side things. (The word "idea" includes sense-perceptions, as used by Descartes.) Ideas seem to be
of three sorts: (1) those that are innate, (2) those that are foreign and come from without, (3) those
that are invented by me. The second kind of ideas, we naturally suppose, are like outside objects.
We suppose this, partly because nature teaches us to think so, partly because such ideas come
independently of the will (i.e., through sensation), and it therefore seems reasonable to suppose
that a foreign thing imprints its likeness on me. But are these good reasons? When I speak of
being "taught by nature" in this connection, I only mean that I have a certain inclination to believe
it, not that I see it by a natural light. What is seen by a natural light cannot be denied, but a mere
inclination may be towards what is false. And as for ideas of sense being involuntary, that is no
argument, for dreams are involuntary although they come from within. The reasons for supposing
that ideas of sense come from without are therefore inconclusive.
Moreover there are sometimes two different ideas of the same external object, e.g., the sun as it
appears to the senses and the sun in which the astronomers believe. These cannot both be like the
sun, and reason shows that the one which comes directly from experience must be the less like it
of the two.
But these considerations have not disposed of the sceptical arguments which threw doubt on the
existence of the external world. This can only be done by first proving the existence of God.
Descartes's proofs of the existence of God are not very original; in the main they come from
scholastic philosophy. They were better stated by Leibniz, and I will omit consideration of them
until we come to him.
When God's existence has been proved, the rest proceeds easily. Since God is good, He will not
act like the deceitful demon whom Descartes has imagined as a ground for doubt. Now God has
given me such a strong inclination to believe in bodies that He would be deceitful if there were
none; therefore bodies exist. He must, moreover, have given me the faculty of correcting errors. I
use this faculty when I employ the principle that what is clear and distinct is true. This enables me
to know mathematics, and physics also, if I remember that I must know the truth about bodies by
the mind alone, not by mind and body jointly.
The constructive part of Descartes's theory of knowledge is much