig and Morar (East Linton). Robertson, W., 1798, An Index, drawn up about the year 1629, of many Records of Charters (Edinburgh). Robinson, C., 1988, Report of the contents and estimate rents of part of the annexed estate of Lochiel, 1772, by William Morison – copy in Fort William Library. 158 Denis Rixson Roy(PC): Roy, Protracted Copy + No. – Manuscript Maps in British Library. Saerheim, I., 2001, ‘Settlement Names of Two Millenniums. The Dating of the -land names and the Semantics of the Ending -land’, Northern Studies 36, 91–107. Saerheim, I., 2005, ‘Norse Settlement Names in -land in Shetland and Orkney’, in P. Gammeltoft, C. Hough and D. Waugh, eds, Cultural Contacts in the North Atlantic Region: The Evidence of Names (Lerwick), 216–229. Stuart, J., et al., eds, 1878–1908, Exchequer Rolls of Scotland (Edinburgh). The Clan Campbell Vol. 3, 1915 (Edinburgh). Taylor, S., 2001, Review of Ian A. Fraser, The Place-Names of Arran(The Arran Society of Glasgow: Glasgow, 1999), Nomina 24, 111–15. Thomson, T., ed., 1811–1816, Inquisitionum ad Capellam Domini Regis Retornatarum (Retours) (London). Thomson, T., and C. Innes, eds, 1814–75, Acts of Parliament of Scotland (London); on-line at . Thomson, W.P.L., 1995, ‘Orkney farm-names: a re-assessment of their chronology’, in B. E. Crawford, ed., Scandinavian Settlement in Northern Britain (London), 42–62. Watson, W.J., 1904a, ‘The Study of Highland Place-Names’, The Celtic Review 1, 22–31. Watson, W.J., 1904b, Place-Names of Ross and Cromarty (repub. 1976). Watson, W.J., 1926, The History of the Celtic Place-Names of Scotland (Edinburgh; repub. 1993). Waugh, D., 1989, ‘The Scandinavian Element Staðir in Caithness, Orkney and Shetland’, Nomina 11, 61–74. Varia

0 10 20 40 60 M

0 10 20 40 60 80 100Km Aberkarf1 Jacob King, Ainmean-Àite na h-Alba

Aberkarf 2 is an aber-name which has hitherto gone largely unnoticed in the study of Scottish toponymy. It is mentioned only once in an MS dated 1336 from Cotton MS Vespasian FVii f. 12 in the British Library.3 This name was brought to the author’s attention by two works of Professor Geoffrey Barrow, the only person to the author’s knowledge who has discussed this name.4 In both these works, he tentatively identifies it with Cromdale, although he also concedes he ‘has not been able to identify [it]’ (Barrow 1967, 78 note 4). This varia takes a closer look at the manuscript, with a view to locating Aberkarf.

The source and identification The manuscript is a Latin document describing the movements of Edward III on a military excursion to Scotland. It is a relatively detailed itinerary of places and distances, which Barrow describes as ‘having the highest literary authority’ (ibid., 77). A number of translations and editions have been produced.5 The actual context for Aberkarf is:

Sed die Martis sequente idem dominus noster unam modicam dietam videli- cet octo leucarum usque ad Aberkarf ‘But on the following Tuesday our same lord (went) eight miles, a mod- erate day’s ride to Aberkarf. ’

Aber- as an element means a ‘confluence’ and derives from a P-Celtic stratum in Scotland (and Wales). Since other existing Scottish aber- names have as their second element a river-name from the P-Celtic or earlier linguistic stratum, and the names themselves are located at the confluence of that watercourse, it is reasonable to assume that we are looking for a place located thus. The context makes it clear the starting point is Kincardine Kirk in Badenoch

1 This was first given as a paper at the Scottish Place-Name Society conference in Glasgow in 2009 and a synopsis appears in the Spring 2010 SPNS Newsletter. 2 Not to be confused with Abercarf c.1124 Glas. Reg. vol. i, p. 4, probably now Wiston LAN. See also Beveridge 1926, 6. 3 The formAberkarf has been checked against the original. 4 Barrow 1967, 78 and 1973, 58 with note 252. 5 Ferrerio 1839, xix-xxii; Ellis 1846, 33–39; Rogers 1999, 48–50.

The Journal of Scottish Name Studies 4, 2010, 159–168 160 Jacob King (NH 938155) – ecclesiam de Kynkardyn in Badenau – and that Edward and his men are travelling north up the Spey towards Lochindorb Castle. The term translated as ‘mile’ here is Latin leuca ‘league’ which varied in distance at different times and places. To find the length intended in this MS, other distances were measured for comparison. While several distances occur in the MS, the ones used need to (a) have a known beginning and end point and (b) have a known route, i.e. be connected by a road which was known to have existed in medieval times. Five such distances fulfil these criteria; these are:

A B leuca miles ratio Perth Blair Atholl 28 33 1.2 Forres Elgin 8 11.5 1.4 Elgin Cullen 18 21 1.2 Aberdeen Dunnottar Castle 16 16 1 Dunnottar Castle Forfar 24 34 1.4

As can be seen, the ratio ranges from 1 to 1.4, i.e. a leuca ranged between being the equivalent of one mile and 1.4 miles.6 This would suggest that Aberkarf is between 8 and 11.2 modern miles north-east of Kincardine Kirk. There are six watercourses which have confluences on the Spey in this range: Allt an Fhithich (NJ 065289), Allt a’ Choire Odhair (NJ 061270), Kylintra Burn (NJ 029267), Glenbeg Burn (NJ 026259), Auchernack Burn (NJ 022249) and the River Dulnain (NJ 003238).7 Of these, two can be removed since they have appropriate places at their foot which are on record from before 1336. The first of these is Allt a’ Choire Odhair which has at its foot Congash (NJ 055266):

Cuneneges 1226 Moray Reg. no. 70 Coninges, Conynges 1281 × 1298 Fraser, 1883 vol. iii no. 11 Kunnyngais 1464 Fraser, 1883 vol. iii no. 33 Connygais 1468 Fraser, 1883 vol. iii no. 35 Cwnygaiss 1468 Fraser, 1883 vol. iii no. 36 Connageis 1489 Fraser, 1883 vol. iii no. 45 Connygais 1493 Fraser, 1883 vol. iii no. 49 Konigash c. 1591 Pont map 6r Conegais 1608 Fraser, 1883 vol. iii no. 180

6 This is perhaps meant to denote a Roman league, which is 1.4 modern miles. 7 It is possible that Aberkarf is situated not at the confluence of a watercourse with the Spey, but at the confluence of a smaller watercourse within the Spey tributary system. It is hard to see, however, what possible site this could include, except perhaps a tributary of the Dulnain. Varia 161 The second is Glenbeg Burn, which has at its foot Inverallan:8

Inueraldeny 1124–1242 Moray Reg. no. 62 Inueraldem 1187–1203 Moray Reg. no. 16 Inueralyen 1224 × 1242 Moray Reg. no. 95 Inneralien 1389 Moray Reg. no. 170

Although it is of course possible for either of these places to have had a second, otherwise unrecorded, name, it is unlikely. In no other case in Scotland is there a record of an aber- name superseding an earlier name. Moreover, since the date of this recorded name is 1336 – presumably at least 300 years after Pictish (the language from which Aberkarf must have been coined) was spoken – the name could not have been new at the time of writing.

The Dulnain / Curr Hypothesis The derivation of the formkarf * , if taken as face value, would be fairly straightforward and relate to Proto-Celtic karwo- ‘deer’. It is well attested in all (known) P-Celtic languages: Old Welsh caru, Middle Welsh carw; Middle Breton caru; Cornish caruu; Gaulish Caruus [place name] (Matasović 2009, 192). It would not be unreasonable to propose that this word existed in Pictish; indeed, it probably occurs in River and Glen Carvie in Aberdeenshire:

terras de Glencarwe 1426 RMS ii, 56 terras de Glencarvy 1497 RMS ii, 2356 Glenkervy 1505 RMS ii, 2812 Glencarvy 1632–54 Gordon map 27 Glencarvy 1654 Blaeu Aberdonia & Banff

There is also a possibility that the underlying form is a Pictish name related to Welsh garw ‘rough, rugged’, re-analysed as Gaelic garbh of the same meaning. This is the most likely derivation for Abercarf mentioned in note 2, since the most suitable river in the context is the River Garf, making Abercarf (now Wiston LAN) the settlement at its foot (Barrow 1999, 61), presumably with a derivation from Welsh garw or Gaelic garbh. As for our Aberkarf, there is no watercourse in Badenoch which resembles or could be attributable in any way, either currently or historically, to the form *karf. This would lead one to look for a river name which may have been renamed. 8 Cf. Fraser 1883 vol. i, xxxiv: ‘The stream which flows down from Tobair-Alline, where it takes its rise, through Glenbeg, evidently in earlier days bore the name of the Allan ... It is also sometimes called the Craggan Burn and the Inverallan Burn.’ 162 Jacob King It is proposed here that, on the balance of evidence, it is most likely that the River Dulnain is the river in question. On modern maps this has at its foot two places, Ballintomb and Curr (itself divided into Skye of Curr, Easter Curr, Mid Curr and Curr Wood). Neither of these places have forms going back further than Pont (c. 1591). By its position on Pont maps, Ballintomb (NJ 004246) was probably formerly *Inverdulnan:

Innerqulden or Innergulden c. 1591 Pont map 6r Innertulnan c. 1591 Pont text 137r

Old forms for Ballintomb only reach back to 1611:

Belnatolme 1611 Fraser, 1883 vol. iii no. 256 Bellintome 1676 Fraser, 1883 vol. iii no. 277 Ballantomb 1695 RPS [1695/5/166] Wr. Ballintom 1747–55 Roy’s Map [Ballintom east of Spey opposite Gaick, current Ballintomb marked as Ballintruan]

Other evidence for the Dulnain in old forms are:

Stradulnen c. 1591 Pont Map 6v Dulnen Water c. 1591 Pont Map 6v River Tulnan, river Tulnen c. 1591 Pont text 137r Water of Dullan 1667 Fraser, 1883 vol. ii, 465 water of Dullan c. 1680 Illustrations of the Topography of Aberdeen and Banff ii, 298

The Gaelic form of Dulnain isTuilnean. 9 This is a Gaelic river name relating to a root tuil ‘flood’ with a-n(e)an suffix. Compare Shesgnan Burn < G.seasgnan < seasg ‘dry’ + suffix-nan. It is possible that Curr, on the south side of the Dulnain, at its foot, represents Aberkarf:

Curre 1379 RMS vol i, no. 674 Corroo 1491 Fraser, 1883 vol. ii no. 47 Cur 1532 Fraser, 1883 vol. ii no. 82 Kurr or Karr c. 1591 Pont map 6v

9 Shaw 1882 vol. ii, 63 gives Tuilenan, but Forsyth 1900, 288 uses the form Tuilnean, as does MacBain 1922, 239. Varia 163 Wester Cure, Easter Cure 1611 Fraser, 1883 vol. ii no. 256 Curris 1627 Fraser, 1883 vol. ii no. 350 Cur c. 1640 Gordon map 5 Wester Curr, Eister Cur 1662 Fraser, 1883 vol. ii no. 273

However, to suggest that this is the same place as Aberkarf, one must first posit the loss of the aber- in the name, and an equivalent can be seen in present-day Cawdor, which shows the loss of this element within the same time frame as that of the speculative Aberkarf / Curr loss.10 The logical development from a Pictish *Carw, following the first proposed derivation above, would be Gaelic *Carbh. Of particular interest here is the 1491 form Corroo. The existing Gaelic termtarbh ‘bull’ is pronounced [tharu]11 in this area and therefore *carbh is most likely to have been pronounced [kharu]. Although not a regular phonological development, this might have changed to something like [khuru] by analogy with the Gaelic term curr (now cùrr) ‘corner, end’, via a dative plural curraibh [khuu]12 – this dative plural form is perhaps what is signified byCorroo .13 Gaelic curr could also represent a strip of land and, being a suitable element for the location, may have been a tempting re-analysis for Gaelic speakers faced with the obscure term *Carbh. At some point around this time, the river (*Carbh) would have been renamed Tuilnean, further divorcing the settlement name from the river name. Clearly this is a tentative explanation, but it does show how such a process might be possible. It is noteworthy that, assuming a river name *Carw would have been transparent in Pictish as meaning ‘deer’, there is a Pictish standing-stone called the Grantown Stone that was discovered by the banks of Allt an Fhithich at NJ 045301, some five miles north of the foot of Dulnain.14 It is perhaps the best example of a deer within Pictish art. This does not necessarily mean that Allt an Fhithich was the river *karf, but it could suggest regional totemic imagery of the deer, to which Aberkarf was somehow related.

10 Cf. Barrow 2003, 48 note 251: Abbircaledouer 1238 Moray Reg. no. 40; Caldor x2 1380 Moray Reg. no. 159; Caldore 1421 Moray Reg. no. 188; Caldor 1455 Moray Reg. no. 195. 11 With an original lenited non-palatal r. Peadar Morgan, pers.comm. 12 With an original unlenited non-palatal r. Cf. Murray 1960–63, 36; for dative plurals in place-names, compare also Diack’s form (1954, 197) for the Streens: G. na strianu, i.e. dat. plur. Na Srianaibh. Thanks to Professor Richard Cox for many of these points. 13 The only Gaelic reflex gathered for the name Curr was from Diack 1954, 194: skei churr, skei (for Skye of Curr), representing G. *Sgeith Churr. Murray 1960–63, 37 proposes Sgiath Curr ‘the wing-shaped piece of land of Curr’. 14 It is currently at the National Museum of Scotland in Edinburgh. Cf. Mack 1997, 33 and Canmore ID 15737. 164 Jacob King It may seem odd that Curr is on the opposite side of the Spey to Kincardine Church, especially since Edward sent some men into Mar to the south-east directly after arriving in Aberkarf.15 Little is known about available crossings over the Spey in this period and the MS does not go into any detail on this matter, although, from the context, it must have been traversed on quite a number of occasions. In more recent times, a number of crossings have been documented in the area (MacGregor 1993, 318), not least the Boat of Garten, the crossing between Kincardine Church and the foot of the Dulnain, which would have ideally suited Edward III’s men. Under any proposed scenario, placing Aberkarf on either side of the Spey, the number of crossings of the Spey (and presumably those of its major tributaries) must have been considerable.

The Glencarnie Hypothesis Another possibility exists: that aberkarf is an MS error for *aberkar(ny).16 The course of the Dulnain was known once as Glencarnie.17 As pointed out by Ross (2003, 160), the extent of Glencarnie (at least in 1456) seems to have been up a stretch of land on the west side of the Spey, between Craigellachie and Tullochgorum, reaching to lands on both sides of the Dulnain, between Gallovie and Carrbridge.

Glantarnin c. 1206 RRS ii no. 474 ?Kyncarny 1232 Fraser, 1883 vol. iii no. 518 Glencarny 1274 x 1329 RMS i, app. 1, 31 Glencherny 14th c. RMS i, app. 2, 1423 Glenkerny 1280 Fraser, 1883 vol. iii no. 10 Glenkerny 1306 Fraser no. 12 Glencharny 1362 RMS i, 116 Glenchernin, Glenchernyn 1367 Fraser, 1883 vol. iii no. 19 Glencharny ×3 1338 Gordon Papers, 1 Glencherny 1398 Fraser, 1883 vol. iii no. 20 Glencharny 1456 ER vi 213 Glencharny ×2 1457 ER vi 216 Glencharny ×4 1458 ER vi 218

15 gentes suas in fortitudines de Marr destinavit. 16 The MS makes another error in the final letters of place-names in Lochindorb, which is given as Loghendorm. 17 Cf. Ross 2003, Watson 1926, 449 and Murray 1960-63, 35-36. 18 Ross 2003, 160, however, identifies this as Kinchurdy (NH935159). Varia 165 Under this proposal, this glen name would have derived from a previous name for the Dulnain: *Carnie, running through Glencarnie with *Abercarnie at its confluence. The meaning may be a metaphorical one relating to a P-Celtic version of Gaelic ceatharnach ‘hero, soldier’. Although the derivation of this word is not well attested, it is perhaps cognate with Welsh cadarn, cedeirn ‘strong’, from British *catarno (Geiriadur Prifysgol Cymru).19 This name already exists as Abercairnie elsewhere in Perthshire, in the parish of Fowlis Wester:20

villa mea de Abercharni 1220 x 1225 Inchaff. Chrs. no. 46 tenementi de Abbircarnych c. 1268 Inchaff. Chrs. no. 35

One can also compare semantically the names Aberargie and River Farg, coming from a P-Celtic cognate of G. fearg ‘anger’ or ‘hero, champion’ (Watson 1926, 462-63). Circumstantial evidence exists to suggest that the lower part of what is now called the Dulnain was once called the *Carnie, with the upper part being called the Dulnain / Tuilnean. It is noteworthy that the extent on the north side of the Dulnain seems to be delimited by a small burn called Allt nan Ceatharnach,21 joining the Dulnain at NH 899224.22 This is the area where Pont marks a settlement called Stradulnen, i.e. *Srath Tuilnean, ‘the strath or broad valley of the Dulnain’. This scenario might suggest an encroachment of the Gaelic name of the area down towards the Spey, finally ousting the *Carnie names such as Glencarnie and *Aberkarnie and installing the Gaelic forms of Tuilnean and *Inbhir Tuilnean.

Conclusion The conclusions drawn here are necessarily tentative: any number of the above assumptions could be incorrect; the form Aberkarf could be corrupt; the itinerary and distances mentioned in the MS could be inaccurate and the exact range of leuca itself is uncertain. Equally, Aberkarf may have been an alternative Pictish name for an existing Pictish settlement that lasted into the Gaelic era.

19 It is, of course, possible that the Gaelic form ceatharnach is a later interpretation of an earlier unrelated P-Celtic homonym. 20 Thanks to Simon Taylor for these old forms. 21 This was Allt a’ Cheatharnaich in the 1875 OS 6 inch 1st edn, but Allt nan Ceatharnach by the time of the 1902 OS 1st revision. Watson 1926, 449 calls this Allt Cheatharnaich, also citing Gleann Cheatharnaich, implying this was the course of the burn in question, not the Dulnain. 22 Somewhat east of this area lies the possibly related Auchernack (NJ027246) –the c.1591 Pont map 6r gives Achacheirnach for Auchernack and Ald Achacheirnaig for Auchernack Burn. 166 Jacob King The above conclusions are, therefore, a set of possibilities and it is hoped more information will come to light in the future that will shed further light on this tantalising name.

Post script Two other place-names are mentioned in the 1336 MS that do not seem to be attested elsewhere and are of interest:

BOGHWAN 23 This is 16 leuca (between 16 and 22.4 modern miles) from Cullen, in the direction of Aberdeen (‘leaving the castle of Kildrummy off to the side’24). This would presumably mean somewhere along the course of the present A96, between approximately Newton (NJ 479451) and Hillhead (NJ 578355). Although the precise site is unknown, this is certainly bog territory, with Gaelic bog appearing numerous times. The name might reflect Gaelic (Am) Bog Bàn ‘the white bog’; compare Bogbain (NH 709418).

FYCHAWYN 25 From the context, this would be near Invereshie. This is in modern Gaelic Inbhir Fheisidh, the confluence of the Feshie River. Fythawyn would, on the face of it, appear to reflect *Feisidh Abhainn; the word order however is not acceptable in Gaelic and, unless other similar forms appear, this form must be considered corrupt. (Also note the first certain form for Invereshie, before the 1336 date of this MS: Invercessy 1224–1233 Moray Reg. no. 76.)

Bibliography Barrow, G.W. S., 1967, ‘The Wood of Stronkalter: A Note on the Relief of Lochindorb Castle by Edward III in 1336’, The Scottish Historical Review 46, 77–79. Barrow, G.W. S., 1973, The Kingdom of the Scots (London). Barrow, G.W. S., ed., 1999, The Charters of David I: The Written Acts of David I King of Scots, 1124–53, and of his son Henry, Earl of Northumberland, 1139– 52 (Woodbridge). Beveridge, Erskine, 1923, The Abers and Invers of Scotland (Edinburgh). Blaeu, Johan, 1654, Atlas Novus, reprinted as The Blaeu Atlas of Scotland, with

23 Not Doghwan as transcribed in Rogers. Although the first letter is not clear, when a comparison is made between the initial letters in Badenau (Badenoch) and Dascheles (a surname), Boghwan is clearly the correct reading. 24 castrum de Kyldromyn quod a latere dimisit pertransiens. 25 Not Fythawyn as appearing in Ellis. Varia 167 translation by Ian Cunningham (Edinburgh 2006); also on-line at – accessed 15/09/2010. Canmore, on-line at – accessed 03/11/2010. Diack, Francis C., 1954, The Inscriptions of Pictland, ed. William M. Alexander and John MacDonald (Aberdeen). ER: The Exchequer Rolls of Scotland, ed. J. Stuart et alia (Edinburgh 1878–1908). Ellis, Henry, ed., 1846, Original Letters, Illustrative of English History Third Series vol. i (London). Ferrerio, Giovanni, 1839, Ferrerii Historia Abbatum de Kynlos (Edinburgh). Forsyth, Rev. W., 1900, In the Shadow of the Cairngorm (Inverness). Fraser, William, 1883, Chiefs of Grant, 3 vols. (Edinburgh). Geiriadur Prifysgol Cymru, on-line at – accessed 30/11/2010. Gordon map 27, on-line at – accessed 30/11/2010. Gordon Papers: papers from the Charter Chest at Gordon Castle, in J. Stuart, ed., The Miscellany of the Spalding Club iv (Aberdeen 1849). Illustrations of the Topography of Aberdeen and Banff ii (Aberdeen 1847). Inchaffray Chrs.: Charters, Bulls and other Documents relating to the Abbey of Inchaffray (Scottish Historical Society 1908). MacBain, Alexander W., 1922, Place Names, Highlands and Islands of Scotland (Stirling). Mack, A., 1997, A Field Guide to the Pictish Standing Stones (Balgavies). MacGregor, N., 1993, ‘Gaelic Place-names in Strathspey’, The Transactions of the Gaelic Society of Inverness 58, 299–370. Matasović, Ranko, 2009, Etymological Dictionary of Proto-Celtic (Boston). Moray Reg.: Registrum Episcopatus Moraviensis, ed. Cosmo Innes (Edinburgh 1827). Murray, D. C., 1960–63, ‘Notes on the Parish of Duthil’, Transactions of the Gaelic Society of Inverness 43, 24–45. Pont map 6r, on-line at – accessed 08/11/2010. Pont map 6v, on-line at – accessed 08/11/2010. Pont text 137r, on-line at – accessed 08/11/2010. RPS: Records of the Parliaments of Scotland to 1707, on-line at – accessed 03/11/2010. RMS: Registrum Magni Sigilli Regum Scottorum, ed. J. M. Thomson et alia ­(Edinburgh 1882–1914). RRS vi: Regesta Regum Scottorum vol. vi (Acts of David ii), ed. B. Webster (Ed- inburgh 1982). Retours: Inquisitionum ad Capellam Domini Regis Retornatarum, quae in publicis archivis Scotiae adhuc servantur, Abbreviatio, 3 vols., ed. T. Thomson (1811– 1816). Rogers, Clifford J., 1999, The Wars of Edward III: Sources and Interpretations (Bury St. Edmunds). Ross, A., 2003, ‘The Lords and Lordship of Glencarnie’, in S. Boardman and A. Ross, eds, The Exercise of Power in Medieval Scotland c.1200–1500 (Dublin), 159–74. Shaw, Lachlan, 1882, The History of the Province of Moray 3 vols. (Glasgow). Watson, W. J., 1926, The History of the Celtic Place-Names of Scotland (Edin- burgh and London; repr. with an introduction by Simon Taylor, 2004). Reviews

Life on the Edge. The Cistercian Abbey of Balmerino, Fife (Scotland), ed. Terry Kinder, Cîteaux: Commentarii cistercienses 59, fasc. 1–2, 2008. 168 pp. ₤25. ISBN 978-2-9600647-1-1

The Abbey of Balmerino in Fife, on the south shore of the Tay, was founded by King Alexander II and his mother, Queen Ermengarde, at the beginning of the second quarter of the 13th century. Property started to be acquired for the foundation around 1225 and the first group of monks, with their abbot, arrived there from Melrose, Balmerino’s mother-house, in December 1229. The abbey survived until the Reformation, the last abbot dying in 1561 and being succeeded by a series of lay commendators, after which the estates were erected into a temporal lordship for Sir James Elphinstone, Lord Balmerino, in 1603. Life on the Edge is the fruits of a session assembled by Julie Kerr and sponsored by the journal, Cîteaux: Commentarii cistercienses, at the International Medieval Congress at the University of Leeds in 2007. Six authors, including Kerr, contributed papers to the volume and one of them, Richard Oram, also provided a brief introduction. The papers explore variously the circumstances of the foundation of the abbey by Queen Ermengarde and the implications of its dual dedication to the Virgin Mary and St Edward the Confessor; the personnel associated with the abbey; the abbey’s endowment, its property portfolio and other physical resources; the architecture of what survives today of the abbey buildings; and the place-names of the monastic landscape of the mid-13th century. There is also a short note from RCAHMS about the resurveying of the abbey site in 2007–2008. Each section concludes with a concise abstract in French, English and German. The papers, although representing disparate disciplines, relate very closely to one another and illuminate each other. Furthermore, whether to the credit of the editor or of the contributors, each of the main papers is a real pleasure to read and they are generously provided with illustrations, maps and tables, facilitating digestion of the information under discussion. Although Balmerino’s charter-book survived and was published in 1841, it is incomplete and contains no documents later than the mid-1300s, but the authors have gathered evidence from numerous contemporary sources, including, amongst others, the charter-books of other related and neighbouring foundations, the Register of the Great Seal, monastic chronicles and the records of the Scottish universities, which help to illuminate something of the scope and limitations of the abbey and its life.

The Journal of Scottish Name Studies 4, 2010, 169–72 170 Henry Gough-Cooper Matthew Hammond’s paper, 'Queen Ermengarde and the Abbey of St Edward, Balmerino’, on the personal and political circumstances of the foundation and its aftermath highlights the importance of Queen Ermengarde as the driving force behind the project. Ermengarde’s husband, William I, had died in 1214 and was buried in the Tironensian abbey he had founded at Arbroath. For whatever reason, Ermengarde preferred not to be interred beside her husband but in her own foundation. Hammond examines in detail the probable motives for Ermengarde’s curious choice of St Edward as the joint patron of Balmerino and, in the second half of the 13th century, the rapid phasing out of references to the English saint. Julie Kerr’s paper, ‘Balmerino of Fife: Cistercians on the East Coast of Fife’, examines not only what is known about the succession of abbots and their family backgrounds, but also what little can be detected about the composition of the population of the abbey, ecclesiastical and lay. It also discusses the evidence for Balmerino monks’ attendance at universities, both St Andrews and Glasgow, and further afield at Cologne, and for their involvement in the political and ecclesiastical affairs of the day. Particularly poignant is the story of an abbot of Balmerino who had the privilege of attending the marriage in Norway between Margaret, daughter of Alexander III, and King Erik in 1281 but who was drowned with most of his companions on the voyage home. Kerr also examines Balmerino’s relationship with the wider Cistercian community, the problems of attendance at the General Chapter, financial demands on and visitations to the abbey by the Cistercian order and James V’s vain attempts to introduce reforms to the Cistercian abbeys in Scotland in the second quarter of the 16th century. Both Hammond’s and Kerr’s papers are provided with clear tabulations of the documentary evidence discussed in the text, including, in the case of Kerr’s article, a very useful hand-list of the known abbots and lay commendators of Balmerino from its inception with Alan of Melrose in 1229 to the resignation of Robert Auchmutie in 1604, when the temporal lordship was created. These two papers serve as an excellent introduction to the question of how the foundation actually operated and survived within the constraints of its original endowment and the fierce competition for new resources from other more powerful institutions, both lay and ecclesiastical, in the vicinity. This seems to have been a particular problem at Balmerino as it was a comparatively late and modestly endowed foundation, in which the royal family soon lost interest after the deaths of Ermengarde and her English daughter-in-law. In his paper, ‘A Fit and Ample Endowment? The Balmerino Estate, 1228–1603’, Richard Oram discusses the scope and limitations of Balmerino’s real estate provision, its arable and grazing lands, its fisheries and (lack of) salt, its quarry at Nydie (not obtained until the 1240s) and the vexing question of a fuel Reviews 171 supply with which to advance its industries. There is no evidence that woodland comprised part of Balmerino’s original endowment and the paper suggests that such precious patches of woodland as are evidenced in the 16th-century feus were prized as small hunting reserves rather than as a source of timber. Nor did Balmerino seem to have access to coal supplies and it seems to have remained dependent on peat for the duration of its existence. This paper is well illustrated with maps and photographs, although the map on p. 65 is the first in the volume to actually show the geographical location of the abbey. Unfortunately, the discussion of its woodland holdings includes a reference to a map (Fig. 7, p. 78) which shows the location of the abbey’s farmlands, quarry, peat moss and fisheries, but not the location of the woodlands under discussion. Richard Oram’s paper connects naturally with Gilbert Markus’s investigation of the place-name evidence, but first there is a digression by way of Richard Fawcett’s exemplary study of the architecture of the abbey buildings, ‘Balmerino Abbey: The Architecture’. This commences with an interesting review of various 19th century antiquarian attempts to come to grips with the fragmentary remains which now consist only of half the church nave and the north transept and what little remains of the rest of the conventual buildings: the sacristy, the chapter house, the day stair and parlour and a couple of detached structures. The whole of the cloister range has vanished. Fawcett gives a thorough, if not exhaustive, description of each of these buildings in turn, including a discussion of the later re-use as domestic accommodation, and provides a revised (but still partly conjectural) plan of the site, although it would have been helpful to have a short glossary of the technical architectural terms employed in the text. The paper is illustrated with a series of excellent photographs, both of substantial structures and of architectural details. It is to be wondered, if the monks did not acquire their quarry at Nydie until the 1240s, what structures existed at Balmerino in the 1230s and what facilities were in place to welcome Dom Alan and his brethren when they arrived in the mid-winter of 1229. A report is forthcoming from the excavators of the farm steading at Balmerino, which may reveal some further details of the history of the settlement (J. Lewis and S. Scott, ‘Excavations at Balmerino Farm Steading, Fife, 1996–2007’, Tayside and Fife Archaeological Journal , forthcoming). Gilbert Márkus’s paper, ‘Reading the place-names of a monastic landscape: Balmerino’, limits itself to the names associated with the abbey in the first few decades of its existence, but 23 major place-names are analysed, as well as a few incidental minor names. In many ways, the first part of his paper is the most intriguing, dealing with the shift from the earlier medieval foundation of Abernethy to the carve up of its lands in the later middle ages. Of the Balmerino estate itself, the conclusion seems to be that the mainly Gaelic, perhaps partly 172 Henry Gough-Cooper Pictish, place-name landscape in which it found itself was largely fossilised by the 13th century and the monks had little reason not to adopt the earlier names or to generate their own; but there are fascinating details here, and not a few unresolved puzzles. Márkus considers the toponymic evidence for the influence of Abernethy and St Andrews Priory in the area before the Balmerino foundation, as well as the nature of the names of the lands of local secular land-owners and places held by the abbey in other parishes. It is perhaps to be especially welcomed that his analysis of the name Balmerino disposes of the derivation from the folkloric ‘St Merinac’ ( – accessed 12th October 2009) and proposes the more prosaic, but plausible, ‘farm abounding in sea-bent’. Once again, the paper is excellently presented with maps and tables listing each name’s early forms with their dates and documentary sources. The volume closes with a cursory note, ‘Balmerino Abbey: resurvey and topographic analysis’, by Piers Dixon of RCAHMS. It is illustrated with two attractive, but largely uninformative, aerial photographs of the abbey site and limits itself to stating that, following the conversion of the Balmerino farmstead in 2007–08 for housing, the conventual buildings were resurveyed and the results incorporated into a new plan, and that topographical analysis has revealed details of the abbey’s water supply and management system. Disappointingly, neither the results of the resurveying nor a location map of the abbey mills and mill lades is provided. The plan of the abbey complex that appears on p. 87, illustrating Richard Fawcett’s paper, is that of the author and not RCAHMS. Henry Gough-Cooper, independent researcher Bibliography 2006–2009

Simon Taylor University of Glasgow

This is the first such Bibliography in The Journal of Scottish Name Studies (JSNS ). It is an attempt to present all work on Scottish name-studies which has appeared since the beginning of 2006: articles, chapters in edited books, monographs, CDs, e-books and PhDs (some of which are now available on- line). It draws heavily on the Bibliographies which I compile for Scottish Place- Name News (see below, s.n.), the twice-yearly Newsletter of the Scottish Place- Name Society, and I would like to express my gratitude to all who have helped both with those Bibliographies and this more extensive one.1 It is therefore somewhat biased towards place-names rather than personal names. However, the much higher proportion of place-name related material is also a reflection of the relative state of each of these disciplines in Scotland. Included here, too, are several works on the insular cult of saints, since these cults had such a profound effect on naming of both persons and place. For more extensive bibliographies of name-studies in Britain and Ireland, and, less comprehensively, other parts of northern Europe, see the bibliographic sections in the relevant issues of Nomina, the journal of the Society for Name Studies in Britain and Ireland: ‘Bibliography for 2006’, compiled by Carole Hough, Nomina 30 (2007), 149‒60; ‘Bibliography for 2007’, compiled by Carole Hough, Nomina 31 (2008), 157‒74; and ‘Bibliography for 2008’, compiled by Carole Hough, Nomina 32 (2009), 195‒207. The material in the Nomina Bibliographies is set out thematically and includes reviews which have appeared in the given year. An extensive, though by no means exhaustive, bibliography of Scottish toponymics, set out thematically, and regionally, can be found on-line at . I would be very pleased to hear from anyone who spots any omissions or errors in the following bibliography. I can be contacted via the JSNS website, or by post c/o Clann Tuirc. Also, I would be glad to receive notice of anything published in 2010 for inclusion in JSNS 5. In order to make it easier for the reader to find their way around this document, I have put in bold not only authors’ surnames but also some of the key places, persons or elements discussed.

1 Special thanks are due to Richard Cox, Carole Hough, Jacob King, Peter McNiven, Gilbert Márkus, Maggie Scott and Doreen Waugh.

The Journal of Scottish Name Studies 4, 2010, 173–86 174 Simon Taylor Ahronson, Kristjan, 2007, Viking-Age Communities: Pap-names and Papar in the Hebridean Islands (BAR Oxford). Ansell, Michael, 2008, ‘Carsphairn and Dalmellington Re-visited’, JSNS 2, 1–10. Ballin Smith, Beverley, Taylor, Simon and Williams, Gareth, ed., 2007, West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300 (Leiden and Boston). [Relevant chapters under individual authors.] Barrow, Geoffrey, 2008, ‘The Lost Place-names of Moray’, JSNS 2, 11–18. Boardman, Steve, Davies, John Reuben, and Williamson, Eila, ed., 2009, Saints’ Cults in the Celtic World (Woodbridge). [Relevant chapters under individual authors.] Boardman, Steve, 2009, ‘The Cult of St George in Scotland’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 146‒59. Bonar Bridge History Society, 2007, Looking in to Creich (including a CD- ROM of the place-names spoken). [Unfavourably reviewed in Scottish Place- Name News 24 (Spring 2008).] Bramwell, Ellen, 2009, ‘Names in Multicultural Scotland’, in Names in Multi-Lingual, Multi-Cultural and Multi-Ethnic Contact: Proceedings of the 23rd ICOS, ed. Wolfgang Ahrens, Sheila Embleton and André Lapierre (York University, Toronto), 158–63. [Deals with personal naming practices of Pakistani Muslim migrants in Glasgow.] [Available free of charge on: ] Breeze, Andrew, 2006, ‘Three Celtic names: Venicones, Tuesis and Soutra’, Scottish Language 25, 71‒79. Breeze, Andrew, 2006, ‘Historia Brittonum and Arthur’s Battle of Tribruit’, Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society 80, 53‒58. Breeze, Andrew, 2006, ‘The Names of Blantyre, Carluke, and Carnwath, near Glasgow’, Scottish Studies 34 (2000‒2006), 1‒4. Breeze, Andrew, 2007, ‘Tacitus, Ptolemy and the River Forth’, Classical Quarterly 57, 324‒28. Breeze, Andrew, 2009, ‘Bede’s castella and the journeys of St Chad’, Northern History 46, 137‒39. Breeze, Andrew, 2009, ‘Where was Historia Brittonum’s Mare Frenessicum?’, Northern History 46, 133–36. Breeze, Andrew, 2009, ‘The name ofBernicia ’, Antiquaries Journal 89, 73–79. Broderick, George, 2007, ‘Goedelic-Scandinavian language contact in the place-names of the Isle of Man’, in Cavill and Broderick, eds, 1‒26. Bibliography 2006–2009 175 Broderick, George, 2009, ‘The names for Britain and Ireland revisited’,Beiträge zur Namenforschung 44, 151–72. Broun, Dauvit, 2008, ‘The property records in the Book ofDeer as a source for early Scottish society’, in Studies on the Book of Deer, ed. K. Forsyth (Dublin), 313–60. Butter, Rachel, 2007, ‘Cill-names and Saints in Argyll: a way towards understanding the early church in Dál Riata?’, unpublished PhD thesis, University of Glasgow. Caldwell, David, 2008, Islay: The Land of the Lordship (Edinburgh) [with much toponymic input from Alan Macniven; Macniven contributed Chapter 2 ‘Prehistory and Early History’; while Chapter 8, ‘Continuity and Change – Place-Names and Extents’ draws heavily on Macniven 2006 (‘The Norse in Islay: A Settlement Historical Case-Study for Medieval Scandinavian Activity in Western Maritime Scotland’, unpublished PhD Thesis, University of Edinburgh); Appendix 1 is a useful compilation and examination of Islay surnames; while Appendix 2, entitled ‘Islay Lands, Recorded Prior to 1722’ is also drawn from Macniven 2006, giving suggested etymologies, as well as sources, but sadly no early forms.] Campbell, Robin, 2009, ‘“Charge of the Temporalitie of Kirk Landis” and the Parish of Lesmahagow, Lanarkshire’, JSNS 3, 1‒14. Cavill, P., and Broderick, G., ed., 2007, Language Contact in the Place-Names of Britain and Ireland, EPNS Extra Series 3 (Nottingham: English Place-Name Society). [Papers from the International Conference on Language Contact in the Place-Names of Britain and Ireland, organised by the Institute for Name Studies, University of Nottingham, in conjunction with the Centre for Manx Studies (University of Liverpool), Douglas, Isle of Man, held in Douglas, Isle of Man, 17–18 September 2004. For review article, see James 2009, below.] Cavill, Paul, 2007, ‘Coming back to Dingesmere’, in Cavill and Broderick, eds, 27–42. Clancy, Thomas Owen, 2007, ‘Spaghetti Junction: OI rót, roüt, ScG ròd, rathad, Scots rod, rode, Eng. road and some other minor by-ways’, in Fil súil nglais: A Grey Eye Looks Back, ed. Sharon Arbuthnot and Kaarina Hollo (Ceann Drochaid / [Brig o’ Turk]), 17–28. Clancy, Thomas Owen, 2008, ‘The Gall-Ghàidheil and Galloway’, JSNS 2, 19–50. Clancy, Thomas Owen, 2008, ‘Two Ayrshire Place-names [Pulprestwic and Trearne]’, JSNS 2, 99–114. Clancy, Thomas Owen, 2008, ‘Deer and the early church in North-Eastern Scotland’, in Studies on the Book of Deer, ed. K. Forsyth (Dublin), 363–97. 176 Simon Taylor Clancy, Thomas Owen, 2009, ‘The cults of Saints Patrick and Palladius in early medieval Scotland’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 18–41. Coates, Richard, 2006, ‘Maiden Castle, Geoffrey of Monmouth and Hārūn al-Rašīd’, Nomina 29, 5–60. Coates, Richard, 2007, ‘Yell’, JSNS 1, 1–12. Coates, Richard, 2007, ‘Bordastubble, a Standing-stone in Unst, Shetland, and Some Implications for English Toponym’, JSNS 1, 137–39. Coates, Richard, 2007, ‘Invisible Britons: the view from toponomastics’, in Cavill and Broderick, eds, 43–56. Coates, Richard, 2009, ‘A Glimpse through a Dirty Window into an Unlit House: Names of Some North-West European Islands’, in Names in Multi- Lingual, Multi-Cultural and Multi-Ethnic Contact: Proceedings of the 23rd ICOS, ed. Wolfgang Ahrens, Sheila Embleton and André Lapierre (York University, Toronto), 228–42. [Includes, e.g., Arran, Uist, Iona, Seil, Islay, Mull, Hebrides, Bass Rock, Coll.] [Available on-line on: .] Colville, Duncan and Martin, Angus, 2009, The Place-Names of the Parish of Campbeltown, Kintyre Antiquarian and Natural History Society, Campbeltown. [First published 1943 by the Kintyre Antiquarian Society; original list compiled by Duncan Colville; this version revised and supplemented by Angus Martin.] Colville, Duncan and Martin, Angus, 2009, The Place-Names of the Parish of Southend, Kintyre Antiquarian and Natural History Society, Campbeltown. [First published 1938 by the Kintyre Antiquarian Society; original list compiled by Duncan Colville; this version revised and supplemented by Angus Martin.] Cox, Richard, 2007, ‘The development of Old Norse-rð(-) in Scottish Gaelic’, in Cavill and Broderick, eds, 57–96. Cox, Richard A. V., 2007, ‘The Norse Element in Scottish Place-names: syntax as a chronological marker’, JSNS 1, 13–26. Cox, Richard A.V., 2007, ‘Notes on the Norse Impact on Hebridean Place- names’, JSNS 1, 139–44. Cox, Richard A.V., ‘Old Norse loan-words in Scottish Gaelic’, Proceedings of the Eighth Symposium of Societas Celtologica Nordica, ed. Jan Erik Rekdal and Ailbhe Ó Corráin, Acta Universitatis Upsaliensis, Studia Celtica Upsaliensis 7, 2007, 51–60. Cox, Richard A.V., 2008, ‘Tamhnaraigh ~Tamnabhagh: the development of Old Norse -fn(-) in (Scottish) Gaelic’, JSNS 2, 51–68. Bibliography 2006–2009 177 Cox, Richard A. V., 2008, ‘The syntax of the place-names’, inStudies on the Book of Deer, ed. K. Forsyth (Dublin), 309–12. Cox, Richard A.V., 2008, ‘Indicators of Ecclesiastical and Norse Settlement and History in the Toponymic Material in Appendixes H1–H8 of the Papar Project’, The Papar Project, Dr Barbara E. Crawford, University of St Andrews, et al., 17 pp. . Cox, Richard A.V., 2008, ‘Old Norse words for “boat” in Scottish Gaelic: Revisiting Henderson’s list’, Scottish Gaelic Studies XXIV, 169–80. Cox, Richard A.V., 2008, ‘Review Article: George Broderick, Placenames of the Isle of Man’, Nomina 31, 99–119. Cox, Richard A.V., 2009, ‘Goatfell, Gaoitbheinn, Gaoth Bheinn’, Scottish Gaelic Studies XXV, 303–29. Cox, Richard A. V., 2009, ‘Towards a Taxonomy of Contact Onomastics: Norse Place-names in Scottish Gaelic’, JSNS 3, 15–28. Crawford, Barbara E., 2006, ‘Houseby, Harray and Knarston in the West Mainland of Orkney. Toponymic indicators of administrative authority?’, in Names through the Looking Glass: Festschrift in Honour of Gillian Fellows-Jensen, ed. P. Gammeltoft and B. Jørgenson (C. A. Reitzels Forlag A/S, Copenhagen), 21–44. Crumplin, Sally, 2009, ‘Cuthbert the Cross-Border Saint in the Twelfth Century’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 119–29. Davies, John Reuben, 2009, ‘Bishop Kentigern among the Britons’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 66–90. Ditchburn, David, ‘The ‘McRoberts thesis’ and patterns of sanctity in late medieval Scotland’, in The Cult of Saints and the Virgin Mary in Medieval Scotland, ed. S. Boardman and E. Williamson (Woodbridge), 177–94. Dixon, Norman, 1947, ‘The Place Names of Midlothian’, unpublished PhD, Edinburgh University. Published 2009 in digital form on SPNS website with introductory notes by S. Taylor, . Drummond, Peter, 2007, ‘The Hill Names of Southern Scotland: a work in progress studying name change’, JSNS 1, 27–36. Drummond, Peter, 2009, ‘Close Compound Mountain Toponyms in Islay and Jura’, in A Land that Lies Westward: Language and Culture in Islay and Argyll, ed. J. Derrick McClure, John M. Kirk and Margaret Storrie (John Donald, Edinburgh), 50–61. Drummond, Peter, 2009, ‘Place-name losses and changes – a study in 178 Simon Taylor Peeblesshire: A comparative study of hill-names and other toponyms’, Nomina 32, 5–17. Drummond, Peter, and Tempan, Paul, 2009, ‘Close Compound Place- names in Ireland and Scotland’, in A Land that Lies Westward: Language and Culture in Islay and Argyll, ed. J. Derrick McClure, John M. Kirk and Margaret Storrie (John Donald, Edinburgh), 48–49; see also above under Drummond and below under Tempan. Edmonds, Fiona, 2009, Whithorn’s Renown in the Early Medieval Period: Whithorn, Futerna and Magnum Monasterium, 16th Whithorn Lecture 13 September 2008 (Whithorn). Edmonds, Fiona, 2009, ‘Personal Names and the Cult of Patrick in Eleventh-Century Strathclyde and Northumbria’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 42–65. Fellows-Jensen, Gillian, 2007, ‘The Scandinavian element gata outside the urbanised settlements of the Danelaw’, in West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300, ed. Beverley Ballin Smith, Simon Taylor and Gareth Williams (Brill: Leiden and Boston), 445–59. Fellows-Jensen, Gillian, 2007, ‘Some thoughts on English influence on names in Man’, in Cavill and Broderick, eds, 97–110. Filppula, M., Klemola J. and Paulasto, H., 2008, English and Celtic in Contact (New York and London: Routledge). Findlater, Alex Maxwell, 2008, ‘Another Look at Bagimond’, Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society 82, 59–75 [Bagimond in Dumfries and Galloway.] Fox, Bethany, 2007, ‘The P-Celtic Place-Names of North-East England and South-East Scotland’, The Heroic Age (an on-line Journal of Early Medieval Northwestern Europe) 10, . Fraser, Iain, and Halliday, Stratford, 2007, ‘The Early Medieval Landscape’, in In the Shadow of Bennachie: A Field Archaeology of Donside, Aberdeenshire, RCAHMS and Society of Antiquaries of Scotland (Edinburgh), 115–35. [Section ‘Place-names and the Pattern of Settlement’, 130–33; other chapters, by Iain Fraser, Piers Dixon and Angela Gannon, with much else of relevance to toponymics, historical geography, and parish and estate formation.] Fraser, James E., 2007, ‘Picts in the West in the 670s? Some thoughts on AU 673.3 and AU 676.3’, JSNS 1, 144–48. Bibliography 2006–2009 179 Fraser, James E., 2009, ‘Rochester, Hexham and Cennrígmonaid: the Movements of St Andrew in Britain, 604–747’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 1–17. Gammeltoft, Peder, 2006, ‘Scandinavian influence on Hebridean island names’, in Names through the Looking Glass: Festschrift in Honour of Gillian Fellows-Jensen, ed. P. Gammeltoft and B. Jørgenson (C. A. Reitzels Forlag A/S, Copenhagen), 53–84. Gammeltoft, Peder, 2007, ‘Scandinavian Naming-Systems in the Hebrides – A way of Understanding how the Scandinavians were in Contact with Gaels and Picts?’, in West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300, ed. Beverley Ballin Smith, Simon Taylor and Gareth Williams (Brill: Leiden and Boston), 479–95. Graham-Campbell, James, 2006, ‘Some reflections on the distribution and significance of Norse place-names in northern Scotland’, in Names through the Looking Glass: Festschrift in Honour of Gillian Fellows-Jensen, ed. P. Gammeltoft and B. Jørgenson (C. A. Reitzels Forlag A/S, Copenhagen), 94–118. Grant, Alexander, 2007, ‘Lordship and Society in Twelfth-Century Clydesdale’, in Power and Identity in the Middle Ages. Essays in Memory of Rees Davies, ed. Huw Pryce and John Watts (Oxford), 98–138. [Section on Lesmahagow place-names.] Hammond, Matthew H., 2007, ‘The Use of the Name Scot in the Central Middle Ages Part I: Scot as a by-name’, JSNS 1, 37–60. Hammond, Matthew H., 2008, ‘Queen Ermengarde and the Abbey of St Edward, Balmerino’, Cîteaux: Commentarii cistercienses, t. 59, fasc. 1–2 (Life on the Edge: The Cistercian Abbey of Balmerino (Scotland), 119–62. Hanks, P., Hardcastle, K., and Hodges, F., 2006, A Dictionary of First Names (2nd edition, Oxford). Henery, Robert and Taylor, Simon, 2007, ‘Pitmiclardie in Fife’, JSNS 1, 148–50. Higham, Mary, 2007, Of Names and Places: Selected Writings of Mary Higham, ed. Alan G. Crosby, published jointly by EPNS and SNSBI, with editor’s Introduction and an Appreciation by Margaret Gelling. [Chiefly on north- west England, with much of relevance to the place-names and historical geography of southern Scotland.] Hough, Carole, 2009, ‘Eccles in English and Scottish Place-Names’, in The Church in Place-Names, ed. E. Quinton (English Place-Name Society, Nottingham), 109–24. 180 Simon Taylor Hough, Carole, 2009, ‘“Find the Lady”: The Termlady in English and Scottish Place-Names’, in Names in Multi-Lingual, Multi-Cultural and Multi-Ethnic Contact: Proceedings of the 23rd ICOS, ed. Wolfgang Ahrens, Sheila Embleton and André Lapierre (York University, Toronto), 511–18. [Available on-line at ] Hough, Carole, 2009, ‘The Role of Onomastics in Historical Linguistics’, JSNS 3, 29–46. James, Alan G., 2008, ‘A Cumbric Diaspora?’, in A Commodity of Good Names: Essays in Honour of Margaret Gelling, ed. O. J. Padel and David N. Parsons (Donington), 187–203. James, Alan G., 2009, ‘*Eglēs / Eclēs and the formation of Northumbria’, in The Church in Place-Names, ed. E. Quinton (English Place-Name Society, Nottingham), 125–50. James, Alan G., 2009, ‘A Note on the Place-name Dreva, Stobo, Peeblesshire’, JSNS 3, 121–26. James, Alan G., 2009, review article on Paul Cavill and George Broderick, ed., Language Contact in the Place-Names of Britain and Ireland (2007), JSNS 3, 135–58. Jennings, Andrew, and Kruse, Arne, 2009, ‘One coast – three peoples: names and ethnicity in the Scottish west during the early Viking period’, in Scandinavian Scotland – Twenty Years After (The Proceedings of a Day Conference held on 19 February 2007), ed. Alex Woolf (St John’s House Papers No. 12, St Andrews), 75–102. Jesch, Judith, 2009, ‘The Norse gods in Scotland’, in Scandinavian Scotland – Twenty Years After (The Proceedings of a Day Conference held on 19 February 2007), ed. Alex Woolf (St John’s House Papers No. 12, St Andrews), 49–73. Kerr, John, 2007, The Atholl Experience [extensive collection of materials relating to the parish of Blair Atholl, the result of 40 years’ research; see Scottish Place-Name News 23 (2007) for more details]. King, Jacob, 2007, ‘Endrick and Lunan’, 150–56. King, Jacob, 2007, review article on George Broderick Placenames of the Isle of Man (7 volumes, 1994–2005), JSNS 1, 157–68. King, Jacob, 2008, ‘Analytical Tools for Toponymy: Their Application to Scottish Hydronymy’, unpublished PhD Thesis, University of Edinburgh. Available on-line at . King, Jacob, 2009, ‘Haberberui: An Aberration?’, JSNS 3, 127–34. [On a problematic early form of Inverbervie KCD.] MacIlleathain, Ruairidh, 2007, A’ Ghàidhlig air Aghaidh na Tìre: Ainmean- Bibliography 2006–2009 181 àite ann an Iar-thuath na Gàidhealtachd / Gaelic in the Landscape: Place names in the North West Highlands (Dualchas Nàdair na h-Alba / Scottish Natural Heritage, Battleby). McKay, Patrick, 2007, A Dictionary of Ulster Place-Names, second edition (Belfast; 1st edition Belfast 1999). McKay, Patrick, 2009, ‘Scots Influence on Ulster Townland Names’, Ainm: A Journal of Name Studies 10, 1–26. McNiven, Peter, 2007, ‘The Gart-names of Clackmannanshire’, JSNS 1, 61–76. MacQueen, John, 2008, Place-Names of the Wigtownshire Moors and Machars, Stranraer and District Local History Trust, Stranraer. [Covering the parishes of Glasserton, Kirkcowan, Kirkinner, Mochrum, Penninghame, Sorbie and Wigtown.] Márkus, Gilbert, 2007, ‘Gaelic under Pressure: a 13th-century charter from East Fife’, JSNS 1, 77–98. Márkus, Gilbert, 2008, ‘Saints and Boundaries: the Pass of St Mocha and St Kessog’s Bell’, JSNS 2, 69–84. Márkus, Gilbert, 2008, ‘Reading the Place-Names of a Monastic Landscape: Balmerino Abbey’, Cîteaux: Commentarii cistercienses, t. 59, fasc. 1–2 (Life on the Edge: The Cistercian Abbey of Balmerino (Scotland), 119–62. Márkus, Gilbert, 2009, ‘Balinclog: A Lost Parish in Ayrshire’, JSNS 3, 47– 64. Márkus, Gilbert, 2009, ‘Dewars and relics in Scotland: some clarifications and questions’, Innes Review 60, no. 2 (Autumn), 95–144. Morrison, Jenni, Oram, Richard, and Ross, Alasdair (with a contribution by Julie Franklin), 2009, ‘Gogar, archaeological and historical evidence for a lost medieval parish near Edinburgh’, PSAS 139, 229–55. Muhr, Kay, 2009, ‘Place-names and Scottish Clan Traditions in North-East County Antrim’, in A Land that Lies Westward: Language and Culture in Islay and Argyll, ed. J. Derrick McClure, John M. Kirk and Margaret Storrie (John Donald, Edinburgh), 79–102. Nicolaisen, W. F. H., 2007, ‘Gaelic sliabh revisited’, in Fil súil nglais: A Grey Eye Looks Back, ed. Sharon Arbuthnot and Kaarina Hollo (Ceann Drochaid /[Brig o’ Turk]), 175–86. Nicolaisen, W. F. H., 2007, ‘The change from Pictish to Gaelic in Scotland’, in Cavill and Broderick, eds, 111–22. Nicolaisen, W. F. H., 2008, ‘On river-names in the Scottish landscape’, in A Commodity of Good Names: Essays in Honour of Margaret Gelling, ed. O. J. Padel and David N. Parsons (Donington), 233–38. 182 Simon Taylor Oftedal, Magne, 1954, ‘The Village Names of Lewis in the Outer Hebrides’, Norsk tidsskrift for sprogvidenskap 17 (Oslo), 363–409, reproduced 2009 in booklet form with the title The Village Names of Lewis, The Islands Book Trust, Kershader, Lewis, with Foreword by John Randall, price £6. O’Grady, Oliver J. T., 2008, ‘The setting and practice of open-air judicial assemblies in medieval Scotland: a multidisciplinary study’, unpublished PhD thesis, University of Glasgow, available on-line at [of especial toponymic interest is Chapter 4, ‘Place-name evidence and the setting and distribution of early medieval assemblies’, pp. 125–224.] Ó Maolalaigh, Roibeard, 2008, ‘The Scotticisation of Gaelic: a reassessment of the language and orthography of the Gaelic Notes in the Book of Deer’, in Studies on the Book of Deer, ed. K. Forsyth (Dublin), 179–274. Orr, David G., 2008, Kirriemuir: Its Streets & Place Names (Kirriemuir) [see Scottish Place-Name News 26, Spring 2009.] Owen, Hywel Wyn, and Morgan, Richard, 2007, Dictionary of the Place- Names of Wales (Llandysul). Padel, O. J., and Parsons, David N., ed., 2008, A Commodity of Good Names: Essays in Honour of Margaret Gelling (Donington) [over 400 pages of toponymic riches, a fitting tribute to an outstanding scholar; much that is relevant to Scotland, beyond the chapters with explicit Scottish content, noted separately in this bibliography under James, Nicolaisen, Taylor and Waugh]. Puzey, Guy, 2009, ‘Opportunity or Threat? The Role of Minority Toponyms in the Linguistic Landscape’, in Names in Multi-Lingual, Multi-Cultural and Multi-Ethnic Contact: Proceedings of the 23rd ICOS, ed. Wolfgang Ahrens, Sheila Embleton and André Lapierre (York University, Toronto), 821–27. [Deals with cases from northern Norway, northern Italy, Ticino (Switzerland), Caithness and Dingle (Ireland).] [Available free of charge at .] Reid, John, 2009, The Place Names of Falkirk and East Stirlingshire (Falkirk Local History Society: Falkirk) [The extensive raw data underlying this rich and detailed study is now on-line at .] Ross, Alasdair, 2006, ‘Scottish environmental history and the (mis)use of Soums’, Agricultural History Review 54, Part II, 213–28. Ross, Alasdair, 2006, ‘The Dabhach in Moray: A New Look at an Old Tub’, in Landscape and Environment in Dark Age Scotland, ed. Alex Woolf (St Andrews), 57–74. Sandnes, Berit, 2006, ‘Toponyms as settlement names – of no relevance in Bibliography 2006–2009 183 settlement history?’, in Names through the Looking Glass: Festschrift in Honour of Gillian Fellows-Jensen, ed. P. Gammeltoft and B. Jørgenson (C. A. Reitzels Forlag A/S, Copenhagen), 230–53. Sandnes, Berit, 2007, ‘Describing language contact in place-names’, in Cavill and Broderick, eds, 123–36. Scott, Margaret, 2006, ‘Previck and Leckprivick : Onomastic Connections in South-West Scotland’, Nomina 29, 115–28. Scott, Margaret, 2008, ‘Words, Names and Culture: place-names and the Scots language’, JSNS 2, 85–98. Scott, Margaret, 2008, ‘Unsung Etymologies: lexical and onomastic evidence for the influence of Scots on English’, in Yesterday’s Words: Contemporary, Current and Future Lexicography, ed. M. Mooijaart and M. van der Wal (Newcastle [Cambridge Scholars Publishing]), 187–98. Scott, Margaret, 2008, Scottish Place Names (‘Say It In Scots’ Series, Edinburgh). Scottish Place-Name News: twice-yearly newsletter of the Scottish Place- Name Society, available on-line at . [Summaries of conference papers, with short articles and reviews; individual items not covered in this bibliography.] Sharples, Niall, and Smith, Rachel, 2009, ‘Norse settlement in the Western Isles’, in Scandinavian Scotland – Twenty Years After (The Proceedings of a Day Conference held on 19 February 2007), ed. Alex Woolf, St John’s House Papers No. 12, Committee for Dark Age Studies, University of St Andrews: St Andrews, 103–30. Stempel, Patrizia de Bernardo, 2007, ‘Pre-Celtic, Old Celtic layers, Brittonic and Goidelic in ancient Ireland’, in Cavill and Broderick, eds, 137–64. Storrie, Margaret, 2009, ‘Settlement and Naming in the Southern Hebridean Isle of Islay’, in A Land that Lies Westward: Language and Culture in Islay and Argyll, ed. J. Derrick McClure, John M. Kirk and Margaret Storrie (John Donald, Edinburgh), 17–47. Stuart-Murray, John, 2006, ‘Differentiating the Gaelic Landscape of the Perthshire Highlands’, Scottish Studies 34 (2000–2006), 159–77. Taylor, Simon, 2006, ‘The Early History and Languages of West Dunbartonshire’, in Changing identities: ancient roots – the history of West Dunbartonshire from earliest times, ed. Ian Brown (Edinburgh University Press), 12–41. Taylor, Simon, 2007, ‘The Rock of the Irishmen: an early place-name tale from Fife and Kinross’, in West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300, ed. Beverley Ballin Smith, Simon Taylor and Gareth Williams (Brill: Leiden and Boston), 497–514. 184 Simon Taylor Taylor, Simon, 2007, ‘Sliabh in Scottish Place-names: its meaning and chronology’, JSNS 1, 99–136. Taylor, Simon, 2007, ‘Gaelic in Glasgow: the Onomastic Evidence’, in Glasgow: Baile Mòr nan Gàidheal / City of the Gaels, ed. Sheila M. Kidd (Glasgow), 1–19. Taylor, Simon, 2008, ‘Pilkembare and Pluck the Craw: verbal place-names in Scotland’, in A Commodity of Good Names: Essays in Honour of Margaret Gelling, ed. O. J. Padel and David N. Parsons (Donington), 274–85. Taylor, Simon, 2008, ‘The toponymic landscape of the Gaelic Notes in the Book of Deer’, in Studies on the Book of Deer, ed. K. Forsyth (Dublin), 275–308. Taylor, Simon, 2009, ‘Ayrshire Place-Names: a rich seam still to mine’, Ayrshire Notes 38 (Autumn 2009), 4–18 [with an attempt at a complete bibliography of articles on individual Ayrshire place-names]. Taylor, Simon, 2009, ‘The Trenches at Falkland, Fife: a Legacy of Royal Deer-management?’, in Carmarthenshire & Beyond: Studies in History and Archaeology in Memory of Terry James, ed. Heather James and Patricia Moore (Carmarthenshire Antiquarian Society: Carmarthen), 235–44. Taylor, Simon, 2009, ‘Place-names of Lesmahagow’, JSNS 3, 65–106. Taylor, Simon, with Gilbert Márkus, 2006, Place-Names of Fife Vol. 1 (West Fife between Leven and Forth) (Donington) [5-volume series]. Taylor, Simon, with Gilbert Márkus, 2008, Place-Names of Fife Vol. 2 (Central Fife between Leven and Eden) (Donington) [5-volume series]. Taylor, Simon, with Gilbert Márkus, 2009, Place-Names of Fife Vol. 3 (St Andrews and the East Neuk) (Shaun Tyas: Donington) [5-volume series]. Tempan, Paul, 2009, ‘Close Compounds in Irish Place-names’, in A Land that Lies Westward: Language and Culture in Islay and Argyll, ed. J. Derrick McClure, John M. Kirk and Margaret Storrie (John Donald, Edinburgh), 62–78. Tempan, Paul, 2009, ‘Towards a Chronology of Topographical Elements in Irish Place-Names: Some Strategies for Establishing Relative Chronology’, in Names in Multi-Lingual, Multi-Cultural and Multi-Ethnic Contact: Proceedings of the 23rd ICOS, ed. Wolfgang Ahrens, Sheila Embleton and André Lapierre (York University, Toronto), 943–51. [Available free of charge at .] Tempan, Paul, 2009, ‘Sliabh in Irish place-names’, Nomina 32, 19–42. Thomson, William P. L., 2007, ‘The Orkney Papar-names’, in West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300, ed. Beverley Ballin Smith, Simon Taylor and Gareth Williams (Brill: Leiden and Boston), 515–37. Thomson, William P. L., 2008, Orkney Land and People (The Orcadian Limited, Kirkwall Press). [Two chapters specifically about place-names: Bibliography 2006–2009 185 Chapter 1: ‘Orkney Farm-names’; Chapter 13: ‘The Place-names of the Crofter Pioneer Fringe’.] Waugh, Doreen, 2006, ‘The -by / -bie names of Shetland’, in Names through the Looking Glass: Festschrift in Honour of Gillian Fellows-Jensen, ed. P. Gammeltoft and B. Jørgenson (C. A. Reitzels Forlag A/S, Copenhagen), 298–321. Waugh, Doreen, 2006, ‘Place-Name Evidence for Portages in Orkney and Shetland’, in The Significance of Portages: Proceedings of the First International Conference on the Significance of Portages, 29th September–2nd October 2004, in Lyngdal, Vest-Agder, Norway, arranged by the County Municipality of Vest- Agder, Kristiansand, ed. C. Westerdahl (BAR International Series 1499), 239–49. Waugh, Doreen, 2007, ‘Placing Papa Stour in context’, in West Over Sea: Studies in Scandinavian Sea-Borne Expansion and Settlement before 1300, ed. Beverley Ballin Smith, Simon Taylor and Gareth Williams (Brill: Leiden and Boston), 539–53. Waugh, Doreen, 2007, ‘From the ‘banks-gaet’ to the ‘hill-grind’: Norn and Scots in the place-names of Shetland’, in Cavill and Broderick, eds, 165–83. Waugh, Doreen, 2008, ‘‘A hōh! My kingdom for a hōh!’’, in A Commodity of Good Names: Essays in Honour of Margaret Gelling, ed. O. J. Padel and David N. Parsons (Donington), 410–15. Waugh, Doreen, 2009, ‘Caithness: Another Dip in the Sweerag Well’, in Scandinavian Scotland – Twenty Years After, ed. Alex Woolf (St Andrews), 31–48. [The proceedings of a day conference held in St Andrews on 19 February 2007.] Waugh, Doreen 2009, ‘Neglected Topographic Names: ness-names in Orkney and Shetland’, JSNS 3, 107–20. [This also appeared in New Orkney Antiquarian Journal 4 (2009)]. Williamson, Eila, 2009, ‘The Cult of the Three Kings of Cologne in Scotland’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 160–79. [Includes a discussion of rocks known as The Three Kings of Cullen BNF.] Williamson, May G., 1942, ‘The Non-Celtic Place-Names of the Scottish Border Counties’, unpublished PhD, Edinburgh University. Published 2009 in digital form on SPNS website with introductory notes by W. Patterson, . Wooding, Jonathan M., 2009, ‘The Medieval and Early Modern Cult of St Brendan’, in Saints’ Cults in the Celtic World, ed. Steve Boardman, John Reuben Davies and Eila Williamson (Woodbridge), 180–204. 186 Simon Taylor Woolf, Alex, 2006, ‘Dún Nechtain, Fortriu, and the geography of the Picts’, Scottish Historical Review 85 (2), 182–201. Woolf, Alex, 2007, Where was Govan in the Early Middle Ages? (Govan). Woolf, Alex, 2007, ‘The Cult of Moluag, the See of Mortlach and Church Organisation in North Scotland in the Eleventh and Twelfth Centuries’, in Fil súil nglais: A Grey Eye Looks Back, ed. Sharon Arbuthnot and Kaarina Hollo (Ceann Drochaid / [Brig o’ Turk]), 299–310. Young, Sheila, 2009, ‘Oil and gas field names in the central and northern sectors of the North Sea: their provenance, cultural influence, longevity and onshore migration’, Nomina 32, 75–112. Notes on Contributors

George Broderick is currently Professor of Celtic Studies at the University of Mannheim, Germany. His main area of research over the years has been Manx Gaelic language and literature and he has published widely on that subject, notably A Handbook of Late Spoken Manx 3 vols. (Tübingen: Niemeyer 1984– 86 ), Language Death in the Isle of Man (Tübingen: Niemeyer 1999), and Place- Names of the Isle of Man 7 vols. (Tübingen: Niemeyer 1994–2005). In recent years he has branched out into other aspects of place-names, publishing his ‘Names for Britain and Ireland Revisited’ (Beiträge zur Namenforschung 44.2, 2009) and ‘Kelten und Nicht-Kelten in Britannien und Irland ...’, Keltische Forschungen (Allgemeine Buchreihe) 1, 2009.

Eileen Brooke-Freeman’s interest in place-names stems from her childhood in Shetland learning local names from family members and an introduction to Scottish ethnology, through the Department of Celtic and Scottish Studies, whilst studying geography at Edinburgh. She trained as an archivist and worked in Lincoln and Chester for 13 years before returning to Shetland in 2001 to establish and run the Shetland Place Names Project.

Henry Gough-Cooper lives and works in Dumfriesshire. He has a special interest in the history and archaeology of early medieval Scotland and its neighbours. He is a member of the Scottish Place-Name Society and helps maintain the Society’s web site. He is currently preparing new editions of the Annales Cambriae B and C texts.

Professor Richard A.V. Cox’s interests include Gaelic language and literature, including modern poetry and prose writing, publishing, lexicography, linguistics, onomastics and Norse-Gaelic contact. Formerly at the Departments of Celtic at Glasgow and then Aberdeen, he is currently Professor in Gaelic language and literature at Sabhal Mòr Ostaig, Isle of Skye.

Dr Alan G. James read English philology and medieval literature at Oxford, then spent 30 years in school-teaching, training teachers and research in modern linguistics. He maintained his interest in place-name studies through membership of the English Place-Name Society, the Society for Name Studies in Britain and Ireland, and the Scottish Place-Name Society. After retiring, he spent a year as a Visiting Scholar in Cambridge University’s Department of Anglo-Saxon, Norse and Celtic, studying Celtic philology. Since then,

The Journal of Scottish Name Studies 4, 2010 188 The Journal of Scottish Name Studies 188 he has been working on the linguistic history of Northumbria and the Old North. ‘The Brittonic Language in the Old North: a guide to the place-name elements’ is currently being digitised to appear on the SPNS web-site, .

Dr Jacob King completed a PhD in 2008 entitled ‘Analytical Tools for Toponymy: Their Application to Scottish Hydronymy’ at the Department of Celtic and Scottish Studies, University of Edinburgh. He works for Ainmean- Àite na h-Alba as a researcher of Scottish place-names. He is a committee member of the Scottish Place-Name Society. Aside from his involvement in various digitisation projects, his interests include the study of Scottish aber-names and the use of Geographical Information Systems and database systems in place-name studies.

Denis Rixson is a retired schoolteacher living in Mallaig. He has written a number of books on aspects of Highland history: The West Highland Galley (Birlinn), Knoydart (Birlinn), The Small Isles (Birlinn), Arisaig and Morar (Tuckwell Press), The Hebridean Traveller (Birlinn). An interest in early levying arrangements led him into a study of land assessment structures on the west coast and in the Hebrides. This has occupied much of the last decade but is now complete and he is extending his research into other parts of Scotland.

Dr Simon Taylor has been working in various aspects of Scottish place- name studies since the early 1990s. His last contract was as lead researcher on the AHRC-funded project, ‘The Expansion and Contraction of Gaelic in Medieval Scotland: the onomastic evidence’, at the then Department of Celtic and Gaelic (now part of the School of Humanities), University of Glasgow. Part of this project’s outcome was the completion of a five-volume series, The Place-Names of Fife, four volumes of which are now published. He is also Convener of the Scottish Place-Name Society, which he helped found in 1995. Editor of JSNS since its inception in 2007, he is now co-editor with Richard Cox. 189 The Journal of Scottish Name Studies 189

Scottish county abbreviations (pre-1975)

ABD Aberdeenshire ELO East Lothian PEB Peeblesshire ANG Angus FIF Fife PER Perthshire ARG Argyll INV Inverness-shire RNF Renfrewshire AYR Ayrshire KCB Kirkcudbrightshire ROS Ross and Cromarty BNF Banffshire KCD Kincardineshire ROX Roxburghshire BTE Bute KNR Kinross-shire SHE Shetland BWK Berwickshire LAN Lanarkshire SLK Selkirkshire CAI Caithness MLO Midlothian STL Stirlingshire CLA Clackmannanshire MOR Moray SUT Sutherland DMF Dumfriesshire NAI Nairnshire WIG Wigtownshire DNB Dunbartonshire ORK Orkney WLO West Lothian

Journal of Scottish Name Studies 4, 2010 Scottish Place-Name Society Comann Ainmean-Àite na h-Alba

If you are interested in Scottish place-names, wherever you live, this is the society for you. Founded in 1996, we have grown in numbers to over 350 members, comprising everyone involved or interested in Scottish place- name study, from academics to amateurs to those who are simply fascinated by place-names.

We publish two copies of the Newsletter a year, each illustrated in full colour and now 16 pages long, with articles summarising talks at our conferences and other material, including book reviews. We hold day conferences twice a year – one in spring, one in autumn – at venues all round Scotland, usually with a local theme to its four or five talks. We run a website with lots of pages about the Society, its past and future conferences, notes on place-names arranged by county, a bookshelf, and links to other websites; you can find it at http://www.spns.org.uk/.

Membership costs just £6 per year (£7 if you live outside the UK), and includes the two Newsletters, invitations to the conferences, voting rights at our AGM, which is part of the spring conference, and a discount on subscribing to The Journal of Scottish Name Studies.

You can contact us via the web site, or by writing to the Society, c/o Department of Celtic and Scottish Studies, University of Edinburgh, EH8 9LD, or by contacting the Treasurer, Pete Drummond, at peter. [email protected], or 8 Academy Place, Coatbridge ML5 3AX.

We look forward to welcoming you. AINM A Journal of Name Studies published by the Ulster Place-Name Society, Belfast

edited by Gregory Toner

The next volume of Ainm: a Journal of Name Studies (Volume XI) will be published in December 2011. Ainm is devoted primarily to the study of Irish place- and personal names. Contributions on names in Scotland and the Isle of Man are also welcome, particularly where they interface with areas of Irish interest, as are contributions on onomastics in general.

All articles submitted for publication are peer-reviewed and the membership of the editorial board reflects expertise in Irish, Manx and Scottish place- and personal names in both the medieval and modern periods, and in related areas of language, literature and historical studies.

Articles for publication and books for review should be sent to the The Editor, Ainm, Irish & Celtic Studies, School of Languages, Literatures & Performing Arts, Queen’s University, Belfast BT7 1NN.

Subscriptions should be sent to The Treasurer of the Ulster Place-Name Society (UPNS), c/o Irish & Celtic Studies at the same address.

Ainm is free to members of UPNS, £15 to non-members and institutions.