Magnificent Publications is often asked to create publications that present data clearly and effectively. Over the next few weeks, we’ll be discussing some of the classic guidelines on the subject from a leading authority on data presentation, Edward Tufte.
“Parallelism is a stylistic arrangement in which similar syntactic patterns repeat, thus allowing reader or listener to rely on the grammatical repetition to echo the logical similarity of the thought and thus improving the clarity and efficiency of the passage.” So says the Columbia Guide to Standard American English.
Just as in prose, Tufte writes, parallel constructions in displays of visual information can do this and, in addition, impart forcefulness, rhythm, and balance. Think of before-and-after photos, for example. To create a stronger effect, the “before” and “after” models strike the same pose.
As with prose, however, the user needs to avoid faulty parallelisms, that is, parallel constructions that mix structures: “She likes hiking, swimming, and to ride a bicycle.” In the visual realm, faulty parallelisms misconnect paired images.
As example, Tufte offers the following:
Two unparallelisms—one effective, one inaccurate—appear in these graphs on the major causes of death from 1982 to 1994 among men (left) and women (right) aged 25 to 44. First, the epidemic of HIV-AIDS infection is shown to surge upward against the steady background levels of other causes of death, as the dark solidity of the HIV line contrasts to the dot-dash lines. … There is, however, a regrettable lack of parallelism in the vertical scales of death-rates for men and for women. Equal vertical distances represent different quantities, which makes visual comparisons of slopes (rates of changes) between the two graphs most difficult and also obscures the fact that men in this age group have an overall risk of death more than double that of women.
These faulty parallelisms can often creep into a publication by accident. Often two images will be placed near each other by happenstance, and the creator of the publication won’t notice the implication created in the reader’s mind that they must be related. A careful editor is on the lookout for such problems.
