Arthur C. Clarke left us in early 2008 at the age of 90, a monumental legacy behind him that any dozen authors would be proud to own a fraction of. Frederik Pohl was himself pushing 90 as The Last Theorem hit the stands later that summer. That two of the genre's giants should come together in the twilight of their careers — indeed, this is Clarke's first posthumous publication — seems a fitting tribute to an era of SF whose most prominent names are now almost all gone. Since 2007 we've lost Lloyd Alexander, Robert Asprin, Algis Budrys, Colin Kapp, Sterling Lanier, Madeleine L'Engle, Fred Saberhagen, Kurt Vonnegut, and Robert Anton Wilson, among others. The march of time is an inexorable thing. But the books will always be with us, keeping alive SF's rich past, inspiring imaginations well into the future.

After such an encomium, what else could I possibly do but be a complete jerk and bitch about how disappointing The Last Theorem is? Well, it can't be helped. The mantra of every critic must be complete honesty and openness about one's opinions. And it's not as if the bibliographies of even the greatest veteran writers aren't full of books good, bad, and indifferent in equal measure. That Frank Herbert's Dune sequels are considered by SF fandom at large to be inferior to one degree or another does not diminish the accomplishment of the original. And I always cringe when I see a critic fall into the pandering trap of gushing over everything an artist has touched, even if the most recent work isn't up to par, merely because the artist has recently passed on. Call me a bitter middle-aged cynic, but I have a sneaking suspicion that had Stanley Kubrick not died right as Eyes Wide Shut was being released, the majority of critics hailing it as his final masterpiece would have taken chainsaws to it instead.

The Last Theorem is a book I really resent not liking, because there's so much about it that deserves to be liked. For its first two-thirds, in fact, it's so good (vastly more satisfying than Clarke's final collaborations with Stephen Baxter, the underwhelming A Time Odyssey trilogy, to be certain) that it almost, but doesn't quite, make up for the nonsense that overtakes the whole affair at the climax. There are, essentially, two stories being told here. And it's fairly easy for readers familiar with the work of both authors to demarcate which is Clarke's and which is Pohl's. The good story is largely Pohl's: it's the compelling character study of a young Sri Lankan mathematical prodigy, Ranjit Subramanian, who, while imprisoned in India following a tragic series of misadventures and misunderstandings, manages to solve the Holy Grail of mathematics, constructing a proof of Fermat's Last Theorem.* (No, of course we aren't told what it is, as that would involve the authors' pulling off the earth-shaking feat themselves. For the purposes of the story, we are asked to suspend our disbelief Ranjit has accomplished this, and the story is so skillfully woven that we can.)

The bad story would appear to be mostly Clarke's, as it recycles a premise he's featured prominently in most of the major works of his post-2001 career: that of ineffable, godlike aliens watching over us and passing judgment upon us. It isn't just that Clarke has been to this well too many times already. It isn't just that the aliens as conceptualized here are almost wincingly pulp-retro (they're called the Grand Galactics, which frankly sounds like a roller derby team or '70s disco band or something similarly naff). It isn't just that their motivations for deciding we need to be wiped out for the good of the universe aren't the least convincing, particularly as it's made clear that all our nuclear bombs and such cannot possibly threaten them. It's all of those things put together, then topped off by the entire invasion plot wrapping up as the biggest shaggy dog story ever. What can you say about an invading alien armada coming light years across the galaxy solely to "sterilize" the Earth, whose commanders all at once change their minds upon reaching the solar system to become our greatest benefactors instead, gifting us with new energy sources and immortality? And wasn't this all supposed to be about the Last Theorem, anyway?

It's so easy to see where this all went wrong. Pohl is the math geek here. His enthusiasm for the subject made even a lousy math student like me utterly engrossed in the arcane (to me) world of number theory in which Ranjit obsessively works. Pohl's little trick for counting to 1,023 on your fingers made me smile hugely, though you math whizzes out there might not be all that impressed. Once Ranjit publishes his proof, he gets drawn into the tempestuous political arena of the mid-21st century world in which the book is set. A United Nations force called the Pax per Fidem has accomplished a shaky form of world peace, through the application of a new EMP-zapping superweapon they call Silent Thunder. After knocking out North Korea and some trouble spots in South America, most of the rest of the world's regional conflicts simmer down. But while many celebrate that the geopolitical waters have been calmed, some, like Ranjit's wife Myra, an AI researcher, are deeply skeptical about the "new world order" (for lack of a better term) that has arisen. She makes ominous comparisons to the superpowers of Orwell's 1984. Meanwhile, Ranjit finds himself pressured to take on highly secretive work for the US government — depicted here as run entirely by reactionary, shoot-first-ask-later loose cannons, no doubt inspired by two terms of Bush.

All of this is so well done, so suspenseful, so compelling, that there's enough in it for a perfectly good novel right there. (In a nod to current affairs, there's a pretty harrowing waterboarding scene.) As it is, grafting a horde of cartoon aliens onto the thing comes off as an act of artistic miscalculation bordering on temporary insanity, as if Leonardo had given the Mona Lisa a once-over and decided that what the old girl really needed was a big yellow sun with a smiley face peeking over her shoulder.

Perhaps the thinking was that a futuristic political techno-thriller starring a mathematician might not be sufficiently SFnal. But I can't imagine either of these two grand masters thinking anything so senseless. I think what happened in this meeting of minds is that two disparate storytelling sensibilities meshed in a way far less harmonious and compatible than Clarke and Pohl may have sincerely thought. The Last Theorem is not simply an uneven novel. It's a deeply disjointed one, a book that introduces a host of wonderful ideas and sympathetic, believeable characters, only to decide it doesn't trust them to carry off the story in the end, ultimately falling back — disastrously — on dated and dubious formula. I'm sad that The Last Theorem is Clarke's last book (I earnestly hope it won't be Pohl's). But I'm reassured that its shortcomings won't have a deleterious effect on the greatness of his legacy.

* One of the things about a book like The Last Theorem that is praiseworthy is that it sends me on a major Googlequest to fill in gaps in my education. So I'm aware of Wiles' 1995 proof of Fermat's Theorem, which those of you who do have a mathematical background were probably wondering about while reading the review. Pohl states, in an appendix, that "hardly anyone, however, was satisfied" with Wiles' proof, as it was too long, dense, and utilized mathematical concepts unknown in Fermat's time (so Wiles' proof couldn't have been Fermat's). The novel has Ranjit using the work of Wiles and other mathematicians like Sophie Germain in constructing his proof.

If you're a mathematician and think Pohl is full of it about Wiles, please, take it up with him, don't send me an email harangue. Okay? I was lucky to pass college algebra!