The first part consists of puzzling and amusing phenomena in soft condensed matter (polymers, surfactants, liquid crystals, wetting, bubbles and surfaces), explained verbally and pictorially (no Feynman diagrams!), and demanding no more than a basic understanding of atoms, molecules and electricity to follow. It is good, after its kind --- i.e. the examples are genuinely interesting, the explanations correct and free of condescension and ``then a miracle happens'', the experiments described are both simple and clever (``the Benjamin Franklin spirit'', as he puts it) and they genuinely lead to larger issues --- but not outstanding. The second part is made up of de Gennes's reflections on the trials and rewards of a life in scientific research, illustrated by anecdotes of his own career. His considerations on when to switch fields are particularly interesting, not least because he's done it at least three times.
The third, and to my mind the most interesting part, contains de Gennes's views on science education in France (dismal), and what should be done to improve it (lots): ``When I welcome a freshman class at our Institute of Physics and Chemistry, I insist on concepts that most math majors are not at all familiar with.... This `reeducation' phase takes a minimum of two years. I am constantly astounded by how little common sense recently degreed engineers have.'' If de Gennes is to be believed, it is perfectly possible to obtain a physics degree in France having only the barest acquaintance with the inside of a laboratory, and no idea of how to do simple quantitative estimates or even Fermi problems --- but be letter-perfect in the properties of self-adjoint matrices. (His story of a Polytechnic graduate who, when faced with a perfectly straight-forward back-of-the-envelope problem, was reduced to crying, ``But, sir, what Hamiltonian should I diagonalize?'' is one I will use on my own students.) De Gennes blames the lingering influence of Auguste Comte, and, much more plausibly, the excessive importance of entrance examinations (which test, essentially, math), the conservatism and insularity of academic departments, especially at universities, and the unwillingness of French students (particularly those who have gotten into the better schools) to exert themselves or sully their hands. Prior to reading Fragile Objects, I hadn't thought that it was possible to do much worse at teaching than American physics departments; I am happy to see I was wrong.
Having won recognition for an exceedingly specialized contribution to human knowledge, the true content of which it takes a good many pages to explain, it was only natural that de Gennes's views be solicited on all manner of subjects, like the environment and ethics, and some of that material has made its way into Fragile Objects. These passages are entirely predictable and unoriginal, but he doesn't claim anything more for them, and they are mercifully short. There is some repetition between different essays, but not so much as to aggravate. The biggest error I found was the confusion, on p. 115, of The Romance of the Rose, a 13th century allegorical love poem, with The Name of the Rose, a 20th century allegorical detective novel.
This will make a fine book for libraries, particularly those used by high school students, but it's hard to see why anyone not obsessed by either Nobel Laureates or by soft condensed matter physics would buy this book, at least in hardcover.