In: Psychology
This is the final question on a Link word lab in cognitive psychology
Although mental imagery and visual working memory both involve the ability to represent and manipulate visual information, research on the two topics has diverged into two separate literatures that rarely reference one another . Because of the different behavioral measures and tasks used, it has proved challenging to establish the degree of commonality between the two functions.
When participants in visual working memory experiments are asked to describe the strategies they use to complete the memory task, they tend to describe one of two different strategies. One involves creating a mental image to compare with the subsequent test stimuli the other strategy involves picking out particular details of a scene or array and encoding them phonologically or verbally, which is then compared with the test stimuli.
Recent behavioral work supports these subjective reports of different strategies . This behavioral work directly compared the sensory strength of mental imagery and different measures of visual working memory. Individuals with stronger mental imagery had greater precision and higher capacity in visual working memory tasks, but not in iconic or verbal working memory tasks . Furthermore, only those participants with strong sensory imagery were perturbed by the passive presence of uniform background luminance during visual working memory storage, but not in a verbal working memory task. Importantly, the creation of visual mental images is also perturbed by the presence of uniform passive luminance . In addition, in a similar vein to visual imagery, the content of visual working memory can bias perception and can facilitate detection in the neglected hemifield of visual extinction patients .
Taken together, these behavioral data suggest that those with relatively strong mental imagery utilize it to perform a visual working memory task whereas those with weaker imagery tend to rely on nonvisual strategies.
Brain imaging work has demonstrated overlap in the neural representation of visual working memory and mental imagery. For example, in one study , on some trials, participants were required to hold an oriented grating pattern in visual working memory until their memory performance was tested with a probe stimulus; on other trials, the same participants had to form and rotate a mental image of the same grating in accordance with a given cue. BOLD activity patterns in area V1 enabled accurate decoding of which pattern was being held in visual working memory and in the mental rotation (imagery) condition. When the classifier was trained on data from the working memory condition and then applied to decode data from the imagery condition, performance was just as high. This generalization of decoding from memory to imagery is evidence for commonalities in the spatial pattern of BOLD activity during the two tasks. This in turn is evidence for representational overlap between mental imagery and visual working memory. Recent results also show that both visual working memory capacity and imagery strength and precision are associated with the surface size of V1.
The combination of behavioral and brain imaging data shows that, despite clear task differences (‘Hold this visual information in memory and we will subsequently test you on it’ vs ‘Create a mental image of this’), mental imagery and visual working memory can share common neural mechanisms in the sensory cortex. In many tasks, participants have to decide for themselves how best to maximize their memory performance. Depending on the ‘mental tools’ at hand, this might be with mental imagery or a propositional strategy. Recent work suggests that imagery strength and the neural networks underlying imagery may play a role in how individuals perform such tasks .
The key to unlocking the mechanistic relationship between visual imagery and visual working memory may lie in the individual differences across the population in visual representational strength, physiology, and even anatomy . If a subset of the population tends to utilize imagery to aid memory performance, as the evidence suggests, whereas another subset of people who lack strong imagery utilize a different strategy, collapsing across these two groups could induce inconsistencies in visual working memory data. Separating participants into these groups, based the strength of their imagery, may be a good starting point for gaining clarity on the neural machinery used in visual working memory.