Book Review: Ancillary Justice
By Ann Leckie (goodreads link)
Quite an interesting premise, but this book didn't move me, for all that it was well written. I'm going to speculate on why, since this book had almost all the things going for it that I normally like: a mystery type plot, in space, with fantastical elements, and a fairly high standard of writing. The speculation does not indicate I didn't like the book, merely that I didn't as much as I could have. Lots of spoilers ahead!
Maybe it was the oddly flat characterizations (common in sci-fi, but usually redeemed by very complex world-building, somewhat lacking in this case, or very intricate plots, also lacking). Maybe it was the lack of tension: the way the revelations and crises of the book felt oddly deflated.
Maybe it was that it used a premise I know something about, but did it in a way at odds with real world knowledge - if one is going to write from the viewpoint of an AI, write it like an AI, not an emotionally unaware human being! I would have loved to see concurrency and true parallelism of thought done well, for example, rather than the subject briefly described and then ignored. This is, btw, the classic "physicist discussing Star Trek" problem - although when the technical material is done well (i.e. Neal Stephenson's Cryptonomicon, or Catherine Asaro's description of relativistic effects), it is an immense joy and pleasure.
Maybe it was some of the silliness in the language: seriously, an AI that can be trained to shoot perfectly and imitate a human being has trouble recognizing biological gender? This came across as a hackneyed device, used to please a particular sub-audience, one that I suspect would not have been at all pleased if, say, a generic male pronoun had been used for every character instead.
Quite an interesting premise, but this book didn't move me, for all that it was well written. I'm going to speculate on why, since this book had almost all the things going for it that I normally like: a mystery type plot, in space, with fantastical elements, and a fairly high standard of writing. The speculation does not indicate I didn't like the book, merely that I didn't as much as I could have. Lots of spoilers ahead!
Maybe it was the oddly flat characterizations (common in sci-fi, but usually redeemed by very complex world-building, somewhat lacking in this case, or very intricate plots, also lacking). Maybe it was the lack of tension: the way the revelations and crises of the book felt oddly deflated.
Maybe it was that it used a premise I know something about, but did it in a way at odds with real world knowledge - if one is going to write from the viewpoint of an AI, write it like an AI, not an emotionally unaware human being! I would have loved to see concurrency and true parallelism of thought done well, for example, rather than the subject briefly described and then ignored. This is, btw, the classic "physicist discussing Star Trek" problem - although when the technical material is done well (i.e. Neal Stephenson's Cryptonomicon, or Catherine Asaro's description of relativistic effects), it is an immense joy and pleasure.
Maybe it was some of the silliness in the language: seriously, an AI that can be trained to shoot perfectly and imitate a human being has trouble recognizing biological gender? This came across as a hackneyed device, used to please a particular sub-audience, one that I suspect would not have been at all pleased if, say, a generic male pronoun had been used for every character instead.