- Email: [email protected]
On Being Importantly Necessary for Consciousness
Consciousness and Cognition 8, 152–154 (1999) Article ID ccog.1999.0381, available online at http://www.idealibrary.com on
COMMENTARY On Being Importantly Necessary for Consciousness Valerie Gray Hardcastle1 Department of Philosophy, Virginia Tech, Blacksburg, Virginia 24060-0126; and Department of Philosophy, University of Cincinnati, Cincinnati, Ohio 45221-0374
Engel et al. (1999) claim that ‘‘appropriate synchronization among cortical neurons may be one of the necessary conditions for the buildup of perceptual states and awareness of sensory stimuli.’’ The question I wish to examine in this brief comment is exactly what they mean by a ‘‘necessary condition’’ for consciousness. For example, gravity is surely necessary for human consciousness, for without gravity life itself would not be possible. On the other hand, when singling out the items of study for consciousness in the brain, gravity does not rank high on the list of possible suspects since there are other things also necessary that somehow seem closer to being sufficient for consciousness as well. I want to examine where on the continuum of items of interest for the study of consciousness neuronal synchronization is likely to rank: Does it resemble gravity or does it resemble substance(s) C, the sufficient factor(s)? Let me put my query another way. When we devise scientific theories to explain phenomena in the world, we single out some causal relations among the vast web of interactions as the important connections for understanding some event. Some causal influences are too trivial to matter much in a general theory; others are too far removed from the phenomenon in question to fall within the scope of a finite theory. Under normal circumstances, UV radiation would fall in the former category for any general theory of consciousness and breathable air would fall into the latter. Though my skull is constantly bombarded by ultraviolet radiation, its effects on my conscious experiences are quite minimal. Hence even a complete theory of consciousness can ignore UV effects, for successful theories highlight only the most important components and interactions in getting to the specified end state. Mutatis mutandis, oxygen supports mammalian life and being alive is a prerequisite for my being conscious. However, that I and creatures like me are alive and consuming oxygen are not going to be facts included in a theory explaining our conscious awareness. We can take life sustained in part by oxygen as a given background assumption and build a theory of consciousness on top of such facts. The question is what sort of phenomenon is neural synchrony? Is it too trivial or too far removed from conscious processing to matter in theory, even if it is necessary? Or is it more important than that? Commentary on A. K. Engel, P. Fries, P. Ko¨nig, M. Brecht, and W. Singer (1999). Temporal binding, binocular rivalry, and consciousness. Consciousness and Cognition, 8(2), 128–151. This commentary is part of a special issue of this journal on Temporal Binding, with James Newman, guest Co-Editor. 1 To whom correspondence and reprint requests should be addressed. E-mail: [email protected]. 152 1053-8100/99 $30.00 Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
COMMENTARY
153
Interestingly enough, Engel and his colleagues give two answers to these questions in their discussion. When they outline the theoretical framework into which they plan to fit their data, it looks as though whatever necessary-for-consciousness role synchronous firing plays in the brain isn’t likely to fall within the scope of a theory explaining consciousness. Instead, it would appear in a neurophysiological theory of attention or memory or the like. However, when they describe the synchronous firing data themselves, it appears that these are much more likely to be relevant to a theory of consciousness per se. Needless to say, determining the appropriate scope of a theory and sorting out the precise relation between data and theory are important for advancing understanding. When discussing their theoretical orientation and assumptions, Engel et al. assert that most likely synchronous firing among neurons unites them as computation units. In particular, neural synchrony of the appropriate type is identical to feature binding in visual perception—or at least it is the neural correlate for it. They further hold that feature binding is necessary for consciousness because the act of bundling the primitive units of perception together into activated wholes allows us to become aware of external stimuli via some other mechanism later in the processing stream. Binding facilitates attention, working memory, structuring representations into coherent units, or some combination process created by bits and pieces of all three. These things—attention, working memory, and so forth—then set the stage of consciousness. If what Engel and his colleagues assume is correct, then synchronous neural firing is pretty clearly importantly necessary in our understanding of attention or working memory or structuring representations. That is, a good theory of attention in the brain will include neural synchrony as a variable because its causal influence is both nontrivial and proximate to the phenomenon. It is less clear how relevant neural synchrony would be to a theory of consciousness. If Engel’s assumptions are correct, then a theory of consciousness might refer to attention or working memory as a component of a consciousness-producing process (or it might not, depending upon the scope of the theory), but it need not then also include the neural underpinnings of attention or memory itself. That information might be too distal from the phenomenon to be explained to be relevant for the explanation. (Of course, the theory of consciousness might very well include the neural firing patterns underwriting attention or memory, depending on how it goes—I don’t want to prejudge anything here— but it isn’t obvious that it should.) In short: Engel et al.’s theoretical sketch of the proposed ties between neural synchrony and psychological events renders nebulous exactly how neural synchrony is supposed to fit into an explanation of consciousness. Their data provide a different story, however. If neural synchrony in areas 17 and 18 increases when an organism consciously perceives a stationary grating and decreases when it does not, even though it is still processing the stimulus visually, then what do we know? First, we know that the synchrony indexes consciousness itself. This fact is not enough for us to conclude that we have found the neural correlate of consciousness, since it might reflect the activity of some other mechanism importantly necessary and possibly sufficient for at least one type of visual perceptual awareness. Much in the same way that the falling of the penultimate domino is necessary and sufficient for the final domino topple in a domino chain, the neural synchrony Engel has uncovered might feed directly into whatever it is that produces consciousness.
154
VALERIE GRAY HARDCASTLE
Second, the mechanisms that Engel et al. list as being computationally relevant for consciousness to occur do not seem to have a place in their experiment. Cats were shown simple gratings in each eye. Whether they paid attention to the visual stimuli was not tested, though they need not in order to run the experiment. The most parsimonious assumption is that sometimes the cats were paying attention and sometimes they were not. Hence, any attentional effects would be washed out in averaging the brain waves as noise. Similarly, working memory need not be involved either in any interesting sense, for the cats do not have to recognize the gratings as gratings in order to perform their cognitive task (which is to see the gratings), nor do they have to remember the gratings for any length of time. They simply have to see the stimuli. Of course, it might be the case that consciousness mandates mnemonic processing (as I have argued that it does elsewhere (Hardcastle, 1995)). But, with the sort of evidence that Engel and his colleagues present, there is no reason to assume that it does. Finally, the cats need not structure how they represent the stimuli in any way to perform their task. They could see the gratings as a jumbled series of line segments, as a random set of pixels, or as a full-blown grate and still succeed. We cannot help but see a grate as a grate (and also interpret it as such), but how our visuocognitive processes compare to a cat’s is relatively unstudied. Assuming that attention, working memory, or structuring representations are necessary (or sufficient) for consciousness plays no role in understanding the experimental task or in interpreting the resultant data. Engel et al. conclude that their data indicate that ‘‘only strongly synchronized neuronal responses can contribute to awareness and conscious phenomenal states. . . . [A]ctivation of feature-detecting cells is, as such, not sufficient to grant access of the encoded information to consciousness. . . . Rather, to be functionally effective and to be selected for perception, neurons have to be strongly synchronized and bound into assemblies.’’ They are arguing that the correlate or the index of the engine driving sensory awareness is something importantly necessary for consciousness. Thus, it should be included in any brain-based theory of awareness. And indeed their data suggest these things. However, what is lacking is an appropriate framework for explaining consciousness into which we can fit these results, for the framework they use does not cohere with the data. We do know that the neural synchrony indexes conscious visual perception, but without an appropriate theoretical framework, we have no idea what that observation is supposed to explain about conscious processing. Ideally, whatever theoretical framework we adopt in science and whatever experimental data we uncover should mutually inform and support one another. We are not there yet in our attempts to integrate the phenomenon of neural synchrony into consciousness studies, for at best we have data searching for a theory. But without the theory, we cannot determine how important neural synchrony is in our understanding of conscious processes. It might be vitally important—the component C— or it may be more reminiscent of breathable air. The jury is still out. REFERENCES Engel, A. K., Fries, P., Ko¨nig, P., Brecht, M., & Singer, W. (1999). Temporal binding, binocular rivalry, and consciousness. Consciousness and Cognition, 8(2), 128–151. Hardcastle, V. G. (1995). Locating consciousness. Amsterdam: Benjamins.
COMMENTARY On Being Importantly Necessary for Consciousness Valerie Gray Hardcastle1 Department of Philosophy, Virginia Tech, Blacksburg, Virginia 24060-0126; and Department of Philosophy, University of Cincinnati, Cincinnati, Ohio 45221-0374
Engel et al. (1999) claim that ‘‘appropriate synchronization among cortical neurons may be one of the necessary conditions for the buildup of perceptual states and awareness of sensory stimuli.’’ The question I wish to examine in this brief comment is exactly what they mean by a ‘‘necessary condition’’ for consciousness. For example, gravity is surely necessary for human consciousness, for without gravity life itself would not be possible. On the other hand, when singling out the items of study for consciousness in the brain, gravity does not rank high on the list of possible suspects since there are other things also necessary that somehow seem closer to being sufficient for consciousness as well. I want to examine where on the continuum of items of interest for the study of consciousness neuronal synchronization is likely to rank: Does it resemble gravity or does it resemble substance(s) C, the sufficient factor(s)? Let me put my query another way. When we devise scientific theories to explain phenomena in the world, we single out some causal relations among the vast web of interactions as the important connections for understanding some event. Some causal influences are too trivial to matter much in a general theory; others are too far removed from the phenomenon in question to fall within the scope of a finite theory. Under normal circumstances, UV radiation would fall in the former category for any general theory of consciousness and breathable air would fall into the latter. Though my skull is constantly bombarded by ultraviolet radiation, its effects on my conscious experiences are quite minimal. Hence even a complete theory of consciousness can ignore UV effects, for successful theories highlight only the most important components and interactions in getting to the specified end state. Mutatis mutandis, oxygen supports mammalian life and being alive is a prerequisite for my being conscious. However, that I and creatures like me are alive and consuming oxygen are not going to be facts included in a theory explaining our conscious awareness. We can take life sustained in part by oxygen as a given background assumption and build a theory of consciousness on top of such facts. The question is what sort of phenomenon is neural synchrony? Is it too trivial or too far removed from conscious processing to matter in theory, even if it is necessary? Or is it more important than that? Commentary on A. K. Engel, P. Fries, P. Ko¨nig, M. Brecht, and W. Singer (1999). Temporal binding, binocular rivalry, and consciousness. Consciousness and Cognition, 8(2), 128–151. This commentary is part of a special issue of this journal on Temporal Binding, with James Newman, guest Co-Editor. 1 To whom correspondence and reprint requests should be addressed. E-mail: [email protected]. 152 1053-8100/99 $30.00 Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
COMMENTARY
153
Interestingly enough, Engel and his colleagues give two answers to these questions in their discussion. When they outline the theoretical framework into which they plan to fit their data, it looks as though whatever necessary-for-consciousness role synchronous firing plays in the brain isn’t likely to fall within the scope of a theory explaining consciousness. Instead, it would appear in a neurophysiological theory of attention or memory or the like. However, when they describe the synchronous firing data themselves, it appears that these are much more likely to be relevant to a theory of consciousness per se. Needless to say, determining the appropriate scope of a theory and sorting out the precise relation between data and theory are important for advancing understanding. When discussing their theoretical orientation and assumptions, Engel et al. assert that most likely synchronous firing among neurons unites them as computation units. In particular, neural synchrony of the appropriate type is identical to feature binding in visual perception—or at least it is the neural correlate for it. They further hold that feature binding is necessary for consciousness because the act of bundling the primitive units of perception together into activated wholes allows us to become aware of external stimuli via some other mechanism later in the processing stream. Binding facilitates attention, working memory, structuring representations into coherent units, or some combination process created by bits and pieces of all three. These things—attention, working memory, and so forth—then set the stage of consciousness. If what Engel and his colleagues assume is correct, then synchronous neural firing is pretty clearly importantly necessary in our understanding of attention or working memory or structuring representations. That is, a good theory of attention in the brain will include neural synchrony as a variable because its causal influence is both nontrivial and proximate to the phenomenon. It is less clear how relevant neural synchrony would be to a theory of consciousness. If Engel’s assumptions are correct, then a theory of consciousness might refer to attention or working memory as a component of a consciousness-producing process (or it might not, depending upon the scope of the theory), but it need not then also include the neural underpinnings of attention or memory itself. That information might be too distal from the phenomenon to be explained to be relevant for the explanation. (Of course, the theory of consciousness might very well include the neural firing patterns underwriting attention or memory, depending on how it goes—I don’t want to prejudge anything here— but it isn’t obvious that it should.) In short: Engel et al.’s theoretical sketch of the proposed ties between neural synchrony and psychological events renders nebulous exactly how neural synchrony is supposed to fit into an explanation of consciousness. Their data provide a different story, however. If neural synchrony in areas 17 and 18 increases when an organism consciously perceives a stationary grating and decreases when it does not, even though it is still processing the stimulus visually, then what do we know? First, we know that the synchrony indexes consciousness itself. This fact is not enough for us to conclude that we have found the neural correlate of consciousness, since it might reflect the activity of some other mechanism importantly necessary and possibly sufficient for at least one type of visual perceptual awareness. Much in the same way that the falling of the penultimate domino is necessary and sufficient for the final domino topple in a domino chain, the neural synchrony Engel has uncovered might feed directly into whatever it is that produces consciousness.
154
VALERIE GRAY HARDCASTLE
Second, the mechanisms that Engel et al. list as being computationally relevant for consciousness to occur do not seem to have a place in their experiment. Cats were shown simple gratings in each eye. Whether they paid attention to the visual stimuli was not tested, though they need not in order to run the experiment. The most parsimonious assumption is that sometimes the cats were paying attention and sometimes they were not. Hence, any attentional effects would be washed out in averaging the brain waves as noise. Similarly, working memory need not be involved either in any interesting sense, for the cats do not have to recognize the gratings as gratings in order to perform their cognitive task (which is to see the gratings), nor do they have to remember the gratings for any length of time. They simply have to see the stimuli. Of course, it might be the case that consciousness mandates mnemonic processing (as I have argued that it does elsewhere (Hardcastle, 1995)). But, with the sort of evidence that Engel and his colleagues present, there is no reason to assume that it does. Finally, the cats need not structure how they represent the stimuli in any way to perform their task. They could see the gratings as a jumbled series of line segments, as a random set of pixels, or as a full-blown grate and still succeed. We cannot help but see a grate as a grate (and also interpret it as such), but how our visuocognitive processes compare to a cat’s is relatively unstudied. Assuming that attention, working memory, or structuring representations are necessary (or sufficient) for consciousness plays no role in understanding the experimental task or in interpreting the resultant data. Engel et al. conclude that their data indicate that ‘‘only strongly synchronized neuronal responses can contribute to awareness and conscious phenomenal states. . . . [A]ctivation of feature-detecting cells is, as such, not sufficient to grant access of the encoded information to consciousness. . . . Rather, to be functionally effective and to be selected for perception, neurons have to be strongly synchronized and bound into assemblies.’’ They are arguing that the correlate or the index of the engine driving sensory awareness is something importantly necessary for consciousness. Thus, it should be included in any brain-based theory of awareness. And indeed their data suggest these things. However, what is lacking is an appropriate framework for explaining consciousness into which we can fit these results, for the framework they use does not cohere with the data. We do know that the neural synchrony indexes conscious visual perception, but without an appropriate theoretical framework, we have no idea what that observation is supposed to explain about conscious processing. Ideally, whatever theoretical framework we adopt in science and whatever experimental data we uncover should mutually inform and support one another. We are not there yet in our attempts to integrate the phenomenon of neural synchrony into consciousness studies, for at best we have data searching for a theory. But without the theory, we cannot determine how important neural synchrony is in our understanding of conscious processes. It might be vitally important—the component C— or it may be more reminiscent of breathable air. The jury is still out. REFERENCES Engel, A. K., Fries, P., Ko¨nig, P., Brecht, M., & Singer, W. (1999). Temporal binding, binocular rivalry, and consciousness. Consciousness and Cognition, 8(2), 128–151. Hardcastle, V. G. (1995). Locating consciousness. Amsterdam: Benjamins.