What is a back plate to networks

The Blank Slate Internet (BSI) is a concept that has developed based on our experience with the current Internet and current software development. This experience suggests that certain key enhancement to our economic and social system might be easily developed.

There are very hard issues to be sure. So most of the time we frame our discussion as conjecture, particularly when talking about human consciousness or social system phenomenon like economic realities. The world economic system is a marvelous system but may produce unsustainable pressures on our social systems and the world’s environmental systems. An evolution of a framework and infrastructure for information exchanges is suggestive of changes that would seem to be positive in regards to modifying wasteful consumption patterns. So our discussion is about how this evolution might be aided.

Core to the current economic system is the information that flows in the Internet. The observed un-sustainability is resilient because expectations are established based on a specific type of business context. It seems not to be a closely held secrete that social values are seen, in this context, as secondary to the specific type of private economic gain supported by current property law. The results of this context are everywhere evident. As such, a new system is sought that establishes a blank slate for computing and communication using microprocessors. If this system is sufficient to allow and enable a cultural shift, then positive benefits may be found in supporting precisely such a blank slate.

The “second school” may conceivably enable a new economic model that is resilient and sustainable. One key to the second school is based on a correspondence between design principles and natural science. Several elements are present, including a paradoxical principle of transparency and encapsulated informational security. This principle may be realized using the back plate as described in this paper.

Our current design work shows precisely how transparency and security can be mutually supportive. A social agreement over what is to be publicly transparent, and transparent within social units, may be made and provable security over information content provided. Such an agreement has to be politically empowered and also to have a technology that is neutral to ownership issues. This means that the software itself has to be optimal and provisioned by our social institutions, and the agreements have to have grounding in Constitutional law.

Social agreements, of the type we envision, simply extend constitutional law and are thus enforceable in cases of attempts to infer and gather private information. However, public information may become more open and transparent. In particular real time public information about the composition of all manufacturing and all commodity use must be seen as a matter of national security and public well-being. For without this clear information there cannot be a market in the sense of Adam Smith’s theory of market forces.

It is not just a question of the technology. The mechanism of the back plate could in fact provide the provable security while algorithms derived from link analysis could create a market place where consequences of social or economic decisions are well understood. Whatever reality is, it should be seen by everyone, unless the reality is private, and then this privacy must be perfectly protected. Knowing the difference and being able to enforce this difference is essential to a conjectured future market place.

Complete and perfect public clarity about the real cost of all goods is one element of the envisioned consequences of back plate systems. Carbon management, and the management of other elements can be instrumented using a back plate designed for U. S. Customs, but not implemented as yet. [1] [2] The question of private information, when this information is about commodity use by manufacturing processes, is not the same as the issue of private information about personal lives or individual human being. The differences are not so easy to delineate. In one case, there is an agreement about the rules of economic interactions. In the other case, there are the constitutional mandated rights of privacy.

Section 1: Definition of “back plate”

What is a back plate? The concept can be applied to any type of finite state machine having a stratification of processing layers. Stratification is seen as a consequence of the organization of physical systems. When a communication system realizes this organization, we are able to establish appropriate correspondence between natural intelligence and the communication architecture.

The Internet has the FTP stack and this certainly is one way to implement stratification. However, the stratification potential has not been optimally used, in the way that the back plate concept suggests.

The back plate, as defined by Prueitt, is a system of compression/encryption dictionaries that communicate in the background. The background communication is minimally sufficient so that a generative capability comes to exist at a number of small computing nodes. The specification of these nodes is addressed in several design documents, and there is on going work on these designs. In essence one has a very small operating system that continually takes background information to maintain a generative cover, discussed below, as well as an ability to enfold and express digital objects. The operating system is mobile in the sense that it can be activated by any micro-processor, such as exists in cell phones.

Internal to the back plate nodes are rather simple optimization algorithms. These algorithms are useful because of the input / output relationship is handled in iterated action-perception cycles with what are called utility functions. Examples of this type of architecture are ubiquitous in the mathematical models of neural function as well as in various types of automated control. A longer discussion is required here. A summary of this discussion must point out that computational systems have up to now only allowed the modeling of biological function using programming languages that separate the processing of data by the program from a measurement process. The issue is that current programming environments are designed for purposes other than simplification. Without a simplification of the computer science, the natural scientist cannot do the work that needs to get done. On the other hand, great strides in natural science seem just on the horizon.

The back plate nodes are designed to simplify the input / output relationship specifically supporting the use of utility functions over an aggregation of invariance. [3]

Several types of optimization process can be computing using a modified steepest descent algorithm. Such algorithms are well known and understood in machine learning, artificial neural network, genetic algorithms, and numerical analysis disciplines. The category, of all steepest descent algorithms [4], includes systems that extract meaningful patterns from unstructured input. In the back plate, as defined by Prueitt, semantic cover generation is maintained at each of many virtual machines. The generators are each one equipped with one of the category of human mediated algorithms where human inspection of results is often required.

The notion of semantic cover is itself a difficult one, but in essence the notion implies a type of minimal sufficiency. The conceptual work for specifying semantic cover generators is founded by Prueitt in an area of formal systems theory called topological logic (Victor Finn, 1982 – 1994). Prueitt describes this conceptual work in chapter six of the book, “Foundations for the Knowledge Sciences”. [5] This work relies on stratification and a simplification of the processing architecture so that the logic described in “Foundation” can be applied directly to input / output relationships.

Semantic Cover Generator

In practical terms, any specific semantic cover generator is defined in terms of sufficiency. Sufficiency is a result of a specific type of well-specified utility function. In the back plate we are concerned about the materialization of objects at a distance. Can any object of a certain class be encoded at one node, the compression tokens sent to another node, and the object be generated? If so, there we have some form of structural cover over the class of objects. If the structural cover has a logic that predicts or anticipates the function/behavior of any generated object, then we have some type of semantic cover generator. If the semantic cover can be processed by a specific logic system, the one suggested is call “quasi axiomatic theory” [6], then a minimization of the cover will be identified. In the simplest form, this architecture is easy to realize, and has been realized in many compressed transmissions.

Even in this simplistic form, there are features related to the back plate model that can be realized. Object ownership in the Second Life virtual community software system already has many of these features. Thus there are actual models of what back plates will produce. The issue we repeat is that current software design is far to complicated, in nature, to realize many of the benefits that will arise once processing architecture shifts to the second school perspective.

Two categories of back plates are the multiple user domains (MUDs) like the visual chat systems, Palace and Manor; and Second Life; and virtual operating environments like bitTorrent and derivatives from bitTorrent. The most advanced specification of a virtual operating environment is perhaps CoreSystem. [7] CoreSystem is a virtual operating environment that is extensively specified but does not exist as a real system as yet. A simulation of CoreSystem is available. In our discussions about the second school, we will make reference to CoreSystem often, capitalizing on a now ten-year experience working with the CoreSystem innovator, Sandy Klausner.

Prueitt’s architecture for managing all commodity transactions across all national borders represents an early effort at producing the back plate design. [8]

As we began to make an investigation into back plate phenomenon; we found many systems that might be considered to be an optimal architecture, where “optimal” is formally defined in the context of a utility function employed by a steepest descent algorithm. In all of these systems, data is not just moved about but rather each whole, e.g., object, is treated as an object and encoded using a dictionary. Transmission then is not of the whole object, but rather is a transmission of a linear series of symbols that when expressed in the presence of the dictionary generates the whole object. The required bit transmission can be up to, in theory, 1/700 of the original bit transmission, and would generally be about 1/40 th the size in normal compression.

The difference between structural cover generators, such as compression dictionaries, and semantic cover generators is now seen clearly. The compression can be in fact encrypted, and the compression/encryption tokens may have some nature that allows a knowledge management function. It is important to acknowledge that this additional feature has not been traditionally associated with compression/encryption paradigms. It is also noted that the work by Prueitt and Adi on language generation from ontologic primes may indicate a direct correspondence between topological covers and the every day generation of the contents of mental awareness. [9] This is a large topic, to be discussed in another settings.

Funding to support the next steps

The demonstration of functional systems of generators is the key element of innovation that is now being considered for funding. The demonstration may provide a watershed, and may induce a long awaited shift in R&D efforts at major IT providers.

Structural cover generators are commonplace but not organized to produce a functioning back plate. The difference between semantic covers and structural covers is not so much a difference that cannot be over come, but rather is a paradigmatic shift potentially impacting the self-limitation of the IT providers. We believe that this limitation has to do with business models, not with technical feasibility. Self limitation is discussed in Section B and D.

If it were not for the manifestation of back plate principles in natural intelligence, we would have very little possibility of over coming the self-limitation of the IT providers. Architecturally, we do make the case that a back plate with semantic generation capability is physically manifest in brain systems. So far; however, the business capital fund managers have not appreciated the description of possibility based on the principles of natural science, as enumerated by our group.

Please note that this architecture is not designed to gain venture funding, rather is being specified as part of foundational science. So the reader should not judge our presentation based on the understanding demonstrated by current practices by the business innovators. Our audience is the community of scholars and individuals who see the necessities involved in shifting control of communication, entertainment and computing activities away from narrowly focused business processes and placing more control in the hands of individual consumers. In the new economic model scholars replace entrepreneurs in the IT development function.

What is conjectured to be potential can be seen in an analogy. Suppose that at the very beginning of the automobile age, venture capital focused on owning all paved roads. Suppose further that a monopoly was developed whereby if one wanted to use one’s automobile, one had to pay the investor groups for the right to drive on a paved road. If the federal government wanted to build Interstate Highways, the federal government would have to pay excessive amounts of money, perhaps as much as 50% of total construction costs, to the initial investor groups. Suppose, further, that patent law has now established a non-ending ownership to the concept of a paved road.

Now, imagine the specific natures of industry and commerce that would have arisen. Imagine if the initial investor’s group governed the political system itself. At one point, the desirability of doing things that could only be imagined, such as we experience every day in our real social reality, would come into conflict with the notion of private ownership of all paved roads. This is the conflict we now see with respect to the information highways.

The shift from first to second school, were it to occur, means a decentralization of control over private information spaces. A new economic model governing information infrastructure would arise. The new model may benefit everyone and democratize the control over commodity production and consumption patterns. The optimality we discuss comes with provable security over information, and thus new business innovation is possible as a consequence of back plate systems. Business does not suffer as a consequence, but is removed as the commanding authority over production and consumption. Anticipatory technology creates a replacement for the type of manipulative advertising as a means to control the production in expectation of consumer demands. Wastefulness is directly addressed by market forces rather than be driven by market forces.

In order that back plate information systems to arise we need to be objective about current information technology sub-optimality. An ultra stable and provable secure distributed operating system cannot be build on XML, RDF and other current technologies, but can be realized using the back plate concept. The optimality of a pure back plate might be understood as a feature of distributed computing governed by utility functions. Part of this optimality has to do with an ability to use smaller pipes for digital data movement, or “re-location”. Small pipe transfer of video transmission is to be realized in wireless environments.

So a strong capitalization argument is present in an area where the producers of digital content in the entertainment markets feel great pain. The second school position is; however, that a shift in the markets is required as a consequence of the imbalances arising from industry driven consumerism. The ubiquity of new application potentials seems to require that a back plate consisting of semantic cover generators be developed by the academic community and made available as un-owned infrastructure. This provision empowers new market strategies.

The optimality of the back plate technology might be relative to the non-optimality of the first school technology. Optimality arguments seem possible, but requiring of some reasoning based on category theory and foundational concepts in mathematics and computing theory (Harold Szu, unpublished and communicated to Prueitt 1998; Prueitt, unpublished). A simple demonstration of first principles may be seen in the physical organization of material reality, and thus the corresponding “feature” is seen as having a number of metaphors.

The theory of back plates suggests metaphors between computing systems and natural systems. For example, the quantum layer of physical reality is a theoretical construct. No one has seen this reality. More could be said on this. Various “structural” issues are delineated, including a back plate model of quantum mechanical phenomenon like Bell’s inequality. [10]

We are suggesting that many of the Second Life properties might be slightly modified so that back plate theory would have a means to be implemented as part of a living interaction between computing environments and human communities. Many of the predicted features of structural covers exist in the current software system, except there is not a full specification that scholars can discuss. Also the semantic cover is not present. A comparison between Second Life realities and CoreSystem specification is valuable but requires familiarity with both systems. In both cases, the supporting system is distributed and manifests digital materials when a small amount of information is delivered via a regular Internet connection.

The most important aspect of our work is on economic models. Many scholars have envisioned an economic model where all digital property released as back-plate content will have license agreements that the contents cannot be copied. Simple! This means that only a back-plate generation is legal, and all back-plate generation is measured according to license agreements. In Second Life, the back plate is called the inventory control. In one’s inventory are all things that one has in SL that are properties, including one’s body, hair, etc. We should be clear; however, that the model we are discussing involves the use of back plate systems that are off the Internet and used as micro-process control systems in production plants.

How will it work?

A back plate works by updating all virtual nodes (these are small virtual engines which MAY be similar in nature to the CoreSystem virtual engines) on an on going basis. All transference of information, such as the information needed to provide an intuitive (topic map) interface as well as all data will NOT be sent from point to point, rather the information will be generated at one point because at some other point there was some event that "caused" the generation. Again, we need to demonstrate this in a small lab, but it is essentially something that has been done a long time with regular compression tables. We just need to show that this is a "back-plane" that has generative capability. This generative capability is to be shown to instrument a digital object management system. The subject of digital object management is addressed by technical discussion in Section A.

Section II: The second school position on computing environments

Over the past decades, computer science has become grounded on a view of natural science that is reductionistic in nature. This grounding has created great economic value, but has also been done in such a way as to also create a narrow viewpoint about information. We conjecture that the hard limits that we experience from the current information technology can be by-passed by properly aligning computer science to natural science. The capital investments required establishing proof of this by-pass is not as significant as the investment that is required to control, e.g., inhibit, the evolution of a new ecosystem of markets. [11] It is thus possible to predict that at one point there will be a shift from first school paradigm implementations to second school implementations.

Some background is needed. Information as defined by Shannon is data oriented. We feel that this is actually not a correct formulation of the nature of information, as commonly understood by an average human. An alternative view of information is oriented towards the interpretive process. This viewpoint is perhaps most associated with the work of Charles Sanders Peirce. [12] Consistent with the Peircean viewpoint, human interpretation transforms symbol systems into the contents of awareness.

Data processing can produce symbol systems that are then either interpreted by humans or systems of humans or other living systems. These symbol systems may be designated to produce a context and within that context a mechanism may be used to control other mechanisms. The specification should follow a service-oriented standard using principles given by Thomas Erl. [13]

The second viewpoint makes other distinctions. One of these is about the meaning of complexity. This distinction may be used to see the difference between artificial intelligence and natural intelligence, between the first school and the second. [14] When this difference is seen, we are given a new and proper understanding of natural intelligence and computing. This understanding is essential for those of us trying to understand “what is next?”

The complete system of mechanisms involved in computing environments is always a system of finite state machines. The first and second schools agree. However; in the second school viewpoint these machines can be extremely complicated, but never complex, in the sense that Robert Rosen’s work has defined. [15]

Rosen’s work in category theory takes off from where Pierce’s work ends. In both Pierce’s work and in Rosen’s work, a core notion is the notion of an “living” interpretant acting as a consequent of being aware of both symbol systems and a range of possible meanings for constructions from these systems. The act of interpretation is part of a chain of many embedded cycles, within cycles, of perception followed by action, as discussed by J.J. Gibson [16] and others. The first school claims that these cycles are ultimately mechanistic. The second school simply asserts that these cycles are not purely mechanistic. There are points of complexity, Rosen complexity.

Complexity is neither simple nor complicated, rather is it an indication of indeterminacy, a quality exhibited by natural systems. A finite state machine, by nature, can never by complex. This is the position that the first school seems not to be able to understand. To see why this understanding seems to be difficult for some people, we need to look at the notion of logical coherence and translatability. Some things are not understandable unless one comes from the right viewpoint. Self-limitation, as discussed in Section B and D, is constructed from the mechanisms enforcing and experience of viewpoint

Linguist Benjamin Whorf [17] developed the notion of “non-translatability”. The concept of “complexity” is illustrative of the notion of non-translatability. The issue of complexity is not translatable into first school thinking, due in part to the, what we would regard as, polemical definitions of words like “intelligence”, “complexity”, “free will” and others by the first school. However, for those in the second school the understanding that complexity entails causes other than Newtonian causes seems justified by empirical observation and a growing body of scientifically grounded theory.

Can the action-perception cycle be “reduced” to data and mechanisms? The question must be informed by a first principle based on examination of natural science. The first school is self limiting, suggestive of an “artificial” notion of intelligence. Natural science can and should examine the mechanisms supporting human consciousness. A careful examination of words such as awareness, living, and interpretation is possible based on natural science. So the path is open to develop a second school of thought about the nature of computing environments.

If nature does not yield completely to reductionism then such an examination should produce computing interfaces that are not now anticipated by the current mainstream of information technology. When a back plate links these interfaces we have new types of social networks.

The first school asserts that one must always keep the question of artificial intelligence open and defined to be a type of intelligence that is superior to human intelligence and having all properties of human intelligence. The second school suggests that first principles derived from natural science provides increasing evidence that natural intelligence has specific properties that are not reducible to data and deterministic mechanism.

Implications of second school thought on human computer design

The human action perception cyclic involves an awareness, of state, in real time and the production of a cognitive map. The awareness is of “particulars” and the production of cognitive means is a production, via induction and abduction, of universals. At least this is second school dogma. First school sees the particular arising from universal. The second school sees the particular as always having some part of its essence not expressible as any set of universals, at least not as human language. Thus a difference of second school thought from first school thought is seen between linguistic expression and biological expression. This is only one of the many distinctions between the first and second school.

In modern information technology paradigms, such as “knowledge management” and “service event analysis” the cognitive map is then used in ways that are reducible to discrete data structure and mechanism in the sense defined by Shannon. In the second school viewpoint, our designs follow an action-perception cyclic in the production of universals from particulars. The mapping symbols can be encoded via the induction of specific symbols and the reification of meaning using a convolution over particulars to produce universals (Mills, Finn, Pospelov, Prueitt). However, this must not “wag the tail”. The particulars cannot be forced as some expression of a fixed set of pre-defined universals. To do so is to create a top down hierarchical control system that cannot instantly acknowledge novelty. These hierarchical control systems are first school in nature.

Such systems are very utilitarian but are subject to unexpected, and often costly, failures when the natural world changes. These issues were addressed in Soviet era efforts to place control over complex systems; such as cities and social systems, but such control systems were never achieved (Prueitt, Finn and Pospelov – private communications 1997).

Second school computer-human interface design

We have developed a computing paradigm that is designed to achieve a number of features not now available. In particular, a bi-lateral protection of intellectual property is made available. The bi-lateral interface is between the “system” of a single human, family, social group, or other “encapsulated” system and another system of the same type.

The interface is “complex” in the Rosen sense, not because of a technology feature, but because of the natural intelligence of any living systems. To talk about the interface involves the use of concepts that do not exist in the first school, and thus have non-translatability issues in the context of a discussion between first school proponents and seconds school proponents. The context does not mean that the second school does not have language, only that this language is not understood by the first school.

Specific language is used in the complex systems general theory literatures that will be used only very carefully here in the description of interfaces. The back plate interfaces are not reducible to data and mechanism, at least there is not any complete reduction process that has been found and is widely known. Rather than dealing with our language in a way that is consist with the well developed school of thought, we have been calling this school the “first school”, we use instead the specific language of the second school of thought about the nature of information.

In the first school, we develop the notion of an encapsulated digital object. This can be done completely within the language of finite state machines and what is called by the information technology sector, “object oriented programming”.

The notion of an encapsulated digital object is further extended to treat the objects as services defined within a computing environment. Service Oriented Computing (Thomas Erl) is then born from the object oriented programming and design literatures and efforts. However, the “object” seen as a “service” is still first school in nature, because the distinction between data and information is not clear. Data works with mechanism, deterministic mechanism; and information works with an interpretive process that is not, according to second school first principles, reducible to deterministic mechanism.

Encapsulated Digital Objects and the back plate

The encapsulated digital object (EDO) is a step along a path defined by object oriented programming (see works by Brad Cox [18]) and now service oriented computing. CoreSystem (developed by Sandy Klausner) provides a view down this path, but Klausner has already taken a different direction then those who are in the first school, a direction that takes one directly to the first principles of the second school. The key word here is “generative”.

To build a market context for the scientific work I am proposing, we need to introduce a bit a jargon. This jargon only roughly approximates what might or might not be actual present in the marketplace, due to non disclosure agreements and classified R&D. CoreSystem relies on the generation of a set of computational primitives and the use of what are called frameworks to generate digital objects. John Sowa, Richard Ballard, and John Zackman all have versions of generative semantic primitives, although a general theory of semantic primitives has not been published. The development of a general theory is attempted in my private work, but the issues have to do with paradigmatic viewpoints and how multiple viewpoints might be represented in a single “logical” system.

The CoreSystem framework is called “cubism” and is related to the art history movement called “Cubism”. [19] The same distinctions made by Thomas Erl regarding the transition between digital objects and digital services (defined within computing environments) can and does get made in the CoreSystem architecture. Other concepts from these that are “behind” XML, Topic Maps, and web ontology languages (OWL) are present, for example “namespaces”. However, the implementation of these concepts is at a more advanced and simpler level. The difference in actual software architecture design and implementation features from one software system to the next may be compared with the existence of different geometries. One may be able to represent very different theories of computing science, each one having specific types of features, capabilities and consequences.

The implementation in CoreSystem is second school in nature. Namespaces become contextual devices where the mechanistic data structures are given meaning based on pre-defined contextualization measured by a specific implementation design for a back plate. Context computing, what ever that means, is then the primarily book keeping task of CoreSystem.

So what might “context computing” properly mean? This bookkeeping is accomplished using an induction of form from the experience of structure. The bookkeeping provides the content for semantic cover generators and cause an induction of universals from particulars, in second school; and a use of pre-defined universal to represent particulars in the first school.

The second school’s intent is to observe and to produce universals from direct observation. This is not an easy task, unless first principles are actually aligned to real natures and to the reality of natural systems. These first principles include the recognition of location and individuality. The computing theory that we are suggesting makes as an assertion the need for real time involvement by everyday users in the behaviors of the software code.

We assert that location has a local and distributed nature, and this nature can be accounted for using stratification. Of course this assertion is a core second school assertion about physical and cognitive reality. The first principles related to stratification can be understood as necessary if complexity is to be acknowledged. The stratification allows symbols system to evolve within a number of layers, layers that are tied together by what the Soviets called “Mill’s logic” and what Prueitt and Kugler extended (1996 – 1998) to produce a tri-level architecture for treating the particular to universal induction. [20]

A stratification of symbol systems also creates the means to generate service objects for provably optimal transmission using compression/encryption dictionaries. Mathematics for provably optimal security over the key encoding to a service object is given in private work, but an outline can be given here. The bottom, or substructural, layer roughly corresponds to physical atoms, which are then aggregated together to produce “compounds”. [21] These atoms are “found” using stochastic means. The compounds are then the digital objects that de-materialize at one place and re-materialize at a different place. The generation process is thus seen to use a table of atoms, to “compress” [22] the signal thus producing gains to effective transmission rates. The categories of data transmission compression patents all work on precisely this principle; however without there being a purposeful refinement of the atoms into what is in essence period tables for the expression of semantics. This is done by CoreSystem and by the Mark III system [23] as well as a number of classified systems.

A number of practical advantages are derived from second school first principles. The generation of services from “service objects” can be instrumented so that all generation events are communicated to an internal data structure (based on a key-less hash (Paul Prueitt)) that then must communicate to a service organization as a means to manage bi-lateral issues with respect to intellectual property. This communication and instrumentation is described in Brad Cox’s book “SuperDistribution” as occurring within a micro-banking system. The generated service is not communicated, only agreed on data is transmitted about the use of the generative service object. The service fulfillment then uses the semantic cover generators.

The technology is then a generative technology having optimal compression and encryption, as well as an evolutionary architecture that creates stable substructural tables having great expressive capability. Individual communities, or processes, may evolve distinct substructural tables for which non-translatability becomes an issue. However, the reality of non-translatability requires a feature supporting terminological reconciliation technology [24] .

On the usability of systems that have a back plate

The picture of interacting gEDO (generative Encapsulated Generative Objects) systems, having well defined interfaces to actual living systems is a picture that can be understood, because this picture is in fact similar to how humans use natural language.

The argument is simple. We use our familiarity with life to understand what cannot be formalized into data and mechanism. We produce second school words and meanings that communicate this familiarity, not by “placing” all of the “knowledge” into a symbol set, but by evoking shared awareness.

The language is not reduced to data and mechanism, so the part of this picture that is hardest to understand is the computing environment. The computing environment is vastly simplified into provably optimal compression and encryption; as well as both localized and global gEDO management environments. These environments each have bi-lateral and uni-lateral capability, and these capabilities produce information security as well as intellectual property management.

Section IV: History

In November 2007, a group of information scientists made the observation that a “back-plate” to the Internet is emerging, and predicted that this phenomenon will foster a new economic model. A precursor Internet back plate is, in fact, evolving via a collective process having no central control, with strengths analogous to the wiki concept. As in other collective and distributed processes, there is an adaptation due to a certain set of principles in order to meet anticipation.

This adaptation is not governed by centralized control. However, where the adaptation is going may now be visible. Economic decisions regarding approaches funded by government and private sources have been involved in the development of the Internet. Some less than optimal work is to be expected; we are in a trial and error phase in the development of the Semantic Web. This sub optimal work is now easily recognized. Optimality is defined with respect to some viewpoint and viewpoints shift, sometimes suddenly. Because the difficulties involved in software use, and the failure of the current system of software development; one can predict this shift in the market in the near future. This prediction cannot be precise, but is based on long standing theory in social sciences and in economic science that an established system that become disconnected from reality will produce an appearance of reality only for a certain period of time.

To see how the collective effort is progressing, we may focus on economic motivation and how this motivation supports, or inhibits, inventions of precursor technology such as semantic extraction, generative-encryption, ontological modeling and the like. Such a focus provides insights into how to capitalize on one or more element of the emerging phenomenon, and in this way more fully participate in the rewards. Because there is a potential shift in the economic model, these precursor technologies may be understood by anyone wishing to invest successfully. Investment however creates its own reality, and thus the success will likely be seen by those who are lucky and whose insights have suggested caution.

Many large-scale projects have anticipated the back-plate, including some in patent evaluation, pharmaceutical and medical literature identification, medical research, drug design and manufacture. In each case, underlying precursor technology is used and used in a way that shows similarity in how the technology is conceived and deployed in other projects. These projects have not produced the critical mass required to break down the old computer science paradigm. We are still, January 2008, in a pre-shift era.

The reason that IP mapping (patent evaluation) would become important was clear sixteen years ago in 1992 when the author gave a talk at a private conference at Georgetown University. The talk covered mapping IP evolution and potential technical means available to automate the communication of IP evolution between the university and the marketplace. These concepts led to the BCNGroup Charter mechanism (1994), for mapping IP and distributing the compensation for university based research. Later, in 1996, associated concepts combined with Brad Cox's concept of super-Distribution as a means to provide transparency for the IP universe. The original architectures in 1993 involved “neural models of cognitive behavior”; like selective attention and orientation. Over the years, the concept embraced work derived from Soviet cybernetics and semiotics (1995-1998); and then ontological modeling (1999-2004). In spite of this history, and similar histories involving other innovators, IP mapping is still misused and not performed in any optimal fashion.

Evidence suggests that this evolution of concepts and related technology is similar to many other projects, some highly funded in private or classified settings; but none clearly visible. In the material presented below, the author attempts to suggest to the reader some of the core principles on which these concepts and technologies have depended. We leave side the questions related to the critical mass and the shift.

Ontology emergence and merging

We observe that human interpretation of linguistic patterns is highly situational, and based on context. Nothing technical about this, but the underlying mechanisms have been a mystery until recently. The natural properties involved in memory, awareness and anticipation suggests a mediation of linguistic parsing by dynamic ontology. These properties have been lifted into an abstraction and realized as computational process, all the while fully understanding that the computational system is neither intelligent nor alive, but is merely an extension of natural processes involved in natural intelligence. The mediation fits each situation with relationships and associations that are semi-automatically constructed.

A new school of thought about human information exchanges has been born, and is called “the second school”. [25] A model is suggested that allows computational support for a natural process, of which we are all quite familiar with.

Our approach merges ontology from past analysis with a type of category theory that applies nuances of nouns, verbs and objects. Taxonomies, controlled vocabularies and web ontology are soon to provide easily understood situational analysis of particulars. The key is that the technology allows normal action perception cycles as humans interact with the computer. Humans do use selective attention and orientation to features as an interpretation and modification behavior. Modification engages the human and results in high quality learning. These principles have computing correlates.

On universals and particulars

The most critical feature of the new environments is in getting an ontological model reified from particulars; e.g. universals from particulars. These universals are seen as "not being everywhere realized".

A short discussion about particulars and universals starts with a question. In the moment, what time scale are we in?

Is this even a reasonable question? For reasons that appear hidden, the nature of the particular and its composition from elements of the universal has been a subject of inquiry in all civilizations and in all times. The current time is not an exception.

The consequences of the investigation, in our times, results in almost every type of belief system and in every form of our science, in what ever system of science one inspects.

As in string theory, there may be more than one conceptual system that accounts well for the phenomenon that manifest in the various scales of physical, biological and social activities. Also like string theory, the development of evidence about biological and social science may only now just begin to be available. The condition of non-translatability may be expected to separate any one of these “systems of thought” from each other. One can have the position that non-translatability has something to do with a failure to find the set of universals that apply to everything. On the other hand, one may take the position that human knowledge always has an illusionary nature, and then very timidly suggest that non-translatability between human conceptual systems has a non-removable truth. The paradox in this timid statement escapes no one’s attention.

In the language of systems theory, we may say that the expression of ontology in time is or appears to be fractal in nature. What this means in pure mathematics is precise in works on scholars like Mandelbrot. What it means to me is that the particular is attempting to expression in the patterns that have formed at slower time scales, and is being required to make that expression with patterns expressed in the fast times scales. The particular is sandwiched between universals at two different scales of expression.

Lines of affordance form and create an event horizon with the present moment appearing to be in the center.

There is a contextual frame to abstracted ontology. Context can be managed using terminological reconciliation and in fact a kind of terminological science that Fiona Citkin pioneered in the Soviet Union (late 1980s and early 1900s, mostly classified). It is quite natural to realize that terminological context is ultimately determined by a consensual relationship between individual humans and the collective agreement. In both the individual and the collective cases, explanatory coherence, i.e. rationality, is involved (Paul Thagard).

A model that corresponds to natural process

Our previous work integrates the above principle into a tri-level architecture designed to reify (create universals from particulars) in real time. The resulting ontological models are complex, in the sense that Rosen defined. Rosen defined simple as any system that had a formal nature similar to Hilbert mathematics. Much of the mechanisms of reality are well modeled by Hilbert mathematics, but a significant number of the mechanisms involved in human knowledge are not. For creating knowledge about these mechanisms we may need to use ontological modeling. The reason may be simple. The evolution of these systems involves the emergence of wholes whose function fits into larger ecosystems of processes. The function can be achieved in many ways, using many different groups of compositional elements. Recognition, processing of stimulus using some “internal” model, and intentionality seem to be involved even in metabolic processes.

The tri-level architecture is designed to create models of complex phenomenon. However, the principles involved in the design of the tri-level have to be justified based on some type of verification principles and on consistency with classical science, for example Pribram’s work on neuro architecture.

The three levels are each composed of a set of abstractions. The lower level is a set of semantic primitives, defined statistically and heuristically as semantic frameworks. The upper level is a set of categories, whose definition is a consequence of prior description of how things evolve. Situational parsers measure the lower category structure and update semantic cover generators. The upper level has to be constructed by some means, and for this we suggest the Mill's logic (Prueitt, 1996).

Once this tri-level architecture is seen and it has been on the OntologyStream web since the late 1990s, the programming is simple (four months). Within this period of time we will produce represent