," or "was borne
in on me," "I was forced to do so and so," and so on, and in this
manner we tend to assimilate internal to external mental phenomena.

Much the same thing shows itself in our customary modes of describing
our internal feelings of pleasure and pain. When a man in a state of
mental depression speaks of having "a load" on his mind it is evident
that he is interpreting a mental by help of an analogy to a bodily
feeling. Similarly, when we talk of the mind being torn by doubt or worn
by anxiety. It would seem as though we tended mechanically to translate
mental pleasures and pains into the language of bodily sensations.

The explanation of this deeply rooted tendency to a slightly illusory
view of our mental states is, I think, an easy one. For one thing, it
follows from the relation of the mental image to the sense-impression
that we should tend to assimilate the former to the latter as to its
nature and origin. This would account for the common habit of regarding
thoughts, which are of course accompanied by representatives of their
verbal symbols, as internal voices, a habit which is probably especially
characteristic of the child and the uncivilized man, as we have found it
to be characteristic of the insane.

Another reason, however, must be sought for the habit of assimilating
internal feelings to external sensations. If language has been evolved
as an incident of social life, at once one of its effects and its
causes, it would seem to follow that it must have first shaped Itself to
the needs of expressing these common objective experiences which we
receive by way of our senses. Our habitual modes of thought, limited as
they are by language, retain traces of this origin. We cannot conceive
any mental process except by some vague analogy to a physical process.
In other words, we can even now only think with perfect clearness when
we are concerned with some object of common cognition. Thus, the sphere
of external sensation and of physical agencies furnishes us with the one
type of thinkable thing or object of thought, and we habitually view
subjective mental states as analogues of these.

Still, it may be said that these slight nascent errors are hardly worth
naming, and the question would still appear to recur whether there are
other fully developed errors deserving to rank along with illusions of
sense. Do we, it may be asked, ever actually mistake the quality,
degree, or structure of our internal fe