[Except for quotation, the following is summary. Unless otherwise noted, all underlining, boldface, and bracketed commentary are my own. Proofreading is incomplete, so mistakes are still currently present.]
Memory gives us data of the past, but sense seems to only tell us about the present. Thomas Reid writes:
It is by memory that we have an immediate knowledge of things past. The senses give us information of things only as they exist in the present moment; and this information, if it were not preserved by memory, would vanish instantly, and leave us as ignorant as if it had never been. (1855: 211) [Reid, T. 1855. Essays on the Intellectual Powers of Man. Ed. Walker. Derby: Boston.]
Memory plays a vital role in how we operate within time. Our actions at one moment, like going to the post office, might result from us remembering from before that we needed to go, and also remember from a while back the route to get there. In the short term, we would for example remember a bell chiming three times only be remembering the first two when the third strikes. But what about the immediate present? “Reid is right about all this. However, his claim that immediate experience is confined to the present is, on reflection, less innocuous that it might seem.” (Dainton 362)
Dainton now notes a strict or ‘proper’ conception of the present, which would regard it as a durationless instant. And since change requires duration, we would not in the present be able to perceive change.
Strictly (or properly) conceived, the present has no duration: it is simply the interface between what is future and what is past. Hence if our experience were confined to the present, it would be impossible for us to experience phenomena which require duration. More specifically, we would not be able | to directly experience change, succession or persistence. Reid draws just this conclusion [Dainton quoting Reid]:
. . . if we speak strictly and philosophically, no kind of succession can be an object of either the senses or of consciousness; because the operations of both are confined to the present point of time (1855: 235) [Dainton 361-362, quoting Reid, T. 1855. Essays on the Intellectual Powers of Man. Ed. Walker. Derby: Boston]
Yet, although this seems common sense, our everyday experiences tell us that we do directly experience succession in present moments of awareness. Reid thinks this only apparently so and that we are not really aware of succession in the present.
Aren’t there occasions (many of them) when we see a horse crossing a finishing line, or a car turning round a corner? Can’t we hear the explosive roar of a crowd, or the barking of a dog? Can’t we feel shivers running down our spine? Reid accepts that we do indeed often think and talk in these terms. He insists, however, that the apparent contradiction between strict philosophical truth and common sense (and common experience) is apparent rather than real [Dainton quoting Reid]:
. . . philosophers and the vulgar differ in the meaning they put upon what is called the present time, and are thereby led to made [sic] a different limit between sense and memory . . .. though in common language we speak with perfect propriety and truth when we say that we see a body move, and that motion is an object of sense, yet when as philosophers we distinguish accurately the province of sense from that of memory, we can no more see what is past, though but a moment ago, that we can remember what is present; so that philosophically speaking, it is only by the aid of memory that we discern motion, or any succession whatsoever. We see the present place of the body; we remember the successive advance it made to that place: the first can, then, only give us a conception of motion, when joined to the last. (1855: 236-7) [Dainton 363, quoting Reid, T. 1855. Essays on the Intellectual Powers of Man. Ed. Walker. Derby: Boston]
Dainton now wonders if Reid’s distinction between sense and memory really holds. For do we not actually perceive motion and change?
[…] a serious worry remains. How plausible is Reid’s proposed (precise, philosophical) delineation of the respective provinces of sense and memory? Motion certainly can be inferred in the way he describes, but can’t it also be perceived? If I move my hand towards my face aren’t I seeing my hand getting closer? Aren’t I seeing it moving closer? If I hear a fast succession of notes, played on a piano say, don’t I hear each note giving way to the next? Reid’s claim that we only ever see ‘the present place’ of a moving body—and so never actually see a body in motion—does not ring true.Might it not be that while Reid’s diagnosis does full justice to our temporal talk, it does not do full justice to our temporal experience? (363)
There are some philosophers who would agree that we experience change and passage in our temporal experiences. Russell says that “Succession is a relation which may hold between two parts of one sensation for instance between parts of a swift movement which is the object of one sensation. (1913: 65)” [Russell, B. 1913/1984. ‘On the Experience of Time’, in The Collected Papers of Bertrand Russell, vol. 7. London: Allen and Unwin., qtd in Dainton 364] Broad as well claims that we do not just see that something has moved or changed, but we we in fact see it moving or changing. “Broad remarked: ‘it is a notorious fact that we do not merely notice that something has moved or otherwise changed; we also often see something moving or changing. This happens if we look at the second-hand of a watch or look at a flickering flame. These are experiences of a quite unique kind; we could no more describe what we sense in them to a man who had never had such experiences than we could describe a red colour to a man born blind.’ (1923: 351)” [Broad, C.D. 1923. Scientific Thought. London: Routledge and Kegan Paul., qtd in Dainton 364]. Husserl also writes that we perceive succession while something is happening [and not after the fact in retrospect]. “In an auditory vein Husserl writes: ‘The evidence that consciousness of a tonal process, a melody, exhibits a succession even as I hear it is such as to make every doubt or denial appear senseless.’ (1964: 23)” [Husserl, E. 1964. The Phenomenology of Internal Time-Consciousness. (Tr. J. Churchill.) The Hague: Marinus Hijhoff., qtd in Dainton 364].
Dainton now examines empirical findings which support the claim that we perceive not states of things but their changes between states. It so far seems we do not have a sense organ for perceiving time, but there is research which suggests we have “perceptual sub-systems which specialize in the detection of motion.” (364) For example there is a part of our visual cortex that do not sensitive to color or shape but are highly sensitive to the motion of visible objects.
Empirical findings support these claims. If we possess a specific organ for the detection of time per se it has yet to be discovered, but we do possess perceptual sub-systems which specialize in the detection of motion. The area of the visual cortex known as V5 is one such: evidence suggests that the neurons in this region are insensitive to colour and shape, but highly attuned to large-scale motions, of the sort associated with medium-sized physical objects. Stroke-damage to this region in neurological subject ‘L.M.’ resulted in the onset of cerebral akinetopsia: the severely degraded ability to perceive motion (Zeki, 1991, 2004; Rizzo et al, 1995). Her predicament was characterized thus [Dainton quoting Zihl et al.]:
The visual disorder complained of by the patient was a loss of movement in all three dimensions. She had difficulty, for example, in pouring tea or coffee into a cup because the fluid appeared to be frozen, like a glacier. In addition, she could not stop pouring at the right time since she was unable to perceive the movement on the cup (or a pot) when the fluid rose. . . . In a room where more than two people were walking she felt very insecure and unwell, and usually left the room immediately, because ‘people were suddenly here or there but I have not seen them moving.’ (Zihl, von Cramon & Mai, 1983: 315) [Zihl, J., von Cramon, D. & Mai, N. 1983. Selective Disturbance of Movement Vision after Bilateral Brain Damage. Brain, 106: 313–340, qtd in Dainton 364].
[[Dainton is pointing to a perception sub-system which could be a candidate for what allows us to perceive time’s flow. I will here point out that change was perceived without flow, and it was done so with memory it seems. The subject does not say that people appeared suddenly here and there but suddenly here or there. In other words, they were not in two places in the same specious present. In each present, something occupies one location. How is that shocking? It also would not be shocking if in the just prior present she saw the object in a neighboring location. What is shocking is the difference between them, and that difference is given now as a phenomenon but it involves a memorial relation to what has just past. This example can in fact strongly support the model of time that Dainton does not advocate, the retentionalist model as he elsewhere calls it. It also allows us to further divide this phenomenon of flowing passage Dainton is here analyzing into two distinct phenomena: the phenomenon of passage and the phenomenon of flow. This woman perceives passage but not flow. If she did not perceive passage or change, then she would not be shocked by each new location of the moving objects. So the change of location is phenomenally given to her in each present moment. What is not given is the phenomenon of the motional and temporal flow between the changing positions of the objects. What this suggests is that the phenomenon of passage can be given without the direct perception of succession. And in terms of Dainton’s distinction between extensional and retentional models, what we might suggest instead is an intensional model. In this model phenomenal time is the result of the synthesis of before and after. To the degree that before and after are similar time in its extensivity is synthesized and the phenomenon of temporal flow is more evident. To the extent that before and after are dissimilar, time in its intensity is synthesized and the phenomenon of temporal passage or change is more apparent.]]
Dainton then turns to further empirical evidence which supports his claim that we directly perceive motion. There is something called the phi phenomenon, which is the pure phenomenon of motion without seeing something moving. Exner demonstrated it by showing two illuminated spots set at a distance from one another and flashed in succession. If flashed in succession very quickly, they appear simultaneous. If very slowly, they seem independent to one another. A little bit faster than that, it appears as if the one flash itself moves from one to the distant location. That case already exhibits the illusion of motion. But there is a very curious case that happens when the flashes are given just a bit slower than this optimal case. When at this slightly slowed speed, they go fast enough that they seem to be motionally connected but slow enough that no visual evidence illusory or not appears to connect their positions. Instead, the subjects perceive that there is motion between the flashes’ positions, but that motion itself is not visible. It is a pure phenomenon of motion. [[Wertheimer’s similar demonstration shows this also to not be a result of the persistence of vision. The marks are separated enough that the prior location does not superpose in overlap to the new one. And the flashes were too fast to be explained by eye motions that may have caught the prior flash, kept it as a persistent visual impression, then subsequently overlaid it with the new perception. So movement is perceptible without it resulting from the persistence of vision.]] This phenomenon explains why film strips can show us a succession of images but we perceive a fluid motion. [[Dainton is using this example it seems to make the point that we have sub-systems in our brain that perceive motion. It is not clear to me yet if this example also illustrates his point that the present must be extensive. So let’s consider the case of watching a film. The retentionalist would say that we only ever see a single image (or the flicker of black between), but we retain impressions of prior images showing the object in different but spatially continuous positions (continuous relative to the positions of other objects), which blur together to seem as though the position is continuously changing. The extensionalist would say that we perceive two images in one present but in their actual succession, and the phi phenomenon discerns the motional connection between them. And this is further supported by Wertheimer’s experiment which shows that it is not on account of superposed retentional images that we sense the motion of the image. However, the retentionalist could still say that motion results from the phi phenomenon, because our minds perceive the distant locations of the flashed object and discerns a motional connection between them. However, it seems Dainton might be making a different point, which is that motion itself is a phenomenon, and we directly in the present perceive an object’s motion and not retrospectively by means of processing memories and comparing them with the present. But the retentionalist might still claim that in any given instant we are impressed with the sense that something has moved, because right now we perceive in retentional memory both the prior and the current position; this proceeds over a series of moments, and thus during each moment we perceived motion.]]
Other findings point in the same direction. As the American psychologist Exner may have been the first to notice, if two brief visual stimuli are shown repeatedly in rapid succession—e.g., two illuminated spots on a screen a few inches apart, flashing on and off alternately—rather than seeing a succession of flashing spots, we see a single spot moving smoothly back and forth. The lesson? Our visual systems are not only capable of producing dynamic visual content, they are prepared to do so on the flimsiest of excuses. Perhaps we should not be surprised, for given the limitations of our optical systems, our eyes provide our brains with partial and fragmentary evidence of our surroundings, and our surroundings often contain objects that are in motion. | In any event, the effect is a happy one: substitute a rapid sequence of cinematic stills for the alternating spots—static images which depict a moving object at a succession of neighbouring locations—and the result is the clean, smooth motion apparent on cinema (and television) screens. Often called the phi phenomenon, this effect is also known as ‘illusory motion’. In one respect this is appropriate, given that our perceptual experience is dynamic whereas the images on the screen are static, but there is also a sense in which it is misleading: in most cases on-screen movements are indiscernible from the real thing. (Dainton 364-365)
So because this scientific evidence suggests we have the capacity to presently perceive motion and change, then we must reject Reid’s position that our consciousness never perceives directly outside a durationless present and instead take up the position that present consciousness occupies a specious present, meaning that we are somehow aware of an interval of time. (365) But there are very different ways to explain on what basis we are able to perceive this tiny interval of specious presence.
Dainton will proceed to explain his own model of specious presence, what he calls the Overlap Model, and he will argue for its advantages over the other models. “Although the Overlap Model has an agreeable simplicity, it can also achieve what more complex competing accounts fail to achieve: it can make sense of how our experience seems to be.” (365) But first he will address confusions and make some important distinctions.
§2 Conflations, Confusions, Etc.
There are not many who find it problematic to say that we perceive spatial extension. But many have argued that it is problematic to say that we perceive temporal extension. (365-366)
In the 15th chapter of his Principles of Psychology, James provides an account of time perception that has been influential but also misleading. (366)
We are trying to explain how we can be aware of an extent of present time. There are two main approaches to doing this. One says that consciousness itself spans a brief extensive temporal interval, and the other says that consciousness only seems to span an interval, but in actuality it does not. Each one accounts for temporal depth as though it ran along a different axis of co-givenness. The extensionalist approach thinks that the specious present is only given in an actual succession that is perceived directly and not by means of representations like memories, while the retentional approach thinks that the specious present is apprehended instantaneously by means of superposing in one instantaneous act of present awareness a series of retentions of prior moments all perceived simultaneously and together seeming to present to our awareness a thick present when in fact it is durationless in actuality. [[In the following quotation from which this summary is based, we find Dainton discussion a phenomenon that we might add to the phenomenon of passage and of flow. This would be the phenomenon of temporal depth, which is that time has a thickness to it. It would result from taking a succession together or distant moments together in the same act. One example of the phenomenon of temporal depth would be seeing a very young relative at one age and then again years later after they have grown up a good bit. In that moment, we do not merely sense the passage of the present or its flow, but also its depth, that there is content that stacks up and gives time a thickness.]]
When it comes to explaining how our awareness seemingly manages to extend through time there are two main schools of thought. Some maintain that consciousness itself spans a brief temporal interval; others maintain that although consciousness seems to embrace a brief temporal interval, it does not really do so. As the latter view is typically developed, the appearance of temporal depth is a consequence of a momentary cross-section of actual experience being apprehended together (and simultaneously) with representations of a temporal spread of phenomena. In the absence of any widely accepted labels for these approaches I will refer to them as follows:
The Retentional Model: a specious present consists of a combination of two simultaneously occurring ingredients: (i) a momentary direct experience and (ii) representations (or retentions) of the recent past.
The Extensional Model: individual specious presents consist of temporally extended episodes of experience that are apprehended as wholes. (Dainton 366)
Fig.1 from p. 366 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18: 362-384.
In figure 1, Dainton provides a diagram showing how proponents of the retentional model regard the specious present as one instant within which are stacked other retained representations of prior moments and how those using an extensional model regard the specious present as occupying a small extent of ongoing present time. In the diagram, letters C and D stand for a short succession of heard tones. (366)
Dainton then gives a passages by James which can be seen as taking both approaches. (367) In one case, James writes that the present cannot be regarded as a knife-edge but rather as a saddle-back which straddles a breadth of time, or as James also calls it, a “duration-block”. James continues: “We do not first feel one end and then feel the other after it, and from the perception of the succession infer an interval of time between, but we seem to feel the interval of time as a whole, with its two ends embedded in it. (1890: 609–10)” [James, William 1890. The Principles of Psychology. New York: Dover, qtd in Dainton 367]. However there are a number of other passages where James portrays the present as exclusive to past and future [rather than including them a little in one present succession]. (367)
James’ lack of clarity on this issue allow for these dual interpretations, and other philosophers are likewise dually interpretable. (367)
There are also problems with James’ assessment of the length of the specious present, which he says is about 12 seconds and includes fringes lasting around a minute. However, if we tap our finger on the table at one second intervals, we do not still hear the prior one when the current one is sounding. (367-368)
When the specious present is construed as having a long duration that extends into future and past, this seems to suggest we can see the future and the past. But our normal perceptions do not allow us to do this. James offers an example of how such perceptions would imply we respond with certain behaviors, when in fact we never would.
Just suppose James’ estimates are along the right lines. Further suppose that you are about to embark on a timed 1500m run. Since you are eager to set a good time, you are straining to hear the starter’s gun; since you become perceptually aware (even if only dimly) of the sound of the gun around 30 seconds before it actually occurs, you could—if you were so inclined—get a half-minute head-start. Since the official at the finishing line will see you crossing the line 30 seconds before you actually do, you might easily find yourself taking up to a minute off the usual sort of time for a 1500m—and do so without even trying. (368)
[[We take special interest in this example, because it will be similar to another sort of example that might locate a problem with the extensionist overlap model. Instead of the runner, starting gun, and finish line, consider a bowling ball rolling to strike a set of pins. Dainton will suggest that the extensivity of the phenomenal specious present tells us about the temporality of the real world around us. So for Dainton, every present moment of the ball’s motion spans a short amount of time, across which the ball moves some extent of motion. Perceptually speaking we might think of this also in terms of the blur of the ball’s motion that we see. The ball in fact occupies a series of locations all in one singular present. Does that mean when it first strikes the pins that it does so not just from position of contact but also from prior distant positions? The problem we are raising with Dainton’s model and its implications for the real physical world are problems regarding how we account for causal relations between physical bodies. If we say that the distant location is less directly related causally to the impact, that it does not directly impact the pin but only indirectly by means of mediary moments, then what is the sense of including it in the present? What does it mean to say that a distant position of the ball is indirectly yet presently causally related to the collision of the ball? And also, if the extensive specious present has an order of succession and a direction of passage, that means there is a future-most part on one end and a past-most part on the back end. Does that not imply there is a present-most part in the center, relative to the ends? Dainton seems to imply such a thing in his finding the absurdity of a long extensive specious present. He says we might hear the sound 30 seconds before it actually occurs. This means it does not actually occur but it is still present. Now by extension consider Dainton’s much shorter specious present of less than a second. When the ball strikes the pins, does it somehow presently strike the pins before it actually strikes it? Is there any phenomenological evidence that we perceive things this way?]]
This example of the race comes from Plumer who tries to show that the sensory present construed as an interval leads to absurd conclusions. However, Dainton says that these conclusions do not actually result merely from the assumption that the present is extensional, because extensionalists do not need to claim that we see into the future and also extensionalists might better avoid these problems by saying that the specious present is only a second or so in length. Moreover, extensionalists still have the better account for how we directly perceive motion and change. (368-369)
Dainton then notes how some find the notion of the extensional specious present absurd, but this is because they mistake presence for simultaneity. Le Poidevin for example writes:
If we have a single experience of two items as being present, then, surely, we experience them as simultaneous. Suppose we are aware of A as preceding B, and of B as present. Can we be aware of A as anything other than past? (2007: 87) [Le Poidevin, R. 2007. The Images of Time. Oxford: Oxford University Press, qtd in Dainton 369]
So extensionalists must face the challenge of accounting for successivity in a real present and retentionalists have the problem of explaining successivity in an instant. In the first case you have a temporal order that given singularly in the present, and in the second case you have a singular moment which contains an ordered succession. [Or perhaps: In the first case, you have the problem explaining how a multiplicity of successive moments can combine into a singular present, and in the second case you have the problem of explaining how a singular present can be seen as divided into a multiplicity of successive moments.]
It concerns the manner in which advocates of the specious present construe the relationship between (what we might term) presence and simultaneity. However, given that there are two fundamentally very different ways of conceiving of the specious present itself, there are two quite different worries in this connection. Extensionalists have the problem of explaining how it is that contents spread over an interval of ordinary clock time can appear successive if they also all seem present. Retentional theorists face no less of a challenge: according to them, the contents of the specious present are actually simultaneous (with regard to ordinary clock time), so how can it be that they appear successive? Or in slightly more formal guise:
Extensional Simultaneity Problem: how is it possible for contents which are (i) experienced together, and (ii) experienced as present, to be experienced as anything other than simultaneous?
Retentional Simultaneity Problem: how is it possible for a collection of contents which occur simultaneously to seem successive? (369, boldface is Dainton’s)
So one problem is accounting for how the parts of one specious present are combined. The other problem is explaining how different presents are put together to form the larger streams of our consciousness that run for very long periods of time.
We also need a plausible account of the manner in which distinct specious presents combine together to form continuous streams of consciousness. We can remain continually conscious for hours at a stretch, and generally speaking each neighbouring brief phase of our stream of consciousness slides seamlessly into its successor—our streams of consciousness are not divided into discrete pulses. (369).
So there are is an additional requirement that these models of the specious present need to satisfy.
The Dynamic Requirement: change, succession, movement, persistence are all directly experienced over short intervals.
The Continuity Requirement: specious presents must be able to combine to form a phenomenally continuous stream of consciousness. (370, boldface is Dainton’s)
[[We might think of these two requirements in terms of our distinction between the phenomenon of change or temporal passage and the phenomenon of flow. The model needs to explain how it is that we are directly aware of how in one specious present we can perceive motion or change, that is, how we can perceive succession within that present. It also needs to explain how many such presents flow continuously into one another to form a larger flowing stream of phenomenal time.]]
§3 Extensional Approaches
Dainton supplies figure 2 to put James’ “duration block” into extensional terms. The whole block P (the square) is one single specious present. It extends through time, as P is “a temporally extended phase of a stream of consciousness.” (370)
Fig.2 from p. 370 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18: 362-384.
Dainton then notes three things. 1) P’s contents, here depicted as the circle changing position, are “truly dynamic,” because “they possess the character (familiar from our own experience) of a ball moving and falling.” 2) the phases of the movement are both successive (as the arrows serve to indicate) while at the same time “ these successive experienced phases are also parts of a single (extended) experience that is sensed as a whole.” They obtain their unity on account of the fact that they are “connected by the relationship of diachronic co-consciousness.” [Dainton 370]. 3) Although these phases of the object’s movement are perceived as successive, they are also “experienced as belonging to the sensory present: all parts of the movement (in this brief interval) are equally and vividly there in the manner typical of directly perceived phenomena.” (Dainton 370-371) [[Here we have Dainton defining presence in terms of contents having full vivacity]].
The dynamic requirement calls for an account for the direct perception of change, which this model fulfills: “since each phase of P is experienced as vividly present, change is presented in as immediate a way as it possibly could be.” (371)
If in this model presence was the same as simultaneity, then it would not be extensive. Presence instead should not mean a temporal location but instead it should be a phenomenal quality of the content. It is present because it appears in immediacy with the full vivacity of present things. In order to grasp how it is that within one extensive specious present all the phases can equally have the phenomenal quality of presence, Dainton has us think of pain sensations. When the pain sensation is occurring, it has phenomenal presence. A week later we can again have another pain that as well when happening will equally bear this phenomenon of presence. Dainton then reasons, if distant phenomena can have the have this phenomenal quality of presence to an equal extent, then of course two phases of the same pain can have have presence equally as well. [[Again it is unclear how to sharply distinguish Dainton’s concept of presence from the retentionalist’s concept of presentedness. Why could the retentionalist not say that whatever is perceived as present is either the content present in that instant or the retentional contents in the retentional background whose vivacity either equals the actively present one or whose vivacity is indiscernibly less vivacious? If our phenomenal evidence tells us that there is a specious present made up of a series of phases appear equally vivacious, could that not be evidence both for an extensionalist model where present contents are actually equally present as well for a retentionalist model whose immediate retentions whose vibrancy are either equal or differ indiscernibly? The only distinction I can see so far is that retentionalists say that the retentions diminish in vibrancy yet are still in the specious present where for Dainton the contents in the specious present do not diminish in vibrancy. But this could be explained by the fact that Dainton’s own phenomenological evidence tells him that there is not a diminishing trail after each immediate perception. A retentionalist might share this with Dainton and have a model where all retended contents in the specious present have equal vivacity even though they are not presently active.]]
But by ‘present’ the Extensional theorist means (or should mean) something else: not a temporal location but rather a phenomenal characteristic. Contents that are experienced ‘as present’ in the relevant sense possess what we might term phenomenal presence: they possess the immediacy and vivacity that are characteristic of all phenomenal properties as and when they occur. A pain sensation has phenomenal presence while it is actually being experienced; if at some future time it is remembered then this phenomenal presence is lacking—though of course memory-images have their own distinctive (but different and less vivid) phenomenal presence. There is nothing problematic in supposing that instantiations of phenomenal properties at different times can have presence in this sense—my headache last week had just as much phenomenal force and vivacity at the time as my current headache. Given this, to suppose that the successive phases of a single specious present can all possess phenomenal presence is not in the least puzzling or problematic either. (371)
So Dainton just explained how an extensionalist model can account for how the contents of the extensional specious present can be both really successive (and not simultaneous) while also being co-present: they all share an equal degree of appearing to be have the full force and vivacity of present things. [[Is this definition circular or problematically tautological? Q: What determines whether something is phenomenally present? A: Whether or not it has the properties of phenomenally present things. The only way I can make sense of this is if one phase in the succession within the specious present is ‘actually’ present or presently present and the others are no less present yet are more specifically pastly present and futurely present.]] [[Another problem with Dainton’s model might be that he still implies the distinction where the present contents are most vibrant and the non-present ones are less vibrant, just as in the retentional models based on presentedness. If what gives a content presence is that it has the full force and vivacity of presently perceived things, that means when they are not present they have less force and vivacity. Such non-present items would then have to be memories, as they are not present. Again, if retentionalists regard the most recent retentions as equally vibrant or indiscernibly less vibrant, then both models would be equally supported by phenomenological evidence.] Now Dainton will account for how these specious presents combine to form our stream of consciousness. He considers first one option, which lays out duration blocks end-to-end to form a continuous series, as depicted in figure 3.
Fig.3 from p. 372 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18: 362-384.
That contents CD are combined in P1 and EF in P2 was already explained in terms of their sharing the full force of presence when they occurred [and presumably less when they form a memory that represents past experience.] But the combination of blocks CD and EF are not accounted for here. Dainton distinguishes succession of experiences from experience of succession. We experience CD as successive and EF as successive, because their successivity is given in their specious present. But what is lacking is the connection between D and E. Diachronic co-consciousness is a shared awareness of successive contents, and this is lacking between the blocks.
But if under the current hypothesis D and E do not fall within a single specious present they are not diachronically co-conscious, and so are not experienced as successive. From a phenomenal perspective, in the absence of any experiential connection between the two, they might as well belong to distinct subjects, or different universes. (371d)
To introduce this connection between blocks, what is needed is an intermediary block, P3, which is co-conscious of D and E.
Recognizing the existence of P3 does not bring with it a commitment to any new experiences (or phenomenal contents); we have the same four experienced tones as previously. All we are recognizing are additional phenomenal relationships among these experiences. Under the initial hypothesis only the pairings (C-D) and (E-F) were phenomenally unified; under the new hypothesis the pairing (D-E) is also phenomenally unified. Hence phenomenal continuity is secured at very little additional cost. (372)
This gives us the basic structure of the overlap model, in which successive specious presents are combined continuously by means of a continuous overlap of contents from one moment to the next. However, this model is an oversimplification of our actual experience. For one, the phases of our awareness are not as simple as merely hearing one tone and another as if each were a solid block. There is much more heterogeneous detail within the experiences of each tone. Nonetheless, richly complex phases are no less able to share common parts. “But this additional complexity does not affect the essentials: P1 could overlap with P3 by way of sharing a common part even if the parts in question possess highly complex and varied contents.” (372)
There is then a second issue to deal with. Dainton notes how in the simplified diagram the middle present overlaps the others by one half. He then says there is reason to believe that although the presents here are given a determinate length, that the do not move in such a disjointly segmented way that there is only exclusively one or another block. It would seem that the present moves in smaller steps. He depicts this in figure 4. (372)
Fig.4 from p. 373 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18.
[His way of presenting this is not entirely clear to me. He says that “Isn’t the final third of C experienced with the third phase of E?” If each specious present keeps the same length, then even smaller increments would not allow for this. If we move P3 back a third (or have a P4 that is one third back) it would cover the final third of C but not the final third of E. And as the diagram already shows, the final third of E is not experienced with the final third of C, because the specious present is not long enough to do so. Perhaps he is suggesting something more dynamic, for example saying something like, ‘Does not the specious present move into the final third of E just while exiting the final third of C?’ Perhaps then the multiplying of smaller increments implies this more continuous motion.]
The overlap model will need to be developed, which would be done depending on one’s view of consciousness and it contents. So far we can say that the overlap model satisfies the continuity and the dynamic requirements. (373)
Dainton will examine another way for an extensionalist model to to treat individual specious presents. It is found more in retentional models and its merits will be evaluated. (373)
So far Dainton has assumed that the contents of duration-blocks all have equal phenomenal presence. Another view says that the contents presence diminishes the further they are from the present. To say that the contents diminish is to say that they are given under different modes of presentation. So Dainton distinguishes non-modal models, which say there is no change in presence, from model models, which say there is.
According to the alternative view, the contents within a single specious present systematically vary in the degree of presence they possess, with the more recent contents seeming more present (or less past) than earlier contents. James sometimes spoke in such terms, and Brentano was explicit about it: ‘only someone capable of a presenting with different modes and of a continuously changing mode of presentation can have a presentation of rest and motion, of continuing to exist or of proceeding in time.’4 (1988: 82) We can mark this difference thus:
Non-Modal Models: all parts of the specious present possess phenomenal presence to an equal degree; there are no temporal modes of presentation.
Modal Models: not all parts of the specious present possess phenomenal presence to the same degree; there are temporal modes of presentation. (373, Boldface is Dainton’s)
Dainton wonders if in fact our consciousness is even able to perceive such variations within a mere second. (374)
Dainton will now show why extensionalists should avoid taking this position. He shows the options for how they would do so in figure 5.
Fig.5 from p. 374 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18.
In figure 5, the darker the regions, the greater the presence of the content.
As we can see, model A which strings the presents together portrays the quality of presence as varying in pulses, but that is unlike how we experience the present as being continuously fresh. So consider model B. Here each present overlaps a little with the prior one such that the freshest content of the prior moment is second fresh in the current, and so on within the span of the specious present. With this rendition we have the problem of lingering contents. Dainton thinks that contents do not lingering in our awareness like this model suggests. We should be able to hear E without C lingering in the retentional background, and if we wave our hands slowly in front of our eyes, it should not be the case that we see a trail behind it. [[Dainton also uses the moving hand example to also demonstrate the blur that can result, but it seems the difference has to do with speed. First note that in his “The Experience of Time and Change,” Dainton discusses how at a certain distance a moving train will appear as blur. In his talk “Passage in Experience and Reality,” at around 11.00 minutes or so, he discusses motion in visual experience. One sort are “objects which disappear into a blur (fast car, waving hand).” However, in his book The Stream of Consciousness he refers again to waving the hand, but in this case he seems to say he does not see a trail, and also he again says to move the hand slowly (skip to summarized pages 156 ff). So whether or not the moving hand makes a trail seems to result from the speed. But this will not help his argument, in fact it would seem to provide evidence against it. When it is moving quickly, it passes through very many more positions within the short span of the specious present. So perhaps as we move it slower, the trail contracts more and more around the hand, and at a certain decrease in the speed the trail is there but becomes indiscernible. Even if not, we have the problem of explaining the fasts moving hand without the trail being less vibrant than the hand, for else the specious present would have modal contents which it is not supposed to have. When I wave my hand quickly before my eyes, it is not as though it stretches from one place to the next. The trail behind the hand is not as vibrant. How does Dainton explain this? That trail appears within a specious present of the hand’s motion.]]
The problem now is with lingering contents. Note the way the content C in P1 continues to be perceived—albeit with gradually diminishing presence—in P2 and P3. If the contents of our consciousness were organized in this way then it would be impossible for us to hear a sequence of notes C-D-E without C still sounding when E is heard—but clearly, this is possible. In the case of vision, we would be unable to see objects move cleanly in the way that we do: if you were to move your hand slowly across your field of vision, you would invariably see it pursued by a ghostly contrail of fast-fading but still visible remnants of earlier perceptions. (374)
Dainton’s overlap model solves this problem because the specious presents share common parts. But if they are numerically identical in all cases, then they cannot have different phenomenal characters in each present. This is why they must be equally vibrant. [[But it might also lead to what we might call the problem of stretched contents. If the different positions of a moving object in one specious present are all equally vibrant, moving objects would appear as though they were being stretched spatially rather than moving from place to place]]
§4 Retentional Models
Those who take up the retentional model must explain our experience of change by means of these retentions [since change cannot be given in the instant of present awareness].
There can be modal and non-modal retentional models. Husserl’s retentional triangle can demonstrate.
Fig.6 from p. 376 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18.
Horizontal line C-G is the experience of a single violin tone rising gradually in pitch [which each intervening letter being momentary phases of the consciousness of this change]. The virticle lines are each distinct specious presents, which are instantaneous and thus themselves are not long or deep enough to themselves be able to perceive any change. However, since they contain retentions going back a couple moments, we can nonetheless still in this model perceive changes that happen over extents of time.
Objectively speaking, each of these specious presents has zero (or minimal) duration; subjectively speaking, each possesses sufficient apparent temporal depth to manifest change and duration as it is immediately experienced. (376)
Retentional theorists often adopt the Model conception. In this diagram, that means points closes to the horizontal have the greatest amount of phenomenal presence, while as we move away from the horizontal line the contents diminish in presence and seem more and more in the past. (376) The primal impression for Husserl is the most present moment. [In the next moment it becomes the most vibrant retentional content now pushed now vertically and] contents and phases of contents move gradually down into the retentional past at pace with the incoming present contents.
tone-phase D-E is experienced first as E-D* and then later as E*-D**: the same extended sound is thus experienced as sliding smoothly into the past. (376)
The lines extending above the horizontal are protentions. Broad does not have this in his model. (377)
Dainton will show that because the Retentional model encounters problems with the Dynamic requirement (explain passage) and Continuity requirements (explain flow) it is not a viable model. (377)
Non-modal retentional models fulfill the dynamic requirement because a spread of contents of equal presence is given retentionally. However it has the (Retentional) Simultaneity Problem: the model does not explain how simultaneous contents (of equal modality) can appear successive. Retentionalists can point out that it is not obvious that it is logically impossible for successive contents to be condensed [while maintaining their phenomenal traits of successivity]. We do not know enough about the relationship between the physical and the phenomenal in order to rule it out [because perhaps the physically successive can still somehow find its way into the phenomenon of successivity given in an instant.] Also, there is no guarantee that the property of consciousness as a vehicle will have the same property of its contents [so an act of awareness with the property of presence can contain contents somehow with the properties of succession.]
Here the (Retentional) Simultaneity Problem rears its head: how is it possible for a collection of contents which occur simultaneously to seem successive? By way of a reply, the Retentionalist can point out that if it is logically impossible for such contents to be temporally condensed, as it were, it is by no means obviously so. Do we know so much about the relationship between the phenomenal and the physical to be able to rule this out? Retentionalists (of both persuasions) can also argue that the divergence between the temporal properties of the contents carried by specious presents, and the specious presents themselves, is simply another manifestation of the widely recognized fact that there is no guarantee that the properties which feature in the content of a representation will also be possessed by the vehicle of a representation. (377)
So the non-modal retentional model might not necessarily suffer from the simultaneity problem. But it has another issue to deal with [in another case grouped with the problem of surplus contents. Here he considers it surplus-to-requirement, which elsewhere was about repeating contents.]. Each specious present when it is happening gives us a second of time through the retentions. So at C we seem to also hear B and A in our present. But at moments B and A we heard a second of sound. So there are two seconds of experience packed into one. And if the presents are divided into intermediary phases, this just multiplies the retained experience of time into a short brief experience of the present. This is not plausible. (377)
Because of these problems retentional theorists often advocate the modal approach, as for example was the case for Brentano, Husserl, and the later Broad. While solving these problems it confronts the theorists with “an awkward dilemma” (278).
On the one side of the dilemma, the danger is saying that retentions are like memory and lack the quality of presence. This then make it difficult to explain how retentional contents appear present to us when they also have the quality of memory or pastness. On the other side of the dilemma, the danger is to say that the retentional contents are all like present ones, then we have the problem of surplus contents [explained above when discussing non-modal models, which this would be close to]. “If at each instant we are aware of a second or more of immediately experienced change, we experience far more than we seem to” (378).
No retentionalist has been able to solve this problem (to Dainton’s knowledge) in a “clear and clearly satisfactory manner.” (378)
There is also the problem of fulfilling the continuity requirement. Each present is distinct, so how are they connected when none overlap?
Recall the example given earlier: if I hear an evenly spaced succession of tones that are of such a duration that only two can be experienced within a single specious present, I also hear each tone flowing into its successor. And what applies for individual tones applies equally to entire brief phases of our streams of consciousness. It may well be that there are relationship of causal dependency between one specious present and the next, but we are seeking | phenomenal relationships, which are something altogether different and more distinctive. (278-279)
[[If each moment is thought of like a differential, the tiniest possible partition of a change, then contained with it already is the link to the other moments.]]
The only way to satisfy the continuity requirement is to connect moments by diachronic co-consciousness. This can only result from the overlap model, but Dainton will consider if there is an alternative.
On reflection, it is difficult to see how the problem can be solved without allowing neighbouring specious presents to be connected by the relationship of diachronic co-consciousness. For unless we allow that the successive brief phases of our streams of consciousness are experienced together how are we to accommodate the phenomenological datum that these phases are phenomenally continuous? But anyone who takes this step is, in effect, abandoning the Retentional approach in favour of the Overlap form of Extensionalism. But is there any alternative? (279)
§5 Objections and Responses
Even though Dainton’s overlap model overcomes the other model’s problems, it has received criticism, and Dainton will examine those remarks. (379)
Sean Kelly for example finds three problems with the extensional approach. 1) How can we be perceptually aware of something no longer taking place? 2) How do we perceive duration directly? and 3) How can the overlap model explain continuity connections without retentions?
Regarding point 3, the overlap model has overlap by sharing common parts, and thus there is always continuity of consciousness. (379)
Regarding the second point, Kelly is skeptical that we can perceive intervals of time in themselves [like perceiving not just the contents of one second but perceiving that one second itself as if it could be experienced independently of its contents.] However, extensional theorist need not claim we have this ability. [[In Kelly’s “Temporal Awareness,” where he makes these objections, he writes, “In the first place, it is hard to understand how I could now be perceptually aware of something that is no longer taking place.” He considers as a possibility only that the time lag between an event and our perception of it allows for a perception of the past. But this is not what advocates of the specious present and specifically of the extensionalist model are saying. The more stubborn problem I think lies in how Dainton realizes he cannot have modal variation in the extensional specious present, but he is unable to formulate a model which explains passage without these valencies being implicit in the model’s structure. There are three phenomenal features of the specious present with regard to their temporal character that suggest the phases of one present carry with them a heterogeneity of temporal values. We will use some of Dainton’s own examples to help us locate some of these temporal features in our own experiences. And corresponding to these phenomenal features of our actual experience are structural features in Dainton’s overlap model which would explain the variance in modal values of speciously present phases even though Dainton insists these features are absent from his model.
Phenomenon 1: only one phase at a time appears to have direct causal influence on other present phases
Structural feature 1: strict linearity of the phases
Phenomenon 2: there appears to be a future-most and past-most part of each specious present.
Structural feature 2: on one side of the series are parts that have been carried over the most throughout a continuous progression of presents and on the other side are parts that have been carried over the least
Phenomenon 3: something in each moment feels fresh, new, original, and thus not-carried over from prior specious presents (by means of shared parts)
Structural feature 3: future-most phase has not appeared in any prior present. (From ABC to BCD, D is unique in that it is the only part that has been experienced for the first time.)
Before we look at how this temporal valency is found in our experience and in Dainton’s model, let first review the reasons why he says extensionalist models should not such modal variation in the phases’ temporal characters. In the A-model there was no overlap of presents but rather there were sequences of extensive presents each with their own phases but put into a chain. This portrayed time as though its freshness renewed in 1 second pulses, which we do not experience. In an overlap by superposition model (the ‘B-model’), there is the problem of lingering contents. (Although as a side note it is unclear how he explains the blur behind a fast moving hand or train if those are not lingering contents, and since there is a fuzzy blur and not stretch, it is apparent from our experiences that if there is lingering in the specious present then it is not contents that linger at full vibrancy. Nonetheless,) Dainton regards the lack of lingering contents to a slow moving hand to be evidence that we cannot have modal variation in an overlap by superposition model. The reason he cannot have it in an overlap by sharing common parts model is that the parts are numerically the same from moment to moment, which means that a phase that is fully present in one specious present and found in a succeeding one must also have the same phenomenal character. This is already called into question by the very phenomenological data Dainton puts forward in various cases, which we just mentioned parenthetically with regard to lingering contents. When speaking of his hand moving slowly within his vision, there is no trail behind it. But when it moves quickly, there is such a blur, just like how we see a blurry trail behind trains and automobiles moving at a certain speed. We see the train move a certain distance within a certain time. We do not see it stretch, just as something vibrating does not enlarge as if under magnification. The outlines might extend, but not at equal vibrancy, for otherwise we would see funhouse mirror-like distortions when things move rather than blurs. So somehow Dainton needs to explain some of his own examples using the overlap model. For now we will just take a couple of his examples, the fast-waving hand and speeding train. Why is it we see a blur behind the train and hand if phases of our awareness do not linger within the specious present? Whether he intends it nor not, his examples of fast moving objects do contain lingering contents, and he needs to explain why the blur is not the result of many positions of a moving object being perceived simultaneously with the most recent one as the most vibrant. But let’s for now just suppose that we see a succession of positions of an objects movement and put aside the question of whether older phases linger. He describes this as the directedness of the phenomenal flow of time, which in one diagram of the specious present he depicts as an arrow, and he writes: “The same three phenomenal presents are shown in the upper expansion, the single-headed arrows show the direction in which experience seems to be flowing, this apparent directedness is a consequence of the contents of the phenomenal presents; these contents have the form of temporal fields or spreads of content possessing inherent directedness or ‘flow’ (e.g. a ball moving to the right).” (Dainton “Time in Experience” p18d) Let’s continue with his assumption that the specious present is about a second or so long, and let’s imagine that we are watching the flow of motion of a ball moving to the right then striking a wall and bouncing backward to the left. There will be one specious present of our experience of this motion right when the ball strikes the wall, and since the specious present is about a second, let’s choose the specious present where the contact is made in the middle phase of the sequence, and including in our present awareness is it moving to the right and afterward to the left. Even though its rightward motion is equally present as the leftward movement, we do not perceive the rightward movement as being directly related causally to the leftward. When seeing any object moving one way we would not anticipate a sudden reversal in its motion unless we also perceived some other influence on its movement. So in our own experiences of present motion, we still seem to consider their being differences among the phases which makes certain ones appear most active and the others less active. Saying that we are mistaking presence with simultaneity does not remove this difference in value between active and inactive. If each phase has its turn at being active at the exclusion of the others, then each phase also has its turn at being inactive, and thus differing temporal values are given for each phase in each specious present. Whether or not we experience the less active parts of one present as having less presence in Dainton’s sense of full force and vivacity might be debatable. However, we still need Dainton to explain how a phase that is experienced as active seems just as present as the phases that are experienced as inactive (or how within one present the same phase has equally the phenomenal character both when it is taking its turn being active and also when some other phase has that turn instead). If it be argued that when we perceived the ball moving toward the wall we anticipated its reversal of motion and thus the activity of the forthcoming phase of the event was already given from the beginning, then imagine instead of the wall there is pane of glass that we were unable to see. Suddenly and inexplicably the ball changes direction. In that case there is really nothing about the older phase of the present experience which is indicative of the newer phase.
This sudden surprise of the ball’s reversed motion illustrates another modal difference in the phenomenal quality of the phases of one present. They are not only ordered in a strictly linear way but they are also ordered according to proximity to the past (memory) or future (the not yet experienced). Imagine Dainton’s melody example with this series of presents, each with three phases: ABC, BCD, CDE. By the third one, C will seem the oldest and past-most, if only because it is the most redundant part of that experience; it is the part that is most experienced in previous presents. There is a most stale and least stale part of each present of our experience. This is also found in the structure of Dainton’s model of the extensive present. The past-ward side has phases that have been around longer and their exit from the present will happen sooner. So perhaps Dainton could clarify how the fact that surprising new contents seem fresher than redundant older ones does not present within one specious present modal variation in the phases’ temporal qualities.
This is similar to the third temporally phenomenal feature in our experience of the specious present which is that each moment does not just seem relatively new but something about each moment is absolutely new and never before experienced. In the present when we first witness the ball reverse motion after hitting the glass, this reversal was nothing we experienced in previous presents. We see this in the structure of the model too, because each moment in the series ABC, BCD, CDE has a phase which was not found in the prior ones. Dainton also speaks of this newness or renewal that is a part of each present. “The fact that we directly experience both change and continuity suggests that contents spread over a brief interval of time can be co-conscious; the fact that our experience consists of a continuously renewed flow of content, a flow within experience itself, suggests that diachronic co-consciousness plays a key role in the generation of streams of consciousness” (Dainton The Stream of Consciousness 2003, 114). Also recall how he writes of tapping a finger over and over. Despite each tap having very similar phenomenal traits with regard to their sonic qualities, instead of hearing the same tap, we hear a “new” tap. “Tap a table with your fingers, at a regular intervals of about a second; after each new tap, ask yourself if you can still hear its immediate predecessors.” (Dainton “Sensing Change” 367). So we ask again for Dainton to clarify his model so we can understand how it is that each present has a part that is absolutely new and unshared unlike the prior redundant phases and yet none of these parts are experienced as having modal variations in their temporal qualities.]] Kelly’s idea that we are seeing for example the past state of the star when we see its light here on earth does not change the fact that when we see it, it seems phenomenally present. But the question is, “How can a single experience, of a perceptual sort, include phases which seem anything other than fully present?” Retentionalists claim that earlier moments in the specious present are given as being past, but the non-modal extensional model says they are all experienced as having equal presence.
Retentionalists have an answer to this—the earlier phases are presented as past—but what has the Extensional theorist to offer? In fact, as I hope is clear, the non-Modal form of Extensionalism outlined in §3 has no difficulty here whatsoever. The earlier and later phases of a duration-block all appear equally present (they all see [sic] equally vivid), even though they are also experienced as a succession. (Dainton “Sensing Change” 380).
In the next paragraph, Dainton seems to suggest that his overlap model is more compatible with the theories of time used in contemporary physics and particularly in Einstein’s relativity theories. [[In another entry we examined parts of Dainton’s book Time and Space to see why he might think this is the case and also to what extent the overlap model is compatible with these theories in physics. What we concluded was that while these theories do understand time as extensive, like the overlap model does, they do not seem to consider there being a moving present of about a length of a second or so during which the phases or contents are present and those outside it are not present.]] Dainton writes:
Of course, anyone who believes that concrete reality is itself confined to what the present instant contains will find the Extensional approach problematic. If the past simply does not exist, contents located in the past cannot be co-conscious with contents located in the present. (This thought may well underlie the Retentionalist doctrine that only items that are momentary and simultaneous can be phenomenally unified.) But while this view of time—Presentism—may have some appeal at the level of commonsense, it does not sit easily with the four-dimensional space-time ontology that (many believe) Einstein’s theories of relativity require. By allowing phenomenal unity to connect contents that are separated by time as well as space, it is the Extensional approach that is more in tune with the findings of contemporary physics. (Dainton “Sensing Change” 380)
Dainton then wonders if the overlap model is compatible with findings in contemporary psychophysics. He looks at Grush’s ‘trajectory-estimation model.’ It is like Husserl’s tripartite structure of time consciousness: retention, intention, and protention. [Husserl also talks about retroactive repainting of retentions as new information comes in.] In Grush’s model our perceptual systems give us momentary updates of our environment.
In simple terms, Grush holds that our perceptual systems generate a continuously updated series of momentary representations of our environment. Each of these representations is a model of a subject’s environment over a brief interval— corresponding to the duration of the specious present—and in Husserlian fashion each model comprises three components: a representation of what is occurring at the present moment, a representation of the immediate past, together with anticipations of the immediate future. Since the data our perceptual systems receive from our sense-organs is often fragmented and ‘noisy’, these internal models usually embody a good deal of extrapolation and educated guesswork. As a consequence, as new data is forthcoming, the content of the models—the ‘story’ they tell of what is happening over a particular interval—may change accordingly. (380)
One thing in Grush’s model that is interesting for us here is backward masking. What happens most recently in a specious present can alter the appearance of what happened previously in that specious present. In the phi phenomenon, the second flash makes the first seem to move. Had there not been a second flash, the first would have appeared stationary. The second flash retroactively went back to the retended image of the first one and made it move to the location of the second.
Grush argues that the trajectory estimation model’s revisionary abilities allows it to make good sense of what is going on in a certain temporal | illusions. The relevant cases feature different forms of what is often called backward masking. Backward masking takes place when what is perceived as occurring over a brief interval (of up to a few hundred msec) is influenced, in often surprising ways, by what is perceived in the final phases of the interval. In the case of the phi phenomenon we encountered earlier, if a single flash of light A is followed, a short time later, by another flash B a short distance away, instead of perceiving a single flash at A and another at B, we perceive a single spot of light moving from the location of A to that of B; if the second flash doesn’t occur, we see only A, an immobile solitary flash. In such cases it seems, bizarrely, that later experiences can exert an influence on what is perceived prior to their occurrence. Applying the trajectory estimation model to the phi phenomenon, Grush suggests the following. Prior to the time at which data pertaining to the second flash reaches our perceptual systems, the latter work on the (reasonable) assumption that just one flash has occurred, and so only a single flash features in the internal models (or specious presents) that are generated up to that point. As soon as data pertaining to the second flash arrives, however, our perceptual systems revise their assessments as to the likely environmental causes of the incoming sensory data. Since the new assessment is ‘an object moving between the locations of A and B’ this is the content which features in subsequent specious presents. As for the earlier representations featuring a single motionless flash, these are immediately forgotten and play no further role. (380-381)
[Recall that the overlap model holds that contents remain numerically the same in different presents, so they would not be revisable it would seem.] Dainton notes that trying to account for this seemingly retroactive alteration in contents presents a problem for the overlap model.
If we try to make sense of the same case in Extensional terms we are forced to ascribe inconsistent contents to the same segment of experience, or so Grush maintains. To see why, consider how an Overlap theorist will interpret the phi phenomenon. We can focus on two overlapping specious presents: the first of these, S1, includes the initial seeing of the flash at A and what immediately precedes this; the second, S2, includes this flash at A and what immediately follows, which includes the seeing of the flash at B. Or more schematically, S1 = [. . .-A] and S2 = [A-B]. Now, if S1 and S2 overlap by sharing common parts, the shared part—in this case A—must have the same intrinsic phenomenal properties in each of S1 and S2. But in this case, this requirement is not met: in S1 the A-flash is motionless, whereas in S2 it is perceived as moving. Evidently, given this difference we cannot coherently hold that the final phase of S1 is numerically identical with the final phase of S2. The Overlap Model thus breaks down (see Grush, forthcoming, §5). [Dainton “Sensing Change” 381]
Dainton’s reply to this is that extensional theorists might hold that there is a processing delay. In the end this would mean that flash A is never perceived by itself. [[It would still seem to be the case that there is a specious present preceding the appearance of B when A also appears. Dainton seems to be saying that in this prior present, A is not perceived by itself, perhaps this is because it was too fast or because the next moment with B in it comes too fast, but still it is a bit unclear why it is that even with the delay there is not a moment with exclusively A in it.]]
We can agree that experiences cannot have inconsistent contents. But it would be wrong to suppose the Extensional theorist has no option but to interpret such cases in this way. Grush seems to be assuming that the contents featuring in Extensional specious presents reflect their environmental causes in an immediate and entirely unmediated manner. But there is no need for Extensionalists to embrace this view of the perceptual process. It is arguably more plausible to construe perceptual contents as representations that are generated in the brain only after a good deal of processing. This processing | makes for a delay—50-100 msec, say—but our brains put this to good use: they try to work out a single, coherent version of events on the basis of the fragmentary and (at times) conflicting data available to them. Only this ‘final draft’, as it were, reaches consciousness. Hence in the phi case, our perceptual systems reach the (in fact erroneous) conclusion that A is in fact a moving light, and this is the only way in which it features in our experience. While the initial solitary, static A-flash may well register in our perceptual systems, it does so only at a pre-conscious level. Since this flash in this form is not experienced, the problem of inconsistent perceptual contents does not arise. This way of construing matters is on display in the upper portion of Figure 7; Grush’s interpretation is shown in the lower part of the diagram. (281-282)
Fig.7 from p. 382 of Dainton, Barry (2008) "Sensing Change." Philosophical Issues, 18.
Dainton continues by examining the flash-lag effect, which does not seem to pose a big challenge to the overlap theory but might need further consideration.
Dainton then concludes by noting “Time will tell which of these models most closely approximates the truth about these puzzling effects,” but there are at least three things we can say for sure at the moment. The first is that the doctrine of the specious present is not certainly untenable. The second is that psychophysical research might determine which models are correct but interpreting their results is not always straightforward. “The third and final lesson can be stated more succinctly: in the debates to come the Extensional approach may not hold all the aces, and it may not ultimately prevail, but it does at least start with several distinct and significant advantages” (383)
Barry Dainton (2008) “Sensing Change.” Philosophical Issues, 18: 362–384.