DEATH IN MIND

When the Virginia opossum feels threatened, she plays dead. After falling to one side, she immediately adopts the bodily and facial expression of a corpse: curled up into something resembling the foetal position, she lies still with her eyes slightly open and the corners of her mouth retracted. Her tongue, which hangs from her open mouth, adopts a blue hue. Her bodily functions are reduced: her breathing and heart rate drop dramatically; so too does her body temperature. She urinates, defecates, and her glands expel a green goo that simulates the smell of rot. In this state, she stops responding to the world. If you came across an opossum playing possum and you didn’t know about her trick beforehand, you would for sure be fooled by it.

The opossum might be aware that she’s playing dead. However, we believe this is unlikely. This behaviour appears to be innate and highly stereotyped, meaning that it is likely automatic and not under cognitive control, as when fear makes our pupils dilate or our hair stand on end. And yet the opossum’s behaviour can tell us something interesting about animal minds—not her own, but those of her predators.

To see this, let’s begin by thinking about a different case that is often discussed in biology: the peacock’s tail. This is a highly disadvantageous trait in some regards: it’s massive, conspicuous, and heavy, thus significantly increasing the peacock’s chances of getting eaten. But it has the one big advantage that peahens find it sexy, and so it makes it likely that the peacock will find a peahen willing to reproduce with him. The peacock need not be aware that his tail is so irresistible, but if we want to give a satisfactory explanation for why this animal has evolved such an otherwise burdensome trait, we need to postulate that in the peahen there is a psychological tendency to find this tail attractive. Thus, the peacock’s tail gives us a window into the mind of the peahen.

Similarly, we believe that the opossum’s death-feigning behaviour gives us a window into the minds of her predators. The opossum’s behaviour is highly complex and made up of several heterogeneous components that have no reason to occur together, apart from the fact that they are all signs that make it more convincing as a display of death. If we assume—which we think is warranted—that the behaviour is automatic and unlearned, we need to postulate a selection pressure that gave shape to it throughout evolutionary history and allowed it to end up as the very impressive corpse disguise it is today. We argue in our BJPS paper that the best available explanation for the opossum’s death-feigning display comes from postulating a concept of death in her deceived predators. This encompasses different ideas, which we shall shortly spell out, but in a nutshell, it means the following: Because predators have a concept of death, this allows them to be deceived by the opossum into believing that she is dead. In turn, feigning death gives the opossum a chance to escape, because her predator might discard her for a fresh kill or reserve her for later consumption.

When we say that the opossum’s death display gives us evidence that her predators have a concept of death, we mean two different things: first, that it gives us evidence that these predators have concepts and, second, that the semantic content (or meaning) of the concept that drove the evolution of the opossum’s behaviour likely corresponds to a minimal concept of death. Let’s unpack this a bit more.

The opossum’s death display is firstly evidence that her predators have concepts or, more specifically, that they manifest five different markers of concept possession. The first of these markers is what we call distance. This refers to the idea that individuals who have concepts don’t merely offer rigid responses to concrete sensory stimuli, but instead can respond in different ways to the same perceptual stimulus or in the same way to different perceptual stimuli. Thus, concepts offer some distance from what our senses deliver.

We have two main reasons for thinking that these predators satisfy the distance requirement. The first is that death-feigning displays encompass signs from different sensory modalities. This suggests that they have been shaped by a representation that integrates perceptual information of various kinds and tracks the idea that dead individuals have a specific look, feel, smell, sound, and so on. Such a representation would have allowed the heterogeneous components of death-feigning displays to cluster together. The second reason has to do with the fact that the opossum’s predators are all generalists, meaning they don’t specialize on any specific kind of prey. When they encounter an opossum playing dead, they might be coming across this animal for the first time, and she will display a very different perceptual pattern to corpses of other species. In order for the predators to be fooled by the death display of an animal they have never seen before, they must be capable of having a representation of death that is independent of concrete sensory stimuli, and that can be applied to animals of unrelated species that appear very different when dead.

The second of these conceptual markers we call inferentiality. This is a capacity to draw inferences from the fact that an individual is classified in a certain way to certain conclusions that follow. When the opossum’s trick works, this will often be because she’s triggering an inferential process in the predator. The triggered inference might be something of the sort ‘if an animal looks, feels, or smells this way, I ought not to eat it’, which would occur when the predator decides that the opossum is not fit for consumption because she appears to be long dead, and the predator’s previous experiences with long-dead animals have taught her that they upset her stomach. Or it might be an inference like ‘if the opossum looks, feels, or smells this way, then she will not escape’, which will occur if the predator does not discard the opossum as a meal, but does relax her attention on the supposed corpse because her past experiences have taught her that dead animals don’t run away.

The third conceptual marker exemplified by these predators is aspectuality. This is the capacity to think about one and the same entity under different aspects. We believe these predators likely possess this capacity because opossums tend to play dead once the predator has already come into contact with them, when fight or flight isn’t an option anymore. If the predator changes her behaviour in that moment (and we know they do this in a non-negligible proportion of times, otherwise this behaviour wouldn’t have offered enough of a survival advantage to have been selected), this is because the predator has switched from representing the opossum under one aspect (alive) to another one (dead).

The fourth conceptual marker is generality. This is the capacity to recombine different ideas in different ways to form different thoughts. We have already hinted at why these predators probably satisfy generality. A predator can only recognize a death display in a newly encountered animal if that predator is able to apply stored information about what to expect from prey animals at different stages of the hunt. This suggests that predators have a concept of being dead that they can combine with representations of different animals, and this allows the opossum to deceive them into thinking she’s one of the dead ones.

The final conceptual marker that one can expect these predators to display is a minimal semantic net. This is the idea that concepts are never held in isolation, but always exist in webs of interrelated concepts. If these predators satisfy the aspectuality requirement, this means that they can think of the same animal as exemplifying different properties, such as ‘edible’, ‘alive’, ‘dead’, ‘injured’. This already points to their possession of a minimal semantic net. But, in addition, many of these properties can only be made sense of in contrast with others. For the opossum to be classified as an animal that-will-no-longer-attack, that-will-not-run-away, that-will-taste-bad, the predator must have some grasp of the contrary notions: prey that-can-attack, that-can-run-away, that-will-taste-good, and so on. This suggests they don’t just have isolated concepts, but webs of interrelated concepts.

In addition to giving us evidence of conceptual abilities in the deceived predators, the opossum’s death display also offers us some clues as to the semantic content of the concept that drove its evolution. In particular, we believe that there are reasons to think that this concept corresponds to a minimal concept of death. We understand a minimal concept of death as composed of two separate ideas: non-functionality (or the idea that dead individuals don’t do things) and irreversibility (or the idea that dead individuals don’t come back to life). The opossum’s death-feigning behaviour appears to incorporate signs of non-functionality with its stillness, reduction of vital functions, absence of sounds, and low temperature. But it also seems to go beyond this, incorporating cues of irreversibility, as though attempting to prove that she’s not asleep, but dead. Thus, she shows signs of putrefaction with her blue tongue and her foul smell; she adopts the specific bodily and facial expression of a corpse; and unlike sleeping individuals, she refrains from responding when picked up off the ground, shaken, or even cut.

We therefore believe that there are strong reasons for thinking that the predators who have attempted to feed on the opossum throughout evolutionary history and whose cognition has shaped her defence behaviour have tended to have conceptual abilities, and that the opossum’s death display targets their minimal concept of death. While this is an interesting story in itself, there are also broader methodological implications to be drawn from it. In particular, we believe that this story exemplifies how the philosophy of animal cognition ought to proceed.

Philosophers have often theorized about animal minds in very abstract terms, posing very broad questions such as ‘do animals possess concepts?’. It seems to us that this method runs the risk of delivering anthropomorphic projections or sweeping generalizations that don’t capture the cognitive, behavioural, and ecological diversity that we see in the animal kingdom. Instead, we believe that philosophy of animal minds needs to make more use of case studies, and focus on what the behaviour of specific species tells us about their particular minds. Thus, the opossum’s death-feigning can teach us something about animal minds, and also something about how we should philosophize.