Like Chargaff's essay, the critique by Margulis and Sagan, constructed of chapters steeped in history, attacks reductionism. Our authors present ideas that challenge the reader to expand his intellectual horizons in order to include some concepts that in the end, unfortunately, seem downright silly. Margulis states, "a totally preposterous idea requires absolute unflinching faith." Is the Gaia hypothesis preposterous? Perhaps no more so than the idea that variation is stored in nitrogenous bases. But one takes pause at the presumption that reproduction is one of the attributes of the 'living Earth,' just as one might flinch at the conceit of a 'selfish gene.' Let the reader decide.
Although it is radical, the critique of Margulis and Sagan is less unexpected than Chargaff's. Our authors have been writing "essays on Gaia, symbiosis, and evolution" for a long time. The news about metascience, women in science, and funding policies of the National Science Foundation is not good news, but it's not new news. The authors understand that profound scientific ideas are worth fighting for. The pen is mightier than the sword, but why is the writing (and the thinking) in this book so sloppy? Consider this sample from Sagan's essay, "What Narcissus Saw."
That humankind is currently the only tenable midwife for Gaian reproductive expansion is a gauge of our possible evolutionary longevity and importance - provided that the violently phallic technology that promises to carry life starward does not destroy its makers first.
Here is the view from "Futures," co-authored by Margulis and Sagan, who seem to have been carried away with their vision of the technological paradise to come. Have you ever read anything like it?
Traditional printed books will become as extravagant - and as expensive - to people of future as first editions or hand-printed manuscripts seem to us. Books will appear to be immensely laborious undertakings. Each bulky mass of ink-spotted paper will take on the antiquated aspect of the Mainz Bible...
I do my bibliographic work at the Farlow Reference Library at Harvard, and I agree that those "bulky masses" were immensely laborious undertakings. But antiquated? Our authors overlook the possibility that inspiration might be drawn by holding an illuminated manuscript in one's hands.
While on the subject of illustrations, and as a segue to Wainer's Visual Revelations, I am obligated to comment on the figures in Slanted Truths. Let all prospective authors know: Difficult ideas require excellent illustrations! The harder it is to get an idea across, the better the pictures should be. Actually one of the best scientific illustrations I ever saw, a cartoon of the evolution of endosymbiosis, was presented at a conference by Lynn Margulis. As best I remember it, a fuzzy little prokaryotic cell with human-like appendages was riding on a rocket ship, which represented another cell. As the rocket ship-cell went faster (and the atmosphere got thinner), the little rider held on harder and harder. As the speed (and selective pressures?) increased, things got blurry until at last, he was safely inside the rocket ship. The intended message of the evolution of endosymbiosis came across loud and clear. The illustration, while it was perfectly "unscientific," was charming, funny, and took itself none too seriously. Most of the illustrations in Slanted Truths lack these qualities. The photos are indistinct, the half-tones are too gray, and the thematic drawings are heavy-handed. The SEM photomicrograph on page 163 was printed with no editing (the scale bar and other data are off to the far right, partly cut off). Most of these illustrations would be edited out of a professional paper. How have they stayed on as a part of this book?
To her credit, Margulis writes in a challenging and thought-provoking style. Her revealing autobiographical essay, "Sunday with J. Robert Oppenheimer," is a portrait of an uncomfortable ugly duckling who comes of age in a dazzling scientific milieu. "Science Education, USA" says many of the right things about how wrong we are in our system of training scientists. I have put it on reserve, along with Chargaff's essay on reductionism, for my non-major students. However, in all honesty I have to say that I wouldn't reserve a space for this volume in my botany library.
Questions of explanation, understanding, and perception are central to Visual Revelations: Graphical Tales of Fate and Deception from Napoleon Bonaparte to Ross Perot. Howard Wainer addresses what may be the most important type of reduction outside of culinary sauces-reducing information into graphical form. The author, whose tenth book this is, has concocted a bubbling cassoulet of topics that range from preparing overheads to making maps. The book reflects a range of experience and thought that is clearly the product of many years of professional work. Almost every page of the book is taken up by diagrams, but most of them are ugly, poorly constructed, horribly reproduced graphs. I think this goes beyond Wainer's intent in the first chapter, a how-to on displaying data badly. The book appears to be a cross between the marvelous How to Lie with Statistics (Huff, 1982) and the recent spate of sublime books of E. R. Tufte, for example, Envisioning Information (Tufte 1990). Unfortunately, this volume lacks characteristics that make the others "must" reading. The lightness of How to Lie with Statistics is missing here and this volume lacks the clarity of Tufte's work. The text is difficult to wade through. The illustrations look greasy and grimy. The collection of poorly constructed graphs may have been meant in part to be a study in bad taste, but somehow it went too far. The distortion that characterizes so many of the diagrams finds its way too deeply into the book itself. The reader has to dig through the visual mud for some really interesting concepts like the Wainer's "lie factor" (one Washington Post graphic is assigned a whopping 'near record' lie factor of 131,000%)! But what bothers me is that I can't see it and the author hasn't given me a really good handle on how to interpret it. Here is explanation that doesn't provide understanding. How can I find the lie factor in the next paper I review? Maybe the problem is with the illustrations themselves. Perhaps Copernicus Press (the division of Springer-Verlag responsible for this and the Margulis-Sagan volume) needs to perfect its reproduction techniques. Wainer understands human cognition and presents many good discussions of cognition and understanding vis a vis his material. And he has drawn that material from a great variety of sources. But the presentation is so ungainly that the message is lost. - Samuel Hammer, College of General Studies, Boston University.