Harvard physicist's take on the VMS

[Jorge_Stolfi]

[history and backwards compatibility] can be seen everywhere, not only in scripts: why is the layout of a standard keyboard so weird and sub-optimal? Why is the machine-level language and registry organization of today's Intel processor so strange (and sub-optimal) (*)?

This story is somewhat famous in engineering schools:

• Nasa's Space Shuttle used two disposable solid-fuel boosters during lift-off.  They were made in sections that were manufactured in Utah, shipped to Florida by rail, and bolted together at Cape Kennedy.
  
• Ideally the boosters should have been wider, but their diameter was limited by the width of the railroad tunnels that they had to go through.
  
• Railroad tunnels were just wide enough to allow two trains to go through at the same time.
  
• The trains, in turn, were only as wide as possible for two trains on adjacent tracks to stay clear of each other.
  
• But parallel tracks were laid down as close as possible.  Thus the width of a train was just a bit more than the spacing between the rails of a track.
  
• The spacing between rails on the other hand had to be standardized, equal to the spacing between the wheels, and vice-versa.
  
• That standard spacing had been the same since the first steam trains were built in England.
  
• Those first trains used the same wheels and axles as horse-pulled carriages.
  
• The spacing of carriage wheels had to be standardized because they often created ruts on dirt roads, and a carriage with a different wheel spacing would topple over.
  
• Thus the spacing between wheels of carriages in England had been the same since the Romans built roads there.
  
• Like train tunnels, Roman roads were just as wide as necessary for two carriages to go past each other.
  
• That meant carriages could not be wider than the pair of horses that pulled them.
  
• And the wheels had to be as far apart as possible, for maximum stability,
  
• Thus the diameter of the Shuttle's boosters was determined by the width of the ass of a Roman horse.
  

(Some  steps of this story have been disputed, but it makes sense on the whole...) 

There is also the story of why the TAB key is called that, and why it behaves in that funny way.  And why ASCII character \015 is called "carriage return" (CR), and why Microsoft Windows uses two characters (\015\012), instead of just one, to separate lines of text files.  And why ASCII \177 is called "delete" (DEL).  And much more...

A freshly designed script, of course, could be much more 'rational', but this surely requires a lot of work and especially a lot of knowledge about phonetics and linguistics, and it would be hard to get people to accept it (even if there are examples where this happened, but starting from no script at all, as for Cree, or in the case of Hangul from a script which was based on a foreign script and was replaced for nationalistic reasons, if I'm not mistaken).

Hangul did not borrow anything from the Chinese script that was used before in Korea, except the square outline of the characters (each being a phonetic encoding of a syllable), the writing direction (originally top-down with columns right to left) and the assumption that the characters would be written with a brush.  Unlike Chinese, where each syllable is a word with some generic meaning, Korean has polysyllabic words, thus each character is purely phonetic -- a combination of consonant signs and vowel signs -- without any specific meaning. From Wikipedia:

You are not allowed to view links. Register or Login to view. to be graphically simple, and traditionally consisted of only straight lines, dots, and circles. The shapes of five basic consonants are based on those of You are not allowed to view links. Register or Login to view.. Most of the other basic consonants, which are considered to correspond to "harsher" sounds than those five, are derived by adding additional lines to those letters to indicate progressively harsher sounds. There are a You are not allowed to view links. Register or Login to view., but these are still debated.

By the way, the Japanese phonetic scripts You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. have that "logical" notation for the voiced/unvoiced bit: き = ki,ぎ = gi;  た = ta, だ = da,  etc.  Except that there is a glitch on the h/p/b columns: ひ = hi, ぴ = pi, び = bi.

All the best, --stolfi