When investigating the neural correlates of consciousness, neuroscientists distinguish between ‘‘conscious state’’ (being awake as opposed to asleep or in a coma), which is regulated by brainstem and thalamic nuclei, and ‘‘conscious representation’’ (awareness of specific phenomenal experience). The advent of functional brain imaging techniques, especially fMRI, enables the noninvasive inquiries of the mechanisms underlying conscious experience, particularly in the visual system. To date, most experimental paradigms designed to study consciousness contrast the response during conscious visual experience with the response during unconscious visual experience (the so-called blindsight phenomenon), or record patterns of brain activation during binocular rivalry, perception of bistable figures, and visual mental imagery. The current data suggest that activity in high-level areas of the ventral visual pathway, but not in V1, are necessary for conscious visual experience. Moreover, visual awareness requires parietal and prefrontal regions (for a recent review, see Rees, Kreiman, & Koch, 2002). The neural correlates of conscious vision, therefore, parallel the distributed cortical networks that modulate visual attention and visual imagery. As most researchers confuse ‘‘awareness’’ with ‘‘consciousness,’’ the reported differential activity during consciousness is currently indistinguishable from that of other higher cognitive functions. In his article, ‘‘Functional fMRI and the Study of Human Consciousness,’’ Dan Lloyd uniquely combines a conceptual analysis of consciousness with neuroscientific methods, in order to characterize the neural manifestations of consciousness (Lloyd, 2002). Lloyd adopts Husserl’s criteria, according to which the phenomenology of consciousness is based on three essential principles: intentionality (the external world as it is experienced and not as it is); superposition (sensory and nonsensory properties are present in perception); and temporality (all objects share perception of present, past, and anticipated future). If indeed these aspects of consciousness are implemented in the brain, the empirical evidence should include temporal flux (with passing time, the multivariate differences between images should increase) and superposition (images sharing task or stimulus conditions should be similar). Lloyd’s methodological approach includes three constraints. First, time points in a scan series are considered individually, because temporality implies that consciousness at each point in time is distinct from the preceding and the proceeding points. Second, subjects are considered individually, because intersubject averaging could eliminate individual expression of consciousness. Finally, brain states are considered globally, seeking distributed patterns of activation that encompass large cortical areas, rather than assuming localized responses. To test his predictions, Lloyd reanalyzes four data sets (Hazeltine, Poldrack, & Gabrieli, 2000; Ishai, Ungerleider, Martin, & Haxby, 2000; Mechelli, Friston, & Price, 2000; Postle, Berger, Taich, & D’Esposito, 2000) provided by The fMRI Data Center. The studies, published in the December 2000 issue of the Journal of Cognitive Neuroscience, included a variety of cognitive tasks (target tracking, passive viewing, delayed matching, reading, and spatial working memory), stimuli (faces, objects, words, pseudowords, 2-D arrays of squares, colored circles), and motor responses (button presses and saccades). Needless to say, none of the original studies was designed for or aimed at underpinning the neural mechanisms of ‘‘neurophenomenology.’’ Nevertheless, Lloyd preprocesses and reanalyzes the raw data to test his predictions about the general structures of consciousness, which by their nature are task and stimulus independent. Using multivariate distance analysis and artificial neural networks, he shows a time–distance effect (i.e., as the time series progressed, the distance between images increased) and that the past and future brain states, retention and protention, respectively, are embedded in present brain states. As time passes, suggests Lloyd, the brain is changing ‘‘globally, incrementally, and monotonically.’’ Previous fMRI studies of consciousness compared one state of awareness with another, assumed localization, and ignored the temporal flux. Lloyd’s original approach proposes methodological and conceptual advantages. Traditionally, fMRI data analysis focused on two parameters, namely the spatial extent of the activation and the amplitude of the response within an activated region. The data are usually displayed as statistical maps indicating the location and size of significant activation, and graphs or histograms showing the percent fMRI signal change. Given the spatial and temporal resolution of the technique, extracting temporal information about the ‘‘tripartite temporality’’ (i.e., the experienced present of National Institutes of Health