The audiobook experience differs radically from traditional silent reading in several respects, one of them being mental imagery. While mental imagery in silent reading taps equally into the referential (story content; all sensory modalities) and verbal (narrative qua discourse; auditory and kinesthetic) domains, listening activates the auditory circuitry to a degree that leaves little need or capacity for verbal mental images. In traditional silent reading, continuous shifts between referential and verbal imaging coincide with the general distribution of one's attention between story and discourse. But how is attention distributed in narrative audiobook processing? On the one hand, one might speculate that the medium, liberating the listener from the visually taxing burden of reading, allows richer possibilities for referential imaging and a deeper focus on story content. On the other hand, the overt voicing of the discourse can make people attentive to linguistic qualities that otherwise would go unnoticed. Moreover, people use audiobooks widely while performing physical tasks or during transportation, thus receiving a flux of sensorimotor stimuli from the environment concurrent with the spoken narrative. This may interfere with, rather than encourage, referential imagery, or even any deeper focus whatsoever. In my talk I will try to systematize and reconcile some of these contradictions from an aesthetic and psychological perspective. Reviewing extant comparative data on literary listening and reading, I will thus elaborate on recent attempts to define the phenomenology of the audiobook as a specific medium. I will support my conclusions with anecdotal evidence from my own experience of listening to vs. reading Ernest Hemingway's The Garden of Eden, a novel high in referential imageability as well as linguistic craftsmanship.