For bidirectional control of a DC motor you typically need a full bridge made up of four FETs or transistors. Each lead of the motor is connected to a half bridge where the upper transistor can take it to the positive supply or the lower transistor can take it to the negative. By setting either side high and the opposite low, you can make the motor spin in either direction.
For unidirectional control, you really only need a single FET or transistor. You can leave one side of the motor permanently connected to the supply, and use the transistor to connect or disconnect the other.
Half bridges that are not packaged in pairs as full bridges are somewhat associated with AC motors particularly because of the situation of 3-phase motors where you end up needing 3 half bridges. In that case you have three wires coming out of the motor, each of which gets driven by its own half bridge so it can be connected to either side of a DC supply. But rapidly pulse-width-modulating the half bridges, something approximating an AC sine wave can be produced, and if the three bridges do this in an appropriate phase relationship (0, 120, 240 degrees) then you can make the motor spin just as if it were connected to a 3-phase AC dynamo - but with the added benefit that you can vary the frequency of the synthesized AC to control speed.
Finally most bipolar stepper motors have two independent coils and require two full bridges.