As a linguist, I must say that one of the basic premises of the book – that a race of aliens can’t use metaphors/signifiers (or even think about the concept of the content of the metaphor) unless they refer to things that happened in reality first – is totally unbelievable from a linguistic point of view.
There are a number of reasons for this, one being that the aliens are able to communicate about concepts in need of a new signifier, so that they stage events in reality to use a figure of speech later. If they are able to communicate about them, the new signifier isn’t really needed in the first place, and they are able to think about them too. Logically, this short-circuits.
Also, figures of speech in human language that are based on metaphors start out as referring to reality too, and only in a later stadium get their additional meaning as figure of speech. There is nothing complicated or deep about this, nor does it alter the way we think.
Sadly, since to book is a bit pretentious about its concepts, and if you know a thing or two about the principles of language, it makes the story fall flat. It’s not only back cover posturing, but also the heavy-handedness of the theme in the book itself, the kinda silly wordplay with TwoNames, etc.
On the other hand, a big enough part of the rest of the story, its world building, the depiction of the aliens (as well as their double voiced language) is often interesting, with some original ideas. Characterization isn’t great though, and some parts of the book are pretty boring.
Conclusion: Embassytown is an okay book, but it is not to be taken too seriously as a philosophical/linguistic sci-fi treatise on the Saphir-Worph hypothesis.
originally written on the 27th of January, 2015