3/29/2010

Do any languages mark the use/mention distinction in speech?

In written English, we mark the distinction between using an expression and mentioning it using quotation marks: 'Chicago' has 7 letters, but Chicago does not. In other words, we can use quotation marks to disambiguate between use and mention.

For various reasons, I am interested in ambiguity tests: ways to diagnose whether a particular expression really is ambiguous or not. One common, widely-accepted (I think) test appeals to different languages: if the e.g. English expression you're interested in is in fact ambiguous, then it should be translated by two unrelated words in at least some other languages. For example, 'bank' is ambiguous in English, and it is translated by 'Bank' and 'Ufer' in German.

After reading a batch of student papers about the definition of 'health,' I was struck by their complete lack of inclination to distinguish use and mention. So then I thought about the ambiguity test -- does any spoken language disambiguate use and mention morphologically (without saying 'the word...' or 'the sentence...')? I posted the question at Ask A Linguist, and the three people who responded said they did not know of any. Does anybody out there know of one? (And it would be strange if it were just one or two languages...)

If there aren't (m)any such languages, then it looks like this could be a counterexample to that particular ambiguity test.

7 comments:

Bryan said...

When the word is a verb, don't English speakers sometimes make the distinction by demanding the infinitive verb tense? Compare:

- "Playing is fun" (present participle)
- "Play is from the old-Dutch pleyen, or to rejoice" (infinitive)

P.D. Magnus said...

This doesn't answer your question, but - surely that ambiguity test is only meant to supply a sufficient condition for ambiguity, right? So this wouldn't really be a counterexample.

Greg Frost-Arnold said...

Thanks Bryan. That may be the answer.

P.D. -- I don't have much confidence in what I'm about to say, but I think it's not sufficient condition, actually. E.g. in Spanish, 'conocer' and 'saber' are both translations of English 'to know,' but 'know' isn't ambiguous -- it is just more 'general in sense' than Spanish knowledge concepts: the Spanish words distinguish knowing a fact from being acquainted with something or someone.

Jonathan Livengood said...

Greg,

I'm confused by your response to PD. If a word has two or more different senses, could be applied in different ways corresponding to those senses, and could be understood in different ways corresponding to those senses, isn't that word ambiguous? I would have thought that "know" in English was paradigmatically ambiguous. Or is it that you think it covers a lot of rather vague ground -- passing acquaintance to deep acquaintance to *very* intimate acquaintance (as in the Biblical sense)?

As to whether there are other languages that mark the use-mention distinction with morphological differences, I wonder why the difference has to be morphological? What about the use of air-quotes or finger-quotes, or word emphasis or other tricks that people use to mark the distinction in English itself?

Kenny said...

I would think that what's going on here is that for every English word, there is another homophonic word that is a name for it. The other word is spelled differently (since it uses quotation marks), but it's pronounced the same. Technically speaking, the name for a word is "zero-derived" from the word itself. You also see this in English with various nouns and verbs that are spelled and pronounced the same way. However, in other languages (for some reason, Hungarian is the one that comes to mind), there is a suffix that transforms nouns into verbs. So if my zero-derivation story were really plausible for quotation names, then there should probably be some language out there that pronounces the morpheme transforming a word into its name, and yet no one can come up with an example of such a language.

Anonymous said...

Lojban has quotation marks that are pronounced

Greg Frost-Arnold said...

Cool, thanks. 1. Is it mandatory? 2. Do you know of any natural languages that do this?