Fred Dretske. Photo courtesy of Stanford University.

In common parlance, sensing an object entails being aware of it. Dretske (1995), who professes to employ "entirely standard" language, which reflects "how ordinary folk talk about these mental states", writes:

For the purposes of this discussion and in accordance with most dictionaries I regard "conscious" and "aware" as synonyms. Being conscious of a thing (or fact) is being aware of it.

I assume, furthermore, that seeing, hearing, smelling, tasting and feeling are specific forms - sensory forms - of consciousness. Consciousness is the genus; seeing, hearing, and smelling are species (the traditional five sense modalities are not, of course, the only species of consciousness). Seeing is visual awareness... You may not pay much attention to what you see, smell, or hear, but if you see, smell or hear it, you are conscious of it.

Anticipating familiar objections that we can sense things and not know what they are, and sense things without being aware that we are doing so, Dretske (1995) clarifies his position:

So, once again, when I say that if you see, hear, or smell something you must be conscious of it, the "it" refers to what you are aware of (burning toast, a french horn), not what it is you are aware of or that you are aware of it. To be conscious of an F is not the same as being conscious that it is an F and certainly not the same as being conscious that one is conscious of an F. Animals (not to mention human infants) are presumably aware of a great many things (they see, smell, and feel the things around them). Nonetheless, without the concept of awareness, and without concepts for most of the things they are aware of, they are not aware of what they are aware of nor that they are aware of it...

Let an animal - a gazelle, say - who is aware of prowling lions - where they are and what they are doing - compete with one who is not and the outcome is predictable. The one who is conscious will win hands down. Reproductive prospects, needless to say, are greatly enhanced by being able to see and smell predators. That, surely, is an evolutionary answer to questions about the benefits of creature consciousness. Take away perception - as you do, when you remove consciousness - and you are left with a vegetable. You are left with an eatee, not an eater. That is why the eaters of the world (most of them anyway) are conscious.

On methodological grounds alone, Dretske's common-sensical assertion that consciousness must have a practical function is to be preferred to the oft-heard suggestion that it is epiphenomenal. The hypothesis that an organismic trait has a biological value is more fruitful for scientific investigation than the hypothesis that it has none. However, there are a number of practical reasons why consciousness might have evolved. Dretske's suggestion that it arose because it enabled animals to sense predators is not entirely convincing. Being able to sense predators is of little use unless you can avoid them - e.g. by running away. Most animals are very good at doing this, at short notice. This is one thing that differentiates them from "vegetables". The ability to react rapidly to sudden, unexpected changes in one's environment is another possible reason for the evolution of consciousness.

Additionally, Dretske's claim that animals possess "perception" while vegetables lack it is empirically flawed: as we have seen, even bacteria have sensory capacities and (as we shall see) so do plants. From this, one could either conclude that all organisms (with the exception of viruses) have conscious mental states, or that Dretske is wrong in linking sensations to mental states.

Finally, Dretske's appeal to "how ordinary folk talk" creates problems for his argument that sensing an object entails being conscious of it: in fact, we often speak of artifacts (e.g. motion detectors, heat detectors) as being able to sense objects, yet we do not credit them with awareness or consciousness. Of course, Dretske could claim that for organisms (which are profoundly different from non-living artifacts), sensing an object is a way of being conscious of it, but such a position requires argumentative justification.

Daniel Dennett. Photo courtesy of University of California.

Writing from an opposing standpoint, Dennett (1997) distinguishes between "sensitivity" and "sentience", but his application of this distinction is questionable. He allows that the sensitivity exhibited by bacteria can be described using the intentional stance, whereby each sensor is regarded as a micro-agent. However, he prefers not to call this sensitivity "sentience", principally because animals possess body-maintenance systems with similar functions, that operate even when they are asleep or comatose (1997, pp. 87-88). There is, however, a disanalogy here. Even the humblest bacterium is capable of performing the activities required in order to protect itself, flourish and reproduce, in accordance with its telos - such as seeking food, avoiding harmful stimuli, and exchanging chemical messages with other individuals - whereas an animal is incapable of performing even basic actions such as feeding or reproducing while asleep, and can only react to a limited range of stimuli. Dennett's exclusion of bacteria from the domain of sentient beings requires further argumentative support.

Aristotle.

Our modern distinction between sensations and perceptions reflects a post-Cartesian standpoint, with its own philosophical biases; so it may be instructive to step back and examine what Aristotle wrote on this subject. Aristotle was untroubled by the modern question of whether the possession of sensory capacities is a sufficient condition for having "mental states", because he did not have a single word that covered perception, emotion, imagination, memory, reason and choice, as our term "mind" does, while excluding nutrition and growth. However, there can be little doubt that he envisaged aisthesis in mentalistic terms, as shown by his remarks (De Anima 3.9, 432a16) that both perceiving (aisthesis) and thought belong to a faculty of discernment (krinein), and that since all animals possess some form of perception (i.e. touch), they must also be capable of pleasure, pain and desire (De Anima 2.3, 414b1ff).

Using the two intentional stances that were distinguished above, we can see that Aristotle has conflated two distinct meanings within the term aisthesis: a capacity to discriminate between stimuli (which can be described in terms of a mind-neutral third-person intentional stance), and the subjective impression of a stimulus (which can only be described in terms of a first-person intentional stance). His two definitions of anger (De Anima 1.1, 403a - line number!) as a desire for revenge (first person) and a boiling of blood around the heart (third person), suggest that he regarded sensations and perceptions as the same thing, described from the outside (third-person stance) and inside (first-person stance). However, the very possibility of using a mind-neutral intentional stance to describe an organism's sensory behaviour shows that the first- and third-person accounts do not have to accompany one another: the former may be redundant. An organism's ability to discriminate between beneficial and harmful stimuli, which enables it to move towards the former and away from the latter, can be described using a mind-neutral intentional stance, which explains the organism's behaviour in terms of its information about its built-in goals. There is no need to invoke subjective states (such as pleasure or pain), let alone desires. Only if aisthesis is taken as a subjective impression can we ask whether a stimulus is experienced as pleasant or painful (words used to describe "feelings"), and whether the organism desires it or not.

In another famous passage (De Anima 3.2, 425b11-25), Aristotle also argued that since we perceive that we are seeing, it must either be by sight that we do so, or some other sense. Since the latter involves us in a potential infinite regress, Aristotle suggested that sight is not a simple sense, but comprises both seeing and perceiving that one is seeing - which suggests that he envisaged the activity of sensing as some kind of subjective state. He does not seem to have addressed the issue of whether there could be organisms that can see but cannot see that they are seeing, which is what is being proposed here for bacteria. Additionally, Aristotle's use of the term "perception" to describe my being aware that I am seeing obscures the fact that this "perception" is a high-level cognitive state - perhaps "realising" would be a better word - which requires that I have a self-concept, as well as a concept of my cognitive states (as distinguished from their objects). It is debatable whether any non-human animal (let alone a bacterium) is capable of such feats.

The upshot of this discussion is that sensation and perception are indeed distinct, because we can describe an organism's sensory capacities using mind-neutral terminology: a third-person intentional stance (or, alternatively, a goal-centred stance) is adequate for the task. Following the methodology that I proposed above, the only valid reason for preferring mentalistic terminology here would be that it allows us to better explain the behaviour of organisms with sensory capacities - i.e. to make new predictions that a neutral account could not make. If we can explain and predict the repertoire of a sensitive organism's behaviour just as well by using neutral terminology, then it serves no scientific or philosophical purpose to introduce mentalistic terminology and ascribe perceptions to it.

We can now formulate another conclusion regarding the conditions for ascribing mental states to organisms:

S.5 The possession by an organism of sensors which encode information about its surroundings is an insufficient warrant for saying that the organism is capable of cognitive mental states.

In organisms, sensors are thus a necessary (Conclusion N.2) but not sufficient condition for having mental states.

Despite bacteria's impressive sensory abilities, we do not have to resort to mentalistic terminology in order to account for their capacity to encode and respond to information about their environment. We have no grounds for supposing that they have perceptions - unless there turn out to be other aspects of their behaviour that are more appropriately explained in mentalistic terms.