Friday, February 11, 2005

the way we speak

I've always been very sensitive about the way people speak. I often identify voices not by their tone or pitch, but the way they pronounce words and their inflection. If only I could consciously reproduce these subtleties, I would be great at doing impersonations.

Anyways, over the last few months, I've noticed that I speak like a theoretical computer scientist, or, at the very least, an MIT theory of CS person. In everyday conversation I talk normally, but when I start explaining something, or stating an opinion, I start introducing pauses, emphasis, and sentence inflections in a way that people in my lab do. I've always been slightly conscious of this, but as I sit here in the Theory of Cryptography Conference, I'm acutely aware of how distinct a way the people here have of explaining things. There are some variations of this "Theory accent" and I'm quite sure that the MIT version of the accent has been heavily influenced by the Italian accent of Silvio Micali.

This, I think, is due to the facts that Silvio a) has the strongest accent in the department and b) is easily the most interesting person to listen to for any period of time. Regardless of the cause, however, it's rather surreeal to be talking to someone about something utterly non-geek-related and realize I'm speaking with a mild Italian inflection (not pronunciation, mind you, inflection).

For example: In normal English, when you pose a rhetorical question, you do it much in the same way you would pose a normal question: your voice rolls up and down as you ask the question, with natural peaks at both the key word in the question (i.e. who/what/etc.) and the sentence end. With an MIT TOC accent, however, your voice naturally ramps up during the entire question, raising in urgency until you reach the end of your rhetoric question. Then you have a pause, where it is clear that no one should interject an answer, then you give your answer with strong emphasis.

I guess it's natural for people who spend a lot of time talking each other to speak in a similar way (e.g. siblings), but it's just so strange to think of that happening to you in someplace like the crypto group at MIT....


  1. it's been pointed out to me that the "crypto accent" is closely related to the hebrew accent, due to the very large number of Israelis in the field.

    also, a lot of them use comic sans in their presentations. now that's a real head scratcher...

  2. haha, you talk like a Jew.

    Not that there's anything wrong with that.


Note: Only a member of this blog may post a comment.