A,B,C’s, Kanji and Orthogonality

Most people recognize that HL7 is non standard and difficult. Nothing terribly insightful to repeat that statement.

What people have a harder time with is really understanding why. One big problem is its lack of:

Orthogonality.

Uh oh. Into PhD speak.

As an old boss I had used to say – please just give me the hood of anorak – not the entire coat.

Fortunately there is a simple example of orthogonality which is easy for anyone to grasp even if you don’t have a technical background.

Compare the alphabet to kanji.

Kanji is the name of the symbol system used to write Japanese. Kanji is a part of a family of writing systems called logographies where each symbol represents a single word. There are 3000 common symbols in Kanji and by some estimates up to 100,000 less commonly used symbols.

I’ll assume if you managed to read to this point that you know the alphabet. It has just 26 symbols which roughly correspond to sounds in spoken language. By combining these sounds together any word can be represented.

The alphabet is much more orthogonal in its design than kanji. i.e., a smaller set of building blocks that are flexibly combined into words. Because of this:

  • Learning the entire alphabet is easier.
  • It’s simpler to design software around the alphabet since there is less effort required to represent it, index it, search etc.
  • It’s less effort to extend languages based on the alphabet to create new words

So breathe a sigh of relief. It’s not a complex concept. It’s just the idea of system that has fewer, more fundamental building blocks that can be combined together with more flexibility. This will make it much easier to discuss that lack of orthogonality within HL7 2.x.

Incidentally after World War II the Japanese government did carry out a set of orthographic reforms in an attempt to reduce the number of symbols to facilitate learning for children and simplifying kanji use in literature and periodicals. Native Japanese speakers could probably give a much better perspective on their language than I since I regret I am only (somewhat) fluent in English.

Eliot Muir – CEO of iNTERFACEWARE

Leave A Comment?

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.