[19] The only thing in the grammar which it might be reasonable to
learn by rote is the numerals.
[20] The story goes that a Swedish dialectologist who was on a tour to
investigate how extensively the strong form _dog_ (died) was in use,
asked a peasant: do you people here say “jag dog” or “jag döde”? The
peasant was not a grammarian; he answered sensibly: well, when we are
dead we generally do not say anything.
[21] Kr. Nyrop informs me that he has found “Mais je mourus hier” in
Mairet, La Silvanire, v. 2, 175, and I myself have come across it in a
short story by Zola about the sensations felt by a person who has been
buried alive after his apparent death--but that does not make the form
more “living.”
[22] The dots which are given in the printed book between the two words
disappear in oral recitation; so they play no part in the minds of the
pupils.
[23] The former “redundant” words are now the most important ones,
indeed in reality the only important ones, since _Pas du tout_ etc.,
where there is no verb, is fully recognized, and sentences like _Je
veux pas_ are becoming more and more common in colloquial language.
[24] Cf. my remarks on “schaffende und erhaltende analogiebildung,”
in Techmer’s _Internat. Zeitschr. f. allgem. Sprachwissensch._, iii.
(1887), p. 191 ff.
[25] _Slawo-deutsches und Slawo-italienisches_, Graz, 1885.
[26] i.e. Translations from the mother-tongue, beginning with single
sentences of the usual kind.
[27] _Aus meinem leben_, II. vi. Goethes werke, Cotta’sche bibl. d.
weltlitteratur, 20. 218.