Writing in the Age of LATEX

878 NOTICES OF THE AMS VOLUME 42, NUMBER 8 Technical writing, and mathematical writing in particular, is a complex and difficult task. While a number of excellent articles have been devoted to mathematical writing, few if any address issues concerning the use of computer technology. The aim of the present article is to draw the reader’s attention to existing works on mathematical writing, to offer some suggestions on writing in general which are tangentially computer-related (such as concerns that arise in LTEX typing but not in plain typing), and to voice some opinions on the proper use of computers in typesetting (not necessarily in this order). LTEX is a powerful tool which can be used creatively or detrimentally, but the young writer who has never encountered or pondered issues of technical writing (e.g., this author five years ago) is far more likely to exploit the detrimental capacity. There is danger of becoming infatuated with the capabilities of LTEX, especially the bewildering variety of fonts and symbols, and neglecting less glamorous but more important aspects of writing such as style, diction, clear communication of ideas, and accurate citation of existing work, not to mention the more subtle aspects of typography such as proper line breaking, spacing, and numbering. In the age of LTEX especially, authors must devote thought to the mechanics of writing and formatting or suffer the consequences of producing literature that is ugly as printed work and is difficult to read, cite, and use. Computer technology has revolutionized the production of mathematical literature. From an author’s point of view, the two most important developments are editing (in the sense of word processing) and typesetting capabilities, including previewing. The inconvenience of typing is evoked by Krantz [7], who describes Stephan Bergmann’s hilarious modus operandi, and Halmos [3, pp. 143–144], whose tribulations simply need not occur today. More importantly, however, computer typesetting has brought the formatting—as opposed to writing—of mathematical papers into the author’s realm of control: LTEX gives the author complete freedom to locate symbols on a printed page. The flipside of this freedom is the responsibility to act competently as a typographical compositor. In order to produce material that is attractive and easy to read, it is helpful to have some idea of the nature of typographical composing. The gist can be gleaned from Halmos [2, 3] and Doob et al. [1], who describe the pre-LTEX typesetting process. Many things have changed since [1], a manual of mathematical style, was written, but most of the advice is quite relevant if the reader substitutes “TEX typist” for “compositor”. Chapters 11 and 12 of Knuth [5] describe more systematically the mechanics of typographical composition and merit perusal. The writings [1, 2, 3] and the articles of Mermin [10, 11; reprinted also in his book Boojums all the way through, Cambridge Univ. Press, 1990, pp. 63–73] raise many important issues involving technical writing. (Mermin is a physicist, and his views are relevant to mathematical writing as well.) The beginning writer who seeks out these works will better appreciate the perils and pitfalls of technical writing and will vividly appreciate the debt owed to Knuth, Lamport, Spivak, and other contributors to TEX and LTEX. To paraphrase Halmos [3], it is the author’s responsibility, not the referee’s, to see that the content of a paper is factually correct. Nonetheless, in order to eliminate ambiguous or incorrect wording, journals might routinely submit accepted articles to copyediting (distinct from refereeing) by someone with mathematical train-