Towards an Ideal Social Simulation Language

Relating Structure and Dynamics in Organisation Models

“…The framework presented here contributes directly to the foundations of some of the criteria proposed by [4] (e.g., the criteria regarding the identification of constraints and flexible analysis) for an ideal social simulation language (ISSL), and it illustrates for some of the other criteria how such an ISSL can come about (e.g., the criteria regarding modeller' s intentions, compositionality, and emergence) in the context of the structure and dynamics of organizations. Furthermore, a shared basis of the work of [4] and the work presented here lies in the view that the different simulation processes need to be specified separately from the significant outcomes of the simulation, and that the processes that emerge from the simulation are typically more interesting than the resulting states of the simulation.…”

“…• temporal logic with operators that are specific for elements of the organisational structure or multi-agent system; e.g., [9], [4].…”

“…Often these simulation models are based on some implementation environment and not specified in an implementationindependent manner using a formally defined conceptual language. The Strictly Declarative Modelling Language SDML [22] (see also [4]) , and the use of the agentoriented modelling approach DESIRE in social simulation as presented in [3] are some of the few exceptions. Both modelling approaches focus on specification and simulation; however, they do not offer dedicated support for a specific type of organisational structure.…”

Validation and Verification in Social Simulation: Patterns and Clarification of Terminology

“…Verification -Most researchers define verification as referring to the performance of the program code, for instance, as "the process of checking that a program does what it was planned to do" (Gilbert and Trotzsch, 1999, p.21), or "checking that the representation is faithful to the simulator's intentions" (Edmonds, 2003, p.108); Validation -According to Gilbert and Troitzsch (1999, p.22), "While verification concerns whether the program is working as the researcher expects it to, validation concerns whether the simulation is a good model of the target", or, according to Edmonds (2003, p.108), that "the expression of the simulation in terms of outcomes is faithful to the relevant social phenomena. "…”

“…Indeed, this would return us to the epistemological debate, considered elsewhere, where adequateness can be explained according to a number of epistemological perspectives, such as the notions of formal, empirical and intentional knowledge. 8 Our goal is essentially terminological. A minimal definition could be the following: adequateness means that the execution behaviours of the computerized model coincide with the semantics of both the pre-and the postcomputerized models, in accord with the researcher's intentions.…”