The notion that a major adaptive function of human cognition is our ability to imagine the future has been rumbling for some time. Ideas about brain processes directed toward future thinking – often called prospection – have emerged from early studies using brain imaging (Ingvar Reference Ingvar1979; Reference Ingvar1985), work on human memory (Tulving Reference Tulving1985; 2005), and theories in social cognition (Gilbert Reference Gilbert2006; Klein et al. Reference Klein, Cosmides, Tooby and Chance2002a). These rumblings have erupted over the past year (for reviews, see Buckner & Carroll Reference Buckner and Carroll2007; Schacter & Addis Reference Schacter and Addis2007a). Recent empirical observations using brain imaging reveal that frontal and medial temporal memory systems participate in envisioning the future in ways that parallel their role in remembering the past (Addis et al. Reference Addis, Wong and Schacter2007; Okuda et al. Reference Okuda, Fujii, Ohtake, Tsukiura, Tanji, Suzuki, Kawashima, Fukuda, Itoh and Yamadori2003; Szpunar et al. Reference Szpunar, Watson and McDermott2007). Characterization of the deficits in patients with brain lesions shows that amnesia is linked to impairments of prospection and imagination (Klein et al. Reference Klein, Cosmides, Tooby and Chance2002a; Hassabis et al. Reference Hassabis, Kumaran, Vann and Maguire2007).
The central idea emerging from this recent flurry of research activity is that humans possess a brain system, or interaction among brain systems, which functions to construct mental simulations that represent possible upcoming situations – a form of “life simulator.” While forms of these abilities exist in other animal species, the capacity for mental exploration of future possibilities is extraordinarily well developed in humans. Planning, social cognition, remembering, moral reasoning, and even daydreaming may depend on these brain systems (Buckner & Carroll Reference Buckner and Carroll2007).
Building upon their earlier seminal work (Suddendorf & Corballis Reference Suddendorf and Corballis1997), the target authors Suddendorf & Corballis (S&C) again return to the question of what foresight is and whether it is unique to humans. This is a well-timed review. Much of its value is to renew attention and revisit some of the central ideas raised in their earlier work. A central message is that, as a field, we should consider the adaptive functions of capacities such as episodic memory (mental time travel) in the context of how they aid decisions and future-oriented behavior. A lingering debate is to what degree other animals possess these capacities that are so well developed in humans. In a recent opinion piece, Buckner and Carroll (Reference Buckner and Carroll2007) concluded that (1) certain animal species such as the scrub jay have adapted to overcome similar challenges that prospection has emerged to benefit in human; (2) prospection and related abilities are more developed in humans than in other animals, perhaps with qualitative differences that emerge from our self-awareness; and (3) other animals possess behavioral and neural patterns that might represent proto-forms of prospection and related abilities. I do not repeat the arguments here.
What I do want to address is whether we are beginning to stall in our progress by only exploring the question at the behavioral level – something that can be considered a weakness (or an incompleteness) in S&C's present argument. They make an explicit separation between their ideas and hypotheses about neural implementation. When introducing the pursuit of their theory of components of mental time travel, they write that no claims are made “about simulation or any other mechanisms by which this may be instantiated in the brain” (sect. 4, para. 2). The target article discusses brain evidence in the section on “neurophysiological evidence” (sect. 2.3), so it is clear that brain data are seen as relevant. Nonetheless, the developed framework proceeds independent of information from studies of the brain.
This orientation has two limitations. The first limitation is that an opportunity is missed by not providing an integrated brain-behavior discussion, because the present theory does not help conceptualize much of the emerging empirical data on the neural implementation of prospection including the important roles of frontal and medial temporal systems. The second, more poignant, limitation is that information about neural architecture will almost certainly help us to understand both how the capabilities are implemented in the human and also to what degree the underlying processes are similar or different across species. The theater metaphor in the target article and reframing of Tulving's models of memory into contemporary future-oriented versions are extremely useful frameworks for generating ideas about the components involved in flexible future-oriented human cognition. But how can we test that these processes really exist? And, relevant to the question of comparative cognition, how are we to explore whether similar mechanisms are at work in other animals? I suspect the answer will come in pairing cognitive and behavioral theories with directed exploration of their implementation within the brain.
Specifically, in the target article a novel framework for understanding component processes involved in prospection is proposed using a theater metaphor. Within the theater metaphor a critical component process is the “stage” – a representational space whereby simulations of the future can be constructed and explored. A proposed detail of this constructional space is that it allows a representation of current reality and also secondary representations that explore future possibilities. This intriguing hypothesis implicitly specifies a mechanism of representation that can be informed by underlying neural implementation.
Neural investigation may be particularly relevant to explore similarities and differences in how information is represented across species. For example, it may be possible to identify neural sequences, in particular in the hippocampus of behaving rats, that “preplay” upcoming decision choices (such as left turns in a maze, versus right turns). One could then ask to what degree such representations are adaptive in rats, how they are similar or different from the kinds of representations observed in humans, and what kinds of limitations arise from any observed differences. Rats are mentioned here as an example because paradigms that probe representation in rats are particularly well developed experimentally (e.g., Ferbinteanu & Shapiro Reference Ferbinteanu and Shapiro2003; Foster & Wilson Reference Foster and Wilson2006). The same approach could be applied to other nonhuman species, barring the great apes. It will be particularly interesting to understand neural implementation of convergent behaviors in distant animal species, such the recent work on scrub jays. I suspect we will find significant differences in implementation. From these differences we may learn about the power of natural selection to find convergent paths to complex neural processes much in the same way wings for flight and fins for swimming have evolved on multiple occasions. Put simply, understanding neural implementation will almost certainly help to clarify the component processes used by humans for prospection and their relations to other species.
