Consciousness and apparent motion: Paradox resolved

The problem

All too rarely do I find colleagues who will assent to the proposition (which I find irresistible) that the very ground-rules of science, its concern only for public knowledge, preclude its finding an explanation for my consciousness, the one phenomenon of which I am absolutely certain. (Pippard, 1992, p. 29)

The discovery of discrepancies between subjective perception and the ostensibly real physical phenomena has been one of the leitmotifs of consciousness research. Such discrepancies are most often found in perceptual illusions—situations in which stimuli produce perceptions that are incompatible with their objective descriptions. From the outset, psychologists have been interested in illusions and for good reasons (Fechner, 1860). First, illusions are somewhat mysterious in the sense that there are no easy explanations as to why they should be occurring. Second, they seem to offer the promise of a deeper understanding of perception and cognition through the use of controlled experimentation.

One important family of perceptual illusions concerns the so-called apparent motion illusions. Apparent motion is created by means of discrete events which, when arranged appropriately in space and time, give the appearance of movement that is not present in the original (static) stimuli. First reported by one of the founders of Gestalt psychology, Max Wertheimer (1912/1961), apparent motion has been researched intensively by a number of psychologists (see Kolers, 1972). Apparent motion phenomena are generally divided into four categories: Alpha—expansion and contraction caused by sequential presentation in the same location, of two objects of different size; Beta—movement is perceived between two distinct, sequentially presented objects; Gamma—expansion/contraction caused by sequential presentation, in the same location, of two objects of different luminance; Delta—motion in the direction opposite to the stimulus sequence (the second light needs to be much brighter than the first). The Beta phenomenon comprises three temporal stages. When lights are presented simultaneously or with a very small lag (roughly under 50 ms for a visual angle of few degrees), they are perceived as distinct (simultaneity). When the lag exceeds about 200 ms, the perception is of a regular sequence (succession). Between these extremes lies the region of apparent motion within which one light is perceived as moving from left to right and back. Original research (Korte, 1915) indicated that in order to maintain smooth motion, increase in the distance between stimuli should be accompanied by an increase in the inter-stimulus interval (coupling; Korte’s third law of motion). This was challenged by subsequent studies (e.g., Burt & Sperling, 1981), which found that increase in spatial distance required a decrease in temporal distance and vice versa (trade-off). The contradiction was resolved by Gepshtein and Kubovy (2007). While Korte’s law obtains at higher speeds (high ratios of spatial and temporal distance), the trade-off is observed at lower, more commonly encountered speeds.

Apparent motion has been studied in many contexts and under varied experimental conditions. A particular form of the effect named the Colour phi phenomenon (Kolers & von Grünau, 1976) has been widely discussed in the consciousness literature because it vividly exposes the temporal paradox. ¹ Assume a red-coloured light is flashed for approximately 150 milliseconds (ms). After a 50-ms interval, a green-coloured light, placed to the right of the first light, flashes for the same amount of time. From this description it appears that an observer should perceive two differently coloured lights flashing sequentially—or at least movement without the premature change in colour. While the former can be achieved by increasing the time interval between the lights, under optimal motion conditions the appearance is of a light trace moving continuously from left to right and changing colour abruptly mid-way. When both shape and colour are varied together, a smooth change in shape is still accompanied by an abrupt change in colour at the centre of the display.

The paradox of apparent motion can be stated as follows: How is it possible to perceive motion from stimulus A to stimulus B before the onset of the latter? In the example of Colour phi, the question is how can we observe the second colour before the onset of the second stimulus (Goodman, 1978)? What makes apparent motion unique is the involvement of time which is assumed to have a special ontological status (Phillips, 2014). While puzzling, spatial illusions and distortions can at least be described in terms of geometrical transformations (e.g., non-Euclidean geometry; Watson, 1978) and/or neural processes (lateral inhibition, excitation, and masking; Eagleman, 2001). By contrast, apparent motion presents a direct challenge to the notions of causality and temporal order. The philosophical implications of the paradox are serious. In order to explain it, one must acknowledge a fundamental disagreement between two descriptions of the phenomenon. This in turn opens up a wide explanatory gap which is made worse by the lack of understanding of the relationship between perception and consciousness. The assumption of simultaneity of perception and consciousness appears to break down and the two have to be treated as separate entities. ²

Dennett (1991) proposed two interpretations for the paradox—Stalinesque and Orwellian (see Akins, 1996 for a detailed analysis; Dennett & Kinsbourne, 1992). According to the former, ³ the perception of the two discrete events is veridical. It is the post-processing that occurs afterwards that somehow revises the perceptual input, generates the illusion, and presents it to consciousness. Since there is a delay between perception and conscious registration, both stimuli have been processed and the motion signal encoded by the time the scene reaches consciousness. One example of the Stalinesque account is Tye’s (2003) backward-looking model according to which, both events are perceived and processed before being presented to consciousness in a compressed form to create a momentary experience within a “specious” present. ⁴ This somewhat mysterious process ensures that the delay between the perception and conscious experience is cancelled out and the assumption of temporal order is maintained (see also Dainton, 2010).

By contrast, in the Orwellian scenario, the brain’s intervention takes place in memory. Veridical representations of the stimulus (a sequence of stationary light flashes) is encoded and stored in memory but the same form of perceptual tampering hypothesized above changes the original perceptual representation. This distorted representation is stored in memory erasing the original record of the event. One example of the Orwellian revision is Grush’s (2007) account of the cutaneous rabbit illusion (Geldard & Sherrick, 1972), which has the same postdictive character as the apparent motion phenomenon. ⁵

To summarize, according to the Stalinesque account, observers perceive, and according to the Orwellian, remember the illusion after a brief delay (of the order of hundreds of milliseconds), during which the brain supposedly engages in a series of operations which, for unknown reasons, change the veridical percept into an illusory one. The notion of backward “reconstruction” or postdiction (Eagleman & Sejnowski, 2000) is not a new one. In the introduction to his review of apparent motion literature, Kolers (1972) discussed Zeno’s paradoxes of motion thus:

One interpretation of Zeno is that the perception of motion is based not on the current sensory information, but on memory for position and time; hence on comparison, guess, or inference. This interpretation alleges that what our visual system actually detects are objects in different locations at different times; noting the disparity, we create a sense of motion to resolve it. Perception of objects, memory of their position, and delusion are therefore the main components, according to this theory, of our perception of motion. (p. 1)

This idea underpins the classical physicalist view according to which perception of motion is caused not by an object undergoing motion, but by the limitation of the mind which is incapable of processing the infinity of discrete spatial positions as slices of reality. According to this view (e.g., Helmholtz, 1855/1896), the mind fails to keep up with the veridical perceptual processing and consequently creates solutions which, in the final analysis, do not correspond with reality as defined by science. The question of why a static interpretation should be considered veridical given the facts of experience is left unanswered. Kolers stated that the discovery of apparent motion by the German physiologist Exner (1875) appeared to dispose of the notion of memory interference; successive events occur so rapidly and the perception of motion is so immediate that memory cannot be implicated in the process. Rather, in opposition to Helmholtz, Exner suggested that motion was a basic element of perception and cognition.

It is worth noting that in over a century of research there has been little progress in understanding perceptual illusions, especially the family of apparent motion phenomena. A number of theories have been put forward (see Arstila, 2015) and while differing in perspective and emphasis, they all face an unavoidable (and as yet unresolved) problem. In keeping with consciousness research, attempts at explaining the apparent motion paradox have only led to more questions. While this does not necessarily imply the futility of the quest, it does bring forth the question similar to that posed by some philosophers in the context of physics, namely, is there a point at which the enquiry is likely to reach an end? I propose that if not the end, then the end of the beginning might be reached not by a careful scientific examination of the relationship between the subjective and objective accounts, or by probing ever smaller chunks of matter using sophisticated technology, but by addressing the question that eludes most of the current theories, namely, why should the brain indulge in falsifying/modifying a veridical percept? What compels the brain to warp what is a simple physical event (two lights blinking in succession) into a mysterious causality-defying paradox? Further, if the brain indeed does this, why is it that the experimenter remains unaffected? For if one is subject to the cerebral subterfuge, how is one able to detect it in another person?

Irrespective of the hypothetical cause of apparent motion, each theory must explain the fundamental disagreement between the objective description (stationary lights flashing in succession) and the subjective perception of continuous motion. Apart from exceptions that challenge the veracity of subjective reports, no theory seems to disagree on the need for some form of retrospective inference, filling in, reconstruction, or completion on the part of the brain. The only point of contention is the precise mechanism responsible for this legerdemain. Most postdictive accounts of apparent motion (as well as theories of consciousness) contain at least one of the following: (a) a delay between perception and conscious awareness and/or (b) temporal disparity between the two. The former implies two commensurate, if misaligned, temporal frameworks running in parallel whereas the latter posits a fundamental disagreement between the two temporal metrics.

The idea of a temporal delay between sensory registration and consciousness has been prominent in psychology and neuroscience (Eagleman & Sejnowski, 2007; Libet, Wright, & Gleason, 1982). Although superficially, the presence of a brief delay does not seem to upset the common temporal order, reflection shows that this is not so, for if the observer experiences a delay between stimulus presentation and consciousness, what places the experimenter at a privileged vantage point from which they are able to detect it? If delayed consciousness is a universal property of minds, the experimenter’s mind should be subject to it and consequently unable to detect the delay in another individual. Alternatively, the idea creates infinite regress where everyone’s consciousness is delayed as soon as they exchange the role of experimenter for that of a subject. No appeal to precise experimentation or sophisticated mathematical analysis can side-step this problem.

What about temporal disparity so reminiscent of the relativistic Twin paradox? The idea here is that at least temporarily the temporal metrics of the experimenter and the participant are substantively different. For example, 500 ms on the experimenter’s clock is experienced as an instant by the participant. Different solutions have been suggested—from the warping of the mental time line or multiple drafts (Dennett & Kinsbourne, 1992) and time markers (Grush, 2004), to the contraction of an extended interval into a brief moment (Tye, 2003), different processing speeds for location and motion in the brain (Arstila, 2015), and quantum entanglement (Hameroff & Penrose, 2014). Perhaps the most enduring has been Libet’s (e.g., Libet, Wright, Feinstein, & Pearl, 1979) “backward referral” hypothesis according to which consciousness backdates its records of events which are initially registered subconsciously. The apparent paradox of backward causation (Churchland, 1981) generated a number of interpretations—from refutation (Pockett, 2002) to evidence for non-material mind (Eccles, 1985). Yet the same logic applies—if the observer’s cognition operates under a “different clock” even for an instant, what special quality enables the experimenter to observe the mismatch in someone else’s mind? One possibility is that motion perception is subject to a form of relativistic contraction of space–time, underpinned by a non-Euclidean geometry.

The study of spatial perception (Fernandez & Farell, 2009; Luneburg, 1947), Gestalt grouping phenomena (Arnheim, 1960), and optical illusions (Watson, 1978) suggests that Euclidean geometry cannot adequately describe the dynamic interplay between elements of a perceptual scene (Aksentijevic, Elliott, & Barber, 2001). ⁶ Thus, describing spatial relationships (global or local) within the framework of a non-Euclidean geometry could provide a better description of visual experience. While they require more complex mathematics, non-Euclidean geometries can offer a better fit between theory and data. The question arises as to why this should not be possible in the case of apparent motion. Although in vision, introducing non-Euclidean space involves computational complications, it complements and refines the existing linear models. In other words, visual grouping does not challenge the tenets of Newtonian physics concerning the relationship between time and space. Introducing non-linearity to explain apparent motion would involve contraction in space–time similar to that described by Einstein (1916/2001). A mathematical model of apparent motion by Caelli, Hoffman, and Lindman (1978) proposed a relativity-like Lorenzian space–time contraction of the experimental space–time to account for the inability of cortical firing to keep up with the speed of stimulus presentation. In what is essentially a neural Stalinesque model, the authors suggested that the non-linearity of apparent motion should be described by a positive-curvature elliptic geometry. Similarly, an approximation of Lorenzian contraction has been invoked to explain the cutaneous rabbit illusion (Goldreich & Tong, 2013). While both studies evoke a special-relativity-like framework to account for the temporal-order paradox, neither takes the crucial step, namely, treating the experimenter’s spatio-temporal metric as different from that of the observer. This of course would be necessary to explain why the observer’s time disagrees with that of the experimenter (and presumably other observers as well). The idea stretches the boundaries of credibility and even if it were to be successfully implemented, it would still have to explain which of the two reference frames was privileged and why.

Apparent motion represents a particularly vivid example of a lack of isomorphism between the objective, scientific worldview and its subjective counterpart because in addition to the mystery of non-existent motion it presents us with the puzzle of an event being perceived before it occurs. Since we cannot deny the reality of our perception, a rational explanation must involve some form of post-hoc reconciliation between its content and the objective description. Understandably perhaps, no theory so far has considered the simplest and most obvious possibility, namely, that at the bottom of it all there is no mystery—that the paradox of precognition reflects reality and that the experimental setup (and accompanying line of reasoning) used to elicit it, are somehow in error. I say “understandably” because the scientific paradigm offers the promise of a steady, fruitful progress. Starting from a set of observations and through abstract elaboration, reality is dissected into apparently more and more fundamental mechanisms and units. This form of reductionist description of the world frees the mind from having to consider inconvenient irregularities. In order to be successful, to move forward, a reductionist epistemology must appear self-contained and self-sufficient, that is, a closed sui generis system of knowledge which at least formally is independent of anything outside of its remit. It is this process of deliberate and presumably necessary isolation from a broader context that gives science its power. It also underpins the supposition that the scientific description of subjective experience is in some sense true. Arguably, as noted by Pippard above, it also robs science of the ability to explain first-person phenomena.

Let us examine this on the example of apparent motion. Wertheimer (1912/1961), who conducted the first systematic investigation and proposed the first theory of apparent motion, doubted that the sensory atomism or “elementarism” of von Ehrenfels, Wundt, and Titchener according to which perception and cognition were built from a set of interacting or “intermixing” fundamental sensations, could answer important questions about perception. This insight led him towards a new and highly influential approach to psychology, namely, Gestalt. The importance of Gestalt approach to the study of perception cannot be overestimated. It was the first (and most successful) modern attempt to highlight the importance of relationships between elements for perception. Wertheimer was right to doubt elementarism. What escaped his attention was that the perceptual and cognitive processes that were responsible for creating the experimental apparatus were exactly the same as those perceiving the illusion. In other words, there is no reason to believe that the scientific framework responsible for arriving at the objective description of the phenomenon (two stationary lights flashing in succession) was in any way different from or superior to, the subjective perception which it was attempting to study. Gradual and insidious separation and alienation of the two perspectives led to a situation in which subjective perception (which ultimately gave rise to the very experimental setup Wertheimer was using) assumed the mantle of objectivity and became its own investigator and judge. What appears to me as a trace of light moving from left to right and changing colour mid-way must appear so to Wertheimer too. Yet, he possesses an ostensibly superior grasp of the situation and knows (without perceiving) that the two lights are separated because he did not approach the problem from perception but from the analytical description of the experiment.

Under the weight of accumulated scientific knowledge and technological progress, the experimental setup was taken as the objective benchmark against which phenomenology was to be judged. Yet, the eyes (and more importantly, the brain) that perceive a travelling light smudge are the same as those that have created the stimulus presentation and measuring apparatus (rulers and clocks) used to investigate the illusion. How is it that these same eyes do not see the illusion while designing the experiment? If they did, how would they know that the lights were really separate? If they did, they would not have discovered the static nature of the display. What caused the same mind to come to view the world in two such mutually contradictory ways?

The solution

Although interesting in itself, apparent motion together with other temporal illusions reveals something important about the relationship between subjective experience and its scientific explanation. Apparent motion does not represent a pathway to understanding how the brain creates consciousness. Rather, it sheds light on the chasm that exists between the two ways of knowing, which has led to a great deal of conceptual confusion in consciousness research. The problem was stated succinctly and beautifully by Democritus: “According to common speech, there are colours, sweets, bitters; in reality however only atoms and emptiness. The senses speak to the understanding: ‘Poor understanding, from us you took the pieces of evidence and with them you want to throw us down? This down throwing will be your fall [emphasis added]’” (Dahlin, 2000, fragment 125).

Commonly, only the first sentence is quoted in physics textbooks. Viewed in isolation, it is supposed to demonstrate the sophistication of the great pre-Socratic philosopher, whose insight chimes in well with modern science. However, read in its entirety, Democritus’s statement reveals a deeper truth. Rather than an affirmation of the superiority of atomism, it can be seen as a melancholy realization that an unhealthy split between experience and abstraction was underway, as well as a warning issued to the “understanding.” There is no way of knowing if Democritus truly meant what he was saying or whether he was being ironic given that his words run counter to his philosophy, which saw true knowledge as being that achieved by means of intellect. Either way, the statement highlights one of the most pressing problems in epistemology and philosophy of science. My interpretation of Democritus’s words agrees with Dahlin’s: increasingly emboldened by its triumphs, understanding forgot its perceptual roots and turned on its progenitor. It started to view subjective experience as unreliable and fallible, constantly seeking ways to improve and ultimately explain it. Moving away from its origins, it accumulated power to the point of becoming the ultimate arbiter of reality, which was now described in terms of abstract equations and invisible forces. Subjective experience was ignored or even denigrated. Like a parvenu ashamed of their humble roots, abstract thought has projected its weaknesses and limitations back onto subjective perception (and ultimately consciousness) while refusing to question its own limitations or its indebtedness to experience. It has constructed theories and metaphors which ultimately justify its triumph. The experienced dynamic interplay of natural phenomena which evades mathematical description is treated as an inconvenience which obscures the truth—the universe as a void populated by particles, which on careful inspection are not even particles but probability waves or abstract 10-dimensional strings oscillating at the speed of light—or quantum foam. It appears that Democritus was aware of the point that tends to escape many modern thinkers. Abstraction might bring many advantages but it must also involve loss of information and ultimately, meaning. There is something paradoxical in the realization that a search for deepest meaning must result in the loss of meaning.

The disparity between the subjective and objective domains could be ignored for as long as the apparatus of science was directed towards ever more remote phenomena well outside human experience. The problem escalated the moment the mind became the central target of enquiry. Classical psychophysics established lawful relationships between physical and mental quantities. However, knowing that measures used in arriving at these originate in subjective experience, one can ask, at precisely what point did these same measures become “objective” and fit to assess subjective perception and cognition (from which they emerged)? Taken further, this line of reasoning suggests that psychophysical laws do not reflect the truth about the mind but describe its relationship with the Euclidean linear metric which itself represents the result of a long-term cleaving of (non-linear) perception and (linearized) measurement.

However, the tension between the first- and third-person perspectives is laid bare in the study of consciousness and more specifically studies of temporal order. It does not come as a surprise that today there are several theories of temporal order illusions ⁷ and very little consensus on what causes these illusions and why. After a century of effort by scientists and philosophers, we are no closer to understanding why two lights flashing sequentially produce an illusion of motion that challenges causality and the complexity of proposed explanations appears to be increasing. The reason is that researchers have been seeking the answer in ever smaller and more remote realms—from behaviour, brain hemispheres and regions, neural modules, assemblies and networks, to individual neurons and sub-neuronal structures, all the way to quantum fields. While requiring increasingly more abstract mathematics and more computational power in order to relate the data back to the original macroscopic observation, this kind of strategy cannot answer the obvious question of how consciousness, which enables us to perceive, think, feel, and create science, can be studied by an approach that denies it special status and treats it as an independent phenomenon, detached from everything including the conscious thoughts and actions of the scientist. Even if this were possible, how could it be achieved by methods which deliberately seek to disavow and “throw down” their subjective origins? ⁸

How does analytical dissection of experience create paradoxes? In the case of apparent motion, the principal culprit is discretization of time (and space), which was contemplated as early as Zeno’s time (Glazebrook, 2001). Subjective consciousness perceives time as continuous. ⁹ The need to represent time as a sequence of discrete intervals has little to do with objective truth and everything to do with the need to capture the fluid, ever-changing experiential reality within a static framework accessible to the limited cognitive grasp of the analyst. Decomposing motion into static frames or equations “freezes” the analogue, constantly changing phenomenon and preserves it in a form that permits further analysis and elaboration. At the same time the act of discretization deprives the resulting description of the dynamic essence of movement. Even if movement is reconstituted by a careful manipulation of static stimuli, this tells us nothing about why the stimuli should be causing the motion or how subjective perception is related to them—beyond the fact that motion is experienced. The correspondence between the two descriptions of the world has been lost, perhaps irretrievably.

The scientific study of consciousness is about providing static descriptions and models for phenomena that are inherently dynamic. Even this statement is potentially circular in the sense that the dynamic nature of perception and cognition can only be detected against a static background of a derivative scientific description. The subjective perception of motion eludes science in part because it contains information that cannot be captured by static mental and mathematical models. No equation can reconstruct the phenomenal essence of motion because it itself lacks that which it describes. While this is not obvious in the case of simple forms of real motion, it becomes fully evident in the case of apparent motion where the static description clearly fails to capture the quality of the phenomenal experience.

The resolution of the paradox lies in understanding and acknowledging the epistemological and methodological consequences of the gradual historical separation of the first- and third-person perspectives. ¹⁰ Over time, the crucial importance of subjective experience for the development of science (and mathematics) was gradually airbrushed out of the history of science in a manner one could perhaps label “Orwellian.” Increasing reliance on sophisticated apparatus and mathematical models pushed perception into the background to the point of turning it into a poor cousin of science. Perception did not fight back. It accepted its subservient status and allowed abstract theorizing and technological prowess to dominate the discourse. By the time Exner and Wertheimer began studying apparent motion, there was no way of “rewinding the film” and restoring the unity of subjective experience and scientific explanation. Fortunately, the unintended pun is appropriate here given that this was the time of the first flowering of cinema—the new art form based entirely on apparent motion (e.g., Steinman, Pizlo, & Pizlo, 2000; but see Hoerl, 2012). While the ability of film technology to reconstruct motion from static images hints at magic, this does not imply that the integrated experience of motion should be viewed as a flawed representation of a static reality (as proposed by Helmholtz, 1855/1896, and cinematographic models of perception; Dainton, 2010).

In common with other perceptual illusions, apparent motion seemed like a new and exciting avenue of psychological research. It is not surprising therefore that original researchers did not consider the possibility that the illusion did not lie in the percept but in the analytical mode of reasoning which disassembles experience into simple components and then attempts to reassemble it Humpty Dumpty-like. One hundred and three years after Wertheimer’s seminal investigation, perhaps the time has come to reassert the importance of subjective experience which gave Gestalt psychology its enduring appeal. There are no mysterious mechanisms that erase the veridical perceptual image only to replace it with a false memory of the event. Perception is not tampering with sensation in order to fulfil some hidden agenda. There are no covert processes and complex interactions that create temporal mirages or take the observer outside the common space–time reference frame. Instead, there is the reality of experience which precedes analysis and is superior to it in terms of abundance and quality of information. Paradoxes and distortions arise because the description of the world offered by the analytical mode of enquiry is quarantined from its origins and treated as the sole standard of truth. Once the artificial barrier between experience and scientific knowledge is removed, paradoxes disappear, the apparent motion effect is properly seen as a primary datum of experience and the objective stationary-lights explanation as a simplified description that lacks the wherewithal to reconstruct the percept. The key point is that this debility is self-imposed. A static, linear description of the world was created in order to overcome the impermanence and unpredictability of real experience which it was now supposed to study. What the static experimental setup (and the accompanying intellectual apparatus) gives the scientist in terms of controllability of the problem space and precision, it takes away in terms of ability to explain the richness of subjective experience.

Given that a great deal of scientific effort has been devoted to improving the precision of measurement relative to unaided perceptual judgement, it is not surprising that consciousness appears fuzzy, imprecise, and even paradoxical when subjected to the rigour of a physical experiment. It is equally unsurprising that the precise physical description of the apparent motion experiment does not accord with perceptual experience. Rather than asking why consciousness refuses to agree with the physical description of the phenomenon, we should ask why it is that the physical description does not behave in accordance with the perceptual experience. While, as demonstrated here, attempts at answering the first question trigger paradoxes, the answer to the second one is straightforward–if unappealing. Conscious perception is the beginning and end of science. If we did not implicitly accept the validity of our perception at the very beginning, scientific progress would be impossible. We must treat perception as veridical in order to record and describe different phenomena such as regularities in nature. This requirement becomes even more important when it comes to the two pillars of science, namely, systematic observation and construction of a measurement system. Doubting perceptual experience from the start would rob science of its claim to objectivity and reduce it to just one of a number of competing accounts of reality. Science is based on the conviction that our experience of patterns and regularities in nature corresponds in some way to the truth about the world. Yet, as Wertheimer and his colleagues demonstrated so forcefully, it is the mind that seeks out patterns and regularities in order to cope with the complexity of the environment. It follows that the discovery of troubling discrepancies between the subjective and objective descriptions of certain phenomena such as the apparent motion illusion brings into question the foundations of science. The dilemma is as follows: either conscious perception is untrustworthy in which case the validity of science must be in doubt (because science itself is based on subjective experience), or the discrepancy must be due to the divorce between science and its subjective origins. Since the first option is unpalatable, the second one seems unavoidable.

What are the consequences of such a conclusion? If the analytical framework of enquiry is flawed vis-à-vis consciousness, then reductionist/materialist theories must by definition be limited in their capacity to solve the hard problem (Chalmers, 1995). The move away from direct experience is always accompanied by information loss. The increase in precision and clarity at a lower level of explanation is bought at a cost. The rich, dynamic, contextual information is deliberately discarded in favour of austere abstract generalization. As the subjective and objective modes of knowledge drifted apart throughout history (Castillejo, 1982), this fact was either ignored or forgotten. If explaining apparent motion has proved virtually impossible, producing a successful scientific theory of consciousness must be that much harder. The “cause” or “source” of consciousness is unlikely to be found in some miraculous property of action potentials or neural networks. Nor can the ineffable conscious experience be created ex-nihilo through emergence of complex properties from an inert (in the sense of intentionality) substrate blindly obeying physical laws. In all cases, the explanans is too denuded of the qualities characterizing the explanandum.

In conclusion, I agree with Dennett and Kinsbourne (1992) that the temporal paradox exemplified by apparent motion is important—not because it tells us something substantive about how the brain works but because it exposes a serious epistemological problem at the heart of consciousness research. Paradoxes studied by psychologists do not reflect a conspiracy by the brain aimed at depriving us of reality but a generally ignored disengagement and alienation of subjective and objective domains that has been taking place for at least several centuries. Although this separation does affect natural science it is when the scientific apparatus is turned on consciousness that a deeper paradox is revealed. Scientific method which has done everything to purge itself from the vagueness and uncertainty of subjective experience from which it originates is used to explain that to which it is opposed in principle. Furthermore, and despite all methodological refinements, scientific enquiry originates and ends in subjective experience. Without implicit belief in the veridicality of perception, humans would never have undertaken the journey of exploration of the world around them. Consequently, if perception and cognition are flawed, science must be flawed too. The only way to avoid this meta-paradox is to realize that temporal illusions do not reflect some peculiarity of perception but the inability of the objective description to capture the quality of subjective experience. They are a testament to a long-term civilizational process that has secured the triumph of logic and science while creating a deep epistemological rupture. Perhaps, a better understanding of consciousness cannot be gained without understanding how and why this rupture occurred.