Pay — a proxy for peer recognition
Innumerable sources talk about how peer recognition matters for more people than money… though the latter itself has some importance too. 
I imagine that at most companies, pay is assigned at approximately market levels based on the responsibilities incurred. As such, CEO pay tends to drift higher than junior engineer pay, because the market has been assigning a higher value to the former — despite the many examples where this has proven incorrect. 
For their part, markets are mechanisms for establishing the price of a good through transactions, based on perceived value of the participants. If not enough buyers feel a banana is worth $3, they will tend not to buy until the price drifts to levels they’re willing to pay; if not enough sellers feel a banana is being properly valued at 3 cents, they will tend divert their productive elsewhere, until the price rises.
If we call the grouping of people in a market “peers” we could easily say market prices are a peer phenomenon — a peer assessment of value of the service / good in question, such as a year’s worth of junior engineering time. Market prices are a form of peer recognition-of-value, or to be more precise, peer perception-of-value. Presumably , if Ahab makes more money than Baal, that means he’s perceived to be more valuable.
In which case pay serves as a proxy for peer recognition (how you’re recognized among your peers, and/or by your peers, depending how hierarchical one’s firm is). Which would go a long way towards explaining why it can be such a sensitive subject…