2 of 2 people found the following review helpful
on 3 February 1997
Casti is a lucid and entertaining writer and has sufficient depth in his own subject and breadth in other disciplines as to be well placed to write a book on modelling and simulation. This is an excellent introduction to some of the mathematical, logical, and philosophical problems of and raised by the increasing use of simulation to investigate 'real-world' problems. It assumes no background beyond a general interest in science.
If this book has weaknesses they tend to be of the variety that afflicts much modern writing in popular science. Editors rather than authors, perhaps? A sizeable portion of the plates and diagrams add nothing to the book ("gee, John, this is a popular science book - we gotta have some pictures") plus a preference for rushing to mine the next source of the 'gee-whiz' factor rather than grappling with consequences.
The strapline on the cover is 'how simulation is changing the frontiers of science' and I'm not at all sure that this question is answered. The last chapter makes an unconvincing attempt to be a summary, but is in fact the least interesting of the five trailing off rather weakly into a 'what we really need is a theory of complex systems' position.
But, the acid test - did I learn anything from this book? Most definitely, yes. - Ashley Oliver
1 of 2 people found the following review helpful
on 11 October 2011
It always worries me when the cover of a book bears a legend such as 'Critical acclaim for John Casti's previous book', followed by glowing extracts from said 'acclaim'. Personally I'd rather read critical acclaim for the book I'm about to buy, so are they trying to hide something? The clear implication is that this book is going to be every bit as good as the last one, and that all the comments we see here will no doubt apply equally to both. I'm not convinced.
Casti's book is a nominally a tour of computer simulation, focusing on 'the science of surprise' (complexity and emergence), but it's really little more than a shallow survey of other people's work cemented together with some dry philosophising about models and simulations. It's a reasonable idea, but at a little over 200 pages it's clear that there isn't enough meat here to warrant a whole book. It's a highly episodic read, which gives the impression that it's been compiled from shorter works; most annoyingly, though, the sort of fleshing-out that would have made it truly interesting is entirely absent and we are repeatedly directed to the references after only the briefest introduction to a piece of research.
The book reads very like A K Dewdney's 'Computer Recreations' anthologies but with far less substance and a somewhat narrower theme. As an account of the field it purports to cover it offers little that hasn't been done much better in books such as Steven Levy's Artificial Life , James Gleick's Chaos and Roger Lewin's Complexity, to name an arbitrary few.
To borrow a phrase from the anonymous New Scientist correspondent quoted on the cover, I'd suggest that Would-be Worlds is little more than a tour d'horizon.