Postpositions?

[keenir]

I thought the whole point of an idiom (and why conlangers & other translators are advised not to translate idioms literally) was because they were fixed.

Some are, some aren't. You have to look.

Let's look at some idioms. Here's the first list I could find via Google. Some basic observations:
[*] Not that this is actual frequency information.

understood; thank you for clarifying and explaining.

that explanation also massively cleared up what confusion i had had about "productivity" when referring to idioms.

[Moose-tache]

Good points about how idioms vary in their productivity. The original point was that the unproductive ones aren't useful for establishing syntactic rules, and I'd like to go back to that.

If we take a sentence like "What can I do you for?", it's obviously breaking the rules, deliberately. We can't say based on this that prepositions can freely move to the end of the sentence. However, it's just as clear that this sentence is playfully referencing syntactic rules that do exist. English speakers are familiar with numerous ways that words similar to "for" can be moved to the end of a sentence. It's exactly the fact that it isn't randomly and thoughtlessly violating all the rules that gives the idiom its charm. It's like writing a sonnet but doing five couplets followed by a single ABAB quatrain. It's... not a sonnet, but it's still based on what the reader knows a sonnet should be.

So I think idioms are still valuable, because they still tell us something about the rules (or more accurately principles and generalizations) that speakers are thinking about when formulating sentences, even if they're willing to break some of them.

[zompist]

So I think idioms are still valuable, because they still tell us something about the rules (or more accurately principles and generalizations) that speakers are thinking about when formulating sentences, even if they're willing to break some of them.

Good point. I think idioms can always tell us something. We have to be careful not to over-interpret, but the thing with linguistics is there's always more damn data, so over-interpretation is usually corrected by looking at more data.

[KathTheDragon]

As just one example, "I'm not as fit as a fiddle" or "Are you as fit as a fiddle?" sound odd to me-- they are grammatical, they just arguably lose the idiom.

These are both fine for me as examples of "fit as a fiddle"

[Moose-tache]

As just one example, "I'm not as fit as a fiddle" or "Are you as fit as a fiddle?" sound odd to me-- they are grammatical, they just arguably lose the idiom.

These are both fine for me as examples of "fit as a fiddle"

"How fit do you take me for? As a fiddle?"
"I prefer metaphors to similes. I am a fit fiddle."
"I need to train more if I'm going to get into full-on fiddle fitness."
"There's only so many fiddles I'm fit as, and that ain't one of them."
"She made a scene when we forced her to wear the violin costume. She had a fit as a fiddle."

[CreativityTheEmotion]

As just one example, "I'm not as fit as a fiddle" or "Are you as fit as a fiddle?" sound odd to me-- they are grammatical, they just arguably lose the idiom.

These are both fine for me as examples of "fit as a fiddle"

"How fit do you take me for? As a fiddle?"
"I prefer metaphors to similes. I am a fit fiddle."
"I need to train more if I'm going to get into full-on fiddle fitness."
"There's only so many fiddles I'm fit as, and that ain't one of them."
"She made a scene when we forced her to wear the violin costume. She had a fit as a fiddle."

Most of these are on the line between "not grammatical" and "you're just dealing with a very creative novelist".