← Journal18 / 23

Typography

The Rivers Nobody Measured

·6 min in full, in the book·PagePerfect Journal

Also available as a typeset book

In 1970 Gregory and Poulton tested 86 readers on text set two ways, justified and unjustified. For the good readers the setting made no difference, and the poorer readers did worse with justification only in short lines.1 That is about as close as the research comes to the pale channels that open up in a badly justified column.1 Their reputation is far larger: a severe physiological impediment to reading. When we checked the sources, the reputation was the part that did not hold.

The rest of this essay is in the book

It goes on to cover:

  • The slack has to go somewhere
  • Short lines, weaker readers
  • The return sweep will not carry it
  • Hyphenation is there, switched off
  • Boxes, glue and penalties
  • The letters stay as they are
  • Justify with a composer, or set it ragged

About the book

Notes

  1. Gregory and Poulton 1970, Abstract: 86 participants; no difference for good readers; justified worse for poorer readers at seven-word lines, p at or below 0.02; no disadvantage at 12-word lines; explanation in terms of spacing irregularities. One study, read in abstract only; the full paper was not obtained. 86 participants in the first experiment and 32 of the poorer readers retested in the second. The bounds matter more than the result: the effect appears for weaker readers at a short measure and not otherwise. Evidence B: one credible study, or evidence with limits.

Sources

  • Gregory, M. & Poulton, E. C. (1970). Even versus uneven right-hand margins and the rate of comprehension in reading. Ergonomics, 13(4), 427-434. doi.org

Further reading