I recently read J. Scott Miller’s Adaptations of Western Literature in Meiji Japan (New York: Palgrave, 2001) and am full of Thoughts on Meiji writers, literature, zeitgeist, continuity, and adaptation. Let me express some of them here.
It’s recently come to my (belated) attention that there is an amazing resource provided by the Japan Foundation: the Japanese Literature in Translation Search.
This database contains full bibliographic information on the translations and lets you search by year, which is crucial for me when I’m trying to find what’s recently been translated. There are some shortcomings, however, and the biggest one is that there is no bibliographic data for the original Japanese publication. If I want to search on what’s been translated that was published between, say, 2000-2012, I can’t do it. My search for what was published in 2000-2012 yielded over 50 pages of results, which varied from contemporary works to The Tale of Heike.
My ideal database would have the publication information for the Japanese originals, and the ability not just to search by that, but also to browse: for example, by reign year (Heisei literature, 1989-present) or by Western year (1990s). More metadata would also be helpful. For example, what if I wanted to find out what Naoki prize winners had been translated? Akutagawa prize winners? Detective fiction? Horror? None of this is possible.
I’m not trying to denigrate this database in any way, because it’s an invaluable resource and one I’ll be using from now on. But there is so much potential for things to do with this data, and it’s unfortunate that the dataset itself is so limited. I could see some cool visualizations resulting from it, but they’re along the lines of “in a given year, what works were translated that were from which decade or reign era?” A simple set of sized bubbles showing the 2010s, 2000s, 1990s, and 1980s in a given year would be a fantastic way to quickly see trends in translation and gauge how much very recent fiction is being translated – in other words, how fast the turnaround is.
Given the focus of publishers such as Vertical, too, it would be interesting to see trends in what recent fiction is being translated in terms of genre and author.
Creating another database that does have this metadata does not seem feasible or reasonable, however. This post is about a case of wishful thinking, and praise for a huge undertaking that does a great service to the community of Japanese literature scholars and students.
I’ve come across a few sites lately – commercial and academic – that offer “translation” into several languages. I click on them out of curiosity. The link does not take me to a translation.
It takes me to Google Translate!
I have a few things to say about this:
- It’s unprofessional. It looks unprofessional too. That bar at the top of the site that says it’s being translated by Google? Kind of ruins the effect.
- If you’re just sending your site text to Google, why not offer all languages? I just visited a site that offers about 7; another (Rakuten) offers a handful. Google can handle more than 7 languages.
- It’s deceptive. Listing only a few languages makes it look like you’re offering actual human translations that you made or commissioned yourself. In fact, this is what I thought about all of the sites that I’ve visited recently that “offer” translations via Google. There’s no button that says “powered by Google Translate,” only a few flags that the user assumes lead to a real translation. They don’t. They lead to Google.
- Last but not least – it’s laughable. The translations aren’t so much inaccurate as hilarious. The only reason I can use the English Rakuten.jp is that I know Japanese and can guess at the original phrasing that produced such funny English. I have to double-check everything with the original Japanese text that it helpfully supplies a link to.*
This is the impression that your “serious” commercial or academic site is leaving: you’re naive about machine translation; you’re too cheap or unimaginative to get your own local translations; you’re out to deceive your users; and you’re an Engrish generator.
Is this the impression you want to leave?
Machine translation isn’t here yet. It may never be. Machine translation is hard. At the very least, get a native speaker of each language to read each end of the translation. You can identify the parts that need to be fixed simply by watching them making funny faces.
Google Translate may be okay in a pinch if you need to, for example, order something from a storefront or service that isn’t in your language. If you want to be taken seriously, find a human.
* The reason I use Rakuten’s international English site is to narrow down the stores to ones that ship internationally. I wish I could use the Japanese site with this filtering.