That depends on the person doing the word game. Generally, most people who do these things would like to do them with only uncapitalized words, but if they can’t find an appropriate lowercase word, capitalized words will be usually be used.
It also has the definition of something in the shape of the letter. This definition applies to all the letters. I wouldn’t consider such usages to be proper nouns, since they’re not names, but they’re often capitalized anyway.
OK, but the trouble with proper nouns is that there are so many of them. If you want to use “wana”, say, no doubt there is a village in Mali or somewhere called that. Or, you could register on a message board under the user name Wana, and hey presto, proper noun. Common nouns have to get past teams of crack lexicographers to get into reputable dictionaries.
Then you need to be consistent. Every letter is listed as a common noun in most dictionaries.
Another consideration is that by excluding all but three of the letters, you’re preemptively excluding about 85% of all words from this particular form of word play.
As for the other way that Annie-Xmas wants, Borgmann also did those. He called them successive beheadments, since he was looking words getting shorter, but it comes down to the same thing. He came up with a couple long ones:
aspirate
spirate
pirate
irate
rate
ate
te
e
prestates
restates
estates
states
Tates (a people who live in Transcaucasia)
ates
tes
es
s
I’m not to sure about the “te” and “tes” in the above. In the first, he actually put all the words in a poem, but the “te” was in the phrase te Deum. In the second, he claims “tes” is some kind of sol-fa syllable or perhaps the plural of one.
OK, we’ve already established that these words are called curtailments, though I don’t think anyone has yet answered your question about what the longest one is in English. According to the Word Ways article I linked to upthread, logologist extraordinaire Dmitri Borgmann found the longest (10 letters) in 1965:
All the above are in the second edition of Webster’s New International Dictionary, except for ALBER, which is the native surname of German poet Erasmus Alberus.
Again, all the words are in Webster’s, except ANG (the hairy part of an ear of barley; English Dialect Dictionary), ANGE (trouble; Oxford English Dictionary, 2nd Edition), and ANGELI (a town in Finland, and a common Italian surname).
I just did an exhaustive search myself, and came up with three further ten-letter examples, though nothing longer. All the words are in the OED2 (though BARRETT appears only as the attribution for certain quotations):
The Oxford English Dictionary (2nd edition) lists a noun “Ate” meaning “infatuation; made impulse” and an obsolete noun “ate” meaning “food”. Possibly Borgmann’s “ATE” is one of these.
Well, if you admit dialectal and obsolete forms, then the OED has plenty of non-abbreviation entries for “co”. You could say that your “co” refers to a jackdaw, or a young boy, or a form of the verb “come”.
Well, I finally got around to subscribing about a year and a half ago, and in perusing the back issues have read the dozen or so very interesting articles you yourself penned. Given all the language-related questions you answer here on the Dope, I’m surprised you haven’t published anything there in the last ten years, though. Can I ask why you stopped?
Thanks. I used to have most of them on my own website, but that was at a previous ISP. I no longer have a site.
It’s a matter of enthusiasm and energy. It takes a lot more of that to research an article than just to be a reader. About 10 or 12 years ago I somewhat lost interest in word play. I still read the magazine, but just can’t get up enough enthusiasm to do the research.
I’m not sure I put that article on the website. At least I don’t have an html version of it on my current computer. It may be on an older computer.
A computer program to find them would be useful. Back in the 90s about the time I wrote that article, Ross Eckler requested a program to find if a word was queen-graphable. I wrote a simple brute-force program and then a simple modification converted it to king-graphable. I didn’t use the program for that article (the request came after I wrote it, IIRC), but I wrote the program in a way that would not have been helpful for that particular problem, anyway (it was easier to write that way).