Dear Xueshan, Bob, Gunther and other FIS colleagues, Thank you for your post presenting your views on three different levels of information studies. From an IS point of view an interesting area for further research is to explore the possibility that there could be some similarities in the processes that lead to the emergence of human language and other rule-making “alphabetic catalysts” at different levels. These are ideas that have puzzled me for a while, and I thought they could benefit from comments and guidance from this group. Bob Logan has suggested that human language is an outcome of evolution in communication that accompanies growing complexity in kinship based social organizations in heterotrophic species [Logan- The Extended Mind (2007)]. Kinship based social groups in heterotrophic species are known to arise before non-kinship based social groups, like formal businesses emerge in human society. Over a period of time kinship based social groups are believed to have evolved and gradually increased in complexity. It is suggested that that this growing complexity in kinship based social groups created an evolutionary pressure that gave rise to the capability for conceptual thinking [Logan (2007)]. According to Logan, conceptual thinking arose when precept based communication was no longer sufficient to handle the cognitive complexities in social living, [Logan (2007)]. Further, conceptual thought and language form an autocatalytic set, where language catalyzes conceptual thought and conceptual thought catalyzes language, [Logan (2007)].
In this line of argument, communications between individuals within social groups themselves operate at two different temporal scales. Precept based communications create an impact in the here-and-now, whereas conceptual communication involves a level of abstraction and their impact extends over a longer time and over a larger group of individuals. The emergence of language is believed to enable communications between individuals to move from the level of precepts to the level of conceptual communications. I quote from Bob’s work, “Assuming that language is both a form of communication and an information processing system I came to the conclusion that the emergence of speech represented the actual transition from percept-based thought to concept-based thought. The spoken word, as we shall see, is the actual medium or mechanism by which concepts are expressed or represented. We must be very careful at this juncture to make sure that we do not formulate the relationship of spoken language and conceptual thought as a linear causal one. Language did not give rise to concepts nor did concepts give rise to language, rather human speech and conceptualization emerged at exactly the same point in time creating the conditions for their mutual emergence, which is a form of autocatalysis. Language and conceptual thought form an autocatalytic set because language catalyzes conceptual thought and conceptual thought catalyzes language.” [Bob Logan- The Extended Mind (2007)]. Hence language extends the impact of communication in space and time by enabling communication between individuals in social groups to move from the "here-and-now" precept based communication to conceptual communication with impact over a longer duration, and facilitates the generation of "collective conceptual knowledge" in a group of interacting and communicating individuals. According to Knight, human language is believed to have emerged simultaneously with the emergence of “social contracts” that are collective rules enabling higher order social organization in humans [Chris Knight- The Human Revolution (2005)]. The above arguments suggest five broad steps in the evolutionary emergence of human language: First step is the evolutionary emergence of sociality in the form of kinship based social groups (i.e.- coupled-composite systems, see paper II- Joshi (2015). Evolution in kinship based social groups leads to an increasing complexity in their social organization. A phase transition gives rise to the emergence of conceptual communication and language (i.e.- alphabetic catalysts- see paper II- Joshi (2015). The emergence of language enables communication of concepts and facilitates the generation of "collective conceptual knowledge" that has an impact over a larger number of individuals, across the social group and over a longer duration in time. This facilitates the emergence of non-kinship based social groups ((i.e.- decoupled-composite systems, see paper II- Joshi (2015)) based on the capability to develop collective language-based rules that support a higher level of complexity in social organization. Using a different line of research, Honkela and others have shown that under certain conditions, the dynamics within communities of autonomous agents can lead to the emergence of conceptual knowledge, and language like forms [Honkela et al (2006 and 2008), see Joshi (2015) paper II for a reference]. They have used a framework of “unsupervised learning” from machine learning to study the emergence of conceptual knowledge and language like forms in interacting autonomous agents. In machine learning terms unsupervised learning is learning that does not involve a environmental feedback [Litman (1996), see Joshi (2015) paper II for a reference]. Such learning is hence based on exchanges between interacting agents themselves. It is known to produce abstract knowledge that can be represented in the form of self-organizing maps [Honkela et al (2006 and 2008), see Joshi (2015) paper II for reference]. An interesting question is that if the two frameworks: (1) Evolution of sociality in kinship based social groups [Logan] and unsupervised learning in autonomous interacting agents [Honkela and others] both give rise to conceptual knowledge and the emergence of language like forms. Could there be equivalence between these two frameworks? In other words, can the learning in the course of evolution of sociality in kinship based social groups as described by Logan and others be represented as unsupervised learning in a machine-learning framework of Honkela and others? Another more speculative idea- kinship based social groups have been shown to belong to a larger class of coupled-composite systems, a class that also includes multicellular autotrophic species, and reduced molecules in the metabolic core in living cells [see Joshi (2015) paper II ). (2) Is this above process that begins with the emergence of sociality (coupled-composite systems) and evolves to the emergence of conceptual knowledge and language (i.e.- alphabetic catalysts), and finally to the emergence of non-kinship based social groups (decoupled-composite systems) also applicable at other levels in self-organization? Hence, could the evolution of cellular sociality [ref- Witzany’s work on RNA sociality in living cells] in evolution of early multicellular autotrophic species have given rise to “conceptual knowledge” at a cellular level, leading to the emergence of the next higher-level alphabetic catalysts- the DNA? could the emergence of DNA have facilitated the emergence of exchange networks between autotrophic and heterotrophic species (decoupled-composite systems)? I admit that these are very early thoughts and are highly speculative, but nonetheless could provide new avenues to pursue in developing a multilevel view of living systems by integrating different research streams. I humbly put forth these ideas for comments, and reflections from the FIS group. Your views and references to other work, will help assess these ideas and are greatly appreciated. Warm regards, Nikhil Joshi > On 01-Dec-2015, at 8:10 pm, Xueshan Yan <y...@pku.edu.cn> wrote: > > Dear Joshi, > > No matter what topic/title you used, no matter what goal you want to reach, > your post has raised a very important theory which can decide the future of > information science: Three Level Theory: Molecular (level1), Cellular > (level2), Social (level3). (Please excuse my minor modification). > > The FIS colleagues can easily recollect the theory of Cell, Brain, Firm which > was advocated by Pedro about 10 years ago, but I think this hierarchy is > could be much better spent taking some positive action. > > Social (level3): It can indicate the all human/social information studies. > Cellular (level2): It can indicate the all cellular/biological information > studies. > Molecular (level1): It can indicate the all molecular/chemical information > studies. > XXXXXXX (level0): Particlate/physical information studies?? > > As we know, due to the Technological Information Science (It includes > computer science and telecommunications) is not self-organizational, or > antipoetic, so we generally don't consider it as a real information science. > > With my best regards! > > Xueshan > Peking University > > From: fis-boun...@listas.unizar.es [mailto:fis-boun...@listas.unizar.es] On > Behalf Of Nikhil Joshi > Sent: Tuesday, December 01, 2015 7:35 PM > To: fis@listas.unizar.es >> fis@listas.unizar.es > Cc: Nikhil Joshi > Subject: Re: [Fis] Sustainability through multilevel research: The Lifel, > Deep Society Build-A-Thon - 1 > > Dear Joseph and Stan, > > Both of you mention about earlier work on isomorphisms, and you also mention > hetero-organization. If it is not inconvenient, may I request more > information on this? You also mention that the use of self-organisation could > be a distracting, could you recommend an alternate formulation? > > At this time, I must clarify that I am not suggesting a hierarchical > relationship between the three levels. I am referring to hierarchical > organisation within the species at each level - molecules (level1), cellular > species (level2) and social groups (level3). > > Coming to your question- how does the concept of hierarchy affect the > analysis? > The common multilevel organisational pattern presented here suggests that a > core element in human social organisation involves exchange networks based on > flow of human resources between kinship based social groups (like families) > and non-kinship based social groups (like businesses). This implies that > evolution of social organisation is based on the emergence of two species > classes with greater complexity and greater compositional hierarchy- kinship > based social groups and non-kinship. > > The question then are- why and how do living species give rise to exchange > networks between species with increasing complexity (and compositional > hierarchy) ? Will this pattern continue at the next higher level? > > Bob Logan and others point to the role of human language and the generation > of conceptual knowedge in the emergence of non-kinship based social groups. > It is interesting that Timo Honkela and Kohonen generalise these ideas and > describe processes that gives rise to conceptual knowledge in systems of > interacting agents. Do Alphabetic catalysts like DNA and Proteins play a > similar role as human language in the emergence of exchange networks at two > different levels? (see section 4.4., paper II in this kick off email). > > While many theoretical perspectives have been presented on the evolution of > such systems (Stanley Salthe- Evolving Hierarchial Systems, Ch 8, John > Holland- aggregate agents, Eric Chaission- growing energy rate density, and > others) what is most interesting here is that the CMOP provides opportunity > to examine processes that give rise to such organisation in much greater > details. This could provide more insights into the emergence and evolution of > such organisations. > > Given the diverse research interests and great depth in this group, I would > love to get your views on these questions. Your views are greatly > appreciated. > > Thanking you, > Regards, > Nikhil Joshi > > > > > Given the wide > --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- > Dear Nikhil, > > I think it is a very interesting exercise to see how a consensus might be > reached on your work by both adding to and subtracting from the different > perspectives. Thus, I agree with Stan that we are looking at instances of > isomorphism at different levels, and this for me is entirely logical (;-). > Levels of reality exist and the rules that apply in them are not identical, > and this constitutes a discontinuity between them. Also, within a given level > involving three elements, even if they all influence one another, it should > be possible to decompose the interactions into those between A and B, the > resultant of which interacts with C. This is Pedro's comment in somewhat > different terms. > > On the other hand, as I have argued elsewhere, the use of the term > 'self-organization' does not bring any additional knowledge. It diverts > attention from the dynamics of the different flows, which are also affected > by such a multitude of external factors, actual and potential, that the > process could equally well be called hetero-organization. Also, and I really > just ask this as a question, how does the concept of hierarchy affect the > analysis? If as you write there are different species involved in exchange > networks across ascending levels, what would be important to know are the > details of these exchanges. Here, the above discontinuity between levels > seems to be replaced by a degree of continuity. Your statement implies to me > interactions /between/ different levels, but are these interactions > bi-directional reactions? How would the rates of forward and back reactions > be related? > > I look forward to your comments on the above which I assure you is intended > to be constructive. > > Best wishes, > > Joseph > > _______________________________________________ > Fis mailing list > Fis@listas.unizar.es <mailto:Fis@listas.unizar.es> > http://listas.unizar.es/cgi-bin/mailman/listinfo/fis > <http://listas.unizar.es/cgi-bin/mailman/listinfo/fis>
_______________________________________________ Fis mailing list Fis@listas.unizar.es http://listas.unizar.es/cgi-bin/mailman/listinfo/fis
