Meta-linguistics

It's a funny old thing, researching language, because you've got to use your object of study to describe your object of study. You have to talk about language using language. Mostly, this is OK, because we can distinguish between metalinguistic mentions of language and actual use of language.

One thing that often happens (and this one isn't linguistics-specific) is that you find yourself using the non-technical version of a word more when you're talking about the technical term (or maybe you just notice it more). I'm researching questions, and I catch myself using question all the time: 'the question is how this can be applied to X' and so on.

We linguists have a fun extra game to play, however. We can use the very linguistic things that we are talking about in the language we use to talk about them. Sometimes this happens by accident, similarly to the above example. But sometimes, you see the opportunity to slip one in as a little in-joke for your readers who are paying attention. I read this sentence today:

The idea... is supported by the fact that only in embedded finite clauses is it possible to front an XP.*

This is a classic example of its type. It's talking about fronting XPs (moving phrases to the start of the clause) in embedded (subordinate) clauses, and in doing so, does just that itself. that only in embedded finite clauses is an embedded clause - it's the complement of fact (it tells you what the fact is). And within it, we have a fronted phrase, only in embedded clauses - it would normally be at the end:

It is posible to front an XP [only in embedded clauses].

Linguist humour. There are whole blog posts to be written about humorous example sentences, comedy names for new generalisations and the like.

*Reference: Breul, C. 2004. Focus structure in Generative Grammar: An integrated syntactic, semantic and intonational approach. Amsterdam: John Benjamins Publishing Company.