Spoiler alert

Inspired by MKR's comment, here's my cruder version of an analysis of secondary sense:

1. words are used with their usual, primary meaning
2. but in a novel, perhaps seemingly meaningless, context
3. in such a way that no paraphrase is possible without loss of meaning or sense  

Wittgenstein gives the example of saying that 'Wednesday' is fat and 'Tuesday' thin. The idea is that the word 'Wednesday' really is thought to be fat, even though it cannot be literally fat, and that in saying it is fat one does not simply mean that it is a long word, or involves forming the mouth into a fattish shape, or anything else. One means simply (i.e. solely) that 'Wednesday' is fat.

There is a good chance that this will just strike you as nonsense, but consider the Kinks' singing about Lola's "dark brown voice." Lola's voice is not literally dark brown--it couldn't be. But the words 'dark brown' convey something that would be lost if they had sung instead of a 'masculine in a 1970s kind of way voice' instead. Whoever first used the word 'blue' to mean sad seems to me to have been doing the same kind of thing. If it's true that all reading used to be done aloud, then the first person to read silently was perhaps using the word 'reading' in a secondary way as well (Wittgenstein says that calculating in the head involves a secondary use of the word 'calculating'). Once you get the idea, I think, there's a tendency to see secondary sense (or secondary uses of words) everywhere. 

This makes me want to interrogate cases that I am tempted to think of as secondary. It also makes me wonder how we can distinguish between nonsense and secondary sense. Is secondary sense just nonsense that we like? When Wittgenstein talks about excluding certain forms of words from our language by calling them nonsense, would he equally say that one can welcome a form or use of words into the language by calling it secondary? And where else might we find nonsense/secondary sense?

Consider the spoiler. According to Wikipedia: 

 A spoiler is an automotive aerodynamic device whose intended design function is to 'spoil' unfavorable air movement across a body of a vehicle in motion. Spoilers on the front of a vehicle are often called air dams, because in addition to directing air flow they also reduce the amount of air flowing underneath the vehicle which reduces aerodynamic lift. Spoilers are often fitted to race and high-performance sports cars, although they have become common on passenger vehicles as well. Some spoilers are added to cars primarily for styling purposes and have either little aerodynamic benefit or even make the aerodynamics worse.    

Spoilers on passenger vehicles are meant to make them look more sporty by adding something that race cars have. So they are meant to be the very thing that race cars have. Yet they are used in a context in which they have little to none of the intended effect, and can even have the opposite effect. They make the aerodynamics worse. This is all only like a use of language, and I'm not sure what the automotive styling equivalent of a metaphor would be, but it seems to me possible to regard the use of spoilers on passenger vehicles as a kind of nonsense. And I can just about imagine a lover of such things insisting that it was instead an inspired secondary use. Perhaps sense is in the eye of the beholder, so to speak.

One more example: the stars on the crest of Manchester City Football Club:   

The crest was introduced in 1997 and has always been controversial (i.e. mocked) because of the stars. Usually stars like these indicate major trophies won. For instance, England have a single star on their shirts to commemorate the time they won the World Cup. Manchester City's stars, on the other hand, "constitute a design element that relate a more continental feel to the design," according to the club's official statement when the crest was unveiled. Wikipedia sheds light on this continental feel and its likely origin: 

The first team to adopt a star was Juventus, who added one above their crest in 1958 to represent their tenth Serie A title. This was an extension of the existing convention by which the reigning champions are entitled to display the scudetto on their shirts for the following season. The star was later formally adopted by some organisations as a symbol for ten titles.

Brazil added three stars above their crest after winning their third World Cup in 1970. Italy did likewise in 1982. All world champions have since followed suit.

It seems likely that the club liked the "winny" look of the stars. So (if this is right) they wanted the significance of the stars, or that the stars convey, without claiming to have literally won anything of that significance. Again, this looks like a case of 'aesthetic nonsense'. But I can imagine a City fan insisting that the stars are apt (perhaps because the team "are winners to me" or some such thing).

I expect that this kind of analysis of aesthetics could be (and quite possibly is) done quite widely. It wouldn't prove that people who like spoilers or City's crest-designers have bad taste. But it would allow for a case to be made that went beyond "I just don't like it."