Picture a field of grass after the wind has passed: every blade leans a certain way. Drop a leaf on the field and let it slide along the blades, and it traces a path. A slope field is that picture for dxdy=f(x,y). At each point you draw a short dash whose slope is f(x,y), and a solution curve is a path that is tangent to the dashes everywhere it goes.
Drawing one is pure arithmetic. For dxdy=x−y, the dash at (0,0) has slope 0, the dash at (2,0) has slope 2, and the dash at (0,2) has slope −2. Curves of constant slope, called isoclines, speed things up: along the line y=x every dash is flat, and along y=x−1 every dash has slope 1. In fact y=x−1 is itself a solution, because its own slope, 1, matches every dash along it, and other solution curves bend toward it as x grows.
A slope field is not the graph of a function. The dash at (2,0) has slope 2; that does not mean a curve passes through the height 2 there. The value of f is a steepness, not a y-coordinate. Nor is a slope field a formula: it shows the shape of every solution at once, but an exact expression such as y=x−1+Ce−x needs an algebraic method.
Think of it like this. A slope field is a weather map of wind directions: an arrow at every town shows which way the wind blows there, and a balloon released anywhere drifts along a path that follows those arrows. Different launch points give different paths, and for most equations two paths never cross. Unlike wind arrows, slope dashes show direction only, not speed.
Core idea. A slope field draws the slope dxdy=f(x,y) as a short segment at each point, so solutions are curves that follow the arrows.