![]() |
![]() |
![]() |
![]() |
|||
LOGOLOG
a weblog of wordplay by Eric Harshbarger "Longest" and "Shortest" wordsIf I were to ask you, "what is the longest eight letter word?" you might spend a few seconds thinking of words such as "pneumonoultramicroscopicsilicovolcanoconiosis", before realizing that I've asked you a trick question. I asked for the longest eight letter words...But I'm actually not trying to trick you; I've just worded my question poorly. Rather, maybe I should have asked, "what is the widest eight letter word?" Meaning, when typed out, which eight letter word extends farthest along its line?
This is a question I thought about recently and decided to figure out which words in the NWL23 Scrabble lexicon are Now, of course, there are some important parameters that need to be defined before this search is done. Foremost is: what font is being used to "print" the words. I'll stick with a fairly traditional choice: Times New Roman. Also, I will measure the width of the words as they are typed in lowercase letters, not uppercase. Both of these choices above are rather arbitrary, but choices have to be made, and since I'm doing this search, I get to make those choices. Finally, there's another choice that needs to be made which is not at all obvious: at what size will I print this font? One would think that that should not really make a difference, but I found that because of subtle kerning issues, rounding calculations and such, the results can vary depending upon how large I make the font. So, I decided to use a 36 point size with the Times New Roman font. Again, a choice had to be made, and 36pt was about as large as I could make the output images (see below) so that they would easily fit on this webpage. So, how did I actually calculate the width of all of the words? Easy enough: I simply wrote a quick Java program that would read in all of the words in the lexicon, create a FontMetrics object associated with a Graphics object, and use the .stringWidth() method to determine the width of each word (the internal workings of this method are what caused some of the kerning and rounding issues I mentioned above; there's a little "magic" going on here, but I'll trust the FontMetrics object to know what it's doing).
So, with all of that preamble out of the way, what are the 2-LETTERS: mm 3-LETTERS: mmm 4-LETTERS: hmmm 5-LETTERS: mommy 6-LETTERS: mummed, powwow, whammo (tied) 7-LETTERS: mugwump 8-LETTERS: powwowed 9-LETTERS: mammogram 10-LETTERS: mammograms 11-LETTERS: mammography 12-LETTERS: homomorphism 13-LETTERS: homeomorphism, monomorphemic 14-LETTERS: newspaperwoman, newspaperwomen 15-LETTERS: morphophonemics, recommencements
Of course, the above list, displayed in a monospaced font, is kind of missing the point. So, here are the actual words, output by the program as rendered in 36 point, Times New Roman font:
It's obvious the that "m" character controls the conversation here ("emdashes", anyone?), though the "w" certainly made its mark with the POWWOWED entry. And what if we flip the question? What are the narrowest words of each length? Well, it wasn't that hard to adjust the program to output those. And, again, we'll see certain letters dominate the results: li ill jill trill illite illicit, illituc, titlist titlists illicitly titillates, triliteral fritillaria fritillarias, fritillaries, risibilities fallibilities, irritabilities, infallibilities And here they are rendered:
Finally, let me re-list all of the words juxtaposing the widest and narrowest words, so one can visualize the difference between their printed lengths:
Anyway, that was a fun way to spend an hour or so. And the next time someone asks me what the longest word I've ever played in Scrabble is, I'll have to ask them to clarify their question... -- Eric
[16 April 2025]
|
![]() Archive
|
|||||
![]() |
![]() |
|||||
![]() |
||||||
![]() |
![]() |
|||||
Leave a comment on LOGOLOG about the article: "Longest" and "Shortest" words |
||||||
![]() |
![]() |
|||||
![]() |
||||||
![]() |
![]() |
|||||
|
||||||
![]() |
![]() |
![]() |
![]() |