A major impediment to writing physics gracefully comes from the need to embed in the prose many large pieces of raw mathematics. Nothing in freshman composition courses prepares us for the literary problems raised by the use of displayed equations. Our knowledge is acquired implicitly by reading textbooks and articles, most of whose authors have also given the problem no thought. When I was a graduate teaching assistant in a physics course for nonscientists, I was struck by the exceptional clumsiness with which extremely literate students who lacked the exposure even to such dubious examples treated mathematics in their term papers. The equations stood out like droppings on a well-manicured lawn. They were invariably introduced by the word “equation,” as in “Pondering the problem of motion, Newton came to the realization that the key lay in the equation

F = ma”.

To these innocents equations were objects, gingerly to be pointed at or poked, not inseparably integrated into the surrounding prose.

Clearly people are not born knowing how to write mathematics. The implicit tradition that has taught us what we do know contains both good strands and bad. One of my defects of character being a preference for form over substance, I have worried about this over the years, collecting principles that ought to govern the marriage of equations to readable prose. I present a few of them here, emphasizing that the list makes no claim to be complete. We are constantly assaulted by so many egregious violations of even these simple precepts that I offer them in the hope that a few sinners—not only writers, but copy editors, publishers of journals, and even the authors of the mathematics subsections of literary style manuals—may read them and repent the error of their ways, or even be inspired to further beneficial studies of the sadly neglected field of mathematico-grammatics.

Rule 1 (Fisher's rule). This rule, named after the savant who reprimanded me for abusing it when I was young and foolish, simply enjoins one to number all displayed equations. The most common violation of Fisher's rule is the misguided practice of numbering only those displayed equations to which the text subsequently refers back. I call this heresy Occam's rule.