Why square the control variable?

I m reading a paper...in its regression, it include a control (independent) variable of firm AGE for the proxy of firm reputation. However, it further said that square of the firm's age is also controlled because reputation building may continue to increase BUT at a slower rate. Why squaring the age is needed??? PLEASE HELP...THANKS!!!Why square the control variable?
AGE is squared because you're dealing with diminishing returns. Think about what the square root function looks like. It increases at a decreasing rate. Since the coefficients have to be linear in a regression model, you change the variables (in this case AGE) to better fit the data when you run the regression. So you square the variable to get the y-variable (reputation?) to have diminishing returns.