The etymology of 'cennetig'.
16th August 2008; revised 17th January 2012
Iain Kennedy
Copyright © 2012 Iain Kennedy
For some time I have been puzzled by two things. Why is there so much confusion as to whether the second part of 'Kennedy'/'cennetig' means 'helmet' or 'ugly/'grim';
and why do these words not seem apparent in my own Gaelic dictionary? (I use Malcolm MacLennan's 'A Pronouncing and Etymological Dictionary of the Gaelic
Language' in its 2005 reproduction form published jointly by Acair Ltd of Stornoway and Mercat Press of Edinburgh, the original having been published in 1925. This
in turn was based on Neil MacAlpine's 1831 'A Pronouncing Gaelic Dictionary'). A third question has formulated in my mind; just how much Gaelic schooling
do the authors of Scottish and Irish surname dictionaries actually have, and to what extent do they rely on Gaelic language dictionaries?
The first part of the word, 'ceann', proves uncontroversial. Any dictionary will tell you it means head, chief etc. If you are in any doubt, take a look at the
railway station sign at Kingussie in the Scottish Highlands... It also appears in multiple locations as 'head of the loch' ie Kinloch, including Kinloch Rannoch.
The RIA dictionary entry for 'cenn' runs to 8 page columns. It even mentions cenn-etig 'freq. as n. pr.' but without discussing its precise meaning.
The second part of the name 'etig' (or countless variations) is more tricky. This is what Black summarises, by way of reference to the available dictionaries
of the time:
'The modern G. form of the name is Ceannaideach, Ir. Cinneididh, from earlier Cinneide or Cinneidigh, Mid. Ir. Cendetig (Book of Leinster), literally 'ugly headed' (Kuno Meyer) or 'grim headed' (Watson). cf the Gaelic name of Loch Etive, Loch Eitigh, 'ugly, horrid loch'. '
Turning to the Irish surname dictionaries and we find Patrick Woulfe, writing in Kilmallock in 1906, giving the meaning as 'descendant of Cinneidig (helmeted-head)'. Woulfe was an Irish speaking native of Co. Limerick and gathered much of his material talking to old men in the Limerick and Kilmallock workhouses. Kilmallock is a mere 25 miles from Nenagh in northern Tipperary where the name has its stronghold. Some decades later Edward MacLysaght was to take a different view, citing the meaning as 'ceann head, eidigh ugly'. MacLysaght, a native of England, appears to have acquired his Irish language skills as a young man when he moved to Co. Clare. He spends some time in his book arguing for a revision of Woulfe, based in part on the availability of the RIA dictionary. Unfortunately, whilst he scorns George Black for accepting Woulfe's work on Irish names in Scotland, he commits a number of howlers on the Scottish Kennedys of whom he clearly has no knowledge whatsoever. I am not going to comment on whose command of modern or middle Irish in general was better. But I do feel that MacLysaght owed his readers an explanation and source for his amendment of Woulfe, and we are thus left to guess whether this was his personal opinion or that of someone (possibly) more learned.
Since George Black did tell us his expert sources, let's take those references first. Kuno Meyer was an eminent Celtic scholar around the turn of the 20th century and held such posts as Professor of Celtic at the Royal Irish Academy. Unfortunately Black doesn't appear to state which of Meyer's many works he is referring to. He did produce part of a Gaelic dictionary but only the first few letters of the alphabet and the National Library copy appears to be missing. William J. Watson held the chair of Celtic at Edinburgh University, the same post later held by Kenneth Jackson; he was a native Gaelic speaker from Easter Ross. The specific reference is to 'Celtic placenames of Scotland' (first published 1926, reprinted 1973 Irish University Press, Shannon).
etig
unnatural, unseemly, uncomely; ugly, repulsive
examples quoted inc.
etich
etig
edidh
n-eitig
n-eitigh
eitig
Note carefully that at least one of these is 'spelt' or transliterated with a 'd' and not a 't'.
eidigthe
cf etigud
clad in armour; equipped
da miledhaib armtha eidigthi YBL 159 41
(Yellow Book of Lecan ria facs.)
forlion na nGall n-edighthe FM iii 276
(Annals of the Four Masters)
coicc ced decc laech armtha eidithe hugh roe 48.31 (f 26)
(O Cleary's life of Hugh Roe O'Donnell)
ba heidighthi siumh tra cenmota an t-ionadh sin trias ro gaottha FM vi 1978 20
cf vc eidedach vc gan eided, Caith. CC 52 where in the first clause eidigthe is probably meant
(caithreim chellachain chaisil)
To also quote from MacBain's
'An Etymological Dictionary of the Gaelic Language' (Alexander MacBain LLD. published in Stirling 1911); 'eitigh' means 'fierce, dismal (O. Ir. 'etig');
in his etymological notes he adds
'scarcely 'anteg' unwonted (Zim) for G. would be 'eidigh'.
The distinction then appears to reduce to the consonant in the middle, with a 'd' giving us the 'armour/helmet' meaning and 't' giving us the 'fierce/grim/ugly' meaning. But is this distinction meaningful? What are the rules on pronouncing these letters in Gaelic? MacLennan tells us this:
d flanked by e or i: as in Eng. j eg 'jerk'
t flanked by e or i: as in Eng. 'chin'
So we can see that you can't actually get a pure 'd' or 't' sound if flanked on either side by an 'e' or 'i', although both letters work as in modern English when flanked by the other vowels. (The vowels are divided into slender (e/i) and broad (a/o/u) and if one precedes a consonant, the same class of
vowel must follow it. This rule has a special name 'caol ri agus leathann ri leathann' - slender to slender and broad to broad).
When we also note that some sources spell or transliterate the name as 'Cennedig' rather than 'Cennetig', we need to bear in mind that fixed spellings of names are a recent phenomenon. For example in Scotland the spelling 'Kennedy' only finally stablised after the introduction of civil birth registrations in 1855. So even if we were to go back to the Old Irish originals and survey all the ways the name was spelt, this would not necessarily give us the definitive answer we are seeking. The CELT text version of the Book of Deer uses the spelling 'Cennedig' (which then appears as 'Cenneitech' in the English translation).
If we now turn to the original spellings, the most reliable source I can select are the early contemporary entries from the Annals of Innisfallen, which is believed to have been copied around 1092, just before Kennedy coalesced into a recognisable surname. Let us recognise in doing so that many of the key annal entries were rendered in Latin prose not Irish, including all the Kennedy entries from the 12th century, although in theory this should not effect proper names. Out of 5 Latin and the 5 later Irish entries, I find 6 spell the name '-edig' (all 5 Irish entries and one Latin) with the remaining 4 Latin entries spelling it '-etig'. Furthermore all 14 of the 'copyup' entries prior to 1092 are spelt '-etich' (13 with the exact spelling cennetich) or '-etig' (just one). Unfortunately we can only speculate whether this was one scribe standardising as he copied, or whether the earlier source(s) followed the same practice. The most likely interpretation of this data I can present is that one scribe copying in c. 1092 standardised the name as cennetich; and that later generations changed preference to '-edigh' which dominated spellings from the late 1100s to the early 1300s. Although I make no claims to be an Irish scholar the conclusion seems inescapable; that cennetig and cennedig meant the same thing to the few literate Irish in the early medieval period, and that both spellings represented the same phonetic name.