err..not quite. the inverse of a matrix A, denoted by A^(-1), can only be found for regular square matrices (i.e. square matrices whose determinant does not equal 0).
when you have a matrix equation such as AX = B. in order to get X = some junk, we need to be able to move the matrix A over to the right hand side. now, in normal arithmetic, you'll recall that we can simply divide both sides by A. however, this is not a valid move in matrix arithmetic. there is however an equivalent. by pre-multiplying both sides by A^(-1), you are effectively dividing everything by A, since now you get A^(-1)AX = A^(-1)B, and by definition A^(-1)A = I, which is the matrix equivalent of 1, and so X = A^(-1)B.
you are right in saying that the determinant of a matrix can only be found in a square matrix. not sure what you mean by "anything that's not in a square matrix".
don't quite understand your second question...