Because the instruction "Add A, R1" is defined that way OR it's better written as "Add (A), R1"
Which again, is architecture specific. "Add (A), R1" in THIS CASE would mean add R1 to the the address pointed to by A.
Notations like: R1, (R1), -(SP) and (SP)+ are common:
R1 = Contents of R1
(R1) = address pointed to by R1
-(SP) = Use something from the stack and decrement the stack pointer
(SP)+ = Put something onto the stack and increment the stack pointer when done