Benzoate said:
This method seems circular.
no, it's no circular. He was explaining to you that you must use the definition of the (left) inverse. I.e., for any element g^{} of the group, the inverse g^{-1} satisfies
<br />
g^{-1}\cdot g = e<br />
where the symbol e^{} stands of the identity element of the group.
Maybe the notation g^{-1} for the inverse of g is too familiar for its own good. You could try a different notation. For example, you could denote the inverse of g as
\bar g. Then, the definition of the inverse says that
<br />
\bar g \cdot g = e<br />
To prove the equality you mentioned, first simply rename the element g to \bar a (which is an element of the group if a is, by definition) to see that
<br />
\bar \bar a \cdot \bar a = e<br />
(which is just using the definition of inverse for \bar a).
Next, rename the element g to a to see that
<br />
\bar a \cdot a = e<br />
(which, is just using the definition of inverse for a).
If you multiply the above equation on the left by \bar \bar a and the equation two-above on the right by a you will see that the LH sides are equal and thus the RH sides are also equal which gives you the equality you desire.