Alphabetization

I hate alphabetical sorting! It is:

• arbitrary
• useless
• much more difficult than you might think, especially when taking different languages and writing systems into account.

Arbitrary and Useless

• alphabetical ordering conveys no useful information
• to look something up in an alphabetical list you need to know what it's called. Why not just search for it, then?
• maintaining the order has a cost

Difficult

• what do you do about punctuation?
• what do you do about differently-cased letters?
• what about accented letters?
• what about other writing systems?
• how do you know if your list is in English or Swedish?

There are libraries that will mostly handle all of this for you. But since alphabetical ordering is useless, even the relatively small effort of using the library hardly seems worth it.

Why does it persist?

• objective (in theory)
• stable
• the legacy of paper dictionaries

Better Alternatives

• priority order
• reading order
• chunking into subcategories of 5–10 items
• random order (really!)

Summary

Lexicographic ordering and the binary search lookups it enables are fundamental to many software algorithms. But they should be done by and for machines, not humans. Even written English has outgrown the alphabet: in its quest for world domination it has swallowed not only foreign characters but punctuation like / and #. It probably won't be long before there are English lexemes with emoji in them.

We should forget the alphabet.

This article is a stub.