I still think it's circular because it's not making a qualitative distinction between different kinds of prices.

The simplest social definition of inflation is that it decreases the choices available to participants in the economy.

This means not all prices are equal. If food becomes 50% more expensive, I have little or no choice but to buy it. (Assuming I'm on a standard and not on a luxury diet.) That means my choices decrease - I have less money to spend on non-essentials, and I have no flexibility in this.

If luxury yachts become 50% more expensive, no one's choices are seriously diminished. Yacht wannabes may feel the inflationary sting of not being able to afford the model they want, but yacht buying remains a discretionary activity. And if a yacht becomes unaffordable, it's always possible to spend an equivalent amount on something equally weighed down with bling.

So even though increased yacht prices may count towards inflation using the GDP definition, the price increase is voluntary, the decision to buy is voluntary, and the profit will most likely be recycled into the economy.

But with food prices, the price increase may not be voluntary (because it may be going on overheads), the decision to buy isn't voluntary (or is at least tightly constrained) and the money effectively disappears. (I suppose technically some of it could be recycled as profit, but that's not how it seems to work out.)

So - some prices affect freedom of choice much more than others. And price inflation in some parts of the economy is much more damaging than in other parts. A proper analysis would pick out price elasticity and capital mobility as elements too. (E.g. capital tied up in house prices is, at least partly, lost to the economy, and therefore house price rises are highly inflationary - even thought they're not counted as such today.)