I
136 CEANIA BEITANNICA.
m
P i
[CHAP. V.
places belonging to Central and Belgio Gaul and Britain, which occur in the Latin and Greek
writers*. As we proceed, it will be found probable that the difference between the language of
Belgic and Celtic Gaul was more than one of mere dialect. Of the languages of North Britain
and of Ireland there is no notice iu the classical writers.
In the British islands and in Prance there are at present five languages or dialects, admitted
to be closely related to each other and to the ancient Celtic tongues. A sixth, the Cornish, ceased
to be spoken in the last century. Of these six, three are peculiar to South Britain and Prance,
viz. the Welsh, Cornish, and Armorican ; and three to Ireland and North Britain, viz. the Irish,
Scottish-Gaelic, and Manx. The former group is by Zeuss called collectively the British,
the latter the Irish language : by English writers, they are more commonly designated the
Cymric and Gaelic. Each comprises three cognate dialects, which, when written or spoken
distinctly, are more or less mutually intelligible ; but with the British and Irish respectively
this is not the case : they are indeed two distinct languages, and not mere dialects of the same
tongue. Prichard calls them sister languages, which " perhaps resemble each other as nearly
as the English and Germant." Others have compared the difference between them to that
between the ancient Norse or Icelandic and the other German dialects Î, or to that between
Greek and Latin §. The close relationship of the two was proved, as regards the vocabulary, by
the researches of Edward Lhuyd || ; and more clearly by the Celtic scholars of the present
century, Garnett, O'Donovan, Zeuss, and others. It appears that " fully two-thirds of the vocabulary
of the Irish are identical with Welsh^; " which language it resembles in many points of
grammatical structure, and in that remarkable system of initial mutations of consonants which
distinguishes the Celtic from all other languages of Europe **. Dr. O'Donovan shows that, in
consequence of the different mode of writing their language adopted by the "Welsh and Irish
respectively, the identity of their inflections is not apparent to one acquainted only with the
written languages. The Welsh write phonetically, exactly as the words are pronounced ; whilst
the Irish preserve the radical letters, which suffer, when inflected, what they call aspiration ; the
difference is in the system of orthography, considered orally the languages are in this respect
the samett. On the other hand, the Irish differs from the Welsh and Armorican dialects
of the British, in having a distinct genitive and dative case, the latter in the plural possessing an
evident analogy with the Sanscrit and Latin ; to which it also approximates in many affixes and
other formatives unknown in Welsh. This last, as well as the Armorican, entn-ely wants the
cases of nouns Î Î . In words radically the same, the Irish systematically substitute gutturals
for the labial and sibilant forms of the Welsh.
As to the dialects of the Gaelic, the Irish is the parent tongue, which, even in the four provinces
of Ireland, presents marked dialectal differences. The Scottish-Gaelic is Irish stripped of
* A comparison of this kind will be found in Prichard,
Phys. Hist. Mankind, 3rd ed. 1841, vol. iii. p. 113-137.
f Phys. Hist. Mankind, vol. iii. p. 52. See also Niebuhr,
" Lectures on Ancient Ethnography," Eng. ed. vol. ii. p. 305.
X Gram. Celt. pr. p. v. Zeuss adds that the difference is
scarcely so great as that between the Lithuanian or ancient
Prussian and the Slavonian.
§ Latham, "Ethnology of the British Islands," 1852, p. 85.
Norris, "Cornish Drama," 1859, vol. ii. p. 457.
11 Archseologia Britannica, 1707.
f Rev. K. Williams, " Origin of the Welsh." Arch. Cambrensis,
vol. vi. Third Series, 1860, p. 203.
** Garnett, "Philological Essays" (ed. 1859), p. 202.
•ft Grammar of the Irish Language, 1845, p. Ixxx. O'Donovan
illustrates this difference by the word fieaw, a woman
(Irish), and bemjn, a woman (Welsh), which with the pronoun
My before it becomes do b-ean (Irish), dy vmyn (Welsh) ; the
changes in pronunciation being almost the same. These
changes, called aspiration by the Irish, are termed infection
by Zeuss.
i t Prichard, "Eastern Origin of Celtic Nations," 1831,2nd
ed., by Latham, 1859, p. 343. Garnett, loe. cit. pp. 202, 204.
The Cornish, as pointed out by Lhuyd, has a genitive case like
the Irish.
CHAP. Y. ] mSTOEICAL ETHNOLOGY OP BBITAIN. 137
a few inflections ; the two difi'ering little more than English and Lowland Scotch. The Manx is
a rude GaeUc disguised by a corrupt phonetic orthography, and in other respects exhibits an approximation
to the Welsh It has no literary remains of importance. The difference between
the three dialects of the Gaelic is about as much as that between the modern Norse, Swedish
and Danish t- The dialects of the Cymric do not resemble each other so closely as those of the
Gaelic; but with attention, a Welshman may understand a continental Breton, speaking on any
familiar subject. The difl-erence between the Welsh and the two other dialects is compared to
that between the Prench and Spanish; that between Cornish and Armorican, which stand m a
closer relation to each other, to that between Spanish and Portuguese, or Dutch and German,.
These three dialects of the British are reasonably regarded as the modern representatives of
the ancient dialects of the same and adjacent districts of Britain and Gaul. The Welsh can
hardly be other than the direct descendant of the dialect spoken in Eoman times in west and
central Britain; the Cornish, as thought even by Giraldus Cambrensis§, that of the south of
Britain at the same period; whilst the Ai-morican may represent one of the dialects of
Ancient Gaul |1. Dr. Arnold observed that the Celtic languages of these islands are in all likeUhood
the solitary survivors of many others1[; and it is highly probable that there were other
dialects of the Cymric which have left no modern representatives.
It has commonly been thought that the ancient language of Gaul and South Britain, in the
time of the Eomans, bore a much closer resemblance to the present dialects of the British than
to the Irish ** ; and this view has been supposed to be finally established by Zeuss t t - It would
however appear, from more recent researches, that the question cannot be considered as decided.
Grimm indeed regards the opposite conclusion, viz. that the ancient Gaulish was more allied to
the Gaelic than to the British, as very possible J J. M. Pictet is of opinion that though the
arguments adduced by Zeuss are important, they do not decide the entii-e problem, and only
estabUsh a large proportion of the Cymric element in the Gaulish. Prom an exposition of
certain lapidary inscriptions from the central and southern parts of Gaul, it is probable that the
language of this district at least was a dialect of GaeUc. No inscriptions in ancient Gaulish
have liitherto been found within the limits of Belgic Gaul; but as the same judicious writer
observes, should such be discovered and prove to have a Cymric stamp, we should be forced to
* Garnett, loe. dt. p. 204. Norris, he. cit. p. 457. O'Donovan,
loc. cit. p. Lxxviii.
t J. F. Campbell, " Popular Tales of the West Highlands,"
1860, Introduction, p. ex. Erse (Irish) is but another name
for Scot t ish-Gael ic—A l b a n a i c h .
t Latham, "Descriptive Ethnology," 1859, vol. ii. p. 49.
Norris, foe. cii.p. 45 7. Mr. Williams maintains that the Cornish
and Breton are more closely related to the Welsh than they are
to each other. Arch. Cambrensis, 3rd ser. vol. v. pp. 148, 297.
§ See this remarkable passage, for a writer of the 11th century,
in Descript. Cambr. c. 6; quoted by Thierry, vol. i. p. 80.
II Garnett, loc. cit. p. 151. Norris, loc. cit. p. 457. Sir
Gardner Wilkinson, " Journal of the British Archseological
Association," 1860, vol. xvi. p. 125. The Armorican, or dialect
of Brittany, is by some supposed to have been imported
from Britain in the fourth or fifth century ; but though it may
have beeu modified by immigrations of insular Britons at
that period, it is doubtless a proper descendant of the old
Gaulish, and probably of the Belgic dialect.
f History of Rome, 1838, vol. i. p. 524.
** Prichard, Phys. Hist. Mankind, Tol. iii. p. 135.
t t Gram. Celt. pr. p. v. Of the principal arguments of
Zeuss, the following is an abstract:—1. The resemblance of
certain sounds in the ancient Gaulish and the Cymric ; that of
P for e.tample, for which the Irish substitute C (K) ; cenn,
cether, ech, &c.; instead oSpenn, petuar, ep. 2. The community
of certain terminations ; en, on, the plural ct •, unknown
to the Irish. 3. The existence in the Cymric and in the ancient
Gauhsh of certain elements which are absent in the Irish.
Among these are the prefixes -gwr, -gur, -gor (as in Vercingetorix,
&c.), and the vocable cun (as in Cunobelinus, &c.).
J t Marcelliu. Formeln. von J . Grimm und A. Pictet, 1855,
p. 52. This work is devoted to an examination of the medical
formulas of Marcellus of Bordeaux, written in the fourth century,
and which appear to be susceptible of explanation from
the Irish. See Ulster Journal of Archaiology, vol. iv. 1856,
p. 264.