A control is how the participants or subjects react naturally to the situation.
Take the age old psychology example: depression. You have three groups: two treatment groups (drug and placebo) and the control group.
The control group ALWAYS receives NO treatment. It is the natural state of the participants.
It doesn't always have to be a group though. If you have are questioning something to test say honesty on some subject, you would ask the pertinent questions and ask a completely unrelated question to test their honesty. Stupid example, I know, but it was all I could think of. Lol. Good luck!