Turns out its really more "Sohu-y". Oops.

My good friend Sumner Lemon, who used to cover China technology for IDG News Service out of Beijing but is now leading the tropical life down in Singapore, has been covering a developing scandal concerning Google's new Chinese input method editor (IME). It turns out that the word database may have been lifted from Sohu's Sogou pinyin IME without --how shall we put it?-- their express written consent.

Sumner has written two articles. In the first he reported on Sohu's request that Google cease distributing the new IME:

"We have requested they stop offering the software for download as quickly as possible. The second request is they make an apology, and the third is to discuss compensation for Sohu," said Wang Xiaochuan, Sohu's vice president of technology and head of the company's research and development center, in a telephone interview.

Google China executives, who are understood to be discussing the matter internally, had not responded to Sohu's letter by Saturday evening local time and Google's Pinyin IME remained available for download. A Google China spokeswoman acknowledged receipt of an e-mail request for comment, but had not responded at the time of writing.

If Google doesn't meet the letter's deadline, the next step for Sohu will be to "follow the normal course of legal proceedings," Wang said.

A day later, in Sumner's follow-up story, a Google's spokeswoman conceded that Google's IME "was built leveraging some non-Google database resources". As a PR man I can really appreciate the clenched teeth that statement must have been delivered through.

The evidence seems pretty compelling that the "non Google database resources" in question included Sogou's IME dictionary. Among other things, some characteristic misspellings and the names of some Sohu engineers are telling clues.

At the moment, Google's Chinese IME (updated) is still available for download here. If you want to compare and contrast, Sogou's is here.

Forum posters who use IMEs were noting the similarities pretty quickly after Google's IME was made available. Said one poster on Chinese Forums on April 5th:

First impressions: It is extremely similar to the already-existing Sogou pinyin IME. You could take the Sogou IME, give it a Google theme, switch the Sogou search button with a Google search button, and I wouldn't be able to tell the difference. The only difference is that the Google IME offers a "suggestive English" IME which is of basically no use to English speakers.

Similar indeed.

I will be interested to see how Google responds to this situation. Did Google's Chinese engineers lift some of the Sogou IME without permission? If so, this is a serious credibility issue for Google's China R&D operations, and for the idea of doing development in China. Everyone knows that China is not the first country that comes to mind when one thinks of heartfelt respect for intellectual property or for primary R&D (more D, less R is what is often said of China R&D operations behind closed doors). But innovation-driven companies like Google are supposed to be changing all that, creating top-flight labs in China.

An IME is hardly rocket science, but Google has a very valuable reputation to defend. A brief apology is on the Google China blog here (Chinese). Similar to the quote in Sumner's second article, it says that the "pilot phase contains some non Google data sources". But does that mean licensed or lifted? They've also updated the software.

High expectations are easily dashed. Will Google's moves be enough to save them from a scorching?

See also:
China Herald
China Web 2.0 Review