one
country to another. My eldest boy spoke English in childhood as well as
any other English child of his age. He was taken to the south of France,
and in three months he replaced his English with Provencal, which he
learned from the servants about him. There were two ladies in the house
who spoke English well, and did all in their power, in compliance with
my urgent entreaties, to preserve the boy's native language; but the
substitution took place too rapidly, and was beyond control. He began by
an unwillingness to use English words whenever he could use Provencal
instead, and in a remarkably short time this unwillingness was succeeded
by inability. The native language was as completely taken out of his
brain as a violin is taken out of its case: nothing remained, _nothing_,
not one word, not any echo of an accent. And as a violinist may put a
new instrument into the case from which he has removed the old one, so
the new language occupied the whole space which had been occupied by
English. When I saw the child again, there was no means of communication
between us.

After that, he was removed to the north of France, and the same process
began again. As Provencal had pushed out English, so French began to
push out Provencal. The process was wonderfully rapid. The child heard
people speak French, and he began to speak French like them without any
formal teaching. He spoke the language as he breathed the air. In a few
weeks he did not retain the least remnant of his Provencal; it was gone
after his English into the limbo of the utterly forgotten.

Novelists have occasionally made use of cases similar to this, but they
speak of the forgotten language as being forgotten in the manner that
Scott forgot the manuscript of "Waverley," which he found afterwards in
the drawers of an old writing-desk when he was seeking for
fishing-tackle. They assume (conveniently for the purposes of their art)
that the first language we learn is never really lost, but may be as it
were under certain circumstances _mislaid_, to be found again at some
future period. Now, although something of this kind may be possible when
the first language has been spoken in rather advanced boyhood, I am
convinced that in childhood a considerable number of languages might
succeed each other without leaving any trace whatever. I might have
remarked that in addition to English, Provencal, and French, my boy had
understood Gaelic in his infancy, at least to some e