Chinese phone books and dictionaries

I did see a Beijing Business phone book a few years back, but I can't quite remember how it was organised! Dictionaries, though, are an old friend… Most PRC dictionaries are organised in alphabetical order by Hanyu pinyin (romanisation). However, words are still organised under head characters, so all words consisting of more than one character are displayed in alphabetical order under their head character, eg all words starting with Tou in the first tone, (to steal) are listed under Tou1 (as are any other hazi pronounced tou1), before Tou2 and any words which start with it. So they are not purely alphabetical, the only purely alphabetical dictionaries are the Hanyu Pinyin Cihui, John DeFrancis' ABC Chinese English Dictionary, and a small Japanese learners dictionary.
Alphabetical order is of course a modern method. Some dictionaries, eg the Far East dictionary and most of the large scale dictionaries like Hanyu Dacidian, Zhongwen Dacidian etc, use the Kangxi radicals or varieants thereof. The Kangxi dictionary was a very large Imperial dictionary, which organised the characters by their components. These are the radicals, significs, call them what you will. The Kangxi index of 214 radicals, and Chinese Academy of Social Sciences catalogue of (I think) 192 radicals are the two most common interpretations of this concept nowadays, but it is an ancient one, going back to the Shuowen Jiezi by Xu Shen. I hope none of my teachers come across this, as they would be most disappointed that I can't remember how many radicals this had ( I think about 4000) and when it was written ( I think later Han).
Anyway, these radicals, however many, are arranged by the number of strokes taken to write them. So first, you decide which part of the character is the radical, then you look at the list of radicals with that many strokes. Finding the radical there then tells you what page all the characters with that radical are on. Turn to that page, where the characters are arranged again by the number of strokes in them (all in ascending order). Once you find the right number of strokes, then locate the character, you have the page number, and can find whatever word or quote or telephone number(!) you were looking for.
I prefer the alphabetical way myself…


edit…I forgot to add, justt to scare you really, the hanyu Dazidian, by including old (and modern) variants of Chinese characters including some Korean and Japanese forms, comes up with about 55,000 characters total:eek:
But thankfully there is a list of 3500 commonly used characters these days, and 7000 characters "in regular use". The others are mostly met by Scholars!