
Regarding the new translation, I have to say that by the end of the book I wasn't sure how good it was. In a 2011 interview with The Guardian, Johnston says, "Much is lost when a book is re-translated from an intermediary translation into English, but I'm shocked at the number of places where text was omitted, added, or changed in the 1970 version." He also talks about how names were changed, and some of those changes (e.g. Snaut to Snow) do seem egregious. However, as prose, I found Johnston's translation pretty plain and at times just plain flat. Compared to Michael Kandel's translation of The Futurological Congress, which I read fairly recently, it has no snap, pop, or crackle. Johnston says, "Lem's characteristic semi-philosophical, semi-technical language is also capable of flights of poetic fancy and brilliant linguistic creativity, for example in the names of the structures that arise on the surface of Solaris," but I didn't get much poetry out of his translation. Interestingly, he does specifically reference a passage that I posted to Facebook as a QOTD:
''When all's said and done, though, no terms can convey what goes on on Solaris. Its "dendromountains," its "extensors," "megamushrooms," "mimoids," "symmetriads," and "asymmetriads," its "vertebrids" and "rapidos" sound terribly artificial, but they do give some idea of Solaris even to those who've seen nothing but a few blurry photographs and poor quality films.'
Here's the Kilmartin/Cox version of that passage: '... it has to be admitted that no semantic system is as yet available to illustrate the behavior of the ocean. The "tree-mountains," "extensors," "fungoids," "mimoids," "symmetriads" and "asymmetriads," "vertebrids" and "agilus" are artificial, linguistically awkward terms, but they do give some impression of Solaris to anyone who has only seen the planet in blurred photographs and incomplete films.'
I'll give Johnston the edge on "dendromountains" over "tree-mountsins" and "megamushrooms" over "fungoids," but it's remarkable how similar all the other made-up words are. I don't see much improvement in "rapidos" over "agilus". And I actually prefer "semantic system" to "terms," because it seems more in keeping with Lem's theme that humanity's systems of knowledge as a whole are seemingly helpless to understand the alien entity that is Solaris, but I have no idea which translation is more accurate.
Switching to commenting on Lem's novel rather than the translation, I have to say that this time through I found the parts of the story involving Kelvin and Harey a little bit less than compelling. Harey is a fascinating character because she's a simulacrum created by Solaris from Kelvin's memories of a real person, and she becomes aware that that's what she is. There's an existential cruelty to her situation that would bear some exploration. However, I think Lem's imagination failed him here, and instead she's treated as a kind of passive, self-hating burden on Kelvin's conscience rather than an agent of her peculiar consciousness. The fact that she ultimately commits suicide, just like the real Harey did, is a powerful piece of pessimism in what is a very pessimistic story, but it also felt misogynistic to me, especially considering that the original, human Harey committed suicide because Kelvin was emotionally cruel to her. Ultimately it's all about Kelvin, and while Lem is upfront about Kelvin's failings as a human being, that too feels rote. If you want to make your novel a serious work of literature rather than a piece of genre fluff, you make your protagonist a jerk.
Not that Kelvin is completely unsympathetic, and the story of how he reacts to Harey reveals a tender side in the way that his horror at what she is turns to what he's capable of (not actually all that much) love. I suppose that Lem might have been suggesting similarities between the failures of scientist to understand the living ocean on Solaris and Kelvin's failure to connect with Harey, both the original and the simulacrum. Perhaps Lem is just more pessimistic about human relationships than I am. But Kelvin's inability to rise to the occasion presented by the mysterious creation of Harey doesn't feel grounded to me. If Lem does mean us to see the similarities with the failures of scientific understanding, I guess I'm still not seeing it after three readings. Instead it just reads like bog standard literary romantic miserabilism.

The fact that we never learn what simulacrums visit Snaut and Sartorius (the other scientists on the scientific mission) and that we see the simulacrum that visited Gibarian but never learn who she is reinforces the sense of mystery pervading the novel. Solaris is about the impossibility of knowing anything or connecting or communicating with anybody, human or alien. It's a very dark, disturbing novel that way, as funny as some of the descriptions of scientific theories about the living ocean are. Another thing that struck me this time was how religious the discussion becomes at times, especially at the end. Kelvin proposes a specifically religious theory about Solaris in the last chapter, comparing it to an imperfect, defective god, and he evokes the concept of the watchmaker in his final thoughts about the living ocean, a creator of clocks (the simulacra) that repeat the traumas of the past over and over with only mechanical purpose. Lem was frequently a critic of Western science fiction, which he found too frivolous, and Solaris could stand as a lance harpooning the concept of the sense of wonder, with Kelvin expressing in the final line a kind of religious faith: "I abided in the unshaken belief that the time of cruel wonders was not yet over." (Or as Kilmartin/Cox have it for better or wose: "I persisted in the faith that the time of cruel miracles was not past.")