I got to spend much less time in Seattle than I did in Boston, though I did manage to walk around the city a bit. I visited Pike Street Market, which looked like it would have been awesome if I could have gotten there much earlier. In the evening when I got there, things were closing down. I also walked to the Space Needle and kicked in the $16 to ride to the top. It’s a bit expensive for the ability to ride an elevator and take some pictures, but it was so peaceful at the top (even with the tourists) that I think it was worth it.
I have been told by many people that Seattle in the summer is spectacular and I have to agree. The temperature was very moderate, the weather was mostly good, and there were people everywhere. It’s a very clean and vibrant city. The customer service at most places was consistently the nicest I’ve ever encountered, which was really surprising. I hadn’t heard anything bad, I just didn’t expect it. The only drawback I saw was that there were a crapload of homeless people, and they were very forward with asking for money. There were also a lot of street musicians, and I’m not so sure most of them weren’t homeless. One guy with no shirt was channeling Kurt Cobain. It was so stereotypically Northwestern grunge that I wanted to laugh.
I managed to finish two books on the journey: The Coming Race
And now for the pictures.
-
- The space needle from the ground in Seattle, Washington.
-
- Some harbor I don’t know the name of in Seattle
-
- Seattle, Washington from atop the Space Needle
-
- Pike Street Market at sunset when everything is over
-
- Protest against Chinese human rights violations in Tibet
On a whim, I saw a late showing of Mummy 3. I was simultaneously happy and mad at Rachel Weisz for not being in it. If she were in it, it may not have sucked so badly, but I’m glad she wasn’t since it was such a giant turd. Brendan Frasier still can’t act and the over-the-top-Mummy-style just broke the fourth wall too much to ever achieve any sort of story immersion. I never cared for any of the characters.
The next Mummy will be about a long dead Eskimo king and a band of dire wolves or woolly mammoths will save the gallant treasure seekers. The next one after that will be about a Mayan king and some sabre-toothed tigers or maybe chupacabra. Oh and they’ll all be seeking to bring the world under their dark dominion. Rinse and repeat.
Night watch sounds very interesting. I want to read the book now and see the movie. I’m most likely going to try and read Stanislaus Lem’s Solaris on my upcoming vacation. I have the English translation from the French translation of a Polish novel (double translation). I saw the first film version from the Russian director Andrei Tarkovsky and this is amazing – haven’t seen the George Clooney version but he is pictured on the cover of the translation I have. I do hope the double translation is not too off the original but who knows. I would have simply liked a straight polish–>English rather than the translation of a translation. I wonder what kind of computational probabilistic linguistic statistics could be generated from this – Monte Carlo Methods. Translation of a translation (ear whispering childs game) with say online language to language translators – good way to mess up Shakespeare I suppose but I wonder what its done with Stanislaus Lem. Remix, remodel
Yeah Night Watch is great, I’m really liking it. I only saw the George Clooney version of Solaris, I wasn’t aware of the previous movie or book until afterwards.
I’m not sure how much you could say in terms of the ear whispering game without having the Polish to English translation, and even then any comparison would probably be difficult. This is same as the general problem with machine translation evaluation. The only way you can really know if a translation is good is if you have a machine capable of determining the correct translation. But then if you have a machine capable of doing that, you have no need to evaluate MT because you’ve already solved it.
I do know that translations of translations has been tried in terms of evaluating MT output, and it sucked. The problem is you have an additional factor that you can’t control. In normal MT evaluation you are looking at one language to another. But with translations of translations, you have two systems perturbing the output (different statistical models are necessary for each (source, target) ordered pair: so (English, French) is a different system altogether from (French, English)). Still, it’s a fun game to play, and there might be some way of doing it that hasn’t been thought of.