Abbreviations

[dr_whom]

So I'm listening to Standing Room Only, and they played "14 Dwight Ave., Natick, Massachusetts", from William Finn's Elegies—a song I'm very fond of.

The title of the song, which is repeated several times in the song itself, is "14 Dwight Ave., Natick, Massachusetts". Afterwards, the DJ announced the title of the song as "14 Dwight Avenue, Natick, Mass."

It's kind of interesting to me that she got wrong which word in the title is abbreviated, since it suggests she wasn't actually listening to the song very much and doesn't know it that well (anecdotally supporting my hypothesis that Standing Room Only has less knowledgeable hosts than it used to).

But it's also interesting that she got right that exactly one word was abbreviated—as if she picked that much up subconsciously, without being able to remember the (seemingly much more obvious) fact of which it was. You see something similar in spelling errors sometimes, where people remember how many double letters a word has but not which letters are doubled, so they're more likely to misspell Jennifer as Jeniffer than Jenniffer.

Comments

[qnmark.livejournal.com]

The word Mitsubishi doesn't work this way. There's an old post on Language Log looking for incidences of each misspelling/mispronunciation: each vowel can be either an < i> or a < u>, for 16 combinations of which 15 are wrong. The most common error in English is if I remember correctly Mitsubushi, followed by Mitsibushi. (I think in Hebrew, Mitsibushi is the more common one - the number of < u> s is preserved, but the location is moved, possibly because mitsi is a word in Hebrew and mitsu isn't.)

[dumble.livejournal.com]

Alternate hypothesis: People know things about the distribution of abbreviations in location names/double letters in words. This is admittedly probably more plausible for the double letters case than for the song title. So, e.g. it's not that people remember that Jennifer in particular has one double letter, but rather they know that words in general (or words of a certain length) are more likely to have a single doubled letter than multiple doubled letters.

This could also conceivably explain the seeming counterexample of Mitsubishi, as people might have learned a more balanced distribution of s and s.

[dr_whom]

Good point! This also probably better accounts for the frequency of misspellings of accommodate.

[dumble.livejournal.com]

Oh hey! http://www.tandfonline.com/doi/abs/10.1080/17470218.2013.846392#.UoiKmmTF3OY

[dr_whom]

Ah ha!

This also gets points for the new-to-me word graphotactic.