Drawing on principles of perception: The line plot

This paper draws attention to a display type that outperforms many competitors in the face of complexity: the line plot. Though familiar to most (if not all) linguists, this graph type is commonly associated with time series data. The true potential of line plots arguably unfolds in settings with multiple discrete (categorical) predictors, however. In such analytic settings, which are often encountered in corpus-based work, line plots are currently outnumbered by bar charts in the research literature. Our aim is to draw attention to this alternative tool and discuss its added value. Drawing on theoretical and empirical insights into visual perception, design recommendations are set forth, ranging from the choice of visual attributes to the arrangement of variables in the display. Their application to corpus data is demonstrated by redesigning selected graphs from the literature. While the drawbacks of this visual form deserve careful consideration, this paper argues for a more widespread usage of line plots for the communication of (corpus) linguistic data.