Spacing Printed Text to Isolate Major Phrases Improves Readability

Abstract

Three-liguistically-motivated algorithms for assigning between-word space sizes were compared for their impact on text readability: a computer-implemented heuristic analysis assigned extra spaces between word groups corresponding to major phrases; a phrase-structure analysis assigned each space a size proportional to the depth of the phrase structure at that point; a prosodic analysis assigned space sizes proportional to the between-word pauses indicated if the sentences were spoken; finally, an even-spacing algorithm, assigned a constant amount of space between each word on a line. The readability of the formats were contrasted using the Cook-Chapman find-the-odd-word test in a paragraph version. The readability results showed the following significant ordering of increasing difficulty: heuristic—>phrase-structure=prosodic=even-spaced. The reason that spacing based on the heuristic parser results in better comprehension than based on the complete phrase structure may be that good readers guide their eye movements by a similarly crude initial parse of texts. These results suggest that the readability of text can be improved with the aid of a rudimentary automatic parser.