Languages: ambiguous parsing

There is one reason computers are great at numbers and awful at languages: the latter are difficult to parse. While complex mathematical operations can be carried out in a well-known order, parsing text can be exruciating difficult even for humans.

This is especially true for languages — such as English — that allow long sequences of words to be joined together without prepositions, and that use the same word both as a noun and as a verb.

Take for instance this news story from New Zealand. The headline is “Police chase driver in hospital.” There are two ways to parse it:

  1. Policemen have chased a driver within a hospital’s premises
  2. A driver who was chased by the police was hospitalized

(Note that in Australian/NZ English, just like in British English, collective nouns are usually conjugated with verbs in the third plural person, unlike in American English.)

Such ambiguous phrases don’t even require verbs. The noun phrase “The beautiful girls’ school” could be interpreted as:

  1. The beautiful school that is for girls only
  2. The school that is only for beautiful girls

There is no solution for this issue, except for rewriting such phrases in a more explcit way. For purely educational purposes, using parentheses  may come in handy to identify the building blocks: “(Police chase) (driver)” vs. “(Police) (chase driver).”

Or we could all switch to Lojban