The Biotope Issue in Meme Media Implementations

The concepts of memetics and the development of meme media implementations have set the stage for knowledge media evolution by which externalized human knowledge is becoming subject to a process of growth which can be boosted by technological means. Computers acting in computer networks in co-operation with humans will constitute interactive knowledge media – another goal of Artificial Intelligence. The key question is whether or not we will be able to establish proper meme pools in which memetic evolution finds an opportunity to take place. And if so, it is the question what conditions we have to provide for a dynamic evolution of meme media based knowledge. This is the biotope issue. Properly dealing with the biotope issue is essential for a future development of meme media going beyond the limits of conventional software technologies. Thus, the biotope issue is not only just a philosophical or methodological foundation of meme media development, but a core problem which is frequently underestimated. The aim of the present publication is to make the biotope issue seen in its right perspective – for the benefit of memetics and meme media development as a whole. 1 Motivation of a Seemingly Esoteric Discussion Memetics is seen as outlined in the truly exciting books by Richard Dawkins [3], Susan Blackmore [1] and Yuzuru Tanaka [16]. Richard Dawkins has attracted the world’s attention to the phenomena of cultural inheritance and has introduced his seminal concept named meme. Susan Blackmore has taken the initiative to discuss the relevance of Dawkins’ perspective from a psychological and from a somehow philosophical point of view telling all of us that we are affected by Dawkins’ work. It is, naturally, up to you whether or not you feel personally affected by memetics, and this might easily become a slightly esoteric discussion. But Yuzuru Tanaka, fortunately, has seized Dawkins’ suggestion and developed it toward concepts, implementations and applications in computer science. He has coined the key term meme media. The present approach relies on Tanaka’s trend-setting work taking Dawkins’ and Blackmore’s contributions seriously. We want to contribute to the endeavour of enabling computer systems to foster true knowledge evolution – the benefit for humans will be paramount. As Mark Stefik [15] has foreseen, there is a tendency towards externalization of knowledge into interactive electronic media, a trend which is considerably boosted by the spread of the Internet. Meme media pools are mushrooming. Meme media objects or meme media, for short, are the inhabitants of these pools. How are they? Are they doing well? Taking Dawkins’ and Blackmore’s contributions [3] and [1] seriously means to ask for the living conditions of meme media in their respective meme media pools. In the spirit of Tanaka [16], every pool of meme media is bringing with it the potential of knowledge evolution. The work is properly done, only if there is some progress recognizable and knowledge does evolve. Otherwise, we have set up just another heap of data, like the Internet which is currently more a data cemetory than a knowledge source. Similar problems are known in every larger enterprise. They all have huge databases, but suffer from a lack of knowledge. The biotope issue in meme media implementations is about making data more beneficial for humans, in general. Moreover, it stresses the issue of knowledge evolution which can potentially go beyond the limits of human expectations and imaginations, in particular. More technically speaking, this is the question of what to do to enable knowledge evolution. When Tanaka’s book [16] did appear, it seemed that the time has come for our data – or, at least, for some of them – to wake up and begin to evolve. But does this really happen? Where are the pools of evolving knowledge media? It seems that we have some severe deficiencies. The life in our meme pools, not to mention our databases, is far behind its potentials. The biotope issue has been widely underestimated. To say it very explicitly: We have to take care of the living conditions of our meme media. We have to think and work about technological measures which foster knowledge evolution. A first step is to establish awareness of the biotope issue. This is a minimal aim of the present publication. One of the reviewers of the present paper’s submitted version has seen the paper as an outline of a future research project. This, for sure, was not the author’s intention. Instead of any separate project dealing with the biotope issue, everyone dealing with meme media theory, engineering and application should become aware of this issue and act appropriately. So, the biotope issue may disappear through the infusion into the meme media community’s daily research and development practice. The present paper will have been successful as soon as there is no further need to refer to it. In other words, considering the biotope issue should become folklore in the meme media community. Last but not least, there is another motivation for the author to raise a discussion about the biotope issue in meme media implementations – the quite serious difficulties in communicating the essentials of meme media concepts and meme media implementations. The author has had numerous talks with representatives from industries about the essentials and the potentials of meme media, system demonstrations included. In general, meme media and IntelligentPad implementations have been seen as just another middleware approach. There is some quite obvious reluctance to getting engaged in meme media applications. Just another middleware does not seem exciting enough to invest time or money. From this perspective, the biotope issues does possess a commercially relevant dimension. 2 Perspectives at the Biotope Issue Recall Richard Dawkins’ impressive report about a colony of birds, so-called saddlebacks, living quite isolated on islands near New Zealand [3]. There is a rather detailed documentation of the evolution of songs these birds are singing. To say it as explicit as possible: The songs are forming a meme pool in which evolution takes place, provably. These birds on the island are only the habitat in which the songs are living. The birds are forming the biotope for the songs. Having a little closer look, it may be reasonable to include the island, the wheather and the like into the biotope considerations. Those birds are somehow building a biotope to their songs. We humans are similarly being biotopes to our thoughts. This is, at least, a possible perspective at the memes we are breeding in our brains. For humans, it is not easy to step back and agree with something else being more central than the self. Chris Fileds is surveying and summarizing current work in cognitive science, the neurosciences and evolutionary psychology to come up with an answer to the question ”Why do we talk to ourselves?” [8]. The summary is consistent with other work about evolutionary side effects of the humans’ cognitive architecture. Fields concludes with the insight that we humans ”talk to ourselves because we have to”. There is a space in the human brain which works somehow like a buffer. Components of the brain write into this space and others read from it. Fields summarizes: ”If our cognitive apparatus allowed different functional modules to easily access each other’s representations in their native form, ... we might not need an inner voice at all.” So, there are biological reasons for breeding memes. In the present chapter of the paper, so far, we dealt only with memes, not yet with meme media. Whenever and whereever we externalize knowledge into modular objects, especially on computers and computer networks, we are getting close to meme media. However, in many cases, the underlying implementation principles are not appropriate to stimulate evolution of the knowledge externalized. It turns out to be literally dead. For already a dozen of years or so, Yuzuru Tanaka is striving hard to establish meme media implementations which allow for memetic evolution [16], at least in principle. All these implementations have their respective conditions under which the meme media objects can be subject to evolution – the biotope issue. To sum up briefly, is this again getting an esoteric discussion, to some extent? Not in the author’s opinion! The brief discussion of the phenomenon of memes in biological reality has revealed that there are reasons for the existence of memes, and there is a variety of preconditions fostering the evolution of those memes. Analogously, we have to think about the conditions of meme media evolution, if we are interested in exploiting the potentials of knowledge evolution in computers and computer networks which goes beyond human expectations and estimations. 3 The Biotope Issue in Application Attempts For the sake of a focussed discussion, we confine ourselves to IntelligentPad implementations, exclusively. The biotope issue discussion can not serve as an overview on meme media and memetics. Readers not satisfied with such a narrow view are directed to the underlying literature already cited, especially to [16]. The implementations envisaged here are of the type reported in [11], e.g. The present volume in which this paper appears will contain further examples galore. 3.1 Fundamental IT Constituents of IntelligentPad Biotopes When knowledge is mapped onto computer systems and computer networks, minimal requirements concerning the hardware and the software are inevitable. Though this is absolutely trivial, is has to be seen as part of the biotope issue: One needs something before pads can begin to duplicate, to combine, to mutate, to cross-over and the like, i.e. before knowledge evolution can take place. We all know the old and ever returning lie about platform-independend code. In contrast, we need to make explicit which assumptions are taken for granted that a certain meme media implementation can be launched. In addition to a certain operating system, there may be the need for some environment like Microsoft’s .NET framework. All this is agreeable, but is has to be made explicit. 3.2 Software Paradigms and Implementation Principles IntelligentPad systems are complex IT systems that are never implemented from scratch. Every individual system approach takes advantage of some programming paradigm and exploits work done earlier. It usually takes some developing environment as a basis and adds some of its own basics. For instance, libraries are taken and extended according to the particular needs of the approach. For illustration, in object oriented programming, classes are defined and particular pads are introduced as instances of those classes inheriting information such that the programming effort for individual pads is minimized. Whatever approach is used, for many good reasons and in accordance with overwhelming experience, there is a need to make the assumptions explicit. In most cases, there will be a tradeoff between independence of meme media objects and efficiency in their development process. Predefining functionalities upwards in a hierarchies and using functionalities located in libraries makes pads more lightweight, but makes them dependent on the presence of related files. In general, there seems to be a deficiency in clarity about those implementation decisions together with a lack of discussion among developers and scientists. Most decisions seem to be ad hoc, instead of driven by a certain methodology. 3.3 IntelligentPad Application Development When IntelligentPad technology is spread all over the world, there will be an increasing request for development support. At a first glance, features of any meme media development environment do not really interfere with the biotope issue, as making the pads can be clearly separated from using the pads, i.e. from the pads’ life. But a closer look reveals that this is simply not true – it can not be true by the very nature of memetics. Birth and death can not be separated from life. More technically speaking, when pads are generated dynamically according to unforeseeable needs of a dynamic domain – when evolution beyond human administration is going to take place – the way in which pads can be generated or not is an essential aspect of the biotope issue. This can be seen in even more detail when currently widely independent approaches meet each other. Take, for instance, the dynamic plan generation approach from [7] and combine it with the application development approach of [6], where a so-called cockpit is proposed to support control of pad generation and maintenance. At the moment, the cockpit is really a tool supporting pad generation, but separated from the generated pads’ lifecycle. In the future, more advanced cockpit functionality will be desirable. The medical therapy planning approach of [7] will lead to dynamic pad generation where details of monitoring and control of the generation process have impact on the direction of the ongoing meme media evolution. 3.4 Embedding IntelligentPad into Application Environments It is with IntelligentPad application as it is with all other IT applications; they have to be embedded into some usually a priori given environment. Embedding requires interfaces. From a strictly memetic point of view, the problem might be paraphrased as follows. On the one hand, meme media objects need to be able to meet each other, to communicate with each other and to do all those things that are necessary to enable evolution. Beyond this, on the other hand, they have to live in an environment and, therefore, need to sense the environment somehow. Whenever we maroon our pads, we should carefully equip them with some ability of perception or, more generally, of communication. There are, at least, to different approaches to embedding of pads into given IT infrastructures which do not mutually exclude each other. First, one may develop particular pads serving the whole community of inhabitants of some meme pool by arranging the communication with the surrounding world. Ito’s so-called service wrapper pad [11] is a good representative of this approach. There are other solutions like, e.g., proxy pads to relational databases. Second, one may equip every pad with the ability to understand and speak some language. Still, we are far from releasing pads with natural language skills. For the moment, languages like HTTPmay do (cf. [10]). In another environment, another language may be more appropriate. Are pads going to learn XML? 3.5 Domains of Evolving Externalized Knowledge Clearly, in areas where no considerable evolution of knowledge takes place, there is no need for an in-depth discussion of the biotope issue. Those areas are, perhaps, inappropriate for meme media technologies, as there is no remarkable advantage over other middleware approaches. In contrast, domains where a considerable part of the knowledge is subject to a dynamic evolution are potentially attractive to attempts of invoking meme media technologies. Naturally, one still needs to find out whether knowledge is sufficiently modular and whether there are natural ways to map knowledge units to IntelligentPad objects. Assume such an application domain has been identified, like medical therapy planning, see [7], e.g. What are particular aspects of the biotope issue to be taken into account, especially those beyond the aspects mentioned above? In those domains, the evolution of knowledge takes place in accordance with disciplinary regulations and customs. Even local peculiarities come into play. For illustration, hospitals have usually their own procedures how decision makers are contacted and how plans are made. When IntelligentPad technologies are in use, the evolution of pads does reflect those peculiarities. In other words, the culture of the environment including the non-IT environment has impact on the pads’ culture. At least, there will be an interference. It might be highly interesting to see whether or not certain human cultural environments can be distinguished by a different impact on knowledge evolution in meme media pools. Most probably, there will be more influential and less influential domains. One may expect that research and development will, as a side effect, reveal unwelcome insights. Phenomena of the latter type, however, are beyond the scope of the present investigation and also far beyond the investigations planned for the future. Instead, we go into a little more detail to illustrate what it means in practice to take the biotope issue into account. Pads of the CHIP system [10] are adapted to become classes and instances of medical therapy actions [7]. Plugging pads together results in therapy plans. But what does it mean to plug pads together? Therapy knowledge may be located in distributed meme pools. Plugging pads together means to establish a certain inter-pad communication. The communication with some underlying patient data record pad is of a particular importance, because executability constraints have to be checked with respect to current patient data. According to the CHIP architecture, every pad has its own Web server. Pads are ‘talking to each other’ by means of http-get and http-post. Whenever changes occur in a pad somewhere down in the hierarchy of a composite pad, these changes are propagated upwards through the whole hierarchy. The communication abilities of the CHIP pads are centrally located in some dynamic link library, thus being part of the biotope of these pads. It turns out that peculiarities of the application domain of medical therapy planning result in special communication requirements. To mention just one, changes in the patient data record pad usually affect only a few pads within a large therapy plan. Instead of propagating changes through the whole hierarchically structured plan according to the chip.dll, it is more efficient and secure if one establishes a direct data exchange between therapy actions pads and the underlying patient data record pad. In other words, certain peculiarities of the medical therapy planning domain lead the authors to changes in the biotope – the biotope issue has been recognized a central one. 4 Perspectives at Knowledge and Human Learning There is recent interest in bridging the gap between meme media technologies and technology enhanced learning (e-learning, for short) like in [13] and [17], e.g. Work about so-called subjunctive interfaces [14] seems to be relevant as well. Meme media meet e-learning. In the author’s opinion, this is an enormously promising domain of research, development and applications, because human memes meet meme media and compliment each other. The key question for the conditions of both meme types’ co-existence and, in particular, co-evolution has never before been made explicit. In e-learning, human knowledge is not simply encoded in media objects, distributed over digital networks, and decoded at the other end. Knowledge is the result of construction processes when humans deal with content and actively engage in exploration, communication, experiment, and the like [2, 5]. It requires in-depth studies of different types (consult [4], e.g.) to reveal the mechanisms of learning. On a more technological level, one may work about meme media technologies and tools for technology enhanced learning [13, 17]. Another approach may be to sort out didact principles to which meme media technologies seem to be especially suitable. Explorative learning, e.g., may be substantially supported by subjunctive interfaces as proposed and developed by Aran Lunzer [14]. As we know it from the arts (see [18] for an invaluable source of case studies), affecting the human means to find some way to get the human’s memes engaged. The approach of e-learning concentrates on the art of connecting meme media to the human memes. From this perspective, storyboarding [12] comes into play. In the future, we will ponder, perhaps, about the biotope issue of human and meme media. 5 Summary and Conclusions Instead of repeating the arguments from before, four theses are intended to summarize the author’s viewpoint and – hopefully – to provoke some discussion among interested scientists, developers and users. – The biotope issue is essential for memes in the biological reality. It is usually taken for granted. To meme media implementations, the biotope issue is of similar importance, but it is widely underestimated. A primary goal of the present paper is to attract attention to the biotope issue and to establish an awareness of the biotope issue in the scientific and engineering community. – The most difficult point in the biotope issue discussion is that humans have to step back, have to give priority to the memes and have to consider themselves as part of the memes’ and the meme media objects’ biotope, only. Success of meme media implementations and applications requires some fundamental rethinking. – There is a large variety of technical and technological details in meme media implementations which turn out to be substantial aspects of the biotope issue. Seeing the biotope issue in its right perspective is the only way to find a guideline for dealing with those seemingly minor implementational questions. – The question for the biotope issue may be considered as a crucial criterion to distinguish those application domains that are appropriate to meme media approaches from those that are not. Simpler software technological problems that ask for a certain middleware solution do not require meme media, if there does not appear any biotope problem. – Certain meme media implementations are best distinguished from concurring developments of information technologies by their evolutionary potentials. Only emphasis on the evolutionary potentials can make the meme media implementations successful. Thus, respecting the biotope issue is of strategic importance and has certain commercial relevance. For a greater impact on meme media implementations and their usage, the present paper should result in some guideline or, at least, in some checklist to take the biotope issue appropriately into account. The reader should be aware of a rather delicate aspect of memetics which complicates not only writing about the biotope issue, but also finding agreements about the essentials of memetics. Thinking about memetics thoroughly leads to the question about the human self. From a very strict memetic point of view (cf. [1] and [8]), there is no other human self than a story told us by our brain. Last but not least, it is worth to point to the fact that a deeper discussion of the aspects systematized in chapter 3 of this paper does need more IntelligentPad development competence. The author depends very much on other colleagues and on ongoing experience with application attempts.