Feeling unhappy? Try ligatures.

The question of whether applying proper typographic rules really makes text more legible or aesthetically pleasing to anyone other than typography geeks has no doubt been debated to death.

Certain aspects of text presentation, such as line width, leading and anti-aliasing have been shown to cause differences in reading speed and/or comprehension. However, do the more subtle aspects that typographers pay attention to, such as ligatures and kerning (allowing the space occupied by two characters to overlap), really make any difference?

As part of a series of studies, a group of researchers around Microsoft’s Kevin Larson tested the use of advanced typographic features of OpenType (kerning, ligatures, small caps, non-lining numerals, subscript and superscript) against text without these features.

OpenType illustration

The result was that they made no significant difference to reading speed or comprehension, and in fact not even to subjective ratings: about half the people preferred the non-OpenType version of the text.

However, they then went on to determine participants’ affect, or emotional state. One way they did this was by measuring activation of the facial corrugator muscle. Surprisingly, participants turned out to frown less, and could therefore be said to have been “happier”, when reading text with the enhanced typography.

In another test, people were given creative problem solving tasks after they had done the reading. It had previously been shown that performance on these correlates with positive affect, so it was hoped that the outcome would capture aesthetic appeal. Indeed, participants who read text with good typography did perform better on the tests.

These results are interesting in themselves, but proving the merit of good typography wasn’t the study’s only goal. Another main motivation was to find new ways of measuring the effect of aesthetic factors. These are often too subtle to be noticed consciously, and therefore can’t be tested through questionnaires. It looks like measuring facial muscle activation and creative cognitive task performance may be sensitive and reliable enough to do the job.

(These results were presented at the British HCI 2006 conference, but the paper, Measuring the Aesthetics of Reading, is not yet available online. However, you can get a precursory paper that covers part of the work.)