Chapter 13: The Consciousness Hypothesis

[hypnagogue]

Having constructed a new theory of causation that culminated with the introduction of the Central Thesis in chapter 12, we are in position to offer a promising new account of what human consciousness is and what causal relationships it shares with the physical. Recall the definition of the Central Thesis:

The Central Thesis: Things in the world are natural individuals if, and only if, they are capable of experiencing phenomenal individuals. (p. 241)

According to the Central Thesis, the existence of subjects that experience phenomenal properties is coextensive with the existence of naturally individuated causal systems (as discussed in chapters 9-11). This thesis thus creates a rudimentary kind of bridging law between causal systems and experiencing subjects: Phenomenally speaking, all natural individuals are subjects of experience, and causally speaking, all subjects of experience are natural individuals. If this thesis is right, then whenever we can ascertain the existence of a natural individual, we can infer the existence of a corresponding experiencing subject; likewise, whenever we can ascertain the existence of an experiencing subject, we can deduce the existence of a corresponding natural individual.

Of course, one can readily ascertain the existence of an experiencing subject in one's own first person case, and one can reasonably judge other cognitively normal humans to be experiencing subjects as well. Thus, an application of the Central Thesis predicts that some causal system in the normally functioning human brain is a natural individual whose existence corresponds to the bounded qualitative field that constitutes a human's first person subjective experience. This prediction is stated with more detail and rigor in the Consciousness Hypothesis:

The Consciousness Hypothesis: Each individual consciousness carries the nomic content [i.e., the effective and receptive properties] of a cognitively structured, high-level natural individual. Conscious experience is experience of the total constraint structure active in the receptive field of such an individual. (p. 248)

A word of caution
Note that the term "consciousness" is being used here to refer only to the rich, intentionality laden, cognitively structured kind of phenomenal experience with which humans are acquainted. The Central Thesis does predict that even extremely simple causal systems that cannot plausibly support complex cognitive functions such as thought, representation, and so on should be experiencing subjects in some minimal sense, provided that they are natural individuals. For instance, if an electron is a natural individual, it follows from the Central Thesis that in some very minimal and alien sense, it is like something to be an electron. But by Rosenberg's lights, the phenomenal experience of such an unsophisticated, low-level natural individual would be "pure feeling too simple to support anything worthy of the name 'consciousness'" (p. 248). (For more on experiencing without cognition, see chapters 5 and 6.)

Thus, on this reading, "consciousness" is not a catch-all term for all forms of subjective experience, but only picks out a subset of possible kinds of phenomenally experiencing systems. To use Ned Block's terminology, one might say that the Consciousness Hypothesis applies to systems that feature both phenomenal consciousness and access consciousness, whereas the Central Thesis deals more generally with systems that have some form of phenomenal consciousness but not necessarily anything like access consciousness.

With the Consciousness Hypothesis in place, we are ready to reap the payoff from our work on causation. We have created a new metaphysical context within which to place human consciousness, and this new place for consciousness will allow us to redress the many conceptual puzzles, mysteries, and paradoxes that swirl around it and make it so profoundly problematic. In particular, the key features of our revised causal picture of the world that will pay big dividends are receptivity and the notion of phenomenal/experiential carriers. The introduction of receptivity into the causal story provides new layers of causal structure with which we can explain some of consciousness's more puzzling structural aspects and reconcile them with the apparently incompatible physical structure of the brain. The casting of phenomenal and experiential properties as carriers for effective and receptive properties will allow us to coherently address the fundamental mystery of why a physical brain is even associated with an experiencing subject in the first place, and will furthermore allow us to talk about the nature of the causal relationship between the two without lapsing into epiphenomenalism or interactionism.

In part I of this book, we explored some of the problematic aspects of consciousness in detail. We argued that consciousness cannot be satisfactorily accounted for by physicalism, and that rejecting physicalism is likely to lead to some form of panexperientialism. We went on to examine a number of conceptual puzzles and paradoxes about consciousness and its relationship to the physical, including a puzzle about the unity of consciousness, the paradoxical simultaneity of the subjective instant, the knowledge paradox, the apparent superfluity of consciousness, the grain problem, and the boundary problem.

In this chapter, we will revisit each of these tensions and puzzles from the point of view of the Consciousness Hypothesis. With our new metaphysical framework in tow, we will be able to resolve some of these seemingly intractable problems outright and garner significant new insights into others. In fact, not only will be able to melt away many of the fundamental mysteries surrounding consciousness, but we will also be able to comprehend why those particular kinds of mysteries have arisen at all; as Rosenberg puts it, "The end result is the striking discovery that the fundamental carriers of effective and receptive causation would have predictable properties that parallel the troublesome properties of consciousness" (p. 253). In sum, we will be able to weave once hopelessly fractured and disjoint concepts about consciousness and the physical into a single, neat, coherent tapestry. By demonstrating the new explanitory powers granted to us by the Consciousness Hypothesis, we will have shown it to be a fruitful thoeretical tool in redressing and unifying our concepts about the subjective and objective aspects of nature, and this fruitfulness will in turn lend further support and credibility to the Liberal Naturalist paradigm we have thus far outlined. (Rosenberg names this overarching framework-- including the Theory of Causal Significance, the Carrier Theory of Causation, the Central Thesis and the Consciousness Hypothesis-- the Theory of Natural Individuals. For convenience, I will refer to the Theory of Natural Individuals as "TNI" for the remainder of this summary.)

A High Level Sketch

It is worth taking a moment to reflect on the kind of metaphysical picture that the Consciousness Hypothesis paints. A useful way of thinking about the Consciousness Hypothesis is in terms of internal and external points of view. The Consciousness Hypothesis roughly states that the subjective, conscious mind and some subsystem of the objective, physical brain are not separate entities, but rather are the same thing as seen from two different vantage points. Roughly, the conscious mind is what some portion of the brain looks like "from the inside," and likewise, some portion of the brain is what the conscious mind looks like "from the outside."

The internal point of view is privileged in the sense that it allows direct knowledge of the intrinsic natures of an individual's carriers. Unfortunately, the only way to 'see' a natural individual from the inside is to actually be that individual, since phenomenal properties can only be experienced by the single experiencing subject (the single instance of receptivity) within which they are bound. For this reason, the nature of an individual's intrinsic carriers is lost to the external view. From the outside, all we can see is the structured system of abstract relationships into which an individual's phenomenal properties enter. These abstract relationships, of course, are nothing but the effective properties bound within the individual, and effective properties are exactly the objects of study for the sciences. Thus, much of the individual's substantial inner being is opaque to third person investigation. Although it can teach us much, the external, third person view of an individual provides us an impoverished, abstract, indirect, and ultimately incomplete way of understanding it. (For a discussion on why effective properties can be observed from the third person, see the section titled 'effective properties' in chapter 9. This discussion should make it clear why receptivity and carriers are not likewise amenable to third person investigation.)

Avoiding the Failures of Physicalism

There are a number of arguments against the thesis that consciousness supervenes upon the physical, including the logical possibility of inverted spectra, the logical possibility of zombies, the epistemic asymmetry between facts about phenomenal consciousness and other facts, the knowledge argument, and the absence of analysis. Each of these arguments points to a systematic, in principle failure of explanation of experiential facts using only physical facts. Turning these arguments against TNI serves as a healthy check to ensure that TNI is not plagued by the same fundamental explanitory shortcomings of physicalism, with respect to phenomenal consciousness.

The absence of analysis argument and the knowledge argument essentially boil down to the objection that phenomenal consciousness cannot be fully analyzed into the schematic structural and functional facts available to physicalism. This objection is not at odds with TNI, since the phenomenal and experiential carriers of the world's effective and receptive properties are explicitly understood to evade complete systematic analysis. In the terminology of chapter 12, these carriers are extrinsic within all schematic systems, and so cannot be fully characterized by merely schematic facts. So TNI's ontological building blocks are not limited to structural and functional facts, but also include fundamental intrinsic facts-- and indeed, these intrinsic facts figure prominently into the TNI account of consciousness.

The argument from epistemic asymmetry observes that our only reason for believing in consciousness in the first place is our private, subjective experiences rather than the public, external facts available from a physicalist perspective. Since third person facts give us no reason to believe in consciousness, and since physicalism only endows us with third person facts, physicalism seems to overlook a crucial piece of evidence regarding consciousness. However, evidence from subjective experience is not neglected by TNI, but rather plays a significant role in its construction. Recall that when we argued in chapter 12 that phenomenal and experiential properties could be pressed into service as the world's fundamental, intrinsic carriers, we justified the argument by analogical appeal to features of consciousness that can only be observed from the first person.

The argument from the logical possibility of zombies asserts that physicalism fails because a complete physicalist account of the world is logically consistent with the absence of phenomenal consciousness. The zombie argument fails against TNI because phenomenal consciousness is built into it from the ground floor by the Central Thesis (and, by extension, the Consciousness Hypothesis). A complete TNI account of the world will explicitly include facts about phenomenal and experiential carriers, and thus will not be logically consistent with the absence of phenomenal consciousness.

The inverted spectrum argument claims that the physical facts about our brains are logically consistent with an inversion of the qualitative color spectrums we experience in the first person. If it is true that physical facts do not logically fix phenomenal facts, then any attempt at an explanitory reduction of phenomenal facts to physical facts is undermined. However, TNI does not attempt to reduce phenomenal facts to physical facts, but rather explains physical facts in terms of the internal contrasts that obtain among fundamental phenomenal facts. So, at worst, the logical possibility of inverted spectra entails that diverse sets of phenomenal facts might realize the same set of physical facts. This presents possible epistemological dangers to TNI in trying to deduce phenomenal facts from physical facts, but does not present grave metaphysical problems as it does to physicalism.

The arguments against physicalism all fail to gain traction against TNI. Physicalism runs into difficulties trying to explain phenomenal consciousness because it casts only schematic, third person facts as the world's fundamental building blocks, and then tries to reduce the apparently intrinsic properties of phenomenal consciousness to these schematic facts. The arguments against physicalism seem to be variations on the theme that you can't derive intrinsic facts from extrinsic ones. TNI evades this conceptual difficulty by turning the physicalist approach on its head. Where physicalism takes schematic facts to be fundamental and tries to explain facts about consciousness in terms of such schematic facts, TNI explicitly acknowledges the intrinsic facts about phenomenal and experiential carriers to be the most basic components of the world's ontological tool kit, and seeks to explain the world's schematic causal facts in terms of the inherent relationships that hold among these carriers. While physicalsim seems to lack an appropriate metaphysical account of how it could be that intrinsic or phenomenal facts can be grounded upon schematic facts, TNI features a plausible analysis by which the intrinsic properties of the carriers can necessarily determine the various extrinsic relationships that hold among them (see chapter 12).

 

[hypnagogue]

summary, part 2

Resolving the Puzzles, Paradoxes, and Tensions

More in-depth analyses of the following conceptual difficulties can be found in earlier chapters. For more on the boundary problem, see chapter 4; for more on panexperientialism, see chapters 5 and 6; for the rest of the topics discussed here, see chapter 7.

Ubiquity and Fundamentalness

Chapters 5 and 6 argued that if physicalism is false, then it is likely that p-consciousness is a fundamental and surprisingly ubiquitous aspect of the world that can only be understood by 'getting under' physics. All of these expectations are met by TNI.

In chapter 12, we identified a conceptual need for ultimate carriers that would be extrinsic within all schematic systems. Such carriers would be fundamental properties in nature and would 'get under' physics by providing its schematic facts with a categorical, intrinsic basis. We went on to argue that phenomenal and experiential properties are ultimately the things that play this fundamental carrier role in nature, thus fulfilling our earlier expectations.

The Central Thesis implies that phenomenal experience is not only surprisingly widespread in nature, but that it is literally ubiquitous, since wherever there is causation (in the strict sense defined by the Theory of Causal Significance) there is also phenomenal experience. The relevant question for TNI, then, is not "Under what conditions do experiencing subjects 'spark' into existence?", but rather, "How are the causal boundaries delineating the boundaries of experiencing subjects drawn in nature?"

For instance, according to TNI, even a simple and innocuous physical object like a rock is most definitely associated with phenomenal experience in at least one sense. We should expect that the world's fundamental physical entities are natural individuals (see chapter 9), and as such it follows from the Central Thesis that the fundamental physical constituents of rocks are bounded subjects of experience, albeit only in some unimaginably minimal and alien sense that we can only understand by way of analogy (see chapter 5). (Note that this analysis extends to the fundamental microphysical constituents of any physical system.) However, it is probably not plausible that a whole rock is itself a high-level natural individual, and so it is not plausible that a rock as a whole supports the existence of a high-level subject of experience. For rocks (and probably most other physical systems), experiencing subjects exist only on the microphysical level of causation.

By contrast, according to the Consciousness Hypothesis, the causal structure of a normally functioning human brain supports the existence of a high-level, cortical natural individual. The causal boundaries of this high-level natural individual trace the phenomenal boundaries of a high-level subject of experience which experiences the causal interactions of billions of neurons as a single, 'macroscopic' experiential unit.

The Unity of Consciousness

Intuitively, the phenomenal qualities experienced in consciousness seem to have a kind of holistic dependence on the whole qualitative field. For instance, it seems as if a patch of phenomenal red in the qualitative visual field cannot exist outside of the context of the whole experiencing subject. Contrast this with a brick in a wall; although the brick helps to constitute the wall, its existence by no means depends on the context afforded by the wall as a whole. So it seems as if consciousness is not merely a conglomerate of various phenomenal properties in the same way that a brick wall is just a conglomerate of bricks; with phenomenal consciousness there seems to be some substantial condition of dependence existing between the parts and the whole, a kind of metaphysical unity that eludes straightforward reductive analysis. This intuition about the unity of consciousness presents us with a challenge to articulate more clearly what is meant by 'the unity of consciousness,' and a problem of how to reconcile the unity of p-consciousness with the apparently reductive, 'brick in the wall' character of the brain's physical structure and behavior as seen from the third person.

Using TNI, we can appeal to the nature of causation to characterize the unity of consciousness more fully and explicitly. Recall from chapter 10 that a natural individual's effective properties are identified with the constraints that its effective states place upon the possible joint states of a causal nexus. Thus, effective properties require the context of a causal nexus in order to exist; without a nexus, there can be no placement of causal constraint, and thus there can be no effective properties. There is thus a kind of holistic interdependence, or compositional circularity, that obtains between effective properties and the receptivity by which they are bound together into a common causal nexus. If we now recall that phenomenal properties carry effective properties and experiential properties carry receptivity, we can see that the same compositional circularity must hold between phenomenal and experiential properties. Without the causal context afforded by a whole experiencing subject (receptivity), there can be no phenomenal (effective) properties.

If the above is an adequate explanation of the phenomenal unity of p-consciousness, then how can it be that when p-consciousness is viewed from an external point of view as a brain, it seems to have an entirely reductive, conglomerate character? The answer lies in the structure of natural individuals. High-level individuals consist of a collection of lower-level individuals that are bound together by a common receptivity. A higher-level individual X cannot be completely reduced to the lower-level individuals bound within it, since such an analysis would fail to take into account X's receptive structure. Nonetheless, X is at least partly reducible to its lower-level constituents. If X were to be observed from the third person, the lower-level individuals directly and indirectly bound to it would stand out as apparently independent entities, whereas their status as components of a receptively bound, holistic causal unit would be far more subtle.

The Subjective Instant

There seems to be a mismatch between the temporal ordering and duration of events as experienced in consciousness and the temporal ordering and duration of events as measured in the brain by third person methods. How can the temporal structure of the two points of view be reconciled? Rosenberg acknowledges that addressing this problem would require a full treatment of spacetime within TNI, but offers a provisional sketch of how the problem might be solved using the idea that spatiotemporal structure can be reduced to receptive structure, as discussed in chapter 10.

In chapter 10, it was argued that time might be constructed from a series of asymmetrically connected individuals. For a high-level cortical individual C in a human brain, then, each subjective instant would correspond to C's reception of effective constraint within an asymmetrically structured nexus. However, note that bound within C would be a large and complex series of lower-level individuals, each of whose own causal properties would be largely masked from the point of view of C. (See the discussion on the grain problem below for an explication of why this would be the case.) The number and complexity of the receptive connections linking these lower-level individuals would likely ensure that, on each successive lower level, there is a temporal structure that is differently structured and finer-grained than on the level above. Thus there could be no clean mapping from the high-level subjective instant to lower-level instants; what is a single instant on a high level would likely be a multiplicity of instants from the point of view of a lower level. A paradox does not ensue since, on this proposal, there would not be one overarching set of temporal relations obtaining among all individuals. Rather, it would be more accurate to say that each causally delineated level of nature has its own semi-insular temporal structure, such that measures of temporal order and duration across various levels would not necessarily agree.

The Knowledge Paradox and Knowledge by Acquaintance

The knowledge paradox arises from a tension between the knowledge claims we make about consciousness and the causal role that consciousness plays in producing such claims. It seems that phenomenal consciousness cannot be attributed causal responsibility for the formation of the brain events that ultimately cause us to utter statements about phenomenal consciousness. If this is so, then it seems our knowledge claims about consciousness cannot be justified. Even on the TNI account, it seems at first glance that causal responsibility can only be attributed to effective properties, and thus the caual role of p-consciousness as a carrier would not be the sort of causal role that could justify claims made about it.

However, questions of what sorts of things can accrue causal responsibility are subtle, even for more prosaic matters. For instance, we regularly attribute causal responsibility to negative facts, e.g. "Bob's disappointment was caused by his girlfriend's failure to show up for their date." Negative facts are not physical things, and so if we attribute causal responsibility to them, it follows that causal responsibility need not hang from physical causal powers. If this is the case, then phenomenal and experiential carriers might also earn causal responsibility for our utterances about consciousness, even though their qualitative natures do not directly participate in the effective dynamics of the brain.

Even if we deny that negative facts can have causal responsibility, they still seem to figure into the story by providing reasons for, and by justifying the occurrence of, certain events ("Bob was disappointed-- and rightfully so-- because his date didn't show up"). In this case, it follows that reasons need not be causes and that facts can justify mental states even if they don't cause those mental states. This allows the possibility that carriers could be reasons for, or could justify, our utterances about consciousness despite not effectively causing such utterances.

If we accept the above analysis, then the failure of p-consciousness to participate in the effective dynamics of the brain does not entail that our claims about p-consciousness cannot be justified. However, a positive account of exactly how it might be that our claims about p-consciousness are indeed justified is still needed.

Rosenberg proposes such an account, but first clarifies some issues in the vicinity. In addition to propositional knowlegde, as expressed by "knowing that..." clauses (e.g. knowing that a day is 24 hours long), and procedural knowledge, as expressed by "knowing how..." clauses (e.g. know how to ride a bike), Rosenberg observes that we have a third form of knowledge, which is knowledge of our internal experiential states. He calls this form of knowledge "knowledge by acquaintance," and it can be expressed in clauses of the form "knowing what..." (e.g. knowing what it is like to see a sunset, knowing what vanilla ice cream tastes like, knowing what it is like to feel as if you are being watched, etc.). Cast in these terms, solving the knowledge paradox comes down to explaining how instances of "knowing what" can justify instances of "knowing that." Before this can be done, we need to analyze knowledge by acquaintance and propositional knowledge of conscious states more closely.

According to TNI, knowledge by acquaintance is a kind of direct acquaintance with the intrinsic phenomenal and experiential carriers of a natural individual's nomic content. It is an intimate, immediate, and primitive form of knowledge that is underscored primarily by the intrinsic natures of the carriers themselves, rather than by a complex system of effective relationships (such as a high-level cognitive mechanism) that might obtain among those carriers. The 'knower' in this case is the experiential property carrying receptivity, and the 'known' is the various phenomenal properties bound to the experiencing subject. Because the binding of effective and receptive properties involves a direct infusion of their natures into one another (one literally becomes part of the other, and vice versa-- see chapter 9), there is no appreciable metaphysical distance between the knower and known, the experiencer and experienced.

The relevant kinds of propositional knowledge claims that need to be analyzed here are first-order, second-order, and third-order phenomenal judgments. First-order phenomenal judgments concern the use of phenomenal properties to represent external, publically available information (e.g. "that truck is red," "this pillow is soft," etc.). Second-order phenomenal judgments are about phenomenal experiences themselves (e.g. "I hear a faint tone," "I'm feeling a sharp pain," etc.). Third-order judgments are about experience as a type (e.g. "consciousness exists," "phenomenal qualities are not reducible to structure and function," "orange resembles red more than it does blue," etc.). The representational contents of these phenomenal judgments will be our targets of analysis. In particular, we need to account for how the representational content of such judgments are indeed justified, how it is that second- and third-order phenomenal judgments really are about conscious experience, and how it is that each kind of judgment is susceptible to its own characteristic types of fallibilities.

Recall the first-oder phenomenal judgment, "that truck is red." In this case, the judgment is about an objective property of the truck itself (its surface reflectance), and thus there is nothing particularly interesting or troublesome about its representational content. It is not a judgment about consciousness, and it is susceptible to being wrong in a rather straightforward manner. This kind of judgment is of note here because it is made via the subjective experience of phenomenal redness. In other words, while the representational content is about a physical property of an external object, the representational vehicle for this content is phenomenal redness. Under TNI, we could argue that because phenomenal redness carries the cognitive stream of information that encodes the representational content about the truck, phenomenal redness itself attains that representational content. More generally, within cognitive contexts, effective constraints are arranged in such a way as to have representational content. Thus, within cognitive contexts, it follows that phenomenal properties are carriers of representational content, and thus can be said to have representational content.

Rosenberg proposes that our ability to make second-order phenomenal judgments is based upon a cognitive mechanism that uses phenomenal properties to represent themselves. In this case, we would have the same representational vehicle-- e.g., phenomenal redness-- being interpreted or acted upon differently by different cognitive mechanisms, and thus being endowed with different representational contents in different cognitive contexts. For instance, suppose that within a cognitively structured natural individual, there exists a neural vector coding of visual information that is carried by an instance of phenomenal redness (call it R). Suppose that R is poised to play the role of a representational vehicle within the cognitive engine.

- For first-order phenomenal judgments, R would serve as input to a cognitive decoding mechanism C1 that 'interprets' the information stored within R as representative of some feature of the external world, such as the surface reflectance of a truck. In the context of the cognitive processing of C1, R's representational content is the truck's surface reflectance.

- For second-order phenomenal judgments, R would serve as input to a separate cognitive decoding mechanism C2 that 'interprets' the information stored within R as representative of R itself. In the context of C2's cognitive processing, R's representational content is the information stored in R; in other words, R's representational content is its own underlying representational vehicle, itself.

It is plausible that decoding mechanisms like C2 could exist in the human brain. Such mechanisms would allow a subject to track the features of the representational vehicles (the phenomenal properties) it uses to track the features of the external environment. This capability could prove advantageous in situations where the representational vehicles become degraded in some way and thus lose their efficacy as accurate or useful representations of the environment. For instance, a second-order phenomenal judgment along the lines of "my vision has become blurry" could indicate to the subject that first-order visual judgments are no longer to be trusted as accurate representations of the environment, and the subject could then adjust its behavior appropriately. In addition, the strategy of using a representational vehicle R1 to represent itself presents several design benefits over using a second representational vehicle R2 to represent R1, including parsimony of resources and reduction of possible error. Therefore, the self-representation strategy would be the one more likely to be selected for by evolutionary pressures.

If second-order phenomenal judgments are implemented in this fashion, then fallibility in such judgments could be attributed to processing limitations in the phenomenal decoding mechanisms that underlie second-order phenomenal judgments (e.g. C2 as described above). Such limitations could include limitations in bandwidth, storage capacity, and the resolution with which differences in information structures could be detected. Whenever the resources of the decoding mechanism are strained, second-order phenomenal judgments become prone to error. For instance, if a human subject attends closely to the music of a radio while driving, then bandwidth limitations are strained and second-order phenomenal judgments about the qualitative visual field are more likely to be impaired.

However, when the relevant cognitive resources are not strained, second-order phenomenal judgments can be held to be infallible and certain. For instance, when a cognitively normal subject can pay undivided attention to a stop sign from a close distance, the subject can make the second-order phenomenal judgment that he is experiencing phenomenal redness with a certainty that is beyond reproach. This certain propositional knowledge about qualitative experience can be accounted for by noting that (1) there is no metaphysical distance between the phenomenal knower and the phenomenally known thing (as discussed above), and (2) there is no epistemic distance between representational content and representational vehicle, since the vehicle is used to represent itself. In effect, the total causal situation is arranged in such a way that there is no 'room,' no opportunity for any error of judgment to arise.

Third-order phenomenal judgments are derived from repeated second-order observations of conscious experience over time. They are not nearly as susceptible to cognitive limitations as are second-order phenomenal judgments, since they need not be held victim to performance-compromising computational loads that could arise and interfere at any given moment. Rather, they can be formed on the basis of practically innumerable second-order phenomenal judgments made under favorable conditions over the course of a lifetime. For this reason, third-order phenomenal judgments deliver our most certain knowledge about consciousness.

 

[hypnagogue]

summary, part 3

The Superfluity of Consciousness

If consciousness were epiphenomenal, then our faith in the parsimony of nature would be undermined. Even worse, we would have to abandon any notions that science can tell us what the world is really like-- after all, if consciousness were entirely superfluous to our understanding of the world's causal structure, how could we rule out the possibility that nature is positively overflowing with such superfluous phenomena whose natures we could never know?

Fortunately, TNI avoids this dilemma entirely. Phenomenal and experiential properties are not causally irrelevant in TNI, but rather play the essential causal role of fundamental, intrinsic carriers. All causal processes depend upon the phenomenal and experiential carriers to form the categorical, intrinsic basis upon which they rest; without such carriers, effective and receptive properties-- in essence, the entire world as we know it-- could not exist. So phenomenal and experiential properties are not causally superfluous, but rather causally indispensible. (As an aside, their causal role qua carriers is not one that involves direct effective influence on the dynamics and evolution of causal systems, and so the carrier role wins them causal relevance without resorting to any form of interactionism. As we will explore further in the next chapter, the causal role of carriers qua carriers is not one of efficient causation, but rather is akin to traditional concepts of material causation.)

We can furthermore observe that high-level natural individuals are not causally superfluous with respect to lower-level individuals. The determination problem is not solved at the world's microphysical level, but rather, constraints on an individual's effective state are enforced within and across many levels of nature at once (see chapters 9 and 10). This implies that the cognitively structured, high-level consciousness of humans is not causally superfluous with respect to the brain's lower-level physical states. That is, there is not a unilateral arrow of influence going from microphysical brain state to 'macroscopic' conscious states, but rather, both levels (and all levels inbetween) place various kinds of causal constraints upon each other, both directly and indirectly. Just as the microphysical individuals in the brain help to determine the effective states of the high-level cortical individual corresponding to the human experiencing subject, the causal structure of that high-level cortical individual likewise enforces causal constraints on possible states of the lower-level individuals.

So in TNI, human consciousness taken as a whole has causal significance for the lower-level physical happenings in the brain, and this adds another sense in which consciousness can be said not to be epiphenomenal. However-- once again-- this is not a form of interactionism. It is not being claimed that phenomenal consciousness spookily 'interacts' with the physical. Nor, as we will discuss in more detail in chapter 14, is this a form of downward causation. Rather, the claim is that human consciousness carries the nomic content of a high-level individual in the brain, and that this high-level individual places causal constraints upon lower-level individuals through the same general mechanism by which all causal processes operate, according to the Theory of Causal Significance-- the activation of effective constraints via a shared receptive connection.

The Grain Problem

The structure of phenomenal properties in human consciousness seems to be inordinately homogenous and coarse-grained with respect to the complex and finely detailed physical structure of the brain. How can we reconcile the two? In chapter 7, it was suggested that a level-encapsulated, functional account of cognitive processes could be the key to resolving the grain problem. An entity's causal role in a functional system is characterized by the functional relationships it shares with other entities within a canonical context, and can thus be abstracted away from lower-level details of how that role is implemented. For instance, computer programs are defined functionally, and so the same computer program can be implemented by a PC, or an abacus, or many other kinds of physical systems; the details of how the program is implemented is ultimately irrelevant, so long as the proper high-level functional behavior can be sustained.

It follows that if the structure of phenomenal properties mirrors the structure of high-level cognitive functions rather than low-level physical details, the grain problem will have been resolved. The challenge is to find a naturalistic account upon which there could be level-encapsulated functional roles acting in a canonical causal context, and to explain how the activity of such a functional system could account for the qualitative aspects of phenomenal experience. TNI delivers on both accounts.

According to the Theory of Causal Significance, the world's receptive structure stratifies causal systems into a heirarchy of causal levels. Within any given level n, a natural individual is afforded a canonical causal context by being receptively bound to other level n individuals by a level n+1 receptive connection. Causal nexii thus provide a basis for the existence of level-encapsulated functional roles within a canonical context. Importantly, this basis is a natural one, one to which nature is sensitive; the functional character of causal processes is actual, rather than being a mere projection of human concepts onto nature's workings.

The functional character of the Theory of Causal Significance, in conjunction with the Consciousness Hypothesis (and more generally, the Central Thesis), can thus explain why low-level physical detail is lost in the structure of the high-level phenomenal properties. The reason is that the high-level effective properties are not identical to conglomerations of lower-level properties, but are in a sense 'pinched off' and encapsulated at their own level of nature, according to the functional roles they play at that level. For instance, recall the discussion from chapter 10 that an individual could have multiple, distinct effective states that contribute the same effective constraint to a causal nexus. Since effective properties are identified with the constraints placed on a nexus by effective states, it follows that some of the information found in an individual's effective states could be lost to that individual's effective properties. The only information that would be preserved is the information that makes a difference to the constraint structure of the nexus in which the individual is a member. Since phenomenal properties carry effective properties, we can see immediately how it could be that the structure of a phenomenal property is more coarse-grained than its lower-level basis.

The Boundary Problem for Experiencing Subjects

The boundary problem is often overlooked, but is vitally important. How can we account for the fact that human consciousness is individuated on a middle level of nature? On the face of it, it seems that the causal mechanisms available to physicalism can only support individuation at the microphysical level, where each fundamental physical entity is a basic individual, and at the highest level possible, where the universe as a whole can be regarded as a completely closed causal system and thus as a kind of individual.

The boundary problem is solved rather straightforwardly, at least in principle, by TNI. The boundaries of natural individuals at middle levels of nature are traced by receptive structures-- each causal nexus is a bounded individual. The inductive definition of natural individuals provided in chapter 9 suggests a clear way in which receptive connections could come to exist at middle levels, rather than being restricted to the smallest or largest possible scales. Because experiential properties carry receptivity, and because phenomenal properties carry the effective properties that are bound together by instances of receptivity, phenomenally experiencing subjects exist at the middle levels of nature wherever causal nexii exist at middle levels. The natural condition that supports the individuation of experiencing subjects at middle levels of nature is just the receptive structure of natural individuals.

Although the theoretical principles of TNI sate the driving concerns behind the boundary problem, there still remains a substantial question: What is the receptive structure of the world really like, and how can we discover it? What systems in nature are natural individuals? This is a highly important and non-trivial question for TNI; it deserves an in-depth treatment that would require substantial further work, both theoretical and empirical. For now it is left as an open question, although in chapter 14 Rosenberg presents some preliminary theoretical work on the topic.

Making Peace with Receptivity

Although a substantial metaphysical account of receptivity has been put forth, one might retain real misgivings about its existence due to the way it seems to elude one's attempts to conceive of it, and because of its seeming epistemic opacity. However, what is striking is that we have constructed a theory of causation incorporating receptivity from first principles, entirely independent from considerations of the mind-body problem, and this theory has proven to be extremely fruitful in addressing deep philosophical problems regarding consciousness. The kinds of characteristics that TNI predicts receptivity and natural individuals should have seem to exactly mirror the classically problematic features of consciousness-- e.g., natural individuals should have a kind of internal, partless unity that nonetheless appears to be a composite system from an external point of view (phenomenal consciousness seems to have a kind of internal unity but nonetheless appears from an external point of view to be a composite system of neurons); natural individuals require carriers that are intrinsic tout court (phenomenal qualities are intrinsic); the intrinsic nature of the carriers of a natural individual should be epistemically private, rather than being susceptible to third person investigation (the qualities of conscious experience are epistemically private and cannot be discovered or shared in a third person, public medium); and so on. Rosenberg sums this up nicely: "Consciousness is strange in just the way a carrier of nomic content has to be strange" (p. 271).

The substantial metaphysical account of receptivity not only predicts that something with the strange features of consciousness should exist, but also complements, validates, and enriches our metaphysical account of consciousness itself. Consciousness's troublesome characteristics (unity, intrinsic-ness, etc.) are epistemically transparent to us, insofar as we are immediately acquainted with them and experience them directly. However, trying to place these characteristics into a coherent metaphysical worldview invariably raises problems-- why should something with these odd properties even exist, and what place do they have within the natural world? Consciousness is so deeply foreign to our concepts of the physical world, and can sometimes seem so far removed from and ultimately irrelevant to the causal workings of the physical, that some are even tempted to severely doubt or abandon outright their beliefs in what seems, in some ways, to be our surest and most immediate object of knowledge and way of knowing. The account of causation provided by TNI, which hinges crucially on the properties of receptivity, quells these metaphysical misgivings about consciousness. We see that consciousness's features are not superfluous or absurd or mystical or antyhing of the sort; rather, they are precisely the sorts of features one would need to construct a coherent, robust, and explicitly causal realist account of the natural world.

While receptivity is metaphysically robust and consciousness is metaphysically problematic, the exact opposite holds true for the epistemology of these things: Consciousness is epistemically transparent and receptivity is epistemically opaque. Just as we used the metaphysical account of receptivity to undercut concerns about the metaphysics of consciousness, we can our first-hand knowledge of the nature of consciousness to undercut partially our concerns about the epistemology of receptivity. If the Consciousness Hypothesis is correct, then it follows that each of us has epistemic access to at least one instance of receptivity (and the phenomenal individuals it binds), in the form of our own immediate subjective experiences.

Receptivity is one of the central pillars of the Theory of Natural Individuals, a theory that makes significant and startling conceptual advances on what have long seemed to be intractable tensions, puzzles, and paradoxes about consciousness. Using this theory, we can begin to explain with coherence, clarity, and cogency why these conceptual difficulties have arisen at all, and how they might be resolved. We can find a place for consciousness in the natural order of things that does not deny, but rather validates and embraces what we know about consciousness from direct experience. We can pick up the disjoint pieces of ontologies past and reshape them into a single, unified, coherent, and even beautiful new worldview. It seems clear that the enormous promise of the Theory of Natural Individuals should temporarily bestill, if not overwhelm, any skeptical urges that might take hold at this early stage.

 

[Canute]

Wow. My brain hurts after reading all that. As far as I can tell it seems to make sense, but I must admit that quite a lot of it is beyond me. However, I'm beginning to see what Rosenberg is getting at, thanks to your clarifications. I still think he needs to read Spencer Brown and the mystics, but he goes further down that road than I realized at first, even if I still think he should go further.

I'll stop posting on this immediately if I'm out of order or disrupting the discussion, but just to illustrate what I mean I'd like to make one or two comments. No need to respond.

I would prefer that Rosenberg used the term "knowledge by identity" rather than "by acquantance," thus sticking to the usual terminology. Sometimes known as "non-intuitive immediate knowledge," although this phrase seems to be used with rather variable meanings. Also known as 'mystical' knowledge. The source of Descartes certainty of 'cogito'.

Did you know that Freeman Dyson used to argue that electrons are conscious? As you'll know there have even been a few recent and articles in the consciousness lit. arguing for microphenomenalism. Empedocles wrote, "If thou shouldst plant these things in thy firm understanding and contemplate them with good will and unclouded attention, they will stand by thee forever every one, and thou shalt gain many other things from them; … for know that all things have wisdom and a portion of thought." Note that he does not say "it is possible that..." or "Let us assume that ..."

I know that you're not sympathetic to all this 'mystical' stuff about knowledge and reality. But I wonder if you realize that what you say in these posts about knowledge, identity, receptivity, the reductive analysis of 'moments' and so on is entirely consistent with that view. Notable is your reduction of effective and receptive properties and knower and known to 'nonduality'.

It is an intimate, immediate, and primitive form of knowledge that is underscored primarily by the intrinsic natures of the carriers themselves, rather than by a complex system of effective relationships (such as a high-level cognitive mechanism) that might obtain among those carriers. The 'knower' in this case is the experiential property carrying receptivity, and the 'known' is the various phenomenal properties bound to the experiencing subject. Because the binding of effective and receptive properties involves a direct infusion of their natures into one another (one literally becomes part of the other, and vice versa-- see chapter 9), there is no appreciable metaphysical distance between the knower and known, the experiencer and experienced.

"intimate, immediate, and primitive" is spot on, and the reason Lao-Tsu says "Knowing the ancient beginnings is the essence of Tao". Anne Rice writes this, in a passage from one of her vampire tales.

"Very few seek knowledge in this world. Mortal or immortal, few really ask. On the contrary, they try to wring from the unknown the answers they have already shaped in their own minds – justifications, explanations, forms of consolation without which they can’t go on. To really ask is to open the door to the whirlwind. The answer may annihilate the question and the questionner."

This is the 'mystical' view of knowledge. By "annihilate question and questionner" is meant precisely the annihilation of the "metaphysical distance between knower and known".

Sri Aurobindo writes:

"Knowledge can only come by conscious identity, for that is the only true knowledge, - existence aware of itself. We know what we are so far as we are consciously aware of ourself, the rest is hidden; so also we can come really to know that with which we become one in our consciousness, but only so far as we can become one with it."

And on the "Unknowable,"

". When we speak of It as unknowable, we mean, really, that It escapes the grasp of our thought and speech, instruments which proceed always by the sense of difference and express by the way of definition; but if not knowable by thought It is attainable by a supreme effort of consciousness. There is even a kind of Knowledge which is one with Identity and by which, in a sense, It can be known."

... "In this consciousness the knower, knowledge and the known are not different entities, but fundamentally one. Our mentality makes a distinction between these three because without distinctions it cannot proceed; losing its proper means and fundamental law of action, it becomes motionless and inactive. Therefore, even when I regard myself mentally, I have still to make this distinction. I am, as the knower; what I observe in myself I regard as the object of my knowledge, myself yet not myself; knowledge is an operation by which I link the knower to the known. But the artificiality, the purely practical and utilitarian character of this operation is evident; it is evident that it does not represent the fundamental truth of things. In reality, I the knower, am the consciousness which knows; the knowledge is that consciousness, myself, operating; the known is also myself, a form or movement of the same consciousness. The three are clearly one existence, one movement, indivisible though seeming to be divided, not distributed between its forms although appearing to distribute itself and stand separate in each."

Do you see why I persist in annoying everyone by quoting from the mystics? Consciousness can be explored fully from the inside, without hypotheses and conjectures and, as one would expect, the results of such explorations are consistent with what we find when we examine it from the outside.

It seems to me that consciousness might be defined as the ability to know, since this is its defining property, direct contact with noumenal through the annihilation of knower and known as distinct categories, and thus the capability of acquiring knowledge by identity, immediate knowledge of what we are. Really the now more-or-less orthodox definition of consiousness from Nagel is short for "knowing what it is like".

Ignore this if it takes the discussion off-track. But really, despite my various objections, I feel Rosenberg is on to something. At least he disposes of the mind-brain antimony and other traps. from what you say even effective and receptive properties reduce to a false distinction in the end, a point I tried to make earlier when I mistakenly thought he did not reach this conclusion. But I'd argue there's a better way of researching into consciousness, causation and knowledge than this, which will always remain hypothesis and theory rather than knowledge, namely directly contacting the "intimate, immediate, and primitive" as you put it.

 

[hypnagogue]

I believe we can learn important things from first person experience, but I think such knowledge has its limits. In particular, as regards consciousness, what we can know with certainty from first person experience is just the character and quality of phenomenal experience itself. For instance, we can come to a phenomenal understanding of particular instances of qualitative experience (e.g. what this looks like), and more generally, I believe we can make certain judgments about experience as a type (e.g. experience is not reducible to structure and function). An appreciation that the phenomenal knower is not truly distinct from the phenomenal known, or at least that no sharp dividing line can be drawn between them, might be another kind of knowledge that falls under this category.

As far as this sort of knowledge goes, I tend to doubt that unusual states of experience can deliver knowledge of experience as a type that is not available to careful examination of normal, waking consciousness. At best, I think what is delivered from unusual or transcendent experience is knowledge of novel instances or combinations of phenomenal qualities, and perhaps a more focused attention on experience qua experience than is normally exerted. But knowledge of experience as a type is what we need to begin to form a picture of how experience fits into the natural world, i.e., it is the philosophically important kind of knowledge as regards consciousness. So I'm not sure mystical, transcendent, or otherwise unusual experiences are really needed to get an appreciation for the philosophically important features of consciousness, although they may be useful in drawing one's attention to those features.

I have much stronger doubts that we can know how consciousness is situated in the grand metaphysical scheme of things just from first person experience. The reason for this is basically that such an analysis introduces metaphysical and epistemic 'distance' between immediate experience on the one hand, and the world as a whole on the other. This distance introduces substantial room for error and doubt. I.e., "just because it seems this way doesn't make it so" becomes a valid critique. For instance, I would not accept the claim that one can know panexperientialism is true just from careful introspection of one's own experience. Such a claim goes beyond the scope of the first person evidence and therefore needs further justification. (There are no metaphysical or epistemic distances when making second- and third-order phenomenal judgments, which is what places them on much surer footing. So for example, the deflationary claim "Just because it seems that we have qualitative experiences doesn't make it so" loses its bite, because in this case we are talking about the seeming of the immediate experience itself, not what the seeming might indicate about something else.)

For this reason, I am not comfortable putting all my eggs in the introspectionist basket. Obviously we must pay heed to what we can know from phenomenal introspection (and Rosenberg's account does indeed do this), but we must also recognize the limits of such an approach. I think Rosenberg's claims are much stronger than those that are made entirely on the basis of first person experience because he recognizes and avoids overstepping the epistemic limitations of such claims.

In particular, he provides a metaphysical account of causation entirely independent from the mind-body problem, and this account still brings consciousness in at the end as an integral and indispensable element. So not only do his wide metaphysical claims about consciousness have stronger justification than pure first person accounts, but they are also much clearer about how consciousness fits into place in the natural world-- i.e. what the causal role of phenomenal experience is, how it is perfectly suited to play this role, and why such a role is even needed in the first place. So what we end up with is (IMO) a quite compelling and elegant account of how consciousness is related to causation, something that to my knowledge has not been provided yet by the mystics (or anyone else for that matter). Using this analysis of the relationship between consciousness and causation, we can derive views such as panexperientialism on a solid theoretical basis (one that is still crucially informed by the first person evidence) without overstepping the epistemic limitations of first person experience.

 

[hypnagogue]

Please note that Canute's reply to the above post can be found here. The discussion was becoming too tangential from this thread's intended topic, so it was moved into a new thread.

 

[antfm]

The issue ot time keeps being a difficult one to handle for me, in spite of the new precisions made about it in this chapter (mainly in the section The Subjective Instant).

"In the proposed model, natural individuals provide frames of reference for constructing spacetime, and distance in space and time between events is a projection of dependency and inmediacy of interaction between individuals"(p. 255)

First, I feel a bit confused because events, as usually understood as occurences in a spacetime background, are different from events according to Rosenberg's ontology. I understand, then, that events in the context of the quote above refer to that usual sense of events. Because events, in the sense of Rosenberg's events ontology, wouldn't need to involve, in principle, a spacetemporal framework.

"The mistery of the subjective instant can be clarified by understanding why, when the higher level individual (with its experiencing receptivity) is chosen as the frame of reference for determining a spacetime mapping of events, projecting this situation into a coherent spacetime scheme could require mapping the instantaneous state for the higher level individual onto a duration of states in the existence of the lower level individuals"(p.255)

Well, this is difficult for me. Again, are events in this context the events of the proposed Rosenberg's ontology? I do not think so, since, again, events, in principle, do not require a temporal construction in the physical sense. It would be the particular operating of events what, creating cascades of asymmetric constraints among individuals, would implement a temporal structure in the world, as suggested in ch.10.

In any case, this would be a subjective, or intersubjective, type of time - the types of time considered by Rosenberg (we'll see that distinction in next chapter). As Tychic pointed out in previous chapter discussion:

But even if "events" like quantum entanglement are simultaneous they perhaps endure one moment at least in the phenomenal sense. When I think of actualization I think of at least one point of a moment of this event. I'm not sure wether this differentiation between phenomenal and physical time is correct. It would be only physical time that supervenes on cascades of events. The structure of phenomenal time would be self displaying in the experience itself.

I'd agree very much with this view and its concerns. However, Rosenberg's proposal still does not need a physical time. A physical time would rather be a theoretical construct, implemented as a cognitive tool, a metric so to say, in the world, and constructed as a projection of the subjective time.

What would remain unclear to me, anyway, would be the status of the subjective time within Rosenberg's scheme.
I can't get out of some kind of circularity, since the physical time has been proposed as a projection of a subjective time, and the subjective time, in occasions, seems to me to be modeled out of the time of a physical framework.
"An individual with such a complex effective state could produce an enormous number of subtly characterized and simultaneously occurring effective properties"(p.257)
How can we understand this "simultaneously occurring" without some kind of temporal reference?. Out of that sense of simultaneity, however, I can understand the following sentences in the same paragraph:
"Additionally, because overlapping individuals in the higher level process may partially share members, the constraints associated with these members may come to exist at multiple moments in the subjective time of the process. From this we can deduce that any such elements within experience would have duration "(p.257)

Anyway, I also see difficult to locate that 'simultaneity', or the consequent subjective time, within Rosenberg's proposed ontology, that is as if an effective or receptive property, or something else (individual, emergence law, etc) that at the moment I can't work out.

 

[hypnagogue]

My understanding of Rosenberg's discussion on the subjective instant is as follows. Suppose we have two high-level cortical individuals at level n that are asymmetrically connected. We can represent this in the notation from chapter 10 as follows:

[C_n.1 => C_n.2]

By the Consciousness Hypothesis, the nomic content of each individual C is carried by a human-like consciousness. Following Rosenberg's speculative proposal about spacetime in chapter 10, we can also say that the asymmetric receptive connection found here serves as the causitive basis for the construction of a temporal mapping of events occurring at level n (call this n-time). Each individual C represents an instant or a brief duration in n-time, such that C_n.1 occurs before C_n.2.

Putting the two approaches together, it is natural to conjecture that the subjective instant corresponds to a high level cortical individual's reception of causal constraint in an asymmetrically structured causal nexus. So for instance, by the Consciousness Hypothesis, C_n.2 should experience the effective constraints placed on it by C_n.1; furthermore, C_n.2's experience should constitute a subjective instant, since C_n.2's place in the greater causal structure makes it a temporal instant in n-time. To state this slightly differently, Rosenberg's reasoning seems to go like this: C_n.2's experience corresponds to its causal structure and context at level n; C_n.2's causal structure and context at level n corresponds to an instant in n-time; therefore, C_n.2's experience corresponds to an instant in n-time, i.e. it is experienced as a subjective instant in n-time.

Now observe that each high level C is itself the receptive binding of a group of level n-1 individuals S, which themselves could possesses intricate internal causal structures and cascades. E.g., it could be that C_n.2's internal causal structure looks like this:

C_n.2: [S_n-1.1 => S_n-1.2 => ... => S_n-1.j]

So although C_n.2's causal structure and context within a level n cascade makes it occupy an instant of n-time, C_n.2's own internal structure could itself be constituted by a cascade of individuals on the immediately lower level. By the proposal in chapter 10, such a cascade at level n-1 would form the basis for the construction of a temporal framework with respect to individuals at that level (call this m-time, where m=n-1), one that would not necessarily map cleanly onto timeframes on other levels. For instance, notice that in the hypothetical example above, the one instant of n-time grounded by C_n.2 maps onto j instants of m-time at the immediately lower level. It follows that what is an instant at level n looks like an extended process from level m. And of course, it could be that the internal structure of each S features its own rich connections and cascades, and so on all the way down to level one. Considerations like this ground the general theoretical explanation for how it might be that what is experienced as an instant (or an irreducible duration) in subjectively experienced time in fact maps onto multiple instants or extended durations of lower level causal processes.

(Some words of caution: Note that the above does not necessarily imply that e.g. there is a subject of experience on level m that experiences time flowing faster or slower than the level n cortical individual. It seems very likely to me that our subjective sense of enduring, flowing time depends on the integration of high-level cognitive mechanisms such as short and long term memory, and such mechanisms might not be appropriately poised to create such a sense of enduring time even for cortical individuals at the immediately lower level. Also, Rosenberg makes no commitment to any direct quantitative mapping of lower-level instants to higher-level ones, either in terms of 'subjective' or 'objective' (inter-subjective) time; he only makes the general observation that instants on a high level could map to multiple instants at lower levels, and this could begin to explain the observed temporal asynchronies between subjective, experiential instants and inter-subjective, physical instants.)

Finally, we should keep in mind that this treatment of the subjective instant is at most a speculative rough sketch of how the problem might be solved within framework like Rosenberg's. In particular, the discussion here depends heavily on the treatment of spacetime in chapter 10, which itself is (by Rosenberg's own admission) a largely speculative and rough proposal. The core concepts of Rosenberg's Theory of Natural Individuals are not firmly committed to the discussion of spacetime given in chapter 10, and so a completely different treatment of the subjective instant than is found here could be given without abandoning TNI's theoretical foundations. By contrast, note that the proposed solutions to e.g. the superfluity of consciousness and the boundary problem issue directly from the core tenets of Rosenberg's framework, and so should be regarded as having more theoretical weight.

 

[Tychic]

perspective and incompatibilistic freedom

Some questions:
1) What is wrong when I call a receptive property that constitutes a natural individual a perspective? I do not know what experience is in the non mental case, but would it be possible that it is no perspective?
A Leibnizian monad is a centre or mirror of the world, as far as I now, (a mirror not in the special visual sense). I think a natural individual is something similar only because the receptive property excludes some possibilities to interact with other natural individuals. But the access to some other individuals may be open (as far as effective connection reaches), but at least the receptivity in which it is included is a frontier with which it is impossible for it to interact. (Rosenberg’s proposal is that each individual can only interact with individuals of the same level of nature, we have discussed and enhanced this position yet, see PF entry 3 to chapter 9.)
An effective property corresponds more to a fact than a perspective. So there could be third parallel for the phenomenal-experiential distinction: besides, effective and receptive, object and subject: perspective and fact.
2)

Without the causal context afforded by a whole experiencing subject (receptivity), there can be no phenomenal (effective) properties.

Is it right that there have to be two levels of causal context, the level integrating some effective properties to a natural individual and the higher level in which the natural individual itself takes part? Only the highest level individuals take no independent part in the causal history of the world, have no independent interaction with the world and are no candidates for a conscious individual.

3) Rosenberg's proposal to the problem of the subjective instant solves the following problem:

There seems to be a mismatch between the temporal ordering and duration of events as experienced in consciousness and the temporal ordering and duration of events as measured in the brain by third person methods.

An example of this mismatch would be the famous Libet experiment for the impossibility of free will (that can be solved also otherwise). I like Rosenberg’s solution that is better illustrated in the old internet version of the book, see:
http://ai.uga.edu/~ghrosenb/chptr13.htm

As far as I see Rosenberg’s proposal is neutral about the question about compatibilism of determinism and free will. Rosenberg’s indetermination of low level physics suggests incompatibilism. Liberarianism requires a subject that has sometimes two opportunities. But does it make sense that not the level above the human mind makes a decision/determination between two options but the mind itself? If it makes sense, would this mean that there are two “similar” receptivities and only one of these comes to existence? In this case, would there be a subject that makes the decision/determination? Another decision means other phenomenal qualities and another receptivity and another subject. But could there be a plausible sense to speak of only one subject in both cases?

 

[hypnagogue]

An effective property corresponds more to a fact than a perspective. So there could be third parallel for the phenomenal-experiential distinction: besides, effective and receptive, object and subject: perspective and fact.

It depends exactly what you mean by 'perspective.' There are some ways in which it seems appropriate to think of receptivity/experience in terms of perspective. The most obvious way is that each experiencing subject intrinsically has a sort of first person perspective with respect to its bound phenomenal properties, as opposed to a third person perspective. Rosenberg's proposed treatment on spacetime also grants each receptivity a kind of perspective, insofar as each receptivity can be seen as providing a unique, privileged kind of reference frame or pivot from which to construct mappings of events into a spacetime framework. There may be other ways as well in which receptivity/experiencing subjects can be fruitfully thought of as providing an inherent basis for some sort of perspective.

However, there also seems to be a way in which receptivity/experience is not appropriately thought of as perspective. I think the usual sense of the word 'perspective' involves a kind of interest-relative component; perspectives are just convenient ways to see things, but don't necessarily correspond to how nature itself is structured. That is, in the normal usage, perspectives seem to correspond to relatively arbitrary mental constructs used to organize information, whereas facts cannot be arbitrarily put together, but are rigidly constrained by nature-- they are the things that are true of nature itself. In this sense, receptivity/experience is not perspectival, but is quite factual. There is nothing arbitrary or interest-relative about the world's receptive structures or experiential properties; they are actual features of the world, and whatever perspectives they might provide therefore are not arbitrary, but have an inherent basis in actual features of the world.

Without the causal context afforded by a whole experiencing subject (receptivity), there can be no phenomenal (effective) properties.

Is it right that there have to be two levels of causal context, the level integrating some effective properties to a natural individual and the higher level in which the natural individual itself takes part? Only the highest level individuals take no independent part in the causal history of the world, have no independent interaction with the world and are no candidates for a conscious individual.

According to the view Rosenberg puts forth in the book, the receptivity of a level n natural individual X experiences the carriers of the effective constraints that are placed upon X by other level n individuals. But such constraints cannot be placed upon X unless X shares a common receptivity with the individuals that are constraining it. So yes, there are two receptivities that enter the story: X's receptivity (to experience the carriers of effective constraint), and the receptivity to which X is bound (to create a causal nexus within which X can receive effective constraint).

For instance, suppose we have an individual at level n+1 that binds two level n individuals, e.g.

I_n+1.1: [I_n.1 => I_n.2]

Suppose that I_n.1 is determinate, and has some effective property E that places causal constraint upon the effective state of I_n.2. Then we could say that I_n.2's receptivity experiences the phenomenal carriers of E. And furthermore, the receptivity of I_n+1.1 is required to 'deliver' the phenomenal carriers of E to I_n.2. Both receptivities are thus needed in order for I_n.2 to experience any phenomenal properties.

However, in e-mail correspondence Rosenberg has expressed some misgivings to me about this view. An alternate view is that individuals at level n experience the effective properties of the level n-1 individuals that they receptively bind together. On this view, I_n.2 would not experience the effective constraints placed on it by I_n.1; rather, it would experience the carriers of the effective properties of the level n-1 individuals that it binds. Likewise, I_n+1.1 would experience the phenomenal carriers of I_n.1's effective properties. (Since I_n.2 is not placing causal constraint in this nexus, it has no effective properties within the context of this nexus.) Notice that with this view, only one receptivity is needed for experience of phenomenal properties to occur.

Both views have their strengths and weaknesses. I find the latter view to be somewhat more intuitive and more consonant with the rest of Rosenberg's theory. Rosenberg originally held the latter view, but switched to the former for the writing of his book. He might be leaning back towards the latter, though.

As far as I see Rosenberg’s proposal is neutral about the question about compatibilism of determinism and free will. Rosenberg’s indetermination of low level physics suggests incompatibilism. Liberarianism requires a subject that has sometimes two opportunities. But does it make sense that not the level above the human mind makes a decision/determination between two options but the mind itself? If it makes sense, would this mean that there are two “similar” receptivities and only one of these comes to existence? In this case, would there be a subject that makes the decision/determination? Another decision means other phenomenal qualities and another receptivity and another subject. But could there be a plausible sense to speak of only one subject in both cases?

You're right that Rosenberg's theory is neutral on the topic of free will. One could probably construct free will and non-free will versions of the theory if one wished. I don't know if there's anything in the basics of the Theory of Natural Individuals itself that should make one lean one way or the other. Of course, the critical question is what exactly you mean by 'free will.'

(I myself don't make much of the free will topic, since it seems the only way in which a choice could be made is either by some deterministic process or by some random process, and neither of these seems amenable to supporting the stronger senses of the term 'free will.' I'm happy to be a compatibilist on this one-- I think an action is free just in case it has been conducted voluntarily. It seems to me free will has more to do with intentions than with more complicated metaphysical issues.)

As for the question of the subject-- I would worry here that the sense of the word 'subject' used in the discussion of free will is not coextensive with the experiencing subjects discussed in Rosenberg's theory. When talking about free will, the subject is the thing to which choice is attributed. The thing to which we attribute choice seems to be the cognitive self-construct, i.e. one's own mental construct/representation of oneself. It's not clear to me that the cognitive self-construct is exactly the cortical individual in the brain whose existence grounds the human experiencing subject at any given moment. It could even turn out that there is no one natural individual to which choice can be attributed, but rather a whole system of individuals is involved in the making of a choice. So your concerns about how to reconcile the metaphysics of the experiencing subject with the subject-as-chooser might not be as pressing as they seem. (My own suspicion is that the cortical experiencing subject is not identical to the cognitive self-construct, and also that the process of choice involves the interaction of many natural individuals in the brain and so cannot be attributed to a single 'subject,' in the natural individual sense of the word.)

 

[hypnagogue]

Evaluating the application of the Theory of Natural Individuals

Here is my take on the manner in which the Theory of Natural Individuals comes to bear on the conceptual difficulties examined in the first half of the book.

Ubiquity and fundamentalness: I find the arguments in chapters 5 and 6 to be convincing: Not only is panexperientialism logically coherent, but in fact it even seems to be the most appealing kind of theory of consciousness we could come across. Granted that physicalism must be rejected, we would need some sort of fundamental (irreducible, ground-level) bridging laws that describe the relationships between a system's physical state and its conscious state.

We might have some strong prior intuitions that only certain kinds of properties pertaining to biology, complexity, or cognition should be linked to conscious experience by such a bridge law. However, to the degree that each of these is ambiguously defined and interest relative, the corresponding bridge laws would have to be inordinately complex and arbitrary, to a degree we do not expect to find in nature's fundamental laws. Therefore, we should prefer the simplest and most clearly defined kinds of bridge laws, and the bridge law proposed by the Central Thesis certainly meets this demand. The relationship it posits between causation and experience is simple and is grounded on features that clearly and straightforwardly inhere to nature itself, rather than to our interest-relative concepts of nature. For instance, thinking of certain systems as being cognitive is probably just a convenient conceptual tool for humans; natural laws most probably do not recognize a category such as 'cognitive' and differentially apply to systems on this basis. However, if the Theory of Causal Significance is right, then effective and receptive properties are most certainly natural categories to which natural laws could plausibly be sensitive. So the Theory of Natural Individuals succeeds in providing us with a plausible and elegant kind of bridge law between nature's schematic and experiential facts.

The superfluity of consciousness: The deep manner in which the Theory of Natural Individuals ties together causation and phenomenal experience also dispatches worries about the superfluity of consciousness quite easily and straightforwardly. Phenomenal experience is given a definite causal role to play, and this causal role is such that it naturally entails the strange epistemological and causal conundrums we run into when trying to place consciousness in the natural order.

The boundary problem, the unity of consciousness, and the grain problem: I group these problems together because they all can be solved with treatments that have the flavor of functionalism. The boundary problem can be solved by appealing to some causal or functional system whose structure inherently and naturally delineates it as a kind of causal or functional unit; the unity of consciousness can be understood as resulting from a functional interdependence of parts and whole; and the grain problem can be resolved if phenomenal contents are understood to correspond to high-level functional roles in the cognitive economy, since functional roles are encapsulated at a high level such that they can be abstracted away from the low-level details of how they are implemented.

However, just giving a purely functionalist account of these problems is insufficient, for two reasons. One problem is that a pure functionalist account is essentially a physicalist account, and cannot explain why the various functions it describes are associated with phenomenal experience. The other problem is that a pure functionalist account refers to functional entities that we only recognize in interest-relative ways. What is it to play a functional role in a canonical context? If we cannot find a naturalistic basis for such things-- if we cannot provide an account whereby functional roles in canonical contexts are categories to which nature is sensitive and responsive-- then the functionalist explanations of the features of consciousness lose their force. We need to show that functional roles in canonical contexts are things that actually exist in nature, rather than being relatively arbitrary conceptual projections onto the world.

The Theory of Natural Individuals provides a compelling resolution to the unity of consciousness, boundary problem, and grain problem precisely because it can import the explanitory strengths of functionalism while simultaneously covering up its weaknesses. The Theory of Causal Significance not only lends itself naturally to functionalist concepts, but also provides a naturalistic basis (receptive connections as the world's causal infrastructure) on which functional roles and canonical contexts can be said to exist in nature, rather than only in our heads. The Central Thesis provides the necessary connection from functional/causal systems to experiential systems by giving phenomenal and experiential properties the role of intrinsic carriers.

The subjective instant: The subjective instant can be thought of as a temporal analogue of the grain problem, and thus could be grouped in the previous section. However, I list it separately because I'm not quite as convinced that the subjective instant needs a treatment in terms of receptivity. I think Rosenberg's discussion on the subjective instant is one plausible way in which the problem might be solved, but I also think that the problem might yield to simpler considerations. For instance, it could be that there is some neural system N whose activity in a single instant reflects or encodes information that is fed into it over some brief duration of time in the past. If experience of the subjective instant corresponded to activity in N, then the problem of the subjective instant seems to be solved rather straightforwardly.

(Rosenberg argues in chapter 7 that the central problem of the subjective instant is not the timing of neural events vs. the timing of experienced events per se, but that the latter seems to have a kind of privileged frame of reference while the former, according to relativity, does not. However, I'm not convinced that this is really a problem, since practically speaking the neurons in the skull always occupy the same reference frame. What matters about the neural events here seems not to be their relative timings with respect to some outside observor, but rather their relative timings with respect to each other. I think Rosenberg's proposed problem might have more bite e.g. if we could somehow accelerate the elements of a conscious system with respect to each other and see how this effects the system's first person experience, but obviously that's a rather exotic kind of experiment that can't be conducted right now.)

The knowledge paradox: This seems to be the toughest nut in the consciousness jar to crack, and arguably it's the most important. I think Rosenberg's initial treatment on the knowledge paradox and his proposal for knowledge by acquaintance makes significant headway. However, this is the one conceptual difficulty addressed in this chapter with which I am still fairly uncomfortable. Rosenberg's treatment has chipped away at my intuition that phenomenal experience must make a direct difference to the brain's effective dynamics in order to justify second- and third-order phenomenal judgments, but it has not completely eroded those intutions.

Rosenberg's analysis provides a causal link between phenomenal experience and utterances about experience, but it's not as intimate as my intuitions might like-- it seems somewhat too loose and coincidental. For instance, take the case where some cognitive decoder C uses the information in some phenomenal carrier R to represent R itself. It seems that C's access to information about R is underwritten by effective causal interactions between the two. If that is the case, there seems to be a disconnect between the intrinsic phenomenal information in R, and the effective, structural information in R. C seems to be accessing R's effective structure rather than its intrinsic phenomenal content, and we've argued at length that the former cannot entail the latter. So it seems that even though C is using R to represent R itself, this still does not put C in position to send out neural signals on the basis of R's intrinsic phenomenal content.

The danger that follows from this consideration can be illustrated thusly. Rosenberg has argued that it could be the case that two distinct phenomenal carriers might implement the same physical, effective structure. Suppose that there is some phenomenal carrier S whose intrinsic phenomenal nature is different from R, but that S's and R's effective structures are identical. Now suppose that S takes the place of R in the greater cognitive context, such that C is now reading off information from S rather than R. Even though S and R have distinct phenomenal properties, C will fail to differentiate the two and will treat them identically. If C is susceptible to such a trick, then it seems that the second-order phenomenal judgments that C produces cannot be justified. The resulting picture looks a bit like epiphenomenalism, if not with respect to the superfluity of consciousness, then with respect to the knowledge paradox.

I'd be interested to hear other posters' takes on Rosenberg's treatments of the conceptual difficulties in this chapter, particularly about the knowledge paradox.

 

[antfm]

One of my main concerns along the reading has been to grasp a better idea of how causation, in Rosenberg's sense, operates. We have drawn a broad view of causation and finally associated it with 'experiencing' as carrier of receptive and effective properties. I was trying to understand now how it could work more in detail. Rosenberg himself declares that his proposals in this respect cannot be but speculative, and that they need further investigation.

I say this because concerning the problem of the 'knowledge paradox' I wonder if it is too different from the question of how causation works in general. As Rosenberg puts it:

"What we need is a way of explaining the intimacy between the phenomenal subject and its physical states"
"Although it seems like a problem about consciousness on the surface, it may really be a problem about how we understand causation and its place relative to physical science" (p.121)

In particular, I do not see the 'knowledge paradox' as very different, for instance, from the question of mental causation (the decision of rising my arm, and rising my arm).

"The knowledge paradox stems from our knowing that we are conscious even though experiencing seems not to be causally responsible of our brain states" (p.258)

But, why can't we say that experiencing is responsible of our brain states? This question is related to the hypothesis of the causal closure of the world under physics:

"These brain states are solidly physical, and we are asuming the causal closure of the physical, meaning that nothing nonphysical can make a causal difference" (p. 120)

However, as Rosenberg also says, carriers of receptivities are crucial components supporting our total causal situation, in a general, metaphysical level. I think that in this sense 'experiencing' (carrier of receptivities) also acts upon physical causation itself. But perhaps in a sense that what might be subtly modified were the conditions to implement physical causation itself, and not the way that physical causation works and we are familiar with.
As an example, we have seen that the evolving of causal connections might be responsible of the emergence of a physical framework, a spacetime. If this is so, the consequent physical causation would be derived over that framework.

In any case, I am not sure why we should think that the Carriers Theory of Causation rules out interactionism (consciousness-brain states) if that interaction might take place in a way that respected the way physical causation works.

 

[hypnagogue]

One of my main concerns along the reading has been to grasp a better idea of how causation, in Rosenberg's sense, operates. We have drawn a broad view of causation and finally associated it with 'experiencing' as carrier of receptive and effective properties. I was trying to understand now how it could work more in detail.

The general picture is that causation is not a process of production, but rather a process of constraint. Level one individuals, considered independently of any higher-level causal nexii, exist in multiple effective states at once and become progressively more determinate by binding with other individuals by means of receptivity. Layers of receptivity act as a hierarchical filtering mechanism of possible effective states; causal constraint is enforced by receptive binding at higher and higher levels of nature, piling on more and more possibility filters, until finally all but one possible state have been filtered out and determinacy is achieved. Effective and receptive properties are only schematic systems of relationships, and ultimately phenomenal and experiential properties form the intrinsic categorical basis that carries effective and receptive properties.

Of course, you already know all that, but a quick summary might be helpful. Perhaps we could address your concerns more directly if you can think of some specific questions to ask whose answers you think would be elucidating.

I say this because concerning the problem of the 'knowledge paradox' I wonder if it is too different from the question of how causation works in general.

Well, Rosenberg's view of how causation works certainly informs his analysis of the knowledge paradox. Still, the knowledge paradox presents special difficulties because it demands that we explain how our knowledge of subjective experience is justified. It's one thing to argue that phenomenal and experiential properties are the carriers of causal processes, and quite another to argue how it is that those carriers actually license utterances we make about them.

I think the core difficulty of the knowledge paradox, at least for Rosenberg's framework, is to provide some sort of plausible link between the quite different causal roles of the carriers (roughly, material causation) and their dynamic internal contrasts (effective causation). How is it that the trajectory of an effective process could somehow depend upon a non-effective entity? And if the trajectory of effective processes do not depend upon a non-effective entity, how exactly can they be in position to make justified judgments about that entity?

Rosenberg attempts to tackle this problem first by eroding our notions that only causally effective things can accrue causal responsibility, or that only causally effective things can count as reasons or justifications for mental states. I think that attempt is somewhat successful, but at the critical moment when he describes how a second-order phenomenal decoding mechanism could read off properties from an intrinsic phenomenal carrier, I think the familiar old problem of how that could be crops up again with full force. Of course, I might just be missing some key concept from the book that could help me steer clear of such worries.

In particular, I do not see the 'knowledge paradox' as very different, for instance, from the question of mental causation (the decision of rising my arm, and rising my arm).

I think our sense of causal agency is more a case of a first-order phenomenal judgment than a second- or third-order phenomenal judgment. That is, when we perform a voluntary action, we have some concomitant experience (perhaps only dimly felt) of causal agency, the feeling that "this is me who is deciding to do this." The representational content of this feeling of causal agency is usually not the feeling itself; rather, the feeling represents some causal process. For that reason, the experience of agency seems to me to be a first-order phenomenal judgment, i.e. a judgment whose representational content does not pertain to phenomenal consciousness per se, and thus a judgment that is straightforwardly susceptible to being wrong (the feeling of agency might just be a persistent illusion). So I wouldn't put agency in the same category as the knowledge paradox, since the central problem of the knowledge paradox revolves primarily around second- and third-order phenomenal judgments.

In any case, I am not sure why we should think that the Carriers Theory of Causation rules out interactionism (consciousness-brain states) if that interaction might take place in a way that respected the way physical causation works.

Interactionism is fundamentally ruled out for the Theory of Natural Individuals simply because of the manner in which TNI describes the causal mesh and phenomenal experience's place in that mesh. The carriers are the intrinsic, categorical basis of all the world's phenomena, and the world's physical interactions are nothing but the systems of internal contrasts obtaining among those carriers. On this picture, p-consciousness cannot interact with brain states because p-consciousness essentially already is the brain states, or rather, p-consciousness provides the intrinsic basis that allows the 'bare difference' structure of the physical brain to exist in the first place. Saying that the one could interact with the other, in the straightforward physical sense of the word 'interact,' amounts to a category error.

Perhaps an imperfect analogy would be of help. Carriers perform a causal role somewhat like material causation, while their internal contrasts (the effective properties) perform the causal role of effective causation. Imagine that the causal mesh is a tapestry that has a regular pattern of colors upon it. On this view, the carriers are something like the fabric, the actual material from which the tapestry is made. The effective properties are more akin to the colors on the tapestry, and the rules that govern what forms the patterns take are like the laws of physics. We might say that some parts of the pattern 'influence' or 'interact' with other parts of the pattern, insofar as their mutual presence entails the existence of some other pattern elsewhere, according to rules by which the overall pattern is governed.

But note that the role of the fabric itself is not to influence the shape of the pattern. Rather, the fabric is what gives those abstract patterns a concrete existence; it is that which implements the pattern, while not directly participating in the rules governing the pattern themselves. The patterns of color exist only insofar as they are properties of the fabric, but the rules that govern the form of the patterns are only sensitive to the colors of the fabric. There is not, and indeed there cannot be, any form of 'influence' by which the fabric itself directly informs the shape of the pattern. By contrast, one part of the pattern must naturally 'influence' another part to be such-and-such a way, since we know that there is a greater, overall pattern whose form follows certain rules.

I'm not sure if that analogy will be helpful or confusing, so please think it over a bit and tell me how it comes across.

 

[antfm]

Thanks always, Hypnagogue.
Your post is, as usual, full of helpful explanations and also of new insights. I'll try to readress the points that I mentioned. As I said, all of them are more or less related to the question of how causation (in Rosenberg's sense) works. With this I'm referring to the suggestions that Rosenberg himself makes:

"In our fundamental ontology, we will have to postulate all the physical facts and natural laws and, in adition to this, some set of metaphysical laws or metaphysical constraints that rule out such things as the zombie world."(p. 68)

"I cannot propose a concrete law for the emergence of configurations at higher levels. Yet it seems natural to suppose that the right law might be a function involving nature's dual concerns for maximizing entropy and completenes." (p.182)

I'm aware of how difficult it must be to try to hypothezise how causation in its deepest levels operate. But I feel that the above suggestions are not enough for me to get a more solid standpoint to understand how causation works.

As far as for the 'knowledge paradox', I think I understand the subtleness of its meaning. But I still keep thinking that there might remain a problem related to mental causation in general. If I gave the example that first came to my mind ("the thinking of rising my arm, and the rising of my arm"), it's not becuse I'd tried to involve the sense of agency, but rather the entailment of a conscious event and a physical outcome. I could have suggested many different examples: the feeling of a pain, and the taking of an aspirine; the thirst, and the search for a glass of water; the funny joke, and the outburst of laughter; the hallucinating of seeing a ghost, and the running away in fear; etc.
So, I don't understand well why focusing just on the 'kowledge paradox' as if the rest of features of conscious life were already solved. Even when analyzing this problem, we consider the case of a negative fact as causally responsible ("the girl didnt show up"), but, before that, the situation involves an expectation, a prediction of a future event (hopefully "the girl will show up"), that is, of a non actual fact (but a conscious event), as causally responsible of a whole physical behavior.

Of course, we can think that mental causation is illusory, or consider that it is the case of just physical events entailing new physical events. But it seems to me that there still remains a lot of explanation to be given.

On this picture, p-consciousness cannot interact with brain states because p-consciousness essentially already is the brain states, or rather, p-consciousness provides the intrinsic basis that allows the 'bare difference' structure of the physical brain to exist in the first place.

I can't help seeing a difficult gap between consciousness being the brain states, and consciousness providing the intrinsic basis for them. I really fail to see how those two domains effectively communicate or even less identify.

I am surely making a category error when thinking of a possible influence (though without violating the laws of natural causation) between conscious events and brain states, but I do not understand either when Rosenberg says:

"Even though consciousness is not physical, its activity underlies our physical nature as a carrier of our nomic content. Our physical states, although not causally interacting with our conscious states, track and therefore represent those states." (p. 265)

How does this 'tracking' and 'representing' take place? Is he not making a distinction between physical states and conscious states?

And here it is where another worry appears to me, because I do not know what exactly 'consciousness' is within TNI.

Is consciousness the resulting effective state of a natural individual? I do not think TNI says so. The effective state seems to me to be just a purely physical outcome of the deeper process of metaphysical causation.

Is consciousness the aggregate of receptivities that constitute a natural individual? I think TNI says so. But then, consciousness is not something identifiable with a physical outcome but just associated to it. To express it more crudely, that kind of consciousness cannot be located in a physical framework (though we might observe its effects, or correlates, in such a framework). It seems to me that in this picture consciousness outruns the physical description of the brain states, because it lies on receptivities rather than on effectivities. Then, consciousness is open to new connections, and so it keeps being causative - and in the long run physically causative.

Another whole story is if we are able to elucidate how that causative process takes place, particularly in the domain of the causative story where it is non-physical.

And finally, Hypnagogue, I really appreciate your efforts for trying to make this whole subject more understandable to me. I think that the analogy of the tapestry is a good one. But I think I get lost again when I try to apply it to the description of the functioning of consciousness, and to concrete exemplifications of consciousness, brain states, and their underlying receptivities.

But, please Hypnagogue, this is not a failure of your communicating skills, that are just great. It is obviously a clear proof of my intellectual shortcomings. Thanks again.

 

[Tychic]

one or two receptivities?

Hi Hypnagogue, thank you for your very interesting answer (8/8/2005), I had not yet time to read the following thread but here is my response.
Rosenberg's connection between receptivity and phenomenal properies is the following:

For instance, suppose we have an individual at level n+1 that binds two level n individuals, e.g.

I_n+1.1: [I_n.1 => I_n.2]

Suppose that I_n.1 is determinate, and has some effective property E that places causal constraint upon the effective state of I_n.2. Then we could say that I_n.2's receptivity experiences the phenomenal carriers of E. And furthermore, the receptivity of I_n+1.1 is required to 'deliver' the phenomenal carriers of E to I_n.2. Both receptivities are thus needed in order for I_n.2 to experience any phenomenal properties.

It seems to me that you are right, when you think that this implementation of the carrier theory is less easy as it could be. I remember the two models of receptivity pictured in charpter 9, p. 157: receptivity as one side of the properties of an individual and receptivity as connection. Both of this conceptions are implemented in the theory of the book, but also the alternative conception of receptivity you propose:

However, in e-mail correspondence Rosenberg has expressed some misgivings to me about this view. An alternate view is that individuals at level n experience the effective properties of the level n-1 individuals that they receptively bind together. On this view, I_n.2 would not experience the effective constraints placed on it by I_n.1; rather, it would experience the carriers of the effective properties of the level n-1 individuals that it binds. Likewise, I_n+1.1 would experience the phenomenal carriers of I_n.1's effective properties. (Since I_n.2 is not placing causal constraint in this nexus, it has no effective properties within the context of this nexus.) Notice that with this view, only one receptivity is needed for experience of phenomenal properties to occur.

Both views have their strengths and weaknesses. I find the latter view to be somewhat more intuitive and more consonant with the rest of Rosenberg's theory.

“More intuitive”: I agree – “more consonant with the rest”: I decline.
I think also that it would be preferable that an individual of the highest level would have phenomenal properties. But the price of this alternative view is that the level 0 individual cannot have phenomal properties even if it is determined. A second argument against this alternative view is that knowledge by acquaintance could not be further identified with a phenomenal property.
What are the misgivings of Rosenberg's view that carry more weight?

(My own suspicion is that the cortical experiencing subject is not identical to the cognitive self-construct, and also that the process of choice involves the interaction of many natural individuals in the brain and so cannot be attributed to a single 'subject,' in the natural individual sense of the word.)

Why do you think that a cognitive self-construct is no natural individual? The self-construction is a conscious act and each conscious human action/event with the duration of a moment of thinking is (as far as I understand Rosenberg’s theory) a prototype of a natural individual. There is intrinsitiy and unity when I think something of myself.
I suppose your position in this question is a consequence of your position about receptivity.

 

[hypnagogue]

"In our fundamental ontology, we will have to postulate all the physical facts and natural laws and, in adition to this, some set of metaphysical laws or metaphysical constraints that rule out such things as the zombie world."(p. 68)

"I cannot propose a concrete law for the emergence of configurations at higher levels. Yet it seems natural to suppose that the right law might be a function involving nature's dual concerns for maximizing entropy and completenes." (p.182)

I'm aware of how difficult it must be to try to hypothezise how causation in its deepest levels operate. But I feel that the above suggestions are not enough for me to get a more solid standpoint to understand how causation works.

The first quote is pointing out that our metaphysical framework will have to include some laws or constraints above and beyond those recognized by physicalism. This is required because, when we tell the final causal story about the human brain and consciousness, we do not want our explanation of the causal dynamics of the brain to be logically consistent with an absence of consciousness. If our explanation is logically consistent with the absence of consciousness, that surely implies that we have missed some crucial part of the story. The metaphysical law that Rosenberg ultimately proposes to address this point is, of course, the Central Thesis. Introducing the Central Thesis into the causal story eliminates the specter of logically possible zombie worlds; any causal story about the world told in terms of the Theory of Causal Significance and the Central Thesis will not be logically consistent with an absence of phenomenal consciousness.

The second quote is some speculation for what features might guide nature's construction of higher-level natural individuals. Given that the Causal Theory of Significance is true, what governs the manner in which nature chooses to create higher-level individuals? This is an important question for the causal framework, and one that must go unanswered for now. Rosenberg offers speculation that higher-level individuals might be constructed in such a way as to maximize entropy, or completeness, or both. But essentially, this remains a gap in the greater theory that cannot yet be patched up with a high degree of confidence. So if your concern is that TNI, as developed in the book, is not yet a complete causal model of the world, then you are quite correct.

As far as for the 'knowledge paradox', I think I understand the subtleness of its meaning. But I still keep thinking that there might remain a problem related to mental causation in general. If I gave the example that first came to my mind ("the thinking of rising my arm, and the rising of my arm"), it's not becuse I'd tried to involve the sense of agency, but rather the entailment of a conscious event and a physical outcome. I could have suggested many different examples: the feeling of a pain, and the taking of an aspirine; the thirst, and the search for a glass of water; the funny joke, and the outburst of laughter; the hallucinating of seeing a ghost, and the running away in fear; etc.
So, I don't understand well why focusing just on the 'kowledge paradox' as if the rest of features of conscious life were already solved.

I think cases like these are already covered, in a general sense, by the Consciousness Hypothesis. To illustrate a few examples:

* Taking aspirin modifies the behavior of neural systems P in the brain that register pain. If the effective properties of P are modified, it follows that the phenomenal carriers of P's effective properties are likewise modified, since the effective properties are just features of the carriers themselves. (Presumably, the phenomenal carriers are modified in such a way as to reduce the degree of experiened pain.)

* We know from brain imaging that thinking of raising one's arm is correlated with heightened activation in certain motor systems in the brain. The TNI explanation is that the phenomenal experience of thinking of raising one's arm carries the effective properties that are shown to be active in the brain imaging. That is, the activation detected in the brain imaging is just the outward, physical manifestation of the corresponding phenomenal experience. So it's not that the experience effectively caused the brain activation, or vice versa; rather, it is the case that the two are in some sense the very same entity, merely looked at from two different viewpoints (subjective vs. objective, internal vs. external).

So in general, TNI does not have problems explaining the correlation between phenomenal events and physical events. This correlation follows directly from the Consciousness Hypothesis. On this account, we do not have to puzzle about how it could be that a physical event effectively causes a phenomenal event, or how it could be that a phenomenal event effectively causes a physical event. TNI says that the link between the two is not one of effective causation, but rather one of material causation. So it is not surprising at all to find that phenomenal and physical events are so tightly correlated, since they are essentially the same thing, seen from two different viewpoints. A phenomenal pain P, as seen to an external, third person observor, literally looks like a neural manifestation N. So insofar as N literally is just P as seen through the obscuring epistemic veil of the third person, it is of course the case that if N changes, P must change also.

Here is another metaphor that might be useful. Suppose we think of a neural event N and its phenomenal correlate P as two aspects of a puzzle that we desire to put together in order to come to a natural explanation of consciousness. If we view the link between N and P as one of effective causation, we will think of N and P as two separate puzzle pieces that we must somehow fit together. Of course, it is notoriously the case that we cannot do this; conceived as puzzle pieces, the two do not have the appropriate types of edges required in order to fit together seemlessly.

The TNI approach rejects this way of viewing the puzzle, holding that the appropriate link between phenomenal and physical events is not one of effective causation, but rather one of material causation. Instead of conceiving of physical events and conscious events as two separate puzzle pieces, TNI conceives of one single puzzle piece, whose different features correspond to phenomenal and physical aspects. Specifically, phenomenal properties correspond to the material from which the puzzle piece itself is made, and physical properties correspond to the shape of the puzzle piece, which is what defines how it might interact and link up with other puzzle pieces. (To complete the metaphor, we might suppose that there are some manufacturing rules of the puzzle pieces describing what kinds of shapes might be constructed from what kinds of materials.)

The knowledge paradox poses special problems because the knowledge paradox is not just about why certain phenomenal events are correlated with certain physical events. Rather, the knowledge paradox asks us how it can be that certain phenomenal events justify certain physical events, such as utterances about consciousness. This is a difficult problem indeed, and one that goes beyond merely explaining why phenomenal and physical events are correlated at all.

Of course, we can think that mental causation is illusory, or consider that it is the case of just physical events entailing new physical events. But it seems to me that there still remains a lot of explanation to be given.

Surely it is not the case that the Theory of Natural Individuals is already complete and battle tested and checked by rigorous empirical methods and so forth. We should expect that there are some gaps in the theory, such that not everything is capable of explanation in any great degree of detail or confidence. But at the same time, TNI does appear to offer enormous explanitory promise. At this early stage we should fully appreciate the conceptual problems it can clear up and the novel predictions it can offer, while also being fully aware of its limitations and areas that need further development. There is indeed a lot of further work that needs to be done should TNI become a full-fledged theory of causation, but at the same time, even in the span of 300 or so pages Rosenberg has managed to cogently and elegantly address many of the deep, central problems regarding consciousness and causation. So while we should be realistic about what work remains to be done, we should also not overlook the profoundly significant advances that have been made.

I can't help seeing a difficult gap between consciousness being the brain states, and consciousness providing the intrinsic basis for them. I really fail to see how those two domains effectively communicate or even less identify.

Well, the technically correct claim is that consciousness provides the intrinsic basis for the existence of physical brain events. The claim about identity is mine-- I don't think Rosenberg has advanced that way of seeing things very much (if at all) in the book. Strictly speaking, it is true that TNI does not depict consciousness as being the brain states, but I think it's still a useful way to think of the relationship between the two, so as to emphasize the significant degree to which they are metaphysically unified. A more qualified way of putting it would be that certain brain events are 'surface features' of consciousness which are amenable to third person investigation.

To return to the tapestry metaphor, it would be incorrect to say that the tapestry is the colored pattern displayed upon it. Still, it's clear that the relationship between the pattern and its underlying tapestry is a significantly intimiate one that is more like identity than it is like the relationship between two separate objects. The technically correct way of putting it would be that the material of the tapestry carries, or provides a basis in existence for, the abstract patterns displayed upon it. Perhaps it would be best to say that the patterns literally are a subset of the properties tapestry as a whole, just as we might say that high-level brain events literally are a subset of the properties of consciousness as a whole.

"Even though consciousness is not physical, its activity underlies our physical nature as a carrier of our nomic content. Our physical states, although not causally interacting with our conscious states, track and therefore represent those states." (p. 265)

How does this 'tracking' and 'representing' take place? Is he not making a distinction between physical states and conscious states?

I have some conceptual difficulties with this claim too, as I pointed out in a previous post, so I'm probably not in a good position to try to clarify it or expound further.

And here it is where another worry appears to me, because I do not know what exactly 'consciousness' is within TNI.

Is consciousness the resulting effective state of a natural individual? I do not think TNI says so. The effective state seems to me to be just a purely physical outcome of the deeper process of metaphysical causation.

Is consciousness the aggregate of receptivities that constitute a natural individual? I think TNI says so. But then, consciousness is not something identifiable with a physical outcome but just associated to it. To express it more crudely, that kind of consciousness cannot be located in a physical framework (though we might observe its effects, or correlates, in such a framework). It seems to me that in this picture consciousness outruns the physical description of the brain states, because it lies on receptivities rather than on effectivities. Then, consciousness is open to new connections, and so it keeps being causative - and in the long run physically causative.

According to TNI, consciousness is neither effective properties nor receptive properties. Rather, (human) consciousness is the phenomenal and experiential carriers of the high-level effective and receptive properties of some natural individual in the brain. It's important not to identify effective and receptive properties (E+R) with their carriers. You can think of the terms "effective properties" and "receptive properties" as referring to the functional behaviors and interactions of phenomenal and experiential properties (Ph+Ex), respectively. Those features of Ph+Ex that are not catalogued by E+R are the intrinsic, qualitative aspects of Ph+Ex-- what it's like to experience a sunset and so on.

You are correct to point out that on TNI, consciousness outruns (is not logically entailed by) physical descriptions of the brain, but this is not due to considerations from receptivity. The picture looks something like this:

* Receptivity outruns effective properties
* Phenomenal properties outrun effective properties
* Experiential properties outrun receptive properties

The qualitative features of consciousness do outrun physical descriptions of the brain, but they also outrun the pure effective and receptive descriptions given by the Theory of Causal Significance. So the qualitative features of consciousness are brought back into the causal story not by the Theory of Causal Significance, but rather they are introduced by the Central Thesis.

But, please Hypnagogue, this is not a failure of your communicating skills, that are just great. It is obviously a clear proof of my intellectual shortcomings. Thanks again.

Nonsense, don't feel bad if everything doesn't click right away. This is pretty difficult material, and it's not unusual to have difficulty with it, especially if one doesn't already have a fairly extensive background in the philosophy of consciousness. I encourage you to keep at it, because you will be greatly rewarded once everything does click into place.

 

[Tychic]

The subjective instant:
There are some difficulties to face the remaining problem after the functional part of solved.

The subjective instant can be thought of as a temporal analogue of the grain problem, and thus could be grouped in the previous section. However, I list it separately because I'm not quite as convinced that the subjective instant needs a treatment in terms of receptivity. I think Rosenberg's discussion on the subjective instant is one plausible way in which the problem might be solved, but I also think that the problem might yield to simpler considerations. For instance, it could be that there is some neural system N whose activity in a single instant reflects or encodes information that is fed into it over some brief duration of time in the past. If experience of the subjective instant corresponded to activity in N, then the problem of the subjective instant seems to be solved rather straightforwardly.

(Rosenberg argues in chapter 7 that the central problem of the subjective instant is not the timing of neural events vs. the timing of experienced events per se, but that the latter seems to have a kind of privileged frame of reference while the former, according to relativity, does not. However, I'm not convinced that this is really a problem, since practically speaking the neurons in the skull always occupy the same reference frame. [...]).

I do also not think that transformations of reference frames are a problem - but reference frames are a problem for an other reason: In physics there is (as far as I know) no frame of reference that two events occur simultaneously that do not occur simultaneously in another frame of reference. But it is possible and corresponds to natural assumption that physically non simultaneous (brain) events can appear to consciousness simultaneously in an subjective instant. And a subjective instant of consciousness is the prototype for a natural individual. In my opinion all ideas about natural individuals have an analog (and have to be compatible with) subjective instants of a human person.

Your proposal for an subjective instant does not respect this prototype and proposes that it is no natural individual that experiences the subjective instant. So I would think that Rosenberg’s proposal and not yours is more simple. Receptivity is needed to make some physically not synchronal events simultaneously. But perhaps there are some misunderstandings of my side. Do you think that N could or should be a natural individual?
Your proposal also presupposes that duration exists. This is implemented in Rosenberg’s proposal in the lover level cascades. Perhaps it could be implemented in an other way with another theory of spacetime, but Rosenberg’s example corresponds to the most simplest case with his spacetime-model.
The problem with Rosenberg’s proposal is to find the subcortical individuals and to argue for his theory of time. Do you know a comparable simple theory of time that fits into his theory?

Knowledge Paradox:
It could be that I do neither understand Rosenberg’s proposal nor your critic, hypnagogue, but I try to tell the schematic story I made for myself with an example:
Suppose that I_n.1 is a conscious cognitive individual, an instance of the conscious live of a human person particularly. I_n.1 has a phenomenal property R, phenomenal redness. Perhaps there are individuals of lower levels that are lower level parts of I_n.1 which have the same (or a very similar) property R. Let I_n-1.1 be an individual with the property R. Each of I_n.1 and I_n-1.1 shares a receptive binding with some individuals that are not part of I_n.1 so that phenomenal red is experienced of each of the two individuals. It may be that the same reactions happen physically to I_n.1 and I_n-1.1 in respect of R because the causal chain starting from the red object and ending in this brain of a human person does work on a lower level of natural individuals as n. No human person can look inside I_n-1.1 what R is phenomenally like. On the first look I imagined that the carrier on level n-1 is the same as the carrier on level n, but it is not necessary, perhaps even misleading. Now I think that it is wrong: Each effective property depends on the receptivity to which it is bound and the two receptivities of the two levels differ. So let me speak instead of one phenomenal property R of the two: R_n and R_n-1. I hope that I now have said something to the concept of the phenomenal vehicle: it is the common phenomenal content of these properties which they share with many other individuals of both levels.
The following situation would be possible: The receptivity/context of I_n-1.1 allows only knowledge by acquaintance of the phenomenal red: the effective part of the property R on this level is not cognitive, perhaps not even representational. The context of an human conscious instant I_n.1 allows cognitive effective sides of the property E: a first order phenomenal judgment “this is red”.
On Rosenberg’s proposal it is another receptivity on level n which makes the assumption “I’m having a red sensation” and a third receptivity which makes third order phenomenal judgments. The representation vehicle remains always the same.
Question: What are the decoders C? Is my solution, that they are receptivities that bind level-(n-1) individuals to level-n individuals too simple?

My evaluation: I do not think that the causal link between phenomenal experience and utterances about experiences is too loose, there is no closer connection as between effective and receptive properties, resp. their carriers. When phenomenal redness can be bound in different levels of nature in different ways, why shall it not be bounded in the same level also in different ways?
My problem is the constancy of the representation vehicle that is not motivated and guaranteed by nothing. Furthermore I ask how there can be a continuity of a human consciousness (and a human person) if the receptivity of her or his instances is always changing?

Your objection:

The danger that follows from this consideration can be illustrated thusly. Rosenberg has argued that it could be the case that two distinct phenomenal carriers might implement the same physical, effective structure. Suppose that there is some phenomenal carrier S whose intrinsic phenomenal nature is different from R, but that S's and R's effective structures are identical. Now suppose that S takes the place of R in the greater cognitive context, such that C is now reading off information from S rather than R. Even though S and R have distinct phenomenal properties, C will fail to differentiate the two and will treat them identically. If C is susceptible to such a trick, then it seems that the second-order phenomenal judgments that C produces cannot be justified. The resulting picture looks a bit like epiphenomenalism, if not with respect to the superfluity of consciousness, then with respect to the knowledge paradox.

Why do you think that this phenomenon covers more than the fact that verbal expressions cannot fully satisfy the richness of phenomenal experiencing?

 

[hypnagogue]

An alternate view is that individuals at level n experience the effective properties of the level n-1 individuals that they receptively bind together. On this view, I_n.2 would not experience the effective constraints placed on it by I_n.1; rather, it would experience the carriers of the effective properties of the level n-1 individuals that it binds. Likewise, I_n+1.1 would experience the phenomenal carriers of I_n.1's effective properties. (Since I_n.2 is not placing causal constraint in this nexus, it has no effective properties within the context of this nexus.) Notice that with this view, only one receptivity is needed for experience of phenomenal properties to occur.

Both views have their strengths and weaknesses. I find the latter view to be somewhat more intuitive and more consonant with the rest of Rosenberg's theory.

“More intuitive”: I agree – “more consonant with the rest”: I decline.

Let's call the view of how receptive structure is related to experiencing developed in the book the "dual receptivity" view, or DR for short; likewise, let's call the opposing view described above the "single receptivity" view, or SR for short.

There are several related reasons why I think SR is more consonant with the rest of Rosenberg's framework than is DR; for simplicity and clarity, I will only explain what I think is the most straightforward and powerful objection.

The following two premises are important ones in Rosenberg's development of the Central Thesis in chapter 12:

1) All instances of receptivity are carried by experiential properties.
2) Experiential properties are compositionally circular with phenomenal properties in the same way that receptive properties are compositionally circular with effective properties.

(1) follows from the Central Thesis, and (2) is a crucial premise used to argue that the Central Thesis is true. If (1) is false, then so is the Central Thesis. If (2) is false, then much of the resonance between phenomenal and experiential properties and effective and receptive properties is lost-- we lose some of our motivation for formulating and justifying the Central Thesis in the first place, and the Central Thesis itself becomes significantly less elegant and compelling. Clearly, we want both (1) and (2) to be true, and so both premises should be consistent with whatever other claims we can make within the framework.

The problem with DR is that it is not consistent with both (1) and (2). Suppose that there exists a level n individual I_n.1 such that I_n.1 is not receptively bound within a causal nexus (such as individual 3.1 in figure 12.3). We know from the Central Thesis that I_n.1's receptivity is carried by an experiential property. We also know from the Theory of Causal Significance that I_n.1's receptivity is compositionally circular with the effective properties of the level n-1 individuals it binds together.

However, DR predicts that I_n.1 will not experience any phenomenal properties. If this is the case, then I_n.1's experiential property cannot be compositionally circular with any phenomenal properties, and premise (2) from above is violated-- although I_n.1's receptivity is compositionally circular with the lower level effective properties, I_n.1's experiential property is not compositionally circular with anything.

An easy way around this difficulty is just to reject DR and instead embrace SR. On SR, I_n.1 would experience the effective properties E of the level n-1 individuals it binds, and thus its experiential property would be compositionally circular with the phenomenal properties carrying E. In general, SR guarantees that for all receptive and effective properties that are compositionally circular, the corresponding experiential and phenomenal carriers are also compositionally circular in precisely the same way; it is impossible on SR to have an instance of receptivity R that is bound to lower-level effective properties such that the experiential property that carries R is not also 'bound' to some phenomenal properties.

I think also that it would be preferable that an individual of the highest level would have phenomenal properties. But the price of this alternative view is that the level 0 individual cannot have phenomal properties even if it is determined.

Level 0 individuals can only be determinate if they are bound together within a common receptivity R into a level 1 individual. In this case, SR predicts that R should experience the phenomenal carriers of the level 0 effective properties.

Even if the level 0 individuals are not bound together within a common receptivity, they are still carried by phenomenal properties-- it's just that those phenomenal properties are not experienced by anything. This claim follows directly from the Central Thesis, so whether we choose to adopt SR or DR is irrelevant to this particular issue.

A second argument against this alternative view is that knowledge by acquaintance could not be further identified with a phenomenal property.

Could you explain what you mean by this some more?

Why do you think that a cognitive self-construct is no natural individual? The self-construction is a conscious act and each conscious human action/event with the duration of a moment of thinking is (as far as I understand Rosenberg’s theory) a prototype of a natural individual. There is intrinsitiy and unity when I think something of myself.

In the thread for chapter 12, I argued that the experiencing subject described by TNI should be taken to be distinct from the cognitive self-construct:

Before I proceed, I must mention an important reminder: The "subject" Rosenberg talks about here is something of a special term, and does not reflect colloquial usage of the word. In particular, the relevant subject in question is not co-extensive with a human's cognitive construct of self. For instance, if we were to talk about a human as a subject in this context, we would not talk about a fellow who has such-and-such personality, memories, interests, skills, worldviews, and so on. Rather, we would talk about a thing that experiences a bounded unit of phenomenal qualities-- an experiential manifold, or a "qualitative field" as the term was introduced in chapter 5. This sense of the word "subject" is quite remote from the sense of the word where "subject" means "a cognitive self-construct."

A good exercise here might be to think of oneself in terms of one's own cognitive self-construct-- something we do naturally and reflexively all the time-- and then to think of oneself as that which experiences a qualitative field. There is a pretty stark contrast, I think, between the two. In particular, thinking of oneself purely as an experiencing thing-- a system for which it is like something to be-- immediately robs one of the familiar cognitive self-construct, or at least places the cognitive self in remote territory, now viewed externally and from afar rather than comfortably and transparently lived in. Consideration of these differences should begin to point to the ways in which the two senses of the word "subject" described above are substantially different.

The cognitive self-construct is a kind of cognitive model each of us creates about ourselves. It is something that informs our actions and beliefs probably every moment of every day, but something that works largely implicitly, that is, unconsciously. It is not identical to our moment-to-moment unitive qualitative field, but rather is a kind of largely invisible cognitive structure built around that conscious core.

Although the cognitive self-construct has a ubiquitous effect on our everyday lives, I believe it can be thought of similarly to any other unconscious cognitive mechanism, such as the mechanism that controls the fine motor movements of my fingers as I type-- essentially, it poses an 'easy' problem rather than the hard problem. It might be that the cognitive mechanism that encodes the self-construct is a natural individual, but I see no more compelling reason to believe this than I do reason to believe that the mechanism that controls the fine-grained details of muscle contractions in my body, or any other unconscious cogntive mechanism, is a natural individual.

 

[Tychic]

single receptivity (SR) or dual receptivity (DR)

Thank you, hypnagogue, for clarifying my concepts and making me see the problematic points in Rosenberg’s terminology.
According to the central thesis all natural individuals can experience other individuals. This is true in SR (they really do in all higher level individuals, they can for level-0 individuals by becoming part of higher level individuals). But in DR this is only true for indeterminate individuals which have a qualitative field. Concerning figure 12.3 (p. 244) in the book Rosenberg writes that “the top-level individuals [individuals like Individual 3.1] do not experience because they do not belong to causal nexii from which they can obtain a receptive field” (p. 245-6). But can they experience other individuals in DR? I see no promising positive answer. It could be that each determined individual could be determinable in some respect and that the law of trans-level-receptivity allows in principal that for all levels there could exist a level above which includes symmetric bindings. But I do not think that the “can” in the Central Thesis comprises these kinds of contrafactual possibilities. A similar problem arises for Individual 2.1 in Rosenberg’s example (similar problem as I_n.1 in your example).
Therefore I think that Rosenberg really accepts only a modified Central Thesis:

All natural individuals with a receptive field (= individuals that are bound in a higher level receptivity symmetrically or as a consequence) can experience other individuals.​

Double receptivity (individual constituting receptivity of the parts and receptive field with the environment) is presupposition for experiencing phenomenal contents. I accept that the original Central thesis would be preferable and SR seems to be compatible with the original Central Thesis.

The following two premises are important ones in Rosenberg's development of the Central Thesis in chapter 12:

1) All instances of receptivity are carried by experiential properties.
2) Experiential properties are compositionally circular with phenomenal properties in the same way that receptive properties are compositionally circular with effective properties.

Even with this modified Central Theses premise (1) of your proof has to be accepted. I also accept your argumentation. So I think I have to deny premise (2). It would be possible to complement “at least below the level of completion:”

The price of DR is as you write that

much of the resonance between phenomenal and experiential properties and effective and receptive properties is lost-- we lose some of our motivation for formulating and justifying the Central Thesis in the first place, and the Central Thesis itself becomes significantly less elegant and compelling.

Perhaps I can alleviate the problem: I would deny that there is only one compositional circularity in the definition of effective and receptive properties. The hierarchy of individuals is also included in the definitions of effective and receptive properties. With a modified premise (2) the mirror functionality on the middle levels is preserved. Does SR avoids this second circularity? I think it cannot avoid it without changes in Rosenberg's theory.

My first argument against SR is invalid as you have shown and I withdraw it.
Concerning my second argument:
I do not like SR because of my idea of the contact of an individual with the environment. It seems to me that in SR a individual can only experience (the properties of) it’s own parts and no part of the environment.

A perceiving individual Pete may be level n, the parts may be level n-1 and below. There could be in principle two possibilities for influence from the environment in SR.
(a) Pete is bound in an individual of level n+1. But this bounding is not experienced in the level n individual. So this binding cannot bear any contribution to perceiving of Pete.
(b) A level n-1 sub-individual of Pete is also bound in a n individual Pam that is not Pete and shares some individuals that are not part of Pete. An individual like Pam could connect some parts of Pete to the environment. But the connection could only put an effective constraint to the n-1 individual because the phenomenal part of this constraint belongs to the n individual Pam, not to the n-1 individual. Hence no phenomenal carrier of the environment can be experienced by Pete who experiences the n-1 individual.​

I propose a third possibility. An individual can experience individuals of the same level and the level below. But I have to reflect more on this topic.

Thank you for explaining the idea of the self as a qualitative field again. The cognitive self construct is really sometimes entirely unconscious and in usual cases only some parts of it make some contribution to an actual subjective instant. Anyway, I'm not very happy about this thesis but it fits in Rosenbergs framework.

 

[hypnagogue]

According to the central thesis all natural individuals can experience other individuals. This is true in SR (they really do in all higher level individuals, they can for level-0 individuals by becoming part of higher level individuals). But in DR this is only true for indeterminate individuals which have a qualitative field.

The last sentence should read, "But in DR this is only true for indeterminate individuals which have a receptive field." Your mistake might have been a slip of the tongue (so to speak) rather than a real misunderstanding, but it's important to clarify it in either case.

Perhaps I can alleviate the problem: I would deny that there is only one compositional circularity in the definition of effective and receptive properties. The hierarchy of individuals is also included in the definitions of effective and receptive properties. With a modified premise (2) the mirror functionality on the middle levels is preserved.

That is an interesting response. The relationship between effective properties and receptivities at the same level of nature is one that involves the completion (determination) of those properties through transitivity (as discussed in chapter 9), where the effective and receptive properties mutually contribute to each other's completion. I can see that being a compositional circularity of a sort.

However, at least on my reading of the text, the dominant compositional circularity is that which obtains between level n effective properties and level n+1 receptive properties. This circular relationship is required for the very existence of the relevant effective and receptive properties, rather than merely their completion. And it is also the case (as we have discussed) that the intra-level transitivity of effective and receptive properties is dependent upon the existence of inter-level binding of level n individuals into a level n+1 receptivity, and in this way the former is ontologically subordinate to the latter.

I do not like SR because of my idea of the contact of an individual with the environment. It seems to me that in SR a individual can only experience (the properties of) it’s own parts and no part of the environment.

Agreed, but I don't think this is a problem-- in fact, I think it is more in agreement with the evidence. Does a brain really experience its environment? If so, it only does so indirectly, by means of an extended causal chain (e.g. light reflecting off a surface and striking the retina, photons being transduced into neural signals and then being sent off to the thalamus for further processing, etc). What the brain seems to experience directly is just its own internal states. Stimulate some part of the brain in the proper way and the subject will have an experience as if some visual event were occurring in the external world, when in fact the relevant cause was limited in scope to the brain itself.

I propose a third possibility. An individual can experience individuals of the same level and the level below. But I have to reflect more on this topic.

This possibility has struck me too, as a kind of compromise between SR and DR. It might be a viable third option.

 

[Rade]

Recall the definition of the Central Thesis:The Central Thesis: Things in the world are natural individuals if, and only if, they are capable of experiencing phenomenal individuals. (p. 241)

I find that your Central Thesis is falsified by quantum mechanics. Thus, let the deuteron (a very stable cluster of proton + neutron) be a "natural individual", given that it is naturally present in the universe, and it is clearly an individual wavefunction as described by quantum mechanics. But, of course the deuteron has no desire or need based on energy considerations to move to a state other than "deuteron" (it is already at a very, very stable state) , likewise, it has no desire or need to "experience" (e.g., bond) with any other similar nucelon clusters (your concept of the other phenomenal individuals). It may such bond or experience, but it also may not. Thus your Central Thesis is falsified--because the deuteron is clearly a "natural individual" that has no need or desire to experience other "phenomenal individuals".

 

[hypnagogue]

I find that your Central Thesis is falsified by quantum mechanics. Thus, let the deuteron (a very stable cluster of proton + neutron) be a "natural individual", given that it is naturally present in the universe, and it is clearly an individual wavefunction as described by quantum mechanics. But, of course the deuteron has no desire or need based on energy considerations to move to a state other than "deuteron" (it is already at a very, very stable state) , likewise, it has no desire or need to "experience" (e.g., bond) with any other similar nucelon clusters (your concept of the other phenomenal individuals). It may such bond or experience, but it also may not. Thus your Central Thesis is falsified--because the deuteron is clearly a "natural individual" that has no need or desire to experience other "phenomenal individuals".

With all due respect, I get the distinct impression that you are not really familiar with the material. Have you read the book up to this point, or at least have you read the chapter summaries? If not, please familiarize yourself with the material before making claims about it.

 

[Rade]

All due respect back, but what would move me to read a book where I know a priori that its Central Thesis is based on a false premise ? I was just scanning the various posts on this forum, read about your hypothesis you called the "Central Thesis", then made a logical reply showing how it can be falsified using an example from quantum mechanics. Such is the way of science and philosophy--to falsify--sorry if you were offended. I look forward to you falsifying my comments.

 

[hypnagogue]

Well, for one thing, it is not a given that a deuteron is a natural individual. (You assume a generic interpretation of the term "natural individual" when in fact it is a special term within this framework.)

For another, it is not clear that you are using the words "experience" or "bind" in the same ways these are used in Rosenberg's framework. (It seems that, in fact, you are using them differently.)

Aside from the incorrect usage of these terms, it is also apparent that you are not familiar in general with the underlying theory of causation on which the Central Thesis is founded (e.g. you assume "binding" refers to something like the binding of atomic nuclei, when in fact it refers to something entirely different and more general within this framework). Without understanding this conceptual foundation, one cannot understand the Central Thesis. And without understanding the Central Thesis, one cannot make any meaningful claims about it.

 

[Rade]

To hypnagogue. OK, I will read your book and look into the following issues:
1. What exactly is a "natural individual" and does the deuteron of physcial reality meet or not meet the definition proposed.
2. How does Rosenberg define "experience" and "binding" and why is this definition different from that used in quantum mechanics.

So, where exactly on this forum do I access your book to read ?

 

[hypnagogue]

It's not my book, it's Gregg Rosenberg's book. You can't access the book online, however I have written chapter summaries that are available in this forum. To see them all you may have to change the forum display options. In the box titled "Display Options" at the bottom of this forum's main page, click on the drop-down box titled "From the:," and select the option titled "Beginning."

edit: Also, please see the stickied thread titled https://www.physicsforums.com/showthread.php?t=59689 for more general information on what the book is about and what guidelines to keep in mind for this forum.

 

[Tychic]

SR, DR and other possibilities

The last sentence should read, "But in DR this is only true for indeterminate individuals which have a receptive field." Your mistake might have been a slip of the tongue (so to speak) rather than a real misunderstanding, but it's important to clarify it in either case.

Thank you, hypnagogue for your exact reading. In DR is the qualitative field of an individual falls together with its receptive field (p. 253), but not in SR.

What effective/phenomenal properties can be experienced by an individual of level n? There are five simple possibilities (and additional the combinations of these)

a) the effective properties of its level n-1 parts which put causal constraint in the nexus
b) that the level n receptivity puts a constraint on the causal connections of the level n-1 parts
c) the experience of being a unity besides other individuals of the same level
d) the effective constraints working on the level n individual by other level n individuals
e) that these effective properties are constraint by a level n+1 receptivity (if existing).​

The included individuals in the five levels are beside the experiencing level n individual:

a) its level n-1 parts
b) no other individuals (Context: the level n-1 parts)
c) no other individuals (Context: other level n individuals)
d) other level n individuals
e) a level n+1 individual from which the n individual is part​

The options (b) and (c) correspond to experiencing/receptive dimension of the individual. What is experience more like? Be (b) a unity out of parts or to be (c) a unity in contact with other unities? I’m not sure about this and think that both aspects are relevant.
The possibilities for the phenomenal experiencing are:

• Single Receptivity SR, i.e. (a, b) or (a, b, c)
• Dual Receptivity DR, i.e. (c,d) or (b, c, d)
• the proposed third solution, let’s call it dual phenomenality DP, i.e. (a, b, c, d).
• and the last two option with an addition MMM of the experience of the individual of the level above (e).​

DP proposes two kinds of receptivity. What duality of phenomenal properties could correspond two the duality of receptivity?
- the duality of active and passive: It is implemented in the system jet (via asymmetric constraints) and it is therefore no good candidate for a duality of phenomenal properties
- the duality inside/outside: As you have noticed in your last post concerning these themes, even all perceptions are inside in some sense so that there are no experiences of the outside. But something in the outer world has to correspond to some experiences if you want to give realism a good foundation.
- the duality whether the contents of experiences are accuracy conditions from the inside or the outside of the level n individual. But does this fit to the causal story of e.g. a perception?

Thank you, hypnagogue, for strengthening my argument with the second compositional circularity that could save DR against the argument that there is no parallel of circularity in the effective/receptive distinction of DR. I describe the second circularity that remains on one level as follows:
For an (indeterminate) natural individual to exist, it must have effective properties; and for effective properties to exist, there must be individuals that bear these properties.

The remaining defence of your argument against DR is the following:

However, at least on my reading of the text, the dominant compositional circularity is that which obtains between level n effective properties and level n+1 receptive properties. This circular relationship is required for the very existence of the relevant effective and receptive properties, rather than merely their completion. And it is also the case (as we have discussed) that the intra-level transitivity of effective and receptive properties is dependent upon the existence of inter-level binding of level n individuals into a level n+1 receptivity, and in this way the former is ontologically subordinate to the latter.

One can characterize this ontological dominant compositional circularity as part/whole-circularity and the second as subject-attribute-circularity. In my opinion the examples of compositional circularity that Rosenberg detects in the sciences (p. 235-6) belong to the second kind of circularity: E.g. in economics “goods and services” are more properties of “consumers and producers” as they are there parts of them. A compositional circularity of the kind you prefer would be between an economy and its goods and services. But Rosenberg chooses not your “circularity”. (Your "circularity" is not a circularity at all in the economical analogy.)

P.S. The whole economy puts some constraints on the consumers and producers. Do they experience them less then the concrete goods and services that are too less or too much? If not, this analogy would argue that (e) has to be integrated in the scope of phenomenal experience. But I know that this analogy is only a week argument.

Are there some further good arguments for SR or DP?

again to the knowledge paradox:
1) Let us make the assumption that the physical world is causally closed. Chalmers' intuition is that a Zombie would make the same second order phenomenal judgments (in “The Conscious Mind”). In TNI no Zombies exist and the phenomenal and experimental properties are the carriers which allow widening their responsibility. I like your anti-Chalmersian intuition

that phenomenal experience must make a direct difference to the brain's effective dynamics in order to justify second- and third-order phenomenal judgments, but it has not completely eroded those intuitions.

2) You share with Chalmers the intuition that effective causality is not a link intimate enough for second-order judgments.

Rosenberg's analysis provides a causal link between phenomenal experience and utterances about experience, but it's not as intimate as my intuitions might like-- it seems somewhat too loose and coincidental. For instance, take the case where some cognitive decoder C uses the information in some phenomenal carrier R to represent R itself. It seems that C's access to information about R is underwritten by effective causal interactions between the two. If that is the case, there seems to be a disconnect between the intrinsic phenomenal information in R, and the effective, structural information in R. C seems to be accessing R's effective structure rather than its intrinsic phenomenal content, and we've argued at length that the former cannot entail the latter. So it seems that even though C is using R to represent R itself, this still does not put C in position to send out neural signals on the basis of R's intrinsic phenomenal content.

I two another possibilities: C can be an aspect of a receptive property that constitutes the (human) individual that has R as phenomenal property or the individual of the level above. Hence no causal interaction is necessary for an second order judgment “I experience R” but only actualization.
One can answer that the effective side of the representational vehicle R is something like “the truck is red”, “is red” or something similar but it is impossible that a phenomenal property carries more then one effective properties. But this possibility is at least in a DR-like model (if your refutation were wrong) realized in an easy way. One can have the same phenomenal quality as a result of a real perception or as a result of an illusion. Hence there are two possibilities for different knowledge of the phenomenal carrier R. (1) The individual that realizes R does sometimes integrate C in its receptivity and sometimes not. (This would be only a shifting of your argument against Rosenberg’s solution of the knowledge paradox.) (2) The individual of the level above (roughly: the environment) makes different constraints which both help to constitute an effective constraint in the human individual which is carried by R. And these implementations in the environment correspond to different kinds of knowledge on the side of the human individual.

May this all be as I propose: Is this knowledge that is known with certainty and how can second order-judgments be known with more certainty as first-order judgments? The certainty comes from the acquaintance with phenomenal properties in both kinds of judgments. I shifted the difficulty of certainty to different kinds of receptivity. Of course I cannot solve all problems but perhaps the problem has become smaller.

 

[hypnagogue]

Sorry for another long delay-- I've been quite busy gearing up for the school year and doing various other projects here on PF, among other things. I'll have a summary for chapter 14, and will be able to resume discussion in this thread, within the week.

 

[hypnagogue]

Hey all-- Thanks to those of you who are still following along. My apologies, but I've been very busy and haven't had much time to write up a chapter summary to my satisfaction. As it stands I won't be able to have one up until Wednesday at the earliest. In the meantime, if you wish, remember that you can always start up a new thread here-- you don't have to wait for me. But I will do my best to have the next summary up before next weekend.