Are we able to address capitophones?

[jsalsman]

This showed up in my YouTube feed just now and to be completely honest I'm totally flabbergasted:

https://www.youtube.com/shorts/bsdkOJl8taE

I note from a quick look at http://www.speech.cs.cmu.edu/cgi-bin/cmudict?in=polish that we simply don't discriminate on capitalization.

@Alexir @lenzo-ka @dhdaines Do you have any idea how we might be able to handle these?

[jsalsman]

Maybe we need a CMUDICT v1.0.0 where the heading words are both upper and lower case?

[jsalsman]

Here are some others from the video's comments:

• job/Job (i.e. the biblical Book of Job)
• Lima (Peru) & lima (bean).
• There’s a street in Portland OR. It’s spelled “Couch” but it’s pronounced “cooch”
• Tangier and tangier [I don't know the meanings of either]
• "The poem Job's job by Richard Lederer is a showcase of capitonyms -- August/august and many others: https://digitalcommons.butler.edu/cgi/viewcontent.cgi?article=4232&context=wordways
• https://en.wikipedia.org/wiki/Capitonym#List_of_capitonyms_in_English (has six with different pronunciations)

[jsalsman]

@dhdaines @lenzo-ka @Alexir Can we make a new version 1 with lowercase?

[Alexir]

In cmudict, words can have different pronunciations, handled by the () notation, i.e. WORD(1), WORD(2), etc. Pronunciation may vary by meaning, context, and word position, perhaps by capitalization. You can annotate such features. But that's a different kind of dictionary.