Writing Figure Captions
Olivia Wilkins
I just finished reading my students’ perspective articles (I’m teaching Scientific Writing at Caltech for the tenth time!). I love reading them because I always learn something new. Scientific Writing is a chemistry course, and since my research falls more within the realm of astronomy, I rarely (if ever) read about the electrochemistry or chemical biology topics that typically pop up as the subject of the perspectives.
The perspectives are supposed to include at least one figure, preferably one that is accessible to an out-of-field scientist reader. As an astrochemist, I am often that out-of-field scientist reader. The figures are usually great, with schematics that illustrate how cells work or how different systems are interconnected. What they show is generally described well in the body text as well, making it easy to understand what is going on.
Something I’ve noticed over the years is that I give a lot of feedback on writing captions. Thinking back, the only time I ever learned how to write a caption was probably in some social studies class where I was describing a picture in no more than five words. And when you’re captioning a photo of George Washington, just writing “Portrait of George Washington” below the image is perfectly acceptable. For figures in research, however, you need to have more.
It’s good practice to write figure captions that can be understood independently from the body text. Thinking about some of the figures I’ve made, a caption that just says “Abundance versus temperature” doesn’t sufficiently explain a figure that is a histogram with bins of 50 pixels (where each pixel is a region on a map) showing the relationship between molecular abundance and gas temperature. That is, abundance and temperature are what are shown on the axes, but they aren’t what the graph is about. Instead, graphs are typically about some kind of relationship.
Write Like a Chemist (Robinson and Stoller, 2008) includes some nice examples of figures, including figures with common errors and their revised and improved counterparts. In Chapter 16, the book goes over the different components of a figure, including captions.
Okay, this tells us how to format a figure caption, but what actually goes into that caption?
I went to the American Chemical Society (ACS) Guide to Scholarly Communication to see what it says, and it offers a little bit more about what makes for a good caption than Write Like a Chemist.
The advice to be concise is more about what goes in the figure itself than it is about the actual caption, but it is advice that can be carried over to caption writing. You don’t want to have paragraphs of text of a figure; the point of a figure is that it illustrates something more clearly than what you could say in words. It complements the text, but it should not compete with it.
Captions should also be concise. You should have enough information in the caption to allow your reader to know what they are looking at and the types of information they should glean from the plot or image. However, any results or discussion of what’s in that figure should be left to the body text.
Being complete is important. If you use symbols or colors, explain them. If I have a contour map with contours in both blue and red, I need to tell you what those blue and red contours represent.
It’s also important to explain any abbreviations. Chemical formulas or abbreviations that are commonly understood are fine, but anything else should be defined in the caption (as well as in the body text). For instance, if my axes are labeled N and T, I should mention in my figure that I’m showing the relationship between “column density N and rotational temperature T.” If I just write “N and T” or “column density and rotational temperature,” the reader isn’t guaranteed to interpret my axes correctly, especially if they are out of field or if I am using unconventional labels.
A piece of sage advice from the ACS that perfectly summarizes how to approach writing captions is this:
If you have any tips about writing captions, or if you want to share some mistakes you’ve seen, I’d love to read them in the comments!