If it's so important to measure and talk about real prices, if nominal prices are as irrelevant as they seem, why not cut to the chase and do our economic transactions directly in some "real" quantity? Scientists do so all the time: They measure length in real meters, weight in real kilograms, etc.; the notion of a purely nominal measure with a varying relationship to anything real would be inconvenient and pointless*. The problem in economics, however, is that it's impossible to immediately determine real prices; whereas nominal prices are by definition immediate; they measure the immediate relations of particular commodities and factors to each other. It is only in retrospect, when we can identify how all these ever-changing relations have sorted themselves out, that we can actually identify real prices. (And economists cannot really identify real prices; they can only estimate real prices.) We can know nominal prices exactly, although we cannot know what they "really" mean; real prices by definition mean something real, but we cannot know them at all immediately, and only imperfectly in retrospect.
*I am not a scientist; it might be the case that scientists do use purely nominal measures. If so, I would very much like to know about these measures, and how scientists go about relating them to real quantities.
A house requires a certain amount of lumber, wires, pipes, shingles, etc., all of which require raw materials; everything, from the extraction of the raw materials, to the manufacture and transport of the intermediate components to the final assembly of the house require a certain amount of human labor, fixed by the natural world and current levels of technology. So long as none of these components change substantially — the cost in labor of extracting materials, or manufacturing wires or pipes, etc. — change, the real value of the house will not change, even if prices overall rise or fall, changing the nominal value of the house. Of course, if the fundamental natural or technological conditions change — wood becomes more scarce, requiring more labor to satisfy marginal demand, or efficiency of production lowers the total labor cost of producing wires and switches — then the real value of the house — irrespective of the nominal value — will change.
If all wealth were held in purely real terms — in things such as houses, factories, cans of food, or even ingots of gold — then there would be no nominal prices, and inflation and deflation would be meaningless. The problem comes in when we start to socially construct assets with purely nominal value. I'll talk about the effect of these assets in another post.