This came up in a conversation between myself, Aleatory and Pezer, among some others. It was my contention that non-human animals do not fear death, that they are incapable of fearing anything abstract-conceptual and derivative such as would be an understanding of the fact of one's own eventual non-existence. Fear is a response in an organism to certain stimuli, perceived/sensed objects which have, either through learned conditioning or past genetic conditioning, been associated with detrimental effects upon the organism itself (of course it is also the case that once the fear drive-system exists it can 'malfunction' or respond to incorrect stimuli, causing fear where it is otherwise unwarranted or unjustified, but that is an entirely different subject and does not impose meaningfully upon this conversation and its thesis). In the first case learned conditioning takes advantage of an already-existing 'fear drive-system', the ability to pair a perceived stimuli with an emotional-affective response that we call "fear". This is simply a case of classical or operant conditioning, and requires first that a functional affective system able to produce the fear affective response is in existence within the organism. In the second case we see that genetic conditioning has produced this fear drive, through natural selection and based on the clear correlation between possessing such a fear drive and an increased survivability potential, with respect reproductive survival. Organisms which possessed versions of a fear drive, of an affective impelling motivation-cause able to attach to potentially harmful stimuli gained an increased survival potential over organisms not possessing such a system, or possessing a less efficient or functional version of the fear drive.
To fear something, which is to have an affective-emotional reaction that we call "fear", requires only that some stimuli exists which triggers this system into action. No amount of cognitive or conceptual understanding is implied here. However, what we humans typically call fear is already bound up within many conceptual and cognitive associations. This is because our fear becomes an object itself, and is conceptualized and abstracted within the mind. Fear itself operates as a powerful object of consciousness---because of the fact that we are beings able to conceptualize, abstract, derive and cognize we experience the fear drive under these faculties, thus coloring our experience of fear in a manner different than beings without these conceptual-abstracting faculties. When we fear we experience a whole host of associated images, ideas, logical implications and derivations, imagined possibilities, all of these depending for their possibility upon the fact that as humans we employ a complex ability to abstract-extract conditions in the imagination, separating these from immediate sensory perception, and through symbolic-representational langauge make of these sensations an object of consciousness, a highest form of abstraction. In short, we utilize reason. When a non-human animal fears something, however, lacking this power of reason, of abstract consciousness, of imagination divorced from the immediate sensory perceptive, and of symbolic-representational language able to objectify sensations into a new register of experiential meaning, it remains, for the non-human animal, a mere automatic reactivity, an emotional response to an immediately perceived stimuli which has been associated, either through a genetic-physiological structuring preconditioning or a learned classical or operant conditioning, with something threatening or harmful. Once the animal's fear drive engages in the presence of a fear-inducing sensation, the form/s of this sensation, in a more or less generalizable manner, is imprinted upon the fear drive's memory and will produce a similar fear response if encountered in the future.
A non-human animal cannot fear death, because such an animal has no conceptual understanding of its own existence. This is to say, the fact that it exists is not meaningful to it, does not legislate upon its behaviors, thoughts or feelings. Such an animal's thoughts, feelings and behaviors are immediate reflections of current environmental conditions and do not attain to a sufficient distance and objectification/abstraction in the realm of possibility in order that they may become under conscious consideration at a time of their not saliently obtaining presently to the animal from within its environment. This is not to say that such an animal has no memory or capacity to remember/learn/associate, but rather that this capacity is again based in a more or less "simple" stimulus-response mechanism, and, most importantly, such cognitive or affective experiences depend for their possibility upon an immediacy of stimulated-necessitated causation from the environment. Even while memory serves a function of internal environment and thus possible environmental necessitating cause here, this does not raise to the level of abstract possibility or conceptual-objectified form which would allow the animal to react to situations that are merely possible but not actual.
What the human language and conceptual capacity does is create a frame/medium in which such "merely possible but not actual" in effect does become actual, becomes able to enter into the present moment of environmental causation. In order to achieve this feat a new realm of reality must be developed, what we call imagination, which is able to construct objects and stimuli within itself which lack immediate correlate to any present entities in the (non-imaginary) environmental fields. To a lesser extent than imagination, socialization also operates in this regard here, becoming sufficiently "deep" and complex enough to store information in such a way so as to act as a sort of "repository" and temporary storage-house for potentially necessitating causes that are otherwise absent from the general environment. The human, armed with a conceptual-abstract possibility through symbolic-representational language, an imagination strongly able to be divorced from "reality", and a social milieu deep and subtle enough to act as a storage-house of information, now possesses a repitoire of abilities and faculties that grant it new means of experiencing, among many other things, the fear drive and the fear response. Conceptualized objects gain embedded meaning from the conceptual framework in which they are situated, a large influx of associational, conditioning meaning; these objects appear in the imagination and are extracted from immediate environmental necessity; and are rendered as objects under the symbolizing-representing possibility of language. Not only this, but the fear response itself, the affect-emotion, flows through each of these channels as well, drawing from them a certain amount of their essence, qualia and character. Fear is an object, this objects becomes a conceptual category the input into which falls all manner of extra-associational relations and objects, from past, present or future. Because of this, and because the human has the 'ego' of self-identity, which is to say Da-sein, "The being for whom its own being has become an issue," (for more on this, we must get into the structure of consciousness, which we can certainly do. Myself or Parodites could elaborate this more here, I am sure, were we inclined to spend the time) the meaningful understanding of "self" in a manner which is akin to "I exist" having direct impact upon the subjective structure and overt behaviors of the entity as a whole, after a long chain of inference and learned conditioning comes to understand that this "I exist" is conditional and temporary, and will at some point cease. The awareness of death goes hand in hand with the awareness of self. And once this assocation occurs, which we could call the moment of the birth of the self-consciousness in the sense above defined ("I exist" having a direct impact upon the subjective structure and overt behaviors of the entity as a whole), this understanding is, like all understanding, objectified in some manner and entered into the conceptual-linguistic field of meaning-relations. The fear-drive may now engage in response to the presence of this object, since the object's character is such that it clearly represents a potential threat and detriment to the organism itself (i.e. that of its own non-existence).
Here we may appropriately ask why it is that the understanding of one's own eventual non-existence is such to stimulate the fear drive. Here we must remember that the fear drive itself is based in the pre-human animal past, and functions largely automatically, simply causally even where the objects to which it has habituated itself to respond have grown more complex, deep and objectified-abstract. However there is still what seems to be a disconnect between the understanding of death, conceptually speaking, and the fear of it. Here we might affirm that man has been taught to fear death, that perhaps this did not arise "naturally" or at the same moment of the understanding of the fact of death, but some time later. This would be an interesting area of exploration, but falls outside the necessary purview of this topic, since we are making the claim that the non-human animal does not fear its own death because this animal does not possess the conceptual-cognitive-linguistic appartus to make of the fact of death a meaningful object, a possible subject or sensation of consideration/affectation. Thus whether or not man learned to fear death or this fear was immediately aroused at the very moment of his awareness of the fact of death is irrelevant to the question of whether or not the non-human animal fears death.
The issue of the mirror test was raised in defense of the idea that non-human animals fear death. This was done so in order to demonstrate that some non-human animals possess self-awareness, and thus an understanding of their own existence, and thus, finally, an understanding of the fact of the eventual termination of this existence. The logical mis-steps along this path of thinking become obvious here, and I barely think I need explicate them in great detail. Firstly it does not follow that the mere recognition of one's body in a reflective surface, as an object to which one's own behaviors are correlated necessarily, constitutes self-awareness. Properly speaking all that the mirror test can necessarily affirm is that the animal which passes it is self-recognizing. This means the animal possesses a complex enough sensory-conceptual apparatus so as to learn-habituate to the association between reflection of itself and its own behaviors as a consequent of this reflection. What we mean by "self-aware" is precisely that one possesses a direct, meaningful sense and understanding of the fact of one's own existence, that this understanding is such that is impacts directly the entire manner of the being's subjective judgment and behavior. Why is this? Because awareness of self in this manner adds an entire new tier of possible computation and relation to all experience. It sets up an internal reciprocity and reflexivity between sensation and effect, this flow now passing through a derivative and secondary stage of self-application under the form of projected possible interpretation, i.e. imagination of the possible. Self-identity, self-awareness internally cleaves the subject, introducing distance into this subject which allows for the juxtaposition of current stimuli-meaning with non-present imagined stimuli and possible non-actual meaning. This is of course a far cry from the mere fact of self-recognition that occurs when an ape grooms itself in a mirror, or behaves in such a way so as to touch its own body while viewing the reflection of this body in the mirror. These behaviors only require that the ape is able to form a basic relation of connection between the image reflected and the movements of its arm, for instance in the case of grooming. The ape does not "understand" that "I am that reflection in the mirror", rather its cognitive-behavioral and perceptive faculties are able to arrange in such a way so as to form the relation, "arm moves here" based on the combination of the information passed on by the reflected image combined with the proprioceptive sense of the ape's own head and arm, and the relation between them. It is NOT necessary that an ape "know" that "That is me" when looking in a reflection, it merely learns, over time (and this is the case with the mirror test, which gives the animals time to habituate to the mirror, to experiement with it's presence) that, and in an unconscious-instinctive manner, "If I do this, that happens." In the case of the mirror, the arm is able to gain a bit of useful information on how/where to move in order to achieve a certain goal (grooming, whatever), and the relation of this new information afforded by the reflected surface is incorporated into the causation of the arm automatically, instinctively, after the manner of rudimentalry conditioning-learning. There is no reason to, and many reasons not to, make the leap from "ape grooms itself based on the reflection of its image" to "the ape understands, 'That is me' ".
Now, there are great apes which can learn signs in sign language and combine these into novel meaning-strings, "sentences" which they were not taught to do. This would be a better argument for "self-awarness" than the mirror test. But even here, even IF we accept that the mirror test or the use of such sign language meaning-creation is a true sign of self-awareness in the sense which we fully mean it, it does NOT follow that such an animal is able to understand the idea of it's eventual death. A further level of logical derivation and abstraction is required to move from, "I exist" to "I will die". The one implies the other only where a sufficient enough possibility for logical inference and deduction exists. Can we claim that such a thing exists in apes, even great apes? Perhaps, but it remains highly doubtful. And yet, even here, the ape would need to be able to form the meaningful thought-object of death, the IDEA of non-existing, as an object, in order for it to associate the meaning of this with, finally, itself and its "I exist". This is an entirely new level of inference and reasoning which we have no reason to think any animal but the human capable of doing.
Effectively, it is a mere anthropomorphizing and logical error to assume that merely because an animal passes the mirror test it "knows that it exists", has a meaningful-direct understanding of the fact of its own existence as a being and in a way that centrally and impositionally informs and organizes the subjectivity of this being. Equally it is anthropomorphic and irrational to conclude that this animal, in possession of a complex enough sensory-conceptual relay system to form relations between its behaviors and proprioceptive senses and goals through the information afforded by the reflected image in the mirror, somenow now must "understand death", the idea of non-existence, let alone fear this. So much less then to claim that "My dog fears death," which is an entirely unjustified and irrational claim. Animals fear individual stimuli and the more or less generalized forms of these stimuli, they do not fear "ideas" or "abstract concepts". Only man possesses these, only man possesses a meaningful, direct understanding of the fact of his own existence as a separate being, and in a way where this understanding informs directly and powerfully the nature of this very being itself. Likewise, only many employs symbolic-representational language and thus the capacity for abstraction, logical inference, conceptual objectification and unreal imagination. The idea that non-human animals fear death is absurd, or at best, totally unfounded.