Their research examines previously unnoticed systematic aspects of English spelling and explains how the system emerged. One suffix that the authors report on is -ous, found in words like nervous and hazardous, which turns nouns like nerve and hazard into adjectives. They discovered that the final letter sequence ous may serve as a flag, informing readers that a word must be an adjective. The sound of this suffix is a short vowel (linguists call it schwa) followed by s: ?s. The authors found that any written English word that ends in the three letter ous is an adjective and conversely that a word that that ends in the sound ?s but is not an adjective is never spelled with ous: service, genius, menace. They also have found similar patterns for other adjective suffixes, including the -ic of allergic, the -al of final, and the -y of funny.
For each of these affixes, the authors analyzed a large sample of written English documents dating back close to a thousand years. For every word that follows each of the regular pattern nowadays (e.g., hazardous, allergic, final, and funny), the linguists looked at every instance in their sample, keeping track of how it was spelled. They found a number of spellings for each suffix over time (e.g., ose, ows, is, owse, ys, es, ouse, us, and ous for modern ous). For every suffix, though, one spelling eventually won out and each suffix followed the sort of pattern that is known from biological competition between species.
This is a striking example of self-organization: No one is or was in charge of English spelling. As opposed to countries like Italy and France and Israel, where national academies oversee the written language, no English-speaking country has a language academy. And yet, somehow, the written language slowly but gradually evolved a system of marking word categories in cases like nervous - despite the fact that this system was never purposely designed. What is apparently a nuisance - we can spell one word ending in more than one way - is actually the trade-off for the grammatical 'flag' that says (in case of -ous), 'This word is an adjective'.
As a follow-up to this research, the authors are now testing their findings experimentally on fluent readers of English. Do they use the regularities that we found when they read? Furthermore, can these be used to help children and adults to learn to read more quickly and more fluently?