(2) 'What are those human shapes which appear in dreams and
visions?'[11]

Here it should be noted that Mr. Tylor most properly takes a distinction
between sleeping 'dreams' and waking 'visions,' or 'clear vision.' The
distinction is made even by the blacks of Australia. Thus one of the
Kurnai announced that his _Yambo_, or soul, could 'go out' during sleep,
and see the distant and the dead. But 'while any one might be able to
communicate with the ghosts, _during sleep_, it was only the wizards who
were able to do so in waking hours.' A wizard, in fact, is a person
susceptible (or feigning to be susceptible) when awake to hallucinatory
perceptions of phantasms of the dead. 'Among the Kulin of Wimmera River a
man became a wizard who, as a boy, had seen his mother's ghost sitting at
her grave.'[12] These facts prove that a race of savages at the bottom of
the scale of culture do take a formal distinction between normal dreams in
sleep and waking hallucinations--a thing apt to be denied.

Thus Mr. Herbert Spencer offers the massive generalisation that savages do
not possess a language enabling a man to say 'I dreamed that I saw,'
instead of 'I saw' ('Principles of Sociology,' p. 150). This could only be
proved by giving examples of such highly deficient languages, which Mr.
Spencer does not do.[13] In many savage speculations there occur ideas as
subtly metaphysical as those of Hegel. Moreover, even the Australian
languages have the verb 'to see,' and the substantive 'sleep.' Nothing,
then, prevents a man from saying 'I saw in sleep' (_insomnium_,
[Greek: enupnion]).

We have shown too, that the Australians take an essential distinction
between waking hallucinations (ghosts seen by a man when awake) and the
common hallucinations of slumber. Anybody can have these; the man who sees