In order to maintain this critical attitude to select and reject suggestions with reference to a goal, the suggestions as they come cannot be accepted as units and followed.  Such a procedure is possible only when the mental process is not controlled by an end.  Control by a goal necessitates analysis of the suggestions and abstraction of what in them is essential for the particular problem in hand.  It is because no complete association at hand offers a satisfactory response to the situation that the need for thinking arises.  Each association as it comes must be broken up, certain parts or elements emerge, certain relationships, implications, or functions are made conscious.  Each of these is examined in turn; as they seem to be valueless for the purpose of the thinker, they are rejected.  If one element or relationship seems significant for the problem, it is seized upon, abstracted from its fellows, and becomes the center of the next series of suggestions.  A part, element, quality, or what not, of the situation is accepted as significant of it for the time being.  The part stands for the whole—­this is characteristic of all thinking.  As a very simple illustration, consider the following one reported by Dewey: 

“Projecting nearly horizontally from the upper deck of the ferryboat on which I daily cross the river, is a long white pole, bearing a gilded ball at its tip.  It suggested a flag pole when I first saw it; its color, shape, and gilded ball agreed with this idea, and these reasons seemed to justify me in this belief.  But soon difficulties presented themselves.  The pole was nearly horizontal, an unusual position for a flag pole; in the next place, there was no pulley, ring, or cord by which to attach a flag; finally, there were elsewhere two vertical staffs from which flags were occasionally flown.  It seemed probable that the pole was not there for flag-flying.

“I then tried to imagine all possible purposes of such a pole, and to consider for which of these it was best suited:  (a) Possibly it was an ornament.  But as all the ferryboats and even the tugboats carried like poles, this hypothesis was rejected. (b) Possibly it was the terminal of a wireless telegraph.  But the same considerations made this improbable.  Besides, the more natural place for such a terminal would be the highest part of the boat, on top of the pilot house, (c) Its purpose might be to point out the direction in which the boat is moving.

“In support of this conclusion, I discovered that the pole was lower than the pilot house, so that the steersman could easily see it.  Moreover, the tip was enough higher than the base, so that, from the pilot’s position, it must appear to project far out in front of the boat.  Moreover, the pilot being near the front of the boat, he would need some such guide as to its direction.  Tugboats would also need poles for such a purpose.  This hypothesis was so much more probable than the others that I accepted it.  I formed the conclusion that the pole was set up for the purpose of showing the pilot the direction in which the boat pointed, to enable him to steer correctly."[8]