现在的初学者最好别选五笔输入法

有人喜欢白色，振振有词——我呢，就喜欢纯洁；有人喜欢黑色，也同样振振有词——我就喜欢纯粹！这叫偏好，没什么需要争论的，当然，喜欢归喜欢，因偏好而标榜的理由——纯洁抑或纯粹——实际上不仅没道理且十分装蛋。

Source: https://web.archive.org/web/20111026160223/http://www.lixiaolai.com:80/archives/8338.html

Beginners Today Had Better Not Choose the Wubi Input Method

> By Li Xiaolai · 2010-01-02 · [Original (web archive)](https://web.archive.org/web/20111026160223/http://www.lixiaolai.com:80/archives/8338.html) (Chinese)
> Our translation from the Chinese original.

Some people like white, and argue for it with great conviction—me, I just like purity; some people like black, and argue for it with just as much conviction—I just like being pure and unmixed! This is called preference, there's nothing that needs arguing about. Of course, liking is liking, but the reasons paraded on account of a preference—purity or unmixedness—are actually not only unreasonable but extremely pretentious bullshit [Translator's note: the original writes 装蛋, a coarse euphemism for 装逼, "showing off, being pretentious"]. Preferences need no reasons, because they are determined by genes, not by thinking.

However, some other things are choices rather than preferences. For example, choosing a pinyin input method rather than the Wubi input method, choosing Google rather than Baidu, choosing modern medicine rather than traditional Chinese medicine. Genes can't decide these, so they can only be choices.

In the early era of "ten thousand codes galloping" (万码奔腾) [Translator's note: a pun on the idiom 万马奔腾, "ten thousand horses galloping," referring to the era when countless Chinese input encodings competed], the pinyin input method was almost the least favored input method. Because intuitively, a single Chinese character might, even after five keystrokes, still present the problem of choosing among homophones (for example "账"—zhang4). As it looked at the time, Wubi had at least two advantages over pinyin input: 1) fewer keystrokes; 2) no duplicate codes.

However, **Wubi has an innate defect, because it doesn't really count as an input method; in essence it is a "copying method."** It only has some advantages when "typing whatever you see," and when "typing whatever you think of," it is clearly at a disadvantage. When Wubi users write while thinking, their thinking is often interrupted by "recalling the code" (perhaps the codes that are all too common have long since been "remembered by the fingers," but certainly some codes take time to recall). All along, people have said (and Wubi has also advertised it this way) that Wubi is the fastest input method; in fact it is pulling a fast one: it is only fastest when copying; once you're off script, its speed drops to half of its copying speed, or even a quarter[[1](#footnote_0_8338 "This is data obtained from the many user tests done during research on language recognition by those developers who eventually chose to develop pinyin input methods. Perhaps someone will say that because they were the ones developing pinyin input methods, this survey is biased. I don't think so—these people did the surveys in order to make decisions, and the eventual outcome of a decision meant investing time, energy and money, so they're unlikely to gamble with their own future.")].

About two years ago (2008.2), [someone made a statistic based on user download counts](http://download.zol.com.cn/download_order/ime_order.html):

> [Translator's note: quoted from a download site's ranking list (product names), not Li Xiaolai's words.]
> Download popularity TOP10
> 1 Sogou Pinyin Input Method 3.1 official release
> 2 Microsoft Pinyin Input Method 2007
> 3 Ziguang Huayu Pinyin Input Method (Ziguang
> 4 Intelligent ABC V5.23
> 5 Google (谷歌) Pinyin Input Method
> 6 Wangma Wubi Input Method 86&98
> 7 Pinyin Jiajia 5.0 official release
> 8 Universal Wubi 2006 New Year edition
> 9 Jidian Wubi V6.1 official release
> 10 Intelligent Chen Qiao Wubi V5.802

The pinyin input methods held the absolute upper hand. And in fact, pinyin input methods are also the input methods bundled with the Windows operating system, so a considerable number of users don't necessarily need to go and download an input method.

There are many reasons why the pinyin input method eventually gained the upper hand. Double-pinyin input is basically two keys per character, and the language models now commonly built into input methods (commonly called "word libraries"; even simplified, they are still quite powerful) mean that abbreviated pinyin (incomplete pinyin) can in many cases even achieve "one key per character"; and because characters are chosen on the basis of words, or even on the basis of sentences, the rate of duplicate codes is greatly reduced, to the point that input methods other than pinyin, relatively speaking, no longer have any advantage worth mentioning. (Interested readers can read this article: [The Beauty of Mathematics series (23): How many keystrokes does it take to input a Chinese character — on Shannon's first theorem](http://www.googlechinablog.com/2007/12/blog-post.html) (数学之美系列（二十三）输入一个汉字需要敲多少个键 — 谈谈香农第一定律).) Another reason that has made more people used to pinyin input should be the spread of mobile phone input. Things like Wubi have almost no use at all on mobile phones.

Compared with pinyin, Wubi is very hard to learn. Perhaps precisely because of this, people who have learned it with great difficulty consider it a "skill," even an advanced skill. Of course, ten or twenty years ago, it really was a skill that could be used to find a job. So, from what I've observed, very few skilled Wubi users switch over to pinyin input, even though all kinds of Wubi software have looked more or less out of date over the course of Windows' continual upgrades (it's so even just from the interface), and are even quite hard to find. But this is easy to understand: Wubi users' earlier costs were too high (the input method was bought, learning it relied on taking a training course, and at the same time the time cost invested was many times higher)—that is, the so-called "sunk cost is too large."

But for beginners now, choosing a pinyin input method is the best deal. Because in the long run, pinyin input is the most likely to become the kind of input method that "users can't feel they are using."

Footnotes:

1. This is data obtained from the many user tests done during research on language recognition by those developers who eventually chose to develop pinyin input methods. Perhaps someone will say that because they were the ones developing pinyin input methods, this survey is biased. I don't think so—these people did the surveys in order to make decisions, and the eventual outcome of a decision meant investing time, energy and money, so they're unlikely to gamble with their own future. [[↩](#identifier_0_8338)]