# Generating Knowledge in a Connected World  ## Metadata - Author: [[Max Boisot]] - Full Title: Generating knowledge in a connected world: The case of the ATLAS experiment at CERN - Category: #Source/articles - [Article link](x-devonthink-item://62A9E1F5-AB31-45FF-B706-76B681D6B70F) ## Summary - Bottom up organising is possible, especially with IT systems supporting. - This is about network co-ordination, trust substituting for hierarchical control. - The conceptual framework is the i-space - Technology allows for more diffusion of information (this is the diffusion effect). This allows bureaucracies to work more like markets. Technology also allows for information to reach individuals at a lower level of codification and abstraction- i.e. towards clans and fiefs (this is the bandwidth effect). This has impacts on what is possible without bureaucracy. - The clans and fiefs are similar to that present in a bureaucracy. - ATLAS involved over 3000 physicists working in 174 universities and laboratories, over 38 countries. - Co-ordination is bottom up and consensus driven. No formal management structure/ no way to give orders. - The detector itself acts as a 'boundary object' that creates a common vision for all participants in the team. It moves through the I-Space as it gets more and more developed, becoming more codified and concrete. - to do this, clan values and outstanding (IT enhanced) communication must be maintained. >[!REFLECTION]+ > Do charities have a better understanding of a 'boundary object' because they are charities with that key focus? Is that a differentiator, especially when Universities cannot decide whether or not students are partners, customers or clients? This becomes part of our VRIO capabilities? ### Concepts - The I-Space. - Relates the speed and extent of information flows within a population of agents to the possibilities for information structuring. - The axis are codified vs uncodified, abstract vs concrete, and undiffused vs diffused. - Bureaucracies are codified and abstract, but are not diffused. Markets are codified and abstract, but ARE diffused. - Clans are not codified or abstract, nor are they diffused. This means trust and shared values are essential. Fiefs are the same but power is consolidated in one individual ## Highlights - Although the proportion of bottom-up activity has increased, it has not displaced the top-down bias in the governance structures of firms and the formal processes that give them effect. Yet (Page 2) - How is network coordination actually achieved? How does trust evolve to the point where it can substitute for hierarchical control? Finally, what level of task complexity are such networks capable of managing in this distributed fashion? (Page 2) - If the essence of coordination is to bring situations, intentions and behaviours into alignment, the essence of managerial coordination is to do so through the agency of others (Barnard, 1938) (Page 2) - An information environment characterized by high degrees of codification, abstraction, and diffusibility, for example, will deliver impersonal networks where agents do not have to know each other in order to interact: markets, where the diffusion of information is uncontrolled; bureaucracies, where it is subject to some degree of central control an information environment characterized by low degrees of codification, abstraction—and hence diffusibility—will deliver highly personalized networks in which trust and shared values are essential: clans, where the information is diffused face-to-face within a group of limited size; fiefs, where undiffused information remains confined within a single head to become a source of personal power. (Page 3) - The diffusion effect—at all levels of codification and abstraction, the new ICTs can process and transmit more data to more people per unit of time than hitherto (Page 6) - bandwidth effect—a given target population in the I-Space can now be reached at a lower level of codification and abstraction than hitherto (Page 6) #### The case of the ATLAS experiment at CERN - It involves the coordination of over 3000 p sities and laboratories, and spread across over 38 countries ATLAS Collaboration is organized around both the components and the need for integrating these (Page 8) - Although the Collaboration has a project management team, it manages with a light touch with little formal managerial authority to draw upon. The glue that binds participating institutions together is not contracts but Memoranda of Understanding. Team members are paid for by their respective participating institutions and do not readily ‘take orders’ from other members of the Collaboration (Page 8) - Coordination is therefore mostly a bottom-up, consensus-driven affair, achieved numerous by face-to face meetings within and between the different teams, with many participants who cannot be physically present taking part virtually (Page 8) - Whatever structures guide action, emerge gradually from the interactions of the players themselves and remain provisional and subject to change— the outcome of what Mintzberg and Waters (1985) describe as emergent strategies (Page 9) - the detector itself, acting as a boundary object (Page 9) - A boundary object acts as a common reference point that allows different actors to coordinate their actions without interacting directly with each other (Page 9) - No boundary object, however, could coordinate an adhocracy of the size and complexity of the ATLAS Collaboration unaided. We therefore hypothesize that two further conditions must be met: (Page 9) - The maintenance of clan valuesThe coherence that the ATLAS detector has been able to achieve in the course of its evolution draws upon deeply held values, motivations and beliefs, tacitly shared by all members of the collaboration throughout the project (Page 10) - The potential of the new ICTs must be fully exploited (Page 10) - To summarize, we believe that the adhocracy that is the ATLAS Collaboration is held together first by a common focus on the ATLAS detector acting as a coordinator of the collaboration’s members; second, by the shared values and beliefs characteristic of clans that maintain the focus; and finally, by the enabling role played by the new ICTs in maintaining the necessary global connectivity between the collaboration’s members (Page 10) - Three avenues for such research suggest themselves: (1) What might be a minimal specification for a boundary object? (2) What might be the maximum size and geographical spread of the adhocracies that different types of boundary object could help coordinate? How complex are the tasks that different boundary objects could help to coordinate? (Page 11) - Carlile Paul R (2002) A pragmatic view of knowledge and boundaries: boundary objects in new product development. Organization Science 13(4): 442–455 (Page 11) - Mintzberg H (1979) The Structuring of Organizations: A Synthesis of the Research. Englewood Cliffs, NJ: Prentice-Hall (Page 11) ## APA Reference Boisot, M. (2011). Generating knowledge in a connected world: The case of the ATLAS experiment at CERN. Management Learning, 42(4), 447–457. https://doi.org/10.1177/1350507611408676