5 Irish Monumental Sculpture: the DatingEvidence Provided by Linguistic Forms

Catherine Swift

J

In 1987, Dr Isabel Henderson published a paperentitled 'Early Christian monuments of Scotlanddisplaying cro!'>sesbut no other ornament'. In thisshe emphasised the testimony of cross-inscribedstones as indicators of the existence of Christian faithin areas for which we often lack documentaryevidence. She also stressed the major difficulties indating these monuments. In both of these obser.vations, as she herself makes clear, she was precededby Dr Ann Hamlin in her 1982 article on similarstones from Ireland. Within the Irish corpus, how-ever, is a sub-group consisting of those stones whichboth display crosses and aTe carved with inscriptionsin the ogam script. About these, Hamlin wrote 'it isfrom language that the best hope of dating comes'(Hamlin 1982, 283). My purpose in this article is tolook at the particulars of such a dating technique inthe hopeful expectation that the results from such astudy might be applied to monuments withoutin"Criptions - either in Ireland or elsewhere.For historical reasons, the class of ogam stones

ornamented with cross forms, amounting to rathermore than 10%of R. A. S. Macalister's 1945corpus,has been underestimated, partly, it must be said,lx-cause of the approach of Macalister himself. Inboth the ogam stone corpus (1945, iv-ix) and in hisbook TheArchaeology of Jreland (1949,328-43) Mac-alister argued that the ogam alphabet was inventedby druids who had learnt the Chalcidic version ofthe Greek alphabet in the sixth and fjfth centuriesBC, probably in northern Italy. At this stage, how-ever, the alphabet consisted of a series of handsignals based on the five fingers. For Macalister, itwas not until approximately the sixth century ADthat ogam began to be inscribed on stones (1949,334) but even at this stage, the primary purpose ofsuch inscriptions was a magical invocation of thedeceased rather than a simple commemoration. Inaddition, he accepted EoinMacNeill's argument that

a number of inscriptions referred to gods rather thanto human ancestors (Macalister 1949, 339; cf Mac-Neill 1906).According to Macalister's hypothesis, as Chris-

tianity began to spread, the missionaries of the newfaith occasionally accommodated the pagan script.More commonly, the Christians came into conflictwith the pagans and with what Macalister termed'implacable hostility' (1949,338) set about neutral-ising the pagan magic of the stones by carving thesign of a cross upon them. Macalister did acknow-ledge that in some cases the cross was carvedcontemporaneously with the inscription (a featurehe saw as a later development in the ogam stone-carving tradition) and in such instances, he felt, thecross would be found at the head of the stone. Inothers, however, where the cross is cut about 4 ft(1.25m) from the ground, he saw this as subsequentcarving on an already erect stone. In still others, hefelt that the stone must have been thrown down, thecross carved on the butt-end and then re-erectedwith the inscription reading downwards into theground and the cross triumphant above it (Mac-alister 1949,338-39).Several features of this model were criticised by

both linguists and archaeologists of Macalister's ownday, including Kenneth Jackson (1946), DanielBinchy (1946)and M. J. O'Kell}' (1945).In 1%1, EoinMacVVhitesuggested that some ogam letters mightderive from Christian symbols and drew attentionto the existence of what linguists had identified asan Irish vernacular version of the Christian hie iacitformula - the word ><01 or KGI (MacWhite 1960-1;Marstrander 1911; Pokorny 1915). In 1982, Hamlinargued that 'in at least ten cases' of the some 44cross-omamentcd ogam stones, there was evidencethat the cross had been added later. 'In others, thereis no indication or priority of ogham or cross andthey may be contemporary' (Hamlin 1982,285).

Catherine Swift50

Amongst historians, Macalister's views still re-main influential and have recently been endorsedby Daibhf 6 Cr6inln in his 1995 book Early MedievalIreland, the first general textbook on early Irishhistory to be written in 25 years. 6 Cr6inin states(1995,35),

In some instances, of course, crosses appear onmonuments but the evidence suggests that most, ifnot all, of these Wl;"re added much later to monu-ments erected in the fifth and sixth centuries,monuments which originally only bore inscrip-tions ... But though the evidence cannot be pushedtoo far, there is no escaping the fact that the ogamstone-sreflect a society in which the pre-Christianpractice of memorial in stone has remained totallyimpervious to the influence of Christianity and thisdespite the fact that the epigraphic habit is usuallybelieved to have derived from contact with theRoman world.

In archaeological circles, in contrast, there has beena tendency to follow Hamlin's lead: both MichaelHerity and John Sheehan have published dates forcertain cross forms deriving from the existence ofogam inscriptions upon the same stone. Sheehan(1990,168) sees the presence of ogam as indicating asixth-century date while Henty (1995, 90, 150, 156,310) argues for the first half of the seventh century.Nancy Edwards (1990, 103-4) quotes Hamlin inextenso in her account of ogams in her 1990 textbookwhile Harold My tum (1992, 38, 54-6, 66-9, 96),without referring explicitly to either Hamlin orMacalister, places the ogam stones within the tran-sitional phase between paganism and Christianityand agrees that in some cases the crosses arecontemporary with the inscriptions. In an importantarticle, which adds much new data to the debate,Fionnbarr Moore endorses Hamlin's views whilepointing out that 'linguistically early ogham stonesare, in the main, not cross-inscribed', and, contrari.wise, that the absence of explicitly Christian featureson a slab 'does not necessarily indicate a monumentwith a pagan background' (Moore 1998, 26-7).

The position of Charles Thomas in relation to thepaganism or otherwise of ogam monuments hasvaried over the last fifteen years although it shouldbe said that his view that the invention of the ogamscript was a pagan onc has remained constant. (Ideal with this issue below.) In 1987 Thomas wrote:

The proposition now is that the single, marked,memorial stone by an actual burial (or not) arose infifth-eentury Ireland because of the spread of Chris-tianity. Theoverall distribution of the Ogham stonesis in the broad region most affected by romanilasand the non-Patrician spread of Christianity inIreland. The common formulaic epitaph with nameand origin - 'Of A, of the ~n of B' - fulfilled aneed, perhaps no more th.m social assertion, arising

•

out of the Christian and Roman cont,act...(fhomas1987,8)

He then points out that in Roman Britain, theinscribed tombstone was by no means universal andonly four (and those dubious) have ever beenclaimed as Christian. This he sees as underminingthe proposition that the appearance of Latin-in-scribed memorials, which he views as making theirappearance in western and northern Britain by 500,automatically come from Gaul. He continues:

...Reconsidering this puzzle, the v,.riterwould air asuspicion that use of Christian memorial stones inlate fifth- and si:dh-eentury Atlantic Britain mayactually have been inspiretl from ireland and notvice versa ... Is it just possible that the late Romanpersonal memorial, tenuously conveyed as an ideato Ireland, took root there for particular reasons andwas later reconveyed to the provinces where it hadoriginated. (fhomas 1987,8-9)

And finally:

in those parts of Ireland where Ogham memorialsare least common (that is, the north and e,,;tremewest) we encounter another and ultimately moreimportant form: the uninscribed cross-omamentedstone. Its genesis is obscure as is the chronology ...By the sixth century we should have Ogham-inscriW stones that, like Britishmemorials, exhibitsimple crosses. (Thomas 1987,9)

These various statements are not (to me) entirelyexplicit - what are the precise connotations of theword 'should' in the last sentence for example? -and they involve certain very large assumptionsabout a range of issues which are the subject of muchscholarly debate. It does appear, however, that inthis article, Thomas is following Hamlin's line inassuming that a percentage of the ogam stones hadcrosses which were carved contemporaneously withthe inscriptions. In 1994, however, in his book AndShall These Mute Stones Speak?, Thomas reverts to theMacalister doctrine: Irish ogam stones are too earlyto be affected by Christianity and 'anything else onthese stones, like incised crosses, represents lateradditions unconnected with the epitaphs' (1994, 69).A different formulation, found in his latest article onthe subject, appears to represent a mixture ofMacalister's views on the role of pagan literati increating ogam mOnuments and Jane Stevenson'sarticle (1989) on the evidence for pre.Christianliteracy in Ireland:

The first applications of ogham to stone, in the formof simple memorials to individual dead, saying nomore than A sonof B;of A. of Q son of llle tribe of B; Asonof 8 of lhe tribe C (and further variations thereof)are not in themselves datable; only context maysometimesprovide a due ... If the inferred and partlyrecorded minor settlements of Munster Deisi in

Irish MQnumrntal Sculpture 51

o

a

e

u

?

T

D

Q

c

X!K c

?

1\1

?

G

R

L

v

B

sN

oFig. 5.] The ogam alpllabet

distinction between the date for the invention of thealphabet and the date of the first stones. The earliestversion of the alphabet on the stones is one whichhas already been modified from the presumedoriginal sequence of four groups of five strokes (Fig .5.1). Some groups do not occur at all and already onthe earliest pre-apocope stones, there are forftda orletters which have been added. This would indica Iea lapse of some, unquantifiable, time between theinvention of the alphabet and the earliest monu-ments as we have them. In other words, it is perfectlyplausible that Thomas's dating of the invention ofogam script to a pre-Otristian period is correct; I donot accept, for reasons well rehearsed elsewhere,that the bulk of the surviving stones themselvesbelong to a pre-Christian period (Swift 1996; 1997,27-69). This is not, of course, to say that the stonesnecessarily commemorate Christians; merely, thatChristianity had been introduced to Ireland and hadmade an impact on the Irish language by the timethe surviving ogam stones were created. We see thisimpact in the adoption of Christian Latin loan wordssuch as axal 'apostle', case 'Easter' or cruimther'priest' which by their (arm indicate that they wereborrowed into Irish before the various linguistic

Our growing awareness of the impact of romanitason early Ireland must reinforce the view that thevery idea of the inscribed memorial, the stoneproclaiming the name and filiation of the deceasedfor those who could read to read and appreciate,was taken from the pagan Roman empire (andvarious ways in which this could have happPOedcan be left aside) ... Ogham was surely invented byfili (sic - presumably for O. Ir. nom. pI. filid 1, theIrish literati, before and beyond Christianity andoriginally for this precise end ... All in all. one couldvery well argue that Ireland's literary revolutionbegan in Munster, within the Roman period andthat only such a revolution could have enabled thesurprisingly early eremitic monasticism of the fifthcentury to ha,...e taken root (Thomas 1998, 15).

As I understand it, this last interpretation sees theinvention of ogam as being the result of (pagan)contacts between the Roman world and Ireland andthat the pattern of ogam stone distribution reflects apagan Irish world affected by romanitas. At somesubsequent point, the 'romanisation' of this arearesults in its early adoption of an extreme form ofChristianity. The argument seems to imply a beliefthat the majority of all Irish ogam stones precedeand are separated chronologically from the arrivalof Christianity, an event which is said to haveoccurred in the fifth century or before. Such asuggestion would imply an extremely early date forthe creation of the vast bulk of Irish ogam stonesand one which is at variance with all other recentscholarship on the subject (see, in particular Mc.Manus 1991, 40-1 where the most probable date forthe vast bulk of ogam inscriptions is said to be thefifth and sixth centuries). It is tantalising. therefore,that Thomas has left the 'various ways in which thiscould have happened' to the reader's speculations .

I should (Xlint out here a (Xltential source of con-fusion; Anthony Harvey (1987) (by implication) andJane Stevenson (1989) have argued that the inventionof the ogam alphabet might be as early as the thirdcentury AD. Whether or nol this is the case (and theheavy degree of Roman influence in Ireland whichboth authors (Xlstulate was not substantiated by theconclusions of a recent 1997 conference on thesubject of Romans in Ireland), there is a clear

south-west Wales began near the end of the fourthcentury AD, there are a few menhir-like pillars inPembrokeshire with ogham script (alone) anddemonstrably Irish names thai are probably for non.Christians and should be assigned 10 the early fifthcentury ... I want to present ogham as, to all intents,a parallel to Roman writing; and the latter as aninnovation already entrenched in a large part ofpre-400 Ireland. Any distribution of literacy is notautomatically also a distribution of contemporaryChristianity (Thomas 1998, 10-11).

And in conclusion:

,,

•n

oemno

I;A

"'~n.yrUyi in

n-,it

rI,

,,

n,chininad.thndherlyonler9).onofinn',.•n

Y,I

n,.dd

Iy,.

52 Catherine Swift

changes which occurred during the period of ogamproduction and which are traceable in the ogaminscriptions (McManus 1983, 48).

My own view on the subject of contemporaneityor otherwise of cross forms on ogam stones is aslightly modified version of Hamlin's thesis, ex-panding on some points which she dealt with onlysummarily and discussing evidence which hasbecome apparent since her article was written. Iwould concur that a percentage of ogam stones havecrosses which are contemporary or indeed earlierthan the ogam inscriptions. Amongst these, the mostimportant is probably the stone from Church Islandwhere the ogam inscription clearly overlies the cross(O'Kelly and Kavanagh 1954). There aTe also anumber of stones in which, though it is impossibleto detect a physical relationship between the crossand the inscription. the existence of the Irish ver-nacular version (><01) of the largely Christian hieianf burial fonnula or the reference of a Latin-namedindividual on a stone, with both cross and in-scription, increases the likelihood that the cross iscontemporary (Swift 1997, %, 107). I would alsoagree that there is some degree of overlap betweenthe ogam period and that of the cross-earved pillarstones (Hamlin 1982,. 285; Swift 1997, 70-83).

In parenthesis, it should, perhaps be noted herethat any argument based on Macalister's corpuswithout examination of the stones themselves cannotbe a conclusive onc. It is clear from the fieldworkdone to date that Macalister's identification of crossforms is not always convincing. The stone fromLecan, Co. Kilkenny, for example, is identified byMacalister as having a Latin cross at its head (1945,38-9); examination suggests that while there is somepossible pocking running in a north-south line, thetransverse arm is a natural fissure.

There appear to be cases where ogam stones havebeen reused at a later date as cross-ornamentedpillars, with no cognisance being taken of theiroriginal function. An example of this would beKilfountan in the Dingle peninsula. A pillar onwhich the ogam EQO(?)DD is carved on one of itsnarrow sides was subsequently decorated with across and Latin letter inscription 'Fintan'.' This, tomy mind, is a matter of practicality rather thanideology; many of the ogam stones are of the samesize and basic shape as the cross-marked pillars and,as Hamlin points out (1982, 285), many ogams,amounting to some 34% of Macalister's corpus, arefound on ecclesiastical sites and would thus be aready resource for later stone-workers. Moore (1998,23) has now refined these figures: a national figureof 133 stones from 65 ecclesiastical sites and aMunster figure of 108 from 54 ecclesiastical sites.

Without going into details which I have arguedelsewhere (Swift 1997, 34-48), I would see the

function of the cross-marked ogam stones as beingfrequently (though not consistently) different fromthat of the cross-marked pillar stones withoutinscriptions. The former I see as being for the mostpart grave markers and as such influenced bycontemporary Christian burial inscriptions in north-west Europe. The latter - or at least a percentage ofthem - I see as Christian estate-markers, the in-spiration for which stems from Old Testamentreferences to boulders being used as estate andboundary markers.

As Hamlin pointed out, there are a large numberof stones where the cross is not clearly associatedwith the inscription and many of these have no otherindication of Christianity such as ><01 or Latin-named individuals. On such stones, distinctionsbetween crosses which might mark the Christianfaith of the deceased or landowner, and crosseswhich represent reuse of earlier orthostats whoseoriginal purpose had been forgotten, are not easy todraw. They can only be established through examin-ation of the stones themselves and not always then(Hamlin 1982, 293).

The corpus of Irish cross-carved stones for whichHamlin called in 1982 has not yet materialised andis an increasingly urgent priority given at least tworecent cases when early medieval stones, lyingunsupervised in rural sites, were removed bythieves.2 Giant steps, however, have been made inrecent years in the systematic linguistic analysis ofogam inscriptions. Unlike Latin,.the Irish languagegoes through a chronological series of identifiablelinguistic changes during the period in which ogaminscriptions were written. The evidence for thesechanges has been explored in detail by DamianMcManus in AGuide toOgam (1991), developing andoccasionally supersroing Kenneth Jackson's (1953)similar work on British stones. Though written forlinguists in a manner which can be somewhatopaque for those outside a highly-specialisro field,McManus's work is crucial for archaeologists andart-historians. In it, he provides clear sub-divisionsof the Irish ogam inscription corpus, allowing us forthe first time to say, not merely that a cross formbelongs to the period of the ogarn stones, but that aparticular cross fonn is attested early in that se-quence, on what are probably fifth-century stones,while another cross fonn could appear late in thesequence, probably late sixth- or early seventh-century. In these islands, where so many of our earlyChristian monuments displaying crosses but noother ornament are divorced from their originalcontext, this provides a fundamental chronologicalbench-mark of a kind often lacking.

In brief, the three basic linguistic sub-divisions ofthe ogam-inscription corpus are entitled: pre-apoc-ope, pre-syncope and post-syncope and I outline the

________________ I_"_.,_h_M_O_"u_m_'"_'_'_I_S_cu_'_p_,u_'_, '1 5_3

review of the Corpus states (1945, 152) that of the 30readings he checked, he felt eight were incorrect anda further ten were dubious. Some carvers appear tohave been conservative and kept to the old spellingwhile their contemporaries were moving on to newfonns. Formula words such as the word for 'son' -pre-apocope MAQQI, subsequently MAQ(Q) orMAC(C) - which occurs on the vast majority ofinscriptions, seem to have retained an old spellinglong after personal names had evolved, Moreover,while the system is a relatively convincing methodof sub-dividing the ogam stone corpus into early,middle and late, it is at its weakest when proposingspecific dates for any particular stone. Furthermore,although the end of the sequence can be establishedfrom the forms extant in seventh-century texts, thebeginning of the sequence is very difficult to estab-lish, All we can say to date is that the stratum ofIrish found in the earliest or pre-apocope stones isone in which a number of ecclesiastical Latin tennshas already been borrowed into the Irish language.This would imply a late fourth. or fifth-centurystarting date; this allows time for Constantine'sestablishment of Christianity as a major religion inthe early fourth century to percolate through to anisland beyond the north-western Roman frontier.Despite these complications, McManus's lin-

guistic dating system provides us with the only clearsubdivisions of the Irish corpus and, in the absenceof any excavated Irish ogam stone in an originalcontext, it provides our only hope for dating thecross forms with which they are sometimes associ-ated. Elsewhere I have argued that it is probablyalso a better method of sub-dividing the whole ogamcorpus than the rather too specific dates sometimesput forward on the basis of epigraphical analysis ofLatin-alphabet inscriptions on the Welsh ogamstones by Jackson (1953) and Nash-Wi11iams (1950)(given the large number of unprovable assumptionson which their chronology is predicated; Swift 1997,56-62), For the rest of this paper, therefore, I wouldlike to examine some of the archaeological conclu-sions which one might draw, using the linguisticdating system, .I begin with the stone from Emlagh East, Co.

Kerry, which is first recorded by Edward Lhuyd atthe beginning of the eighteenth century when itstood upright in a field near the beach where it nowlies (Brash 1879, 173-74). The inscription runs up theleft-hand side of the stone to a point at the headwhere the edge of the stone is somewhat rough.MacaHstcr argued (1945, 173; Fig. 5.3) that this wassubsequent damage which eliminated part of alonger inscription; I would agree with the DingleArchaeological Survey (Cuppage 1986, 255) andDamian McManus (1991, 66) who saw no suchindication on the stone. The extant inscription reads

. ..

Pre-apocope

Post-syncope

Pre-syncope

•fig. 5.2 TIlt three basic linguistic sub-divisions of theogam-inscripfion corpus: pre-apoeopt, pre-syncope and

post-syncope

sequence in Fig. 5.2. Apocope is the term for the lossof the final syllable of words and it is the first majorlinguistic development which we can detect in theinscriptions; McManus dates its onset to the bcgin~ning of the sixth century although it takes some timeto be accepted. by all carvers. Syncope is the loss ofmiddle syllables in three-syllable words: somethingwhich begins around the middle of the sixth centuryand which spreads during the second half of thatcentury. Inscriptions where there is no trace of thepre-syncope spelling are termed post-syncope andare dated to the end of the sixth and early seventhcentury and it is Irish of this post-syncope typewhich is revealed in the earliest manuscript evidenceof later seventh-century Old Irish (such as the place-names and personal names in the Patrician lives byTlrechan and Muirchu: McManus 1991, 92-7).There are complications, of course. The system is

predicated on the accuracy of Macalister's readingswhich have still to be consistently rechecked; it mustbe a worrying consideration that M. J, O'Kelly in his

COLOMANN

COLMAN

COLOMAGNI

-

Ciltflerinl:' Swift

Fig. 5.3 Inscription on tile stone from Emlagh East, Co. Kerry

Primitive Irish word ><01 - seen by McManus andhis predecessors as a vernacular reflection of theChristian hie iacit burial formula. There is, therefore,good evidence for assuming that this cross is broadlycontemporary with the inscription. Again it is a pre-apocope formulation which suggests a fifth-eenturydate for the cross.Another two stones from Ballintaggart have

equal-armed crosses in the middle of the face. Thereis no physical relationship between the inscriptionsand the crosses to indicate their relative date but oneinscription includes ><01 (Macalister 1945,152, 155;McManus 1991, 65). In both instances, the inscrip-tions are pre-apocope or probably fifth century indate.The stones, collected together today in the re-

cently walled enclosure at 8allintaggart, were foundin and around this site and all are of the local, water-rolled boulders from Minard beach, immediately tothe south. Given this, the two stones with ><01inscriptions, the two pre-apocope stones with plainLatin or Greek crosses and a fourth stone to bediscussed below, with a different cross type and alsoassociated with a pre-apocope inscription, it seemsclear that at Ballintaggart, we have evidence of aChristian community of the fifth century. The gravesof this community could be marked by boulderswith ogam inscriptions and plain Latin or Greekcrosses. It seems likely that the fifth-eentury crossand inscription at Emlagh East, the next townland tothe east of Ballintaggart, is another indication of theexistence of fifth-eentury Christians on the southcoast of the Dingle peninsula. This would coincide

54

BRUSCCOS MAQQI CALUACI or in English 'be-longing to 8-, son of C'. All three words are pre-apocope Le. they have not lost their final syllables.Thus it belongs to the earliest sub-division of theogam corpus, a group which probably belongs tothe fifth century.The importance of this date (or the inscription lies

in the relationship between the cross and the ogamscores of the letter L. It seems quite clear on examin-ation of the stone (Macalister 1945, 173; Cuppage1986, 255; McManus 1991, 54; Swift 1997, lOB)thatthe second score of the letter L has been shortened inorder not to run into the arm of the cross, a featurewhich indicates that the cross is earlier than theogam (though how much earlier we cannot tell.)Here, then, we have a plain Latin cross whichappears to be fifth century or earlier.A similar cross is found on one of the stones from

Ballintaggart, also in Co. Kerry. Macalister argued(1945,157) that the cross was inverted with respectto the writing - in other words the upper arm islonger than the lower one. Examination of the stoneindicates that Macalister was not making allowancesfor a natural fissure in the face of the stone whichruns along its entire length (see Cuppage 1986,fig.147;Swift 1997, 109, pl.). The carver apparentlyused this fissure to guide him in creating the north-south line and exact dimensions of the latter aretherefore extremely difficult to ascertain. The scoresrepresenting N do not run into the cross, indicatingsome care on the part of the carver of either theinc;cription or the cross given their close proximity.Most importantly, the inscription includes the

Iris" Monummtal Sculpture 55

with the evidence from Reask towards the north ofthe Dingle peninsula where the earliest radiocarbondate, apparently contemporaneous with the firstphase of the cemetery was cal. AD 410-600 (1 sigma:UB 2167, 1565:190BP) (Fanning 1981, 79-86. i13-15.121,164).This is not to say that all plain Latin or Greek

crosses found on ogam stones are necessarily of fifth-century date. An ogam stone from Whitefield. nearKillarney has a pre-apocope inscription with a Greekcross which might, therefore. correspond in date tothe fifth-century Ballintaggart stones, but there arealso four stones from Coolineagh. Co. Cork,Curraghmore West. Killogrone and Ratass. Co.Kerry, all of which include the word ANM in theirinscription. This is a post-apocope version of the OldIrish word ainm, 'name'; therefore these four in-scriptions are at least of sixth-century date (Moore1998, 28-9). The use of this word in ogam in-scriptions has been compared by Vendryes withChristian inscriptions, particularly those from NorthAfrica. but possibly also two examples from Wales(Yendryes 1955).The position of the crosses on these ANM stones

differ. At Killogrone, in south Kerry. the cross is atthe base of the stone. reading the inscription up theright-hand side of the cross. Brash tells us that onfirst recording, the pillar was inverted with the crossat the top and stood as the headstone of a grave(Brash1879. 239-41). It is not clear that Brash himselfhad seen the stone in this location. Here, then, theremight be evidence of Macalister's theory that thecros~ were added subsequent to the inscription.Having said this. however. there are examples oflater Irish Latin-letter inscriptions where crossesoccur below the writing (Lionard 1960-1, 102).Looking at the Killogrone monumcnt with the crossin the middle of the broad face at its base. theinscription runs from the bottom of the right-handside up towards the top. Normally, the inscriptionsrun from the viewer's left hand to the top (McManus1991, 47) but if one reversed the pillar so that theinscription read in this fashion (with the cross nowpositioned at the top of the pillar) the scores of theogam would be reversed - reading Q for N and soforth (Fig. 5.1). The resultant reading would notmake se~. It cannot be stated conclusively. there-fore, that the position of a cross towards the base ofa pillar must automatically be scen as post-datingthe accompanying inscription. The only date that canbe put forward for the Killogrone stone is a date forthe inscription which can simply be dated to the endof the sixth century or beginning of the seventh.At Ratass, in north Kerry. Thomas Fanning also

argued that the cross was inverted with respect tothe inscription and should therefore post-dale it. Inthis case, however. there was evidence that the pillar

was used for sharpening and polishing bladessubsequent to its use as an ogam memorial, and thatthe cross in turn overlay some of the polished area(Fanning and 6 CorrAin 1977). The inscription ispost-apocope but pre-syncope. indicating a datetowards the middle of the sixth century for theinscription and a somewhat later date for the cross.In the case of the Curraghmore West inscription.

in south Kerry. there is no relationship between theinscription and the cross, but the cross is located inthe middle of the broad face about two thirds of theway up, with the ogam inscription running up theleft edge. Moore (1998,29) takes it that the two arecontemporary. The inscription is certainly post-apocope and may possibly be post-syncope. How-ever the last person to publish a reading (McManus1991,176) had considerable doubt about Macalister'sreading which, if it were correct, would indicate adate well into the seventh century. Making allow-ances for this possibility. it seems best to argut' thatsimple Latin and Grft'k crosses, as witnessed onthese various stones. probably occur throughout thedate range of the ogam corpus, from the fifth centuryto the seventh.Other cross forms can be more distinctive and

may, therefore, have had a shorter lifespan. Theshape of the stone from Drumconwell, Co. Armagh,makes it fairly clear that the monument was de-signed to be looked at from the same side as iscurrently on view. Here. a simple ringed cross withstem extending below the lower line of the circle,lies in the centre of the upper face (Fig. 5.4). Theinscription is unusual in layout, running in two linesfrom bottom to top: first on the left-hand side andthen on the right. There is no physical relationshipbeho.'eencross and inscription. My reading agreeswith that of all commentators other than Macalisterin ways which are crucial for dating purposes;Macalister's reading (1945, 298) would indicate apre-apocope or fifth-century reading; I found thatlike William Reeves, John Rhys, Ann Hamlin andRichard Warner. I could not see the crucial finalletter in QETAI(S) (see Warner 1991. 45 for refer-ences to all of the above). The lack of such a letterindicates the onset of apocope (McManus 1991. 82.85-7; Swift 1997,51). implying a sixth-century date.r should add that I found no evidence in favour ofRichard Warner's suggested emendation of sevenscores of the inscription to produce the name of theeponymous Ulster hero ConmAel whose name isincorporated in that of the townland (Warner 1991.45-6). As Warner himself states, the consonantscores are all clear and while wholesale spellingerrors on the part of an Archaic Irish carver tran-scribing his native tongue are of course possible, itdoes not appear probable to me.The fact that this cross type is associated with

56

,

Catherine Swift

••••••

Uupffo

Llan ••••• el.

Port TMt><>!

Fig. 5.4 Crossfonns occurring on stones from Ireland and Wales

what is probably a sixth-century inscription isparticularly interesting, in that it is a form which iswidespread on cross-marked pillars from bothIreland and western Britain. From Ireland there arcexamples from Aghacarrible (from a souterrain sitewhich also contained two ogam stones), Ahane,Ballydarrig, Ballymorcreagh (a church site withogam stones), Beginish, Illauntannig, Skellig andClio North.. all in Co. Kerry (Cuppage 1986, 103-4,268-69, 290-92. 295-96; O'Sullivan and Sheehan1996. 251, 259, 261-63, 281) as well as a closelyrelated type from Inishmore, Co. Galway (Higgins1987, ii, 125). From Wales there are examples of

similar stones from Liantrisant and Port Talbot inGlamorgan, LJawhaden, Morvil and St David's inPembrokeshire amongst others (Nash. Williams1950, nos 219, 262, 342, 350, 372).1There is also anexample from Maughold on the Isle of Man(Kennode 1994, pI. VII:10) and another from lona(RCAHMS 1982, 180-81).

In one example. from Trallwng in Brecknockshirethe cross is accompanied by both ogam and Latin-letter inscriptions; Nash-Williams suggested that thecross post-dates the Latin letters (Nash-Williams,1950, no. 70). Looking at the stone with the circle inthe middle of a broad face, at the top of the shaft, the

Irish MOllumental Sculpture 57

inin

"'>nIn

la

re,-••n

J

ogam inscription funs along the right-hand sidefrom top to bottom while the latin.Jetter inscription,apparently referring to the same person, runs 'up-ward facing right', i.e. with the bottoms of the lettersto the right.hand side of the viewer. (The termin-ology is that of Okasha 1993, 28.) The arrangementof both the agam-scores and the Latin-Jetter in-scription is not the norm and {or the ogam scores toparallel the Latin letters in their pr<.>sentposition onemust reverse the scores, reading C for S, N for Q etc.(Such a reading would not make sense.) If oneassumed, however, that the stone was inverted andthe cross added later, the arrangement of both theogam and the Latin-letter inscription would be bothnormal and consistent with each other, In otherwords, as Nash-Williams pointed out. the ogaminscription lies on the original left-hand edge,reading upwards with the Latin inscription in twovertical Jines reading downwards, facing left. Su~sequently, the stone was inverted and reused. withthe upper part being incised with a ring-eross (ibid.;Lewis and Redknap, in prep.).

If the father's name refers to an Irishman. it is apre-apocope form of fifth-eentury date; if it is British,Jackson (1953,185)suggested the early sixth century(McManus 1991, 98, 113).A dating range later thanthe early sixth century is congruent with thatsuggested for Drumconwell (before end of the sixthcentury). A third example of this cross form. fromL1anllyr in Cardiganshire, is on an opposing face toan inscription which includes post-apocope andprobably post-syncope forms of Irish names (Nash-Williams 1950, no.124). If this inscription is to beassociated with the cross. on the other side of thestone, it would extend the dating range of this crossfonn into the seventh century and beyond.

A cross on the Ballynahunt stone, Co. Kerry. iscarved in double lines with square terminals andwhat appears to be a triangular base (Macalister1945.164-65; Cuppage 1986,253-54). It seems to bea case where the cross is not contemporary with theinscription. for if it is right to read the triangle asbase, the inscription is inverted. running from thetop downwards along the right-hand side andcontinuing around what is the bottom of the stone inthis position nus is not a case where a small crossmay Ile either above or below the inscription asargued above for Killogrone, Co. Kerry; rather, theinscription is unreadable when viewed with thecross apparently upright and the reader must re-verse all scores in order to make the two compatible(reading 0 for L, N (or Q etc.). Such a 'reversed'reading does not make sense. Thus Macalister isprobably right (1945, 164) when he states that thecross is later than the inscription. The words stillhave final endings and the inscription is, therefore,of pre-apocope or probable fifth-«,ntury type. This

would mean the accompanying cross could be datedto some period after that date _ a rather open-endeddating.

A cross potent with rectangular terminal atDromkcare, Co. Kerry, is associated with a cross o{similar form. Here the inscription runs from bottomleft to top as normal and is of post-syncope form. Le.late sixth or early seventh century. As withDrumconwell and Curraghmore West, the cross ispositioned in the middle o{a broad {ace,about two-thirds of the way up the shaft and - without beingassociated physically with the inscription - there isnothing problematical about seeing the two ascontemporaneous. This single example, with theambiguous parallel at Ballynahunt, may mean thatwe should see the cross-potent on this Irish stone asbeing a late sixth or early seventh century phenom-enon.

Another example from the collection o{ stones atBallintaggart, Co. Kerry"has a more complex cross,with trident-shaped trifid terminals (Fig. 5.4) thanthe plain Greek or Latin crosses {rom that sitediscussed previously. This is a stone with twoinscriptions. Looking at the {ace with the cross in thecentre of a broad face in the upper third o{ the shaft,one of the inscriptions runs from boltom left tomiddle of head according to the normal conventionand is a unique memorial to the three sons o{*Mailagnas; the {arm o{ the name suggests a pre.apocope date in the fifth century. The other runsfrom bottom right to middle of head and, as withBallynahunt, demands that the reader reversesscores, reading C {orapparent 5 and so forth, if it isto be made compatible with an upright cross {acingthe onlooker. It is a single name inscription but thename form *Curcitti is not clearly diagnostic andcould be of any period within the ogam-inscriptionrange. The two inscriptions do not appear to berelated to each other. On the whole, it appears morelikely that the cross is to be associated with the pre-apocope Mailagna..'i inscription which would suggesta fifth-century date; if it is to be associated withCurcitti, one must either reverse the scores as alreadymentioned or assume the cross is carved on the backo{ the monument.

John Sheehan (1994, 28) has drawn attention tothe Mcrovingian parallels for this particular crossform, citing the evidence o{ the fifth- to seventh.century sarcophagi {rom Poitou, discussed byEdward James (1977, 71). It is also a style whichoccurs on a variety of cross-inscribed boulders both{rom Co. Kerry and Co. Galway (Sheehan 1994,29),from Cloontuskert, Co. Longford (Farming and 6h~i1idhe 1980,15)and {rom elsewhere in the BritishIsles (see, {orexample, Nash-Williams 1950,no. 118(rom L1anddewi-brefi, Cardiganshire). A Welshexample, {rom L1andeilo in Pembrokeshire is ac-

,

58 Cathen'ne Swift

,I

,,

I,I'

I

companied by both an ogam and a Latin-letterinscription (Nash.Williams 1950, no. 313). If thecross is viewed in the middle of the broad face, atthe upper end of the shaft, the ogam inscription runsup the left-hand side from bottom to top. The latin-letter inscription is also vertical but reads 'down-wards facing left'. Both of these positions representthe norm for ogam and Latin-letter inscriptionsrespectively. It is inscribed in Roman capitals forwhich Nash-Williams suggested a fifth- or earlysixth--eentury date; the names - (A)NDAGELLIMACU CA V(ETI) - include one, ANDAGELU,which is pre-syncope and possibly pre-apocope.Another stone from the same churchyard is dedi.cated to another son of Caz.>etus, called .Coimagnasbut inscribed only in Latin letters. The name on thissecond stone is a pre-apocope form of the Old Irishname, Coffman and is probably fifth century.

In short, the names associated with this particularcross form indicate that this type of cross wasapparently being produced in these islands by theend of the fifth century. Furthennore, there is noclear distinction in age between the Welsh exampleand that from Co, Kerry, indicating the probabilityof contemporary contact between the two areas. Themanuscript spelling system of the Old Irish vernacu-lar, as found in late seventh- and eighth-eenturytexts, differs from that used in ogam and insteadderives from the spelling conventions of VulgarLatin as spoken by British speakers, It has been along-standing dictum of early Irish ecclesiasticalhistory that the manuscript spelling system whichhad replaced ogam orthography by the classical OldIrish period in the eighth century was brought toIreland by Welsh missionaries, during the later fifthand sixth centuries (MacNeill 1931; O'Rahilly 1957,40-6; Stevenson 1989, 144-47). Archaeologically,however, we have not, as yet, investigated in anydetail the possibility of links between the two islandsin the cross forms on the memorial stones.

This brings me my last category of cross formsassociated with ogam stones and that is the crosstype known as the Maltese cross or cross-of.arcs. In1991, Peter Harbison suggested that the Irish versionof this cross type was associated with pilgrimageand possibly belonged to an eighth- or ninth-eenturydate. The pilgrimage associations were based in thefirst instance on the Maumanorig inscription fromDingle which Macalister read as ANM COLMANAIl.ITHER or 'the name of COl.MAN the pilgrim'(Harbison 1991, 75, 84; Macalister 1945, 191). Un.fortunately this is one of Macalister's odder tran.scriptions; subsequent commentators, such as theDingle Archaeological Survey (Cuppage 1986, 333-34) and Damian McManus (1991, 67) are agreed thatthere is no evidence (or recognisable letters beyondANM COl. ... The evidence (or an eighth- or ninth-

century date range is also based on what is no ••.•outdated Jinguistic analysis by Donnchad~Corrain of the Ratass stone in 1977 (Fanning an,Corrain 1977, 18; see now McManus 1991, 'Finally, it is worth noting that Harbison's malthis cross type (1991, 193) includes the simple crlof.arcs together with a number of highly develoJand ornate forms whose chronological relationsto the plainer (orms is not clear.

In a recent study (1997, 70-83) I argued that wl1the simple cross-o(.an:s is (ound associated ••.•ogam inscriptions, they are all associated with I<inscriptions of later sixth- or seventh-eentury d;the classic example being that of the pre-sync(stone at Arraglen, Co. Kerry, which is a rare examo( an ogam memorial where the role of the ccmemorand is recorded: 'of the priest, •Ronan" sor.Comogann', I suggested that the origin of Sl

designs might be sought in the Christian memo!stone tradition as represented in Iberia, MerovingFrance and as far away as Egypt. Since this suggtion was made, Derek Craig's publication of 1Whitham stones has indicated that not only is thoa collection of these stones from Whithom,. bufragment of one such was found in a mid. to I.seventh.century grave (Craig 1997, 439), datthrough stratigraphical analysis. The excavarsuggested this may have been debris from tadjacent shrine o( an earlier seventh-eentury datia date which is arrived at totally independentlythe linguistic analysis of the ogam associations, tone which has produced a remarkably similar resL

Craig's work also draws attention to thecollecti,of similar cross-of-arc stones from Man and the hstones from Wales while other examples o( this crctype are (ound in St Boniface's Church, PalWestray, Orkney (see current exhibition in tlMuseum of Scotland, Edinburgh). In Ireland, tltraditional picture of the distribution of these crosshas been one which has (ocussed very strongly (the west coast, although Uonard pointed out ththere was a collection o( the type from Gallen (19611, 110-12) and other stones have been identifi(from both the north and east midlands. To this 01

might add that the current Duchas inventory of son700 graves labs from Clonmacnoise includes at leaseven examples of the cross-o(-arcs as well as oth.examples of the associated marigold type. On one I

these, an uncial inscription in horizontal linebeneath a cross-of-arcs, calls for a prayer (or sombody called Muirethach; the medial TH indicattthat this is likely to be of late seventh. or earleighth-century date preceding the later spelling witmedial D (Lionard 1960-1, 111). Two of thesClonmacnoise monuments, including the Muinthach stone, are pillar-shaped; possibly indicatingcontinuity of (arm from the ogam period into a phas

60 Catherine Swift

Ireland. Studiespr~ntl'd to C. A. Ral~h RD.dJord (- BAR.Brit. Set. 102), 283-96. Oxford.

HARBISON, P. 1991, Pilgrimage in Ireland: the Monumentsand till' Prop/e. London.

HARVEY, A. 1987, 'Early literacy in Ireland; the evidencefrom ogam', CAmbridgE' Mrdieval Cdtic Stud. 14, 1-15.

HENDERSON. I. 1987, 'Early Christian monuments ofScotland displaying crosses but no other ornament' inA. Small (00.), The Picts:a New Look at Old Problems. 45-59. Dundee.

HERITY, M. 1995, Studits in the 1Ayvut, Buildings Qnd Artin Stone of Early Irish Monas/mes. London.

HIGGINS, J. 1987, fu E.arly Christian Cross-Slabs, PillarSlone$ and Rdaled Monummts ofCormty Galway, Ireland,2 vols (- BAR Supp!. Ser. 375). Oxford.

JACKSON. K. H. 1946, Review of Corpus lnscriplionumlnsularum Ctlticarum Vol. 1. by R. A. S. Macalister,Speculum 26, 521-23.

JACKSON, K H. 1953, Lmguagt and History in EarlyBritain. A Chronological Survey of the 8rittonic umguagesFirst to Twelfth Century A.D. Edinburgh.

JAMES, E. 1977, TheMmwingian Arc1lJJrology of South-Wt'StCalj/, 2 vols (- BAR. Suppl. Set. 25). Oxford.

KENNEY, ]. F. 1929, The Sources for the Early History ofIreland, Vol. 1. Ecclesiastical. New York, reprint 1979,Dublin.

KERMODE, P. M. C. 1994, Manx CrosstS (reprint of 1907vol). B.l!gavies.

LEWIS, J. M. and REDKNAP, M. in prep .. Corpus of EarlyMedieval Inscribtd Stones and Stont Sculpturt in Walts,fascicule 1, South and Soljth-East Walrs.

L10NARD, P. 1960-1, 'Early Irish grave-slabs', Proc. Ray.Ir. Acad. 61C, 95-170.

MACALISTER, R. A. S. 1945, CorpljS lnscriptionum In-sularum Celticarum VoU. Dublin.

MACAUSTER. R. A. S. 1949, rep. 1996, TheArchaeology ofIreland, 2nd ed. Dublin.

MACNEILL. E. 1906, 'The ancient Irish genealogies', TheNew Ireland Review 26, 129-45.

MACNEILL, E. 1931, 'Beginnings of Latin culture inIreland part r, Studies 20, 39-48.

MACWHITE, E. 1960-1, 'Contributions to a study of ogammemorial stones', Uitschri.jt fir Celtische Philmogk 28,294-305.

MCMANUS, D. 1983, 'A chronology of the Latin loan-words in early Irish', t.riu 34, 21-n.

MCMANUS, D. 1991, A Guide to Ogam. Maynooth.MARSTRANDER, C. 1911, 'Ogham XOI', t.riu 5, 144.MONK. M. A. andSHEEHAN,J. (eds.) 1998, urlyMedievalMunster; Archarology, History and Society. Cork.

MOORE, F. 1998, 'Munster ogham stones: siting. contextand function', in Monk and Sheehan.. 23-32.

MYTUM, H. 1992, TIle Origins of ftlrly Christian Ireland.London, New York.

NASH-WILLIAMS, V. E. 1950, 111e Early Christian Monu-mm/s of Wales. Cardiff.

6 CR6INfN, D. 1995, Early Mediroal Ireland 400-1200.London, New York.

OKASHA, E. 1993, Corpus of Early Chrish'an Inscribed Stonesof South-West Britain. leicester.

O'KELLY, M.J. 1945, Review of R. A. S. Macalister's CorpusInscriptionum Insularljm Cdticarum Vol. 1, J. CorkArchRroI. Hist. Soc. SO, 152-53.

O'KELLY, M. J. and KAVENAGH, S. 1954, 'An ogaminscribed cross-slab from, County Kerry', I. CorkArcharol. Hist. Soc. 59, 101-10.

O'RAHILLY, T. 1957, The Two Patricks: a Lectljre on theHistory ofOlristianity in Fifth-Century Ireland. Dublin.

6 RIA IN, P. 1981, 'The Irish element in Welsh hagin-graphical tradition', in D. 6 Corrain (ed.), Irish An-tiquity: Studies Presented to Professor M. f. O'Kelly, 291-303. Cork.

O'SULLIVAN, A. and SHEEHAN, J. 1996, Tht IveraghPeninsljla: an Archaeological Survey of South Kerry.Dublin

POKORNY, J. 1915, 'Ogom CI "hierN " Zeitscllrift furCLltische Philowgie 10, 403.

RCAHtwlS 1982, Royal Commission on the Ancient andHistorical Monuments of Scotland, Argyll, An InUffltoryof the Monuments, Vol. 4, lona. Edinburgh.

SHEEHAN,]. 1990, 'Some early historic cross-forms andrelated motifs from the Iveragh peninsula', f. KerryArchaeol. His!. Soc. 23, 157-74.

SHEEHAN, J. 1994, 'A Merovingian background for theArdmoneel stoner I. Cork Arclmeol. Hist. Soc. 99, 22-31.

STEVENSON, J. 1989, 'The beginnings of literacy inIreland', Prrx. Ray. Ir. Acad. 89C, 127-65.

SWIFT, C. 1996, 'Christian communities in fifth- and sixth-century Ireland', Trowel 7, 11-19.

SWIFT, C. 1997, Ogam Stones and the EArliest Irish Christians.Maynooth.

THOMAS, A. C. 1987, 'The earliest Christian art in Irelandand Britain', in M. Ryan, (ed.), Ireland and Insular ArtAD 500-1200, 7-11. Dublin.

THOMAS, A. C. 1994, And Shall These Mute St()l1ts Speak?Post-Roman Inscriptions in Western Britain. Cardiff.

maMAS, A. c. 1998, 'Early medieval Munster; thoughtson its primary Christian phase', in Monk and Sheehan,9-16.

VENDRYES, J. 1955, 'Sur un emploi du mot AINM 'nom'en Irlandais', t.tudt'S ultiques 7, 139-46.

WARNER, R. 1991, 'The Drumconwell ogham and itsimplications', Emania 8, 43-50.