[20] Cf. Rowland’s Grammar of the Welsh Language, 4th edition, (Wrexham, Hughes), p. 23, § 132, where more instances, and also some from Armorican, are cited.
[21] Raoul de la Patisserie: De la Psychologie du Langage. Paris, 1889, pp. 22, 41.
[22] So again, ‘brung’ can often be heard from children, and in German, ‘gebrungen’ appears as a humorous form, probably in imitation of an original blunder.
[23] Cf. Studies in Classical Philology, No. II., B. I. Wheeler: Analogy, and the Scope of its Application in Language (Ithaca, N.Y., 1887), p. 7. Much of what follows is taken from this little work, which contains an admirable discussion of analogy, besides a highly useful bibliography of the subject. See also Jespersen’s article in the Internationale Zeitschrift für Allgemeine Sprachwissenschaft, Zur Lautgesetzfrage, (1886). Professor Wheeler, however, ranges under ‘Analogy-formation’ much that we should prefer to consider separately under ‘Contamination.’
[24] The personal influence, or ‘magnetism,’ of the speaker or speakers who engender the ‘mistake’ is also an important element in determining its propagation. We, parrot-like, imitate the speech, like the manners, of some more readily than of others.
[25] Cf. C. Goeders, Zur Analogiebildung im Mittel-und Neuenglischen. (Kiel, 1884.) Dr. Goeders has collected an enormous mass of illustrative material. Some of his examples, however, may not prove as new as he thinks. Our posterity will be able to decide this point if Dr. Murray’s Dictionary has made greater progress than at present. This apprehension, however, does not detract from the value of Goeders’ work, nor from the truth of the proposition which he illustrates.
[26] Henry, Étude sur l’Analogie en général et sur les Formations de la Langue Grecque. Paris, Maisonneuve, 1883.
[27] Professor Almkvest kindly informs us that there are rules about the grave accent in the Swedish, but that they are difficult to investigate. The grave accent, as it occurs in Swedish, is quite peculiar, and nothing similar exists in other languages.
For instance, the first syllable in brä́der (pl. of brä́de = board) and sånger (pl. of sång = song) has the accent, but is musically lower than the second syllable, which has a feeble secondary accent, and is musically higher. This is different—in contradiction to breder (pres. of breda = to spread), where the first syllable has the accent, and is musically higher than the second syllable, which is quite without accent.
It is the first-named pronunciation, brä́dè, brä́dèr; góssè (a boy), góssàr, which has nothing corresponding to it in other languages.