Some Methodological and Conceptual Considerations in Studies of Auditory Imagery

ABSTRACT

Methodological and conceptual issues in the study of auditory imagery are discussed, and these issues include the use of different measures in experiments on auditory imagery (questionnaires, behavioral tasks, physiological recordings), proliferation and ambiguity of terminology associated with auditory imagery (multiple terms for the same construct, failure to define many constructs), and distinctions regarding different aspects of auditory imagery (vividness, clarity, and control; inner ear and inner voice; format and activity; voluntary imagery and involuntary imagery; auditory information and nonauditory information). Caution in generalizing between different potential types, components, mechanisms, and processes of auditory imagery (e.g., inferring properties of involuntary auditory hallucinations based on characteristics of voluntary auditory imagery) is suggested. Whether findings previously attributed to auditory imagery necessarily resulted from auditory imagery or could have resulted from some other representational format is discussed, and ways to address this representational ambiguity are proposed (e.g., obtaining convergent evidence of imagery generation and use, reclassifying participants into imagery or non-imagery conditions based on debriefing). Parallels with analogous issues in visual imagery literature are noted. These issues are relevant to future studies of auditory imagery and also have implications for other topics within psychological science (e.g., motor theories of perception, working memory, embodied cognition).

KEYWORDS:

• Auditory imagery
• inner speech
• earworms
• inner ear and inner voice
• methodology
• representational ambiguity

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. This possibility was suggested by an anonymous reviewer, and it assumes that most participants would agree on what the desired or expected response would be. However, it is possible in principle that different participants might believe different outcomes are desired or expected, and such a case would lead to a relatively larger range and larger variance. Relatedly, the presence of significant individual differences in auditory imagery might also lead to a larger range or larger variance even in the absence of any demand characteristics.

2. Although not referring to auditory imagery per se, Keller (2012, p. 206) suggested a definition of musical imagery as “a multimodal process by which an individual generates the mental experience of auditory features of musical sounds, and/or visual, proprioceptive, kinesthetic, and tactile properties of music-related movements, that are not (or not yet) necessarily present in the physical world.” Importantly, such a definition does explicitly include auditory (and other sensory) features.

3. Kosslyn (1976) asked one group of participants to form an image of an object (e.g., “cat”) and then answer a question based on that image, and a second group of participants were asked to think about an object (with no reference made to imagery). If participants based their answer on inspection of an image, then response time was based on the size and ease of seeing the queried feature in the image (e.g., “does a cat have a head?” was answered more quickly than “does a cat have claws?”). However, if participants were just instructed to think about an object, then response time was based on the strength of semantic association (e.g., “does a cat have claws?” was answered more quickly than “does a cat have a head?”). Such a dissociation provides evidence consistent with the use of imagery in the former condition and a lack of imagery in the latter condition.

4. Subsequent research has suggested that speech is actually not necessary in order to create an apparent irrelevant speech effect (e.g., Jones & Macken, 1993) and that the irrelevant speech effect might be more accurately characterized as an irrelevant sound effect (Beaman & Jones, 1997). However, the irrelevant sound effect and articulatory suppression are not necessarily equivalent (Larsen & Baddeley, 2003; Page & Norris, 2003), and so the importance and specificity of the effects of irrelevant speech and irrelevant sound are not entirely clear.

5. Such an embodied or grounded cognition approach would be consistent with arguments in philosophy of mind that humans cannot imagine the sensory experiences of animals with different sensory systems than ours (e.g., what is it like to be a bat, Nagel, 1974).

Additional information

Funding

This work was not supported by any funding agency or grant program.