A Post by Max Jones and Tom Schoonen.

If we want to use our imagination to acquire knowledge, it has to be constrained (Kind 2016). After all, if the imagination were completely unconstrained then it would be equally likely to generate truths as falsehoods; possibilities as impossibilities. Some of the constraints on the imagination are within our control. When we’re imagining whether we can visit each other this weekend, it serves us well to constrain our imagination to be reality-oriented. Yet, on other occasions, we might choose to imagine being able to instantly transport ourselves from Amsterdam to Leeds and back. Importantly, these constraints cannot be too tightly under our control: if we always imagine what we intend to imagine, imagination can’t give us anything more than was already part of our intentions (Langland-Hassan 2016).

While the focus in the recent literature has primarily been on ways that we can impose constraints on the imagination, some constraints result from the kinds of creature we are and the bodies we have. To use Nagel’s (1974) classic example, it seems impossible for us to successfully imagine what it’s like to be a bat. The fact that we lack wings and sonar means that no amount of trying will provide us with an imaginative window into bat-like experiences. These embodied constraints are not chosen, they simply result from the kind of body we have.

At first sight, embodied constraints on our imagination seem to limit the extent to which the imagination can provide us with knowledge. Cognitive limitations of imagination raise issues for any account on which imagination is used as a tool for gaining modal knowledge. Any time we are unable to imagine a situation, this inability could either be due to the impossibility of the situation or due to some unchosen constraint on our imagination. With embodied constraints, we might not be in a position to tell which of these is the case. For example, we cannot imagine something being both red and green all over, which might indicate that such a situation is impossible. However, it could equally be a result of our limited embodied perspective. A creature with two visual systems might think otherwise. The problem stems from the fact that we are poor at introspectively identifying embodied constraints on imagination. As such, there is a temptation to downplay the role of the imagination in generating knowledge.

This move may be too hasty, as we are not restricted to understanding embodied constraints from the inside. We can study the mechanisms that support imagination from the outside using the tools of the cognitive sciences. A particularly fruitful avenue for understanding the way in which the body constrains imagination lies in turning to the theory of embodied cognition. In particular, moderate forms of embodied cognition claim that higher cognitive processes -- such as conceptual thought, propositional reasoning, and, importantly, imagination -- involve activation of the same systems that we use to interact with the world through perception and action (Barsalou 2008). If imagination involves reusing perceptual and motor systems for other purposes, then the limitations and constraints on those systems are likely to have a significant impact on the constraints on our imagination.

The idea that the body places constraints on the imagination is not new and has already been explored in the literature on mental imagery. For example, the time it takes to rotate or scan across a mental image depends on the spatial extent of the rotation or scan (see Kosslyn 1980). However, an embodied account would suggest that imagination is based on reuse of perceptual and motor systems regardless of whether it involves any conscious image-like phenomenology. Much of the most interesting evidence in favour of embodied cognition goes beyond cases where offline cognitive processes have a similar phenomenological feel to perceiving and acting. Even cases of purely propositional imagination should be seen as supported by perceptual and motor systems. In such cases, we may experience nothing akin to perceptual phenomenology, yet perceptual and motor systems may still play an important role in both supporting and constraining our imaginative capacities.

Rather than looking towards our own experiences for the consequences of our embodiment, we should look to work in embodied cognitive science that explains the nature of imagination on the basis of its evolution and function. Thankfully, a range of recent work (e.g. Grush 2004; Jeannerod 2006; Pezzullo 2011) offers just the right kind of picture. Our capacity to imagine can be seen as an offshoot of anticipatory mechanisms that are needed for the online control of action. In order to successfully execute goal-directed actions we need to be able to anticipate the sensory consequences of those actions. Once this capacity is in place, we can anticipate the sensory consequences of actions that don’t actually take place. In doing so, we move from predicting the immediate future to imagining mere possibilities. This capacity for ‘offline’ motor anticipation can gradually become more and more detached from the here and now. The same capacity that enables us to pick up a glass can be used to imagine rearranging the furniture in our apartment or lying on a beach in the Bahamas. Yet, despite this freedom, imagination, on this picture, is constrained by the type of mechanisms that support it; sometimes in surprising ways that we could not predict by simply introspecting on our own imaginative episodes. If we want to understand the constraints on the imagination, we need to study the nature of our systems for planning, controlling, and executing action.

So, what does this all mean for the question regarding acquiring knowledge using the imagination? On one hand, it may suggest that the kind of knowledge that we can gain via the imagination is limited. Our perceptual and motor systems evolved to control action given the laws of physics in this world. As such, we may have reason to distrust our imagination in cases that involve too much of a departure from the kind of environments that we tend to act in. On the other hand, since our systems for controlling action evolved to control action in this world, they may operate by implicitly taking into account many aspects of our world to which we lack conscious introspective access. Simulating interactions with the world may bring this implicit content to the surface so that we can come to know of it. We are physical objects, so by imaginatively exploring what we can do in the world, we can come to know about what physical objects can do, by learning about what the world will let us do to it.

References

Barsalou, L. W. (2008). Grounded cognition. Annu. Rev. Psychol., 59, 617-645.

Grush, R. (2004). The emulation theory of representation: Motor control, imagery, and perception. Behavioral and brain sciences, 27(3), 377-396.

Jeannerod, M. (2006). Motor cognition: What actions tell the self. Oxford University Press.

Kind, A. (2016). Imagining under constraints. Knowledge through imagination, 145-159.

Kosslyn, S. M. (1980). Image and mind. Harvard University Press.

Langland-Hassan, P. (2016). On choosing what to imagine. Knowledge through imagination, 61-84.

Nagel, T. (1974). What is it like to be a bat? The philosophical review, 83(4), 435-450.

Pezzulo, G. (2011). Grounding procedural and declarative knowledge in sensorimotor anticipation. Mind & Language, 26(1), 78-114.