1. Introduction

To paraphrase the eloquence of Borges, we are made of time and consumed by time. The continuity of experience is one of the defining features, if not the defining feature of consciousness. To be more specific, as first explained by Karl Lashley, each thought or sensation is stable, but each is immediately present after the other, fully formed, with no experience of the underlying processing that led each to become conscious (Lashley, 1956).

Another way to think about the continuity is through a thought experiment of the inverse condition. Start by imagining a lower state of continuity, in which each individual thought is stable, but gaps are apparent. Each. Word. For. Example. In. This. Sentence. Is. Separated. The extra breaks affect your experience of reading the sentence, because you have to think about the flow of the words to get the meaning. Now jump to a complete loss of continuity. Each. Word. Is. Completely. Frozen. In. Time. For. A. Moment. Each. Pops. Into. Cognition. And. Is. Replaced. By. Another. Without temporal integrity, we are repeatedly frozen in time. Frozen memories can inform us where we've been, but not where we are going. We become biological computers without sentience.

The concept of separating information processing from sentience has been proposed in a much more colorful manner by David Chalmers, who describes a philosophical zombie that roughly appears to be human, but otherwise has no awareness (Chalmers, 1996). This is not as abstract as it sounds, because one aspect of this concept has been demonstrated by Hakwan Lau and Richard Passingham (Lau & Passingham, 2006). These researchers used a variant of the well-known paradigm of masking. In a simple version, subjects are briefly shown an image, known as the target, followed quickly by a second brief image, known as the mask, and if the timing falls into well-defined parameters, the target is eliminated from conscious awareness (Koch, 2004a). Lau and Passingham used a more complex version known as a type II metacontrast masking, but the underlying principle is the same, and they tested the accuracy of identifying the target, in this case a square or a diamond, followed by asking the subjects to press keys to indicate whether they actually saw the identity of the target or simply guessed what it was. By using different lengths of time between the presentation of the target and the mask, the researchers were able to identify two conditions in which the accuracy of identifying the target was statistically the same, but the subjective assessments of awareness were significantly different (Lau & Passingham, 2006). Cognition (defined as high-level biological computation) and awareness can be separated, although presumably only under certain circumstances and for brief periods of time.

Taken together, the thought experiment and the actual experiment suggest that the output of cognitive processing is transferred into consciousness in a continuous or seemingly continuous process. The normal limit for information transfer is the speed of light, which is clearly faster than human perception, but not instantaneous. True continuity would presumably require a mechanism based on quantum mechanics. The possibility that microtubules mediate coherent quantum states across large populations of neurons was proposed by Stuart Hameroff and Roger Penrose (1996). This hypothesis has received indirect support in recent years. Physicists at the University of Geneva, for example, demonstrated quantum entanglement by observing two-photon interferences well above the Bell inequality threshold (Salart et al. 2008), but this was with isolated pairs of photons. Moreover, physicists at the University of California at Santa Barbara were able to coax a mechanical resonator into two states at once, which showed for the first time that quantum events could be observed in complex objects, but this required cooling to near absolute zero (Cho, 2010; O'Connell et al. 2010). Notwithstanding this recent progress, whether microtubules, which undergo constant remodeling, can be islands of quantum events in the biochemical and electrical cauldron of the human brain at 37 degrees Celsius remains to be observed. I propose an alternative hypothesis that, rather than true continuity, consciousness is broken up into discrete cycles of cognition and that the sense of continuity is the result of determining the magnitude and direction of changes between cycles.

2. Consciousness Is Likely Discontinuous

Even though the experience of continuity is a defining characteristic of consciousness, it is likely an illusion. The experimental evidence for the discontinuity is largely based on the delay between sensory perception and conscious awareness. Briefly, the pioneering work on the delay was performed by Benjamin Libet, in which he and his colleagues showed an undetectable stimulus could become conscious after approximately 500 msec (Libet et al. 1964; Libet et al. 1967; Libet et al. 1991), but it is not clear whether the delay was due to the processing time to reach consciousness or the time for summation of the stimulus to reach threshold, and others have criticized Libet's conclusions (for example, see Gomes, 1998; Pockett, 2002). A better study was designed by Marc Jeannerod and colleagues, in which subjects were trained to grasp one of three dowels following the appropriate signal and performed the task with a reaction time of 120 msec. When the subjects were asked to verbalize when they first became aware of the signal, the response time was 420 msec, or 300 msec longer (Castiello et al. 1991). Even allowing 50 msec for the required muscle contraction for verbalization, the delay is still a quarter of a second (Koch, 2004b). Thus, according to this experiment, the frequency of cycles of cognition is roughly 4 per second. A larger and arguably more compelling body of evidence for the delay comes from the well-established phenomena of masking, as described above, in which a mask eliminates and replaces the awareness of a target. The elimination is not the result of interfering with the sensory input because if a second mask is presented, the first mask can be eliminated and the awareness of the target can be restored (Dember & Purcell, 1967). The elimination is only possible with a delay between sensory perception and consciousness. The delay, in turn, indicates that consciousness is discontinuous (Koch, 2004c; Libet, 1999).

3. Continuity and Consciousness Vectors

If consciousness is discontinuous, but appears to be continuous, then the problem of understanding consciousness becomes better posed: what creates the continuity? I propose that the sense of continuity is the result of determining the magnitude and direction of changes between cycles (Bodovitz, 2008). These putative consciousness vectors, however, are largely undefined. They presumably track the magnitudes and directions of multiple changes in parallel and/or in aggregate. Moreover, they presumably track changes in inherently qualitative information, such as words and concepts. While these open questions leave the key tenet of this hypothesis unsubstantiated, they create opportunities for breakthroughs by experts in advanced mathematics, physics and/or computer science. At the very least, efforts to model consciousness vectors may provide insights into the value of using the flow of information as feedback for better organizing complex and dynamic data.

The most significant substantiation of consciousness vectors is through analogy to motion vectors, which add motion to a series of otherwise discrete images. The standard speed for movies based on film is 24 frames per second, but rather than strobing, we perceive smooth motion. This is because motion vectors are calculated by the visual system, most likely in visual area V5, also known as visual area MT (middle temporal). Without motion vectors, vision becomes a series of still images, a condition known as akinetopsia or visual motion blindness (Shipp et al. 1994; Zihl et al. 1983; Zihl et al. 1991). The strobe effect makes otherwise simple tasks, such as crossing a street, extremely difficult. The cars are a safe distance away, then bearing down, without any sense of the transition. Even though there is memory of where the cars were, there is no sense of where they are going. Likewise, without consciousness vectors, simple cognitive tasks involving even a limited series of steps would be extremely difficult.

Motion vectors, like consciousness vectors, appear relatively simple at first approximation. In fact, the retina is arguably even a two-dimensional, Euclidian array of detectors, and most objects in motion follow standard trajectories. Yet, the exact neural computations to generate motion vectors have been difficult to determine and are the subject of decades of debate (for review, see Born & Bradley, 2005), although there is consensus that they involve the mapping of retinal activity onto higher-order visual processors such as those in V1 and V5. Thus identifying the calculations for motion vectors may provide the ideal model system for identifying the calculations for the more complex consciousness vectors.

4. Continuity and the Awareness of Change

If consciousness vectors and a sense of continuity are necessary for consciousness, then the corollary is that changes in cognition are necessary for awareness. This corollary is supported by the fact that we only see changes in our visual field. Even though an image may be static, our eyes never are, even during fixation. Our eyes are in constant motion with tremors, drifts and microsaccades. If these fixational eye movements are eliminated, then visual perception fades to a homogenous field (Ditchburn & Ginsborg, 1952; Riggs & Ratliff, 1952; Yarbus, 1967). The significance of these fixational movements for visual processing has long been debated, and no clear consensus has emerged (for review, see Martinez-Conde et al. 2004). The best correlation of neuronal responses to fixational eye movements are specific clusters of long, tight bursts, which might enhance spatial and temporal summation (Martinez-Conde et al. 2004); in addition, a more recent study showed that fixational eye movements improve discrimination of high spatial frequency stimuli (Rucci et al. 2007). But a lack of enhancement or improved discrimination does not explain the complete loss of visual perception. A better explanation is that changes in cognition are necessary for awareness.

5. Discussion

If cognition is broken up into discrete cycles and consciousness vectors create the illusion of continuity, then conscious feedback has pitfalls. It is slow, such that any action that occurs in less than approximately 250 msec will be over before reaching consciousness. Moreover, if events are happening too quickly and/or you are thinking too fast, your conscious feedback will miss changes in cognition, and, to make matters worse, you will have no immediate awareness of any deficiencies and will only be able to deduce the errors afterwards (see figure 1).

In addition, according to this theory, conscious feedback is inherently subtractive. The feedback tracks the magnitude and direction of what has already happened, thereby constraining the introduction of new ideas. All things otherwise being equal, turning down the feedback should inspire more creativity by allocating more energy to new information. Of course, all things otherwise being equal, without feedback, thoughts will be much more disorganized.

Ideally, these practical benefits are only the beginning. If consciousness vectors are real, and if we can begin to understand how they are calculated, we will have a much deeper knowledge of the highest functions of the human brain and possibly be able to apply our insights to artificial intelligence. Unlocking consciousness vectors may unlock human consciousness.

Acknowledgment: I would like to thank Aubrey Gilbert for her insights and careful review. Some of the material in this review was previously published in Bodovitz, 2008. Whereas the previous review included an argument about the possible localization of the brain region responsible for calculating consciousness vectors, this review is focused more on the significance of continuity and the role of putative consciousness vectors.