Historical studies of objectively measured differences in mental imagery ability
This page was created to curate "forgotten" scientific articles on the objective measurement of mental imagery differences. It will be updated casually as I find more relevant articles. If you have access to an article that is not shown here and would like to share it, please send the pdf to email@example.com.
The following three papers appeared in the British Journal of Psychology in 1927, forming a symposium on "the relevance of visual imagery to the process of thinking". Long before the Kosslyn/Pylyshyn debate about whether mental representations are depictive vs. descriptive, there was quite some insight that this depended on the individual*: T. H. Pear clearly had dominantly depictive representations, F. Aveling clearly had dominantly descriptive representations, and F. C. Bartlett proposed that most people (including himself) probably use mixed or alternating representations as task demands or preference dictates. This symposium, and the research that contributed to it, formed the theoretical basis for the objective measurement of individual differences in mental imagery.
T. H. Pear
British Journal of Psychology, 1927
Summary: Pear proposes that any experiment of imagery ability needs to take certain points into account: that there are many dimensions of imagery; that it can refer to experiences at different levels of consciousness (e.g., dream imagery); that people may have different mental representation strategies that are not related to imagery ability; that it is important to consider whether imagery is required or useful for a given task; and that many tasks can be performed without imagery, even if the subject is a visualizer. Particularly, many scientists attempting to measure imagery in that era used tasks requiring manipulation of - or memory for - geometric patterns. These, the author pointed out, are quite abstract, and subjects may use abstract strategies to perform them even if instructed to activate a mental image. Voluntary vs. involuntary imagery is briefly touched on.
British Journal of Psychology, 1927
Summary: Aveling states that introspection is necessary, but not sufficient, to elucidate the functional significance of mental imagery, and calls for the use of objective measurements in determining the "presence" or "absence" of mental pictures. He points out that the vividness of a mental image is judged compared to an original experience, which is itself, in that moment, now a thought and not a percept. One study that required subjects to describe and draw their mental images, found that descriptions did not often match drawings, even in visualizers - this led the author to conclude that imagery is not needed to draw objects from memory. Aveling concludes that if one sets out to measure mental imagery, they should design a task that differentiates conditions in which the subject experiences imagery vs. no imagery.
F. C. Bartlett
British Journal of Psychology, 1927
Summary: Bartlett proposes that all thoughts are expressed symbolically, and mental imagery is a type of symbol that can be used in thought. He also points out that a sense of "knowing" something does not require any symbol or image. He refers to 15 years of research he has conducted on this topic; although no study is specifically referenced there are a few from this time that seem relevant, including: Bartlett, 1916 (click for a pdf) and Bartlett 1921 (click for a pdf), presenting a series of experiments in which he asked subjects to introspect about the clarity of percepts and mental images using geometric stimuli of different complexity (potentially the source of Pear's criticisms).
*This symposium was already part of a second wave of discussions on individual differences in mental imagery -- the first wave goes at least as far back as 1880 with Francis Galton's article Statistics of Mental Imagery (click for pdf), in which 100 men reported on the clarity of their mental imagery of their breakfast tables. In addition to providing insight into the lavish meals these men enjoyed, this article aptly summarizes a breadth of individual differences in mental imagery abilities: some reported imagery as clear as real life that could be manipulated and projected into the external environment; others reported only abstract thoughts about what they had for breakfast. Galton briefly discusses the possibility that imagery ability can diminish over the lifespan due to lack of use in favor of more efficient (abstract) mental representation formats.
Gary E. Schwartz
American Scientist, 1975
Summary: Briefly reviews studies that investigate the association between cognitive events (including mental imagery) and physiological responses, including: heart rate, EEG, and facial EMG.
Golla, Hutton, & Walter
Journal of Mental Science, 1943
Summary: Uses a combination of respiration and alpha-EEG to objectively measure differences between "M" (reported visual imagery), "P" (reported verbal and kinaesthetic imagery), and "R" (mixed use of imagery types) groups. The M group showed more regular breathing compared to the P group. Furthermore, the M group showed "less energy" in resting occipital alpha signals compared to the P group, who showed high-energy occipital alpha across rest (both eyes open and shut) and task (mental arithmetic) conditions. The R group showed low-energy alpha at rest with eyes-open and high-energy alpha with eyes-shut, with a diminution of alpha activity during the task.
P. L. Short
British Journal of Psychology, 1953
Summary: Uses a combination of respiration and alpha-EEG to objectively measure differences between visualizers and verbalizers. Visualizers showed regular breathing and "blocking" of alpha rhythms, while verbalizers showed irregular breathing and persistent alpha rhythms during task performance. Individuals rarely reported using a combined verbal-sensory strategy, most preferring one or the other.
Alan D. Price
Journal of Research in Personality, 1975
Summary: Visual imagery tasks elicit more regular heart rate and shorter respiratory cycles compared to non-visual imagery tasks. Visualizers have more regular heart rate and respiration compared to non-visualizers during a visual imagery task.
Schwartz, Fair, Salt, Mandel, & Klerman
Psychosomatic Medicine, 1976
Summary: Investigated facial muscle activity in response to imagined emotions (happy, sad) in depressed and non-depressed individuals. The authors found an attenuation of facial EMG activity for happy emotional imagery in depressed patients.
Golla & Anotonovitch
Summary: Found differences in breathing regularity in individuals who use auditory imagery (verbal rehearsal), visual imagery, or a combination of the two during mental arithmetic. Subjective aspects of imagery supplement these findings. Irregular breathing is associated with dominantly verbal imagery, and regular breathing is associated with dominantly visual imagery.
Chowdhury & Vernon
British Journal of Psychology, 1964
Summary: Split individuals by more than two subtypes, including visual (passive picturing of static objects/scenes) and auditory imagery (including speech but also sounds and music), as well as "visual-active" (visualizing a dynamic scene), verbal (speech-imagery only, exclusive of kinaesthetic imagery), and kinaesthetic imagery (including eye and muscle movements of the mouth, throat, tongue, hands, etc.). They found more regular breathing among individuals with dominantly visual-passive imagery, and more irregular breathing among individuals with dominantly kinaesthetic imagery. There was no difference in breathing regularity between visual-active vs. verbal imagers (even though these two groups showed almost perfect negative correlations on various introspective and behavioral tasks). More regular breathing in low-auditory imagers did not reach significance.
Perceptual and Motor Skills, 1969
Summary: Measured dominant direction of gaze while thinking/problem solving and found that individuals with greater leftward bias in gaze direction show higher clarity of mental imagery and susceptibility to hypnosis compared to those who showed a rightward bias (although results are modest).
Psychological Bulletin, 1932
Summary: Reports on a study by E.R. Jaensch who found that children with eidetic imagery show an "eidetic pupillary response", in that imagining light stimuli constricts the pupils and imagining dark stimuli dilates the pupils. However, at the time of writing I cannot find this study (either in its original German or translated), and this effect could not be replicated in high vs. low imagers (Paivio & Simpson 1968, click for pdf); Paivio (1973; click for pdf) suggests imagined stimuli can only dilate and not constrict pupils, although he admits that they never managed to test eidetic imagers.
David F. Marks
Perception and Psychophysics, 1973
Summary: Found that rate of eye movements while recalling pictures is lower during imagery than perception, especially in vivid visualizers compared to weak visualizers. The main finding was allegedly replicated a couple of times, but could only be found reproduced in a book chapter (in The Function and Nature of Imagery, Ed. P. W. Sheehan, 1972, pp. 100-104; click here for pdf). In context, this is one significant finding among dozens of other (non-significant) tests in this study. There was no difference in imagined scene scanning between the two groups.
Barbara B. Brown
Summary: Found a relationship between self-reported trait imagery vividness and both rate and amplitude of eye movements while imagining a metronome moving back and forth. Visualizers showed more eye movements and higher amplitude eye movements compared to non-visualizers. Eye movements were independent of reported contents of imagery, and similar group differences were also found during mental arithmetic and instructed visualization of concrete objects.
W. Grey Walter
The Living Brain, 1953
Summary: Describes Alpha EEG differences between visualizers and non-visualizers: non-visualizers belong to the P type, for "persistent alpha", which means their alpha levels do not fluctuate with opening or closing the eyes or performing different tasks; visualizers belong to the M type, for "minus alpha", and they show the opposite effect - that is, low alpha regardless of condition. Grey Walter proposed that these differences are robust, but these extremes (P and M types) are rare. He briefly describes that for individuals of the P type, their mind's eye is almost blind, as if they have only a feeling, when asked to imagine a visual scenario.
Kenneth H. Slatter
Electroencephalography and Clinical Neurophysiology, 1960
Summary: Both task-based and resting EEG alpha is associated with individual differences in "habitual" use of mental imagery and mental imagery effort. Specifically, low-amplitude resting and task-associated alpha rhythms occur in vivid and habitual visualizers, whereas high-amplitude alpha rhythms are associated with low and effortful imagery (verbalizers). These associations are not as clear in individuals who do not show a strong visual or non-visual preference.
Gale, Morris & Lucas
British Journal of Psychology, 1972
Summary: Vivid imagers have a higher mean dominant occipital alpha frequency compared to weak imagers (but not medium imagers) during rest. This difference is attenuated (both show lower frequency occipital alpha) when instructed to produce voluntary imagery, or when presented with high-imageable versus low-imageable words.
Simpson, Paivio, & Rogers
Psychonomic Science, 1967
Summary: High imagers (measured by a behavioral task, a strategy questionnaire, and an imagery questionnaire) have generally higher mean occipital alpha amplitude compared to low imagers during rest, which is opposite to earlier findings by Golla, Hutton, & Walter (1943; click for pdf), although this could be due to procedural differences and results were determined to be inconclusive.
The American Journal of Psychology, 1958
Summary: Could not replicate the 3 imagery types reported by Golla, Hutton, & Walter (1943), although the author admits subjects could be using non-imagery strategies for the task (in fact, it was a spatial manipulation task of combining shapes into more complex forms). He found more compelling evidence from two case studies of individuals with self-professed vivid imagery (and from their reports, could even be classified as having hyperphantasia). In both cases, lower occipital alpha rhythms corresponded to more vivid images within-subject.
John R. Suler
Journal of Mental Imagery, 1985
Summary: Measured electrodermal activity in high-visualizers and low-visualizers during rest (mind wandering) and during a free-associative task in which participants were asked to make associations to presented words, focusing on either pictures or other words/phrases that came to mind. High-visualizers showed higher arousal in both mind wandering and free associative conditions compared to low-visualizers.