All things stigmergy
Special Issue of Cognitive Systems Research
Join the Linkedin Stigmergy group:
Agents interact with others and their environment to circumvent cognitive limitation. Stigmergy is a property supervening on informational amplification and decay in complex adaptive social environments, environments that have sets of “search space” constraints. Stigmergic cognition is a socio-cognitive view (an externalist view) of knowledge and mind, necessarily dual aspect, that shows how individual cognition subject to cognitive and epistemic constraints, resolves the so-called ”coordination” paradox: that is, how does one reconcile the prima facie “chaotic” behavior of individuals with collective achievement, be it an anthill or an economy. Intelligence on a stigmergic account is distributed and manifest in the network of interactions among the individual agents and between individuals and their environment. The term stigmergy (originating in entomology and popularized through computational intelligence) is a compound term: stigma (στίγμα) meaning “sign” and ergon (ἔργον) meaning “work.” Agents observe signals that trigger an action or response, which in turn may reinforce or modify a given signal, thereby influencing the actions of others – there is an ongoing cybernetic relation between mind and milieu.
Though the concept of stigmergy has been associated with ant- or swarm-like “agents” with minimal cognitive ability, stigmergy offers a powerful analytical tool to be deployed in the human domain. Stigmergic systems are a ubiquitous feature of human sociality and include classic examples of complex adaptive systems such as stock markets, economies, traffic patterns, supply logistics and resource allocation, urban sprawl, and cultural memes. A more recent array of stigmergic activities afforded by digital technology includes wiki, the Google algorithm, recommendation algorithms, social networking and open-source ware, data mining, and telecommunications routing. The typical features of a stigmergic system (or a complex adaptive system if you like) are:
• A context or environment:
o Comprised by an indefinite number of local environments
o Only partially perceivable through an internal dynamics that govern its temporal evolution
o There are a multiplicity of agents populating with no one individual or clustering of individuals having global knowledge
o Rationality is bounded
o Behavior is self-organized
o Behavior is stochastic
o Behavior is dynamical
• Novel features:
o Arise from interactions that are neither predictable nor reducible to simpler constituents.