OVERALL STYLE & TONE

Avoid the First Person singular. You discover the cosmic schmutz in your petri dish not because you happen to be the researcher, but because the schmutz simply exists in your petri dish as a phenomenon of Truth. Moreover, if you say "I", it makes your work sound like a "My summer camp project was . . .".

Write to the audience. Target your paper towards someone in your field, but not the specific sub-discipline. Explain abbreviations and technical terms, but use the technical terms in the paper (don't use simplistic substitutes).

Beware the weak transition. The most critical thing in any paper you ever write will be your ability to give smooth transitions from each paragraph to the next. Without clear transitions from each paragraph to the next, even the best of paragraphs are but mere oases in a desert of forsaken meaning. (It just breaks your heart, doesn't it?) One of the more common examples of this is:

X is really important because blah blah blah. Therefore, X was studied to blah. Y is really important because blah blah blah. Therefore, Y was studied to blah.

Instead, you might want to try:

X is really important because blah blah blah. Therefore X was studied to blah. It was also critical to study Y, as previous studies have shown that blah blah blah.

Or:

X is really important because blah blah blah. Integrally related to this blah, is Y, which blahs. Therefore, this study measured both X and Y, while Zing across a really big blah.

• Read your paper to yourself, either out loud or subvocalizing. If a sentence doesn't sound like something you'd say out loud, it can probably be written more clearly (usually by making it shorter). If an audience wouldn't immediately understand why you've just said what you've said in a given sentence, then you can probably find a better transition from the sentence before it. If your sentences flow one after the other like a series of falling dominoes, then you're done. (Life rarely grants us the time to be "done" in that sense, but it's a pretty good measuring point.)
• Edit your paper many times, and have other people (especially people with writing or technical writing backgrounds) edit it several times as well.
• Avoid semi-colons unless you're really sure you need them. Before you use a semi-colon, think really seriously about using a period instead. Also, when you're using a colon, bear in mind that you're actually not supposed to capitalize the first word after the colon. You only do that after a period.
• Using longer words and sentence structures simply to sound scientific might not always be the wisest course. There's nothing wrong with short, punchy sentences that get your thoughts across without sounding like a primer for the SAT. All that matters is that your reader understands you. If your sentence spans more than 2 1/2 lines, the reader probably won't understand you.

Back to Top

DEFINING YOUR TERMS

• Use as much jargon as you need and do not define terms that people in the field will know. Model your style along a Science or Nature paper, but with more detail in the materials and methods section than they use.
• Make sure that you define any uncommon term before using it.
• When proving theorems, provide an intuitive explanation of what you plan to prove before diving into the details.

Back to Top

RESULTS & DISCUSSION

• There is often too much summarization of past results. Summaries of past data should be restricted to one or two pages.
• Use "data" correctly ("The data support...", "The data are...")
• Extreme care should be given to the discussion section, the heart of the paper. The importance of the data presented should not be exaggerated; if the paper is truly significant the results will speak for themselves. If the work is not significant the reader will be put off by sweeping conclusions.
• Be sure to properly cite others' work. Acknowledge helpful discussions with your mentor or others in the program. For example, if someone taught you how to use C or Matlab, thank them.

Back to Top

FIGURES / CHARTS / HEADINGS

• Figures and charts should have descriptive captions and be numbered in order.
• Use better figures, especially graphs. Don't clutter up the graph with chart junk, and be sure to label everything. Try to economize on space and combine similar graphs, or put up a representative graph and simply summarize the other results. Don't use a graph when a table will do. An excellent source of more information is a series of books by Edward Tufte of Yale, The Visual Display of Quantitative Information, Envisioning Information, and Visual Explanations.

Back to Top

SPACE: THE FINAL FRONTIER

• My first advice is to be concise. My experience has been that most papers, especially abstracts, are needlessly wordy.
• Be clear and concise. Avoid unnecessary and uninteresting detail. People are usually interested only in the solution that works, not in the twenty-three failed attempts you had to make to find the solution that works. Having other people critique the paper can help to eliminate such problems and can aid you in developing a good flow of ideas.
• Write succinctly. Take out useless words (for example, replace "in order to" with "to"). Sentences should be straightforward and should not be confusing.

Back to Top

HELPFUL BOOKS TO READ

"The Elements of Style", William Strunk Jr. and E. B. White

A classic. Very helpful tips for improving your writing. Not aimed at technical writing, but certainly the tips suggested are universal. There is a companion book "The Elements of Grammar" by Margaret Shertzer, but at 168 pages it's maybe a little too dense to be all that useful except as a reference book.

"Writing Mathematics Well", Leonard Gillman (Mathematical Association of America (c) 1987, 64 pages)

An excellent guide for writing mathematics. Plus, the author has a good sense of humor, so it's fun to read, and at only 64 pages, not very time-consuming. Unfortunately, you are unlikely to find this book at your local bookstore; it is available through the MAA for $10 ($8 for MAA members) at (800) 331-1622.

Back to Top