What we can know through perception depends on what we get in perception. Barry Stroud1 and Fred Dretske2 espouse a view of perception that allows us to get much more in perception than has traditionally been thought possible. I will call their view “propositional perception” or PP for short. I call it this because what we see “is described in a sentence in which the compliment of the perceptual verb ‘see’ [for example] is a sentence with a truth-value, not a singular term referring to an object” (Stroud, 93). Many skeptical problems can be reduced by this new understanding of perception. But it also seems to imply that we can perceive the future. In this paper, I will show how PP comes to this result. I will then offer a possible response that Stroud or Dretske might give in order to avoid this result. Finally, I will offer some criticisms of this possible response. Thus, I will show that PP is not (at least now) able to avoid saying that we can perceive future events before they happen.
PP was developed in response to what can be called the restricted view of perception. Accounts involving sense data or mere appearances fall under this description. The idea is that strictly speaking, in experience we only get how things appear to be and that we need to add some other assuring element to this appearance in order to have knowledge of how things are. This view of perception led to major skeptical problems because nothing seemed to be able to offer enough assurance to such bare appearances. It could be, in any case, that we are presented with the exact same appearance that we now perceive and yet the world does not in fact match that appearance. Instead, PP holds that what we get in perception is much more than just an appearance of what may or not actually be the case.
The main insight of PP is that there is more going on in perception than just passively receiving appearances. PP holds that in perception we judge our experience to be the instantiation of some concept or other. For example, Dretske traces our difficulty in answering the problem of other minds to the fact that a person’s mental states are not able to be captured by this ‘appearance-only,’ view of perception. We run into a problem when we emphasize our inability to see someone’s pain. Instead, he thinks we should say that we see that someone is in pain. He thinks that with this description, we can better understand that we can directly perceive that someone is in a certain mental state. Similarly, Stroud writes that we can see (and thereby know) things “by perception alone” (92). These claims do not mean that our beliefs and knowledge do not play a role. Quite the opposite, Stroud and Dretske think that the major skeptical problems that have plagued philosophy for the past several centuries come about precisely by denying that beliefs are part of perception. The traditional view of perception is restricted in that it does not include our faculty for recognizing concepts in our experience.
PP also allows that we can come to incorporate more beliefs and knowledge into perception. For example, Dretske writes that he can tell what mood his wife is in better than others. “Knowing her as I do I can also see when she is tired, bored, uncomfortable, frustrated, and interested” (34). An example from Stroud is that he can see that his neighbor is home just by seeing her car in her driveway. He knows that she rarely takes the bus and so forth, so he is able to come to see this fact even when he cannot see her in her house. Under the restricted view of perception, these beliefs can only be added inferentially to the way things appear in order to become the corresponding belief. It may be the case that at first Stroud had to use his belief that his neighbor always travels by car inferentially, but over time the belief has been built into his perceptual faculty. Neither Stroud nor Dretske write much about the process of building beliefs into perception. They may even believe that simply holding a belief allows it to be used in perception. Regardless, it is this capability of using belief and knowledge non-inferentially in perception that allows PP to overcome many skeptical worries.
Stroud has also argued that PP is essential in a proof of the external world. Roughly, the idea is that the concepts we use in perception must be gained and mastered by actually experiencing the concept instantiated at some time prior. So, even though we cannot prove in a particular case that the concept we employ is actually instantiated, we must have correctly recognized the concept at some point. Dretske has argued that the problem of other minds is no different from the problem of other worlds, so it is easy to imagine a similar proof for the existence of other minds.
This same ability to incorporate beliefs into perception—which aids in solving these skeptical problems—seems to allow that we can see that future events will happen. For example based on our belief that unsupported bodies fall, we might say that we see that a rolling ball will fall once it reaches the edge of the table. Or we might say that we see that someone will respond when she is asked a question because we see that she is listening. Dretske writes that we can “tell by a person’s enthusiasm that cocktails are about to be served” (43). In general, if we are able to build our belief in the laws of physics into perception, PP would say that we are able to see that the events these laws predict will occur. Even without a belief in laws, these concepts themselves seem to imply future states of the things we ascribe them to. The concept of ‘rolling’ involves that a rolling object is in motion and therefore will soon be in a new position. Likewise, a person who is ‘listening’ will soon be thinking about whatever is said next. If we are able to incorporate these concepts into perception such that we are able to recognize them directly and non-inferentially in our experience, then we are able to perceive these events that are included in them.
I suppose it is possible that Stroud and Dretske might just accept that their view entails that we can see the future, but since that view differs so greatly from the common sense they believe to be vindicated by PP, I will assume that this is a problem for them. If they choose to reject that we can see that future events will happen, one way they may respond is that there are certain concepts that cannot be built into perception. Some concepts involve temporal extension of a certain state, like rolling or dancing or listening. Peter Strawson makes such a distinction with what he calls ‘P-predicates’ which are used in attributing intentions to persons. They “imply intention or a state of mind or at least consciousness in general, and which indicate a characteristic pattern, or range of patterns, of bodily movement” (111).3 Strawson realizes that these kinds of predicates give us reason to believe that future movements of a body will take place because the future movement is included in the predicate. So, the response would say, in cases where we use a temporal concepts like Strawson’s P-predicates, we can only make an inference about what will happen in the future. Our experience of the rolling ball or the dancing person should be described as seeing reason to believe that the ball will fall or that dancing person will continue dancing. After all, the ball might be picked up before it reaches the edge, but that would not negate our judgement that it was rolling. Other predicates that don’t include some future state can be built into perception.
This response seems incorrect for two reasons. First, as Dretske himself noticed, perceptual knowledge cannot be gained through inference. This is the reason PP was developed in the first place. Dretske spells this out and calls it the paradox of mediate knowledge. “Either P is the sort of thing that can be known immediately (non-inferentially, directly, with no evidential basis), or it is the sort of thing that can’t be known at all” (40). If P is not knowable directly then there is no way to know that something else can help us make a valid inference to know it. PP hoped to show that we can know things directly, rather than by inference as the restricted view of perception held. PP is based on the insight that nothing can validate inferences of this kind. Thus, future events must either be known directly, or they cannot be known at all.
The second reason why I think this response is flawed is that we cannot really distinguish temporal predicates from non-temporal ones. Tables continue to be tables in the same way that rolling things continue to roll. Our belief that the future will resemble the past leads us to believe that static states will continue just as much as it leads us to believe that dynamic states will alter something in time. If I saw that something is red and I also believe that nothing has changed its color, when I see that it is now green I will question our past perception. Did I really see that it was red? That I will predict it to remain red shows we can make a prediction even with predicates that do not seem to contain a temporal element. Even if we could distinguish temporal predicates, we predict future states based on non-temporal predicates just as much as we might based on temporal ones.
Stroud and Dretske developed PP because they believed that the traditional view of perception allowed us to gain too little from perception. But PP allows us to gain too much. Simply receiving appearances from perception does not allow any knowledge since all knowledge requires the application of a concept. But denying the separation of the faculties of perception and that of conceptual application allows knowledge that we are uncomfortable accepting. Unless there is some way to distinguish the concepts that can be incorporated into perception, PP in too powerful of a view of perception.
1 Stroud’s views on this topic are taken from “Seeing What Is So” Printed in Perception, Causation, and Objectivity. Roessler J. et al (eds). 2011. pp. 92-102.
2 Dretske’s views are taken from “Perception and Other Minds” Nous 7 (1). 1973. pp. 34-44.
3 Individuals: An Essay in Descriptive Metaphysics. Methuen & Co. 1959.