So there is this building for lease that I occasionally pass on West Burnside. It is called The Weave Building and has a logo that is almost, but not quite, an ambigram:
I like how they were a little liberal on the letter usage in the almost-ambigram of “building”, leaving part of the descender off the G and still dotting the i’s.
I wondered how many words in the English language were naturally ambigrams using these fast and loose rules. I started with a list of matching letters as rules, trying to be very relaxed with the rules. For example I considered the A and V to be equal in much the same way the Weave Building considered g and b to be — it’s easy to overlook the horizontal line in the A. A lowercase h could look like a y upside, or when uppercase it could look like another H. And so on. The final list of rules looks a little something like this:
- A => V
- b => gq
- d => q
- g => b
- h => Hy
- i => Ii
- l => 7l
- m => Ww
- n => Nu
- o => Oo
- p => d
- q => b
- s => Ss
- t => t
- u => n
- v => A
- w => Mm
- x => Xx
- y => h
- z => Zz
I then wrote a quick C program to scan the Unix dictionary. A few seconds later, I had a measly 28 words that were dictionary ambigrams according to my rules:
- big
- bog
- bung
- gib
- gob
- hoy
- i
- l
- mow
- nu
- o
- oto
- otto
- pod
- s
- sis
- sooloos
- t
- tit
- toot
- tot
- tst
- un
- unsun
- wim
- woom
- x
- z
And, admittedly, some of those aren’t even proper words but single letters! But I did catch great words like bung and toot!