Composition and Deformance: Measuring Imageability with a Text-to-Image Model
This work addresses the challenge of quantifying imageability in linguistics and psychology, offering computational tools for researchers, though it is incremental as it builds on existing text-to-image models.
The authors tackled the problem of measuring imageability in both single words and connected text using a text-to-image model, finding high correlation with human judgments for words and more consistent responses to compositional changes than baselines.
Although psycholinguists and psychologists have long studied the tendency of linguistic strings to evoke mental images in hearers or readers, most computational studies have applied this concept of imageability only to isolated words. Using recent developments in text-to-image generation models, such as DALLE mini, we propose computational methods that use generated images to measure the imageability of both single English words and connected text. We sample text prompts for image generation from three corpora: human-generated image captions, news article sentences, and poem lines. We subject these prompts to different deformances to examine the model's ability to detect changes in imageability caused by compositional change. We find high correlation between the proposed computational measures of imageability and human judgments of individual words. We also find the proposed measures more consistently respond to changes in compositionality than baseline approaches. We discuss possible effects of model training and implications for the study of compositionality in text-to-image models.