748
AN GAODHAL.
LECTURE IV., DECLENSION.
Modern Systems of Declension.
Having cleared the ground by the definition and
explanation of terms, and having got our tools into
working order by the enunciation of principles, I
will go to the proper subject of this lecture, name¬
ly, Irish declension. Declension, you will remem¬
ber, is a handy term for expressing the modes by
which relations of case and number are marked.
each mode being a decension. Thus, if there are
five modes, there are five declensions, if there is
but one mode, there is but one decension. How
many such modes are there in Irish? Those
that are curious to see the opinions of the older
grammarians, may consult the College Irish Gram¬
mar §47. For us it is sufficient to know that
O'Donovan, followed by Bourke and Joyce, recog¬
nize five. They say that in Irish there are five
modes of forming cases and number — five rows
of case — endings. The genesis of this idea is eas¬
ily seen when we compare these five declensions
with the five given in Latin grammars, but the
genesis of the idea should not stand in its way if
it is, as Canon Burke says, the most philosophi¬
cally correct. This is the question which is be¬
fore us now — does this theory that there are in
Irish five modes of forming cases, fully account
for the facts of the language? Taking only the
broad outlines of the subject, one would be inclin¬
ed to answer that if the theory be correct, then the
language is most barbarous and unsettled. The
reason for this answer may be seen in nearly every
page of Bourke's and Joyce's treatises on the de¬
clensions, e. g., the College Irish Grammar §67
gives baile, a noun of IV dec, with a
plural bailte, manifestly of Il., and
with another form bailteacha, if any¬
thing, of III. A paragraph above
dorus, doruis, dorusa, and doirse
skip gaily through I., II. and III. dec.
while deil, a page over, jumps from II.
clean into V. deileanna. To say the
least, these changes are rather embarrassing, and
if the declension system be all right then the
language is full, as no other language is full, of
irregularities. But if the language be not so bad
after all, then there must be something wrong with
the theory, and taking the probabilities, the theory
is far more likely to be wrong than the language.
But a graver question than the mere theory of
this or that writer is at issue. Now more than
ever before, the principle of analogy is at work.
The tendency of literature is to reduce the num¬
ber of what are called exceptions, and to make the
great classes of words form their inflections
after certain models. In the classic Irish, that
we hope, is to be, what models are we to take? It
might be handy, say for a poet, to
have bailte and bailidh and bailteacha,
but, if we are to judge from what has happened in
all other languages, some of them will have to
go. Which of them is the question, and a very
important question too. The following a false
analogy may lead us far from the fountains unde¬
filed of Gaelic, the following, the true analogy
will not only conduct us to the pure sources of our
speech, but will bring us thereto with surer step and
by easier stages. The true solution can only be
discovered when the true theory of declension is
known, and it by these words of mine I shall have
moved our Irish scholars to investigate the ques¬
tion, I feel confident that I shall have done no
small service to the cause of the old tongue.
From even a general view of the declensions
we see there is some reason to doubt the "philo¬
sophical correctness" of the present system. We
shall now see what light a detailed examination of
each declension will throw on our subject.
1. Take up the fifth. Canon Bourke says that
"this declension, like the fourth, comprises nouns
that end in a vowel and is distinguished by a pe¬
culiar inflection n or nn in the genitive
singular." Dr. Joyce adds occasion¬
ally d or t." The example given
pearsa, a person, gen. pearsan, dat.
pearsain. Now this statement, when
explained by the definitions given in my last lecture
means that n is a termination added on to a base
or stem pearsa, to express the relation origin. The
word however, pearsa, at very first
sight is suspicious. There is hardly a fact better at¬
tested than that we have borrowed many terms from
Latin. * That pearsa is one of those
I think there can be no doubt, when we consider
the etymology of its representative, Persona. It
is derived from the words, per, through, and sono
to sound, and was applied to the masks worn by act¬
ors. From this it came to mean the part played or
person represented by the actor and thus, at last,
signified what we understand in English by person.
Leaving out of consideration, what I mentioned in
my second lecture, about the absence of the p
from Old Irish, and that tre is the
Gaelic representative of Latin per, it is very un¬
likely that another language and another civiliza¬
tion would have evolved the same idea from the
same constituents; and it is just as unlikely that
such similar sounds could have come to mean the
same thing from different roots. Add to this that
pearsa agrees in gender with its Latin
representative and that Windisch §39 unhesitat¬
ingly put it down as a loan-word, and
I think you will come to the conclu¬
sion that pearsa is a broken form for
persona, and thus that the stem is not
pearsa but pearsan. Let us now de¬
cline this stem according to the second declension
and we get the following highly suggestive result.
Sing. N. Pearsan
G. Pearsaine
D. Pearsain
Pl. N. Pearsana
G. Pearsan
D. Pearsanaibh
Only the genitive singular differs from the example
in the grammar, and the reason of that difference is
very likely that when the n had been worn away in
the nominative, and restored in the genitive in the
same ways the n of the English indefinite article
a is restored before a vowel, such restoration was
considered sign enough of the case relation, and the
e accordingly dropped off with its attenuation
* Max Mueller Lectures on the science of Lan¬
guage, V.
