Chapter
Five
The theatre
Decades later, this great ‘mental imagery debate’ contin-
ues (Pylyshyn, 2003; Kosslyn, Thompson and Ganis, 2006),
though with some signs of a truce emerging (Pearson and
Kosslyn, 2015). When Shepard and Metzler’s experiments
were first carried out, no one knew where in the brain the
processing was taking place, although there was specula-
tion that the same areas might be used for imagining an
object as for seeing it. With the advent of MRI scans and
other ways of measuring brain activity, it is now clear that
this is correct. When we mentally scan a visual image, simi-
lar areas of the visual cortex are activated as when we look
at a similar object (Cohen et al., 1996; Pearson and Kosslyn,
2015). But learning more about the similarities between
seeing and imagining does not get us very far in adjudicat-
ing between pictorialism and propositionalism as compet-
ing theories of imagery, since both accept these similarities
but draw different conclusions from them.
Both the pictorialist and propositionalist positions have
more recently been challenged by a third way of thinking
about mental imagery: rather than having either pic-
ture-like or language-like images (discrete entities some-
where in the head), in enactivist and sensorimotor theories,
we engage in acts of imagining. These are closely related
to the activity of perceiving the real world (as described in
Chapter 3), but enactivist theories hold that in the case of
imagining, the sensory exploration is performed without
any interaction with the environment. And in the senso-
rimotor framework, even potential for such exploration is enough (Thomas, 2014;
Foglia and O’Regan, 2015). These kinds of theory are supported by evidence
showing how fundamental action is to both seeing and imagining, for example
in a close correspondence between the eye movements we make when we see
and when we imagine (Johansson, Holsanova, and Holmqvist, 2006), and even in
changes to the thickness of the lens as we see or imagine things close up or far
away (Ruggieri and Alfieri, 1992). While pictorialism and propositionalism both
rely on the idea that we see and imagine by means of mental representations of
the things being seen or imagined, theories within this third camp suggest that
action and interaction play much of the role traditionally attributed to represen-
tation (Troscianko, 2014, pp. 86–92).
Despite differing interpretations, findings about the timing and the behaviours
of imagining and their connections to seeing do show that there is something
measurable going on when people have private imaginings – imagery is not
something mysterious and unamenable to scientific study. They do not show
either that consciousness is needed to do the imagining, or that there must be a
mental screen on which the ‘images’ are projected.
First, mental rotations and other manipulations can happen unconsciously, and
indeed do so all the time. When we insert the front door key in the lock, reach
out with an accurate grasp to pick up a cup by its handle, or manoeuvre a car
into a tight parking space, we deal with rotated imagined objects, but we are
FIGURE 5.3 • In Shepard and Metzler’s (1971)
classic experiment, participants
had to decide whether the pairs of
figures showed the same object
rotated or two different objects.
The time they took increased with
the length of time it would take to
rotate real 3D objects.